Wang, Chu-Kia - Reinforced Concrete Design-Oxford University Press, USA (2017 - 2018)

i
REINFORCED CONCRETE DESIGN
ii
iii
REINFORCED
CONCRETE DESIGN
CHU-K IA WANG
CHARLES G. SALMON
JOSÉ A. PINCHEIRA
GUSTAVO J. PARRA-M ONTESINOS
University of Wisconsin–Madison
EIGHTH EDITION
New York Oxford
OXFORD UNIVERSITY PRESS
iv
Oxford University Press is a department of the University of Oxford. It furthers

the University’s objective of excellence in research, scholarship, and education
by publishing worldwide. Oxford is a registered trade mark of Oxford University
Press in the UK and certain other countries.
Published in the United States of America by Oxford University Press

198 Madison Avenue, New York, NY 10016, United States of America.
© 2018 by Oxford University Press

© 2007 by John Wiley & Sons, Inc.
© 1997 by Addison Wesley Publishing Company
For titles covered by Section 112 of the US Higher Education

Opportunity Act, please visit www.oup.com/us/he for the latest
information about pricing and alternate formats.
All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by license, or under terms agreed with the appropriate reproduction
rights organization. Inquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above.
You must not circulate this work in any other form

and you must impose this same condition on any acquirer.
Library of Congress Cataloging-in-Publication Data

Names: Wang, Chu-Kia, 1917–author.
Title: Reinforced Concrete Design / Chu-Kia Wang, Charles G. Salmon, José A. Pincheira,
Gustavo J. Parra-Montesinos.
Description: New York: Oxford University Press, [2018] |
Includes bibliographical references and index.
Identifiers: LCCN 2017000252 | ISBN 9780190269807 (hardcover) |
ISBN 9780190647049 (looseleaf) | ISBN 9780190269852 (eISBN)
Subjects: LCSH: Reinforced concrete construction.
Classification: LCC TA683.2 .W3 2018 | DDC 624.1/8341—dc23
LC record available at https://lccn.loc.gov/2017000252
987654321
Printed by Edwards Brothers Malloy, United States of America
v
CONTENTS IN BRIEF
Preface xxi
About the Authors xxv
Conversion Factors xxvii
1 INTRODUCTION, MATERIALS, AND PROPERTIES 1
2 DESIGN METHODS AND REQUIREMENTS 31
3 FLEXURAL BEHAVIOR AND STRENGTH OF BEAMS 43
4 T-S ECTIONS IN BENDING 94
5 SHEAR STRENGTH AND DESIGN FOR SHEAR 110
6 DEVELOPMENT OF REINFORCEMENT 176
7 ANALYSIS OF CONTINUOUS BEAMS AND ONE-WAY

SLABS 239
8 DESIGN OF ONE-WAY SLABS 254
9 DESIGN OF SLAB–B EAM–G IRDER AND JOIST FLOOR

SYSTEMS 266
10 MEMBERS IN COMPRESSION AND BENDING 321
11 MONOLITHIC BEAM-C OLUMN CONNECTIONS 380
12 SERVICEABILITY 403
13 SLENDERNESS EFFECTS ON COLUMNS 463

vi
vi C ontents in B rief
14 STRUT-A ND-T IE MODELS—D EEP BEAMS, BRACKETS,

AND CORBELS 528
15 STRUCTURAL WALLS 569
16 DESIGN OF TWO-WAY FLOOR SYSTEMS 622
17 YIELD LINE THEORY OF SLABS 720
18 TORSION 748
19 FOOTINGS 799
20 INTRODUCTION TO PRESTRESSED CONCRETE 844
21 COMPOSITE MEMBERS AND CONNECTIONS 897
Index 949
vi
CONTENTS
Preface xxi
About the Authors xxv
Conversion Factors xxvii
1 INTRODUCTION, MATERIALS, AND PROPERTIES 1
1.1 Reinforced Concrete Structures 1
1.2 Historical Background 2
1.3 Concrete 4
1.4 Cement 5
1.5 Aggregates 6
1.6 Admixtures 6
1.7 Compressive Strength 9
1.8 Tensile Strength 12
1.9 Biaxial and Triaxial Strength 14
1.10 Modulus of Elasticity 14
1.11 Creep and Shrinkage 16
1.12 Concrete Quality Control 18
1.13 Steel Reinforcement 19
1.14 Fiber-Reinforced Concrete 25
1.15 Units 26
Selected References 26
2 DESIGN METHODS AND REQUIREMENTS 31
2.1 Structural Design Process—General 31
2.2 ACI Building Code 31

vi
viii C ontents
2.3 Strength Design and Working Stress Methods 32
2.4 Working Stress Method 33
2.5 Strength Design Method 33
2.6 Safety Provisions—General 34
2.7 Safety Provisions—ACI Code Load Factors and Strength Reduction Factors 36
2.8 Serviceability Provisions—General 38
2.9 Serviceability Provisions—ACI Code 39
2.10 Handbooks and Computer Software 39
2.11 Dimensions and Tolerances 40
2.12 Accuracy of Computations 41
3 FLEXURAL BEHAVIOR AND STRENGTH OF BEAMS 43
3.1 General Introduction 43
3.2 Flexural Behavior and Strength of Rectangular Sections 44
3.3 Whitney Rectangular Stress Distribution 47
3.4 Nominal Flexural Strength Mn—Rectangular Sections Having Tension

Reinforcement Only 48
3.5 Balanced Strain Condition 51
3.6 Tension-and Compression-Controlled Sections 52
3.7 Minimum Tension Reinforcement 58
3.8 Design of Rectangular Sections in Bending Having Tension Reinforcement

Only 60
3.9 Practical Selection for Beam Sizes, Bar Sizes, and Bar Placement 64
3.10 Nominal Flexural Strength Mn of Rectangular Sections Having Both Tension

and Compression Reinforcement 72
3.11 Design of Beams Having Both Tension and Compression Reinforcement 78
3.12 Nonrectangular Sections 84
3.13 Effect of As, As′ , b, d, fc′ , and fy on Flexural Behavior 86
Problems 88
ix
C ontents ix
4 T-S ECTIONS IN BENDING 94
4.1 General 94
4.2 Comparison of Rectangular and T-Sections 95
4.3 Effective Flange Width 95
4.4 Nominal Moment Strength Mn of T-Sections 97
4.5 Design of T-Sections in Bending 105
Problems 108
5 SHEAR STRENGTH AND DESIGN FOR SHEAR 110
5.1 Introduction 110
5.2 Shear Stresses Based on Linear Elastic Behavior 111
5.3 Combined Normal and Shear Stresses 113
5.4 Behavior of Beams without Shear Reinforcement 114
5.5 Shear Strength of Beams without Shear Reinforcement—ACI Approach 119
5.6 Function of Web Reinforcement 122
5.7 Truss Model for Reinforced Concrete Beams 125
5.8 Shear Strength of Beams with Shear Reinforcement—ACI Approach 128
5.9 Deformed Steel Fibers as Shear Reinforcement 129
5.10 ACI Code Design Provisions for Shear 130
5.11 Critical Section for Nominal Shear Strength Calculation 135
5.12 Shear Strength of Beams—Design Examples 136
5.13 Shear Strength of Members under Combined Bending and Axial Load 146
5.14 Deep Beams 151
5.15 Shear Friction 152
5.16 Brackets and Corbels 157
Problems 172
x
x C ontents
6 DEVELOPMENT OF REINFORCEMENT 176
6.1 General 176
6.2 Development Length 177
6.3 Flexural Bond 179
6.4 Bond Failure Mechanisms 180
6.5 Flexural Strength Diagram—Bar Bends and Cutoffs 182
6.6 Development Length for Tension Reinforcement—ACI Code 185
6.7 Modification Factors ψt, ψe , ψs, and λ to the Bar Development Length
Equations—ACI Code 190
6.8 Development Length for Compression Reinforcement 194
6.9 Development Length for Bundled Bars 195
6.10 Development Length for a Tension Bar Terminating in a Standard Hook 195
6.11 Bar Cutoffs in Negative Moment Region of Continuous Beams 198
6.12 Bar Cutoffs in Positive Moment Region of Continuous Beams 201
6.13 Bar Cutoffs in Uniformly Loaded Cantilever Beams 202
6.14 Development of Positive Reinforcement at Simple Supports and at Points

of Inflection 209
6.15 Development of Shear Reinforcement 211
6.16 Tension Lap Splices 213
6.17 Welded Splices and Mechanical Connections in Tension 215
6.18 Compression Lap Splices 216
6.19 End Bearing Connections, Welded Splices, and Mechanical Connections

in Compression 217
6.20 Splices for Members under Compression and Bending 217
6.21 Design Examples 217
Problems 236
7 ANALYSIS OF CONTINUOUS BEAMS AND ONE-WAY

SLABS 239
xi
C ontents xi
7.2 Analysis Methods under Gravity Loads 240
7.3 Arrangement of Live Load for Moment Envelope 241
7.4 ACI Code—Arrangement of Live Load and Moment Coefficients 246
7.5 ACI Moment Diagrams 247
7.6 Shear Envelope for Design 250
Selected Reference 252
Problems 252
8 DESIGN OF ONE-WAY SLABS 254
8.1 Definition 254
8.2 Analysis Methods 254
8.3 Slab Design 255
8.4 Choice of Reinforcement 258
8.5 Bar Details 264
Problems 265
9 DESIGN OF SLAB-B EAM-G IRDER AND JOIST FLOOR

SYSTEMS 266
9.2 Size of Beam Web 267
9.3 Continuous Frame Analysis for Beams 270
9.4 Choice of Longitudinal Reinforcement in Beams 274
9.5 Shear Reinforcement in Beams 285
9.6 Details of Bars in Beams 287
9.7 Size of Girder Web 294
9.8 Continuous Frame Analysis for Girders 297
9.9 Choice of Longitudinal Reinforcement in Girders 300
9.10 One-Way Joist Floor Construction 306
9.11 Design of Joist Floors 307

xi
xii C ontents
9.12 Redistribution of Moments—Introduction to Limit or Plastic Analysis 312
Problems 318
10 MEMBERS IN COMPRESSION AND BENDING 321
10.2 Types of Columns 322
10.3 Behavior of Columns under Pure Axial Load 322
10.4 Safety Provisions for Columns 325
10.5 Concentrically Loaded Short Columns 326
10.6 Strength Interaction Diagram 326
10.7 Slenderness Effects 328
10.8 Lateral Ties 329
10.9 Spiral Reinforcement and Longitudinal Bar Placement 330
10.10 Limits on Percentage of Longitudinal Reinforcement 332
10.11 Maximum Strength in Axial Compression—ACI Code 333
10.12 Balanced Strain Condition 333
10.13 Nominal Strength of a Compression-Controlled Rectangular Section 336
10.14 Nominal Strength of a Rectangular Section with Eccentricity e Greater than That
at the Balanced Strain Condition 340
10.15 Design for Strength—Region I, Minimum Eccentricity 342
10.16 Design for Strength—Region II, Compression-Controlled Sections

(emin < e < eb ) 345
10.17 Design for Strength—Region III, Transition Zone and Tension-Controlled

Sections (e > eb ) 351
10.18 Circular Sections Under Combined Compression and Bending 354
10.19 Combined Axial Tension and Bending 357
10.20 Combined Axial Force and Biaxial Bending 359
10.21 Design for Shear 368
Problems 374
xi
C ontents xiii
11 MONOLITHIC BEAM-C OLUMN CONNECTIONS 380
11.2 Beam-Column Joints Actions 381
11.3 Joint Transverse Reinforcement 383
11.4 Joint Shear Strength 387
11.5 Column-to-Beam Moment Strength Ratio 389
11.6 Anchorage of Reinforcement in the Joint Region 390
11.7 Transfer of Column Axial Forces through the Floor System 391
11.8 Examples 391
11.9 Additional Remarks 399
Problems 401
12 SERVICEABILITY 403
12.2 Fundamental Assumptions 403
12.3 Modulus of Elasticity Ratio, n 404
12.4 Equilibrium Conditions 404
12.5 Method of Transformed Section 407
12.6 Deflections—General 410
12.7 Deflections for Linear Elastic Members 411
12.8 Modulus of Elasticity 414
12.9 Effective Moment of Inertia 414
12.10 Instantaneous Deflections in Design 417
12.11 Creep Effect on Deflections under Sustained Load 428
12.12 Shrinkage Effect on Deflections under Sustained Load 431
12.13 Creep and Shrinkage Deflection—ACI Code Method 435
12.14 Creep and Shrinkage Deflection—Alternative Procedures 436
12.15 ACI Minimum Depth of Flexural Members 439

vxi
xiv C ontents
12.16 Span-to-Depth Ratio to Account for Cracking and Sustained Load Effects 441
12.17 ACI Code Deflection Provisions—Beam Examples 446
12.18 Crack Control for Beams and One-Way Slabs 451
12.19 Side Face Crack Control for Large Beams 455
12.20 Control of Floor Vibrations—General 456
Problems 459
13 SLENDERNESS EFFECTS ON COLUMNS 463
13.1 General 463
13.2 Buckling of Concentrically Loaded Columns 465
13.3 Effective Length Factor 468
13.4 Moment Magnification—Members with Transverse Loads—Without Joint

Lateral Translation (i.e., No Sidesway) 470
13.5 Moment Magnification—Members Subject to End Moments Only—Without

Joint Lateral Translation (i.e., No Sidesway) 472
13.6 Moment Magnification—Members with Sidesway—Unbraced (Sway) Frames 477
13.7 Interaction Diagrams—Effect of Slenderness 479
13.8 ACI Code—General 480
13.9 ACI Code—Moment Magnifier Method for Columns in Nonsway Frames 482
13.10 ACI Code—Moment Magnifier Method for Columns in Sway Frames 485
13.11 Alignment Charts for Effective Length Factor k 490
13.12 Second-Order Analysis—ACI Code 493
13.13 Minimum Eccentricity in Design 493
13.14 Biaxial Bending and Axial Compression 494
13.15 ACI Code—Slenderness Ratio Limitations 494
13.16 Amplification of Moments in Beams 495
13.17 Examples 495
Problems 526
xv
C ontents xv
14 STRUT-A ND-T IE MODELS—D EEP BEAMS, BRACKETS,

AND CORBELS 528
14.2 Deep Beams 542
14.3 Brackets and Corbels 559
14.4 Additional Remarks 565
Problems 567
15 STRUCTURAL WALLS 569
15.1 General 569
15.2 Minimum Wall Dimensions and Reinforcement Requirements—ACI Code 569
15.3 Design of Nonbearing Walls 573
15.4 Design of Bearing Walls 573
15.5 Design of Shear Walls 576
15.6 Lateral Support of Longitudinal Reinforcement 596
15.7 Retaining Structures 597
Problems 620
16 DESIGN OF TWO-WAY FLOOR SYSTEMS 622
16.1 General Description 622
16.2 General Design Concept of the ACI Code 624
16.3 Total Factored Static Moment 625
16.4 Ratio of Flexural Stiffnesses of Longitudinal Beam to Slab 633
16.5 Minimum Slab Thickness for Deflection Control 637
16.6 Nominal Requirements for Slab Thickness and Size of Edge Beams, Column
Capital, and Drop Panel 639
16.7 Direct Design Method—Limitations 644
16.8 Direct Design Method—Longitudinal Distribution of Moments 645

xvi
xvi C ontents
16.9 Direct Design Method—Effect of Pattern Loadings on Positive Moment 647
16.10 Direct Design Method—Procedure for Computation of Longitudinal

Moments 647
16.11 Torsion Stiffness of the Transverse Elements 651
16.12 Transverse Distribution of Longitudinal Moment 656
16.13 Design of Slab Thickness and Reinforcement 662
16.14 Size Requirement for Beam (If Used) in Flexure and Shear 669
16.15 Shear Strength in Two-Way Floor Systems 671
16.16 Shear Reinforcement in Flat Plate Floors 676
16.17 Direct Design Method—Moments in Columns 686
16.18 Transfer of Moment and Shear at Junction of Slab and Column 687
16.19 Openings and Corner Connections in Flat Slabs 697
16.20 Equivalent Frame Method for Gravity Load Analysis 698
16.21 Equivalent Frame Models 710
16.22 Equivalent Frame Method for Lateral Load Analysis 711
Problems 718
17 YIELD LINE THEORY OF SLABS 720
17.2 General Concept 720
17.3 Fundamental Assumptions 723
17.4 Methods of Analysis 724
17.5 Yield Line Analysis of One-Way Slabs 725
17.6 Work Done by Yield Line Moments in Rigid Body Rotation of Slab
Segment 728
17.7 Nodal Forces at Intersection of Yield Line with Free Edge 729
17.8 Nodal Forces at Intersection of Three Yield Lines 732
17.9 Yield Line Analysis of Rectangular Two-Way Slabs 736
17.10 Corner Effects in Rectangular Slabs 742

xvi
C ontents xvii
17.11 Application of Yield Line Analysis to Special Cases 743
Problems 747
18 TORSION 748
18.1 General 748
18.2 Torsional Stress in Homogeneous Sections 749
18.3 Torsional Stiffness of Homogeneous Sections 751
18.4 Effects of Torsional Stiffness on Compatibility Torsion 752
18.5 Torsional Moment Strength Tcr at Cracking 755
18.6 Strength of Rectangular Sections in Torsion—Skew Bending Theory 757
18.7 Strength of Rectangular Sections in Torsion—Space Truss Analogy 761
18.8 Strength of Sections in Combined Bending and Torsion 765
18.9 Strength of Sections in Combined Shear and Torsion 767
18.10 Strength Interaction Surface for Combined Bending, Shear, and Torsion 768
18.11 Torsional Strength of Concrete and Closed Transverse Reinforcement—ACI

Code 770
18.12 Combined Torsion with Shear or Bending—ACI Code 772
18.13 Minimum Requirements for Torsional Reinforcement—ACI Code 773
18.14 Examples 775
Problems 796
19 FOOTINGS 799
19.1 Purpose of Footings 799
19.2 Bearing Capacity of Soil 799
19.3 Types of Footings 800
19.4 Types of Failure 800
19.5 Shear Strength 802
19.6 Flexural Strength and Development of Reinforcement 803

xvii
xviii C ontents
19.7 Proportioning Footing Areas for Equal Settlement 804
19.8 Investigation of Square Spread Footings 804
19.9 Design of Square Spread Footings 809
19.10 Design of Rectangular Footings 814
19.11 Design of Plain and Reinforced Concrete Wall Footings 818
19.12 Combined Footings 822
19.13 Design of Combined Footings 823
19.14 Pile Footings 841
Problems 842
20 INTRODUCTION TO PRESTRESSED CONCRETE 844
20.2 Historical Background 844
20.3 Advantages and Disadvantages of Prestressed Concrete Construction 845
20.4 Pretensioned and Post-tensioned Beam Behavior 846
20.5 Service Load Stresses on Flexural Members—Tendons Having Varying Amounts

of Eccentricity 849
20.6 Three Basic Concepts of Prestressed Concrete 853
20.7 Loss of Prestress 856
20.8 Nominal Strength Mn of Flexural Members 866
20.9 Cracking Moment 871
20.10 Shear Strength of Members without Shear

Reinforcement 873
20.11 Shear Reinforcement for Prestressed Concrete Beams 881
20.12 Development of Reinforcement 883
20.13 Proportioning of Cross Sections for Flexure When No Tension is Permitted 885
20.14 Additional Topics 894
Problems 895
xi
C ontents xix
21 COMPOSITE MEMBERS AND CONNECTIONS 897
21.2 Composite Action 897
21.3 Concrete Composite Flexural Members 901
21.4 Concrete-Steel Composite Columns 916
21.5 Concrete-Encased Steel Composite Columns 918
21.6 Concrete-Filled Tube Columns 934
21.7 Moment Connections with Composite Columns 943
Problems 947
Index 949
x
xxi
PREFACE
The eighth edition of this textbook has been substantially updated to incorporate the
changes introduced by the publication of the 2014 American Concrete Institute (ACI)
Building Code and Commentary for Structural Concrete, as well as to reflect changes in
construction and design practices that have occurred in the last few years.
APPROACH
This new edition follows the same philosophical approach that has gained wide acceptance
of users since the first edition was published in 1965. Herein, as in past editions, consider
able emphasis is placed on presenting to the student, as well as to the practicing engineer,
the basic principles of reinforced concrete design and the concepts necessary to understand
and properly apply the provisions of the ACI Building Code. Numerous examples are pre-
sented to illustrate the general approach to design and analysis. The material is incorpo-
rated into the chapters in a way that permits the reader to either study in detail the concepts
in logical sequence or obtain a qualitative explanation and proceed directly to the design
process using the ACI Code.
NEW TO THIS EDITION

The eighth edition of this book incorporates the changes arising from the publication of the
2014 American Concrete Institute Building Code and Commentary (ACI 318-14). While
past editions of the ACI Code were largely structured around member actions (e.g., flex-
ure, shear, and axial load), ACI 318-14 is organized primarily by structural elements
(e.g., beams, columns, walls). As a result, virtually all design provisions have changed
in format and number, and are located under a new chapter designation in the Code.
Accordingly, all chapters and example problems have been revised to conform to the for-
mat and reorganization of the 2014 ACI Building Code (ACI 318-14). In addition, content
has been reorganized within existing chapters, moved to other chapters, or relocated as new,
stand-alone chapters for better continuity and presentation of the material. Main revisions,
updates, and new material include the following.
1. A new chapter on Structural Walls (Chapter 15) has been added. This chapter includes
the design of Non-Bearing and Bearing Walls, as well as the design of Shear Walls.
The design of Cantilever Retaining Walls (formerly Chapter 12) has been revised and
is included at the end of the new Chapter 15.
2. The chapter on composite construction (Chapter 21) has been substantially revised
and renamed “Composite Members and Connections” to better reflect its new scope.
The first part covers the design of concrete-concrete composite flexural members,
including calculation of deflections for shored and unshored construction. In addi-
tion, the chapter now includes sections on Concrete-Encased Steel Columns and
Concrete-Filled Tubes, along with a new section on Moment Connections between
Composite Columns and Steel Beams.
xxi
xxii P R E F A C E
3. The material on the Strut and Tie Method and its application to the design of Deep
Beams, Brackets, and Corbels (previously included in Chapter 5, Shear Strength) has
been updated and is now presented as a stand-alone, separate chapter (Chapter 14).
4. The material on Rectangular Sections in Bending under Service Load Conditions
(formerly Chapter 4) and Deflections (formerly Chapter 14) was revised and com-
bined into a single chapter dealing with Serviceability (new Chapter 12).
5. The design of T-Sections in Bending (formerly Chapter 9) was relocated as Chapter 4,
immediately after Chapter 3, Flexural Behavior and Strength of Beams, for better
flow and continuity of the material on beam design.
6. The chapter on Slenderness Effects on Columns, now Chapter 13 (formerly
Chapter 15) has been revised to include additional content and example problems for
better understanding of the ACI Code procedures established to account for second-
order effects in column design.
7. The alternative provisions of Appendix B and the alternative load factors of Appendix
C included in past editions of the ACI Code have been removed from the Code.
Accordingly, discussion and example problems corresponding to these appendices
are not included in this edition of the textbook.
8. All sections have been revised to improve the flow and continuity of the material in
accordance to the changes in ACI 318-14.
In addition to the content revisions indicated above, all the examples and the problems
at the end of each chapter have been revised and updated to conform to the current ACI
Code. Examples and problems have also been updated, and new examples have been added
to reflect the strengths of the materials most commonly used in current practice. A few
examples, however, use less common values in order to emphasize specific aspects of the
design process that students might otherwise overlook.
To aid instructors, a solutions manual has been prepared for the end-of-chapter prob-
lems. Many problems are solved in Mathcad®, allowing alternative solutions to be easily
arrived at by modifying a few parameters, either as suggested in this textbook or at the
choice of the instructor.
COURSE SUGGESTIONS
Depending on the proficiency required of the student, this book may provide material for
two courses of three or four semester-hours each. It is suggested that the beginning course
in concrete structures for undergraduate students contain all or most of the material in
Chapters 1 through 6, and Chapters 8 through 10.
The second course may begin with Chapter 10, using that topic (members in compres
sion and bending) to review many of the subjects in the first course, followed by Chapter 12
on serviceability, Chapter 13 on slenderness effects on columns, and Chapter 16 on two-
way slab systems. In addition, one or two of the following may be included in a second
course: the remaining sections of Chapter 5 on shear strength affected by axial force;
Chapter 15 on structural walls; Chapter 18 on torsion; Chapter 14 on strut-and-tie models,
deep beams, brackets, and corbels; and Chapter 20 on prestressed concrete.
Chapters on beam- column connections (Chapter 11), yield line theory of slabs
(Chapter 17), footings (Chapter 19), and composite members and connections (Chapter 21)
may serve as contents for a third course.
UNITS
This edition continues the modest treatment of SI units used in previous editions. The
2014 ACI Code has an SI version (known as ACI 318-14M), and the SI versions of the
ACI Code equations appear in this book as footnote equations with the same equation
number. According to the ACI Code, the designer must use in its entirety either the
Inch-Pound units version (ACI 318-14) or the SI version (ACI 318-14M), although the
xxii
PREFACE xxiii
Inch-Pound units version is the official version of the Code. The authors believe that
sufficient metrication should be included in a text on reinforced concrete to permit the
reader to gain some familiarity with SI units, but suspect that too much would interfere
with learning the basic concepts of concrete design; constant conversion back and forth
between Inch-Pound and SI units is more confusing than using either one exclusively.
The text provides data on reinforcing bars in accordance with the American Society for
Testing and Materials Inch-Pound units, and also ASTM SI units (the “soft” conversion
of the bar sizes and strengths approved in 1996). Some design tables are provided for
bars and material strengths in SI units, a few numerical examples are given in SI units,
and some problems at the ends of chapters are given with an SI alternate in parentheses
at the statement concluding the problem.
In all parts of this book that use metric units, force is expressed in the newton (N) or
kilonewton (kN) unit. The SI unit of stress is the pascal (Pa), or newton per meter squared,
which because of its typically large numerical value is usually expressed in megapascals
(MPa): that is, 106 pascals. A few diagrams show, along the stress axis, the kilogram force
per centimeter squared (kgf/cm2) in addition to Inch-Pound and SI units. For the conven-
ience of the reader, some conversion factors for forces, stresses, uniform loading, and
moments are provided on a separate page following this Preface. It is noted that through-
out the textbook, conversion factors (for forces, stresses, and dimensions) used in example
problems (when needed) are shown with a smaller font so as to not interfere with the values
of the parameters actually involved in the calculations and to facilitate understanding of the
problem solution.
ACKNOWLEDGMENTS
The authors continue to be indebted to students, colleagues, and other users of the first seven
editions of this book, who have suggested improvements of wording, identified errors, and
recommended items for inclusion or omission. The authors are pleased to acknowledge
the following reviewers, to whom they owe special thanks: Mohammad Azarbayejani,
University of Texas–Pan American; Abdeldjelil Belarbi, University of Houston; Sergio
F. Breña, University of Massachusetts– Amherst; Norbert Delatte, Cleveland State
University; Apostolos Fafitis, Arizona State University; Susan Faraji, University of
Massachusetts Lowell; Catherine French, University of Minnesota; David Garber, Florida
International University; Roberto Leon, Virginia Tech University; John B. Mander, Texas
A&M University; Fatmir Menkulasi, Louisiana Tech University; Gregory K. Michaelson,
Marshall University; Levon Minnetyan, Clarkson University; Ayman M. Okeil, Louisiana
State University; Nima Rahbar, Worcester Polytechnic Institute; Michael Seek, Old
Dominion University; Ahmed Senouci, University of Houston; Lisa Spainhour, FAMU–
Florida State University; Andreas Stavridis, University at Buffalo; Jale Tezcan, Southern
Illinois University–Carbondale; Robin Tuchscherer, Northern Arizona University; Baolin
Wan, Marquette University; Paul Ziehl, University of South Carolina. Their comments and
suggestions have been carefully considered and the results of our review are reflected in
this complete revision.
Users of this eighth edition are urged to communicate with the authors regarding all
aspects of this book, particularly on identification of errors and suggestions for improvement.
We are indebted to late Professors Chu-Kia (CK) Wang and Charles (Chuck) G. Salmon,
who originated this textbook and entrusted us to carry on their legacy. Much of the new and
expanded material presented in this eighth edition would not have been possible without
their work in earlier editions of this book.
Special thanks are due to the Higher Education Group, Oxford University Press—in
particular, Dan Kaveney, Executive Editor, Christine Mahon, Associate Editor, Claudia
Dukeshire, Production Editor, Megan Carlson, Assistant Editor, and Nancy Blaine, former
Senior Acquisitions Editor.
We acknowledge the long-time continuing patience and encouragement from our fami-
lies and especially from our wives, Rebeca and Connie, throughout the preparation of this
edition of the book. Nicole and Gabriel Parra, with their frequent smiles and unbounded
vxi
xxiv P R E F A C E
love, were a continuous source of inspiration to their father. We also owe a special recog-
nition to our parents, Paulina Peña, Hernán Pincheira, Gustavo Parra Pardi and Yolanda
Montesinos Soteldo, who instilled in us from an early age the importance of learning, edu-
cation, and hard work. To all of them we wholeheartedly dedicate this book.
José A. Pincheira
Gustavo J. Parra-Montesinos
xv
ABOUT THE AUTHORS
CHU-KIA WANG* was Professor of Civil Engineering at the University of Wisconsin–

Madison for more than 30 years. A devoted teacher throughout his career, he was the author
or coauthor of many textbooks in the field of structural engineering as the outgrowth of
lectures he prepared for his classes. The University of Wisconsin–Madison recognized
his contribution to the education of future engineers with the College of Engineering’s
Benjamin Smith Reynolds Award for Excellence in Teaching. A fellow and lifetime mem-
ber of the American Society of Civil Engineers, Professor Wang was also a member of
the American Concrete Institute (ACI), the American Society for Engineering Education
(ASCE), and other professional societies.
CHARLES G. SALMON* was Professor Emeritus of Civil Engineering at the University
of Wisconsin–Madison. An accomplished author, educator, researcher, and professional
structural engineer, Professor Salmon received numerous honors in recognition of his
contributions to the field, including the Western Electric Award for excellence in teach-
ing from the American Society for Engineering Education, the University of Wisconsin’s
Emil H. Steiger Distinguished Teaching Award, and the American Concrete Institute’s Joe
W. Kelly and Delmar L. Bloem Awards. He was a long-time member of the ACI Building
Code Committee for Structural Concrete (ACI 318), Committee 340 (Design Aids), and
Committee 435 (Deflections of Concrete Structures). Professor Salmon was also an honor-
ary member of ACI, an honorary member of the American Society of Civil Engineers; and
a life member of the American Society for Engineering Education.
*Deceased
JOSÉ A. PINCHEIRA is Associate Professor of Civil and Environmental Engineering
at the University of Wisconsin–Madison. His main research interests include the behavior
and design of reinforced concrete structural systems subjected to earthquakes, as well as
the seismic rehabilitation of concrete structures. Dr. Pincheira is a fellow of the American
Concrete Institute, a member of subcommittee 318-R (High Strength Steel Reinforcement)
of the Building Code for Structural Concrete (ACI 318), and former member of subcom-
mittee 318-D (Flexure and Axial Loads). He is past chair and current member of ACI
Committee 369, Seismic Repair and Rehabilitation of Concrete Buildings; member of ACI
Committee 374, Performance-Based Seismic Design of Concrete Buildings; and former
member of the Committee on Seismic Rehabilitation of the American Society of Civil
Engineers. Professor Pincheira has received several prestigious awards in recognition of his
contributions to research and teaching, including the CAREER Award from the National
Science Foundation; the James M. Robbins Excellence in Teaching Award from Chi
Epsilon, the Civil Engineering Honor Society; the Martin P. Korn Award from the Precast/
Prestressed Concrete Institute; and the Wason Medal from the American Concrete Institute.
GUSTAVO J. PARRA-MONTESINOS is the C. K. Wang Professor of Structural
Engineering in the Department of Civil and Environmental Engineering at the University
of Wisconsin–Madison. Professor Parra’s main research interests include the behavior
and design of reinforced concrete, fiber-reinforced concrete, and hybrid steel-concrete
structures. He is a member of the ACI Building Code Committee 318 and Chair of its
xvi
xxvi A B O U T T H E AU T H O R S
subcommittee 318-J (Joints and Connections). He is also Chair of ACI-ASCE Committee

352 on Joints in Monolithic Reinforced Concrete Structures. In addition, he is a member
of ACI-ASCE Committee 335 on Composite and Hybrid Structures and of the Fédération
Internationale du Béton (fib) Task Group T4.1 on fiber-reinforced concrete. Professor Parra
has received several prestigious awards for his contributions to research in the field of
reinforced concrete, including the Wason Medal, the Chester Paul Siess Award, and the
Charles S. Whitney Medal from the American Concrete Institute. In addition, he is a recip-
ient of the Arthur J. Boase Award from the ACI Foundation, the Walter L. Huber Research
Prize from the American Society of Civil Engineers, the Shah Family Innovation Prize
from the Earthquake Engineering Research Institute, and the ACI Young Member Award
for Professional Achievement. Professor Parra is also a fellow of the American Concrete
Institute.
xxivi
CONVERSION FACTORS
Some Conversion Factors, between Inch-Pound and SI Units, Useful in Reinforced

Concrete Design
To Convert To Multiply by
Forces kip force kN 4.448
lb N 4.448
kN kip 0.2248
Stresses ksi MPa (i.e., N/mm2) 6.895
psi MPa 0.006895
MPa ksi 0.1450
MPa psi 145.0
Moments ft-kip kN · m 1.356
kN · m ft-kip 0.7376
Uniform Loading kip/ft kN/m 14.59
kN/m kip/ft 0.06852
kip/ft2 kN/m2 47.88
psf N/m2 47.88
kN/m2 kip/ft2 0.02089
Density pcf kg/m3 16.01846
Basis of Conversions: 1 in. = 25.4 mm; 1 lb force = 4.448 newtons.
Basic SI units relating to structural design:
Quantity Unit Symbol

length meter m
mass kilogram kg
time second s
Derived SI units relating to structural design:
Quantity Unit Symbol Formula

force newton N kg · m/s2
pressure, stress pascal Pa N/m2
energy, or work joule J N·m
xxivi
xi
REINFORCED CONCRETE DESIGN
x
CHAPTER 1
INTRODUCTION, MATERIALS,
AND PROPERTIES
1.1 REINFORCED CONCRETE STRUCTURES

The most common materials from which most structures are built are wood, steel, rein-
forced (including prestressed) concrete, and masonry. Lightweight materials such as alumi-
num and advanced composite materials, such as fiber-reinforced polymers (FRP), are also
becoming more common in use. Reinforced concrete is unique in that two materials, rein-
forcing steel and concrete, are used together; thus the principles governing structural design
in reinforced concrete differ in many ways from those involving design in one material.
Many structures are built of reinforced concrete: buildings, bridges, viaducts, retain-
ing walls, tunnels, tanks, conduits, and others. This text deals primarily with fundamental
principles of behavior and design of reinforced concrete members subjected to axial force,
bending moment, shear, torsion, or combinations of these. These principles are applicable
to the design of any structure, as long as information is known about the variation of axial
force, shear, moment, and so on, along the length of each member. Although analysis and
design may be treated separately, they are inseparable in practice, especially in the case of
reinforced concrete structures, which are usually statically indeterminate. In such cases,
reasonable sizes of members are needed in the preliminary analysis that must precede the
final design; so the final conciliation between analysis and design is largely a matter of trial,
judgment, and experience.
Reinforced concrete is a logical union of two materials: plain concrete, which possesses
high compressive strength but little tensile strength, and steel, in the form of bars embed-
ded in the concrete, which can provide the needed strength in tension and ductility to the
member. For instance, the strength and deflection capacity of the beam shown in Fig. 1.1.1
are greatly increased by placing steel bars in the tension zone. Without steel reinforcement,
the beam would undergo a brittle failure once the tensile stress at the bottom of the beam
reached the tensile strength of the concrete. Adding sufficient longitudinal steel reinforce-
ment in the tension zone, however, allows the beam to sustain additional load beyond the
formation of a transverse (flexural) crack. As shown in Fig. 1.1.1, several flexural cracks
will likely develop as the load is increased, providing some degree of warning prior to
failure. Since reinforcement steel is capable of resisting compression as well as tension, it
is also used to provide part of the carrying capacity in reinforced concrete columns, and
frequently in the compression zone of beams to increase ductility and control deflections.
Also, reinforcement is needed transversely to resist shear, to provide lateral support to lon-
gitudinal reinforcement, and to confine the concrete.
2
2 C H A P T E R 1 I N T R O D U C T I O N , M A T E R I A L S , A N D P R O P E R T I E S
Stratosphere Tower, Las Vegas; the tallest free-standing observation tower in the United States,
1149 feet high. A three-legged concrete tower is topped by a ring beam that supports the steel dome,
completed in 1996 (Photo by C. G. Salmon).
Steel and concrete work readily in combination for several reasons: (1) bond (interac-
tion between bars and surrounding hardened concrete) allows transfer of forces between
the two materials; (2) proper concrete mixes provide adequate impermeability of the con-
crete against water intrusion and bar corrosion; and (3) sufficiently similar rates of ther-
mal expansion—that is, 0.0000055 to 0.0000075 for concrete and 0.0000065 for steel per
degree Fahrenheit (ºF), or 0.000010 to 0.000013 for concrete and 0.000012 for steel per
degree Celsius (ºC)—introduce negligible forces between steel and concrete under atmos
pheric changes of temperature.
1.2 HISTORICAL BACKGROUND
Joseph Monier, the owner of an important nursery in Paris, is generally given the credit
for making the first practical use of reinforced concrete. In 1867, Monier recognized
many of its potential uses and successfully undertook to expand the application of the
new method [1.1].1 Prior to his work, however, the method of reinforcing concrete with
1 Numbers in brackets refer to the Selected References at the end of the chapter.
3
1.2 HISTORICAL BACKGROUND 3
P Neutral axis
A Concrete
compression zone
tension
zone Steel bars
A
Cracks Steel bars
Section A–A
Figure 1.1.1 Use of steel bars as tension reinforcement in a reinforced concrete beam.
iron was known and in some cases was protected by patents. Ancient Grecian structures
show that much earlier builders knew something about the reinforcing of stonework for
added strength [1.2].
In the mid- 1800s, Lambot in France constructed and later exhibited at the Paris
Exposition of 1854 a small boat, on which he received a patent in 1855. In Lambot’s pat-
ent was shown a reinforced concrete beam and a column reinforced with four round iron
bars. Another Frenchman, François Coignet, published a book in 1861 describing many
applications and uses of reinforced concrete. In 1854, W. B. Wilkinson of England took out
a patent for a reinforced concrete floor.
Monier acquired his first French patent in 1867 for iron-reinforced concrete tubs. This
was followed by his many other patents, such as for pipes and tanks in 1868, flat plates
in 1869, bridges in 1873, and stairways in 1875. In 1880–1881, Monier received German
patents for innovations that included railroad ties, water feeding troughs, circular flower
pots, flat plates, and irrigation channels. Monier’s iron reinforcement was made mainly to
conform to the contour of the structural element and generally strengthen it. He apparently
had no quantitative knowledge regarding its behavior or any method of making design
calculations [1.1].
In the United States, the pioneering efforts were made by Thaddeus Hyatt, originally
a lawyer, who conducted experiments on reinforced concrete beams in the 1850s. In a
perfectly correct manner, the iron bars in Hyatt’s beams were located in the tension zone,
bent up near the supports, and anchored in the compression zone. Additionally, transverse
reinforcement (known as vertical stirrups) was used near the supports. However, Hyatt’s
experiments were unknown until 1877 when he published his work privately.
Built in 1870, the William Ward house in Port Chester, New York, is generally cred-
ited as the first cast-in-place reinforced concrete structure in the United States [1.3]. E. L.
Ransome, head of the Concrete–Steel Company of San Francisco, apparently used some
form of reinforced concrete in the early 1870s. He continued to increase the application
of wire rope and hoop iron to many structures and was the first to use and patent in 1884
the deformed (twisted) bar. M. K. Hurd has provided an interesting biographical sketch of
Ernest L. Ransome [1.4].
In 1890, Ransome built the Leland Stanford Jr. Museum in San Francisco, a reinforced
concrete building two stories high and 312 ft (95 m) long. Since that time, development
of reinforced concrete in the United States was rapid. During the period 1891–1894, vari-
ous investigators in Europe published theories and test results; among them were Moeller
(Germany), Wunsch (Hungary), Melan (Austria), Hennebique (France), and Emperger
(Hungary), but practical use was less extensive than in the United States.
Throughout the entire period 1850–1900, relatively little was published, because the
engineers working in the reinforced concrete field considered construction and computa-
tional methods to be trade secrets. One of the first publications that might be classified as a
textbook was that of Considère in 1899. By the turn of the century, there was a multiplicity
of systems and methods with little uniformity in design procedures, allowable stresses, and
systems of reinforcing. In 1903, with the formation in the United States of a joint commit-
tee of representatives of all organizations interested in reinforced concrete, uniform appli-
cation of knowledge to design was initiated. The development of standard specifications is
discussed in Chapter 2.
The earliest textbook in English was that of Turneaure and Maurer [1.5], published
in 1907. In the first decade of the twentieth century, progress in reinforced concrete was
4
rapid. Extensive testing to determine beam behavior, compressive strength of concrete, and
modulus of elasticity was conducted by Arthur N. Talbot at the University of Illinois, by
Frederick E. Turneaure and Morton O. Withey at the University of Wisconsin, and by Bach
in Germany, among others. From about 1916 to the mid-1930s, research centered on axially
loaded column behavior. In the late 1930s and 1940s, eccentrically loaded columns, foot-
ings, and the ultimate strength of beams received special attention.
Since the mid-1950s, reinforced concrete design practice has made the transition from
that based on elastic methods to one based on strength. Prestressed concrete (Chapter 20),
wherein the steel reinforcement is stressed in tension and the concrete is in compression
even before external loads are applied, has advanced from an experimental technique to a
major structural composite material. There has been a transition from cast-in-place rein-
forced concrete to elements precast at a manufacturer’s plant and shipped to the job site
for assembly. A summary of concrete building construction in the United States is given
by Cohen [1.3].
Understanding of reinforced concrete behavior is still far from complete; building codes
and specifications that give design procedures are continually changing to reflect latest
knowledge.
1.3 CONCRETE
Plain concrete is made by mixing cement, fine aggregate, coarse aggregate, water, and
frequently admixtures (see Fig. 1.3.1). When reinforcing steel is placed in the forms
and wet concrete mix is placed around it, the final solidified mass becomes reinforced
concrete. The strength of concrete depends on many factors: notably the proportion of
the ingredients and the conditions of temperature and moisture under which it is placed
and cured.
Subsequent sections contain brief discussions of the materials in and the properties
of plain concrete. The treatment is intended to be only introductory; an interested reader
should consult standard references devoted entirely to the subject of plain concrete [1.6–1.8].
Figure 1.3.1 Cross section of concrete. Cement-and-water paste coats each aggregate particle and
fills space between particles. (Photo by José A. Pincheira.)
5
1.4 CEMENT 5
1.4 CEMENT
Cement is a material that has adhesive and cohesive properties enabling it to bond mineral frag-
ments into a solid mass. Although this definition can apply to many materials, the cements of
interest for reinforced concrete construction are those that can set and harden in the presence
of water—the so-called hydraulic cements. These consist primarily of silicates and aluminates
of lime made from limestone and clay (or shale) which is ground, blended, fused in a kiln,
and crushed to a powder. Such cements chemically combine with water (hydrate) to form a
hardened mass. The usual hydraulic cement used for reinforced concrete is known as port-
land cement because of its resemblance when hardened to Portland stone found near Dorset,
England. The name originated in a patent obtained by Joseph Aspdin of Leeds, England,
in 1824.
Concrete made with portland cement ordinarily requires several days to attain strength ade-
quate to allow forms to be removed and construction and dead loads carried. The design or
specified compressive strength of such concrete is typically assumed to be reached at about
28 days. This ordinary portland cement is identified by ASTM (American Society for Testing
and Materials) C150/C150M [1.9] as Type I. Other types of portland cement and their intended
uses are given in Table 1.4.1. ACI Committee 225 provides a guide for selection and use of
hydraulic cements [1.10].
There are also several categories of blended hydraulic cements (ASTM C595/C595M
[1.11]), such as portland blast-furnace slag cement (Type IS), portland-pozzolan cement (Type
IP), portland-limestone cement (Type IL), and ternary blended cement (Type IT). Ternary
blended cements are defined in ASTM C595 as those “consisting of portland cement with
either a combination of two different pozzolans, slag cement and a pozzolan, a pozzolan and
a limestone, or a slag cement and a limestone.” Pozzolan is a finely divided siliceous or sili-
ceous and aluminous material that possesses little or no inherent cementitious property; in the
powdery form and in the presence of moisture, however, it will chemically react with calcium
hydroxide at ordinary temperatures to form compounds possessing cementitious properties.
Portland blast-furnace slag cement has lower heat of hydration than ordinary Type I
cement and is useful for mass concrete structures such as dams; and because of its high sul-
fate resistance, it is used in seawater construction. Portland-pozzolan cement is a blended
mixture of ordinary Type I cement with pozzolan. Blended cements with pozzolan gain
strength more slowly than cements without pozzolan; hence they produce less heat during
hydration, and thus are widely used in mass concrete construction.
Air-entraining portland cement contains a chemical admixture finely ground with the
cement to produce intentionally air bubbles on the order of 0.002 in. (0.05 mm) diameter
uniformly distributed throughout the concrete. Such air entrainment will give the concrete
improved durability against frost action, as well as better workability. Air-entraining port-
land cement for Types I, II, and III, given in Table 1.4.1, is designated IA, IIA, or IIIA.
Air-entraining agents may also be added to the blended hydraulic cements in ASTM C595/
C595M [1.11] at the time the concrete is mixed.
TABLE 1.4.1 TYPES OF PORTLAND CEMENTa
Type Uses
I Ordinary construction where special properties are not required

II Ordinary construction when moderate sulfate resistance or moderate heat of
hydration is desired
III When high early strength is desired; has considerably higher heat of hydration
than Type I cement
IV When low heat of hydration is desired
V When high sulfate resistance is desired
a According to ASTM C150/C150M [1.9].
6
1.5 AGGREGATES
Since aggregate usually occupies about 75% of the total volume of concrete, its properties
have a definite influence on the behavior of hardened concrete. Not only does the strength
and stiffness of the aggregate affect the strength and stiffness of the concrete, its properties
also greatly affect durability (resistance to deterioration under freeze-thaw cycles). Since
aggregate is less expensive than cement, it is logical to use the largest percentage feasible.
In general, for maximum strength, durability, and best economy, the aggregate should be
packed and cemented as densely as possible. Hence, aggregates are usually graded by size
and a proper mix specifies percentages of both fine and coarse aggregates.
Fine aggregate (sand) is any material passing through a No. 4 sieve2 [i.e., less than
about 3 16 -in. (5-mm) diameter]. Coarse aggregate (gravel) is any material of larger size.
The nominal maximum size of coarse aggregate permitted (ACI-26.4.2.1)3 is governed by
the clearances between sides of forms and between adjacent bars and may not exceed (a) 1 5
the narrowest dimension between sides of forms, nor (b) 1 3 the depth of slabs, nor
(c) 3 4 the minimum clear spacing between individual reinforcing bars. Additional informa-
tion concerning aggregate selection and use is to be found in a report of ACI Committee
221 [1.13].
Natural stone aggregates conforming to ASTM C33 [1.14] are used in the majority of
concrete construction, giving a unit weight for such concrete of about 145 pcf (pounds per
cubic foot) or 2320 kg/m3 (kilograms per cubic meter). When steel reinforcement is added,
the unit weight of normal-weight reinforced concrete is taken for calculation purposes as
150 pcf or 2400 kg/m3. Actual weights for concrete and steel are rarely, if ever, computed
separately. For special purposes, lightweight or extra-heavy aggregates are used.
Structural lightweight concretes are usually made from aggregates conforming to ASTM
C330 [1.15] which are produced artificially in a kiln, such as expanded clays and shales.
The unit weight of such concretes typically ranges from 70 to 115 pcf (1120–1840 kg/m3)
(see Fig. 1.5.1). Lightweight concretes ranging down to 30 pcf (480 kg/m3), often known as
cellular concretes, are also used for insulating purposes and for masonry units. When light-
weight materials are used for both coarse and fine aggregates in structural concrete, it is
termed all-lightweight concrete. When only the coarse aggregate is of lightweight material
but normal weight sand is used for the fine aggregate, it is said to be sand-lightweight con-
crete. Often the term “sand replacement” is used in connection with lightweight concrete.
This refers to replacing all or part of the lightweight aggregate fines with natural sand.
Steiger [1.16] provides historical background for the use of lightweight aggregate concrete,
Mackie [1.17] has discussed uses of lightweight concrete, and ACI Committee 213 has a
guide for the use of structural lightweight aggregate concrete [1.18].
Heavyweight, high-density concrete is used for shielding against gamma and X radia-
tion in nuclear reactor containers and other structures [1.19]. Naturally occurring iron ores,
titaniferous iron ores, “hydrous iron ores” (i.e., containing bound and adsorbed water), and
barites are crushed to suitable size for use as aggregates. Heavyweight concretes typically
weigh from 200 to 350 pcf (3200–5600 kg/m3).
1.6 ADMIXTURES
In addition to cement, coarse and fine aggregates, and water, materials known as admix-
tures, mineral or chemical, are often added to the concrete mix immediately before or
during the mixing. Admixtures are used to modify the properties of the concrete to make it
better serve its intended use or for better economy.
2 4.75 mm according to ASTM Standard E11.

3 Numbers refer to sections in the “ACI Code,” officially 318-14, Building Code Requirements for Structural Concrete [1.12].
7
1.6 ADMIXTURES 7
Low-density Moderate strength Structural

concrete concrete concrete
Expanded slag
Sintering grate expanded

shale, clay, or fly ash
Rotary kiln expanded
shale, clay, and slate
Scoria
Pumice
Perlite
Vermiculite
kgf/m3 400 600 800 1000 1200 1400 1600 1800 kgf/m3
pcf 20 40 60 80 100 120 pcf

28-day air dry unit weight
Figure 1.5.1 Approximate unit weight and use classification of lightweight aggregate concrete. (From
Ref. 1.114.)
Mineral Admixtures
Mineral admixtures are finely divided materials including pozzolans such as fly ash, cement
slag, and silica fume. These admixtures are often used as cement replacement, but can also
be used in addition to cement, as replacement of sand. In general, mineral admixtures
reduce the heat of hydration and improve workability and durability. General information
about mineral admixtures is available in [1.7] and [1.20]. Mielenz has given a history and
background on mineral admixtures [1.21].
Fly ash
Fly ash is a by-product from the combustion of coal in power plants. Fly ash used in
concrete shall meet ASTM C618 [1.22]. Since its cost is substantially lower than that of
cement, it is typically used as cement replacement for economic reasons. The use of fly
ash in concrete improves workability, reduces the heat of hydration, and increases dura-
bility. Albinger [1.23] provides general information on when to use and what to expect
from fly ash concrete, and Ravina [1.24] discusses slump retention of fly ash concrete (see
Section 1.7) with and without chemical admixtures.
Cement Slag
Blast-furnace slag is a by-product of iron production. This by-product is first granulated
and then ground to achieve particle sizes similar to those of cement. For use as mineral
admixture in concrete, slag shall meet ASTM C989 [1.25]. Cement slag in concrete reduces
the heat of hydration and provides increased durability by decreasing concrete permeabil-
ity and increasing resistance to sulfate attacks. Although the strength gain of concrete with
cement slag is lower in the first few days, compressive strength after that is typically greater
than or comparable to that of concrete without slag cement. Additional information on the
use of cement slag in concrete can be found in a report by ACI Committee 233 [1.26].
Silica Fume
Silica fume is the finely divided solid-microsilica material collected from the fumes of elec-
tric furnaces that produce ferrosilicon or silicon metal. When used in concrete, it shall con-
form to ASTM C1240 [1.27]. In addition to its use as a pozzolan, silica fume in the concrete
mix produces a more impermeable concrete, able to resist chloride intrusion into concrete
exposed to deicing chemicals. Silica fume is an important admixture in high-performance
8
concrete to achieve high strength and excellent durability. Further information about silica
fume is available from Cohen, Olek, and Mather [1.28] and Durning and Hicks [1.29].
Chemical Admixtures
A wide variety of chemical admixtures for concrete are available, most commonly air-
entraining, water-reducing, and set-controlling admixtures. A history and background on
chemical admixtures can be found in Durning and Hicks [1.30]. A report by ACI Committee
212 [1.31] provides an essential guidance for use of chemical admixtures. In this report,
chemical admixtures are classified into categories that include air-entraining admixtures,
accelerating admixtures, water-reducing and set-retarding admixtures, admixtures for flow-
ing concrete, and admixtures for self-consolidating concrete. A brief discussion on these
chemical admixtures follows.
Air-Entraining Admixtures
These chemicals, meeting the requirements of ASTM C260 [1.32], can be added either
to the hydraulic cement or as an admixture to the concrete mix. The chemical causes air in
the form of minute bubbles (often 0.004 in. or 0.1 mm diameter or smaller) to be dispersed
throughout the concrete mix, with the purpose of increasing workability and resistance to
deterioration that results from both freeze-thaw action and ice-removal salts.
Air-entraining admixtures are probably the most widely used type of chemical admix-
ture. In addition to resistance to freeze-thaw cycles and to the corrosiveness of deicing
chemicals, air entrainment improves plasticity and workability, permitting a reduction
in water content. Uniformity of placement with little bleeding and segregation can be
achieved. In addition, air-entrained concrete is more watertight and increases resistance to
sulfate action. For exposed concrete, the possible reduction in strength (approximately 5%
for each percent of entrained air) is far less important than the improved durability in terms
of resistance to freeze-thaw action and deicing chemicals.
Accelerating Admixtures
Accelerating admixtures modify the properties of concrete, particularly in cold weather,
to (a) accelerate the rate of early-age strength development; (b) reduce the time required
for proper curing and protection; and (c) permit earlier start of finishing operations.
Accelerators must not be used as antifreeze agents for concrete. Accelerators must meet the
requirements of Type C or E in ASTM C494 [1.33]; calcium chloride, the best known and
most common accelerator, must also meet the requirements of ASTM D98, Specifications
for Calcium Chloride. Limits on water-soluble chloride, however, are specified in ACI 318
(ACI-19.3.2.1) to reduce potential for corrosion of reinforcement.
Water-Reducing and Set-Retarding Admixtures

Water-reducing admixtures are used to reduce the amount of water required for a given
slump or, when used without water reduction, to increase concrete workability. Some
of these admixtures also increase the setting time for concrete. Water-reducing and set-
retarding admixtures must meet the requirements of ASTM C494 [1.33], where they are
classified as water-reducing admixtures (Type A); retarding admixtures (Type B); water-
reducing and retarding admixtures (Type D); water-reducing and accelerating admixtures
(Type E); water-reducing, high-range admixtures (Type F); and water reducing, high-range,
and retarding admixtures (Type G). Water-reducing, high-range admixtures are sometimes
referred to as “superplasticizers,” meaning that the quantity of mixing water is reduced by
12% or more. The last two classifications (Type F and Type G) are also covered by ASTM
C1017 [1.34].
A report by ACI Committee 212 [1.31] lists seven general categories for materials used
as water-reducing admixtures, including lignosulfonic acids and their salts, hydroxylated
carboxylic acids and their salts, carbohydrate-based compounds and polysaccharides, and
9
1.7 COMPRESSIVE STRENGTH 9
polycarboxylates. Materials used for water-reducing, high-range admixtures, on the other

hand, include sulfonated naphthalene condensates, sulfonated melamine condensates,
modified lignosulfonates, and polycarboxylates. Ramezanianpour, Sivasundaram, and
Malhotra [1.35] have discussed superplasticizers and their effect on strength properties
of concrete.
Set-retarding admixtures are used primarily to offset the accelerating and damaging
effect of high temperature, to keep concrete workable during placement, and to minimize
form-deflection cracks. A variety of water-reducing admixtures also serve as set-retarding
admixtures.
Admixtures for Flowing Concrete

Flowing concrete is defined [1.34] as “concrete that is characterized as having a slump
greater than 7 in. (190 mm) while maintaining a cohesive nature … .” These admixtures
are classified by ASTM C1017 [1.34] into two types: Type I—Plasticizing, and Type II—
Plasticizing and Retarding. Flowing concrete is commonly used where high rates of casting
are required or in highly reinforcement-congested members [1.31].
Admixtures for Self-Consolidating Concrete

Self-consolidating concrete is highly flowable concrete that requires no vibration. Given
its high flowability, a flow slump, rather than slump, is measured in self-consolidating con-
cretes. In general, flow slumps between 22 and 30 in. (550 and 750 mm) are associated with
these concretes. Two primary types of chemical admixtures are used in self-consolidating
concrete, high-range, water-reducing admixtures (typically polycarboxylate based) and
viscosity-modifying admixtures (polymer based or in the form of fine particles). High-
range, water-reducing admixtures are used to lower the yield stress of the material, while
viscosity-modifying admixtures serve to increase cohesion and plastic viscosity when a
concrete with low yield stress and high plastic viscosity is desired [1.31]. More information
about self-consolidating concrete can be found in a 2007 ACI publication [1.36].
Besides the admixture categories listed above, ACI Committee 212 [1.31] lists the fol-
lowing categories for chemical admixtures: cold-weather admixture systems, admixtures
for very high-early strength concrete, extended set-control admixtures, shrinkage-reducing
admixtures, corrosion- inhibiting admixtures, lithium admixtures to reduce deleterious
expansion from alkali-silica reaction, permeability-reducing admixtures, and miscellaneous
admixtures. Miscellaneous admixtures are categorized as bonding admixtures; coloring
admixtures; flocculating admixtures; fungicidal, germicidal, and insecticidal admixtures;
rheology-and viscosity-modifying admixtures; and air-detraining admixtures. Details are
available in the ACI Committee 212 Report [1.31].
1.7 COMPRESSIVE STRENGTH
The strength of concrete is primarily controlled by the proportioning of cement, coarse
and fine aggregates, water, and various admixtures. In reinforced concrete design, “con-
crete strength” means uniaxial compressive strength measured by a compression test, typ-
ically of a standard test cylinder. The most important variable in determining concrete
strength is the water to cement (w/c) ratio, as shown in Fig. 1.7.1. The lower the water/
cement ratio, the higher the compressive strength. This relationship has been recognized
since the 1920s.
In the past decades, with the increasing use of admixtures, many of which contain cemen
titious materials, researchers have confirmed that any cementitious admixtures should be
included with the cement in determining the proper mix to obtain a specified strength. This
is recognized in ACI 318, where a water to cementitious ratio (w/cm) shall be calculated,
including the weight of fly ash and other pozzolans (ASTM C618 [1.22]), slag cement (ASTM
C989 [1.25]), and silica fume (ASTM C1240 [1.27]). Popovics [1.37] and Popovics and
10
50
7000 500
For Type I
portland cement 45
450
6000
Compressive strength, psi

40
400
N
on
-a 35
ir-
kgf/cm2
5000 en 350
Ai
MPa
r-e tra
nt ine
ra dc 300 30
in on
4000 ed c re
co te
nc
ret 250 25
e
3000 200 20
15
4 5 6 7 8
Water–cement ratio, gallons per bag
0.4 0.5 0.6 0.7

Water–cement ratio, by weight
Figure 1.7.1 Effect of water-to-cement ratio on 28-day compressive strength. Average values for
concrete containing 1.5 to 2% trapped air for non-air-entrained concrete and no more than 5 to 6%
air for air-entrained concrete. (Curves drawn from data in Ref. 1.42, Table 6.3.4a.)
Popovics [1.38, 1.39] have reviewed the validity of strength based on the water to cemen-
titious material ratio.
A certain minimum amount of water is necessary for the proper chemical action in the
hardening of concrete; extra water increases the workability (the ease of concrete flow) but
reduces strength. A measure of the workability is obtained by a slump test. A truncated
cone-shaped metal mold 12 in. (300 mm) high is filled with fresh concrete, the mold is then
lifted off, and a measurement is made of the distance to the top of the wet mass “slumps”
from its position before the mold was removed. The smaller the slump, the stiffer and less
workable the mix. In building construction, a 3-to 4-in. (75-to 100-mm) slump is common.
Vibration of the concrete mix will greatly improve workability and even very stiff no-slump
concrete can be placed [1.40].
Proportioning of concrete mixes can be done in accordance with Design and Control of
Concrete Mixtures [1.41], as well as ACI Standard 211.1 for normal-weight, heavyweight,
and mass concrete [1.42], ACI Standard 211.2 for structural lightweight concrete [1.43],
and ACI Standard 211.3 for no-slump concrete [1.40]. Strength of concrete in place in the
structure is also greatly affected by quality control procedures for placement and inspec-
tion. Details regarding good practice are available in ACI Standard 304 [1.44] and in the
ACI Manual of Concrete Inspection [1.45].
Durability, long recognized as an important quality of concrete, is related to the w/cm
ratio and compressive strength, among other factors. Durability requirements for con-
crete can be found in Chapter 19 of the ACI Code (Concrete: Design and Durability
Requirements). A source for obtaining durable concrete is the Committee 201 Guide to
Durable Concrete [1.46].
The strength of concrete is denoted in the United States by fc′, which is the compres-
sive strength in psi of test cylinders with diameter and height of either 6 in. (150 mm) and
12 in. (300 mm) or 4 in. (100 mm) and 8 in. (200 mm), typically measured at 28 days after
casting. In many parts of the world, the standard test unit is the cube, frequently measuring
8 in. (200 mm) to a side.
Since nearly all reinforced concrete behavior is related to the standard 28-day compres-
sive strength fc′, it is important to note that such strength depends on the size and shape of
the test specimen and the manner of testing [1.47]. Properties such as tensile strength of
concrete and size of contact area of the testing machine have more effect on cube strength
than on cylinder strength. As an average, the 6 × 12 in. (150 × 300 mm) cylinder strength
1
1.7 COMPRESSIVE STRENGTH 11
is 80% of the 6-in. (150-mm) cube strength and 83% of the 8-in. (200-mm) cube strength
[1.48]. For lightweight concrete, cylinder strength and cube strength are nearly equal.
Given the effect on cylinder compressive strength of numerous test variables (e.g., load
rate, specimen dimensions, casting and curing conditions), it is clear that such strength will
differ from the in-place concrete strength in a structure. Results from a thorough investi-
gation of in-place versus molded cylinder concrete compressive strengths were reported
by Bloem [1.49]. Compressive strengths obtained from tests of drilled cores were less
than those of cylinders. Compared with strengths of field-cured cylinders, the compres-
sive strength of drilled cores averaged between 11 and 21% less, depending on curing
conditions.
An interesting discussion of the cylinder test is given by Shilstone [1.50], and Tait [1.51]
has discussed the use of test results. When an assessment of the strength of in-place con-
crete is desired, procedures ranging from tests of cylindrical cores cut from the structure to
the use of nondestructive tests [1.52–1.58] are available. ACI Committee 214 [1.59] has a
recommended practice for evaluating strength test results from concrete cores.
Stress-Strain Relationship
The stress-strain relationship for concrete depends on its strength, age at loading, rate of
loading, aggregates and cement properties, and type and size of specimens [1.60, 1.61].
Typical curves for specimens (6 × 12 in. cylinders) loaded in compression at 28 days using
normal testing speeds are shown in Fig. 1.7.2. The rate of applying strain during testing
influences the shape of the stress-strain curve, as shown in Fig. 1.7.3, particularly the por-
tion after the maximum stress has been reached.
Note from Fig. 1.7.2 that lower-strength concrete has greater deformability (ductility)
than higher-strength concrete, as evidenced by the length and smaller slope of the descend-
ing portion of the curve after the maximum stress has been reached at a strain between
0.002 and 0.0025. Ultimate strain at crushing of concrete often varies from 0.003 to as high
as 0.008.
12
si
800 80
0p
11
,00
12
10,000
10 700 70
fc’ =
Concrete compressive stress, fc, ksi
9
600 60
8000
8
7 500 50
kgf/cm2
MPa
6 6000
400 40
5000
5
4000 300 30
4
3000
3 20
200
2
100 10
1
0 0
0.001 0.002 0.003 0.004
Strain, in./in. (mm/mm)
Figure 1.7.2 Typical stress-strain curves for concrete in compression under short-time loading.
(Curves represent a compromise adapted from curves and results given by Wang, Shah, and Naaman
[1.61], Bertero [1.62], Naaman [1.63], and Nilson [1.64].)
12
1.00 Strain rate 0.001 per 100 days
Ratio of concrete stress to

0.0
01
cylinder strength
0.75 0. per
00 day
0. 1
00 pe
1 rh
0.50 pe ou
Cylinder strength rm r
fc’ = 3000 psi in
ut
0.25 e
at 56 days
0.00
0.001 0.002 0.003 0.004 0.005 0.006 0.007
Concrete compressive strain, εc
Figure 1.7.3 Stress-strain curves for various strain rates of concentric loading. (From Rüsch [1.65].)
In usual reinforced concrete design, specified concrete strengths fc′ of 3500 to 5000 psi
(24–35 MPa) are used for nonprestressed structures, and 5000 to 8000 psi (35–56 MPa) are
used for prestressed concrete. For special situations, particularly in columns of tall build-
ings, concretes ranging from 6000 to 14,000 psi (42–97 MPa) have been used [1.66–1.68].
On the Pacific First Center in Seattle, the specified strength was 14,000 psi (97 MPa) at
56 days [1.66]. The average strength obtained throughout the project was about 18,000
psi (124 MPa). Research continues on high-strength concrete (often referred to as “high-
performance” concrete) because in addition to high strength, the concrete must have other
excellent characteristics [1.69, 1.70].
1.8 TENSILE STRENGTH
The strength of concrete in tension is an important property that greatly affects the extent
and size of cracking in structures. Tensile strength is usually determined by using the split-
cylinder test in accordance with ASTM C496/C496M [1.71]. In this test, the same size
cylinder used for the compression test is placed on its side in the testing machine so that the
compression load P is applied uniformly along the length of the cylinder in the direction of
the diameter. The cylinder will split in half when the tensile strength is reached. The stress
is computed by 2P/[π(diameter)(length)] based on the theory of elasticity for a homogene-
ous material in a biaxial state of stress.4 Tensile strength is a more variable property than
compressive strength and is about 10 to 15% of it. The split-cylinder tensile strength fct has
been found to be proportional to fc′ ,5 such that
fct = 6 fc′ to 7 fc′ psi for normal-weight conrete 6
fct = 5 fc′ to 6 fc′ psi for lightweight conrete 6
In the ACI Code, fct = 6.7 fc′ psi is assumed for normal-weight concrete. The lower split-
cylinder tensile strength exhibited by lightweight concrete compared to normal-weight
concrete is accounted for in the ACI Code through the use of a factor λ that multiples the
4 See, for example, S. Timoshenko and J. N. Goodier, Theory of Elasticity, 3rd ed., McGraw-Hill, 1970, pp. 122–123.
5 fc′ is in psi units; thus fc′ = 3000 psi, fc′ = 54.8 psi. When fc′ is in kilogram-force per square centimeter (kgf/cm2), the
constant in front of fc′ is to be multiplied by 0.265; when fc′ is in newtons per square millimeter, that is, megapascals (MPa),
the constant in front of fc′ is to be multiplied by 0.083.
6 In SI, with fc′ and fct in MPa,
fct = 0.5 fc′ to 0.6 fc′ for normal-weight conrete

fct = 0.4 fc′ to 0.5 fc′ for lightweight conrete
13
1.8 TENSILE STRENGTH 13
strength associated with normal-weight concrete (see, e.g. Lightweight Concrete in Section
5.5). This factor λ may be determined from results of split-cylinder tests as (ACI-19.2.4.3)
fct
λ= ≤ 1.0 (1.8.1)
6.7 fcm
where fcm is the measured average compressive strength of the concrete. Alternatively to the
use of results from split-cylinder tests, λ may be taken as 0.75 for all-lightweight concrete
( )
i.e., fct = 5 fc′ and 0.85 for sand-lightweight concrete i.e., fct = 5.7 fc′ (ACI-19.2.4.2). ( )
For lightweight-fine blend concrete, λ ranges between 0.75 and 0.85 depending on the
absolute volume of normal-weight fine aggregates as a fraction of the total absolute volume
of fine aggregates. Similarly, for sand-lightweight, course blend concrete, λ ranges between
0.85 and 1.0 based on the absolute volume of normal-weight course aggregate as a fraction
of the total absolute volume of course aggregate.
Tensile strength in flexure, known as modulus of rupture, measured in accordance with
ASTM C78 [1.72], is also important when considering cracking and deflection of beams.
The modulus of rupture fr , computed from the flexure formula f = Mc / I , gives higher
values for tensile strength than the split-cylinder test, primarily because the concrete com-
pressive stress distribution is not linear when tensile failure is imminent, as is assumed in
the computation of the nominal Mc / I stress. It is generally accepted (ACI-19.2.3.1) that
an average value for the modulus of rupture fr may be taken as 7.5λ fc′ 0.62 λ fc′ MPa , ( )
where λ = 1.0 for normal-weight concrete, while λ = 0.85 and 0.75 for sand-lightweight
concrete and all-lightweight concrete, respectively, as discussed above. Because of the
large variability in modulus of rupture, as shown in Fig. 1.8.1, the selection of the coeffi-
cient 7.5, or even the entire expression 7.5λ fc′, should be viewed as a practical choice for
design purposes.
One may note that neither the split-cylinder nor the modulus of rupture tensile strength
is correctly a measure of the strength under axial tension. However, axial tension is difficult
to measure accurately and, when compared with the modulus of rupture or split-cylinder
strength, it does not give better correlation with tension-related failure behavior such as
inclined cracking from shear and torsion or splitting from interaction of reinforcing bars
with surrounding concrete.
f’c (MPa)
10 20 30 40 50 60
1000
900
6.0
800
Modulus of rupture, fr (psi)
5.0
700
600
fr (MPa)
4.0
500
3.0
400
fr = 7.5 f’c
(ACI-19.2.3.1)
300 2.0
200
1000 3000 5000 7000 9000

Concrete compressive strength, f’c (psi)
Figure 1.8.1 Comparison of test results for modulus of rupture of normal-weight concrete with
ACI Code expression. (Adapted from Mirza, Hatzinikolas, and MacGregor [1.73].)
14
1.9 BIAXIAL AND TRIAXIAL STRENGTH

Concrete is seldom subjected to only uniaxial compressive or tensile stress. For example,
the presence of shear in flexural members generates biaxial stresses in the concrete and
the restraint against lateral expansion provided by transverse reinforcement in members
under axial compression leads to triaxial compressive stresses. Fig. 1.9.1 shows a biaxial
stress interaction diagram from Ref. 1.74. Under biaxial compression, concrete compres-
sive strength is greater than the uniaxial compressive strength, being approximately 30%
higher on average for a 1:0.5 ratio of biaxial compressive stresses. Under combined tension
and compression, the strength interaction is approximately linear with strengths lower than
both the uniaxial compressive and tensile strength. Under equal biaxial tension, concrete
strength is approximately equal to the uniaxial tensile strength.
Compressive strength and deformation capacity are greatly increased by the presence
of lateral compressive stresses. From tests of cylinders subjected to triaxial compression
with the largest stress applied in the longitudinal direction, Richart, Brandtzaeg, and Brown
[1.75] showed that concrete strength increases at a rate of approximately 4.1 times the
magnitude of the lateral (confining) stress. Strain capacity also increased with an increase
in lateral pressure, with strain at peak stress ranging between 0.5 and 7% for the range of
lateral pressures considered (Fig. 1.9.2).
1.10 MODULUS OF ELASTICITY
The modulus of elasticity of concrete varies, unlike that of steel, with strength. It also
depends, though to a much lesser extent, on the age of the concrete, the properties of
the aggregates and cement, the rate of loading, and the type and size of the specimen.
Furthermore, since concrete exhibits some permanent set even under small loads, there are
various definitions of the modulus of elasticity.
Figure 1.10.1 represents a typical stress-strain curve for concrete in compression. In the
figure, the initial modulus (tangent at origin), the tangent modulus (at 0.5 fc′ ), and the secant
modulus (also at 0.5 fc′ ) are noted. Usually the secant modulus at 25 to 50% of the compres-
sive strength fc′ is considered to be the modulus of elasticity. For many years the modulus
was approximated adequately as 1000 fc′ by the ACI Code; but with the rapidly increasing
fc
(psi)
fc2/fc’ fc’ = 3660 psi
1.2 20,000 fℓ = 4090 psi
1.0 16,000
fc2 P
fc1
0.8 fℓ = 2010 psi
12,000
fℓ fℓ
0.6 f c2
=
f c1 8000 fℓ = 1090 psi
0.4 P
fℓ = 550 psi
4000
0.2
fc1/fc’ 0.01 0.02 0.03 0.04 0.05 0.06 εc

0.2 0.4 0.6 0.8 1.0 1.2
Figure 1.9.1 Strength of concrete under biaxial stress. Figure 1.9.2 Stress-strain response of concrete
(Adapted from Fig. 6 in Ref. 1.74.) under triaxial compression. (Adapted from Fig. 23 in
Ref. 1.75.)
15
1.10 MODULUS OF ELASTICITY 15
fc’
Concrete compressive stress

Tangent modulus at 0.5 fc’
Ultimate strain
typically varies
Initial modulus (tangent at origin) from 0.003 to
0.5fc’ 0.004
Secant modulus at 0.5 fc’
0 0.001 0.002 0.003 0.004

Concrete strain
Figure 1.10.1 Definitions of the modulus of elasticity for concrete in compression. Values of the
modulus of elasticity for various concrete strengths appear in Table 1.10.1.
TABLE 1.10.1 VALUES OF MODULUS OF ELASTICITY (USING EC = 33

w1.5
c f c′ FOR NORMAL-WEIGHT CONCRETE WEIGHING 145 PCF)
Inch-Pound Units SI Unitsb
fc′ (psi) Ec (psi) fc′ (MPa) Ec† (MPa)
3000 3,150,000 21a 21,500

3500 3,400,000 24 23,000
4000 3,640,000 28 24,900
4500 3,860,000 31 26,200
5000 4,070,000 35 27,800
6000 4,460,000 41 30,100
8000 5,150,000 55 34,900
a These metric values are rounded values approximating concrete strengths in Inch-Pound units; actual equivalents for 3000,
3500, 4000, 4500, 5000, 6000, and 8000 psi are 20.7, 24.1, 27.6, 31.0, 34.5, 41.3, and 55.1 MPa, respectively.
b Multiply MPa values by 10.2 to obtain kgf/cm2.
† Using Ec = 4700 f c′ as per ACI 318-14M.
use of lightweight concrete, the variable of density needed to be included. As a result of a

statistical analysis of available data, the empirical formula given by ACI-19.2.2.1
Ec = 33w1c .5 fc′ (1.10.1)7
was developed [1.76] for values of wc between 90 and 155 pcf. Equation (1.10.1) is represen-
tative of the secant modulus at a compressive stress of 0.45 fc′ . Reviews of the applicability
of Eq. (1.10.1) have been made by Shih, Lee, and Chang [1.77] and also Oluokun, Burdette,
and Deatherage [1.78]. For normal-weight concrete weighing 145 pcf, Eq. (1.10.1) gives
Ec = 57,600 fc′ . For normal-weight concrete, ACI-19.2.2.1 suggests
Ec = 57, 000 fc′ (1.10.2)8
7 For SI, with wc in kg/m3 and Ec and fc′ in MPa,

Ec = 0.043w15
c fc′ (ACI 318-14M) (1.10.1)
8 For SI, Ec and fc′ in MPa,
Ec = 4700 fc′ (ACI 318-14M) (1.10.2)
and with Ec and fc′ in kgf /cm2.
Ec = 15, 000 fc′ (approximate)
16
1.11 CREEP AND SHRINKAGE

Creep and shrinkage are time-dependent deformations that, along with cracking, cause a
great concern for the designer because of the inaccuracies and unknowns that surround
them. Concrete may behave essentially elastic only under loads of short duration; and,
because of additional deformation with time, the effective behavior is that of an inelastic
material. Deflection after a long period of time is therefore difficult to predict, but its con-
trol is needed to assure serviceability during the life of the structure.
Creep
Creep is the property of concrete (and other materials) by which it continues to deform with
time under sustained loads at unit stresses within the accepted elastic range (say, below
0.5 fc′. This inelastic deformation increases at a decreasing rate during the time of loading,
and its total magnitude may be several times as large as the short-time elastic deformation.
Frequently, creep is associated with shrinkage, since the two occur simultaneously and
often provide the same net effect: increased deformation with time. As may be noted by
the general relationship of deformation versus time in Fig. 1.11.1, the “true elastic strain”
decreases, since the modulus of elasticity Ec is a function of concrete strength fc′, which
increases with time.
Although creep is separate from shrinkage, it is related to it. Detailed information is
available for estimating creep [1.79, 1.80]. The internal mechanism of creep, or “plastic
flow” as it is sometimes called, may be due to any one or a combination of the follow-
ing: (1) crystalline flow in the aggregate and hardened cement paste, (2) plastic flow of the
cement paste surrounding the aggregate, (3) closing of internal voids, and (4) the flow of
water out of the cement gel due to external load and drying.
Factors affecting the magnitude of creep are (1) the constituents—such as the compo-
sition and fineness of the cement, the admixtures, and the size, grading, and mineral con-
tent of the aggregates; (2) proportions, such as water content and the water/cement ratio;
(3) curing temperature and humidity; (4) relative humidity during period of use; (5) age at
loading; (6) duration of loading; (7) magnitude of stress; (8) surface to volume ratio of the
member; and (9) slump.
Accurate prediction of creep is complicated because of the variables involved; however,
a general prediction method developed by Branson [1.80] gives a standard creep coefficient
equation (4 in. or less slump, 40% relative humidity, moist cured, and loading at 7 days or
more),
creep strain
Ct =
initial elastic strain (1.11.1)
t 0.60
= Cu
10 + t 0.60
shown in Fig. 1.11.2, where t is the duration of loading (days) and Cu is the ultimate
creep coefficient. (Branson [1.80] suggests using an average of 2.35 for Cu under standard
Creep
Strain
Shrinkage
Nominal
True elastic elastic strain
strain
t0 Time
Figure 1.11.1 Change in strain of a loaded and drying specimen; t0 is the time at application of load.
17
1 . 1 1 C R E E P A N D S H R I N K AG E 17
conditions, but the range is shown to be from 1.3 to 4.15.) Correction factors are given for
relative humidity, loading age, minimum thickness of member, slump, percent fines, and
air content. For practical purposes, the only factors significant enough to require correction
are humidity and age at loading.
The effect of unloading may be seen from Fig. 1.11.3, where at a certain time t1 the load
is removed. There is an immediate elastic recovery and a long-time creep recovery, but a
residual deformation remains.
Creep of concrete will often cause an increase in the long-term deflection of members.
Unlike concrete, steel is not susceptible to creep. For this reason, steel reinforcement is
often provided in the compression zone of beams to reduce their long-term deflection.
Ct = 0.78Cu Ct = 0.90Cu
at 1 year at 5 years
Cu
Ct = elastic strain
creep strain
Eq. (1.11.1)
100 200 300 400 500 600

Duration of loading, days
Figure 1.11.2 Standard creep coefficient variation with duration of loading (for 4 in. or less
slump, 40% relative humidity, moist cured, and loading at 7 days or more).
Elastic recovery
Deformation
Creep
Creep recovery
Elastic
t1
Time after load application
Figure 1.11.3 Typical relationship between creep and recovery with time.
Shrinkage
Shrinkage, broadly defined, is the volume change during hardening and curing of the con-
crete. It is unrelated to load application. The main cause of shrinkage is the loss of water as
the concrete dries and hardens. It is possible for concrete cured continuously under water to
increase in volume; however, the usual concern is with a decrease in volume. A discussion
of the mechanisms of shrinkage may be found in Ref. 1.7. In general, the same factors have
been found to influence shrinkage strain as those that influence creep—primarily those fac-
tors related to moisture loss.
The Branson general prediction method [1.80] gives a standard shrinkage strain equa-
tion (for 4 in. or less slump, 40% ambient relative humidity, and minimum member dimen-
sion of 6 in. or less, after 7 days moist cured)
 t 
ε sh =  (ε )
 35 + t  sh u
(1.11.2)
shown in Fig. 1.11.4, where t is time (days) after moist curing, and (ε sh )u is the ulti-
mate shrinkage strain. (Branson [1.80] suggests using 800 × 10 –6 for average conditions,
18
εsh = 0.91(εsh)u
εsh = 0.72(εsh)u at 1 year
at 3 months
(εsh)u
Shrinkage strain, ε sh
Eq. (1.11.2)
100 200 300 400 500 600

Time after moist curing, days
Figure 1.11.4 Standard shrinkage strain variation with time after 7 days of moist curing (for 4 in.
or less slump, 40% ambient relative humidity, and minimum member dimension of 6 in. or less).
but the range is from approximately 400 to more than 1000 × 10 −6 .) Correction factors are
given with the primary one relating to humidity H,
correction factor = 1.40 − 0.01H for 40% ≤ H ≤ 80%
correction factor = 3.00 − 0.03H for 80% ≤ H ≤ 100%
Shrinkage, particularly when restrained unsymmetrically by reinforcement, causes defor-

mations generally additive to those of creep. For proper serviceability, it is desirable to
predict or compensate for shrinkage in the structure.
1.12 CONCRETE QUALITY CONTROL

In reinforced concrete design, concrete sections are proportioned and reinforced using a
specified compressive strength fc′. The strength fc′ for which each part of a structure has
been designed should be clearly indicated on the design drawings. In the United States, as
indicated in Section 1.7, fc′ is based on cylinder strength (6 × 12 or 4 × 8 in. cylinders).
Because concrete is a material whose strength and other properties are not precisely
predictable, test cylinders from a mix designed to provide, say, 4000 psi (roughly 28 MPa)
concrete will show considerable variability. Therefore, mixes must be designed to provide
an average compressive strength greater than the specified value fc′.
ACI-26.4.3.1(b) allows concrete to be proportioned to achieve the specified compres-
sive strength following ACI 301 [1.81], based on either field test data or trial mixes. When
the ready-mix plant or other concrete production facility has a field test record based on at
least 15 consecutive strength tests, or two groups of consecutive strength tests with a total
no less than 30 tests and at least 10 tests in a group, for materials and conditions similar to
those expected, the standard deviation s can be computed based on those tests to establish
how variable the concrete strength is. These records shall correspond to a concrete with
compressive strength within 1000 psi (6.9 MPa) from the specified concrete strength and
obtained within the past 12 months, from a period no less than 60 calendar days.
When at least 30 consecutive strength tests are the basis for computing the standard
deviation, s, ACI 301 indicates that the required average compressive strength fcr′ used for
proportioning the mix must be taken as the larger of Eq. (1.12.1) and the appropriate one
of either (1.12.2) or (1.12.3):
fcr′ = fc′ + 1.34 s (1.12.1)
and when fc′ ≤ 5000 psi,
fcr′ = fc′ + 2.33s − 500 (1.12.2)

19
1.13 STEEL REINFORCEMENT 19
or when fc′ > 5000 psi,
fcr′ = 0.90 fc′ + 2.33s (1.12.3)
For example, if the designer has used a specified strength fc′ of 4000 psi, and the concrete
producer has shown field test data with a standard deviation of 450 psi, the mix should
be designed for an average strength of 4600 psi [i.e., the larger of 4000 + 1.34 ( 450 ) and
4000 + 2.33 ( 450 ) − 500 ]. When fewer than 30 tests are available, a modification factor will
require using a higher value of fcr′ .
When data are not available to establish a standard deviation, the required average com-
pressive strength fcr′ is calculated as follows
when fc′ ≤ 3000 psi,
fcr′ = fc′ + 1000
when 3000 psi ≤ fc′ ≤ 5000 psi,
fcr′ = fc′ + 1200
when fc′ > 5000 psi,
fcr′ = 1.1 fc′ + 700
According to ACI-26.12.3.1(b), concrete compressive strength is considered acceptable if

both conditions below are satisfied.
1. Every arithmetic average of any three consecutive strength tests9 equals or exceeds fc′.
2. No individual strength test falls below fc′ by more than 500 psi when fc′ is 5000 psi
or less; or by more than 0.10 fc′ when fc′ exceeds 5000 psi.
Statistical variations are to be expected, and strength tests failing to meet the above
criteria will occur perhaps once in 100 tests even though all proper procedures have been
followed. A discussion on the risks inherent in the consideration of limited test data is pro-
vided by Tait [1.82]. ACI-26.12.4 specifies steps to be taken in case low-strength test results
are obtained.
The foregoing discussion of concrete strength variation should merely give an aware-
ness of the fact that concrete having a specified compressive strength fc′ cannot be expected
to provide precisely known actual strength and other properties.
Quality control in the broader sense for reinforced concrete construction is a subject of
great importance, but generally lies outside the scope of this text. The ACI Committee 121
Report [1.83] and the papers by Tuthill [1.84], Mather [1.85], Newman [1.86], and Scanlon
[1.87] provide an excellent overall treatment of this subject.
1.13 STEEL REINFORCEMENT
Steel reinforcement may consist of bars, welded wire reinforcement, wires, or discrete
fibers.
Deformed Bars
For usual construction, bars (called deformed bars) having lugs or protrusions (defor-
mations) are used (Fig. 1.13.1). Such deformations serve to deter slip of the bar relative
to the concrete that surrounds it resulting from tension or compression in the bar. These
9 According to ACI 301-4.2.2.8a, a strength test is the “average of at least two 6 × 12 in. cylinders or the average of at least
three 4 × 8 in. cylinders made from the same concrete sample.”
20
deformations can have different patterns depending on the bar producer (see Fig. 1.13.1),
but they all have to meet minimum requirements of spacing, height, and gap according
to ASTM standards. These deformed bars are available in the United States in sizes 3 8 to
2 1 4 in. (9.5–57 mm) nominal diameter.
Sizes of ASTM bars (in Inch-Pound units) are indicated by numbers (see Table 1.13.1).
For sizes #3 through #8, they are based on the number of eighths of an inch included in the
nominal diameter of the bars. Bars designated #9 through #11 are round bars corresponding
to the former 1 in. square, 1 1 8 in. square, and 1 1 4 in. square sizes, and bars designated #14
and #18 are round bars having cross-sectional areas equal to those of 1 1 2 and 2 in. square
sizes, respectively. The nominal diameter of a deformed bar is equivalent to the diameter of
a plain bar having the same weight per foot as the deformed bar. For metric units, ASTM
standards use “soft” conversion, as given in Table 1.13.2.
Reinforcing bar steel in the United States is covered under ASTM designations as shown
in Table 1.13.3 [1.88–1.92]. The “Grade” of steel is the minimum specified yield stress10
expressed in ksi for Inch-Pound reinforcing bar Grades 40, 50, 60, 75, 80, 100, and 120 and
in MPa for SI reinforcing bar Grades 280, 350, 420, 520, 550, 690, and 830. Both Grades
40 and 60 exhibit the well-defined yield point and elastic-plastic strain behavior shown in
Fig. 1.13.2(a). The overall relationships are shown in Fig. 1.13.2(b). Higher grade steels,
on the other hand, often exhibit little or no yield plateau and lower ductility compared to
Grades 40 and 60 steels.
To ensure sufficient ductility for use in earthquake-resistant structures as well as ade-
quate weldability, ASTM A706/A706M [1.91] has more restrictive mechanical and chemi-
cal properties than the other types of steel. Minimum elongation measured over a length of
8 in. (200 mm) ranges between 10 and 14% depending on the grade of steel and bar size.
Also, restrictions apply to the ratio between tensile and actual yield strength, as well as the
actual yield strength (see footnote d of Table 1.13.3). Deformation requirements for ASTM
A615 and A615M steel [1.88] are lower than those for ASTM A706/A706M steel, ranging
between 6 and 9% for Grades 60, 75, and 80 steel. Higher deformation requirements apply
to smaller bar sizes. No special requirements are specified for enhanced weldability. Axle
and rail steel bars, both of which are rarely used now, are rerolled from old axles and rails
and are generally less ductile than bars satisfying ASTM A615/A615M.
Grade 60 is the most widely used grade for reinforcing bars. However, higher-strength
steels are gaining popularity, since the ACI Code allows the use of a design yield strength
of 80 ksi (550 MPa) for longitudinal reinforcement in structural members other than those
of “special seismic systems” and of 100 ksi (690 MPa) for confinement reinforcement in
earthquake-resistant members and spirals in columns.
Figure 1.13.1 Deformed reinforcing bars. (Courtesy of Concrete Reinforcing Steel Institute.)
10 The term “yield stress” refers to either yield point, the well-defined deviation from perfect elasticity, or yield strength, the
value obtained by a 0.2% offset strain for material having no well-defined yield point (ACI-20.2.1.2).
21
120 120
800 800
Tensile
Strain hardening begins
700 strength 700
100 at strains as low as about 100
0.004 Grade 60
600 600
Nominal stress, MPa
Nominal stress, MPa

Larger bars,
Nominal stress, ksi
Nominal stress, ksi

80 Grade 60 80
say #11 and larger
steels 500 500
Tensile
60 60
Smaller size bars, say 400 strength 400
#8 and smaller 300 Grade 40 300
40 40
Grade 40 Strain hardening 200 Yield stress Grade 40 200
20 steels begins at about 20
100 100
0.012 to 0.020
0 0.002 0.006 0.010 0.014 0 0.05 0.10 0.15 0.20

0.004 0.008 0.012 Strain
Strain
(a) Enlarged portion in design (b) Entire curve to rupture
Figure 1.13.2 Typical stress-strain curves for Grade 60 and Grade 40 steel reinforcing bars in tension.
TABLE 1.13.1 ASTM STANDARD REINFORCING BAR DIMENSIONS AND

WEIGHTS (BARS IN INCH-POUND UNITS)
Nominal Dimensions
Diameter Area Weight
Bar Number (in.) (mm) (sq in.) (cm2) (lb/ft) (kg/m)

3 0.375 9.5 0.11 0.71 0.376 0.559
4 0.500 12.7 0.20 1.29 0.668 0.994
5 0.625 15.9 0.31 2.00 1.043 1.552
6 0.750 19.1 0.44 2.84 1.502 2.235
7 0.875 22.2 0.60 3.87 2.044 3.041
8 1.000 25.4 0.79 5.10 2.670 3.973
9 1.128 28.7 1.00 6.45 3.400 5.059
10 1.270 32.3 1.27 8.19 4.303 6.403
11 1.410 35.8 1.56 10.06 5.313 7.906
14 1.693 43.0 2.25 14.52 7.65 11.38
18 2.257 57.3 4.00 25.81 13.60 20.24
TABLE 1.13.2 1996 ASTM (“SOFT” METRIC) REINFORCING BAR

DIMENSIONS AND WEIGHTS IN SI UNITS
Metric Bar Inch-Pound Diameter Mass Area

Number Bar Number (mm) (kg/m) (mm2)
10 3 9.5 0.560 71
13 4 12.7 0.994 129
16 5 15.9 1.552 199
19 6 19.1 2.235 284
22 7 22.2 3.042 387
25 8 25.4 3.973 510
29 9 28.7 5.060 645
32 10 32.3 6.404 819
36 11 35.8 7.907 1006
43 14 43.0 11.38 1452
57 18 57.3 20.24 2581
2
TABLE 1.13.3 REINFORCING BAR STEELS
Minimum Minimum
Yield Stressa fy Tensile Strength, fu
ASTM Designation Grade Bar Sizes ksi MPa ksi MPa

A615/A615Mb 40 #3–#6 40 70
(Carbon steel) 60 #3–#18 60 90
75 #3–#18 75 100
80 #3–#18 80 105
280 10–19 280 500
420 10–57 420 620
520 10–57 520 690
550 10-57 550 725
A955/A995M 60 #3–#18 60 90
(Stainless steel) 75 #3–#18 75 100
420 10–57 420 620
520 10–57 520 690
A996/A996M 40 #3–#8 40 70
(Rail steel,c 50 #3–#8 50 80
axle steel) 60 #3–#8 60 90
280 10–25 280 500
350 10–25 350 550
420 10–25 420 620
A706/A706Md 60 #3–#18 60 80
(Low-alloy steel) 80 #3–#18 80 100
420 10–57 420 550
550 10–57 550 690
A1035/A1035M 100 #3–#18 100 150
(Low-carbon, 120 #3–#18 120 150
chromium steel) 690 10–57 690 1030
830 10–57 830 1030
a The term “yield stress” refers to either yield point, the well-defined deviation from perfect elasticity, or yield strength, the value obtained by a 0.2% offset
strain for material having no well-defined yield point (ACI-20.2.1.2).
b Metric (SI) specification applies to bars designated numbers 10 through 57.
c Although rail steel is no longer considered a practical source of bar reinforcement, its use is still permitted.
dIn addition to the yield and tensile strength limits given in the table, the tensile strength shall be at least 1.25 times the actual yield strength, where actual
yield strength shall not exceed 78 and 98 ksi (540 and 675 MPa) for Grade 60 and Grade 80 steel, respectively.
Wire Reinforcement
Welded wire reinforcement is used in thin slabs, thin shells, thin webs of T-beams, and
other locations where available space would not permit the placement of deformed bars
with proper cover and clearance. Welded wire reinforcement shall conform to either
ASTM A1064 [1.93] for carbon steel or ASTM A1022 [1.94] for stainless steel. Welded
wire reinforcement consists of cold-worked wire, cold-rolled or hot-rolled from steel rod,
in orthogonal patterns, square or rectangular, resistance welded at all intersections. The
wires may be smooth or deformed. The wire is specified by the symbol W (for smooth
wires) or D (for deformed wires), followed by a number representing the cross-sectional
area in hundredths of a square inch, varying from 1.5 to 45. On design drawings such
reinforcement usually is indicated by the spacings of wires in the two orthogonal direc-
tions, followed by the type and wire sizes. Thus, 6 × 8—W5 × W5 indicates welded
wire reinforcement with 6- in. longitudinal wire spacing, 8- in. transverse wire spac-
ing, and both sets of wires smooth and having a cross-sectional area of 0.05 sq in.
Unlike most hot-rolled steel bars, the wire used in welded wire reinforcement does
not generally have a well-defined yield point and is less ductile. Figure 1.13.3 shows
23
100 100 Tensile strength

Tensile strength
Nominal stress, ksi
Nominal stress, ksi

80 80
60 60
40 40
20 20
1 2 3 4 5 1 2 3 4 5
Strain, % Strain, %
(a) Smooth (b) Deformed
Figure 1.13.3 Typical stress-strain curves for welded wire reinforcement.
typical stress-strain curves for welded wire reinforcement. Additional information about
welded wire reinforcement is available from the Wire Reinforcement Institute [1.95].
Wires in the form of individual wires conforming to either ASTM A1064/A1064M or
A1022/A1022M can also be used as reinforcement for certain purposes. When deformed,
wires can be used for confinement and as spiral reinforcement, among others. Plain wires,
on the other hand, can be used only in the form of spirals.
Prestressing Reinforcement
Wire reinforcement in the form of groups of wires forming strands or as individual wires
are used for prestressing concrete. Wire and strands are available in great variety. The
most prevalent strand is the 7-wire strand, either stress relieved (i.e., normal relaxation)
or low relaxation, conforming to ASTM A416 [1.96]. Low-relaxation reinforcement is
now regarded as the standard type. These strands have a center wire enclosed by six
helically wound outside wires (see Fig. 1.13.4). Usual nominal diameters for 7-wire
strand are 1 4 , 3 8 , and 1 2 in. The minimum tensile strength for strands of Grade 250 is
250,000 psi (1725 MPa), and that of Grade 270 is 270,000 psi (1860 MPa); there
is no well-defined yield point. A typical stress-strain curve for stress-relieved strands
is shown in Fig. 1.13.5. ASTM A416 requires that the yield strength, measured at a 1%
extension under load, should be at least 85% of the tensile strength for stress-relieved
strand and 90% of the tensile strength for low- relaxation strand. Typically, under
service conditions, these prestressed strands have a stress of 150,000 to 160,000 psi
(1030–1100 MPa).
250 1724
1500
200
Unit stress, MPa
Unit stress, ksi
fpu = 250 ksi 1000
100
500
0
1 2 3 4 5
Unit strain, %
Figure 1.13.4 Typical 7-wire strand used in prestressed Figure 1.13.5 Typical stress-strain curve for
concrete construction (Photo by José A. Pincheira.) stress-relieved strands.
24
Additionally, for prestressing, uncoated stress-relieved and low-relaxation wire under

ASTM A421 [1.97] and uncoated high-strength steel bars under ASTM A722 [1.98]
are used.
The modulus of elasticity for all nonprestressed steel is permitted to be taken (ACI-
20.2.2.2) as 29,000,000 psi (200,000 MPa). For prestressing steel, the modulus of elasticity
is lower and more variable; therefore it must be obtained from the manufacturer or from
tests. A value of 27,000,000 psi (186,000 MPa) is often used for 7-wire strands conforming
to ASTM A416 [1.96].
Coated Reinforcement
Corrosion of the steel reinforcement can occur when the structure is subjected to severe
environmental conditions, such as in structures exposed to marine environments or bridge
decks or parking garages subjected to deicing salts. The corrosion of a reinforcing bar
embedded in concrete is a slow process that eventually will lead to cracking and spalling of
the concrete cover. The repair of corrosion-induced damage is often expensive and difficult.
A proper concrete cover can effectively delay the corrosion of steel bars and it is generally
agreed that larger covers will provide better protection. In severe environments, however,
large concrete covers alone will not be effective.
A common method to prevent or ameliorate corrosion of the reinforcement in concrete
structures is the use of epoxy-coated reinforcing bars. The surface of these bars is protected
with a thin coat of epoxy (between 7 and 12 mils) to isolate the steel from the oxygen, mois-
ture, and chlorides that will induce corrosion. Although the performance of epoxy-coated
bars has been the subject of some controversy in the past, many studies have shown that
epoxy-coated bars can effectively reduce corrosion of the reinforcement and extend the
service life of concrete structures. Care must be exercised during transportation, handling,
storage, and placing of the bars to prevent damage to the coating. Current practice requires
that any damage to the coating (cracking, nicks, and cuts) be repaired before bar placement
in the forms. The manufacturing requirements of these bars are presently covered by ASTM
A775, Specification for Epoxy-Coated Reinforcing Steel Bars [1.99], and ASTM A934,
Specification for Epoxy-Coated Prefabricated Steel Reinforcing Bars [1.100]. It must be
noted that epoxy-coating of bars will result in reduced slip resistance. The ACI Code con-
tains specific provisions that account for the reduced ability of an epoxy-coated bar to
transfer the force in the reinforcement to the surrounding concrete. These provisions are
discussed in detail in Chapter 6.
Zinc-coated (galvanized) bars are sometimes specified to reduce corrosion of steel rein-
forcement. Similar to epoxy-coated bars, galvanized bars are protected with a thin layer of
zinc on the surface. Although zinc coating will protect the steel bar, zinc will corrode in
concrete [1.101]; and eventually, corrosion of the steel bar will also occur. While the use
of galvanized bars can delay the onset of concrete spalling, it will not significantly extend
service life in a severe chloride environment [1.102]. In this case, the use of zinc and
epoxy dual-coated bars would be more advantageous. A layer of zinc alloy is first applied,
followed by a layer of epoxy. Requirements for the manufacture of galvanized bars are
given in ASTM A767 Specification for Zinc-Coated (Galvanized) Steel Bars for Concrete
Reinforcement [1.102] and in ASTM A1055/A1055M Standard Specification for Zinc and
Epoxy Dual-Coated Steel Reinforcing Bars [1.103] for zinc and epoxy dual-coated bars.
Fiber Reinforcement
Another type of steel reinforcement, permitted for use for the first time in the 2008 ACI
Code, is deformed steel fibers. According to ACI-26.4.1.5.1, deformed steel fibers, when
used for shear resistance, shall conform to ASTM A820/A820M [1.104] and have a length-
to-diameter ratio between 50 and 100. These fibers are typically 1 to 2.4 in. (25–60 mm)
in length with diameters between 0.015 and 0.04 in. (0.38 and 1 mm). Deformations are
introduced to the fibers in order to increase bond with the surrounding concrete. Typical
deformations in steel fibers include hooks, flat ends, and waves and twists along the fiber
length. The tensile strength of wire used to manufacture deformed steel fibers typically
ranges between 150 and 350 ksi (1030 and 2400 MPa).
25
1.14 FIBER-REINFORCED CONCRETE 25
1.14 FIBER-R EINFORCED CONCRETE

Fiber-reinforced concrete consists of concrete reinforced with short, typically randomly ori-
ented, fibers. Fibers used in structural applications are often made of steel, with deformations
to provide mechanical bond with the concrete (see section on fiber reinforcement above).
Contrary to regular bar-type reinforcement, however, fibers are generally expected to pull
out rather than yield (except locally at or near fiber deformations). The behavior of fiber-
reinforced concrete is thus controlled by the bond stress versus slip response of the fibers.
Typically, steel fibers are used in dosages ranging between 0.5 and 1.5% by volume. For steel
fibers, this corresponds to approximately 65 lb and 200 lb cu yard (40 and 120 kg/m3). Fibers
can be premixed with cement and aggregate or added after the concrete has been mixed.
Fiber reinforcement provides the concrete with post-cracking tensile resistance. Fiber-
reinforced concrete subjected to tension typically exhibits a softening response. In some
cases, however, a pseudo-strain-hardening response in tension can be obtained, which is
accompanied by multiple cracking and strains at peak stress typically greater than 0.5%.
These strain-hardening materials are often referred to as high-performance fiber-reinforced
concrete (HPFRC). Cracking strength (or first cracking strength in the case of HPFRC) is
little sensitive to the presence of fibers, given the relatively small volume fractions used
in practice. Fig. 1.14.1 shows qualitative tensile stress-strain responses for plain, strain-
softening, and strain-hardening fiber-reinforced concretes.
Since there is no commonly accepted tensile test method for fiber-reinforced concrete,
tensile behavior is generally evaluated indirectly, through a standard beam test. The most
commonly used flexural test in the United States is that specified in ASTM C1609 [1.105],
where a beam, typically 4 × 4 or 6 × 6 in. in cross section (100 × 100 or 150 × 150 mm)
and with a span equal to three times its cross-sectional side dimension, is tested under four-
point bending up to a midspan deflection of 1150 times the span length. In Europe, flexural
tests are often conducted on a notched beam, for which the crack mouth opening displace-
ment is often measured to obtain an estimate of tensile stress versus crack width response.
Fiber reinforcement also has an effect on concrete compressive behavior. While strength
is little sensitive to the addition of fibers in typical amounts used in practice, ductility
is increased by the presence of fibers [1.106], translating into a lower post-peak stress
decrease and higher strain capacity (Fig. 1.14.2). In the case of HPFRCs, a compressive
stress-strain response similar to that of well-confined concrete can be obtained.
The addition of fibers leads to a decrease in concrete workability, which should be taken
into account when concrete mixes are selected. Typically, an increase in fiber volume frac-
tion or fiber length to diameter ratio leads to a decrease in concrete workability. However,
with the use of high-range, water-reducing admixtures and viscosity-modifying agents,
ft
fcr Fiber-reinforced concrete fc/fc’

(Strain-hardening)
1
Vf = fiber volume
Fiber-reinforced concrete fraction
(Strain-softening) Fiber-reinforced
0.5 concrete
Plain Vf = 2%
Plain concrete concrete Vf = 1%
εcr εt 1 2 3 4 5 εc/εo
Figure 1.14.1 Qualitative tensile stress-strain response of Figure 1.14.2 Effect of fiber reinforcement on

plain and steel fiber-reinforced concretes. compressive post-peak response of concrete. (Data from
Ref. 1.106.)
26
self-consolidating concrete can be manufactured, even with steel fibers in a 1.5% volume
fraction (200 cu yard or 120 kg/m3) [1.107].
The enhancement of tension and compressive behavior of concrete with the addition of
fibers makes the use of fiber-reinforced concrete attractive in cases when additional shear
resistance or enhanced bond between reinforcement and concrete is required, or where
large amounts of confinement reinforcement are needed. Examples include the use of fibers
as minimum shear reinforcement in beams and as punching shear reinforcement in slabs,
confinement steel in beam-column connections, and shear walls and coupling or link beams
of earthquake-resistant structures. Information about the effect of fiber reinforcement on
shear resistance, bond between reinforcement and concrete, and confinement requirements
can be found in References 1.108 through 1.112.
Structural applications of fiber-reinforced concrete in the United States have been lim-
ited, primarily because of lack of Code design provisions. It was not until 2008 that the
ACI Code included provisions for the use of deformed steel fibers (as minimum shear
reinforcement). Steel fibers have been used for a few years, however, in precast segmental
linings and, more recently, in coupling or link beams of earthquake-resistant coupled wall
structures [1.112].
1.15 UNITS
U.S. Customary units are the primary units used throughout this text. In some cases, values
or equations are also expressed in SI units. Although not SI, the MKS (meter-kilogram-
second) system is used in most western hemisphere countries (the United States and
Canada excepted), where instead of the kilogram (kg) as a unit of mass, as in SI, the kilo-
gram (kgf) is used as a unit of force. Thus, curves involving stresses have sometimes three
parallel axes: pounds per square inch (psi) or kips per square inch (ksi) for Inch-Pound
units, megapascals (MPa) for SI units, and kilogram-force per square centimeter (kgf/cm2)
for the other metric (MKS) system. The authors believe that some familiarity with met-
ric units is essential; on the other hand, overemphasis on units in design calculations and
procedures detracts from concepts. In general, throughout the remaining chapters, design
equations that involve units will have an SI version (usually ACI 318-14M) given in a
footnote. The reader interested in proper use of SI units should consult the IEEE/ASTM
American National Standard for Metric Practice [1.113].
SELECTED REFERENCES
1.1. Hans Straub. A History of Civil Engineering. Cambridge, MA: MIT Press, 1964 (pp. 205–215);
also London: Leonard Hill Ltd., 1952.
1.2. R. S. Kirby, Sidney Withington, A. B. Darling, and F. G. Kilgour. Engineering in History.
New York: McGraw-Hill, 1956.
1.3. PCA. “An Historical Look at Reinforced Concrete Structures Through the Eyes of Eli

W. Cohen,” Engineered Concrete Structures, 8, 2 (August 1995). Skokie, IL: Portland Cement
Association, 3 pp.
1.4. M. K. Hurd. “Ernest L. Ransome—
Concrete Designer, Constructor, Inventor,” Concrete
International, 18, May 1996, 50–51.
1.5. Frederick E. Turneaure and Edward R. Maurer. Reinforced Concrete Design.

New York: Wiley, 1907.
1.6. Adam M. Neville. Properties of Concrete (4 ed.). New York: John Wiley & Sons, 1996.
1.7. Sidney Mindess, J. Francis Young, and David Darwin. Concrete (2nd ed.) Upper Saddle River,
NJ: Prentice-Hall, 2003.
1.8. P. Kumar Mehta. Concrete Structure, Properties, and Materials. Englewood Cliffs,
NJ: Prentice-Hall, 1986.
1.9. ASTM. Standard Specification for Portland Cement (C150/C150M–12). West Conshohocken,
PA: ASTM International, 2012.
27
SELECTED REFERENCES 27
1.10. ACI Committee 225. Guide to the Selection and Use of Hydraulic Cements (ACI 225R-99
[Reapproved 2009]). Farmington Hills, MI: American Concrete Institute, 1999, 30 pp. Also ACI
Journal, Proceedings, 82, November–December 1985, 901–928.
1.11. ASTM. Standard Specification for Blended Hydraulic Cements (C595/C595M-13). West
Conshohocken, PA: ASTM International, 2013.
1.12. ACI. Building Code Requirements for Structural Concrete (ACI 318–14) and Commentary (ACI
318R-14). Farmington Hills, MI: American Concrete Institute, 2014.
1.13. ACI Committee 221. Guide for Use of Normal Weight and Heavyweight Aggregates in Concrete
(ACI 221R- 96 [Reapproved 2001]). Farmington Hills, MI: American Concrete Institute,
1996, 28 pp.
1.14. ASTM. Standard Specification for Concrete Aggregates (C33/C33M-13). West Conshohocken,
PA: ASTM International, 2013.
1.15. ASTM. Standard Specification for Lightweight Aggregates for Structural Concrete (C330/
C330M-13). West Conshohocken, PA: ASTM International, 2013.
1.16. Richard W. Steiger. “Development of Lightweight Aggregate Concrete,” Concrete Construction,
30, June 1985, 519–525.
1.17. George K. Mackie II. “Recent Uses of Structural Lightweight Concrete,” Concrete Construction,
30, June 1985, 497–502.
1.18. ACI Committee 213. Guide for Structural Lightweight-Aggregate Concrete (ACI 213R-03).
Farmington Hills, MI: American Concrete Institute, 2003, 38 pp.
1.19. ACI Committee 304. Heavyweight Concrete: Measuring, Mixing, Transporting, and Placing
(ACI 304.3R-96 [Reapproved 2004]). Farmington Hills, MI: American Concrete Institute,
1997, 8 pp.
1.20. V. M. Malhotra. “Use of Mineral Admixtures for Specialized Concretes,” Concrete International,
6, April 1984, 19–24.
1.21. Richard C. Mielenz. “History of Chemical Admixtures for Concrete,” Concrete International, 6,
April 1984, 40–53.
1.22. ASTM. Standard Specification for Coal Fly Ash and Raw or Calcined Natural Pozzolan
for Use as a Mineral Admixture in Concrete (C618-12a). West Conshohocken, PA: ASTM
International, 2012.
1.23. John M. Albinger, “Fly Ash for Strength and Economy,” Concrete International, 6, April
1984, 32–34.
1.24. Dan Ravina. “Slump Retention of Fly Ash Concrete With and Without Chemical Admixtures,”
Concrete International, 17, April 1995, 25–29.
1.25. ASTM. Standard Specification for Slag Cement for Use in Concrete and Mortars (C989-13).
West Conshohocken, PA: ASTM International, 2013
1.26. ACI Committee 233. Slag Cement in Concrete and Mortar (ACI 213R-03 [Reapproved 2011]).
1.27. ASTM. Standard Specification for Use of Silica Fume for Used in Cementitious Mixtures
(C1240-14). West Conshohocken, PA: ASTM International, 2014.
1.28. Menashi D. Cohen, Jan Olek, and Bryant Mather. “Silica Fume Improves Expansive-Cement
Concrete,” Concrete International, 13, March 1991, 31–37.
1.29. T. A. Durning and M. C. Hicks. “Using Microsilica to Increase Concrete’s Resistance to
Aggressive Chemicals,” Concrete International, 13, March 1991, 42–48.
1.30. T. A. Durning and M. C. Hicks. Concrete Construction, 30, April 1985 (entire issue devoted to
admixtures).
1.31. ACI Committee 212. Chemical Admixtures for Concrete (ACI 212.3R-10). Farmington Hills,
MI: American Concrete Institute, 2010, 61 pp.
1.32. ASTM. Standard Specification for Air-Entraining Admixtures for Concrete (C260/C260M-10a).
West Conshohocken, PA: ASTM International, 2010.
1.33. ASTM. Standard Specification for Chemical Admixtures for Concrete (C494-99ae1). West
Conshohocken, PA: ASTM International, 1999.
1.34. ASTM. Standard Specification for Chemical Admixtures for Use in Producing Flowing Concrete
(C1017-98). West Conshohocken, PA: ASTM International, 1998.
1.35. A. A. Ramezanianpour, V. Sivasundaram, and V. M. Malhotra. “Superplasticizers: Their Effect
on the Strength Properties of Concrete,” Concrete International, 17, April 1995, 30–35.
1.36. ACI Committee 237. Self-
Consolidating Concrete (ACI 237R- 07). Farmington Hills,
1.37. Sandor Popovics. “Analysis of the Concrete Strength versus Water–Cement Ratio Relationship,”
ACI Materials Journal, 87, September–October 1990, 517–529.
1.38. Sandor Popovics and John S. Popovics. “The Foundation of a Computer Program for the
Advanced Utilization of w/c and Air Content in Concrete Proportioning,” Concrete International,
16, December 1994, 21–26.
28
1.39. Sandor Popovics and John S. Popovics. “Computerization of the Strength Versus w/
c*
Relationship,” Concrete International, 17, April 1995, 37–40.
1.40. ACI Committee 211. Guide for Selecting Proportions for No-Slump Concrete (ACI 211.3R-02
[Reapproved 2009]). Farmington Hills, MI: American Concrete Institute, 2002, 26 pp..
1.41. PCA. Design and Control of Concrete Mixtures (15th ed.). Skokie, IL: Portland Cement
Association, 2011.
1.42. ACI Committee 211. Standard Practice for Selecting Proportions for Normal, Heavyweight, and
Mass Concrete (ACI 211.1-91 [Reapproved 2009]). Farmington Hills, MI: American Concrete
Institute, 1991, 38 pp.
1.43. ACI Committee 211. Standard Practice for Selecting Proportions for Structural Lightweight
Concrete (ACI 211.2- 98 [Reapproved 2004]). Farmington Hills, MI: American Concrete
1.44. ACI Committee 304. Guide for Measuring, Mixing, Transporting, and Placing Concrete (ACI
304R-00 [Reapproved 2009]). Farmington Hills, MI: American Concrete Institute, 2000, 44 pp..
1.45. ACI Committee 311. ACI Manual of Concrete Inspection (SP-2), 10th ed. (ACI 311.1R-07).
Farmington Hills, MI: American Concrete Institute, 2007.
1.46. ACI Committee 201. Guide to Durable Concrete (ACI 201.2R-
008). Farmington Hills,
1.47. Walter H. Price. “Factors Influencing Concrete Strength,” ACI Journal, Proceedings, 47,

February 1951, 417–432.
1.48. UNESCO. Reinforced Concrete: An International Manual. London: Butterworth, 1971
(pp. 19–22).
1.49. Delmar L. Bloem. “Concrete Strength in Structures,” ACI Journal, Proceedings, 65, March
1968, 176–187.
1.50. James M. Shilstone, Jr. “The Cylinder Test—Reliable Informer or False Prophet,” Concrete
International, 2, July 1980, 63–68.
1.51. J. Bruce Tait “Making the Most of Concrete Strength Test Results,” ACI Journal, Proceedings,
83, May–June 1986, 383–388.
1.52. John R. Smith. “Estimating Later Age Strengths of Concrete,” ACI Journal, Proceedings, 81,
November–December 1984, 609–612.
1.53. A. P. Keiller. “Assessing the Strength of In Situ Concrete,” Concrete International, 7, February
1985, 15–21.
1.54. Robert S. Jenkins. “Nondestructive Testing—An Evaluation Tool,” Concrete International, 7,
February 1985, 22–26.
1.55. William C. Stone and Bruce J. Giza. “The Effect of Geometry and Aggregate on the Reliability
of the Pullout Test,” Concrete International, 7, February 1985, 27–36.
1.56. Thomas J. Parsons and Tarun R. Naik. “Early Age Concrete Strength Determination by Maturity,”
Concrete International, 7, February 1985, 37–43.
1.57. Kal R. Hindo and Wayne R. Bergstrom. “Statistical Evaluation of the In-Place Compressive
Strength of Concrete,” Concrete International, 7, February 1985, 44–48.
1.58. V. M. Malhotra. “Testing Early-Age Strength of Concrete In-Place,” Concrete International, 7,
April 1985, 39–41.
1.59. ACI Committee 214. Guide for Obtaining Cores and Interpreting Compressive Strength Results
(ACI 214.4R-10). Farmington Hills, MI: American Concrete Institute, 2010, 17 pp.
1.60. Hjalmar Granholm. A General Flexural Theory of Reinforced Concrete. New York: Wiley, 1965
(pp. 23–36).
1.61. P. T. Wang, S. P. Shah, and A. E. Naaman. “Stress–Strain Curves of Normal and Lightweight
Concrete in Compression,” ACI Journal, Proceedings, 75, November 1975, 603–611.
1.62. Vitelmo V. Bertero. “Inelastic Behavior of Structural Elements and Structures,” High Strength
Concrete, Proceedings of a Workshop, University of Illinois at Chicago Circle, December 2–4,
1979 (pp. 102–103, 153).
1.63. Antoine E. Naaman. Prestressed Concrete Analysis and Design Fundamentals.
New York: McGraw-Hill, 1982 (p. 54).
1.64. Arthur H. Nilson. Design of Prestressed Concrete. New York: Wiley, 1978 (p. 45).
1.65. Hubert Rüsch. “Researches Toward a General Flexural Theory for Structural Concrete,” ACI
Journal, Proceedings, 57, July 1960, 1–28.
1.66. Vaughan Randall and Kenneth Foot. “High-Strength Concrete for Pacific First Center,” Concrete
International, 11, April 1989, 14–16.
1.67. Nathan L. Howard and David M. Leatham. “The Production and Delivery of High-Strength
Concrete,” Concrete International, 11, April 1989, 26–30.
1.68. Gregory J. Smith and Franz N. Rad. “Economic Advantages of High-Strength Concretes in
Columns,” Concrete International, 11, April 1989, 37–43.
29
1.69. S. P. Shah, editor. High Strength Concrete, Proceedings of a Workshop, University of Illinois at
Chicago Circle, December 2–4, 1979 (Sponsored by National Science Foundation) (226 pp).
1.70. Anthony E. Fiorato. “PCA Research on High-Strength Concrete,” Concrete International, 11,
April 1989, 44–50.
1.71. ASTM. Standard Test Method for Splitting Tensile Strength of Cylindrical Concrete Specimens
(C496/C496M-11). West Conshohocken, PA: ASTM International, 2011.
1.72. ASTM. Standard Test Method for Flexural Strength of Concrete (Using Simple Beam with
Third-Point Loading) (C78/C78M-02). West Conshohocken, PA: ASTM International, 2010.
1.73. Sher Ali Mirza, Michael Hatzinikolas, and James G. MacGregor. “Statistical Descriptions of
Strength of Concrete,” Journal of the Structural Division, ASCE, 105, ST6 (June 1979), 1021–
1037. Disc. 106, ST7 (July 1980), 1659–1660.
1.74. Helmut Kupfer, Hubert K. Hilsdorf, and Hubert Rusch. “Behavior of Concrete under Biaxial
Stresses,” ACI Journal, Proceedings, 66, August 1969, 656–666.
1.75. Frank E. Richart, Anton Brandtzaeg, and Rex L. Brown. A Study of the Failure of Concrete under
Combined Compressive Stresses. Bulletin 185, Engineering Experiment Station, University of
Illinois, Urbana, November 1928 (104 pp).
1.76. Adrian Pauw. “Static Modulus of Elasticity of Concrete as Affected by Density,” ACI Journal,
Proceedings, 57, December 1960, 679–687.
1.77. T. S. Shih, G. C. Lee, and K. C. Chang. “On Static Modulus of Elasticity of Normal-Weight
Concrete,” Journal of Structural Engineering, ASCE, 115, 10 (October 1989), 2579–2587.
1.78. Francis A. Oluokun, Edwin G. Burdette, and J. Harold Deatherage, “Elastic Modulus, Poisson’s
Ratio, and Compressive Strength Relationships at Early Ages,” ACI Materials Journal, 88,
January–February 1991, 3–10.
1.79. ACI Committee 209. “Prediction of Creep, Shrinkage and Temperature Effects in Concrete
Structures,” Designing for Creep and Shrinkage in Concrete Structures (SP-76). Farmington
Hills, MI: American Concrete Institute, 1982 (pp. 193–300).
1.80. Dan E. Branson. Deformations of Concrete Structures. New York: McGraw-Hill, 1977 (pp. 11–
27, 44–55).
1.81. ACI Committee 301. Specifications for Structural Concrete (ACI 301-10). Farmington Hills,
1.82. J. Bruce Tait. “Concrete Quality Assurance Based on Strength Tests,” Concrete International, 3,
September 1981, 79–87.
1.83. ACI Committee 121. Guide for Concrete Construction. Quality Systems in Conformance with
ISO 9001 (ACI 121R-08). Farmington Hills, MI: American Concrete Institute, 2008 (35 pp).
1.84. Lewis H. Tuthill. “Obtaining Quality in Concrete Construction,” Concrete International, 8,
March 1986, 24–29.
1.85. Bryant Mather. “Selecting Relevant Levels of Quality,” Concrete International, 8, March

1986, 30–36.
1.86. Ken Newman. “Common Quality in Concrete Construction,” Concrete International, 8, March
1986, 37–49.
1.87. John M. Scanlon. “Quality Control During Hot and Cold Weather,” Concrete International, 1,
September 1979, 58–65.
1.88. ASTM. Standard Specification for Deformed and Plain Carbon- Steel Bars for Concrete
Reinforcement (A615/A615M-13). West Conshohocken, PA: ASTM International, 2013.
1.89. ASTM. Standard Specification for Deformed and Plain Stainless-Steel Bars for Concrete
1.90. ASTM. Standard Specification for Rail-Steel and Axle-Steel Deformed Bars for Concrete
1.91. ASTM. Standard Specification for Low-Alloy Steel Deformed and Plain Bars for Concrete
1.92. ASTM. Standard Specification for Deformed and Plain, Low-Carbon, Chromium, Steel Bars
for Concrete Reinforcement (ASTM A1035/A1035M-14). West Conshohocken, PA: ASTM
1.93. ASTM. Standard Specification for Carbon- Steel Wire and Welded Wire Reinforcement,
Plain and Deformed, for Concrete (A1064/A1064M-13). West Conshohocken, PA: ASTM
1.94. ASTM. Standard Specification for Deformed and Plain Stainless Steel Wire and Welded
Wire for Concrete Reinforcement (A1022/ A1022M- 13). West Conshohocken, PA: ASTM
1.95. WRI. Structural Welded Wire Reinforcement—Manual of Standard Practice (8th ed.). Hartford,
CT: Wire Reinforcement Institute (942 Main Street, Suite 300, Hartford, CT 06103), 2010.
1.96. ASTM. Standard Specification for Steel Strand, Uncoated Seven-Wire for Prestressed Concrete
(A416/A416M-12a). West Conshohocken, PA: ASTM International, 2012.
30
1.97. ASTM. Standard Specification for Uncoated Stress- Relieved Steel Wire for Prestressed
Concrete (A421/A421M-10). West Conshohocken, PA: ASTM International, 2010.
1.98. ASTM. Standard Specification for Uncoated High-Strength Steel Bar for Prestressing Concrete
(A722/A722M-12). West Conshohocken, PA: ASTM International, 2012.
1.99. ASTM. Specification for Epoxy- Coated Steel Reinforcing Bars (A775/A775M-07b
[Reapproved 2014]). West Conshohocken, PA: ASTM International, 2007.
1.100. ASTM. Specification for Epoxy-Coated Prefabricated Steel Reinforcing Bars (A934/A934M-
13) West Conshohocken, PA: ASTM International, 2013.
1.101. ACI Committee 222. Protection of Metals in Concrete Against Corrosion (ACI 222R-01
[Reapproved 2010]). Farmington Hills, MI: American Concrete Institute, 2001, 41 pp.
1.102. ASTM. Specification for Zinc-Coated (Galvanized) Steel Bars for Concrete Reinforcement
(A767/A767M-09). West Conshohocken, PA: ASTM International, 2009.
1.103. ASTM. Standard Specification for Zinc and Epoxy Dual- Coated Steel Reinforcing Bars
(A1055/A1055M-10e1) West Conshohocken, PA: ASTM International, 2010.
1.104. ASTM. Standard Specification for Steel Fibers for Fiber Reinforced Concrete (A820/A820M-
11) West Conshohocken, PA: ASTM International, 2011.
1.105. ASTM. Standard Test Method for Flexural Performance of Fiber-Reinforced Concrete (Using
Beam with Third- Point Loading (C1609/ C1609M- 12) West Conshohocken, PA: ASTM
1.106. Duane E. Otter and Antoine E. Naaman. “Properties of Steel Fiber Reinforced Concrete under
Cyclic Loading,” ACI Materials Journal, 85, July–August 1988, 254–261.
1.107. W.-C. Liao, S.-H. Chao, S.-Y., Park, and A. E. Naaman. “Self-Consolidating High-Performance
Fiber-Reinforced Concrete (SCHPFRC)—Preliminary Investigation.” Report No. UMCEE 06-
02. University of Michigan: Ann Arbor, 2006.
1.108. G. J. Parra-Montesinos. “Shear Strength of Beams with Deformed Steel Fibers,” Concrete
International, 28, November–December 2006, 57–66.
1.109. M.-
Y. Cheng and G. J. Parra- Montesinos. “Evaluation of Steel Fiber Reinforcement for
Punching Shear Resistance in Slab-Column Connections. Part I: Monotonically Increased
Load,” ACI Structural Journal, 107, January–February 2010, 101–109.
1.110. S. H. Chao, A. E. Naaman, and G. J. Parra-Montesinos. “Bond Behavior of Reinforcing Bars in
Tensile Strain-Hardening Fiber Reinforced Cement Composites,” ACI Structural Journal, 106,
1.111. G. J. Parra-Montesinos, S. Peterfreund, and S.-H. Chao. “Highly Damage Tolerant Beam-
Column Joints through Use of High-Performance Fiber-Reinforced Cement Composites,” ACI
Structural Journal, 102, May–June 2005, 487–495.
1.112. G. J. Parra-Montesinos, J. K. Wight, C. Kopczynski, R. D. Lequesne, M. Setkit, A. Conforti, and
J. Ferzli. “High-Performance Fiber Reinforced Concrete Coupling Beams: From Research to
Practice,” Proceedings of the 10th National Conference in Earthquake Engineering, Earthquake
Engineering Research Institute, Anchorage, AK, 2014.
1.113. IEEE/ASTM. American National Standard for Metric Practice (IEEE/ASTM SI 10-2010).
West Conshohocken, PA: ASTM International, 2010.
1.114. ACI Committee 213. “Guide for Structural Lightweight Aggregate Concrete,” Concrete

International, 1, February 1979, 33–62.
CHAPTER 2
DESIGN METHODS AND
REQUIREMENTS
2.1 STRUCTURAL DESIGN PROCESS—G ENERAL

The main objective of the structural design process is to ensure that the structure is able
to withstand the load effects with an appropriate margin of safety against failure. In this
context, failure is defined as that condition at which the structure ceases to fulfill its intended
function. This condition may occur because the structure has achieved its maximum resist-
ance and can no longer carry the imposed loads or, although it can still carry the loads,
because it has undergone excessive deformations that have rendered the structure unusable.
In other words, the structure must be designed so that it is both safe and functional (or serv-
iceable). Strength and serviceability considerations such as a limit on deflection may be
thought of as “limit states.” MacGregor [2.1] has provided an excellent treatment of limit
states design as applied to reinforced concrete.
Ensuring structural safety (i.e., protecting the life of the occupants) is a matter of public
concern and is clearly the most important of these objectives. Nevertheless, serviceabil-
ity can be as important in design to ensure that the structure is usable and can serve its
intended purpose. To satisfy the structural safety requirement, structural components must
be adequately proportioned to ensure that the structure is provided with enough strength
to resist the effects of the anticipated loads. On the other hand, serviceability issues (such
as control of deflections, vibrations, and cracking) are mostly a function of the stiffness of
members. In this context, the proper design of a reinforced concrete structure consists in
the judicious selection of member size and material strengths to provide the structure with
enough strength and stiffness to ensure that both safety and serviceability are satisfied.
Throughout the book, structural safety and serviceability are emphasized. However, other
aspects, such as the proportioning of members for cost effectiveness and ease of construc-
tion, are also addressed. Maintenance and aesthetics are also considered where appropriate.
In practice, the engineer must satisfy all these aspects while ensuring an economical design
of the structure.
2.2 ACI BUILDING CODE

When two different materials, such as steel and concrete, act together, the analysis for
strength of a reinforced concrete member clearly must be, in part, empirical. These prin-
ciples and methods are being constantly revised and improved as results of theoretical and
experimental research accumulate. The American Concrete Institute, serving as a clearing-
house for these changes, issues building code requirements, the most recent of which is
32
32 C H A P T E R 2 D E S I G N M E T H O D S A N D R E Q U I R E M E N T S
Lake Point Tower, Chicago; 70-story apartment building (Photo courtesy of Magnusson
Klemencic Associates).
Building Code Requirements for Structural Concrete (ACI 318-14), hereafter referred to as
the ACI Code [1.12].1
The ACI Code is a Standard of the American Concrete Institute. To achieve legal sta-
tus, it must be adopted by a governing body as a part of its general building code. The ACI
Code is partly a specification-type code, which gives acceptable design and construction
methods in detail, and partly a performance code, which states desired results rather than
details of how such results are to be obtained. A building code, legally adopted, is intended
to prevent people from being harmed; therefore, it specifies minimum requirements to pro-
vide adequate safety and serviceability. It is important to realize that a building code is not a
recommended practice, nor is it a design handbook, nor is it intended to replace engineering
knowledge, judgment, or experience. It does not relieve the designer of the responsibility
for having a safe, functional, and economical structure.
2.3 STRENGTH DESIGN AND WORKING STRESS METHODS

Two philosophies have been used in the design of reinforced concrete structures in the
past. The working stress method, the main approach used from the early 1900s until the
early 1960s, focused on conditions at service loads (i.e., at the maximum loads under the
1 The reader is advised to have the ACI Code as a ready reference while using this text.
3
2.5 STRENGTH DESIGN METHOD 33
intended use of the structure). Today, with few exceptions, the strength design method is
the prevalent procedure used in practice. The strength design method, which focuses on
conditions at loads greater than service loads, when failure may be imminent, is deemed
conceptually a more realistic approach to establish structural safety. Both approaches, how-
ever, provide approximately the same level of safety and serviceability.
2.4 WORKING STRESS METHOD

In the working stress method, a structural element is so designed that the stresses resulting
from the action of service loads (also called working loads) and computed by the mechan-
ics of elastic members do not exceed some predesignated allowable values.
Service load is the load (e.g., dead, live, snow, wind, earthquake) that is assumed actu-
ally to occur when the structure is in service.
The working stress method may be expressed by the following:
f ≤ fallow (2.4.1)
where
f = an elastic stress, such as by using the flexure formula f = Mc/I for a beam, com-
puted under service load
and
fallow = a limiting or allowable stress prescribed by a building code, for example, as a
percentage of the compressive strength fc′ for concrete, or of the yield stress fy for
the steel reinforcing bars.
Some of the obstacles to the working stress method are as follows:
1. Since the limitation is on the stress under the total service load, there is no sim-
ple way to account for different degrees of uncertainty of various kinds of loads.
Generally the dead load (gravity load due to weight of structural elements and per-
manent attachments) is known more accurately than the live load, which may have
unknown and variable distribution.
2. Creep and shrinkage of concrete, which contribute major time-dependent effects on
a structure, are not easily accounted for by calculation of elastic stresses.
3. Concrete stress is not proportional to strain up to its crushing strength, so that the inher-
ent safety provided is unknown when a percentage of fc′ is used as the allowable stress.
2.5 STRENGTH DESIGN METHOD

In the strength design method (formerly called ultimate strength method), the service loads
are increased by factors to obtain the load at which failure is considered to be “imminent.”
This load is called the factored load or factored service load. The structure or structural ele-
ment is then proportioned such that the strength is reached when the factored load is acting.
The strength design method may be expressed by the following,
strength provided ≥ strength required ( to carry factored loads) (2.5.1)
where the “strength provided” (such as moment strength) is computed in accordance with
the provisions of a building code, and the “strength required” is that obtained by perform-
ing a structural analysis using factored loads.
The “strength provided” is commonly referred to by practitioners as “ultimate strength.”
The computation of this strength takes into account the nonlinear stress-strain behavior of
concrete and reinforcing steel. However, it is a code-defined value for strength and is not
necessarily “ultimate” in the sense of being a value that cannot possibly be exceeded. The
ACI Code uses a conservative definition of strength; thus the modifier “ultimate” is not
entirely appropriate.
34
When the strength design method is used, the comparison of provided strength with
required strength (i.e., axial force, shear, or bending moment, caused by factored loads)
does not imply that any material “yields” or “fails” under service load conditions. In fact,
at service loads, the behavior of the structure is essentially elastic. The use of the term
“imminent failure” under factored loads is only a mechanism for establishing adequate
safety parameters.
Comments on Strength Design Methods

Historically, “ultimate” strength was the earliest approach to design, since the failure load
could be measured by tests without knowledge of the magnitude or distribution of internal
stresses. With the interest in and understanding of the elastic methods of analysis in the
early 1900s, the elastic working stress method was adopted almost universally by codes
as the best for design. As more detailed understanding of the actual behavior of reinforced
concrete structures subjected to loads in excess of the service loads developed, adjustments
in the theory and in the design procedures were made.
The first modification of the elastic working stress method resulted from the study of
axially loaded columns in the early 1930s. By 1940, the design of axially loaded columns
was based on ultimate strength. Next, the working stress method was modified to account
for creep of concrete in beams with compression steel and in eccentrically loaded columns.
The early history of the ACI Code has been summarized by Kerekes and Reid [2.2]. An
excellent discussion of what is involved in writing the ACI Code is given by Siess [2.3].
The 1956 ACI Code was the first that officially recognized and permitted the strength
design method, the result of work by ACI-ASCE Committee 327 [2.4]. The 1963 ACI Code
treated the working stress method and the strength design method on an equal basis; but,
actually, the major portion of the working stress method was based on strength. With the
relegation of the working stress method to a small section referred to as the “alternate
method,” the 1971 ACI Code entirely accepted the strength design method. Between 1971
and 1999, the ACI Code had the “alternate design method” in an appendix, but it was
removed from the 2002 and subsequent editions of the ACI Code.
2.6 SAFETY PROVISIONS—G ENERAL

Structures and structural members must always be designed to carry some reserve load
above what is expected under normal use. Such reserve capacity is provided to account
for a variety of factors, which may be grouped into two general categories: factors relat-
ing to overload and factors relating to understrength (i.e., less strength than computed by
acceptable calculating procedures). Overloads may arise from changing the use for which
the structure was designed, from underestimation of the effects of loads by oversimpli-
fication in calculation procedures, and from effects of construction sequence and meth-
ods. Understrength may result from adverse variations in material strength, workmanship,
dimensions, control, and degree of supervision, even though individually these items are
within required tolerances.
Conventionally, the term “safety factor” has been used in working stress design to desig-
nate the ratio between the yield stress (real, as for steel; nominally defined, as for concrete)
and the allowable working stress. Such use has resulted in structures and structural elements
having the same “safety factor” but considerably different variance in their strength to ser
vice load ratio. Thus the term “safety factor” as conventionally applied has little meaning
with respect to the prediction of strength.
The variability in the ratio of the strength to service load in the working stress method
was an important reason for the change to the strength design method. To distinguish
between the term “safety factor,” as used in working stress design, and the ratio of strength
to service load, the term “load factor” was adopted for the latter.
The purpose of a safety provision is to limit the probability of failure and yet permit
economical structures. Obviously, if cost is no object, it is easy to design a structure whose
35
2 . 6 S A F E T Y P R OV I S I O N S — G E N E R A L 35
probability of failure is nil. To arrive properly at a suitable degree of safety, the relative
importance of various items must be established. Some of those items are
1. Seriousness of a failure, either to humans or goods.

2. Reliability of workmanship and inspection.
3. Expectation of overload and to what magnitude.
4. Importance of the member in the structure.
5. Chance of warning prior to a failure.
By assigning percentages to those five items and evaluating the circumstances for any
given situation, proper values for overload factors U and strength reduction factors φ can be
established. The background and practicalities leading to the present strength design proce-
dure are discussed later in this section.
The ACI Code strength design method has traditionally divided the safety provisions into
two parts; U factors to account for the probability of overload, and φ factors to account for
probability of understrength. The requirement for strength design may be expressed as follows:
design strength ≥ required strength (2.6.1)
or
φ Sn ≥ Su (2.6.2)
where Sn represents the code-defined strength of the member computed by using stand-
ard assumptions and specified material strengths; it is referred to as the “nominal”
strength. The term φ Sn is called the “design strength.” The “required strength,” Su, cor-
responds to the member actions (or load effects) computed from a structural analysis
using the U factors. In terms of member actions (or load effects), Eq. (2.6.2) may be
written as follows:
φ Pn ≥ Pu (2.6.3)
φ M n ≥ M u (2.6.4)
φ Vn ≥ Vu (2.6.5)
φ Tn ≥ Tu (2.6.6)
where Pn, Mn, Vn, and Tn are the “nominal” strengths in axial compression (or tension),
bending moment, shear, and torsion, respectively. Similarly, Pu, Mu, Vu, and Tu are the mem-
ber actions (or load effects) computed under factored loads in axial compression (or ten-
sion), bending moment, shear, and torsion, respectively.
For many years the ACI Code used U and φ factors that resulted from combined expe-
rience and historical precedent to arrive at the appropriate numerical values. Over time,
attention focused on using the theory of probability as a basis for a design code, thus pro-
viding a more rational basis for the components comprising the U and φ factors. A series of
papers [2.5–2.9] gives an excellent overview of this approach, and an interesting collection
of discussion and opinion is also available [2.8].
One U factor combination that has been used for many years is that for gravity load:
U = 1.2 D +1.6 L (2.6.7)
where D and L are the service dead load and live load (or dead load and live load effects,
such as axial force, bending moment, or shear). In this load combination, the service live
load is assigned a larger load factor than that of the service dead load to account for the
greater uncertainty that exists in computing the value and distribution of the actual service
live load. In other words, it is implied that dead loads can be computed with greater cer-
tainty than live loads.
36
The previous example provides some rationale for the use of different load factors with
loads of different types (live loads, snow loads, wind loads, etc.), but it gives no insight into
the appropriateness of the assigned values—that is, whether the load factors provide for
adequate safety. Benjamin and Lind [2.6] have stated the following five safety conditions,
which form the basis for the current practice in structural analysis and design:
1. The probability of a real loading in excess of the nominal service load D + L must be
satisfactorily small.
2. The probability of a real loading in excess of the factored loading, say U = 1.2D +
1.6L, must be very small or near to zero during the life of the structure.
3. The probability of unsatisfactory performance at the factored load U must be
satisfactorily small.
4. The probability of unsatisfactory performance under a load test must be very small
or near to zero.
5. The probability of unsatisfactory performance at the service load D + L must be
practically zero.
It is noted that conditions 1 and 2 relate to the overload factors U along with analysis
methods, whereas conditions 3, 4, and 5 relate to the factor φ for understrength as well as
the methods for computing strength.
The probability of overload is certainly independent of the material (i.e., concrete, steel,
wood, or masonry) supporting the load. Detailed discussions of the implications of using
design reliability and probability-based load criteria are given by Ellingwood [2.10, 2.11],
MacGregor, Mirza, and Ellingwood [2.12], and MacGregor [2.13]. The subject is treated
in general by Galambos, Ellingwood, MacGregor, and Cornell [2.14, 2.15]. Corotis [2.16]
and Israel, Ellingwood, and Corotis [2.17] reflect thinking regarding unified load criteria
for all materials.
Today, the load factors and load combinations used in design are based on probabilistic
analysis and survey data elicited from design experts and are specified by the ASCE/SEI 7-
10 Standard, Minimum Design Loads for Buildings and Other Structures [2.18]. Developed
for use with design specifications for traditional structural materials, including structural
concrete, steel, masonry, and wood, the ASCE/SEI 7 standard forms the basis for the load
factors and load combinations specified by model codes such as the International Building
Code [2.19]. Accordingly, the 2014 ACI Code uses the ASCE/SEI 7-10 [2.18] load factors
and load combinations U along with appropriate φ factors in the main body of the Code.
2.7 SAFETY PROVISIONS—A CI CODE LOAD FACTORS

AND STRENGTH REDUCTION FACTORS
Load Factors and Load Combinations U
The basic load factors and load combinations U as given by ACI-5.3.1 are
U = 1.4 D (2.7.1)
U = 1.2 D + 1.6 L + 0.5 ( Lr or S or R ) (2.7.2)
U = 1.2 D + 1.6 ( Lr or S or R ) + (1.0 L or 0.5W ) (2.7.3)
U = 1.2 D + 1.0W + 1.0 L + 0.5 ( Lr or S or R ) (2.7.4)
U = 1.2 D + 1.0 E + 1.0 L + 0.2 S (2.7.5)
U = 0.9 D + 1.0W (2.7.6)
U = 0.9 D + 1.0 E (2.7.7)

37
2 . 7 S A F E T Y P R OV I S I O N S 37
where D is dead load; L is live load; Lr is roof live load; S is snow load; R is rain load;
W is wind load; and E is the earthquake-induced load. The load factor on the live load L
in Eqs. (2.7.3), (2.7.4), and (2.7.5) is permitted to be reduced to 0.5 except for garages,
areas occupied as places of public assembly, and all areas where the live load L is greater
than 100 psf.
The load factors in Eqs. (2.7.1) through (2.7.7) assume that D, L, Lr, S, and R are all
unfactored, service-level loads. However, the wind load factor in Eqs. (2.7.3), (2.7.4) and
(2.7.6) assumes that the wind load W is based on strength-level loads, computed in accord-
ance with the provisions of ASCE/SEI 7-10 [2.18]. If the wind load W is based on service-
level loads, then 0.8W must be used in place of 0.5W in Eq. (2.7.3), and 1.6W instead of
1.0W in Eqs. (2.7.4) and (2.7.6). Similarly, the earthquake load factor of 1.0 in Eqs. (2.7.5)
and (2.7.7) assumes that the earthquake-induced load E is based on strength-level seismic
forces. If the earthquake-induced load E is computed using service-level seismic forces, a
higher load factor on the earthquake-induced load E must be used.
If present, loads due to weights and pressures of fluids, F, loads due to lateral earth pres-
sure, H, and structural actions (or load effects) due to the restraint of volume change caused
by temperature changes, by creep and shrinkage, and by differential settlement, T, must
also be included in the basic load combinations given above. Appropriate load factors for
these and other loads are given in ACI-5.3.6 through ACI-5.3.12.
It is noted that in applying the load combinations, the effect of one or more loads not
acting simultaneously also must be investigated. Also note that loads must always be taken
to cause the more severe effect and that their effect can be either additive or subtractive.
Comments on Load Factors and Load Combinations

The load factors and combinations are based on the notion that in addition to the dead load
(considered to be permanent), one of the variable loads takes its maximum value, while the
rest take an arbitrary value less than the maximum. Consider, for example, the load combi-
nation given by Eq. (2.7.2). In this load combination, the live load, L, is considered to take
its maximum value, with a load factor of 1.6, while Lr, S, or R takes an arbitrary value less
than the maximum, and thus is assigned a lower load factor. In contrast, in the load combi-
nation given by Eq. (2.7.3), Lr, S, or R is assumed to be acting at its maximum value, with a
load factor of 1.6, while the live load, L, (or the wind load, W) takes an arbitrary value less
than the maximum, and thus has a lower load factor.
The term 1.2D is common to all load combinations except in Eqs. (2.7.1), (2.7.6) and
(2.7.7). The load combination in Eq. (2.7.1) considers the dead load only with a load
factor of 1.4 and represents an extreme case. The load combinations in Eqs. (2.7.6) and
(2.7.7) have a reduced load factor of 0.9 on the dead load, D, because they are intended
to capture the cases in which structural actions (or load effects) from wind, W, or earth-
quake, E, counteract those from gravity loads. A common example is a roof girder, where
the effects from wind loads may cause uplift, thereby counteracting the actions from the
girder self-weight.
Strength Reduction Factors φ

The factors φ for understrength are called strength reduction factors according to Chapter 21
of the ACI Code. These factors account for the probability of understrength due to varying
material properties, varying dimensions and tolerances, imperfections, and different failure
modes. All other things being equal, brittle failures are assigned lower strength reduction
factors to safeguard the structure against failures that may be catastrophic, or that may
occur with little or no warning. Conversely, noncatastrophic, ductile failure modes, where
member strength can be reached with ample warning of failure, are assigned larger strength
reduction factors.
38
The φ factors in ACI-21.2.1 and ACI-21.2.2 are as follows:

φ Factor
1. Flexure (with or without axial force)a
Tension-controlled sections 0.90
Compression-controlled sections
Spirally reinforced 0.75
Other 0.65
2. Shear and torsion 0.75
3. Bearing on concrete 0.65
4. Post-tensioned anchorage zones 0.85
5. Struts, ties, nodal zones, and bearing areas in strut-and-tie models 0.75
6. Brackets and corbels 0.75
7. Plain concrete elements 0.60
aFor combined axial load and flexure, both axial load and bending moment are subject to the same φ factor, which may be
variable as discussed in Chapter 10.
2.8 SERVICEABILITY PROVISIONS—G ENERAL

As noted earlier in this chapter, the structure must be designed so that it protects the life of
the occupants (i.e., it is safe), but also so that it can serve its intended purpose throughout its
service life (i.e., it must provide for serviceability). Some serviceability issues that can be
as important as strength in the design of reinforced concrete structures are excessive deflec-
tion, detrimental cracking, excessive amplitude or undesirable frequency of vibration, and
excessive noise transmission. Any one, or a combination, of the strength and serviceability
factors may provide a criterion for the limit of structural usefulness. For example, service-
ability can govern the design of beams and slabs in long span floors, where the minimum
depth of these members is often dictated by deflection or vibration limits, rather than by
strength considerations. While serviceability issues (e.g., excessive deflection) do not gen-
erally pose a threat to the safety of the occupants, they can have severe economic conse-
quences. Thus, the designer must adequately address them in design.
The design requirement for serviceability may be expressed as follows:
allowable limit ≥ performance under service loads (2.8.1)
or
Sallowable ≥ Sservice_loads (2.8.2)
where Sallowable represents the specified limit not to be exceeded to ensure adequate
behavior and functionality of the structure during its service life. The term Sservice_loads is the
calculated performance of the structural member or system under service-level loads.
In terms of the service conditions that are commonly considered in the design of rein-
forced concrete structures, Eq. (2.8.2) may be expressed, for example, as follows:
∆ allowable ≥ ∆ service_loads (2.8.3)
wallowable ≥ wservice_loads (2.8.4)
Fallowable ≥ Fservice_loads (2.8.5)
where Δallowable represents the allowable limit on the deflection of a member or on the lateral
drift of the structural system; wallowable is the maximum permitted crack width; and Fallowable is
the limit on the frequency of vibration of the floor system. Similarly, Δservice_loads, wservice_loads,
39
2 . 1 0 H A N D B O O K S A N D C O M P U T E R S O F T WA R E 39
and Fservice_loads are the deflection, crack width, and frequency of vibration of the member
or structural system, respectively, computed from a structural analysis under unfactored,
service loads.
2.9 SERVICEABILITY PROVISIONS—A CI CODE

Serviceability limits depend primarily on the intended function and occupancy of the struc-
ture, and on the type of structural system. It is thus difficult to specify allowable service-
ability limits that apply to all types of structures. Chapter 24 of the ACI Code provides
minimum requirements for serviceability in reinforced concrete structures. For example,
maximum allowable deflections [i.e., Δallowable in Eq. (2.8.3)] in roofs and floor systems are
given in Table 24.2.2 of the ACI Code. This table provides limits on the immediate and
long-term deflections for elements that support or are attached to non-structural elements,
as well as for elements that do not support or are not attached to structural elements. These
limits, though largely based on engineering judgment and years of experience, are not nec-
essarily adequate in every case. Depending on the functional needs of the structure, tighter,
more restrictive deflection limits may be required.
It must be emphasized that the performance of the structural member or system (e.g.,
Δservice_loads) is computed using service-level, unfactored loads: that is, using the loads
expected from normal use. Service-level load combinations are not provided in the main
provisions of the ACI Code, though some guidance and recommendations are given in
the ACI Code Commentary. In addition, the commentary to Appendix C of ASCE/SEI
7-10 [2.18] provides recommendations for load combinations to be used in serviceability
checks. Note that the wind load W used in the load combinations of the safety provisions
(Section 2.7) is based on strength-level loads and should not be used for checking serv-
iceability; its use for evaluating serviceability would be too conservative. The selection
of the appropriate wind load for evaluating the performance of the structure under service
conditions is, in part, a matter of engineering judgment. The commentary to Appendix C
of ASCE/SEI 7-10 [2.18] provides guidance for the calculation of the wind load, Wa,
based on service-level wind speeds that are appropriate for use in checking serviceability
requirements.
Design for serviceability (control of deflections, cracking, and vibrations) in accor
dance with Chapter 24 of the ACI Code is treated in detail later in Chapter 12 of this
textbook. However, basic design considerations and recommended practices for adequate
serviceability are discussed throughout the text, together with strength requirements where
appropriate.
2.10 HANDBOOKS AND COMPUTER SOFTWARE

This textbook does not devote much space to the use of design aids. Once concepts and
principles are thoroughly understood, the use of curves and tables can greatly speed up
design. Several handbooks are in common use: The Reinforced Concrete Design Manual
[2.20], published by ACI; CRSI Design Handbook 2008 [2.21], published by the Concrete
Reinforcing Steel Institute; and PCI Design Handbook [2.22], published by the Prestressed
Concrete Institute. These publications contain useful tables and charts that can speed up
design for the experienced designer.
The importance of correct and clear detailing work cannot be overemphasized. For
this the reader is referred to ACI Detailing Manual (SP-66) [2.23], which contains typical
detailing of steel reinforcement, engineering and placing drawings, and other reference
data regarding materials and sizes.
Today, computer software is used for a considerable portion of the design computations.
The designer, especially the inexperienced engineer, should be wary of using commercial
software (for analysis or design, or for both) to do computations that the designer/engineer
does not thoroughly understand and would be unable to formulate in the absence of a com-
puter program.
40
2.11 DIMENSIONS AND TOLERANCES

Although the designer may tend to think of dimensions, clearances, and bar locations as
exact, practical considerations require that there be accepted tolerances. These tolerances
are the permissible variations from dimensions given on drawings.
Overall dimensions of reinforced concrete members are usually specified by the engi-
neer in whole inches for beams, columns, and walls, sometimes half-inches for thin slabs,
and often 3-in. increments for more massive elements such as plan dimensions for foot-
ings. Formwork for the placing of these members must be built carefully so that it does
not deform excessively under the action of workmen, construction machinery loads, and
wet concrete [2.24]. Based on accepted practice, ACI Committee 117 [2.25] has provided
recommended tolerances for concrete construction and materials. Examples of accepted
tolerances for variation in cross-sectional dimensions of columns and beams and in
the thickness of walls are + 1 2 in. and –3 8 in. when the specified dimension is greater
than 12 in. but not exceeding 3 ft [2.25]. For concrete footings, accepted variations in
plan dimensions are +2 in. and –1 2 in. [2.25], whereas the thickness has an accepted
tolerance of –5% of the specified thickness [2.25]. The strength reduction factor φ is
intended to account for (among other reasons) a case in which several acceptable toler-
ances might adversely combine to reduce the strength from that computed using specified
dimensions.
Reinforcing bars are normally specified in 3-in. length increments. Placement tolerances
for reinforcement are given in the ACI Code [ACI-26.6.2.1(a)] in terms of the effective
depth d (distance from compression face of concrete to center of tension steel), and in terms
of the specified concrete cover in flexural members, walls, and compression members, as
follows:
Effective Depth, d Tolerances
On Effective Depth d On Specified Concrete Cover
(in.) (mm) (in.) (mm) (in.) (mm)

Smaller of (–1/3) × specified cover or
d≤8 200a ± 3 8 ±10a – 3 8 –10a

d>8 200a ± 1 2 ±13a – 1 2 –13a
These values are from ACI 318-14M, and are not hard conversions from ACI 318-14.
a
Since the effective depth and the clear concrete cover are both components of total depth,
the tolerances on those dimensions are directly related. When the tolerances on bar place-
ment and cover accumulate, the overall dimension tolerance may be exceeded; thus field
adjustment may have to be made [2.25]. This may be particularly important for very thin
sections such as in prestressed, precast, and shell structures.
For location of bars along the longitudinal dimension, and of bar bends, the tolerance is
±2 in. (±50 mm) except at discontinuous ends of brackets and corbels, where the tolerance
shall be ± 1 2 in. (±13 mm), and ±1 in. (±25 mm) at the discontinuous ends of other members
[ACI-26.6.2.1(b)].
41
2.12 ACCURACY OF COMPUTATIONS

When one understands that variations in material strength, for both steel and concrete, and
in dimensions are inevitable (and acceptable), it becomes clear that design calculations for
reinforced concrete structures do not require a high degree of precision.
The designer should place highest priority on determining proper location and length of
steel reinforcement to carry the tension forces. Failures, when they occur, generally result
from gross underestimation of tensile forces or failure to identify ways in which the struc-
ture or element will behave under loads. Failures are rarely the result of carrying too few
significant figures in the design computations. However, significant figures may be lost in
arithmetic operations, and gross errors sometimes result from sloppiness.
Results indicated on the display of a calculator or computer, no matter how many digits,
rarely indicate a precision of more than two or three significant figures. Recording of results
on computation sheets should not exceed four significant figures, primarily for systematic
control and checking of computations. While the extra digits recorded may not at first seem
harmful, the resulting many-digit values make difficult the scanning of computations to
detect gross errors (in addition to being a waste of time).
SELECTED REFERENCES
2.1. J. G. MacGregor. “Safety and Limit States Design for Reinforced Concrete,” Canadian Journal
of Civil Engineering, 3, December 1976, 484–513.
2.2. Frank Kerekes and Harold B. Reid Jr. “Fifty Years of Development in Building Code Requirements
for Reinforced Concrete,” ACI Journal, Proceedings, 50, February 1954, 441–472.
2.3. Chester P. Siess. “Writing the Code—More Than 40 Years on Committee 318” [interview con-
ducted by Nancy L. Gavlin], Concrete International, 20, November 1998, 37–42.
2.4. ACI–ASCE Committee 327. “Ultimate Strength Design,” ACI Journal, Proceedings, 52, January
1956, 505–524.
2.5. Robert G. Sexsmith and Mark F. Nelson. “Limitations in Application of Probabilistic Concepts,”
ACI Journal, Proceedings, 66, October 1969, 823–828.
2.6. Jack R. Benjamin and N. C. Lind. “A Probabilistic Basis for a Deterministic Code,” ACI Journal,
Proceedings, 66, November 1969, 857–865.
2.7. C. Allin Cornell. “A Probability- Based Structural Code,” ACI Journal, Proceedings, 66,
December 1969, 974–985.
2.8. R. C. Reese, D. E. Allen, C. A. Cornell, Luis Esteva, R. N. White, R. G. Sexsmith, and George
Winter. “Probabilistic Approaches to Structural Safety,” ACI Journal, Proceedings, 73, January
1976, 37–49.
2.9. Haresh C. Shah and Robert G. Sexsmith. “A Probabilistic Basis for the ACI Code,” ACI Journal,
2.10. Bruce Ellingwood. “Reliability of Current Reinforced Concrete Designs,” Journal of the

Structural Division, ASCE, 105, ST4 (April 1979), 699–712.
2.11. Bruce Ellingwood. “Reliability Based Criteria for Reinforced Concrete Design,” Journal of the
2.12. J. G. MacGregor, S. A. Mirza, and B. Ellingwood. “Statistical Analysis of Resistance of
Reinforced and Prestressed Concrete Members,” ACI Journal, Proceedings, 80, May–June
1983, 167–176.
2.13. James G. MacGregor. “Load and Resistance Factors for Concrete Design,” ACI Journal,

Proceedings, 80, July–August 1983, 279–287.
2.14. Theodore V. Galambos, Bruce Ellingwood, James G. MacGregor, and C. Allin Cornell.

“Probability Based Load Criteria: Assessment of Current Design Practice,” Journal of the
Structural Division, ASCE, 108, ST5 (May 1982), 959–977.
2.15. Bruce Ellingwood, James G. MacGregor, Theodore V. Galambos, and C. Allin Cornell.

“Probability Based Load Criteria: Load Factors and Load Combinations,” Journal of the
Structural Division, ASCE, 108, ST5 (May 1982), 978–997.
2.16. Ross B. Corotis. “Probability-
Based Design Codes,” Concrete International, 7, April
1985, 42–49.
42
2.17. Morris Israel, Bruce Ellingwood, and Ross Corotis, “Reliability-Based Code Formulations for
Reinforced Concrete Buildings,” Journal of Structural Engineering, ASCE, 113, 10 (October
1987), 2235–2252.
2.18. ASCE. American Society of Civil Engineers, Minimum Design Loads for Buildings and Other
Structures (ASCE/SEI 7-10). Reston, VA: American Society of Civil Engineers, 2010.
2.19. International Code Council. International Building Code. Country Club Hills, IL: International
Code Council, Inc., 2012.
2.20. Ronald Janowiak, Michael Kreger, and Antonio Nanni, Editors. The Reinforced Concrete Design
Manual [SP-17(11)1)]. American Concrete Institute, Farmington Hills, MI, 2012, 338 pp.
2.21. CRSI Design Handbook 2008 (10th ed.), Concrete Reinforcing Steel Institute, Schaumberg,
IL, 2008.
2.22. PCI Design Handbook—Precast and Prestressed Concrete (7th ed.), Prestressed Concrete
Institute, Chicago, 2010, 804 pp.
2.23. ACI Committee 315. ACI Detailing Manual–2004 [SP-66(04)]. American Concrete Institute,
Farmington Hills, MI, 2004, 212 pp.
2.24. ACI Committee 347. Guide to Formwork for Concrete. American Concrete Institute, Farmington
Hills, MI, 2014, 36 pp.
2.25. ACI Committee 117. Specification for Tolerances for Concrete Construction and Materials (ACI
117-10) and Commentary, (ACI 117-10). Farmington Hills, MI: American Concrete Institute,
2010, 78 pp.
2.26. CRSI. “Construction Tolerance Conflicts in Reinforced Concrete,” Engineering Data Report
Number 40. Schaumberg, IL: Concrete Reinforcing Steel Institute, 1995.
2.27. P. R. Morgan, T. E. Ng, N. H. M. Smith, and G. D. Base. “How Accurately Can Reinforcing
Steel Be Placed? Field Tolerance Measurement Compared to Codes,” Concrete International, 4,
October 1982, 54–65.
CHAPTER 3
FLEXURAL BEHAVIOR AND
STRENGTH OF BEAMS
3.1 GENERAL INTRODUCTION
Reinforced concrete beams are primarily subjected to a combination of flexure and shear.
Axial forces, if present, are generally negligible. If properly designed, a reinforced con-
crete beam should be able to undergo large, noticeable deformations prior to failure, pro-
viding ample warning to occupants about the risk of a potential failure. It is thus of utmost
importance to have a thorough understanding of the behavior of beams and the factors that
control their flexural response. The material presented in this chapter is intended to serve
this purpose.
University of Wisconsin Stadium, Madison; rectangular tapered beams, cantilevers, and rigid
frame (Photo by C. G. Salmon).
4
44 C H A P T E R 3 F L E X U R A L B E H A V I O R A N D S T R E N G T H O F B E A M S
3.2 FLEXURAL BEHAVIOR AND STRENGTH

OF RECTANGULAR SECTIONS
Consider in Fig. 3.2.1 a simply supported reinforced concrete beam with two concentrated
loads on top and supported at the bottom. Under such loading and support conditions,
flexure-induced stresses will cause compression at the top and tension at the bottom of the
beam. Concrete, which is strong in compression but weak in tension, resists the stresses
in the compression zone, while steel reinforcing bars are placed in the bottom of the beam
to resist tension once flexural cracking has occurred. As the applied load is gradually
increased from zero to failure of the beam (ultimate condition), a well-designed beam may
be expected to behave in the following manner:
1. Initially, when the applied load is low, the stress and strain distribution is essen-
tially linear over the depth of the section. The tensile stresses in the concrete are low
enough that the entire cross section remains uncracked. The flexural strain and stress
distribution is as shown in Fig. 3.2.1(a).
2. On increasing the applied load, the tensile stress at the bottom of the beam reaches the
tensile strength of the concrete, at which point a transverse (flexural) crack forms. Right
after cracking, the neutral axis moves upward and a sudden increase in tensile stress
in the steel reinforcement occurs. Just below the neutral axis, tensile stresses remain
below the tensile strength of the concrete, leaving a small depth of the beam uncracked.
These tensile stresses in the concrete, however, offer only a small contribution to flexural
strength. At this stage, concrete behavior in compression is essentially linear.
3. As load is further increased, the neutral axis moves farther upward and the steel rein-
forcement reaches its yield strength [Fig. 3.2.1(b)]. After yielding, flexural cracks
widen as the beam deflection increases. The concrete stress distribution in the com-
pression zone becomes highly nonlinear [see Fig. 3.2.1(c)]. It is possible for the steel
reinforcement to develop large tensile strains and reach strain hardening. A relatively
modest increase in load beyond first yield is to be expected. Beam deflections, how-
ever, are large at the point of maximum load, with strains in the steel several times
the yield strain.
4. Failure occurs when the concrete strain at the top of the beam reaches the crushing
strain of concrete. Prior to failure, a small decrease in load is expected as the concrete
becomes highly nonlinear and the neutral axis depth starts to increase, leading to a
decrease in the lever arm between the internal compressive and tensile forces.
Basis of Flexural Strength

Consider the rectangular beam cross section shown in Fig. 3.2.2(a). Such a beam is assumed
to reach its flexural strength when the extreme concrete fiber in compression reaches the
crushing strain of the concrete εcu. The assumed strain and stress distributions for comput-
ing flexural strength are shown in Fig. 3.2.2. Two main assumptions have been made in
drawing this figure:
1. It has been assumed that plane sections have remained plane after bending up to fail-
ure of the beam. This traditional beam theory assumption, strictly valid for elastic,
homogeneous beams, has been verified by tests and found to be a good assumption
for reinforced concrete beams loaded up to failure [3.1]. This assumption permits the
use of a linear strain distribution over the beam depth [see Fig. 3.2.2(b)].
2. The steel reinforcing bars are assumed to be perfectly bonded to the surrounding con-
crete such that there is no slip of bars relative to concrete. This assumption implies
that the change in strain in the steel reinforcement and that in the concrete surround-
ing the steel are the same. While locally at a crack this assumption is not true, the
average strain measured over several cracks indicates the concrete and the steel bars
work together reasonably well.
45
3. 2 F L E XU RAL B E H AV I O R A N D S T R E N G T H O F R E C TA N G U L A R S E C T I O N S 45
b
εc fc
h d
εs
fs
(a)
εc fc
εs = εy
fs = fy
(b)
εc
εs > εy
fs ≥ fy
(c)
Figure 3.2.1 Strain and stress distributions in a reinforced concrete beam under increasing load.
From Fig. 3.2.2 it may also be noted that tensile stresses in the concrete are ignored, as
their contribution to flexural strength is negligible. This implies that the portion of the beam
cross section below the neutral axis does not affect the flexural strength. Thus, the impor-
tant depth dimension for computing strength is the effective depth d rather than the overall
depth h. The effective depth is defined as the distance from the extreme fiber in compression
to the centroid of the tension steel area. When the tension steel comprises bars in several
layers, the effective depth d is taken as the distance from the extreme compression fiber to
the centroid of the combined area, with all bars assumed to have the same strain for calcula
tion of tensile stress and force in the reinforcement.
Although the compressive stress distribution in a beam may be expected to have the
same general shape as that obtained from a test cylinder (or cube) specimen [1.60, 3.1,
3.2], the strain at crushing and the peak compressive stress will not be necessarily the same.
A cylinder or a cube specimen is subjected to a uniform strain distribution during a standard
test, while the compression zone of a beam has a strain gradient [Fig. 3.2.2(b)]. Also, the
in-place concrete compressive strength will likely differ from that obtained using a cylinder
compressive test (see Section 1.7). The exact shape of the stress distribution in the compres
sion zone of a beam cannot thus be determined, but it can be expressed in terms of three
coefficients ( k1 , k2 , k3 ) that define the magnitude and position of the internal compression
force resultant, as shown in Fig. 3.2.2(c). The k3 coefficient represents the ratio between
the peak stress in the compression zone of a beam to the cylinder compressive strength
fc′. The coefficient k1 represents the ratio between the average stress and the peak stress k3 fc′.
The k2 factor, on the other hand, defines the location of the resultant compressive force C
from the extreme compressive fiber, as a fraction of the neutral axis depth c. According to
Hognestad [3.1, 3.2], this approach was first proposed by Stüssi in 1932. Further develop-
ments occurred during the 1950s, led by Hognestad himself and others [3.1–3.6].
The stress-strain response of the reinforcing steel is typically assumed to be the same as
that obtained from a tension test. Thus, if the strain in the steel is known, the tensile stress
and corresponding tensile force can be calculated. In many cases, such as in the calculation
46
k3 f’c
b k2 c
εcu = crushing strain
MR
c
N.A.
C = k1k3 f’c bc
d
As T = Asfy (for εs ≥ εy)

εs
d = effective depth; distance from compression face to centroid of tension steel
(a) Singly (b) Strain condition (c) Stress condition

reinforced when MR is when MR is reached
beam reached
Figure 3.2.2 Strain and stress conditions when flexural strength is reached.
of the nominal flexural strength (see Section 3.4), any increase in stress beyond yield due
to strain hardening is ignored.
When concrete crushing occurs (commonly a sudden occurrence), the strain in the tension
steel may be either larger or smaller than the yield strain ε y. If the steel area As relative to the
area of the beam cross section is low enough, the steel will yield prior to crushing of the con-
crete; the result will be a ductile failure mode in which there is large deformation. On the other
hand, the reinforcing steel, when used in large quantities, would remain elastic at the time of
crushing of the concrete, causing a brittle or sudden mode of failure. Sections in which the steel
yields prior to the concrete reaching the strain εcu are said to be underreinforced. Otherwise, the
section is overreinforced. The ACI Code has provisions that limit the strain in the tension steel,
with the intent of ensuring a ductile mode of failure (see Section 3.6).
The internal compressive force C [Fig. 3.2.2(c)] is the resultant of the compressive stresses
acting on the compression concrete area, which may be considered a “stress solid” volume as
C = favg cb = k1 k3 fc′ cb (3.2.1)
Assuming that the steel yields prior to crushing of the concrete, the tensile force T is
T = As f y (3.2.2)
Equilibrium requires C = T , from which

As f y
c= (3.2.3)
k1 k3 fc′ b
The flexural strength, M R , may then be expressed as
M R = T (arm ) = T ( d – k2 c ) = As f y ( d – k2 c ) (3.2.4)
Substituting Eq. (3.2.3) for c into Eq. (3.2.4) gives
 k As f y 
M R = As f y  d − 2 (3.2.5)
 k1 k3 fc′ b 
One may note that if the flexural strength MR is the quantity of interest, it is readily
obtainable from Eq. (3.2.5) if the quantity k2 / ( k1 k3 ) is known. It is thus not necessary
to have values for k1, k2, or k3 individually if the value for the combined term is known.
Experimental results [1.59, 3.3] have established values for the combined term, as well
as the individual k values, with some of the results shown in Fig. 3.2.3. From this figure,
k2 / ( k1 k3 ) ranges from about 0.55 to 0.63. It should be mentioned that these values, which
were experimentally determined when crushing of the concrete occurred at the compress-
ion face, necessarily involved variation in the crushing strain εcu for the various tests.
47
3 . 3 W H I T N E Y R E C TA N G U L A R S T R E S S D I S T R I B U T I O N 47
fc’, MPa
20 25 30
1.0
k1
Coefficients k2/(k1k3)
0.5
k2
2500 3000 3500 4000 4500 5000

f’c , psi
200 250 300 350

f’c , kgf/cm2
Figure 3.2.3 Values of coefficients that define compression zone in a flexural member. (Adapted
from Hognestad, Hanson, and McHenry [3.3].)
3.3 WHITNEY RECTANGULAR STRESS DISTRIBUTION

Flexural strength based on the approximately parabolic stress distribution of Fig. 3.2.2(c)
may be computed by using Eq. (3.2.5) with given values of k2 /(k1 k3 ). However, it is desir-
able for the designer to have a simple method in which basic static equilibrium is used.
Although the use of a stress block in lieu of a more sophisticated inelastic stress distribu-
tion has been proposed since the early 1910s [3.1], the most widely used stress block is that
developed by Whitney in the 1930s [3.7, 3.9]. Whitney realized that it was not necessary to
know the real stress distribution in the compression zone of a beam to accurately predict its
flexural strength. He further added [3.7]: “Whatever it actually is, it must have an average
intensity, fc, and an effective depth, a.” Assuming a peak concrete compressive stress equal
to fc′ (i.e., k3 = 1.0), Whitney proposed the use of an average stress of 0.85 fc′ over a depth
a [Fig. 3.3.1(c)].
f’c 0.85 f’c

b a
k2 c
2
C Mn
a = β1 c
c C
d a
d–
h N.A. 2
As
T T
d = effective depth; distance from compression face of concrete to centroid of tension steel
(a) Beam (b) Assumed stress distribution (c) Whitney rectangular stress block
Figure 3.3.1 Definition of Whitney rectangular stress distribution.

48
A relationship between the depth of the stress block a and the neutral axis depth c
was proposed by Mattock, Kriz, and Hognestad [3.6], where a = β1c, determined so that
a /2 = k2 c. For fc′ ≤ 4000 psi, β1 = 0.85, decreasing at a rate of 0.05 for each 1000 psi of
fc′ in excess of 4000 psi.1 The value of β1 is not to be taken less than 0.65. The ACI Code
explicitly accepts the Whitney rectangular stress block (ACI-22.2.2.4.1) in combination
with the β1 factor proposed by Mattock, Kriz, and Hognestad (ACI-22.2.2.4.3) for the deter-
mination of nominal flexural strength (see Section 3.4).
3.4 NOMINAL FLEXURAL STRENGTH M n —

RECTANGULAR SECTIONS HAVING TENSION
REINFORCEMENT ONLY
The quantities defining a rectangular section with tension reinforcement only are b, d, and As
[Fig. 3.3.1(a)]. Such a section is said to be singly reinforced. The steel area As is, of course,
furnished by the combined area of an actual number of reinforcing bars. Protective concrete
cover is necessary around the bars not only to make the steel and concrete act together but
also very importantly, to provide protection against corrosion and fire. Minimum cover
requirements are generally prescribed by code (see ACI-20.6.1).
For computation of nominal flexural strength Mn, the following assumptions (ACI-22.2)
are made:
1. The strength of members shall be based on satisfying the applicable conditions of

equilibrium and compatibility of strains.
2. Strain in the steel reinforcement and in the concrete shall be assumed directly propor-
tional to the distance from the neutral axis (except for deep members covered under
ACI-9.9).
3. The maximum usable strain εcu at the extreme concrete compressive fiber shall be
assumed equal to 0.003 (Fig. 3.4.1.).
4. The tensile strength of the concrete is to be neglected (except for certain prestressed
concrete conditions).
Steel reinforcement
fy
εy = 0.00207
for fy = 60 ksi
Stress
Concrete
εy = fy /Es
f’c εcu = ACI Code-defined crushing strain
1 2 3 4 5 6 7 8 9 10
× 10–3
Strain
Figure 3.4.1 Stress-strain curves for concrete and steel reinforcement, and definitions of εy
and εcu.
1 For SI, ACI 318-14M gives β1 = 0.85 for fc′ ≤ 28 MPa and reduces by 0.05 for each 7 MPa of fc′ in excess of 28 MPa, but
not less than 0.65.
49
3.4 NOMINAL FLEXURAL STRENGTH 49
5. The stress-strain relationship for nonprestressed steel shall be assumed to be elastic–

perfectly plastic (i.e., strain hardening is ignored) (Fig. 3.4.1). The modulus of elas-
ticity of such steel reinforcement may be taken as 29,000,000 psi (200,000 MPa or
2,040,000 kgf/cm2).
6. For practical purposes, the relationship between the concrete compressive stress dis-
tribution and the concrete strain when nominal strength is reached may be taken as an
equivalent rectangular stress distribution (ACI-22.2.2.4), wherein a concrete stress
intensity of 0.85 fc′ is assumed to be uniformly distributed over an equivalent com-
pression zone bounded by the edges of the cross section and a straight line located
parallel to the neutral axis at a distance a = β1c from the fiber of maximum com-
pressive strain. The distance c from the fiber of maximum strain to the neutral axis
is measured in a direction perpendicular to that axis. The value of β1 is given by the
following equations:
For fc′ ≤ 4000 psi,
β1 = 0.85 (3.4.1)
For fc′ > 4000 psi,
 f ′ − 4000 
β1 = 0.85 − 0.05  c ≥ 0.65 (3.4.2)
 1000 
It should be noted that assumption (6) describes the Whitney rectangular compressive

stress distribution with the provisions for determination of neutral axis depth by Mattock,
Kriz, and Hognestad [3.6] (see Section 3.3). However, other shapes of stress solids, such as
the trapezoid and the parabola, have been used [3.10] and are acceptable for use according
to ACI-22.2.2.3.
The nominal flexural strength M n, using the equivalent rectangular stress block, is
obtained from Fig. 3.3.1(c) as follows:
C = 0.85 fc′ ba (3.4.3)
T = As f y (3.4.4)
where the use of f y assumes that the steel yields prior to crushing of the concrete (required
for beams, as discussed in Section 3.6).
Equilibrium requires that C = T ; thus
As f y
a= (3.4.5)
0.85 fc′ b
Taking moments about the compressive force resultant, C
 a  a
M n = T  d −  = As f y  d −  (3.4.6)
 2   2
which, on substituting Eq. (3.4.5) into Eq. (3.4.6), gives
 As f y 
M n = As f y  d − 0.59 (3.4.7)
 fc′ b 
Note that 0.59 corresponds to k2 /( k1 k3 ) of Eq. (3.2.5). One may also note that β1 is needed
only to establish the neutral axis location for determining the steel strain. As long as the
steel strain exceeds ε y , as shown in Fig. 3.2.2, the nominal flexural strength is not affected
by the value of β1 .
50
EXAMPLE 3.4.1
Determine the nominal flexural strength Mn of the rectangular section shown in Fig. 3.4.2.
given fc′ = 5000 psi, f y = 65, 000 psi, b = 14 in., d = 21.5 in., and As = 4–#8 bars.
0.85f’c
a = 0.80c a
14”
εcu = 0.003
2
4.32” C
a
21.5” d–
17.18” 2
As = 3.16
sq in.
T
εs
(a) Cross section (b) Assumed strain and

stress distribution
Figure 3.4.2 Singly reinforced beam of Example 3.4.1.
SOLUTION
Assume the steel has already yielded when the nominal strength is reached. From
Fig. 3.4.2 the internal forces are

C = 0.85 fc′ ba = 0.85(5)(14)a = 59.5a

T = As f y = 3.16(65) = 205 kips
For equilibrium, C = T ; therefore
205
a= = 3.45 in.
59.5

β1 = 0.80 for fc′ = 5000 psi
The neutral axis position is
a 3.45
c= = = 4.32 in.
β1 0.80
The strain in the tension steel when the compressive strain 0.003 is reached at the
extreme concrete fiber is, by straight-line proportion,
d −c 17.18
εs = (0.003) = (0.003) = 0.012
c 4.32
fy
εy = = 0.0022
Es
When the nominal flexural strength is reached, ε s is five times ε y , which means that large
deformation has occurred before the crushing of concrete. The assumption that the steel
yields has been validated.
The nominal flexural strength is
 a  a
M n = C  d −  or T  d − 
 2  2
= 205 ( 21.5 − 1.73)

1
= 339 ft-kips
12
51
3.5 BALANCED STRAIN CONDITION 51
3.5 BALANCED STRAIN CONDITION

At the balanced strain condition [Fig. 3.5.1(b)], the maximum strain at the extreme concrete
compressive fiber just reaches the crushing strain εcu when the tension steel reaches a strain
εy. As mentioned earlier, the ACI Code requires the use of ε cu = 0.003 ( ACI − 22.2.2.1).
The balanced strain condition will exist for an amount of tension steel Asb such that the
neutral axis depth will be cb, as shown in Fig. 3.5.1(b). If the actual As were greater than
Asb , equilibrium of internal forces ( C = T ) would require an increase in the depth a of the
compression stress block (and thereby would also make c exceed cb), so that the strain εs
would be less than εy when the extreme fiber in compression reaches ε cu (= 0.003 accord-
ing to the ACI Code). The failure of this beam will be sudden when the concrete reaches
the strain 0.003, with little deformation (steel does not yield) to warn of impending failure.
On the other hand, when the actual As is less than Asb , the tensile force decreases so
that internal force equilibrium leads to a reduction in the depth a of the compression stress
block (and thereby makes c less than cb), giving a strain εs greater than ε y . In this case, with
the steel having yielded and if As is substantially less than Asb , the beam will have noticea-
ble deflection prior to the concrete reaching the crushing strain of 0.003.
Therefore, the relative amount of tension steel with respect to that in the balanced strain
condition can be used to determine whether the failure is ductile (gradual, giving warning)
or brittle (sudden, without warning).
Reinforcement Ratio at Balanced Strain Condition for Rectangular Beam

Having Tension Reinforcement Only
The symbol ρ , the tension reinforcement ratio (often called tension reinforcement percent-
age), may be conveniently used to represent the relative amount of tension reinforcement
in a beam. Thus, using the dimensions of Fig. 3.5.1,
As (3.5.1)
ρ=
bd
The reinforcement ratio in the balanced strain condition, ρb , may be obtained by applying
the equilibrium and compatibility conditions. From the linear strain condition, Fig. 3.5.1(b),
cb ε cu 0.003 87, 000 (3.5.2)
= = =
d ε cu + ε y 0.003 + f y / 29, 000, 000 87, 000 + f y
The compressive force Cb is

Cb = 0.85 fc′ bab = 0.85 fc′ bβ1cb
The tensile force Tb is
Tb = Asb f y = ρb bdf y
0.85f’c
b
εcu = 0.003
ab = β1cb Cb
cb
Asb = ρbbd
Tb = Asbfy
εs = εy = fy /Es
(a) Cross section (b) Strain diagram (c) Internal forces
Figure 3.5.1 Balanced strain condition.

52
Equating Cb to Tb gives
0.85 fc′ bβ1cb = ρb dbf y

0.85 fc′  cb  (3.5.3)
ρb = β1  
fy  d
which on substitution of Eq. (3.5.2) gives
0.85 fc′  87, 000  2

ρb = β1   (3.5.4)
fy  87, 000 + f y 
where the stresses f y and fc′ are in psi.

It may be noted that ρb depends only on the material properties and is independent of
the beam size. In other words, given fc′ and f y , the balanced reinforcement ratio ρb can be
readily determined. Values of ρb for typical concrete and steel yield strengths are shown in
Table 3.6.1. It can be seen that ρb increases with an increase in fc′, while it decreases with an
increase in f y . For the common case of fc′ = 4000 psi and f y = 60 ksi, ρb = 2.85%, increas-
ing to 3.77% for fc′= 6000 psi . When the yield strength is increased to 80 ksi while the
same concrete strength is maintained, ρb becomes 66% of that for fy = 60 ksi. On the other
hand, if Grade 40 steel is used, ρb increases by about 70% compared to ρb for fy = 60 ksi.
3.6 TENSION- AND COMPRESSION-C ONTROLLED

SECTIONS
The ACI Code classifies reinforced concrete sections as either tension- or compression-
controlled (ACI-21.2.2) depending on the magnitude of the net strain in the reinforcement
closest to the tension face, ε t , as the concrete reaches the assumed crushing strain ε cu of
0.003 as follows (see Fig. 3.6.1):
εt ≥ 0.005
Tension-controlled sections:
Compression-controlled sections: εt ≤ εy
For sections having more than one layer of tension steel, the strain limits must be checked
for the layer of reinforcement closest to the tension face of the member (at a distance dt
from the extreme compressive fiber; see Fig. 3.6.1). For a typical Grade 60 reinforcing bar,
fy 60
εy = = = 0.0021
Es 29, 000
Thus, the ACI Code allows the compression-controlled strain limit εy to be taken as 0.002
for Grade 60 reinforcement. Sections with tension steel strains between the compression-
and tension-controlled strain limits are considered to be in a transition region.
Minimum Net Tension Steel Strain εt, min

To have reasonable assurance of a ductile mode of failure, ACI-9.3.3.1 requires that the
net tension steel strain εt be equal to or greater than 0.004 in beams subjected to a factored
axial load less than 0.1Ag fc′ (where Ag is the gross area of the cross section). For Grade
60 steel reinforcing bars, this requirement corresponds to about twice the yield strain ε y .
2 For SI,
0.85 fc′  600 
ρb = β1   (3.5.4)
fy  600 + f y 
with fc′ and fy in MPa, and β1 as given in footnote 1 in Section 3.3.
53
3.6 TENSION- AND COMPRESSION-CONTROLLED SECTIONS 53
The writers, however, recommend that beam sections be designed to be tension controlled
(i.e., ε t ≥ 0.005), to achieve a more ductile design with ample warning prior to failure. As
discussed in Example 3.6.1, while the area of steel corresponding to the tension-control
limit is slightly less than that for ε t = 0.004, the design flexural strength will be approxi-
mately the same, the beam with larger steel area being less ductile.
When reinforcement with yield strength higher than 60 ksi is used (e.g., Grade
80 steel), the tension-controlled strain limit may need to be increased above 0.005 because
the amount of plastic deformation in the steel reinforcement when ε cu = 0.003 will be less.
For example, for Grade 80 steel, the plastic deformation for a net tensile strain of 0.005
would be only 1.8ε y compared to 2.5ε y for Grade 60 steel. In such cases, to ensure adequate
inelastic deformation prior to failure, it may be desirable to design for a higher net tensile
strain.
Reinforcement Ratio Corresponding to Strain Limit

for Tension-Controlled Sections ρtc
Similar to the concept of balanced reinforcement ratio, there is a unique amount of rein-
forcement that will cause the tension steel to reach the net tensile strain ε t of 0.005 for
tension-controlled sections just as the extreme concrete fiber in compression reaches ε cu of
0.003. For sections with one layer of tension reinforcement, dt is equal to d and, using the
procedure discussed in Section 3.5, the following expression for the reinforcement ratio ρ
can be obtained
0.003 + ε y
ρ (εt ) = ρb (3.6.1)
0.003 + ε t
For the strain limit corresponding to tension-controlled sections, Eq. (3.6.1) becomes
0.85β1 fc′ 0.003 β f′

ρtc = ρ ( ε t = 0.005) = = 0.319 1 c (3.6.2)
fy 0.008 fy
This value represents the maximum reinforcement ratio that ensures a tension-controlled
section design. In other words, beam sections reinforced with a reinforcement ratio ρ ≤ ρtc
will develop a net tensile strain ε t greater than or equal to 0.005. It should be noted that
because Eq. (3.6.2) was derived assuming all tension reinforcement in a single layer (i.e.,
dt = d), the reinforcement ratio obtained from this equation will lead to a strain ε t slightly
greater than 0.005 for sections with more than one layer of tension reinforcement. Further,
it should be kept in mind that Eq. (3.6.2) applies only to rectangular sections or nonrectan-
gular sections with a rectangular compression zone (see Section 4.4).
b
εcu = 0.003
dt = distance from extreme

dt fiber in compression to
centroid of the layer of
reinforcement closest to
the tension face.
εt ≥ 0.005 Tension controlled

εt ≤ εy Compression controlled
Figure 3.6.1 Definition of tension-and compression-controlled sections (ACI-21.2.2).

54
TABLE 3.6.1 REINFORCEMENT RATIO ρ AT VARIOUS STRAIN CONDITIONS FOR SINGLY

REINFORCED RECTANGULAR BEAMS
fc′ = 3000 psi 4000 psi 5000 psi 6000 psi 8000 psi
fy β1 = 0.85 β1 = 0.85 β1 = 0.80 β1 = 0.75 β1 = 0.65

40,000 psi ρb = 0.0371 0.0495 0.0582 0.0655 0.0757
ρtc = ρ (ε t = 0.005) = 0.0203 0.0271 0.0319 0.0359 0.0414
ρ (ε t = 0.0075) = 0.0155 0.0206 0.0243 0.0273 0.0316
ρ (ε t = 0.01) = 0.0125 0.0167 0.0196 0.0221 0.0255
60,000 psi ρb = 0.0214 0.0285 0.0335 0.0377 0.0436

ρtc = ρ (ε t = 0.005) = 0.0135 0.0181 0.0212 0.0239 0.0276
ρ (ε t = 0.0075) = 0.0103 0.0138 0.0162 0.0182 0.0210
ρ (ε t = 0.01) = 0.0083 0.0111 0.0131 0.0147 0.0170
80,000 psi ρb = 0.0141 0.0188 0.0221 0.0249 0.0288

ρtc = ρ (ε t = 0.005) = 0.0102 0.0135 0.0159 0.0179 0.0207
ρ (ε t = 0.0075) = 0.0077 0.0103 0.0121 0.0137 0.0158
ρ (ε t = 0.01) = 0.0063 0.0083 0.0098 0.0110 0.0128
Values of ρ corresponding to net tensile strains ε t of 0.005, 0.0075, and 0.01 for typical
concrete and steel yield strengths are shown in Table 3.6.1. As discussed earlier, it is desir-
able to use a higher net tensile strain limit when reinforcing steel with yield strength higher
than 60 ksi is used.
Strength Reduction Factors φ

A tension-controlled section will exhibit a ductile failure mode with visible flexural cracks
and deflection. For these types of members, the strength reduction factor φ is 0.90 (see
Section 2.7). In contrast, compression-controlled sections are expected to fail suddenly,
with little or no warning of impending failure. This lack of ductility is accounted for by
using a lower strength reduction factor. The ACI Code specifies two strength reduction fac-
tors for compression-controlled sections depending on the type of transverse reinforcement
used. Sections containing spirals as transverse reinforcement (spirally reinforced) have a
φ factor of 0.75, while those with stirrups or ties have a φ factor of 0.65. In practice, spiral
reinforcement is used almost exclusively in columns. Because spirals provide better con-
finement to the concrete (see Chapter 10), a larger φ factor is assigned to spirally reinforced
sections. For sections that fall in the transition region between a compression-controlled
and a tension-controlled section, the φ factor must be computed by using linear interpola-
tion, as illustrated in Fig. 3.6.2. The linear equations in terms of the extreme tensile strain
εt are as follows:
For sections with stirrups or ties,
φ = 0.65 + 0.25
(ε t − εy ) ≤ 0.90 (3.6.3)
(0.005 − ε ) y

5
For spirally reinforced sections,
φ = 0.75 + 0.15
(ε t − εy ) ≤ 0.90 (3.6.4)
(0.005 − ε ) y

where ε y is the yield strain of the tension reinforcement layer closest to the tension face,
which for Grade 60 steel can be taken as 0.002.
The φ equations can also be expressed in terms of the ratio c / dt , where c is the neutral
axis depth measured from the compression face of the member and dt is the distance from
the extreme compressive fiber to the layer of reinforcement closest to the tension face.
When Grade 60 bars are used for the longitudinal reinforcement, the strength reduction
factor φ can be obtained as follows:
For sections with stirrups or ties,
 1 5
φ = 0.65 + 0.25  −  ≤ 0.90 (3.6.5)
 c / dt 3 
For spirally reinforced sections,

(3.6.6)
 1 5
φ = 0.75 + 0.15  −  ≤ 0.90
 c /d t 3 
It should be noted that beam sections designed with a reinforcement ratio ρ less than or
equal to ρtc will be tension controlled and thus, they will be assigned a strength reduction
factor φ = 0.90.
εt, min = 0.004 for beams with

factored axial load less than
0.1Agf’c (ACI-9.3.3.1)
φ
0.90
ρ ≤ ρtc (Eq. 3.6.2)
Spiral
0.75
0.65
Other
Compression Tension
controlled Transition controlled
εt = εy εt = 0.005
Figure 3.6.2 Variation of φ with net tensile strain εt. (Adapted from ACI-R21.2.2.)
EXAMPLE 3.6.1
The beams shown in Fig. 3.6.3 have the same cross-sectional dimensions but have been
reinforced with different amounts of tension steel. Both have two layers of reinforce-
ment, but the beam of Fig. 3.6.3(a) has 3–#8 bars for the upper layer and 3–#9 bars for
the bottom layer. The beam of Fig. 3.6.3(b) has 3–#8 bars in both layers. Compute the
nominal flexural strength, the corresponding φ factor, and the design flexural strength
φMn. Use fc′ = 4500 psi and f y = 60, 000 ksi.
(Continued)
56
Example 3.6.1 (Continued)
0.85fc’
14”
εcu = 0.003
a/2
c = 7.29” a = 6.02”
C = 322 K
dt = 17.5” d = 16.40”
3 – #8
εs T=C
2.5”
εt = 0.0042
3 – #9
(a)
14” 0.85f’c
εcu = 0.003
a/2
c = 6.44” a = 5.31” C = 284 K
dt = 17.5” d = 16.25”
3 – #8
εs T=C
2.5”
εt = 0.0052
3 – #8
(b)
Figure 3.6.3 Beam sections and strain and stress diagrams at nominal condition for Example 3.6.1.
SOLUTION
(a) Referring to Fig. 3.6.3(a), the effective depth d, which is measured to the centroid of
the reinforcement, is calculated as
3 (1.0 )(17.5) + 3 ( 0.79)(15)
d= = 16.40 in.
3 (1.0 ) + 3 ( 0.79)
Because there is more than one layer of tension steel, the distance dt is calculated to the
layer of steel closest to the extreme tension face. In this case, dt = 17.5 in.
Assuming the steel yields before the concrete reaches the strain ε cu, the internal ten-
sile and compressive forces, as well as the depth of the stress block a, are calculated as
T = As f y = 3 ( 0.79) + 3 (1.0 ) (60 ) = 322 kips = C
C 322
a= = = 6.02 in.
0.85 fc′ b 0.85 ( 4.5)(14 )
For = 4500, β1 = 0.825. Thus, the neutral axis depth c is
a 6.02
c= = = 7.29 in.
β1 0.825
Once the depth of the neutral axis c is known, the strain at the level of the centroid of the
steel and at the layer of steel closest to the extreme tension face can be readily calculated:
0.003 0.003
εs = (d − c) = (16.40 − 7.29) = 0.0037
c 7.29
0.003 0.003
εt =
c
( dt − c ) =
7.29
(17.50 − 7.29) = 0.0042
(Continued)
57
The strain at the level of the centroid of the steel reinforcement εs is greater than
the yield strain for Grade 60 steel (0.0021). Thus, the assumption that the steel yields
prior to the concrete reaching the strain εcu was correct. While the strain at the centroid
of the steel reinforcement εs is less than 0.004, the strain at the layer of steel clos-
est to the extreme tension face (ε t = 0.0042) is greater than the minimum strain ε t,min
of 0.004 required by ACI-9.3.3.1. Because εt is less than 0.005, the section is in the
transition region for the purpose of determining the strength reduction factor φ. Using
Eq. (3.6.3), a value of φ = 0.83 is obtained.
The nominal flexural strength of the section M n and the design flexural strength φ M n
can now be calculated as
 a  a  6.02  1
M n = T  d −  = C  d −  = 322  16.40 −
  = 359 ft-kiips
2  2  2  12

φ M n = 0.83 (359) = 300 ft-kips
(b) The analysis of the section shown in Fig. 3.6.3(b) follows the same procedure as for
the section in Fig. 3.6.3(a). Since the area of steel in the layer closest to the extreme
tension fiber is smaller, the effective depth d needs to be calculated again. The depth dt ,
however, remains the same as the location of this layer of steel has not changed. Thus,
3 ( 0.79)(17.5) + 3 ( 0.79)(15)
d= = 16.25 in.
6 ( 0.79)
Since the area of steel has been reduced compared to the section analyzed in part (a), it
is already known that the steel will yield prior to the concrete reaching the strain ε cu. The
internal tensile and compressive force resultants, as well as the depth of the stress block
a and of the neutral axis c, are then calculated as follows:
T = As f y = 6 ( 0.79) (60 ) = 284 kips = C
C 284
a= = = 5.31 in.
0.85 fc′ b 0.85 ( 4.5)(14 )
a
c= = 6.44 in.
β1
With the neutral axis depth c known, the strain at the level of the centroid of the
steel (ε s) and at the layer of steel closest to the extreme tension face (ε t ) can be readily
calculated:
0.003 0.003
εs = (d − c) = (16.25 − 6.44) = 0.0047
c 6.44
0.003 0.003
εt =
c
( dt − c ) = 6.44 (17.50 − 6.44) = 0.0052
In this case, the section is tension controlled (ε t > 0.005 ) and φ = 0.9 [compared with
φ = 0.83 in part (a)]. The nominal and design flexural strengths are then calculated as
 a  a  5.31 1
M n = T  d −  = C  d −  = 284  16.25 −
  = 322 ft-kiips
2  2  2  12

φ M n = 0.90 (322 ) = 290 ft-kips
(Continued)
58
On comparing the design flexural strength φ M n for both sections, it can be seen that
while the area of steel in the beam of part (b) was 88% of that used in part (a), the design
flexural strengths of the sections differed by only 3%. The result is due to the lower
φ factor assigned to beam sections in the transition region, such as the beam of part (a),
compared to that assigned to beam sections that are tension controlled. From an eco-
nomical and performance viewpoint, it is thus better to design beam sections to be ten-
sion controlled (ε t ≥ 0.005) rather than in the transition region (0.004 ≤ ε t < 0.005 ).
3.7 MINIMUM TENSION REINFORCEMENT

When steel reinforcement in a flexural member is only a small amount because the factored
moment Mu is low, the beam may perform uncracked at service load. The computation of
nominal moment strength M n, in accordance with Section 3.4, assumes the tension concrete
to be cracked. Thus, the computed nominal strength M n for a section having a small amount
of reinforcement could be less than the strength of the same section of plain uncracked
concrete (i.e., no reinforcement). Since a ductile failure mode is desired, the lowest amount
of steel permitted should be at least the amount that would equal the strength of an unrein-
forced section. The desired relationship then becomes
strength of reinforced  strength of plain 

concrete beam, φ M  ≥ concrete beam, M  (3.7.1)
 n   cr 
The flexural strength of a plain concrete beam, known as the cracking moment M cr , is
achieved when the extreme fiber in tension reaches the modulus of rupture fr (see Section
1.8). For normal-weight concrete, ACI-19.2.3.1 uses
fr = 7.5 fc′ (3.7.2)3
As discussed in Section 1.8, this modulus of rupture is meant to approximately represent an

average value from test results. Modulus of rupture, however, is highly variable, as shown
in Fig. 1.8.1.
Assuming plain (nonreinforced) concrete as an elastic homogeneous material, the flex-
ure formula gives Mcr as
Ig
M cr = fr = fr S t (3.7.3)
yt
where Ig = moment of inertia of the gross concrete cross section

yt = distance from the neutral axis to the extreme fiber in tension
St = elastic section modulus with respect to extreme fiber in tension; Cbwh2/6
C = coefficient to account for flanges of T-sections; C = 1.0 for rectangular section
bw = width of section; width of web for T-section
h = overall depth of the section
On expanding, Eq. (3.7.3) becomes
bw h 2
M cr = 7.5 fc′ C (3.7.4)
6
3 For SI, with fr and fc′ in MPa,

fr = 0.62 fc′ (3.7.2)
59
3.7 MINIMUM TENSION REINFORCEMENT 59
For a reinforced concrete section, using Eq. (3.4.6),
 a
φ M n = φ As f y  d −  (3.7.5)
 2
Substituting Eqs. (3.7.4) and (3.7.5) into Eq. (3.7.1) gives
 a 7.5 fc′ Cbω h 2

φ As f y  d −  ≥ M cr = (3.7.6)
 2 6
Estimating a/2 as 0.05d, and using φ = 0.9 for flexure, Eq. (3.7.6) gives
 7.5 fc′   h  2  C 
As ,min =      bω d (3.7.7)
 f y   d   5.13 
Rectangular Sections—For rectangular sections, C = 1.0 and h/d varies from about 1.05 to
1.2. For such sections, Eq. (3.7.7) becomes
1.6 fc′ 2.1 fc′

As ,min = bω d to As ,min = bω d (3.7.8)
fy fy
T-Sections Having Slab in Compression—For this case, C will vary from about 1.3 to 1.6
for practical ranges of ratios of flange thickness to overall depth and of flange width to web
width. Taking C = 1.5 along with h/d from 1.05 to 1.2, Eq. (3.7.7) becomes
2.4 fc′ 3.2 fc′

fy fy
The minimum area of tension reinforcement required in the ACI Code is (ACI-9.6.1.2)
3 fc′ bw d
As ,min = bw d ≥ 200 (3.7.10)4
fy fy
Note that 200bw d / f y in Eq. (3.7.10) will govern for concrete strengths fc′ less than 4444 psi.
For Grade 60 steel and fc′ ≤ 4444 psi, As ,min = 0.0033bw d . For higher concrete strengths,
As,min increases, being As ,min = 0.0045bw d for fc′ = 8000 psi. This minimum area of ten-
sion reinforcement is required “at every section where tensile reinforcement is required by
analysis …” (ACI-9.6.1.1), unless the area of reinforcement provided is at least one-third
greater than the area required by analysis (ACI-9.6.1.3).
T-Sections Having Slab in Tension—For this case, C will vary from about 3.0 to 4.0 for
a practical range of ratios of flange thickness to overall depth and of flange width to web
width. Taking C = 3.5 along with h/d from 1.05 to 1.2, Eq. (3.7.7) becomes
5.6 fc′ 7.4 fc′

fy fy
ACI-9.6.1.2 states that minimum reinforcement “For a statically determinate beam with
a flange in tension, … ” is to be computed using Eq. (3.7.10), with bw replaced by either
2bw or the width of the flange, whichever is smaller.
4 For SI, with fy and fc′ in MPa, and bw and d in mm,

fc′
As ,min = bw d (3.7.10)
4 fy
60
3.8 DESIGN OF RECTANGULAR SECTIONS

IN BENDING HAVING TENSION
REINFORCEMENT ONLY
In the design of rectangular sections in bending with tension reinforcement only, the prob-
lem is often to determine b, d, and As , from the required value of M n = Mu / φ, and the given
material properties.
The two conditions of equilibrium are
C = T (3.8.1a)
and
 a  a
Mn = T  d −  = C  d −  (3.8.1b)
 2  2
Since there are three unknowns but only two conditions, there are many possible solutions.
If the reinforcement ratio ρ is preset, then, from Eq. (3.8.1a),
C = 0.85 fc′ ba = ρ bd f y = T
 fy 
a = ρ d (3.8.2)
 0.85 fc′
Substituting Eq. (3.8.2) into Eq. (3.8.1b),
 ρ  fy  
M n = ρ bd fy  d −  d (3.8.3)
 2  0.85 fc′ 
2
A strength coefficient of resistance Rn is obtained by dividing Eq. (3.8.3) by bd and letting
fy
m= (3.8.4a)
0.85 fc′
Thus
Mn  1 
Rn = = ρ f y  1 − ρ m (3.8.4b)
bd 2  2 
The relationship between ρ and Rn for various values of fc′ and fy is shown in Fig. 3.8.1.
In some situations the values of b and d may be preset, which is equivalent to having
Rn preset; then ρ may be determined by solving the quadratic equation Eq. (3.8.4b). Thus
 1 
Rn = ρ f y  1 − ρ m
 2 
from which
1 2 mRn 
ρ= 1 − 1 −  (3.8.5)
m fy 
The procedure (neglecting certain practical decisions for now) to be used in the strength
design of rectangular sections with tension reinforcement only involves the following steps.
1. Assume a value of ρ equal to or less than the reinforcement ratio corresponding to the
tension control limit, ρtc [see Eq. (3.6.2), repeated here, or Table 3.6.1], but greater
61
3 . 8 D E S I G N O F R E C TA N G U L A R S E C T I O N S 61
than the minimum of ACI-9.6.1.2 (Eq. 3.7.10). As the section would then be tension
controlled, the minimum strain limit ε t = 0.004 of ACI-9.3.3.1 will be satisfied.
β1 fc′
ρ ( ε t = 0.005) = ρtc = 0.319
f y [3.6.2]
where
for f c′ ≤ 4000 psi β1 = 0.85

 f c′ − 4000 
for f c′ > 4000 psi β1 = 0.85 − 0.05  ≥ 0.65
 1000 
2. If b and d are unknown, determine the required bd 2 from
required M n Mu / φ
required bd 2 = =
Rn Rn
in which
 1 
Rn = ρ f y  1 − ρ m
 2 
with
fy
m=
0.85 fc′
As the section is being designed with ρ ≤ ρtc , φ = 0.9.
1800
f’c = 8000 12
120
f’c = 8000 fy = 60,000
1600 fy = 80,000 11
110
10 100
1400
f’c = 6000
9
fy = 80,000 f’c = 6000 90 Coefficient of Resistance, Rn, kgf/cm2
Coefficient of Resistance, Rn, MPa
Coefficient of Resistance, Rn, psi
1200 fy = 60,000
8 80
f’c = 4000
1000 7 70
fy = 80,000
f’c = 4000 6 60
800 fy = 60,000
5 50
600 4 40
3 30
400
Mn = Rn bd 2 2 20
200
1 10
Min ρ controls
0 0 0
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035
Reinforcement Ratio, ρ
Figure 3.8.1 Strength curves (Rn vs ρ) for singly reinforced rectangular sections. Upper
limit of curves is at ρmax, that is, reinforcement ratio corresponding to ε t = 0.004.
62
3. Choose a suitable set of values of b and d so that the bd 2 provided is approximately

equal to the bd 2 required. (Note: Actually, d is not chosen; rather, the overall depth h
is chosen and d is computed from h while maintaining the desired minimum protec-
tive cover.)
4. Determine the revised value of ρ after computing Rn = M n / bd 2 for the selected sec-
tion using one of the following methods:
a. By formula (most exact),
1 2 mRn 
ρ= 1 − 1 − 
m fy 
b. By curves (Fig. 3.8.1).
c. By approximate proportion (on the safe side if revised Rn is smaller than original
Rn), the revised ρ is
(revised Rn )
ρ ≈ (original ρ)
(original R n )
It may be noted from Fig. 3.8.1 that the relationship between Rn and ρ is approximately
linear at low values of ρ , even though the actual equation is a quadratic one.
5. Compute As from
As = (revised ρ) (actual bd )
6. Select reinforcement and check the strength of the section to be certain that
φ M n ≥ Mu
EXAMPLE 3.8.1
Determine a set of values b, d, and As that will carry a factored bending moment Mu of
430 ft-kips. Use fc′ = 5000 psi and f y = 60, 000 psi.
SOLUTION
(a) Establish the limits within which the reinforcement ratio ρ can be chosen. From
Eq. (3.6.2), the reinforcement ratio corresponding to the tension control limit is
β1 fc′
ρtc = 0.319
fy
where β1 = 0.80 for fc′ = 5000 psi. Thus
ρtc = 0.319
(0.80)(5) = 0.0212
60
From Eq. (3.7.10) (ACI-9.6.1.2), the minimum percentage area of tension reinforce-
ment is
3 fc′
As ,min = bw d
fy
3 fc′ 3 5000
ρmin = = = 0.00354
fy 60, 000
As 3 5000 is greater than 200 psi, ρmin = 0.00354 .

(Continued)
63
3 . 8 D E S I G N O F R E C TA N G U L A R S E C T I O N S 63
(b) Choose reinforcement ratio ρ, determine corresponding required beam size, and
select actual beam size. Arbitrarily assume that ρ = 0.016 , which is within the
prescribed limits of ρmin and ρtc . Also note that for the chosen value of ρ = 0.016 ,
ε t = 0.0075 (see table 3.6.1.). Thus,
fy 60
m = = = 14.12
0.85 fc′ 0.85 ( 5)
 1   1 
Rn = ρ f y  1 − ρ m = 0.016 (60, 000 ) 1 − ( 0.016 )(14.12 )
 2   2 

= 852 psi
As the selected ρ ≤ ρtc , the section will be tension controlled and φ = 0.90.
Mu 430
required M n = = = 478 ft-kips
φ 0.9
required M n 478 (12, 000 )

required bd 2 = = = 6733 sq in.
Rn 852
Try b = 16 in., d = 6733 / 16 = 20.5 in. The effective depth d is usually 2 3 8 to 2 5 8 in. less
than overall beam depth h if one layer of steel can be used. Assuming that d is 2.5 in. less than
the overall beam depth, h = 23 in. is obtained. However, even numbers are usually preferred
for overall beam depths; thus, h = 24 in. is selected with d = 21.5 in. and b = 16 in.
(c) Compute required As and select actual steel bars, using actual beam size. The
effective depth exceeds that corresponding to ρ = 0.016 . Thus, the required Rn and
required ρ must be revised.
required M n 478 (12, 000 )

required Rn = = = 775 psi
b d2 16 ( 21.5)
2
1 2 mRn 
ρ = 1 − 1 − 
m fy 
1  2 (14.12 )( 775) 
= 1 − 1 −  = 0.0144
14.12  60, 000 

As = ρ bd = 0.0144 (16 )( 21.5) = 4.95 sq in.
Select 5–#9 bars, all in one layer, As = 5 (1.0 ) = 5.0 sq in., which is only slightly larger
than the required As. Other bar combinations are possible, but they may not fit in one
layer or give a total area much larger than that required. The practical selection of bars
is discussed in Section 3.9.
(d) Check design. A review of the correctness of the above computations in which for-
mulas are used should be made using the basic statics shown in Fig. 3.8.2.

T = As f y = 5 (60 ) = 300 kips
T or C 300
a= = = 4.41 in.
0.85 fc′ b 0.85 ( 5)(16 )
 a  4.41 1
M n = (T or C )  d −  = 300  21.5 −  = 482 ft-kips
 2  2  12
(Continued)
64
Since the chosen reinforcement ratio was less than ρtc , the section will be tension
controlled (i.e., ε t ≥ 0.005). For illustration purposes, however, the strain εt will be
calculated. From Fig. 3.8.2, the depth of the neutral axis is
a 4.41
c= = = 5.52 in.
β1 0.80
and the strain in the tension steel is
d −c 21.5 − 5.52
εs = εt = (0.003) = (0.003) = 0.0087 > 0.005
c 5.52
as was anticipated.
(e) Final decision and design sketch. Every design requires a clear decision as its
conclusion, along with an appropriate design sketch.
Use 5–#9 bars in the 16 × 24 in. (d = 21.5 in.) beam, as shown in Fig. 3.8.2.
0.85f’c
16”
εcu = 0.003
C = 300 K
c = 5.52” 4.41”
24” 21.5” 19.29”
5 – #9
εs = εt = 0.0087 T = 300 K
Figure 3.8.2 Section for Example 3.8.1.
3.9 PRACTICAL SELECTION FOR BEAM SIZES,

BAR SIZES, AND BAR PLACEMENT
In the preceding section, the procedure and example for the design of rectangular sections
in bending with tension reinforcement only were based on the assumption that the factored
moment Mu was known. This is rarely the case, however, because the factored moment
must include the effect of the weight of the beam itself, which has not yet been designed. In
reality, then, the dead weight of the beam must be assumed at the outset; a trial beam size
is then obtained and may be readjusted if its effect on the factored moment significantly
differs from the assumed value. For typical design situations, a preliminary estimate for the
self-weight of the beam could be obtained by assuming that the beam depth is 10 to 20%
greater than the minimum depth for which deflections are assumed to be satisfactory (ACI-
9.3.1.1) and a beam depth-to-width ratio of 1.5. These minimum depths are L /16 for simply
supported spans, L /18.5 for spans with one end continuous, L /21 for spans with both ends
continuous, and L /8 for cantilever beams, where L is the span length. Calculation and con-
trol of deflections in flexural members is covered in Chapter 12.
The choice of the steel reinforcement ratio ρ is very much dependent on the desired level
of ductility and limitation on the deflection of the beam. The selection of high reinforce-
ment ratios, while allowing the use of smaller beam sections, may lead to limited ductility
and unacceptable deflections. Years of experience with the working stress method showed
that deflection problems were rarely encountered with beams having a steel reinforcement
ratio ρ not more than one-half the maximum value permitted in the 2002 and earlier edi-
tions of the ACI Code (i.e., 0.75 ρb). For Grade 40, 60, and 80 steel, this corresponds
65
3.9 PRACTICAL SELECTION FOR BEAM SIZES 65
approximately to 0.7ρtc , 0.6ρtc and 0.5ρtc , respectively, where ρtc is the reinforcement ratio
at the tension control limit [see Eq. (3.6.2)]. In general, a tension reinforcement ratio for
singly reinforced sections between 1.0 and 1.5% represents a good preliminary choice to
ensure adequate ductility and reduce the likelihood of encountering deflection problems.
Abendroth and Salmon [3.11] have presented a sensitivity study to show how the variables
in simply supported beam design affect the economics of the design.
TABLE 3.9.1 TOTAL AREAS FOR VARIOUS NUMBERS OF REINFORCING BARS
Nominal Number of Bars

Bar Diameter Weight
Size (in.) (lb/ft) 1 2 3 4 5 6 7 8 9 10
#3 0.375 0.376 0.11 0.22 0.33 0.44 0.55 0.66 0.77 0.88 0.99 1.10
#4 0.500 0.668 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00
#5 0.625 1.043 0.31 0.62 0.93 1.24 1.55 1.86 2.17 2.48 2.79 3.10
#6 0.750 1.502 0.44 0.88 1.32 1.76 2.20 2.64 3.08 3.52 3.96 4.40
#7 0.875 2.044 0.60 1.20 1.80 2.40 3.00 3.60 4.20 4.80 5.40 6.00
#8 1.000 2.670 0.79 1.58 2.37 3.16 3.95 4.74 5.53 6.32 7.11 7.90
#9 1.128 3.400 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00
#10 1.270 4.303 1.27 2.54 3.81 5.08 6.35 7.62 8.89 10.16 11.43 12.70
#11 1.410 5.313 1.56 3.12 4.68 6.24 7.80 9.36 10.92 12.48 14.04 15.60
#14a 1.693 7.65 2.25 4.50 6.75 9.00 11.25 13.50 15.75 18.00 20.25 22.50
#18a 2.257 13.60 4.00 8.00 12.00 16.00 20.00 24.00 28.00 32.00 36.00 40.00
a #14 and #18 bars are used primarily as column reinforcement and are rarely used in beams.
TABLE 3.9.2 MINIMUM BEAM WIDTH (INCHES) ACCORDING TO THE ACI CODEa
Number of Bars in Single Layer of Reinforcement

Add for Each
Size of Bars 2 3 4 5 6 7 8 Added Bar
#4 6.8 8.3 9.8 11.3 12.8 14.3 15.8 1.50

#5 6.9 8.5 10.2 11.8 13.4 15.0 16.7 1.63
#6 7.0 8.8 10.5 12.3 14.0 15.8 17.5 1.75
#7 7.2 9.0 10.9 12.8 14.7 16.5 18.4 1.88
#8 7.3 9.3 11.3 13.3 15.3 17.3 19.3 2.00
#9 7.6 9.8 12.2 14.3 16.6 18.8 21.1 2.26
#10 7.8 10.4 12.9 15.5 18.0 20.5 23.1 2.54
#11 8.1 10.9 13.8 16.6 19.4 22.2 25.0 2.82
#14 9.1 12.5 15.9 19.3 22.6 26.0 29.4 3.40
#18 10.8 15.3 19.8 24.3 28.8 33.3 37.9 4.51
Table shows minmium beam widths when #3 stirrups are used for bars #11 and smaller, and #4 stirrups are used for bars #14 and #18.
For additional bars, add dimmension in last column for each added bar.
For bars of different size, determine from table the beam width for smaller size bars and then add last column figure for each larger bar used.
aAssumes maximum aggregate size does not exceed three-fourths of the clear space between bars (ACI-25.2.1).
1 B
A = 1 -in. clear cover to stirrup
2 A C D
B = stirrup bar diameter

C = For #11 and smaller bars, use twice the diameter of #3 stirrups
(i.e., C = 0.75 in.). For #14 and #18 bars, use C = 0.5db
D = clear distance between bars = db,1 in. or 4 the maximum

aggregate size, whichever is greater 3
Diameter of corner bar is
(db = diameter of largest adjacent longitudinal bar) assumed to be located to
intersect the horizontal
tangent to stirrup bend
6
For the selection of an actual number of bars to meet a total steel area requirement, it is
desirable to tabulate the combined area of several bars at a time. Table 3.9.1 gives bar areas
for up to 10 bars of different sizes. Note that once the area of one bar, say of #8, has been
set at 0.79 sq in., the area of 10 bars becomes 7.90, not 7.85; this practice has been carried
on by tradition.
For placement of bars in reinforced concrete members, the ACI Code specifies a min-
imum concrete cover for the reinforcement based on member and exposure types (ACI-
20.6.1.3). For cast-in-place beams not exposed to weather or in contact with ground, the
minimum specified cover for the longitudinal reinforcement and stirrups is 1 1 2 in. Since
beams are nearly always provided with stirrups surrounding the longitudinal reinforce-
ment, the minimum clear cover to the stirrups will govern. In addition to minimum cover
requirements, ACI-25.2.1 specifies the clearance needed between bars in a single layer to
permit proper concrete placement around them. This clearance for bars in a layer is 1 in.,
or the nominal diameter of the largest bar, or 4 3 the maximum aggregate size, whichever
is greater. When two or more layers of bars are required, the minimum clearance between
layers is 1 in. (ACI-25.2.2). The bars in different layers must be located directly above one
another and not staggered (ACI-25.2.2). Table 3.9.2 gives minimum beam widths for var-
ious numbers of equal-sized bars, computed in the manner described above, but assuming
that aggregate size does not control minimum clear distance between bars.
Developing the tensile force in the reinforcing steel requires the bars to be bonded
with the surrounding concrete. Larger than minimum spacing and clear cover of bars will
improve the bond between the bars and the concrete. Since this chapter has treated only the
selection of cross section and bars thus far, suffice it to say that shorter bar lengths can be
used when 2db clear lateral spacing (minimum width in Table 3.9.3) and 3db clear lateral
spacing (minimum width in Table 3.9.4) can be provided.
Bars are supported from the bottom of the forms, and layers of bars are separated by bar
supports of various types; these are known as bolsters and chairs, some of which are shown
in Table 3.9.5. Bar supports may be made of concrete, metal, or other approved materials.
Usually they are standard factory-made wire bar supports. These remain in place after the
concrete is cast and must have special rust protection on the portions nearest the face of the
concrete.
TABLE 3.9.3 MINIMUM BEAM WIDTH (INCHES) TO SATISFY 2 BAR

DIAMETERS CLEAR SPACING
Number of Bars in Single Layer
Bar Size 2 3 4 5 6 7 8
#4 6.8 8.3 9.8 11.3 12.8 14.3 15.8

#5 7.1 9.0 10.9 12.8 14.6 16.5 18.4
#6 7.5 9.8 12.0 14.3 16.5 18.8 21.0
#7 7.9 10.5 13.1 15.8 18.4 21.0 23.6
#8 8.3 11.3 14.3 17.3 20.3 23.3 26.3
#9 8.6 12.0 15.4 18.8 22.2 25.6 28.9
#10 9.1 12.9 16.7 20.5 24.3 28.1 31.9
#11 9.5 13.7 17.9 22.2 26.4 30.6 34.9
#14 10.8 15.9 20.9 26.0 31.1 36.2 41.2

#18 13.0 19.8 26.6 33.3 40.1 46.9 53.7
Table assumptions:
a. Side cover 1.5 in. each side.
b. #3 stirrups for bars #11 and smaller.
c. #4 stirrups for bars #14 and #18.
d. Since stirrups are bent around 4 stirrup bar diameters, the distance from centroid of bar nearest side face of beam to inside
face of #3 stirrup is taken as 0.75 in. for bars #11 and smaller; and equal to the bar radius for #14 and #18 bars.
67
TABLE 3.9.4 MINIMUM BEAM WIDTH (INCHES) TO SATISFY 3 BAR DIAMETERS CLEAR SPACING
Number of Bars in Single Layer
Bar Size 2 3 4 5 6 7 8
#4 7.3 9.3 11.3 13.3 15.3 17.3 19.3

#5 7.8 10.3 12.8 15.3 17.8 20.3 22.8
#6 8.3 11.3 14.3 17.3 20.3 23.3 26.3
#7 8.8 12.3 15.8 19.3 22.8 26.3 29.8
#8 9.3 13.3 17.3 21.3 25.3 29.3 33.3
#9 9.8 14.3 18.8 23.3 27.8 32.3 36.8
#10 10.3 15.4 20.5 25.6 30.7 35.7 40.8
#11 10.9 16.5 22.2 27.8 33.5 39.1 44.7
#14 12.5 19.2 26.0 32.8 39.6 46.3 53.1
#18 15.3 24.3 33.3 42.4 51.4 60.4 69.5
Table assumptions:
a. Side cover 1.5 in. each side.
b. #3 stirrups for bars #11 and smaller.
c. #4 stirrups for bars #14 and #18.
d. Since stirrups are bent around 4 stirrup bar diameters, the distance from centroid of bar nearest side face of beam to inside face of #3 stirrup is taken as
0.75 in. for bars #11 and smaller; and equal to the bar radius for #14 and #18 bars.
TABLE 3.9.5 STANDARD TYPES AND SIZES OF BAR SUPPORTS (ADAPTED FROM REF. 2.23)
Symbol Bar Support Illustration Type of Support Standard Sizes

SB
Slab bolster 3
, 1,1 1 2, and 2 in. heights in 5-ft
4
5” and 10-ft lengths
SBU Slab bolster upper Same as SB

5”
BB Beam bolster 1, 1 1 2, 2, over 2 to 5 in. heights in

increments of 1 4 in. in lengths of 5 ft
2 21 ”
BBU Beam bolster upper Same as BB
2 21 ”
BC Individual bar chair 3

4 , 1, 1 1 2, and 1 3 4, in. heights
HC Individual high chair 2 to 15 in. heights in increments of

1
4 in.
CHC Continuous high chair Same as HC in 5-ft and 10-ft lengths
8”
CHCU Continuous high chair Same as CHC

upper
8”
68
The following guidelines are suggested to assist the designer further in making choices
for beam sizes, bar sizes, and bar placement. These suggestions may be regarded as
accepted practice; they are not ACI Code requirements. Undoubtedly, situations will arise
in which the experienced designer will, for good and proper reasons, make a selection not
conforming to the guidelines.
For Beam Sizes
1. Use whole inches for overall dimensions; slabs, however, may be in 1 2-in.
increments.
2. Beam stem widths are most often in multiples of 2 or 3 in., such as 9, 10, 12, 14, 15,
16, and 18 in.
3. Minimum specified clear cover is measured from outside the stirrup or tie to the face
of the concrete. (Thus beam effective depth d has rarely, if ever, a dimension to the
whole inch.)
4. An economical rectangular beam proportion is one in which the overall depth-to-
width ratio is between about 1.5 to 2.0.
5. For T-shaped beams, the flange thickness typically represents about 20% of overall
depth (see Chapter 4 for treatment of T-shaped sections).
For Reinforcing Bars
6. Maintain bar symmetry about the centroidal axis, which lies at right angles to the
bending axis (i.e., symmetry about the vertical axis in usual situations).
7. Use at least two bars wherever flexural reinforcement is required. Use bars #11 and
smaller for usual sized beams. The use of a larger number of bars of smaller diame-
ter leads to improved bond and better crack control than can be achieved with fewer
bars of larger diameter.
8. Use no more than two bar sizes and no more than two standard sizes apart for steel
in one face at a given location in the span (e.g., #7 and #9 bars may be acceptable,
but #9 and #4 bars would not).
9. Place all bars in one layer if practicable. Try to select bar size so that no less than
two and no more than five or six bars are put in one layer.
10. Follow requirements of ACI-25.2.1 and 25.2.2 for clear distance between bars and
between layers, and for arrangement between layers.
11. When different sizes of bars are used in several layers at a location, place the largest
bars in the layer nearest the face of the beam.
EXAMPLE 3.9.1
Select an economical rectangular beam size and select bars using the ACI Code
requirements. The beam is a simply supported span of 40 ft; it is to carry a live load of
1.38 kips/ft and a dead load of 1 kip/ft (not including beam weight). Without actu-
ally checking deflection, use a reinforcement ratio ρ such that excessive deflection is
unlikely. Use fc′ = 4000 psi, and fy = 60,000 psi.
SOLUTION
(a) Decide on a reinforcement ratio ρ to use. To have reasonable expectation that deflection
will not be excessive, choose ρ at about 0.6ρtc = 0.0108 as recommended. Use ρ = 0.011.
(b) Determine the desired Rn (corresponding to the desired ρ) using Eq. (3.8.4).
fy 60
m= = = 17.65
0.85 fc′ 0.85(4)
1
Rn = ρ f y (1 − ρm)
2
= 0.011(60,000)[1 –0.5(0.011)(17.65)] = 596 psi
(Continued)
69
(c) Determine required Mn using ACI-5.3.1 and a φ factor of 0.9.

M u = 1.2 M D + 1.6 M L
1.38 ( 40 )
2
ML = = 276 ft-kips
8
As the beam span is 40 ft long and simply supported, the minimum required depth in
ACI-9.3.1.1 to avoid the need for deflection check would be L /16 = 40(12)/16 = 30 in.
For the purpose of estimating the beam weight, assume a depth of 32 in. and a width of
22 in. (approximately two-thirds of beam depth). This leads to a beam weight estimated
at 0.733 kip/ft, based on a unit weight of reinforced concrete of 0.15 kips/cu ft.
MD =
(1.0 + 0.733)( 40)2 = 347 ft-kips
8
thus

M u = 1.2 (347) + 1.6 ( 276 ) = 416 + 442 = 858 ft-kips
Mu
required M n = = 953 ft-kips
φ
(d) Determine required bd2 from desired Rn. Use required Mn = 953 ft-kips,
M n 953 (12, 000 )

required bd 2 = = = 19,190 cu in.
Rn 596
(e) Establish beam size. Select width b and determine the corresponding required value
for effective depth d. Make a table of possibilities.
Chosen b Required d
12 40.0
15 35.8
18 32.7 ← try
20 31.0
Selecting the 18-in. width will give a beam whose overall depth is between 1 1 2 and
2 times its width (suggested guideline).
Verify the assumed weight by determining overall depth. Assuming that the bars
to be selected will fit in one layer, the minimum overall depth may be computed (see
Fig. 3.9.1). Assume #8 bars for the longitudinal reinforcement and #3 bars for stirrups.

h = d + 1 12 in. cover + 83 in. diameter stirrup + bar radius, say 12 in.

= d + (2 83 to 2 12 in.)

= 32.7 + 2.5 = 35.2 in.
The overall depth would be in whole inches; a depth of either 36 or 34 in. may thus
be selected (even numbers are often preferred for overall beam depths). Since the guide-
line value of Rn = 596 psi for ρ = 0.011 is not a rigorous requirement, the overall depth
selected could be somewhat less or somewhat more than the computed requirement and
still obtain a desired dimension. In this case, the overall depth (h) of 34 in. is selected
(Continued)
70
Small bars for holding

stirrup in place
d
h
Bar radius
3„ 1
Stirrup, usually to „ diam.
8 2
1
Clear cover, 1 min.
2
(ACl-20.6.1.3.1)
Figure 3.9.1 Quantities added to effective depth d to get overall depth h for beams having one
layer of tension steel.
which, as it is less than the required overall depth for the target reinforcement ratio, will
lead to a slightly larger reinforcement ratio than originally assumed. The stirrup is rein-
forcement to provide shear strength for the beam and should always be allowed for at this
stage of the design. The actual size and spacing of stirrups is determined after the cross
section and tension bars have been selected, as explained in Chapter 5. Try h = 34 in.
(f) Check weight, revise Mu, and select reinforcement.
18 (34 )
w = (0.15) = 0.637 kip / ft
144
(1.0 + 0.637)( 40 )
2
revised M D = = 328 ft-kips

8

revised Mu = 1.2 (328) + 1.6 ( 276 ) = 835 ft-kips
revised required M n = 835 = 927 ft-kips

0.90
Compute the value of d from the overall dimension h, actual d = h – ( ≈ 2.5 in.) for one
layer of bars = 34 – 2.5 = 31.5 in.
When the overall depth is increased or decreased, the clear cover distance (ACI-
20.6.1.3.1) is the one that is held constant; thus the effective depth will shift.
M n 927 (12, 000 )

bd 2 18 (31.5)2
The steel requirement may be determined from Eq. (3.8.5) or from the curves of
Fig. 3.8.1, or approximated by straight-line proportion. Using the latter and knowing that
R = 596 psi for ρ = 0.011
n
find approximate ρ for Rn = 623 psi,

 623 
approximate ρ = 0.011  = 0.0115
 596 
[using Eq. (3.8.5) would have given about the same value]
approximate As = ρ bd = 0.0115 (18)(31.5) = 6.52 sq in.
(Continued)
71
Select 4–#10 and 2–#9 bars, As = 7.08 sq in. (Table 3.9.1). (Note that 3–#10 and 3–#9
having As = 6.81 sq in. provide a smaller As , but the bars cannot be placed in one layer
and still have symmetry about the vertical beam axis, and 7–#9 will not fit in one layer.)
Check whether 4–#10 and 2–#9 will fit into an 18-in. width in one layer.
18 − 2(1.5) − 2(0.375) − 4(1.27) − 2(1.128)

approx. clear spacing between bars =
5
= 1.38 in. > [1.27in. = diameter of # 10] OK
Subtracted from the overall width are the combined values of the minimum clear cover
on both sides (3.0 in.), one stirrup diameter on both sides (0.75 in.), 4–#10 bar diameters
(5.08 in.), and 2–#9 bar diameters (2.26 in.). The result is divided by the number of
spaces between bars, and this is the approximate clearance that must exceed the diame-
ter of the larger bar (ACI-25.2.1).
Note that the above clearance computation is approximate because it assumes the #3
stirrup may be bent tightly around the corner longitudinal bar. ACI-25.3.2 requires the
inside diameter of bends for stirrups to be not less than four stirrup bar diameters for
#5 stirrups and smaller; thus for #3 stirrups the actual curve of the stirrup at the corner
has a radius of 3 4 in., which is larger than the longitudinal bar radius for #11 bars and
smaller. Table 3.9.2 is based on the conservative assumption that the diameter of the
corner longitudinal bar is located to intersect the horizontal tangent to the stirrup bend
(see sketch accompanying Table 3.9.2). Using Table 3.9.2, the minimum required width
for 4–#10 and 2–#9 is

min b = 7.6 + 4 ( 2.54 ) = 17.76 in.
The 7.6 in. is the minimum width for 2–#9 and the 2.54 in. applies to each of the addi-
tional #10 bars (bar diameter plus a spacing of one bar diameter).
(g) Check strength and provide design sketch. Using computed d = 31.5 in.,
C = 0.85 fc′ ba = 0.85(4)18a = 61.2 a

T = As f y = 7.08(60) = 425 kips
425
a= = 6.95 in.
61.2
 a 1
M n = As f y  d −  = 425(31.5 − 3.47) = 994 ft-kips
 2 12
The net strain in the tension steel is
d −c d − a / β1
εs = (0.003) = (0.003)
c a / β1
31.5 − 6.95 / 0.85
= (0.003) = 0.0086 > 0.005
6.95 / 0.85
Thus, the section is tension controlled and the φ factor is 0.90, as assumed.

[φ M n = 0.90(994) = 895 ft-kips] > [ Mu = 835 ft-kips] OK
Use 18 × 34 beam with 4–#10 and 2–#9 bars, as shown in Fig. 3.9.2.
(Continued)
72
0.85f’c a
2 = 3.47”
#3 stirrup C = 61.2a
(assumed) a = 6.95”
d = 31.5”
34”
#10
#9 T = 425k
1 12 ” 18”
Figure 3.9.2 Design for Example 3.9.1.
3.10 NOMINAL FLEXURAL STRENGTH M n OF

RECTANGULAR SECTIONS HAVING BOTH
TENSION AND COMPRESSION REINFORCEMENT
Rectangular sections having both tension and compression reinforcement are also called
“doubly reinforced” sections. Because the compressive strength of concrete is high, the
need for compression reinforcement to obtain adequate strength is not great. In beams
where compression reinforcement might be used to reduce the size of the cross section,
deflections may be excessive and it may be difficult to place all the tension reinforcement
within the width of the beam, even if two or more layers of bars are used. In addition, the
shear stress will become high, possibly calling for a large amount of shear reinforcement.
In fact, the usual reason for the use of compression reinforcement is for deflection control
(to reduce the creep- and-shrinkage-induced deflection).
Another reason to provide compression reinforcement is to increase the ductility of the
beam when, for example, the tension steel yields but the strain εt is less than the minimum
required by the ACI Code (ε t,min = 0.004). Adding compression steel will raise the neutral
axis and decrease the depth of the compression stress block, resulting in an increase in the
strain in the tension reinforcement. Similarly, compression reinforcement can be used to
achieve a tension-controlled section design when the required amount of tension steel in a
singly reinforced beam would lead to a compression-controlled section design.
An area of compression steel equal to one-quarter that of the tension steel can be considered
a reasonable amount for the purpose of increasing ductility and controlling deflections. For
members subjected to earthquake effects, where large inelastic deformations may be expected,
the use of areas of compression steel greater than one-half that of the tension steel is common.
The nominal moment strength Mn of a doubly reinforced section such as that shown in
Fig. 3.10.1 involves using variables b, d, d′, As, As′, fc′, and fy. The computation is similar to
that for the singly reinforced beam except the compressive force C consists of two parts,
one in the concrete and the other in the steel. The compression steel at nominal strength M n
of the beam may or may not be at yield depending on the position of the neutral axis. The
stress in the compression steel used in the nominal strength computation must be compati-
ble with the strain diagram at crushing of concrete.
In Fig. 3.10.1, the area of the concrete under compression in a beam with compression
steel is given by
Aconc = ab − As′
and the corresponding compression resultant equals 0.85 fc′Aconc. The point of action
of this resultant requires, however, the calculation of the centroid of the shaded area in
73
ε cu = 0.003 0.85f’c 0.85f’c

Compression face
d’ d’ Cs = As’ (fs’ – 0.85fc’) Cc Cs
As’ c εs’ a
a/2 Cc = 0.85fc’ ab
= +
h d Aconc
(d – a–2) Mnc (d – d’) Mns
As
T = As fy
εs Asc fy (As – Asc)fy
b
Tension
face
(a) Cross section (b) Strain diagram (c) Stress diagram (d) Moment carried as (e) Moment carried by
a singly reinforced the compression
beam, Mnc steel, Mns
As’
ρ’ = f’s = Compression steel stress = Es ε’s ≤ fy
bd = Compression steel reinforcement ratio
As
ρ= = Tension steel reinforcement ratio Asc = Part of the tension steel to balance Cc
bd
Figure 3.10.1 Strain and stress condition at nominal flexural strength for doubly reinforced beam.
Fig. 3.10.1(a), which is often cumbersome. This calculation, however, can be avoided by

treating the concrete compression zone as rectangular and adjusting the stress in the com-
pression steel, as illustrated next.
Equilibrium requires that
C = T
or
0.85 fc′Aconc + As′ fs′ = As f y
0.85 fc′ ( ab − As′ ) + As′ fs′ = As f y
0.85 fc′ ab + As′ ( fs′− 0.85 fc′ ) = As f y
or
Cc + C s = T
where
Cc = 0.85 fc′ ab
and
Cs = As′ ( fs′− 0.85 fc′ )
The line of action of Cc is at a /2, while that of Cs is at d ′ from the extreme fiber in com-
pression; both distances are known or easily determined.
Nominal flexural strength is then calculated with respect to the centroid of the tension
steel (Fig. 3.10.1)
 a
M n = M nc + M ns = Cc  d −  + Cs ( d − d ′ )
 2
It is emphasized that the actual stress in the compression steel is fs′ and that the term
( fs′− 0.85 fc′) is fictitious but mathematically convenient. The components Cc and Cs are
hereinafter computed as shown above [see Fig. 3.10.1(c)]. Note that the use of this ficti-
tious stress is valid as long as a ≥ d ′. When the depth of the stress block a is sufficiently
small that a < d ′ but c ≥ d ′, the compression steel would fall outside the compression stress
block and Cs would equal As′ fs′ . In other cases, however, c < d ′ (i.e., “compression” steel is
in tension). In such situations, the area of steel As′ can be ignored and the section treated as
a singly reinforced section.
74
The stress in the compression steel fs′, and thus the resultant force Cs , will depend on
the depth of the neutral axis, which is not known. If both the tension and compression steel
have the same yield strength, the compression steel will yield if the following inequality is
satisfied
f y ( As − As′ )
≥ γ d′
0.85 fc′ bβ1
where γ is equal to 1.85 and 3.22 for Grade 40 and Grade 60 steel, respectively. Equilibrium
requires
Cc + C s = T
For the case of compression steel that has yielded,
( )
0.85 fc′ ab + As′ f y − 0.85 fc′ = As f y
a=
(
As f y − As′ f y − 0.85 fc′ )
0.85 fc′ b
On the other hand, when the strain in the compression steel is less than the yield strain
(i.e., ε s′ < ε y ),
Cc + C s = T
or
0.85 fc′ bβ1c + As′ [ fs′− 0.85 fc′] = As f y
where
fs′ = Es ε s′ < f y
and
 c − d′
ε s′ = 0.003 
 c 
Replacing ε s′ and fs′, and rearranging terms leads to the following quadratic equation for
the neutral axis depth, c,
0.85 fc′ bβ1c 2 +  As′ ( 0.003Es − 0.85 fc′ ) − As f y  c − 0.003Es As′d ′ = 0 (3.10.1)
Once c is known, the following procedure can be used to calculate the nominal flexural
strength.
a = β1c
 c − d′
fs′ = Es ( 0.003)  c 
Cs = As′ ( fs′− 0.85 fc′) for a ≥ d ′
or Cs = As′ fs′ for a < d ′ and c ≥ d ′
 a
M n = Cc  d −  + C s ( d − d ′ )
 2
75
EXAMPLE 3.10.1
Determine the nominal moment strength Mn of the rectangular section shown in

Fig. 3.10.2, given fc′ = 5000 psi, fy = 60,000 psi, b = 14 in., d = 26 in., d′ = 3 in.,
As′ = 2–#8 and As = 8–#10 bars. Use 1-in. clear spacing between layers.
14” 3”
εcu = 0.003 0.85f’c = 4.25 ksi
Cs = 88k
a = 8.77”
c = 10.96”
A’s = 2 – #8 Cc = 522k
εs‘ = 0.00218
26”
A’s = 8 – #10 εs2 =

0.0038
1”
T = 610k
εs1 = 0.0044
Figure 3.10.2 Strain and stress condition, and internal force resultants for section of Example 3.10.1.
SOLUTION
(a) Determine the neutral axis distance c at nominal strength. From Fig. 3.10.2, use the
requirement of equilibrium (i.e., C = T). Compression steel will yield if
f y ( As − As′ )
≥ γ d ′
0.85 fc′ bβ1
where γ is equal to 3.22 for Grade 60 steel. Thus
60 (10.16 − 1.58)
= 10.82 ≥ γ d ′ = 3.22 (3) = 9.66
0.85 ( 5)(14 )( 0.80 )
Therefore, compression steel yields.
T = f y As = 60(10.16) = 610 kips
Cc = 0.85 fc′ ba = 0.85(5)(14)a = 59.5a

Cs = ( fs′− 0.85 fc′ ) As′ = (60 − 4.25)1.58 = 88 kips
C=T
T − Cs 610 − 88
a= = = 8.77 in.
59.5 59.5
a 8.77
c= = = 10.96 in.
β1 0.80
(Continued)
76
(b) Determine whether the amount of tension steel complies with the minimum net ten-
sion steel strain per ACI-9.3.3.1. From the strain diagram, the strain in the bottom
layer of reinforcement is
 1 1.27 
 26 + + − 10.96

2 2
ε s1 = 0.003 = 0.0044 > 0.004 OK
10.96
Similarly, the strain in the second layer of reinforcement is
 1 1.27 
 26 − − − 10.96

2 2
ε s2 = 0.003 = 0.0038 > ε y OK
10.96
Thus, both layers of steel have yielded.
(c) Compute Mn
M n = Cc ( d − a / 2 ) + C s ( d − d ′ )
1 1
= 522[26 − 8.77 / 2] + 88(26 − 3) = 940 + 169 = 1109 ft-kips
12 12
Noted that while ε t = 0.0044 > 0.004 , the section is in the transition region. In this case,
it is advisable to increase the area of compression steel so that the section is tension
controlled (i.e., ε t ≥ 0.005).
EXAMPLE 3.10.2
Repeat the solution for Example 3.10.1, except that As = 4–#11 bars instead of As = 8–#10
bars. Also, determine the percentage increase in the strength (and the net tensile strain)
of the beam over the same beam without compression steel.
SOLUTION
(a) Check if the compression steel yields at nominal strength.
Compression steel will yield if
f y ( As − As′ )
≥ γ d ′
0.85 fc′ bβ1
where γ is equal to 3.22 for Grade 60 steel. Thus
60 (6.24 − 1.58)
= 5.87 < γ d ′ = 3.22 (3) = 9.66
0.85 ( 5)(14 )( 0.80 )
Compression steel does not yield.

(Continued)
7
14” 3” εcu = 0.003 0.85f’c = 4.25 ksi

Cs = 66.9k
c = 6.46” a = 5.17”
A’s = 2 – #8 Cc = 308k
εcu
εs’ = (c – 3)
c
26”
As = 4 – #11
T = 374k
εs
Figure 3.10.3 Strain and stress condition, and force resultants for section of Example 3.10.2.
(b) Determine actual c location. Because compression steel does not yield, solve
Eq. (3.10.1).
4.25 (14 ) 0.80c 2 + 1.58 (0.003 ( 29, 000 ) − 4.25) − 6.24 (60 ) c
− 0.003 ( 29, 000 )(1.58) 3 = 0

c = 6.46 in.

a = β1c = 0.80 (6.46 ) = 5.17 in.
Then

Cc = 4.25 (14 )( 5.17) = 308 kips
 6.46 − 3.00 
ε s′ =   0.003 = 0.00161 < ε y (as previously confirm
med)
 6.46 
Cs = 29, 000 ( 0.00161) − 4.25 (1.58) = 66.9 kips

 

Cc + Cs = 308 + 66.9 = 375 kips

T = As f y = 6.24 (60 ) = 374 kips (Check)
M n = 308 26 – 0.5 ( 5.17) (26 – 3) 12

1 1
+ 66.9
12
M n = 600 + 128 = 728 ft-kiips
Compute the net tensile strain and check ACI-9.3.3.1,
d −c 26 − 6.46
εt = εs = (0.003) = (0.003) = 0.0091 > 0.004 OK
c 6.46
(Continued)
78
(c) Determine Mn if the compression steel had not existed.

C = 0.85 fc′ba = 4.25 (14 ) a

T = 6.24 (60) = 374 kips
374
a = = 6.29 in.
4.25 (14 )
M n = 374 26 – 0.5 (6.29) = 713 ft-kips

1
12
Compute the net tensile strain and check ACI-9.3.3.1,
6.29
c= = 7.86 .
0.80

26 − 7.86
εt = εs = (0.003) = 0.0069 > 0.004 OK
7.86
Thus, the addition of the compression steel to the singly reinforced beam increased
the nominal flexural strength only 2.1%. The net strain in the tension steel, however,
increased 31% when compression steel with an area about 25% that of the tension steel
was added, which shows the effectiveness of compression steel to increase the ductility
of the beam at nominal strength.
3.11 DESIGN OF BEAMS HAVING BOTH TENSION

AND COMPRESSION REINFORCEMENT
As discussed above, the main reason for using compression reinforcement in a beam, in
addition to reducing long-term deflections caused by creep and shrinkage, is to increase
ductility. For singly reinforced sections in which the amount of tension reinforcement
required is such that the strain in the layer of extreme tension reinforcement is less than
permitted (ε t < 0.004 ) or desired, the use of compression steel may provide an acceptable
solution without the need for increasing section dimensions. It should be noted that the
addition of compression steel will lead to only a slight increase in flexural strength (see
Example 3.10.2 and Section 3.13). However, it allows an increase in area of tension steel
beyond that permitted for a singly reinforced beam, with the associated increase in flexural
strength.
Having decided that compression steel is to be used, be it required for ductility or desir-
able for deflection control, the designer now needs to select the appropriate tension steel
As and compression steel As′. For cases in which the flexural strength of a singly reinforced
section with the maximum permitted amount of steel is less than the required strength,
the minimum area of compression steel can be calculated based on the additional area of
tension steel needed to achieve the required flexural strength. It is thus useful to calculate
the flexural strength of the section as the summation of the strength of a singly reinforced
section with an area of tension reinforcement Asc [see Fig. 3.10.1(d)] and the moment
79
3.11 DESIGN OF BEAMS 79
corresponding to a couple consisting of the compression force Cs and a tension force of

equal magnitude [Fig. 3.10.1(e)] as follows:
M n = M nc + M ns
or
 a  a
M n = Cc  d −  + Cs ( d − d ′ ) = Asc f y  d −  + ( As − Asc ) f y ( d − d ′ )
 2  2
′
The required area of tension steel As and the minimum area of compression steel As,min
needed for a tension-controlled section can be calculated as follows
Cc 0.85 fc′ ab
Asc = =
fy fy
The neutral axis depth c for the tensile strain that defines the limit of tension-controlled
sections (i.e., ε t = 0.005) is (see Fig. 3.6.1)
 0.003 
c= d
 0.008  t
Thus
 0.003 
a = β1  d
 0.008  t
and
 a M
M n = Asc f y  d −  + ( As − Asc ) f y ( d − d ′ ) ≥ u
 2 φ
where φ = 0.90 because the section is designed to be tension controlled.

Solving for As
Mu  a
− Asc f y  d − 
φ  2
As = + Asc
fy (d − d ′ )
The minimum area of compression steel required such that the section is tension controlled
is then
( As − Asc ) fy
As′,min =
fs′
where
fs′ = 0.003
(c − d ′ ) E = 0.003
( a / β1 − d ′ ) E ≤ fy
c
s
a / β1
s
Typically, the minimum amount of compression steel required for the section to be tension
controlled should not be greater than about one quarter of As . This is because larger areas
would indicate that the section dimensions are too small, which would lead to reinforce-
ment congestion, excessive deflections, or both.
80
It should be noted that the required area of tension and compression steel can be calcu-
lated for any target strain in the extreme layer of tension steel εt by following the same pro-
cedure outlined above, the only difference being in the calculated value of the stress block
depth a, which will decrease as the target tensile strain εt is increased (see Example 3.11.2).
The procedure for designing a doubly reinforced section is illustrated in Examples 3.11.1
and 3.11.2.
EXAMPLE 3.11.1
Determine the steel areas As and As′ required to carry a service live load moment of 414
ft-kips and a service dead load moment of 234 ft-kips. It is desired that the section be
tension controlled. Use the ACI code and b = 14 in., d = 26 in., d′ = 3 in., fc′ = 5000 psi,
and f y = 60, 000 psi, as shown in Fig. 3.11.1(a).
εcu = 0.003 4.25 ksi

14” d’ = 3”
a = 7.80”
c = 9.75”
d = 26” C
T
εs = εtc = 0.005
(a) Beam size (b) Singly reinforced section with εs = εtc = 0.005
εcu = 0.003 4.25 ksi εcu = 0.003 4.25 ksi
a = 4.80”
c = 6.00”
a = 7.80”
εs’ εs’ Cc
c = 9.75”
Cs
Cs
Cc
T T
εs = εtc = 0.005 εs = εtc = 0.01
(c) Doubly reinforced section with εs = εtc = 0.005 (d) Doubly reinforced section with εs = εtc = 0.01
Figure 3.11.1 Section, strain and stress condition, and force resultants for Examples 3.11.1 and
3.11.2.
SOLUTION
(a) Determine the required nominal strength, using ACI-5.3 for calculating Mu and
21.2.2 for φ factors.

M u = 1.2 ( 234 ) + 1.6 ( 414 ) = 281 + 662 = 943 ft-kips
Assuming a φ factor of 0.90 (i.e., tension-controlled section),
Mu 943
φ 0.90
(Continued)
81
(b) Determine the nominal flexural strength and maximum reinforcement allowed for a
tension-controlled singly reinforced section. The location of the neutral axis for this
condition may be obtained as in Fig. 3.11.1(b);
0.003 0.003
c = dt = 26 = 9.75in.
0.008 0.008

a = β1C = 0.80 ( 9.75) = 7.80 in.

C = 0.85 fc′ ab = 0.85 ( 5)( 7.80 )(14 ) = 464 kips
Thus, the maximum As in a tension-controlled, singly reinforced section is 464/60 =

7.74 sq in. and the corresponding nominal moment is
 7.80  1
M n = 464  26 −  = 855 ft-kips
 2  12
The required Mn exceeds the maximum strength obtainable without compression steel
for a tension-controlled section. In this case, therefore, compression steel is needed to
increase the amount of tension steel enough to achieve the required strength while main-
taining εt ≥ 0.005.
(c) Determine the minimum compression reinforcement required. Maintain c at εt = 0.005,
which is 9.75 in. [see Fig. 3.11.1(c)].
Let

M nc = 855ft-kips [from part(b)]

M ns = M n − M nc = 1048 − 855 = 193ft-kips
193 (12 )
required Cs = = 101 kips
26 − 3
Will compression steel yield when c = 9.75 in.?
ε s′ = 0.003
(c − d ′ ) = 0.003 (9.75 − 3) = 0.0021 > ε
y
c 9.75
Thus, compression steel will yield.
( )
Cs = As′ f y − 0.85 fc′ = 101 kips
101
required As′ = = 1.81 sq in.
60 − 4.25

T = Cc + Cs = 464 + 101 = 565 kips
565
required As = = 9.42 sq in.
60
This amount of tension steel corresponds exactly to the tension-controlled limit for the
beam with the compression reinforcement area calculated.
In this example the remaining steps of selecting actual bars and making a final check
of strength are not shown.
82
EXAMPLE 3.11.2
Redesign the section of Example 3.11.1 so that strain in the extreme layer of tension
reinforcement ε t is approximately 0.01. From Example 3.11.1, Mu = 943 ft-kips.
SOLUTION
(a) Determine M nc corresponding to locating the neutral axis such that ε t = 0.01. Start
by referring to Fig. 3.11.1(d); the location of the neutral axis at this strain condition
may be obtained as
ε cu 0.003
c = dt = 26 = 6.00 in.
ε cu + 0.01 0.003 + 0.01
Once the neutral axis location is known, the depth of the stress block a, internal force Cc,
and Mnc can be determined as follows

a = β1c = 0.80 (6.00 ) = 4.80 in.

C = 0.85 fc′ ab = 0.85 ( 5)( 4.80 )(14 ) = 286 kips

C = T1 = 286 kips
M nc = 286  26 − 4.80  1 = 562 ft-kips

 2  12
Let Asc be the part of tension steel to match the concrete compressive force; then
Asc = T1 = 286 = 4.76 sq in.

fy 60
(b) Determine steel requirements for both faces of the beam. Because the section
is tension-controlled, φ = 0.90 and the required nominal moment strength M n is
943 / 0.90 = 1048 ft-kips. Thus,

M ns = 1048 − 562 = 486 ft-kips
486 (12 )
Cs = T2 = = 254 kips
26 − 3
ε s′ = 0.003
(c − d ′ ) = 0.003 (6.00 − 3.0) = 0.0015 < ε
y
c 6.00
Compression steel does not yield. If c = 6.00 in. is still to be maintained, then
fs′ = 0.0015(29,000 ksi) = 43.5 ksi
so that the stress used is consistent with the strain on the compression steel.
T2 254
As′ = = = 6.46 sq in.
fs′− 0.85 fc′ 43.5 − 4.25
Ass = As − Asc = T2 = 254 = 4.23 sq in.

fy 60

As = Asc + Ass = 4.76 + 4.23 = 8.99 sq in.
(Continued)
83
Select 4–#10 and 4–#9 in two layers for tension steel (As = 9.08 sq in.), and 4–#11
as compression reinforcement ( As′ = 6.24 sq in.). The 4–#11 is less than the amount
required, but it is the maximum steel that will fit into one layer.
(c) Check the strength of the section. The above solution of As = 8.99 and As′ = 6.46 sq in.
is the correct one for ε t = 0.01. When As′ = 6.24 sq in. is used instead, both c and As
will have to change. A trial-and-error procedure is presented below to determine the
stress fs′ acting in the compression steel, an alternative to that used in Example 3.10.2,
where a quadratic equation was solved for the neutral axis location c.
Since a smaller amount of compression steel is used than that required for ε t = 0.01,
the actual neutral axis will be lower if the section is to carry the same bending moment.
Also, the stress fs′ must exceed 43.5 ksi, as it was calculated for c = 6.00 in. (c for
ε t = 0.01). Estimate the compression steel stress to be 45 ksi.

Cc = 0.85 fc′ ba = 0.85 ( 5)(14 ) a = 59.5a

Cc = As′ ( fs′ − 0.85 fc′ ) = 6.24 ( 45 − 4.25) = 254 kips

T = As f y = 9.08 (60 ) = 545 kips

C c + Cs = T
545 − 254 4.89

a = = 4.89 in.; c = = 6.11 in.
59.5 0.80
6.11 − 3.00
ε s′ = (0.003) = 0.00153
6.11

fs′ = 0.00153(29, 000) = 44.3 ksi
Since this does not agree with the 45 ksi assumed, make a new assumption, say,
fs′ = 44.5 ksi :
revised Cs = 6.24(44.5 – 4.25) = 251 kips
a = 545 − 251 = 4.94 in.

59.5
c = 4.94 = 6.18 in.

0.80
ε s′ = 0.003
(6.18 − 3.0) = 0.00154
6.18
fs′ = 0.00154(29,000) = 44.8 ksi
Repeated trials may be made until computed fs′ agrees as closely as desired with the
assumed value. In this case, assume that the present agreement is close enough.
Cc = 59.5a = 59.5 ( 4.94 ) = 294 kips

Cs = 251 kips
1 1
M n = 294[26 − 0.5(4.94)] + 251(26 − 3) = 576 + 481 = 1057 ft-kips
12 12
φ Mn = 0.90(1057) = 951 ft-kips ≥ Mu = 943 ft-

kips OK
(Continued)
84
It may be observed that when there is a large amount of compression steel that does not
yield, varying the amount of compression steel has essentially the effect of changing
only the proportions in M nc and M ns. The choice of 4–#10 and 4–#9 (tension steel) and
4–#11 (compression steel) is acceptable.
3.12 NONRECTANGULAR SECTIONS
When nonrectangular beams are used, the shape of the compression zone dictates whether
the formulas developed in this chapter are applicable. (T-sections are treated in Chapter 4.).
The concept of using the Whitney rectangular compressive stress distribution, however,
may be used in accordance with ACI-22.2.2.4.1 for any shape of the compression zone
in the cross section. Determination of stress block depth a and neutral axis depth c is per-
formed as for rectangular sections by using equilibrium of forces (i.e., C = T ). Since the
shape of the compression zone is no longer rectangular, the resultant concrete compressive
force, which is needed for calculation of moment strength, will not be located at a/2 from
the extreme compressive fiber.
Since ε t or c /dt is used to determine the appropriate φ factor to use, the design proce-
dures presented earlier are identical for all shapes and all kinds of reinforcements, includ-
ing compression steel.
EXAMPLE 3.12.1
For the trapezoidal beam section shown in Fig. 3.12.1, determine the required tension
reinforcement to carry a factored bending moment Mu of 155 ft-kips. Use fc′ = 4500 psi
and f y = 60, 000 psi.
0.85f’c
8”
εcu = 0.003
c = 5.16”
C = 144 K
a = 4.25”
d = 17.5”
15.31”
20”
4 – #7
εs = εt = 0.0072 T = 144 K
16”
Figure 3.12.1 Section, strain and stress condition, and force resultants for Example 3.12.1.
SOLUTION
(a) Determine the required nominal moment strength assuming the section is tension
controlled.
Mu 155
φ 0.90
(Continued)
85
3 . 1 2 N O N R E C TA N G U L A R S E C T I O N S 85
(b) Determine the area of tension steel. As the compression zone is not rectangular,
it is more convenient to first estimate the area of tension steel based on an esti-
mate of the moment arm. Given that the section is narrowest at the level of the
extreme compressive fiber, a moment arm of 0.80d should be a conservative
assumption.
M n = T (arm ) = As f y ( 0.80 d ) = As (60 )( 0.80 )(17.5) 1

≥ 172 ft-kips
12
172 (12 )
As ≥ = 2.46 sq in.
(60)(0.80)(17.5)
Calculate depth of stress block a using equilibrium of forces. The area on which the
compression stress block acts is taken as the summation of a rectangular area equal to
8a, and the summation of the area of two triangles, equal to a2/5.

T = As f y = 2.46 (60 ) = 148 kips
 a2   a2 
C = T = 148 kips = 0.85 fc′ 8a +  = 0.85 ( 4.5) 8a + 
 5  5
Solving for a gives

a = 4.35 in.
for which the resultant compressive force is located at 2.25 in. from the extreme com-
pressive fiber. The corresponding moment arm is then

arm = d − 2.25 in. = 15.25 in. = 0.87d
As the arm is nearly 10% greater than the originally assumed value of 0.80d, recalculate
the required area of steel.
172 (12 )
As ≥ = 2.26 sq in.
(60)(15.25)
Select 4–#7 bars (As = 2.4 sq in.) and calculate nominal flexural strength. From
Table 3.9.2, a minimum width of 10.9 in. is required to satisfy bar spacing requirements,
and this is less than the section width at the level of the tension reinforcement (15 in.).

T = As f y = 2.4 (60 ) = 144 kips
 a2   a2 
C = T = 144 kips = 0.85 fc′ 8a +  = 0.85 ( 4.5) 8a + 
 5  5
Solving for a gives

a = 4.25 in.
(Continued)
86
For this value of a, the resultant compressive force is located at 2.20 in. from the extreme
compressive fiber, which leads to a moment arm of 15.31 in. Thus
M n = T (arm ) = 144 (15.31)

1
= 184 ft-kips > 172 ft-kips
12
(c) Verify that the section is tension controlled. For fc′ = 4500 psi, β1 = 0.825. Thus
a 4.25
c= = = 5.16 in.
β1 0.825
and
ε t = ε s = 0.003
( d − c ) = 0.003 (17.5 − 5.16) = 0.0072 > 0.005
c 5.16
Section is tension controlled.
(d) Determine minimum area of tension reinforcement. Given that the width of the web
varies and is maximum in the extreme tension fiber, the use of bw = 16 in. is recom-
mended. From Eq. 3.7.10 (ACI-9.6.1.2), the minimum percentage area of tension
reinforcement is
3 fc′ 3 4500
As ,min = bw d = (16)(17.5) = 0.94 sq in.
fy 60, 000
As 3 4500 is greater than 200 psi, As,min = 0.94 sq in. < 2.4 sq in.
(e) Final decision and design sketch. Every design requires a clear decision as its con-
clusion, along with an appropriate design sketch.
Use 4–#7 bars with d = 17.5 in. as shown in Fig. 3.12.1.
3.13 EFFECT OF A S , As′ , b, d, fc′ AND f y ON

FLEXURAL BEHAVIOR
The flexural behavior of reinforced concrete sections, and the effect of variables such as
percentage of tension and compression steel, concrete compressive strength, and yield
strength of reinforcement can be better understood through the use of a moment-curvature
analysis. In such analysis, a measure of section deformation is selected, typically strain in
the extreme compressive fiber, εc, and the corresponding moment and curvature (ϕ = εc /c)
are calculated. This process is repeated for concrete compressive strains ranging from those
corresponding to elastic behavior up to the concrete crushing strain. In a moment-curvature
analysis, it is assumed that plane sections prior to loading remain plane after loading (i.e.,
linear strain distribution over section depth) and that the steel reinforcement is perfectly
bonded to the concrete.
The relationship between compressive strain and stress in the compression zone of a
beam may be expected to have the same general shape as that obtained from a test cyl-
inder. A parabolic shape for the ascending portion of the stress-strain response is typi-
cally assumed, with a linear post-peak descending branch up to the concrete compressive
strain capacity ε cu. The peak concrete compressive stress is then assumed equal to k3 fc′ (see
Fig. 3.2.2). For beams, however, k3 is typically taken as 1.0.
The stress-strain behavior for the steel reinforcement is assumed to be the same as that
obtained from a standard tensile test. Thus, increases in stress beyond the yield strength due
to strain hardening of the steel are often considered.
87
3 . 1 3 E F F E C T O F F L E X U R A L B E H AV I O R 87
To illustrate the effect of design variables on flexural behavior, Fig. 3.13.1 plots results
from a series of moment-curvature analyses. To facilitate comparison, the moments and
curvatures were normalized by the maximum moment and maximum curvature of the sin-
gly reinforced rectangular section shown in the figure (Control). Also, the stress-strain
response of the steel was assumed to be bilinear, with a constant stress f y beyond the yield
strain ε y (i.e., no strain hardening). Failure was assumed when the strain in the extreme
compressive fiber reached 0.004.
The moment-curvature responses shown in Fig. 3.13.1 are characterized by three main
regions: uncracked, linear-elastic behavior; cracked, elastic behavior; and flexural yielding.
The maximum curvature for all the sections shown in Fig. 3.13.1 was governed in each case
by the assumed compressive strain capacity of concrete of 0.004. The singly reinforced
control section, with a reinforcement ratio of approximately 1%, exhibited an ultimate
curvature approximately 6 times that at first yield of the longitudinal steel. Depending on
the changes made to the section dimensions and reinforcement, changes in strength and/or
curvature capacity occurred with respect to that of the control section.
A 50% increase in beam width led to only a little increase in flexural strength. This
can be explained by noting that while the depth of the compressive zone decreased,
the change in distance between the resultant tensile and compressive forces was small.
Deformation capacity, however, increased because for the same maximum strain in the
extreme compressive fiber, the reduction in neutral axis depth led to larger strains in the
tension steel, and thus to a higher curvature. The same situation occurs with the addition
of compression steel ( As′ = 0.25 As ) or with an increase in concrete compressive strength
[ fc′ = 2( fc′ ) control].

A 25% increase in yield strength of steel led to a proportional increase in flexural
strength. However, a larger neutral axis depth is needed to balance the larger tensile force
in the reinforcement, leading to a decrease in curvature for a given maximum compressive
strain, and thus to a decrease in ductility.
Increasing the section effective depth by 50% also led to a similar increase in flexural
strength. No change in curvature capacity occurred, however, because the depth of the
neutral axis depth remained unchanged since the tensile force remained the same (no
strain hardening considered). The strain in the tension steel, however, increased in com-
parison to the control section, as the distance from the tension steel to the neutral axis
increased.
1.6
d = 1.5dcontrol
14”
1.4
f’c = 2(f’c)control
fy = 1.25(fy)control
1.2
d = 21.5”
1
M/(Mmax)control
b = 1.5bcontrol 4#8
0.8 2.5”
Control As’ = 0.25As
0.6 fc’ = 4000 psi
fy = 60,000 psi
0.4 As = 3.16 in2
ρ = 0.0105
0.2 Mmax = 310 ft-kips

ϕmax = 9.9 × 10–4 rad/in
0
0 0.5 1 1.5 2
ϕ/(ϕmax)control
Figure 3.13.1 Effect of As, As′, b, d, fc′, and fy on moment-curvature response.

8
The responses shown in Fig. 3.13.1 can be summarized as follows:
1. Flexural capacity is rather insensitive to changes in section width, concrete compres-

sive strength, and addition of compression steel.
2. Maximum curvature increases with an increase in beam width, concrete compressive
strength, and the addition of compression steel.
3. An increase in yield strength of steel leads to an increase in strength but a reduction
in curvature capacity.
4. An increase in effective depth leads to an increase in strength but no or little change
in maximum curvature.
SELECTED REFERENCES
3.1. Eivind Hognestad. A Study of Combined Bending and Axial Load in Reinforced Concrete
Members. Bulletin Series No. 399. Engineering Experiment Station, University of Illinois,
November 1951.
3.2. Eivind Hognestad. “Fundamental Concepts in Ultimate Load Design of Reinforced Concrete
Members,” ACI Journal, Proceedings, 48, June 1952, 809–828.
3.3. Eivind Hognestad, N. W. Hanson, and Douglas McHenry. “Concrete Stress Distribution in
Ultimate Strength Design,” ACI Journal, Proceedings, 52, December 1955, 455–479.
3.4. Jack R. Janney, Eivind Hognestad, and Douglas McHenry. “Ultimate Flexural Strength of
Prestressed and Conventionally Reinforced Concrete Beams,” ACI Journal, Proceedings, 52,
February 1956, 601–620.
3.5. Eivind Hognestad. “Confirmation of Inelastic Stress Distribution in Concrete,” Journal of the
Structural Division, ASCE, 83, Paper No. 1189, ST2, March 1957.
3.6. Alan H. Mattock, Ladislav B. Kriz, and Eivind Hognestad. “Rectangular Concrete Stress
Distribution in Ultimate Strength Design,” ACI Journal, Proceedings, 57, February 1961, 875–928.
3.7. Charles S. Whitney. “Design of Reinforced Concrete Members under Flexure or Combined
Flexure and Direct Compression,” ACI Journal, Proceedings, 33, 1937, 483–498.
3.8. Charles S. Whitney. “Plastic Theory of Reinforced Concrete Design,” Transactions ASCE, 107,
1942, 251–326.
3.9. C. S.Whitney and Edward Cohen. “Guide for Ultimate Strength Design of Reinforced Concrete,”
ACI Journal, Proceedings, 53, November 1956, 455–475.
3.10. ACI-ASCE Joint Committee. “Report on ASCE–ACI Joint Committee on Ultimate-Strength
Design,” ASCE, Proceedings, 81, Paper No. 809. October 1955. See also ACI Journal,
Proceedings, 52, January 1956, 505–524.
3.11. Robert E. Abendroth and Charles G. Salmon. “Sensitivity Study of Reinforced Concrete Simple
Beams,” ACI Journal, Proceedings, 83, September–October 1986, 764–771.
PROBLEMS
All problems* (except Problems 3.1 and 3.2), are to be done in accordance with the ACI Code; all loads
given are service loads, unless otherwise indicated. Wherever possible, basic principles are to be used for
solutions, avoiding the direct use of formulas. All design problems require a clear statement of the final
choice at the end of the calculations, along with a design sketch drawn to scale.
Review Problems in Mechanics of Materials

3.1 An elastic homogeneous beam of material of supports. The beam carries a uniformly
capable of carrying both tension and com- distributed load of 2 klf (kips per linear ft)
pression, has the dimensions shown in the (30 kN/m) in addition to a concentrated load
figure for Problem 3.1; it is simply supported of 10 kips (45 kN) located at 5 ft (1.5 m) from
over a span of 20 ft (6 m) center-to-center the right end of the span.
* Most problems may be solved as problems stated in Inch-Pound units, or as problems in SI units using quantities in paren-
theses. The metric conversions are approximate to avoid implying higher precision for the given information in metric units than
that for the Inch-Pound units.
89
PROBLEMS 89
Reinforced Concrete Problems

Case h b1 t1 tw b2 t2
3.3 For the case (or cases) assigned, do the following.
(Dimensions in Inches) (a)
Compute the nominal moment strength
Mn of the cross section shown in the fig-
1 30 12 2 1 12 2 ure for Problem 3.3 using basic statics (i.e.,
2 20 40 3 4 40 3 no formula) with the Whitney rectangular
3 20 40 4 4 20 4 stress block and the internal couple. Metric
4 20 60 4 12 12 0 bar dimensions for cases 6 and 7 are in
5 24 50 4 14 14 0 Table 1.13.2.
6 28 40 6 14 14 0
7 30 30 4 12 12 0
8 30 10 3 2 10 3 Case Beam- Beam Bars fs′ fy Shape
9 20 10 4 20 20 0 Effective Width
Depth b
(Dimensions in millimeters) d
10 750 300 50 25 300 50
11 500 1500 100 300 300 0 (Dimensions in inches) (Stresses, psi)
1 19.5 12 3–#7 3500 60,000 Rectangular

b1 2 19.5 12 3–#10 3500 60,000 Rectangular
3 19.5 10 6–#7 3500 40,000 Rectangular
(Two layers of three)
4 14.7 10 5–#9 3500 40,000 Rectangular
(3–#9 outside layer and
2–#9 inside layer)
t1
5 36.25 18 8–#11 4000 60,000 Rectangular
tw h
(Two layers of four)
(Dimensions in (Stresses, MPa)
Millimeters)
6 495 300 3–#19M 25 400 Rectangular
b2 7 495 300 3–#32M 25 400 Rectangular
t2
(b)
Compute the service moment capacity
Problem 3.1
(M D + M L ) using the factored gravity dead
(a) Compute independently the maximum flex- load plus live load combination U of ACI-
ural tensile and compressive stresses on the 5.3.1 along with the φ factors of ACI-21.2
section shown in the accompanying figure (assume M L = 1.5 M D ).
(use the case assigned by the instructor),
using basic statics involving internal forces
and internal couple, that is, C = T and
M = (C or T) times (moment arm between
points of action of C and T).
(b) Check the correctness of part (a) by com-
puting both extreme fiber stresses using the d
flexure formula, f = Mc/I.
3.2 The cross section of Problem 3.1 (use the case
assigned by instructor) is used for an elastic
homogeneous beam on a simply supported span
of 25 ft (7.5 m); the beam carries a uniformly dis- As
tributed load of 1 kip/ft (15 kN/m) plus a concen-
trated load of 9 kips (40 kN) at 8 ft (2.4 m) from b
the left end of the span. Compute the stresses as
required by parts (a) and (b) of Problem 3.1. Problem 3.3
90
3.4 For the rectangular section shown in the figure 3.6 For the case assigned in Problem 3.5, design
for Problem 3.4, prove by calculating strains the largest practical size beam using the limits
in the tension steel whether there is yield of the of ACI-9.6.1.2. This time, assume that the given
tension steel when the nominal strength M n is dead load does not include the beam weight; and
reached (i.e., when the extreme fiber in com- make the practical design decisions, including
pression reaches a strain of 0.003). Use basic the choice of overall dimensions, selection of
statics with the Whitney rectangular stress the bars, and checking of the strength. The beam
block and the internal couple (i.e., no formu- is to be designed using tension steel only (i.e.,
las). Is the beam section tension controlled? Use no compression steel). Use the U factors of ACI-
fc′ = 4000 psi and fy = 58,000 psi ( fc′ = 30 MPa 5.3.1 along with the φ factors of ACI-21.2.
and f y = 400 MPa). 3.7 For the case assigned in Problem 3.5, design the
smallest practical size for a tension-controlled sec-
tion. All other requirements are as in Problem 3.6.
(350 mm) 3.8 Design a rectangular cross section having tension
steel only of a size that would obviate the need to
14”
check deflections. The beam is simply supported

(4 – #32M) and is to carry given uniformly distributed dead
4 – #10
and live loads, in addition to the weight of the
beam. Make all the practical decisions for size of
14”
(350 mm)
cross section, select the reinforcement, and check
the strength using statics and basic principles. Do
Problem 3.4 the case assigned by the instructor.
3.5 For a rectangular section, determine the required

fc′ fy wD wL Span Approx
effective size (width b rounded to whole inches Case (psi) (psi) (kips/ft) (kips/ft) (ft) d/b
and effective depth d) and theoretical steel area
As to carry uniformly distributed dead and live 1 3500 40,000 1.0 2.0 30 1.5–2.0
loads on a simply supported span as given. Do 2 3000 60,000 0.8 1.8 30 1.5–2.0
the case assigned by the instructor. Do not make 3 3500 60,000 1.3 2.5 30 1.5–2.0
the practical decisions of choosing actual over- 4 4000 60,000 1.2 2.5 30 1.5–2.0
all beam dimensions and selecting bars. Assume 5 4000 60,000 0.8 1.5 30 1.5–2.0
the dead load includes the beam weight. Use 6 4000 60,000 1.2 2.2 32 1.5–2.0
basic principles (i.e., no formulas) to obtain 7 4000 60,000 1.6 2.5 28 1.5–2.0
minimum (ACI- 9.6.1.2) and maximum (ACI- 8 4000 60,000 1.8 2.7 26 1.5–2.0
9.3.3.1) reinforcement ratio limits, as well as the 9 3000 60,000 0.8 2.8 24 1.5–2.0
maximum reinforcement ratio corresponding to
a tension-controlled section (verify the latter by
comparing with Table 3.6.1). Derive the formula 3.9 For the case in Problem 3.8 assigned by the
for Rn in terms of ρ and then use it to get the instructor, design the smallest practical size for
limiting values. a tension- controlled rectangular cross section
(a) Obtain the largest effective size permitted having tension steel only. The beam is simply
by ACI-9.6.1.2. supported and is to carry given uniformly dis-
(b) Obtain the smallest effective size permitted tributed dead and live loads, in addition to the
by ACI-9.3.3.1. weight of the beam. Make all the practical deci-
sions for size of cross section, select the rein-
forcement, and use statics and basic principles
fc′ fy wD wL Span Approx to check the strength.
Case (psi) (psi) (kips/ft) (kips/ft) (ft) d/b 3.10 Design a rectangular beam to carry a live load of
1.5 kips/ft and a dead load of 1.5 kips/ft, in addi-
1 3000 60,000 1.3 1.0 24 1.75 tion to the beam weight, for a simple span of 32
2 4000 60,000 1.3 1.0 24 2.0 ft. Use a steel ratio ρ corresponding to a net ten-
3 4000 60,000 1.3 1.0 24 1.25 sile strain ε t of approximately 0.0075, and also
4 5000 60,000 1.0 1.6 24 1.5 satisfy ACI-Table 9.3.1.1. Use fc′ = 4000 psi and
5 5000 60,000 0.8 1.2 18 1.5 f y = 60, 000 psi. (Live load = 22 kN/m; dead
6 3000 40,000 0.8 1.4 18 2.0 load = 22 kN/m; span = 9.8 m; fc′ = 30 MPa;
f y = 400 MPa.)
91
PROBLEMS 91
3.11 Select an economical reinforcement for a beam 3.14 For the double overhanging cantilever beam
12 in. wide by 22 in. deep overall to carry dead shown, design the smallest practical rectangu-
load and live load moments of 15 and 60 ft-kips, lar cross section (without compression steel)
respectively. Use fc′ = 5000 psi, f y = 60, 000 psi. within the ρ limits of ACI-9.3.3.1. The same
(Width = 300 mm; depth = 560 mm; dead overall size is to be used for the entire beam
load moment = 20 kN·m; live load moment = length (including cantilevers). The given dead
80 kN·m; fc′ = 35 MPa; f y = 400 MPa.) load does not include the beam weight. Select
3.12 Select economical reinforcement for a beam reinforcement for both positive moment and
20 in. wide by 40 in. overall depth to carry a negative moment regions. Do the case assigned
live load moment of 500 ft-kips and a dead load by the instructor. Assume there is no deflec-
moment (including beam weight) of 300 ft-kips. tion limit, and refer to the note at the end of
Use a singly reinforced section. Though a check Problem 3.13.
of deflection cannot be made with the given
information, would you expect such a check to
show that deflection is excessive? Explain your
answer. Use fc′ = 4000 psi and f y = 60, 000 psi. Main span
(Beam size = 500 mm × 1000 mm; live load
moment = 700 kN·m; dead load moment = 400
Cantilever spans
kN·m; fc′ = 30 MPa; f y = 400 MPa.)
3.13 Assuming no deflection limitation, and without Problem 3.14
using compression steel, select a rectangular sec-
tion having the smallest practical size permitted
by ACI-9.3.3.1; use the same size for the entire
length (main span plus cantilever) of beam, as Main Cantilever
shown in the figure for Problem 3.13. The loads fc′ fy wD wL Span Span
given are in addition to the beam weight. Select Case (psi) (psi) (kips/ft) (kips/ft) (ft) (ft)
reinforcement for both the positive and negative
1 3500 60,000 1.0 1.5 30 10
moment regions. Use the case assigned by the
2 3500 60,000 1.0 1.8 28 10
instructor. For SI, use the reinforcing bars from
3 3500 60,000 1.3 2.5 28 9
Table 1.13.2. (Note: Live load is always to be
4 4000 60,000 1.2 2.5 26 9
applied in the locations and distributions that
5 4000 60,000 0.8 1.5 32 8
will cause the most severe effects; spans may
6 4000 60,000 1.2 2.2 32 12
be fully loaded, partially loaded, or unloaded, as
7 4000 60,000 1.6 2.5 24 8
necessary to obtain maximum effects.)
8 4000 60,000 1.8 2.7 24 6
9 3000 60,000 0.8 2.8 24 8
Main span Cantilever

span
3.15 For the loading and beam span conditions of

Problem 3.13
Problem 3.14, design the beam as part of a floor
system supporting nonstructural elements not
Main Cantilever
likely to be damaged by large deflections. (Hint:
fc′ fy wD wL Span Span
Refer to ACI-Table 9.3.1.1 and Problem 3.8.)
Case (psi) (psi) (kips/ft) (kips/ft) (ft) (ft)
Do the case assigned by the instructor.
1 3500 40,000 1.0 1.5 20 8 3.16 Use principles of statics with internal couple
2 3500 40,000 1.25 2.0 24 8 to compute the nominal strength M n for the
3 3500 60,000 1.0 1.5 16 10 assigned case of a beam having compress-
4 4000 60,000 1.0 2.0 18 8 ion steel. As a normal part of the procedure,
5 4000 60,000 0.8 1.5 20 8 verify by means of basic principles whether
6 4000 60,000 1.2 2.2 20 8 compression steel has yielded when nomi-
7 4000 60,000 0.8 1.5 22 10 nal strength is reached; if it does not yield,
8 4000 60,000 1.2 2.2 22 10 use a compression steel stress proportional
9 4000 60,000 1.2 2.5 24 10 to the strain in the compression steel. Verify
that tension steel does not exceed the maxi-
(MPa) (MPa) (kN/m) (kN/m) (m) (m)
mum permitted by ACI- 9.3.3.1. Determine
10 25 300 15 20 6 2.5
the appropriate φ factor to use according to
11 25 300 17 30 7.3 2.4
ACI-21.2.
92
b
Compression
fc′ fy b d d′ As As′ face d’
Case (psi) (psi) (in.) (in.) (in.) Bars Bars
1 4000 40,000 18 36.1 2.5 10–#11 6–#7 A’s

2 3000 60,000 16 27.3 3.0 4–#11 3–#11
3 4000 60,000 16 27.3 3.0 4–#11 3–#11
4 5000 60,000 16 27.3 3.0 4–#11 3–#11
5 3500 60,000 12 21.5 2.5 3–#10 2–#10 d
6 3500 40,000 12 20.5 2.44 3–#10, 3–#9 3–#9
7 3500 60,000 12 17.6 2.38 3–#9 3–#8
8 3500 60,000 14 19.5 2.5 2–#11, 2–#9 2–#10
9 4000 60,000 20 28.4 2.38 6–#10, 2–#9 2–#8
As
(MPa) (MPa) (mm) (mm) (mm) Bars Bars
10 30 300 460 917 64 10-36Ma 5-25M
11 30 400 400 693 76 4-36M 3-35M Tension
face
Use reinforcing bars from Table 1.13.2.
a
Problem 3.16
3.17 A rectangular section with b = 14 in . and effec- Compression face

tive depth d = 21.5 in. has 4–#10 as tension
reinforcement and 2–#10 as compression rein-
forcement centered 2.5 in. from the face of the 3 – #7
#3
beam. Determine the strength M n for this sec-
tion. How much can the strength be increased
22”
by adding tension steel only when the case is
governed by ACI-9.3.3.1? At the point where
compression steel is needed to further increase
1 12 ” clear
strength by adding tension steel, what ratio cover all As = ?
of As′ to As would require to be added? Use sides
fc′= 5000 psi and f y = 60, 000 psi (b = 350 mm;
d = 546 mm; tension steel, 4–#32M; compress- 10”
ion steel, 2–#32M centered 63.5 mm from face;
fc′ = 35 MPa; f y = 400 MPa). Problem 3.20
3.18 For the cross section obtained for the assigned
data case of Problem 3.13, redesign as a beam
having compression steel such that the net rein- NONRECTANGULAR SECTION PROBLEMS
forcement ratio (ρ – ρ′) is about that correspond- 3.21–3.22 For the beam cross sections shown in the
ing to a net tensile strain ε t = 0.0075 for a singly figures for Problems 3.21 and 3.22, assuming
reinforced beam. fc′ = 3000 psi, f y = 60, 000 psi, and the clear
3.19 For the conditions of Problem 3.14, Case 1, use cover from the bottom (tension) face to the
a rectangular cross section 14 × 22 in. over- bars = 2 in., perform the following:
all, and control creep and shrinkage deflection (a) Compare the given tension reinforcement
by designing compression steel such that As′ is with the maximum permitted by ACI-9.3.3.1.
about 0.5As. (b) Using basic principles with the Whitney
3.20 For the beam shown in the figure for Problem rectangular stress distribution, compute the
3.20, it is desired to utilize 3–#7 bars as com- nominal strength M n for the cross section.
pression reinforcement. The factored moment (c) Neglecting the reinforcement and assuming
Mu to be carried is 210 ft-kips; fc′ = 4500 psi, that the concrete is a homogeneous elastic
and f y = 60, 000 psi. Determine the tension material, compute the cracking moment M cr
steel required, select the bars, and check the sec- when the extreme fiber in tension reaches
tion strength. the modulus of rupture value given by ACI
(i.e., 7.5 fc′ for normal-weight concrete).
93
PROBLEMS 93
(d) Using the result in part (c), determine the (e)

Assuming that compression steel of an
reinforcement ratio As ,min / bw d that makes amount equal to the given tension steel were
φ M n for a reinforced concrete beam (ignore to be used and located with 2-in. clear cover
the given bars for this part) equal to M cr . to the top (compression) face of the beam,
Compare with ACI-9.6.1.2. determine the maximum amount of tension
steel that ACI-9.3.3.1 would permit.
4” 6” 4” 8”
8”
8”
20”
5 – #6
24”
16”
3 – #8
Problem 3.22
14”
Problem 3.21
CHAPTER 4
T-S ECTIONS IN BENDING
4.1 GENERAL
Chapter 3 covered the analysis and design of isolated rectangular beams and briefly dis-
cussed nonrectangular sections (Section 3.12). In most building frame structures, however,
beams are built monolithically with the slab, leading to a T-shaped beam section; hence the
name T-beams. Beams with a T shape may also be individually built as such, as is com-
monly done in precast construction, where the flange may later serve as support for a slab
topping or floor finishing.
For the purpose of analysis and design, each beam cast monolithically with the slab is
assumed to include a portion of the slab projecting as a flange from each side of the stem.
For negative bending moment the flange is on the tension side of the neutral axis, thus the
T-section is in effect a rectangular section with a compression area of width equal to the
width of the stem or web. For positive bending moment the flange does provide considerably
more compression area than in the negative bending moment portion of the span. Analysis
and design of T-sections subjected to bending are discussed in sections 4.2 through 4.5.
City Hall, Boston. (Photo by C. G. Salmon).

95
4.3 EFFECTIVE FLANGE WIDTH 95
b b
N.A. N.A.
d d
N.A.
bw
(a) (b)
Figure 4.2.1 Two equivalent sections in bending. Figure 4.2.2 T-section under positive bending.
4.2 COMPARISON OF RECTANGULAR AND T-S ECTIONS

A comparison of the two sections shown in Fig. 4.2.1 indicates that the flexural strength
of a rectangular section is the same as that of a T-section as long as both sections have
compression zones of equal width and depth, and the same steel area at the same effective
depth. Thus, as far as bending is concerned, any T-section with a rectangular compression
area, such as shown in Fig. 4.2.1(b), may be regarded as a rectangular section regardless of
the shape of the tension zone. When the compression zone of a T-section subjected to pos-
itive bending extends below the flange, as shown in Fig. 4.2.2, computation of its flexural
strength requires different treatment than for a rectangular section; this is because the shape
of the compression zone will no longer be rectangular.
4.3 EFFECTIVE FLANGE WIDTH
Very wide beams do not conform in behavior to the assumption of the elementary theory
of bending. In ordinary theory, bending stresses are assumed not to vary across the beam
width. Simple theory, therefore, would dictate a constant stress at, say, the extreme fiber
over the entire flange width of a T-section, no matter how great the overhang from the stem.
In reality, the stress, based on the plate theory, decreases the more distant a point is from the
stem of the beam. Thus, for a flange of infinite width, the compressive stress in the flange
varies as shown in Fig. 4.3.1.
Theoretical investigations for an infinitely long continuous beam on equidistant sup-
ports, with an infinitely large flange width and a thickness that is small in comparison to the
beam depth, have led to the determination of an effective flange width bE over which the
compressive stress may be considered constant. The total compressive force carried by
the equivalent system is the same as that carried by the actual system. For such assump-
tions, the equivalent width of overhang λ depends only on the type of loading and the
span length of the beam. The theory, first developed by T. von Kármán, along with certain
results, is summarized in Timoshenko and Goodier [4.1] and Girkmann [4.2].
Whereas the aforementioned theory gives the equivalent width bE for an infinite flange
width as a function of the span length L, in practical situations other variables are impor-
tant as well. These variables are the spacing of the beams, the width of the stem or web of
the beam bw, and the relative thickness of the slab with respect to the total beam depth t/h.
To illustrate the effect of loading, several cases (from Girkmann [4.2]) showing the varia-
tion of effective flange projection for a flange of zero thickness between beams (t/h = 0) are
given in Fig. 4.3.2.
In the case of a floor slab built monolithically with floor beams, there is transverse bend-
ing in the slab between beams, which also tends to reduce the effectiveness of the slab in
carrying compression at points remote from the beam web. Thus, there is a valid reason for
using a conservatively low effective flange width. The effective section of a T-beam in a
floor system is shown shaded in Fig. 4.3.3.
96
96 C hapter 4 T- S ections in B ending
λ = equivalent width Actual extreme fiber

for uniform stress compressive stress fc
and same for infinitely wide flange
compressive force bE λ
as actual stress
distribution
bW
Figure 4.3.1 Actual and equivalent stress distribution over flange width.
x w sin πx/L
w
L L
0.191L
bw +0.363L
bw
bw
bw +0.194L
0.191L
w P
L/2 L
L
Max moment
3L/8
bw +0.764L
bw +0.550L
bw bw
bw +0.152L
bw +0.36L bw +0.196L
Figure 4.3.2 Equivalent flange widths for beams with infinitely wide flanges and a rib cross-
sectional area of 0.1 tL. (Adapted from Girkmann [4.2].)
The effective width bE of the flange is an essential factor in the neutral axis location.
Available information indicates that as the strain εc at the extreme compressive fiber
increases toward its ultimate value εcu, the width of the effective compression flange
increases [4.4]. Therefore, it is safe to use the smaller effective widths applicable under
service load conditions based on linear elastic behavior.
97
4 . 4 N O M I N A L M O M E N T S T R E N G T H M n O F T-S E C T I O N S 97
bE
t
h
2bo
bw
Figure 4.3.3 Effective section of a T-beam with the slab in compression in a beam-slab floor system.
The ACI Code (ACI-6.3.2.1) prescribes a limit on the effective flange width bE of inter
ior T-sections to the smallest of the following:
1. bE = bw + Ln /4 (4.3.1a)

2. bE = bw + 16t (4.3.1b)
3. bE = center-to-center spacing of beams (4.3.1c)
where Ln is the clear span length of the beam, and bw and t are as shown in Fig. 4.3.3.
For exterior T-sections (slab on only one side of web), ACI-6.3.2.1 prescribes for effec-
tive width bE the smallest of the following:
1. bE = bw + Ln /12 (4.3.2a)

2. bE = bw + 6t (4.3.2b)
3. bE = bw + 1 / 2 ( clear distance to next beam ) (4.3.2c)
For isolated T-sections, ACI-6.3.2.2 requires
bE ≤ 4bw (4.3.3a)
and
1
t≥ bw (4.3.3b)
2
Since the true effective width depends greatly on the type of loading and on the relation-
ships (Fig. 4.3.3) t/h, L /bw, and L /b0, the ACI criteria are very much simplified. They seem
properly adequate for certain loading cases and are unduly conservative for others. Detailed
procedures for determining the values of effective flange width are presented in [4.3, 4.4].
4.4 NOMINAL MOMENT STRENGTH M n OF T-S ECTIONS

In computing the nominal moment strength Mn of a T-section under positive bending, the
neutral axis location determines whether the compression zone is T-shaped or rectangular.
Since the option of using Whitney rectangular stress distribution (Section 3.3) is also avail-
able for nonrectangular sections, use of the rectangular stress distribution would mean that
as long as its depth a does not exceed the flange thickness t, the moment strength Mn would
be the same as for a rectangular section having b = bE. Strictly speaking, rectangular sec-
tion behavior should be based on the depth of the neutral axis c and not of Whitney’s stress
block. However, the use of a = t as the boundary between rectangular section and T-section
behavior leads to adequate results in terms of flexural strength.
Two cases are possible for computation of the positive nominal moment strength Mn of a
T-section: (1) the depth a of the rectangular stress distribution is equal to or less than t, and
(2) the depth a is greater than t. As for most cases a ≤ t, it is convenient to first calculate the
depth of the stress block assuming that Case 1 governs.
98
Case 1: Rectangular section behavior, a ≤ t [Fig. 4.4.1(a)]

For this case, a is calculated taking b = bE . Thus, the analysis is exactly as presented for
rectangular sections in Chapter 3, using bE for b. The following discussion is limited to
singly reinforced T-sections. In general, the effect of compression steel reinforcement, if
provided, can be neglected because it will have little effect on moment strength. Further,
εt >> 0.005 in typical singly reinforced T-sections with rectangular section behavior
(φ = 0.90), and thus there is no need to account for the beneficial effect of compression
steel on section ductility in determining the strength reduction factor φ. If compression
steel must be included in the calculation of moment capacity, it can be done follow-
ing the same procedure outlined in Section 3.10. The effect of compression reinforce-
ment on the flexural strength and normal strain distribution in T-sections is illustrated in
Example 4.4.4.
The depth of the stress block a is then calculated accordingly and compared with the
flange thickness t as follows:
As f y
a= ≤t (4.4.1a)
0.85 fc′ bE
obtained from equating C to T, where
C = 0.85 fc′ bE a (4.4.1b)
T = As f y (4.4.1c)
bE
εc = 0.003 c = a/β1 0.85fc’
C
t
arm = (d – 21 a )
N.A.
a≤t
Moment
As
εs ≥ εy T = Asfy
bw
(a) a ≤ t
bE
εc = 0.003
1 1
2
A2 2
A2 t C2
A1 C1
c
arm = (d – 21 t)
arm < (d – 21 t)
Moment
d
Moment
a>t
N.A.
As
εs ≥ εy T = C1 + C2 = Asfy
bw
(b) a > t
Figure 4.4.1 Strain and stress conditions, and force resultants of T-sections in bending: a) a ≤ t
and b) a > t.
9
Case 2: a > t [Fig. 4.4.1(b)]

For this case, the area on which the uniform stress 0.85 fc′ acts is T-shaped. Except for cases
in which compression steel is needed for ductility so that the section is tension controlled
(not common), its presence can be neglected in the calculation of moment strength as in
Case 1. The discussion that follows deals with singly reinforced T-sections. Calculation of
nominal moment strength in T-sections with compression reinforcement and a > t is illus-
trated in Example 4.4.5.
It is desirable to separate the total compressive force into forces C1 and C2, with C1
resulting from the stress on area A1 and C2 on area A2. The moment arm for C2 is equal to
d – t/2, but that of C1 is less than d – t/2. Thus
 a  t
M n = C1  d −  + C2  d −  (4.4.2a)
 2  2
in which
C1 = 0.85 fc′ bw a (4.4.2b)
C2 = 0.85 fc′ (bE − bw )t (4.4.2c)
and since
C1 + C2 = As f y = T
then
T − C2
a= (4.4.2d)
0.85 fc′ bw
The tensile force T and the tension steel area As may also be separated into T1 and T2 and
As1 and As2, respectively.
EXAMPLE 4.4.1
Determine the nominal moment strength Mn within the span of a floor beam (Fig. 4.4.2)
whose projection below a 4 1 2 -in. slab is 13 × 24 in. (effective depth is 25 in. for two
layers of steel). Tension reinforcement is 8–#8 bars. The span length of the beam, meas-
ured to the center of the supporting columns, is 24 ft, and the beams are centered 13 ft
apart. Supporting columns have a square cross section with 28-in. side dimension. Use
fc′ = 4500 psi and fy = 60,000 psi.
4 21 ”
1 a
Effective width bE = 78” 0.85fc’ 2
c C
a
25”
24”
8 – #8 As = 6.32 sq in.
T
13”
Figure 4.4.2 T-section for Example 4.4.1.

(Continued)
10
SOLUTION
(a) Determine the effective flange width. Following ACI-6.3.2.1, the effective flange
width bE from Eqs. (4.3.1) is the smallest of 13 + ( 24 )(12 ) − 28 / 4 = 78 in.,
13 + 16 ( 4.5) = 85 in ., or (13)(12 ) = 156 in. Thus bE = 78 in.
(b) Assume rectangular section behavior (a ≤ t) and yielding of the reinforcement.
As f y (6.32)(60)
a= = = 1.27 in. ≤ t = 4.5 in.
0.85 fc′ bE (0.85)(4.5)(78)
Assumption of rectangular section behavior is correct. For such a small value of a,
it is clear that εs > εy.
(c) Treat as a rectangular section with a = 1.27 in.
a
moment arm = d − = 25 − 0.64 = 24.36 in.
2
1
M n = (6.32)(60)(24.36) = 770 ft-kips
12
EXAMPLE 4.4.2
Determine the nominal moment strength Mn of the isolated T-section shown in Fig. 4.4.3
when As = 12.48 sq in. (8–#11). Use fc′ = 4000 psi and f y = 60, 000 psi.
1 a
30” εc = 0.003 0.85fc’ 3 21 ” 2
1 1
A2 A2 7” C2
2 A1 2
c a
C1
N.A.
36”
As
T
εs ≥ εy
14”
Figure 4.4.3 T-section for Examples 4.4.2 and 4.5.1.
SOLUTION
(a) Check requirements for isolated T-sections. According to ACI-6.3.2.2, bE cannot
exceed 4bw = 56 in., and t must be at least bw / 2 = 7 in. Thus, flange thickness is sat-
isfactory and bE = 30 in. is less than 4bw.
(b) Calculate stress block depth a assuming rectangular section behavior.
As f y (12.48)(60)
a= = = 7.34 in. > t = 7 in.
0. 85 f c′ E
b ( 0 .85)(4)(30)
Assumption of rectangular section behavior is not correct. Section behaves like a
T-section.
(c) Treat T-section by the two-couple method (Fig. 4.4.3).
T = As f y = 12.48(60) = 749 kips
C = C1 + C2 = 0.85 fc′ A1 + 0.85 fc′ A2
(Continued)
10
749 = 3.4(14a ) + 3.4(16)(7)

a = 7.73 in.
C1 = 3.4(14)(7.73) = 368 kips
C2 = 3.4(16)(7.0) = 381 kips
1 1
M n = C1 [36 − 0.5(7.73)] + C2 (36 − 3.5)
12 12
M n = 985 + 1032 = 2017 ft-kips
EXAMPLE 4.4.3
For the T-section shown in Fig. 4.4.4, determine the design strength φMn in accordance
with the ACI Code. The tension steel is 9–#11 (As = 14.04 sq in.) with three bars in each
of three layers. Assume #4 stirrups are used, and there is one inch between layers. Use
fc′ = 4000 psi and fy = 60,000 psi.
30”
c = 11.40” 0.003
7”
d dt
40”
9 – #11
εt = 0.0068
14”
Figure 4.4.4 T-section and strain distribution for Example 4.4.3.
SOLUTION
(a) Determine the effective depth d and calculate the stress block depth a, assuming
rectangular section behavior.
For a clear cover of 1.5 in. and a #4 stirrup,
d = 40 – 1.5 ( i.e., cover ) – 0.5 ( i.e., stirrup )
– 1.5 (1.41) ( i.e., 1.5 bar diameters) – 1 ( i.e., between layers) = 34.9 in.
The depth of the compressive stress block is

As f y (14.04)(60)
a= = = 8.26 in. > t = 7 in.
0.85 fc′ bE (0.85)(4)(30)
Thus, the compression zone is T-shaped and the neutral axis is in the web. The internal
forces are
T = As f y = 14.04(60) = 842 kips
C1 = 0.85 fc′ bw a = 0.85(4)(14)a = 47.6 a

C2 = 0.85 fc′ (bE − bw)t = 0.85(4)(16)7 = 381 kips
C = C1 + C2 = T
(Continued)
102
842 − 381
a= = 9.69 in. > t
47.6
a
c= = 11.40 in.
β1
(b) Compute the nominal moment strength Mn. Using the two-couple method,
 a  t
M n = C1  d −  + C2  d − 
 2  2
 9.69  1 1
M n = 47.6(9.69)  34.9 −  12 + 381(34.9 − 3.5) 12
 2
M n = 1155 + 997 = 2152 ft-kips
(c) Compute the net tensile strain εt at the extreme tension steel. The distance dt to the
extreme tension steel is
dt = d + 1.41 / 2 + 1 + 1.41 / 2 = 34.9 + 2.41 = 37.3 in.
dt − c 37.3 − 11.40
ε t = 0.003 = 0.003 = 0.0068
c 11.40
Because εt exceeds the limit 0.005 for tension-controlled sections (see Fig. 3.6.2),
the appropriate φ factor is 0.90 (ACI-21.2.2). Thus,
φ M n = 0.90 ( 2152 ) = 1937 ft-kips
EXAMPLE 4.4.4
Determine the design moment strength φMn in accordance with the ACI Code of the
same section used in Example 4.4.3 but with 5–#8 bars ( fy = 60 ksi) as compression
reinforcement (Figure 4.4.5).
30” 0.003
d’ = 2.5”
c = 7.29”
7”
5 – #8
d dt
40”
9 – #11
εt = 0.0123
14”
SOLUTION
(a) Determine depth of stress block and internal forces. Assuming the compressive steel
yields and rectangular section behavior at nominal strength condition,
As f y − As′ ( f y − 0.85 fc′ ) (14.04)(60) − 3.95[60 − 0.85(4))]
a= = = 6.07 in. < t = 7 in.
0.85 fc′ bE (0.85)(4)(30)
(Continued)
103
Note that the stress in the compression steel has been adjusted to account for the area
of displaced concrete in compression.
Verify that for the calculated stress block depth, the compression steel reinforcement
yields as assumed.
a
c= = 7.14 in.
β1
c − d′ 7.14 − 2.5
ε s′ = 0.003 = 0.003 = 0.00195 < ε y = 0.00207
c 7.14
Compression reinforcement does not yield. Rather than developing a closed- form
solution, it is easier to assume a value for the depth of the stress block and verify that
equilibrium is satisfied. Since the compression steel is near yield, assume only a small
increment in the depth of the stress block a. Thus, taking a = 6.20 in.,
a
c= = 7.29 in.
β1
T = As f y = 14.04(60) = 842 kips
Cc = 0.85 fc′ bE a = 0.85(4)(30)(6.20) = 632 kips
c − d′ 7.29 − 2.5
ε s′ = 0.003 = 0.003 = 0.00197
c 7.29
Cs = As′ (ε s′ Es − 0.85 fc′ ) = 3.95 [(0.00197)29000 − 0.85(4)] = 212 kips
C = Cc + Cs = 844 kips ≈ T
By adding compression reinforcement with an area approximately equal to one-third

that of the tension steel, the section now exhibits a rectangular section behavior (Case 1,
a ≤ t).
(b) Compute the nominal moment strength Mn. Taking moments with respect to the cen-
troid of the tension steel,
 a
M n = Cc  d −  + Cs (d − d ′ )
 2
  6.20  1
M n = 632  34.9 −  + 212(34.9 − 2.5)
  2   12
M n = 1676 + 573 = 2249 ft-kips
This nominal moment strength is only 5% greater than that calculated without com-
pression reinforcement in Example 4.4.3.
(c) Compute the net tensile strain εt at the extreme tension steel.

dt − c 37.3 − 7.29
ε t = 0.003 = 0.003 = 0.0123 > 0.005
c 7.29
Thus, φ = 0.90 and
φ M n = 0.90 (2249) = 2024 ft-kips
Note that the net tensile strain εt nearly doubled compared to that of the same section
without compression steel. Therefore, the addition of compression steel, though having
little effect on flexural strength, did have a significant effect on section ductility.
104
EXAMPLE 4.4.5
Determine the design moment strength φMn in accordance with the ACI Code of the
same section used in Example 4.4.4 but using a flange thickness of 5 in. (Figure 4.4.6).
30” 0.003
d’ = 2.5”
c = 8.58”
5”
5 – #8
d dt
40”
9 – #11
εt = 0.010
14”
SOLUTION
(a) Determine depth of stress block and internal forces. Assuming rectangular section
behavior and yielding of the compression steel, the depth of the compressive stress
block is
As f y − As′ ( f y − 0.85 fc′ ) (14.04)(60) − 3.95[60 − 0.85(4))]

a= = = 6.07 in. > t = 5 in.
0.85 fc′ bE (0.85)(4)(30)
Section falls within Case 2 (i.e., T-section behavior). The internal forces, assuming
yielding of compression steel, are
T = As f y = 14.04(60) = 842 kips
C1 = 0.85 fc′ bw a = 0.85(4)(14)a = 47.6 a
C2 = 0.85 fc′ (bE − bw )t = 0.85(4)(16)5 = 272 kips
Cs = As′ ( f y − 0.85 fc′ ) = 3.95 [60 − 0.85(4)] = 224 kips
C = C1 + C2 + Cs = T
842 − 272 − 224

a= = 7.29 in.
47.6
C1 = 0.85 fc′ bw a = 0.85(4)(14)(7.29) = 347 kips
a
c= = 8.58 in.
β1
Verify that the compression steel reinforcement yields as assumed.
c − d′ 8.58 − 2.5
ε s′ = 0.003 = 0.003 = 0.00212 > ε y = 0.00207
c 8.58
Compression reinforcement yields as assumed.
(Continued)
105
4 . 5 D E S I G N O F T- S E C T I O N S I N B E N D I N G 105
(b) Compute the nominal moment strength Mn. Taking moments with respect to the cen-
troid of the tension steel,
 a  t
M n = C1  d −  + C2  d −  + Cs (d − d ′ )
 2  2
  7.29   5.00  1
M n = 347  34.9 −  + 272  34.9 −  + 224(34.9 − 2.5) 12
  2 2 
M n = 904 + 734 + 605 = 2243 ft-kips
(c) Compute the net tensile strain εt at the extreme tension steel.

dt − c 37.3 − 8.58
ε t = 0.003 = 0.003 = 0.010 > 0.005
c 8.58
Thus, φ = 0.90 and
φ M n = 0.90(2243) = 2019 ft-kips
4.5 DESIGN OF T-S ECTIONS IN BENDING

The design of T-shaped isolated beams involves the dimensions of the flange and web and
the area of tension steel, or a total of five unknowns—one more if compression steel is used.
Thus, there are many possible solutions to the problem. The more common T-sections are
those in the region of positive bending moment in continuous monolithic slab-beam-girder
systems. The size of the available flange is therefore known after the slab thickness has
been selected, and only the web size needs to be designed. The selection of web size for
such beams is treated in Chapter 9, and it is often based on negative moment requirements
at the supports, where the web will be in compression, or on shear strength requirements.
Once the overall dimensions have been established, the design problem is to determine
the amount of positive moment reinforcement. Thus, it is important first to ascertain the
location of the neutral axis associated with the positive moment. This may be done by first
computing the moment strength at which the depth a of the rectangular compressive stress
block is equal to the flange thickness t. If the depth a is less than t, then the compression
zone is rectangular and the beam is designed as a rectangular section. If a is greater than t,
then the compression zone is T-shaped and the two-couple method can be used.
EXAMPLE 4.5.1
Determine the amount of tension steel required in the T-section of Fig. 4.4.3 to carry
a dead load moment of 560 ft-kips and a live load moment of 700 ft-kips, using
fc′ = 4000 psi and fy = 60,000 psi.
SOLUTION
(a) Determine the factored moment Mu.
Mu = 1.2(560) + 1.6(700) = 1792 ft-kips
(Continued)
106
Assuming φ = 0.90,
Mu 1792
φ 0.90
(b) Determine whether the depth a of the rectangular stress distribution will be greater
than t = 7 in. For a = t,
C = 0.85 fc′ bE t = 0.85(4)(30)(7) = 714 kips
 t  7 1
M n = C  d −  = 714  36 −  = 1934 ft-kips
 2   2  12
Since the required Mn exceeds 1934 ft-kips, the actual a must exceed t.
(c) Use the two-couple method (Fig. 4.4.3) to obtain As.

 a  t
M n = 0.85 fc′ A1  d −  + 0.85 fc′ A2  d − 
 2   2
 a
1991(12) = 3.4(14 a )  36 −  + 3.4(16)(7)(36 − 3.5)
 2
a 2 − 72 a = −484
a = 7.50 in. > t
C1 = 0.85 fc′ bw a = 0.85(4)(14)(7.50) = 357 kips
T1 357
As1 = = = 5.95 sq in.
fy 60
C2 = 0.85 fc′ (bE − bw) t = 0.85(4)(16)(7) = 381 kips
T2 381
As 2 = = = 6.35 sq in.
fy 60
As = As1 + As 2 = 5.95 + 6.35 = 12.30 sq in.
Check minimum tension reinforcement requirement. Since in this case 3 fc′ < 200 psi,
the minimum required amount of tension reinforcement is (see Section 3.7)
200 200
As ,min = bw d = (14)(36) = 1.68 sq in. < 12.30 sq in.
fy 60, 000
Try 8–#11 bars in two layers (As = 12.48 sq in.). From Table 3.9.2, the minimum beam
width required for 4–#11 is 13.8 in., which does not exceed the 14 in. available, and is
acceptable. Assuming 1-in. clear cover between layers, the distance dt to the extreme
tension steel is
dt = d + 0.5 + 1.41 / 2 = 36 + 1.21 = 37.2 in.
dt − a / β1 37.2 − 7.50 / 0.85

ε t = 0.003 = 0.003 = 0.0096
a / β1 7.50 / 0.85
Since εt exceeds the limit of 0.005 for tension-controlled sections (see Fig. 3.6.2), then
φ = 0.90 as assumed.
Use 8–#11 bars in two layers.
107
4 . 5 D E S I G N O F T- S E C T I O N S I N B E N D I N G 107
EXAMPLE 4.5.2
Design the steel reinforcement for the section of Fig. 4.4.2 to carry a factored moment
Mu of 735 ft-kips. Use fc′ = 3000 psi and f y = 50, 000 psi.
SOLUTION
(a) Determine whether the depth a of the rectangular stress distribution will be greater
than t. For a = t,
C = 0.85 fc′ bE t = 0.85(3)(78)(4.5) = 895 kips
 t  4.5  1
M n = C  d −  = 895  25 −  = 1697 ft-kips
 2  2  12
M n 735
φ 0.90
Since the required Mn is less than the amount necessary to cause a = t, the section can be
designed as a rectangular section.
(b) Design as a rectangular section. The computation of the coefficient of resistance

Rn = Mu /(φ bE d 2) and the solving for ρ may be made exactly as described in Chapter 3
using Eq. (3.8.5). However, because of the wide flange (bE = 78 in.) and the likeli-
hood of a relatively small value for a, a trial procedure may be preferred. First use
0.9d as a conservative (i.e., low) trial value for the moment arm (d – a/2); then
required M n 817(12)
required As = = = 8.71 sq in.
( d − a / 2) f y 0.9(25)50
because in this case 3 fc′ < 200 psi, the minimum required tension reinforcement is (see
Section 3.7)
200 200
As ,min = bw d = (13)(25) = 1.08 sq in. < 8.71 sq in.
fy 60, 000
Try 4–#9 and 4–#10, As = 9.08 sq in. (minimum width = 12.9 in. from Table 3.9.2).
Check:
C = 0.85 fc′ bE a = 0.85(3)(78)a = 198.9a
T = As f y = 9.08(50) = 454 kips
454
a= = 2.28 in.
198.9
arm = 25 − 0.5(2.28) = 23.86 in. (i.e., 0.95d )
A second trial gives
817(12)
23.86(50)
Revise to 4–#8 and 4–#10, As = 8.24 sq in. Check:

C = 198.9a (as before)
T = 8.24(50) = 412 kips
412
a= = 2.07 in.
198.9
arm = 25 − 0.5(2.07) = 23.96 in. ≈ 23.86 in. used
(Continued)
108
No further saving can be made.

Using a 1-in. clearance between layers,
dt = d + 0.5 + 1.27 / 2 = 25 + 1.14 = 26.1 in.
d − a / β1 26.1 − 2.07 / 0.85
ε t = 0.003 t = 0.003 = 0.029
a / β1 2.07 / 0.85
Since εt > 0.005 for tension-controlled sections, then φ = 0.90 as assumed; thus,

1
M n = 412(23.96) = 823 ft-kips
12
[φ M n = 0.90(823) = 740 ft-kips ] > [ Mu = 735 ft-kips] OK
Use 4–#8 and 4–#10.

Note that the strain in the extreme tension steel (2.9%) is large compared to that in a
typical rectangular section. This result is usual for floor T-beams and is due to the wide
flange, which requires only a small depth of the compression zone to balance the tensile
force in the steel.
SELECTED REFERENCES
4.1 S. Timoshenko and J. N. Goodier. Theory of Elasticity (3rd ed.). New York: McGraw-Hill, 1970
(pp. 262–268).
4.2 Karl Girkmann. Flachentragwerke (3rd ed.). Vienna: Springer-Verlag, 1954 (pp. 116–123).
4.3 Franco Levi. “Work of the European Concrete Committee,” ACI Journal, Proceedings, 57,
March 1961, 1049–1054.
4 .4 Gottfried Brendel. “Strength of the Compression Slab of T-Beams Subject to Simple Bending,”
ACI Journal, Proceedings, 61, January 1964, 57–76.
4.5 Antoine E. Naaman. “Rectangular Stress Block and T-Section Behavior,” PCI Journal, 47,
September–October 2002, 107–112.
PROBLEMS
All problems* are to be done according to the ACI (b) Determine the maximum tension reinforce-
Code. All loads given are service loads unless other- ment As permitted for this beam by ACI-
wise indicated. 9.3.3.1. What is the maximum As permitted
if the section is to be tension controlled?
4.1 (a) Determine the nominal moment strength Mn (c)
Determine and compare the design
for the beam cross section shown in the figure moment strengths φ M n corresponding to
for Problem 4.1. Use fc′ = 4000 psi and the two areas of tension steel determined in
f y = 60, 000 psi. ( fc′ = 30 MPa; f y = 400 MPa; part (b).
slab t = 125 mm; beam h = 900 mm; 4.2 (a)
Design the reinforcement for the beam
bw = 380 mm.) shown in the figure for Problem 4.2, if the
* Problems may be solved as problems stated in Inch-

Pound units, or as problems in SI units using the quantities in paren
theses. The SI values are approximate conversions to avoid implying higher precision for given information in metric units than
for Inch-Pound units.
109
PROBLEMS 109
5”
d
36” Inch-Pound SI
As = 9.08 sq in. 4 – #30M
(4 – #9, 4 – #10, 4 – #35M
two layers)
Span = 30 ft
15” 4” (100 mm)
8’ – 0”
(2.4 m)
Problem 4.1
dead load moment is 90 ft-kips and the live Use fc′ = 5000 psi and f y = 60, 000 psi.
load moment is 140 ft-kips. (M D = 122 kN ⋅ m ; M L = 190 kN ⋅ m ; fc′ =
(b) Determine the maximum tension reinforce- 35 MPa; f y = 400 MPa.)
ment As permitted for this beam by ACI- 4.3 Repeat Problem 4.2 for a dead load moment
9.3.3.1 and for a tension-controlled section. of 65 ft- kips and a live load moment of
100 ft- kips. Use fc′ = 3500 psi and f y =
48” (1.2 m) 4” (100 mm)
60,000 psi. (M D = 88 kN ⋅ m; M L = 135 kN ⋅ m ;
fc′ = 24 MPa; f y = 400 MPa.)
24” (610 mm)
As
15”
(380 mm)
Problem 4.2
CHAPTER 5
SHEAR STRENGTH AND
DESIGN FOR SHEAR
5.1 INTRODUCTION
This chapter treats the shear strength of nonprestressed one-way flexural members.
Effects of axial compression and tension are included in Section 5.13. Deep beams
behave differently than slender beams, and so their design is usually performed using
strut-and-tie models, as covered in Chapter 14. Consideration is also given to the shear-
friction concept (Section 5.15) and its application to the design of corbels and brackets
(Section 5.16). Shear strength of prestressed concrete members is treated in Chapter 20.
Shear effects arising from torsion and its interaction with shear due to bending are treated
in Chapter 18.
For the simple beam shown in Fig. 5.1.1, the bending moment M at section A-A causes
compressive stresses in the concrete above the neutral axis and tensile stresses in the rein-
forcement (and in the concrete below the neutral axis if it has not yet been cracked). To
satisfy vertical force equilibrium, the resultant of the vertical shear stresses across the sec-
tion must be equal to the shear force V. In an element located at the neutral axis, there is
a state of pure shear as shown in Fig. 5.1.2, which gives rise to a principal tensile stress
Shear failure of reinforced concrete beam without web reinforcement. (Photo by Gustavo
J. Parra-Montesinos).
1
5 . 2 S hear S tresses B ased on L inear E lastic B eha v ior 111
A
A
N.A.
V
A
A
Figure 5.1.1 Shear force and bending moment at a section in a simply supported beam.
ft (max) = v
v v v
45°
v v
Figure 5.1.2 State of pure shear stress (i.e., no tensile or compressive stresses on the faces of the
element).
equal to the shear stress, and acting on a 45° plane from the longitudinal beam axis. This
diagonal tension will cause diagonal cracking, which can lead to premature failure of the
beam (i.e., prior to development of the full flexural strength with ample deformation). Thus
the failures in beams commonly referred to as “shear failures” are often actually diagonal
tension failures caused by the formation of inclined cracks. Among the earliest to recog-
nize this phenomenon were Emil Mörsch [5.1] in Germany and Arthur Talbot in the United
States [5.2] in the early 1900s.
Bresler and MacGregor [5.3] have presented an excellent systematic treatment of the
various situations in which shear-related cracks develop. Their work was expanded by
ACI-ASCE Committee 426 [5.4]. Collins [5.5], Collins and Mitchell [5.6], Marti [5.7],
Vecchio and Collins [5.8, 5.10], Schlaich, Schäfer, and Jennewein [5.9], Ramirez and
Breen [5.11], Al-Nahlawi and Wight [5.12], Collins, Mitchell, Adebar, and Vecchio [5.13],
and Tureyen and Frosch [5.14], among others, have presented alternative models for esti-
mating the shear strength of beams. A good review of additional approaches for the shear
design of reinforced concrete members has been published by ASCE-ACI Committee 445
[5.15].
This chapter begins with the concepts of horizontal and vertical shear, and the resulting
principal tensile stress, so that the reader may gain insight into the potential direction of
inclined cracks. Then, the variables affecting shear strength, leading up to the current ACI
Code provisions, are presented.
5.2 SHEAR STRESSES BASED ON LINEAR

ELASTIC BEHAVIOR
A simply supported, uniformly loaded rectangular beam is shown in Fig. 5.2.1. Consider
the free body of the elemental block abcd, as shown in Fig. 5.2.1(b), where kd is the elastic
neutral axis distance. Horizontal force equilibrium requires
vy b dz = C2 − C1 (5.2.1)
12
112 C hapter 5 S hear S trength and D esign for S hear
1 2
w
a b N.A.
c d d
e f y
g h
b
dz
1 2
(a)
w vb dz
fc1 a fc2 e f
b
C1 C2
d T1 T2
c kd
fc1y fc2y
y vy b dz
g h
N.A.
(b) (c)
Figure 5.2.1 Horizontal shear stress in a beam.
in which vy is the unit horizontal shear stress on a plane at a distance y from the neutral
axis. Under service loads, the flexural stress distribution in the compression region may be
assumed to vary linearly, thus
1
C1 = ( fc1 + fc1 y )b(kd − y)
2
y
f c1 y = f c1
kd
1  y 1   y  2 
C1 = fc1  1 +  (kd − y)bb = fc1bkd 1 −    (5.2.2)
2  kd  2   kd  
Dividing the bending moment M1 on section 1-1 by the arm of the internal couple gives the
full internal compressive force on section 1-1, or
M1 1
= fc1b(kd )
arm 2
Substitution of fc1 from the above expression into Eq. (5.2.2) gives
M1  y2 
C1 = 1 −  (5.2.3)
arm  (kd )2 
Similarly,
M2  y2 
C2 = 1 −  (5.2.4)
arm  (kd )2 
Substituting Eqs. (5.2.3) and (5.2.4) into Eq. (5.2.1) gives
 M − M1  1  y2 
vy =  2 1−
 dz  b (arm)  (kd )2 
 
(5.2.5)
V  y2 

= 1 − 
b (arm)  (kd )2 
Equation (5.2.5) is valid from y = 0 at the neutral axis to the extreme concrete compression
fiber, where y = kd . Proceeding from the extreme compressive fiber to the neutral axis,
13
5 . 3 C o m bined N or m al and S hear S tresses 113
the differential horizontal force C2 – C1 increases to a maximum. Thus at the neutral axis
( y = 0), Eq. (5.2.5) gives the maximum shear stress
V
v= (5.2.6)
b(arm)
The horizontal shear stress v of Eq. (5.2.6) is also the vertical shear stress, because shear
stresses on two perpendicular planes must be equal. For a homogeneous, linear elastic rectan-
gular beam of overall depth h, the arm is equal to (2 / 3) h, which leads to a maximum shear
stress at the location of the neutral axis equal to 1.5 times the average shear stress V / ( bh ).
For strength design, the shear force V is the factored shear force and the distribution of
flexural stress is no longer linear in the compression region. The shear stress v of Eq. (5.2.6)
is then not the actual shear stress. Distribution of shear stresses over the depth of a cracked
section is difficult to estimate. However, the relative magnitude obtained from Eq. (5.2.6) is
still a measure of the potential for inclined cracking. In the ACI Code, the denominator in
Eq. (5.2.6) is thus replaced by b times the full value of the effective depth d.
5.3 COMBINED NORMAL AND SHEAR STRESSES

If at a certain point below the neutral axis in a homogeneous beam the tensile stress is ft
and the shear stress is v [Fig. 5.3.1(a)], the principal tensile stress ft ( max ) is given by
2
1 1 
ft (max) = ft +  ft  + v 2 (5.3.1)
2 2 
The derivation of Eq. (5.3.1) is available in most textbooks on mechanics of materials, but
because of its importance, it is shown here, as well.
Using equilibrium of the forces acting on the free body in the directions of ft ′ and v ′
shown in Fig. 5.3.1(b) and calling the width of the beam b,
 bdz   bdz   bdz 

ft ′  = ft  cos α + v(bdz ) cos α + v  sin α
 sin α   tan α   tan α 
 bdz   bdz   bdz 

v′  = ft  sin α + v(bdz )sin α − v  cos α
 sin α   tan α   tan α 
from which
1
ft ′ = ft (1 + cos 2α ) + v sin 2α
2
1
v ′ = ft sin 2α − v cos 2α
2
dz
v v v
ft ft ft ft v αmax
α v’ v’ = 0
dz αmax
v v tan α v ft’
ft (max)
dz
v sin α
(a) (b) (c)
Figure 5.3.1 Stress condition on an infinitesimal element block.

14
ft (max)
Figure 5.3.2 Directions of potential cracks in a simply supported beam under uniform loading.
The value of the principal angle α max (i.e., the angle α that makes ft ′ maximum and at
the same time makes v′ zero) may be found either by differentiating the expression for ft ′
with respect to α or by setting the expression for v′ to zero. This leads to the following
expression, for which the principal angle can be determined if the shear stress v and normal
stress ft are known.
v
tan 2α max = (5.3.2)
1
f
2 t
Equation (5.3.1) may then be obtained by substituting Eq. (5.3.2) into the expression for ft ′.
The principal tensile stress ft ( max ), which is at an angle α max with the beam axis, is
at least as large as either ft or v. It is nearly equal to the longitudinal tensile stress ft if
the shear stress v is small (i.e., near the extreme tension fiber of a beam) and its direction
is nearly horizontal. It is nearly equal to the shear stress v if the longitudinal tensile stress
ft is small, and its direction is nearly at 45° with the beam axis. Since concrete is weak in
tension, these principal tensile stresses are undoubtedly correlated to inclined cracking, as
shown in Fig. 5.3.2.
5.4 BEHAVIOR OF BEAMS WITHOUT SHEAR

REINFORCEMENT
As discussed in Section 5.3, diagonal tension caused by shear stress on a beam may
result in the formation of inclined cracks. Typically the propagation and width of inclined
cracks are controlled by using transverse reinforcement (known as “shear reinforcement”)
in the form of closed or U-shaped stirrups; normally the stirrups are placed perpendicular
to the beam longitudinal axis to enclose the main longitudinal reinforcement along the
faces of the beam (see Fig. 5.6.1 in Section 5.6). Before discussing the shear behavior and
strength of beams with shear reinforcement, however, it is necessary to have a clear under-
standing of shear behavior of beams without transverse reinforcement.
Inclined cracking in the webs of reinforced or prestressed concrete beams may develop
either in the absence of flexural cracks or as an extension of a previously developed flexural
crack. An inclined crack occurring in a beam with no cracks due to flexure is known as a
web-shear crack, as shown in Fig. 5.4.1(a). An inclined crack originating at the top of and
becoming an extension to a previously existing flexural crack is known as a flexure-shear
crack, as shown in Fig. 5.4.1(b). In this case, the critical flexural crack is referred to as the
“initiating crack.”
Flexure-shear cracks are the usual type found in both reinforced and prestressed con-
crete. Web-shear cracks are relatively rare in nonprestressed beams. These cracks occur
in thin-webbed I-shaped beams having relatively large flanges, common only in pre-
stressed concrete construction. This is discussed further in Chapter 20, which is devoted
entirely to prestressed concrete. Web-shear cracks may also occur near the inflection
points or bar cutoff points on continuous reinforced concrete beams subjected to axial
tension [5.15, 5.16].
15
5 . 4 B eha v ior of B ea m s W itho u t S hear R einforce m ent 115
Section A–A
A
(a) Web-shear crack
Flexure-shear
crack
Initiating crack
Secondary crack
(b) Flexure-shear crack
Figure 5.4.1 Types of inclined cracks. (From Bresler and MacGregor [5.3].)
Va = aggregate interlock (interface shear)
C
Vcz = shear
resistance
Arm, ha
of uncracked
concrete
T
Vd = dowel
force
w
z
Figure 5.4.2 Shear resistance components after formation of inclined crack.
Shear Resisting Mechanisms

The transfer of shear in reinforced concrete members without shear reinforcement occurs
by a combination of the following mechanisms [5.4], as shown in Fig. 5.4.2:
1. Shear resistance of the uncracked concrete in the compression zone, Vcz .

2. Aggregate interlock (or interface shear transfer) force Va, tangentially along a crack
[5.18–5.20], and similar to a frictional force due to irregular interlocking of the
aggregates along the rough concrete surfaces on each side of the crack.
3. Dowel action, Vd, the resistance of the longitudinal reinforcement to a transverse
force [5.21–5.26].
4. Arch action [see later: Fig. 5.4.5(a)] on relatively deep beams.
The ability of a beam to carry additional load after an inclined crack has formed depends
on whether the portion of shear formerly carried by uncracked concrete can be redistributed
across the inclined crack and the uncracked compression zone. Mechanisms 1 through 4
listed above all participate in the redistribution, the success of which determines the shear
strength and the degree of seriousness of the crack formation.
For rectangular beams without shear reinforcement, it is reported [5.4] that after an
inclined crack has formed, the proportion of the shear transferred by the various mecha-
nisms is roughly as follows: 15 to 25% by dowel action, 20 to 40% by the uncracked con-
crete compression zone, and 30 to 50% by aggregate interlock or interface shear transfer.
As diagonal cracks increase in length and width, however, the relative contributions of
these mechanisms will change. For example, the contribution of aggregate interlock to
16
the beam shear strength will decrease as diagonal cracks widen with increasing load. This
reduction in shear resistance contributed by aggregate interlock (and also by dowel action)
with increasing deformation may lead to a shear failure even after the beam has reached
its flexural strength, as progressive flexural yielding causes cracks to widen. It is therefore
critical that a beam be capable not only of reaching its flexural strength, but also of exhibit-
ing substantial flexural deformations prior to the development of a shear failure should the
beam ultimately fail in shear.
Inclined cracks begin and grow depending on the relative magnitudes of shear stress v
and flexural stress ft , as shown in Section 5.3. These controlling stresses may be expressed as
V
v = k1 (5.4.1a)
bd
M
ft = k2 (5.4.1b)
bd 2
where k1 and k2 are proportionality constants. The discussion in Section 5.2 may serve to
justify the shear stress as proportional to V / bd ; in the presentation of Chapters 3 and 4,
the flexural capacity M was related to bd 2 through the use of a coefficient of resistance R,
which has the same units as the flexural stress.
From Section 5.3 one may note that the principal tensile stress is a function of the ratio
ft / v. Another expression for ft /v may be obtained from Eqs. (5.4.1); thus
ft k2 M M
= = k3 (5.4.2)
v k1 Vd Vd
For a simply supported beam symmetrically loaded with two equal concentrated loads (see
Fig. 5.4.3), the ratio M / V may be thought of as the distance a over which the shear is con-
stant. This distance a is known as the shear span. For the general case where the shear is
continually varying, the “shear span” may be expressed as
M
a= (5.4.3)
V
which varies along the length of the beam.

Using Eq. (5.4.3) in Eq. (5.4.2), the ratio ft / v becomes
ft  a
= k3   (5.4.4)
v  d
The ratio of shear span to effective depth a / d has also been shown experimentally to be a
highly influential factor in establishing shear strength [5.3, 5.4, 5,17, 5.27]. When factors
P P
a a
V = +P
V = –P
M = Va
Figure 5.4.3 Definition of shear span a.

17
5 . 4 B eha v ior of B ea m s W itho u t S hear R einforce m ent 117
other than a / d are kept constant, shear capacity tends to decrease with an increase in a / d
ratio. The variation in shear capacity, expressed in terms of the failure moment Va, with a / d
ratio, may be illustrated by Fig. 5.4.4 using the results for rectangular beams.
Member depth has also been found to significantly affect shear strength of flexural
members without shear reinforcement. Results from tests of large flexural members
[5.28–5.30] have shown that an increase in member depth results in a decrease in shear
stress at failure. This phenomenon has often been attributed to an increase in crack spac-
ing with member depth, which results in wider cracks and reduced aggregate interlock
[5.10, 5.29, 5.30]. Thus, this so-called shear size effect has been linked not only to mem-
ber depth, but also to aggregate size. Shear size effect has also been reported to be more
pronounced in members constructed with high-strength concrete [5.29]. With the use of
shear reinforcement, however, shear size effect is greatly diminished or even eliminated,
as transverse steel contributes to a better cracking distribution and controls the growth of
diagonal cracks.
Failure Modes
From Fig. 5.4.4, four general categories of failure may be established: (1) deep beams with
a /d < 1; (2) short beams with a /d ratios from 1 to about 2 1 2 , in which the shear strength
exceeds the inclined cracking capacity; (3) usual beams of intermediate length having a /d
ratios from about 2 1 2 to 6, in which the shear strength equals the inclined cracking strength;
and (4) long beams with a /d ratios greater than 6, whose flexural strength is typically less
than their shear strength.
Deep Beams, a/d ≤ 1

For a deep beam, shear stress has the predominant effect. After inclined cracking has
occured, a simply supported beam such as that shown in Fig. 5.4.5(a) tends to behave
like a tied arch wherein the load is carried by direct compression extending around the
shaded area and by tension in the longitudinal steel. Once the shear-related crack devel-
ops, the beam transforms quickly into a tied arch which exhibits considerable reserve
capacity.
Possible modes of failure are indicated in Fig. 5.4.5(b) [5.3]; they are (1) an anchor-
age failure (i.e., pullout of the tension reinforcement has occurred at the support) (2) a
crushing failure at the reactions; (3) a “flexural failure” arising from either crushing of
concrete near the top of the arch or yielding of the tension reinforcement; and (4) failure
of the arch rib due to an eccentricity of the arch thrust, resulting in either a tension crack
Flexural Compression
Shear-compression strength arch
strength
Inclined cracking
Failure moment = Va
strength, Vc Tension tie
(a) Arch action
Shear-tension and
Deep shear-compression Flexural
4 3 1. Anchorage failure
beams failures failures 2. Bearing failure
5
2 3. Flexural failure
1 4. & 5. Arch-rib failure
Diagonal tension failures 3
0 1 2 3 4 5 6 7
a/d (b) Types of failure
Figure 5.4.4 Variation in shear strength with a/d for Figure 5.4.5 Modes of failure in deep beams, a/d
rectangular beams. (Adapted from Bresler and MacGregor [5.3].) ≤ 1.0. (Adapted from Bresler and MacGregor [5.3].)
18
over the support at region 4 of Fig. 5.4.5(b) or crushing of concrete on the underside of

the rib at region 5.
Short Beams, 1 < a/d ≤ 2 12

Like deep beams, short beams have a shear strength that exceeds the inclined cracking
strength. After an inclined crack develops, the crack extends further into the compression
zone as the load increases. It may also propagate as a secondary crack toward the tension
reinforcement and then progress horizontally along that reinforcement. Failure eventually
results, either by (1) an anchorage failure of the tension reinforcement, called a “shear-
tension” failure [Fig. 5.4.6(a)]; or (2) a crushing failure in the concrete near the compress-
ion face, called a “shear-compression” failure [Fig. 5.4.6(b)].
Intermediate Length Beams, 2 12 < a/d ≤ 6

For intermediate length beams, vertical flexural cracks are the first to form, followed by
the inclined flexure-shear cracks. At the beginning, several flexural cracks tend to incline,
creating beam segments between cracks. Kani [5.31] proposed a “comb” analogy, where
the “teeth” of the comb are the concrete pieces in between flexural cracks (Fig. 5.4.7). He
proposed that, as a result of the increasing number of flexural cracks, when the root of the
“tooth” is so reduced in size that it becomes unable to carry the moment arising from ΔT,
it breaks to form the inclined flexure-shear crack. At the sudden occurrence of the inclined
crack, the beam is not able to redistribute the load, as is the case in beams with smaller
a/d ratio. In other words, the formation of the inclined crack represents the shear strength of
beams in this category, for which the term “diagonal tension failure” has been given [5.3].
This is the most common category for beam design. Moreover, for beams in this category,
a shear failure may develop before or after the beam has reached its flexural strength. As
discussed above, widening of cracks caused by flexural yielding may lead to a shear failure
soon after flexural yielding begins due to a decrease in the shear resistance provided by
aggregate interlock and dowel action.
Long Beams, a /d > 6

The failure of long beams starts with yielding of the tension reinforcement and ends by
crushing of the concrete at the section of maximum bending moment. In addition to the
nearly vertical flexural cracks at the section of maximum bending moment, prior to failure,
slightly inclined (from the vertical) cracks may be present between the support and the
section of maximum bending moment. Nevertheless, the strength of the beam is entirely
dependent on the magnitude of the maximum bending moment, as the shear force required
for the beam to reach its flexural strength is less than its shear capacity.
In summary, shear tends to cause inclined cracks. When no such cracks form before
the nominal flexural strength is reached, the effect of shear is often negligible, although it
is possible for a beam to fail in shear after the initiation of flexural yielding, since aggre-
gate interlock and dowel action decrease with increasing deformation. Beams may fail
on formation of an inclined crack, as for the so-called diagonal tension failures when
a/d is between about 2 1 2 and 6, or there may be considerable reserve strength, as is the
case for shorter beams. Experimental results also indicate a reduction in shear strength
with increased member depth. For beams with reserve capacity to maintain a state of equi-
librium, forces must be redistributed after an inclined crack has formed. For the design of
all but deep beams, however, the shear strength of beams without shear reinforcement is
assumed to be reached when the inclined crack forms.
19
5.5 STRENGTH OF BEAMS WITHOUT SHEAR REINFORCEMENT 119
Crushing of concrete
Loss of bond due to crack
(a) Shear-tension failure (b) Shear-compression failure
Figure 5.4.6 Typical shear failures in short beams, a / d = 1 to 2 1 2 . (Adapted from Bresler and
MacGregor [5.3].)
Tooth
Figure 5.4.7 “Diagonal tension failure” or “tooth-cracking failure” as suggested by Kani [5.31]

on intermediate length beams; a/d about 2 1 2 to 6.
5.5 SHEAR STRENGTH OF BEAMS WITHOUT SHEAR

REINFORCEMENT—A CI APPROACH
Except for deep beams (see Section 5.14), the strength at which an inclined crack forms
[usually a flexure-shear crack as in Fig. 5.4.1(b)] is taken to be the shear strength of a beam
without shear reinforcement, according to the intent of the ACI Code. After establishing
on a rational basis the variables involved, the relationship between them was statistically
determined from test results [5.17].
It is assumed that the strength is reached when the principal tensile stress in Eq. (5.3.1)
reaches the tensile strength of concrete, which is assumed to be proportional to fc′ .
Although the exact distributions of the flexural and shear stresses in a cross section are not
known, it may be assumed that flexural tensile stress ft varies as Ec / Es times the tensile
stress in the reinforcement and that v varies as the average shear stress. Assume that Ec is
also proportional to fc′ .
As already shown by Eq. (5.4.1a), the average shear stress when the inclined crack forms
may be written
V
v = k1 (5.5.1)
bd
The stress in the steel is proportional to M /( As d ), and the tensile stress ft in the concrete
then becomes
Ec f s EM M fc′ M  fc′ 
ft ∝ ∝ c ∝ ∝ 2 
Es Es dAs Es dAs bd  ρEs 
The above expression for ft may be written as
k4  fc′  M
ft =   (5.5.2)
Es  ρ  bd 2
in which k4 is a dimensionless constant and Es has a known value. Also, the tensile strength
of concrete may be represented by
ft (max) = k5 fc′ (5.5.3)

120
Substituting Eqs. (5.5.1) to (5.5.3) into the principal stress equation, Eq. (5.3.1),
  1 k M fc′ 
2 
V  1 k4 M fc′ 2 
k5 fc′ = + 
4
 + k
bd  2 Es Vd ρ
1 
  2 Es Vd ρ  
or
 2 
V
= k5  1 k4 M fc′ +  1 k4 M fc′  + k 2  (5.5.4)
 2 E Vd ρ   1 
bd fc′  2 Es Vd ρ 
 s

In Eq. (5.5.4) the variables are observed to be V / (bd fc′) and M fc′ / ( Es ρVd ). It may
be noted that these two variables are nondimensional quantities because fc′ has units of
force per unit area. In the statistical study, the shear force V was defined as that causing
the critical inclined crack and M as the corresponding moment at the location of diagonal
cracking which, for shear spans longer than 2d, was taken at a distance d from the section of
maximum moment. On the basis of 440 tests [5.17], as shown in Fig. 5.5.1, the relationship
between these two variables, on using the numerical value of Es , was obtained as follows:
V ρ Vd
= 1.9 + 2500 ≤ 3.5 (5.5.5)
bd fc′ M fc′
Note that in Fig. 5.5.1, V on the left-hand side is taken as the nominal shear strength, while V
and M on the right-hand side are taken as the applied shear and corresponding moment at the
critical section. Equation (5.5.5) is generally considered [5.4, 5.28, 5.31] to be acceptable for
predicting the flexure-shear cracking load, particularly for M / (Vd ) (i.e., shear span/depth)
ratios of about 2 1 2 to 6, with considerable conservatism for lower M / (Vd ) values. Thus the
favorable factors for the shear strength of beams without shear reinforcement are a high per-
centage ρ of longitudinal reinforcement and a high ratio of Vd to M, that is, a low a / d ratio.
Since 1963, the ACI Code has accepted the relationship of Eq. (5.5.5) as the shear
(inclined cracking) strength of beams without shear reinforcement. Thus defining Vc as the
nominal strength of such beams, using the web width1 bw for b and adding a multiplier λ to
fc′ to account for the use of lightweight concrete (see Section 1.8), Eq. (5.5.5) becomes
 ρ V d
Vc =  1.9λ fc′ + 2500 w u  bw d ≤ 3.5λ fc′bw d (5.5.6)2
 Mu 
which is given in Table 22.5.5.1 of the ACI Code. For normal weight concrete, λ = 1.0.
Values for λ when lightweight concrete is used are discussed shortly in this section (see subsec-
tion on lightweight concrete). Note that in Eq. (5.5.6), the factored shear force Vu and the fac-
tored moment Mu acting concurrently with Vu at the section under consideration are used. Also
note that the reinforcement ratio ρw = As / ( bw d ) is used in the ACI Code formula, where bw is
the web width for a T-section rather than the flange width. The ACI Code defines bw as “web
width.” For tapered webs, however, such a definition is unclear. In general, when the web is
subject to flexural tension, the “average web width” should be used as bw in Eq. (5.5.6). On the
other hand, when the web is subject to flexural compression (as for negative moment regions)
the use of average web width may be unsafe [5.32]. For such negative moment regions, the
use of a value lower than the average, perhaps the minimum, web width, is more appropriate.
The value of Vu d / Mu in Eq. (5.5.6) shall not be taken greater than 1.0 except when
axial compression is present [see Eq. (5.13.2), where a modified moment replaces Mu ],
which has the effect of limiting Vc at and near the points of inflection.
1 The term “web” refers to the width dimension of a beam at its neutral axis. For T-shaped beams the flange width b is dis-
tinguished from the stem, or web, width bw.
2 For SI, ACI 318-14M, with fc′ in MPa, gives
 ρ V d
Vc =  0.16 λ fc′ + 17 w u  bw d ≤ 0.29λ fc′ bw d (5.5.6)
 M 
u
12
5.5 STRENGTH OF BEAMS WITHOUT SHEAR REINFORCEMENT 121
6.0
Vn = Vc
5.0
4.0
bwd fc’
3.0
V
Vn
=2
bw d fc’
2.0
Vn ρ Vu d
= 1.9 + 2500 ≤ 3.5
bw d fc’ Mu fc’
1.0
Inverse scale
0.2 0.4 0.6 0.8 1.0 1.5 2.0 5
1000
(
ρ Vu d
Mu fc’ )
Figure 5.5.1 Comparison of shear stresses calculated using Eq. (5.5.5), with test results (shown as
dots) for beams without shear reinforcement, that is, the nominal shear strength Vn = Vc. (Adapted
from ACI-ASCE Committee 326 Report [5.17].)
In lieu of Eq. (5.5.6), the ACI Code (ACI-22.5.5.1) allows the use of a simplified expres-
sion, as follows:
Vc = 2 λ fc′ bw d (5.5.7)
Note that except for a few experimental results, this equation represents a lower bound of
the test data shown in Fig. 5.5.1. More recent work [5.29, 5.30], however, has shown that
large beams without web reinforcement may fail at shear stresses significantly lower than
those obtained using either Eq. (5.5.6) or (5.5.7).
Continuous Beams
The application of Eq. (5.5.6) for continuous beams has been subject to question [5.4]. The
strut action on a continuous beam is shown in Fig. 5.5.2. The analogy in Fig. 5.4.3 that
M / V equals the shear span a implies that at the point of zero moment there is a support to
accommodate a diagonal strut [see Fig. 5.4.5(a)]. For a continuous beam as in Fig. 5.5.2,
the distances a1 and a2 are analogous to a in Fig. 5.4.3; however, there is no support to resist
a strut reaction at the inflection point. ACI-ASCE Committee 426 [5.16] recommended that
Eq. (5.5.6) [ACI Table 22.5.5.1] no longer be used in such situations; instead Vc should be
taken as 2λ fc′ bwd [Eq. (5.5.7)].
Lightweight Concrete
It has been shown [5.33–5.35] that the same general relationships as those for normal-
weight concrete are valid for lightweight concrete when a multiplier λ , typically less than
1.0, is applied to fc′ [see Eq. (5.5.6)]. This multiplier accounts for the fact that for light-
weight concrete, the tensile strength fct based on the split-cylinder test (see Section 1.8)
provides a better correlation with inclined cracking strength than does fc′ .
Tests have shown [5.33] that inclined cracking strength for lightweight concrete varies
from about 60 to 100% of the values for normal-weight concrete of the same nominal com-
pressive strength, depending on the particular aggregates used. Studies [5.35] have shown
that using 0.75 to 0.85 of the tensile strength–related term fc′ for normal-weight con-
crete in shear strength equations is a reasonable and generally conservative approach. Thus,
alternatively to the use of results from split-cylinder tests, λ may be taken as 0.75 for all-
lightweight concrete and 0.85 for sand-lightweight concrete (ACI-19.2.4.2) (see Section).
12
t
stru
a1 = Mneg /V a2 = Mpos /V
Mpos
– Point of
inflection
Mneg
Figure 5.5.2 Crack pattern, diagonal strut, and bending moment diagram near support on a
continuous beam.
High-Strength Concrete
In response to the increasing use of concrete having fc′ greater than 8000 psi, studies
[5.36–5.43] have indicated that shear strength does not continue to increase in proportion
to fc′ . Though a definitive relationship for the shear strength of beams using high-
strength concrete has not been established, the ACI Code Committee (ACI-22.5.3.1) limits
fc′ to 100 psi in shear strength calculations of members with fc′ above 10,000 psi.
However, values of fc′ greater than 100 psi are permitted in computing Vc as per ACI-
22.5.3.2 when the area Av of shear reinforcement (see Section 5.8) satisfies the minimum
shear (and torsion per ACI-9.6.4.2; see Chapter 18) reinforcement required in ACI-9.6.3.3
[see Section 5.10, Eq. (5.10.8)].
Size Effect
As mentioned in Section 5.4, shear stress at failure has been reported to decrease with an
increase in member depth and a decrease in aggregate size. This size effect, however, is not
accounted for in Eqs. (5.5.6) and (5.5.7). Research [5.29, 5.30] has shown that the use of
these equations may lead to unconservative strength predictions in large flexural members
without shear reinforcement. It is thus recommended that at least a minimum amount of shear
reinforcement (see Section 5.10) always be used in members with depths greater than 24 in.
5.6 FUNCTION OF WEB REINFORCEMENT

The types of shear reinforcement recognized for reinforced nonprestressed concrete by the
ACI Code (ACI-22.5.10.5.1) are (1) vertical stirrups (Fig. 5.6.1), ties or hoops perpendicular
to the longitudinal axis of the member; (2) welded wire fabric (see Section 1.13) with wires
located perpendicular to the axis of the member; and (3) spiral reinforcement. Inclined stir-
rups, where the plane of the stirrup makes an angle of 45° or more with the longitudinal rein-
forcement, are also permitted (ACI-22.5.10.5.2), as well as longitudinal tension bars bent
toward the compression zone of the member so that the axis of the bent portion makes an
angle of 30° or more with the axis of the longitudinal portion (22.5.10.6.1). Vertical stirrups,
ties, or hoops (as opposed to inclined stirrups and bent bars) are used almost exclusively for
shear reinforcement in current design practice.
The most generally accepted model for the behavior of reinforced concrete beams con-
taining shear reinforcement is the truss model, originated by Ritter [5.44] and Mörsch [5.1].
The current thinking related to the truss model is well presented by Schlaich, Schäfer, and
Jennewein [5.9]. In a simply supported steel truss, as shown in Fig. 5.6.2(b), the upper and
123
5 . 6 F u nction of W eb R einforce m ent 123
A Stirrup
Vertical stirrups
Section A–A
A
Single-leg U stirrup with U stirrup with Closed stirrup Closed stirrup

stirrup 90° hooks 135° hooks with 90° hooks with 135° hooks
Figure 5.6.1 Vertical stirrups as shear reinforcement.
(a) Reinforced concrete beam with vertical shear reinforcement
(b) A steel truss
Concrete Web reinforcement
(c) Truss action in a reinforced concrete beam
Figure 5.6.2 Truss analogy.
lower chords are in compression and tension, respectively, and the diagonal and vertical
members, called web members, are in compression and tension. In the reinforced concrete
beam of Fig. 5.6.2(c), the concrete performs the task of carrying the compressive forces
while steel reinforcement is used to resist tensile forces. Since the vertical transverse rein-
forcement (stirrups) acts in a reinforced concrete beam like the tension web members of a
steel truss, the term “web reinforcement” is often used for shear reinforcement. The shear
reinforcement wraps around the longitudinal tension reinforcement and must be anchored
in the compression zone, usually by hooking it around the top longitudinal bars. (These lon-
gitudinal bars either are compression reinforcement or are provided solely to hold in place
and anchor the shear reinforcement.)
While shear reinforcement provides shear strength, its contribution to the strength
occurs only after inclined cracks form. Prior to the formation of inclined cracks, the con-
crete performs the task of carrying the shear. However, shear reinforcement is necessary to
allow a redistribution of internal forces across any inclined crack that may form, and thus
prevent a sudden failure upon formation of the crack.
124
stirrups
Va = aggregate interlock (interface shear)
C
Vcz = shear
s resistance of
Arm, ha
uncracked concrete
Fv
Fv Fv
dowel force = Vd T
Fv = Avfv
w
z
Figure 5.6.3 Shear resisted by stirrups crossed by an inclined crack.
Splitting along longitudinal

bars
Vcz = shear carried Loss of interface shear
by concrete transfer
Internal resisting shear
Vd = dowel
Va = aggregate interlock
action
(or interface shear transfer)
Vs = shear carried
by web reinforcement
Vc Stirrups yield
Vs
Failure
Vs
Applied shear
Vc
Inclined crack forms
Flexural cracking occurs
Figure 5.6.4 Distribution of internal shears in beams with shear reinforcement. (Adapted from
ACI-ASCE Committee 426 Report [5.4].)
The role of web reinforcement in resisting shear is also illustrated in Fig. 5.6.3, which is
the same as Fig. 5.4.2, except for the presence of stirrups. In addition to the contributions to
shear strength assigned to the concrete (i.e., Vcz , Va, and Vd), the stirrups crossing the inclined
crack contribute a resistance Vs equal to the number of stirrups intersected by the crack
(w / s ) times the force in each stirrup (Fv = Av fv , where Av is the cross-sectional area of all
legs in each stirrup and fv is the stress in the stirrup).
Therefore, shear reinforcement has several functions that combine to allow a redistribu-
tion of the internal forces upon the formation of an inclined crack. The primary functions
[5.4] (see Figs. 5.6.3 and 5.6.4) are (1) to carry part of the shear Vs ; (2) to restrict the growth
of the inclined crack and thus help maintain aggregate interlock (also known as interface
shear transfer) strength Va; and (3) to tie the longitudinal bars in place and thereby increase
their strength Vd to resist transverse forces (known as dowel action). In addition, dowel
action on the stirrups may transfer a small force across a crack, and the confining action of
the stirrups on the compression concrete may slightly increase its strength.
If the amount of shear reinforcement is too little, it will yield immediately at the for-
mation of an inclined crack and the beam will fail without warning. If, on the other hand,
the amount of shear reinforcement is too large, a web-crushing failure will occur prior to
yielding of the transverse reinforcement. The ideal amount of shear reinforcement should
be such that the shear reinforcement and the compression zone of the beam each continues
to carry increasing shear after the formation of the inclined crack, allowing the shear rein-
forcement to yield and providing some warning prior to shear failure.
125
5 . 7 T r u ss Model for R einforced C oncrete B ea m s 125
5.7 TRUSS MODEL FOR REINFORCED

CONCRETE BEAMS
The reinforced concrete beam shown in Fig. 5.7.1(a) contains vertical stirrups spaced at a
constant spacing s. The behavior of a cracked reinforced concrete beam can be represented,
in its simplest form, by the truss model shown in Fig. 5.7.1(b). In this model, the diagonals,
such as gd, carry compressive forces and are represented by concrete struts. The top chord,
such as cd, would carry the internal flexural compressive force. The bottom chord (i.e., the
main tension reinforcement) would carry the internal flexural tensile force.
The vertical members, such as gc, represent the stirrups and carry tensile forces. Rather
than representing a single stirrup, each vertical member represents stirrups distributed over
the tributary width of the vertical member [i.e., w in Fig. 5.7.1(b)]. Note that the contribu-
tions to shear strength (see Fig. 5.6.4) from Vcz , Va and from dowel action in the longitudinal
reinforcement Vd are ignored in the truss model. In other words, the shear resistance is
assumed to be provided entirely by the stirrups. This drawback in the use of truss mod-
els for determining shear strength of beams had already been recognized in the 1920s by
Slater, Lord, and Zipprodt [5.45] and Richart [5.46], who proposed the addition of a shear
contribution as a function of the concrete compressive strength fc′ to the strength obtained
through the use of a truss model.
Figure 5.7.1(c) shows a refined truss model for the same beam, except that the vertical
member spacing has been reduced to s. In this case, each vertical member represents a
single stirrup. In this model, it is noted that the diagonals, such as fc, f ′c ′, and f ′′c ′′, do not
have to extend from the top of one vertical member to the bottom of the next vertical mem-
ber. This idealization stems from the observation that, in reality, the shear is resisted by a
continuous field of diagonal compressive stresses, often called a compression field, rather
than by discrete diagonal struts. The angle of the compression field depends primarily on
the size and spacing of the stirrups, amount of longitudinal reinforcement, and axial load,
but it is often taken between 30 and 45°.
In the vicinity of concentrated loads [under the concentrated load and at supports in
Fig. 5.7.1(c)], the traditional beam theory, which assumes that plane sections remain
plane, is not applicable, and the shear transfer mechanism cannot be assumed to be
a continuous compression field. In such regions, the load will spread or “fan out” to
more than one vertical tension member to equilibrate the force. In the model, “fanning
out” of the concentrated load at midspan is represented by diagonals gd, jd, kd, and id.
Similarly, diagonals eb′, eb, and ec′ represent the “fanning out” of the reaction force at
the support.
Once a truss model has been conceived, the truss member forces can be obtained from
statics. The member forces in the diagonals, verticals, and top and bottom chords are then
checked against prescribed limits. For example, the compressive stresses in the concrete
struts (top chord and diagonals in Fig. 5.7.1) are limited by an effective concrete stress fc
which is less than fc′. The value of fc depends on the stress field in the region where the
member is located. For instance, a diagonal concrete strut that crosses a stirrup in a region
with shear cracks has a lower strength than a concrete strut in the compression zone of the
beam. The strength of the vertical members and bottom chord is usually taken as the area of
the reinforcement times its yield strength. It is noted that truss models can be used to design
not only the shear reinforcement, but also the longitudinal reinforcement. Marti [5.47]
provides several practical examples of the use of truss models in the design of reinforced
concrete beams. Strut-and-tie models, which are in effect truss models, and their applica-
tion to the design of structural members, are discussed in Chapter 14.
The truss analogy has served as the basis for the development of other approaches, such
as the variable-angle truss models of Marti [5.47] and Nielsen [5.48], the compression field
theory of Mitchell and Collins [5.49], the modified compression field theory of Vecchio and
Collins [5.8], and the softened truss model of Hsu [5.50].
The zone in the vicinity of a concentrated load, or at an abrupt discontinuity in the mem-
ber (such as an abrupt change in member depth) is called a D-region, where D stands for
discontinuity, disturbance, or detail [5.9]. In such regions, the plane section theory is not
126
P
Stirrups
Main tension steel

P s P
2 2
(a) Reinforced concrete beam
P
1
a b c d
arm
α
P e f g h j P
1
2 2
6 @ w = 6w (w = panel width)
(b) Truss model for concrete beam
a b’ b c’ c c” d
P e f’ f f” g j h k i P
s
2 2
(c) Concrete beam showing possible compression fields
s Requirement for main tension

reinforcement with stirrups
n spaced at s
PL
4(arm)
m o
(d) Internal tensile or compressive force from bending moment, M/arm
Figure 5.7.1 Truss model for reinforced concrete beam.
applicable, and the true forces are not those obtained by shear and moment diagrams and
first-order elastic beam theory, particularly when the sections are cracked. In general, the
reinforcement design of the D-region has relied on “past experience” or “good practice.”
However, truss models or strut-and-tie models have become accepted tools for the design
of D-regions.
In the regions of a beam where the assumption of plane sections remaining plane after
loading (the Bernoulli hypothesis) is appropriate, such as panel bc of Fig. 5.7.1(b), and
127
5 . 7 T r u ss Model for R einforced C oncrete B ea m s 127
therefore design for the forces at each section along the member is appropriate, the term
“B-region” is used (where B stands for beam or Bernoulli).
The truss model also shows that the use of forces from the homogeneous beam bending
moment diagram, such as mno in Fig. 5.7.1(d), to design the main tension reinforcement is
not correct. Rather, since the shear reinforcement is provided at discrete intervals (stirrups
at spacing s), the tensile force T in the main reinforcement obtained from a truss model cor-
responds to the stepped diagram of Fig. 5.7.1(d), with the tensile force being constant over
the panel distance (i.e., the tensile force changes only at the location of a vertical member).
Since the real situation is not a pin-jointed truss, and the angle of the compression field
may vary, the requirement for the main tension steel would actually be somewhere between
the M/arm [based on ordinary bending moment diagram) and the stepped diagram for a
truss, as shown in Fig. 5.7.1(d)].
Shown in Fig. 5.7.2 is a segment along a beam, together with the internal forces acting
on that segment. At the section x, the forces are Ccx (concrete compression chord), Cwx (the
diagonal “strut”), and Tsx (the tensile force carried by the main tension steel). Equilibrium
in the vertical direction requires that the shear force Vx be resisted by the vertical component
of the diagonal strut. Thus
Vx = Cwx sin θ (5.7.1)
The horizontal forces Tsx and Ccx can be calculated as
Mx (a − z )
Tsx = + Vx (5.7.2)
arm arm
Mx Vz
Ccx = − x (5.7.3)
arm arm
where z is the distance measured from the vertical member of the truss to section x (see
Fig. 5.7.2).
The important observation is that the truss model recognizes that the force in the tension
steel is often larger than that obtained from the bending moment diagram.
In general, this effect has been accounted for in ACI Code design by extending the lon-
gitudinal bars a distance d beyond the point where they are no longer required. This is dis-
cussed further in Chapter 6, which is devoted to determining bar lengths and the distance
along the span that bars must extend. However, the truss model also shows that the shear
reinforcement behavior and the forces in longitudinal bars cannot be entirely separated.
The truss model also illustrates the importance of reinforcement at the “joints” of the
truss. The behavior of the stirrups, their proper detailing, and their effectiveness have been
studied by Hsiung and Frantz [5.51], Johnson and Ramirez [5.52], Anderson and Ramirez
[5.53], Mphonde [5.54], Mphonde and Frantz [5.55], and Belarbi and Hsu [5.56].
(arm) cos θ
Ccx1
Mx1 Mx
Ccx
Vx1 Cwx
arm Cwx1
θ θ
Tsx
Vx Tsx1
a a = (arm) cot θ
x1 x x1 x
z
Figure 5.7.2 Beam actions and internal forces in truss model for equilibrium of beam element.
128
5.8 SHEAR STRENGTH OF BEAMS WITH SHEAR

REINFORCEMENT—A CI APPROACH
The traditional ACI approach to design for shear strength is to consider the total nominal
shear strength Vn as the sum of two parts,
Vn = Vc + Vs (5.8.1)
in which Vn is the nominal shear strength; Vc is the shear strength of the beam attributable
to the concrete, and Vs is the shear strength attributable to the shear reinforcement (see
Figs. 5.6.3 and 5.6.4, and Section 5.5). Note that regardless of whether shear reinforce-
ment is provided, the ACI Code uses the same expression for Vc . As discussed earlier, Vc
is intended to represent the shear corresponding to inclined cracking in members without
shear reinforcement. In the case of members with shear reinforcement, however, Vc is meant
to account for the contribution to shear strength of the concrete compression zone, aggre-
gate interlock, and dowel action once diagonal cracking has occurred. For simplicity, these
two quantities are assumed to be equal.
An expression for Vs may be developed from the truss analogy. Consider the truss model
shown in Fig. 5.8.1, where the shear reinforcement (stirrups) is assumed to be inclined at
an angle α with respect to the horizontal and spaced at a distance s. The angle θ of the diag-
onal struts is assumed as shown. Equilibrium of joint A of the truss shows that the vertical
component of the tensile force developed in the stirrups must balance the shear force on the
cross section. Assuming that the web reinforcement yields prior to crushing of the concrete
(i.e., fv = f yt in Fig. 5.8.1, where f yt is the yield strength of the shear reinforcement), its con-
tribution to shear strength can be obtained as
Vs = Av f yt sin α (5.8.2)
where Av is the area of the shear reinforcement within a distance s.

From the geometry of the truss,
s = arm(cot θ + cot α ) (5.8.3)
The force per unit length in a stirrup is
Av f yt Vs
= (5.8.4)
s sin α (arm)(cot θ + cot α )
Thus
Av f yt (arm)sin α (cot θ + cot α )

Vs = (5.8.5)
s
Since the early twentieth century, the angle θ of the diagonal struts has typically been
assumed to be 45°, though in reality it will vary depending on the amount and spacing of
C
Avfv
A Vs arm
α θ α
T
s s
Figure 5.8.1 Truss model for a beam with inclined stirrups.

129
5 . 9 D efor m ed S teel F ibers as S hear R einforce m ent 129
the stirrups, the amount of longitudinal steel, and the axial load, if any. Assuming a 45°
angle for the diagonal struts, Eq. (5.8.5) becomes
Av f yt d (sin α + cos α )
Vs = (5.8.6)
s
where the arm of the internal couple has been approximated as the effective depth d. For a
beam containing stirrups perpendicular to its longitudinal axis, that is, α = 90°,
Av f yt d
Vs = (5.8.7)
s
Equation (5.8.7) can also be obtained by summing the forces in the stirrups crossed by
a diagonal crack (Fig. 5.6.3) whose projection along the longitudinal axis of the beam
(w in Fig. 5.6.3) is equal to d. The force in each stirrup is equal to Av f yt , while the num-
ber of stirrups crossed by a crack within a projection along the beam axis of d is equal
to d / s.
5.9 DEFORMED STEEL FIBERS AS SHEAR

REINFORCEMENT
Steel reinforcement in the form of randomly oriented deformed fibers (see Section 1.14)
has been shown to be effective in increasing shear strength of reinforced concrete beams
[5.62]. Fibers contribute to shear resistance by transferring tensile stresses across diagonal
cracks and restraining their widening, which in turn increases aggregate interlock. The use
of deformed steel fibers as shear reinforcement was recognized in the ACI code for the first
time in the 2008 edition. Although no shear strength equation was provided then or in the
2014 ACI Code, the contribution of fibers is implicitly recognized by allowing their use in
lieu of minimum stirrup reinforcement in normal-weight concrete beams with depths not
greater than 24 in., fc′ ≤ 6000 psi, and Vu ≤ φ 2 fc′ bw d (ACI 9.6.3.1), where φ is the strength
reduction factor for shear, equal to 0.75.
Deformed steel fibers shall comply with ASTM A820 [1.104] and have a length-
to-diameter ratio between 50 and 100 (ACI- 26.4.1.5.1). Because many types of deformed
steel fibers can be used in various dosages, performance criteria are provided in the ACI
Code, based on the use of a third-point load beam test as defined in ASTM 1609 [1.105].
To be considered as minimum shear reinforcement, fiber-reinforced concrete must satisfy
the following [ACI-26.12.5.1 and ACI-26.4.2.2(d)]:
• Residual flexural strength at a midspan deflection of 1/300 of the span length of at

least 90% of the first peak strength.
• Residual flexural strength at a midspan deflection of 1/150 of the span length of at
least 75% of the first peak strength.
• A minimum fiber content of 100 lb/cu yard of concrete.
Residual flexural strength is determined assuming linear elastic, uncracked behavior

(i.e., using an elastic modulus of bh 2 / 6). First peak flexural strength is not to be taken
less than 7.5 fc′.
The minimum required fiber content was determined based on the test results of
fiber-reinforced concrete beams, which indicated a minimum shear failure stress of
3.6 fc′ in beams with at least 100 lb/ cu yard (fiber volume fraction of 0.75%)
of deformed fibers [5.62]. Similarly, the performance criteria based on ASTM 1609 tests
were determined from performance of materials similar to those used in experimental
investigations.
130
Shear failure of fiber-reinforced concrete beam without stirrups. (Photo by Gustavo J. Parra-Montesinos.)
5.10 ACI CODE DESIGN PROVISIONS FOR SHEAR

In the ACI design method for shear, it is required that
φVn ≥ Vu (5.10.1)
where Vu is the factored shear force and φ Vn is the design strength in shear. The strength
reduction factor φ is 0.75 for shear (ACI Table 21.2.1). The nominal shear strength Vn is
Vn = Vc + Vs (5.10.2)
where Vc and Vs are the portions of the shear strength attributable to the concrete (see
Section 5.5) and to the shear reinforcement (see Section 5.8), respectively.
Strength Vc Attributable to Concrete

The development of the detailed equation, Eq. (5.5.6), for Vc was shown in Section 5.5 (see
also Fig. 5.5.1). Since that equation is not easy to use as a design equation, and because of
the wide scatter of test results, ACI-22.5.5.1 permits using either of the following:
(1) For the simplified method,
Vc = 2 λ fc′ bw d (5.10.3)3
where bw is the web width, d is the effective depth, and λ is 0.75 for all-lightweight con-
crete and 0.85 for sand-lightweight concrete (ACI-19.2.4.2) (see Section 5.5.5, subsection
on lightweight concrete and Section 1.8 for more information).
From Fig. 5.5.1 the value obtained using Eq. (5.10.3) appears to be conservative; how-
ever, studies [5.3, 5.59–5.61] have shown otherwise when ρw is below about 0.012. For val-
ues of ρw (that is, As / bw d ) lower than 0.012, the following, more conservative expression,
is suggested [5.59], where the λ factor has been added to account for the use of lightweight
concrete:
Vc = (0.8 + 100ρw )λ fc′ bw d (5.10.4)

Vc = 0.17λ fc′ bw d (5.10.3)
13
5 . 1 0 A C I C ode D esign P ro v isions for S hear 131
(2) For the more detailed method,
 ρ V d
Vc =  1.9λ fc′ + 2500 w u  bw d ≤ 3.5λ fc′ bw d
 Mu  (5.10.5)4
Equation (5.10.5) is identical to Eq. (5.5.6). The value of Vu d / Mu may not exceed 1.0, and
Mu is the factored moment occurring simultaneously with the Vu for which shear strength
is being provided.
ACI–ASCE Committee 426 recommended [5.16] against further use of Eq. (5.10.5) (see
Section 5.5, subsection on continuous beams). The primary practical use of that equation is
and has been to justify slightly larger stirrup spacings in high shear regions where spacings
using Eq. (5.10.3) are small (say, less than 3 in.).
Strength Vs Attributable to Shear Reinforcement

The contribution of shear reinforcement, as developed in Section 5.8, is (ACI-22.5.10.5.4)
Av f yt d
Vs = (sin α + cos α ) (5.10.6)
s
where fyt is the yield strength of the shear reinforcement. When vertical stirrups (i.e., stir-
rups perpendicular to the axis of the member) are used (α = 90°) (ACI-22.5.10.5.3),
Av f yt d
Vs = (5.10.7)
s
Lower and Upper Limits for Amount of Shear Reinforcement

As noted in Section 5.6, the amount of shear reinforcement should be neither too low nor
too high in order to prevent a sudden failure if diagonal cracking develops and to ensure
yielding of the transverse steel (i.e., fv = f yt) when the failure strength in shear is reached.
When shear reinforcement is used, a minimum area Av is required for nonprestressed beams
(ACI-9.6.3.3), slabs (ACI-7.6.3.3), and columns (ACI-10.6.2.2), equal to
bw s 50bw s
min Av = 0.75 fc′ ≥ (5.10.8)5
f yt f yt
in which bw is the member web width and s is the spacing of the shear reinforcement in
inches, f yt is the yield strength of the shear reinforcement in psi, and fc′ is the specified
compressive strength of the concrete in psi.
From Eq. (5.8.7) and using the lower limit of Eq. (5.10.8), this minimum amount
corresponds to
Av f yt d f yt d  bw s 
Vs = =  50  = 50bw d (5.10.9)6
s s  f yt 

 ρ V d
Vc =  0.16 λ fc′ + 17 w u  bw d ≤ 0.29λ fc′ bw d (5.10.5)
 Mu 
5 For SI, ACI 318-14M, for Av in mm2, bw and s in mm, f y and fc′ in MPa, ACI-9.6.3.3 gives
bw s 0.35bw s
min Av = 0.062 fc′ ≥
f yt
(5.10.8)
f yt
6 For SI, ACI 318-14M, for bw and d in mm and Vs in newtons (N), gives
Vs = (0.35MPa ) bw d (5.10.9)
132
or in terms of nominal unit stress on area bw d ,

Vs b d
vs = = 50 w = 50 psi (5.10.10)
bw d bw d
To ensure that the amount of shear reinforcement is not too high, the designer will usually
keep Vs below the following range:
Vs ≤ 6 fc′ bw d to 8 fc′ bw d
ACI-22.5.1.2 gives the upper limit for Vs as 8 fc′ bw d . Although this requirement is
imposed on the shear reinforcement, it is intended to limit the maximum stress in the diag-
onal concrete struts, which may cause a sudden and premature crushing of the web, as well
as to limit the width of diagonal cracks under service loads.
Design Categories and Requirements

For design, the shear envelope for Vu is the starting point. From a practical point of view it is
better to plot the Vu diagram rather than the “required Vn” diagram (equal to Vu / φ ). The basic
diagrams used should be those of Mu and Vu owing to factored load; the φ factor, which is
different for moment and for shear, should not be included in the load-related diagrams.
The design for shear may be separated into the following categories:
1. Vu ≤ 0.5φ Vc (5.10.11)
In this category, no shear reinforcement is required (ACI-9.6.3.1). However, because of the

potential for large members to exhibit shear strengths lower than those obtained from ACI
Code strength equations (see Section 5.5), it is recommended that members with depths
of 24 in. or greater be provided with at least minimum shear reinforcement satisfying
Eq. 5.10.8 (ACI-9.6.3.3).
2. 0.5 φ Vc ≤ Vu ≤ φ Vc (5.10.12)
Minimum shear reinforcement is required in beams (ACI-9.6.3.1) except for
(a) beams where the total depth does not exceed 10 in.

(b) beams with depth less than the greater of 2 1 2 times the flange thickness for
T-shaped sections, or one-half of the web width, but in all cases beam depth shall
not exceed 24 in.
(c) normal-weight fiber-reinforced concrete beams with depth no greater than 24 in.,
fc′ ≤ 6000 psi, and Vu ≤ φ 2 fc′ bw d . The fiber-reinforced concrete used shall satisfy
the performance criteria in ACI-26.12.5.1 and contain at least 100 lb/cu yard of
deformed steel fibers (see Section 5.9).
(d) one-way joist systems.
(e) one-way slabs and footings (ACI-7.6.3.1).
For this category, the shear reinforcement must satisfy ACI-9.6.3.3 and ACI-9.7.6.2.2,
as follows:
required φVs = min φVs = φ 0.75 fc′ bw d ≥ φ (50)bw d (5.10.13)7
and
d
maximum spacing s ≤ ≤ 24 in. (5.10.14)
2
7 For SI, ACI 318-14M, for bw and d in mm and Vs in N, ACI-9.7.6.2.2 gives
( ) (
min φ Vs = φ 0.062 fc′ bw d ≥ φ 0.35 fc′ ) (5.10.13)
13
5 . 1 0 A C I C ode D esign P ro v isions for S hear 133
The requirement of minimum shear reinforcement may be waived if test results show that
the required flexural and shear strengths can be developed (ACI-9.6.3.2).
3. φ Vc < Vu ≤ [φ Vc + min φ Vs ](5.10.15)
For all flexural members, including those exempted from shear reinforcement in Category
2, shear reinforcement must be provided satisfying Eqs. (5.10.13) and (5.10.14).
4. [φ Vc + min φ Vs ] < Vu ≤ [φ Vc + φ (4 fc′ ) bw d ](5.10.16)8
For this category, the computed shear reinforcement requirement will exceed the min φVs
requirement, and the shear reinforcement must satisfy ACI Code Eq. (22.5.10.1), ACI-
22.5.10.5.3 and ACI-9.7.6.2.2, as follows:
required φ Vs = Vu − φ Vc (5.10.17)
φ Av f yt d
provided φVs = (for α = 90° ) (5.10.18)
s
d
maximum s = ≤ 24 in. (5.10.19)
2
Note that in terms of nominal stress, vs = Vs / ( bw d ) = 4 fc′ psi is the maximum vs for which
the d/2 maximum spacing limit applies.
5. [φ Vc + φ (4 fc′ )bw d ] < Vu ≤ [φ Vc + φ (8 fc′ )bw d ](5.10.20)9
The difference between Categories 4 and 5 is that for all regions of a beam where the nomi
nal stress vs to be taken by shear reinforcement is between 4 fc′ and 8 fc′ the maximum
shear reinforcement spacing s may not exceed d /4 nor 12 in. These reduced spacing limits
are intended to provide adequate diagonal cracking control at high shear stresses, which
helps improve aggregate interlock and dowel action. The shear reinforcement provided
in this category must satisfy ACI Code Formula (22.5.10.1), ACI-22.5.10.5.3 and ACI-
9.7.6.2.2, as follows
required φ Vs = Vu − φ Vc (5.10.21)
φ Av f yt d
provided φ Vs = (for α = 90°) (5.10.22)
s
d
maximum spacing s ≤ ≤ 12 in. (5.10.23)
4
In addition, factored shear Vu may not exceed the upper limit of Eq. (5.10.20) according to
ACI-22.5.1.2.
The ACI Code provisions for shear strength of beams described in this section are sum-
marized in Table 5.10.1.
8 For SI, ACI 318-14M gives in place of 4 fc′ psi , fc′ / 3 when fc′ is in MPa.
9 For SI, ACI 318-14M gives in place of 4 fc′ and 8 fc′ psi, fc′ /3 and 2 fc′ /3 , respectively, when fc′ is in MPa.
134
TABLE 5.10.1 SHEAR STRENGTH OF MEMBERS UNDER BENDING ONLY—ACI CODE
Item Strength Design (φ = 0.75) Code
φ Vn ≥ Vu 9.5.1.1, 9.4.3.2
1
Maximum Vu at a distance d from face of support in usual
situations (three exceptions; see Section 5.11)
2 Vn = Vc + Vs Formula (22.5.1.1)
3 fc′ ≤ 100 psi 22.5.3.1

bw s 50bw s
unless Av ≥ 0.75 fc′ ≥
f yt f yt
4 Simplified method:
Vc = 2 λ fc′ bw d Formula (22.5.5.1)
More detailed method:

 ρ V d
 Mu 
Vu d/Mu not to exceed unity Table 22.5.5.1
Allow 10% increase for one-way joists 9.8.1.5

Lightweight concrete factor λ when fct is specified:
fct 19.2.4.3
λ= ≤ 1.0
6.7 fcm
Lightweight concrete factor λ when fct is not specified: 19.2.4.2
Use λ between 0.75 and 0.85 depending on the absolute
volume of normal-weight fine aggregates as a fraction of the
total absolute volume of fine aggregates. For sand-lightweight,
coarse blend concrete, λ ranges between 0.85 and 1.0 based on
the absolute volume of normal-weight coarse aggregate as a
fraction of the total absolute volume of coarse aggregate.
5 Av f yt d (sin α + cos α )
Vs = Formula (22.5.10.5.4)
s
Vs = Av f y sinα ≤ 3 fc′ bw d
(single bar or single group of 22.5.10.6.2
bent-up bars)
f yt not to exceed 60,000 psi, except for welded deformed wire Table 20.2.2.4a
reinforcement not to exceed 80,000 psi
Vs not to exceed 8 fc′ bw d 22.5.1.2
6 For 0.5 φ Vc < Vu ≤ φ Vc
min φ Vs = 0.75 φ fc′ bw d ≥ φ 50 bw d 9.6.3.1, 9.6.3.3
Av f yt Av f yt d
use s = ≤ (max s ≤ ≤ 24 in.)
50bw 0.75 fc′ bw 2
except for slabs; footings; joists; small beams shallower than
10 in., 2 1 2 times flange thickness, or bw / 2, but with depth no
greater than 24 in.; steel fiber-reinforced, normal-weight concrete
beams with depth no greater than 24 in., fc′ ≤ 6000 psi, and
Vu ≤ φ 2 fc′ bw d
(Continued)
135
5 . 1 1 N O M I NA L S H E A R S T R E N G T H CA L C U L AT I O N 135
TABLE 5.10.1 (Continued)
Item Strength Design (φ = 0.75) Code
7 For φ Vc < Vu ≤ [φ Vc + minφ Vs ]

9.6.3.1,
min φ Vs = φ 0.75 fc′ bw d ≥ φ 50 bw d 9.6.3.3
Av f yt Av f yt  d 
use s = ≤  max s ≤ ≤ 24 in.
50bw 0.75 fc′ bw  2 
8
 (
For [ φ Vc + min φ Vs ] < Vu ≤ φ Vc + φ 4 fc′ bw d 
 )
Design shear reinforcement. For α = 90°, Formula (22.5.10.1),
Formula (22.5.10.5.3),
Av f yt d Vu
Vs = ≥ − Vc 9.7.6.2.2
s φ
d
max s = ≤ 24 in.
2
9 For [φVc + φ (4 fc′ bw d)] < Vu ≤ [φVc + φ (8 fc′ bw d)] Formula (22.5.10.1),
Formula (22.5.10.5.3),
Design shear reinforcement. For α = 90°, 9.7.6.2.2
Av f yt d Vu
Vs = ≥ − Vc
s φ
d
max s = ≤ 12 in.
4
5.11 CRITICAL SECTION FOR NOMINAL SHEAR

STRENGTH CALCULATION
In experimental work the critical section for computing the nominal shear strength was the
location of the first inclined crack. Since most testing was made on simply supported beams
under simple loading arrangements, it was difficult to extend such results to generalized
loadings on continuous structures.
For nonprestressed beams subjected to distributed loading near their top surface and
supported at the bottom such that the support reaction in the direction of shear introduces
compression, ACI-9.4.3.2 permits taking the first critical section at a distance d from the
face of support. Thus the Code recognizes that the portion of the distributed loading applied
over a distance d from the face of the support will transfer directly to the support, without
requiring additional shear reinforcement. Shear reinforcement must be provided, however,
between the face of support and the distance d therefrom, using the same requirements as
at the critical section.
The critical section must be taken at the face of support when one of the following
occurs:
1. Factored shear Vu does gradually decrease from the face of support, but the support
is itself a beam or girder and therefore does not introduce compression into the end
region of the member (see Fereig and Smith [5.58]).
2. Loads are not applied at or near the top of beam.
3. A concentrated load occurs between the face of support and a section at a distance d
therefrom.
4. Any loading that may cause a potential inclined crack to occur at the face of support
or extend into instead of away from the support (see Fig. 5.11.1).
136
Figure 5.11.1 Inclined crack when the reaction induces tension in the member. Critical
section is at the face of support.
5.12 SHEAR STRENGTH OF BEAMS—D ESIGN EXAMPLES

Three examples are presented to illustrate the basic procedure of designing vertical stirrups.
In the first example, the complete design procedure for flexure and shear is shown for a
simply supported beam using the simplified procedure. Some of the more practical aspects
are discussed in the second example. The third example illustrates the use of metric units.
Design of shear reinforcement in the continuous spans of a slab-beam-girder floor system
is treated in Chapter 9.
EXAMPLE 5.12.1
For the given beam [Fig. 5.12.1], first determine the maximum uniform dead and live
loads under service condition permitted by the ACI Code. Then use those maximum
service loads to design the shear reinforcement using vertical stirrups and the simplified
procedure using constant Vc . Assume that the service live load to dead load ratio is 1.5,
fc′ = 4000 psi (normal-weight concrete), and f yt = 60, 000 psi.
SOLUTION
(a) Compute the nominal moment strength M n of the given section. Assume the com-
pression steel yields at nominal strength M n. Then from statics,
Cc = 0.85 fc′ ba = 0.85(4)(14)a = 47.6 a
Cs = ( f y − 0.85 fc′ ) As′ = (60 − 3.4)2.40 = 136 kips
T = f y As = 60(8.0) = 480 kips

Cc + C s = T
(480 − 136) a 7.23

a= = 7.23 in.; c= = = 8.51 in.
47.6 β1 0.85
Check to see whether compression steel yields; compute strain ε s′ at the location of the
compression steel,
c − d′ 8.51 − 2.50
ε s′ = ε cu = 0.003 = 0.00212
c 8.51
The compression steel yields as assumed because

 fy 60 
[ ε s′ = 0.00212] >  ε y = = = 0.00207
 E s 29, 000 
(Continued)
137
5 . 1 2 S hear S trength of B ea m s — D esign E xa m ples 137
4 – #7 throughout
8 – #9 throughout
23’–0” clear
24’–0” center to center
2.5” As’ = 2.40 sq in. εcu = 0.003
εs’
4 – #7
26” d = 22.5” dt c = 8.51 in.
8 – #9
εs
bw = 14” As = 8.00 sq in.
(a) Cross section (b) Strain diagram at Mn
Figure 5.12.1 Beam for Example 5.12.1.
Since there is no axial load, check that the net tensile strain is at least 0.004 (ACI-
9.3.3.1). To obtain dt to the layer of steel reinforcement closest to the tension face, add
0.5 in. (assuming 1 in. between layers) plus the bar radius (0.564 in. for a #9 bar) to the
effective depth d. Thus,
dt = d + 0.5 + 0.564 = 22.5 + 0.5 + 0.564 = 23.6 in.
The strain εt at the extreme tension steel is

dt − c 23.6 − 8.51
ε t = 0.003 = 0.003 = 0.0053 > 0.004 OK
c 8.51
Since the tensile strain exceeds 0.005, the section is tension controlled (ACI-21.2.2),
and the appropriate φ factor is 0.90.
The nominal moment strength M n is
M n = Cc ( d − a / 2 ) + C s ( d − d ′ )
1 1
= 47.6(7.23)(22.5 − 7.23 / 2) + 136(22.5 − 2.5)
12 12

= 542 + 227 = 769 ft--kips
(b) Compute the maximum permitted service loads wD and wL using the U and φ factors
of ACI-5.3 and ACI-21.2.2.
1
Mu = wu L2 = φ M n
8
8φ M n 8(0.90)769
wu = = = 9.61 kips/ft
L2 (24)2
wL = 1.5 wD
wu = 1.2 wD + 1.6 (1.5wD )
9.61
service dead load wD = = 2.67 kips//ft
1.2 + 2.4
service live load wL = 4.00 kips/ft
(Continued)
138
(c) Design of shear reinforcement using the simplified method with a constant value for
Vc . The factored shear to be designed for must be the maximum that may possibly
act at each point along the span; that is, an envelope of maximum shear is needed.
A knowledge of influence lines tells the designer that for bending moment on a
simply supported span, the maximum value occurs at every point along the span
when the full span is loaded with live load. However, for shear, the maximum shear
occurs with partial span loading for every point along the span except at the sup-
ports. Unless the designer can justify other treatment, the live load should always
be treated as variable position loading acting wherever it may cause the greatest
effect, whereas the dead load would be a fixed position loading. For most ordinary
situations an approximate shear envelope may be acceptable, using a straight-line
relationship between the maximum shear at the support and the maximum shear at
midspan. Such a procedure will always be conservative.
CL of support
True envelope for maximum Vu
115.2 Straightline approximation for

Vu = 1.2VD + 1.6 VL’ kips
maximum shear envelope Vu

due to factored loads
Shear diagram for Vu
+ 19.2 with full dead plus live
load on entire span
19.2
–
Midspan
115.2
24’–0”
Figure 5.12.2 Factored shear Vu diagram for Example 5.12.1.
For this beam (see Fig. 5.12.2), the maximum factored shear at the centerline of support is
wu L [2.67(1.2) + 4.00 (1.6)] 24

Vu = = = 115.2 kips
2 2
The maximum factored shear possible at midspan occurs with live load on half the span,
wL L 4.00(1.6)(24)
Vu = = = 19.2 kips
8 8
The dead load shear (with full span loaded) is zero at midspan.
The critical section for determining the closest stirrup spacing may be taken, accord-
ing to ACI-9.4.3.2, at a distance d from the face of support; in this case, d = 22.5 in. and
the support width is 12 in., making the critical section (22.5 + 6 = 28.5 in.) 2.38 ft from
the center of the support. By linear interpolation, Vu at d from the face of support is
 115.2 − 19.2 
Vu = 115.2 −   2.38 = 96.2 kips
 12
The design requirement between the face of support and the critical section is consid-
ered to be constant; in this case, 96.2 kips.
(Continued)
139
The factored shear Vu diagram for the left half of this symmetrical structure is shown
enlarged in Fig. 5.12.3. The design may be done primarily on this factored shear Vu
diagram.
The design shear strength attributable to the concrete is
(
φ Vc = φ 2 λ fc′ bw d )
 
φ Vc = 0.75[2(1) 4000 ](14)(22.5)  1  = 29.99 kips
 1000 
The difference (the portion cross-hatched in Fig. 5.12.3) between Vu and φ Vc must be

provided for by shear reinforcement. For most beams (ACI-9.6.3.1) shear reinforce-
ment may not be terminated when Vu equals φ Vc, but rather must be continued until
Vu = 0.5φ Vc (Fig. 5.12.3). The distance over which shear reinforcement is required from
the face of support is in this case the entire span because 0.5φ Vc never exceeds Vu .
11 @ 3” 7 @ 4” 4 @ 6” 5 @ 10” # 3 stirrups
12”
3”
12’ –0“
Face of support
115.2k Critical section
Vu = 1.2VD + 1.6 VL, kips
96.2k
φVn = 85.6k for s = 4”
d s = 5”
s = 6”
2.38’ s = 8”
s = 10”
s = 11.25”
29.9k
19.2k
φVc 15.0k
0.5φVc
4” 5” 6” 8” 10” 11.25”
CL of support Midspan
Figure 5.12.3 Design of shear reinforcement in Example 5.12.1 using constant φ Vc.
Examination of Fig. 5.12.3 shows that Vu exceeds φ Vc over most of the span, putting
most of the design in Categories 3 through 5 of Section 5.10. At the critical section,
required φ Vs = Vu − φ Vc = 96.2 − 29.9 = 66.3 kips
Check limits on φ Vs ,
 1 
min φ Vs (ACI − 9.6.3.3) = 0.75(0.75 4000 )14(22.5)  = 11.2 kips
 1000 
but not less than
1
0.75(50)14(22.5) = 11.8 kips
1000
(Continued)
140
Thus, the 11.8 kip limit controls.

 d   1 
max φ Vs  for s = ≤ 24 in. = 0.75(4 4000 )14(22.5) 
 2   1000 
= 59.8 kips
(ACI-9.7.6.2.2)
 d   1 
max φ Vs  for s = ≤ 12 in. = 0.75(8 4000 )14(22.5) 
 4   1000 
= 119.5 kips
(ACI-9.7.6.2 .2)
Since the required φ Vs equals 66.3 kips at the first critical section (which is the maximum
for the beam) and lies between 59.8 and 119.5 kips, design for the portion of the beam
between the face of support and the critical section falls into Category 5 (max s = d /4).
Using #3 vertical U stirrups, in which two stirrup bar areas comprise Av in Eq. (5.10.18),
φ Av f yt d 0.75(2)(0.11)(60)(22.5)
s (at critical section) = = = 3.4 in.
φ Vs 66.3
This means that #3 stirrups may not be spaced farther apart than 3.4 in. for the
region from the face of support to the critical section at a distance d from the face of
support.
The determination of max s at the critical section usually controls the size of stirrups
to be used. If the spacing for #3 stirrups were required to be too small, say less than
about 3 in., the stirrup bar size would be increased to #4, or occasionally #5. In this case,
a spacing of 3 in. is considered acceptable.
As the Vu requirement decreases toward midspan, the stirrup spacing can be increased.
A table of potentially acceptable spacings (s vs. φ Vs ) should be made, as in the first two
columns of Table 5.12.1. Values are chosen until the spacing reaches the maximum per-
mitted. Since φ Vs does not exceed 4 fc′ bw d beyond a distance of 3.19 ft (38.25 in.) from
the center of the support, the maximum spacing beyond this section is the lesser of d /2
(11.25 in. for this beam) or 24 in. (ACI-9.7.6.2.2). In the region where a minimum area
TABLE 5.12.1 SPACING AND STRENGTH RELATIONSHIPS

FOR VERTICAL STIRRUPS IN EXAMPLE 5.12.1
z = Distance from face of

support to intersection of
φ Vn (i.e., φ Vc + φ Vs ) with Vu
φ Av f yt d 222.8 66.3 − φ Vs
φ Vs = = = 22.5 + (144 − 28.5)
s s (1000) s 96.2 − 19.2
3.4 in. 66.3 kips (max) 0 to 22.5 in.
4 55.7 38
5 44.6 55
6 37.1 66
8 27.9 80
10 22.3 89
11.25 max 19.8 92–138 in.
(Continued)
14
of shear reinforcement is required (ACI-9.6.3.3), the spacing required for #3 stirrups is

the smaller of
0.22(60, 000) 0.22(60, 000)
s= = 19.9 in. or s= = 18.9 in.
0.75 4000 (14) 50(14)
Therefore, the maximum spacing of 11.25 in. based on beam depth controls.
In Table 5.12.1, the d /2 spacing limit results in a shear strength of 49.7 kips (φ Vc + φ Vs)
at a distance z = 92 in. from the face of the support. This means that between
z = 92 in. and midspan, the spacing of the stirrups cannot be increased even though Vu
continues to decrease. Thus, a stirrup spacing of 11.25 in. (or less) should be used for this
region of the span until stirrups are no longer required. As noted earlier, however, Vu is always
larger than φ Vc /2 and, therefore, stirrups are required over the entire span in this case.
To determine the set of spacings to actually be used, the designer may prefer to
draw on the Vu diagram the lines representing the φ Vn provided by different spacings, as
shown in Fig. 5.12.3, and then scale from the diagram the locations z where the various
spacings are permissible (marked along the baseline of Fig. 5.12.3); or, alternatively, the
locations z may be computed as in the third column of Table 5.12.1. Note that in this
column, the quantity (144 –28.5)/(96.2 –19.2) is the distance in which the value of Vu
drops 1 kip. Whether the location z is computed or scaled, a table of φ Vs versus s should
be used. The limiting spacings of 3.4 and 11.25 have already been explained. The inter-
mediate spacings of 4, 5, 6, 8, and 10 in. are those chosen by the designer as practical
possibilities.
Since the spacing of stirrups cannot be varied continuously, they are varied by using
discrete increments. One conservative policy is to use a spacing, say, 6 in., only beyond
the theoretical point at which a 6-in. spacing may be used (in this case, 66 in. or more
from the face of support). In a less conservative manner, the next larger spacing may be
used somewhat before the point at which this spacing may be used. With this in mind,
the set of spacings to be used is
3 in. | 11 @ 3 in. | 7 @ 4 in. | 4 @ 6 in. | 5 @ 10 in. | = 138 in.
3 36 64 88 138 (midspan)
To make uniform spacing at midspan, the first stirrup is placed at 3 in. from the sup-
port, more than the usual half-space. Many designers prefer to place the first stirrup a
full space from the face of support for closely spaced stirrups. Note that in the adopted
set of spacings, 4-in. spacing is used after z = 36 in. (z = 38 in. required); 6-in. spacing
is used after z = 64 in. (z = 66 in. required); and 10-in. spacing is used after z = 88 in.
(z = 89 in. required).
EXAMPLE 5.12.2
Design the locations of #3 vertical U stirrups to be used in the beam of Fig. 5.12.4.
The beam is to carry service dead and live loads of 5.2 and 6.0 kips/ft, respectively.
Use fc′ = 3500 psi (normal-weight concrete) and f y = f yt = 60, 000 psi. Use the alternate
approach of plotting vn (vn is the nominal shear stress, which is the required nominal
shear strengthVu / φ divided by bw d ) diagram and then providing the vc attributable to the
concrete and vs attributable to shear reinforcement.
(Continued)
142
SOLUTION
(a) The factored shear force Vu and required nominal shear stress vn diagrams (envel-
ope of maximum values for different loading conditions) for design are as shown in
Fig. 5.12.4. The effect of partial span live load on shear is approximated by passing
a straight line through the maximum shear values at the support and at midspan.
1.2 wD = 1.2(5.2) = 6.24 kips/ ft
1.6 w L = 1.6(6.0) = 9.60 kips/ft
At the center of the support,

1
Vu = (6.24 + 9.60)(12) = 95 kips
2
Setting vn = vu / φ
Vu 95, 000
vn = = = 490 psi
φ bw d 0.75(12)21.5
At midspan,
1
Vu = (9.60)(12) = 14.4 kips
8
Vu 14, 000
vn = = = 74.4 psi
φ bw d 0.75(12)(21.5)
Note that the shear at midspan due to dead load is zero; but positive live load shear at
midspan is largest when only the right half of the span is loaded.
(b) Simplified method.
vc = 2 λ fc′ = 2(1) 3500 = 118 psi
The required nominal shear stress at the distance d from the face of the support is
(6 + 21.5)
vn (at d ) = 490 − (490 − 74.4) = 331 psi
72
required vs = vn − vc = 3311 − 118 = 213 psi
4 fc′ = 237 psi (ACI − 9.7.6.2.2)
Where vs > 4 fc′ , the maximum spacing of vertical stirrups may not exceed d /4;
otherwise, d /2 (ACI-9.7.6.2.2). In this case, the spacing limitation d /2 applies wherever
stirrups are needed. Note that this simple check on whether required vs exceeds 4 fc′
often gives advantages over using the criterion shown in item 9 of Table 5.10.1.
Try #3 vertical U stirrups.
Av f yt = 2(0.11)(60, 000) = vs bw s
13, 200
vs s = = 1100 psi. in.
12
which for the vs of 213 psi at the critical section permits

1100
max s (at critical section) = = 5.2 in.
213
(Continued)
143
d = 21.5”
4 – #9
2” #3
5 @ 5” 3 @ 9” 12”
2 @ 6” U stirrups
12”
12’ –0” center to center of supports

Symmetrical
(a) The beam about CL of
beam
95
Critical section
Straightline approximation
Vu
True maximum shear envelope
shear force,
for uniform live load
kips
d = 21.5”
14.4
Face of support
12”
(b) Shear force diagram
490
d = 21.5” Shaded portion indicates
stress to be carried by
331
Vu s = 6” shear reinforcement
s = 8”
φbwd s = 10”
shear stress, Required vs
psi
74.4
vc/2
vc = 118 psi carried by concrete
(c) Required shear stress vn diagram
Max. s (d/2) = 10.75”
6” 8” 10”
12
Face of support
s
stirrup 8 Permitted
spacing, Provided
inches 4 #3 stirrups
5 @ 5 = 2’ –1” 2 @ 6”
2” 3 @ 9”
(d) Stirrup requirement diagram
Figure 5.12.4 Stirrup placement for Example 5.12.2.
The maximum spacing for shear reinforcement to provide for a minimum vs of 50 psi or
0.75 fc′, whichever is greater, is
1100
max s (for min Av ) = = 24.8 in.
.
0 75 3500
1100 d 
but not less than = 22 in. >  = 10.75 in.
50 2 
Thus the spacing cannot be more than 10.75 in. even when vs gets small. If the above
computation had given a value less than d /2 instead of 22 in., then that lesser value
would be the maximum s for the region wherever stirrups are required (vn > vc /2).
(Continued)
14
The placement of stirrups will be done by scaling from the diagram of Fig. 5.12.4(c).
To facilitate this, Table 5.12.2 is computed using spacings considered desirable by the
designer. The shear stress vs values are plotted as horizontal lines marked s = 6, 8, and
10 in. on Fig. 5.12.4(c); then their intersections with the required shear stress vn line are
projected downward to the base line. Stirrups are then laid out with a scale beginning a
distance s /2 from the face of support. (Some designers place the first stirrup a full space
from the face of support for small spacings.) The same spacing necessary at the critical
section is specified by ACI-9.4.3.2 to be used between the face of support and the crit-
ical section. Thus one may start at s /2 from the face of the support with a 5-in. spacing
until within s /2 of the capacity line for the next desired spacing, which in this case is
slightly beyond the vertical line projected from s = 6 in. The ACI Code requires shear
reinforcement until the stress vn ≤ vc / 2. In this case, the entire beam must be provided
with stirrups, because the smallest vn ( 74.4 psi ) is larger than vc /2 (59 psi).
To illustrate clearly what has been done, a diagram showing permitted (actual spacings
must be below permitted line) and provided stirrup spacings is given in Fig. 5.12.4(d).
In this case, some shifting of spacing has been made to avoid leaving a small fragmental
space at midspan. One more space at 5 in. was used than necessary and the three spaces
adjacent to midspan were reduced to 9 in. from the permitted 10 in. in order to eliminate
a small space near midspan.
The final design details are shown in Fig. 5.12.4(a).
TABLE 5.12.2 SPACING OF VERTICAL STIRRUPS (EXAMPLE 5.12.2)
1100
vs =
s s vu /φ = vn = vc + vs
5.2 213 331

6 183 301
8 138 256
10 110 228
max 10.75 102 220
EXAMPLE 5.12.3
Determine the vertical stirrup requirement for the beam of Fig. 5.12.5. Use #10M
bars (see Table 1.13.2) for U stirrups, fc′ = 25 MPa (normal-weight concrete), and
f y = f yt = 300 MPa. The service live load is 30 kN/m and the service dead load is 43 kN/m
(including beam weight).
SOLUTION
(a) Determine maximum factored shear Vu envelope.
wu = 1.2(43) + 1.6(30) = 99.6 kN/ m
(Continued)
145
For maximum shear at d (480 mm) from face of support, place live load on remainder
(3.2 − 0.15 − 0.48 = 2.57 m) of the span.
(2.57)2
Vu at critical section =1.2(43)(1.6 − 0.15 − 0.48)+1.6(30)
2(3.2)
= 50. 0 + 49. 6 = 99.6 kN

1
Vu at midspan = (1.6)(30)(3.2) = 19.2 kN
8
The straight-line approximation maximum shear envelope is given as Fig. 5.12.5(c).

(Note that, in the previous example, the straightline approximation was made between
centerline of support and midspan. The present approximation is closer to the “exact”
maximum shear curve.)
(b) Determine stirrup spacing. If the factored shear Vu diagram is used,
fc′
vc = (from ACI 318 − 14M with fc′ in MPa)
6
= 25 / 6 = 0.833 MPa
1
φ Vc = φ vc bw d = 0.75(0.833)(280)(480) = 84.0 kN
1000
At the critical section,
required φ Vs = Vu − φ Vc = 99.6 − 84.0 = 15.6 kN
The limiting φ Vs = φ ( fc′ / 3)bw d for d /2 stirrup spacing limitation can be conveniently
obtained from 2φVc ; thus
limiting φ Vs = 2φ Vc = 2 (84.0 ) = 168 kN > [required φ Vs = 15.6 kN ]
Thus maximum spacing cannot exceed d /2.

For #10M U stirrups, Av = 2 ( 71) = 142 mm 2 . The spacing requirement for strength is
φ Av f yt d (0.75)(142)(300)480 1 15, 228
s= = =
φ Vs φ Vs 1000 φ Vs (kN)
The stirrup spacing requirements can be summarized as follows:

1. Maximum spacing for strength requirement at the critical section,
15, 228
max s = = 976 mm
15.6
2. Maximum spacing d /2,

d 480
max s = = = 240 mm Controls
2 2
(Continued)
146
280 mm
Support width = 300 mm
#10M
stirrup
3.2 m
d = 480 mm
(a)
4 – #19M
40 mm cover
Support width = 300 mm
(b)
Face of support
Stirrups required = 1.17 m
99.6
Factored shear Vu, kN
84.0
Midspan
d
φVc 42.0
φVc
2 19.2
4 @ 240 mm #10M
120 mm
U stirrups
(c)
Figure 5.12.5 Beam and stirrup design for Example 5.12.3.
3. Maximum spacing for minimum shear reinforcement (ACI-9.6.3.3),
(
min φ Vs = φ ( 0.35 MPa ) bw d > φ 0.062 fc′ bw d (ACI 318-14M) )
1
= 0.75 (0.35)(280)(480) = 35.3 kN
1000
15, 228 d
max s = = 431 mm >
35.3 2
4. Conclusion. Use 240-mm spacing for the portion where stirrups are required (i.e.,
from face of support to location where Vu = 0.5φ Vc).
The final stirrup arrangement is shown in Fig. 5.12.5(c).
5.13 SHEAR STRENGTH OF MEMBERS UNDER

COMBINED BENDING AND AXIAL LOAD
The presence of an axial compressive load on a reinforced concrete flexural member
decreases the longitudinal tensile stress and the resulting tendency for inclined cracking.
Conversely, the addition of an axial tensile load increases the longitudinal tensile stress and
the tendency for inclined cracking. Thus, for the same bending moment, the shear strength
of a member is increased by the addition of an axial compressive load and decreased by an
147
5.13 SHEAR STRENGTH OF MEMBERS 147
axial tensile load. Some experimental work is available on shear strength in the presence of
axial load [5.4, 5.63–5.67].
Axial Compression
Since inclined cracking is dependent on the combination of tensile or compressive stress
due to flexure and shear stress, as discussed in Sections 5.3 through 5.5, the addition of
axial compression tends to delay the formation and opening of inclined cracks and limit
their propagation due to an increase in the size of the compression zone.
When the simplified method is used with axial compression, a linear increase with axial
compression is given by ACI-22.5.6.1:
 Nu 

Vc = 2  1 +  λ fc′bw d (5.13.1)10
 2000 Ag 
Note that Nu is the factored axial compressive load (positive quantity), Ag is the gross area
of the concrete cross section, and N u / Ag is expressed in psi.
A more detailed equation for Vc is also permitted by 22.5.6.1,
 
 Vu d  Nu
Vc =  1.9λ fc′ + 2500ρw  bw d ≤ 3.5λ fc′bw d 1 + (5.13.2)11
( 4h − d ) 500 Ag
 Mu − N u 
 8 
( 4h − d )
This expression, however, is not applicable when Mu − N u ≤ 0.
8
( 4h − d )
The rationale for the use of the term Mu − N u in lieu of Mu in Eq. (5.10.5) may
8
be seen by referring to Fig. 5.13.1, which shows a free body of a short length of beam dz.
Moment equilibrium about point A, the line of action of the internal compressive force
C, gives
 a  h a
T  d −  = Mu − N u  −  (5.13.3)
 2  2 2
When the moment arm (d – a /2) is approximated as 7d /8, thereby making a /2 = d /8, the
( 4h − d )
right side of Eq. (5.13.3) becomes Mu − N u , and the left side represents an equiva-
8
lent moment. This equivalent moment procedure is to be used in the more detailed method
for combined axial compression and bending only. Experience [5.63] has shown the method
to be unsafe for axial tension.
The upper limit for the nominal shear strength Vc of members without shear reinforce-
ment subject to bending only is given in Eq. (5.10.5) as 3.5λ fc′ bw d . In the presence of
Nu
axial compression, however, this upper limit is increased by a factor 1 + , whose
500 Ag
rationale is discussed next.
10 For SI, ACI 318-14M, for Nu /Ag and fc′ in MPa, ACI-22.5.6.1 gives
 N 
Vc = 0.17  1 + u  λ fc′ bw d
14 Ag 
(5.13.1)

11 For SI, ACI 318-14M, with Nu /Ag and fc′ in MPa, gives
 
 Vu d  Nu
Vc  0.16λ f c′ + 17ρw  bw d ≤ 0.29λ f c′ bw d 1 +
3.5 Ag
(5.13.2)
 M u − Nu
( 4 h − d ) 
 8 

148
a
2
h A C
Mu
2
h d– a
2
Nu
dz
Figure 5.13.1 Actions and internal forces in a member under combined axial compression and
bending moment.
Nu/Ag, MPa
5 2.5 0 –2.5 Curve (a): ACI Table 22.5.6.1(a)
Eq. (5.13.2)
[Mu/(Vud) = 2, ρw = 0.03,
and fc’ = 2500 psi]
6 Vc Curve (b): ACI Table 22.5.6.1(a)
(d)
Eq. (5.13.2)
bwd fc’
5 [Mu/(Vud) = 5, ρw = 0.005,
(a) and fc’ = 5000 psi]
Curve (c): ACI Eq. (22.5.6.1)
4 Eq. (5.13.1)
Curve (d): ACI Table 22.5.6.1(b)
(c) Eq. (5.13.5)
3
Curve (e): ACI Eq. 22.5.7.1
Eq. (5.13.6)
2 Vc = Vu
(b) d = 0.9h
1
(e)
Compression Tension
1000 500 0 –500
Nu/Ag, psi
Figure 5.13.2 Effect of axial load on ACI-calculated inclined cracking shear stress. Lightweight
concrete factor λ = 1. (Adapted from MacGregor and Hanson [5.63])
The formula for the principal tensile stress ft ( max ) in terms of the tensile stress ft and
the shear stress v was derived in Section 5.3 as
2
1 1 
ft (max) = ft +  ft  + v 2
2 2 
Solving this equation for v,
ft
v = ft (max) 1 − (5.13.4)
ft (max)
It may be seen from Eq. (5.13.4) that the shear strength of a member under bending only
becomes a constant if ft is zero, that is, if the bending moment approaches zero. Empirically,
this constant is the upper limit 3.5λ fc′. For a member under an axial compressive load N u
without bending moment, ft is a constant that is equal to – N u / Ag . Substituting this value of ft in
Eq. (5.13.4) and using an average value of 500 psi for ft ( max ) inside the square root, the upper
limit of strength Vc in members under combined bending and axial compression becomes
Nu
Vc (upper limit) = vbw d = 3.5λ fc′ bw d 1 + (5.13.5)
500 Ag
where N u / Ag is to be expressed in psi (see Fig. 5.13.2).

149
5.13 SHEAR STRENGTH OF MEMBERS 149
Axial Tension
When a flexural member is subject to axial tension, a simple linear reduction for Vc is speci-
fied in ACI-22.5.7.1 as follows:
 Nu 

Vc = 2  1 +  λ fc′ bw d (5.13.6)12
 500 Ag 
Note that N u is negative for tension. It is recommended, however, that Vc = 0 for members
in which the axial tensile load is considered significant.
Figure 5.13.2 shows the relationship between ACI formulas and experimental results.
The cross-hatched portion on the compression side represents a reasonable range when the
more detailed procedure is used.
Parts of the ACI Code are summarized in Table 5.13.1 so that the shear strength of mem-
bers under bending only may be compared with that of members under combined bending
and axial load.
TABLE 5.13.1 EFFECT OF AXIAL LOAD ON THE SHEAR STRENGTH OF MEMBERS

WITHOUT SHEAR REINFORCEMENT—ACI CODE
Simplified Method More Detailed Method
Bending Formula (22.5.5.1) Table 22.5.5.1

only
Vc = 2 λ fc′bw d  ρ V d
 Mu 
Vu d / Mu not to exceed unity
Bending and 22.5.6.1 Table 22.5.6.1

axial
compression  Nu   
Vc = 2 1 +  λ fc′ bw d  Vu d 
 2000 Ag  Vc =  1.9λ fc′ + 2500ρw b d
( 4h − d )  w
 Mu − N u 
 8 
Nu
≤ 3.5λ fc′ bw d 1 +
500 Ag
Nu is positive for compression
and Nu / Ag is in psi Nu is positive for compression and Nu / Ag is in psi
Bending and Although not required, it is 22.5.7.1
axial tension recommended to take Vc = 0 if
axial tension is considered  Nu 
significant. Vc = 2  1 +  λ fc′bw d
 500 Ag 
N u is negative for tension and N u / Ag is in psi
12 For SI, ACI 318-14M, for N u / Ag and fc′ in MPa, ACI-22.5.7.1 gives
 Nu 
Vc = 0.17  1 +  λ fc′ bw d
3.5 Ag 
(5.13.6)

150
Example 5.13.1
Show the effect of axial load on the ACI shear strength Vc for the beam of Example 5.12.1
(Fig. 5.12.1) when it contains no shear reinforcement. Compute Vc for the critical section
at d from the face of support. As in Example 5.12.1, use normal-weight concrete with
fc′ = 4000 psi, b = 14 in, d = 22.5 in, h = 26 in., and 8–#9 bars for the tension steel.
SOLUTION
(a) Axial compression. Using the simplified method, Eq. (5.13.1) or ACI-22.5.6.1
 Nu 
Vc = 2  1 +  λ fc′ bw d
 2000 Ag 
Vc  Nu 
= 2 1 + λ
fc′ bw d  2000 Ag 
The right side of the above equation is 2 for N u / Ag = 0; it varies linearly to 2.8 for
N u / Ag = 800 psi. This linear expression is plotted on the left side of the vertical axis in
Fig. 5.13.3.
Using the more detailed equation, Eq. (5.13.2) or ACI Table 22.5.6.1,
As 8(1.00)
ρw = = = 0.0254
bw d 14(22.5)
Vc 2500(0.0254) Vu d Nu
= 1.9(1.0) + ≤ 3.5(1.0) 1 +
fc′ bw d 4000 [ 4(26) − 22.5] 500 Ag
Mu − N u
8(12, 000)
where Mu is in ft-kips and N u is in lb.
vc
Vud fc’
vc = 1.9 fc’ + 2500ρw 6.0
Mm
Mm = Mu – Nu
8(
4h–d
) 5.0
vc = 3.5 fc’ 1+ 0.002Nu /Ag
4.0
3.0
2.0 vc = 2(1+ 0.002Nu /Ag) fc’

vc = 2(1+ 0.0005Nu/Ag) fc’
1.0
800 600 400 200 100 200 300 400 500

Compression (Nu = +), psi Tension (Nu = –), psi
Figure 5.13.3 Shear strength variation with axial load: Example 5.13.1, with fc′ = 4000 psi
(normal-weight concrete), ρw = 0.0254, Mu = 247 ft-kips, Vu = 96.2 kips, d = 22.5 in., h = 26 in.,
and Ag = 364 sq in. Lightweight concrete factor λ = 1 (not shown in equations). (Note that
vc = Vc / bwd.)
(Continued)
15
5 . 1 4 D eep B ea m s 151
At the critical section, considering live load applied to the entire span, Vu = 92.5 kips and
wu 9.61
Mu = (2.38)(24 − 2.38) = (2.38)21.62 = 247 ft-kips
2 2
Thus,
Vc 92.5 (1.88) Nu
= 1.9 + ≤ 3.5(1.0) 1 +
fc′bw d 247 − 0.00085 N u 500 Ag
where Ag = (14 )( 26 ) = 364 sq in.

Note that Vu and Mu at the critical section must correspond to the same loading condi-
tion, rather than the maximum value for each.
Values for the detailed formula as well as the upper limit equation, Eq. (5.13.5) [ACI
Table 22.5.6.1], are tabulated in Table 5.13.2. The results are shown in Fig. 5.13.3.
(b) Axial tension. The simple linear expression, Eq. (5.13.6) [ACI 22.5.7.1], is plotted
on the right side of Fig. 5.13.3.
TABLE 5.13.2 SHEAR STRENGTH WITH AXIAL COMPRESSION—

EXAMPLE 5.13.1
N u / Ag Mm 1.88Vu / M m Vc / ( fc′ bw d ) 3.5 1 + 0.002 N u / Ag

(psi) (ft-kips)a
50 231 0.75 2.65 3.5(1.05) = 3.67

100 216 0.81 2.71 3.5(1.095) = 3.83
200 185 0.94 2.84 3.5(1.183) = 4.14
400 123 1.41 3.31 3.5(1.342) = 4.70
600 61 2.83 4.73 3.5(1.484) = 5.19
800 0 — — 3.5(1.612) = 5.64
a M = M − N (4 h − d ) .
m u u
8

5.14 DEEP BEAMS
Deep beams are flexural members in which the assumption of plane sections remaining
plane after loading is not appropriate because of significant nonlinearity in the longitu-
dinal strain distribution over the member depth. In the ACI Code (ACI-9.9.1.1), deep
beams are defined as members loaded on one face and supported on the opposite face such
as to allow the development of diagonal struts. For a member to qualify as a deep beam
in the ACI Code, it must also have a clear ratio of span to overall depth not exceeding
4.0 and/or be subjected to concentrated loads within twice the member depth from the face
of the support.
Since shear resistance in deep beams is primarily provided through diagonal compress-
ion between the loads and supports, the procedures developed previously are not applicable
for deep beams. The design of deep beams, both for shear and flexure, is typically per-
formed using the strut-and-tie method, as permitted in ACI-9.9.1.3. Design of deep beams
is discussed in Chapter 14.
152
5.15 SHEAR FRICTION
Even though uncracked concrete is relatively strong in shear, and shear-related cracks
in usual beams are inclined cracks (diagonal tension cracks), such shear-related cracks
become more vertical as the member becomes deeper in comparison to the shear span,
as discussed in Section 5.4. The ACI Code design procedures for beams as discussed in
Sections 5.5 through 5.13 are intended either to prevent inclined cracking (diagonal ten
sion cracking) or to allow a redistribution of stresses and an increase in load beyond that at
diagonal cracking through the use of web reinforcement.
For situations in which a crack may form and slippage along that crack interface might
occur if no steel reinforcement crosses the crack, and the usual design procedures for shear
reinforcement are inappropriate (such as for a / d less than about 1.0), the shear-friction con-
cept of shear transfer represents a valid tool for design. The use of the shear-friction con-
cept is appropriate for providing a shear transfer mechanism in such cases as the following:
1. At the interface between concretes cast at different times.

2. At the junction of a corbel (bracket) with a column, such as shown in Fig. 5.15.1(a).
3. At the junction of elements in precast concrete construction, such as the bearing
detail shown in Fig. 5.15.1(b).
4. At an interface between steel and concrete, such as the steel bracket attachment to a
concrete column shown in Fig. 5.15.1(c).
The crack for which shear-friction reinforcement is required may not have been caused by
shear. It could, for instance, have been caused by shrinkage. However, once the crack has
occurred (from whatever source), a shear transfer mechanism must be provided. The design
approach is to assume that a crack will occur and then provide reinforcement across the
assumed crack to resist relative displacement along the crack.
Consider a block of concrete as in Fig. 5.15.2(a) acted on by collinear shear forces V such
that a failure plane would form along plane a-a. Since the crack along a-a will tend to be
rough, the sliding motion will produce a separation, as in Fig. 5.15.2(b). One might imagine
slippage along a sawtooth that forces the crack to open; as it opens, the reinforcement is
Nuc
An =
φfy
a Bars welded
to angle
An + part of Avf Vu Nuc
Assumed crack
Remainder
of Avf Assumed crack
and shear plane Avf
Bars welded to angle
Vu
(a) Corbel (see Sec. 5.16 for details) (b) Bearing region of beam
Vu
Embedded
studs, Avf
Studs welded to
plate, steel angles
bracket welded to
Assumed crack face plate
(c) Column face plate
Figure 5.15.1 Examples of applications of shear-friction concept.

153
5 . 1 5 S hear F riction 153
V V
a Section a–a
with
exaggerated
roughness
Reinforcement
a in which
tension is
induced
V V
(a) (b)
Figure 5.15.2 Idealization of the shear-friction concept.
put in tension, with a resulting compression or clamping force on the concrete. A frictional
force is then developed equal to the compression in the concrete (or the tension in the bars)
times the coefficient of friction. If one may assume that the separation and the slip are suf-
ficient to load the steel reinforcement to its yield stress, the shearing resistance then equals
the frictional force; thus (ACI-22.9.4.2) states
Vn = Avf f y µ (5.15.1)
where Avf is the area of reinforcement extending across the potential crack at 90° to it, and
µ is the coefficient of friction between materials along the potential crack. This concept of
shear friction has been verified experimentally [5.68–5.81].
ACI-22.9.1.1 states that shear-friction provisions apply “where it is appropriate to con-
sider shear transfer across any given plane, such as an existing or potential crack, an interface
between dissimilar materials, or an interface between two concretes cast at different times.”
If the shear-friction reinforcement is inclined at an angle to the assumed crack, such that
the shear force produces tension in the shear-friction reinforcement, as shown in Fig. 5.15.3,
the shear strength Vn becomes (ACI-22.9.4.3)
Vn = Avf f y ( µ sin α f + cos α f ) (5.15.2)
where α f is the angle between the shear-friction reinforcement and the shear plane. This
equation does not apply if the shear-friction reinforcement is oriented such that it is sub-
jected to compression.
T sin αf Assumed
crack
T cos αf
T
Force (and its horizontal

and vertical components) Applied shear
that develops in shear-
friction reinforcement
Shear-friction
reinforcement,
C = T sin αf Avf
µC
Compressive force αf
acting to keep
crack from opening T
Figure 5.15.3 Action of shear-friction reinforcement when inclined to shear plane.

154
The logic of Eq. (5.15.2) maybe observed from Fig. 5.15.3, where the shear strength
provided along the potential crack consists of two parts: the vertical component Tcos α f
of the force in the reinforcement Avf and the frictional force µC, which is the same as
µT sin α f . Thus,
Vn = T cos α f + µC (5.15.3)
One should note that the component of the tensile force in the shear-friction reinforcement
normal to the potential crack causes a compression at the crack interface, giving rise to the
friction force µC.
The required nominal shear-friction strength Vn is Vu / φ, in which case Eq. (5.15.1)
becomes
Vu
required A vf = (5.15.4)
φ fy µ
and when α f is less than 90°, the required Avf from Eq. (5.15.2) is
Vu
required Avf = (5.15.5)
φ f y (µ sin α f + cos α f )
Note that Eq. (5.15.5) becomes Eq. (5.15.4) when α f = 90°. Just as for regular stirrups, f y
may not be taken greater than 60,000 psi. The strength reduction factor φ is taken to be 0.75
for shear in Eqs. (5.15.4) and (5.15.5).
The coefficient of friction µ is to be taken (ACI Table 22.9.4.2) as follows:
1. Concrete cast monolithically. µ = 1.4λ

2. Concrete placed against hardened concrete with surface intentionally
roughened to a full amplitude of approximately 1 4 in. µ = 1.0λ
3. Concrete placed against hardened concrete not intentionally roughened. µ = 0.6λ
4. Concrete placed against as-rolled structural steel where shear transfer
is achieved through headed studs or welded bars or wires, with
as-rolled steel “clean and free of paint.” µ = 0.7λ
In the above expressions for µ, the multiplier λ shall be 1.0 for normal-weight concrete, 0.85
for “sand-lightweight” concrete, and 0.75 for “all-lightweight” concrete (ACI-19.2.4.2).
For other lightweight concretes, λ varies depending on the volumetric ratios of lightweight
and normal-weight aggregate, but is not to exceed 0.85.
When normal-weight concrete is cast against hardened concrete intentionally roughened
to a full amplitude of approximately 1 4 in., the maximum nominal shear force that can be
transferred may not exceed the least of [ACI Table 22.9.4.4(a)–(c)]:
a) 0.2 fc′Ac
b) (480 psi +0.08 fc′) Ac
c) (1600 psi) Ac
where Ac is the area of concrete section over which shear transfer is achieved (sq in.).
For all other cases, the nominal shear that can be transferred may not exceed 0.2 fc′Ac nor
(800 psi)Ac [ACI Table 22.9.4.4(d) and (e)]. When concretes with different compressive
strengths are used, the lesser fc′ value should be used.
Since the steel Avf across the potential crack as determined by Eqs. (5.15.4) and (5.15.5)
is only that necessary to provide the clamping action that produces friction, any exter-
nal direct tension on the assumed crack must be provided for by additional reinforcement
(ACI-22.9.4.6). On the other hand, any permanent net compression may be added to the
clamping force provided by the reinforcement (ACI-22.9.4.5).
To ensure development of the required clamping force, shear-friction reinforcement
should be properly anchored on both sides of the shear plane to develop its yield strength
15
5 . 1 5 S hear F riction 155
(ACI-22.9.5.1). Shear-friction reinforcement should also be uniformly distributed along

the shear plane unless a moment is present, in which case it is recommended that shear-
friction reinforcement be placed primarily on the tension side (ACI-R22.9.5.1).
The application of the shear-friction provisions to brackets and corbels appears in
Section 5.16, which is devoted entirely to that topic.
For guidance in the application of shear friction to bearings and other special situations
commonly encountered in precast concrete construction, the designer should consult the
PCI publication Design and Typical Details of Connections for Precast and Prestressed
Concrete [5.82], as well as PCI Design Handbook [2.22]. The following example presents
an application for cases other than brackets and corbels.
EXAMPLE 5.15.1
Design the reinforcement needed at the bearing region of a precast beam 14 in. wide
by 28 in. deep supported on a 4-in. bearing pad. The factored shear Vu is 95 kips. The
horizontal force resulting from restraint of volume change movements due to creep,
shrinkage, and temperature effects is 0.3 of the factored shear Vu . Grade 60 steel is to be
used for reinforcement. A rolled structural steel angle is used for confinement across the
width of the beam at the support.
SOLUTION
(a) Identify the potential crack location. One should assume that a crack will form in the
most undesirable manner. According to the PCI Design Handbook [2.22], a vertical
crack is likely to develop [i.e., crack angle θ ≈ 0° in Fig. 5.15.4(a)]. The crack may
intersect the bottom of the beam immediately adjacent to the bearing pad, which in
this example is taken as 4 in. from the end of the beam.
(b) Determine the shear-friction reinforcement Avf required. Presumably, it would be
appropriate to resolve the Vu and N uc in Fig. 5.15.4(a) into components parallel and
perpendicular to the potential crack when θ is assumed other than zero. Even in such
a case, it will be simpler and more practical to assume that all of Vu will act parallel
to the crack. Thus, using Eq. (5.15.4),
Vu 95
required Avf = = = 1.51 sq in.
φ f y µ 0.75(60)1.4
(c) Determine the additional reinforcement An to provide for the net tension across the
potential crack. It will be conservative not to use the sum of components Vu and N uc
perpendicular to any assumed nonvertical crack, but rather to use N uc as if for a ver-
tical crack. According to ACI-22.9.4.6,
N uc 0.3(95)
required An = = = 0.63 sq in.
φ f y 0.75(60)
Note that N uc was given as 0.3 of Vu , presumably as the result of an analysis for volume
change effects. It is recommended [2.22, 5.79] (and required by ACI-16.5.3.5 in the case
of brackets and corbels) that unless all tensile force N uc can be eliminated by appropriate
design, the value of N uc should not be taken less than 0.2Vu . The φ factor of 0.75 for shear
is considered appropriate for the above calculation, even though 0.90 for axial tension is
indicated by ACI-21.2.2. For brackets and corbels, in the similar situation of reinforce-
ment for the tensile force N uc , ACI-21.2.1 specifies taking φ as 0.75.
(d) Total reinforcement to restrain primary crack [Fig. 5.15.4(a)]. The total reinforce-
ment required is
As = Avf + An = 1.51 + 0.63 = 2.14 sq in.
(Continued)
156
Possible crack
Asv
θ ≈ 0°
Avf + An
15°
Nuc Nuc
Main longitudinal
reinforcement
Vu Vu
(a) Primary possible crack and (b) Secondary possible crack
reinforcement to provide and reinforcement Asv to
shear transfer provide shear transfer
Figure 5.15.4 Shear-friction concept applied to bearing region of a beam.
Use 5–#6, As = 2.20 sq in.

Distribute as shown in Fig. 5.15.5, place at the recommended [5.82] maximum angle
of 15° with the horizontal, weld to the steel angle on one end, and embed the other end
into the beam to develop the tensile strength of the #6 bars beyond the potential crack.
(Development length requirements are treated in Chapter 6.)
(e) Reinforcement for the potential secondary horizontal crack that may form as shown
in Fig. 5.15.4(b). If a vertical crack begins near the corner region where the main
shear-friction reinforcement terminates, then either with or without the tensile force
N uc acting, there would be a potential horizontal crack owing to the tensile force
developed in the main shear-friction reinforcement. The maximum shear that could
act along such a failure plane would be the horizontal shear-friction force arising
from the tensile capacity of the main shear-friction reinforcement. Thus the required
vertical stirrup shear-friction reinforcement Asv [Fig. 5.15.4(b)] is
tensile capacity of shear - friction reinforcement
required Asv =
µ fy
2.20(60)
= = 1.57 sq in.
1.4(60)
Use 4–#4 U stirrups. Asv = 4 ( 0.4 ) = 1.60 sq in.
1 – #4 closed 4 – #4
stirrup U stirrups
@ 4“ spacing
2 – #3 Main
U stirrups reinforcement 28”
15°
Nuc
2’ – 0”
3 – #6
Vu 2 – #6
Figure 5.15.5 Final design for Example 5.15.1.

(Continued)
157
5 . 1 6 B rac k ets and C orbels 157
(f) Additional confinement reinforcement. A conservative approach, recommended by

Mast [5.69], is to provide reinforcement to prevent splitting in the vertical plane of
the beam equal to 25% of the support reaction. This confinement reinforcement is
divided equally into horizontal Ach and vertical Acv portions. Thus,
Vu 95
Ach = Acv = = = 0.20 sq in .
8 f y 8(60)
Use 1-#4 vertical closed stirrup and 2-#3 U stirrups horizontal. The final design is shown
in Fig. 5.15.5.
5.16 BRACKETS AND CORBELS

Brackets and corbels projecting from the faces of columns are widely used in precast con-
crete construction to support beams and girders, as shown in Fig. 5.16.1. It is inappropriate
to design brackets and corbels as cantilever beams using the customary beam provisions for
shear as described in Sections 5.5 through 5.8 and 5.10 through 5.13. As discussed in Section
5.4 and shown in Fig. 5.4.4, when a / d is less than about 1.0, deep beam theory, rather than
simple flexural theory, should apply. Brackets and corbels, furthermore, differ from deep
beams in that design calculations for horizontal forces must also be made. Because the beams
are attached to the bracket, the restraint on the beams due to creep, shrinkage, and tempera-
ture deformations give rise to horizontal forces in the bracket or corbel (N uc in Fig. 5.16.1).
Typically in the past, reinforcement for brackets or corbels consisted of several bars
across the width of the bracket, bent as shown in Fig. 5.16.2(a). When minimum bend
radii are considered, the actual arrangement is as in Fig. 5.16.2(b), where the dashed
line indicates a potential failure surface. When the outer face is too shallow, the critical
inclined crack will form in the location shown in Fig. 5.16.2(c). When the bracket is deep
enough, the crack will tend to extend back into the column [Fig. 5.16.2(d)], with the portion
between the crack and the sloping face acting as a compression element. If the strut can be
developed, the bracket will have reserve capacity after the crack forms; if the strut cannot
develop [as in Fig. 5.16.2(c)], failure will be instantaneous upon formation of the crack.
Vu
a
Asc = ρbd
Nuc
h d
Ah
Figure 5.16.1 Bracket or corbel.
V V V V
N N N N
Potential
failure
surface
(a) As shown on drawing (b) As bent (c) Cracking in corbel (d) Cracking in corbel
with too shallow an with outer face of
outer face sufficient depth
Figure 5.16.2 Corbel details and possible failure modes.

158
Research by Mattock et al. [5.74, 5.75] has shown that the shear-friction concept can
be applied to bracket (corbel) design for a /d as high as 1.0. The design recommendations
of Mattock [5.83], the suggestions of ACI-ASCE Committee 426 [5.16], and the further
discussions of MacGregor and Hawkins [5.87] are the basis for the traditional provisions of
ACI-16.5. Additional discussions of the subject are provided by Shaikh [5.84] and Solanki
and Sabnis [5.85]. Starting with the 2002 edition, the ACI Code allows the design of
brackets and corbels using strut-and-tie models. These provisions are applicable for shear
span-to-depth ratios a /d less than 2; thus, they can be used for larger a /d ratios than the
traditional provisions, which are limited to a /d less than 1. A design example of a corbel
using a strut-and-tie model is given in Chapter 14.
Basic Equilibrium Equations

Using the shear-friction concept and referring to Fig. 5.16.3, the strengths in shear Vn and
in tension N nc are related to the internal forces such that statics is satisfied. From vertical
force equilibrium,
Vn = µC (5.16.1)
From horizontal force equilibrium,
Nnc = T – C (5.16.2)
and from moment equilibrium, taken about point A,
 a   a 
Vn a + N nc  h − 1  = T  d − 1  (5.16.3)
 2   2 
Substituting Eq. (5.16.1) for C into Eq. (5.16.2), and taking T = Asc f y ,
Vn
N nc = Asc f y −
µ
or
Vn N nc
Asc = + (5.16.4)
fy µ fy
Vn
a
T = Ascfy
Nnc
Asc
a1 a1
d– h–
2 2
h d
a1 µC A
C
a1 0.85fc’ Forces contributed by

2 horizontal stirrups or
ties are neglected
Figure 5.16.3 Equilibrium of forces acting on a bracket or corbel.

159
Substitution of Asc f y for T in Eq. (5.16.3) gives
 a   a 
Vn a + N nc (h − d ) + N nc  d − 1  = Asc f y  d − 1 
 2  2 
and solving for As gives
Vn a + N nc (h − d ) N nc (5.16.5)
Asc = +
 a1 fy
fy  d − 
 2 
Equations (5.16.4) and (5.16.5) give the formulas for Asc to provide the required strengths
Vn and N nc.
For design, the factored loads Vu and N uc divided by φ are the required strengths Vn and
N nc, respectively. Thus, Eq. (5.16.4) becomes
Vu N
required Asc = + uc (5.16.6)
f fy µ f fy
and Eq. (5.16.5) becomes
Vu a + N uc (h − d ) N uc
required Asc = + (5.16.7)
 a1  φ fy
φ fy  d − 
 2 
Note that Nuc /(φ fy) is the reinforcement An required for axial tension (using the symbol An in
Example 5.15.1) and Vu /(φ fy µ) is the shear-friction reinforcement Avf given by Eq. (5.15.4).
Furthermore, observe that if the numerator of the first term in Eq. (5.16.7) is treated as an
“equivalent” moment, the first term would represent the reinforcement A f required for a
beam, corresponding to As of Eq. (3.4.6).
To summarize the steel area requirements for brackets and corbels, the following may
be stated:
required Asc = Avf + An (5.16.8)
or
required Asc = Af + An (5.16.9)
in which
Vu
Avf = (5.16.10)
φ fy µ
N uc
An = (5.16.11)
φ f y
equivalent Mu
Af = (5.16.12)
φ f y (d − a1 / 2 )
and
equivalent Mu = Vu a + Nuc(h – d)(5.16.13)
Minimum Horizontal Stirrups

In addition to the steel Asc in Fig. 5.16.3, stirrups in the horizontal plane are needed across
the vertical potential crack to prevent a premature failure. These must be closed stirrups
or ties, or hoops, having a total area Ah. It was conservative to neglect this steel Ah in the
160
development of Eqs. (5.16.8) and (5.16.9) because Ah could have been deducted from the
right side of those equations. Thus Eq. (5.16.8) could become
required Asc = Av f + An – Ah (5.16.14)
Tests [5.76] on brackets (corbels) indicate that minimum Ah for the horizontal stirrups
must be
1
min Ah ≥ Af
2 (5.16.15)
and
1
min Ah ≥ Av f
3 (5.16.16)
ACI Code Provisions

ACI-16.5.5.1 requires the area of primary tension reinforcement Asc to be the larger of the
following:
required Asc = A f + An (5.16.17)
2
required Asc = Avf + An (5.16.18)
3
but not less than
 f ′
Asc = 0.04  c  bw d (5.16.19)
 fy 
The minimum Asc in Eq. (5.16.19) is intended to prevent a sudden failure should
cracking occur.
According to ACI-16.5.5.2, closed stirrups or ties parallel to Asc must also be used, hav-
ing a total area Ah not less than the following:
required Ah ≥ 0.5 ( Asc − An ) (5.16.20)
The area Ah is to be uniformly distributed within two-thirds of the effective depth from Asc
(ACI-16.5.6.6).
Note that the ACI requirements expressed by Eqs. (5.16.17), (5.16.18), and (5.16.20)
automatically satisfy the basic equilibrium requirements of Eqs. (5.16.9) and (5.16.14), as
well as the minimum Ah requirements of Eqs. (5.16.15) and (5.16.16). That Eq. (5.16.16) is
always satisfied can be proved by substituting Eq. (5.16.18) into Eq. (5.16.20), or
2 
required Ah ≥ 0.5  A ν f + An − An 
3 
1
required Ah ≥ A ν f
3
Equation (5.16.18) satisfies Eq. (5.16.14) in recognition that Ah is at least 1

3 Av f .
16
In addition to Eqs. (5.16.17) to (5.17.20), ACI Code limitations and requirements in the
design of brackets and corbels, including those described in the introductory material, are
as follows.
1. Shear span-to-depth ratio a /d may not exceed 1.0 (ACI-16.5.1.1).

2. Factored tensile force N uc may not exceed factored shear Vu (ACI-16.5.1.1).
3. Factored tensile force N uc may not be taken less than 0.2Vu unless special precautions
are taken to avoid tensile forces (ACI-16.5.3.5).
4. Critical section is at face of support, where the effective depth d is to be measured
(ACI-16.5.2.1), as shown in Fig. 5.16.4. The effective depth d may not be more than
twice the depth d1 at the outer edge of the bearing area (ACI-16.5.2.2).
5. The strength reduction factor φ is to be taken as 0.75 for all calculations relating to
the design of brackets and corbels (ACI-21.2.1).
6. The maximum strength Vn (or Vu / φ) for which brackets and corbels may be designed
using normal-weight concrete is (ACI-16.5.2.4)
max Vn = 0.2 fc′ bw d ≤ (480 psi + 0.08 fc′ ) bw d ≤ (1600 psi)bw d (5.16.21)
according to ACI-11.9.3.2.1. For “all-lightweight” or “sand-lightweight” concrete, the

maximum (ACI-16.5.2.5) is
 a  (280 psi) a
max Vn =  0.2 − 0.07  fc′ bw d ≤  800 psi −  bw d (5.16.22)
 d  d
7. Primary reinforcement Asc at front face must be anchored (a) by a structural weld to
a transverse bar of at least equal size to develop a force of Asc f y or (b) by bending
Asc bars back to form a horizontal loop or (c) by some means of positive anchorage
(ACI-16.5.6.3).
8. Bearing area must not project beyond straight portion of Asc bars, nor beyond the
interior face of the transverse anchor bar if one is provided (ACI-16.5.2.3).
Additional Recommendations for Detailing

Kriz and Raths [5.86] provide several recommendations for detailing, as shown in
Fig. 5.16.5. Detail A in that figure is essentially that of ACI-16.5.6.3. Alternatively, a con-
finement angle as in Fig. 5.16.6 can be used, to which the main tension bars are welded
at the underside. The use of the confinement angle is one of the recommendations in PCI
Design Handbook [2.22]. Kriz and Raths also recommended that the outer edge of a bear-
ing plate resting on a corbel should be placed not closer than 2 in. from the outer edge of the
Outer edge Vu
of the
a
bearing area
Asc
Nuc Asc ρ=
bd
b = corbel width
d1
d ≤ 2d1
Figure 5.16.4 Effective depth of bracket or corbel.

162
Note:
Distance x should be large 1 Asc
enough to prevent contact 2
” min
between outer corbel edge Same
and beam due to possible bar Weld must develop the
rotations size yield strength of Asc
Detail A
Unrestrained Restrained
precast beam precast beam
Beam
reinforcement
Elastomeric pad Steel PL not shown
x x
2” min 5
min 2” min
8 ++++++++
++++++++++++
5
min see A s 5
min see A s
8” 8”
3 h/2 min s 3 h/2 min s

4 ” min 4 ” min
d d
h
h
4” min 4” min
(a) Corbels subject to vertical load only (b) Corbels subject to vertical load and
restrained creep and shrinkage force.
Steel PL ‘s welded or not welded.
Figure 5.16.5 Recommended corbel details. (From Kriz and Raths [5.86].)
Face of column
Steel angle Vu
say 6 × 4 × 1
2 Asc
Main tension steel

welded to angle
along contact between
bars and inside of angle
Figure 5.16.6 Anchorage of main steel provided by welding to a confinement angle.
corbel. PCI Design Handbook recommends a 1-in. minimum setback when no confinement
angle is used, but does not indicate a setback when a confinement angle is used. Further,
for good practice, when corbels are designed to resist horizontal forces, steel bearing plates
welded to the tension reinforcement should be used to transfer the horizontal forces directly
to the tension reinforcement. Details and design aids for brackets and corbels are given in
the PCI Handbook [2.22].
163
EXAMPLE 5.16.1
Design a typical interior bracket that projects from a 14-in. square tied column. It must
support a dead load reaction of 26 kips and a live load reaction of 51 kips, resulting from
gravity loads. Assume that suitable bearings are provided for the supported prestressed
concrete girder to avoid the development of horizontal restraint forces. The tolerance
gap between the beam end and column face is 1 in. Use fc′= 5000 psi, f y = 60, 000 psi,
and the ACI Code.
SOLUTION
(a) Factored loads.
Vu = 1.2 VD +1.6 VL = 1.2 ( 26) +1.6 (51) = 113 kips
(b) Preliminary bracket size. The shear span a is dependent on the bearing length
required to support the reaction on the concrete. ACI-22.8.3.2 gives nominal bear-
ing strength as 0.85 fc′A1, where A1 is the bearing area, so that
Vu = φ (0.85 fc′ ) A1
using φ = 0.65 (ACI-21.2.1),

113, 000
bearing plate width = = 2.9 in.
0.65(0.85)(5000)14
Use 3 in. for bearing plate width. Allowing the tolerance gap of 1-in. clear between face
of column and beam for possible overrun in beam length and also because the beam
might be 1 in. too short, the shear span is
1
a = 2 + (bearing plate width) = 2 + 1.5 = 3.5 in.
2
(c) Determine depth of bracket. Based on the maximum strength Vn [Eq. (5.16.21),
repeated below] permitted by ACI-16.5.2.4,
max Vn = 0.2 fc′ bw d ≤ (480 psi + 0.08 fc′ )bw d ≤ (1600 psi)bw d [5.16.21]
Since 0.2 fc′ = 1000 psi, max vn = Vn / ( bw d ) = 480 + 0.08 ( 5000 ) = 880 psi ; then
Vu 113, 000
min d = = = 12.2 in.
φ bw (max vn ) 0.75(14)(880)
Select overall h = 15 in., d ≈ 13.5 in. (allowing 1-in. cover). Check:

a 3.5
= = 0.26 < 1.0 (ACI-16.5.1.1)
d 13.5
(d) Determine the shear-friction reinforcement Av f . Using Eq. (5.16.10) according to

ACI-16.5.4.4,
Vu 113
Aν f = = = 1.79 sq in .
φ f y µ 0.75(60)1.4
using µ = 1.4 for monolithically cast concrete.

(Continued)
164
(e) Determine main tension reinforcement Asc. First calculate the requirement A f for
flexure (ACI 16.5.3.1).
Mu = Vu a + Nuc(h – d)
= Vu a = 113 (3.5)
1
= 33.0 ft-kips
12
Mn 33.0(12,000)
φ bd 2
0.75(14)(13.5)2
From Eq. (3.8.5),

required ρ = 0.0035
From ACI-16.5.5.1, the main steel Asc requirement is the largest of Eqs. (5.16.17) through
(5.16.19), as follows:
Asc = A f + An ( zero in this example ) = ( 0.0035)(14 )(13.5) = 0.66 sq in.
or
 2  2
Asc =   Avf + An ( zero in this example ) =   (1.79) = 1.119 sq in. Controls
 3  3
or
fc′  5
Asc = 0.04 bw d = 0.04   (14)(13.5) = 0.63 sq in.
fy  60 
Use 4–#5 (As = 1.24 sq in.).

(f) Design closed stirrups or ties. In accordance with ACI-16.5.5.2, Eq. (5.16.20)
requires
required Ah = 0.5( Asc − An ) = 0.5 (1.19) = 0.60 sq in.
Note that in this example, with small a / d = 0.28, the flexure requirement A f does not
affect the design; the shear-friction requirement Avf for Vu controls, with Avf = Asc + Ah
[Eq. (5.16.14)].
Use 3–#3 closed stirrups [Ah = 2(3)0.11 = 0.66 sq in.]. The stirrups must be placed
within the upper two-thirds of the effective depth (ACI-16.5.6.6).
(g) Final design. Referring to Fig. 5.16.7, the overall depth at the face of column is
h = 13.5 +1( cover ) + 0.3125 ( bar radius ) = 14.8 in.
Use h = 15 in. The 1-in. cover is used here but it could be as little as the 5 8 -in. bar
diameter for precast columns (ACI Table 20.6.1.3.3), but would presumably have to be
1 1 2 in. for cast-in-place members (ACI Table 20.6.1.3.1).
At the outer edge of the bearing area, the effective depth d1 must be at least half of
that used at the face of the column (see Fig. 5.16.4). In this case, making the outer face
8 in. will satisfy the minimum required d1 of 13.7/2.
Another important aspect of the bracket design is the provision of adequate anchor-
age into the column so that the tensile force Asc f y is available at the face of the column.
Development of reinforcement is treated in Chapter 6.
(Continued)
165
a = 3.5”
Available embedment length
= 14 – 1.5 – 0.375 = 12.1”
#5 bars welded (see Chapter 6)
to underside of 3” 1
steel angle 1 2 ” clear cover
4 – #5
8”
15”
#3 closed stirrups @ 3” spacing
#4
7”
14” square column
EXAMPLE 5.16.2
Design a bracket that is to support gravity dead and live loads of 15 and 25 kips, respec-
tively. The vertical reaction is 10 in. from the face of a 14-in. square column. The bracket
is also subjected to a horizontal reaction of 9.5 kips due to creep and shrinkage of a
restrained beam. Use fc′ = 5000 psi, f y = 40, 000 psi, and the ACI Code.
SOLUTION
(a) Factored loads.
Vu = 1.2 (15) +1.6 ( 25) = 18 + 40 = 58 kips
ACI-16.5.3.4 states that N uc is to be regarded as live load when it results from creep,
shrinkage, or temperature change. Thus
N uc = 1.6 (9.5) = 15.2 kips
Also, Nuc must be greater than or equal to 0.2Vu (ACI-16.5.3.5).

N uc 15.2
= = 0.26 > 0.20 min OK
Vu 58
(b) Depth of bracket for shear.
max Vn = 0.2 fc′ bw d ≤ (480 psi + 0.08 fc′ )bw d ≤ (1600 psi)bw d [5.16.21]
Since 0.2 fc′ = 1000 psi, max vn = Vn / ( bw d ) = 880 psi; then

Vu 58, 000
min d = = = 6.3 in.
φ bw (880) 0.75(14)(880)
(Continued)
16
Since this is very small, perhaps the flexure requirement will require a larger d (this is
a good possibility because the load on the bracket is 10 in. from the face of column).
(c) Depth of bracket for flexure.
M u = Vu a + N uc (h − d )
= 58 (10 ) + 15.2(h − d )
Estimating (h – d) at 2 in.,
Mu = 58 (10 ) +15.2 ( 2 ) = 610 in. -kips
Using the minimum reinforcement ratio,

fc′  5
min ρ = 0.04 = 0.04   = 0.005
fy  40 
which corresponds to minimum Rn = 193 psi (see Section 3.8),
Mu 610, 000
required d = = = 17.3 in.
φ Rn b 0.75(195)14
For a reinforcement ratio of 3%, maximum Rn = 1031 psi, which gives
610, 000
required d = = 7.5 in.
0.75(1031)14
(d) Select bracket depth. Since the provisions of ACI-16.5 for bracket and corbel design
apply only when a / d does not exceed 1.0,
min d = a = 10 in.
Try a bracket somewhat deeper, say 15 in. overall. This would make d ≈ 13.5 in .
(e) Determine the shear-friction reinforcement Avf . Using Eq. (5.16.10) according to
ACI-16.5.4.4,
Vu 58
required A ν f = = = 1.38 sq in.
φ f y µ 0.75(40)1.4
where µ = 1.4 for monolithic concrete.

(f) Determine the flexure reinforcement A f .
Mu
required Rn =
φ bd 2
where (ACI-16.5.3.1)
Mu = Vu a + N uc (h − d ) = 58 (10 ) +15.2 (1.5) = 603 in.-kips
603, 000
required Rn = = 315 psi
0.75(14)(13.5)2
required ρ = 0.0082 [from Eq. (3.8.5) or Fig. 3.8.1]

required Af = 0.0082(14)13.5 = 1.55 sq in.
(Continued)
167
(g) Determine additional reinforcement An for axial tension. In accordance with ACI-
16.5.4.3, using Eq. (5.16.11),
N uc 15.2
required An = = = 0.51 sq in.
φ f y 0.75(40)
(h) Total main tension reinforcement Asc. From ACI-16.5.5.1, the main steel Asc require-
ment is the largest of Eqs. (5.16.17) through (5.16.19), as follows:
required Asc = Af + An = 1.55 + 0.51 = 2.06 sq in. Controls
or
2 2
required Asc = A ν f + An = (1.38) + 0.51 = 1.43 sq in.
3 3
or
fc′  5
Asc = 0.04 bw d = 0.04   (14)(13.5) = 0.63 sq in.
fy  60 
The required Asc = 2.06 sq in.
Use 5–#6 for main tension steel, Asc = 2.20 sq in.

(i) Determine stirrup (or tie) requirements. According to ACI-16.5.5.2,
required Ah = 0.5( Asc − An ) = 0.5 ( 2.06 − 0.51) = 0.78 sq in.
Use 3–#4 closed stirrups, Ah = 0.40 (3) = 1.20 sq in., the spacing of which should be
(2/3)[13.5/3] = 3.0 in. (ACI-16.5.6.6). Use 3-in. spacing.
Steel plate 3” 14” square column

welded to
main steel 2” 10”
Welded 5 – #6
#7 welded
7 21”
15”
#4 ties @
3” spacing
13 21 #4
(Continued)
168
(j) Overall bracket dimensions. Assuming that a 1-in.-thick bearing plate is to be

welded to the main tension reinforcement, the overall depth is
h = bearing plate + bar radius + effective depth d
= 1 + 0.375 + 13.5 = 14.9 in., say 15 in.
Vu
bearing plate length =
φ 0.85 fc′ (column width)
58, 000
= = 1.50 in.
0.65(0.85)5000(14)
Use a 3-in. plate length as the practical minimum.

length of bracket projection = 2 in. + 12 bearing plate + shear span, a
= 2 + 1.5 + 10 = 13.5 in
depth of outer face of bracket = 1

2
overall depth = 7 12 in.
Final design is shown in Fig. 5.16.8.
SELECTED REFERENCES
General
5.1. Emil Mörsch. Concrete- Steel Construction (Der Eisenbetonbau), The Engineering News
Publishing Company, New York, 1909, 368 pp.
5.2. Arthur N. Talbot. “Tests of Reinforced Concrete Beams: Resistance to Web Stresses,”
Engineering Experiment Station, University of Illinois, January 1909, 85 pp.
5.3. Boris Bresler and James G. MacGregor. “Review of Concrete Beams Failing in Shear,” Journal
of the Structural Division, ASCE, 93, ST1 (February 1967), 343–372.
5.4. ACI-ASCE Committee 426. “The Shear Strength of Reinforced Concrete Members—Chapters 1
to 4,” Journal of the Structural Division, ASCE, 99, ST6 (June 1973), 1091–1187.
5.5. Michael P. Collins. “Towards a Rational Theory for RC Members in Shear,” Journal of the
5.6. Michael P. Collins and Denis Mitchell. “Shear and Torsion Design of Prestressed and Non-
Prestressed Concrete Beams,” PCI Journal, 25, September–October 1980, 32–100. Disc. 26,
5.7. Peter Marti. “Basic Tools of Reinforced Concrete Beam Design,” ACI Journal, Proceedings, 82,
5.8. Frank J. Vecchio and Michael P. Collins. “The Modified Compression- Field Theory for
Reinforced Concrete Elements Subjected to Shear,” ACI Journal, Proceedings, 83, March–April
1986, 219–231. Disc. 84, January–February 1987, 87–90.
5.9. Jörg Schlaich, Kurt Schäfer, and Mattias Jennewein. “Toward a Consistent Design of Structural
Concrete,” PCI Journal, 32, May–June 1987, 74–150.
5.10. Frank J. Vecchio and Michael P. Collins. “Predicting the Response of Reinforced Concrete
Beams Subjected to Shear Using Modified Compression Field Theory,” ACI Structural Journal,
85, May–June 1988, 258–268.
5.11. Julio A. Ramirez and John E. Breen. “Evaluation of a Modified Truss-Model Approach for
Beams in Shear,” ACI Structural Journal, 88, September–October 1991, 562–571.
5.12. Khaled A. Al-Nahlawi and James K. Wight. “Beam Analysis Using Concrete Tensile Strength in
Truss Models,” ACI Structural Journal, 89, May–June 1992, 284–289.
5.13. Michael P. Collins, Denis Mitchell, Perry Adebar, and Frank J. Vecchio. “A General Shear
Design Method,” ACI Structural Journal, 93, January–February 1996, 36–45.
169
S elected R eferences 169
5.14. A. Koray Tureyen and Robert J. Frosch. “Concrete Shear Strength: Another Perspective,” ACI
Structural Journal, 100, September–October 2003, 609–615.
5.15. ASCE-ACI Committee 445. “Recent Approaches to Shear Design of Structural Concrete,”
Journal of Structural Engineering, ASCE, 124, 12 (December 1998), 1375–1417.
5.16. ACI–ASCE Committee 426. Suggested Revisions to Shear Provisions for Building Codes.
Detroit: American Concrete Institute, 1979, 82 pp.
5.17. ACI-ASCE Committee 326. “Shear and Diagonal Tension,” ACI Journal, Proceedings, 59,
January, February, and March 1962, 1–30, 277–344, and 352–396.
Aggregate Interlock
5.18. T. Paulay and P. J. Loeber. “Shear Transfer by Aggregate Interlock,” Shear in Reinforced
Concrete, Vol. 1 (SP-42). Detroit: American Concrete Institute, 1974 (pp. 503–537).
5.19. Theodossius P. Tassios and Elizabeth N. Vintzeleou, “Concrete-to-Concrete Friction,” Journal
of Structural Engineering, ASCE, 113, 4 (April 1987), 832–849.
5.20. Sandro Dei Poli, Pietro G. Gambarova, and Cengiz Karakoc. “Aggregate Interlock Role in R.C.
Thin-Webbed Beams in Shear,” Journal of Structural Engineering, ASCE, 113, 1 (January
1987), 1–19.
Dowel Action
5.21. David W. Johnston and Paul Zia. “Analysis of Dowel Action,” Journal of the Structural Division,
ASCE, 97, ST5 (May 1971), 1611–1630.
5.22. Rafael Jimenez, Richard N. White, and Peter Gergely. “Cyclic Shear and Dowel Action Models
in R/C,” Journal of the Structural Division, ASCE, 108, ST5 (May 1982), 1106–1123.
5.23. Parviz Soroushian, Kienuwa Obaseki, Maximo C. Rojas, and Jongsung Sim. “Analysis of
Dowel Bars Acting Against Concrete Core,” ACI Journal, Proceedings, 83, July–August 1986,
642–649.
5.24. E. N. Vintzeleou and T. P. Tassios. “Behavior of Dowels under Cyclic Deformations,” ACI
Structural Journal, 84, January–February 1987, 18–30.
5.25. P. Soroushian, K. Obaseki, M. Rojas, and H. S. Najm. “Behavior of Bars in Dowel Action
Against Concrete Cover,” ACI Structural Journal, 84, March–April 1987, 170–176.
5.26. Elizabeth N. Vintzeleou and Theodossius P. Tassios. “Eccentric Dowels Loaded Against Core
of Concrete Sections,” Journal of Structural Engineering, ASCE, 116, 10 (October 1990),
2621–2633.
Beams Without Shear Reinforcement

5.27. JoDean Morrow and I. M. Viest. “Shear Strength of Reinforced Concrete Frame Members with-
out Web Reinforcement,” ACI Journal, Proceedings, 53, March 1957, 833–869.
5.28. G. N. J. Kani, “How Safe Are Our Large Reinforced Concrete Beams?” ACI Journal,

Proceedings, 64, March 1967, 128–141.
5.29. Michael P. Collins and Daniel Kuchma. “How Safe Are Our Large, Lightly Reinforced Concrete
Beams, Slabs, and Footings?” ACI Structural Journal, 96, July–August 1999, 482–491.
5.30. Edward G. Sherwood, Evan C. Bentz, and Michael P. Collins. “Effect of Aggregate Size on
Beam-Shear Strength of Thick Slabs,” ACI Structural Journal, 104, March-April 2007, 180–190.
5.31. G. N. J. Kani. “The Riddle of Shear Failure and Its Solution,” ACI Journal, Proceedings, 61,
April 1964, 441–467.
Lightweight Aggregate
5.32. John M. Hanson. Square Openings in Webs of Continuous Joists (RD001.01D). Skokie,
IL: Portland Cement Association, 1969.
5.33. J. A. Hanson. “Tensile Strength and Diagonal Tension Resistance of Structural Lightweight
Concrete, ACI Journal, Proceedings, 58, July 1961, 1–40.
5.34. E. Hognestad, R. C. Elstner, and J. A. Hanson. “Shear Strength of Reinforced Structural
Lightweight Aggregate Concrete Slabs,” ACI Journal, Proceedings, 61, June 1964, 643–656.
5.35. Don L. Ivey and Eugene Buth. “Shear Capacity of Lightweight Concrete Beams,” ACI Journal,
Proceedings, 64, October 1967, 634–643.
170
High-Strength Concrete
5.36. Andrew G. Mphonde and Gregory C. Frantz. “Shear Tests of High-and Low-Strength Concrete
Beams without Stirrups,” ACI Journal, Proceedings, 81, July–August 1984, 350–357.
5.37. Ashraf H. Elzanaty, Arthur H. Nilson, and Floyd O. Slate. “Shear Capacity of Reinforced
Concrete Beams Using High-Strength Concrete,” ACI Journal, Proceedings, 83, March–April
1986, 290–296.
5.38. Shuaib H. Ahmad, A. R. Khaloo, and A. Poveda. “Shear Capacity of Reinforced High-Strength
Concrete Beams,” ACI Journal, Proceedings, 83, March–April 1986, 297–305.
5.39. Ashraf H. Elzanaty, Arthur H. Nilson, and Floyd O. Slate. “Shear Capacity of Prestressed
Concrete Beams Using High-Strength Concrete,” ACI Journal, Proceedings, 83, May–June
1986, 359–368.
5.40. Shuaib H. Ahmad and D. M. Lue. “Flexure-Shear Interaction of Reinforced High-Strength
Concrete Beams.” ACI Structural Journal, 84, July–August 1987, 330–341. Disc. 85, May–June
1988, 354–358.
5.41. M. Keith Kaufman and Julio A. Ramirez. “Re-evaluation of the Ultimate Shear Behavior of
High-Strength Concrete Pre-stressed I-Beams,” ACI Structural Journal, 85, May–June 1988,
295–303.
5.42. Miguel A. Salandra and Shuaib H. Ahmad. “Shear Capacity of Reinforced Lightweight High-
Strength Concrete Beams,” ACI Structural Journal, 86, November–December 1989, 697–704.
5.43. John J. Roller and Henry G. Russell. “Shear Strength of High-Strength Concrete Beams with
Web Reinforcement,” ACI Structural Journal, 87, March–April 1990, 191–198.
Truss Models and Shear Reinforcement

5.44. W. Ritter. “Die Bauweise Hennebique,” Schweizerische Bauzeitung, 33, February 1899,

pp. 59-61.
5.45. Willis A. Slater, Arthur R. Lord, and Roy R. Zipprodt. “Shear Tests of Reinforced Concrete
Beams,” Technologic Papers of the Bureau of Standards, No. 314, 20, April 1926, pp. 389-495.
5.46. Frank E. Richart. “An Investigation of Web Stresses in Reinforced Concrete Beams,” Bulletin
No. 166, University of Illinois Engineering Experiment Station, June 1927, 105 pp.
5.47. Peter Marti. “Truss Models in Detailing,” Concrete International, 7, 12 (December 1985), 66–73.
5.48. M. P. Nielsen. Limit Analysis and Concrete Plasticity (2nd ed.). Boca Raton, FL: CRC
Press, 1998.
5.49. Denis Mitchell and Michael P. Collins. “Diagonal Compression Field Theory—A Rational
Model for Structural Concrete in Pure Torsion,” ACI Journal, 71, August 1974, 396–408.
5.50. Thomas T.C. Hsu. “Softened Truss Model Theory for Shear and Torsion,” ACI Structural
Journal, 85, November–December 1988, 624–635.
5.51. Wayne Hsiung and Gregory C. Frantz. “Transverse Stirrup Spacing in R/C Beams,” Journal of
Structural Engineering, ASCE, 111, 2 (February 1985), 353–362. Disc. 113, 1 (January 1987),
174–177.
5.52. Mark K. Johnson and Julio A. Ramirez. “Minimum Shear Reinforcement in Beams with Higher
Strength Concrete,” ACI Structural Journal, 86, July–August 1989, 376–382. Disc. 87, May–
June 1989, 362–364.
5.53. Neal S. Anderson and Julio A. Ramirez. “Detailing of Stirrup Reinforcement,” ACI Structural
Journal, 86, September–October 1989, 507–515.
5.54. Andrew G. Mphonde. “Use of Stirrup Effectiveness in Shear Design of Concrete Beams,” ACI
5.55. Andrew G. Mphonde and Gregory C. Frantz. “Shear Tests of High and Low-Strength Concrete
Beams with Stirrups,” High-Strength Concrete (SP-87). Detroit: American Concrete Institute,
1985 (pp. 179–196).
5.56. Abdeldjelil Belarbi and Thomas T. C. Hsu. “Stirrup Stresses in Reinforced Concrete Beams,”
ACI Structural Journal, 87, September–October 1990, 530–538.
5.57. CSA. Design of Concrete Structures for Buildings (CAN3-A23.3-04). Rexdale, Ontario: Canadian
Standards Association (178 Rexdale Blvd., Rexdale, Ontario, Canada M9W IR3), 2004.
5.58. S. M. Fereig and K. N. Smith. “Indirect Loading on Beams with Short Shear Span,” ACI Journal,
Proceedings, 74, May 1977, 220–222.
5.59. K. S. Rajagopalan and P. M. Ferguson. “Exploratory Shear Tests Emphasizing Percentage of
Longitudinal Steel,” ACI Journal, Proceedings, 65, August 1968, 634–638. Disc. 66, 150–154.
5.60. Barrington deV. Batchelor and Mankit Kwun. “Shear in RC Beams without Web Reinforcement,”
Journal of the Structural Division, ASCE, 107, ST5 (May 1981), 907–921.
5.61. Michael N. Palaskas, Emmanuel K. Attiogbe, and David Darwin. “Shear Strength of Lightly
Reinforced T-Beams,” ACI Journal, Proceedings, 78, November–December 1981, 447–455.
17
S elected R eferences 171
Fiber Reinforced Concrete

5.62. Gustavo J. Parra-Montesinos. “Shear Strength of Beams with Deformed Steel Fibers,” Concrete
International, 28, November 2006, 57-66.
Axial Load Effect
5.63. James G. MacGregor and John M.Hanson. “Proposed Changes in Shear Provisions for

Reinforced and Prestressed Concrete Beams,” ACI Journal, Proceedings, 66, April 1969,
276–288. Disc. 849–851.
5.64. Alan H. Mattock. “Diagonal Tension Cracking in Concrete Beams with Axial Forces,” Journal
of the Structural Division. ASCE, 95, ST9 (September 1969), 1887–1900.
5.65. Munther J. Haddadin, Sheu-
Tien Hong, and Alan M. Mattock. “Stirrup Effectiveness in
Reinforced Concrete Beams with Axial Force,” Journal of the Structural Division, ASCE, 97,
ST9 (September 1971), 2277–2297.
5.66. Shrinivas B. Bhide and Michael P. Collins. “Influence of Axial Tension on the Shear Capacity
of Reinforced Concrete Members,” ACI Structural Journal, 86, September–October 1989,
570–581.
5.67. Alan H. Mattock and Zuhua Wang. “Shear Strength of Reinforced Concrete Members Subject to
High Axial Compressive Stress,” ACI Journal, Proceedings, 81, May–June 1984, 287–298.
Shear Transfer; Shear Friction; Corbels

5.68. Philip W. Birkeland and Halvard W. Birkeland. “Connections in Precast Concrete Construction,”
ACI Journal, Proceedings, 63, March 1966, 345–368.
5.69. R. F. Mast. “Auxiliary Reinforcement in Precast Concrete Connections,” Journal of the Structural
Division, ASCE, 94, ST6 (June 1968), 1485–1504.
5.70. J. A. Hofbeck, I. O. Ibrahim, and Alan H. Mattock. “Shear Transfer in Reinforced Concrete,”
ACI Journal, Proceedings, 66, February 1969, 119–128. Disc. 66, August 1969, 678–680.
5.71. A. H. Mattock and N. M. Hawkins. “Research on Shear Transfer in Reinforced Concrete,” PCI
Journal, 17, March–April 1972, 55–75.
5.72. Bjorn R. Hermansen and John Cowan. “Modified Shear-Friction Theory for Bracket Design,”
ACI Journal, Proceedings, 71, February 1974, 55–60.
5.73. A. H. Mattock. Disc. of “Modified Shear-Friction Theory for Bracket Design,” by B. R.
Hermansen and J. Cowan, ACI Journal, Proceedings, 71, August 1974, 421–423.
5.74. A. H. Mattock, “Shear Transfer in Concrete Having Reinforcement at an Angle to the Shear
Plane,” Shear in Reinforced Concrete, Vol. 1 (SP-42). Detroit: American Concrete Institute,
1974 (pp. 17–42).
5.75. Alan H. Mattock, L. Johal, and H. C. Chow. “Shear Transfer in Reinforced Concrete with
Moment or Tension Acting Across the Shear Plane,” PCI Journal, 20, July–August 1975, 76–93.
5.76. Alan H. Mattock, W. K. Li, and T. C. Wang. “Shear Transfer in Lightweight Reinforced
Concrete,” PCI Journal, 21, January–February 1976, 20–39.
5.77. Joost Walraven, Jerome Frenay, and Arjan Pruijssers. “Influence of Concrete Strength and
Load History on the Shear Friction Capacity of Concrete Members,” PCI Journal, 32, January–
February 1987, 66–84.
5.78. Thomas T. C. Hsu, S. T. Mau, and Bin Chen. “Theory of Shear Transfer Strength of Reinforced
Concrete,” ACI Structural Journal, 84, March–April 1987, 149–160.
5.79. Robert A. Bass, Ramon L. Carrasquillo, and James O. Jirsa. “Shear Transfer Across New and
Existing Concrete Interfaces,” ACI Structural Journal, 86, July–August 1989, 383–393.
5.80. G. Annamalal and Robert C. Brown, Jr. “Shear Transfer Behavior of Post-Tensioned Grouted
Shear-Key Connections in Precast Concrete-Framed Structures,” ACI Structural Journal, 87,
5.81. Nijad I. Fattuhi. “Reinforced Corbels Made with High-Strength Concrete and Various Secondary
Reinforcements,” ACI Structural Journal, 91, July–August 1994, 376–383. Disc. 92, May–June
1995, 386–387.
5.82. PCI Committee on Connection Details. Design and Typical Details of Connections for Precast
and Prestressed Concrete (2nd ed). Chicago: Prestressed Concrete Institute, 1988.
5.83. Alan H. Mattock, “Design Proposals for Reinforced Concrete Corbels,” PCI Journal, 21, May–
June 1976, 18–42. Disc., 22, March–April 1977, 90–109.
5.84. A Fattah Shaikh, “Proposed Revisions to Shear-Friction Provisions,” PCI Journal, 23, March–
April 1978, 12–21.
172
5.85. Himat Solanki and Gajanan M. Sabnis. “Reinforced Concrete Corbels—
Simplified,” ACI
5.86. L. B. Kriz and C. H. Raths, “Connections in Precast Concrete Structures—Strength of Corbels,”
PCI Journal, 10, 1 (February 1965), 16–47.
5.87. J. G. MacGregor and N. M. Hawkins. “Suggested Revisions to ACI Building Code Clauses
Dealing with Shear Friction and Shear in Deep Beams and Corbels,” ACI Journal, Proceedings,
74, November 1977, 537–545. Disc., 75, May 1978, 221–224.
PROBLEMS
All problems13 are to be worked in accordance with the ACI Code unless otherwise indi-
cated, and all stated loads are service loads. Use the “correct” shear envelope (i.e., live load
must be appropriately applied such as to cause maximum effect). All shear diagrams are to
be drawn to scale in terms of Vu directly below the diagram of the beam, also drawn to scale.
Use the Vu diagram for design by scaling values from it; also scale from it the locations
where stirrup spacings are permitted. Computation of Vu values and locations of stirrup
spacings may be made to confirm scaled information. Unless otherwise indicated, use the
load factors U of ACI-5.3.1 and the φ factors of ACI-21.2.
5.1 The simply supported beam of 16-ft span, shown draw a smooth curve through these points
in the figure for Problem 5.1, is to carry a uni- for design use.
form dead load of 1.6 kips/ft (including beam (b) Draw the curve for required stirrup spacing
weight) and a uniform live load of 2.6 kips/ft. using the simplified method for calculating
Use fc′ = 3500 psi and f yt = 60, 000 psi. Vc (ACI Formula 22.5.5.1); show on the
(a) Determine the adequacy of the #3 U stirrups same diagram the spacings provided.
that are spaced at 8 in. Use the simplified (c) Use the simplified method of constant Vc
method with constant Vc . (ACI F ormula 22.5.5.1) to evaluate whether
(b) Draw superimposed on the factored shear Vu the spacings used satisfy the ACI Code.
diagram the diagram of strength provided (d) Repeat part (b) using the more detailed pro-
φVn. From the comparison of φVn with Vu, cedure of ACI Table 22.5.5.1; show also the
determine if the stirrups are adequate for the spacings provided.
entire beam. If not, make recommendations to (e) Evaluate whether the spacings used are satis-
obtain a beam having adequate shear strength. factory according to the analysis of part (d).
5.2 The beam in the figure for Problem 5.2 carries a 5.3 The beam in the figure for Problem 5.3 is to carry
uniform live load of 3.6 kips/ft in addition to its 1.6 kips/ft live load and 0.90 kip/ft dead load
own weight. Assume a support width of 12 in., (including beam weight). Using fc′ = 3000 psi
and use fc′ = 4500 psi and f yt = 60, 000 psi. and f yt = 40, 000 psi, investigate the beam for
(a) Draw the maximum factored shear Vu envel- stirrup adequacy according to the simplified
ope. Calculate the value at the critical sec- method using constant Vc . If design is not ade-
tion, at the 1 4 point, and at midspan; then quate, indicate what revision is necessary.
18.5”
12” 4” 5 – #8 14”
4 @ 8” 3 @ 9” 5 – #8
# 3U stirrups Sym. abt CL
8’ – 0”
Problem 5.1
13 Some problems may be solved as problems stated in Inch-Pound units, or as problems in SI units using quantities in
parentheses. To avoid implying higher precision for the given information in SI units than that given for the Inch-Pound units, the
metric conversions are approximate.
173
P roble m s 173
5 – #9 25”
6 @ 10” = 5’–0” Same stirrups

#3 U stirrups
as at other 12”
8” support
5 @ 6” = 2’–6”
20’–0”
Problem 5.2
Symmetrical about CL span bE = 60”
4” 22” 3 – #7
3 – #9
12” supports d = 18”
12”
#3 U stirrups
9” 9 @ 6”
22’–0”
Problem 5.3
d = 19.5”
12”
12”
9’–0”
58k CL of span
Factored shear Vu
18k
Problem 5.4
5.4 For the portion of the continuous beam shown 5.5 For the beam shown in the figure for Problem
in the figure for Problem 5.4, with the given 5.5 and the case assigned by the instructor, use
portion of the factored shear Vu envelope, the simplified method of ACI Formula 22.5.5.1
determine the spacings to be used for #3 U to design vertical U stirrups (i.e., specify their
stirrups. Dimension and show the stirrups on size, dimension their locations, and show them
the given portion of the beam. Use the simpli- on a side view of the beam). Use whole inches
fied method of constant Vc , with fc′ = 3500 psi for spacings. The beam is simply supported and
and f yt = 60, 000 psi. (Beam: b = 300 mm; has a support width of 18 in.; and the given dead
d = 530 mm; support width = 300 mm; half- load does not include the beam weight. Choose
span = 2.7 m; Vu at support = 260 kN; Vu at mid- the stirrup size to ensure that the spacing will not
span = 80 kN; use #10M stirrups; fc′ = 25 MPa; be closer than 3 in. Explicitly state the length of
f yt = 400 MPa.) the beam over which stirrups are required.
Dead Load Live Load

Case Span (ft) bw (in.) d (in.) As (sq in.) (kips/ft) (kips/ft) fc′ (psi) fyt (psi)
1 20 18 36.3 5-#10 6 10 4000 40,000

5-#11
2 26 18 29.6 6-#9 1.6 2.7 4000 60,000
3 30 16 24.6 9-#9 1.2 2.5 4000 60,000
4 28 18 29.6 6-#9 1.6 2.5 4000 60,000
174
h d
As
bw
18”
Symmetrical about
CL of span
L/2
Problem 5.5
5.6 For the simply supported beam shown in the compression steel effect. Use a “correct” fac-
figure for Problem 5.6 having a span of 32 ft, tored shear Vu envelope. Using the simplified
with uniform dead load of 2.3 kips/ft (includ- procedure of constant Vc , design and detail the
ing beam weight) and live load of 3.7 kips/ft, spacings for #3 U stirrups for the beam. Use
design #4 vertical U stirrups using whole inches whole inches for spacings.
for spacings. Show the stirrups on a side view 5.9 Repeat Problem 5.8, except use an 18-ft main
of the beam and dimension their locations. Use span and a 7-ft cantilever, with dead load of
fc′ = 3750 psi and f yt = 50, 000 psi, and the sim- 1.5 kips/ft (including beam weight) and live load
plified method with constant Vc . The support of 1.7 kips/ft. All other dimensions and reinforc-
width is 12 in. ing bars are the same as in Problem 5.8.
5.10 The beam shown in the figure for Problem 5.10
is to carry a dead load of 55 kN/m (including
d = 24.5” beam weight) and a live load of 72 kN/m. Use
fc′ = 25 MPa and f yt = 400 MPa. Using the sim-
plified procedure of constant Vc, design and
15” detail on the beam 20-mm increment spacings
of U stirrups for the beam. Choose a stirrup size
Problem 5.6 to ensure that the 20-mm increment spacing will
not be closer than 80 mm.
5.7 A reinforced concrete simply supported beam 5.11 For a simply supported beam of 16-ft span, hav-
(b = 250 mm, d = 410 mm) must carry on a span ing support widths of 12 in., design and detail
of 5.5 m a single concentrated moving load of on the beam U stirrups (use no spacing less than
45 kN plus a uniform dead load of 30 kN/m 3 in.). The dead load is 1.6 kips/ft (including
(including beam weight). Design and detail weight of beam) and the live load is 3.0 kips/ft.
the stirrup spacing (use 20- mm increments) Use fc′ = 3000 psi and f yt = 60, 000 psi. Use the
for #10M vertical U stirrups; support width is ACI Code simplified procedure with constant
300 mm; fc′ = 25 MPa; f yt = 400 MPa. Apply the Vc . Assume bw = 14 in. and d = 21.5 in.
simplified method using a constant value for Vc. 5.12 A reinforced concrete simply supported beam
See SI footnote to Eq. (5.10.3). of span 5.5 m carries a concentrated dead load
5.8 The beam shown in the figure for Problem 5.8 of 23 kN at 1.8 m from the left support and a
is to carry dead load of 1.4 kips/ft (including uniform dead load of 90 kN/m. The width of
beam weight) and live load of 1.6 kips/ft. Use support is 300 mm. The rectangular beam has a
fc′ = 3500 psi, f yt = 40, 000 psi, and neglect any 300-mm width and a 650-mm effective depth d.
9’–0”
2 – #8 4’–6” 4’–0” 2 – #8
2 – #9 4 – #8
4 – #9 24”
12” supports 12”

2’–6”
Prob. 5.8
20’–0” 8’–0”
18’–0” 7’–0”
Prob. 5.9
Problems 5.8 and 5.9

175
P roble m s 175
2.5 m
3 – #45M
660 mm
300 mm
4 – #30M
Supports
360 mm
5.6 m 2.8 m
Problem 5.10
axial compressive force of 70 kips dead load and

Use fc′ = 30 MPa and f yt = 400 MPa. Design and
120 kips live load is acting on the beam.
specify by dimensioning the spacings to be used
(a) Use simplified method using ACI Formula
for #10M U stirrups. Use the simplified method
(22.5.6.1).
of constant Vc (ACI-Formula 22.5.5.1).
(b) Use more detailed ρVd / M method (ACI
5.13 For a rectangular beam of 14 in. width and
Table 22.5.6.1).
effective depth 22.5 in. with fc′ = 4000 psi and
5.20 Design the details of the bearing shoe on a pre-
f yt = 60, 000 psi, determine the maximum fac-
stressed girder of 12 in. width. Assume a ver-
tored shear Vu for this beam for the following
tical crack forms at the support as shown in
conditions:
Fig. 5.15.4(a). The reaction is 35 kips dead load
(a) When no stirrups are to be used.
and 40 kips live load. The girder concrete has
(b) When minimum amount of shear reinforce-
ment (#3 U stirrups) is used according to fc′ = 6000 psi. Assume no horizontal restraint is
ACI-9.6.3.3; specify the spacing to be used. developed.
(c) When maximum amount of shear rein- 5.21 Design a bracket (corbel) that projects from
forcement is used (#4 U stirrups) accord- one side of a 16 × 16 column to support a verti
ing to ACI-22.5.1.2; specify the spacing to cal load of 35 kips dead load and 65 kips live
be used. load. Assume that suitable bearings are pro-
5.14 Completely design and detail the stirrups for vided to prevent any horizontal restraint. The
the beam of Problem 5.3 (ignore the spacings reaction is located 5 in. from the column face.
given), using the more detailed ρVd / M method Use fc′ = 5000 psi and f y = 60, 000 psi . (Column
of ACI Table 22.5.5.1. size = 400 × 400 mm; dead load = 160 kN; live
5.15 For the case assigned by the instructor, repeat load = 290 kN; reaction 130 mm from column
the requirements of Problem 5.5, except use face; fc′= 35 MPa; f y = 400 MPa.)
the more detailed ρVd / M method of ACI 5.22 Redesign the bracket (corbel) of Problem 5.21 if the
Table 22.5.5.1. reaction is 3.5 in. (90 mm) from the column face.
5.16 Repeat the requirements of Problem 5.8, except 5.23 Design for the conditions of Problem 5.21
use the more detailed ρVd / M method of ACI except take the reaction location 9 in. (230 mm)
Table 22.5.5.1. from the column face.
5.17 If a factored axial compression N uc of 140 5.24 Repeat Problem 5.21 if the reaction is from a
kips is acting additionally on the beam of restrained beam that induces a horizontal ten-
Example 5.13.1, determine the number of stir- sion equal to 50% of the total gravity reaction.
rups that may be eliminated by taking the com- 5.25 Redesign the bracket (corbel) of Problem 5.24
pression into account when computing Vc by if the reaction is 3.5 in. (90 mm) from the
the more detailed procedure involving ρVd / M . column face.
(Note: This compressive force is approximately 5.26 Repeat Problem 5.23 if the reaction is from a
0.1 fc′Ag and might reasonably be neglected in restrained beam that induces a horizontal ten-
designing the section for flexure.) sion equal to 40% of the total gravity reaction.
5.18 Reinvestigate the shear reinforcement for the 5.27 Redesign the bracket (corbel) of Example 5.16.1
beam of Problem 5.1 if an axial tensile force of considering that the supported prestressed girder
35 kips live load is acting. Redesign stirrups for is welded to the bracket. Creep, shrinkage, and
the beam using ACI Formula 22.5.7.1). temperature effects on the restrained girder
5.19 Redesign the stirrups for the beam case of induce a horizontal force of 50 kips (unfactored)
Problem 5.5 assigned by the instructor if an on the bracket.
CHAPTER 6
DEVELOPMENT
OF REINFORCEMENT
6.1 GENERAL
A basic requirement in reinforced concrete construction is the adequate and reliable trans-
fer of the force in the reinforcement to the surrounding concrete. Consider the bar embed-
ded in a concrete block a length L, with an applied tensile force as shown in Fig. 6.1.1. As
the bar is loaded, the tensile force T must be balanced by stresses acting on the bar surface
over the embedded length L. These interacting stresses between the bar and the surround-
ing concrete have been traditionally referred to as bond stresses. If the embedded length of
the bar, L, is too short, the bond stresses may not be able to balance the applied bar force
T. Consequently, the bar will begin to slip and eventually pull out or split the concrete.
The length of embedment necessary to transfer the full bar force into the concrete is called
development length. In design, the development length is usually calculated to develop the
specified yield strength of the bar. In members subjected to seismic actions, however, a
greater development length may be required.
Concrete cantilever beams, Denver (Photo by C. G. Salmon).

17
6.2 DEVELOPMENT LENGTH 177
Steel bar
Concrete
Figure 6.1.1 Bond stresses between a bar and surrounding concrete along the bar length.
The sources of bond stresses include chemical adhesion, friction, and bearing against
the concrete of the raised ribs, or “lugs,” of the deformed bars. Chemical adhesion between
the steel reinforcing bars and the concrete does not offer great resistance. For smooth bars
(i.e., without ribs), very long embedment lengths would be required to develop the bar
yield strength before breaking the adhesion between the bar surface and the surrounding
concrete. Friction can provide added resistance, but it is not always a reliable transfer
mechanism. As the bar is stretched in tension, its diameter will be slightly reduced owing
to Poisson’s effect, but enough to gradually detach it from the surrounding concrete. The
higher the bar force, the greater the reduction in the bar diameter and the smaller the
friction resistance. For these reasons, smooth bars (i.e., without ribs) are not used in prac-
tice (except for confinement of the concrete). Instead, only deformed bars with specially
designed rib patterns, rib angles, and rib gap (see Section 1.13) are specified. The bond
transfer mechanism associated with bearing of the bar ribs against the surrounding con-
crete, as well as the concepts relating to the transfer of force between reinforcement and the
surrounding concrete over the development length, with or without additional mechanical
end anchorage, are presented in the next several sections. The mechanics of the behavior
has been explained by ACI Committee 408 [6.1], Lutz and Gergely [6.2], Orangun, Jirsa,
and Breen [6.3, 6.4], Jirsa, Lutz, and Gergely [6.5], Yankelevsky [6.6], and Kemp [6.7].
6.2 DEVELOPMENT LENGTH
Design of longitudinal and shear reinforcement to accommodate the moment and shear at
sections along a beam has been treated in Chapters 3, 4, and 5. To resist bending moment,
an area of longitudinal steel is provided to carry the tensile force. However, the bars must
be embedded in the concrete sufficiently to allow the tensile force to develop. If there
is inadequate development length, the bars will either pull out or split the surrounding
concrete.
The flexural strength of a beam is, therefore, a three-dimensional relationship involving
not only the cross-sectional properties at a location along the span, but also the embedment
lengths of the steel bars in both directions therefrom.
Consider a uniformly loaded cantilever beam as shown in Fig. 6.2.1(a), which has been
properly proportioned so that at nominal strength Mn the steel force is As fy. To illustrate
the concept of development length, assume the tension reinforcement consists of a single
bar of diameter db. Consider the free-body diagram of the bar segment A-B, as shown in
Fig. 6.2.1(b). The tensile force at B, which is f y (πdb2 / 4) , must be transmitted to the con-
crete by the interaction between the bar and the surrounding concrete over the development
178
178 C hapter 6 D evelopment of R einforcement
length L1 = AB. If us, the failure stress against slippage acting over the nominal surface area
πdb L1, is assumed to be constant over L1, then equilibrium requires that at a slippage failure
db2
us π db L1 = f y π
4
or
fy
L1 = db (6.2.1)
4us
On the other hand, if ub is the failure stress against splitting and Abr is the average bearing
area per unit length (also assumed constant over L1), then, at a splitting failure
db2
ub Abr L1 = f y π
4
or
fy db2
L1 = π (6.2.2)
Abr ub 4
In other words, the bar must be provided with a development length at least equal to
or greater than that given by Eq. (6.2.1) (for the bar to reach yield prior to pullout) or
by Eq. (6.2.2) (to prevent a splitting failure).
The same situation exists in free body B-C, as shown in Fig. 6.2.1(c). Thus the max-
imum tensile force at B has to develop by embedment in both directions from B: that is,
both the A-B and B-C distances. Where space limitations prevent providing the proper
amount of straight embedment, such bars may be terminated by standard hooks (as defined
in ACI-25.3). A standard hook is permitted to be considered as contributing to an equivalent
development length by mechanical action (ACI-25.4.3), thus reducing the total embed-
ment length required. Section 6.10 treats the subject of providing development length with
standard hooks.
Adequate development length must be provided for a reinforcing bar in compression as
well as in tension.
A B C
L1 L2
(a)
L1 L2
u πdb2 u
T= f T
4 y
A B B C
(b) (c)
Figure 6.2.1 Development of reinforcement.

179
6.3 FLEXURAL BOND 179
6.3 FLEXURAL BOND
As bending moment varies along a span, the tensile force in the steel also varies; this
induces flexural bond, that is, the longitudinal interaction between the bars and the sur-
rounding concrete. High localized stress, either at the bar surface or at the bearing of the
steel reinforcement lugs against the concrete, exists at locations along the span where the
rate of change of tensile force in the bars is high.
Consider a segment D - D ′ of the reinforcing bar in the cantilever beam used in Section 6.2.
As shown by the free body of D - D ′ in Fig. 6.3.1(b), TD is slightly greater than TD′. The bend-
ing moment MD equals the internal force C or T times the moment arm between them; thus,
MD MD′
TD = and TD ′ =
arm arm (6.3.1)
Also, from horizontal force equilibrium,
us π db dz + ub Abr dz = TD – TD ′ (6.3.2)
in which us is the localized surface stress over the nominal contact area between the steel
bar and the concrete, db is the diameter of the single bar, and ub is the localized bearing
stress over the area Abr per unit length between the lugs and the concrete.
Whether the action represented by the left side of Eq. (6.3.2) is more “pullout”
or “splitting,” the magnitude depends on (TD – TD ′ ) / dz , which from Eq. (6.3.1) is
dM / (arm ) dz  = V / (arm ). Thus, this localized interaction (i.e., flexural bond) between
bar and surrounding concrete is proportional in magnitude to the shear.
Even though high localized surface stress on bars resulting from the rate of change of
moment may seem to be important for proper design, research and experience in practice
have shown that prevention of bar “pullout” failure or “splitting” failure by having adequate
development length of embedment is a sufficient criterion. Several cases may be identified
where the local situation has little significance in design, provided the bars have adequate
development length.
1. In a region of low bending moment and the concrete is uncracked: thus the change
in force, say (TD – TD ′ ) in Fig. 6.3.1, in the bars is overestimated because concrete
actually carries part of the tensile force.
2. At a point where high bending moment (and low shear) exists and therefore where a
flexural crack will likely occur, the low rate of change of moment (i.e., shear) would
indicate low (TD − TD ′ ) . Adjacent to the flexural crack (see Fig. 6.3.2), however, the
D D’
A B C
dz
(a)
D u D’
TD TD’
dz
(b)
Figure 6.3.1 Flexural bond in a tension bar.

180
Flexural
cracks
Calculated fs
fs between
(at crack)
cracks
Near max ultimate

surface stress
0 0
Figure 6.3.2 Probable surface stress between cracks when beam shear is zero.
surface stress is likely to be high because the concrete shares in carrying the tension,
whereas at the crack, the concrete flexural tension is zero [6.12, 6.49].
3. In the vicinity of a shear-related inclined crack, such as that of Fig. 5.4.2, not only
does high surface stress exist adjacent to the crack but, in addition, the tensile force T
computed at z from the support actually acts much closer to the support (refer to the
truss model discussion in Section 5.7 and Fig. 5.7.1).
4. At locations where some bars are terminated in a tension zone, the abrupt change in
the distribution of the total tensile force among the bars gives rise to very high sur-
face stresses.
Thus, the local effects related to the rate of change in moment are not directly correlated
with the development-length-related strength of the member. When the bars are properly
anchored—that is, when they have adequate development length provided and continue to
carry their required tensile force—the localized stress condition is not of concern.
6.4 BOND FAILURE MECHANISMS

The term “bond failure” has been given to the mechanism by which failure occurs when the
provided development length is inadequate. Years ago, when plain bars (relatively smooth
bars without lug deformations) were used, slip resistance (“bond”) was thought of as adhe-
sion between concrete paste and the surface of the bar. Yet even with low tensile stress in
the reinforcement, there was sufficient slip immediately adjacent to a flexural crack in the
concrete to break the adhesion, leaving only friction to resist bar movement relative to the
surrounding concrete over the slip length.
Shrinkage can also cause frictional drag against the bars. Typically, a hot-rolled plain
bar may pull loose by longitudinal splitting when the adhesion and friction resistances are
high, or just pull out, leaving a cylindrical hole when adhesion and friction resistances
are low.
Deformed bars were created to change the behavior pattern so that there would be less
reliance on friction and adhesion (though these still exist) and more reliance on the bearing
of the lugs against the concrete. These bearing forces act at an angle to the axis of the bar,
which cause longitudinal and radial outward components against the concrete, as shown in
Fig. 6.4.1. The radial component causes circumferential tensile stresses [Fig. 6.4.1(d)] and
will tend to crack (split) the concrete around the bar.
When inadequate development length is provided, deformed bars in normal-weight con-
crete usually give rise to a splitting mode of failure (i.e., “bond failure”) [6.1, 6.5, 6.7].
A splitting failure occurs when the wedging action of the steel lugs on a deformed bar
18
6.4 BOND FAILURE MECHANISMS 181
Bearing and friction

forces on bar
Adhesion and friction forces

along the surface of the bar
(a) On bar (b) On concrete (c) Components
radial
(d) Radial component tends

to split surrounding concrete
Figure 6.4.1 Forces between bar and surrounding concrete.
Final splitting failure
Whole layer suddenly

splits after initial
horizontal splits at sides
First splitting
(a) (b) (c)
Figure 6.4.2 Splitting cracks and ultimate splitting failure modes. (From ACI Committee 408 [6.1].)
causes cracks in the surrounding concrete parallel to the bar. These cracks occur between
the bar and the nearest concrete face, as shown in Fig. 6.4.2(a) and 6.4.2(b), or between bars
when bars are closely spaced, as in Fig. 6.4.2(c).
When small-size bars are used with large cover, the lugs may crush the concrete by bear-
ing and result in a pullout failure without splitting the concrete. This nonsplitting failure has
also been reported for larger bars in structural lightweight concrete [6.1].
Although splitting is the usual failure mode, an initial splitting crack on one face of
a beam is not considered failure. The distress sign indicating failure is progressive split-
ting. The presence of transverse reinforcement (stirrups, ties, or spirals) will provide con-
finement and restrain the growth of splitting cracks, thereby delaying failure (defined as
an increase in deformation that results in no increase in resistance) until several splitting
cracks have formed.
Originally, development length requirements were based on pullout tests [6.8] of plain
bars, followed by pullout tests [6.9–6.15] of deformed bars, including the related load-slip
data. These tests often consisted of a concrete block with an embedded bar. In such tests,
the surrounding concrete provides ample confinement, and a pullout rather than a splitting
failure occurs. Thus, the results of pullout tests were not representative of splitting failures
commonly observed in beams.
182
s
Cs Cs Cs Cs
db db Cs1 db Cs2 Cs2 db Cs1
Failure plane Cylinder of concrete
tributary to bar
Cb Cb
Cs1 > Cb ... C = Cb
Cb > Cs, ...C = Cs Cs2 > Cb
(a) (b)
Figure 6.4.3 Concrete cylinder hypothesis for splitting failure. (From Orangun, Jirsa, and Breen [6.3].)
The splitting failure mode has been studied in detail by Orangun, Jirsa, and Breen [6.3,
6.4] and many others [6.7, 6.16–6.24]. The studies of Orangun, Jirsa, and Breen [6.3] and
Untrauer and Warren [6.16] have hypothesized that the action of splitting arises from a
stress condition analogous to a concrete cylinder surrounding a reinforcing bar and acted
on by the outward radial components [Fig. 6.4.1(d)] of the bearing forces from the bar.
The cylinder would have an inner diameter equal to the bar diameter db and a thickness
C equal to the smaller of Cb, the clear bottom cover, or Cs, half of the clear spacing to the
next adjacent bar (see Fig. 6.4.3). The tensile strength of this concrete cylinder determines
the resistance against splitting. If Cs < Cb , a side-split type of failure occurs [Fig. 6.4.2(c)].
When Cs > Cb , longitudinal cracks through the bottom cover form first [first splitting cracks
in Fig. 6.4.2(a) and 6.4.2(b)]. If Cs is only nominally greater than Cb, the secondary split-
ting will be side splitting along the plane of the bars. If Cs is significantly greater than Cb,
the secondary splitting will also be through the sides [Fig. 6.4.2(a)] or through the bottom
cover to create a V-notch failure [Fig. 6.4.2(b)].
The current ACI Code provisions for development length of straight bars are based on
a proposal from a 1979 ACI Committee 408 report [6.25] that recognized the cylinder
hypothesis for splitting failure. Although the development length rules in the ACI Code
have changed in format over the years, the basis of the provisions has remained unchanged
since 1979.
6.5 FLEXURAL STRENGTH DIAGRAM—B AR BENDS

AND CUTOFFS
As stated in Section 6.2, the flexural strength of a beam at any section along its length is
a function of its cross section and the actual embedment length of its reinforcement. The
concept of a diagram showing this three-dimensional relationship can be a valuable aid in
determining cutoff or bend points of longitudinal reinforcement. It may be recalled from
Chapter 3 that in terms of the cross section, the nominal flexural strength for a singly rein-
forced rectangular beam may be expressed as:
 a
M n = As f y  d –  [3.4.6]
 2
Equation (3.4.6) assumes that the steel reinforcement comprising As is adequately embed-
ded in each direction by the required development length Ld from the section where Mn is
computed such that the stress fy is reached.
183
6 . 5 F L E X U R A L S T R E N G T H D I AG R A M 183
► EXAMPLE 6.5.1
For the beam of Fig. 6.5.1, compute and draw qualitatively the flexural strength diagram.
SOLUTION
The flexural strength in each region is represented by the horizontal portions of the diagram
in Fig. 6.5.1. In this example, there are five bars of one size in section C–C; thus the flexural
strength provided by each bar is in this case approximately one-fifth of the total capacity.
Actually, the sections with four and two bars will have a little more than four-fifths and
two-fifths, respectively, of the total strength of the section containing five bars, owing to the
slight increase in moment arm when the number of bars in the section decreases.
At point a, the location where the fifth bar terminates, this bar has zero embedment
length to the left and thus has zero capacity. Proceeding to the right from point a, the bar
may be counted on to carry a tensile force proportional to its embedment from point a up
to the development length Ld. Thus, in Fig. 6.5.1, point b represents the point where the
fifth bar is fully developed through the distance Ld and can therefore carry its full tensile
capacity. The other cutoff points are treated in the same way.
A B C
2 bars 4 bars Point a 5 bars
A B C
Note: Showing the tip of a bar bent up is a scheme used throughout
this text to show the bar termination. The bars are actually
straight and lie in a common plane.
CL of span
Section A–A Section B–B Section C–C
Ld
b
Ld
a
Mn of 5 bars
Ld
Mn of 4 bars
Mn of 2 bars
Figure 6.5.1 Flexural strength diagram.

184
► EXAMPLE 6.5.2
Demonstrate qualitatively the practical use of the diagram for the design strength, φ M n ,
for verification of the locations of cutoff or bend points in a design. Assume that the
main cross section with five equal-sized bars provides exactly the required strength at
midspan for this simply supported beam with uniform load, as shown in Fig. 6.5.2.
SOLUTION
(a) Compute the design strength, φ M n , for each potential bar grouping that may be
used; in the present case, for five bars, four bars, and two bars.
(b) Decide which bars must extend entirely across the span and into the support. In beams,
ACI-9.7.3.8.1 states, “At simple supports, at least one-third of the maximum positive
moment reinforcement shall extend along the beam bottom into the support at least 6
in. …” In this example, two bars should extend into the support.
(c) Decide on the order of cutting or bending the remaining bars. The least amount of lon-
gitudinal reinforcement will be obtained when the resulting φ M n diagram is closest
to the factored moment Mu diagram. With that thought in mind, and proceeding from
maximum moment region to the support, cut off one bar as soon as permissible.
(d) Cutoff restrictions. Point A of Fig. 6.5.2 is the theoretical location where the design
strength, φ M n , of the remaining four bars is adequate. To provide for a safety factor
against shifting of the moment Mu diagram (especially in continuous spans) and to pro-
vide partially for the complexity arising from a potential diagonal crack, the ACI Code
Uniformly distributed load
d
2 bars 4 bars 5 bars
Cross section
at midspan
Center of
CL of span
support
Ld interior bars Ld center bar
φMn of 5 bars
φMn
diagram
φMn of 4 bars
Ld corner bars B
A Theoretical
d or 12 diam cutoff point
Factored moment diagram, Mu
C
Theoretical cutoff point φMn of 2 bars
d or 12 diam
Mu is zero at centerline of simple support
At end of bar: no capacity; develops full strength

by “bond” at length Ld
Figure 6.5.2 Factored moment Mu and design flexural strength φ Mn, diagrams for determining bar cutoffs.
(Continued)
185
6.6 DEVELOPMENT LENGTH FOR TENSION REINFORCEMENT 185
states that there must be an extension beyond the point where a bar theoretically may
be terminated, or it may be bent into the compression face. Accordingly, ACI-9.7.3.3
requires that the reinforcement be extended beyond the point at which it is no longer
required to resist flexure for a distance equal to the effective depth of the member or 12
bar diameters, whichever is greater, except at supports of simply supported spans and at
free ends of cantilevers.
(e) Once cutoff or bend points have been located, a check is made by drawing the φ M n
diagram to ensure no encroachment on the factored moment Mu diagram.
(f) Other restrictions. Since points B and C of Fig. 6.5.2 are bar terminations in a ten-
sion zone, the stress concentrations described in Section 6.3 are present (Fig. 6.3.2),
effectively reducing the shear strength of the beam [6.26, 6.27]. Thus, for cutoffs
to be acceptable, one of the three special conditions of ACI-9.7.3.5 must be satis-
fied. These conditions are discussed in Section 6.11. If these bars were bent up and
anchored in the compression zone, no further investigation would be necessary; this,
however, is seldom done in today’s practice.
6.6 DEVELOPMENT LENGTH FOR TENSION

REINFORCEMENT—A CI CODE
The term “development length” was defined in Section 6.2 as the length of embedment
needed to develop the yield stress in the reinforcement. As described in Section 6.4,
the development length requirement is primarily a function of the splitting resistance of
the concrete surrounding the bars rather than a frictional-adhesional pullout resistance. The
splitting resistance is roughly proportional to the bar area, indicated by Eq. (6.2.2), whereas
the pullout resistance is roughly proportional to the bar diameter, indicated by Eq. (6.2.1).
The 2014 ACI Code provisions are based on the basic relationship developed by
Orangun, Jirsa, and Breen [6.3, 6.4] and by a later study by Sozen and Moehle [6.28]. The
general equation for computing the development length of deformed bars is given in ACI-
25.4.2.3 as ACI Formula (25.4.2.3a),
 

3 fy ψ t ψ e ψ s 
Ld =  d (6.6.1)1
 40 λ fc′  cb + K tr   b

 d  
 b 
where
Ld = development length
db = nominal diameter of bar or wire
cb = cover or spacing dimension
= the smaller of (1) distance from center of bar or wire being developed to the
nearest concrete surface, and (2) one-half the center-to-center spacing of bars or
wires being developed
1 For SI, with fy and fc′ in MPa,

 
 
 fy ψt ψeψs 
Ld = d (6.6.1)
 1.1λ fc′  cb + K tr   b
  d  
 b 
186
Ktr = transverse reinforcement index defined as follows:
40 Atr
K tr = (6.6.2)2
sn
where
Atr = total cross-sectional area of all transverse reinforcement which is within the
spacing s that crosses the potential plane of splitting through the reinforcement
being developed
s = maximum center-to-center spacing of transverse reinforcement within develop-
ment length Ld
n = number of bars being developed along the plane of splitting
The transverse reinforcement index, Ktr, reflects the improvement and increase in strength
obtained from the confinement provided by the transverse steel. It is noted, however, that
ACI-25.4.2.3 allows the use of K tr = 0 even if transverse reinforcement is present. This is
conservative.
The modification factors ψ t , ψ e , ψ s , and λ are discussed in detail later in Section 6.7.
In the use of Eq. (6.6.1), the cover and transverse reinforcement terms cannot be taken
greater than 2.5; thus,
 cb + K tr 
 d  ≤ 2.5 (6.6.3)
b
The upper limit on this parameter represents a pullout failure and indicates that beyond a
certain point, increasing the amount of cover or the amount of transverse reinforcement is
no longer effective at increasing the bond stress at failure. Indeed, providing a very large
cover or extremely heavy transverse reinforcement will result in pullout rather than a bond-
splitting failure.
In computing the transverse reinforcement index, Ktr, it is important to identify the
potential plane of splitting and the total area of reinforcement crossing the potential split-
ting plane, Atr. To be fully effective, the transverse reinforcement must be adjacent to the
bar being developed and must cross the splitting plane on the outside of the bar [6.5].
In Fig. 6.6.1, examples showing the potential splitting planes and the effectiveness of the
transverse reinforcement in beams are illustrated. In this figure, it is assumed that all shown
bars are being developed at the same location.
If one-half the center-to-center spacing of the bars is smaller than the distance from
the center of the bar to the nearest concrete surface, then cracking between bars with a
horizontal splitting plane can be expected to occur as shown in Fig 6.6.1(a). For the bar
arrangement and transverse reinforcement shown in Fig. 6.6.1(a), the stirrup is effective
only for the outer bars. In such a case, the designer could choose to calculate a different Ld
for the inner and outer bars, ignore the effect of the transverse reinforcement, or include the
effect of the transverse reinforcement as an average over the bars being developed. In the
last case, the total area of reinforcement crossing the splitting plane would be computed as
Atr = 2 Aleg ( Aleg = area of one leg of the stirrup) for the four bars being developed and thus,
n = 4 in Eq. (6.6.2).
If, on the other hand, the bottom cover controls, vertical splitting would be expected to
occur, as shown in Fig. 6.6.1(b). In this case, the stirrup crosses each splitting crack and the
transverse reinforcement can be considered to be effective for all the bars. Thus, the total
area of transverse reinforcement may be computed as Atr = 3 Aleg with n = 3 in Eq. (6.6.2).
For the bar arrangement shown in Fig. 6.6.1(c), the transverse reinforcement is effec-
tive for only four of the seven bars being developed and thus, Atr = 4 Aleg and n = 7 in
Eq. (6.6.2).
2 For SI,
40 Atr
K tr =
sn (6.6.2)
187
Atr = 2Aleg Atr = 3Aleg

n =4 n =3
Splitting plane Splitting
(a) Atr is effective only for two outer bars (b) Atr is effective for all bars
Atr = 4Aleg Atr = 2Aleg

n =7 n =2
Splitting plane Splitting
(c) Atr is effective only for four bars (d) Atr is effective for all bars
Figure 6.6.1 Transverse reinforcement effectiveness for various bar arrangements and splitting
planes when all bars shown are developed at the same location. (Adapted from Ref. 6.5.)
For other cases, the designer must exercise judgment in identifying the splitting planes
and in computing the contribution of the transverse reinforcement. For example, if the side
and bottom cover are the same (a common case in practice), then vertical or side splitting,
or both vertical and side splitting, can occur in practice as shown in Fig. 6.6.1(d). In fact,
diagonal splitting cracks may also occur, as shown in Fig. 6.4.2. For the purpose of com-
puting the contribution of the transverse reinforcement, the vertical and side splitting cracks
occurring around the same bar are considered as a single splitting plane. Thus, for the
example shown in Fig. 6.6.1(d), the transverse reinforcement is considered effective for
both bars, with Atr = 2 Aleg and n = 2 in Eq. (6.6.2). In any case, it is always conservative to
ignore the contribution of the transverse reinforcement, although in some cases it may be
too conservative and ignoring it will result in development lengths much greater than those
required.
Simplified Equations of ACI-25.4.2.2

The use of Eq. (6.6.1) is a rather involved way of determining development length Ld. For
many practical situations, ACI-25.4.2.2 provides simplified equations that can be also used.
These simplified equations are divided into two categories, as follows.
Category A
Either one of the following two conditions will satisfy this most favorable situation:
1. (a) clear lateral spacing between bars at least db, and

(b) clear cover at least db, and
(c) minimum Code-specified stirrups or ties along the development length
or
2. (a) clear lateral spacing between bars at least 2db, and
(b) clear cover not less than db.
18
Simplification of Eq. (6.6.1) after substituting ( cb + K tr ) / db = 1.5 and a bar size modifi-
cation factor ψ s = 0.8 for #6 and smaller bars (see Section 6.7) becomes:
for #6 and smaller bars:
 fy ψ t ψ e 
Ld =   db (6.6.4)3
 25λ fc′ 
for #7 and larger bars:

Ld =   db (6.6.5)4
 20 λ fc′ 
Category B
Anything not in Category A is in Category B. Simplification of Eq. (6.6.1) after substituting
(cb + K tr ) / db = 1.0 and a bar size modification factor ψ s = 0.8 for #6 and smaller bars (see
Section 6.7) becomes
for #6 and smaller bars:
 3 fy ψ t ψ e 
Ld =   db (6.6.6)5
 50 λ fc′ 
for #7 and larger bars:

 3 fy ψ t ψ e 
Ld =   db (6.6.7)6
 40 λ fc′ 
It is noted that no strength reduction factor, φ, is used in the development length equations.
An allowance for strength reduction has already been included in both the general and sim-
plified equations (Eqs. 6.6.1, 6.6.4–6.6.7) for computing development lengths.
Useful Tables
Development length for common values of fc′ and Grade 60 reinforcement are shown in
Tables 6.6.1 and 6.6.2 for all bar sizes using the simplified equations of the ACI Code [i.e.,
Eqs. (6.6.4)–(6.6.7)]. Note the descriptive heading on the top of these tables. Also note that
3 For SI, with fy and fc′ in MPa, for #20 (see Table 1.13.2) and smaller:
Ld =   db (6.6.4)
 2.1λ fc′ 
4 For SI, with fy and fc′ in MPa, for #22 (see Table 1.13.2) and larger:
Ld =   db (6.6.5)
 1.7λ fc′ 
5 For SI, with fy and fc′ in MPa, for #20 (see Table 1.13.2) and smaller:
Ld =   db (6.6.6)
 1.4 λ fc′ 
6 For SI, with fy and fc′ in MPa, for #22 (see Table 1.13.2) and larger:
Ld =   db (6.6.7)
 1.1λ fc′ 
189
the modification factors ψ t, ψ e, and λ have all been assumed equal to 1.0 (see Section 6.7
for the definition and values of these factors). These tables show that for Grade 60 rein-
forcement and normal-weight concrete with fc′ = 4000 psi, the simplified equations require
a development length of about 40 bar diameters for the smaller bar sizes and about 50 bar
diameters for the larger bars.
Practical Application of the Simplified Development Length Equations

The practicality for applying the simplified equations in ordinary reinforced concrete
construction is that most beams will contain at least Code-specified minimum stirrups
(thereby satisfying Category A, item 1c); in addition, clear spacing must satisfy the larger
of the bar diameter db , 1 in. or 4/3 of the maximum aggregate size (ACI-25.2.1), and
cover must satisfy the minimum specified in ACI-20.6.1.3.1 in any case. Using the min-
imum 1.5 in. of cover on beams will commonly provide the Category A minimum of
db . For slab-like elements without shear reinforcement, clear spacing will usually sat-
isfy the Category A, item 2a, minimum of 2db; and 3/4-in. minimum cover required by
ACI-20.6.1.3.1 will commonly provide at least the db cover required for Category A,
item 2b.
Regarding bar spacing for beams, minimum width tables for satisfying the minimum
of ACI-25.2.1 (db ≥ 1 in.), 2db, and 3db are given in Section 3.9 as Tables 3.9.2, 3.9.3, and
3.9.4, respectively. All bar arrangements in beams must satisfy Table 3.9.2; Category A,
item 2, beams must satisfy Table 3.9.3; and good practice should provide 3db in accordance
with Table 3.9.4.
It must be mentioned that the simplified equations, while practical, generally result in
conservative values of the required development length. Where congestion of reinforce-
ment is anticipated, it is preferable to use the general equation, Eq. (6.6.1).
TABLE 6.6.1 DEVELOPMENT LENGTH Ld IN INCHES FOR GRADE 60 BARS,

CATEGORY A,* EQS. (6.6.4) AND (6.6.5) WITH ψt , ψe , AND λ = 1.0
fc′ (psi)
Bar 3000 4000 5000
#3 16.4 14.2 12.7

#4 21.9 19.0 17.0
#5 27.4 23.7 21.2
#6 32.9 28.5 25.5
#7 47.9 41.5 37.1
#8 54.8 47.4 42.4
#9 61.8 53.5 47.9
#10 69.6 60.2 53.9
#11 77.2 66.9 59.8
#14 92.7 80.3 71.8
#18 124 107 95.8
*(a) Clear spacing and clear cover ≥ db and minimum stirrups, or
(b) Clear spacing ≥ 2db and clear cover ≥ db.
190
TABLE 6.6.2 DEVELOPMENT LENGTH Ld IN INCHES FOR GRADE 60 BARS,

CATEGORY B,* EQS. (6.6.6) AND (6.6.7) WITH ψt , ψe , AND λ = 1.0
fc′ (psi)
Bar 3000 4000 5000
#3 24.6 21.3 19.1

#4 32.9 28.5 25.5
#5 41.1 35.6 31.8
#6 49.3 42.7 38.2
#7 71.9 62.3 55.7
#8 82.2 71.2 63.6
#9 92.7 80.3 71.8
#10 104 90.4 80.8
#11 116 100 89.7
#14 139 120 108
#18 185 161 144
* Everything not in Category A.
6.7 MODIFICATION FACTORS ψ t , ψ e , ψ s , AND λ TO THE

BAR DEVELOPMENT LENGTH EQUATIONS—A CI CODE
The symbols ψ t , ψ e , ψ s , and λ in Eq. (6.6.1) are modification factors to the required bar
development length per ACI-25.4.2.4 that consider the main parameters known to affect
bond strength for various bar surface conditions that may be commonly encountered in
practice.
Casting Position, ψt
Horizontal bars placed so that more than 12 in. (300 mm) of fresh concrete is cast in the
member below the development length or splice region are often referred to as top-cast
bars. Research has shown that top-cast bars have lower bond strengths owing to increased
settlement and bleed of water at the bar as the depth of concrete below the bar increases
[6.1, 6.11, 6.29]. For this reason, whenever the depth of concrete mixture placed in one lift
under the horizontal bar exceeds 12 in., the ACI Code requires a 30% increase in develop-
ment length (i.e., a modification factor ψ t = 1.3). Otherwise, ψ t = 1.0.
Epoxy-Coated Bars, ψe
Epoxy coating of reinforcing bars reduces adhesion and lowers the coefficient of friction
between the bar surface and the surrounding concrete. As a result, bond strength of epoxy-
coated reinforcement is reduced (Cleary and Ramirez [6.30], Choi, Hadje-Ghaffari, Darwin,
and McCabe [6.31], Hamad, Jirsa, and D’Abreu de Paulo [6.32], Hadje-Ghaffari, Choi,
Darwin, and McCabe [6.33], Bartoletti and Jirsa [6.34], Treece and Jirsa [6.50], Johnston
and Zia [6.51], and Mathey and Clifton [6.52]). The effect of epoxy coating on bond strength
is reflected in ACI-25.4.2.4 through three different modification factors ψ e, depending on the
type of coating, concrete cover, and clear spacing of the bars. The values are
1. Epoxy-coated, or zinc and epoxy dual-coated bars:

With clear cover less than 3db , OR with clear spacing less than 6 db ψ e = 1.5
All other conditions ψ e = 1.2
2. Uncoated or zinc-coated (galvanized) bars ψ e = 1.0
19
6 . 7 M O D I F I CAT I O N FAC TO R S 191
When the epoxy-coated bars are in the unfavorable category of ψ e = 1.5 and the bars are
top-cast bars (ψ t = 1.3), ACI-25.4.2.4 states that the product ψ t ψ e need not exceed 1.7 (it
would otherwise be 1.95).
Bar Size, ψs
This factor reflects the results of tests that suggested that shorter development lengths
would be required for smaller diameter bars. Accordingly, the modification factor ψ s = 0.8
for #6 and smaller bars. For #7 and larger bars, ψ s = 1.0.
This factor was obtained based on limited test data. Subsequent data and a reevaluation
of the influence of bar size on bond strength by ACI Committee 408 [6.53] suggests that no
reduction factor should be used. While the subject is still being debated, the latest report of
ACI Committee 408 [6.53] does not support the use of this factor.
Lightweight Aggregate Concrete, λ

There is increased potential for a pullout failure mode in lightweight aggregate concrete.
Thus, a modification factor λ = 0.75 is to be used for lightweight aggregates of all types.
Alternatively, a greater λ factor is permitted based on the composition of the aggregate in
the concrete mixture per ACI Table 19.2.4.2 or when the splitting tensile strength of the
lightweight concrete is specified per ACI-19.2.4.3 (see Section 1.8). For normal-weight
concrete, λ = 1.0.
Excess Reinforcement, αexs
When steel reinforcement in excess of that required has been provided, the designer is per-
mitted to reduce the development length by the ratio (required As / provided As) in limited
situations (ACI-25.4.10.1). However, the development length is not permitted to be reduced
by excess reinforcement in the following situations (ACI-25.4.10.2):
a) when anchorage to develop fy is specifically required. For example, the factor does not
apply for the development of the temperature and shrinkage reinforcement required
by ACI-24.4.3.4 and of the reinforcement required by the structural integrity provi-
sions of ACI-8.7.4.2, 8.8.1.6, 9.7.7 and 9.8.1.6.
b) at noncontinuous supports.
c) where bars are required to be continuous.
d) for headed and mechanically anchored deformed reinforcement.
e) when the flexural member is part of the seismic force resisting system in structures
assigned to Seismic Design Categories D, E, or F in accordance to ASCE 7 [2.18].
Minimum Development Length
After all modifications factors have been applied, the resulting development length Ld can-
not be less than 12 in. (300 mm) in accordance with ACI-25.4.2.1(b).
192
Summary
The modification factors for both the general equation, Eq. (6.6.1), and the simplified equa-
tions, Eqs. (6.6.4) through (6.6.7), are summarized below. For conditions not listed, the
value of the modification factor is 1.
Condition Modification Factors ψ t , ψ e , ψ s, and λ
Casting positiona ψ t = 1.3

Epoxy-coated or zinc and epoxy ψ e = 1.5
dual-coated barsa for bars with clear cover < 3db
or bars with clear spacing < 6db
ψ e = 1.2 otherwise
#6 or smaller bars ψ s = 0.8b
Lightweight aggregate concrete λ = 0.75c
Excess reinforcement α exs = As required / A s provided ≤ 1.0
a The product ψ t ψ e need not exceed 1.7.
b Not recommended by ACI Committee 408 [6.53]. Recommend ψ s = 1.0 for all bar sizes.
c A larger value may be permitted based on the composition of the aggregate in the concrete mixture (ACI-19.2.4.2) or when the
fct
splitting tensile strength is specified, λ = ≤ 1.0 (ACI-19.2.4.3).
6.7 fcm
EXAMPLE 6.7.1
Determine the development length Ld required for the #9 epoxy-coated bars A on the
top of a 15-in. slab, as shown in Fig. 6.7.1. Use f y = 60, 000 psi, and fc′ = 4000 psi with
lightweight aggregate concrete. Use both the simplified and the general equation.
Bars B
Bars A
8”
1 8” 2 3
8”
Plan view #9 @ 8”, long (bars B)

121 ” clear
15”
121 ” clear
CL of span
#9 @ 8”, short (bars A)
alternate with
#9 @ 8”, long (bars B)
Figure 6.7.1 Top bars for Example 6.7.1.
SOLUTION
(a) Determine the development length Ld using the simplified equations. Since the 1.5
in. clear cover exceeds db of 1.128 in., and the 8-in. bar spacing exceeds clear spac-
ing of 2db (i.e., 2.3 in.), the bars satisfy the conditions under Category A, item 2. For
#9 bars Eq. (6.6.5) applies,
Ld =   db [6.6.5]
 20 λ fc′ 
(Continued)
193
6 . 7 M O D I F I CAT I O N FAC TO R S 193
 60, 000 ψ t ψ e   47.4 ψ t ψ e 

Ld =   db =   db
 20 λ 4000   λ
 47.4 ψ t ψ e  ψψ ψψ
Ld =   db = 47.4(1.128) t e = 53.5 t e
 λ  λ λ
Note that 53.5 in. agrees with the value in Table 6.6.1.

In checking the bar spacing, bars A are developed over distance 1-2, while bars B are
developed over the distance 2-3 (Fig. 6.7.1). In such cases, the spacing to be used for
bars A is the spacing of the closest bars that terminate at the same point. In other words,
the spacing for both bars A and B is 8 in.
(b) Modification factor ψ t for top bars. Since the negative moment region bars are cast
with more than 12 in. of fresh concrete below them, a casting position factor ψ t = 1.3
must be used.
(c) Modification factor ψ e for coated bars. Since the bars are epoxy coated, a modifica-
tion factor ψ e may apply. Check clear cover,
1.5 1.5
clear cover = = = 1.3db < 3db
db 1.128
Since clear cover is less than 3db , ψ e = 1.5. However, the maximum value of ψ t ψ e = 1.7
controls in this case.
(d) Modification factor λ for lightweight aggregate concrete. Since the composition of
the lightweight aggregate in the concrete mixture is unknown and the tensile split-
ting strength is not specified, use λ = 0.75.
(e) Final development length Ld.
ψt ψe 1.7
Ld = 53.5 = 53.5 = 121 in.
λ 0.75
(f) Compute development length Ld using the general equation, Eq. (6.6.1).
 

3 f ψ t ψ e ψ s 
Ld = 
y
d [6.6.1]
 40 λ fc′  cb + K tr   b

 d  
 b 
There are no stirrups; thus K tr = 0. The value of cb for Eq. (6.6.1) is the smaller of the
cover (i.e., the distance from the center of the bar to the nearest concrete face) or one-
half the center-to-center spacing of the bars being developed. In this case,
cover = 1.5 +1.128 /2 = 2.06 in. Controls
(center-to-center spacing) / 2 = 8 / 2 = 4 in.
Thus,
 cb + K tr 2.06 + 0 
 d = = 1.83 ≤ 2.5 [6.6.3]
b 1.128 
and from Eq. (6.6.1)

 3 60, 000 1.7(1.0) 
Ld =  d = 88.1db
 40 0.75 4000 (1.83)  b
(Continued)
194
In the above calculation,ψ t ψ e = 1.7, the upper limit of that product, which exceeds the
actual ψ t ψ e = 1.3 (1.5) = 1.95. The bar size factor ψ s = 1.0 for #7 bars and larger. Thus,
Ld = 88.1 db = 88.1(1.128) = 99.4 in.
As noted earlier, the simplified method (Ld = 121 in.) is conservative and yielded a
development length 22% greater than that required by the general Eq. (6.6.1). In the
simplified equations, the term ( cb + K tr ) / db is taken as 1.5 for all cases, whereas Eq.
(6.6.3) yielded a value of 1.83 in this example.
6.8 DEVELOPMENT LENGTH FOR COMPRESSION

REINFORCEMENT
Relatively less is known about the development length for compression bars than for ten-
sion bars, except that the weakening effect of flexural tension cracks is not present and that
there is beneficial effect of the end bearing of the bars on the concrete. ACI-25.4.9.2 gives
the development length of deformed bars and wires in compression as the larger of
 fy ψ r 
Ldc =   db (6.8.1)7
 50 λ fc′ 
which is basically one-third of the minimum development length for tension reinforcement
to prevent a “pullout” mode of failure, or
7
Ldc = 0.0003 f y ψ r db (6.8.2)
which means that only fc′ up to about 4400 psi may be counted on for normal-weight concrete.
The modification factor λ to be used in Eq. (6.8.1) is the same as that described in
Section 6.7 for bars in tension.
A reduction in development length is permitted by the modification factor, ψr, when
reinforcement is enclosed by transverse reinforcement as follows:
i) a spiral (as specified in ACI-25.7.3; see Chapter 10),
ii) a continuously wound circular tie of not less than 1 4 -in. diameter and having a pitch
not exceeding 4 in.,
iii) #4 bar or D20 wire ties (as specified in ACI-25.7.2; see Chapter 10) with center-
to-center spacing not exceeding 4 in., or
iv) hoops (as specified in ACI-25.7.4; see Chapter 10) with center-to-center spacing not
exceeding 4 in.
Under these confinement conditions, Ldc may be reduced by 25% (i.e., ψ r = 0.75).
When excess bar area is provided such that provided As exceeds required As, Eq. (6.8.1)
or (6.8.2), whichever controls, may be reduced by applying the multiplier (αexs = required
As /provided As).
After all modifications, the development length Ldc is not permitted to be less than 8 in.
Thus, in general, for compression reinforcement,
 Eqs. (6.8.1)  required As  0.75 for enclosure

Ldc =     ≥ 8 in. (6.8.3)
 or (6.8.2)   provided As   by spirals or ties 
7 In Eqs. (6.8.1) and (6.8.2), for SI, with Ldc and db in mm, and fc′ and fy in MPa, ACI-25.4.9.2 gives, respectively
 0.24 f y ψ r 
Ldc =   db (6.8.1)
 λ fc′ 
Ldc = 0.043 f y ψ r db (6.8.2)
195
6.10 DEVELOPMENT LENGTH FOR A TENSION BAR 195
6.9 DEVELOPMENT LENGTH FOR BUNDLED BARS

When space for proper clearance is restricted and large steel areas are required, groups
of parallel bars are sometimes bundled. No more than four bars bundled in contact (ACI-
25.6.1.1), enclosed by stirrups or ties, may be arranged with no more than two bars in the
same plane into typical bundle shapes, such as triangular, square, or L-shaped for three-and
four-bar bundles (see Fig. 6.9.1). Bars larger than #11 shall not be bundled in beams or
girders, primarily to ensure proper control of cracking (see ACI Commentary R25.6.1.3).
In flexural members, termination of individual bars within a bundle at points along the span
must be offset by at least 40 bar diameters (ACI-25.6.1.4). Where spacing requirements
and minimum clear cover are based on bar size, one bundle of bars shall be treated as a
single bar of an equivalent diameter derived from the total area of the bars in the bundle. In
applying the crack control provisions of ACI-24.3.2 (see Chapter 12) to bundled bars, this
assumption of treating a bundle of bars as a single large bar may be overly conservative. An
alternative is suggested by Lutz [6.35].
A bundle of bars in the three-bar triangular pattern and the four-bar arrangement will
have 16 2 3 and 25% reduction, respectively, in the total nominal surface contact (consider-
ing all bars cylindrical and using the nominal diameter) of the bars with surrounding con-
crete. Further, to allow for the difficulty of attaining good resistance to splitting or pullout
at the reentrant corner where the bars in the bundle touch each other, additional reduction
in resistance appears logical.
The requirements of ACI-25.6.1.5 specify that the development length for a bundle shall
be based on that for the individual bar in the bundle, increased by 20% for a three-bar bun-
dle and 33% for a four-bar bundle. For determination of all needed modification factors,
ACI-25.6.1.6 states, “A unit of bundled bars shall be treated as a single bar with an area
equivalent to that of the bundle and a centroid coinciding with that of the bundle.”
Experimental results for beams and columns having bundled reinforcement are reported
by Hanson and Reiffenstuhl [6.36]; some practical applications are given by Steiner [6.37].
Figure 6.9.1 Bundled bar arrangements (bars are in contact with each other).
6.10 DEVELOPMENT LENGTH FOR A TENSION BAR

TERMINATING IN A STANDARD HOOK
When straight embedment is inadequate to provide for the necessary development length of
a tension bar, or when it is desired to have the full capacity of a bar available in the shortest
distance of embedment, a standard hook as defined in ACI-25.3.1 and shown in Fig. 6.10.1
may be used for the main reinforcement. Standard hook geometry for stirrups, ties, and
hoops are defined in ACI-25.3.2. Hooks are not considered effective in adding to the com-
pressive resistance of reinforcement.
In general, the resistance of a hook in mass concrete is about the same as that of a straight
bar with the same total length of embedment [6.38, 6.39]. Quite commonly, hooks in structural
members are located close to a free surface where splitting forces proportional to the tensile
capacity of the bar will govern the hook capacity. Because of the tendency for splitting failures,
hooks may have less capacity than provided by an equal length of straight embedment.
db D, diameter of bend D Minimum diameter of bend
T T D ≥ 6db for #3 to #8
D ≥ 8db for #9 to #11
D ≥ 10db for #14 to #18
12db
4db ≥ 2 21 ”
(a) 180° hook (b) 90° hook
Figure 6.10.1 Standard hooks for development of main tension reinforcement.

196
db
T
Critical section
where full tensile
capacity of bar is db 12db straight
available
T
db
4db ≥ ≥ 4db #3 through #8

2 21 ” ≥ 5db #9, #10, #11
≥ 6db #14, #18
Ldh
Figure 6.10.2 Development length Ldh for hooked bar.
ACI Code Procedure

Based on the works of Hribar and Vasko [6.38], Orangun, Jirsa, and Breen [6.3], Minor
and Jirsa [6.39], Marques and Jirsa [6.40], and Pinc, Watkins, and Jirsa [6.41], the ACI
Code prescribes in a direct manner the required embedment (development length Ldh as in
Fig. 6.10.2) to develop fy in a hooked bar, uncoupled from the embedment required (devel-
opment length Ld) to develop fy in a straight bar. A summary of the rationale for this proce-
dure is given in Jirsa, Lutz, and Gergely [6.5].
According to ACI-25.4.3.1, the development length for deformed bars in tension with a
standard hook, Ldh, is computed as
 fy ψ e ψ c ψ r 
Ldh =   db (6.10.1)8
 50 λ fc′ 
but not less than 8db nor less than 6 in.

For epoxy-coated or dual-coated zinc and epoxy reinforcement, ψ e is taken as 1.2 and
λ is taken as 0.75 for lightweight aggregate concrete. For other cases, ψ e and λ are taken
as 1.0.
Increased cover and stirrups that confine the concrete within Ldh will improve the perfor-
mance of the hook. For this reason, ACI-25.4.3.2 allows Ldh to be reduced by the factors ψ c
and ψ r when the conditions described in Table 6.10.1 are met. Similar to the provisions for
development length of straight bars, Ldh is permitted to be reduced when there is reinforce-
ment in excess to that required by analysis (see Table 6.10.1).
Most hooked bars are embedded in joints where other concrete members frame in at
the sides and therefore provide some lateral confinement, in addition to the vertical con-
finement provided by the force in the column. When these other members are not present,
such as at the discontinuous end of a cantilever as shown in Fig. 6.10.3, ACI-25.4.3.3
requires either 2 1 2 in. or more of side and top (or bottom) cover to the hooked bar, or that
the hooked bar be enclosed within stirrups or ties perpendicular to the bar being developed
and spaced not greater than 3db along Ldh. It is important to place the first tie or stirrup as
close to the corner of the hook as possible, and thus ACI-25.4.3.3 requires that the first
tie or stirrup enclose the bent portion of the hook within 2db of the outside of the hook
8 For SI, for Ldh and db in mm, and fc′ in MPa, ACI-25.4.3.1 gives
 0.24 f y ψ e ψ c ψ r 
Ldh =   db (6.10.1)
 λ fc′ 
197
6.10 DEVELOPMENT LENGTH FOR A TENSION BAR 197
TABLE 6.10.1 FACTORS TO MODIFY BASIC DEVELOPMENT LENGTH Ldh FOR A HOOKED BAR
(ACI-25.4.3.2)
Condition Modification Factor ψc and ψr
1. Concrete cover
90 or 180º hooks, #11 bars and smaller having side cover (normal to ψ c = 0.7
plane of hook) ≥ 2 1 2 in.; or
90º hooks, with clear cover on bar extension beyond hook ≥ 2 in.
2. Confining reinforcement
90 or 180º hooks of #11 and smaller bars enclosed within ties or ψ r = 0.8
stirrups perpendicular to the bar being developed and spaced not
greater than 3db (db = diameter of hooked bar) along Ldh; or
90º hooks of #11 and smaller bars enclosed within ties or stirrups
parallel to the bar being developed and spaced not greater than 3db
(db = diameter of hooked bar) along the length of the tail extension
of the hook plus bend. [Note: Multiplier cannot be used if the hook
occurs at the discontinuous end of a member when side and top or
bottom clear cover to hooked bar is less than 2 1 2 in. (ACI-25.4.3.3).]
3. Excess reinforcement
Where anchorage or development for fy is not specifically required, required As
α exs = ≤ 1.0
and there is reinforcement in excess of that required provided As
Clear cover to Ldh

hooked bar < 2 21 ” A
Standard hook at
discontinuous end
of member
First tie or stirrup

must be placed
within 2 diameters
A
of the hooked bar
< 2 21 Spacing
” Hooked bar must be
≤ 3db
enclosed within ties
or stirrups, spaced
< 2 21 ” not farther apart
than 3 diameters
Side and top (or of the hooked bar
bottom) clear
cover to hooked
bar less than 2 21 ”
Section A–A
Figure 6.10.3 Special requirements for standard hooks developed at the discontinuous end of a
member (ACI-25.4.3.3).
tail (db = diameter of the hooked bar) as shown in Fig. 6.10.3. As noted under item 2 of
Table 6.10.1, the reduction factor attributed to ties or stirrups cannot be used if hooks are
used at the discontinuous end of a member with clear side and top (or bottom) cover over
the hook less than 2 1 2 in.
Values of the development length Ldh, according to Eq. (6.10.1), for f y = 60, 000 psi are
shown in Table 6.10.2.
198
TABLE 6.10.2 DEVELOPMENT LENGTH Ldh FOR HOOKED GRADE 60

BARS,a EQ. (6.10.1)
Inch-Pound Bars with Ldh in Inches
fc′ (psi)
Bar 3000 4000 5000

#3 8.2 7.1 6.4
#4 11.0 9.5 8.5
#5 13.7 11.9 10.6
#6 16.4 14.2 12.7
#7 19.2 16.6 14.8
#8 21.9 19.0 17.0
#9 24.7 21.4 19.1
#10 27.8 24.1 21.6
#11 30.9 26.8 23.9
#14 37.1 32.1 28.7
#18 49.4 42.8 38.3
a Assumes normal-weight concrete (λ = 1.0) and ψe = ψc = ψr = 1
6.11 BAR CUTOFFS IN NEGATIVE MOMENT REGION

OF CONTINUOUS BEAMS
The general concept of drawing the flexural strength diagram to investigate the adequacy of
a given beam was presented in Section 6.5. In this and the next two sections, several situa-
tions are discussed in which the ACI Code provisions are applied to establish cutoff points.
Three factors are involved: (1) adequate horizontal offset from theoretical cutoff points in
recognition of the truss action described in Section 5.7, where the tensile force is shown to
be greater than that calculated at the section where Mu is computed due to the presence of
shear, as well as to provide safety against the possible shifting of the moment diagram due
to unusual loading arrangements; (2) adequate development lengths so that full bar capacity
is available where needed; and (3) sufficient relief from stress concentrations when bars are
terminated in a tension zone.
Cutoffs in the negative moment region of continuous beams can best be explained by
means of a specific case. (For an example of a complete factored moment Mu envelope for
a continuous beam, the reader is referred to Fig. 9.6.2.) Referring to the partial envelope for
Mu in Fig. 6.11.1, assume that it is desired to cut two out of four bars as close as possible
(point C) to the support, and then terminate the remaining two as soon as feasible (point
E). In accordance with ACI-9.7.3.8.4 for beams (and ACI-7.7.3.8.4 for one-way slabs), the
area provided by bars R2 must be at least one-third of the total reinforcement area provided
for negative moment. In a general situation, the horizontal distance between points C and
E is greater than the development length Ld for bars R2; the special case wherein this is not
so will be discussed later.
Point A represents the maximum factored moment Mu that must be resisted by the beam
cross section at the face of support, the critical location for bending moment. The design
strength, φ M n , provided by bars R1 and R2 is somewhat greater than required. The theo-
retical cutoff for the two R1 bars is at point B. Point C is then located horizontally to the
right of point B a distance equal to the greater of the effective depth of the member and 12
diameters of bars R1 (ACI-9.7.3.3). Point D is located horizontally to the left of point C a
distance equal to the development length Ld for bars R1. To provide adequate capacity at the
face of support, point D must lie at or to the right of point A.
Point E is next established by extension, according to ACI-9.7.3.8.4, beyond the point
of inflection (point of zero moment) not less than the effective depth of the member, d,
19
6 . 1 1 BA R C U TO F F S I N N E G AT I V E M O M E N T R E G I O N 199
1
Bars R2
4 bars 2 bars, R2
E
C
Bars R1
d
Section 1–1
1 2 bars, R1 cut here
Mu = factored moment diagram
E
Largest of
clear span
d, 12db, (ACl–9.7.3.8.4)
16
Bars R2
Point C must also satisfy ACl–9.7.3.5

φMn
Bars R1 + R2
φMn
B C
F
Larger of
12db or d
(ACl–9.7.3.3) Ld for R2 φMn diagram
A
D Horizontal distance AD ≥ 0
Ld for R1 (i.e., horizontal distance AC ≥ Ld for R1), preferably
Face of support
Figure 6.11.1 Bar cutoffs in negative moment region of continuous beams wherein the distance C-E is greater
than Ld for bars R2.
12db, or one-sixteenth the clear span, whichever is greater. Point F, where the bars R2 will
become capable of carrying their full capacity, is located at a distance Ld horizontally to the
left of point E.
In accordance to ACI-9.7.3.4, continuing flexural tension reinforcement (bars R2) must be
extended at least Ld beyond the point where bent or terminated tension reinforcement is no
longer required to resist flexure (i.e., point B). In the situation shown in Fig. 6.11.1, this require-
ment is readily satisfied, since point F lies to the right of point B. When a longer development
length Ld is required for bars R2, however, point F may lie to the left of point C, as shown in
Fig. 6.11.2. The design strength, φ M n diagram is indicated by EC′F′D. The variation between
the design flexural strengths at C′ and F′ is linear since development of flexural strength is
assumed to be proportional to the amount of embedment up to Ld. Note that φ M n between
point E and point F′ is less than that shown in Fig. 6.11.1. Nonetheless, the requirement that
the horizontal distance B-E be greater than or equal to F-E (i.e., Ld for bars R2) ensures that the
design strength diagram, φ M n , does not encroach the factored moment diagram, Mu.
Cutting Bars in the Tension Zone

When bars are terminated in a tension zone, the shear strength of the member may be
reduced. Since point C in Fig. 6.11.1 is still in the tension zone (though only slightly), the
provisions of ACI-9.7.3.5 for beams (and ACI-7.7.3.5 for one-way slabs) must be checked
in order to minimize the potential for a diagonal crack to develop or to ensure that sufficient
shear reinforcement is provided should a diagonal crack form. One of the following three
conditions must be satisfied at point C.
1. The factored shear Vu at the cutoff point does not exceed two-thirds of the design
shear strength φ Vn, or
2 2 
Vu ≤  φVn = φ (Vc + Vs ) (6.11.1)
3 3 
20
4 bars 2 bars, R2
E
C
2 bars, R1
Factored
moment,
Mu E
φMn’ Bars R2
Hor. dist. BF ≥ 0
(i.e., horizontal distance
φMn’ Bars R1 and R2
BE ≥ Ld for R2, ACI–9.7.3.4)

C’
M
B C
F
F’
Ld for R2
A φMn diagram
D
Ld for R1
Figure 6.11.2 Special case of Fig. 6.11.1 wherein the distance C-E is less than Ld for bars R2.
2. Stirrup or hoop area of at least 60bw s / f yt is provided in excess of that required for
shear and torsion. The excess stirrups are to be used along the terminated bar over a
distance from the termination point equal to three-fourths of the effective depth of the
member, d.
The spacing of such stirrups must not exceed
d
max s = (6.11.2)
8βb
where βb is the ratio of area of longitudinal bars cut off to the total area of bars at the
section.
3. For #11 and smaller bars, the following is satisfied at the cutoff point:
3 3 
Vu ≤  φVn = φ (Vc + Vs ) (6.11.3)
4 4 
and
[φ M n of continuing reinforcement] ≥ 2 Mu (6.11.4)
When it may be impractical or undesirable to satisfy the above-described provisions

of ACI-9.7.3.5, the cutoff point should be extended until it is in the compression zone.
201
6.12 BAR CUTOFFS IN POSITIVE MOMENT REGION 201
6.12 BAR CUTOFFS IN POSITIVE MOMENT REGION

OF CONTINUOUS BEAMS
Bar cutoffs in the positive moment region of continuous beams are explained by reference
to Fig. 6.12.1. Assume that it is desired to cut two of the four bars that are used for resist-
ing the maximum positive factored moment Mu. The area provided by bars R4 must exceed
one-fourth (one-third for simple spans) of the total reinforcement area provided for positive
moment, in accordance with ACI-9.7.3.8.2.
Point G is first located by computing the design strength, φ M n , of the continuing bars
R4. The cutoff point H for bars R3 must lie to the left of point G at least 12 diameters of
bars R3 or the effective depth d, whichever is larger (except at the support) in accordance
with ACI-9.7.3.3. Point I is then located horizontally to the right of point H a distance equal
to the development length Ld for the bars R3.
Bars R4 must extend into the support a minimum distance of 6 in. from the face of the
support (point J) in accordance to ACI-9.7.3.8.2. If the member, however, is part of the pri-
mary lateral load resisting system, the bars must be anchored to develop the tensile yield
strength at the face of the support. Point K is located horizontally to the right of point J, a
distance equal to the development length Ld for the bars R4.
In accordance with ACI-9.7.3.4, continuing flexural tension reinforcement (bars R4) must
have a development length at least Ld beyond the point where bent or terminated tension rein-
forcement is no longer required to resist flexure (i.e., point G). In the general situation shown in
Fig. 6.12.1, this requirement is readily satisfied because point K lies to the left of point G (and
also point H). In some cases, however, a larger Ld may be required for bars R4 and point K may
lie to the right of point H as shown in Fig. 6.12.2. The φ M n diagram is indicated by JH ′ K ′ I ,
where the variation between the design strengths at H′ and K′ is linear because development
of flexural strength is assumed to be proportional to the amount of embedment up to Ld. In this
case, ACI-9.7.3.4 is satisfied by making sure that point K lies at or to the left of point G (i.e.,
horizontal distance KG ≥ 0). Note that the design strength, φ M n , between point J and K′ is
less than that shown in Fig. 6.12.1, but it is still greater than the required factored moment, Mu.
In addition, since the cutoff at point H lies in the tension zone, one of the conditions of
ACI-9.7.3.5 must be satisfied [see Eqs. (6.11.1)–(6.11.4).]
≥ 6”, ACl–9.7.3.8.2 Bars,

d R4
4 bars
2 bars, R4
J H
2 bars, R3 cut here Bars, R3

2
Section 2–2
Ld for R3
Hor. dist. KG ≥ 0
d or 12db, ACI–9.7.3.3
(i.e., hor. dist. JG ≥ Ld,
ACI–9.7.3.4) I φMn diagram
Ld for R4
K H G Factored moment diagram, Mu
φMn, Bars R3 and R4
φMn, Bars R4
J
Point H must also satisfy ACI–9.7.3.5
Figure 6.12.1 Bar cutoffs in positive moment region of continuous beams wherein the distance
J-H is greater than Ld for bars R4.
20
2 bars, R4
2 bars, R3
J H
4 bars
Hor. dist. KG ≥ 0
Ld for R3 (i.e., hor. dist. JG ≥ Ld for R4,
ACI–9.7.3.4)
Factored
moment
diagram,
K′
φMn, Bars R3 and R4

Mu
N K
H
G
φMn, Bars R4
H′
J
Ld for R4
Figure 6.12.2 Special case of Fig. 6.12.1 wherein the distance J-H is less than Ld for bars R4.
6.13 BAR CUTOFFS IN UNIFORMLY LOADED

CANTILEVER BEAMS
Bar cutoffs in uniformly loaded cantilever beams can best be discussed be examining
Fig. 6.13.1. Assume that of the six bars provided for the maximum factored moment Mu, it
is desired to cut two bars R5 as close to the support as possible, then cut two more bars R6,
and run the remaining two bars R7 out to the end of the cantilever.
The following steps illustrate the cutoff determination and check:
1. Locate the theoretical cutoff point A where the design strength φ M n of four bars (bars
R6 and R7) is adequate; extend 12 diameters of bars R5 or the effective depth of the
member, whichever is greater, to arrive at point B.
2. Determine the development length Ld for the bars R5 that are intended to be cut. Full
capacity from the R5 bars is available at the distance Ld to the left of the cutoff point.
3. Since the horizontal distance CB is less than Ld, less than full capacity of the R5 bars
will be available at the support had they been cut at point B; therefore, extend cutoff
location to point D so that the horizontal distance CD equals Ld. (ACI-9.7.3.5 must
also be satisfied, since the proposed cut location lies in a tension zone.)
4. Locate point E, the theoretical location where only the two R7 bars are required for
moment; extend 12 diameters of bars R6 or the effective depth of the member, which-
ever is greater, to arrive at point F. (ACI-9.7.3.5 must also be satisfied for point F,
since it lies in a tension zone.) Check that distance from point E to end of bars R7 is
at least Ld (ACI-9.7.3.4).
5. Determine the development length Ld required for the R6 bars being cut; these two
bars will have their full capacity available at point G, a distance Ld to the left of
point F. Check that point F is at least Ld of bars R6 from point A (ACI-9.7.3.4).
6. In the region from G to D, the flexural strength consists of partial contributions from
the bars R5 and R6 plus the full contribution of bars R7. The combined flexural
203
6 . 1 3 B A R C U T O F F S I N U N I F O R M LY L O A D E D C A N T I L E V E R B E A M S 203
3 Bars, R5
6 bars Bars, R6
2 bars, R7
d
Bars,
R7
2 bars R5 Section 3–3
cut here 2 bars R6
cut here
3
Ld for Factored moment

R6 bars diagram, Mu
φMn, 2 bars
φMn, 4 bars
φMn, 6 bars
(R7)
(R6 and R7)
(R5, R6 and R7)
12db or d
φMn
E F
diagram
H Ld for
A Points H and F R7 bars
D must also satisfy
G
B ACI–9.7.3.5
I
Extra extension so full capacity is
C available at point C
Ld for 12db or d, ACI–9.7.3.3

R5 bars
Figure 6.13.1 Bar cutoff in a uniformly loaded cantilever beam.
strength (represented by line IH) is the sum of the linear contributions. Should any
part of the design strength, φ M n , encroach on the factored moment diagram, Mu, then
the proposed cut locations (points D and F) must be extended toward the free end of
the cantilever.
► EXAMPLE 6.13.1
For the cantilever beam shown in Fig. 6.13.2 determine the distance L1 from the sup-
port to the point where 2–#8 bars may be cut off. Assume the #4 stirrups shown (solid,
not dashed) have been preliminarily designed. Assume there will be at least Ld embed-
ment of the bars into the support. Draw the resulting φ M n diagram for the entire beam.
Use normal-weight concrete with fc′ = 3000 psi and uncoated (black) steel reinforce-
ment with f y = f yt = 60, 000 psi. Assume 1 1 2 in. clear cover to the stirrups.
SOLUTION
(a) Compute the maximum design strength φ M n of the section.
C = 0.85(3)16 a = 40.8a
T = [3(1.27) + 2(0.79)]60 = (3.81 + 1.58)60 = 323 kips
323
a= = 7.92 in.
40.8
1
M n = 323 [28 − 0.5(7.92)] = 647 ft-kips
12
(Continued)
204
First stirrup must be

placed within 2db
L1 = 4’– 6” Ldh = 27.8”
(2.5 in.) of the outside
Revised stirrup spacing bend
to satisfy cutoff in 8” 8 @ 3.5 = 2’– 4” 3 – #10
tension zone #4 stirrups
28” 2 – #8
Given spacing 3 @ 14” #4

2” 4 @ 5” 3 @ 8” 16”
U stirrups
8’–0” 1
12 ”
Factored moment Mu
Improbable that
hook develops
Factored moment, ft-kips
φMn = 433 ft-kips

Mu = 1.2 MD + 1.6 ML
capacity in a
with 3 – #10
linear manner
Moment capacity diagram, φMn
d = 2.33’ Extension to provide adequate

A B C C′ capacity at support
Ldh = 2.32’
D
D Development length
590’ k for #10 hooked bar
Cut 2 – #8 bars
Ld = 4.1’ (Ldh = 27.8”)
Vu = 1.2 VD + 1.6 VL
Factored shear, kips
147.5 k
Extension
to satisfy
ACI–9.7.3.5
71.9 k 64.5 k
Figure 6.13.2 Beam of Example 6.13.1.
Compute the net tensile strain at the steel level

d − a / β1 28 − 7.92 / 0.85
εt = (0.003) = (0.003) = 0.006 > 0.005
a / β1 7.92 / 0.85
Thus, the section is tension controlled and φ = 0.90:

φ M n = 0.90 (647) = 582 ft-kips ≈ Mu = 590 ft-kips OK
(b) Determine the theoretical cutoff point for the 2–#8 bars. The remaining design
strength φ Mn with 3–#10 bars is
C = 40.8a
T = 3.81(60) = 229 kips
229
a= = 5.61 in.
40.8
1
φ M n = 0.90(229)[28 − 0.5(5.61] = 433 ft-kips
12
(Continued)
205
Plot φ M n on the factored moment Mu diagram and locate the theoretical cutoff point A.
Extend to the right 12 bar diameters (of the #8 bars that are to be cut) or the effective
depth of the member, whichever is greater, to arrive at point B.
d = 28 in. ( 2.33 ft ) > [12 db = 12 (1.0 ) = 12 in.]
(c) Use the simplified equations to determine the development length Ld for #8 bars.
Can Category A, the more favorable one, be used? Check the clear spacing of bars.
Assuming the bars, though unequal in size, are uniformly spaced, the clear spacing
between them is
16 − 2 (1.5) − 2 (0.5) − 3(1.27) − 2 (1.0)
clear spacing = = 1.55 in.
4
Since only the 2–#8 bars are being developed, and the 3–#10 are presumed to continue
beyond the #8 cutoff location, it is the spacing between the two #8 bars that determines
the category. The possible failure modes to consider are splitting from a #8 bar to the
side or top face of the member and splitting between the two #8 bars. The ACI Code
rules consider a bar (or bars) as essentially inert when it is not being developed within
the development region of other bars. Thus, when the #10 bars of this example have a
development length from their termination near the free end of the cantilever that is less
than the distance to the #8 bar cut, the #10 bars are considered to have no influence on
Ld for the #8 bars. That is, in this case the full spacing between the #8 bars, 2(1.55) +
1.27 diam of #10 = 4.37 in. The writers believe it appropriate in this case to consider the
spacing of the #8 to be 4.37 in. for the purpose of satisfying a Category A requirement,
assuming Ld for the #10 bars does not overlap the Ld for the #8 bars.
Even if the concrete width between #8 bars were taken as 2(1.55) = 3.10 in., it would
still exceed the 2db for the #8 bars to satisfy category A, item 2(a), given in Section 6.6,
as well as item 2(b), because cover to the top face of the beam is 2 in., which exceeds db
needed for that item.
Thus, Category A applies! Using simplified Eq. (6.6.5) for #7 and larger bars,
Ld =   db [6.6.5]
 20 λ fc′ 
The casting position factor ψ t = 1.3 applies, while both the epoxy-coated bar factor ψ e
and the lightweight aggregate concrete factor λ are 1.0. Thus, Eq. (6.6.5) gives
 60, 000 ψ t ψ e  ψψ
Ld =   1.0 = 54.8 t e
 20 λ 3000  λ
ψt ψe (1.3)(1.0)
Ld = 54.8 = 54.8 = 71.2 in.
λ 1.0
The 54.8 in. can be verified from Table 6.6.1. Thus,
Ld ( for # 8) = 71.2 in. ( 5.9 ft )
(d) Use the general equation, Eq. (6.6.1), to determine the development length Ld for #8
bars. That equation is
 

Ld =  d [6.6.1]
 40 λ fc′  cb + K tr   b

 d  
 b 
(Continued)
206
The cover or spacing dimension cb is the smaller of (1) the distance from the center
of the bar being developed to the nearest concrete surface and (2) one-half the center-
to-center spacing of the bars being developed. The distance cb is the smaller of the fol-
lowing two values:
top cover = 1.5 (i.e., clear) + 0.5 (i.e.,stirrup) + 0.5 ( i.e., bar radius) = 2.5 in.
one-half center-to-center spacing = 4.37 [see part (c)]/2 = 2.19 in.
Thus, one-half the center-to-center spacing governs and cb = 2.19 in. This result indi-
cates a potential failure mode that involves horizontal splitting. In this case, the trans-
verse reinforcement is considered to be effective only for the two #10 corner bars [see
Fig. 6.6.1(a)], which are not being developed at this location. Whether the transverse
reinforcement will provide some confinement to the concrete and help improve the bond
strength of the center bars is a matter of opinion. It is clear, however, that if horizontal
splitting occurs, the stirrups will not cross the horizontal splitting plane adjacent to the
#8 bars being developed. Therefore, the contribution of the transverse reinforcement can
be conservatively ignored in this case (i.e., Ktr = 0).
Evaluating Eq. (6.6.3),
 cb + K tr 2.19 + 0 
 = = 2.19 < 2.5 max
 db 1.0 
Thus, ( cb + K tr ) / db = 2.19. Evaluate Eq. (6.6.1),

 

3 fy ψ t ψ e ψ s   3 60, 000 1.3(1.0)1.0 
Ld =  db =  1.0 = 48.8 in. (4.1 ft )
 40 λ fc′  cb + K tr    40 1.00 3000 2.19 

 d  
 b 
This value is significantly lower than the 71.2 in. obtained from the simplified equation
in part (c). Use Ld = 48.8 in. for the design strength diagram in Fig. 6.13.2.
Since point B, the proposed cutoff point, lies only about 3.5 ft from the support, the
#8 bars would not have full capacity at the support. Therefore, extend the proposed cut-
off to point C, which is located at Ld (for #8) = 4.1 ft from the support.
(e) Check ACI-9.7.3.5 for cutting bars at point C in the tension zone. The shear strength,
including contribution of stirrups, is first computed. Using the simplified method of
constant Vc with λ = 1.0,
1
Vc = 2 λ fc′ bw d = 2(1) 3000 (16)(28) = 49.1 kips
1000
For the 14-in. spaced #4 stirrups in the vicinity of the potential cut point C,
Av f yt d 2(0.20)(60)28
Vs = = = 48.0 kips
s 14
The shear strength φ Vn at point C is

φ Vn = φ (Vc + Vs ) = 0.75(49.1 + 48.0) = 72.8 kips
Vu 71.9
= = 99% > 75% NG
φ Vn 72.8
(Continued)
207
Even when only 50% of the design strength φ M n is used by Mu, Vu /(φVn) cannot exceed
75% [see Condition 3, preceding Eqs. (6.11.3) and (6.11.4)]. Try using one more stirrup
at 8-in. spacing to cover the potential cut at point C, and see whether or not Condition
1, Eq. (6.11.1), is satisfied.
 14 
Vs = 48.0   = 84.0 kips
 8
Vu 71.9
= = 72%
φ Vn 0.75(49.1 + 84)
This is close to satisfying the two-thirds limit of ACI-9.7.3.5(a) [Eq. (6.11.1)]. Extending
the #8 bars to point C′, 4.5 ft from the face of support, would result in
Vu 64.5
= = 65% OK
φ Vn 0.75(49.1 + 84)
(f) Check whether the continuing #10 bars have adequate development length to the
right of point C′. The clear spacing between the continuing three #10 bars is
16 − 2(1.5) − 2(0.5) − 3(1.27)
clear spacing = = 4.1 in.
2
which exceeds the 2db = 2.54 in. required for Category A, item 2(a). Top cover of 2.64 in.
[i.e.,(1.5 + 0.5 + 1.27 / 2 ) = 2.64 in.] to the center of the #10 bars exceeds the db require-
ment of Category A, item 2(b). Thus, the simplified equation, Eq. (6.6.5) for #7 and
larger bars, can be used
 fy ψ t ψ e   60, 000 ψ t ψ e  ψψ
Ld =   db =   1.27 = 69.6 t s
 20 λ fc′   20 λ 3000  λ
ψt ψe 1.3(1.0)
Ld = 69.6 = 69.6 = 90.5 in.
λ 1.0
where a casting position factor ψ t = 1.3 has been used. Both the epoxy-coated bar factor
ψ e and the lightweight aggregate concrete factor λ are 1.0.
Calculate the development length Ld based on the general equation, Eq. (6.6.1). The
distance cb is the smaller of the following two values:
top and side cover = 1.5 (i.e., clear) + 0.5 (i.e., stirrup) + 0.635 (i.e., bar radius) = 2.6 in.
4.1
one-half center-to-center spacing = + 0.635 (i.e., bar radius) = 2.7 in.
2
Although the cover parameter for top and side cover splitting with cb = 2.6 in. would
govern, the difference between the two computed values is too small to predict the split-
ting failure plane with certainty. Here, the parameter will be computed assuming that
both top and side cover splitting and horizontal splitting may occur.
If top and side cover splitting is assumed, then the transverse reinforcement may
be assumed to be effective for all three bars being developed [as in Fig. 6.6.1(b) and
6.6.1(d)]. Thus, the total area of transverse reinforcement is Atr = 3 Aleg for three bars
being developed: that is, n = 3. Using the given 14-in. spacing near the free end of the
cantilever for computation, Eq. (6.6.2) gives
40 Atr 40(3)(0.20)
K tr = = = 0.57
sn 14(3)
(Continued)
208
If horizontal splitting is assumed [as in Fig. 6.6.1(a)], then the transverse reinforcement

is effective for only two of the three bars being developed. Thus Atr = 2 A leg and n = 3.
Eq. (6.6.2) gives then
40 Atr 40(2)(0.20)
K tr = = = 0.38
sn 14(3)
which is more conservative. Use K tr = 0.38.

 cb + K tr 2.7 + 0.38 
 = = 2.43 < 2.5 max
 db 1.27 
Thus, ( cb + K tr ) / db = 2.43.
Evaluate Eq. (6.6.1),
 

Ld =  d
 40 λ fc′  cb + K tr   b

 d  
 b 
 3 60, 000 ψ t ψ e ψ s  ψψ ψ
= 1.27 = 42.9 t e s
 40 λ 3000 2.43  λ
(1.3)(1.0)(1.0)
= 42.9
1.0
= 55.8 in. (4.65 ft )
This embedment of 4.65 ft measured from the end of straight #10 bars would overlap the
development length region of the #8 bars, possibly requiring longer development length Ld
for the #8 bars because the center-to-center spacing then would be the reduced value based
on five bars in the 16-in. width. The #10 bars would satisfy literally the statement of ACI-
9.7.3.4, which requires “Continuing flexural tension reinforcement shall have an embed-
ment length at least Ld beyond the point where bent or terminated tension reinforcement is
no longer required to resist flexure.” In other words, the distance from point A to the free
end of the cantilever must be at least Ld (for #10 bars). The authors believe in a somewhat
more conservative approach, requiring the design strength φ Mn diagram to have an offset
from the factored moment Mu diagram, except at or near a simple support or the free end
of a cantilever, equal to 12 bar diameters or the effective depth d, whichever is greater.
In this case, try standard 90º hooks [see Fig. 6.10.1(b)] on the ends of the #10 bars.
Since the beam has the usual 1.5-in. clear cover and #4 stirrups, the cover to the hooked
bars is 2 in., which is less than the 2 1 2 in. required by ACI-25.4.3.3; thus, the special
provisions of that code section must be satisfied.
The development length Ldh for the #10 hooked bar is given by Eq. (6.10.1) and
Table 6.10.2. Thus, for a #10 hooked bar,
Ldh = 27.8 in.
which exceeds the minimum 8db or 6 in., whichever is greater (ACI-25.4.3.1). The Ldh
of 27.8 in. is dimensioned from the outside face of the tail of the hook, as shown in
Fig. 6.13.2. In accordance with ACI-25.4.3.3, stirrups spaced at not more than 3db (3.81 in.)
must be provided along the 27.8 in. of development distance. Also, the first stirrup must
be provided within 2db (2.54 in.) of the outside bend, as shown in Fig. 6.13.2.
(Continued)
209
6.14 DEVELOPMENT OF POSITIVE REINFORCEMENT 209
(g) Design strength φ M n diagram. The full design strength, φ M n , for the beam with
3–#10 hooked bars will be available at 27.8 in. from the outside of the hook on
the end of the beam. Assuming 1.5-in. cover, full capacity is available at 29.3 in.
(2.44 ft) from end of beam (point D). The dashed line in Fig. 6.13.2 goes from zero
strength at the end of the hook to full strength φ M n = 433 ft-kips at 2.44 ft from end
of beam; however, the drawing is not intended to imply that the hooked bar develops
its strength linearly, since that is highly improbable.
(h) Final decision. Cut 2–#8 bars at 4 ft-6 in. from the support; use 90° standard hooks
on the 3–#10 bars; use #4 U stirrups @ 3.5-in. spacing as confinement over the Ldh
distance, as shown in Fig. 6.13.2 by the dashed stirrups.
The use of #10 bars in this cantilever beam is not a practical design but serves to illus-
trate the need for extending the cut location from B to C and then C′, and the need for
and treatment of hooked bars.
6.14 DEVELOPMENT OF POSITIVE REINFORCEMENT

AT SIMPLE SUPPORTS AND AT POINTS OF
INFLECTION
The concept of requiring the development of reinforcement on both sides of a section
where the bars are to be fully stressed may also be applied to the continuation of positive
moment tension reinforcement beyond either the centerline of a simple support or a point
of inflection.
Simple Supports
Referring to Fig. 6.14.1, consider point A on the factored moment Mu curve near a simple
support, where the factored moment Mu equals the design strength φ M n of the bars continu
ing into the support. The area of the shaded portion of the shear diagram equals the change
in moment between the center of support and point A; thus the ordinate on the factored
moment diagram at A is
Vavg (say, 0.9Vu ) x = φ M n
Thus, the distance x between the point A and the centerline of support in Fig. 6.14.1 is
approximately equal to
Mn
x=
Vu
Therefore, the available embedment length of the bars continuing into the support must
equal or exceed Ld, or
La + x ≥ Ld
or
Mn
La + ≥ Ld (6.14.1)
Vu
210
where
Mn = nominal flexural strength at the section with all reinforcement assumed to be
stressed at fy
Vu = factored shear at the support, and
La = straight embedment length beyond the centerline of support to the end of
the bars
Equation (6.14.1) is given in ACI-9.7.3.8.3(b) for beams [and in ACI 7.7.3.8.3(b) for one-
way slabs]; it applies when the end of the reinforcement is not confined by a compressive
reaction. In recognition of the fact that bars extending into a simple support have less ten-
dency to cause splitting when confined by a compressive reaction [6.42], the ACI Code
allows a 30% increase in the value of x or M n /Vu in such cases [ACI-9.7.3.8.3(a) for beams
and ACI-7.7.3.8.3(a) for one-way slabs], or (ACI-9.7.3.9.3 or ACI- 7.7.3.8.3)
Mn
La + 1.3 ≥ Ld (6.14.2)
Vu
Equations (6.14.1) or (6.14.2) need not be satisfied “If reinforcement terminates beyond
the centerline of supports by a standard hook, or a mechanical anchorage at least equivalent
to a standard hook. …” (ACI 9.7.3.8.3 and 7.7.3.8.3)
Inflection Points
Since an inflection point is a point of zero moment located away from a support (refer to
Fig. 6.14.2), bars in that region are not confined by a compressive reaction; therefore the
1.30 factor is interpreted as not to apply. In this case, the embedment length that must
exceed the required development length Ld may be stated as
 actual La , but 
M
available emdedment length =  not exceeding the  + n ≥ Ld [6.14.1]
Vu
 larger of 12db or d 
Use 1.3 Mn/Vu when reaction causes compression

around bars (i.e., at usual simple supports)
La
Mn/Vu
ACI–9.7.3.8.3 (beams)
≥ Ld ACI–7.7.3.8.3 (one-way slabs)
Factored moment diagram, Mu
A
Design strength of bars at support
φMn
Vu
Vavg
Figure 6.14.1 Development of reinforcement at a simple support.

21
6.15 DEVELOPMENT OF SHEAR REINFORCEMENT 211
For computation, La may not be

taken greater than 12db or d
Actual La
Usable La
Mn/Vu
Available anchorage Must exceed Ld
CL support Factored moment diagram, Mu
A
Based on bars
which extend
φMn
past inflection
point
Inflection point
Vu
Figure 6.14.2 Development of reinforcement at an inflection point.
where Mn and Vu refer to the nominal flexural strength and the factored shear at the point
of inflection. The limitation of the usable La to 12 bar diameters or the effective depth has
been applied because, according to ACI Commentary R9.7.3.8.3, there is no experimental
evidence to show that longer extensions will improve bar anchorage.
Additional development of reinforcement at the face of support is required by ACI-
9.7.3.8.2 when the beam is part of the primary lateral load resisting system.
6.15 DEVELOPMENT OF SHEAR REINFORCEMENT

Reinforcement in the web of a beam, whether it be for shear or for torsion (see Chapter 18),
must be properly anchored so that its full tensile capacity is available at or near the mid-
depth of a beam. For proper function, the shear reinforcement must “extend as close to the
compression and tension surfaces of the member as cover requirements and proximity of
other reinforcement permits and shall be anchored at both ends” (ACI-25.7.1.1). It is espe-
cially important to extend the stirrups as close to the compression face as possible because
diagonal cracks may extend deeply into the compression zone when the nominal strength
of the member is approached.
The ends of single leg, simple U, or multiple U stirrups of deformed bars or wires must
be anchored by means of a standard stirrup hook around longitudinal reinforcement, as
shown in Fig. 6.15.1. Such stirrups may be inclined, but the angle between the stirrups and
the longitudinal bars must be at least 45° in accordance with ACI-22.5.10.5.2.
The requirements for standard hooks for stirrups and ties are somewhat relaxed in
comparison to the standard hooks used for the main tension reinforcement bars shown in
Fig. 6.10.1. Also, hooked stirrups and hooked ties are not permitted for bars larger than #8.
As noted in Chapter 5, stirrups and ties are usually #3 or #4 bars, occasionally #5 or #6, and
rarely, if ever, larger than #8, so the ACI Code limit on size of hooked stirrups should rarely
apply. Note also that the 180° hook is not typically used for stirrups and ties. The require-
ments for standard hooks at ends of ties and stirrups are shown in Fig. 6.15.2.
21
Standard stirrup
hook (see Fig. 6.15.2)
Longitudinal bar required
As close as feasible, As close as feasible,
ACI–25.7.1.1 ACI–25.7.1.1
fyt
0.014db
h λ fc’
2 minimum
(a) For #5 bar and D31 wire, (b) For #6, #7, #8 stirrups
and smaller, or for #6, #7, with fyt > 40,000 psi
and #8 bar with fyt ≤ 40,000 psi (ACI–25.7.1.3b)
(ACI–25.7.1.3a)
Figure 6.15.1 Development of deformed bar or deformed wire stirrups.
6db ≥ 3 in.
135°
12db
6db ≥ 3 in.
(a) 90° hook (b) 90° hook (c) 135° hook
(for #5 bars and smaller) (for #6, #7, and #8 bars) (for #8 bars and smaller)
Figure 6.15.2 Standard hooks for stirrups and ties (ACI-25.3.2).
As close as feasible,
ACI–25.7.1.1
d d
4
max d
max max
4 4
2”
d
2” min
8 wire diameter
bend (min)
(a) (b)
Figure 6.15.3 Development of welded plain wire reinforcement stirrups (ACI-25.7.1.4).
When closed stirrups are desired, one practical procedure is to use a pair of U stirrups
without hooks, placed to form a closed unit. If this is done, ACI-25.7.1.7 requires laps of
1.3Ld for proper splicing. When members are at least 18 in. deep and the yield strength of
the stirrup does not exceed 9 kips per leg, splices are adequate if the legs extend the full
available depth of the member.
For U stirrups of welded plain wire reinforcement, anchorage may be accomplished
(ACI-25.7.1.4) by using either “(a) Two longitudinal wires spaced at a 2-in. spacing along
the member at the top of the U” or “(b) One longitudinal wire located not more than d/4
from the compression face and a second wire closer to the compression face and spaced
not less than 2 in. from the first wire. The second wire shall be permitted to be located on
the stirrup leg beyond a bend, or on a bend with an inside diameter of bend of at least 8db.”
These provisions are illustrated in Fig. 6.15.3.
For single-leg stirrups of welded plain or deformed wire reinforcement, ACI-25.7.1.5
includes the recommendations [6.43] of the PCI/WRI Ad Hoc Committee on Welded Wire
213
6.16 TENSION LAP SPLICES 213
See ACI–25.7.1.1 Two horizontal wires

top and bottom
2” minimum
d
2 Plain or Greater of
deformed d/4 or 2”
d vertical wires
Greater of
d/4 or 2”
Main
reinforcement
2” minimum
“as close to tension
surface as cover
requirements and Outer wire not above
proximity of other lowest main reinforcement
reinforcement will
permit” (ACI–25.7.1.1)
Figure 6.15.4 Development of single-leg plain or deformed welded wire shear reinforcement

(ACI-25.7.1.5).
Fabric for Shear Reinforcement. Single-leg stirrups are practical and common in precast pre-
stressed concrete T-sections. The provisions of ACI-25.7.1.5 are illustrated in Fig. 6.15.4.
For guidance in placing stirrups and ties to obtain proper anchorage, reference should be
made to the (2004) ACI Detailing Manual [2.23].
6.16 TENSION LAP SPLICES

Wherever bar lengths required in a structure exceed the length available or exceed the
length that may be economically shipped or erected, splices are necessary. Splicing may
be accomplished by simple lapping of bars side by side, either in contact or separated, over
a specified length, referred to as the splice length (see Fig. 6.16.1). Alternatively, welded
splices or mechanical connectors (though generally more expensive) may be used to pro-
vide continuity of the reinforcement (see Section 6.17).
In the lapped splice of Fig. 6.16.1, the total force in the tension reinforcement must be
transferred from bars a to bars b over the splice length, Lst. Test data suggest that the force
transfer mechanism for lapped splices is similar to that for single bars [6.3, 6.4], where the
forces from one bar are mainly transferred to the concrete by mechanical locking of the lugs
to the surrounding concrete, and then from the concrete to the adjacent bar. Furthermore,
the concrete cylinder hypothesis for splitting failure of single bars (see Fig. 6.4.3) may be
extended for a lap splice where the cylinders of concrete tributary to each bar interact to
form an oval ring, as shown in Fig. 6.16.2.
Analysis of the data on splice and development length tests suggest that the overlap dis-
tance required in tension lap splices, Lst, should be equal to the development length Ld of
the bar. However, because stress concentrations near the splice tend to produce splitting at
early stages of loading, a splice length greater than the development length Ld is required
unless special precautions are taken [6.3–6.5, 6.23, 6.25, 6.28, 6.44–6.48].
The ACI Code defines two classes of tension lap splices, Class A or Class B, depending
on (1) the percentage of bars being spliced within the required lap distance and (2) the stress
level in the unspliced bars at the splice location (ACI-25.5.2.1). As shown in Table 6.16.1,
Class A and Class B are defined to have overlap distances of 1.0Ld and 1.3Ld, respectively, but
not less than 12 in., where Ld is the development length of the bar calculated in accordance
with Section 6.6. These provisions apply equally to deformed bar or deformed wire splices,
but tension lap splices are not permitted for bars larger than #11 (ACI-25.5.1.1).
The ratio (As required)/(As provided) column in Table 6.16.1 refers to the percentage
of available capacity that is utilized. The ratio may also be considered as the percent of fy
214
Figure 6.16.1 Bar arrangement and force transfer in a typical lap splice.
S’ S’
S’ Cs = S’ 2 2db S’ = 2Cs 2db 2
2 2db S’ 2db 2
Failure plane
Cb
Cb
Cs > Cb
C = Cb
Cb > Cs C = Cs = S’ Cs >> Cb
2
Failure patterns as for single bars
Just before failure

Side split failure
At failure (CS >> Cb) At failure (CS > Cb)

V-notch failure Face-and-side split failure
Figure 6.16.2 Concrete cylinder hypothesis and split failure patterns for lapped spliced bars. (From Orangun, Jirsa, and
Breen [6.3].)
215
6.17 WELDED SPLICES AND MECHANICAL CONNECTIONS 215
TABLE 6.16.1 TENSION LAP SPLICES (ACI-25.5.2.1)
 As required  Maximum Percent Splice Required Lap, Notes

 A provided 
s of As Spliced Class Lsta
over splice length
≤ 0.5 50 A 1.0Ld Desirable

100 B 1.3Ld OK
> 0.5 100 B 1.3Ld Avoid more than 0.5
a 12 in. minimum.
to which the bars are stressed. When the factored moment Mu is only 50% of the design
strength φ M n, the ratio would be considered 0.5. In general, temperature and shrinkage
reinforcement should be considered as fully stressed for the purpose of designing splices.
The modification factors described in Section 6.7 to account for the location of the
reinforcement ψ t , epoxy coating ψ e, bar size ψ s, and lightweight concrete λ, also apply
to splices, except that Ld cannot be reduced if reinforcement in excess of that required by
analysis is provided (ACI-25.5.1.4).
A beam with splices should be as ductile as one without splices. In general, splices
should be located away from points of maximum tensile stress, and splicing should be
staggered along the length of the bars. In other words, all bars should not be spliced at
one location. The ACI Code provisions are intended to prevent splice failure when the full
nominal strength in flexure is reached at the spliced location. Requirements for minimum
clear spacing of contact splices (ACI-25.5.1.2) are to ensure that an adequate amount of
surrounding concrete is provided to resist splitting. In noncontact lap splices, however, the
individual bars should not be spaced transversely too far apart; otherwise, an unreinforced
section is created. Noncontact splices in flexural members may not be farther apart than
one-fifth of the required lap length nor 6 in. (ACI-25.5.1.3).
The lap lengths prescribed in Table 6.16.1 (also stated earlier: Section 6.9) shall be
increased 20% for a three-bar bundle and 33% for a four-bar bundle (ACI-25.6.1.5). For
determining the appropriate modification factors, a unit of bundled bars is to be treated as
a single bar of a diameter derived from the equivalent total area. As stated in ACI-25.6.1.7,
lap splices of individual bars within a bundle must be staggered so that they do not overlap.
6.17 WELDED SPLICES AND MECHANICAL

CONNECTIONS IN TENSION
A welded tension butt splice or mechanical connection is used when large tensile forces
are to be transmitted across a splice or large bars need to be spliced and lapping may be
impractical or prohibited. Bars larger than #11 may not be lap spliced to carry tension
(ACI-25.5.1.1). Tension tie members may not be lap spliced either (ACI-25.5.7.4). A ten-
sion tie is a member (a) carrying a tensile force large enough to cause tension over the
entire section, (b) having a stress level in the reinforcement high enough to require every
bar to be fully effective, and (c) having limited concrete cover on its sides (ACI-R25.5.7.4).
In the case of tension tie members, welded or mechanical splices in adjacent bars shall be
staggered at least 30 inches.
Minimum requirements for mechanical and welded splices are provided in ACI-25.5.7.
These splices are required to develop in tension at least 1.25fy (ACI-25.5.7.1). According to
ACI-R25.5.7.1, “the 25 percent increase above the specified yield strength was selected as
both an adequate minimum for safety and a practicable maximum for economy.” It is noted
that mechanical and welded splices with strengths less than 1.25fy (allowed in the 2011 and
previous editions of the ACI Code) are no longer permitted in the 2014 ACI Code.
The welded splice is intended primarily for relatively large bars (#6 and larger) in
main members (ACI-R25.5.7.1). The tensile capacity required is intended to ensure sound
216
full penetration welds—that is, to produce splices capable of developing 1.25As fy of the
spliced bars.
6.18 COMPRESSION LAP SPLICES

Whereas bars larger than #11 acting in tension are not permitted to be lap spliced, #14 and
#18 bars in compression may be lap spliced to #11 and smaller bars (ACI-25.5.5.3).
The minimum overlap in compression lap splices when fc′ is not less than 3000 psi is the
following (ACI-25.5.5.1):
for f y ≤ 60, 000 psi, Lsc = 0.0005 f y db or 12 in. ( whichever is larger ) (6.18.1)9
(
for f y > 60, 000 psi, Lsc = 0.0009 f y − 24 db ) or 12 in. ( whichever is larger ) (6.18.2)9
When fc′ is less than 3000 psi, the lap length is to be increased by one-third. Also, when
bars of two different sizes are lapped (ACI-25.5.5.4), the lap splice length is to be the larger
of (1) the compression splice length of the smaller bar or (2) the compression development
length Ld of the larger bar (see Section 6.8).
For compression members whose main steel is surrounded by closed ties throughout
the lap length, the required lap may be taken at 0.83 of that otherwise required by Eqs.
(6.18.1) and (6.18.2), but not less than 12 in. (ACI-10.7.5.2.1). A minimum percentage of
column tie area is also required by ACI-10.7.5.2.1. Column ties are discussed in Chapter 10
(Section 10.8).
For members whose main steel is surrounded by a closely wound spiral, the required lap
may be taken at 0.75 of that otherwise required, but not less than 12 in. (ACI-10.7.5.2.1).
Spiral reinforcement is discussed in Chapter 10 (Section 10.9).
The number of bar diameters required for the overlap in compression lap splices is sum-
marized in Table 6.18.1.
TABLE 6.18.1 BAR DIAMETERS REQUIRED FOR COMPRESSION LAP

SPLICES FOR fc′ ≥ 3000 PSI (ACI-25.5.5.1 AND ACI-10.7.5.2.1)
Yield Stress Bar Diametersa

fy (ksi)
Spiral Column Tied Column Others
40 15 16.6 20
50 18.75 20.75 25
60 22.5 24.9 30
75 32.6 36.2 43.5
80 36.0 39.9 48.0
aWhen computing splice length, the minimum to be used is 12 in.
9 Eqs (6.18.1) and (6.18.2), for SI, Lsc and db in mm, fc′ and fy in MPa, give, respectively,
For f y ≤ 420 MPa, Lsc = 0.071 f y db or 300 mm (6.18.1)
For f y > 420 MPa, Lsc = (0.13 f y − 24) db or 300 mm (6.18.2)

217
6.21 DESIGN EXAMPLES 217
6.19 END BEARING CONNECTIONS, WELDED

SPLICES, AND MECHANICAL CONNECTIONS
IN COMPRESSION
End bearing connections are allowed for compression only, wherein the load in the bars is
transmitted by bearing of square cut ends held in concentric contact by a suitable device.
According to ACI-25.5.6.3, bar ends must terminate in flat surfaces within 1.5° of a right
angle to the axis of the bars and must be fitted within 3° of full bearing after assembly. End
bearing splices are only permitted when the member contains closed ties, closed stirrups,
spirals, or hoops (ACI-25.5.6.2).
When welded splices or mechanical connections are used in compression, the require-
ments are the same as for tension splices—that is, they must be capable of developing
125% of the yield strength of the bars (ACI-25.5.7.1).
6.20 SPLICES FOR MEMBERS UNDER COMPRESSION

AND BENDING
For compression members, there are three categories of special splice provisions
(ACI-10.7.5.1).
Compressive Bar Stress Due to Factored Loads

When the factored load causes compression in the bars, any type of splice may be used,
including lap splices according to ACI-25.5.5.1 and 25.5.5.3, mechanical and welded
splices according to ACI-25.5.7, and an “end bearing splice” according to ACI-25.5.6 and
10.7.5.3.1. For lap splices, the lap length must satisfy Eq. (6.18.1) or (6.18.2) and cannot
be less than 12 in.
Tension Bar Stress Not Exceeding 0.5fy Due to Factored Loads

When moderate tension bar stresses under factored loads exist but do not exceed 0.5fy,
lap splices are permitted according to ACI-10.7.5.2.2. Alternatively, mechanical or welded
splices are permitted according to the requirements of ACI-25.5.7. When lap splices are
used (for tension not exceeding 0.5fy), Class B tension lap splices must be used when
more than 50% of the bars are spliced at a section. When fewer than 50% of the bars are
spliced at a section and the splices are staggered by at least Ld, Class A splices may be used
(ACI-10.7.5.2.2).
Tension Bar Stress Greater than 0.5fy due to Factored Loads

These are tension splices. When lap splices are used, they must be Class B splices accord-
ing to ACI-10.7.5.2.2. As an alternative to lap splices, mechanical or welded splices are
permitted according to ACI-25.5.7.
6.21 DESIGN EXAMPLES
Two complete examples on the design of reinforced concrete flexural members are pre-
sented for the purpose of showing the design for flexure, shear, and development of rein-
forcement, all in the same beam.
218
EXAMPLE 6.21.1
Design the simply supported beam shown in Fig. 6.21.1(a). The dead load is 1.08 kips/ft,
not including the weight of the beam. The live load consists of a concentrated load of
13.8 kips at midspan. Use normal-weight concrete with fc′ = 3000 psi and fy = fyt = 40,000
psi steel reinforcement.
SOLUTION
(a) Design a cross section for flexure. Assume that a rectangular section is desired with
tension reinforcement only. A small beam is desired, having a reinforcement ratio
ρ somewhat less than that corresponding to a net tensile strain ε t = 0.005. This will
satisfy the minimum net tensile strain limit of 0.004 (ACI-9.3.3.1). From basic prin-
ciples as illustrated in Section 3.6, or the value from Table 3.6.1,
ρtc = ρ (ε t = 0.005) = 0.0203
Select ρ = 0.02 , and using Eq. (3.8.4b) or strain compatibility and equilibrium, as shown
in Section 3.8, obtain Rn = 675 psi.
(b) Compute the factored loads using ACI-5.3. Assume that the weight of the beam is
0.2 kip/ft,
wu = 1.2 D +1.6 L = 1.2 (1.08 + 0.2 ) +1.6 ( 0 ) = 1.54 kips/ft ( dead load )
Wu = 1.2 ( 0 ) +1.6 (13.8) = 22.1 kips ( live load )
1 1
Mu = (1.54)(20)2 + ( 22.1) (20) = 77 + 111 = 188 ft-kips
8 4
Assuming φ = 0.90 for a tension-controlled section,
Mu 188
required Mu = = = 209 ft-kips
φ 0.90
(c) Determine beam size. For the selected Rn ≈ 675 psi, the corresponding bd 2 is
Mn 209 (12,000)
required bd 2 = = = 3716 cu in.
chosen Rn 675
If b = 12 in.,
3716
d= = 17.6 in.
12
1
required h = d + approx 2 in. for one layer of bars
2
= 17.6 + 2.5 = 20.1 in.
12(20)
weight of beam = (0.15) = 0.25 kip/ft
144
Use 12 × 20 cross section.
(Continued)
219
1’ – 0” #3 to support
4’ – 6” 12”
stirrups
20”
#10
3 12 ” 13 spaces @ 8 12 ” #11
#3 U stirrups Symmetrical about CL

10’– 0”
(a) φMn
Ld = 2.6’
d = 1.46’
191’k
Ld = 3.1’
for #11
145’k
B A
Factored
moment, Mu
φMn with 2 – #11
5’ #10 bar
cut here
(b)
Face of
support
φVn = 30.9 k
φVs = 13.6k for s = 8 12 ”
d
27.1k
Max required φVs = 6.7 k
Factored
19.1k shear, Vu
11.1k
φVc = 17.3k
0.5φVc = 8.7 k
(c)
Figure 6.21.1 Simply supported beam of Example 6.21.1.
(d) Correct the loads for beam weight and select reinforcement.
revised wu = 1.2(1.08 + 0.25) = 1.60 kips/ft (dead load)
1 1
revised Mu = (1.60)(20)2 + (22.1)(20) = 80 + 111 = 191 ft-kips
8 4
Mu 191
required Mu = = = 212 ft-kips
φ 0.90
M 212(12,000)
required Rn = n2 = = 692 psi
bd 12(17.5)2
(Continued)
20
The required steel reinforcement ratio ρ may next be computed from Eq. (3.8.5), or
approximated by straight-line proportion,
 692 
As ≈ 0.020(12)(17.5)  = 4.3 sq in.
 675 
Note that when revised Rn is larger, as in this case, the approximation using straight-line
proportion is slightly nonconservative (i.e., it underestimates the ρ value).
Try 2–#11 and 1–#10 bars (As = 4.39 sq in.). Assuming #3 U stirrups, the minimum
width of beam to accommodate these bars is 10.8 in. (see Table 3.9.2), which is less than
the width of beam being used. [Note that the entry of Table 3.9.2 must be with the two
corner bars (#11); then add for the one smaller bar (#10), and then add the difference
in diameters between the larger and smaller bars, giving 10.64 in. from the table plus
0.14 in., making a total of 10.78 in. (say, 10.8 in.). Alternatively, enter Table 3.9.2 with
3–#11 and subtract the difference in bar diameters.]
(e) Check the strength of the section. Even when tables and design approaches should give
a result satisfying the ACI Code if no errors have been made reading tables and curves,
a check of strength should always be made to verify that no mistakes have been made.
C = 0.85 fc′ ba = 0.85 (3)12 a = 30.6 a
T = As f y = 4.39 ( 40 ) = 176 kips
a = 176 / 30.6 = 5.75 in.
 a
M n = (C or T )  d – 
 2
= 176 17.5 − 0.5 ( 5.75) = 215 ft-kips

1
12
(f) Compute the net tensile strain εt at the extreme tension steel,
β1 = 0.85; c = a/ β1 = 5.75 / 0.85 = 6.76 in.
For this beam having one layer of steel, dt = d = 17.5 in.,

dt − c 17.5 − 6.76
ε t = 0.003 = 0.003 = 0.00477
c 6.76
Since ε t is less than the limit 0.005 of tension-controlled sections (see Fig. 3.6.2), the
appropriate φ factor is slightly less than 0.90 (ACI-21.2.2). For sections with stirrups, the
linear relationship is given by Eq. (3.6.3),
(ε t − ε y )
φ = 0.65 + 0.25 ≤ 0.9
(0.005 − ε y )
where
fy 40, 000
εy = = = 0.00138
Es 29, 000, 000
Thus,
(0.0047 − 0.00138)
φ = 0.65 + 0.25 = 0.88
(0.005 − 0.00138)
[φ M n = 0.88(215) = 189 ft-kips] ≈ [ Mu = 191 ft-kips]
The section is acceptable.

(Continued)
21
(g) Design of shear reinforcement; simplified method with constant Vc.

Vu (at centerline of support) = 1.60 (10) + 11.1 = 27.1 kips
1
Vu (at d from face of support) = 27.1 − 1.60 (6 + 17.5) = 24.0 kips
12
1
φ Vc = φ (2 fc′ bw d ) = 0.75(2 3000 )(12)(17.5) = 17.3 kips
1000

required φ Vs = Vu − φ Vc = 24.0 − 17.3 = 6.7 kips
min φ Vs = φ 0.75 fc′ bw d ≥ φ (50)bw d
1 1
min φ Vs = 0.75[0.75 3000 (12)17.5] ≥ 0.75[50(12)17.5]
1000 1000
min φ Vs = 6. 5 kips ≤ 7.9 kips
thus, the 7.9-kip limit controls.

Since required φ Vs = 6. 5 kips is less than min φ Vs = 7.9 kips to satisfy the require-
ment of ACI-9.6.3.3, use min φ Vs = 7.9 kips to determine the stirrup spacing at the criti
cal section. For #3 U stirrups, using Eq. (5.10.7),
φ Av f yt d
φ Vs =
s
Thus,
φ Av f yt d 0.75(0.22)(40)17.5
required s = = = 14.6 in.
φ Vs 7.9
However, the stirrup spacing may not exceed d / 2 = 8.75 in.

Try #3 stirrups @ 8 1 2 -in. spacing. Stirrups at this spacing must be used until
Vu ≤ 0.5φ Vc = 17.3 / 2 = 8.7 kips, which for this beam [see Fig. 6.21.1(c)] means the
entire span.
(h) Make the preliminary selection of the cutoff point for 1–#10. The remaining design
strength with 2–#11 bars is
C = 30.6 a
T = 2(1.56)40 = 125 kips
125
a= = 4.08 in.
30.6
Compute the net tensile strain εt,

17.5 − 4.08 / 0.85
ε t = 0.003 = 0.00794
4.08 / 0.85
Since εt exceeds 0.005, the section is tension controlled and φ = 0.90.
φ M n = 0.90(125) [17.5 − 0.5(4.08)]

1
= 145 ft-kips
12
(Continued)
2
The value of 145 ft-kips is plotted on the factored Mu diagram to locate the theoretical
cutoff point A in Fig. 6.21.1(b). This point is located 6.73 ft from the center of the sup-
port. The actual cutoff location (point B) is found by extending from point A toward the
support a distance of 12 bar diameters or the effective depth of the member, whichever
is greater (ACI-9.7.3.3).
 1.27 
12 db = 12  = 1.27 ft
 12 
17.5
d= = 1.46 ft (Controls)
12
Therefore, the actual cutoff at point B would be located at (6.73 − 1.46 ) = 5.27 ft from
the center of the support. Make the proposed cut at 5 ft from the support center line.
This cutoff point will be acceptable only if the shear does not exceed two-thirds of the
shear strength at point B [ACI-9.7.3.5(a)]. An alternative is to provide extra stirrups in
accordance with ACI-9.7.3.5(c).
(i) Check cutoff point for satisfying the shear requirement of ACI-9.7.3.5(a) for cut-
ting bars in the tension zone. The shear strength provided by #3 stirrups at 8 1 2 -in.
spacing is
0.75(0.22)(40)(17.5)
φ Vn = φ Vc + φ Vs = 17.3 + = 30.9 kips
8.5
Vu at propo
osed cutoff point B = 19.1kips
Vn 19.1
= = 62% < 67% OK
φ Vn 30.9
Use 1–#10, 10′–0 ″ long, placed symmetrically about midspan.

(j) Use the simplified equations to determine the development length Ld for a #10 bar
that is to be terminated at point B. Since the #11 bars are (presumably) not being
developed over the portion of the beam where the #10 will be developed, the #11
bars do not influence the determination of Category A or B. Since the #10 has nearly
5-in. cover to the side faces of the beam and has cover to the bottom face of 1 7 8 in.
(with #3 stirrups), both items 1 and 2 of Category A are satisfied.
Using simplified Eq. (6.6.5) for #7 and larger bars,
Ld =   db [6.6.5]
 20 λ fc′ 
The modification factors ψ t , ψ e , λ are all 1.0 for this beam. Thus, Eq. (6.6.5) gives
 40, 000(1.0)1.0 
Ld (for # 10) =  1.27 = 46.4 in. (3..9 ft)
 20(1.0) 3000 
(k) Use the general equation, Eq. (6.6.1), to determine the development length Ld for
#10 bars. That equation is
 

Ld =  d [6.6.1]
 40 λ fc′  cb + K tr   b

 d  
 b 
(Continued)
23
The distance cb is the smaller of (a) bottom or side cover and (b) one-half center-
to-center spacing; in this case,
cb = bottom cover = 1.5 (i.e.,clear) + 0.375 (i.e.,stirrup)

+ 0.635 (i.e., bar radius) = 2.51 in.
In other words, the potential splitting plane is expected to be vertical between the cen-
ter bar and the bottom face of the beam, similar to that shown for the center bar in
Fig. 6.6.1(b). To compute Ktr, use the #3 stirrups with spacing s = d / 2 = 8.5 in. along
the development length. The number n of bars being developed is 1. The area of trans-
verse reinforcement that crosses the potential splitting plane within the spacing s is
Atr = 1(0.11) sq in. Thus, evaluating Eq. (6.6.2) gives
40 Atr 40(1)(0.11)  40 
K tr = =   = 0.345
sn 8.5(1)  60 
where the ratio 40/60 has been applied to account for the yield strength of 40,000 psi.
The 2008 and earlier editions of the ACI Code included the yield strength of the trans-
verse reinforcement in the calculation of Ktr. The coefficient in the current equation
for Ktr (i.e., 40) includes, in effect, the yield strength of the transverse reinforcement,
and, for Grade 60 reinforcement, replaces the term f yt / 1500 ( = 60, 000 /1500 = 40 ) in
the preceding expression for Ktr. Whether Ktr should be adjusted for the Grade of steel
is a matter of opinion. According to ACI Commentary R25.4.2.3, the yield strength of
transverse reinforcement was removed from the expression for Ktr “because tests dem-
onstrate that transverse reinforcement rarely yields during a bond failure.” Because a
Grade lower than 60 ksi steel is used in this example, the authors have adjusted the value
of Ktr for Grade 40. This will result in a lower contribution of the transverse reinforce-
ment, Ktr, a result that is conservative.
 cb + K tr 2.51 + 0.345 
 = = 2.25 < 2.5 max
 db 1.27 
Thus, ( cb + K tr ) / db = 2.25. Note that whenever the cover and transverse reinforcement
term (cb + Ktr)/db is greater than 1.5, the general equation will give a smaller Ld than the
simplified equations. Evaluate Eq. (6.6.1),
 3 40, 000 1.0(1.0)1.0 
Ld (for # 10) =  1.27 = 30.9 in. (2.6 ft)
 40 (1.0) 3000 2.25 
which is 33% less than Ld computed with the simplified equation.

(l) Use the general equation, Eq. (6.6.1), to determine the development length Ld for the
#11 bars that extend into the support.
A check of Table 3.9.4 giving minimum beam width such that clear spacing of 2–#11
bars will equal 3db, shows a beam 10.9 in. wide or more will qualify. Since this is less
than the 12 in. used, the #11 bar spacing exceeds 3db, meaning that cover, rather than bar
spacing, will control the dimension cb. Thus,
cb = bottom or side cover = 1.5 (i.e.,clear) + 0.375 (i.e.,stirrup)
+ 0.705 (i.e., bar radius) = 2.58 in.
(Continued)
24
The potential plane of splitting is between the bars and the bottom and/or side face of the
beam [see Fig. 6.6.1(d)]. For the stirrups in the development region, use s = d / 2 = 8.5 in .
spacing for computation of Ktr. The number n of bars being developed is 2, and Atr =
2(0.11) sq in. Thus, evaluation of Eq. (6.6.2) gives
40 Atr 40(2)(0.11)  40 
K tr = =   = 0.345
sn 8.5(2)  60 

 cb + K tr 2.58 + 0.345 
 = = 2.07 < 2.5 max
 db 1.41 
Thus, ( cb + Ktr ) / db = 2.1. Evaluate Eq. (6.6.1),

 3 40, 000 1.0(1.0)1.0 
Ld (for # 11) =   1.41 = 36.8 in. (3.1 ft)
 40 (1.0) 3000 2.1
(m)
Check development length requirement at the support (Section 6.14). According to
ACI-9.7.3.8.3(a), it is required that
Mn
1.30 + La ≥ Ld
Vu
145
M n for 2 – #11 bar = = 161 ft-kips
0.9
Vu = factored shear at centerline of support = 27.1 kips
La = embedment length beyond the center of support; assume zero here
161(12)
1.30 = 92.7 in.> [ Ld (# 11) = 36.8 in.] OK
27.1
Actually, the design strength φ M n diagram provides this same check but more conser
vatively (i.e., without the 1.30 factor), since the horizontal distance from the center of
support to point A exceeds Ld.
(n) Design sketch. The final conclusions are presented in Fig. 6.21.1(a). The crack con-
trol provisions of ACI-24.3 need to be checked. If deflection control is important
to prevent damage to partitions or other construction, deflections must be checked
according to ACI-24.2.2. Computation for crack control and deflections is treated in
Chapter 12.
EXAMPLE 6.21.2
Design the overhanging beam shown in Fig. 6.21.2. The superimposed service uniform
dead load is 4.0 kips/ft. Use fc′ = 4000 psi (normal-weight concrete), f y = f yt = 60, 000 psi.
(Live load is not used in this example because its purpose is to show how flexure, shear,
and development of reinforcement are to be considered together in one simple case.)
(Continued)
25
SOLUTION
(a) Design for flexure. Since tension reinforcement will be required in the top of the over-
hang, it may be desirable to run some bars straight across the top of the entire beam.
This would also help reduce creep and shrinkage deflection under sustained load.
To have reasonable expectation that deflection will not be excessive, choose ρ at
about 0.6ρtc = 0.011 as recommended in Section 3.9. Use ρ = 0.011 .
For the selected ρ, one can use Eq. (3.8.4b) to obtain Rn = 596 psi. Alternatively,
choose Rn ≈ 600 psi from Fig. 3.8.1.
(b) Compute the factored loads and the corresponding shear and moment diagrams.
Estimate the beam weight as 0.3 kip/ft (roughly 2 sq ft, say 12 × 24 in.),
wu = 1.4 D = 1.4 ( 4.0 + 0.3) = 6.02 kips / ft ( dead load )
The factored shear at the left support is

6.02(18) 0.5(6.02)(6.5)2
Vu = − = 47.1 kips
2 18
(47.1)2
max(+ ) Mu = = 184 ft-kips
2(6.02)
1
max( −) Mu = (6.02)(6.5)2 = 127 ft-kips
2
The shear and moment diagrams for factored loads are shown in Fig. 6.21.2.
(c) Determine beam size. For the chosen Rn ≈ 600 psi, the corresponding bd 2 is
Mu 184(12,000)
required bd 2 = = = 4090 cu in.
φ (chosen Rn ) 0.90(600)
where φ = 0.90 is assumed. Thus,
b d h
(in.) (in.) (in.)
12 18.5 21
14 17.5 20
Assuming one layer of tension steel, 2.5 in. has been added to d to give the overall
dimension h. When checking strength, compute d, starting from overall h.
Use 14 × 20 cross section.
(d) Select the reinforcement. Correct the beam weight and update the factored loads,
shears, and moments, if needed.
14(20)
revised weight of beam = (0.15) = 0.29 kip/ft
144
The revised factored load wu is 6.01 kips/ft, which is approximately the same as that ini-
tially assumed. The factored shear and moment diagrams remain unchanged (Fig. 6.21.2).
The corrected effective depth d, assuming one layer of steel, is
revised d = h − cover − stirrup diam − bar radius
d = 20 − 1.5 − 0.375 ( #3 est.) − 0.5 ( est.) = 17.6 in.
(Continued)
26
At section A-A of Fig. 6.21.2, the revised required Rn then becomes

Mn 184(12, 000)
φ bd 2
0.90(14)(17.6)2
The required steel reinforcement ratio ρ may next be computed from Eq. (3.8.5),
obtained from the Fig. 3.8.1 curves, or approximated by straight-line proportion,
 566 
As ≈ 0.011(14)(17.6)  = 2.6 sq in.
 600 
Use 2–#9 and 1–#7 (As = 2.60 sq in.) in the midspan region of the 18-ft span.
wD = 4.0 kips/ft (not including beam weight)
(wu = 6.01 kips/ft with
corrected beam weight)
18’– 0 6’– 6”
Width of supports = 18”
4’– 0” B
1
72 ” A 4 – #6 × 10’– 6”
20”
14”
A 5’
B
1
2 – #9 × 18’– 7 2 ”
d 1 – #7 × 13’
47.1k
d
39.1k
33.8k
23.4k 23.6k
#3 stirrups
Vu, kips
4”
14 @ 7” = 8’– 2”
φVc
0.5φVc 0.5φVc
0.5φVc
7 @ 8” φVc
4” 5 @ 8”
#3 stirrups #3 stirrups
23.4k 4”
7.82’
47.8k
Ld (#9) = 2.7’ Ld (#7) = 2.1’ d 61.2k
187’k Ld (#9) = 2.7’
184
B
Mu, ft-kips
A
Moment
capacity Tensile
diagram, φMn capacity
2 – #9
Min 12db or d Factored φMn = 147’k
moment, 1.46’ (ACI–9.7.3.8.4)
Mu D
Min 12db or d = 17.6”
2 – #6
φMn = 67’k
C
Ld = 2.3’ Ld = 2.5’
127’k
131’k
Ld = 2.3’
18” support width
Figure 6.21.2 Overhanging beam for Example 6.21.2.

(Continued)
27
At section B-B, obtain by proportion

− Mu 127
required As = 2.6 = 2.6 = 1.8 sq in.
+ Mu 184
Use 4–#6 (As = 1.76 sq in.) in the negative moment region.
(e) Check the strength of the section. At section A-A,

C = 0.85 fc′ ba = 0.85(4)14 a = 47.6 a
T = As f y = 2.60(60) = 156 kips
a = 156 / 47.6 = 3.28 in.

M n = (C or T )(d − a /2)
1
= 156[17.6 − 0.5(3.28)] = 207 ft-kips
12
Compute the net tensile strain ε t ,
17.6 − 3.28 / 0.85
ε t = 0.003 = 0.0107
3.28 / 0.85
This far exceeds the limit of 0.005 for tension-controlled sections; therefore, the appro-
priate φ factor is 0.90.
φ M n = 0.90 ( 207) = 187 ft-kips > [ Mu = 184 ft-kips] OK
At section B-B,
C = 0.85 fc′ ba = 0.85(4)14 a = 47.6 a
T = As f y = 1.76(60) = 106 kips
a = 106 / 47.6 = 2.22 in.
M n = 106 [17.6 − 0.5(2.22)] = 146 ft-kips

1
12
Compute the net tensile strain ε t ,
17.6 − 2.22 / 0.85
ε t = 0.003 = 0.0172
2.22 / 0.85
Since εt exceeds 0.005, the section is tension controlled and φ = 0.90.
φ M n = 0.90 (146 ) = 131 ft-kips > [ Mu = 127 ft-kips] OK

The reader is reminded that for a given Rn, the required ρ expression is a quadratic
function; however, for practical use it may be approximated to be linear (see Fig. 3.8.1).
A check of strength should always be made.
The arrangement of main reinforcement finally selected is shown in Fig. 6.21.2.
Note that in the design of the flexural reinforcement for the maximum negative
moment, the center of support value of 127 ft-kips was used instead of the lesser value at
the face of support, as could have been used in accordance with the discussion of the neg-
ative moment on continuous beams (Section 6.11). On this statically determinate beam,
the authors prefer a more conservative approach of using the center of support value. For
beams on wider supports, corrections to the face of support would be appropriate.
(Continued)
28
(f) Shear reinforcement. At d from the face of right support on the 18-ft span,
Vu = 61.2 − 6.01(17.6 + 9)
1
= 47.8 kips
12
φ Vc = φ (2 λ fc′) bwd
= 0.75 2(1.0) 4000  (14)(17.6)

1
= 23.4 kips
1000
where λ = 1.0 for normal-weight concrete
required φ Vs = Vu − φ Vc = 47.8 − 23. 4 = 24. 4 kips
 d
max permitted φ Vs  for s =  = 2φ Vc = 46.8 kips > [ required φ Vs = 24..2 kips] OK
 2
Note that maximum permitted φ Vs calculated is based on the nominal stress vs = 4 fc′
at the maximum spacing of d/2 (ACI-9.7.6.2.2), which gives maximum φ Vs = 2φ Vc when
the simplified Vc expression is used. For minimum percentage of stirrups (ACI-9.6.3.3),
( ) (
min φ Vs = φ 0.75 fc′ bw d = 0.75 0.75 4000 (14)(17.6) ) 1
1000
= 8.8 kips
but not less than

1
φ (50bw d ) = 0.75[50(14)(17.6)] = 9.2 kips
1000
Thus, the 9.2-kip limit controls.
For strength, using #3 U stirrups,
φ Av f yt d 0.75(0.22)(60)(17.6) 174
φ Vs = = =
s s s
or
174
s= = 7.2 in.
24.2
Thus, the strength requirement permits stirrups no farther apart than 7.2 in. and is more
restrictive than the d/2 limit of 8.8 in.
Use #3 U stirrups at 7-in. spacing, as dimensioned on the shear diagram in Fig. 6.21.2.
(Ordinarily the stirrups should be dimensioned on the design sketch, i.e., the side view
of the beam in Fig. 6.21.2.)
For the cantilever portion and to the right of the left support of the beam, it can be ver-
ified that only minimum stirrups at d / 2 = 8.5 in. maximum spacing are needed. Thus,
Use #3 U stirrups at 8-in. spacing in these regions, as shown in Fig. 6.21.2.
(g) Development length using the simplified method of ACI-25.4.2.2. Because the clear
cover for bars being developed is at least db, clear lateral spacing is at least db (see
Table 3.9.2), and at least minimum stirrups are used, Category A, item 1 is satisfied.
Use simplified equations, Eqs. (6.6.4) and (6.6.5). From Table 6.6.1, the following
Ld values are obtained when the modifications ψ t , ψ e , and λ = 1.0:
Ld ( #6 ) = 28.5 in. ( 2.4 ft )
Ld ( #7) = 41.5 in. (3.5 ft )
Ld ( #9) = 53.5 in. ( 4.5 ft )

(Continued)
29
For the #6 bars in the negative moment zone, a casting position factor ψt of 1.3 must be
used. Thus, for the #6 bars,
Ld ( #6 ) = 28.5 (1.3) = 37.1in. (3.1ft)
(h) Development length using the general equation, Eq. (6.6.1).

 

Ld =  d [6.6.1]
 40 λ fc′  cb + K tr   b

 d  
 b 
The distance cb is the smaller of (a) top or side cover, and (b) one-half center-to-center
spacing.
For the #9 corner bars:
bottom and side cover = 1.5 + 0.375 + 1.128 / 2 = 2.44 in. Controls
Assuming that the #7 bar will not be developed at the same location, one half of the bar
spacing dimension is
1 1  1.128  
spacing = 14 − 2(1.5) − 2(0.375) − 2  = 4.6 in.
2 2  2  
Thus, the potential plane of splitting for these corner bars is between the bars and the
bottom and/or side face of the beam [see Fig. 6.6.1(d)]. To the right of the left support,
stirrup spacing is 8 in. Thus, Ktr is computed by using Eq. (6.6.2) as follows:
40 Atr 40(2)(0.11)
K tr = = = 0.55
sn 8(2)
To the left of the right support, stirrup spacing is 7 in. Thus,
8
K tr = 0.55 = 0.63
7
For the #7 center bar:
Assuming that the #9 corner bars will not be developed at the same location, then verti-
cal splitting and thus the bottom cover dimension governs for the #7 bar:
bottom cover = 1.5 + 0.375 + 0.875/2 = 2.3 in. (vertical splitting governs)
Toward the left support, stirrup spacing is 8 in; thus,

40 Atr 40(1)(0.11)
K tr = = = 0.55
sn 8(1)
Toward the right support, stirrup spacing is 7 in; thus,

8
K tr = 0.55 = 0.63
7
For the #6 bars (negative moment region):
cover = 1.5 + 0.375 + 0.75/2 = 2.3 in.
1 1  0.75   1
spacing = 14 − 2(1.5) − 2(0.375) − 2  = 1.6 in. (horizontal splitting governs)
2 2  2   3
(Continued)
230
To the left of the right support, stirrup spacing is 7 in. Since horizontal splitting governs, the
stirrups are effective only for the two corner bars out of the four bars being developed. Thus,
40 Atr 40(2)(0.11)
K tr = = = 0.31
sn 7(4)
For the development of the #6 bars near the free end of the cantilever K tr = 0.
Compute (cb + K tr ) / db for the #9, #7, and #6 bars of this design.
 cb + K tr 2.44 + 0.55 
For #9 bars  = = 2.65 > 2.5 max
 db 1.128 
 cb + K tr 2.30 + 0.55 
For #7 bars  = = 3.26  > 2.5 max
 db 0.875 
In the calculations shown above, the smaller value of Ktr (i.e., larger stirrup spacing) was
used for the #9 and #7 bars. Use of the larger Ktr value will result in an even larger value
of (cb + K tr ) / db and thus the 2.5 limit would still control.
To the left of the right support
 cb + K tr 1.60 + 0.31 
For #6 bars  = = 2.55 > 2.5 max
 db 0.75 
For the development of the #6 bars near the free end of the cantilever K tr = 0; thus,
 cb + K tr 1.60 + 0 
For #6 bars  = = 2.13 < 2.5 max
 db 0.75 
Substituting into Eq. (6.6.1) gives

 3 60, 000 1.0(1.0)1.0 
Ld (# 9) =   1.128 = 32.1 in. (2.7 ft)
 40 (1.0 4000 ) 2.5
 3 60, 000 1.0(1.0)1.0 

Ld (# 7) =   0.875 = 24.9 in. (2.1ft)
 40 (1.0 4000 ) 2.5
At the 18-ft span side of the right support, using ψ t = 1.3 for top bars,
 3 60, 000 1.3(1.0)1.0 
Ld (# 6) =   0.75 = 27.7 in. (22.3ft)
 40 (1.0 4000 ) 2.5
At the free end of the cantilever, using ψ t = 1.3 for top bars,

 3 60, 000 1.3(1.0)1.0 
Ld (# 6) =  0.75 = 32.5 in. (2.7 ft)
 40 (1.0 4000 ) 2.13 
Although ACI-25.4.2.4 permits a bar size factor ψ s = 0.8 to be used for #6 and smaller
bars, it was not applied in the above calculations. In this example, however, the authors
have adopted the recommendations of ACI Committee 408 [6.53] which do not support
the use of this factor (see Section 6.7).
The above computation of development length Ld has assumed #3 stirrups at either
7- or 8-in. spacing for the high-shear regions near the supports. The 8-in. spacing is very
(Continued)
231
near the maximum permitted of d/2 (8.75 in.). Good practice will put stirrups in the
entire beam. Note again that because the cover and transverse reinforcement factor,
(cb + K tr ) / db , is larger than 1.5 in. in all Ld calculations, the simplified equations give
much larger Ld values [part (g)].
(i) Examine the feasibility of cutting the #7 bar in the positive moment zone. The
remaining 2–#9 would provide the following strength:
C = 47.6 a T = 2 (1.0 ) 60 = 120 kips
a = 2.52 in.
 2.52  1
φ M n = 0.90 (120 )  17.6 −  = 147 ft-kips
 2  12
Based on the design strength alone, the #7 bar could be cut at points A and B of
Fig. 6.21.2; however, ACI-9.7.3.5 must also be satisfied to cut bars in the tension zone.
This is examined in part (j).
For the #7 bar, the bar arrangement must be established before development length
can be accurately determined. If the #7 bar is cut as shown in Fig. 6.21.2, then its
development and the #9 bars development are independent. However, if all three bars
(2–#9 and 1–#7) are run into the support, then the center-to-center spacing between
bars might control, requiring increased development length for both the #7 and the
#9 bars. At the point in a design when the final check is made, of which the design
strength diagram is a part, the bar arrangement is known and from that the develop-
ment lengths can be computed.
The shorter development lengths computed using the general equation, Eq. (6.6.1),
were used in drawing the design strength diagram of Fig. 6.21.1.
(j) Check ACI-9.7.3.5 for cutting the #7 bar at points A and B in the tension zone (see
Section 6.11).
i. Since φ Mn for the continuing #9 bars is not twice the factored moment Mu at the
cutoff point, ACI-9.7.3.5(b) cannot be satisfied.
ii. Check if ACI-9.7.3.5(a) can be satisfied for #3 U stirrups at 8 in. at A and 7 in. at
B as provided.
At point A
The distance from the center of the left support to point A is
 1
(4.2 ft − d ) =  4.2 − (17.6)  = 2.7 ft
 12 
Make the cut location 2 ft 8 in. from the left support. Then,
Vu (at A) = 47.1 − 2.66 (6.01) = 31.1 kips
φ Av f yt d 0.75(0.22)(60)17.6
φ Vs (at A) = = = 21.8 kips
s 8
The percent stressed at A is then,
Vu 31.1
= = 69% > 67% max NG
φ (Vc + Vs ) 23.4 + 21.8
The percent stressed is very close to the maximum permitted; thus, a possible strategy
to satisfy this requirement would be to reduce stirrup spacing to, say, 7 in. In such a
(Continued)
23
case, the percent stressed at A would become 64%. Alternatively, check ACI-9.7.3.5, as
shown below.
At point B
The distance from the center of the right support to point B is
 1
(6.6 ft − d ) = 6.6 − (17.6)  = 5.13 ft
 12 
Make the cut location 5 ft from the right support. Then,

Vu (at B ) = 61.2 − 5.0 (6.01) = 31.2 kips
φ Av f yt d 0.75(0.22)(60)17.6
φ Vs (at B) = = = 24.9 kips
s 7
The percent stressed at B is then
Vu 31.2
= = 65% < 67% max OK
φ (Vc + Vs ) 23.4 + 24.9
iii. Check if ACI-9.7.3.5(c) can be satisfied for #3 U stirrups at 8 in. from A to ( 3 4 )d to
the right of A, as provided.
Vu (at A) = 31.1 kips
The required area of stirrups spaced at 8 in. at this location is

(31.1 − 23.4) 8
Av = = 0.08 sq in.
0.75 60(17.6)
The area of stirrups in excess of that required at this location is
excess Av = 0.22 − 0. 08 = 0. 14 sq in.
which is greater than the minimum required,
60 bw s / f yt = 60 (14 )(8) / 60, 000 = 0.11 sq in.
Therefore, the #7 bar could be cut at the proposed point A. Since cutoff point A is close
to the left support (2 ft 8 in.) and only one bar is being cut off, it may be just as cost
effective to simply extend the bar into the support. The latter approach is adopted in this
design example.
Cut 1–#7 bar only at point B and extend the bar into the left support (length = 13 ft).
(k) Determine cut of 2–#6 bars in the negative moment zone. In trying to cut the 2–#6
bars from the negative moment region near section B-B, the bars must be extended
into the span to the left of the support farther than 12 bar diameters or the effective
depth d, whichever is greater, beyond the point where they are no longer needed
(i.e., where φ M n (2 − # 6) = Mu ). Since the potential cut at point C (at Ld from the face
of support) is not in a tension zone, no further investigation of this cutoff location is
required. On determining that the remaining 2–#6 bars can be cut off only a short dis-
tance farther into the span at point D, beyond the point of inflection per ACI-9.7.3.8.4,
the decision is made to cut all 4–#6 bars at point D, that is, at 1.46 + 2.40 = 3.86 ft,
say, at 4 ft from the center of the support. If for some reason, such as the desire to
(Continued)
23
have compression steel for deflection control, two of the #6 bars could be extended
across the 18-ft span instead of being cut at D; then the other 2–#6 would be cut at
point C. On the right side of the support in the negative moment region, any potential
cut location would be so near the end of the cantilever that it was decided to run all
4–#6 bars to the end (say, 1.5-in. clear).
(l) Check ACI-9.7.3.8.3 for development of positive moment reinforcement at the left
simple support and at the point of inflection near the right support (see Section 6.14).
At the inflection point, ACI-9.7.3.8.3(b) requires that
Mn
+ La ≥ Ld
Vu
Assuming that the 1–#7 bar will be terminated (a conservative assumption for this com-
putation), only the 2–#9 bars contribute to Mn. The actual La from the inflection point to
the end of the bars exceeds d (17.6 in.); thus, use La = 17.6 in.
 147(12) 
 0.90(47.1) + 17.6 = 59.2 in. > [ Ld = 32.1 in.] OK
 
For the simple support, ACI-9.7.3.8.3(a) requires

1.30 M n
+ La ≥ Ld
Vu
In this case, La = 7.5 in. since the bars extend this amount beyond the center of the sup-
port. Assuming that the #7 bar does not extend into the support (a conservative assump-
tion for this computation),
1.30(147)12 
 0.90(47.1) + 7.5 = 61.6 in. > [ Ld = 32.1 in.] OK
 
(m) Design strength diagram. Figure 6.21.2 compares the factored moment Mu diagram
and the design strength φ M n diagram provided by the selected reinforcing bars. In
computing the design strength, the steel in the compression side of the beam has
been neglected because it would have a negligible effect and because it was not
required for strength. At the left end of the beam, the #7 and the 2–#9 bars have been
extended as far as is feasible into the support. In drawing the design strength dia-
gram for the left end of the beam, it was assumed that the design strength will vary
linearly with the development length of the #9 bars, although the #7 bar requires a
shorter development length. This is conservative.
(n) Deflection. The reinforcement ratio ρ being used was expected to control the
deflections. Nevertheless, the deflection must be investigated if damage to parti-
tions or other construction is of concern.
(o) Design sketch. The final arrangement of longitudinal steel and stirrups is shown
in Fig. 6.21.2. The stirrup locations are dimensioned on the Vu diagram. Omitted
from the elevation view of the beam are the nominal-sized (say, #3 or #4) longi-
tudinal bars arbitrarily added for the stirrups to wrap around wherever the stirrups
would otherwise have no support to hold them in vertical position. These pairs of
bars would be located in this beam at both top and bottom faces of the beam where
no longitudinal bars are shown in the figure. Cross-sectional views showing bars,
omitted here, are always part of the design sketch.
234
SELECTED REFERENCES
6.1. ACI Committee 408. “Bond Stress—The State of the Art,” ACI Journal, Proceedings, 63,
November 1966, 1161–1190. Disc., 63, 1569–1570.
6.2. LeRoy A. Lutz and Peter Gergely. “Mechanics of Bond and Slip of Deformed Bars in Concrete,”
ACI Journal, Proceedings, 64, November 1967, 711–721. Disc., 65, 412–414.
6.3. C. O. Orangun, J. O. Jirsa, and J. E. Breen. The Strength of Anchored Bars: A Reevaluation of
Test Data on Development Length and Splices, Research Report 154–3F, Center for Highway
Research, University of Texas, Austin, January 1975.
6.4. C. O. Orangun, J. O. Jirsa, and J. E. Breen. “A Reevaluation of Test Data on Development
Length and Splices,” ACI Journal, Proceedings, 74, March 1977, 114–122. Disc., 74, September
1977, 470–475.
6.5. James O. Jirsa, LeRoy A. Lutz, and Peter Gergely. “Rationale for Suggested Development,
Splice, and Standard Hook Provisions for Deformed Bars in Tension,” Concrete International,
1, July 1979, 47–61.
6.6. D. Z. Yankelevsky. “Bond Action Between Concrete and a Deformed Bar—A New Model,” ACI
Journal, Proceedings, 82, March–April 1985, 154–161.
6.7. Emory L. Kemp. “Bond in Reinforced Concrete: Behavior and Design Criteria,” ACI Journal,
Proceedings, 83, January–February 1986, 50–57.
6.8. Herbert J. Gilkey, Stephen J. Chamberlain, and Robert W. Beal. Bond Between Concrete and
Steel. Iowa Engineering Experiment Station Bulletin No. 147, Iowa State College, 1940.
6.9. Arthur P. Clark. “Bond of Concrete Reinforcing Bars,” ACI Journal, Proceedings, 46, November
1949, 161–184.
6.10. Phil M. Ferguson and J. Neils Thompson. “Development Length for Large High Strength
Reinforcing Bars,” ACI Journal, Proceedings, 62, January 1965, 71–93. Disc., 62, 1153–1156.
6.11. Phil M. Ferguson, John E. Breen, and J. Neils Thompson. “Pullout Tests on High Strength
Reinforcing Bars,” ACI Journal, Proceedings, 62, August 1965, 933–950.
6.12. Ervin S. Perry and J. Neils Thompson. “Bond Stress Distribution on Reinforcing Steel in Beams
and Pullout Specimens,” ACI Journal, Proceedings, 63, August 1966, 865–875.
6.13. E. L. Kemp, F. S. Brezny, J. A. Unterspan. “Effect of Rust and Scale on the Bond Characteristics
of Deformed Reinforcing Bars,” ACI Journal, Proceedings, 65, September 1968, 743–756.
Disc., 66, 224–226.
6.14. Saeed M. Mirza and Jules Houde. “Study of Bond Stress–Slip Relationships in Reinforced
Concrete,” ACI Journal, Proceedings, 76, January 1979, 19–46.
6.15. S. Soretz and H. Holzenbein. “Influence of Rib Dimensions of Reinforcing Bars on Bond and
Bendability,” ACI Journal, Proceedings, 76, January 1979, 111–125.
6.16. Raymond E. Untrauer and George E. Warren. “Stress Development of Tension Steel in Beams,”
ACI Journal, Proceedings, 74, August 1977, 368–372.
6.17. E. L. Kemp and W. J. Wilhelm. “Investigation of the Parameters Influencing Bond Cracking,”
6.18. Shiro Morita and Tetsuzo Kaku. “Splitting Bond Failures of Large Deformed Reinforcing Bars,”
6.19. R. Jimenez, R. N. White, and P. Gergely. “Bond and Dowel Capacities of Reinforced Concrete,”
6.20. S. Ali Mirza. “Bond Strength Statistics of Flexural Reinforcement in Concrete Beams,” ACI
6.21. J. P. Moehle, J. W. Wallace, and S.-J. Hwang. “Anchorage Lengths for Straight Bars in Tension,”
6.22. David Darwin, Steven L. McCabe, Emmanuel K. Idun, and Steven P. Schoenekase. “Development
Length Criteria: Bars Not Confined by Transverse Reinforcement,” ACI Structural Journal, 89,
November–December 1992, 709–720. Disc., 90, September–October 1993, 581–583.
6.23. LeRoy A. Lutz, S. Ali Mirza, and Narendra K. Gosain. “Changes to and Applications of
Development and Lap Splice Length Provisions for Bars in Tension (ACI 318–89),” ACI
Structural Journal, 90, July–August 1993, 393–406.
6.24. Shyh-Jiann Hwang, Yih-Ren Leu, and Han-Lin Hwang. “Tensile Bond Strengths of Deformed
Bars of High-Strength Concrete,” ACI Structural Journal, 93, January–February 1996, 11–20.
6.25. ACI Committee 408. “Suggested Development, Splice, and Standard Hook Provisions for
Deformed Bars in Tension,” Concrete International, 1, July 1979, 44–46.
6.26. Phil M. Ferguson and Farid N. Matloob. “Effect of Bar Cutoff on Bond and Shear Strength of
Reinforced Concrete Beams,” ACI Journal, Proceedings, 56, July 1959, 5–23.
6.27. Anthony M. Kao and Raymond E. Untrauer. “Shear Strength of Reinforced Concrete Beams
Terminated in Tension Zones,” ACI Journal, Proceedings, 72, December 1975, 720–722.
235
6.28. Mete A. Sozen and Jack P. Moehle. A Study of Experimental Data on Development and Lap-
Splice Lengths for Deformed Reinforcing Bars in Concrete. Report to The Portland Cement
Association, Skokie, IL, and The Concrete Reinforcing Steel Institute, Schaumburg, IL,
August 1990.
6.29. Paul R. Jeanty, Denis Mitchell, and M. Saeed Mirza. “Investigation of ‘Top Bar’ Effects in
Beams,” ACI Structural Journal, 85, May–June 1988, 251–257.
6.30. D. B. Cleary and J. A. Ramirez. “Bond Strength of Epoxy-
Coated Reinforcement,” ACI
Materials Journal, 88, March–April 1991, 146–149.
6.31. Oan Chul Choi, Hossain Hadje-Ghaffari, David Darwin, and Steven L. McCabe. “Bond of
Epoxy-Coated Reinforcement: Bar Parameters,” ACI Materials Journal, 88, March–April 1991,
207–217.
6.32. Bilal S. Hamad, James O. Jirsa, and Natalie I. D’Abreu de Paulo. “Anchorage Strength of
Epoxy-Coated Hooked Bars,” ACI Structural Journal, 90, March–April 1993, 210–217.
6.33. Hossain Hadje-Ghaffari, Oan Chul Choi, David Darwin, and Steven L. McCabe. “Bond of
Epoxy- Coated Reinforcement: Cover, Casting Position, Slump, and Consolidation,” ACI
6.34. Stacy J. Bartoletti and James O. Jirsa. “Effects of Epoxy Coating on Anchorage and Development
of Welded Wire Fabric,” ACI Structural Journal, 92, November–December 1995, 757–764.
6.35. LeRoy A. Lutz. “Crack Control Factor for Bundled Bars and for Bars of Different Sizes,” ACI
Journal, Proceedings, 71, January 1974, 9–10.
6.36. N. W. Hanson and Hans Reiffenstuhl. “Concrete Beams and Columns with Bundled

Reinforcement,” Journal of the Structural Division, ASCE, 84, ST6 (October 1958), 1–23.
6.37. Frank D. Steiner. “Suggested Applications for Bundled Bars,” ACI Journal, Proceedings, 64,
April 1967, 213–214.
6.38. John A. Hribar and Raymond C. Vasko. “End Anchorage of High Strength Steel Reinforcing
Bars,” ACI Journal, Proceedings, 66, November 1969, 875–883. Disc., 67, 423–424.
6.39. John Minor and James O. Jirsa. “Behavior of Bent Bar Anchorages,” ACI Journal, Proceedings,
72, April 1975, 141–149.
6.40. José L. G. Marques and James O. Jirsa, “A Study of Hooked Bar Anchorages in Beam–Column
Joints,” ACI Journal, Proceedings, 72, May 1975, 198–209.
6.41. Robert L. Pinc, Michael D. Watkins, and James O. Jirsa. Strength of Hooked Bar Anchorages in
Beam–Column Joints. CESRL Report No. 77-3, Department of Civil Engineering, University of
Texas, Austin, November 1977.
6.42. Raymond E. Untrauer and Robert L. Henry. “Influence of Normal Pressure on Bond Strength,”
ACI Journal, Proceedings, 62, May 1965, 577–586.
6.43. Joint PCI/WRI Ad Hoc Committee on Welded Wire Fabric for Shear Reinforcement, PCI
Technical Activities Committee (Leslie D. Martin, Chairman). “Welded Wire Fabric for Shear
Reinforcement,” PCI Journal, 25, July–August 1980, 32–36.
6.44. Phil M. Ferguson and John E. Breen. “Lapped Splices for High Strength Reinforcing Bars,” ACI
Journal, Proceedings, 62, September 1965, 1063–1078.
6.45. John P. Lloyd and Clyde E. Kessler. “Splices and Anchorages in One-Way Slabs Reinforced
with Deformed Wire Fabric,” ACI Journal, Proceedings, 67, August 1970, 636–642.
6.46. M. A. Thompson, J. O. Jirsa, J. E. Breen, and D. F. Meinheit. “Behavior of Multiple Lap Splices
in Wide Sections,” ACI Journal, Proceedings, 76, February 1979, 227–248.
6.47. Telvin Rezansoff, Jim A. Zacaruk, and Rob Topping. “Tensile Lap Splices in Reinforced
Concrete Beams under Inelastic Cyclic Loading,” ACI Structural Journal, 85, January–February
1988, 46–52.
6.48. Tel Rezansoff, Adeniyi Akanni, and Bruce Sparling. “Tensile Lap Splices under Static

Loading: A Review of the Proposed ACI 318 Code Provisions,” ACI Structural Journal, 90,
July–August 1993, 374–384.
6.49. T. D. Mylrea. “Bond and Anchorage,” ACI Journal Proceedings, 44, March 1948, 521–552.
6.50. R. A. Treece and J. O. Jirsa. “Bond Strength of Epoxy-Coated Reinforcing Bars,” ACI Materials
6.51. D. W. Johnston and P. Zia. Bond Characteristics of Epoxy-Coated Reinforcing Bars. Report No.
FHWA/NC/82-002, Department of Civil Engineering, North Carolina State University, Raleigh,
August 1982.
6.52. R. G. Mathey and J. R. Clifton. “Bond of Coated Reinforcing Bars in Concrete,” Journal of the
Structural Division, ASCE, 102, January 1976, 215–228.
6.53. ACI Committee 408. “Bond and Development of Straight Reinforcing Bars in Tension,” ACI
408R-03. American Concrete Institute, Farmington Hills, MI, 2003, 49 pp.
236
PROBLEMS
All problems are to be done in accordance with the ACI Code, and all loads given are
service loads, unless otherwise indicated. Assume that all beams have at least minimum
cover satisfying ACI-20.6.1.3.1, as well as minimum stirrups according to ACI-9.6.3.3 and
ACI-9.7.6.2.2, unless otherwise indicated. All design strength φ M n diagrams used in these
problems must be drawn to scale directly below a side view of the beam drawn to the same
longitudinal scale on the same page. Design strength φ M n diagrams must have critical
numerical values stated on the diagram, and horizontal distances to the critical points must
be dimensioned with numerical values. Use the load factors and load combinations per
ACI-5.3 and the φ factors of ACI-21.2.1.
6.1 Draw the free-body diagram for the 3-in. slice (φ M n ) diagram. Assume that the cutoff loca-
shown cross-hatched on the beam given in the tion 3 ft from the support satisfies the require-
figure for Problem 6.1. Assuming linear elas- ments of ACI-9.7.3.5 and ACI-9.7.3.8.3. What
tic behavior (see Sections 12.2–12.5), compute is the maximum uniformly distributed service
and show on the diagram values for the internal load that the beam may be permitted to carry
forces (tension, compression, and shear) on each (assume 50% live load and 50% dead load)? Use
side of the slice. fc′ = 3500 psi and f y = 40, 000 psi.
(a) Compute the average flexural bond stress on 6.3 The beam of the figure for Problem 6.3 has 3–#7
the bars over the 3-in. slice. bars as the reinforcement at section A-A with a
(b) Compute the average flexural bond stress on 1–#7 bar cut at 4 ft-0 in. from the face of the sup-
the bars over the distance from section A to port. Use fc′ = 4000 psi and f y = f yt = 60, 000 psi
the left end of the beam. to investigate the adequacy for development of
(c) What is the average anchorage bond stress reinforcement if the beam is subjected to a uni-
resisting the tensile force at section A? form live load of 2.02 kips/ft and uniform dead
(d) Explain what happens if the flexural bond load of 1.86 kips/ft (including beam weight).
stress in part (a) is so high that slippage Compare factored moment Mu to the design
occurs over the 3-in. slice. What determines strength φ M n by superimposing both diagrams.
the adequacy of the beam? Assume that Consider all factors involved, including ACI-
loading is within the usual service range. 9.7.3.5 for cutting bars in the tension zone.
6.2 For the simply supported beam shown in the Assume that #3 U stirrups at 8-in. spacing are
figure for Problem 6.2, draw the design strength used in the vicinity of the cutoff point.
A B 5 kips/ft (total dead load)

n=8
17.5”
3”
3 – #9
2’– 0” 3”
12”
10’– 0”
Problem 6.1
A 14”
2 #8 21.5” 24”
1 #7 2 #8 cut here 1 #7
1’– 0” 4 #8
3’– 0” A
Section A–A
16’– 0” c–c of supports

Symmetrical about midspan
Problem 6.2
237
P roblems 237
3 bars 1 bar 3 – #7 20’

17.5”
4’– 0”
2’– 0” 6’– 0”
12”
A Section A–A
Problem 6.3
2 – #8B
2 – #9C
Bars C
Bar A Bars B 1 – #6A
20”
3’– 0” 3’– 0” 2’– 0”
3’– 0” 8’– 0”
#3 stirrup
14”
Problem 6.4
3 – #9 × 30’– 0” 6’– 0” 4’– 3”

2 – #9 × 10’– 3”
2 – #9 × 20’– 9” A B
25 21 ”
12” 4’– 0” 1–#9 4’– 0”

× 11’– 0” 12”
2–#8
20’– 0” 10’– 0” 14”
A Section A–A
Problem 6.5
6.4 If the beam in the figure for Problem 6.4 is to the 20-ft span. The loads are 1.63 kips/ft
carry uniformly distributed loads of 4.25 kips/ft dead load (including beam weight) and
live load and 2.21 kips/ft dead load (including 2.98 kips/ft live load.
beam weight), determine the adequacy of the 6.6 For the beam of Problem 6.5, determine the
bar cutoffs and the development of reinforce- acceptability of the cut locations at points A and
ment. Stirrups are #3 at 6-in. spacing where bars B near the right support.
A are cut, and #3 at 8-in. spacing where bars (a) Assume that #3 U stirrups are spaced at
B are cut. As part of the solution, compare the 10 in. in the vicinity of the cut locations.
factored moment Mu with the design strength (b) Assume that #3 U stirrups are spaced at
φ M n by superimposing the diagrams. Use 12 in. in the vicinity of the cut locations.
fc′ = 5000 psi and f y = f yt = 60, 000 psi. 6.7 A beam 14 in. wide by 24 in. deep (effective
6.5 For the beam shown in the figure for Problem depth = 21.5 in.) is used as the section for a 20-ft
6.5, use fc′ = 3000 psi and f y = f yt = 40, 000 psi. simply supported span having an 8-ft cantilever
(a) Neglecting any compression reinforcement at one end. The positive moment reinforcement is
effect, plot the design strength φ M n diagram 4–#9 bars, and the negative moment reinforcement
(positive moment over the 20-ft span and neg- is 4–#8 bars. The loading to be carried is a live
ative moment over the support). load of 2.01 kips/ft and a dead load of 1.86 kips/ft
(b) Check the development of reinforcement at (including beam weight). Determine the lengths
the simply supported end and at the point of bars (3-in. increments) if two of the four bars in
of inflection closest to the right support on both the positive and the negative moment regions
238
are to be terminated as soon as practicable. The Problems Using Concepts of Chapters 1

remaining bars are to be extended as required by Through 6
the ACI Code. Width of supports is 15 in., and #3
U stirrups spaced at 10 in. are used in any potential For all designs, use at least one bar cutoff, even if such
cutoff region. Verify your design by showing for cutoff seems impractical. Use 0.6 ρtc as a reinforce-
one set of axes the factored moment Mu envelope ment guideline where ρtc is the reinforcement ratio
and the provided design strength φ M n diagram. corresponding to εt = 0.005 (tension-control limit).
Use fc′ = 3500 psi and f y = f yt = 40, 000 psi.
6.8 In the figure for Problem 6.8, a cantilever slab 6.9 Design, including a design sketch, a rectangular
(e.g., for a retaining wall) varies in thickness from cantilever beam 14 ft long to carry a live load of
24 in. at its supported end to 12 in. at its free end, 2.7 kips/ft. The beam size may not exceed 15 in.
with 2 1 2 -in. clear cover over the reinforcement. wide and 24 in. deep. Use fc′ = 4000 psi and
Assuming that #7 bars at 9-in. spacing are effec- f y = f yt = 60, 000 psi.
tive at the supported end, at what distance from the 6.10 The floor system shown in the figure for Problem
supported end may every other #7 bar be cut off? 6.10 consists of 6-in. precast slab sections of 10-ft
No shear reinforcement is used. Verify your result span. The live load is 66 psf, and the maximum
by showing the provided design strength φ M n dia- depth available is 30 in. from the top of the floor
gram superimposed on the factored moment Mu slab. Completely design, as a simply supported
diagram. Use fc′ = 4000 psi and f y = 60, 000 psi. beam, the beam indicated on the figure. Use
Assume that the entire loading is earth pressure whole-inch increments for beam depth and width.
and neglect the beam weight. Use fc′ = 4000 psi and f y = f yt = 60, 000 psi.
0.7 k/ft
0.3 k/ft
12”
24”
13’– 0”
Problem 6.8
Live load = 50 psf

3 @ 10’– 0” = 30’– 0”
6” precast slab
Beam to be designed
10’– 0”
Typical section through
30’– 0” span beams
30’– 0”
Problem 6.10
CHAPTER 7
ANALYSIS OF CONTINUOUS
BEAMS AND ONE-WAY SLABS
7.1 INTRODUCTION
Reinforced concrete building systems commonly consist of floor slabs, beams, girders,
and columns cast to form a continuous, monolithic structure. Consider the plan of a typical
slab-beam-girder floor construction. In Fig. 7.1.1, section A-A through the slab shows that
the slab is supported on 10 beams. Intermediate beams such as B1, B2, and B3 are sup-
ported on the girders, whereas beams on the column lines such as B4, B5, and B6 are built
integrally with and supported directly by the columns. Girders such as G1, G2, and G3 are
also supported directly by the columns.
Rigid-frame bridge piers. (Photo by C. G. Salmon).

240
240 C H A P T E R 7 A N A L Y S I S O F C O N T I N U O U S B E A M S A N D O N E - W A Y S L A B S
G4 G5 G6
B3
B3
B6
G1 G2 G3
A A
B2
B2
B5
G1
B1
B1
B4
G4 G5 G6
Section A–A
B1 B2 B3
G4 G1 G1 G4
Beams monolithic with girders
B4 B5 B6
Beams monolithic with columns
G1 G2 G3
Girders monolithic with columns
Figure 7.1.1 Slab-beam-girder floor system.
7.2 ANALYSIS METHODS UNDER GRAVITY LOADS

An elastic analysis of the structural system under gravity loads shown in Fig. 7.1.1, which
may consist of, say, 10 or 15 stories or more, may be done by building a three-dimensional
model of the entire structure with the aid of standard structural analysis software.
Alternatively, experience has shown that under gravity loads, analysis of the floor system
may be simplified without great loss of accuracy by treating each floor level as a subas-
sembly consisting of the members in the level being considered (slab, beams, and girders)
and the columns above and below that level. Furthermore, for the purpose of gravity load
analyses, the far ends of adjacent columns may be assumed to be fixed (ACI-6.3.1.2), as
A further simplification can be made for regular building frames, by subdividing the struc-
ture into two-dimensional frames in the longitudinal and transverse direction of the building
(e.g., along B4, B5, and B6, and along G1, G2, and G3 in Fig. 7.1.1), and analyzing each
frame under its corresponding tributary load. Intermediate beams, such as B1, B2, and B3,
may be modeled as a continuous beam supported on girders G1 and G4. In this case, it may
be necessary to account for the torsional rigidity of the supporting girders depending on the
magnitude of the flexural stiffness of the beam relative to the torsional stiffness of the girders.
This is discussed later in Chapter 9 (Section 9.3).
Regardless of the chosen analysis approach, it cannot be emphasized enough that the
structural engineer must fully understand the capabilities and limitations of the software,
as well as the effect of the modeling assumptions. In addition, the engineer should be com-
petent in the use of traditional methods of analysis to be able to correctly interpret and
241
7 . 3 A R R A N G E M E N T O F L I V E L OA D F O R M O M E N T E N V E L O P E 241
Figure 7.2.1 Approximate deflected shape of a slab-girder-beam floor subassembly under a

uniformly distributed load. Far ends of columns (above and below the floor) are assumed to be fixed.
corroborate the computer output. Paramount to obtaining a realistic distribution of member

actions is the appropriate selection of the relative stiffness of the members; in particular, the
correct inclusion of the effects of cracking on member stiffness. Guidance for the appro-
priate computation of member stiffness is provided later (see Sections 16.20 and 16.21).
7.3 ARRANGEMENT OF LIVE LOAD FOR MOMENT

ENVELOPE
The live load positions that cause the largest bending moments in slabs, beams, and girders are
discussed in this section. To be able to establish the loading conditions appropriate for obtain-
ing the bending moment and shear envelopes, the reader should review the subject of influence
lines. Bending moments to be used in the design of columns are treated separately later.
Consider the continuous beam ABCDEFGH with its adjacent columns shown in
Fig. 7.3.1(a). The influence lines for bending moment at any point in the central portion
of span CD and at a section an infinitesimal distance to the left of support D are shown in
Fig. 7.3.1(b) and 7.3.1(c), respectively. From these influence lines the following cases for
uniform live load are indicated:
1. For maximum positive moment within a span, load that span and all other alternate
spans.
2. For maximum negative moment within a span, load the two spans adjacent to that
span and all other alternate spans (all the spans not loaded in 1).
3. For maximum negative moment at a support, load the two spans adjacent to that sup-
port and all other alternate spans.
4. For maximum positive moment at a support, load the two spans beyond each of the two
spans adjacent to that support and all other alternate spans (all the spans not loaded in 3).
Note that loading cases 1 and 2 in the preceding list are complementary; that is, their combina-
tion results in all spans being loaded. Loading cases 3 and 4 are also complementary. It may be
noted further that loading cases 1 and 3 are primary; that is, they result in moments of the same
sign as those due to dead load only. Loading cases 2 and 4, on the other hand, are secondary;
they result in moments opposite in sign to those due to dead load. When the secondary live load
moment is numerically larger than the dead load moment, there is moment reversal.
24
A B C D E F G H
(a)
+ +
+ +
– – –
(b) Influence line for moment at section in central portion of span CD
+
+ +
– –
– –
(c) Influence line for moment at section infinitesimally to left of support D
+
+ +
– – – –
(d) Influence line for moment at section between the fixed point and the support
+ +
– –
(e) Influence line for moment at the fixed point near right end of span
Figure 7.3.1 Influence lines for continuous spans.
The loading cases mentioned thus far involve loading entire spans; that is, no partial
loading of a span. These full span loading cases correctly give maximum and minimum
bending moment in the midspan region (roughly the middle 50–60%) of a span as well as
at the support. The correct maximum (or minimum) values are not obtained, however, for
approximately 20 to 25% of the span nearest the supports.
A qualitative examination of Fig. 7.3.1(b) and 7.3.1(c) shows that if an influence line
were drawn for a series of specific points along span CD between the midspan and the
support, the peak ordinate in the positive portion of such influence line would get smaller
and smaller. Also, the slope of the influence line at point C goes upward to the right for the
midspan influence line but downward to the right for the support D influence line.
As successive influence lines are drawn for the points along the span from midspan
to the support, there must be some location for which the slope of the influence line at
C is horizontal. Such point is called a fixed point, giving an influence line illustrated by
Fig. 7.3.1(e). Any loading on spans to the left of the span under study will cause no moment
at this fixed point. The fixed point is the closest location to the support for which full span
loadings give the correct maximum or minimum bending moments. The influence line for
a section between a fixed point and support is as shown by Fig. 7.3.1(d), indicating partial
loading for the span in question to obtain maximum or minimum bending moment.
For practical reasons, partial span loading for maximum or minimum bending moments
is rare in the design of building frames—the effect on the design is too small to justify the
effort. For the design of long spans, such as in highway bridges, partial span loading, as
indicated by influence lines, for locations between the fixed point and the support would
usually be considered.
The application to a six-span continuous beam with upper and lower columns is illus-
trated in the following example.
243
EXAMPLE 7.3.1
Determine the maximum and minimum moments at the middle and the ends of each
span in a beam-column frame with six equal spans as shown in Fig. 7.3.2. Assume that
the uniform live load wL is twice the uniform dead load wD, and that the flexural stiffness
(EI/L) of the columns is twice the stiffness of the beams. Express all moments in terms
of wL2 in which w = wD + wL and L is the span length.
2 2 2 2 2 2 2
A 1 B 1 C 1 D 1 E 1 F 1 G
2 2 2 2 2 2 2
Relative
stiffness
Figure 7.3.2 Six-span continuous frame of Example 7.3.1 with relative stiffnesses.
SOLUTION
All moments will be expressed numerically in terms of wL2 × 10 – 4. Seven loading condi-
tions, as shown in Fig. 7.3.3, need to be investigated. The frame can be analyzed by means
of any classical method of structural analysis (e.g., the moment distribution method) or
with the aid of standard structural analysis software. The moments at the ends of the beams
in each span for each of the seven loading conditions shown in Fig. 7.3.3 are summarized
in Table 7.3.1. In this table, the designer’s sign convention for bending moment, in which
a positive moment causes compression on the top side of the beam, is used.
The controlling values of moments at the left and right ends of each span, M L and M R ,
for the various critical conditions are taken from Table 7.3.1 and entered in Table 7.3.2.
DL
DL + LL
LC1
Max positive moment in spans AB, CD, and EF
LC2
Max negative moment at B
LC3
Min negative moment at B
LC4
Max negative moment at C
LC5
Min negative moment at C
LC6
Max negative moment at D
LC7
Min negative moment at D
Figure 7.3.3 Loading conditions for six-span continuous frame.

(Continued)
24
TABLE 7.3.1 BEAM END MOMENTS FOR EXAMPLE 7.3.1
Joint A B C D E F G
Member AB BC CD DE EF FG
ML MR ML MR ML MR ML MR ML MR ML MR
LC1 –712 –804 –340 –331 –777 –777 –334 –332 –772 –790 –399 –184
LC2 –668 –911 –888 –729 –325 –337 –778 –777 –331 –340 –804 –712
LC3 –227 –293 –241 –374 –785 –774 –333 –332 –772 –790 –399 –184
LC4 –187 –390 –745 –878 –882 –732 –325 –336 –772 –790 –399 –184
LC5 –709 –813 –385 –225 –228 –379 –786 –774 –330 –340 –804 –712
LC6 –712 –805 –343 –323 –731 –883 –883 –731 –323 –343 –805 –712
LC7 –184 –398 –787 –780 –378 –228 –228 –378 –780 –787 –398 –184
Note: Values of moments in wL2 × 10–4 using designer’s sign convention.
The moment at the midspan M s may be determined by superposition of the effect of

end moments with that of the simply supported beam moment due to transverse loading,
1
M s = M 0 − ( M L + M R )
2
where M 0 is the moment at the midspan for a simply supported beam. In the above equa-
tion, M 0, ML, and M R are taken as positive values. When the beam end moments are not
equal, the maximum moment in the span does not occur at midspan, but its value is close
to that at midspan.
Because of symmetry, the controlling values obtained for a given span would also
occur in the symmetric span when the frame is loaded with the mirror image of the load
condition being considered. For example, the controlling values obtained for maximum
positive moment in spans AB, CD, and EF (lines 1, 2, and 3 in Table 7.3.2), would also
occur in spans FG, DE, and BC, respectively, if the frame were to be loaded with the
mirror image of load condition LC1. Similarly, the controlling values obtained for mini-
mum positive or maximum negative moment at midspan in spans BC, DE, and FG (lines
4, 5, and 6 in Table 7.3.2), would also occur in spans EF, CD, and AB, respectively.
For the purpose of illustration, the moment diagram for the maximum and minimum
positive moments at midspan, using the results of the first loading condition, LC1, in
Fig. 7.3.3, and its mirror image is shown in Fig. 7.3.4. First, the simple beam moment
diagrams for the total load and for dead load only are drawn to scale in Fig. 7.3.4(a). Next,
the end moments for each span are taken from Table 7.3.2 and plotted in Fig. 7.3.4(b).
Because of symmetry, only the moment diagrams for the first three spans are shown. The
final moment diagrams in Fig. 7.3.4(b) are drawn by rotating the baselines for zero end
moments in Fig. 7.3.4(a) to those connecting the end moments in Fig. 7.3.4(b).
In this example, the dead load has been applied on all the spans in each of the seven
loading conditions of Fig. 7.3.3. This has been done because an important purpose of the
example is to justify the use of approximate moment coefficients as discussed in the next
section. An alternative approach would be to use eight loading conditions: one condition
for dead load only plus the same seven live load conditions but without the dead load
applied simultaneously. Separation of the effects of dead and live load is desirable, because
if a loading combination that includes wind-or earthquake-induced loads must be consid-
ered, different live load factors must be used according to ACI-5.3.1 (see Section 2.7).
(Continued)
245
TABLE 7.3.2 SUMMARY OF RESULTS—CONTROLLING VALUES OF MOMENTS AT BEAM

ENDS FROM TABLE 7.3.1
Line Span ML MR Value of ML and 1

Number MR from Load M s = M 0 − 2 � ( M L + M R )
Condition No.
1 For maximum AB –712 –804 LC1 +492

2 positive CD –777 –777 LC1 +473 [M0 = +1250]
3 moment at EF –772 –790 LC1 +469
midspan
4 For minimum BC –340 –331 LC1 +81
5 positive or DE –334 –332 LC1 +84 [M0 = +417]
6 maximum FG –399 –184 LC1 +125
negative
moment at
midspan
7 For maximum A of AB –712 (–804) LC1
8 negative B of AB (–668) –911 LC2
9 moment B of BC –888 (–729) LC2
10 C of BC (–745) –878 LC4
11 C of CD –882 (–732) LC4
12 D of CD (–731) –883 LC6
13 For minimum A of AB –184 LC7
14 negative or B of AB –293 LC3
15 maximum B of BC –241 LC3
16 positive C of BC –225 LC5
17 moment C of CD –228 LC5
18 D of CD –228 LC7
Note: Values of moments are shown in wL2 × 10–4 using designer’s sign convention. Numbers in parentheses are to be used in shear envelope computations
(see Section 7.6).
1250 wL2 (10–4) 1250 wL2 (10–4) 1250 wL2 (10–4)

wL
417 wL2 417 wL2 417 wL2

wD +
(10–4) (10–4) (10–4)

w=
1 w
= 3
wD
(a) Simple beam moment diagrams
492 469 473
A B C D
125 81 84
184
340 331 332 334
399
712 804 790 772 777 777

(b) Moment diagrams for maximum and minimum
positive midspan moments
Figure 7.3.4 Moment diagrams for the first three spans under LC1 of Example 7.3.1.
246
7.4 ACI CODE—A RRANGEMENT OF LIVE LOAD

AND MOMENT COEFFICIENTS
For the design of floor or roof systems under gravity loads, ACI-6.4.1 and ACI-6.3.1.2 permit
the assumption that the live load is applied only to the level under consideration, with its upper
and lower columns fixed at their far ends, as discussed earlier in Section 7.2. Furthermore,
for one-way slabs and beams, ACI-6.4.2 allows computation of the (a) maximum positive
moment near midspan by positioning the live load on the span under consideration and on
alternate spans (LC1 in Fig. 7.3.3) and (b) maximum negative moment at a support by posi-
tioning the live load on adjacent spans only (LC2, LC4, and LC6 in Fig. 7.3.3).
Alternatively, ACI-6.5 allows an approximate method for calculating the required fac-
tored moments (and shears—see Section 7.6) for nonprestressed continuous beams and
one-way slabs that satisfy all of the following (ACI-6.5.1):
(a) Members are prismatic.
(b) Loads are uniformly distributed.
(c) The live load does not exceed three times the dead load.
(d) There are at least two spans.
(e) The longer of two adjacent spans does not exceed the shorter by more than 20%.
Table 7.3.2 shows that for the six-span frame of Fig. 7.3.2, which would meet the require-
ments of ACI-6.5.1, critical values of moments may vary within the following limits:
Exterior span:
Exterior end –0.0184wL2 and –0.0712wL2
Midspan +0.0125wL2 and +0.0492wL2
Interior end –0.0293wL2 and –0.0911wL2
First interior span:
Second interior span:
Similar values may be worked out for other values of the relative stiffness between the
columns and the beams and of the wL / wD ratio. It may be observed that the maximum posi
tive moments in the first and second interior spans are similar, that the maximum positive
moment in the exterior span is higher than that in the interior spans, that the maximum
negative moment at the interior end of the exterior span has the largest numerical value,
and that the maximum negative moments at both ends of all interior spans are about equal.
Based on these and similar analyses, ACI-6.5.2 provides approximate moment coef-
ficients for nonprestressed continuous beams and one-way slabs, as shown in Table 7.4.1.
A comparison of these coefficients with the largest possible theoretical values [7.1] is
shown in Table 7.4.1. Certainly, the largest possible theoretical values will be for the case
of wL / wD = 3, which is the limit set forth in ACI-6.5.1. In this instance, secondary live load
moments with signs opposite to that of dead load occur infrequently; if they do occur, their
values are small. Thus, as long as the ratio of live load to dead load does not exceed 3 and
span lengths do not differ considerably, the ACI moment coefficients will be reasonably
close to the theoretical values and, in general, conservative.
Note that the ACI moment coefficients are given in terms of wL2n, where Ln is the clear
span for positive moment and the average of the two adjacent clear spans for negative
moment—negative moments being those at the face of supports, not at the centerline of
support. On the other hand, the theoretical coefficients are in terms of wL2 , in which L is
247
7 . 5 AC I M O M E N T D I AG R A M S 247
TABLE 7.4.1 COMPARISON OF ACI MOMENT COEFFICIENTS WITH LARGEST THEORETICAL

COEFFICIENTS [7.1]
Largest Theoretical Coefficients
Number of
Location of Section ACI Value Spans (ΣKcol)/Kbma wL/wD

Positive moment at
End spans
If discontinuous end is 1
+ +0.094 3 0 3
unrestrained 11
If discontinuous end is 1
+ +0.073 3 0.5 3
integral with the support 14
1
Interior spans: + +0.063 4 or more 0.5 3
16

Negative moment at
Exterior face of first
interior support:
1
Two spans − –0.111 2 0.5 3
9
1
More than two spans − –0.107 4 or more 0.5 3
10
Other faces of interior 1
− –0.092 4 or more 2 3
supports: 11
Face of all supports for:
(a) slabs with spans
not exceeding 10 ft, or
(b) beams and girders
where (ΣKcol)/Kbeam exceeds 1
− –0.083 any number ∞ any ratio
8 at each end of the spana 12
Negative moment at
Interior faces of exterior
supports for members
built integrally with their
supports:
Where the support is a 1 –0.036 4 or more 0.5 3
−
spandrel beam or girder 24 –0.050 4 or more 1 3
Where the support is a 1
− –0.064 4 or more 2 3
column: 16
a K = EI/L
the distance between centerlines of supports, and coefficients for negative moments refer to
those at the centerlines of support. Although span lengths between centerlines of supports
are always used in elastic analysis, the ACI Code states that for beams (ACI-9.4.2.1) and
one-way slabs (ACI-7.4.2.1) built integrally with supports, moments computed at faces of
supports may be used for design.
7.5 ACI MOMENT DIAGRAMS

In designing any span in a multispan continuous frame subjected to live load with the
moment coefficients, two primary sets of shear and moment diagrams are inherently
being assumed. In the general case, one will result from the loading position that causes
248
maximum positive moment within the span, and the other will result from assuming that the
maximum negative moments occur simultaneously at both ends. Actually, the loading pos-
ition that causes maximum negative moment at one end is different from that which causes
maximum negative moment at the other end. However, by assuming that both maximum
negative end moments occur simultaneously, a critical curve having greater magnitude than
either of the two actual curves is obtained.
The ACI moment coefficients (ACI-6.5.2) shown in Table 7.4.1 are the common values
from the two primary conditions as described in the preceding paragraph. No secondary
moment coefficients are suggested by the ACI Code, the reason being that as long as the
design ratio of live to dead load is limited to 3, moment reversal will be unlikely to occur;
that is, there can be only positive moment in the midspan region and only negative moment
in the support region.
Figures 7.5.1 through 7.5.4 show the two primary sets of shear and moment diagrams
to be used in the design of continuous spans in accordance with the ACI moment coeffi-
cients. These diagrams are applicable to the actual clear span, which is also used to com-
pute the positive moment. For negative moment, Ln is the average of adjacent clear spans
(ACI-6.5.2).
The reader should utilize the fundamentals of shear and moment diagrams to verify the
numerical ordinates on these diagrams. For instance, in case of maximum positive moment
in Fig. 7.5.2(a), the distance x from the left support to the point of zero shear may be
1
0.0736 wu Ln2 w L 2
10 u n
(given)
wu wu
0.4264 wu Ln 0.5736 wu Ln 0.4000 wu Ln 0.6000 wu Ln

Ln Ln
0.4264 Ln 0.4000 Ln
0.5736 Ln 0.6000 Ln
+0.4264 wu Ln +0.4000 wu Ln
–0.5736 wu Ln –0.6000 wu Ln
1
+ 11 wu Ln2 (given)
+0.0800 wu Ln2
0.4264 Ln 0.4264 Ln 0.4000 Ln 0.4000 Ln
–0.0736 wu Ln2
1
– 10 wu Ln2 (given)
(a) Maximum in the positive zone (b) Maximum in the negative zone
Figure 7.5.1 Exterior span with discontinuous end unrestrained.

249
7 . 5 AC I M O M E N T D I AG R A M S 249
determined from the relationship that the change of moment between any two sections is
equal to the area of the shear diagram between these two sections. Thus
wu x 2  1 1
=  +  wu L2n
2  24 14 
from which
x = 0.4756 Ln
Also, the distance x between the section of maximum positive moment to the point of
zero moment is
wu x 2 1
= wu L2n
2 14
from which
x = 0.3780 Ln
Any time a designer uses moment coefficients for determining the factored moments, as
permitted by the approximate method of ACI-6.5, the moment diagrams that correspond to
1 1 1
w L 2 (given) 0.0661 wu Ln2 w L 2 (given) w L 2 (given)
24 u n wu 24 u n wu 10 u n

Ln Ln
0.4756 Ln 0.5244 Ln 0.4417 Ln 0.5583 Ln
+0.4756 wu Ln +0.4417 wu Ln
–0.5244 wu Ln –0.5583 wu Ln
1
+ w L 2 (given)
14 u n
+0.0558 wu Ln2
0.3780 Ln 0.3780 Ln 0.3341 Ln 0.3341 Ln

1 2 1
– w L (given) – w L 2 (given)
24 u n 24 u n
–0.0661 wu Ln2
1 w L 2 (given)
–
10 u n
Figure 7.5.2 Exterior span with exterior support built integrally with spandrel beam or girder.
250
1 w L 2 (given) 1 w L 2 (given) 1 w L 2 (given)

16 u n wu 0.0450 wu Ln2 16 u n wu 10 u n
Ln Ln
0.5175 Ln 0.4825 Ln 0.4625 Ln 0.5375 Ln
+0.5175 wu Ln +0.4625 wu Ln
–0.4825 wu Ln –0.5375 wu Ln
+0.0445 wu Ln2
+ 1 wu Ln2 (given)
14
0.2983 Ln 0.2983 Ln
0.3780 Ln 0.3780 Ln – 1 wu Ln2 (given)

16
–0.0450 wu Ln2
– 1 2
w L (given) – 1 wu Ln2 (given)
16 u n 10
Figure 7.5.3 Exterior span with exterior support built integrally with column.
such coefficients should be used when establishing bar bend or cutoff locations. The use of
moment coefficients implies a statically compatible moment diagram.
7.6 SHEAR ENVELOPE FOR DESIGN

Inasmuch as the design of shear reinforcement depends on the variation of shear forces
along the span, it is necessary to compute the maximum factored shear force due to the
combination of dead and live loads at all sections. When the live load consists of important
concentrated loads, such as in the design of highway bridge spans, accurate calculations of
such maximum shears at all sections must be performed. The proper position of live load
for maximum shear at a section can be determined by examining the influence line for shear
at that section along the span. For instance, the influence lines for end shear at C of span CD
and for the shear at midspan of CD in the frame of Fig. 7.6.1(a) are shown in Fig. 7.6.1(b)
and 7.6.1(c), respectively. Figure 7.6.1(b) shows that the position of uniform live load for
maximum end shear at C is identical to that for maximum negative moment at C [see
Fig. 7.3.3 (LC4)]. Likewise, the position of uniform live load for maximum end shear at D
is identical with that for maximum negative moment at D. From Fig. 7.6.1(c), it is apparent
that partial span loading of uniform live load is indicated to give maximum shear at any
point within the span. The partial span loading to obtain the correct maximum shear at loca-
tions within the span has already been used in the shear strength design examples of Section
5.12 for those statically determinate situations.
251
7.6 SHEAR ENVELOPE FOR DESIGN 251
0.0625 wu Ln2 0.0625 wu Ln2 1 w L 2 (given) 1 w L 2 (given)

wu 11 u n wu 11 u n
Ln Ln
0.5000 Ln 0.5000 Ln 0.5000 Ln 0.5000 Ln
+0.5000 wu Ln +0.5000 wu Ln
–0.5000 wu Ln –0.5000 wu Ln
1
+ wu Ln2 (given)
16
+0.0341 wu Ln2
0.3535 Ln 0.3535 Ln 0.2611 Ln 0.2611 Ln
–0.0625 wu Ln2 –0.0625 wu Ln2
– 1 wu Ln2 (given) – 1 wu Ln2 (given)

11 11
Figure 7.5.4 Interior span.
A B C D E F G
(a)
+ +
– – –
(b) Influence line for maximum end shear at C of span CD
+
+ +
– – – –
(c) Influence line for maximum shear at midspan of span CD
Figure 7.6.1 Influence lines for shear in a continuous frame.

25
In buildings of usual types of construction, spans, and story heights, wherein the ide-
alized rigid frame, such as shown in Fig. 7.6.1(a), is taken into consideration, the use of
partial span loading of uniform live load is commonly ignored, although theoretically it
is necessary for the computation of maximum shear at any section within the span. When
partial span loading is not considered necessary, the maximum shears at the ends can be
used alone to establish an approximate shear envelope. Note that the loading condition for
maximum shear at one end is different from that for maximum shear at the other end. These
two critical shear diagrams for each span, when a continuity analysis is performed, may be
easily obtained by using the values of the negative end moments (including those in paren-
thesis) contained in lines 7 through 12 of Table 7.3.2.
When the ACI moment coefficients are used, it is generally assumed that the shear dia-
grams accompanying the critical moment diagrams as shown in Figs. 7.5.1 through 7.5.4
may be used in the design. Alternatively, for beams and one-way slabs satisfying ACI-
6.5.1 [see (a) through (e) in section 7.4], ACI Table 6.5.4 allows the use of a shear force
of 1.15wu Ln /2 at the exterior face of the first interior support. A shear force of wuLn /2 is
permitted at the face of all other supports.
SELECTED REFERENCE
7.1 A. J. Boase and J. T. Howell. “Design Coefficients for Building Frames,” ACI Journal,

Proceedings, 36, September 1939, 21–36.
PROBLEMS
7.1 Compute and draw to scale the bending moment 7.3 For the beams of the frame in the figure, compute
and shear diagrams for the loading condi- and draw the bending moment envelope using
tions 1 through 7 of Fig. 7.3.3; that is, ver- the coefficients of ACI-6.5.2. (If Problem 7.2 has
ify the results given in Tables 7.3.1 by using also been solved, compare the moments by giv-
any structural analysis method assigned by the ing the percentage difference in the maximum
instructor. values obtained by coefficients as compared with
7.2 Compute and draw to scale the envelope of the more exact values of a structural analysis.)
bending moments (diagram showing range 7.4 Consider an equal-span, uniform-section con-
over which bending moment may vary) due tinuous beam over many supports. Compute and
to factored loads for the beams of the frame show diagrams for dead load coefficients of wL2
of the figure for Problems 7.2 and 7.3 using for moments at critical locations in the exterior
any structural analysis method assigned by the and first interior spans. Could the coefficients for
instructor. The uniform factored dead load is the first interior span be applied appropriately to
1 kip/ft, and the uniform factored live load is the other interior spans? Recommend dead load
2 kips/ft. coefficients for equal spans.
Symmetrical
about CL
Relative
stiffness, I/L
2 2 2 2 2
A 5 B 4 C 5 D 6 E
2 2 2 2 2
16’– 0” 18’– 0” 16’– 0” 14’– 0”

253
PROBLEMS 253
7.5 Repeat Problem 7.4 for the case of alternate 7.7 For an equal-span, uniform-section continuous
spans that are 20% longer than the others (1.2L), beam over many supports, compute and show
taking the exterior span as a short one (L). diagrams for live load coefficients in terms of
7.6 For an equal-span, uniform-section continuous wL2 for the maximum positive moment in
beam over many supports, compute and show (a) the exterior span
diagrams for live load coefficients in terms of (b) the first interior span
wL2 for maximum negative moments at (c) a typical interior span
(a) the first interior support
(b) the second interior support Recommend live load coefficients.
(c) a typical interior support
Recommend live load coefficients.

CHAPTER 8
DESIGN OF ONE-WAY SLABS
8.1 DEFINITION
One of the most common types of floor construction is the slab-beam-girder system, as
briefly described in Section 7.1. The slab panel, bounded on its two long sides by the beams
and on its two short sides by the girders, is usually at least twice as long as it is wide. In
such a condition, the dead and live load acting on the slab area may be considered as being
entirely supported in the short direction by the beams (see Section A-A in Fig. 7.1.1), hence
the term “one-way slab.” Two-way floor systems, with or without beams on column lines,
are treated in Chapter 16, and ribbed-joist floor construction is described in Sections 9.10
and 9.11.
The determination of an optimum floor framing plan—that is, the spacing of columns,
beams, and girders—depends on both the functional and the structural requirements. In
most cases preliminary calculations are necessary for several different layouts and, after
comparison, the most suitable and economical plan is chosen.
8.2 ANALYSIS METHODS
Because the loading on a one-way slab is nearly all transferred in the short direction, such
a slab continuous over several supports may be treated as a beam. Because sufficiently
accurate results are obtained, ACI-6.5.1 permits the use of moment and shear coefficients
in the case of two or more approximately equal spans (the larger of two adjacent spans not
exceeding the shorter by more than 20%) with loads uniformly distributed, where the unit
live load does not exceed three times the unit dead load. These coefficients are in terms
of clear span Ln and the values given are for critical locations, that is, faces of support for
shears and negative moments and midspan regions for positive moments.
When the conditions of ACI-6.5.1 are not satisfied, a structural analysis, typically a first-
order elastic analysis, is required (see Section 7.2). Various live load arrangements accord-
ing to ACI-6.4 should be investigated in order to maximize design negative and positive
moments and shear forces (see Section 7.3). ACI-6.4.2 permits the determination of maxi-
mum negative moments at supports by applying the live load on adjacent spans, while max-
imum positive moment near midspan is determined by applying live load on alternate spans.
For one-way slabs built integrally with the supports and with clear spans not exceeding
10 ft, ACI-6.6.2.3 permits the use of a model consisting of a continuous flexural member on
knife-edge supports with span lengths equal to the clear span lengths of the slab. As such a
model would lead to zero moment at the end supports, the application of a moment equal
to wu Ln 2 / 24 is recommended, where wu is the factored distributed load and Ln is the clear
span length. For clear span lengths greater than 10 ft, no explicit guidelines for the use of a
25
8 . 3 S lab D esign 255
Escala Apartments, Seattle, WA (photo courtesy of Cary Kopczynski & Associates).
simplified model are provided in the ACI Code. The writers believe that in lieu of a three-
dimensional analysis of the entire floor system, a simplified model consisting of a continu-
ous beam on knife-edge supports can still be used. Rather than using the clear span length,
however, a span length equal to the distance between centerlines of supporting beams is
recommended. Such a model has been found to be adequate for determination of negative
moments in beams supported by girders [8.1]. Rigid zones may be used to model the inter-
section between the slab and supporting beams, which will facilitate the determination of
slab design moment and shear at the faces of the supporting beams. As in the simplified
model allowed in ACI-6.6.2.3, the application of a moment equal to wu L2n / 24 at exterior
supports is recommended.
8.3 SLAB DESIGN
In designing a one-way slab, a typical imaginary strip 12 in. wide is usually considered.
The continuous slab may then be designed as a continuous beam having a known width of
12 in.; the slab thickness, however, is unknown.
The thickness of the slab depends on the deflection, bending, and shear require-
ments. Deflection requirements are imposed to prevent excessive deformations that might
adversely affect the serviceability of the structure. According to ACI Table 7.3.1.1, one-way
slabs not supporting or not attached to partitions or construction likely to be damaged by
large deflections must have at least a minimum slab thickness (for Grade 60 steel) of L /20,
L /24, L /28, or L /10 depending on whether L is the length of a simply supported, a one-
end-continuous, a both-ends-continuous, or a cantilever span. Thinner slabs are permitted,
however, provided deflections are calculated and determined to be less than the limits in
ACI Table 24.2.2. If the slab supports or is attached to construction likely to be damaged by
large deflections, deflections must also be computed and shown to satisfy the limits of ACI
Table 24.2.2. An extended treatment of deflections can be found in Chapter 12.
256
256 C hapter 8 D esign of O ne - W ay S labs
TABLE 8.3.1 AVERAGE AREA PER FOOT OF WIDTH PROVIDED BY VARIOUS BAR SPACINGS
Bar Size Nominal Spacing of Bars in Inches

Number Diameter (in.)
4 5 6 7 8 9 10 12 14 16 18
3 0.375 0.33 0.27 0.22 0.19 0.17 0.15 0.13 0.11 0.095 0.83 0.074
4 0.500 0.59 0.47 0.39 0.34 0.29 0.26 0.24 0.20 0.17 0.15 0.13
5 0.625 0.92 0.74 0.61 0.53 0.46 0.41 0.37 0.31 0.26 0.23 0.20
6 0.750 1.33 1.06 0.88 0.76 0.66 0.59 0.53 0.44 0.38 0.33 0.29
7 0.875 1.80 1.44 1.20 1.03 0.90 0.80 0.72 0.60 0.52 0.45 0.40
8 1.000 2.36 1.88 1.57 1.35 1.18 1.05 0.94 0.79 0.67 0.59 0.52
In most cases slab thickness is controlled by deflection requirements rather than flex-
ural strength requirements. Minimum thickness required to resist the factored moment Mu ,
however, can be checked in accordance with the principles of Chapter 3. In this case, the
desired coefficient of resistance Rn can be computed for the required effective depth (and
slab thickness) determined using M n = Mu / φ = Rn bd 2 . In equal continuous spans, the nega-
tive moment at the exterior face of the first interior support is the largest; therefore, this neg-
ative moment is used to determine minimum slab thickness based on bending requirements.
Table 8.3.1 lists reinforcement areas per foot of width for various bar sizes and spacings,
which can be used once the required area of flexural reinforcement per 12-in.-wide strip
has been determined.
The shear requirement does not usually control, but it must be checked. Because of
practical space limitations, shear reinforcement is not used in typical slabs; thus the
governing factored shear Vu , which in equal continuous spans occurs at the exterior
face (at a distance d therefrom) of the first interior support, must be kept below φ Vc of
ACI-22.5.5.1.
According to ACI Table 20.6.1.3.1, the concrete protective covering for reinforcement
(#11 and smaller) in slabs shall be not less than 3 4 in. at surfaces not exposed directly to the
ground or to the weather. Also, when the top of a monolithic slab is the wearing surface and
when unusual wear is expected as in buildings of the warehouse or industrial class, it has
been customary to use an additional depth of 1 2 in. of concrete protective covering over that
required by the design of the member. The ACI Code does not require extra slab thickness
for wearing surface, but ACI-7.3.1.2 permits, at the discretion of the designer, a monolithic
floor finish to be considered as part of the structural member. For nonstructural purposes,
any concrete floor finish may be considered as part of the required cover (ACI-20.6.1.2).
EXAMPLE 8.3.1
Establish the thickness of the floor slab shown in the second-floor framing plan of
Fig. 8.3.1 (see Fig. 8.4.2 for a section through slabs 2S1-2S2-2S3) for a superimposed
dead load of 20 psf and a service live load of 100 psf. Use fc′ = 4000 psi, f y = 60,000 psi,
and the ACI Code. Assume an exterior staircase so that no openings are to be made in
the slabs.
(Continued)
257
8 . 3 S lab D esign 257
SOLUTION
Because live load does not appear to exceed three times the dead load, the design will be done
in accordance with the ACI moment and shear coefficients (see Sections 7.4, 7.5 and 7.6).
(a) Minimum thickness based on deflection control requirements. For spans with one
end and both ends continuous, the respective minimum thicknesses h from ACI
Table 7.3.1.1 are L /24 and L /28.
L 13(12)
min h = = = 6.5 in. (for 2S1)
24 24

L 13(12)
min h = = = 5.6 in. (for 2S 2 and 2S 3)
28 28
Assume a 6 1 2 -in.-thick slab.1 The weight of the slab is (6.5/12)(0.15) = 0.081 kips/sq ft.
2S1 2S2 2S3 2S3 2S3 2S3 2S2 2S1
2B1
2G1 2G2 2G2 2G1
3 @ 26’ = 78’
A A
2S1 2S2 2S3 2S3 2S2 2S1
2B2
2S1 2S2 2S3 2S3 2S2 2S1

2B1
4 @ 39’ = 156’
Figure 8.3.1 Second-floor framing plan (for Section A-A, see Fig. 8.4.2).
(b) Flexural strength requirements. Assume width of supporting beams to be 14 in.

w = (0.081 + 0.02) = 0.101kips/ft /ft of width
D
wL = 0.1 kips/ft /ft of width
wu = 1.2(0.101) + 1.6(0.1) = 0.281 kips/ft /ft of width

14
clear span = 13 − = 11.83 ft
12
From Table 7.4.1, the negative moment at the exterior face of the first interior support is
1
Mu = (0.281)(11.83)2 = 3.94 ft-kips /ft of width
10
Strictly speaking, the slab thickness must be checked against bending requirements.
Unless unusually large loads are applied to the slab, however, the minimum thickness
required for deflection control will be adequate. For a tension-controlled section,
Mu 3.94
Mn ≥ = = 4.38 ft-kips /ft of width
φ 0.90
(Continued)
1 Note that a thinner slab may be permitted by ACI-7.3.1.1 if it can be shown that the deflection limits of ACI Table 24.2.2
are satisfied.
258
Assuming #4 bars are used (db = 0.5 in.) and using a minimum cover of 3 4 in. (surface
not exposed directly to the ground or the weather), the effective depth of the slab is
db
d = h− − 0.75 in. = 5.5 in.
2
Required area of tension steel. Following Eq.(3.8.4b), Eq. (3.8.4a) and Eq. (3.8.5)
Mu 4.38(12, 000)
Rn = = = 145 psi
φ bd 2
12(5.5)2
fy 60, 000
m= = = 17.65
0.85 fc′ 0.85(4000)
As 1 2 mRn  1  2(17.65)(145) 
ρ= = 1 − 1 − = 1− 1− = 0.0025
bd m  
f y  17.65  60, 000 
As = ρ bd = 0.0025(12)(5.5) = 0.16 sq in. /ft of width
(c) Shear strength requirement.
wu Ln (0.281)(11.83)
max Vu = 1.15 = 1.15 = 1.92 kips /ft of width
2 2
The design shear strength φ Vc for a member without shear reinforcement is
( )
φ Vc = φ 2 λ fc′ bd = 0.75(2(1.0) 4000 )(12)(5.5)
1
1000

= 6.26 kips / ft off width > 1.92 kips / ft of width
The slab is acceptable for a member without stirrups. Note that the shear force at a dis-
tance d from the face of support could have been used, in which case the factored shear
would have been even less than 1.92 kips/ft of width.
8.4 CHOICE OF REINFORCEMENT
Flexural Reinforcement
The choice of flexural reinforcement depends primarily on the steel area and secondarily on
development length requirements. The steel areas required at the principal sections, namely,
those at the middle and at the ends of each span, are first computed. The calculated areas of flex-
ural reinforcement cannot be less than the minimum specified in ACI Table 7.6.1.1 as follows:
For
f y < 60, 000 psi, As ,min = 0.0020bh (8.4.1a)
For
0.0018(60, 000)
f y ≥ 60, 000 psi, As ,min = bh ≥ 0.0014bh (8.4.1b)
fy
259
8 . 4 C hoice of R einforcement 259
Once a tentative choice of reinforcement has been made based on flexural strength require-
ments, development length requirements (Chapter 6) are examined (see Fig. 8.4.1 for a
possible arrangement of reinforcement). Critical sections for development of reinforcement
are those where maximum stress in the reinforcement occurs, or where bent or terminated
reinforcement is no longer required to resist flexure (ACI-7.7.3.2). Reinforcement should
be extended at least the greater of d or 12 db beyond the point at which it is no longer
required to resist flexure except at simple supports or free ends of cantilevers (ACI-7.7.3.3).
Also, continuing reinforcement must be extended at least Ld beyond the point at which the
bent or terminated reinforcement is no longer required (ACI-7.7.3.4). Further, tension rein-
forcement should not be terminated in a tensile zone. ACI-7.7.3.5, however, allows tension
reinforcement to be terminated in the tensile zone of a slab without shear reinforcement
when Vu < ( 2 3 ) φ Vc , or when continuing reinforcement, consisting of #11 bars or smaller, is
at least twice the required area of tension steel and Vu ≤ ( 3 4 ) φ Vc .
At simple supports, at least one-third of the maximum positive moment reinforcement
should be extended into the support (ACI-7.7.3.8.1). For straight bars (i.e., bars not ter-
minated by a standard hook or mechanical anchorage), the diameter of this reinforcement
must be such that (ACI-7.7.3.8.3)
Ld ≤ α M n /Vu + La (8.4.2)
where α = 1.3 if the reinforcement is confined by a compressive reaction force or α = 1.0

otherwise, and La is the embedment length beyond the center of the support. In Eq. (8.4.2),
M n is to be calculated assuming all tension reinforcement yields. This same requirement for
bar diameter applies to positive moment reinforcement at inflection points, where La in Eq.
(8.4.2) is the embedment length beyond the point of inflection, but not to exceed the greater
of d and 12 db. In this case, α = 1.0, since there is no compressive reaction force that could
confine the bar being developed. For continuous supports, no less than one-fourth of the
maximum positive moment reinforcement must be extended at least 6 in. into the support
(ACI-7.7.3.8.2).
At least one-third of the negative moment reinforcement at supports must be extended
at least the greatest of d, 12 db, and Ln /16 beyond the point of inflection (ACI-7.7.3.8.4).
Spacing s between flexural bars in slabs shall not exceed three times the slab thickness
nor more than 18 in. (ACI-7.7.2.3). Bar spacing, however, must also satisfy the maximum
spacing required for crack control, as discussed in Section 12.18.
Shrinkage and Temperature Reinforcement

Reinforcement perpendicular to the main flexural reinforcement (see Fig. 8.4.1), in
accordance with ACI-7.6.4, is required in one-way slabs to resist stresses and control
cracking due to shrinkage and temperature changes. ACI-24.4.3.2 requires the same min-
imum amount of shrinkage and temperature reinforcement as that for flexure [Eqs. (8.4.1)]
spaced no farther apart than the lesser of 5 times the slab thickness or 18 in. (ACI-
24.4.3.3). This reinforcement must be capable of developing its tensile yield strength
(ACI-24.4.3.4). Note that the Code-required amount of shrinkage and temperature rein-
forcement is a minimum; larger amounts of reinforcement may be needed in certain cases
to ensure adequate control of cracking caused by stress-induced shrinkage and changes
in temperature [8.2].
Shrinkage and temperature reinforcement may be placed near the top or the bottom face
of the slab, or distributed on both faces, the latter option being more appropriate for thicker
slabs. This may be provided in the form of deformed bars, welded wire reinforcement or
welded deformed bar mats.
260
Top flexural bars
Bottom flexural
bars
Shrinkage &
temperature
reinforcement
Figure 8.4.1 One possible arrangement of slab reinforcement.
Example 8.4.1
Choose the arrangement of reinforcement for the 6 1 2 -in. floor slab 2S1, 2S2, and 2S3,
of Example 8.3.1.
SOLUTION
(a) Steel area requirements. The ACI moment coefficient, the bending moment, the steel
area required, and the tentative choice of reinforcement at the critical sections are
shown in lines 1 to 6 of Table 8.4.1. The arrangement of reinforcement is shown in
Fig. 8.4.2. The flexural reinforcement arrangement must satisfy the minimum rein-
forcement and detailing requirements of ACI-7.6.1 and ACI-7.7. Also, reinforce-
ment perpendicular to the flexural bars must satisfy the shrinkage and temperature
reinforcement requirement of ACI-7.6.4, which refers to ACI-24.4. For Grade 60
steel, the minimum required flexural reinforcement ratio based on the gross concrete
area is 0.0018. Thus

As,min = 0.0018 (12 )(6.5) = 0.14 sq in./ft of width
For #4 bars, a spacing of 16 in. would correspond to an area of 0.15 sq in. per foot of
width (Table 8.3.1), which satisfies the minimum reinforcement requirement. Maximum
spacing is given by three times the slab thickness or 18 in.; thus

max s = 3 (6.5) = 19.5 in. or 18 in.
Use s = 16 in.
The selection of the longitudinal steel should begin at the typical interior support. In
this case, #4 bars at 16-in. spacing ( As = 0.15 sq in.) are chosen for all interior supports
except the first. At the first interior support, #4 bars at 14 in. spacing are chosen because
of a slight increase in the design moment. At the middle of the first and typical interior
spans, as well as the exterior support, minimum reinforcement requirements control. At
these locations, the required area is furnished by #4 bars at 16 in. spacing.
Also, it can be easily verified that the tensile strain in the reinforcement far exceeds
the value of 0.005 in all sections. Thus, all sections are tension controlled and the corre-
sponding φ factor is 0.90 as initially assumed.
(b) Development length requirements. To confirm the choice of longitudinal reinforce-
ment made on the basis of the required areas only, it is necessary to review the devel-
opment length requirements, as covered earlier in this section. The requirements of
ACI-7.7.3.8.3 must be checked at the exterior support (point 1 of Fig. 8.4.2) and at
inflection points 2 to 6. Also, positive reinforcement must extend at least 6 in. into
interior supports (ACI-7.7.3.8.2). In addition, embedment equal to the required devel-
opment length must be provided in both directions from the maximum moment points
at the faces of supports (points 7–12). Finally, embedment of the bars at the top of the
slab beyond extreme points of inflection (points 13–18) must satisfy ACI-7.7.3.8.4.
(Continued)
261
TABLE 8.4.1 REINFORCEMENT FOR ONE-WAY SLAB OF EXAMPLE 8.4.1, USING ACI MOMENT COEFFICIENTS
2S1 2S2 2S3
Line Number Support Middle Support Support Middle Support Support Middle Support
1. ACI moment coefficient 1 1 1 1 1 1 1 1 1

– + – – + – – + –
24 14 10 11 16 11 11 16 11
2. Mu = line(1) × 0.281(11.83)2 –1.64 +2.81 –3.94 –3.58 +2.46 –3.58 –3.58 +2.46 –3.58
line(2) × 12,000 60 103 145 132 91 132 132 91 132

3. Required Rn (psi) =
0.9(12)(5.5) 2
8 . 4 C hoice of R einforcement
line(3) × 0.0025 0.0010 0.0018 0.0025 0.0023 0.0016 0.0023 0.0023 0.0016 0.0023
4. Requireda ρ ≈
145
5. Required As = line (4) × 12(5.5) 0.14b 0.14b 0.17 0.15 0.14b 0.15 0.15 0.14b 0.15
6. Provided As #4@16 #4@16 #4@14 #4@16 #4@16 #4@16 #4@16
aAssuming a linear relationship between ρ and Rn (see Example 8.4.1 for ρ = 0.0025 and Rn = 145 psi).
bMinimum per ACI Table 7.6.1.1 controls [As,min = 0.0018(12)(6.5) = 0.14 sq in.].
(Continued)
261
26
8” 2’– 0” 3’ – 7” 3’ – 7” 3’ – 7” 3’ – 7” 3’ – 7”
#4 @ 16 #4 @ 14 #4 @ 16 #4 @ 16
7 1 13 14 2 8 9 3 15 16 4 10 11 5 17 18 6 12
#4 @ 16 #4 @ 16 #4 @ 16
Std. hook #4 @ 16 (S&T)
2S1 2S2 2S3
Figure 8.4.2 Reinforcement in slab of Example 8.4.1.
Positive moment reinforcement. From the ACI shear and moment diagrams in
Figs. 7.5.2 and 7.5.4, the shears at inflection points on the typical 1-ft width of slab are

V1 = V2 = 0.3780 wu Ln = 0.3780 ( 0.281)(11.83) = 1.26 kips

V3 = V4 = V5 = V6 = 0.3535wu Ln = 0.3535 ( 0.281)(11.83) = 1.18 kips
In computing the development length Ld for the bottom #4 bars that extend past the
inflection points into the supports, the cb / db value used in Eq. (6.6.1) is based on cover,
which is clearly smaller than one-half the center-to-center bar spacing. Thus,
 cb 0.75 (i.e.,clear cover) + 0.25 (i.e., bar radius / 2) 1.00 

 = = = 2.00  < 2.5 max
 db 0.50 0.50 
Because there are no stirrups, K tr = 0. Thus, Eq. (6.6.1) gives
 

3 f ψ t ψ e ψ s 
Ld = 
y
d [6.61]
 40 λ fc′  cb + ktr   b
  d  
 b 
 3 60, 000 1.0(1.0)(0.8) 

Ld =   0.5 = 14.2 in.
 40 1.0 4000 2.0
In the above equation, the reinforcement size factor ψ s = 0.8 for bars #6 and smaller.
Note again that the Ld computed above using the general equation has to be smaller
than the Category A simplified equation [Eq. (6.6.4)] value of 19.0 in. from Table 6.6.1,
where ( cb + K tr ) / db is taken as 1.5.
The requirement of ACI-7.7.3.8.3 at inflection points (such as point 2) is
Mn
+ La ≥ Ld
Vu
For #4 @ 16 in.,
C = 0.85(4)12 a = 40. 8 a

T = 0.20(60) = 12.0 kips
12.0
a= = 0.29 in.
40.8
 0.29  1
M n = 12.0  5.5 −  = 5.35 ft-kips / ft
 2  12

Vu = 1.26 kips (computed above)
(Continued)
263
8 . 4 C hoice of R einforcement 263

La = 12 db (max) = 6.0 in.
5.35(12)
+ 6.0 = 57.0 in. > Ld = 14.2 in. OK
1.26
This calculation applies identically to other inflection points (3–6).
At the exterior end, which might be considered a simple support, the requirement of
ACI-7.7.3.8.3
Mn
1.30 + La ≥ Ld
Vu
is satisfied by inspection. This reinforcement is extended 8 in. into the exterior support.
Because the bottom reinforcement is continuous over interior supports (Fig. 8.4.2), there
is no need to check the minimum embedment length of 6.0 in.
Negative moment reinforcement. The distance from the face of support to the cutoff
location (beyond point 14, for instance) must be (1) at least Ld , and (2) adequate to sat-
isfy ACI-7.7.3.8.4.
For the #4 bars, Ld = 14.2 in. (computed above)
To satisfy ACI-7.7.3.8.4, the moment diagram of Fig. 7.5.2(b) corresponding to the
ACI coefficients may be used to locate the point of inflection. The length required from
the exterior face of the first interior support to the cutoff location is
11.83
(0.5583 − 0.3341)(11.83) + = 3.39 ft
16
which compares favorably with the commonly used value of 0.3 of the clear span (3.55 ft),
as suggested in the 2004 ACI Detailing Manual [2.23]. For all other interior supports
[Fig. 7.5.4(b)],
11.83
(0.5 − 0.2611)(11.83) + = 3.57 ft
16
Note that the second term in the two expressions above corresponds to Ln /16, which is
greater than d and 12db. Both requirements above can be satisfied by extending the top
slab bars 3 ft 7 in. from the face of each interior support. Note that while slightly differ-
ent lengths could have been used, such a small difference in length at the first interior
support versus the other interior supports would be impractical.
For the top reinforcement at the exterior support, the minimum extension from the
face of the support is (Fig. 7.5.2b)
11.83
(0.4417 − 0.3341)(11.83) + = 2.0 ft > Ld
16
This reinforcement is anchored into the exterior support through a standard 180° hook.
(c) Shrinkage and temperature reinforcement
Minimum amount of shrinkage and temperature reinforcement is (ACI Table 24.4.3.2)

0.0018bh = 0.0018(12)(6.5) = 0.14 sq in.
at a spacing not greater than 5 times the slab thickness (32.5 in.) and 18 in. Thus, provide
#4 bars at 16 in. spacing perpendicular to the slab flexural reinforcement. Because the
slab is only 6.5 in. thick, there is no need to distribute this reinforcement near the top
and bottom faces of the slab. In this case, the shrinkage and temperature reinforcement
is provided at the bottom of the slab, directly on top of the bottom flexural reinforcement
(labeled as S&T in Fig. 8.4.2).
264
8.5 BAR DETAILS
Consistent with the shear and moment diagrams given in Chapter 7 for the ACI Code coef-
ficients, acceptable standard bar bend distances, extensions, and anchorage lengths have
been developed. For instance, typical bar details in end and interior spans of one-way slabs
as adapted from the 2008 design handbook of the Concrete Reinforcing Steel Institute
[2.21] are reproduced in Fig. 8.5.1.
Although typical bar details are of importance in office practice, they must be used with
discretion and care to ensure that they conform to the prevailing shear and moment diagrams.
ACI standard hook

(tilt from vertical if necessary
to maintain ¾” clearance)
Size and spacing CL Symmetrical
as tabulated about CL
0.25Ln
Temperature – shrinkage
¾” Clear bars
¾” Clear X X X
1½” cover
minimum Extend all bottom X = Bar spacing
Bars into support
Slab thickness
Ln = Clearspan
Single span, simply supported

ACI standard hook
(tilt from vertical if necessary
to maintain ¾” clearance)
Size and spacing Slab thickness
As tabulated
¾” Clear
0.25Ln 0.3Ln or 0.3Ln1 0.3Ln or 0.3Ln1
¾” Clear Greater Greater Bars
0.125Ln
1” cover
0” ¾” Clear X X X
minimum Extend all bottom
Bars into support 6”* X = Bar
Ln = Clearspan Ln1 spacing
End span, simply supported

CL
¾” Clear Symmetrical
about CL
0.3Ln or 0.3Ln1 0.3Ln or 0.3Ln1
Greater Greater Bars
0”
0.125Ln X X X
¾” Clear
X = Bar spacing
Ln1 6”* Ln = Clearspan
Slab thickness
Interior span, continuous

*min. 6”, unless otherwise specified by the engineer
Figure 8.5.1 Typical details for one-way solid slabs. (Adapted from CRSI Design Handbook 2008 [2.21].)
265
P roblems 265
SELECTED REFERENCES
8.1. Phil M. Ferguson. “Analysis of Three-Dimensional Beam-and-Girder Framing,” Proceedings,
Journal of the American Concrete Institute, 47, September 1950, 61–72.
8.2. R. Ian Gilbert. “Shrinkage Cracking in Fully Restrained Concrete Members.” ACI Structural
Journal, 89, March–April, 1992, 141–149.
PROBLEMS
All problems are to be worked in accordance with the 2014 ACI Code. All stated loads are
service loads unless otherwise indicated.
8.1 Design for a warehouse a continuous one-way (For SI: live load = 12.0 kN/m; fc′ = 30 MPa;
slab supported on beams 12 ft on centers as shown f y = 400 MPa.)
in the figure for Problem 8.1. Assume that beam 8.3 Repeat Problem 8.1 using a live load of 300 psf
stems are 12 in. wide. The dead load is 25 psf in with fc′ = 4000 psi and f y = 60, 000 psi.
addition to the slab weight, and the live load is 8.4 Repeat Problem 8.1 using a live load of 350 psf
200 psf. Use fc′ = 3000 psi and f y = 60, 000 psi. with fc′ = 4000 psi and f y = 60, 000 psi.
Use ACI coefficients if permissible. (For an SI 8.5 Design a one-way slab for the conditions shown
problem, use 3.7-m spans, 400-mm beam widths; in the figure for Problem 8.5. Assume that beam
superimposed dead load = 1.2 kN/m; live load = stems are 12 in. wide. The live load is 175 psf
9.6 kN/m; fc′ = 25 MPa; f y = 400 MPa.) and the slab will not be the final wearing sur-
8.2 Repeat Problem 8.1 using a live load of 250 face. Use fc′ = 4000 psi and fy = 60,000 psi.
psf with fc′ = 4000 psi and f y = 60, 000 psi.
Spandrel-type end
12’– 0” 12’– 0” 12’– 0”
Problems 8.1 through 8.4
Cantilever end
8’– 0” 16’– 0” 16’– 0”
Problem 8.5
CHAPTER 9
DESIGN OF SLAB-B EAM-G IRDER
AND JOIST FLOOR SYSTEMS
9.1 INTRODUCTION
The design of rectangular and T-sections was treated in Chapters 3 and 4, respectively;
shear strength along with stirrup design in Chapter 5; development of reinforcement in
Chapter 6; and continuity analysis in Chapter 7. The slab-beam-girder type of floor con-
struction has been described generally; Chapter 8 illustrates the design of one-way slabs.
This chapter presents complete designs of a typical floor beam and girder in a monolithic
slab-beam-girder system. Primarily, what has been developed in the preceding chapters is
applied. Thus the reader may consider the material of this chapter as an integrated review
of the subjects in the aforementioned chapters.
Also included in this chapter is the design of one-way concrete joist floors. On continu
ous spans, the concrete joists should be regarded as having rectangular sections near the
supports and T-sections in the positive moment region within the span.
In some instances, precast slabs, either conventionally reinforced or prestressed, are
placed on a monolithic beam-and-girder framing system. T-sections are not involved in
such systems because the slab rests on, but does not act with, the beams and girders. The
procedures discussed in this chapter are also generally applicable to this simpler system of
rectangular beams.
Noho III office building, North Hollywood, CA (Photo courtesy of Cary Kopczynski & Co).
267
9.2 SIZE OF BEAM WEB 267
Reinforced concrete building design may be a complicated venture, involving irregu-

lar floor plans and intricate structural framing. The simple floor framing plan used in the
examples does not necessarily reflect a typical design practice, but it does serve to illustrate
the basic essentials of design.
9.2 SIZE OF BEAM WEB
The size of the beam web for continuous T-shaped sections is usually controlled by the flex-
ural and shear strength requirements at the exterior face of the first interior support. For typ-
ical conditions of equal spans and uniform dead and live loads, a negative moment of 110 wL2n,
and a shear of 0.60wLn (see Fig. 7.5.1) or 1.15wLn /2 (see ACI-6.5.4) may be used in esti-
mating the size of the beam. The following discussion sets out the detailed considerations
involved in selecting the beam cross section based on the critical bending moment and shear.
Negative Moment Requirement

The section of the beam resisting the negative bending moment at the face of the sup-
port behaves as a rectangular section, even in T-section construction, because compression
occurs in the lower part of the section (i.e., in the lower portion of the web). The tension
reinforcement is most often provided by straight bars extending across the top of the beam,
as required for negative moment (bent bars, though still permitted by the ACI Code, are
rarely if ever used today). In accordance with ACI-9.7.3.8.2, at least one-fourth of the max-
imum positive moment reinforcement must continue along the bottom of the beam into the
support. When this reinforcement is continuous or developed into the support, it may be
utilized as compression reinforcement for increased ductility and deflection control.
The guideline reinforcement ratio ρ to be used for deflection control of singly reinforced
rectangular beams has been suggested (in Section 3.9) to be about 0.6ρtc for Grade 60 steel.
For the negative moment requirement on continuous T-sections, a higher value for ρ should
be acceptable without contributing to higher deflection, because the gross moment of iner-
tia of a typically proportioned T-section is roughly twice that of its rectangular portion.
Thus it may be reasonable to design the negative moment region of a T-shaped section with
about twice the percentage of reinforcement that would be used for deflection control on a
completely rectangular beam (or more if compression steel is utilized).
Positive Moment Requirement

In the positive moment region, the flange of the T-section is in compression. In practice,
the effective flange width is often large and thus the depth a of the Whitney rectangular
stress distribution will rarely extend below the bottom of the flange. On the tension side, the
amount of steel required will be inversely proportional to the depth of the section.
For sizing the beam web, the only consideration utilizing the positive moment is to
establish the width required to maintain adequate clearances for a given number of bars.
This should be examined because a somewhat deeper or shallower beam might still permit
the required steel to fit into one or two layers, as the case may be.
Shear Strength Requirement

The designer may wish to establish the beam web size to limit the maximum nominal shear
stress. This may be desirable for economical stirrup size and spacing. The ranges of five
categories of design for shear reinforcement are given in Section 5.10. The upper limit of
Category 4 (i.e., Vs = 4 fc′bw d ) serves as a practical guideline for ordinary design, typi-
cally permitting stirrup spacing from 3 in. to a maximum of d/2 or 24 in.
268
268 C hapter 9 S lab - B eam - G irder A N D J O I S T F L O O R S Y S T E M S
When Vc is taken as 2 λ fc′ bw d according to the simplified procedure, the total Vn at the
upper limit in Category 4 is
Vn = Vc + Vs
= 2 λ fc′ bw d + 4 fc′ bw d

= (2 λ + 4) fc′ bw d (9.2.1)
or, for a given factored shear Vu, the bwd required to satisfy this guideline is
required Vn Vu
required bw d = = (9.2.2)
(2 λ + 4) fc′ φ (2 λ + 4) fc′
Except for unusual conditions involving short spans, heavy concentrated loads, heavily
doubly reinforced sections, or combinations of these, the maximum nominal shear stress
Vu /(φ bw d ) should be less than 6 fc′ for reasonable stirrup size and spacing.
EXAMPLE 9.2.1
Use the ACI Code provisions to establish the preliminary size for the continuous
floor beams 2B1-2B2-2B1 supported by girders as shown in the floor framing plan of
Fig. 8.3.1. Assume a slab thickness of 5.5 in. Use fc′ = 4000 psi and f y = 60, 000 psi.
SOLUTION
(a) Negative moment requirement. Beam sections located at the face of the support are
designed as rectangular sections with the compression zone located at the lower part
of the section. These sections may be sized for the maximum reinforcement ratio
corresponding to tension-controlled sections because deflections will be governed
primarily by the stiffness (moment of inertia) of the T-sections (with the flange in
compression) within the span. For fc′ = 4000 psi and f y = 60, 000 psi, the maximum
reinforcement ratio recommended in Section 3.6 may be computed from Eq. (3.6.2)
or from Table 3.6.1,

recommended ρmax = ρtc = 0.0181
Estimating the weight of the stem (portion of the web below the slab) at 0.3 kips/ft
(2 sq ft of area), and using the load factors of ACI-5.3.1,
Factored dead load = 1.2 5.5 ( 0.15 /12 )(13) + 0.3 = 1.43 kips/ft

Factored live load = 1.6 ( 0.10 )(13) = 2.08 kips/ft
Using Eqs. (3.8.4a) and (3.84b), the Rn corresponding to ρ = 0.0181 is
fy
m= = 17.65
0.85 fc′
1
Rn = ρ f y (1 − ρ m) = 913 psi
2
Assume width of supporting girders to be 18 in., then the clear span Ln is

Ln = 26 – 1.5 = 24.5 ft
(Continued)
269
9.2 SIZE OF BEAM WEB 269
Using the moment coefficient from ACI-6.5.2 (see Table 7.4.1) and φ = 0.90 (tension-
controlled section),
1
max Mu = (1.43 + 2.08)(24.5)2 = 211 ft-kips
10

M 211(12, 000)
required bd 2 = u = = 3081 cu in.
φ Ru 0.90(913)
If b = 13 in.,
3081
13
The steel area required would then be
As = 0.0181(13)(15.4 ) = 3.62 sq in.
In the negative moment region the flange at the top (slab) is available so that the steel
does not have to fit within the web width.
The shear requirement and the steel requirement for positive moment should be
examined before a decision is made.
(b) Shear requirement. Using the approximate shear coefficient from ACI-6.5.4,
w L   3.51(24.5) 
max Vu = 1.15  u n  = 1.15   = 49.4 kips
 2   2
If it is desired that the nominal shear stress υn = Vu /(φ bw d ) does not exceed
6 fc′ = 379 psi,
49, 400
required bw d = = 174 sq in.
0.75(379)
For b = 13 in., then d = 13.4 in. (Actually the maximum shear may be taken at a distance
d from the face of the support which would result in a smaller required beam depth.)
(c) Positive moment requirement. The effective flange width bE is the smallest of the
following (see Section 4.3):
24.5(12)
1. bw + Ln / 4 = bw + = bw + 73.5 in. Controls!
4
2. bw + 16t = bw + 16(5.5) = bw + 88 in.
3. center-to-center spacing = 13(12) = 156 in.
For bw = b = 13 in., the effective width will be 86.5 in., or about 6.7 times wider than the
web width, bw. Try d = 16 in. and estimate the moment arm at 0.95d ≈ 15 in. (This is a
reasonable assumption for T-beams with the flange in compression.) Using the moment
coefficients from ACI-6.5.2 for an end span with discontinuous end integral with the
support and assuming φ = 0.90 (tension-controlled section),
1 3.51
Mu = wL2n = (24.5)2 = 150 ft-kips
14 14

Mu 150(12)
required As = ≈ = 2.22 sq in.
φ f y (arm ) 0.90(60)(≈ 15)
(Continued)
270
This amount of steel is satisfied with 3 #8 bars (As = 2.37 sq in.), for example, which
easily fit into a 13-in.-wide beam in one layer (see Table 3.9.2).
(d) Minimum depth. According to ACI Table 9.3.1.1, the minimum depth, h, of a beam
with one end continuous (beams 2B1 are more restrictive than the center beam, 2B2)
cannot be less than
L 26(12)
min h = = = 16.9 in.
18.5 18.5
unless the deflection limits given in ACI-9.3.2 are satisfied. The designer must also keep
in mind that if excessive deflection may cause damage, the deflection must be computed
in accordance with ACI-24.2 and must satisfy ACI Table 24.2.2, even if the limits of
ACI Table 9.3.1.1 are satisfied. The above check is only a minimum requirement, where
deflection is not likely to cause damage to nonstructural elements.
(e) Choice of size. Based on the calculations in parts (a) through (c), a beam with a
13-in.-wide web and an effective depth, d, of about 16 in. would be adequate.
Allowing 2.5 in. for the clear cover and stirrups, this would result in a minimum over-
all depth, h, of 18.5 in., or 19 in. Frequently the size is chosen larger than required.
Sometimes this occurs because the designer wants a large number of beams to have
the same external dimensions for economy of forming, or perhaps because ductwork
or pipes are to pass through the beams, necessitating larger sizes.
Use bw = 13 in., h = 22.5 in. (which gives d ≈ 20 in.). It is common to make the stem por-
tion below the flange a whole-inch increment, such as 17 in. for this case. The arbitrary
size selected is larger than necessary; any size at least equal to that indicated by steps (a)
through (d) would serve to illustrate the design procedure.
9.3 CONTINUOUS FRAME ANALYSIS FOR BEAMS

The shear and bending moment diagrams to be used in the design of the floor beams in the
preceding example may be obtained by using the ACI moment coefficients (ACI-6.5.2).
However, because those coefficients are more suitable for use in frames involving, say,
more than four continuous spans, an elastic analysis will be used to illustrate how to obtain
the shear and bending moment envelopes for beams 2B1-2B2-2B1.
Some designers would assume the girders to provide only vertical support to floor beams
resting on them; thus a pure continuous beam analysis would be made. The girders, however,
have torsional stiffness that acts to restrain the end rotation of the beams over these support
girders. The ACI Code permits torsional effects to be ignored whenever the factored torsional
moment, Tu, is less than the design threshold torsion, φ Tth (ACI-9.5.4.1; see Chapter 18).
In the case of a beam supported with relatively large and stiff girders [see Fig. 9.3.1(a)],
the torsional stiffness provided by the girder may reasonably be taken as twice the flexural
stiffness of the beam. An equivalent beam and column frame model for this beam-girder
system is shown in Fig. 9.3.1(b). For beams that frame into columns, however, the combina-
tion of the bending stiffness of the columns plus the torsional stiffness of the girders results
in a relative end restraint greater than in the aforementioned case. To design such beams, the
sizes of the members may be estimated and the (∑Kcol /∑Kbeam) ratio computed therefrom.
With regard to a T-section, opinions as to how its stiffness in a continuous beam should
be computed differ considerably. A T-section certainly provides more stiffness in the posi-
tive moment region where the flange is in compression than it does as a rectangular section
in the negative moment region. For gravity load analyses, it has been common practice to
use the gross moment of inertia, neglecting reinforcement, in computing the flexural stiff-
ness of such elements. For T-sections in the positive moment region, the gross section usu-
ally includes the effective flange width.
271
9 . 3 C O N T I N U O U S F R A M E A N A LY S I S F O R B E A M S 271
A B C D
2B1 2B2 2B2

26’–0” 26’–0” 26’–0”
(a) Actual continuous beam supported by girders
1 1 1
Relative
2 2 2 2
stiffness
(b) Frame model to approximate floor beams supported on girders
Figure 9.3.1 Actual continuous beam and equivalent beam and column model.
One of the accepted methods [9.1] has been to use a T-section moment of inertia equal
 
to 2  1 bw h3  , which is equivalent to using an effective flange width of about 6 times the
 12 
web width. Since the true stiffness is that of a span with variable moment of inertia along its
length with the T-section stiffness over, say, the middle half of the span and the rectangular
section stiffness over the end quarters, the equivalent system is obtained approximately by
using the T-section with a flange width only twice the web width over the entire span. The
ACI Code provides some guidance for calculating section properties to be used in the anal-
yses under factored lateral loads (ACI-6.6.3), but it is silent for the analysis under gravity
loads. In ACI-R6.6.3.1.1, however, it is stated that the moment of inertia of T-beams should
be based on the effective flange width defined in ACI-6.3.2.1 or ACI-6.3.2.2 (see Section
 1 
4.3), or as 2  bw h3  as discussed above. Examination of analyses using various stiffness
 12 
ratios will show that a fairly wide variation in stiffness ratio may be accommodated with
relatively small changes in bending moments.
In the following example, only the analysis of the floor beams supported on girders is
shown, using the analytical frame model in Fig. 9.3.1(b). An overestimation of the beam
stiffness will give a conservative design for the beam; but such an assumption should be
reevaluated if any columns are to be designed.
EXAMPLE 9.3.1
Using an elastic statically indeterminate analysis, obtain the shear and moment diagrams
to be used in compiling the moment and shear envelopes for the design of beams 2B1
and 2B2 in Example 9.2.1. Because the girders are not expected to be significantly larger
and stiffer than the beams, assume the torsional stiffness of the girders to be the same as
the flexural stiffness of the beams.
SOLUTION
Using the beam dimensions chosen in part (e) of Example 9.2.1, the actual weight of
the portion of the web below the slab is 0.15(13/12)(17/12) = 0.23 k/ft, say 0.25 k/ft. The
revised factored load then becomes
Factored dead load = 1.2[5.5(0.15)(13/12) + 0.25] = 1.37 kips/ft
Loading conditions 1 through 5 as shown in Fig. 9.3.2 are established by visualizing the
influence lines for positive moment in the midspan region of each span and for negative
moment at each support (see Section 7.3). The structural analysis may be performed by
any method of statically indeterminate analysis—for example, the moment distribution
method—or by using any standard structural analysis software package. Since the specific
(Continued)
27
WD
1.
A B C D
WL
2.
WL
3.
WL
4.
WL
5.
Figure 9.3.2 Loading conditions for Example 9.3.1.
structural analysis technique is outside the scope of this text, only the end moments
resulting from the analysis for each of the five loading cases are given in Table 9.3.1.
The primary shear and moment diagrams for maximum positive and negative
moments for beams 2B1 and 2B2 are given individually in Figs. 9.3.3 and 9.3.4, respec-
tively. The secondary shear and moment diagrams are shown in Figs. 9.3.5 and 9.3.6,
respectively. The primary and secondary moment diagrams are drawn separately for
clarity in the example. Ordinarily they are not drawn separately but, instead, are super-
imposed to become the moment envelope as shown later (see Fig. 9.6.2).
a
TABLE 9.3.1 END MOMENTS (ft-kips) FOR THE FIVE LOADING
CONDITIONS OF FIG. 9.3.2
Joint A B C D
Member AB BA BC CB CD DC
Case 1, M = –27.1 +94.8 –82.1 +82.1 –94.8 +27.1 Dead load only
Case 2, M = –54.2 +95.2 –35.6 +35.6 –95.2 +54.2 Live load spans 1
and 3
Case 3, M = +13.1 +48.8 –89.1 +89.1 –48.8 –13.1 Live load span 2
Case 4, M = –39.0 +153.2 –143.9 +69.8 –39.6 –11.1 Live load spans 1
and 2
Case 5, M = –2.0 –9.2 +19.2 +54.9 –104.4 +52.2 Live load span 3
aClockwise moments acting on member ends are taken positive.

(Continued)
273
9 . 3 C O N T I N U O U S F R A M E A N A LY S I S F O R B E A M S 273
81.3 190.0 66.1 248.0

3.45 k/ft 3.45 k/ft
A Ln = 24.5’ B A Ln = 24.5’ B
40.7
Vu , kips 37.9
14.20’ 15.00’
11.80’ 11.00’
158.8 49.0 142.1 51.8
2.20’ 4.60’ 1.91’

+ 2 – #7
5.94’
+
2 – #5
Mu , ft-kips
1 2
– –
9.60’ 9.60’ 66.1 9.08’ 9.08’ 2 – #8
81.3
1.45’
190.0 2 – #7
x1 x2
248.0
x3
(a) Maximum in positive region (b) Maximum in negative region at B
(loading conditions 1 + 2) (loading conditions 1 + 4)
Figure 9.3.3 Primary shear and moment diagrams for 2B1.
171.2 3.45 k/ft 171.2 226.0 3.45 k/ft 151.9
B Ln = 24.5’ C B Ln = 24.5’ C
44.9 47.7
Vu , kips
13’ 12.17’
13’ 13.83’
44.9 42.0
2 – #7 121.0
103.8
4.61’ + 4.61’
2 – #4 6.07’ + 4.42’
3 4
Mu , ft-kips
– –
– 8.39’ 8.39’ – 2 – #8 7.76’ 7.76’
171.2 2 – #7 151.9
x4 x4 171.2
226.0
x5
(a) Maximum in positive region (b) Maximum in negative region at B

Figure 9.3.4 Primary shear and moment diagrams for 2B2.
(Continued)
274
14.0 1.37 k/ft 143.6 29.1 1.37 k/ft 86.6
A B A 26’ B
26’
12.8 15.6
16.7’ 14.6’
Vu, kips
9.3’ 11.4’
22.8 20.0
1.17’
45.8 2.05’ 59.7
Mu, ft-kips
7.77’
+ +
–
– 29.1
14.0 8.18’ 8.18’ 9.32’ 9.32’
86.6
143.6
(a) Minimum in positive region (b) Minimum in negative region at B

Figure 9.3.5 Secondary shear and moment diagrams for 2B1.
117.7 1.37 k/ft 117.7 62.9 1.37 k/ft 137.0
B C B C
26’ 26’
17.8 15.0
13.0’ 15.05’
Vu , kips
13.0’ 10.95’
17.8 20.7
5.65’
19.2
Mu , ft-kips
– – 5.3’ 5.3’
–
2.07
62.9
117.7 117.7
137.0
(a) Minimum in positive region (b) Minimum in negative region at B
Figure 9.3.6 Secondary shear and moment diagrams for 2B2.
9.4 CHOICE OF LONGITUDINAL REINFORCEMENT

IN BEAMS
It cannot be overemphasized that the choice of longitudinal reinforcement depends on both
the steel area and development length (or anchorage) requirements. The design of the main
reinforcement in floor beams 2B1-2B2-2B1 is shown in the following example.
275
9.4 CHOICE OF LONGITUDINAL REINFORCEMENT IN BEAMS 275
EXAMPLE 9.4.1
Choose the arrangement of main reinforcement in the floor beams 2B1-2B2-2B1 of

Example 9.3.1.
SOLUTION
(a) Flexural strength requirements. The critical moment at the face of the support, or
at any point along the beam length, can be obtained from the diagrams shown in
Figs. 9.3.3 through 9.3.6. Alternatively, a good estimate for preliminary design may
1
be obtained by subtracting ∆M = [V × (width of support )] from the moment at the
3
center of support. The shear V for this calculation is taken as that at the face of sup-
port. Start with A-A, C-C, and D-D (Fig. 9.4.1),
Section A-A (rectangular section):
38.1(1.5)
Mu at face = 81.3 − = 81.3 − 19.1 = 62.2 ft-kips
3
Assuming φ = 0.90,
Mu 62.2(12, 000)
φ bd 2
0.90(13)(20)2
Using Eq. (3.8.5), or Fig. 3.8.1,
1 2 mRn 
required ρ = 1 − 1 − 
m fy 
fy 60
m= = = 17.65
0.85 fc′ 0.85(4)
1  2(17.65)(159) 
required ρ =  1− 1− = 0.0027
17.65  60, 000 
The minimum reinforcement for a statically indeterminate T-beam section with the
flange in tension is given by
3 fc′
As ,min = bw d [3.7.10]*
fy
but not less than 200bw d /f y . Thus,
3 4000 200(13)20
As,min = (13)20 = 0.82 or = 0.87; 0.87 sq in. Controls!
60, 000 60, 000
The required area based on the applied loads is
required As = 0.0027 (13) 20 = 0.70 sq in. <  As ,min = 0.87 sq in.

Therefore, use As = As,min = 0.87 sq in.

(Continued)
* Equation numbers in brackets indicate material from another section or chapter.

276
A B F C D E
10 11
4 8 2 5 6 3 9
1 7
A B F C D E
24’–6” 12’–3”
2B1 2B2
2 – #8 (A = 2.78) Symmetrical
2 – #7 (As = 1.20) 2 – #7
s about CL
1.04 2.73 2.49

1.78 1.36
2 – #7 (A = 1.82) 2 – #7 (A = 1.60)
s s
2 – #5 2 – #4
Figure 9.4.1 Longitudinal reinforcement areas required (sq in.) and bars selected for beams 2B1
and 2B2 in Example 9.4.1.
Section C-C (rectangular section):
49.2(1.5)
Mu at face = 248.0 − = 248.0 − 24.6 = 223 ft-kips
3

223(12, 000)
0.90(13)(20)2
From Fig. 3.8.1, or Eq. (3.8.5), required ρ = 0.0105,

required As = 0.0105(13)20 = 2.73 sq in. >  As ,min = 0.87 sq in.
Section D-D (rectangular section):

45.1(1.5)
Mu at face = 226.0 − = 226.0 − 22.6 = 203 ft-kips
3

203(12, 000)
0.90(13)(20)2
Computing the required steel area As in proportion to Rn,
required As = 2.73(521 / 572) = 2.49 sq in. >  As ,min = 0.87 sq in.

For sections B-B and E-E (Fig. 9.4.1): The effective flange width is the smaller of:
24.5(12)
1. bw + Ln / 4 = bw + = 13 + 73.5 = 86.5 in Controls!
4
2. bw + 16t = bw + 16(5.5) = 13 + 88 = 101 in.
Thus, bE = 86.5 in.
For T-sections having the flange in compression, the minimum reinforcement is
given by
3 fc′
As ,min = bw d [3.7.10]
fy
(Continued)
27
but not less than 200bw d /fy. In Eq. (3.7.10), bw is the width of the web. Thus,

As,min = 0.87 sq in.
as computed earlier.
Section B-B (T-section with flange in compression):
Estimate moment arm = 0.9d = 18 in.
Mu 158.8(12)
φ f y (arm) 0.90(60)(≈ 18)
Check:
C = 0.85 fc′ bE a = 0.85(4)(86.5)a = 294 a
T = As f y = 1.96(60) = 118 kips
118
a= = 0.40 in.
294
0.40
arm = 20 − = 19.8 in.
2
158.8(12)
revised required As = = 1.78 sq in. > [As ,min = 0.87 sq in.] OK
0.90(60)19.8
Section E-E (T-section with flange in compression):
Mu 121.0(12)
required As = =
φ f y (arm) 0.90(60)(≈ 19.8)
= 1.36 sq in. > [As ,min = 0.87 sq in.] OK
Check:
C = 294 a
T = As f y = 1.36(60) = 82 kips
82
a= = 0.28 in.
294
0.28
arm = 20 − = 19.9 in. OK
2
On the basis of the areas required at sections A-A, B-B, C-C, D-D, and E-E, the arrange-
ment of main reinforcement shown in Figs. 9.4.1 and 9.4.2 is tentatively chosen. Note
that the 2–#5 bottom bars in beam 2B1 and the 2–#4 bottom bars in beam 2B2 furnish the
minimum one-fourth of the positive moment reinforcement that must be extended at least
6 in. into the support in each span (ACI-9.7.3.8.2). Also note that the two 2–#8 top bars
in section E-E are needed to satisfy negative moment strength requirements at midspan
as discussed later, in part (h) of Example 9.6.1. Confirming the choice of main reinforce-
ment calls for checking the design strength φ Mn at the critical sections. The effect of any
compression reinforcement will be small and is thus neglected.
Negative moment at section A-A: 2–#7 top bars (As = 1.20 sq in. > required
As = 0.87 sq in.)
C = 0.85(4)(13)a = 44.2 a T = 2(0.60)60 = 72.0 kips
72.0 1.63
a= = 1.63 in. c= = 1.92 in.
44.2 0.85
(Continued)
278
1
5 2” 2 – #8
2 – #7 2 – #8 2 – #8
2 – #3 2 – #7
Stirrup
17” supports
2 – #5 2 – #4 2 – #5
2 – #5
13” 2 – #5 (C–C) 2 – #7
2 – #7 2 – #7
2 – #4 (D–D)
Section A–A Section B–B Sections C–C and D–D Section E–E Section F–F
Figure 9.4.2 Typical sections for 2B1 and 2B2 in Example 9.4.1. Stirrups not shown for clarity.
The net tensile strain in the extreme tension steel is
d −c 20 − 1.92
εt = (0.003) = (0.003) = 0.028 > 0.005
c 1.92
Thus, φ = 0.90
1
φ M n = 0.90(72.0)(20 − 0.82) = 104 ft-kips [> 62.2 ft-kips] OK
12
Positive moment at sections B-B and F-F: 2–#5 and 2–#7 bottom bars (As = 1.82 sq in.
> required As = 1.78 sq in.)
C = 0.85(4)(86.5)a = 294a T = 1.82(60) = 109 kips
a = 0.37 in. c = 0.44 in.
εt = 0.133 > 0.005
1
φ M n = 0.90(109)(20 − 0.19 ) = 162 ft-kips [> 158.8 ft-kips] OK
12
Negative moment at sections C-C and D-D: 2–#8 and 2–#7 top bars [As = 2.78 sq in. >
required As = max. of (2.73, 2.49) sq in.]
C = 44.2a T = 2(0.79 + 0.6)60 = 167 kips
a = 3.77 in. c = 4.44 in.
εt = 0.0105 > 0.005
1
φ M n = 0.90(167)(20 − 1.89) = 227 ft-kips [> (223 and 203) ft-kkips] OK
12
Positive moment at section E-E: 2–#4 and 2–#7 bottom bars (As = 1.60 sq in. > required
As = 1.36 sq in.)
C = 294a T = 2(0.20 + 0.60)60 = 96 kips
a = 0.33 in. c = 0.38 in.
εt = 0.153 > 0.005
1
φ M n = 0.90(96)(20 − 0.17) = 143 ft-kips[> 121 ft-kips] OK
12
(Continued)
279
In addition to checking the design strength at the above critical sections, the design
strength of the cross sections at locations where some of the longitudinal reinforce-
ment is to be terminated must be computed. The design strength at these locations is
required for later computation of the appropriate bar cutoff locations to ensure that the
φ Mn diagram does not encroach the moment envelope, as will be illustrated later in
Example 9.6.1.
Positive moment near exterior support, after termination of 2–#7 bottom bars:
2–#5 (As = 0.62 sq in.)
It is noted that at this location the provided amount of reinforcement is less than As, min
(= 0.87 sq in.). According to ACI-9.6.1.3, As, min need not be satisfied when As provided is
at least one-third greater than that required by analysis. Assume that the 2–#7 bars will
be cut at a distance such that ACI-9.6.1.3 will be satisfied with 2–#5 bars. Thus,
C = 294a T = 0.62(60) = 37.2 kips
a = 0.13 in.
c = 0.15 in.
εt = 0.40 >>> 0.005
1
φ M n = 0.90( 37.2)(20 − 0. 065) = 56 ft-kips
12
Note that the computed net tensile strain is much greater than the elongation at fracture
for a typical Grade 60 reinforcing bar (see Fig. 1.13.2), which implies that the tension
steel will not only yield but will also strain harden and likely fracture before concrete
crushing. This behavior is due to the small amount of flexural reinforcement provided
at this location and the wide flange (effective width), which results in a very small neu-
tral axis depth (note that even when As, min is provided, the net tensile strains are also
computed to exceed the expected elongations at fracture). It will be shown later that the
design strength is much greater than the required strength and that ACI-9.6.1.3 is satis-
fied by a large margin and thus As, min need not be provided at this location.
Negative moment near sections C-C and D-D, after termination of 2–#7 top bars: 2–#8
(As = 1.58 sq in.)
T = 2 (0.79) 60 = 94.8 kips
C = 44.2a
a = 2.14 in. c = 2.52 in.
εt = 0.021 > 0.005
1
φ M n = 0.90(94.8)(20 − 1.07) = 135 ft-kips
12
Positive moment near section D-D, after termination of 2–#7 bottom bars: 2–#4 (As =
0.40 sq in.)
C = 294a T = 2 (0.20)60 = 24 kips
a = 0.082 in. c = 0.10 in.
εt = 0.60 >>> 0.005
1
φ M n = 0.90(24)(20 − 0.041) = 36 ft-kips
12
These moment capacities are compared later with a moment envelope (see Fig. 9.6.2),
which shows that flexural strength requirements are met with the chosen reinforcement
at all critical sections.
(Continued)
280
(b) Development lengths using the simplified equations, Eqs. (6.6.4) through (6.6.7);
Tables 6.6.1 and 6.6.2 can be used. The development lengths are
Bars Location Category A Category B
#4 3 (bottom) 19.0 in. 28.5 in.

#5 1 and 2 (bottom) 23.7 in. 35.6 in.
#7 4, 5, and 6 (top) 54.0 in. 81.0 in.
7, 8, and 9 (bottom) 41.5 in. 62.3 in.
#8 5, 6, 10, and 11 (top) 61.6 in. 92.6 in.
where a modification factor for the casting position, ψt = 1.3, is applied for the bars
located at the top of the beams. In this example, nearly all the bars meet Category A,
and thus, the values shown in the table could be used for determining bar cutoff loca-
tions. As noted in Chapter 6, however, the simplified equations often result in develop-
ment lengths greater than those required by the general equation, Eq. (6.6.1). Although
the use of Eq. (6.6.1) requires additional calculations and effort, it often yields shorter
development lengths, which could result in important cost savings due to the use of less
material, less congestion of reinforcement, and easier fabrication and erection.
(c) Development lengths using the general equation, Eq. (6.6.1)
i. Development lengths for the #5 bottom bars at locations 1 and 2 (Fig. 9.4.1). The
proposed layout of the reinforcement at these locations is shown as section A-A and
section C-C in Fig. 9.4.2. Here the 2–#5 extend into the support after the 2–#7 bars
are cut off. Assume that the #5 bars are fully developed between the supports into the
span before the cutoff location of the #7 bars. Thus, the 2–#5 are treated as if alone in a
13-in.-wide stem enclosed within at least minimum #3 U stirrups at these locations. The
general equation, Eq. (6.6.1), is reproduced below
 

3 f ψ t ψ e ψ s 
Ld = 
y
d [6.6.1]
 40 λ fc′  cb + K tr   b
  d  
 b 
but no less than 12 in.
The cover or spacing dimension cb is the smaller of (1) distance from center of bar
being developed to nearest concrete surface, and (2) half the center-to-center spacing of
bars being developed. Thus, the distance cb will be the smaller of the following two values:
top and side cover = 1.5 (i.e., clear) + 0.375 (i.e., stirrup)
+ 0.313 (i.e., bar radius) = 2.19 in. Controls!
1 1
c-c spacing = [13 − 2(1.5) − 2(0.375) − (0.625)] = 4.3 in.
2 2
If there were no stirrups (which in fact there are), Ktr = 0 and
(cb + Ktr) /db = (2.19 + 0)/0.625 = 3.5 > 2.5 max
Thus,
 3 60, 000 1.0(1.0)1.0 
Ld =   0.625 = 17.8 in. (1.5 ft)
 40 1.0 4000 2.5
where the modification factors ψt, ψe, and λ are all equal to 1.0. Although a bar size fac-
tor of ψs = 0.8 is permitted by the ACI Code, it was not applied here, in conformity to
the recommendations of ACI Committee 408 [6.53], as discussed in Section 6.7. This
is conservative.
(Continued)
281
ii. Development length for the #4 bottom bars at location 3 extending into the interior
support (Fig. 9.4.1). The proposed layout of the reinforcement is shown as section D-D
in Fig. 9.4.2. Assuming that the 2–#4 are fully developed before the #7 bars are cut off,
they are then treated as if alone in a 13-in.-wide stem enclosed within at least minimum
#3 U stirrups. The distance cb is the smaller of the following two values:
1 1
c-c spacing = [13 − 2(1.5) − 2(0.375) − (0.50)] = 4.38 in.
2 2
If there were no stirrups (Ktr = 0), then
(cb + Ktr)/db = (2.13 + 0)/0.5 = 4.3 > 2.5 max
The general equation using the maximum (cb + Ktr)/db of 2.5 gives
 3 60, 000 1.0 (1.0 )1.0 

Ld =   0.5 = 14.2 in. (1.2 ft )
 40 1.0 4000 2.5
where no bar size reduction factor was applied (i.e., ψs = 1.0).

iii. Development length for the #8 top bars extending from the interior support into
beams 2B1 and 2B2 (near locations 10 and 11 in Fig. 9.4.1). Although the #8 bars are
within the flange, they will be treated as if alone in a 13-in.-wide stem enclosed within
at least minimum #3 U stirrups (see sections E-E and F-F in Fig. 9.4.2). The distance cb
is the smaller of the following two values:
1 1
c-c spacing = [13 − 2(1.5) − 2(0.375) − (1.00)] = 4.13 in.
2 2
Compute
cb + K tr 2.38 + 0
= = 2.38 (without stirrup contribution) < 2.5 max
db 1.00
The general equation, taking ψt = 1.3 for the casting position (i.e., more than 12 in. of
fresh concrete cast below the bar), gives
 3 60, 000 1.3(1.0)1.0 

Ld =  1.00 = 38.9 in. (3.2 ft)
 40 1.0 4000 2.38 
iv. Development length for the 2–#7 top bars at location 4 (Fig. 9.4.1) extending into
the exterior support. The 2–#7 are treated as if alone enclosed within at least minimum
#3 U stirrups in a 13-in.-wide stem. The distance cb is the smaller of the following two
values (section A-A of Fig. 9.4.2):
1 1
c-c spacing = [13 − 2(1.5) − 2(0.375) − (0.875)] = 4.19 in.
2 2
(Continued)
28
Compute
cb + K tr 2.31 + 0
= = 2.64 (without stirrup contribution) > 2.5 max
db 0.875
The general equation taking ψt = 1.3 for the casting position (i.e., more than 12 in. of
fresh concrete cast below the bar) gives
 3 60, 000 1.3(1.0)1.0 

Ld =   0.875 = 32.4 in. (2.7 ft)
 40 1.0 4000 2.5
To develop properly the 2–#7 bars into the support a straight embedment of 32.4 in. is
required, which exceeds the width of the support (girder) of 18 in. Thus, use a standard
90° hook. For a #7 standard hook, the development length Ldh, according to Eq. (6.10.1)
(ACI-25.4.3.1), is
Ldh =   db [6.10.1]
 50 λ fc′ 
but not less than 8db = 8(0.875) = 7 in. nor less than 6 in.

Assume concrete cover of at least 2 1 2 in. normal to the plane of the hook, and assume
that cover on the bar extension beyond the 90° hook is at least 2 in., which would satisfy
Condition 1 of Table 6.10.1. Thus, ψc = 0.7 and Ldh becomes
 60, 000(1.0)(0.7)1.0 
Ldh =  0.875 = 11.6 in. > 6 in. OK
 50(1.0) 4000 
where ψr has been assumed to be 1.0. The available Ldh (see Fig. 6.10.3) is

available Ldh = 18(support width ) − 2 (cover on tail) = 16 in. > 11.6 in. OK
Thus, the 2–#7 bars are assumed to be fully effective at point 4 of Fig. 9.4.1 when 90°
hooks are used. Since a girder frames in at each side of beam 2B1, there will automati-
cally be at least 2 1 2 -in. side cover on the hooks.
v. Development length for the 2– #7 bottom bars near locations 7, 8, and 9
(Fig. 9.4.1). Assume that the 2–#7 are cut as soon as permitted, and that the 2–#5 (or
2–#4) are extended into the support. The proposed layout of the reinforcement for section
B-B of beam 2B1 is shown in Fig. 9.4.2. Assuming the 2–#5 (or 2–#4) are fully developed
before the 2–#7 are cut off, then the calculation of the clear spacing is based on the 2–#7
bars only. Since the 2–#7 bars are placed toward the center of the cross section with the
2–#5 (or 2–#4) bars in the corners, side cover will not govern. Thus, the distance cb is the
smaller of the bottom cover or half the center-to-center spacing between the 2–#7 bars
as follows:
bottom cover = 1.5 (i.e., clear) + 0.375 (i.e., stirrup)
+ 0.438 (i.e., bar radius) = 2.31 in.
The clear spacing between bars may be obtained from the geometry of the cross sec-
tion with the aid of Table 3.9.3. From this table, the minimum width for a beam with
2–#5 bars and 2db clear spacing is 7.1 in. For a beam with 2–#5 and 2–#7, the clear spac-
ing between bars may be computed as follows:
1
clear spacing =
3
{13 − [7.1 − 2(0.625)] − 2(0.875)} = 1.8 in. or 2.11db
(Continued)
283
Therefore, the center-to-center spacing of the 2–#7 bars is

1 1
c-c spacing = [1.8 + (0.875)] = 1.34 in. Governs!
2 2
Assuming no contribution from stirrups, Ktr = 0 and
(cb + Ktr)/db = (1.34 + 0)/0.875 = 1.53 < 2.5 max
Thus,
 3 60, 000 1.0(1.0)1.0 
Ld =  0.875 = 40.7 in. (3.4
4 ft)
 40 1.0 4000 1.53 
Alternatively, if minimum #3 U stirrups spaced at d/2 = 10 in. are assumed to restrain

the opening of the splitting plane, Ktr will be greater than zero. Because the potential
splitting plane is horizontal and two of the four bars are being developed,
40 Atr 40(2)(0.11)
K tr = = = 0.44
sn 10(2)
and
(cb + Ktr)/db = (1.34 + 0.44)/0.875 = 2.03 < 2.5 max
Thus, the development length could be reduced to
1.53
Ld = 40.7 = 30.7 in. (2.6 ft)
2.03
At location 9 (see section E-E in Fig. 9.4.2) where 2–#4 bars instead of 2–#5 bars are
used, it is reasonable and practical to assume that the center-to-center spacing of the #7
bars at the bottom of beam 2B2 is the same as that computed for locations 7 and 8 above.
Thus, use the same development length at all three locations.
vi. Development length for the 2–#7 top bars near locations 5 and 6 (Fig. 9.4.1). Assume
that the 2–#7 are cut as soon as permitted, and that the 2–#8 are extended into the spans
of beams 2B1 and 2B2. The proposed layout of the reinforcement for sections C-C of
beam 2B1 and D-D of beam 2B2 is shown in Fig. 9.4.2. Assuming that the 2–#8 are fully
developed before the 2–#7 are cut off, the calculation of the clear spacing is based on the
2–#7 bars only. Since the 2–#7 bars are placed toward the center of the cross section with
the 2–#8 bars in the corners, side cover will not govern. Thus, the distance cb is the smaller
of the top cover or one-half the center-to-center spacing between the 2–#7 bars as follows:
top cover = 1.5 (i.e.,clear ) + 0.375 (i.e.,stirrup)
+ 0.438 (i.e., bar radius) = 2.31 in.
The clear spacing between bars may be obtained from the geometry of the cross section
with the aid of Table 3.9.3. From this table, the minimum width for a beam with 2–#8
bars and 2db clear spacing is 8.3 in. For a beam with 2–#8 and 2–#7 bars, the clear spac-
ing between bars may be computed as follows:
1
clear spacing =
3
{13 − [8.3 − 2(1.00)] − 2(0.875)} = 1.65 in. or 1.9db
Therefore, one-half of the center-to-center spacing of the 2–#7 bars is
1 1
c-c spacing = [1.65 + (0.875)] = 1.26 in. governs!
2 2
(Continued)
284
Assuming no contribution from stirrups, Ktr = 0 and

(cb + Ktr)/db = (1.26 + 0)/0.875 = 1.44 < 2.5 max
Thus,
 3 60, 000 1.3(1.0)1.0 

Ld =  0.875 = 56.2 in. (4.7
7 ft)
 40 1.0 4000 1.44 
where ψt = 1.3 is applied to account for the casting position of the bar. Alternatively, if
minimum #3 U stirrups spaced at d/2 = 10 in. are assumed to restrain the opening of the
splitting plane, Ktr is computed under the assumption that the potential splitting plane is
horizontal and that two of the four bars are being developed as follows:
40 Atr 40(2)(0.11)
K tr = = = 0.44
sn 10(2)
and
(cb + K tr ) / db = (1.26 + 0.44) / 0.875 = 1.94 < 2.5 max
Thus, the development length could be reduced to
1.44
Ld = 56.2 = 41.7 in. (3.5 ft)
1.94
Based on the general equation, Eq. (6.6.1), the development lengths at the various
locations are
Bars Location Ld (in.)
#4 3 (bottom) 14.2
#5 1 and 2 (bottom) 17.8
#7 4 (top) 32.4
5 and 6 (top) 41.7
7, 8, and 9 (bottom) 30.7
#8 5, 6, 10, and 11 (top) 38.9
A comparison of these values with those computed using the simplified equations
(Categories A and B) in part (b) reveals that the simplified equations can be very con-
servative. For example, for the #8 bars, the simplified equations (Category A) require a
development length that is about 60% greater than that required by the general equation.
Also note that for the #4 and #5 bars, a bar size factor ψs = 1.0 was used in the general
equation, while the simplified equations already include ψs = 0.80.
(d) Crack control requirements. At this point it is useful to check the distribution
and spacing between bars for crack control in accordance with the serviceability
requirements of ACI-24.3. For the purposes of this example it will be assumed that
the provided reinforcement satisfies the requirements for crack control. Crack con-
trol provisions are presented in Chapter 12.
285
9.5 SHEAR REINFORCEMENT IN BEAMS 285
9.5 SHEAR REINFORCEMENT IN BEAMS

The design of shear reinforcement in beams 2B1-2B2-2B1 is shown in the following example.
EXAMPLE 9.5.1
Design the shear reinforcement in the floor beams 2B1-2B2-2B1 of Example 9.4.1. Use
fc′ = 4000 psi and f y = f yt = 60, 000 psi.
SOLUTION
The factored shear Vu envelope for which shear reinforcement will be provided is taken
from Figs. 9.3.3 and 9.3.4 and shown in Fig. 9.5.1. For these beams, the maximum fac-
tored shear may be taken at a distance equal to the effective depth from the face of sup-
port. Let V1, V2, and V3 be the maximum factored shear in regions 1, 2, and 3 as shown
in Fig. 9.5.1; then using wu = 3.45 kips/ft,
26’–0” 13’–0”
Symmetrical
15.00’ about CL
47.7
d V3 = 39.4
40.7 V1 = 32.4 1.67’
12.58’ Region 3
Region 1 or 151.0”
x2 =
φVc = 24.7 43.0” 65.1”
φVc/2 φVc/2
V, kips
φVc/2
x1 = 26.8” 43.0” x3 = 43.0”
51.1”
9.38’ 11.41”
0.75’
or 112.6” d or 136.9”
d Region 2
1.67’ 1.67’
13.83’
11.80’ V2 = 43.4
51.8
2B1 2B2
Figure 9.5.1 Factored shear Vu envelope used for 2B1 and 2B2 in Example 9.5.1.
V1 = 3.45(9.38) = 32.4 kips

V2 = 3.45(12.58) = 43.4 kips
V3 = 3.45(11.41) = 39.4 kips
The concrete contribution to shear strength, φVc, is
( )
φ Vc = φ 2λ fc′ bw d = 0.75 2(1.0) 4000  (13)(20) 1000
1
= 24.7 kips
Thus, the distance from the location where V1, V2, and V3 occur to the location where
Vu = φVc is (see Fig. 9.5.1):
 32.4 − 24.7 
x1 = 112.6   = 26.8 in.
 32.4 
 43.4 − 24.7 
x2 = 151.0   = 65.1 in.
 43.4 
 39.4 − 24.7 
x3 = 136.9   = 51.1 in.
 39.4 
(Continued)
286
The regions spanning from the face of the support to the location where Vu = φVc /2,
which represent the portions of the beam where shear reinforcement is required, are
drawn from Fig. 9.5.1 as regions 1, 2, and 3 in Fig. 9.5.2.
Assume #3 vertical U stirrups. From Fig. 9.5.2 the maximum required φVs anywhere
on the beam is (43.4 –24.7) = 18.7 kips. As shown below, the largest required φVs does
not exceed that based on a nominal stress vs of 4 fc′,

limit φ Vs = φ (4 fc′bw d ) = 2(φVc ) = 2(24.7) = 49.4 kips
max required φ Vs = 18.7 kips < limit φVs
Thus, the upper limit on stirrup spacing (Eq. 5.10.19) is d/2 = 10 in. (< 24 in.)
For a minimum percentage of shear reinforcement (Eq. 5.10.13),
min φVs = φ 0.75 fc′ bw d ≥ φ (50)bw d [5.10.13]

1
= 0.75(0.75 4000 )(13)(20) = 9.2 kips
1000

1
≥ 0.75(50)(13)(20) = 9.8 kips
1000
Thus, the 9.8-kip limit controls.

Face of support Required φVs
= Vu – φVc
Face of support
13 @ 10”, 1 @ 5”
7.7
Vu – φV c , kips
128.1” 14.7
65.1” 20”
26.8”
20” 20” 51.1”
90” 82.6”
18.7
1 @ 5”, 9 @ 10” 1 @ 5”, 8 @ 10”
Required φVs
= Vu – φVc
(a) Region 1 (b) Region 2 (c) Region 3
Figure 9.5.2 Portions of factored shear Vu in excess of φVc (i.e., required φVs diagram) for
beams 2B1 and 2B2 in Example 9.5.1.
The strength requirement to provide for the shear represented by the shaded areas of
Fig. 9.5.2 is
Av f y d
Vs =
s
or

φ Av f y d 0.75(0.22)(60)20 198 kips
s= = = in.
φVs φVs φVs
For region 1, a spacing of d/2 governs because the maximum required φVs = 7.7 kips
is less than the 9.8 kips required for minimum stirrups, and φVs = 9.8 kips permits spac-
ing of 20.2 in., which exceeds d/2.
(Continued)
287
9 . 6 D E TA I L S O F BA R S I N B E A M S 287
Use 1 @ 5 in. and 9 @ 10 in. (95 in. from face of support).

For region 2, the required spacing for the maximum required φVs of 18.7 is 198/18.7 =
10.6 in., which is larger than the maximum permitted spacing of d/2 = 10 in.
Use 1 @ 5 in. and 13 @ 10 in.
For region 3, to satisfy the strength requirement at the critical section,
198
max s = = 13.5 in.
14.7
Thus, d/2 = 10 in. controls.
Use 1 @ 5 in. and 8 @ 10 in.
As a practical matter, many designers would place stirrups by scaling from the fac-
tored shear envelope (in terms of force or unit stress), rather than accurately compute
distances as illustrated here. Further, this example uses the simplified procedure of ACI-
22.5.5.1 for the value of Vc. It is noted that stirrup spacing will be modified on both
beams near the interior support based on development length requirements, as shown in
the next section.
9.6 DETAILS OF BARS IN BEAMS

For typical conditions of equal spans and uniform load where moment and shear coef-
ficients of ACI-6.5.2 are used, standard bar details such as provided in the ACI Detailing
Manual—2004 [2.23] may be used. When the moment and shear envelopes are available,
cutoff locations and embedment lengths into the supports should be determined therefrom,
as illustrated shortly in Example 9.6.1.
Structural Integrity Reinforcement

In addition to satisfying the basic structural safety requirement [design strength ≥ required
strength: (Eq. 2.6.1)], the ACI Code has reinforcement and detailing requirements aimed at
enhancing the overall integrity (i.e., redundancy and ductility) of the structure, in the event
of severe damage to a major supporting element. Such requirements are often referred to as
structural integrity requirements.
For beams located along the perimeter of the structure, a minimum amount of struc-
tural integrity reinforcement must be provided at supports as follows (ACI-9.7.7.1 and
ACI-9.7.7.4):
(a) At least one-quarter of the maximum positive moment reinforcement, but not less
than two bars, must be continuous.
(b) At least one-sixth of the negative moment reinforcement at the support, but not less
than two bars, must be continuous.
(c) At noncontinuous supports, structural integrity reinforcement must be anchored at
the face of the support to develop fy.
This longitudinal integrity reinforcement must be enclosed by closed stirrups in accordance

with ACI-25.7.1.6 (see Fig. 9.6.1) or hoops along the clear span of the perimeter beams.
For nonperimeter beams, the longitudinal reinforcement must be enclosed by closed
stirrups in accordance with ACI-25.7.1.6 or by hoops along the clear span of the beam
(ACI-9.7.7.2). Alternatively, at least one-quarter of the maximum positive moment rein-
forcement, but not less than two bars, must be continuous and, at noncontinuous supports,
28
90-degree hook permitted

if spalling restrained by flange
or slab
Cap tie
U stirrup
with
135-degree 135-degree
hooks hooks
(a) (b)
Figure 9.6.1 Details of closed stirrups used for enclosing the longitudinal structural integrity
reinforcement (ACI-25.7.1.6) (a) Single piece with ends terminating with 135° standard hooks and
(b) two-piece stirrup made of a U stirrup with 135° hooks and a cap tie.
be anchored to develop fy at the face of the support, as for perimeter beams. This require-
ment will usually require a standard hook or a headed bar.
For beams or girders supported on columns or walls, the structural integrity reinforce-
ment is required to pass through the region bounded by the column (or wall) reinforcement
(ACI-9.7.7.3). Also, the structural integrity reinforcement may be spliced if necessary. In
such a case, splices must satisfy the requirements of ACI-9.7.7.5 and 9.7.7.6.
EXAMPLE 9.6.1
Determine the lengths of the main reinforcing bars in 2B1 and 2B2 of Example 9.4.1.
SOLUTION
(a) In determining bar cutoff locations, it is necessary first to locate the theoretical
points where the bars are no longer required. Lines representing full-strength φMn of
the various bar combinations are computed and compared with the factored moment
Mu, as shown in Fig. 9.6.2.
(b) Extension of positive moment reinforcement into supports. To satisfy ACI-
9.7.3.8.2, the #5 straight bars in 2B1 must extend at least 6 in. into the supports. At
the first interior support, the 6-in. minimum embedment is sufficient, since there is
no computed tensile requirement for these bars unless the bars are (1) to be used as
compression reinforcement or (2) are required for “structural integrity” under ACI-
9.7.7. In this design, beam 2B1 is not a perimeter beam and the bottom bars are
not utilized as compression reinforcement. While #3 U stirrups will be provided,
they must be closed by using, for example, the details shown in Fig. 9.6.1. In such
a case, an extension greater than the 6-in. minimum is not required. However, since
the extra cost of providing more ductility and tying the structure together is mini-
mal, the writers recommend that, in general, the positive moment bars required to
extend into the support be capable of developing fy at the face of support, a prop-
erty that often requires a Class A splice when the bars are not continuous through
the support.
At the exterior support, the 6-in. embedment will not be adequate if the support is a
column and the flexural member is part of the primary lateral load resisting system. ACI-
9.7.3.8.2 requires development of the full yield stress in tension at the face of support
for such cases. In this example, the supporting member is a girder and is not a part of
(Continued)
289
the primary lateral load resisting system; thus a 6-in. embedment is sufficient, provided
that closed stirrups are used along the clear span. Similarly, the 2–#4 bars in beam 2B2
are extended into the supports 6 in. The total length of the #4 and #5 straight bars is [24 ft
6 in. + 2(6 in.)] = 25 ft 6 in. (Bar lengths are usually specified in 3-in. increments.)
(c) Theoretical cutoff locations. In establishing the cutoff for the #7 bottom bars near
sections A-A and C-C in beam 2B1, and near section D-D in beam 2B2, the distances
x1, x2, and x4, as marked in Fig. 9.6.2 [also shown as x1, x2 in Fig. 9.3.3(a) and as x4
in Fig. 9.3.4(a)] must be computed. Thus, referring to Fig. 9.3.3(a),
3.45(11.05 − x1 )2 3.45(13.45 − x2 )2
= = 159 − 56 = 103 ft-kips
2 2
which gives:
x1 = 3.31ft; and x2 = 5.72 ft.
And, from Fig. 9.3.4(a),
3.45(12.25 − x4 )2
= 121 − 36 = 85 ft-kips
2
x4 = 5.23 ft
In accordance with ACI-9.7.3.3, the reinforcement must be extended a distance of

12 bar diameters or the effective depth of the member, whichever is larger, beyond the
point at which it is no longer needed to resist flexure. In this case, the effective depth of
the member, 1.67 ft, controls for all bar sizes. Thus, the cutoff locations (from the face
of the support) for the #7 bottom bars could be:
cutoff based on x1 = 3.31 − 1.67 = 1.64 ft


Similarly, to determine the cutoff locations for the #7 top bars near section C-C in
beam 2B1 and near section D-D in beam 2B2, the distances x3 and x5, as marked in
Fig. 9.6.2 [x3 in Fig. 9.3.3(b) and x5 in Fig. 9.3.4(b)] are computed as follows. From
Fig. 9.3.3(b),
3.45(14.25 − x3 )2
= 142 + 135= 277 ft-kips
2
x3 = 1.58 ft
cutoff based on x3 = 1.58 + 1.67 = 3.25 ft
And, from Fig. 9.3.4(b),
3.45(13.08 − x5 )2
= 104 + 135 = 239 ft-kips
2
x5 = 1.31 ft
cutoff based onn x5 = 1.31 + 1.67 = 2.98 ft
The distances x1 through x5 with the effective depth d = 1.67 ft either added or subtracted
locate correctly the theoretical cutoff points. However, additional requirements for bars
terminated in a tension zone may require an extension beyond the calculated cutoff loca-
tions above. These requirements are checked in the next section.
(Continued)
290
Figure 9.6.2 Moment envelope, φMn diagram, and bar arrangement for beams 2B1 and 2B2 of Example 9.2.1, as treated in Example 9.6.2 (all moments Mu and φMn are given in
ft-kips). Stirrups are not shown for clarity; LL = live load.
(Continued)
291
(d) Check cutoff location for the 2–#7 bottom bars in beam 2B1. The proposed cutoff
locations at x1 – d (1.64 ft) and at x2 – d (4.05 ft) are both in a tension zone [see
Figs. 9.3.3(a) and 9.6.2]. Therefore, ACI-9.7.3.5 must be satisfied for the cutoff to
be acceptable at these locations. From Fig. 9.6.2, it is clear that the continuing bars
(2–#5) provide more than double the area required for flexure at the potential cut-
off point. Thus, it is necessary to check only whether the factored shear Vu exceeds
three-fourths of the shear strength φVn [Eq. (6.11.3)].
At x1 – d (= 1.64 ft) [see Fig. 9.3.3(a)]:
Vu = wu (11.80 − 0.75 − x1 + d )

Vu = 3.45(11.80 − 0.75 − 1.64) = 32.5 kips
It is required that
Vu ≤ [0.75φVn = 0.75φ (Vc + Vs )] [6.11.3]
The strength provided, including stirrups, is

198
φ Vc + φ Vs = 24.7 + = 44.5 kips
10
Vu 32.5
= = 73% < 75% permitted OK
φVn 44.5
Had the requirement not been satisfied, the stirrup spacing could have been reduced so
that Vs would increase. Alternatively, the 2–#7 bars may be extended to the inflection
point, where they will no longer be in the tension zone. There is little material saved by
cutting so close to the support, but the case serves to illustrate the procedure.
At x2 – d (= 4.05 ft) [see Fig. 9.3.3(a)]:

Vu = wu (14.20 − 0.75 − x2 + d )

Vu = 3.45(14.20 − 0.75 − 4.05) = 32.4 kips
Since the strength provided by the stirrups is the same as that at x1 – d, it is clear that the
requirement is also satisfied at this location.
(e) Check cutoff for 2–#7 bottom bars in beam 2B2. At the proposed location, x4 – d
(3.56 ft), the bars terminate in the compression zone [see Figs. 9.3.4(a) and 9.6.2].
Therefore, the proposed locations is acceptable.
(f) Check cutoff location for the 2–#7 top bars in section C-C of beam 2B1. The pro-
posed cutoff locations at x3 + d (3.25 ft) is a tension zone (see Fig. 9.6.2). Check
whether the continuing bars (2–#8) provide at least double the area required for flex-
ure at the potential cutoff point (Fig. 9.6.2). At the cutoff location, it is required that
φ M n (2 −# 8) ≥ 2 Mu at (x3 + d )

135 ft-kips ≥ 2(67. 4) = 134.8 ft-kips OK
Therefore, it is necessary only to check whether the factored shear Vu exceeds three-
fourths of the shear strength φVn (Eq. 6.11.3).
At x3 + d (= 3.25 ft) [see Fig. 9.3.3(b)]:
Vu = wu (15.00 − 0.75 − x3 − d )

Vu = 3.45(15.00 − 0.75 − 3.25) = 38.0 kips
(Continued)
29
Thus,
Vu 38.0
= = 85% > 75% permitted NG
φ Vn 44.5
In this case, two possible solutions are as follows: (a) extend the 2–#7 bars to a point
where the required shear, Vu, is equal to or less than 0.75(44.5) = 33.4 kips (i.e., at 4.57 ft
from the face of the support), or (b) reduce the stirrup spacing so that Vs increases in this
region. In this example, alternative (b) is chosen to illustrate the procedure.
Try reducing stirrup spacing to 6 in.
198
φ Vc + φ Vs = 24.7 + = 57.7 kips
6
Vu 38.0
φ Vn 57.7
Thus, make the cutoff at x3 + d = 3.25 ft, but use a 6-in. stirrup spacing between the face
of the support and the cutoff location. It is good practice to extend stirrups beyond the
exact cutoff location.
(g) Check cutoff location for the 2–#7 top bars in section D-D of beam 2B2. The pro-
posed cutoff locations at x5 + d (2.98 ft) is in a tension zone (see Fig. 9.6.2). Check
whether the continuing bars (2–#8) provide at least double the area required for
flexure at the potential cutoff point. At the cutoff location,
φ M n (2 −# 8) = 135 ft-kips < 2 Mu at (x 5 + d ) = 2(72. 2) = 144. 4 ft-kips NG
Therefore, the cut cannot be made at this location unless the factored shear Vu at the cut-
off point does not exceed two-thirds of the design shear strength φ Vn [(ACI-9.7.3.5(a)].
At x5 + d (= 2.98 ft) (see Fig. 9.3.4(b)]:
Vu = wu (13.83 − 0.75 − x5 − d )

Vu = 3.45(13.83 − 0.75 − 2.98) = 34.8 kips
It is required that
2 2 
Vu ≤  φ Vn = φ ( Vc + Vs ) [6.11.1]
3 3 
Vu 34.8
= = 78% > 67% permitted NG
φ Vn 44.5
Again, two possible solutions are suggested: (a) extend the 2–#7 bars to a point where
the required shear, Vu, is equal to or less than 0.67(44.5) = 29.8 kips or (b) reduce the
stirrup spacing so that Vs increases in this region. Using a stirrup spacing of 6 in.,
198
φ Vc + φ Vs = 24.7 + = 57.7 kips
6
Vu 34.8
φ Vn 57.7
Thus, make the cutoff at x5 + d = 2.98 ft, but use a 6-in. stirrup spacing between the face
of the support and the cutoff location (good practice is to extend stirrups at 6 in. beyond
the exact cutoff location).
(Continued)
293
(h) Inflection point extensions.

Top bars (negative moment)
ACI-9.7.3.8.4 (see Section 6.11) requires that at least one-third of the negative moment
reinforcement at a support have an extension beyond the point of inflection a distance equal
to the largest of one-sixteenth of the clear span, the effective depth of the member, or 12
bar diameters. In this case, the effective depth of the member, 1.67 ft, controls. In mem-
bers 2B1 and 2B2, the continuing bars (2–#8) represent more than one-third of the maxi-
mum negative moment reinforcement at the interior support (sections C-C and D-D). From
Fig. 9.6.2 and Fig. 9.3.5(a), the cutoff point for the #8 bars in member 2B1 may be made at
7.77 (inflection point) + 1.67 = 9.44 ft from the face of the support. Make the cutoff at 9 ft
6 in. from the face of the support (see Fig. 9.6.2). In member 2B2, a negative moment of
2.1 ft-kips is required at midspan [Fig. 9.3.6(a)]; make the #8 bars continuous in this span.
At the end support, all of the negative moment reinforcement (2–#7) is to be extended
into the span. From Fig. 9.3.3(a), this reinforcement may be terminated at (2.20 –0.75
+ 1.67) = 3.12 ft. Extend the bars into the span a distance of 3 ft 3 in. from the face of
the support.
Bottom bars (positive moment)
At the inflection points (see Section 6.14), ACI-9.7.3.8.3(b) requires that
Mn
+ La ≥ Ld
Vu
At location 1 in Fig. 9.4.1
For 2–#5, extending 6 in. beyond face of support [actual length from inflection point
(see Fig. 9.3.3) = 2.20 –0.25 = 1.95 ft],
56
Mn = = 62 ft-kips
0.90
Vu = 40.7 − 3.45(2.20) = 33.1 kips
La = d = 1.67 ft < actual length 1.95 ft
Mn 62
+ La = + 1.67 = 3.5 ft > Ld (# 5) = 1.5 ft. OK
Vu 33.1
At location 2 (Fig. 9.4.1)
For 2–#5, since Mn and Vu are identical to their values at location 1, this check is made
by inspection.
At location 3 (Fig. 9.4.1)
For 2–#4,
36
Mn = = 40 ft-kips
0.90
Vu = 44.9 − 3.45(4.61) = 29.0 kips
Mn 40
+ La = + 1.67 = 3.05 ft > Ld (# 4) = 1.2 ft. OK
Vu 29.0
(i) The summary of bar arrangement, lengths, dimensions, and design strength pro-
vided was shown in Fig. 9.6.2. Bar cutoff locations have been rounded off so that
bar lengths are specified in 3-in. increments. It is noted that the 2–#7 top bars are
(Continued)
294
not fully developed at the face of the interior support. However, the design strength
exceeds the required flexural strength at that location, as it does everywhere else
in the spans for both beams. Furthermore, it can be shown that the development
length for these bars could be reduced to 3 ft (instead of 3.5 ft) when the reduc-
tion in stirrup spacing to 6 in., as computed in parts (f) and (g), is accounted for
in the general equation for computing development length in part (c), item vi, of
Example 9.4.1. Regardless of whether the reduction in stirrup spacing is included
in the calculations, the bar arrangement and cutoff locations are acceptable. The
cross sections for the final choice at each designated section of Fig. 9.6.2 were
(j) Stirrup development. For stirrup supports, 2–#3 bars are provided in the central part of
beam 2B1. The dimensions of the stirrups are shown in Fig. 9.6.3. Closed stirrups with
ends terminating in 135-degree hooks must be provided over the entire clear span in
accordance with the structural integrity requirements (ACI-9.7.7.2 and ACI-25.7.1.6).
To facilitate fabrication and erection of the steel reinforcement cage, the stirrup end
closest to the slab may be terminated in a 90-degree hook (Fig. 9.6.1) because poten-
tial spalling may be assumed to be restrained by the slab [ACI-25.7.1.6(b)]. As an
alternative to providing closed stirrups over the entire span, continuing reinforcement
in the bottom of the beam could be made continuous over the interior supports (ACI-
9.7.7.2) and anchored to develop fy at the noncontinuous end support (ACI-9.7.7.4).
10”
This end could terminate

in a 90-degree hook because
19 12” spalling is restrained by the slab
3
8”
Figure 9.6.3 Stirrup details in beams 2B1 and 2B2.
9.7 SIZE OF GIRDER WEB
For the purpose of determining the size of the girder web, the maximum negative moment
at the exterior face of the first interior support may be used, taken as 0.8 of the maximum
positive moment in a simple span with span length equal to the clear span of the girder and
with loadings identical to those on the girder. The maximum shear at the same location may
be taken as 1.20 times the reaction to the simple span described above. Note that because
concentrated loads are involved, the coefficients of ACI-6.5.2 cannot be used in lieu of an
elastic analysis.
The concentrated loads on the girder may be taken as half of the total factored dead
and live loads on the clear span of the beams on both sides of the girder. The dead and
live uniform load on the girder will be the weight of the concrete and the live load on the
floor, respectively, both within the width of the girder. This loading transfer is a sufficiently
good approximation, although it is probable that the slab weight and floor load on a narrow
strip parallel and close to either edge of the girder would act on the girder as uniform load
instead of being carried by the one-way slab to the adjacent beams and then to the girder
as concentrated loads.
295
9.7 SIZE OF GIRDER WEB 295
EXAMPLE 9.7.1
Design the floor girders 2G1-2G2-2G2-2G1 as shown in the floor framing plan of
Fig. 8.3.1. Assume a slab thickness of 5.5 in. Work through the choice of the size of the
girder web in this example according to the ACI Code ( fc′ = 4000 psi; fy = 60,000 psi).
SOLUTION
The concentrated reactions from the beams using the 24.5-ft clear span are
dead load = 1.2(24.50)[(0.069)(13) + 0.25] = 33.7 kips
live load = 1.6(24.50)(0.100)(13) = 51.0 kips

total load = 33.7 + 51.0 = 84.7 kipss
Assuming an 18 × 35 in. web, the uniform loads on the girder are
uniform dead load = 1.2 [18(40.5)(0.150)]

1
= 0.91 kips/ft
144
uniform live load = 1.6(0.100)(1.5) = 0.24 kip/ ft

uniform total load = 0.91 + 0.24 = 1.15 kips/ft
Maximum positive moment on a simple span (see Fig. 9.7.1) is
1
Mu = 84.7(12.25) + (1.15)(37.5)2 = 1240 ft-kips
8
The estimated maximum negative moment in the girder at the exterior face of the first
interior support is 0.8(1240) = 992 ft-kips. From Fig. 9.7.1, estimate
dneg = 40.50 − 4.75 = 35.75 in.

dpos = 40.50 − 3.50 = 37 in.
The negative moment requirement is

Mu 992 (12, 000)
φ bd 2
0.90(18)(35.75)2
From Fig. 3.8.1, ρ ≈ 0.011 which is less than ρtc = 0.0181 (see Table 3.6.1). Thus, the
section is tension controlled and required As for (–M) ≈ 0.011(18)35.75 = 7.08 sq in.
For the positive moment requirement, using the coefficients of ACI-6.5.2 to obtain
the approximate proportion between positive and negative moments,
10 10
estimated (+ M ) ≈ ( − M ) = (992) = 709 ft-kips
14 14
Mu 709(12)
required As ≈ = = 4.77 sq in.
φ f y (arm) 0.90(60)(≈ 33)
This may well fit into one layer (5–#9 require a beam width of 14.3 in. according to
Table 3.9.2).
For the shear requirement,
max Vu ≈ 1.20 [84.7 + 1.15(18.75)] = 128 kips
128, 000
vu = = 265 psi = 4.2 fc′
0.75(18)(35.75)
(Continued)
296
The stem could be made smaller; with this relatively large depth, deflection is unlikely
to be excessive. The 18 × 35 stem appears to be somewhat large, though it could cer-
tainly be used. In this case, because of the large depth, reduce the stem size to 18 × 29.
Estimated effective depths become
dneg = 29.75 in.

dpos = 31 in.
P P
12’–3” 13’–0” 12’–3”
18 × 18 18 × 18
column w column
Clear span = 37’– 6”
202 ft-kips
Simple beam
moments
1035 ft-kips
3 14 ” clear (to allow placement of slab

and beam reinforcement) 4.75”
5 12 ” Measured to
top of girder
steel
34.5” (40.5” initially

assumed)
2” clear
18” 3.5”
Figure 9.7.1 Loading information for floor girders 2G1 and 2G2 in Example 9.7.1.
revised girder weight =1.2 [18(34.5)(0.15)]

1
= 0.78 kip/ft
144
wu = 0.78 + 0.24 = 1.02 kips/ft

1
revised Mu (simple beaam) = 1038 + (1.02)(37.5)2 = 1217 ft-kips
8
0.8(1217)(12, 000)
required Rn = = 814 psi < 913 psi OK
0.90(18)(29.75)2 (max. for tension-
controlled sections)
124, 600
vn at support = = 310 psi = 4.9 fc′
0.75(18)(29.75)
Use 18 × 29 stem section.
297
9 . 8 C O N T I N U O U S F R A M E A N A LY S I S F O R G I R D E R S 297
9.8 CONTINUOUS FRAME ANALYSIS FOR GIRDERS

Although floor girders subjected to large concentrated loads are structural members of
common occurrence, the ACI Code makes no mention of moment coefficients for these
cases. In the following example, elastic statically indeterminate structural analysis is used
for 2G1-2G2-2G2-2G1 (Figure 8.3.1).
EXAMPLE 9.8.1
By the use of an elastic statically indeterminate structural analysis, determine the shear
and moment diagrams to be used in the design of girders 2G1 and 2G2 in Example 9.7.1
( fc′ = 4000 psi; fy = 60,000 psi).
SOLUTION
As discussed in Section 9.3, there are various ideas regarding what constitutes the cor-
rect stiffness for continuous T-sections. In accordance with the thought that the true
stiffness is that of a span with a variable cross section, the effective flange width for an
equivalent uniform moment of inertia section may be assumed to be twice the web width
(Fig. 9.8.1). The centroid of the gross area of the T-section is at
18(29)14.5 − 36(5.5)(2.75) 7025

y= = = 9.76 in.
18(29) + 36(5.5) 720
The gross moment of inertia Ig is
36(5.5)3 18(29)3
Ig = + − 720(9.76)2 = 79, 700 in.4
3 3

79, 700
K girder = = 2040 in.4 /ft
39
If the size of the upper and lower columns were 18 × 18 in. and the column height
were 15 ft,
18(18)3 1
K col = = 583 in.4 /ft
15 12
36” 5 12 ”
y 29”
18”
Figure 9.8.1 T-section dimensions for stiffness computation in Example 9.8.1.

(Continued)
298
2G1 2G2 2G2 2G1
1st span 2nd span 3rd span 4th span
Figure 9.8.2 Equivalent frame model for 2G1-2G2-2G2-2G1.
Analysis of the frame shown in Fig. 9.8.2 yields the girder end moments shown in
Table 9.8.1 under the various loading conditions considered. The controlling moment
envelopes for spans 2G1 and 2G2 are given, respectively, in Figs. 9.8.3 and 9.8.4. In
practice, this composite diagram should be made instead of the individual moment and
shear diagrams for each loading case.
a
TABLE 9.8.1 END MOMENTS OF GIRDERS 2G1-2G2-2G2-2G1
IN EXAMPLE 9.8.1
Joint A B C D E
Member AB BA BC CB CD DC DE ED
Dead Load Only

M= –153 +467 –438 +365 –365 +438 –467 +153
Live Load Only Spans 1 and 3
M= –230 +353 –171 +137 –306 +361 –213 –45
Live Load Only Spans 1, 2, and 4
M= –177 +606 –599 +230 –91 +186 –362 +228
Live Load Only Span 3
M= –8 –40 +67 +214 –353 +346 –205 –43
M= +35 +165 –279 +566 –566 +279 –165 –35
M= –220 +402 –253 –123 +123 +253 –402 +220
M= +45 +213 –361 +306 –137 +171 –353 +230
a Clockwise moments acting on member ends are taken as positive.
(Continued)
29
9 . 8 C O N T I N U O U S F R A M E A N A LY S I S F O R G I R D E R S 299
A B 3 – #11 C
3 – #9 3 – #11
2 – #10 2 – #9 3 – #10 2 – #10

B
A C
39’–0”
742
598
LL spans 1 and 3
18” column 18”

6.8’ column
3.0’
LL spans 2 and 4
A B
LL span 3
108
11.0’
383 3.7’
427
LL spans 680
1, 2, and 4
820
1073
Figure 9.8.3 Moment envelope and longitudinal bar arrangement for girder 2G1 (span 1).
D 4 – #9
E
3 – #11 4 – #9 F
3 – #11 4 – #9 st
2 – #9 4 – #9 2 – #9
E
F
D
39’–0”
557
514
18” column
18” column
LL spans 2 and 4 6.4’
7.2’
LL spans 1 and 3 LL spans
1 and 4
B C
LL span 3
11.6’ 242
14.6’
371
LL spans 502
609 LL spans 2 and 3
671
1, 2, and 4
799
931
1037
Figure 9.8.4 Moment envelope and longitudinal bar arrangement for girder 2G2 (span 2).
(Continued)
30
13’–0” 13’–0” 13’–0”

CL of support
92.4 CL of support
93.2
79.9
Spans 1 and 3
with LL
9”
Face of support
A B
12.4
25.7 9”
(a) Girder 2G1
Face of support
Spans 1, 2, and 4
with LL
114.9
115.7 110.2
102.5 123.5
122.7
18.0
4.7
B C
12.1
(b) Girder 2G2 Spans 2 and 3

with LL
96.6
109.9
109.1
Figure 9.8.5 Shear envelopes for girders 2G1 and 2G2 (full-span loadings only).
For investigating development length requirements at points of inflection, for check-

ing cutoff acceptability in the tension zone, and for designing stirrups, an approximate
shear envelope (only full-span loadings) is used, as given in Fig. 9.8.5.
9.9 CHOICE OF LONGITUDINAL REINFORCEMENT

IN GIRDERS
As with the design of the floor beams, the choice of longitudinal reinforcement in girders
depends on both the steel area and the development length requirements as dictated by the
moment and shear envelopes such as shown in Figs. 9.8.3 through 9.8.5. Attention is called
to ACI-7.5.2.3, which is applicable when the main reinforcement in the slab is parallel to the
longitudinal axis of the girder. For this situation, reinforcement perpendicular to the longitu-
dinal axis of the beam must be provided in the top of the slab to carry the load on the portion
of the slab acting effectively as the flange of the girder. The overhanging flange is assumed
to act as a cantilever. The spacing of the bars may not exceed 5 times the thickness of the
flange or, in any case, 18 in. (ACI-7.7.2.4).
301
9.9 CHOICE OF LONGITUDINAL REINFORCEMENT IN GIRDERS 301
EXAMPLE 9.9.1
Choose the arrangement of the main reinforcement in the floor girders 2G1-2G2-2G2-
2G1 of Example 9.7.1.
SOLUTION
(a) Sections (rectangular sections) A-A, C-C, D-D, and F-F (Figs. 9.8.3, 9.8.4, and 9.9.1).
Section A-A: The moment at the face of the supports may be obtained directly from the
bending moment diagrams or for preliminary design by subtracting ΔM = 1 3 V × (width
of support) from the moment at the center of support. Using this approximation:
92.4(1.5)
Mu at face = 383 − = 383 − 46 = 337 ft-kips
3

Mu 337(12, 000)
φ bd 2 0.90(18)(30.7)2
where d is taken as 30.7 in. (see Fig. 9.9.1). Either from Eq. (3.8.5) or from Fig. 3.8.1,
required ρ ≈ 0.0045

required As = 0.0045(18)30.7 = 2.49 sq in.
Sections C-C and D-D: The larger moment is at section C-C,
122.7(1.5)
Mu at face = 1073 − = 1012 ft-kips
3

1012(12, 000)
0.90(18)(29.1)2
From Fig. 3.8.1, ρ ≈ 0.0175,

required As ≈ 0.0175(18)(29.1) = 9.2 sq in.
Section F-F:
109.1(1.5)
Mu at face = 931 − = 931 − 55 = 876 ft-kips
3

876(12, 000)
0.90(18)(29.6)2
Using straight-line approximation,
 741
required As ≈ 0.0175  (18)29.6 = 7.9 sq in.
 873 
(b) Sections (T-sections) B-B and E-E (Figs. 9.8.3, 9.8.4, and 9.9.1):
Effective flange width bE is the smallest of:
Ln (39 − 1.5)(12)
1. bw + = 18 + = 131 in.
4 4
2. bw + 16t = 18 + 16(5.5) = 106 in. Controls
It is likely that the depth a of the rectangular stress distribution will be less than the
flange thickness t. Estimate a = 2 in. (about one-half flange thickness).
(Continued)
302
3 14 ” clear
3 – #9 6 – #11
d = 30.6” (approx)
d = 29.1” (approx)
d = 30.7” (approx)
2.6” (approx)
18” (typical)
2 – #9
2 – #10 2 – #10 (Section C–C)
2 – #9 (Section D–D)
2’ clear 3 – #10
Section A–A Section B–B Sections C–C and D–D
4 – #9
5 12 ”
d = 31.9” (approx)
4 – #9
d = 29.6” (approx)
29” (typical)
4 – #9 2 – #9
2.6 (approx)
Section E–E Section F–F
Figure 9.9.1 Typical sections for 2G1 and 2G2 (refer to Figs. 9.8.3 and 9.8.4) for Example 9.9.1.
Section B-B: estimated d = 30.6 in.

Mu 742(12)
φ f y (arm) 0.90(60)(30.6 − 1)
Check:
C = 0.85(4)(106)a = 360.4 a
T = 5.6(60) = 336 kips; a = 0.93 in.

742(12)
revised reequired As = = 5.5 sq in.
0.90(60)(30.6 − 0.47)
Section E-E: estimated d = 31.9 in.

557(12)
0.90(60)(31.9 − 1)
Check:
C = 360.4 a; T = 4.0(60) = 240 kips
a = 0.67 in
557(12)
revised required As = = 3.9 sq in.
0.90(60)(31.9 − 0.33)
(Continued)
30
(c) Confirmation of the tentative arrangement of main reinforcement as summarized in Fig.

9.9.2 awaits the check of development of reinforcement, crack control, and deflection.
Note in Fig. 9.9.2 that the requirements indicated in the compression zone at B and C are merely
the minimums of one-fourth of the positive moment steel required to satisfy ACI-9.7.3.8.2.
(As = 8.00)
(As = 3.00) (As = 9.36) 4 – #9
3 – #9 6 – #11 4 – #9
2.5 2G1 9.2 NA 2G2 7.9
5.5 1.4 NA 3.9 1.0
A 3 – #10 B 4 – #9 C
2 – #9 (As = 4.00)
(As = 5.81)
Figure 9.9.2 Longitudinal reinforcement areas (sq in.) required and bars selected for girders
2G1 and 2G2 for Example 9.9.1.
The details of determining bar lengths are not shown; for the general arrangement, see
Figs. 9.8.3 and 9.8.4.
(d) Check design strength at critical sections based on the selected arrangement of steel
reinforcement. (The effect of compression reinforcement is ignored.)
At section A-A, 3–#9:
C = 0.85(4)(18)a = 61.2 a T = 3.0(60) = 180 kips
a = 2.94 in. c = 3.46 in.
d−x 30.7 − 3.46
ε1 = (0.003) = (0.003) = 0.0236 > 0.005
x 3.46
φ = 0.90
 1 
φ M n = 0.90(180)(30.7 − 1.47) 12 = 395 ft-kips  > [ M u = 337 ft-kips] OK
 
At section C-C, 6–#11 (two layers):
C = 61.2 a T = 9.36(60) = 561.6 kips
a = 9.18 in. c =10.80 in.
1.41
dt = 34.5 − 3.25 − = 30.5 in. (distance to the layer of reinforcement
2
closest to the tension face)
30.5 − 10.8
εt = (0.003) = 0.0055 > 0.005; φ = 0.90
10.8
 1 
φ M n = 0.90(561.6)(29.1 − 4.6) 12 = 1032 ft-kips > [ Mu = 1012 ft-kips] OK
 
Similar calculations can be made for the rest of the sections. The results are summarized
in Table 9.9.1. Note that all sections satisfy minimum reinforcement requirements. Also,
they are all tension controlled, and thus φ = 0.9.
(e) Check the development length requirement at the positive moment inflection points.
From analysis of the girders under the various loading conditions considered
(Example 9.8.1), the factored shear values at the inflection points are as follows:
For span 1,
Vu (near left end ) = 89 kips (live load on spans 1 and 3)

Vu (near right end) = 114 kips (live load on spans 1, 2 and 4)
(Continued)
304
TABLE 9.9.1 CALCULATED FLEXURAL STRENGTH FOR THE GIRDERS

OF EXAMPLE 9.9.1
Section As d εt φMn Mu
(sq in.) (in.) (ft-kips) (ft-kips)
A-A 3–#9 (3.00) 30.7 0.0236 395 337 OK

C-C (D-D) 6–#11 (9.36) 29.1 0.0055 1032 1012 OK
F-F 8–#9 (8.00) 29.6 0.0066 924 876 OK
B-B 2–#9, 3–#10 (5.81) 30.6 0.0774 787 742 OK
E-E 4–#9 (4.00) 31.9 0.1184 568 557 OK
For span 2,
Vu (near left end ) = 106 kips (live load on spans 1,2 and 4)

Vu (near right end) = 101 kips (live load on spans 2 and 3)
At the span 1 inflection points, 2–#10 bars continue along the bottom of the girder.
For 2–#10 bars.
C = 0.85(4)(106)a = 360.4 a
T = 2(1.27)(60) = 152.4 kips
a = 0.42 in.

M n = 152.4 [31.9 − 0.5(0.42)]
1
= 402 ft-kips
12
The development length Ld for the 2–#10 in the bottom of span 2G1 depends on the
clear spacing of those bars. From Table 3.9.3, in order to satisfy 2db clear spacing, the
minimum width is 9.1 in. Thus, Category A, item 2 of the simplified development length
equations is satisfied. From Table 6.6.1,
Ld (for # 10) = 60.2 in.(5.0 ft )
The critical location is at the right inflection point, where maximum Vu occurs (i.e.,
Vu = 114 kips). Thus,
 Mn 402 
 + La = + 2.66 = 6.19 ft  > Ld = 5.0 ft OK
 Vu 114 
where La = effective depth, d, governs for this situation (see Section 6.14).
Use of the general equation, Eq. (6.6.1), shows the value of (cb + Ktr)/db to be larger
than 1.5, thus resulting in a shorter requirement for Ld. In this case, the double-check
is not needed because the Ld requirement has been satisfied by the simplified equation.
In span 2, 2–#9 continue past the inflection point and into the support. For 2–#9 bars,
C = 0.85(4)(106)a = 360.4 a
T = 2(1.00)(60) = 120 kips
a = 0.33 in.

M n = 120 [31.9 − 0.5(0.33)]
1
= 317 ft-kips
12
(Continued)
305
The development length Ld for the 2–#9 in the bottom of span 2G1 near the left support
depends on the clear spacing of those bars. From Table 3.9.3, in order to satisfy 2db clear
spacing, the minimum width is 8.6 in. Thus, Category A, item 2 of the simplified devel-
opment length equations is satisfied. From Table 6.6.1,

Ld (for # 9) = 53.5 in. (4.5 ft )
The critical location is at the left inflection point where Vu = 106 kips. Thus,
 Mn 317 
 + La = + 2.66 = 5.7 ft  > Ld = 4.5 ft OK
 Vu 106 
(f) The required steel reinforcement in the top flange, perpendicular to the longitudinal
axis of the girder (ACI-7.5.2.3), may be computed approximately as follows (Fig. 9.9.3):
1
Mu =
2
[0.069(1.2) + 0.10(1.6)] (3.67)2 = 1.64 ft-kips /ft
1.64(12, 000)
0.9(12)(4.06)2
Main slab
steel, #4 6” or 8”
typical
bE
wD = 69 psf; wL = 100 psf
d = 5.5”
–0.75” cover
44” 18” 44” –0.50” bar diam (#4)
Girder #3 @ 11” –0.19” bar rad (#3)
4.06”
(Girder
steel not
shown)
Figure 9.9.3 Steel reinforcement across the top of girder for Example 9.9.1.
Observing Fig. 3.8.1, this Rn indicates ρ ≈ 0.0025, which is less than the minimum
ρ for beams. Since this reinforcement is for a slab of uniform thickness, ACI-7.6.1.1
applies. The minimum amount of reinforcement for Grade 60 deformed bars is 0.0018
bh. Thus, the steel required is the larger of the following:

As = 0.0025 bd = 0.0025(12)4.06 = 0.12 sq in.
or

As = 0.0018 bh = 0.0018(12)5.5 = 0.12 sq in.
The maximum spacing of this reinforcement is the lesser of 3 times the slab thickness
(i.e., 3 × 5.5 = 16.5 in.) or 18 in. (ACI-7.7.2.3).
Use #3 @ 11-in. spacing. (As = 0.12 sq in./ft)
(g) The girder longitudinal bar arrangement is shown in Figs. 9.8.3 and 9.8.4 and cross
sections were shown in Fig. 9.9.1. The shear reinforcement and dimensions of longi-
tudinal reinforcement are not shown, because little new information will be involved
in their presentation.
306
9.10 ONE-W AY JOIST FLOOR CONSTRUCTION

One-way concrete joist floor construction (Fig. 9.10.1), sometimes called ribbed-slab con-
struction, consists of regularly spaced ribs monolithically built with a top floor slab and
arranged to span in one direction. Such a system may also be designed as a two-way system
(waffle slab) according to the procedures for two-way floor systems treated in Chapters 16
and 17. The dimensions of the one-way joist system are usually such that only temperature
and shrinkage reinforcement is required in the slab. The slab is usually in the range of 2- to
4-in. thick but may occasionally be as much as 6 in. The ribs (joists) of at least 4-in. width
are usually tapered [Fig. 9.10.1(b)] and are designed so that the clear spacing between adja-
cent ribs does not exceed 30 in. During construction, removable and reusable form fillers
are inserted into spaces between the joists. Such fillers may be standard-sized steel “pans”
in 20-or 30-in. widths and depths of 6, 8, 10, 12, 14, 16, and 20 in. Sometimes form fillers
are made from hardboard, fiberboard, glass-reinforced plastic, or corrugated cardboard.
Occasionally, permanent fillers are used consisting of lightweight or normal-weight con-
crete blocks or clay tile blocks [see Fig. 9.10.1(c)].
To limit deflection on floor joist construction, the minimum depth requirements of ACI-
9.3.1.1 may be applied. Whenever excessive deflection may cause cracking or other adverse
effects, deflections must be computed even if ACI-9.3.1.1 has been satisfied.
Concrete joist floor construction is referred to in ACI-9.8. Some of the requirements are
as follows.
1. The joists shall not be farther apart than 30 in. face to face. The ribs shall be not
less than 4 in. wide and of a depth not more than 3.5 times the minimum width of
the rib.
2. The vertical shells of permanent fillers in contact with the joists are permitted to be
included in strength calculations involving shear or negative bending moment pro-
vided the filler material has a unit compressive strength at least equal to that of the
concrete in the joists. In this case, the minimum slab thickness is 1 1 2 in. or 112 of the
clear distance between joists, whichever is larger.
3. When removable forms or fillers having less compressive strength than required in
item 2 are used, the thickness of the concrete slab shall not be less than 112 of the
clear distance between ribs, nor less than 2 in.
For more information on concrete joist construction, see Chapter 10 of Manual of

Standard Practice [9.1], Chapter 4 of Rice, Hoffman, Gustafson, and Gouwens [9.2], and
Chapter 8 of CRSI Design Handbook—2008 [2.21].
12
1
20” or 30”
(b)
12” or 16”
(a) (c)
Figure 9.10.1 Concrete joist floor construction.

307
9.11 DESIGN OF JOIST FLOORS 307
9.11 DESIGN OF JOIST FLOORS
The design of concrete joist floors involves the slab, the joists, and the girders. Generally,
temperature and shrinkage reinforcement is placed at right angles to the joists, and the con-
crete slab is treated as if it were of plain concrete. The short clear span between joists may
be considered as being fixed at both ends.
The joist itself may be designed as a floor beam having a rectangular section in the
region of negative bending and a T-section in the region of positive bending. The critical
design moment curves for each span may either follow the ACI moment coefficients or be
determined by a continuity analysis. Largely because of the interaction of the slab with the
closely spaced joists, ACI-9.8.1.5 permits the shear strength Vc provided by the concrete to
be 10% higher than for regular beams.
The girder is designed as a floor girder, but the load from the joists may be considered
as being uniformly distributed along the span.
In the case of joist floors over removable steel pans, tapered end forms are available that
increase the effective joist width 2 in. on each side for 20-in.-wide forms and 2 1 2 in. on each
side for 30-in.-wide forms over a distance of 3 ft from the end. This increased width may be
necessary to take the large shear or negative bending moment near the end of the span.
EXAMPLE 9.11.1
Design the typical interior span of a concrete joist floor, using 30-in.-wide forms, for
a center-to-center span of 26 ft and clear span of 24 ft 6 in. The live load is 80 psf,
and the additional superimposed dead load is 10 psf. Use normal-weight concrete with
fc′ = 3000 psi and fy = 60,000 psi. Use the moment and shear coefficients of ACI-6.5.2.
SOLUTION
(a) Slab design. Assume a 3-in. slab for weight estimation purposes. The factored load is
wu = 1.2(3.0)(0.15)
1
+ 1.2(0.010) + 1.6(0.080) = 0.185 ksf
12
Assuming that the slab is fixed at its junction with the joist,
1 1
Mu ≈ (0.185)(30)2 = 0.096 ft-kip/ft
12 144
The nominal flexural strength Mn of a plain concrete member in accordance with

ACI-14.5.2.1 is
M n = 5λ fc′Sm (9.11.1)
where Sm is the elastic section modulus. Using a strength reduction factor φ of 0.60 for
plain concrete (ACI-21.2.1), the design strength, φ Mn, of the slab is
 bh 2 
φ M n = 0.60(5)(1.0) 3000 
 6 
Setting φ Mn = Mu and b = 12 in. for a typical 1-ft strip, the required slab thickness, h,
is found.
0.096(12, 000)
h= = 1.87 in., say, 2 in.
1
6
(
(0.60) 5(1.0) 3000 12 )
(Continued)
308
Thus, a 3-in. slab will be thick enough; perhaps 2 1 2 in. would be sufficient.
Use h = 3 in.
Shrinkage and temperature reinforcement must be provided per ACI-9.8.1.7 in an
amount equal to (ACI-24.4.3.2)

As = 0.0018(12)(3.0) = 0.065 sq in./ft
Use welded wire reinforcement, 4 × 12—W2.5 × W1.4 (As = 0.075 sq in.).
The selection is made from Table 9.11.1; As in direction perpendicular to joists is 0.075
sq in./ft, and As in direction parallel to joists is 0.014 sq in./ft. The “4 × 12” means 4-in.
spacing of the wires running perpendicular to the joist and 12-in. spacing of the wires
running parallel to the joist.
TABLE 9.11.1 COMMON WELDED WIRE FABRIC FOR TEMPERATURE

AND SHRINKAGE REINFORCEMENT
Designation
Spacing (in.) of Wire Size Designation As Longitudinal As Transverse

Longitudinala and Longitudinal × Direction (sq in.) Direction (sq in.)
Transverse Wires Transverse
4 × 12 W1.4 × W1.4 0.042 0.014

4 × 12 W2 × W1.4 0.060 0.014
4 × 12 W2.5 × W1.4 0.075 0.014
4 × 12 W3 × W2 0.090 0.020
4 × 12 W3.5 × W2 0.105 0.020
a Longitudinal refers to the slab span, which is perpendicular to the joist span.
(b) Joist design. The overall depth of the joist floor must satisfy the minimum require-
ment of ACI-9.3.1.1 unless deflections are computed; thus, for an interior span
L 26(12)
min h = = = 14.9 in.
21 21
If excessive deflection may cause damage to nonstructural elements, deflection must
be computed in all cases. Assuming that deflection must later be computed, it will be
prudent to make joists deeper than indicated by ACI-9.3.1.1. Assume use of joists 12 in.
deep below the bottom of the slab with a width of 5 in. at the bottom (i.e., average 6 in.).
The weight of the joist will then be
1
wD = [(3.0 + 12)6 + 30(3.0)](0.150) = 0.19 kip/ft
144
The factored load is
1 1
wu = 1.2(0.19) + 1.2(0.010)(35) + 1.6(0.080)(35) = 0.636 kip//ft
12 12
The maximum factored negative moment is
1
Mu = (0.636)(24.5)2 = 34.7 ft-kips
11 (Continued)
309
Using the minimum cover of 3 4 in. (ACI-20.6.1.3.2) and assuming #5 bars for the main
steel,

d = 15 – 0.75 – 0.31 = 13.94 in.
The required Rn then becomes
Mu 34.7(12, 000)
φ bd 2
0.90(5)(13.94)2
Using Eq. (3.8.5),
required ρ = 0.009

required As (over support) = 0.009(5)13.94 = 0.63 sq in.
The minimum reinforcement for the negative moment region of a continuous T-section
in a statically indeterminate member where the flange would be in tension is given by
3 fc′
As ,min = bw d [3.7.10]
fy
but not less than 200bw d / f y. Thus,
3 3000
As,min = (6)13.94 = 0.23 sq in.
60, 000
200
but </ (6)13.94 = 0.28 sq in. Governs!
60, 000
which is less than the amount required for factored loads.
The shear at a distance d from the face of support is

Vu = 0.636(12.25 − 1.16) = 7.05 kips
The shear strength of a joist without shear reinforcement (ACI-9.8.1.5) is
(
φ Vc = φ (1.10) 2 λ fc′ bw d )
1
= 0.75(1.10)[2(1.0) 3000 ](5)(13.94)
1000

= 6.3 kips < 7.05 kips NG
Note that the minimum rather than the average width of the web was conservatively
used in the above calculation. To satisfy the shear strength requirement at the ends, it is
common practice to widen the ends of ribs, as shown in Fig. 9.11.1. When such a taper
is used to satisfy the shear strength requirement, the 30-in.-wide form tapers to 25 in. in
the last 3 ft near the support. At d from the face, the beam width then becomes
 3.0 − 1.16 
bw = 5 +  5 = 8.1 in.
 3.0 

 8.1 
φ Vc = 6.30   = 10.2 kips > [Vu = 7.05 kips] OK
 5.0 
The maximum factored positive moment near midspan is
1
Mu = (0.636)(24.5)2 = 23.9 ft-kips
16
(Continued)
310
Joist width
2” for 20” forms
2 12 ” for 30” forms
Usually 20”
or 30” pans
Girder
3’–0”
Figure 9.11.1 Plan view for example Example 9.11.1 showing ends of tapered pan joists.
For the T-section, assume that the depth of the rectangular stress distribution falls within
the flange; estimate arm to be ≈ 0.95d =0.95(13.94) = 13.2 in. Then, by trial,
Mu 23.9(12)
required As ≈ = = 0.40 sq in.
φ f y (arm ) 0.90(60)(13.2)
Determine moment arm more accurately by using bE as 35 in.
C = 0.85 fc′ ba = 0.85(3)(35)a = 89.3a

T = As f y = 0.40(60) = 24.0 kips
a = 24.0/89.3 =0.269 in.

23.9(12)
revised required As = = 0.39 sq in.
0.90(60)(13.94 − 0.13)
Try 2–#4 bottom bars and 2–#5 top bars (see Fig. 9.11.2), thus
provided As over support = 0.62 sq in. (2 − # 5)

provided As at midspan = 0.40 sq in. (2 − # 4)
Distribution rib A 2 – #5
2 – #4
1 – #4
A
24’–6”
3”
12” 12 2 – #4
1
5” 30”
Section A–A
Figure 9.11.2 Concrete joist floor using removable pans for Example 9.11.1.

(Continued)
31
To confirm the choice of main reinforcement, the design strength φ Mn must be checked
at midspan and at the support.
At the support (neglecting the increased width due to tapered ends),
C = 0.85(3)(5)(a ) = 12.75a T = 0.62(60) = 37.2 kips
37.2 2.92
a= = 2.92 in. c = = 3.44 in.
12.75 0.85
d −c 13.94 − 3.44
εt = (0.003) = (0.003) = 0.0092 > 0.005
c 3.44
Thus, φ = 0.90
φ M n = 0.90(37.2) [13.94 − 0.5(2.92)]

1
12

= 34.8 ft-kips ≈ [ Mu = 34.7 ft-kips] OK
At midspan,
C = 89.3a T = 0.40(60) = 24.0 kips
24.0 0.27
a= = 0.27 in. c= = 0.32 in.
89.3 0.85
13.94 − 0.32
εt = (0.003) = 0.128 > 0.005
0.32

φ M n = 0.90(24.0) [13.94 − 0.5(0.27)]
1
12
= 24.8 ft-kips >[Mu = 23.9 ft-kips] OK
Use 2–#4 bottom bars and 2–#5 top bars

(c) Development length requirements. Consistent with a loading for the maximum pos-
itive moment in an interior span, the inflection point is at 0.354Ln from centerline
of span [see Fig. 7.5.4(a)]. Thus the development length requirement at inflection
points (ACI-9.7.3.8.3) must be checked. Only one #4 bar will be extended beyond
the inflection point and into the support at least 6 in.
C = 0.85 fc′ bE a = 0.85(3)(35)a = 89.3a

T = 0.20(60) = 12.0 kips
a = 0.13 in.
M n = 12.0 [13.94 − 0.5(0.13)]

1
= 13.9 ft-kips
12

Vu = 0.636(0.354)(24.5) = 5.5 kips
La = 13.9 in. (larger of 12db or effective depth d)
The equivalent embedment length provided is
Mn 13.9(12)
+ La = + 13.9 = 30.3 + 13.9 = 44.2 in.
Vu 5.5
(Continued)
312
The development length Ld for the #4 that extends beyond its cutoff location depends
entirely on the amount of cover, because there are no other bars in the development
length and no stirrups are used. The bottom cover is the minimum 3 4 in., which exceeds
db but is less than 2db for the #4 bar; thus Category A in the simplified method is not
satisfied. From Table 6.6.2 for Category B,

Ld (for # 4) = 32.9 in.(2.7 ft)
which exceeds the 12-in. minimum.

Thus, the check of ACI-9.7.3.8.3 is completed by showing that 44.2 in., the equiva-
lent embedment provided, exceeds Ld of 32.9 in. and thus is satisfactory.
Frequently a transverse distribution rib (Fig. 9.11.2) is used, having a 4-in. mini-
mum width and containing at least one #4 bar both top and bottom. Such a rib would be
located at the third points of the span for spans greater than 30 ft. For spans between 20
and 30 ft, one rib should be used near midspan; for spans less than 20 ft, the rib may be
omitted [2.21].
9.12 REDISTRIBUTION OF MOMENTS—
INTRODUCTION TO LIMIT OR PLASTIC ANALYSIS
Methods of proportioning beams for flexure, shear, and bar development requirements have
been discussed in Chapters 3, 4, 5, and 6 and further illustrated in Sections 9.1 through
9.11. When the inelastic behavior of concrete and steel has been accounted for in the design
of a cross section based on its strength, it may seem somewhat illogical to have used an
elastic analysis to determine the design moments and shears. However, because the evalua-
tion of the true ultimate strength, or limit strength, of an entire structure requires a difficult
and elaborate inelastic analysis, the procedures presented in the preceding sections and
chapters are used. These are both conservative and safe.
The concept of redistribution of moments introduced into the ACI Code in 1963 was the
result of considerable research [9.3–9.9] into the limit behavior of the entire structure (pri-
marily continuous beams) beyond the elastic range to the point where the collapse load is
reached. The use of plastic theory requires that the sections involved actually behave plas-
tically. Figure 9.12.1(a) shows the ideal relationship of moment, M, to curvature, φ, where
there is a perfectly elastic portion and an ideal plastic portion. Such an ideal relationship
is often referred to as elastic−perfectly plastic relation. In reality, reinforced concrete may
exhibit an M-φ relation as shown in Fig. 9.12.1(b) and Fig. 3.13.1, which may often be
approximated by such an ideal behavior.
Consider the statically determinate, simply supported beam of Fig. 9.12.2 with a concen-
trated load at midspan. The limit load on such a system is reached when the extreme fiber of
concrete in compression reaches the crushing strain εcu (taken by ACI as 0.003) after the lon-
gitudinal reinforcement has yielded. In limit analysis, the nominal moment strength, Mn, is
taken as Mp, the plastic moment in the ideal system. When Mp is reached, the section at mid-
span begins to deform (rotate) into the plastic range without further increase in the applied
moment (load). In other words, after reaching Mp, the cross section at midspan behaves as a
real hinge (with no further moment resistance) with plastic deformations, thus it is referred
to as a plastic hinge. For such simply supported beams, development of a plastic hinge at
midspan causes the beam to become a mechanism with no further load-carrying capacity.
Therefore, the load that causes Mp to develop at one location along the span represents the
maximum load that can be applied to the beam: that is, the limit of the system for such an
ideal section behavior. It will be shown, however, that for statically indeterminate systems,
31
9.12 REDISTRIBUTION OF MOMENTS 313
Plastic Mn
My
Moment M
Moment M
Elastic
Mp = plastic Mn
moment
ϕy ϕu ϕy ϕu
d 2y d 2y
(
Curvature ϕ i.e.,
dx2
) (
Curvature ϕ i.e.,
dx2
)
(a) Ideal elastic-plastic (b) Reinforced concrete
Figure 9.12.1 Moment curvature relationships.
Wu
θup
Plastic hinge
Figure 9.12.2 Limit condition for simply supported beam.
the development of a plastic hinge at one location does not necessarily correspond to the
maximum load that can be carried by the structure.
Though typically shown as a single point, the plastic hinge is actually a region of con-
centrated inelastic deformations (curvatures) that spread out from the point of maximum
moment a certain length, often called the plastic hinge length, lp. This length depends on the
details of the longitudinal and transverse reinforcement, the shear span, among others, but it
has been found to vary in practice between 0.5d and 2.5d on each side of the critical section
[9.4, 9.8, 9.12]. For practical purposes, the beam may be treated as two rigid elements con-
nected by a plastic hinge having a maximum concentrated rotation angle, or plastic hinge
rotation, θup, at ultimate ( just before collapse) as shown in Fig. 9.12.2. This plastic hinge
rotation may be approximately computed by assuming that the plastic curvature is constant
over the length lp on each side of the point of maximum moment, in which case
θup = (ϕu − ϕ y )2l p (9.12.1)
This maximum rotation that can be developed by the plastic hinge is often referred to
as plastic rotation capacity. Most reinforced concrete beams have a reasonable amount of
plastic rotation capacity since they are usually designed with a net reinforcement ratio ρ or
ρ –ρ′ of 50% or less of the balanced amount ρb.
Consider the statically indeterminate, fixed-end beam of Fig. 9.12.3(a) subjected to a
uniformly distributed load, along with its corresponding moment diagram computed from
an elastic analysis. As the load is applied to the beam, the sections with the largest moments
(i.e., at the fixed ends) achieve Mp first. Thus, at the ends [see Fig. 9.12.3(b)],
wy L2 12 M p
= M p and wy =
12 L2
where wy is the load carried when Mp is just reached at an end, accompanied by a curvature
ϕy [see Fig. 9.12.1(a)]. Upon the slightest increase in the applied load, the ends of the beam
314
would behave as plastic hinges and would rotate but would carry no additional moment.
The moments at all other sections along the beam span, however, remain within the elas-
tic range, below Mp. In particular, the section at midspan has a moment of only Mp /2.
Therefore, as the load is increased above wy, the beam will behave as a simply supported
beam with plastic hinges at the ends. In carrying this load, the beam is as stable as a simple
beam and is capable of carrying load above wy. Increasing the load will cause the moments
to increase everywhere along the beam span except at the ends, with maximum increase
in moment occurring at midspan. It can be shown that the limit condition for this beam
requires three plastic hinges (two at the ends and one at midspan), as shown in Fig. 9.12.3.
Also shown in the figure is the corresponding moment diagram at the limit (collapse) con-
dition. Statics may be used in computing the collapse load, for example, by drawing a
free-body diagram of the beam between the end and the plastic hinge at midspan and taking
moments about the beam ends:
wu L L
Mp + Mp − = 0
2 4
wu L2
2M p =
8
16 M p
wu =
L2
Thus the beam may carry 33% additional load after the plastic moment has been reached
at the fixed ends, provided the rotation of the plastic hinges at the ends not exceed the
rotation capacity (θup) prior to developing the plastic hinge at midspan (i.e., prior to reach-
ing the load wu). In other words, rotation capacity permitting, the positive and negative
wL2
24
+
wL2 – – wL2
12 12
(a)
wy
Mp Mp
(b)
wu
Mp Mp
Mp
Plastic hinges
Mp
Mp Mp
(c)
Figure 9.12.3 Limit condition for a fixed-end beam.

315
9.12 REDISTRIBUTION OF MOMENTS 315
Ultimate concrete
εc strain, εcu = 0.003
ϕu
ϕy
Strain
occurring during Steel strain at first yield, εy
redistribution Total steel strain
of moment
Figure 9.12.4 Strain diagram for reinforced concrete beams in which steel reaches εy before
maximum concrete strain εc = εcu.
moments under any particular loading condition tend to equalize (assuming, of course, that
the strength of the section at the two regions is the same). For a thorough treatment of limit
analysis the reader is referred to the work of Baker [9.8].
Reinforced concrete beam sections with a low net percentage of reinforcement ρ – ρ′
(i.e., tension-controlled sections) will develop yielding of the tension steel reinforcement
while the maximum concrete strain is still of low magnitude (see Fig. 9.12.4). Therefore,
sufficient reserve curvature φu – φy, or plastic hinge rotation capacity, θup, will be available,
allowing redistribution of moments to occur before the ultimate concrete strain is reached.
Redistribution of Negative Moments under ACI-6.6.5

As noted earlier, redistribution of moments can occur only when there is enough reserve cur-
vature after first yield of the tension steel. For this condition to exist, the tension steel must
develop strains well above the yield strain before the concrete is crushed in the compression
zone (see Fig. 9.12.4). Accordingly, redistribution of negative moments is permitted only
when εt is equal to or greater than 0.0075 at the section at which moment is reduced [ACI-
6.6.5.1(b)]. Also, there is a limit to the amount of moment redistribution that can be allowed,
irrespective of the net tensile strain at the cross section. ACI-6.6.5.3 states that the percent-
age permitted for redistribution is 1000εt percent, with a maximum of 20%. The strain εt is
the net tensile strain in the extreme tension steel at nominal strength. An easy way to remem-
ber the idea is to recall that a 10% redistribution limit corresponds to a value εt = 0.01.
The limits of applicability of this provision may be summarized as follows.
1. Application is limited to continuous flexural members (i.e., statically indeterminate

systems), ACI-6.6.51(a).
2. No more than 20% reduction of moments for any given loading arrangement may be
applied [ACI-6.6.5.3].
3. Bending moments used in such an adjustment must be obtained by an elastic analysis
(ACI-6.6.5.1); moments from use of coefficients or other approximate methods may
not be adjusted (ACI-6.5.3).
4. Adjustment, when permitted, is made for each given loading condition. The envelope
of the adjusted diagrams from all loading conditions is then used to proportion the
members (ACI-6.6.5.4 and ACI-6.6.5.5).
In approaching a design, it may be advantageous to apply ACI-6.6.5 directly. It is likely,

however, that more frequent use will occur when the design is in progress and the designer
realizes that the conditions of the redistribution provision are met and savings appear to be
possible. The following example illustrates an application of this provision.
316
EXAMPLE 9.12.1
Show the effects of redistribution on the moments obtained by elastic analysis for the
floor beams 2B1 and 2B2 in Example 9.3.1. Use the provisions of ACI-6.6.5.
SOLUTION
The beams, along with the critical moment diagrams, are shown in Fig. 9.12.5. The
elastic end moments Mu are those computed under factored loads and tabulated in
Table 9.3.1 for each loading condition. The elastic moment diagrams are shown in Figs.
9.3.3 through 9.3.6.
(a) Investigate negative moment region at B to determine the maximum percent of
moment that can be redistributed. The net tensile strain at B (sections C-C and D-D)
A B C
26’–0’ 26’–0’
3.45 k/ft (factored load) 1.37 k/ft (factored load)
Spans 1 and 3 with LL

159 ft-kips
Symmetrical
144 about CL of
structure
From elastic
analysis
2
81 14
90
190 118 118
210 130 130
(a)
3.45 k/ft
148 Spans 1 and 2 with LL

105
142
104
From elastic From elastic

66
analysis analysis
73
152
224 205 168
248 226
(b)
1.37 k/ft 3.45 k/ft
Span 2 with LL 121
46 103
From elastic
40 analysis
14
15
144 171 171

159 189 189
(c)
Figure 9.12.5 Redistribution of elastically computed moments for Example 9.12.1 according to

ACI-6.6.5.
(Continued)
317
was computed earlier in part (a) of Example 9.4.1 as 0.0105, which is greater than
0.0075. Thus, moment redistribution is permissible at B.
percent adjustment permitted = 1000 εt = 1000(0.0105) = 10.5%
Since the steel in the compression zone was not fully developed at the faces of the support
(see Fig. 9.6.2), it may not be counted as compression steel. It might well be economical
to develop the capacity of the steel in the compression zone at the support. This would
increase the ductility in that region and allow a higher percentage of moment redistribution.
(b) Make adjustments to elastic moments. The adjustment of the negative moments may
be either an increase or a decrease as long as the positive moments are also adjusted
to satisfy static equilibrium. Examine first the loading for maximum positive moment
in span AB, Fig. 9.12.5(a). Increasing the negative moments by 10.5% reduces the
maximum positive moment for this loading from 159 to 144 ft-kips. The 10.5%
adjustment for Case (b) is made by reducing the negative moment at B and increasing
those at A and C, thus minimizing the effect on the positive moments. In Case (c), the
negative moments are increased 10.5%, thus reducing the positive moment in span 2
from 121 to 103 ft-kips. Note that, after these adjustments, the increased negative
moment at B for Case (a) is still less than the controlling negative moment of 224 ft-kips
occurring under Case (b). The increased positive moment in span 1 for Case (b) is only
3% greater than the reduced controlling positive moment of 144 ft-kips from Case (a).
Similarly, the increased positive moment in span 2 under Case (b) is only 2% greater
than the reduced positive moment of 103 ft-kips under Case (c).
The envelope of adjusted moments would then be used to design the sections. The net
effect on the envelope is a reduction for both positive and negative moments. This is not
actually a reduction in the safety factor below that used for a simply supported beam; it
is a reduction in the excess strength that the system has, by virtue of its continuity, one
span with another.
A simple application of limit design concepts to practical design has been presented
by Furlong [9.10]. Further discussion of Furlong’s proposal has been presented [9.11] as
a proposed alternate to the use of ACI-6.6.5.
It must be noted, however, that such factors as shear, development of reinforcement,
and deflection may still control the design; thus, limit analysis for flexure may offer lit-
tle, if any, economic advantage.
SELECTED REFERENCES
9.1. CRSI. Manual of Standard Practice (27th ed.). Schaumburg, IL: Concrete Reinforcing Steel
Institute, 2001.
9.2. Paul F. Rice, Edward S. Hoffman, David P. Gustafson, and Albert G. Gouwens. Structural
Design Guide to the ACI Building Code (3rd ed.). New York: Van Nostrand Reinhold, 1985.
9.3. A. H. Mattock. “Limit Design for Structural Concrete,” Journal of the Research and Development
Laboratories, Portland Cement Association, 1, May 1959, 14–24.
9.4. W. G. Corley. “Rotational Capacity of Reinforced Concrete Beams,” Journal of the Structural
Division, ASCE, 92, ST5 (October 1966), 121–146. (Also PCA Development Department
Bulletin D108.)
9.5. ACI-ASCE Committee 428. “Progress Report on Code Clauses for ‘Limit Design,’” ACI
Journal, Proceedings, 65, September 1968, 713–720. Disc. 66, 221–223.
9.6. M. Z. Cohn. “Limit Design for Reinforced Concrete Structures: An Annotated Bibliography,”
ACI Bibliography No. 8. Detroit: American Concrete Institute, 1970.
9.7. Harold W. Conner, Paul H. Kaar, and W. Gene Corley. “Moment Redistribution in Precast
Concrete Frame,” Journal of the Structural Division, ASCE, 96, ST3 (March 1970), 637–661.
9.8. A. L. L. Baker. Limit State Design of Reinforced Concrete. London: Cement and Concrete
Association, 1971.
318
9.9. M. Z. Cohn. “Inelasticity of Reinforced Concrete and Structural Standards,” Journal of the
Structural Division, ASCE, 105, ST11 (November 1979), 2221–2241.
9.10. Richard W. Furlong. “Design of Concrete Frames by Assigned Limit Moments,” ACI Journal,
Proceedings, 67, April 1970, 341–353.
9.11. Richard W. Furlong and Carlos Rezende. “Alternate to ACI Analysis Coefficients,” Journal of
the Structural Division, ASCE, 105, ST11 (November 1979), 2203–2220.
9.12. Robert Park and Thomas Paulay. Reinforced Concrete Structures. New York: John Wiley &
Sons, 1975, 769 pp.
PROBLEMS
All problems are to be worked in accordance with the provisions of ACI Code, and all
stated loads are service loads, unless otherwise indicated. A design sketch showing all
design decisions is required for each problem. For verifying bar lengths, draw the φ Mn dia-
gram directly below a side view of the beam showing the bars, as in Fig. 9.6.2.
CONTINUOUS BEAM PROBLEMS

9.1 Design a rectangular beam continuous over three (c) Use only full-span loadings for computing
spans as shown in the figure for Problem 9.1. the shear envelope and use U stirrups of #3
The live load is 2.75 kips/ft, and the dead load size if possible.
is 1.0 kip/ft in addition to the beam weight. The 9.2 Redesign the beam of Problem 9.1 as a mon-
floor is to be a prefabricated system. Assume the olithic T-section floor system. The beams are
supports to be 15 in. wide. Use fc′ = 4000 psi spaced 8 ft on centers and the slab is 6 in. thick.
and fy = fyt = 60,000 psi, and do not apply ACI- The 1.0-kip/ft dead load includes the slab but
6.6.5 for moment redistribution. not the beam stem.
(a) Determine the moment envelope using fac- 9.3 Design the beam ABC of the frame shown in the
tored loads. figure for Problem 9.3 in which the relative stiff-
(b) Determine the bar cutoff locations, directly nesses EI/L are given by the encircled numbers.
from the moment envelope. The beams are T-sections having a 6-in. slab.
The dead load is 0.40 kip/ft (not including beam
stem or slab) and the live load is 3.75 kips/ft.
Rectangular beam
Assume the columns to be 15 in. square. Use
fc′ = 4000 psi and fy = fyt = 60,000 psi.
9.4 Design the transverse beam indicated for the
15’– 0” 21’– 0” 15’– 0”
floor plan given in the figure for Problem 9.4.
Problem 9.1 Assume that a warehouse live load of 375 psf
is to be used. Assume also a 5-in. slab placed
monolithically with beams and girders; the
14’–0” 2 2 2
A 8 B 6 C
14’–0” Transverse
2 2 spacing of 2
frames = 12 ft
30’–0” 40’–0”
Problem 9.3
319
PROBLEMS 319
width of support at longitudinal girders is 18 in., 9.6 Redesign the transverse beams of Problem 9.4,
and only a nominal minimum of moment res- except use spans 24-21-24 instead of original
traint is provided by the exterior wall support spans 21-18-21, and use fc′ = 4000 psi and fy =
(i.e., assume hinge for elastic analysis). Use 60,000 psi. All other details are the same as in
fc′ = 3500 psi and fy = fyt = 60,000 psi. For a Problem 9.4.
comparison of the effect of considering torsional 9.7 Design the four-span longitudinal girder indi-
stiffness of longitudinal girders, divide the class cated for the floor system of Problem 9.4. In lieu
into three parts, each using one of the following of the more accurate loadings from the results
assumptions: of Problem 9.4, use concentrated dead and live
(a) zero torsional stiffness of the two longitudi- loads of 15 and 60 kips, respectively, coming
nal girders to the girder from each side, plus the weight
(b) torsional stiffness equal to 25% of the bend- of the girder. Assume that the columns are
ing stiffness of the 21-ft span beam 18 in. square and 15 ft high and that the beam
(c) torsional stiffness equal to 50% of the bend- receives equivalent restraint from monolithic
ing stiffness of the 21-ft span beam attachment to the 15- in. reinforced concrete
exterior wall. Assume also that the columns are
Suggested additional problems (no solutions manual) fixed at the far ends. Use fc′ = 3500 psi and fy =
9.5 Redesign the transverse beams of Problem 9.4, 60,000 psi.
except use spans 22-20-22 instead of the origi- 9.8 Design the continuous 26-ft-span floor beams
nal spans 21-18-21, and use fc′ = 4000 psi and along a column line in Fig. 8.3.1. Assume that the
fy = fyt = 60,000 psi. All other details are the same floor slab is 5 in. thick. Assume the columns are
as in Problem 9.4. 15 in. square and 10 ft long. Use fc′ = 4000 psi,
21’–10”
21’–0”
Transverse beam
61’–8”
18’–0”
18’–0”
Longitudinal
girder 21’–10”
21’–0”
7’–0” 7’–0” 7’–0”
21’–10” 21’–0” 21’–0” 21’–10”
85’–8”
10”
CL of bearing
Transverse beam
15”
21’–10”
Section at wall
Problem 9.4
320
fy = fyt = 60,000 psi, and a 100 psf live load. 2 1 2 -in. slab, 20-in.-wide × 8-in.-deep forms,
Make a preliminary design without structural and 4-in.-wide joists with 2–#5 bars. No taper
analysis; then, using the preliminary size, make is used. Use fc′ = 3000 psi and fy = 60,000 psi.
the structural analysis to obtain the factored 9.11 Design an end-span joist for a continuous sys-
moment and factored shear envelopes for the tem to carry a live load of 225 psf, using 20-in.-
final design. wide removable pans, for a clear span of 18 ft.
Allow an extra 1 2 in. of thickness for dead load
CONCRETE JOIST PROBLEMS purposes only, since the concrete slab is to serve
as the final wearing surface. Use fc′ = 4500 psi
9.9 Design a concrete joist, using 30- in.-
wide and fy = 60,000 psi.
removable pans, for a typical interior span of
28 ft center to center of supporting girders.
Assume a support width of 18 in. Use a live load MOMENT REDISTRIBUTION PROBLEM
of 100 psf, fc′ = 4000 psi, and fy = 60,000 psi. 9.12 Redesign the beam of Problem 9.1, taking into
9.10 Determine the service live load capacity for account permissible moment redistribution.
a single-span joist of 20-ft clear span, using a Compare with work on Problem 9.1.
CHAPTER 10
MEMBERS IN COMPRESSION
AND BENDING
10.1 INTRODUCTION
Columns subjected to pure axial load rarely, if ever, exist. All columns are subjected to
some bending moment, which may be caused by (1) end restraint arising from the mon-
olithic placement of floor beams and columns; (2) accidental eccentricity from imperfect
alignment and variable materials; (3) asymmetrical floor loads; (4) eccentric loads such as
crane loads in industrial buildings, and (5) lateral loading such as from wind or induced by
an earthquake.
Concrete construction is usually monolithic; thus reinforced concrete frames and arches
(Fig. 10.1.1) are common and advantageous. All sections in the two structures shown in
Fig. 10.1.1 are subjected to combined bending and axial load. The vertical members in
Fig. 10.1.1(a) and sections near the supports in Fig. 10.1.1(b) may be subjected to a high
Reinforced concrete tied columns under construction (Courtesy of Cary Kopczynski & Co.).
32
322 C H A P T E R 1 0 M E M B E R S I N C O M P R E S S I O N A N D B E N D I N G
(a) (b)
Figure 10.1.1 Reinforced concrete rigid frame and arch.
ratio of axial force to bending moment, while the horizontal member in Fig. 10.1.1(a) and
sections near the crown in Fig. 10.1.1(b) may be subjected to a low ratio of axial force to
bending moment.
This chapter deals first with columns having minimal bending moment, commonly
called axially loaded members. The effect of medium and large amounts of bending in
columns is considered later in this chapter. Only the basic strength of short compress-
ion members is considered, however. Slenderness effects in columns are treated in
Chapter 13.
10.2 TYPES OF COLUMNS
Reinforced concrete columns are principally of two types, classified according to the man-
ner in which the longitudinal reinforcing bars are laterally supported. A tied column, usu-
ally of square or rectangular shape, is one in which the longitudinal reinforcing bars are
held in position by separate lateral ties, as shown in Fig. 10.2.1(a) and photo on page 321.
A spirally reinforced column, usually of circular or square shape, is one in which the lon-
gitudinal reinforcing bars are arranged in a circle and wrapped by a continuous closely
spaced spiral, as shown in Fig. 10.2.1(b). For a column to be treated as a spirally reinforced
column, however, the spiral reinforcement must satisfy the requirements of ACI-25.7.3
(Section 10.9). Otherwise, the column is treated as a tied column.
A composite column is one that uses a structural steel shape, pipe, or tubing, with or
without additional longitudinal bars. One common composite column arrangement may
contain a structural steel shape completely encased in concrete, which is further reinforced
with both longitudinal and lateral reinforcement (spiral or ties), as shown in Fig. 10.2.1(c).
In a second kind of composite column, the steel may encase a concrete core, which may
or may not contain longitudinal reinforcing bars, as shown in Fig. 10.2.1(d). Composite
columns are treated in Chapter 21.
10.3 BEHAVIOR OF COLUMNS UNDER PURE

AXIAL LOAD
Many tests were made in the early 1900s on reinforced concrete columns under axial
loads [10.1–10.5], but loading was generally of short duration. As early as 1911, however,
Withey [10.5], at the University of Wisconsin, observed that as load increased beyond
the service load range, a transfer of load from concrete to steel took place. In the early
1930s, ACI Committee 105 reported [10.6] on the results of 564 column tests, primarily
at Lehigh University and the University of Illinois, where attention was given to column
size, quality of concrete, quality and amount of longitudinal and lateral reinforcement,
rate of application of load, and shrinkage and creep under sustained loads. By 1940, code
design procedures for axially loaded columns were based on the ultimate strength results
of the aforementioned extensive investigation. Joint ACI-ASCE Committee 441 has pro-
vided an excellent annotated bibliography of the early studies on reinforced concrete
columns [10.7].
32
1 0 . 3 B E H AV I O R O F C O L U M N S U N D E R P U R E A X I A L L OA D 323
Ties Spiral Steel tubing
Concrete filled
Spiral
Pipe
(a) Tied column (b) Spirally (c) Composite column (d) Composite
reinforced (spiral bound column
column encasement around (steel encased
structural concrete core)
steel core)
Figure 10.2.1 Types of columns.
When concrete and steel act together in compression, the proportion of loading carried
by each changes continuously during the time the load is acting. Initially, the stress in the
steel is Es /Ec times the stress in the concrete, according to the elastic theory. As the time-
dependent effects of creep and shrinkage occur, the steel gradually carries relatively more
load than its elastic share. Creep and shrinkage deformations were introduced in Section
1.11, and are treated in detail with regard to deflections in Chapter 12.
Members that are subjected to axial compression, either alone or in combination with
bending, frequently have a substantial portion of the total load sustained. Consequently, the
transfer of load to the steel from the concrete due to time-dependent deformation is more
pronounced in these members than in beams.
The axial load versus axial deformation behavior of spirally reinforced columns is
significantly different from that of tied columns. Spirally reinforced columns exhibit a
marked “yielding,” followed by considerable axial deformation before failure, as shown
in Fig. 10.3.1. On reaching the first peak point, the shell spalls off (Fig. 10.3.2) and the
spiral begins to restrain lateral expansion, confining the concrete in the core. As discussed
in Section 1.9, concrete compressive strength increases with lateral confinement. Thus, as
the column core expands and the spiral reinforcement provides confinement, the compres-
sive strength of the column core increases. If spiral reinforcement is provided in sufficient
amount, the peak load resisted by the confined column core will be greater than the peak
load prior to spalling of the column shell. The first peak has sometimes been referred to
as the “yield” strength of a spirally reinforced column, while the second peak represents
the ultimate strength. Typical appearance of spirally reinforced columns under concentric
loading at failure is shown in Figure 10.3.2.
Tied columns, particularly those with rectilinear ties, do not exhibit the deformation
capacity of spirally reinforced columns. As shown in Fig. 10.3.1, the axial load versus axial
deformation response shows only one peak, which corresponds to the point at which the
shell begins to spall off and buckling of longitudinal reinforcement between the ties initi-
ates (Fig. 10.3.3). Thus, there is no “yield” point in tied columns; the first peak represents
the ultimate strength.
324
Based on the discussion above, the ultimate strength of a spirally reinforced column
(i.e., after spalling of the column shell) may be expressed as follows [10.6]:
Pult = k3 fc′( Ach − Ast ) + Ast f y + ks ρs f yt Ach (10.3.1)
where
k3 = coefficient (taken equal to 0.85) to account for the difference between concrete
strength in the column and that in a test cylinder
fc′ = standard 28-day cylinder strength
Ach = area of concrete core, taken to the outside edges of the spiral reinforcement (see
later: Fig. 10.9.1)
Ast = total area of longitudinal reinforcement
fy = yield stress of longitudinal reinforcement
ρs = ratio of volume of spiral reinforcement to volume of concrete core calculated as
(see Fig. 10.9.1)
4 Asp
ρs = (10.3.2)
Dc s
ks = constant that varies typically from 1.5 to 2.5

Asp = cross-sectional area of spiral reinforcement (see Fig. 10.9.1)
Dc = diameter of concrete core (see Fig. 10.9.1)
s = spiral spacing or pitch (see Fig. 10.9.1)
fyt = yield stress of spiral steel
Note that the third term in Eq. (10.3.1) represents the contribution from confinement of the
concrete core by the spiral reinforcement after spalling of the concrete shell.
The “yield ” strength of a spirally reinforced column (first peak in Fig. 10.3.1) and ulti-
mate strength of a tied column, on the other hand, can be expressed as
Pult = k3 fc′( Ag − Ast ) + Ast f y (10.3.3)
where Ag is the gross cross-sectional area of the column.

Studies on ductility of spirally reinforced column have been made by Priestley, Park, and
Potangaroa [10.8], Ahmad and Shah [10.9], and Martinez, Nilson, and Slate [10.10]. Studies on
the behavior of tied columns by Sheikh and Uzumeri [10.11], Scott, Park, and Priestley [10.12],
Park, Priestley, and Gill [10.13], Fafitis and Shah [10.14], Moehle and Cavanagh [10.15], Sheikh
and Yeh [10.16, 10.20, 10.21], Özcebe and Saatcioglu [10.17], Yong, Nour, and Nawy [10.18],
and Abdel-Fattah and Ahmad [10.19] show how the deformation capacity of tied columns var-
ies with tie and longitudinal bar arrangements. Razvi and Saatcioglu [10.22] have studied the
effects of confinement using welded wire reinforcement. Much of this work has been concerned
with ductility under cyclic loading, such as for earthquake-resistant design.
As will be shown later, though columns of both types have approximately the same
strength, a higher factor of safety should be provided for the tied column than for the
spirally reinforced column because of the sudden failure and lack of toughness (energy
absorption) exhibited by tied columns.
Spirally reinforced column

shows ability to deform
“Yield” point (spirally reinforced prior to failure
column shell spalls off)
Tied column fails suddenly

Load
Tied and spirally reinforced columns
Deformation (unit shortening)
Figure 10.3.1 Typical load-deformation curves for tied and spirally reinforced columns.
325
1 0 . 4 S A F E T Y P R OV I S I O N S F O R C O L U M N S 325
Figure 10.3.2 Failure of spirally reinforced columns under concentric axial force. (Courtesy of
Portland Cement Association.)
Figure 10.3.3 Failure of tied columns under concentric axial force. (Courtesy of Portland Cement
Association.)
10.4 SAFETY PROVISIONS FOR COLUMNS

The load factors used in the design of columns are the same as for any other type of mem-
ber. The design of columns in regions of little or no seismic risk is often governed by load
combinations that include wind, whereas beams are usually governed by the gravity-load
combinations.
As discussed in Section 2.7, for gravity loads,
U = 1.4 D [2.7.1]
U = 1.2 D + 1.6 L + 0.5( Lr or S or R) [2.7.2]
For loadings including wind,
U = 1.2 D + 1.6( Lr or S or R ) + (1.0 L or 0.5W ) [2.7.3]
U = 1.2 D + 1.0W + 1.0 L + 0.5( Lr or S or R ) [2.7.4]
U = 0.9 D + 1.0W [2.7.6]
The other load combinations are not repeated from Section 2.7.

The strength reduction factor φ for compression-controlled columns (ACI-21.2.2) is
0.75 for spirally reinforced columns and 0.65 for tied columns.
326
The strength reduction factor φ, which is lower for tied columns than for spirally rein-
forced columns, reflects the lower toughness exhibited by tied columns. The difference in
the behavior of tied and spirally reinforced axially loaded columns is further accounted for
by the use of a different maximum nominal compression capacity, Pn(max), as discussed in
Section 10.11.
For combined compression and bending, the φ value is permitted to be variable and may
increase to 0.90 for tension-controlled columns. This variation of φ is treated later (see
Section 10.17).
10.5 CONCENTRICALLY LOADED SHORT COLUMNS

According to the ACI Code, the strength for a concentrically loaded short column, Po, can
be calculated based on Eq. (10.3.3) as follows (ACI-22.4.2.2):
Po = 0.85 fc′( Ag − Ast ) + f y Ast (10.5.1)
where Ag is the gross column area and Ast is the total area of longitudinal reinforcement. For
the rectangular column shown in Fig. 10.5.1, Ag = bh and Ast = (A1 + A2). As discussed in
Section 10.3, the 0.85 factor in Eq. (10.5.1) accounts for the difference between the com-
pressive strength of concrete in a column and that of a concrete cylinder. This factor, thus,
is not the same factor used in Whitney’s stress block. Po may also be expressed as
Po = Ag 0.85 fc′(1 − ρg ) + f y ρg  (10.5.2)
where ρg = Ast /Ag.
When the terms including ρg are combined, Eq. (10.5.2) becomes
Po = Ag 0.85 fc′ + ρg ( f y − 0.85 fc′ ) (10.5.3)
Figure 10.5.1 Internal force resultants in a concentrically loaded column.
10.6 STRENGTH INTERACTION DIAGRAM

When axial compression combined with bending moment acts on a member having low
slenderness ratio (unbraced length Lu to radius of gyration r) such that column buckling
is not a potential mode of failure, the strength of a member is governed by the material
strength (corresponding to yielding in a homogeneous elastic material) of the cross section.
ACI-22.2.2.1 states that the nominal strength of this so-called short column subjected to
combined axial load and bending is achieved when the extreme concrete compression fiber
reaches the strain εcu = 0.003.
327
1 0 . 6 S T R E N G T H I N T E R AC T I O N D I AG R A M 327
For a given cross s ection there is an infinite number of strength combinations at which
Pn and Mn act together. These strength combinations lie on a curve, as shown in Fig. 10.6.1,
which is called the strength interaction diagram or P–M interaction diagram. Depending
on the ratio of Mn to Pn (see Fig. 10.6.1), the strain diagram will exhibit two distinct cat-
egories. There may be (1) compression over most or all of the section such that the com-
pressive strain in the concrete reaches 0.003 before the tension steel reaches the yield strain
εy = fy /Es or (2) tension in a large portion of the section such that the strain εs in the tension
steel exceeds the yield strain εy when the compressive strain in the concrete reaches 0.003.
The balanced strain condition in combined bending and axial load is represented by the
point Pn = Pb and Mn = Mb on this diagram.
The balanced strain condition corresponds to the “compression control” limit (see
Section 3.6 for a discussion on compression- and tension-controlled sections). Thus, mem-
ber sections with a resultant axial force higher than that corresponding to the compression
control limit are said to be compression controlled. In this case, the strength reduction fac-
tor φ is 0.75 for spirally reinforced columns and 0.65 for tied columns (ACI-21.2.2).
The region below the compression control limit is divided into two parts. A section
having a combination of Pn and Mn such that the strain in the extreme layer of tension rein-
forcement ε t ≥ 0.005 , is said to be tension controlled; the corresponding strength reduction
factor φ is 0.90 (ACI-21.2.2). This would be the case of a member subjected to pure flexure
(i.e., no axial load applied), or to flexure and low levels of axial load. Sections such that
f y / Es < ε t < 0.005 are said to be in the transition zone of the P–M diagram for which the
strength reduction factor φ varies linearly with εt between 0.90 and 0.75 for spirally rein-
forced columns and between 0.90 and 0.65 for tied columns.
This maximum strength interaction relationship for short rectangular members has
been verified by research [10.23–10.28]. Nonprismatic members have also been studied
[10.29, 10.30], as well as members subject to nonproportionally varying axial load [10.31].
The major emphasis in this chapter is the analysis and design of sections whose nominal
strength (Pn and Mn) lies at various points on the interaction diagram (Fig. 10.6.1).
In studying Fig. 10.6.1, the reader may note that radial lines from the origin (Pn = 0,
Mn = 0) represent constant ratios of Mn to Pn; that is, they represent eccentricities e of the
load Pn from the plastic centroid, which was defined in Fig. 10.5.1. Under usual conditions
of symmetrical reinforcement, the plastic centroid coincides with the centroid of the gross
section. That e is equal to M/P may be observed from Fig. 10.6.2, where it is shown that a
column subject to an eccentric load is statically equivalent to a member under the combined
action of an axial load and a bending moment. Thus, in Fig. 10.6.1 the vertical axis repre-
sents e = 0 and the horizontal axis represents e = ∞. This concept of replacing axial load and
bending moment by a single eccentric load provides the basis for the practical approach to
analysis and design computations in reinforced concrete.
Pn - axis Axis of bending
Po Mn
e=
Pn
εt < εy 0.003
e=0
Compression-controlled section
Balanced strain
0.003
condition fy
Pb
Transition zone εt = εy = E
s
e = eb
Tension-controlled section
0.003
e=∞
Mb εt = 0.005
M0 Mn - axis
Mn, bending moment
Figure 10.6.1 Typical Pn–Mn interaction diagram for axial compression and bending moment
about one axis. Transition zone is where εy < εt < 0.005.
328
M
e P e P
P P
P
A A
= =
Plastic
centroid Eccentrically
loaded
Axis load and member
bending moment
M P
e= X
P
e
Section A–A
Figure 10.6.2 Eccentrically loaded column statically equivalent to member subject to axial load and
bending moment.
10.7 SLENDERNESS EFFECTS
The effect of slenderness ratio (ratio of unbraced length Lu to radius of gyration r) must be
taken into account in the design of compression members. Design moments may increase
significantly as a result of increased eccentricity of the axial compressive force due to lat-
eral deflections in compression members and, in some cases, buckling may govern the axial
compressive strength of the member (see Fig. 13.2.2 for curves typical of reinforced con-
crete sections). In addition, slender members will deflect more under any primary bending
moment, thus having a larger secondary moment, which is the product of the axial com-
pression and lateral deflection (as shown in Fig. 13.1.1).
The magnitude of the slenderness ratio determines whether the increase in moments
due to lateral deflection, or stability in general, is sufficiently important that it cannot
be neglected in design. The reference condition for slenderness ratio is that of a column
with hinged ends (i.e., no resistance to rotation at either end), as shown in Fig. 13.3.1(a).
Equivalent pin-end lengths of columns with end restraints can be expressed by kLu, where
k is the effective length factor and Lu is the actual unsupported length.
A vital factor in the determination of the equivalent pin-end length is whether the struc-
tural system is nonsway (braced) so that relative movement of the ends transverse to the axis
of the member is prevented [see Figs. 13.3.1, 13.3.3(a) and 13.3.3(c)], or sway (unbraced),
where such relative movement is possible and restraint is provided only by the rigidity of
the joints and the stiffness of interacting beams and columns [see Figs. 13.3.2, 13.3.3(b)
and 13.3.3(d)]. Without general derivation or proof, the following may be stated:
For nonsway systems, k ≤ 1.0

For sway systems, k ≥ 1.0
A qualitative explanation is available from the study of Figs. 13.3.1 to 13.3.3, but a theo-
retical development would require a study of structural stability (see Chapter 13, [13.26]
and [13.27].
329
1 0 . 8 L AT E R A L T I E S 329
ACI-6.2.5 permits neglect of slenderness effects for columns braced against sidesway
(nonsway) when
kLu M 
≤ 34 + 12  1  (10.7.1)
r  M2 
where [34 + 12M1/M2] may not exceed 40. M1 and M2 are numerically the smaller and larger
first order bending moments, respectively, at the ends of the member; the ratio M1/M2 is
negative for single curvature and positive for double curvature. The intent of the ACI Code
is to permit the design of compression members in nonsway systems as short columns,
without considering slenderness effects, when consideration of slenderness effects would
result in up to a 5% increase in moments.
For unbraced or sway systems, ACI-6.2.5 permits neglect of slenderness effects when
kLu
≤ 22 (10.7.2)
r
At the time these limits were established, a study of existing structures indicated [10.32]
that over 90% of the columns in braced frames and over 40% of the columns in unbraced
frames would fall within the limits of ACI-6.2.5, and thus moment increase due to slender-
ness effects could be ignored. Since, however, columns exceeding the ratios indicated by
Eqs. (10.7.1) and (10.7.2) have become relatively more common, a higher percentage of
designs today will likely require consideration of slenderness effects.
The remainder of this chapter treats the design of short compression members exclu-
sively. Slenderness effects in both nonsway and sway systems are considered in detail sep-
arately, in Chapter 13.
10.8 LATERAL TIES
The effect of ties on the behavior of columns is complex. When a tied column is axially
loaded to failure, the first occurrence is the spalling off of the exterior cover, which in turn
causes transfer of load to the concrete core and the longitudinal steel. Lateral ties provide
lateral support so that any tendency to buckling of individual bars would occur only between
the tie supports. The loss of stiffness of the longitudinal steel, which begins to yield or to
buckle outward, causes additional load on the concrete core. Once the core has achieved its
crushing strength, the column, unless heavily confined, fails suddenly. The above sequence
usually takes place rapidly, producing “sudden” failure. A proper cross-sectional arrange-
ment of ties placed at a sufficiently close spacing, however, may provide large core confine-
ment and an increase in the strain at which the concrete core crushes, leading to increased
column ductility and better warning against failure [10.12, 10.40]. Studies [10.11, 10.33–
10.37] have indicated that present tie requirements are conservative for ordinary columns
with Grade 40 reinforcement but may not be conservative for columns having high-strength
reinforcement, large or bundled bars, or unusual dimensions. An overall review of confine-
ment in columns is provided by Sakai and Sheikh [10.38], and the subject of tie requirements
for prestressed concrete columns is provided by Halvorsen and Carinci [10.39].
ACI-10.7.6.1.2 requires that ties used in columns be detailed in accordance with ACI-
25.7.2, which specifies the following.
1. All nonprestressed bars for tied columns shall be enclosed by lateral ties, at least #3
in size for longitudinal bars #10 or smaller, and at least #4 in size for #11 or larger
and bundled longitudinal bars.
2. The center-to-center spacing of the ties shall not exceed 16 longitudinal bar diam-
eters, 48 tie bar diameters, or the least dimension of the column. Further, the clear
spacing between ties must be at least 4 3 of the maximum aggregate size.
30
4 bar Alternate position of hooks

in successive sets of ties
(for all bar arrangements)
6 bar
≤ 6” > 6”
8 bar
≤ 6” > 6”
12 bar
6” max
6” max
6” max
16 bar
Figure 10.8.1 Common column tie arrangements. (Adapted from Ref. 2.23.)
3. Rectilinear ties shall be so arranged that every corner and alternate longitudinal bar
shall have lateral support provided by the corner of a tie having an included angle
of not more than 135° and no bar shall be farther than 6 in. clear on either side from
such a laterally supported bar.
4. Where the bars are located around the periphery of a circle, a complete circular tie
may be used.
The ACI detailing manual [2.23] suggests tie arrangements for various numbers of bars,
some common ones being shown in Fig. 10.8.1.
In regions where longitudinal bars are spliced or offset bent, ACI Commentary
R10.7.6.1.5 says that “it is prudent to provide a set of ties at each end of lap spliced bars,
above and below end-bearing splices, and at minimum spacings immediately below sloping
regions of offset bent bars.”
In addition, ACI-10.7.6.1.5 waives the lateral reinforcement requirements where “tests
and structural analyses demonstrate adequate strength and feasibility of construction.”
10.9 SPIRAL REINFORCEMENT AND LONGITUDINAL

BAR PLACEMENT
The spiral provides the column with the ability to absorb considerable deformation prior
to failure [10.8]. This toughness is the principal gain that is achieved by the use of spirally
31
10.9 SPIRAL REINFORCEMENT 331
reinforced columns. Today’s knowledge of spiral behavior is based on the column research
of the early 1930s [10.6]. Although the spiral does actually contribute strength to the
column through confinement (as early as 1903, Considère [10.1, 10.2] indicated that the
spiral was 2.4 times as effective as longitudinal reinforcement in providing column capac-
ity), the conservative policy of ACI specifications since about 1940 has been to provide
spiral reinforcement sufficient to increase the capacity of the core by an amount equal to
the capacity of the shell, thus maintaining the column axial load capacity when the shell
spalls off.
Using the third term of Eq. (10.3.1) with an average ks of 2, the axial strength P contrib-
uted by the spiral reinforcement is
P = 2ρs f yt Ach (10.9.1)
As discussed in Section 10.3, this strength increase due to the spiral reinforcement was
proposed by ACI Committee 105. Information on the derivation of this term, based on
compressive strength exhibited by concrete under triaxial compressive stress, has also been
presented by Huang [10.41].
Equating Eq. (10.9.1) to the strength of the shell and taking the concrete strength of the
shell as about 90% of that in the core, or 0.75 fc′ ,
2.0 f yt ρs Ach = 0.75 fc′ ( Ag − Ach )
from which
 Ag  f′
ρs = 0.375  − 1 c (10.9.2)
 Ach  f yt
Using an additional factor of safety of 1.20 to assure that the strength provided by the spiral
through confinement exceeds the shell strength, Eq. (10.9.2) becomes
 Ag  f′
ρs = 0.45  − 1 c (10.9.3)
 Ach  f yt
which is specified in ACI-25.7.3.3. It is noted that the yield strength of the spiral, fyt, may
not exceed 100,000 psi (see ACI-Table 20.2.2.4a). A review of the spiral steel requirement
has been made by Gamble [10.42].
According to ACI-25.7.3.1, the clear spacing between spirals must be at least the greater
of 1 in. and 4 3 the maximum aggregate size, but shall not exceed 3 in. Also, the spiral
in cast-in-place construction shall have a diameter not less than 3 8 in. (ACI-25.7.3.2).
db
πDc2
Ach =
4
πh2 Asp = area of
Ag =
4 spiral
4Asp s
ρs =
Dcs
Dc
h
Figure 10.9.1 Spirally reinforced column.

32
Lower 1 12 diam
bar Upper
or 1 12 ” min bar
Lower
bar
Upper
bar
1 12 diam
or 1 12 ” min
Figure 10.9.2 Bar arrangement in spirally reinforced columns.
Anchorage of spiral reinforcement shall be provided by 1 1 2 extra turns of spiral bar or
wire at each end of the spiral unit (ACI-25.7.3.4). Splices, when necessary, shall be either
welded or mechanical splices conforming to ACI-25.5.7, or lap splices in accordance with
ACI-25.7.3.6 for yield strengths not exceeding 60,000 psi. For uncoated deformed bars or
deformed wire, the minimum splice length shall be 48 spiral bar or wire diameters but not
less than 12 in. (ACI-25.7.3.6). The spiral reinforcement is to be protected by the usual 1 1 2
in. minimum clear cover required by ACI-20.6.1.3 for nonprestressed beams and columns.
In spirally reinforced columns with large amounts of longitudinal steel, it sometimes
becomes necessary at splice locations to lap bars (see ACI-10.7.5 for lap splices of longitu-
dinal bars in columns) inside the main circle of bars. This is done to maintain a minimum
clear distance between individual longitudinal bars not less than the greatest of 1 1 2 times
the nominal bar diameter, 1 1 2 in., and 4 3 times the maximum aggregate size (ACI-25.2.3).
The preferred and alternate bar arrangements are shown in Fig. 10.9.2. Bars may also be
butt spliced by welding or mechanical connections (ACI-10.7.5.1.1), thus permitting more
utilization of available space. In a large column, an inner core of bars wrapped with a spiral
or by ties may also be used.
10.10 LIMITS ON PERCENTAGE OF LONGITUDINAL

REINFORCEMENT
The percentage of total longitudinal reinforcement area Ast in terms of the gross cross-
sectional area Ag must be between 1 and 8% (ACI-10.6.1.1): that is, ρg = Ast /Ag between
0.01 and 0.08. However, ACI-10.3.1.2 permits basing the percentage on a reduced concrete
area Ag when the gross concrete area is in excess of that needed for load considerations;
but in no case may the reduced concrete area be less than half the actual concrete area (i.e.,
reinforcement ratio ρg may not be less than 0.005 based on the gross area provided). The
primary purpose of these provisions for minimum steel is to prevent the failure mode from
becoming that of a plain concrete column, which might be more disastrous than the sudden
failure of tied columns previously described; in addition, these provisions are intended to
avoid yielding of longitudinal steel under service loads due to load transfer from the con-
crete to the steel caused by creep. The upper limit on the amount of longitudinal reinforce-
ment is a practical one in that if proper clearances are maintained between bars, little more
than ρg = 0.08 can be put into the section. Thus the maximum ρg is, in a way, a double check
on the minimum spacing restrictions of ACI-25.2.3. The need for amounts of longitudinal
steel close to this upper limit indicates that a small cross-sectional area is being provided
relative to the design loads, which could lead to problems associated with column stability
and excessive shear stresses. A review of the longitudinal steel limits has been made by Lin
and Furlong [10.43].
In order “to permit the use of reinforced concrete columns with small cross sections
in lightly loaded structures, such as low-rise residential and light office buildings,” ACI
makes no restrictions on column dimensions (ACI Commentary-R10.3.1). In such cases,
“there is a greater need for careful workmanship, and shrinkage stresses have increased
significance.”
3
Pn
Pn(max) = 0.80Po for tied columns
Po = 0.85Po for spirally reinforced columns
Pn(max)
e = emin
e= 0
e= ∞ Mn
Figure 10.11.1 Maximum nominal axial strength.
10.11 MAXIMUM STRENGTH IN AXIAL

COMPRESSION—A CI CODE
Since a truly concentrically loaded column is rare, if not nonexistent, some minimal eccen-
tricity should be provided for. Accidental eccentricity may result from end conditions,
inaccuracy of manufacture, or variation in materials even when the load is theoretically
concentric. Therefore, to account for a minimum eccentricity, the ACI Code prescribes
(ACI-22.4.2.1) that the maximum axial load nominal strength Pn(max) may not exceed 0.80Po
for tied columns and 0.85Po for spirally reinforced columns (Fig. 10.11.1), with Po given
by Eq. (10.5.1) or (10.5.2).
10.12 BALANCED STRAIN CONDITION

The balanced strain condition, or compression control limit, represents the dividing point
between the “compression-controlled” and the “transition zone” regions of the strength
interaction diagram (Fig. 10.6.1). Defined in the same manner as in Chapter 3, it is the
simultaneous occurrence of a compressive strain of 0.003 in the extreme fiber of concrete
and the tensile yield strain εy = fy /Es on the layer of steel reinforcement closest to the
tension face.
Referring to the rectangular section in Fig. 10.12.1, the balanced strain condition gives
cb 0.003
=
d f y /Es + 0.003
0.003 87, 000 d

cb = d= (10.12.1)
f y /[29(106 )] + 0.003 f y + 87, 000
Force equilibrium requires
Pb = Cc + Cs − T (10.12.2)
The compressive strength of concrete in a column subjected to pure axial load is k3 fc′ (see
Eq. (10.3.3) where k3 is taken equal to 0.85. Under combined bending and axial load, the
stress intensity of Whitney’s stress block should be taken as k3 (0.85 fc′ ). It has become com-
mon practice, however, to use a stress block intensity in columns of 0.85 fc′ , as in beams
(i.e., k3 = 1.0). Such a difference in stress block intensity will not have significant impact in
the calculated strength of a column below the balanced point; it may, however, lead to appre-
ciable differences as the axial load approaches the pure axial load strength of the column. For
34
e’
d” e = eb
Plastic
N.A. centroid
As As’
Pn = Pb
b
fy
εy =
Es
cb
εs’
εcu = 0.003
d’
d
h Pn = Pb
Actual stress distribution
Average = 0.85k3fc’
T = Asfy Cc Cs
a = β1cb
Figure 10.12.1 Balanced strain condition—rectangular section.
the balanced strain condition, this difference may be neglected and thus, k3 will be taken
equal to 1.0. Based on this, the resultant concrete compressive force Cc in Eq. (10.12.2) can
be expressed as
Cc = k3 (0.85 fc′ ab) = 1.0(0.85 fc′ ab) = 1.0(0.85 fc′β1cb b) (10.12.3)
and the tension force
T = As f y (10.12.4)
If, moreover, compression steel yields at the balanced strain condition,
Cs = As′ ( f y − 0.85 fc′ ) (10.12.5)
Thus Eq. (10.12.2) becomes
Pb = 0.85 fc′β1cb b + As′ ( f y − 0.85 fc′ ) − As f y (10.12.6)
The eccentricity eb is measured from the plastic centroid, which was defined in
Fig. 10.5.1. For symmetrical sections, the plastic centroid is at the middepth of the section.
Moment equilibrium of the forces in Fig. 10.12.1 is satisfied by taking moments about
any point such as the plastic centroid,
 a 
M b = Pb eb = Cc  d − − d ′′ + Cs (d − d ′ − d ′′ ) + Td ′′ (10.12.7)
 2 
35
EXAMPLE 10.12.1
Determine the axial force Pb and eccentricity eb corresponding to the balanced strain
condition (compression control limit) for the section of Fig. 10.12.2. Use fc′ = 3000 psi,
k3 = 1.0, fy = 50,000 psi, and the ACI Code.
Figure 10.12.2 Balanced strain condition for Example 10.12.1.
SOLUTION
(a) Locate the neutral axis for the balanced strain condition.
0.003(21.6)
cb = = 13.72 in.
0.00172 + 0.003
ab = β1cb = 0.85(13.72) = 11.67 in.
The value β1 is to be taken at 0.85 for fc′ ≤ 4000 psi (ACI-22.2.2.4.3).

(b) Compute the forces Cc, Cs, and T.
Cc = 0.85(3.0)(11.67)(15) = 446 kips

T = 50.0(2.37) = 119 kips
 13.72 − 2.4  fy
ε s′ = 0.003   = 0.00248 > compression steel yiields
 13.72  Es
Cs = (2.37)[50.0 − 0.85(3.0)] = 119 − 6 = 113 kips
(c) Compute Pb and eb.

Pb = Cc + Cs − T = 446 + 113 − 119 = 440 kips
(Continued)
36
For moment equilibrium about the plastic centroid,
Pb eb = [ 446(12 − 11.67 / 2) + 113(12 − 2.4) + 119(12 − 2.4)]

1
12
= 229 + 90 + 95 = 414 ft-kips
414(12)
eb = = 11.3 in.
440
On the given section, if Pn > 440 kips (or e < 11.3 in.), the section is referred to as com-
pression controlled; if Pn < 440 kips (or e > 11.3 in.), the section may be in the “transition
zone” (see Fig. 10.6.1) or, when e is large enough, the section would be tension controlled.
10.13 NOMINAL STRENGTH OF A COMPRESSION-

CONTROLLED RECTANGULAR SECTION
When the nominal compression strength Pn exceeds the balanced nominal strength Pb, or
when the eccentricity e is less than the balanced value eb, or when εt at the extreme layer
of steel at the face opposite the maximum compression face is less than the tensile yield
strength εy, the section is “compression controlled.” The tensile force T [see Fig. 10.13.1(c)]
will then be based on a tensile strain less than εy (see Fig. 10.6.1) and may actually be a
compressive force if the eccentricity is small enough.
The nominal strength Pn for a given eccentricity e < eb may be obtained by considering the
strain variation (or neutral axis depth c) as the unknown and using the principles of statics. This
is the most rational approach. As an alternative to the direct solution for the neutral axis as shown
in the following examples, Reed [10.44] has presented an iterative procedure for analysis.
EXAMPLE 10.13.1
Determine the nominal compressive strength Pn for the section of Fig. 10.13.1 for an
eccentricity e = 8 in. Use fc′ = 3000 psi, k3 = 1.0, fy = 50,000 psi, and the ACI Code.
Figure 10.13.1 Strain distribution and internal force resultants in a compression-controlled

section for Example 10.13.1.
(Continued)
37
10.13 COMPRESSION-C
O N T R O L L E D R E C TA N G U L A R S E C T I O N 337
SOLUTION
(a) Determine whether the given eccentricity e is larger or smaller than eb. Since the
balanced strain condition was computed in Example 10.12.1 as

Pb = 440 kips, eb = 11.3 in
it is known that the section is compression controlled for e < 11.3 in. For e = 8 in., how-
ever, the position of the neutral axis c is not known.
(b) Determine the location of the neutral axis. Since the actual c for e = 8 in. should
exceed the value of cb = 13.72 in. and the value of ε s′ exceeds εy at the balanced strain
condition, it is certain that ε s′ > ε y (see strain diagram of Fig. 10.12.2). Thus, refer-
ring to Fig. 10.13.1 for the forces and taking k3 = 1.0,
Cs = As′ ( f y − 0.85 fc′ ) = 2.37(50.0 − 2.55) = 113 kips

Cc = 0.85 fc′ b ( 0.85c ) = 2.55 (15)( 0.85c ) = 32.5c
Because ε s < ε y
(29, 000)(0.003)(21.60 − c) 4450 − 206c

T = As fs = As ( Es ε s ) = 2.37 =
c c
Taking moments about Pn,
 0.85c  4450 − 206c

0 = 113(4.0 − 2.40) − 32.5c  − 4.0 + (21.6 − 4.0)
 2  c
0 = c 3 − 9.41c 2 + 249.6c − 5673
c = 16.0 in.
(c) Compute internal forces and strength Pn.
Cs = 113 kips
Cc = 32.5(16.0) = 520 kips
4450 − 206(16.0)
T= = 72.11 kips
16.0
Pn = 520 + 113 − 72.1 = 561 kips
(d) Check by a moment equation about the plastic centroid.

561(8) ≈ 113(9.6) + 72.1(9.6) + 520(12 − 6.80)

4487 ≈ 4481 OK
EXAMPLE 10.13.2
Repeat the calculation of the nominal strength Pn for the section of Fig. 10.13.1, except
consider that 1–#8 is added in each of the 24-in. faces of the member, at the middepth
12 in. from the compression face of the member (see Fig. 10.13.2). The eccentricity of
Pn is 8 in., fc′ = 3000 psi, and fy = 50,000 psi. Use k3 = 1.0.
(Continued)
38
SOLUTION
(a) Estimate whether the given eccentricity will cause the section to be compression
controlled, in the transition zone, or tension controlled based on Fig. 10.6.1. Since
the extra 2–#8 bars are at the center of the section, they are unlikely to have an impor-
tant effect on eb. Assume that the section will be compression controlled, since eb for
Example 10.13.1 exceeds the 8-in. eccentricity. Note that even if a wrong assumption
is made, the neutral axis location thus obtained will reveal whether ε s′ < ε y or ε s > ε y ,
in which case the revised expression for Cs or T will have to be used in a new solution.
(b) Estimate for each layer of reinforcement whether it is in compression or tension,
and whether the strain at the layer will be greater or smaller than εy = 0.00172. The
assumptions are as follows.
For layer 1, 3–#8 at 2.4 in. from the compression face; bars are in compression and are
assumed to yield

Cs1 = As′1 ( f y − 0.85 fc′ ) = 2.37(50 − 2.55) = 113 kips
For layer 2, 2–#8 at 12 in. from the compression face; bars are assumed to be in com-
pression within the compressive stress block and assumed not to yield:
0.003
ε s′2 = (c − 12.0)
c
 0.003 
Cs 2 = As′2  (c − 12.0)29, 000 − 0.85 fc′
 c 
 87 
= 1.58  (c − 12.0) − 2.55
c 
1650
= 137.5 − − 4.03
c
1650
= 133.5 −
c
For layer 3, 3–#8 at 21.6 in. from the compression face; bars assumed to be in tension
and assumed not to yield
 0.003 
T = As  (21.6 − c)29, 000 
 c 

 87  4454
= 2.37  (21.6 − c) = − 206.2
c  c
(c) Determine the compressive force Cc as a function of the neutral axis depth c. From
Example 10.13.1,

Cc = 32.5c
(d) Compute the neutral axis distance c. Taking moments about Pn,
 0.85c 
0 = 113(4.0 − 2.4) − 32.5c  − 4.0
 2 
133.5c − 1650 4454 − 206.2c
− (8.0) + (21.6 − 4.0)
c c
0 = c 3 − 9.41c 2 + 327.1c − 6629
c = 15.63 in. a = 0.85(15.63)=13.3 in.
(Continued)
39
10.13 COMPRESSION-C
O N T R O L L E D R E C TA N G U L A R S E C T I O N 339
Figure 10.13.2 Analysis of a section containing an intermediate layer of steel for Example 10.13.2.
(e) Verify assumptions (see Fig. 10.13.2).

For layer 1, 3–#8 at 2.4 in. from the compression face,
0.003
ε s′1 = (15.63 − 2.40) = 0.00254 > ε y (as assumed )
15.63
For layer 2, 2–#8 at 12 in. from the compression face,
0.003
ε s′2 = (15.63 − 12.0) = 0.00070 < ε y (as assumed )
15.63
fs′2 = ε s′2 Es = 0.00070(29, 000) = 20.3 ksi
These bars are in compression as assumed, and the depth a is greater than 12 in. Thus,
the correction for displaced concrete (i.e., the −0.85 fc′ in the Cs2 force above) was
appropriate.
For layer 3, 3–#8 at 21.60 in. from the compression face,
0.003
εs = (21.60 − 15.63) = 0.00115 < ε y (as assumed )
15.63
These bars are in tension as assumed.
(Continued)
340
(f) Compute internal forces and strength Pn.
Cs1 = 113 kips
Cs 2 = As′2 ( fs′2 − 0.85 fc′ ) = 1.58(20.3 − 2.55) = 28.0 kips
T = As ε s Es = 2.37(0.00115)29, 000 = 79.0 kips

Cc = 32.5c = 32.5(15.663) = 508 kips
Pn = 113 + 28.0 − 79.0 + 508 = 570 kips
(g) Check by a moment equation about the plastic centroid.

570(8) = 113(9.6) + 28.0(0) + 79.0(9.6) + 508(12 − 6.64)
OK
4556 ≈ 4561
Note that the additional two bars placed at the plastic centroidal axis increase Pn only
from 561 kips (Example 10.13.1) to 570 kips (this example) for the same eccentricity
of 8 in.
10.14 NOMINAL STRENGTH OF A RECTANGULAR

SECTION WITH ECCENTRICITY e GREATER THAN
THAT AT THE BALANCED STRAIN CONDITION
When the eccentricity e is greater than the balanced value eb, or when the nominal compress-
ion strength Pn is less than the balanced nominal strength Pb, the net tensile strain εt at the
extreme layer of steel at the face opposite the maximum compression face is greater than
εy = fy /Es. In such a case, the section behaves closer to a beam than a column (see Fig. 10.6.1).
A rational approach to determine any point in the Pn–Mn diagram below the balanced point
corresponding to an eccentricity e is to designate the actual neutral distance c as unknown
and apply statics. Alternatively, a value for c < cb may be first selected and the corresponding
axial force, moment and eccentricity of the axial force determined through statics.
EXAMPLE 10.14.1
Determine the nominal compressive strength Pn for the member shown in Fig. 10.14.1
for an eccentricity e = 20 in., using fc′ = 3000 psi, fy = 50,000 psi, and the ACI Code.
SOLUTION
From the result of Example 10.12.1 it is known that e = 20 in. exceeds the eccentricity at
the balanced strain condition eb = 11.3 in.; therefore the strain εs on the tension steel exceeds
εy. It is assumed (initially) that the strain ε s′ on the compression steel is at least equal to the
yield strain εy, although the validity of the assumption must be verified before the solution
is accepted. Referring to Fig. 10.14.1 and using k3 = 1.0, the forces T, Cc, and Cs are
T = As f y = 3(0.79)(50.0) = 119 kips
Cc = 0.85 fc′ab = 0.85(3.0)(0.85)(c)(15) = 32.5c

Cs = As′ ( f y − 0.85 fc′ ) = 3(0.79)(50 − 2.55) = 113 kips
(Continued)
341
1 0 . 1 4 A R E C TA N G U L A R S E C T I O N W I T H E C C E N T R I C I T Y e > e b 341
Figure 10.14.1 Nominal strength Pn when e > eb for Example 10.14.1.
Force equilibrium requires

Pn = Cc + Cs − T = 32.5c + 113 − 119 = 32.5c − 6.0
Taking moments arbitrarily about the tension steel, moment equilibrium gives
 d − d′  a
Pn  e +  = Cc  d −  + Cs (d − d ′ )
 2 2
(32.5c − 6.0)(20 + 9.6) = 32.5c(21.6 − 0.425c) + 113(19.2)

c 2 + 18.82c − 169.9 = 0
c = 6.69 in.
Therefore
Cc = 32.5(6.69) = 217 kips

Pn = 217 − 6.0 = 211 kips
Verifying the correctness of the strain condition on the compression steel,
c − d′  6.69 − 2.40 
ε s′ = ε cu = 0.003   = 0.00192
c  6.69 

50
εy = = 0.00172 < 0.00192 OK
29, 000
The nominal moment capacity Mn is then,
1
M n = Pn e = 211(20) = 350 ft-kips
12
(Continued)
342
The complete strength interaction diagram for the section is given in Fig. 10.14.2.
Figure 10.14.2 Strength interaction diagram for the section of Fig. 10.14.1.
10.15 DESIGN FOR STRENGTH—R EGION I,

MINIMUM ECCENTRICITY
The approach to the design of members subjected to combined axial force and bending in
accordance with the strength design method of the ACI Code may be divided into three
categories (see Fig. 10.15.1): (a) design in Region I for a member having small or negligi-
ble bending moment (i.e., maximum permitted axial strength Pn(max) governs); (b) design for
Region II in which a section is compression controlled but the strength Pn is less than max-
imum permitted strength of 0.80Po (tied columns) or 0.85Po (spirally reinforced columns);
and (c) design for Region III in which e > eb or εt > εy.
Designs in Region I, although permitted by the ACI Code, are nevertheless discouraged,
since members subjected to compressive forces close to their pure axial load capacity pos-
sess low deformation capacity, and thus may fail without warning.
Design in Region I occurs under the following conditions.
1. For braced or nonsway members (i.e., k ≤ 1) having low slenderness ratio kLu /r such
that the effects of slenderness may be neglected according to ACI-6.2.5 (see also
Section 10.7):
(a) The member is subject to axial compression where the bending moment is con-
sidered negligible and is not computed.
(b) The bending moment on the member is computed, but the corresponding eccen-
tricity e = Mu /Pu is less than emin (Fig. 10.15.1) corresponding to the maximum
axial capacity.
2. For braced or nonsway members, where the effects of slenderness must be considered,
computation of bending moment is required, with the result magnified by the factor
δns in accordance with ACI-6.6.4 (see Chapter 13) or obtained directly from a second-
order analysis. When the computed eccentricity is less than emin (Fig. 10.15.1), design
is in Region I.
34
10.15 DESIGN FOR STRENGTH—R

EGION I 343
Figure 10.15.1 Design categories for strength of a section under combined axial compression and
bending moment.
An unbraced or sway member (i.e., k > 1.0) will rarely if ever be a design in Region I; details
of the design procedure for unbraced compression members are to be found in Chapter 13.
As all sections designed in Region I are compression controlled, a strength reduction
factor φ = 0.65 (tied columns) or φ = 0.75 (spirally reinforced columns) must be used in
design.
EXAMPLE 10.15.1
Design an axially loaded, spirally reinforced circular column for a gravity dead load of
345 kips and a live load of 405 kips using approximately 3.5% longitudinal reinforce-
ment ratio. The column is of average height, and it will be assumed slenderness effects
can be neglected. Use fc′ = 4000 psi, fy = 60,000 psi, and the ACI Code.
SOLUTION
(a) Determine the required nominal strength Pn.

Pu = 1.2(345) + 1.6(405) = 414 + 648 = 1062 kips
Pu 1062
required Pn = = = 1416 kips
φ 0.75
(b) Determine the column size. ACI-22.4.2.1 gives the maximum nominal strength in
axial compression for spirally reinforced columns as

Pn(max) = 0.85Po
where Po is given by Eqs. (10.5.1) or (10.5.3). Thus
Pn (max) = 0.85 Ag 0.85 fc′ + ρg ( f y − 0.85 fc′ )
1416 = 0.85 Ag [0.85(4) + 0.035(60 − 3.40)]
1416
required Ag = = 310 sq in., h(diameter) = 19.8 in.
4.57
Try h = 20 in. with Ag = 314 sq in.
(Continued)
34
(c) Determine reinforcement. Solve the Pn(max) equation for ρg.
1416 = 0.85(314)[3.40 + ρg (60 − 3.40)]
required ρg = 0.0337
required Ast = ρg Ag = 0.0337(314) = 10.57 sq in.
Use 20-in. diameter column with 7–#11 bars (As = 10.92 sq in.).

No further check is required because no moment has been computed and the maxi-
mum nominal axial strength given by ACI-22.4.2.1 governs in this case.
(d) Design the spiral reinforcement. Using Eq. (10.9.3), which is ACI Formula
(25.7.3.3),
 Ag  f′
ρs = 0.45  − 1 c
 Ach  f yt
Using 1.5-in. clear cover to the spiral
π(20 − 3)2
Ach = = 227 sq in.
4

 314  4.0
ρs = 0.45  − 1 = 0.0115
 227  60.0
Applying Eq. (10.3.2) gives

4 Asp 4 Asp
smax = =
ρs Dc 0.0115(17)
which gives the data in Table 10.15.1.
TABLE 10.15.1a SPACING OF SPIRAL REINFORCEMENT FOR EXAMPLE 10.15.1
Bar Asp smax Maximum Clear Spacing (in.)

(sq in.) (in.)
#3 0.11 2.25 1.87

#4 0.20 4.10 3.59
a Limitations (ACI-25.7.3.1):
1. Clear spacing < 3 in.
2. Clear spacing > 1 in. (assumed to be also greater than 4 3 times the maximum aggregate size).
Use #3 spiral at 2-in. spacing.

345

EGION II 345
10.16 DESIGN FOR STRENGTH—R EGION II,

COMPRESSION-C ONTROLLED SECTIONS
(e min < e < e b )
Sections that fall within Region II in Fig. 10.15.1 are compression controlled, but given the
combination of axial force and bending, the axial strength is less than the maximum of either
0.80Po for tied columns or 0.85Po for spirally reinforced concrete columns. Therefore, the
applicable strength reduction factor φ is that for compression-controlled sections (φ = 0.65
for tied columns and φ = 0.75 for spirally reinforced columns). The design of a column in
Region II is illustrated in Examples 10.16.1 and 10.16.2.
EXAMPLE 10.16.1
Design a square tied column with no more than 3% longitudinal reinforcement ratio for
a dead load axial force of 270 kips and bending moment of 150 ft-kips, and a live load
axial force of 150 kips and bending moment of 65 ft-kips. Use fc′ = 4500 psi, fy = 60,000
psi, and the ACI Code.
SOLUTION
(a) Compute the required nominal strength.
Pu 1.2(270) + 1.6(150)
φ 0.65

M 1.2(150) + 1.6(65)
required M n = u = = 437 ft-kips
φ 0.65
(b) Because the required axial strength for a column design within Region II is above
the balanced point, it is useful to estimate first the size required for a design cor-
responding to the balanced condition. Assuming a total of eight longitudinal bars,
equally spaced around the column perimeter, are used and that the middle layer is in
compression (Fig. 10.16.1),

Pn = Cc + Cs1 + Cs 2 − Ts 3
Disregarding the force in the middle layer of reinforcement, which for the balanced
condition is small relative to the other forces, and assuming that the steel in the extreme
compression steel yields (i.e., εs1 ≥ εy),
3 
Cc = 0.85 fc′ cbβ1b = 0.85(4.5)  d3  (0.825)b = 1.89bd3
5 
( )
Cs1 = As1 f y − 0.85 fc′ = As1 (60 − 3.83) = 56.2 As1
Ts = As 3 f y = 60 As 3
Pn = Pb = 1.89bd + 56 As1 − 60 As 3
For this symmetrical reinforcement layout, As1 = As3. Thus, it can be seen that the axial
strength at balanced condition for a square or rectangular section of given dimensions is
little sensitive to the amount of longitudinal reinforcement provided. The axial force at
the balanced condition can thus be approximated as
Pu
Pb ≈ 1.89bd3 = = 868 kips
φ
(Continued)
346
Figure 10.16.1 Strains and internal force resultants in column section in Example 10.16.1 at the
balanced condition (transverse reinforcement omitted for clarity).
For a square cross section, b = h. Taking d3 as 0.9 h,

Pu
Pb ≈ 1.7h 2 = = 868 kips
φ
which leads to h = 22.6 in.
A square section with side dimension less than 22.6 in. should therefore fall within
Region II (compression-controlled region) of the Pn–Mn interaction diagram.
(d) Select section dimensions and calculate amount of longitudinal reinforcement. Try
a 20-in. square column (assume a cover to center of bars of 2.5 in.).
The required area of longitudinal reinforcement and the neutral axis depth can be calcu-
lated either by solving simultaneously the expressions for axial force and moment about
the column plastic centroid or through trial and error. The first approach, however, could
be time consuming. Thus, in this example, a reinforcement ratio of approximately 2.5%
will first be selected and adjusted based on the calculated capacity.

As ≈ 0.025bh = 0.025(20)(20) = 10 sq in.
Select 8–#10 bars; As = 8(1.27) = 10.16 sq in.
The neutral axis depth for Pn = 868 kips can be calculated from equilibrium of axial
forces. Assuming that the steel in the extreme compression layer yields and that the mid-
dle layer of steel is outside the compressive stress block,
Pn = 0.85 fc′ bβ1c + As1 ( f y − 0.85 fc′ ) + As 2 fs 2 + As 3 fs 3 = 868 kips
where tensile stresses are negative and compressive stresses arre positive.
Pn = 0.85(4.5)(20)(0.825)c + (3.81)(60 − 0.85(4.5)) + (2.54) fs 2 + (3.81) fs 3
= 868 kips
  0.003   0.003 
Pn = 63.1c + 214 + (2.54)  (c − 10) + (3.81)  (c − 17.5)  29, 000
  c   c 
= 868 kips
 0.0762 0.20 
Pn = 63.1c + 214 + 0.0076 − + 0.0114 − 29, 000 = 868 kips
 c c 
Thus,
63.1c 2 − 102c − 8011 = 0
c = 12.1 in.
For this neutral axis depth, the extreme layer of compression reinforcement As1 yields,
as assumed. Also, the middle layer of reinforcement is outside the compressive stress
block, as β1c = 9.98 in.
(Continued)
347

EGION II 347
The nominal flexural strength for Pn = 868 kips can be calculated with respect to the
section plastic centroid as follows (refer to Fig. 10.16.1 for internal forces),
 h β c h  h 
2 2 
( )
M n = Cc  − 1  + As1 f y − 0.85 fc′  − d1  + As 2 fs 2 (0) + As 3 fs 3  − d3 
2  2 
M n = [764(10 − 4.99) + 3.81(56.2)(7.5) + 2.54(15.1)(0)

 1
+3.83( −38.8)(10 − 17.5)]  
 12 
M n = 545 ft-kips > 437 ft-kkips
The calculated nominal moment capacity is approximately 25% greater than the required
capacity, and thus a reduction in the amount of longitudinal steel is possible. Try
8–#8 bars ( ρg = 0.016 > 0.01).
For a design with 8–#8 bars at Pn = 868 kips, c = 12.6 in., and Mn = 452 ft-kips >
438 ft-kips.
Use 20-in. square column with 8–#8 bars, equally distributed in each face (Fig. 10.16.2).
(g) Select lateral ties. Applying the provisions of ACI-25.7.2, try #3 ties. Spacing
limitations:
1. Least lateral dimension = 20 in.
2. 16 longitudinal bar diameters = 16(1.0) = 16 in.
 3
3. 48 tie bar diameters = 48   = 18 in.
 8
Use #3 ties, 1 closed tie and a crosstie in each direction per set, at 16-in. spacing.
The final cross section is shown in Fig. 10.16.2.
#3 crossties @ 16”
Ast = 8 – #8
20” #3 ties @ 16”
1.5”
20”
Figure 10.16.2 Section for Example 10.16.1.
Practical Design Approach

In practice, compression member design is rarely done in the detailed manner illustrated
throughout this chapter. Instead, designers usually use interaction diagram charts in nondi-
mensional format or computer software. Various design aids are available, such as ACI-SP-17
(The Reinforced Concrete Design Handbook) [2.20], (which contains nondimensional
interaction diagrams; the very useful ACI predecessor document SP-7, by Everard and
Cohen [10.46]; the CRSI Design Handbook 2008 [2.21]; and for SI, the Canadian Metric
Design Handbook [10.47]. Typical interaction charts ( fc′ = 4000 psi, fy = 60,000 psi, and
γ = 0.6 and 0.8) from Ref. 2.20 are given as Fig. 10.16.3. For L-shaped columns, Marin
[10.48] has provided design aids. Circular columns are treated in Section 10.18 of this
book. Today, however, most column designs are performed using computer software based
on a sectional analysis.
348
Figure 10.16.3 Strength interaction diagrams for uniaxial bending and compression, symmetrical
reinforcement, fc′ = 4 ksi, fy = 60 ksi, γ = 0.60 and 0.80. (From ACI SP-17 [2.20]).
349

EGION II 349
EXAMPLE 10.16.2
Design a square tied column containing about 2% longitudinal reinforcement to carry a

dead load axial compression of 770 kN and a bending moment of 68 kN·m, and a live
load axial compression of 503 kN and bending moment of 33 kN·m. Use fc′ = 30 MPa,
fy = 400 MPa, and the ACI Code. Assume the longitudinal reinforcement is distributed
in equal amounts on two opposite faces.
SOLUTION
(a) Calculate factored axial force and moment and compute the eccentricity
Pu = 1.2(770) + 1.6(503) = 1729 kN

Mu = 1.2(68) + 1.6(33) = 134 kN ⋅ m
134(1000)
e= = 77.5 mm
1729
(b) Estimate e/h and use the strength interaction charts in Fig. 10.16.3, which are
for φ = 1.0. Note that the chart values of fc′ = 4000 psi and fy = 60,000 psi corre-
spond closely to the given metric data. Alternatively, the Canadian Metric Design
Handbook [10.47] can be used. Try the chart (Fig. 10.16.3) for γ = 0.8. Try e/h = 0.2
(radial line from origin to vertical axis 1.0 intersected with horizontal axis 0.2) and
use ρg = 0.02. From the chart, obtain
Pn
≈ 0.70
fc′ Ag
Note that for e/h = 0.2, Pn > Pb and thus the section is compression controlled. Use
φ = 0.65.
Pu 1729(1000)
Pn φ 0.65
required Ag ≈ = = = 126, 700 mm 2
0.70 fc′ 0.70(30) 0.70(30)
Try a section 360 mm square (Ag = 129,600 mm2). Check assumed value for γ. For a
cover to the center of the longitudinal bar equal to 65 mm,
360 − 2(65)
γ= = 0.64 < 0.8
360
Since the assumed value for γ is not conservative (bars are farther from the column faces
than assumed), either use chart for γ = 0.6 or interpolate between two charts.
Compute chart horizontal axis value ( Pn / fc′Ag ) ( e / h ),
Pu /φ  e  1729(1000) / 0.65  77.5 

  =   = 0.684(0.215) = 0.147
fc′ Ag h 30(129, 600)  360 
Enter Fig. 10.16.3 (interpolating for γ) with ( Pn / fc′Ag ) ( e / h ) = 0.147 and Pn / fc′Ag = 0.684;
find ρg ≈ 0.023.

required Ast = 0.023(129, 600) = 2980 mm 2
Try 6–#25M bars from Table 1.13.2, Ast = 3060 mm2.

(Continued)
350
(c) Check by statics. Based on the e/h ratio and the curves of Fig. 10.16.3, the section is
expected to be compression controlled. Referring to Fig. 10.16.4,
1
Cc = 0.85 fc′ ba = 0.85(30)(360)a = 9.18a
1000
a = β1c = 0.85c (see footnnote, Section 3.3)

Cc = 7.80c
Assume compression steel yields:
Cs = 3(510) [ 400 − 0.85(30)]

1
= 573 kN
1000
Taking Es = 200,000 MPa and d = 296 mm,
 296 − c  1
T = 3(510)  (0.003)(200, 000)
 c  1000

271, 730 − 918c
T=
c
360 mm
#10M
6 – #25M
360 mm 40-mm
clear
cover
0.003
c = 277 mm
77.5 mm
Pn
0.85fc’
64 mm
T
235 mm Cc Cs
Figure 10.16.4 Section, strain distribution, and internal force resultants for Example 10.16.2.
(Continued)
351

EGION III 351
Taking moments about Pn gives
T (296 − 102.5) + Cs (102.5 − 64) + Cc (102.5 − 0.425c) = 0

271, 730 − 918c
(193.5) + 573(38.5) + 7.80c(102.5 − 0.425c) = 0
c
c 3 − 241c 2 + 46, 930c − 15, 861,160 = 0
c = 277 mm
Check strains:
fy 400
εy = = = 0.00200
Es 200, 000
At Cs,
 277 − 64 
ε s′ = 0.003  = 0.00231 > ε y (compression steel yieldds)
 277 
At T,
 296 − 277 
ε s = 0.003  = 0.00021 < ε y
 277 
This steel is in tension but does not yield; the section is compression controlled, as
assumed.
Compute forces:
Cc = 7.80c = 7.80(277) = 2161 kN

C s = 573 kN
T = 3(510)(0.00021)(200) = −64 kN
Pn = 2670 kN
[φPn = 0.65(2670) = 1736 kN] > [ Pu = 1729 kN required] OK
Compute nominal moment capacity by taking moments with respect to the plastic centroid:
1
M n = [2161[180 − 0.5(235)] + 573(116) + 64(116) = 208 kN ⋅ m
1000

φ M n = (0.65)208 = 135 kN ⋅ m > Mu = 134 kN ⋅ m OK
Use 360 mm square column with 6–#25M bars, as shown in Fig. 10.16.4.
10.17 DESIGN FOR STRENGTH—R EGION III,

TRANSITION ZONE AND TENSION-
CONTROLLED SECTIONS (e > e b )
The “transition zone” (see Fig. 10.6.1) corresponds to the region where εt is greater than
εy (e > eb), but less than 0.005. When εt exceeds 0.005, the section is tension controlled and
φ = 0.90.
352
The transition zone provides the variation in strength reduction factor φ from that corre-
sponding to compression-controlled sections to the φ factor for tension-controlled sections,
in accordance with ACI-21.2.2.
The φ factors of 0.65 and 0.75 for tied and spirally reinforced columns, respectively, are
to be used when a section is compression controlled (i.e., when εt < εy). In the transition
zone (Fig. 10.6.1), where the extreme tensile strain εt is between εy and 0.005, the φ fac-
tor is to be obtained by linear interpolation using either Eq. (3.6.3) or (3.6.4), as shown in
Fig. 3.6.2. The linear equations in terms of the extreme tensile strain εt for Grade 60 bars are
1. For tied sections:
 250 
φ = 0.65 + (ε t − 0.002)  ≤ 0.90 (10.17.1)
 3 
2. For spirally reinforced sections:
φ = 0.75 + ( ε t − 0.002 ) 50 ≤ 0.90 (10.17.2)
The φ equations can also be expressed in terms of the ratio c/dt, where c is the neu-
tral axis depth measured from the compression face of the member and dt is the dis-
tance from the extreme compressive fiber to the layer of reinforcement closest to the
tension face.
1. For tied sections:

 1 5
φ = 0.65 + 0.25  −  ≤ 0.90 (10.17.3)
c
 t / d 3 
2. For spirally reinforced sections:
 1 5
φ = 0.75 + 0.15  −  ≤ 0.90 (10.17.4)
 c /d t 3 
EXAMPLE 10.17.1
Design a rectangular tied column, not over 14 in. wide and with about 3% reinforce-
ment, equally distributed on all four faces, to carry a service dead load of PD = 65 kips
and MD = 144 ft-kips and a service live load of PL = 49 kips and ML = 115 ft-kips. Use
fc′ = 4500 psi, fy = 60,000 psi, and the ACI Code.
SOLUTION
(a) Required nominal strengths and eccentricity.
Pu = 1.2(65) + 1.6(49) = 156 kips
Mu = 1.2(144) + 1.6(115) = 357 ft -kips

357(12)
e= = 27.5 in.
156
(Continued)
35

EGION III 353
Assume a square section as a first approximation. Using Fig. 10.16.3 with γ = 0.8 ( fc′ is
not the same as that given in the figure but is close enough; also, for designs below the
balanced point, such a small difference in concrete compressive strength will have a neg-
ligible effect on column design) and e/h = 27.5/14 ≈ 2 (radial line from origin to vertical
axis 0.2 intersected with horizontal axis 0.4) shows that for ρg = 0.03, Pn is well below
Pb and probably εt > 0.005. Assuming φ = 0.85,
156
required Pn = = 184 kips
0.85

357
required M n = = 423 ft-kips
0.85
(b) Check assumed column depth. For e/h = 2 and ρ = 0.03, the values of Kn and Rn in
Fig. 10.16.3 are approximately 0.1 and 0.2, respectively. The required column area
would thus be
Pn 184
Ag ≥ = = 410 sq in.
fc′ K n 4.5(0.10)
which requires a column depth h ≥ 29.3 in. The required column depth is much greater
than the initially assumed value of 14 in. However, as the section depth increases, the
normalized eccentricity e/h decreases, leading to an increase in Kn. Assuming h = 20 in.,
e/h ≈ 1.5. From Fig. 10.16.3 and using ρ = 0.03, the values of Kn and Rn are approxi-
mately 0.14 and 0.21, respectively. The minimum required column cross sectional area
Ag = 292 sq in., which corresponds to h ≥ 20.9 in.
Try a section 14 in. wide and 20 in. depth. For a reinforcement ratio ρg = 0.03 (note
that the actual required reinforcement ratio should be slightly larger than 0.03 because
the section depth selected is slightly smaller than the minimum required for that rein-
forcement ratio),

Ast ≥ ρg Ag = 0.03(14)(20) = 8.4 sq in.
Select 8–#10 bars, Ast = 10.16 sq in., as shown in Fig. 10.17.1.

(d) Check section strength using statics (Fig. 10.17.1). Calculate the neutral axis depth
c corresponding to Pn = 184 kips. Assume first c = 7.5 in.

a = β1c = 0.825(7.5) = 6.19 in.
Strains and stresses in the reinforcing steel are

ε s1 = 0.0019, fs1 = 55.1 ksi (compression)
ε s 2 = −0.0010, fs 2 = −29.0 ksi (tension)

ε s 3 = −0.0039, fs 3 = −60.0 ksi (tension)
Compression force in concrete:

Cc = 0.85 fc′ ba = 0.85(4.5)(14)(6.19) = 331 kips
Net axial force:
Pn = Cc + Ast1 ( fs1 − 0.85 fc′ ) + Ast 2 fs 2 + Ast 3 fs 3

Pn = 331 + 3.81[ 55.1 − 0.85(4.5)] + 2.54( −29.0) + 3.81( −60) = 224 kips
(Continued)
354
The calculated Pn is approximately 20% greater than the required Pn = 184 kips. The
neutral axis depth c will thus be less than 7.5 in. For c = 7.1 in.,

a = β1c = 0.825(7.1) = 5.86 in.
Strains and stresses in the reinforcing steel are
ε s1 = 0.0018, fs1 = 53.3 ksi (compression)
ε s 2 = 0.0012, fs 2 = −35.5 ksi (tension)

ε s 3 = 0.0043, fs 3 = −60.0 ksi (tension)
Compression force in concrete:

Cc = 0.85 fc′ ba = 0.85(4.5)(14)(5.86) = 314 kips
Net axial force:
Pn = Cc + Ast1 ( fs1 − 0.85 fc′ ) + Ast 2 fs 2 + Ast 3 fs 3

Pn = 314 + 3.81[53.3 − 0.885(4.5)] + 2.54( −35.5) + 3.81( −60) = 183 kips ≈ 184 kips
As εt = 0.0043, φ = 0.84 (Eq. 10.17.1) ≈ 0.85, as originally assumed.

Calculate moment strength. Taking moments about column plastic centroid,
 h a h  h  h 
M n = Cc  −  + Ast1 ( fs1 − 0.85 fc′ )  − d1  + Ast 2 fs 2  − d2  Ast 3 fs 3  − d3 
 2 2 2  2  2 
  5.86  1
M n = 314  10 − 2  + 3.81 [53.3 − 0.85(4.5) ] (10 − 2.75) + 0 + 3.81( −60)(10 − 17 ..25) 12
= 437 ft-kkips > 423 ft-kips
Figure 10.17.1 Section, strain distribution, and internal force resultants and statics for Example 10.17.1.
10.18 CIRCULAR SECTIONS UNDER COMBINED

COMPRESSION AND BENDING
The concepts presented for the calculation of the strength Pn (acting at an eccentricity e
from the plastic centroid) for a rectangular section are equally applicable to a circular sec-
tion. The rectangular stress distribution may be applied to the concrete area under com-
pression according to ACI-22.2.2.3 and 22.2.2.4, though the use of statics requires knowing
the area and centroid of area for circular segments. This information may be computed from
formulas or by the use of coefficients for circular sections, as in Fig. 10.18.1. For a value
on the strength interaction diagram for a tension-controlled section, where the portion of
the circular section within the rectangular stress distribution may be small, the relationship
between the centroid of a circular segment and its area given by Fig. 10.18.2 may be useful.
35
10.18 CIRCULAR SECTIONS 355
When the steel reinforcement is in a circular arrangement, each bar may be located and
treated in the same manner as a layer of steel in a rectangular section. The analysis would
require first an assumption regarding whether the bar (or bars) will yield when εcu = 0.003,
as for Example 10.13.2. When about eight or more bars are used, the analysis may treat the
bars as a steel tube.
Design aids for circular columns, both tied and spirally reinforced, are available from
the Portland Cement Association [10.49], the American Concrete Institute [2.20], and the
Concrete Reinforcing Steel Institute [2.21]. Mekonnen [10.50] has provided strength equa-
tions for circular beam-columns. The CRSI has presented the equations to use for the inter-
action diagram using a programmable calculator [10.51], and several computer programs
are available to compute interaction diagrams based on a sectional analysis.
y h cos θ
2
dx = h dθ sin θ
2
(x, y)
dθ h
sin θ
θ 2
α
x
ηh
h
2 α
A of segment = h sin2 θ dθ = h2 α – sin α cos α
2 0 4
α
h3 sin3 α
Q0 of segment = sin2 θ cos θ dθ = h3
4 12
0
α
4
I0 of segment = h sin2 θ cos2 θ dθ = h4 4α – sin4 α
8 256
0
0.60 0.085 0.027
0.50 0.080 0.026
0.40 0.075 0.025
0.30 0.070 0.024
0.20 0.065 0.023
0.10 0.060 0.022
η = 0.3 0.4 0.5 0.6 0.7 η = 0.3 0.4 0.5 0.6 0.7 η = 0.3 0.4 0.5 0.6 0.7
A = (coefficient) h2 Q0 = (coefficient) h3 I0 = (coefficient) h4
Figure 10.18.1 Properties of circular segments.
Figure 10.18.2 Centroid of circular segment (measured from center of circle) versus its area.
356
EXAMPLE 10.18.1
For the circular column shown in Fig. 10.18.3, calculate the axial force and bending
moment corresponding to the balanced condition (i.e., compression control limit). Use
fc′ = 5000 psi, fy = 60,000 psi, and the ACI Code.
Figure 10.18.3 Cross section, strains, and internal force resultants in column of Example 10.18.1.
SOLUTION
Referring to Fig. 10.18.3, the depth of the neutral axis corresponding to the balanced
condition or compression control limit can be calculated as
  1.27  
 20 −  1.5 + 0.5 + 
 d5   2 
cb = 0.003   = 0.003   = 10.28 in.
 ε y + 0.003   0.00207 + 0.003 
 
 
and the depth of the compression stress block, a, is

a = β1cb = 0.80(10.28) = 8.23 in.
which corresponds to an angle α = 79.8° in Fig. 10.18.1 and 10.18.3. Referring to

Fig. 10.18.1, the resultant compressive force in the concrete can be calculated as
Cc = 0.85 fc′ Acomp = 0.85(5)(121.8) = 518 kips
which is located at yc = 5.22 in. from the center of the column.

The tensile strains in the longitudinal steel can be calculated based on the position of
the reinforcement with respect to the neutral axis depth (third column of Table 10.18.1).
Once these strains have been calculated, the stresses and corresponding forces in the
reinforcement can be determined, as shown in Table 10.18.1.
TABLE 10.18.1
Layer i Asti ysi + (cb −h/2) εsi fsi (ksi) Fsi (kips)
(in.2) (in.)
1 1.27 7.66 0.00223 60–4.25a 70.8

2 2.54 5.50 0.00160 46.5–4.25a 107
3 2.54 0.28 0.00008 2.40 6.09
4 2.54 –4.93 –0.00144 –41.7 –106
5 1.27 –7.09 –0.00207 –60 –76.2
a Stress is adjusted to account for displaced concrete. Tensile strains, stresses, and forces are negative.
(Continued)
357
10.19 COMBINED AXIAL TENSION AND BENDING 357
The axial force corresponding to the balanced condition or compression control

limit can then be determined as
i=5
Pn = Cc + ∑ Fsi = 518 + 2.1 = 520 kips
i =1
The corresponding nominal flexural strength can be calculated by taking moments

about the plastic centroid of the column (i.e., column center).
h  i=5
M n = Cc  − yc  + ∑ Fsi ysi = 225 + 183 = 408 ft-kips
2  i =1
10.19 COMBINED AXIAL TENSION AND BENDING

In the relatively uncommon situation of axial tension in combination with bending moment, the
strength interaction diagram can be considered to extend on the negative Pn side of the axis, as
shown in Fig. 10.19.1. The strength analysis for tension values of Pn is similar to that used for
compressive load in the region for a tension-controlled section (Fig. 10.6.1). When the eccen-
tric load is tensile, the neutral axis distance c will be smaller than it is under pure bending (M0).
When axial tension and bending moment exist, the usual tendency will be to proportion
the section in a manner similar to the design as a beam. The subject of axial tension com-
bined with bending moment has been discussed by Harris [10.52], Moreadith [10.53], and
Villalta, Carreira, and Erler [10.54].
For sections subjected to pure axial tension force, the nominal axial tensile
strength of the section is taken as the yield strength of the longitudinal reinforcement
(ACI-22.4.3.1).
Examples 10.19.1 and 10.19.2 illustrate the calculation of points on the tension side of
the interaction diagram (Fig. 10.19.1).
EXAMPLE 10.19.1
Determine the maximum value for the tensile force Pn when no bending moment is act-
ing on the section shown in Fig. 10.19.1.
Figure 10.19.1 Strength interaction diagram showing both axial compression and axial tension
regions.
(Continued)
358
SOLUTION
Since the concrete will crack before the steel yields, only the steel participates in carry-
ing axial tension. Thus, according to ACI-22.4.3.1,

Pn = Ast f y = 6(0.79)50 = 237 kips
This is plotted on Fig. 10.19.1.
EXAMPLE 10.19.2
Determine the axial tensile strength Pn on the section of Fig. 10.19.1 when the eccen-
tricity is 20 in.
SOLUTION
Figure 10.19.2 shows that the eccentric tensile force Pn must be acting on the tension
side of the plastic centroid. In this case, the neutral axis distance c is smaller than its
value for pure bending. For bending alone,
Cc = 0.85 fc′ ba = 0.85(3)(15)a = 38.3a

T = As f y = 3(0.79)50 = 118.5 kips
Assuming the compression steel yields,

Cs = 3(0.79)(50 − 2.55) = 112.5 kips
C=T
118.5 − 112.5 0.16
a= = 0.16 in.;; c= = 0.19 << 2.4 in.
38.3 0.85
The above computation shows that compression steel does not yield under bending
alone. In fact, it would be in tension but because of its proximity to the compression
face, it is not likely to yield in tension when Pn is a tensile force at e = 20 in.
Figure 10.19.2 Free-body diagram and assumed strain diagram for Example 10.19.2.

(Continued)
359
10.20 COMBINED AXIAL FORCE AND BIAXIAL BENDING 359
Thus for axial tension, estimate Cs to be in tension but at less than the yield stress.
 c − 2.4  206c − 495

Cs = As′ε s′ Es = 3(0.79)  (0.003)(29, 000) =
 c  c
Cc = 0.85 fc′ b(0.85c) = 0.85(3)(15)(0.85c) = 32.5c

T = As f y = 3(0.79)50 = 118.5 kips
Taking moments about Pn (refer to Fig. 10.19.2) gives
T (20 − 12 + 2.4) − Cc (20 + 12 − 0.425c) − C s (20 + 9.6) = 0
c 3 − 75.3c 2 − 352c + 1060 = 0

c = 2.10 in.
Check assumptions:
 2.10 − 2.40 
ε s′ =   0.003 = −0.000429 < ε y (tension)
 2.10 
The steel represented by Cs actually is in tension and does not yield; there is no displaced
concrete effect.
Cs = 3(0.79)( −0.000429)29, 000 = −29.4 kips (tension)
Cc = 32.5(2.10) = + 68.3 kips (compression)
T= − 118.5 kips (tension)

Pn = −79.6 kips (tension)
1
M n = 79.6(20) = 133 ft-kips
12
This value is plotted as point A on Fig. 10.19.1. Check by taking moments about the
plastic centroid:
133 = [(118.5 − 29.4)(9.6) + 68.3(12 − 0.89)]

1
12

133 ≈ 134 OK
The reader may note that when the axial tension is high enough to cause the neutral
axis to fall beyond the edge of the section, there is no longer a compression face on the
member. However, the method illustrated in this example seems sufficient to show how
points on the tension–bending moment interaction diagram may be obtained.
Note also that a φ factor of 0.90 is to be used for the entire region of axial tension.
10.20 COMBINED AXIAL FORCE AND BIAXIAL

BENDING
The investigation or design of a square or rectangular section subjected to axial compress-
ion in combination with bending moments about both the x- and y-axes has received con-
siderable attention [10.55–10.81]. One method of analysis is to use the basic principles of
equilibrium and compatibility with the same strength assumptions as were used earlier in
360
this chapter for the case of axial compression and bending about one axis only. This method
essentially involves a trial-and-error process for obtaining the position of an inclined neu-
tral axis corresponding to the desired axial force and eccentricities.
Although today’s practice will likely involve the use of in-house or commercial com-
puter programs in design, which are based on the use of compatibility, equilibrium, and
constitutive relations for concrete and steel, there are a number of approximate procedures
for obtaining reasonable estimates of the capacity of a member under biaxial bending and
compression. Some of these methods, namely, the reciprocal load method [10.55] and the
load contour method [10.55, 10.58], are discussed later in this chapter.
Nominal Strength for Combined Axial Force and Biaxial Bending

The analysis of a section subjected to combined axial force and biaxial bending may be
made by using statics and compatibility in the same way as for uniaxial bending and com-
pression. The compression zone can be modeled using Whitney’s stress block, as defined in
the ACI Code, or any other suitable stress distribution. In biaxial bending and axial force,
the uniform compressive stress 0.85 fc′ in Whitney’s stress block is considered to act on a
compression zone bounded by the edges of the cross section and a straight line at a distance
a = β1c from the fiber of maximum compressive strain (corner of section) and parallel to
the neutral axis, as shown in Fig. 10.20.1.
Each point in the axial force–biaxial bending nominal strength surface (Pn, Mnx, Mny)
will correspond to a given neutral axis depth and angle α in Fig. 10.20.1. For doubly
symmetrical lines, a value of α = 0 corresponds to a section subjected to uniaxial bend-
ing about the x-axis, while a value of α = 90° corresponds to a section subjected to
uniaxial bending about the y-axis. Strains in the longitudinal reinforcement are pro-
portional to the distance from each bar to the neutral axis (measured perpendicularly
to the neutral axis). Once the resultant forces in the concrete and reinforcing bars are
determined, the resultant axial force Pn and corresponding moments Mnx and Mny about
the plastic centroid are calculated. It should be noted that a line drawn from the col-
umn plastic centroid to the location of the axial force at eccentricities ey = Mnx /Pn and
ex = Mny /Pn will seldom be perpendicular to the neutral axis.
The procedure outlined above is illustrated in the following example. Because this pro-
cedure is time consuming, however, it is typically implemented through the use of com-
puter software.
Figure 10.20.1 Strains and stresses in section subjected to combined axial force and biaxial
bending (transverse reinforcement omitted for clarity).
361
EXAMPLE 10.20.1
For the rectangular column section in Fig. 10.20.2, calculate the biaxial flexural
strength for a nominal axial force of 426 kips and a moment ratio Mnx /Mny = 1.60. Use
fc′ = 6000 psi and fy = 60 ksi. Concrete cover to centroid of adjacent bar is 2.5 in. All
bars are equally spaced along each column face.
SOLUTION
The determination of biaxial moment strength for a given axial force involves a trial-and-
error process for calculating the neutral axis depth and angle α. For Pn = 426 kips, this
process led to a neutral axis depth c = 12.5 in. and α = 43.9° (Fig. 10.20.2). Once the values
of c and α are known, the resultant compressive force in the concrete can be determined as

Cc = 0.85 fc′ Acomp
where
(cβ1 )2 [12.50(0.75)]2
Acomp = = = 88..0 sq in.
2 cos α sin α 2 cos( 43.9)sin(43.9)
Thus,

Cc = 0.85 fc′Acomp = 0.85(6)(88.0) = 449 kips
Referring to Fig. 10.20.2, the strains, stresses, and forces in the longitudinal reinforce-
ment are listed in Table 10.20.1.
3
00
0.
Pn falls along this line
”
1.6 y .5
12
1 = ε s2
c
6 4
10 d2
3
20” x 9 x
2
8 α = 43.9º
7 5 1 Ast = 10 – #8 bars
16”
Figure 10.20.2 Cross section and strain distribution for column in Example 10.20.1 (transverse
reinforcement omitted for clarity).
Resultant axial force and moments about x- and y-axes are then determined as
i =10
Pn = Cc + ∑ Fsi = 426 kips
i =1
 h  i =10  h
M nx = Cc  yc −  + ∑ [ Fsi  yi −  = 321 ft-kips
 2  i =1  2
b  i =10 b 
M ny = Cc  − xc  + ∑ [ Fsi  − xi  = 201 ft-kips
2  i =1 2 
(Continued)
362
TABLE 10.20.1 STRAINS, STRESSES, AND FORCES IN LONGITUDINAL

REINFORCEMENT OF COLUMN IN EXAMPLE 10.20.1
Bar di (in.) εsi fsi (ksi) Fsi (kips)
1 −9.47 −0.00227 −60.0 −47.4

2 −5.87 −0.00141 −40.8 −32.3
3 −2.26 −0.00054 −15.8 −12.4
4 1.34 0.00032 9.3 7.40
5 −5.66 −0.00136 −39.4 −31.1
6 5.15 0.00124 35.9–0.85(6)a 24.3
7 −1.84 −0.00044 −12.8 −10.1
8 1.76 0.00042 12.3 9.70
9 5.36 0.00129 37.3–0.85(6)a 25.5
10 8.97 0.00215 60.0–0.85(6)a 43.4
a Adjusted to account for displaced concrete. Tensile strains, stresses, and forces are negative.
_ _
where xi and yi are the x and y coordinates of bar i, and xc and yc are the x and y coordi-
nates of the centroid of the area on which the stress block acts, respectively, all measured
from the bottom left corner of the section.
Verify that the relationship between moments about the x- and y-axes (or eccentrici-
ties ey and ex) is equal to the desired value (i.e., 1.60).
M nx ey 321
= = = 1.60
M ny ex 201
Failure Surfaces
By repeating the procedure discussed in the preceding section for a large number of combi-
nations of axial force and biaxial moments, a failure surface as a function of three variables,
Pn, Mnx, and Mny, such as that shown in Fig. 10.20.3, may be generated. Alternatively, the
failure surface may be expressed in terms of the the axial force Pn acting at eccentricities
ey = Mnx /Pn and ex = Mny /Pn, as shown in Fig. 10.20.4, or in terms of the reciprocal of the
axial force, 1/Pn, as shown in Fig. 10.20.5. These failure surfaces have served as the basis
for approximate methods for determining the nominal strength of sections subjected to
axial force and biaxial bending. Two of these methods, the reciprocal load method and the
load contour method, which are useful for design purposes, are discussed next.
Pn
Mny
Mnx
Figure 10.20.3 Pn-Mnx-Mny failure surface.

36
Pn 1
Pn
Failure
surface
S1(Pn, ex, ey)
Failure surface
S2(1/Pn, ex, ey)
ex ex
ey
ey
Figure 10.20.5 Reciprocal load failure surface
Figure 10.20.4 Pn-ex-ey failure surface. (From Bresler [10.55].) (1/Pn, ex, ey). (From Bresler [10.55].)
Reciprocal Load Method
Bresler, in an attempt to develop a realistic procedure for investigation, suggested [10.55]
approximating a point (1/Pn1, exA, eyB) on the reciprocal failure surface S2 by a point (1/Pi,
exA, eyB) on a plane S2′ passing through points A, B, and C (Fig. 10.20.6). Each point on the
true surface is approximated by a different plane; that is, the entire failure surface is defined
by an infinite number of planes.
The problem then is to determine the strength Pn1 that exists with biaxial eccentricities
exA and eyB by assuming that Pn1 equals the value Pi lying on the plane S2′ specifically estab-
lished for it. The specific plane is defined by passing it through three points A, B, and C
known to lie on the true failure surface S2,
 1 
A  exA , 0, 
 Pny 
 1 
B  0, eyB,
 Pnx 
 1
C  0, 0, 
 Po 
where Po is the nominal strength under axial compression alone without any eccentricity;
Pnx is the nominal strength at the uniaxial eccentricity eyB (= Mnx /Pnx); and Pny is the nominal
strength at the uniaxial eccentricity exA (= Mny /Pny).
In other words, point A represents a point (Pny, Mny) on the uniaxial Pn–Mn interaction
diagram, such as Fig. 10.6.1, for bending about the y-axis; point B represents a point (Pnx,
Mnx) on the uniaxial Pn–Mn interaction diagram for bending about the x axis; and point C is
a point that is common to both uniaxial Pn–Mn interaction diagrams.
The equation of the plane S2′ may be defined in terms of the three points A, B, and C as
follows:
1 1 1 1 (10.20.1)
= + −
Pi Pnx Pny Po
364
1
Pn
S2
S,2
A
B
1
C Pny
1
1 Po ex
1 1
Pnx Pn1 eyB
Pi
exA
ey
Figure 10.20.6 Graphical representation of the reciprocal load method. (From Bresler [10.55].)
The nominal axial strength of the column under biaxial bending, Pn, is then obtained by
taking Pn = Pi.
Bresler [10.55] found computed values of Pi using Eq. (10.20.1) to be “in excellent
agreement with test results, the maximum deviation being 9.4%, and the average devia-
tion being 3.3%.” In his discussion of Ref. 10.55, however, Pannell [10.57] suggests that
Eq. (10.20.1) is inappropriate when the ratio Pn /Po is in the range of 0.06 or less. In such
low axial load cases, it is best to design the member for flexure only.
Load Contour Method
The load contour method involves cutting the Pn–Mnx–Mny failure surface (Fig. 10.20.7) at
a constant value of Pn to give a so-called load contour interaction relating Mnx and Mny. In
other words, the entire failure surface may be considered to include a family of curves (load
contours) corresponding to constant values of Pn, which if drawn superimposed on one
another in a single plane would be analogous to a contour map. A typical plane at constant
Pn along with its load contour is shown in Fig. 10.20.7.
The general nondimensional equation for the load contour at constant Pn may be
expressed [10.55] in the form
α1 α2
 M nx   M ny 
 M  +  = 1.0 (10.20.2)
0x  M0 y 

where
M nx = Pn ey and M ny = Pn ex
M 0 x = M nx at axial load Pn when M ny (or ex ) is zero

M 0 y = M ny at axial load Pn when M nx (or ey ) is zero
and α1 and α2 are exponents that depend on the dimensions of the cross section, the rein-
forcement amount and location, the concrete strength, the steel yield stress, and the amount
of concrete cover.
Bresler [10.55] suggested that it is acceptable to take α1 = α2 = α; then Eq. (10.20.2)
becomes
α α
 M nx   M ny 
 M  +  M  = 1 (10.20.3)
0x  0y 
which is shown graphically in Fig. 10.20.8 for various values of the exponent α.

365
Pn
Pn – Mn interaction curves
Failure surface S3
Plane at constant Pn
M0y
Load
contour
M0x
Pn
Mny
Mnx
Figure 10.20.7 Load contours for constant Pn on Pn-Mnx-Mny failure surface. (Adapted from
Bresler [10.55].)
1.0
α
=
α
3
=
4
0.8 α
=
α
1.
4
=
2
0.6
Mnx ey
α
=
or
M0x e0y
1
0.4
0.2
0 0.2 0.4 0.6 0.8 1.0

Mny ex
or
M0y e0x
Figure 10.20.8 Interaction curves for Eq. (10.20.3). (From Bresler [10.55].)
In using Eq. (10.20.3) or Fig. 10.20.8, however, it is still necessary to have a value of α

that is applicable to the particular column under investigation. Bresler [10.55] reported the
calculated values of α to vary from 1.15 to 1.55. For practical purposes, it seems satisfac-
tory to take α as 1.5 for rectangular sections. The reader is referred to the extension of the
Bresler load contour method by Parme, Nieves, and Gouwens [10.58], in which additional
information is presented for defining the shape of the load contour.
36
EXAMPLE 10.20.2
Use the reciprocal load and the load contour methods to evaluate the adequacy of the
column section analyzed in Example 10.20.1 (Fig. 10.20.2) when subjected to Pu = 440
kips, Mux = 225 ft-kips, and Muy = 150 ft-kips.
SOLUTION
(a) The eccentricities are
Mux 225(12)
ey = = = 6.1 in.
Pu 440

Muy 150(12)
ex = = = 4.1 in.
Pu 440
(b) Use of the reciprocal load method calls for the determination of the values of Pnx and
Pny. Pnx represents the axial force in the Pn–Mnx interaction diagram corresponding to
the intersection between a radial line ey = 6.1 in. and the Pn–Mnx diagram. Similarly,
Pny represents the axial force in the Pn–Mny diagram corresponding to the intersec-
tion between a radial line ex = 4.1 in. and the Pn–Mny diagram. The determination of
Pnx = 925 kips and Pny = 1075 kips is illustrated in Fig. 10.20.9. Note that each of
these two points falls in the compression control region of the corresponding inter-
action diagram.
The common point for both interaction diagrams is the pure axial load strength, Po,
which is computed using Eq. (10.5.1) as follows,
Po = 0.85 fc′( Ag − Ast ) + Ast f y = 0.85(6)[320 − 10(0.79)] + 10(0.79)(60)

Po = 2066 kips
Then, using Eq. (10.20.1),
1 1 1 1 1
≈ = + −
Pn Pi Pnx Pny Po
1 1 1 1 1
= + − =
Pi 925 1075 2066 655
Pn ≈ Pi = 655 kipss
Because both Pnx and Pny fall within the compression control region of the corresponding
interaction diagram, φ = 0.65 will be used for evaluation of the section capacity under
biaxial bending.

φPn = 0.65(655) = 426 kips < 440 kips
Column design is therefore not adequate to resist the applied combination of axial force
and biaxial moments.
(c) Using the load contour method, Mox = Pnx ey = 470 ft-kips and Moy = Pny ex =
367 ft-kips. From Eq. (10.20.3), taking a = 1.5 and φ = 0.65
1.5 1.5
 225 1   150 1 
 ×  + × = 1.13 > 1
0.65 470   0.65 367 
Design is therefore not adequate according to the load contour method.

(Continued)
367
2500
2000
1500
Pn (kips)
Pnx = 925 kips

1000
500 in.
6.1 Mox
ey =
0 100 200 300 400 500 600

Mnx (ft-kips)
2500
2000
1500
Pn (kips)
Pny = 1075 kips

1000
in.
.1
=4
500 e x
Moy
0 100 200 300 400 500 600

Mny (ft-kips)
Figure 10.20.9 Determination of Pnx and Pny for section of Example 10.20.2.
(d) A more exact analysis of this section may be made by following the procedure in
Example 10.20.1. For a nominal axial force Pn = Pu /φ = 440/0.65 = 677 kips and a
moment ratio Mnx /Mny = 1.5, the following values are obtained:
c = 14.55 in.
α = 47.3°
M nx = 324 ft-kips
M ny = 217 ft-kips
For this case, the strain in the extreme tension bar εt = 0.0015, which is less than the
yield strain εy. Thus, φ = 0.65. Even if εt had been slightly larger than εy, the use of φ
corresponding to the compression control limit would still be recommended, since using
a larger φ value based on only one bar reaching a strain greater than εy would be ques-
tionable. The design moments for bending about the x- and y-axes are then
φ M nx = 0.65(324) = 211 ft-kips < 225 ft-kips

φ M ny = 0.65(217) = 141 ftt-kips < 150 ft-kips
The section is therefore inadequate, as already demonstrated through the use of the
reciprocal load method.
368
10.21 DESIGN FOR SHEAR
Design of columns for shear follows the procedures outlined in Chapter 5. Because col-
umns are often subjected to significant levels of axial compression, the shear strength con-
tribution attributed to the concrete, Vc, is typically calculated using Eq. (5.13.1), given in
ACI-22.5.6.1, which is repeated below for convenience.
 Nu 

Vc = 2  1 +  λ fc′bw d [5.13.1]
 2000 Ag 
For members with axial tension, Eq. (5.13.6) may be used. The authors recommend, how-
ever, that the concrete contribution to shear strength be ignored (i.e., Vc = 0) when there is
significant axial tension. In applying Eq. (5.13.1) for columns, a question may arise as to
the value of the effective depth d to be used in the calculations. Because the amount of steel
in tension depends on the eccentricity (e = Mu /Pu) being considered, different values of d
may be computed. For example, at very low eccentricities, it is possible that no reinforce-
ment is in tension (i.e., all reinforcement is in compression), whereas at high eccentricities
several layers of reinforcement might be in tension. For circular sections of diameter D,
ACI-22.5.2.2 allows the use of the same shear strength equations used for rectangular sec-
tions, with d = 0.8D and bw = D. However, the ACI Code is silent on the calculation of d
for rectangular columns. In this case, the authors recommend taking d = 0.8h, similar to the
case of circular columns.
Once Vc has been calculated, the amount of transverse reinforcement required for shear
resistance is obtained by
Vu
Vs ≥ − Vc (10.21.1)
φ
where (ACI-22.5.10.5.3 and ACI-22.5.1.2)
Av f yt d
Vs = ≤ 8 fc′ bw d [5.10.7]
s
In Eq. (5.10.7), Av is the area of shear reinforcement within a spacing s, fyt is the yield
strength of the transverse reinforcement, and bw is the web width (for solid rectangular sec-
tions, bw is equal to the section dimension perpendicular to the direction of shear). However,
wherever Vu > 0.5φ Vc , the area of shear reinforcement Av must be at least (ACI-10.6.2),
bw s 50bw s
min Av = 0.75 fc′ ≥ [5.10.8]
f yt f yt
Maximum spacing of shear reinforcement depends on the shear force required to be resisted
by the transverse reinforcement as follows,
For Vs ≤ 4 fc′ bw d , s ≤ d / 2 ≤ 24 in.

For Vs > 4 fc′ bw d , s ≤ d / 4 ≤ 12 in.
In addition to these requirements, transverse reinforcement in tied columns must satisfy the
requirements for lateral support discussed in Section 10.8. For spirally reinforced columns,
spiral reinforcement must satisfy the minimum volumetric ratio and spacing requirements
discussed in Section 10.9.
369
10.21 DESIGN FOR SHEAR 369
EXAMPLE 10.21.1
The column section shown in Fig. 10.21.1(a) is subjected to a factored shear force Vu = 75
kips parallel to the long direction and a factored axial compressive force Pu = 420 kips.
Design the transverse reinforcement such that shear and lateral support requirements are
satisfied. Use fc′ = 4000 psi (normal weight) and fyt = 60 ksi.
10 – #7 bars #3 ties @ 8 in.
2.5”
6.5”
18”
6.5”
2.5”
2.5” 5.5” 6” 5.5” 2.5” 2.5” 5.5” 6” 5.5” 2.5”
22” 22”
(a) Column section (b) Layout of lateral ties
Figure 10.21.1 Cross section and layout of lateral ties for column in Example 10.21.1.
SOLUTION
(a) Determine the shear strength attributed to the concrete (Vc) by using Eq. (5.13.1) and
the required contribution from shear reinforcement. Taking d = 0.8h = 17.6 in.,
 Nu 
Vc = 2  1 +  λ fc′ bw d
 2000 Ag 
 420, 000  1
= 2 1 +  (1.0) 4000 (18)(17.6) 1000
 2000(18)(22) 
= 61.3 kips

V 75 1
Vs ≥ u − Vc = − 61.3 = 38.7 kips < 8 4000 (18)(17.6) = 160 kips
φ 0.75 1000
(b) Determine maximum spacing of shear reinforcement.
1
Vs = 38.7 kips < 4 fc′bw d = 4 4000 (18)(17.6) = 80.1 kips
1000
Thus, s ≤ d/2 = 17.6/2 = 8.8 in. ≤ 24 in. Select s = 8 in.

(c) Verify that transverse reinforcement spacing satisfies requirement for lateral bar
support. Assume #3 ties.
From Section 10.8, maximum tie spacing shall not exceed the smallest of 16 times the
longitudinal bar diameter [16( 7 8 ) = 14 in.)], 48 times the tie diameter [48( 3 8 ) = 18 in.],
or the minimum cross-sectional dimension (18 in.). The selected spacing s = 8 in. is
therefore adequate for lateral bar support.
(d) Determine required area of shear reinforcement. From Eqs. (10.21.1) and (5.10.7),
Vs s (38.7)(8)
Av ≥ = = 0.29 sq in.
f yt d (60)(17.6)
(Continued)
370
Select #3 ties with three legs parallel to the long direction [Av = 3(0.11) = 0.33 sq in.]
by using a closed tie and a crosstie as shown in Fig. 10.21.1(b). The crosstie is anchored
by a 135° hook on one end and a 90° hook on the other end. The closed tie, on the other
hand, may be anchored by 90° hooks. In this example, however, 135° hooks are used
[Fig. 10.21.1(b)].
(e) Verify that the selected reinforcement satisfies minimum shear reinforcement
requirement. From Eq. (5.10.8) and noting that 0.75 fc′ < 50 psi for fc′ = 4000 psi,
50bw s 50(18)(8)
min Av = = = 0.12 sq in. < 0.33sq in. OK
f yt 60, 000
(f) Provide additional ties (if needed) to satisfy requirements for lateral support of
longitudinal bars.
As discussed in Section 10.8, ties shall be arranged such that every corner and alternate
longitudinal bar is laterally supported by the corner of a tie having an included angle of
not more than 135°. In addition, no longitudinal bar shall be farther than 6 in. clear on
either side from such a laterally supported bar. This will require additional ties parallel
to the short direction of the column, to support the two intermediate longitudinal bars
along each long face of the column. A closed tie enclosing the four intermediate longi-
tudinal bars, as shown in Fig. 10.21.1(b), satisfies this requirement. Maximum spacing
for these ties would be 14 in., assuming that shear reinforcement requirements in that
direction do not control. For practical purposes, however, it is convenient to use the same
spacing for all ties. Thus, select a spacing of 8 in.
Final design for column transverse reinforcement is shown in Fig. 10.21.1(b).
SELECTED REFERENCES
Historical
10.1 A. Considère. “Compressive Resistance of Concrete Steel and Hooped Concrete, Part I,”
Engineering Record, December 20, 1902, 581–583; Part II, December 27, 1902, 605–606.
10.2 A. Considère. “Concrete-Steel and Hooped Concrete.” Reinforced Concrete, 1903, p. 119.
10.3 A. N. Talbot. “Tests of Concrete and Reinforced Concrete Columns,” Bulletin, No. 10, 1906, and
No. 20, 1907, University of Illinois, Urbana.
10.4 M. O. Withey. “Tests of Plain and Reinforced Concrete Columns,” Engineering Record, July
1909, 41.
10.5 M. O. Withey. “Tests on Reinforced Concrete Columns,” Bulletin, No. 300, 1910, and No. 466,
1911, University of Wisconsin, Madison.
10.6 ACI Committee 105. “Reinforced Concrete Column Investigation,” ACI Journal, Proceedings,
26, April 1930, 601–612; 27, February 1931, 675–676; 28, November 1931, 157–158; 29,
September 1932, 53–56: 29, February 1933, 275–284; 30, September–October 1933, 78–90; 30,
10.7 Joint ACI-ASCE Committee 441. Reinforced Concrete Columns (annotated bibliography), ACI
Bibliography No. 5. Detroit: American Concrete Institute, 1965.
Confinement, Spirals, and Ties

10.8 M. J. N. Priestley, R. Park, and R. T. Potangaroa. “Ductility of Spirally-Confined Concrete
Columns,” Journal of the Structural Division, ASCE, 107, ST1 (January 1981), 181–202.
10.9 S. H. Ahmad and S. P. Shah. “Stress–
Strain Curves of Concrete Confined by Spiral
Reinforcement.” ACI Journal, Proceedings, 79, November–December 1982, 484–490.
371
10.10 Salvador Martinez, Arthur H. Nilson, and Floyd O. Slate. “Spirally Reinforced High-Strength
Concrete Columns,” ACI Journal, Proceedings, 81, September–October, 1984, 431–442.
10.11 Shamim A. Sheikh and S. M. Uzumeri. “Strength and Ductility of Tied Concrete Columns,”
10.12 B. D. Scott, R. Park, and M. J. N. Priestley. “Stress-Strain Behavior of Concrete Confined by
Overlapping Hoops at Low and High Strain Rates,” ACI Journal, Proceedings, 79, January–
February 1982, 13–27.
10.13 Robert Park, M. J. Nigel Priestley, and Wayne D. Gill. “Ductility of Square-Confined Concrete
Columns,” Journal of the Structural Division, ASCE, 108, ST4 (April 1982), 929–951.
10.14 Apostolos Fafitis and Surendra P. Shah. “Predictions of Ultimate Behavior of Confined Columns
Subjected to Large Deformations,” ACI Journal, Proceedings, 82, July–August 1985, 423–433.
10.15 Jack P. Moehle and Terry Cavanagh. “Confinement Effectiveness of Crossties in RC,” Journal
of Structural Engineering, ASCE, 111, 10 (October 1985), 2105–2120.
10.16 Shamim A. Sheikh and C. C. Yeh. “Flexural Behavior of Confined Concrete Columns,” ACI
Journal, Proceedings, 83, May–June 1986, 389–404.
10.17 Güney Özcebe and Murat Saatcioglu. “Confinement of Concrete Columns for Seismic

Loading,” ACI Structural Journal, 84, July–August 1987, 308–315.
10.18 Yook-Kong Yong, Malakah G. Nour, and Edward G. Nawy. “Behavior of Laterally Confined
High-Strength Concrete under Axial Loads.” Journal of Structural Engineering, ASCE, 114, 2
(February 1988), 332–351.
10.19 Hisham Adbel-Fattah and Shuaib H. Ahmad. “Behavior of Hoop-Confined High-Strength
Concrete under Axial and Shear Loads,” ACI Structural Journal, 86, November–December
1989, 652–659.
10.20 Shamim A. Sheikh and Ching-Chung Yeh. “Tied Concrete Columns under Axial Load and
Flexure,” Journal of Structural Engineering, ASCE, 116, 10 (October 1990), 2780–2800.
10.21 Shamim A. Sheikh and C. C. Yeh. “Concrete Strength in Tied Columns,” ACI Structural
Journal, 87, July–August 1990, 379–385.
10.22 Salim R. Razvi and Murat Saatcioglu. “Confinement of Reinforced Concrete Columns with
Welded Wire Fabric,” ACI Structural Journal, 86, September–October 1989, 615–623.
Interaction Diagrams
10.23 E. Hognestad. “A Study of Combined Bending and Axial Load in Reinforced Concrete

Members,” Bulletin, No. 399, November 1951, Engineering Experiment Station, University of
Illinois, Urbana (117 references).
10.24 ACI-ASCE Committee 327. “Report on Ultimate Strength Design,” Proceedings ASCE, 81,
October 1955, Paper No. 809. See also ACI Journal, Proceedings, 52, January 1956, 505–524.
10.25 A. H. Mattock, L. B. Kriz, and Eivind Hognestad. “Rectangular Concrete Stress Distribution in
Ultimate Strength Design,” ACI Journal, Proceedings, 57, February 1961, 875–928.
10.26 E. O. Pfrang, C. P. Siess, and M. A. Sozen. “Load–Moment–Curvature Characteristics of
Reinforced Concrete Cross Sections,” ACI Journal, Proceedings, 61, July 1964, 763–778.
Disc., 62, 3, 1673–1683.
10.27 Parviz Soroushian and Kienuwa Obaseki. “Strain Rate-
Dependent Interaction Diagrams
for Reinforced Concrete Sections,” ACI Journal, Proceedings, 83, January–February 1986,
108–116.
10.28 Shamim A. Sheikh and C. C. Yeh. “Analytical Moment–Curvature Relations for Tied Concrete
Columns,” Journal of Structural Engineering, ASCE, 118, 2 (February 1992), 529–544.
10.29 Michael C. Head and J. Dario Aristizabal-Ochoa. “Analysis of Prismatic and Linearly Tapered
Reinforced Concrete Columns,” Journal of Structural Engineering, ASCE, 113, 3 (March
1987), 575–589.
10.30 Abi O. Aghayere. “Design of Nonprismatic Concrete Columns by the Degree of Fixity Method,”
ACI Structural Journal, 87, July–August 1990, 473–478.
10.31 M. Ala Saadeghvaziri and Douglas A. Foutch. “Behavior of RC Columns under

Nonproportionally Varying Axial Load,” Journal of Structural Engineering, ASCE, 116, 7 (July
1990), 1835–1856.
10.32 James G. MacGregor, John E. Breen, and Edward O. Pfrang. “Design of Slender Concrete
Columns,” ACI Journal, Proceedings, 67, January 1970, 6–28.
372
10.33 F. E. Richart, J. O. Draffin, T. A. Olson, and R. H. Heitman. “The Effect of Eccentric Loading,
Protective Shells, Slenderness Ratios, and Other Variables in Reinforced Concrete Columns,”
Bulletin No. 368, Engineering Experiment Station, University of Illinois, Urbana, 1947, 130 pp.
10.34 B. Bresler and P. H. Gilbert. “Tie Requirements for Reinforced Concrete Columns,” ACI
Journal, Proceedings, 58, November 1961, 555–570; Disc., 58, 897–907.
10.35 James F. Pfister. “Influence of Ties on the Behavior of Reinforced Concrete Columns,” ACI
Journal, Proceedings, 61, May 1964, 521–537.
10.36 Edwin G. Burdette and Hubert K. Hilsdorf. “Behavior of Laterally Reinforced Concrete
Columns,” Journal of the Structural Division, ASCE, 97, ST2 (February 1971), 587–602.
10.37 S. T. Mau. “Effect of Tie Spacing on Inelastic Buckling of Reinforcing Bars,” ACI Structural
10.38 Koji Sakai and Shamim A. Sheikh. “What Do We Know about Confinement in Reinforced
Concrete Columns? (A Critical Review of Previous Work and Code Provisions),” ACI Structural
10.39 Grant T. Halvorsen and Craig A. Carinci. “Tie Requirements for Prestressed Concrete Columns,”
PCI Journal, 22, July–August 1987, 46–79.
10.40 J. S. Ford, D. C. Chang, and J. E. Breen. “Behavior of Concrete Columns Under Controlled
Lateral Deformation,” ACI Journal, Proceedings, 78, January–February 1981, 3–20.
10.41 Ti Huang. “On the Formula for Spiral Reinforcement,” ACI Journal, Proceedings, 61, March
1964, 351–353. Disc., 61, 9 1241–1248.
10.42 William L. Gamble. “Re-examination of Spiral Steel Requirements,” Concrete International, 8,
November 1986, 33–34.
10.43 Chien-Hung Lin and Richard W. Furlong. “Longitudinal Steel Limits for Concrete Columns,”
ACI Structural Journal, 92, May–June 1995, 282–287.
Design
10.44 Paul W. Reed. “Simplified Analysis for Thrust and Moment of Concrete Sections,” ACI Journal,
Proceedings, 77, May–June 1980, 195–200.
10.45 Curtis J. Young. “Direct Selection of Concrete Dimensions in Columns,” Concrete International,
3, October 1981, 27–31.
10.46 N. J. Everard and Edward Cohen. Ultimate Strength Design of Reinforced Concrete Columns
(SP-7). Detroit: American Concrete Institute, 1964, 182 pp.
10.47 Metric Design Handbook for Reinforced Concrete Elements in Accordance with the Strength
Design Methods of CSA Standard CAN3-A23.3-M77. (Edited by Murat Saatcioglu). Ottawa,
Ontario, Canada: Canadian Portland Cement Association, 1980.
10.48 Joaquin Marin. “Design Aids for L-Shaped Reinforced Concrete Columns,” ACI Journal,
10.49 PCA. Ultimate Load Tables for Circular Columns. Chicago: Portland Cement Association, 1960.
10.50 Bekele Mekonnen. “Reinforced Concrete Column Design Equations,” ACI Journal,

Proceedings, 81, May–June 1984, 242–250.
10.51 CRSI. Interaction Diagrams for Uniaxial Bending Moment and Axial Force on Reinforced
Concrete Circular Sections Using a Programmable Calculator, Structural Bulletin No. 12.
Schaumburg, IL: Concrete Reinforcing Steel Institute, June 1985.
Axial Tension and Bending

10.52 E. C. Harris. “Design of Members Subject to Combined Bending and Tension,” ACI Journal,
Proceedings, 72, September 1975, 491–495.
10.53 F. L. Moreadith. “Design of Reinforced Concrete for Combined Bending and Tension,” ACI
Journal, Proceedings, 75, June 1978, 251–255, Disc., 75, December 1978, 721–723.
10.54 Fernando Villalta, Domingo J. Carreira, and Bryan Erler. Discussion of “Design of Reinforced
Concrete for Combined Bending and Tension,” ACI Journal, Proceedings, 75, December 1978,
721–723.
37
Biaxial Bending
10.55 Boris Bresler. “Design Criteria for Reinforced Columns under Axial Load and Biaxial Bending,”
ACI Journal, Proceedings, 57, November 1960, 481–490. Disc., 1621–1638.
10.56 Richard W. Furlong. “Ultimate Strength of Square Columns under Biaxially Eccentric Loads,” ACI
Journal, Proceedings, 57, March 1961, 1129–1140.
10.57 F. N. Pannell. “Failure Surfaces for Members in Compression and Biaxial Bending,” ACI Journal,
10.58 Alfred L. Parme, Jose M. Nieves, and Albert Gouwens. “Capacity of Reinforced Rectangular
Columns Subject to Biaxial Bending,” ACI Journal, Proceedings, 63, September 1966, 911–923.
10.59 Donald C. Weber. “Ultimate Strength Design Charts for Columns with Biaxial Bending,” ACI
Journal, Proceedings, 63, November 1966, 1205–1320. Disc., 64, 6, Part 2, Vol.64 No.6, Part 2.
1583–1586.
10.60 Anis Farah and M. W. Huggins. “Analysis of Reinforced Concrete Columns Subjected to

Longitudinal Load and Biaxial Bending,” ACI Journal, Proceedings, 66, July 1969, 569–575.
10.61 J. C. Smith. “Biaxially Loaded Concrete Interaction Curve,” Computers and Structures, 3 (1973),
1461–1464.
10.62 K. N. Smith and W. H. Nelles. “Columns Subjected to Biaxial Bending—Preliminary Selection of
Reinforcing,” ACI Journal, Proceedings, 71, August 1974, 411–413.
10.63 Peter D. Heimdahl and Albert C. Bianchini. “Ultimate Strength of Biaxially Eccentrically Loaded
Concrete Columns Reinforced with High Strength Steel,” Reinforced Concrete Columns (SP-50).
Detroit: American Concrete Institute, 1975 (pp. 93–117).
10.64 S. I. Abdel-
Sayed and N. J. Gardner. “Design of Symmetric Square Slender Reinforced
Concrete Columns under Biaxially Eccentric Loads,” Reinforced Concrete Columns (SP-50).
10.65 A. K. Basu and P. Suryanarayana. “Analysis of Restrained Reinforced Concrete Columns under
Biaxial Bending,” Reinforced Concrete Columns (SP-50). Detroit: American Concrete Institute,
1975 (pp. 211–232).
10.66 Albert J. Gouwens. “Biaxial Bending Simplified,” Reinforced Concrete Columns (SP-50).

10.67 W. F. Chen and M. T. Shoraka. “Tangent Stiffness Method for Biaxial Bending of Reinforced
Concrete Columns,” IABSE Publications, International Association for Bridge and Structural
Engineering, Part 35-I, 1975, 23–44.
10.68 Richard W. Furlong. “Concrete Columns under Biaxially Eccentric Thrust,” ACI Journal,

10.69 M. Pinto de Magalhaes. “Biaxially Loaded Concrete Sections,” Journal of the Structural Division,
ASCE, 105, ST12 (December 1979), 2639–2656.
10.70 L. N. Ramamurthy and T. A. Hafeez Khan. “L-Shaped Column Design for Biaxial Eccentricity,”
Journal of Structural Engineering, ASCE, 109, 8 (August 1983), 1903–1917.
10.71 Michael A. Taylor. “Direct Biaxial Design of Columns,” Journal of Structural Engineering, ASCE,
111, 1 (January 1985), 158–173.
10.72 Cheng-Tzu Thomas Hsu. “Biaxially Loaded L-Shaped Reinforced Concrete Columns,” Journal of
Structural Engineering, ASCE, 111, 12 (December 1985), 2576–2595.
10.73 Cheng-Tzu Thomas Hsu. “Reinforced Concrete Members Subject to Combined Biaxial Bending
and Tension,” ACI Structural Journal, 83, January–February 1986, 137–144.
10.74 David A. Ross and J. Richard Yen. “Interactive Design of Reinforced Concrete Columns with
Biaxial Bending,” ACI Structural Journal, 83, November–December 1986, 988–993.
10.75 Cheng-Tzu Thomas Hsu. “Channel-Shaped Reinforced Concrete Compression Members under
Biaxial Bending,” ACI Structural Journal, 84, May–June 1987, 201–211.
10.76 Troels Brøndum-
Nielsen. “Concrete Sections under Biaxial Bending,” Journal of Structural
Engineering, ASCE, 113, 10 (October 1987), 2137–2144.
10.77 Cheng-Tzu Thomas Hsu. “Analysis and Design of Square and Rectangular Columns by Equation
of Failure Surface,” ACI Structural Journal, 85, March–April 1988, 167–179.
10.78 Issam E. Harik and Hans Gesund. “Flowcharts for Biaxial Bending in R/C Tied Columns,” Journal
of Structural Engineering, ASCE, 114, 6 (June 1988), 1230–1249; Disc., 116, 2 (February 1990),
556–557.
10.79 Chu-Kia Wang. “Solving the Biaxial Bending Problem in Reinforced Concrete by a Three-Level
Iteration Procedure,” Microcomputers in Civil Engineering, 3 (1988), 311–320.
10.80 F. A. Zahn, R. Park, and M. J. N. Priestley. “Strength and Ductility of Square Reinforced Concrete
Column Sections Subjected to Biaxial Bending,” ACI Structural Journal, 86, March–April 1989,
123–131.
10.81 Cheng-Tzu Thomas Hsu. “T-Shaped Reinforced Concrete Members under Biaxial Bending and
Axial Compression,” ACI Structural Journal, 86, July–August 1989, 460–468.
374
PROBLEMS
All problems* are to be worked in accordance with the ACI Code unless otherwise indi-
cated, and all stated loads are service loads. Note that eccentricities ex and ey are measured
along the x- and y-axes, respectively (see Fig. 10.20.1). For all design problems a design
sketch (to scale) of the cross section is required: show section dimensions, location and size
of bars, and tie or spiral size and spacing.
Problems on Rectangular Sections Subject to Uniaxial Bending

10.1 Calculate the nominal strength Pn for the section x
of the figure for Problem 10.1 for an eccentricity
ey (measured along the y-axis) = 0.1h = 1.8 in. 3”
Use fc′ = 3000 psi and fy = 40,000 psi. Use basic
4”
principles of statics to obtain solution, consid-
ering the effect of compression concrete dis-
(500 mm)
placed by steel; compare the result with the
20”
y 6”
maximum Pn given by ACI-22.4.2 (e = 46 mm;
10 – #11
fc′ = 20 MPa; fy = 300 MPa). (10 – #36M)
4”
3”
#3 @ 18
3” 9” 9” 3” (66 mm)
(#10M)
(234 mm) (234 mm)
24”
6 – #9 (600 mm)
18” y
(6 – #29M) (460 mm) Problems 10.3, 10.4, 10.5, and 10.6
bending with respect to the x-axis (h = 24 in.).

To obtain points for the diagram, compute Pn
for the following cases in addition to eb (see
3” 12” 3” (75 mm) Problem 10.3):
(310 mm)
18” (a) e = 0 (P0) (d) e = 0.7h
(460 mm)
(b) e = 0.1h (e) e = h
(c) e = 0.3h (f) e = ∞ (M0)
10.5 Same as Problem 10.3 except use fc′ = 4000 psi
10.2 Same as Problem 10.1 except fc′ = 5000 psi and and fy = 60,000 psi. ( fc′ = 3000 MPa;
fy = 60,000 psi. ( fc′ = 35 MPa; f y = 400 MPa.) f y = 400 MPa.)
10.3 Compute the nominal strength Pn = Pb for the
10.6 Same as Problem 10.4 except use fc′ = 4000 psi
balanced strain condition for bending about the
and fy = 60,000 psi. (For eb, see Problem 10.5.)
strong axis (measured as ey) for the column of
10.7 For the section of the figure for Problem 10.7,
the figure for Problem 10.3. Compute also the
use statics to compute and plot the strength
eccentricity eb. Use basic statics, including the
interaction diagram of Pn–Mn for bending about
effect of the compression concrete displaced
the x-axis. Compute the balanced condition in
by steel, fc′ = 3000 psi and fy = 40,000 psi.
addition to those points indicated for Problem
( fc′ = 20 MPa; f y = 300 MPa.) 10.4. Use fc′ = 3000 psi and fy = 60,000 psi.
10.4 Using basic statics and taking into account con-
( fc′ = 20 MPa; f y = 400 MPa.)
crete displaced by steel in compression, cal-
culate and plot the Pn–Mn strength interaction 10.8 Repeat Problem 10.7 except consider bending
diagram for the section of Problem 10.3. Take about the y-axis.
* Many problems may be solved either as problems stated in Inch-Pound units or as problems in SI units using quantities
in parentheses at the end of the statement. The SI conversions are approximate to avoid implying higher precision for the given
information in metric units than that for the Inch-Pound units.
375
PROBLEMS 375
10.9 For the section of Problem 10.7, compute the 10.11 For the section of the figure for Problem
nominal strength Pn for an eccentricity ey of 5 10.11, compute the nominal strength Pn for
in. with respect to the x-axis. Use fc′ = 5000 psi the eccentricity given below (as assigned by
and fy = 60,000 psi. (e = 125 mm; fc′ = 35 MPa; the instructor) with respect to the x-axis. Use
f y = 400 MPa.) fc′ = 3000 psi and fy = 40,000 psi. ( fc′ = 20 MPa;
fy = 300 MPa.)
#3 ties @ 14” (a) e = 6 in. (150 mm) (e) e = 30 in. (750 mm)
x (#10M ties)
2.5” (66 mm) (b) e = 8 in. (200 mm) (f) e = 40 in. (1000 mm)
1
1 2 ” clear
cover 4.5” (109 mm) (c) e = 12 in. (300 mm) (g) e = 200 in. (5 m)
(40 mm) y 4 – #10 14” (350 mm)
(4 – #32M) 4.5” (d) e = 20 in. (500 mm) (h) e = 2000 in. (50 m)
2.5”
10.12 For the section of the figure for Problem
2.5” 6.5” 6.5” 2.5”(66 mm) 10.12, compute the nominal strength Pn for
(164 mm)
18” the eccentricity given below (as assigned by
(460 mm)
the instructor) with respect to the x-axis. Use
Problems 10.7, 10.8, 10.9, 10.10, 10.39, and 10.40 fc′ = 3000 psi and fy = 40,000 psi.
(a) e = 2 in. (e) e = 40 in.
10.10 For the section of Problem 10.7, compute the
(b) e = 3.6 in. (f) e = 70 in.
nominal strength Pn. Repeat Problem 10.9,
except use an eccentricity ey of 22 in. with (c) e = 10 in. (g) e = 200 in.
respect to the x-axis. (ey = 560 mm.)
(d) e = 15 in. (h) e = 1000 in.
(66 mm) 3”
(117 mm) 4 12 ” 16 – #10 bars

(16 – #32M bars)
1
4 2”
24”
(600 mm) y
1
4 2”
1
4 2”
3”
(66 mm) 3” 7” 10” 10” 7” 3”

(184 mm) (250 mm)
40”
(1000 mm)
Problem 10.11 and 10.46
3”
1
42” 14 – #9
1
42”
24” y
1
42”
1
42”
3”
1 1
22 ” 10” 11” 10” 22 ”
36”
Problem 10.12
376
10.13 For the section of the figure for Problem 10.13, 10.15 For the case assigned by the instructor, design
use basic statics to compute and plot the strength the smallest (in whole inches) square tied col-
interaction diagram of Pn–Mn for bending about umn having not more than 4% reinforcement
the x-axis. Compute the balanced condition, to carry an axial compression (i.e., M = 0 or
in addition to enough other points to plot the neglected). Compute the strength Pn when e =
curve. (Hint: By successively setting the neu- 0.10h. Use fc′ = 3500 psi and fy = 60,000 psi.
tral axis distance c at specific values, the points ( fc′ = 25 MPa; f y = 400 MPa.)
can be computed without solving a cubic equa-
PD (dead load) PL (live load)
tion.) Use fc′ = 4000 psi and fy = 60,000 psi.
( fc′ = 30 MPa; fy = 400 MPa.) (a) 100 kips (450 kN) 80 kips (350 kN)
(b) 200 kips (900 kN) 160 kips (700 kN)
12 – #9 (12 – #29M)
x (c) 350 kips (1500 kN) 150 kips (700 kN)
(d) 400 kips (1800 kN) 320 kips (1400 kN)
(e) 600 kips (2700 kN) 480 kips (2100 kN)
12” (300 mm)
10.16 Repeat Problem 10.15, except use fc′ = 4000 psi
and fy = 60,000 psi. ( fc′ = 30 MPa and
fy = 400 MPa.)
10.17 Repeat Problem 10.15, except use fc′ = 3500 psi
and fy = 40,000 psi. ( fc′ = 25 MPa and
fy = 300 MPa.)
1 1 (64 mm) 10.18 For the case assigned by the instructor, design
22 ” 3” 3” 3” 22 ”
a square tied column having symmetrical
reinforcement of about 3 1 2 % to carry service
(75 mm)
14” loads as given. Use basic statics to check final
(350 mm) answer. Use whole inches with fc′ = 4000 psi
and fy = 50,000 psi.
PD PL MD ML
(dead load) (live load) (dead load) (live load)
10.14 Using basic principles, compute the nominal
strength Pn for an eccentricity ey (as assigned (a) 100 kips 90 kips 40 ft-kips 30 ft-kips
by the instructor) on the column of the figure (b) 200 kips 175 kips 75 ft-kips 65 ft-kips
for Problem 10.14. Use fc′ = 3500 psi and
fy = 40,000 psi. ( fc′ = 25 MPa; fy = 300 MPa.) (c) 300 kips 270 kips 120 ft-kips 100 ft-kips
(a) e = 3 in. (75 mm) (f) e = 30 in. (750 mm) (d) 300 kips 270 kips 250 ft-kips 200 ft-kips
(b) e = 6 in. (150 mm) (g) e = 60 in. (1.5 m) (e) 300 kips 270 kips 500 ft-kips 400 ft-kips
(c) e = 9 in. (220 mm) (h) e = 150 in. (3.8 m) (f) 600 kips 550 kips 125 ft-kips 100 ft-kips
(d) e = 15 in. (380 mm) (i) e = 300 in. (7.5 m) (g) 600 kips 550 kips 250 ft-kips 200 ft-kips
(e) e = 20 in. (500 mm) (j) e = 2000 in. (50 m) (h) 600 kips 550 kips 500 ft-kips 400 ft-kips
εy
#4 ties
(#13M) Pn
16” 2 – #11 4 – #7
y
(420 mm) (2 – #36M) (4 – #22M)
1 cover
12 ” (40 mm)
2.70” 12.30” 12.50” 2.5”

(68 mm) (312 mm) (318 mm) (62 mm)
30”
(760 mm)
Problem 10.14
37
PROBLEMS 377
10.19 Redesign the column of Problem 10.18 using a

Ds = 14.87”
rectangular section not over 15 in. wide instead 1
1 2 ” cover
of a square section. If the depth h would be
greater than 30 in., use h/b = 2.0 and exceed
the 15 in. limit.
#4 spiral
10.20 Design a square tied column with symmetrical 10 – #9 36º
reinforcement of about 4% to carry a dead load
of P = 225 kips and M = 220 ft-kips and a live
load of P = 200 kips and M = 190 ft-kips. Use
basic statics to check final answer. Use whole
inches (20 mm for SI) for size with fc′ = 4000 psi
and fy = 40,000 psi. (DL: P = 1000 kN, 2.30”
M = 300 kN·m; LL: P = 900 kN, M = 260 kN·m; 6.01”
fc′ = 30 MPa ; fy = 300 MPa.) 7.44”
10.21 Redesign the column of Problem 10.20 using
20”
a rectangular section not over 16 in. (400 mm)
wide instead of a square section.
d = 17.44”
10.22 Design a square tied column with about 4% rein-
c = 11.08”
forcement to carry a dead load of P = 250 kips
and M = 90 ft-kips, and a live load of P = 185 εy = 0.00172 0.003
kips and M = 70 ft-kips. Use whole inches
(20 mm for SI) for size with fc′ = 4000 psi and
fy = 60,000 psi. (DL: P = 1100 kN, M = 120 kN·m; Pb
eb
LL: P = 820 kN, M = 95 kN·m; fc′ = 30 MPa;
fy = 400 MPa.)
10.23 Design a square tied column with symmetrical
reinforcement of about 2% to carry a dead load a = 9.41”
of P = 25 kips and M = 125 ft-kips and a live load
of P = 10 kips and M = 50 ft-kips. Use whole 0.85fc’ = 3.4 ksi
inches (20 mm for SI) for size with fc′ = 4000 psi
T1 T2 T3
and fy = 60,000 psi. (DL: P = 110 kN, Cs3 C Cs1
M = 170 kN·m; LL: P = 45 kN, M = 68 kN·m; x
s2
Cc
fc′ = 30 MPa; fy = 400 MPa.)
10.24 Redesign the column of Problem 10.23 as a Problem 10.27, 10.28, and 10.31
rectangular column having a depth-width ratio
of between 1.5 and 2.0 along with unsymmet-
rical reinforcement with respect to the bending distribution as for beams. Use fc′ = 4000 psi
axis. and fy = 50,000 psi.
10.25 Design a square tied column with symmetrical 10.28 Determine the nominal compressive strength Pn
reinforcement of about 2 1 2 % to carry a dead for the section of Problem 10.27 for an eccen-
load of P = 75 kips and M = 125 ft-kips and tricity e = 5 in. by the direct application of stat-
a live load of P = 30 kips and M = 50 ft-kips. ics. Use fc′ = 4000 psi and fy = 50,000 psi.
Use whole inches (20 mm for SI) for size with 10.29 Compute the nominal strength Pn of the spi-
fc′ = 4000 psi and fy = 60,000 psi. (DL: P = 340 rally reinforced column of the figure for
kN, M = 170 kN·m; LL: P = 140 kN, M = 70 Problem 10.29 when the loading has an eccen-
kN·m; fc′ = 30 MPa; fy = 400 MPa.) tricity e = 0.05h = 0.9 in. Use fc′ = 3000 psi
10.26 Redesign the column of Problem 10.25 using a and fy = 40,000 psi. Use basic statics with the
rectangular member but still using symmetrical circular section, including the effect of com-
reinforcement. pression concrete displaced by steel. Compare
with maximum Pn obtained from ACI-22.4.2.
(e = 23 mm; fc′ = 20 MPa; fy = 300 MPa.)
Problems on Circular Sections 10.30 Same as Problem 10.29, except use
10.27 Determine the strength Pn = Pb and the eccen- fc′ = 5000 psi and fy = 60,000 psi. ( fc′ = 35 MPa;
tricity eb for the balanced strain condition on the fy = 400 MPa.)
section of the figure for Problem 10.27 by using 10.31 Use a direct application of statics to determine
the method of statics and the rectangular stress the nominal compressive strength Pn for the
378
1
10.37 Redesign the column of Problem 10.25 as a
1 2 ” cover
spirally reinforced circular column.
(40 mm)
#3 spiral
(#10M) Problems on Biaxial Bending and
6 – #8 Compression
(6 – #25M)
10.38 Compute the nominal strength Pn for the column
of Problem 10.3 under an eccentricity ey of 8 in.
about the strong axis and an eccentricity ex of 5 in.
about the weak axis. Compute strength using (a)
the Bresler reciprocal load method and (b) stat-
18” ics and compatibility. (Hint: for part (a), use the
(460 mm) interaction curve of Problem 10.4 and calculate a
similar interaction curve for the weak axis.) Use
fc′ = 3000 psi and fy = 40,000 psi. (ey = 200 mm;
ex = 130 mm; fc′ = 20 MPa ; fy = 300 MPa.)
section of Problem 10.27 for an eccentricity e = 10.39 Determine the adequacy of the rectangular
20 in. Use fc′ = 4000 psi and fy = 50,000 psi. tied column section of Problem 10.7 when
10.32 For the column of the figure for Problem subjected to an axial load of 90 kips, applied
10.32, determine the nominal strength Pn at with an eccentricity ey of 5 in. with respect to
an eccentricity of 21 in. Use fc′ = 4000 psi the x-axis and with an eccentricity ex of 3 in.
and fy = 50,000 psi. Apply the basic princi- with respect to the y-axis. Use fc′ = 3000 psi
and fy = 60,000 psi (use information developed
in Problems 10.7 and 10.8). Compare results
1
1 2 ” cover using (a) the Bresler reciprocal load method
(40 mm) and (b) statics and compatibility.
10.40 Repeat Problem 10.39, except the axial load
#5 spiral is 45 kips applied with eccentricities of 10 in.
(#16M)
16 – #14 with respect to the x-axis and 6 in. with respect
(16 – #43M) to the y-axis.
10.41 For the case assigned by the instructor, compute
the nominal strength Pn for biaxial bending and
compression for the section of Problem 10.7.
Use both the Bresler reciprocal load method
and statics and compatibility. (If available, the
28” strength interaction diagrams from Problems
(720 mm) 10.7 and 10.8 may be used.) Use fc′ = 3000 psi
and fy = 60,000 psi.
(a) ey = 1.8 in. ex = 1.4 in.
ples of statics. (e = 530 mm; fc′ = 30 MPa; (b) ey = 5.4 in. ex = 1.4 in.
fy = 350 MPa.)
(c) ey = 15 in. ex = 2 in.
10.33 Same as Problem 10.15, except use a spirally
reinforced circular column. (d) ey = 30 in. ex = 3 in.
10.34 Redesign the column of Problem 10.18 as a
(e) ey = 40 in. ex = 6 in.
10.35 Redesign the column of Problem 10.20 as a (f) ey = 40 in. ex = 20 in.
10.42 Design a square tied column with about 4%
10.36 Design a circular spirally reinforced column
reinforcement uniformly distributed around
for the conditions given in Problem 10.22.
its sides. The loads are dead load: P = 225
Make statics check using the designed circular
kips, Mx = 236 ft-kips, My = 108 ft-kips, and
section.
379
PROBLEMS 379
live load: P = 200 kips, Mx = 204 ft-kips, Verify that lateral support requirements for lon-
My = 64 ft-kips. Select sizes in whole inches gitudinal bars are satisfied. Use fyt = 60 ksi.
that are multiples of 2, using fc′ = 4000 psi and 10.46 For the section of the figure for Problem
fy = 40,000 psi. 10.11, design the shear reinforcement for
10.43 Repeat Problem 10.42, except use Vu = 215 kips parallel to the short direction
fy = 60,000 psi. of the column. Verify that lateral support
10.44 Repeat Problem 10.42, except reduce moments requirements for longitudinal bars are satis-
Mx about the x-axis to 215 ft-kips dead load fied. There is no need to use the same trans-
and 162 ft-kips live load, and increase fy to verse reinforcement layout shown in the
60,000 psi. figure. Use fyt = 60 ksi.
10.47 For the section of the figure for Problem 10.32,
determine the required spiral pitch if the sec-
Problems on Shear tion is subjected to a shear force Vu = 165 kips.
10.45 For the section of the figure for Problem 10.13, Verify that the final spiral pitch selected satis-
design the shear reinforcement for Vu = 42 kips fies the requirements for spiral reinforcement
parallel to the long direction of the column. in the ACI Code. Use fyt = 50 ksi.
CHAPTER 11
MONOLITHIC BEAM-C OLUMN
CONNECTIONS
11.1 INTRODUCTION
Chapters 3 through 10 emphasized the design of flexural members (beams) and members
subjected to combined axial force and bending (columns) for bending, shear, and develop-
ment of reinforcement. Some attention has been given to development of reinforcement at
exterior supports, including the use of hooks (Section 6.10). Often in design, not enough
attention is given to the details of connections—how the forces in beams and columns
interact and get transmitted through the joint—that is, the column region over the depth of
the deepest beam framing into the column. A connection, on the other hand, consists of the
joint and adjacent beam, slab, and column regions.
The ACI Code (Chapter 15) provides little guidance specifically directed to detailing
of beam-column joints in regions of no or low seismicity, for which design is typically
governed by gravity and wind loads. On the other hand, provisions for shear strength and
detailing of joints in special moment frames, for which large shear reversals are expected
under a design-level earthquake, are provided in Section 18.8 of the Code.
The state of the art regarding the design of beam-column connections is summarized by
ACI-ASCE Committee 352, Joints and Connections in Monolithic Reinforced Concrete
Structures [11.1]. That report contains detailed provisions for the design of two classes of
beam-column connections:
•
Type 1 connections, primarily for members designed to satisfy ACI Code strength
requirements, where no significant inelastic deformations are expected.
•
Type 2 connections, usually for members subject to earthquake loading, where there is
need for sustained strength under deformation reversals into the inelastic range.
In general, Type 1 connections require only nominal ductility, typical of connections in

a continuous moment-resisting frame structure designed to resist gravity and wind loads.
The primary source of what follows is from the 2002 ACI-ASCE Committee 352 report
[11.1]. These recommendations, which apply to beam-column connections of normal-
weight concrete with strengths up to 15,000 psi, include connections where the beam width
is larger than the column depth, often called wide-beam connections. The recommenda-
tions also apply to eccentric connections (i.e., where the beam centerline does not pass
through the column centroid) provided that all beam reinforcement is anchored or passes
through the column core.
Requirements for transverse reinforcement in Type 2 beam-column connections dis-
cussed in Section 11.3 are those in Section 18.8 of the 2014 ACI Code; these have been
381
11.2 BEAM-COLUMN JOINT ACTIONS 381
Beam-column connection reinforcement. (Photo courtesy of Cary Kopczynski & Co.)
modified with respect to previous editions of the code and may in certain situations be more
stringent than those in the ACI-ASCE Committee 352 recommendations.
The following discussion of some general concepts incorporates three examples. A more
detailed treatment of the subject is outside the scope of this book. Also, some of the more
readily available references are listed at the end of this chapter.
11.2 BEAM-C OLUMN JOINT ACTIONS

Just like the members themselves, beam-column joints need to be designed for all actions
that may act on them: axial load, bending moment, shear, and torsion, caused by external
loads and other actions, such as ground motions, wind, creep, shrinkage, temperature, or set-
tlement of supports. Assuming that the members have themselves been properly designed,
the critical factor in joint design is the transmission into and through the joint of the forces
that are present at the ends of the members. Figure 11.2.1 shows an interior connection with
beams framing into the joint from all sides of a column (the slab is not shown for clarity).
Figure 11.2.2 shows a beam-column connection region of a frame subjected to lateral
forces such that beams framing from opposite sides of the column are subjected to nega-
tive bending and positive bending (right and left beam in Fig. 11.2.2, respectively). The
scenario in which both beams are subjected to negative bending, as is the case in a frame
subjected to gravity loads only, is of less interest because little or no shear is introduced
into the joint.
Referring to Fig. 11.2.2, the forces T1 and C1 are the tension and compression resultants
for negative bending in the beam framing into the joint from the right side; the forces C2
and T2 represent positive bending in the beam framing in from the left side; the forces Vcol
represent the shears in the column just above and below the joint. The horizontal shear
within the joint that potentially may cause the diagonal crack shown may be expressed as
Vu ( joint ) = T1 + C2 − Vcol = T1 + T2 − Vcol (11.2.1)

382
382 C hapter 1 1 M onolithic B eam - C olumn C onnections
or
Vu ( joint ) = α f y Ast + α f y Asb − Vcol (11.2.2)
For Type 1 joints, ACI-ASCE Committee 352 [11.1] recommends that joint shear
demand be computed based on the flexural strength of the beams or the columns, whichever
applies, without strength reduction factors [11.1, Sections 3.1 and 3.3.4.]. For Type 2 joints,
on the other hand, flexural yielding is expected to occur primarily in the beams, and thus
joint shear demand is determined under the assumption that the beams reach their flexural
strength. Thus, T1 and T2 in Eq. (11.2.1) are computed as α fy Ast and α fy Asb rather than the
forces calculated from the bending moment determined by structural analysis at that sec-
tion. The parameter α is a stress multiplier based on the degree of inelastic deformation
expected as follows:
α ≥ 1.0 for Type 1 connections

α ≥ 1.25 for Type 2 connections
A value of 1.0 is permitted for Type 1 connections because only limited inelastic deforma-
tion is expected in the adjacent members. For Type 2 connections, however, a minimum
value of 1.25 is recommended because of the significant inelastic deformations expected in
the members framing into this type of joint. The value of α = 1.25 is intended to account for
the actual yield stress of the bar (typically 10–25% higher than the nominal value) and for
possible strain hardening of the bars at large plastic rotations [11.1, Section 3.3.4].
Note that Ast in Eq. (11.2.2) should include the beam reinforcement plus any effective
flange reinforcement. For Type 1 connections, the effective flange width should be taken as
that prescribed for flanges in tension in accordance with ACI-24.3.4 [11.1, Section 3.3.1].
For Type 2 joints, the effective flange width is based on that prescribed for flanges in com-
pression in ACI-6.3.2.1. However, additional restrictions exist depending on the classifica-
tion of the connection—interior, exterior, or corner connection [11.1, Section 3.3.2].
Pu
Mux
Muy Vux
Vuy
Mux
Vuy
Vux Muy
Pu
Figure 11.2.1 Actions on members framing into a joint.

38
11.3 JOINT TRANSVERSE REINFORCEMENT 383
Column steel Column

steel
Ast
Vcol T1 = Ast αfy s/2
C 2 = T2
Vu (joint) s
Beam
s
steel
T2 = Asbαfy C 1 = T1
Shear crack
Vcol s/2 Ash
Asb
α ≥ 1.0 for Type 1 joints Ties at less than s/2 from

α ≥ 1.25 for Type 2 joints resultant T or C force
in beam are probably not
effective
(a) (b)
Figure 11.2.2 Horizontal shear and area of transverse reinforcement in a beam-column joint (column transverse steel not
shown for clarity).
11.3 JOINT TRANSVERSE REINFORCEMENT

The primary roles of transverse reinforcement in beam-column joints are to provide con-
finement to the concrete, to serve as tie reinforcement for joint shear resisted through a truss
mechanism (see Section 11.4), and to provide lateral support to longitudinal column bars.
Below is a summary of transverse reinforcement requirements for Type 1 and Type 2 joints.
For Type 1 joints, these requirements are primarily based on the ACI-ASCE Committee 352
recommendations [11.1], supplemented with requirements in Chapter 15 of the ACI Code.
For Type 2 joints, transverse reinforcement requirements are primarily those of Section
18.8 of the ACI Code, which have changed with regard to the 2002 ACI Code, to which the
latest ACI-ASCE Committee 352 recommendations refer.
Type 1 Joints
For a Type 1 joint, the transverse reinforcement (ties or spiral) may be omitted when
“adequate” lateral confinement is provided by beams framing into the joint. ACI-ASCE
Committee 352 [11.1, Section 4.2.1.4] has defined confinement as shown diagrammatically
in Fig. 11.3.1. The joint region within the depth of the shallowest member framing into a
column may be considered adequately confined when beams frame in from all four sides
and each beam has a width b at least three-fourths the column width and no more than 4 in.
of column width is exposed on each side of the beam.
Where beams frame in from only two opposite sides, and their widths cover at least three-
fourths of the column width at those sides, and not more than 4 in. of column width is exposed
on each side of the beams, confinement occurs only at those faces. In this case, transverse rein-
forcement (Fig. 11.3.2) is needed through the joint only in the direction parallel to the two sides
of the joint into which the beams frame. This provision has no effect when confinement is from
a spiral or circular hoop, since to provide confinement the spiral or circular hoop must extend
through the joint, thus providing full confinement (all directions) to the joint in any case.
When ties must be provided over the joint depth, at least two layers of transverse rein-
forcement should be used between the top and bottom levels of beam reinforcement in the
deepest member framing into the joint; the center-to-center spacing of tie or circular, con-
tinuously wound bar or wire,1 sh, should not exceed 12 in. [11.1, Section 4.2.1.3]. When the
1 ACI-
25.7.3 has specific requirements in order for a circular, continuously wound bar or wire to be classified as a spiral.
Either a spiral or a circular continuous wound bar or wire not satisfying ACI-25.7.3 may be used in joints. The latter, however,
must be treated as a circular hoop.
384
hc
bc ≥ 34 hc
≥ 34 bc
Beam
Joint B2
(shaded)
b2
Beam B1 b1
x
≤ 4”
Shallowest ≤ 4”
member Note: Not more than 4 in.
of column face width exposed
at any edge
Figure 11.3.1 Confinement at a joint by members framing into all four sides of the column.
Column
Required transverse
reinforcement
(cross-ties shown)
Required transverse
reinforcement must be
used through core at
≤ 4” spacing along column
axis not to exceed 12 in.
or one-half the depth of
shallowest beam for
Type 1 joint.
Figure 11.3.2 Confinement by members framing into two opposite faces of the column only.
joint is part of the nonseismic lateral load resisting system, the spacing should not exceed
6 in. In no case, however, should the spacing exceed one-half the depth of the shallowest
beam framing into the column (ACI-15.4.2.2). Also, ACI-15.4.2 requires a minimum area
of tie reinforcement within a tie spacing sh as follows:
bc sh bs
Ash = 0.75 fc′ ≥ (50 psi) c h (11.3.1)
f yt f yt
with sh not greater than one-half the depth of the shallowest beam framing into the column
(ACI-15.4.2.2).
385
11.3 JOINT TRANSVERSE REINFORCEMENT 385
A Hooked Vertical transverse

Free horizontal face bars reinforcement
Vertical ≥Ld
transverse
reinforcement
Section A–A
A
Elevation
Figure 11.3.3 Transverse reinforcement details in joints with discontinuous columns. (Adapted

from Fig. 4.2 of ACI-ASCE Committee 352 [11.1].)
When required, the joint transverse reinforcement should satisfy the provisions of the
ACI Code for tied or spirally reinforced columns (ACI-10.7.6; see Sections 10.8 and 10.9)
in addition to the joint confinement recommendations presented above.
For Type 1 joints with a free horizontal face at the discontinuous end of a column (such
as at the roof or at a mezzanine level) and with discontinuous beam reinforcement near the
free horizontal face (knee joints), vertical transverse reinforcement is required through the
full height of the joint. At least two layers of vertical transverse reinforcement should be
provided between the outermost longitudinal column bars at a spacing not exceeding 12 in.,
or 6 in. when the joint is part of the nonseismic lateral load resisting system [11.1, Section
4.2.1.5]. An example of the transverse reinforcement details at a discontinuous end of a
column is shown in Fig. 11.3.3.
Type 2 Joints
For Type 2 joints, a minimum amount of transverse reinforcement is required regardless of
whether “adequate” confinement is provided by beams framing into the joint. The trans-
verse reinforcement requirements below are those given for earthquake-resistant structures
in ACI-18.8.3.1.
For columns confined by spirals or circular hoops with factored axial load Pu ≤ 0.3 Ag fc′
and fc′ ≤ 10, 000 psi, the volumetric ratio of transverse reinforcement ρs must be at least
the greater of
 Ag  f′
ρs = 0.45  − 1 c (11.3.2)
 ch  f yt
A
and
fc′
ρs = 0.12 (11.3.3)
f yt
For columns with factored axial load Pu > 0.3 Ag fc′ or fc′ > 10, 000 psi, the volumetric ratio
ρs of spiral or circular hoop reinforcement must be at least the greatest obtained from Eqs.
(11.3.2), (11.3.3), and (11.3.4).
 P 
ρs = 0.35k f  u  (11.3.4)
 f yt Ach 
where
fc′
kf = + 0.6 ≥ 1.0 (11.3.5)
25, 000
386
and
Ag = gross cross-sectional area of compression member
Ach = area of rectangular core measured out-to-out of transversse reinforcement

f yt = yield stress of transverse reinforcement, but not to be taken greater than 100,000 psi
When rectilinear hoops and cross-ties are used as transverse reinforcement in accor
dance with ACI-18.7.5.2, the total required cross-sectional area of hoops and cross-ties
within a spacing sh, Ash (see Fig. 11.3.4), for columns with factored axial load Pu ≤ 0.3 Ag fc′
and fc′ ≤ 10, 000 psi, is
fc′  Ag 
Ash = 0.3sh bc  − 1 (11.3.6)
f yt  Ach 
but not less than
fc′
Ash = 0.09 sh bc (11.3.7)
f yt
where bc is the cross-sectional dimension of the column core measured to the outside edges
of confining reinforcement, perpendicular to the transverse reinforcement area Ash being
calculated.
For columns with factored axial load Pu > 0.3 Ag fc′ or fc′ > 10, 000 psi, area of transverse
reinforcement must be at least the largest from Eqs. (11.3.6), (11.3.7), and (11.3.8)
 P 
Ash = 0.2 k f kn  u  sh bc (11.3.8)
 f yt Ach 
where kf is computed from Eq. (11.3.5) and
nl
kn = (11.3.9)
nl − 2
where nl is the number of longitudinal bars laterally supported by the corner of the hoop or
by a seismic hook. In ACI 318, a seismic hook is defined as a hook with a bend no less than
135° and an extension of six bar diameters, but not less than 3 in.
Transverse reinforcement must be arranged such that longitudinal column bars later-
ally supported by the corner of a hoop leg or a cross-tie are no farther apart than a distance
hx = 14 in. around the perimeter of the column, except for columns with Pu > 0.3 Ag fc′ or
fc′ > 10, 000 psi, where each longitudinal bar must be laterally supported by the corner of a
hoop or by a seismic hook and spaced no farther apart than hx = 8 in.
The center-to-center spacing sh of the transverse reinforcement should not exceed the
following:
a. For joints of special moment frames (ACI-18.7.5.3): 1 4 of the minimum column

dimension, 6 times the diameter of the smallest longitudinal bar, or so
where
  14 − hx  
4 in. ≤  so = 4 +  ≤ 6 in. (11.3.10)
  3  
and hx is taken as the largest distance between laterally supported longitudinal bars
around the column perimeter.
387
11.4 JOINT SHEAR STRENGTH 387
Direction of shear being considered
#3
sh
bc
Ash = 4(0.11) = 0.44 sq in.

dh bc = outside-to-outside of hoops
dh = outside-to-outside of hoops
Ach = bcdh
Figure 11.3.4 Example of rectilinear hoop reinforcement in height sh of joint core.
b. For joints that connect members that are not part of the primary seismic-force-resisting
system, but such members are designed to sustain deformations in the inelastic range
due to lateral displacement compatibility [11.1, Section 4.2.2.4], center-to-center spac-
ing should not exceed 1 3 of the minimum column dimension, or 12 in.
The amount of transverse reinforcement may be reduced by one-half and maximum
spacing increased to 6 in. when the joint is completely confined on all sides by beams
whose width is at least three-fourths the column width (ACI-18.8.3.2).
The ACI Committee 352 report [11.1] contains additional recommendations for beam
bars closest to the free horizontal face in Type 2 joints with a discontinuous column. These
recommendations are based on experimental tests on knee joints [11.47, 11.50, 11.60] and
are similar to those described for Type 1 joints. Because of the inelastic behavior of Type 2
joints, the requirements are, however, more stringent [11.1, Section 4.2.2.8].
11.4 JOINT SHEAR STRENGTH

In a well-designed beam-column joint, shear resistance is assumed to be provided by
a diagonal strut mechanism and a truss mechanism (Fig. 11.4.1). The diagonal strut is
activated by the compression zones of the beams and columns framing into the joint.
The truss mechanism, on the other hand, relies on force transfer to the joint through
bond between the longitudinal beam and column bars and the surrounding concrete.
Transverse reinforcement thus enhances joint shear strength by providing confinement
to the diagonal strut, as well as resisting horizontal tensile forces as part of the truss
mechanism.
As in any reinforced concrete design, design for shear strength must satisfy
φ Vn ≥ Vu (11.4.1)
where Vu is calculated according to Eq. (11.2.2) and φ = 0.75.

It is not possible to separate the contributions to joint shear strength from the strut and
truss mechanisms. Thus, the approach taken by ACI Committee 352 (and ACI Committee
38
Strut
(a) (b)
Figure 11.4.1 Shear-resisting mechanisms in a beam-column joint: (a) diagonal strut (a) and (b) truss.
318) is to define nominal joint shear strength Vn as a function of fc′ as follows [11.1,
Section 4.3.1],
Vn = γ fc′ b j h (11.4.2)
where bj = effective joint width (see Fig. 11.4.2)
 bbeam + bcol
 ,
2

 mh
= smallest of bbeam + ∑ , or
 2
b
 col
The term mh / 2 should not be larger than “the extension of the column beyond
the edge of the beam.” Also, the summation term applies on “each side of the
joint where the edge of the column extends beyond the edge of the beam.”
bbeam = design beam width in the direction of loading. When there is a beam on only one
face of the column in the direction of the load, bbeam is the width of that beam.
When there are beams on opposite faces of the column, bbeam is the average of the
widths of those beams.
bcol = column width perpendicular to the direction of the load being considered
h = depth of column in the direction of the load being considered
m = coefficient that accounts for the joint eccentricity
= 0.3 when the eccentricity between the beam centerline and the column centroid
exceeds bcol / 8
= 0.5 for all other cases
γ = constant that depends on joint type and classification, as shown in Table 11.4.1.
389
1 1 . 5 C O L U M N - TO - B E A M M O M E N T S T R E N G T H R AT I O 389
bbeam + bcol
,
2
bj ≤ bbeam + Σ mh , or
2
bcol
bj
bj
h Column Column h
Direction
bbeam of e
loading
bcol
bbeam
bcol
Plan views e = Eccentricity between

column centroid and
beam centerline
Figure 11.4.2 Determination of effective joint width bj.
TABLE 11.4.1 VALUES OF JOINT SHEAR CONSTANT γ (FROM ACI

COMMITTEE 352 [11.1])
Joint Classification
Joint Type Confineda on All Confineda on Three or Other

Four Vertical Faces Two Opposite Vertical Cases
Faces
With a continuous column 24 20 15

1
With a discontinuous column 20 15 12
2 With a continuous column 20 15 12

With a discontinuous column 15 12 8
a By beams with width at least ¾ the width of the column face into which they frame.
11.5 COLUMN-T O-B EAM MOMENT STRENGTH RATIO

The design of Type 2 connections that are part of the seismic-force-resisting system requires
that the summation of the nominal moment strengths of the columns framing into a joint
be at least 1.2 times the summation of the nominal moment strengths of the beams, calcu-
lated at the faces of the joint [11.1, Section 4.4.2]. This minimum strength ratio, although
too low to prevent the occurrence of column hinging, is considered acceptable to prevent
the formation of a story mechanism. It is not applicable to roof connections, however,
as either column or beam hinging is acceptable in this type of connections. The column-
to-beam moment strength ratio has also been linked to joint performance under shear rever-
sals [11.9], joint behavior improving with an increase in moment strength ratio.
No minimum column-to-beam moment strength ratios are specified for Type 1 joints,
since column hinging is possible in frame design for which Type 1 joints apply.
390
11.6 ANCHORAGE OF REINFORCEMENT IN THE

JOINT REGION
Adequate anchorage of beam and column longitudinal reinforcement in the joint region
is critical because the members must be able to develop the required strength and transfer
forces into and through the joint without unacceptable reinforcement slip.
For Type 1 joints, the ACI Committee 352 recommendations include provisions for
anchorage of beam and column longitudinal reinforcement terminating at a joint (e.g., exte-
rior and roof connections). The critical section for development of beam longitudinal bars
is taken at the face of the column. For longitudinal column bars, on the other hand, the crit-
ical section is taken at the outside edge of the bottom longitudinal reinforcement passing
through (e.g., T-joints) or anchored into (e.g., knee joints) the joint.
When bars are terminated by a standard 90° hook, the development length is determined
as [11.1, Section 4.5.2.3],2
fy ψ c ψ r
Ldh = db (11.6.1)
50 fc′
where ψc is a cover factor equal to 0.7 for #11 bars and smaller with side cover not less than
2.5 in. and cover beyond the hook tail greater than or equal to 2 in.; otherwise, ψc = 1.0. The
confinement factor ψr is equal to 1.0 unless the hook is enclosed by horizontal or vertical
ties along development length at a spacing not greater than 3db of the bar being terminated,
in which case ψr = 0.8.
For Type 2 joints, because there is potential for loss of concrete cover due to inelastic
deformation reversals, the critical section for development of reinforcement is to be taken
at the outside edge of the column core. In this case, the development length for longitudinal
bars terminating at the joint is calculated as [11.1, Section 4.5.2.4],
α fy ψ r
Ldh = db (11.6.2)
75 fc′
where α ≥ 1.25.
In all cases, Ldh must not be taken less than 6 in. or 8db, whichever is greater. Also, the
outside edge of the hook tail should be within 2 in. from the farthest edge of the core, with
the hook oriented toward the inside of the joint such that a diagonal strut can be properly
developed (Fig. 11.6.1). For beams with more than one layer of flexural reinforcement, the
hook tails of each additional layer should be located within 3db of the adjacent hook tail.
To prevent excessive slip of longitudinal beam and column bars passing continuously
through Type 2 joints, the following ratios between member depth and longitudinal bar
diameter must be satisfied:
h(beam)  fy 
≥ 20  (11.6.3)
db ( column )  60, 000 
h(column)  fy 
≥ 20  (11.6.4)
db ( beam )  60, 000 
but no less than 20, where h refers to member depth and db to longitudinal bar diameter.
2 This equation is the same as Eq. (6.10.1) when applied to uncoated reinforcement embedded in normal-weight concrete. The
lightweight concrete factor λ is not included in the ACI Committee 352 recommendations, which apply to normal-weight concrete.
The cover and confinement factors, ψc and ψr, respectively, although not explicitly included in the ACI Committee 352 document,
are used here for consistency with expressions used in ACI 318-14 and Chapter 6 of this book.
391
11.8 EXAMPLES 391
≤ 2 in.
Critical section
(Type 1)
Critical section
(Type 2)
Joint ties
Figure 11.6.1 Anchorage details of flexural reinforcement with 90° hook.
The ACI Committee 352 recommendations also include provisions for the use of headed
reinforcement in lieu of reinforcement terminating in a standard hook for both Type 1 and
Type 2 joints. The reader is referred to these recommendations for further information.
11.7 TRANSFER OF COLUMN AXIAL FORCES

THROUGH THE FLOOR SYSTEM
The use of concretes with different compressive strengths for the columns and floor sys-
tem (slab and beams) is not uncommon. When the specified compressive strength of the
concrete used in the columns is greater than 1.4 times that of the concrete used in the floor
system, the ACI Code requires that transmission of column axial load through the floor sys-
tem be investigated (ACI-15.3). For cases in which the design axial strength of the column
joint calculated using fc′ of the floor system is less than the factored axial force, one of the
following measurements should be taken to increase the axial strength of the connection,
i. Concrete with the same specified compressive strength as that of the column should
be provided within the floor depth and extended beyond the column faces at least 2 ft.
ii. Addition of vertical dowels enclosed by spirals, designed to provide required addi-
tional axial strength.
For beam-column connections confined on all four sides by beams of approximately the
same depth framing into the column and where the specified strength of the column con-
crete does not exceed 2.5 times that of the floor concrete, fc′ used in the calculation of axial
strength can be taken as 75% that of the column plus 35% that of the floor system.
11.8 EXAMPLES
Three examples are presented to illustrate the recommendations as they apply to Type 1
joints used in ordinary building construction. The requirements for Type 2 joints, where
ductility and dissipation of energy through inelastic deformations are required, are more
stringent. For examples illustrating the design of Type 2 joints, the reader is referred to the
ACI Committee 352 recommendations [11.1].
392
EXAMPLE 11.8.1
Design the exterior beam–column joint shown in Fig. 11.8.1 for the case of the main
beam subjected to negative bending. The joint is to be a Type 1 joint, and strength is the
primary criterion. Use fc′ = 6000 psi, fy = fyt = 60,000 psi, and a factored axial load in
the column Pu = 350 kips. Assume the story heights below and above the connection are
equal to 11 ft and 10 ft, respectively.
20”
20”
8 – #9
Column
Main beam
Spandrel beam
24”
4 – #10
20” 3 – #9
2 – #9
2 – #8
15”
10”
Figure 11.8.1 Design Example 11.8.1.
SOLUTION
(a) Evaluate anchorage of #10 bars in the main beam. From Eq. (11.6.1), the required
development length for #10 bars terminating with a standard hook in a Type 1 joint is
fy ψ c ψ r 60, 000(0.7)(1.0)
Ldh = db = (1.27) = 13.8 in.
50 fc′ 50 6000
where ψc is taken as 0.7 given that the minimum side cover is satisfied (column is 5 in. wider
than beam, which ensures a minimum side cover greater than 2.5 in.) and assuming a mini-
mum of 2 in. of cover beyond the hook. The required development length is clearly less than
the available length for anchorage. Also, the hooks for these bars should be located within 2
in. of the furthest edge of the core (i.e., core side closest to exterior face of column).
(b) Examine the shear on the column to be transmitted through the joint. The moment
on the columns may be assumed to be zero at midheight (see Fig. 11.8.2). In this
case, with 12-ft and 10-ft column lengths and assuming that the beam reaches its
flexural strength prior to the columns reaching their strength,
(h1 + h2 )
Vcol = M pr ,b
2
where Mpr,b is the probable flexural strength of the beam. (Note that the moment caused
by the beam shear with respect to the column centroid has been neglected in the deter-
mination of column shear.)
(Continued)
39
11.8 EXAMPLES 393
Midheight
Vcol of column
assumed
h1 inflection
= 5’ 20”
point
2 Vcol = 45.1k
Mpr,b
T = 305k
h2
= 6’ C = 305k
2 45.1k
Vcol
Figure 11.8.2 Column shear at joint in Example 11.8.1.
Because this is a Type 1 connection, Mpr,b may be taken as the nominal flexural
strength using α = 1.0 (see Fig. 11.2.2). Thus, ignoring the bottom compression steel,
C = 0.85 fc′ ba = 0.85(6)(15a ) = 76.5a
T = α Ast f y = 1.0(4)(1.27)(60) = 305 kips

a = 3.99 in.
Taking d = 21.5 in.,
 a  3.99  1
M pr ,b = T  d −  = 305  21.5 −  = 496 ft-kips
 2   2  12
3.99
21.5 −
d −c 0.75 (0.003) = 0.0091
εt = εs = (0.003) =
c 3.99
0.75
Thus, the tension reinforcement yields and the section is tension controlled.
2 496
Vcol = M pr ,b = = 45.1 kips
(h1 + h2 ) 11
From Eq. (11.2.2), the joint horizontal shear is then

joint Vu = 305 − 45.1 = 260 kips
The nominal strength of the joint is given by Eq. (11.4.2). The design beam width bbeam
in the direction of loading is 15 in. The effective joint width bj is the smallest of
bbeam + bcol 15 + 20
(i) = = 17.5 in. Governs!
2 2
mh
(ii) bbeam + ∑
2
where mh / 2 cannot exceed the extension of the column beyond the edge of the beam
on each side. In the direction of loading being considered, the beam centerline and
the column centroid intersect. Thus, the coefficient m is equal to 0.5. Accordingly,
mh 0.5(20)
= = 5 in.
2 2
(Continued)
394
The extension of the column beyond the edge of the beam is (20 –15) / 2 = 2.5 in.
Thus, mh /2 = 2.5 in. and
mh
bbeam + ∑ = 15 + 2(2.5) = 20 in.
2
(iii) bcol = 20 in.
Therefore, bj = 17.5 in.
According to ACI-ASCE Committee 352 [11.1, Section 4.3.2], a joint side is effectively
confined when the beam covers at least three-fourths of the width of the column, and
the total depth of such a member is not less than three-fourths of the total depth of the
deepest member framing into the joint. Here the spandrel beams have a width of 10 in.
and do not cover three-fourths of the column width of 20 in. Because the joint is not
confined on at least two opposite faces, the value γ for this connection, according to
Table 11.4.1, is 15.
The nominal shear strength is then (Eq. 11.4.2)
1
Vn = γ fc′ b j h = 15 6000 (17.5)(20) = 407 kips
1000
and

φ Vn = 0.75(407) = 305 kips > Vu = 260 kips
The joint shear strength is therefore adequate.
(c) Horizontal column ties through the joint. For the purpose of establishing transverse
reinforcement requirements, a core is not considered confined unless beams frame in
from all four sides and cover at least three-fourths of the column width, and not more
than 4 in. of column width is exposed on each side of the beam (see Section 11.3).
In this example the joint is not confined. Thus, according to the ACI Committee 352
recommendations, at least two layers of transverse reinforcement should be pro-
vided between the top and bottom longitudinal reinforcement of the deepest beam
framing into the joint, not spaced farther than 12 in. Also, ACI-15.4.2.2 specifies a
maximum spacing of one-half the depth of the shallowest beam framing into the
column (10 in. in this case).
For unconfined joints, transverse reinforcement should also satisfy the requirements
for columns in ACI-10.7.6.1.2 which, for the case of tied columns, refers to ACI-25.7.2.
For #10 longitudinal column bars or smaller, ACI-25.7.2.2 requires the use of tie bars no
smaller than #3. Maximum spacing for a column with #3 ties is limited to the smallest
of (ACI-25.7.2.1)
16 longitudinal bar diameters = 16(1.128) = 18 in.
48 tie diameters = 48(0.375) = 18 in.
least cross-sectional dimension of column = 20 in.
The above would not control for this joint, since 18 in. is approximately equal to the
clear distance between the top and bottom layers of beam longitudinal reinforcement
and at least two layers should be provided.
Use #3 hoops @ 6 in. spacing, as detailed in Fig. 11.8.3.
(Continued)
395
11.8 EXAMPLES 395
For each layer of transverse reinforcement, provided area of transverse reinforcement

must satisfy (ACI-15.4.2)
bcol sh (20)(6)
Ash,min = 0.75 fc′ = 0.75 6000 = 0.12 sq in. < 2(0.11 sq in.) OK
f yt 60, 000
Note that because fc′ = 6000 psi, 0.75 fc′ > 50 psi.

(d) Transmission of axial load through the connection. Because concrete with the same
compressive strength is used in both the columns and the beams, transmission of
axial force through the connection is not a concern.
6 in. 4 – #10
6 in.
6 in. #3 joint hoops
2 – #9
6 in.
Figure 11.8.3 Joint reinforcement details for Example 11.8.1.
EXAMPLE 11.8.2
Compute the effective joint width bj for the joint of Example 11.8.1 for loading in the
direction parallel to the spandrel beams.
SOLUTION
In this direction, the centerline of the spandrel beams does not pass through the column
centroid (see Fig. 11.8.1). However, all beam bars pass through the column core, and
thus the recommendations of the ACI-ASCE Committee 352 report apply [11.1, Section
2.2.2]. The effective joint width bj is the smallest of
(i) = = 15 in.
2 2
mh
(ii) bbeam + ∑
2
(Continued)
396
The eccentricity between the column centroid and the centerline of the spandrel
beams is (20 – 10) / 2 = 5 in., which is greater than bcol / 8 = 2.5 in. Therefore, the
coefficient m is 0.3. Accordingly,
mh 0.3(20)
= = 3 in. ≤ [10 in.(extension of the column beyond thee edge
2 2

of the beam)]
and
mh
bbeam + ∑ = 10 + 3 = 13 in. Governs!
2
where ∑ mh/2 applies only to the side of the joint where the edge of the column extends
beyond the edge of the beam; the other side of the beam is flush with the column face.
(iii) bcol = 20 in.
Therefore, bj = 13 in.
EXAMPLE 11.8.3
For the interior joint shown schematically in Fig. 11.8.4, check the adequacy of the joint
shear strength and determine the transverse reinforcement required if the joint is Type
1 and the structure is subjected to non-seismic lateral loading in both directions. Use
fc′ = 6000 psi, fy = 60,000 psi, and a factored axial load in the column Pu = 570 kips.
Beam, 15 × 20, 4 – #9 top
11’–0 (assumed distance between
4 – #8 bottom
points of inflection)
Beam, 16 × 24
4 – #9 top
4 – #8 bottom
Column, 20 × 20, 8 – #9
10 ft long above beam
12 ft long below beam
Figure 11.8.4 Joint for Example 11.8.3.
Figure 11.8.5 Forces in direction of 16 × 24 beams in Example 11.8.3.

(Continued)
397
11.8 EXAMPLES 397
SOLUTION
(a) Development of reinforcement. All longitudinal steel is to be extended through
the joint.
(b) Shear in direction of 16 × 24 beams. Due to lateral loading (from left to right) on
the structure, the moments [Fig. 11.8.5(a)] M pr− ,b and M pr+ ,b have the same rotational
direction and give the highest shear through the joint.
Again, it is assumed that the probable moments equal the nominal strengths Mn for
both positive and negative bending. If, however, the shear transferred into the joint is
governed by the nominal moment capacity of the columns, the beam moments will be
less than their nominal capacities. This is checked after calculation of probable moment
capacities in the beams.
The probable negative moment strength, M pr− ,b , based on tension in the 4–#9 bars and
neglecting compression reinforcement, is
C = 0.85 fc′ ba = 0.85(6)(16)a = 81.6 a
T = α Ast f y = 1.0(4)(1.00)(60) = 240 kips
a = 2.94 in.
 a  2.94  1
M pr− ,b = T  d −  = 240  21.5 −  = 400 ft-kips
 2  2  12
2.94
21.5 −
d −c 0.75 (0.003) = 0.0135
εt = εs = (0.003) =
c 2.94
0.75
Thus the tension reinforcement yields and the section is tension controlled.
For bending in the positive direction (tension in the 4–#8 bars), neglecting compression
reinforcement,
T = α Ast f y = 1.0(4)(0.79)(60) = 190 kips

a = 2.32 in.
 a  2.32  1
M pr+ ,b = T  d −  = 190  21.5 −  = 322 ft-kips
 2  2  12
Because the area of positive (bottom) reinforcement is less than that of the negative (top)
reinforcement, the tension reinforcement also yields and the section is tension controlled.
The nominal moment capacity of the column for a factored axial force of 570 kips is
580 ft-kips. Because the summation of the nominal moment capacities of the columns
framing into the joint (1140 ft-kips) is much greater than the summation of the moment
capacities of the beams (722 ft-kips), joint shear is governed by the moment capacity of
the beams, as initially assumed. Thus,
(
Vcol = M pr− ,b + M pr+ ,b ) (h +2 h ) = (40011+ 322) = 65.6 kips
1 2
The joint horizontal shear is then

joint Vu = 240 + 190 − 65.6 = 364 kips
(Continued)
398
The nominal strength of the joint is given by Eq. (11.4.2). The design beam width
bbeam in the direction of loading is the average of the widths of the two beams (i.e., 16 in.,
since both are 16 × 24). The effective joint width bj is the smallest of
(i) = = 18 in. Governs!
2 2
mh
(ii) bbeam + ∑ , where m = 0.5 (no joint eccentricity). Thus,
2
mh 0.5(20)  (20 − 16) 
= = 5 in. >  = 2 in. (column extension beyond edge)
2 2  2 
Thus, mh / 2 = 2 in., and
mh
bbeam + ∑ = 16 + 2(2) = 20 in.
2
(iii) bcol = 20 in.
Thus, bj = 18 in.
Because there are beams on all four sides of the column whose widths (16 in. or 15 in.)
cover at least three-fourths of the width of the column (15 in.) and because their total
depth (20 or 24 in.) is not less than three-fourths of the total depth of the deepest mem-
ber framing into the joint (18 in.), the joint is considered to be effectively confined in all
four vertical faces. Thus, the value of γ is 24 (see Table 11.4.1).
The nominal shear strength is (Eq. 11.4.2)
1
Vn = γ fc′ b j h = 24 6000 (18)(20) = 669 kips
1000
and

φ Vn = 0.75(669) = 502 kips > Vu = 364 kips OK
The joint shear strength is therefore adequate.
(c) Shear in the direction of the 15 × 20 beams. If the lateral loading on the structure
requires moments to be transfered in this direction, similar to the force system in
Fig. 11.5.5, the process used in step (b) would be repeated for this direction. The
same top and bottom reinforcement is used in these beams as in the 16 × 24 beams,
which means that the shear force introduced into the joint will be close to that calcu-
lated for the other direction (there will be a slight difference in column shear force
because the moment capacities of the beams will be smaller due to their shallower
depths). The joint shear strength will also remain approximately the same, since the
effective joint width for this direction is 17.5 in., which is almost the same as that in
the perpendicular direction (18 in.). Given the large difference between design joint
shear strength and the joint shear demand calculated in part (b), joint shear strength
in this direction should also be adequate.
(d) Horizontal column ties through the joint. As discussed in Section 11.3, a core is con-
sidered confined when beams frame in from all four sides and cover at least three-
fourths of the column width and no more than 4 in. of column width is exposed on
each side of the beam. In this case, the joint is confined. Thus, column ties need not
extend through the joint.
(e) Transmission of axial load through the connection. Because concrete with the same
compressive strength is used in both the columns and the beams, transmission of
axial force through the connection is not a concern.
39
11.9 ADDITIONAL REMARKS
The factors relating to the design of joints intended to resist primarily static loads within
the elastic range (Type 1) have been discussed and treated using three examples. For the
design of Type 2 joints requiring greater ductility, the reader is referred to the Committee
352 report [11.1], as well as ACI-18.8. For clarity, the slab reinforcement was omitted from
the calculations of the tensile resultant at the top of the beams in the examples. As recom-
mended by the ACI-ASCE Committee 352 report, slab reinforcement should be included in
the calculations. This is particularly important for thick, heavily reinforced slabs, since this
factor may result in an important increase in joint shear demand.
Beam-to-column connections resisting seismic loads have been given considerable atten-
tion by several researchers, with the subject reviewed by Meinheit and Jirsa [11.10]. The
behavior of beam-to-column connections in precast concrete has been studied by Pillai and
Kirk [11.41] and others [11.15, 11.17, 11.25]. A comprehensive list of references on the
subject of behavior and design of beam-column joints is provided in the References section.
SELECTED REFERENCES
11.1. ACI- ASCE Committee 352. “Recommendations for Design of Beam– Column Joints in
Monolithic Reinforced Concrete Structures” (ACI 352R-02). Farmington Hills, MI: American
Concrete Institute, 2002, 37 pp.
11.2. Norman W. Hanson and Harold W. Conner. “Seismic Resistance of Reinforced Concrete Beam-
Column Joints,” Journal of the Structural Division, ASCE, 93, ST 5 (October 1967), 533–560.
11.3. Norman W. Hanson. “Seismic Resistance of Concrete Frames with Grade 60 Reinforcement,”
Journal of the Structural Division, ASCE, 97, ST 6 (June 1971), 1685–1700.
11.4. L. M. Megget and R. Park. “Reinforced Concrete Exterior Beam-Column Joints Under Seismic
Loading,” New Zealand Engineering (Wellington), 26, 11 (November 1971), 341–353.
11.5. S. M. Uzumeri and M. Seckin, “Behavior of Reinforced Concrete Beam-Column Joints
Subjected to Slow Load Reversals,” Report 74 05, Department of Civil Engineering, University
of Toronto, Ontario, Canada, March 1974.
11.6. John Minor and James O. Jirsa. “Behavior of Bent Bar Anchorages,” ACI Journal, Proceedings,
72, April 1975, 141–149.
11.7. James K. Wight and Mete A. Sozen. “Strength Decay of Reinforced Concrete Columns under
Shear Reversals,” Journal of the Structural Division, ASCE, 101, ST5 (May 1975), 1053–1065.
11.8. José G. L. Marques and James O. Jirsa. “A Study of Hooked Bar Anchorages in Beam–Column
Joints,” ACI Journal, Proceedings, 72, May 1975, 198–209.
11.9. Duane L. N. Lee, Robert D. Hanson, and James K. Wight. “RC Beam-Column Joints under
Large Load Reversals,” Journal of the Structural Division, ASCE 103, 12 (December 1977),
2337–2350.
11.10. Donald F. Meinheit and James O. Jirsa. “Shear Strength of R/C Beam–Column Connections,”
Journal of the Structural Division, ASCE, 107, ST11 (November 1981), 2227–2244.
11.11. Erik Skettrup, Jørgen Strabo, Niels Houmark Anderson, and Troels Brøndum-
Nielson.
“Concrete Frame Corners,” ACI Journal, Proceedings, 81, November– December 1984,
587–593.
11.12. M. R.Ehsani and J. K. Wight. “Effect of Transverse Beams and Slab on Behavior of Reinforced
Concrete Beam-to-Column Connections,” ACI Journal, Proceedings, 82, March–April 1985,
188–195.
11.13. Ahmad J. Durrani and James K. Wight. “Behavior of Interior Beam-to-Column Connections
Under Earthquake-Type Loading,” ACI Journal, Proceedings, 82, May–June 1985, 343–349.
11.14. M. R. Ehsani and J. K. Wight. “Exterior Reinforced Concrete Beam-to-Column Connections
Subjected to Earthquake-Type Loading,” ACI Journal, Proceedings, 82, July–August 1985,
492–499.
11.15. Prabhakara Bhatt and D. W. Kirk. “Tests on an Improved Beam Column Connection for Precast
Concrete,” ACI Journal, Proceedings, 82, November–December 1985, 834–843.
11.16. T. Ueda, I. Lin, and N. M. Hawkins. “Beam Bar Anchorage in Exterior Column–Beam
Connections,” ACI Journal, Proceedings, 83, May–June 1986, 412–422.
11.17. Charles W. Dolan, John F. Stanton, and Richard G. Anderson. “Moment Resistant Connections
and Simple Connections,” PCI Journal, 32, March–April 1987, 62–74.
40
11.18. Ahmad J. Durrani and James K. Wight. “Earthquake Resistance of Reinforced Concrete
Interior Connections Including a Floor Slab,” ACI Structural Journal, 84, September–October
1987, 400–406.
11.19. Daniel P. Abrams. “Scale Relations for Reinforced Concrete Beam–Column Joints,” ACI
Structural Journal, 84, November–December 1987, 502–512.
11.20. S. E. El-Metwally and W. F. Chen. “Moment–Rotation Modeling of Reinforced Concrete
Beam–Column Connections,” ACI Structural Journal, 85, July–August 1988, 384–394.
11.21. Parviz Soroushian and Ki-Bong Choi. “Local Bond of Deformed Bars with Different Diameters
in Confined Concrete,” ACI Structural Journal, 86, March–April 1989, 217–222.
11.22. Olga Velez Ammerman and Catherine Wolfgram French. “R/
C Beam– Column– Slab
Subassemblages Subjected to Lateral Loads,”Journal of Structural Engineering, ASCE, 115, 6
(June 1989), 1289–1308.
11.23. Roberto T. Leon. “Interior Joints with Variable Anchorage Lengths,” Journal of Structural
Engineering, ASCE, 115, 9 (September 1989), 2261–2275.
11.24. Thomas Paulay. “Equilibrium Criteria for Reinforced Concrete Beam–Column Joints,” ACI
Structural Journal, 86, November–December 1989, 635–643.
11.25. Catherine Wolfgram French, Michael Hafner, and Viswanath Jayashankar. “Connections

between Precast Elements— Failure within Connection Region,” Journal of Structural
Engineering, ASCE, 115, 12 (December 1989), 3171–3192.
11.26. Roberto T. Leon. “Shear Strength and Hysteretic Behavior of Interior Beam–Column Joints,”
ACI Structural Journal, 87, January–February 1990, 3–11.
11.27. S. J. Pantazopoulou and J. P. Moehle. “Truss Model for 3-D Behavior of R. C. Exterior
Connections,” Journal of Structural Engineering, ASCE, 116, 2 (February 1990), 298–315.
11.28. M. R. Ehsani and J. K. Wight. “Confinement Steel Requirements for Connections in Ductile
Frames,” Journal of Structural Engineering, ASCE, 116, 3 (March 1990), 751–767.
11.29. M. Seckin and H. C. Fu. “Beam–Column Connections in Precast Reinforced Concrete,” ACI
11.30. Philip K. C. Wong, M. J. N. Priestley, and R. Park. “Seismic Resistance of Frames with
Vertically Distributed Longitudinal Reinforcement in Beams,” ACI Structural Journal, 87,
July–August 1990, 488–498.
11.31. Parviz Soroushian and Ki-Bong Choi. “Analytical Evaluation of Straight Bar Anchorage
Design in Exterior Joints,” ACI Structural Journal, 88, March–April 1991, 161–168.
11.32. Parviz Soroushian, Ki-Bong Choi, Gill-Hyun Park, and Farhang Aslani. “Bond of Deformed
Bars to Concrete: Effects of Confinement and Strength of Concrete,” ACI Structural Journal,
88, May–June 1991, 227–232.
11.33. Ian N. Robertson and Ahmad J. Durrani. “Gravity Load Effect on Seismic Behavior,” ACI
11.34. Mohammad R. Ehsani and Fadel Alameddine. “Design Recommendations for Type 2 High-
Strength Reinforced Concrete Connections,” ACI Structural Journal, 88, May–June 1991,
277–291.
11.35. Fariborz Barzegar, Rainer Echle, and Mehrdad Foroozesh. “Moment Transfer and Slab Effective
Widths in Laterally Loaded Edge Connections,” ACI Structural Journal, 88, September–
October 1991, 615–623.
11.36. A. G. Tsonos, I. A. Tegos, and G. Gr. Penelis. “Seismic Resistance of Type 2 Exterior Beam–
Column Joints Reinforced with Inclined Bars,” ACI Structural Journal, 89, January–February
1992, 3–12.
11.37. Gilson N. Guimaraes, Michael E. Kreger, and James O. Jirsa. “Evaluation of Joint-Shear
Provisions for Interior Beam–Column–Slab Connections. Using High-Strength Materials,” ACI
11.38. A. H. Mattock and J. F. Shen. “Joints between Reinforced Concrete Members of Similar
Depth,” ACI Structural Journal, 89, May–June 1992, 290–295.
11.39. Stavroula Pantazopoulou and John Bonacci. “Consideration of Questions about Beam–Column
Joints,” ACI Structural Journal, 89, January–February 1992, 27–36.
11.40. Richard H. Scott. “Intrinsic Mechanisms in Reinforced Concrete Beam–Column Connection
Behavior,” ACI Structural Journal, 93, May–June 1996, 336–346.
11.41. S. U. Pillai and D. W. Kirk. “Ductile Beam–Column Connection in Precast Concrete,” ACI
11.42. B. Abdel-Fattah and James K. Wight. “Study of Moving Beam Plastic Hinging Zones for
Earthquake-Resistant Design of R/C Buildings,” ACI Structural Journal, 84, January–February
1987, 31–39.
11.43. Sergio M. Alcocer. “R/C Frame Connections Rehabilitated by Jacketing,” Journal of Structural
Engineering, ASCE, 119, 5 (May 1993), 1413–1431.
401
PROBLEMS 401
11.44. Sergio M. Alcocer and James O. Jirsa. “Strength of Reinforced Concrete Frame Connections
Rehabilitated by Jacketing,” ACI Structural Journal, 90, May–June 1993, 249–261.
11.45. O. V. Ammerman and Catherine Wolfgram-French. “R/C Beam–Column–Slab Subassemblages
Subjected to Lateral Loads,” Journal of Structural Engineering, 115, 6 (June 1989), 1298–1308.
11.46. P. C. Cheung, Thomas Paulay, and Robert Park. “Mechanisms of Slab Contributions in Beam–
Column Subassemblages,” Design of Beam–Column Joints for Seismic Resistance (SP-123).
Farmington Hills, MI: American Concrete Institute, 1991 (pp. 259–289).
11.47. P. A. Cote and John W. Wallace. “A Study of RC Knee-Joints Subjected to Cyclic Lateral
Loading,” Report No. CU/CEE-94/04, Department of Civil and Environmental Engineering,
Clarkson University, Potsdam, NY, 1994.
11.48. Catherine Wolfgram-French and Jack P. Moehle. “Effect of Floor Slab on Behavior of Slab-
Beam-Column Connections,” Design of Beam–Column Joints for Seismic Resistance (SP-123).
Farmington Hills, MI: American Concrete Institute, 1991 (pp. 225–258).
11.49. Roberto T. Leon and James O. Jirsa. “Bi-directional Loading of RC Beam–Column Joints,”
Earthquake Spectra, 2, 3 (1986), 537–564.
11.50. S. Mazzoni, Jack P. Moehle, and C. R. Thewalt. “Cyclic Response of RC Beam-Column Knee
Joints: Test and Retrofit,” Report No. UCB/EERC-91/14, Earthquake Engineering Research
Center, University of California, Berkeley, October 1991, 24 pp.
11.51. Stavroula Pantazopoulou, Jack P. Moehle, and B. M. Shahrooz. “Simple Analytical Model for
T-Beam in Flexure,” Journal of Structural Engineering, 114, 7 (July 1988), 1507–1523.
11.52. Robert Park, M. J. N. Priestley, and W. D. Gill. “Ductility of Square-Confined Concrete
Columns,” Journal of the Structural Division, ASCE, 108, ST4 (April 1982), 929–950.
11.53. C. G. Quintero-Febres and J. K. Wight. “Investigation of the Seismic Behavior of RC Interior
Wide Beam-Column Connections,” Report No. UMCEE 97-15, Department of Civil and
Environmental Engineering, University of Michigan, Ann Arbor, September 1997, 292 pp.
11.54. G. S. Raffaelle and James K. Wight. “R/C Eccentric Beam–Column Connections Subjected to
Earthquake-Type Loading,” Report No. UMCEE 92-18, Department of Civil and Environmental
Engineering, University of Michigan, Ann Arbor, 1992, 234 pp.
11.55. E. I. Saqan and Michael Kreger. “Evaluation of U.S. Shear Strength Provisions for Design
of Beam– Column Connections Constructed with High- Strength Concrete,” High-Strength
Concrete in Seismic Regions (SP-176). Farmington Hills, MI: American Concrete Institute,
1998 (pp. 311–328).
11.56. S. Sugano, T. Nagashima, H. Kimura, and A. Ichikawa. “Behavior of Beam–Column Joints
Using High-Strength Materials,” Design of Beam–Column Joints for Seismic Resistance (SP-
123). Farmington Hills, MI: American Concrete Institute, 1991 (pp. 359–378).
11.57. John W. Wallace, S. W. McConnell, P. Gupta, and P. A. Cote. “Use of Headed Reinforcement in
Beam–Column Joints Subjected to Earthquake Loads,” ACI Structural Journal, 95, September–
October 1998, 590–606.
11.58. Catherine Wolfgram-French and A. Boroojerdi. “Contribution of R/C Floor Slab in Resisting
Lateral Loads,” Journal of Structural Engineering, 115, 1 (January 1989), 1–18.
11.59. H. E. Zerbe and A. J. Durrani. “Seismic Response of Connections in Two-Bay R/C Frame
Subassemblies,” Journal of Structural Engineering, 115, 11 (November 1989), 2829–2844.
11.60. S. W. McConnell and John W. Wallace. “Behavior of Reinforced Concrete Beam-Column Knee
Joints Subjected to Reversed Cyclic Loading,” Report No. CU/CEE-95/07, Department of Civil
and Environmental Engineering, Clarkson University, Potsdam, NY, June 1995.
PROBLEMS
All problems are to be worked using the ACI Code and the recommendations of ACI-ASCE
Committee 352 [11.1]. In all cases, assume that joint shear is governed by the flexural
capacity of the beams.
11.1 Check the anchorage of bars and determine Type 1 joint and use fc′ = 4000 psi and fy =
the transverse reinforcement for the exterior fyt = 60,000 psi. Show detail of transverse rein-
beam-column joint of the figure for Problem forcement in joint.
11.1. Assume that lateral loading provides 11.2 Repeat Problem 11.1, except consider the col-
the critical condition at the joint; assume that umn to be 18 in. square, the spandrel 14 × 24
columns are bent in double curvature with with 3–#9 bars (top and bottom), and the beam
an inflection point at midheight. Assume a 18 × 28 with 4–#10 bars (top and bottom).
402
11.3 Repeat Problem 11.1, except consider the joint

Column 24 × 24
to be an interior one with the 21 × 28 beam
on both sides of the column. The spandrel then
Column becomes an interior beam. Assume that all bars
height 11’
are continuous through the column. The critical
loading condition for the joint is with clock-
wise moments from the 21 × 28 beam acting on
11’ Beam, 21 × 28, 4 – #11 bars top the joint at both sides of the column.
4 – #9 bars bottom
Spandrel, 18 × 30, 3 – #10 bars top

3 – #9 bars bottom
Problem 11.1
CHAPTER 12
SERVICEABILITY
12.1 INTRODUCTION
As discussed in Chapter 2, a reinforced concrete structure must be designed not only to pro-
tect the life of the occupants, but also to ensure that it serves its intended purpose through-
out its service life: that is, it must be functional or serviceable. The first requirement is
satisfied by meeting the safety provisions of the ACI Code (see Sections 2.6 and 2.7). The
second requirement, hereafter referred to as serviceability, will be satisfied by meeting the
ACI Code serviceability provisions, which may be expressed as given earlier.
Sallowable ≥ Sservice_loads [2.8.2]
(see Sections 2.8 and 2.9). In this equation, Sallowable represents the specified limit not to be
exceeded in order to ensure satisfactory behavior and functionality of the structure dur-
ing its service life, while the term Sservice_loads is a calculated performance quantity of the
structural member under service-level loads. Note that the allowable serviceability limit,
Sallowable, for satisfactory performance of a given member or structural system is not unique
and will depend on the use of the structure. For example, a floor system supporting highly
sensitive instrumentation must be designed for stricter deflection and vibration limits than
one supporting partitions in a standard residential building.
Common serviceability issues that affect the design of reinforced concrete members are
excessive deflection, detrimental cracking, excessive amplitude or undesirable frequency
of vibration, and excessive noise transmission. While serviceability issues do not generally
pose a threat to the safety of the occupants, they can have severe economic consequences
and thus, the structural engineer must adequately address them in design.
This chapter discusses in detail design for the two most common serviceability con-
cerns: control of deflections and control of crack widths. Vibration and noise control, which
may also be important, lie outside the scope of this book and are not treated here in detail.
Nonetheless, a brief discussion of current design criteria and some guidance for the control
of floor vibrations is provided in Section 12.20.
12.2 FUNDAMENTAL ASSUMPTIONS
Consider the flexural behavior of the reinforced concrete beam section under increasing
applied loads described in Section 3.2. When designed to satisfy the strength require-
ments of the ACI Code, the strain and stress distributions in the concrete and the steel
reinforcement at ultimate (i.e., under factored loads) will be as shown in Fig. 3.2.1(c). This
means that under service-level loads, the strains and stresses in the concrete and the steel
40
404 C H A P T E R 1 2 S erviceability
reinforcement will be much smaller than those at the ultimate condition. Indeed, the behav-
ior of the concrete in the compression zone is expected to be essentially linear and the steel
strains less than the yield strain. For the purpose of computing the behavior of reinforced
concrete members under service loads, it will be assumed that the stresses in the concrete
in compression and the steel reinforcement are proportional to the strains: that is, their
behavior is linear elastic. The assumption that stresses are proportional to strains for the
concrete in compression is reasonably correct for stresses below about one-half fc′, the
28-day compressive strength. This will typically be the case for beams with low and mod-
erate amounts of tension reinforcement. The assumptions that (1) plane sections remain
plane, (2) concrete does not carry tension (it cracks under tension), and (3) no slip occurs
between the steel bars and the surrounding concrete during the development of the tensile
forces in the bars also hold true under service load levels.
12.3 MODULUS OF ELASTICITY RATIO, n

As discussed in Chapter 1, the stress-strain relation for the steel reinforcement is linear
below the yield stress, but that of concrete is only approximately linear, even at or below
the service load stresses. The modulus of elasticity of steel, Es, varies little with its strength,
whereas that of concrete varies with its density and strength. In ACI-20.2.2.2, Es is taken to
be 29,000,000 psi (200,000 MPa); ACI-19.2.2.1 gives Ec as 33w1c .5 fc′, in which wc is the
unit weight of concrete in lb/cu ft. Table 1.10.1 gives values for the modulus of elasticity Ec
for normal-weight concrete of various strengths.
It will be shown that the ratio between Es and Ec, or the modular ratio, n = Es /Ec, is often
used for computing stresses under service loads rather than the actual values of Ec or Es.
These stresses are called service stresses (or working stresses, in reference to the working
stress method used in the past). It is suggested that the values in Table 12.3.1 be used for
normal-weight concrete.
12.4 EQUILIBRIUM CONDITIONS
Two equilibrium conditions apply to a section subjected to bending only: (1) the resultant
internal compressive force must be equal to the resultant internal tensile force; and (2) the
moment of the internal couple, composed of the resultant compressive and tensile forces,
must be equal to the applied bending moment on the cross section. These two equilibrium
conditions must hold true under any applied loads (e.g., under service-level loads and at
the ultimate condition). Under service loads, however, it is assumed that the stress distri-
bution across the depth is linear and proportional to strains as discussed in Section 12.2.
TABLE 12.3.1 PRACTICAL VALUES FOR MODULAR RATIO n

FOR NORMAL-WEIGHT CONCRETE
Inch-Pounds SI
a
fc′ (psi) n fc′ (MPa) n
3000 9 20 9
3500 8.5 25 8
4000 8 30 7.5
4500 7.5 35 7
5000 7 40 6.5
6000 6.5
a For practical use, multiply MPa by 10 to obtain value in kgf/cm2.

405
12.4 EQUILIBRIUM CONDITIONS 405
The resultant compressive force may be entirely from concrete stresses, or it may be from
a combination of the stresses in concrete and those in the compression reinforcement. The
resultant tensile force, of course, is provided entirely by the tension reinforcement, since
the contribution of the concrete in tension is neglected.
EXAMPLE 12.4.1
Using the equilibrium conditions, determine the stresses under service loads in the ten-
sion steel reinforcement and on the extreme compression fiber of concrete for the beam
of Example 3.9.1. The beam is simply supported with a span of 40 ft and was designed to
carry a live load of 1.38 kips/ft and a dead load of 1 kip/ft (not including beam weight),
with fc′ = 4000 psi, and fy = 60,000 psi. The final design called for an 18 × 34 in. cross
section reinforced with 2–#9 and 4–#10 bars as tension steel (see Fig. 3.9.2). Ignore the
presence of the top bars in the compression zone.
SOLUTION
(a) Compute the applied moment under service-level loads. From part (a) of the solution to
Example 3.9.1, the moment under service live loads, ML = 276 ft-kips, and from part (f),
the moment due to service dead loads (including self-weight), MD = 328 ft-kips.
Therefore, the applied moment under service-level loads is

Mservice = M D + M L = 328 + 276 = 604 ft-kips
b = 18”
εc fc = Ec εc
c/3
c C
NA
d = 31.5”
c
31.5 –
3
4 #10
No tensile stresses except
in steel reinforcement
εs fs = Es εs T
2 #9
(a) Section (b) Strains (c) Stresses
Figure 12.4.1 Assumed strain and stress distributions for the beam of Example 12.4.1 under
service loads.
(b) Compute stresses under service loads. The assumptions of linearly varying strains
and stress proportional to strain are shown in Fig. 12.4.1(b) and 12.4.1 (c). The first
step in the solution is to locate the neutral axis (N.A.). The internal compressive
force is obtained by integration of the stress times the area on which it acts. This is
equivalent to computing the volume of the stress solid. Thus, the internal compres-
sive force in the concrete is,
1
C= fc bc = 9.0 fc c
2
The internal tensile force is
T = fS As = fs [ 2(1.00) + 4(1.27)] = 7.08 fs
(Continued)
406
Equating C to T gives
fs 9.0c
=
fc 7.08
The ratio of fs to fc may also be obtained by using the linear strain relationship and
taking stress proportional to strain, as shown in Fig. 12.4.1,
ε s 31.5 − c
=
εc c
fs Es ε s ε
= = n s
fc Ec ε c εc
where n is the modular ratio defined in Section 12.3. From Table 12.3.1, choose n = 8
for fc′ = 4000 psi. Therefore,
fs ε  31.5 − c 
= n s = 8 
fc εc  c 
Note that the actual values of Es and Ec are not needed. Equating the two expressions
for fs /fc,
9.0c 8(31.5 − c)
=
7.08 c
9c 2 = 56.64(31.5 − c)
c 2 + 6.29c = 198.24
c = 11.28 in.
Note that other than the ratio of the moduli of elasticity, only the properties of the
section (depth, width, and steel area) affect the position of the neutral axis. The loading
does not affect the neutral axis location.
The moment arm of the internal couple, or the distance from the centroid of the com-
pressive solid to the centroid of the tension steel, is
c 11.28
arm = 31.5 − = 31.5 − = 27.74 in.
3 3
Then, using the second equilibrium condition,

Mservice = 604 ft-kips = 7248 in.-kips = (C or T ) × arm
Mservice 7248
C=T = = = 261.3 kips
arm 27.74
The stresses are determined from the expressions for C and T,
C 261, 300
fc = = = 2574 psi [or ≈ 0.64 fc′]
9c 9(11.28)

T 261, 300
fs = = = 36, 907 psi [or ≈ 0.62 f y ]
7.08 7.08
(Continued)
407
12.5 METHOD OF TRANSFORMED SECTION 407
The stresses in the tension steel reinforcement are less than the yield strength, and there-
fore the assumption that the behavior of the reinforcement is linear elastic is correct. As
noted earlier, behavior of the concrete is nonlinear even at low stress levels, but it can be
assumed to behave linearly up to stresses of approximately of 0.5 fc′ . In this example, the
maximum stress is greater than this value, and therefore it may be expected that concrete
behavior near the top of the beam will be nonlinear. In reference to Fig. 1.10.1, it may be
seen, however, that while behavior of the concrete at 0.64 fc′ is nonlinear, the assumption
of linear elastic behavior is still reasonable. In effect, it may be shown that an analysis of
the beam cross section using a nonlinear stress-strain distribution for the concrete will
yield results similar to those obtained with the linear elastic assumption.
12.5 METHOD OF TRANSFORMED SECTION

In the method of the transformed section, the cross section composed of steel reinforcement
and concrete is transformed into an equivalent isotropic, homogeneous section of either
concrete or steel alone. Most commonly, the steel is transformed into an equivalent area
(i.e., imaginary area) of concrete where the transformed section is treated as a cross section
made entirely of concrete. In the transformation, two conditions must be satisfied.
Consider in Fig. 12.5.1(a) a reinforced concrete section in a tension zone where the con-
crete is assumed to be cracked. Let As and At be the areas of the actual steel reinforcement
and the transformed, equivalent concrete, respectively. Similarly, let fs and ft be the tensile
stresses in the actual steel and the equivalent concrete, respectively. First, the equilibrium
condition requires that the total tensile force be the same in the actual and the transformed
section, or
As fs = At ft (12.5.1)
where the contribution of the cracked concrete is ignored. Second, compatibility of defor-
mations requires that the strain be the same, or
fs f
= t (12.5.2)
E s Ec
As
a nAs
Actual section Transformed section

(a) Actual and transformed area of steel in a tension zone
(cracked concrete)
As’
a (n–1)As’
b
Actual section Transformed section
(b) Actual and transformed area of steel in a compression zone
Figure 12.5.1 Transformed areas of steel (a) in a tension zone and (b) in a compression zone.
408
Utilizing the modular ratio n = Es /Ec and combining Eqs. (12.5.1) and (12.5.2), the equiva-
lent area of concrete (transformed area of steel) is
At = nAs (12.5.3)
and the corresponding stress is
fs
ft = (12.5.4)
n
Thus, the equivalent concrete area At is n times the actual steel area in tension, and the ten-
sile stress ft (i.e., imaginary stress) of the equivalent area of concrete is 1/n times the actual
tensile steel stress.
Similar expressions may be obtained for a cross section in the compression zone where
the steel is in compression and the concrete is subjected to a stress fc. As before, let As′ and
At be the areas of the actual steel reinforcement and the transformed, equivalent concrete,
respectively [Fig. 12.5.1(b)]. Similarly, let fs′ and ft be the compressive stresses in the actual
steel and the equivalent concrete, respectively. Equilibrium requires that the total compres-
sive force be the same in the actual and the transformed section, or
As′ fs′+ Aconc fc = At ft + abfc
where Aconc = ab − As′

Therefore,
As′ fs′+ (ab − As′ ) fc = At ft + abfc

 f′ 
As′  s − 1 fc = At ft (12.5.5)
f
c 
Compatibility of deformations require that the strain be the same in the concrete and the
steel in the actual section and in the equivalent area of concrete in the transformed section, or
fs′ f f (12.5.2′)
= t = c
E s Ec Ec
Rearranging and substituting Eq. (12.5.2′) in Eq. (12.5.5),
E 
As′  s − 1 = At
 Ec 
or
As′ (n − 1) = At (12.5.6)
and
fs′ (12.5.7)
ft =
n
In other words, the transformed, equivalent concrete area At is (n–1) times the actual steel
area in compression, and the compressive stress ft (i.e., imaginary stress) is 1/n times the
actual compressive steel stress.
By using Eqs. (12.5.3), (12.5.4), (12.5.6), and (12.5.7), a reinforced concrete section
may then be treated as a section of one material (concrete). These equations are very use-
ful whenever elastic properties are needed—for example, to compute deflections under
service-level loads.
409
12.5 METHOD OF TRANSFORMED SECTION 409
EXAMPLE 12.5.1
Use the transformed section method to locate the elastic neutral axis and to compute the
stresses in the tension steel reinforcement and in the extreme fiber in compression in the
concrete for the beam of Example 12.4.1 under service loads.
SOLUTION
(a) Locate the elastic neutral axis using transformed area [Fig. 12.5.2(b)]. For
fc′ = 4000 psi, the modular ratio n may be taken as 8 (see Table 12.3.1). The area
18.0c above the neutral axis is in compression, whereas the equivalent tension area
nAs = 8(7.08) = 56.64 sq in. is assumed to be concentrated at the centroid of the steel
reinforcement at a distance (31.50 –c) below the neutral axis.
Equating the first moments of the compression and tension areas about the neutral or
centroidal axis of the transformed section,
1
(18.0)c 2 = 56.64 (31.50 − c)
2
c = 11.28 in.
Figure 12.5.2 Transformed section for the beam of Example 12.5.1.
(d) Use the transformed section method to find stresses. The transformed cracked sec-
tion moment of inertia about the neutral axis (centroid of the transformed section) is
1
I cr = (18.0)(11.28)3 + 56.64(20.22)2
3
= 8611 + 23157 ≈ 31, 770 in. 4
The applied moment under service loads was computed as 604 ft-kips [part (a) of
Example 12.4.1]. By the flexure formula, the stress in the extreme fiber in compression
in the concrete at the top is
604 (12, 000 )
fc = (11.28) = 2573 psi
31, 770
and in the equivalent area of concrete (transformed area of steel),
604 (12, 000 )
ft = ( 20.22) = 4613 psi
31, 770
The actual stress in the steel, fs, is found from Eq. (12.5.4) as

fs = nft = 8(4613) = 36, 904 psi
The values computed for fc and fs are the same as those computed in Example 12.4.1, as
was expected.
410
EXAMPLE 12.5.2
Compute the moment of inertia Icr of the transformed cracked section to be used in
deflection calculation for the doubly reinforced section shown in Fig. 12.5.3, using
fc′ = 4000 psi (n = 8).
SOLUTION
(a) Locate the neutral axis. The steel is transformed into equivalent concrete by using
At = nAs for the steel in tension [Eq. (12.5.3)] and At = (n − 1) As′ for the compression
steel [Eq.(12.5.6)]. Referring to Fig. 12.5.3, and equating the first moments of the
compression and tension area,
13c 2
+ 10.5(c − 2.5) = 56.64(22 − c)
2
Solving for c,

c = 9.76 in.
(b) Determine the transformed cracked section moment of inertia.
1
I cr = (13)(9.76)3 + 10.5(9.76 − 2.5)2 + 56.64(22 − 9.76)2
3

= 4030 + 550 + 8480 = 13,100 in.4
(n – 1)As’ = 10.5 sq in.
13”
2.5”
c = 9.76”
2 – #6
2 – #5
As’ = 1.50 sq in. 22”
4 – #10 nAs = 56.64 sq in.

2 – #9
As = 7.08 sq in.
Figure 12.5.3 Doubly reinforced section of Example 12.5.2.
12.6 DEFLECTIONS—G ENERAL
Over the years, the use of higher strength concrete has allowed a more efficient design of
reinforced concrete members, which in turn has resulted in smaller, shallower sections.
These smaller sections, while still providing the required strength, have less stiffness and
consequently exhibit larger deflections. In fact, control of deflections rather than strength is
often the governing design criterion for many structures built today.
As noted earlier, the permissible deflection is governed by the serviceability require-
ments for the structure, such as the amount of deformation that can be tolerated by the
interacting components of the structure. Excessive deflection of the member may not, in
itself, be detrimental, but the effect on nonstructural (and structural components) that are
supported by the deflecting member frequently determines the acceptable amount of deflec-
tion. Both the short-term (instantaneous or immediate) and the long-term (time-dependent)
effects must be considered. The acceptable deflection depends on many factors, among
which are the intended use or function of the building (warehouse, school, factory, resi-
dence, etc.), the types of finish (presence of brittle tiles, plastered ceilings, brick veneer,
etc.), the type and arrangement of partitions, the sensitivity of equipment to deflection, and
41
1 2 . 7 D E F L E C T I O N S F O R L I N E A R E L A S T I C M embers 411
the magnitude and duration of the transient loads (i.e., live load, snow load, or rain loads).
Note that even in situations in which excessive deflections are not likely to cause damage,
large noticeable deflections could be psychologically disturbing to humans and thus should
be avoided.
The ACI Code provides guidance for computing the minimum member depth to avoid
excessive deflections, and also provides limits on the maximum permitted deflections for
common situations. For nonprestressed members “not supporting or attached to partitions
or other construction likely to be damaged by large deflections,” ACI-Table 9.3.1.1 pro-
vides the minimum depth required for beams (see Section 12.15). Similar requirements for
the minimum depth of one-way slabs are given in ACI-Table 7.3.1.1. When these minimum
depth requirements are not satisfied or if excessive deflection may cause a problem, the
calculated immediate and time-dependent deflections must satisfy the limits of ACI-24.2.2.
Accordingly, the maximum acceptable immediate deflections due to live load, for flat roofs
or floors that do not support and are not attached to nonstructural elements such as plastered
ceilings or frangible partitions likely to be damaged by large deflections, are prescribed by
ACI Table 24.2.2 to be (a) for flat roofs, L /180, and (b) for floors, L /360. Further, in recog-
nition of the increase of deflection with time, ACI Table 24.2.2 also prescribes limits for the
sum of the time-dependent deflection due to all sustained loads plus any immediate deflec-
tion due to any additional live load. The two stated limits for this deflection combination
are (a) L /480 when nonstructural elements are likely to be damaged, and (b) L /240 when
no damage to nonstructural elements is likely.
Limitations on deflection are somewhat arbitrary. Historically, L /360 has been the
accepted limit to prevent the cracking of plastered ceilings. Other limits should be consid-
ered as guidelines, with the designer having the responsibility for evaluating the possible
adverse effects of excessive deflection in any given situation.
Recommendations for allowable deflections for a great variety of situations are provided
by ACI Committee 435 [12.1–12.6].
The general concepts dealt with in this chapter are applicable to both one-way (beams
and slabs) and two-way systems. Specific examples are applied only to one-way systems.
The reader is referred to Section 16.20 for a brief discussion of two-way system deflections,
along with specific references.
12.7 DEFLECTIONS FOR LINEAR ELASTIC MEMBERS

Various methods are available in structural analysis for computing deflections of members
with uniform or variable moment of inertia in statically determinate and indeterminate
structures. In general, the maximum deflection in an elastic member may be expressed as
ML2
∆ max = β a (12.7.1)
EI c
where
M = a reference value of bending moment such as the maximum positive value
L = span length
E = modulus of elasticity
Ic = moment of inertia of member cross section
βa = a coefficient that depends on the degree of fixity at supports, the variation in moment
of inertia along the span, and the distribution of loading; such as 5 48 for simply sup-
ported uniformly loaded beam; 1 4 for uniformly loaded cantilever
The deflection coefficient βa for simple cases may be found in handbooks, such as the ACI
Design Manual [2.20], or Handbook of Concrete Engineering [12.7].
The following example uses one elastic method, the conjugate beam method, to derive
one of the most useful deflection expressions.
412
EXAMPLE 12.7.1
Using the conjugate beam method, derive the general expression for the elastic midspan
deflection for a uniformly loaded span with unequal end moments, as shown in Fig. 12.7.1.
SOLUTION
In the conjugate beam method, the deflection at a given point equals the bending moment
at that point for a beam loaded with the M/EI diagram. Thus the system of Fig. 12.7.1 may
be regarded as being composed of three separate conjugate beams, as shown in Fig. 12.7.2.
w
Ma Mb
Ms
wL2
Mo =
+ 8
– –
Constant I
Ma
Mb
L/2 L/2
Figure 12.7.1 Typical bending moment diagram for a uniformly loaded span of a continuous beam.
Mo
El
+
Simply supported
uniform load
L/2
L
(a)
Ma
El
– End moment at left

end of span
(b)
Mb
El
– End moment at right

end of span
(c)
Figure 12.7.2 Component conjugate beams. (Continued)

413
1 2 . 7 D E F L E C T I O N S F O R L I N E A R E L A S T I C M embers 413
The midspan deflections, or the bending moments on the three conjugate beams, are
2  L   M 0   L 3  L   5 M 0 L2
uniform load, ∆ s =   −   =
3  2   EI   2 8  2   48 EI
1  − Ma   L   L  1  − Ma   L   1   L 
left end moment, ∆ a =   −    
3  EI   2   2  2  2 EI   2   3   2 

− M a L2
=
16 EI
− M b L2
right ennd moment, ∆ b =
16 EI
The total midspan deflection Δm is
∆m = ∆s + ∆a + ∆b
5 M 0 L2 M a L2 M b L2
= − − (12.7.2)
48 EI 16 EI 16 EI
L2
=
48 EI
[5M0 − 3( M a + Mb )]
The net positive midspan moment Ms is
1
M s = M 0 − ( M a + M b ) (12.7.3)
2
Then, on substituting Eq. (12.7.3) in Eq. (12.7.2), one obtains
L2  5 
∆m = 5 M s + 2 ( M a + M b ) − 3( M a + M b )
48 EI  
(12.7.4)
5L2  1 
=  M s − 10 ( M a + M b )
48 EI  
Equation (12.7.4) may be used with satisfactory results for practically all prismatic (i.e.,
constant EI) beams, even though the absolute maximum deflection will be obtained only
when the loading is uniform and the end moments are equal.
Continuous Beams Having Variable Flexural Rigidity

When significant changes of gross cross section along the span length are involved, such
changes in cross-sectional properties should be included when statically indeterminate
analyses are performed. When deflections are desired for such cases, the best approach is
to account explicitly for the variable flexural rigidity EI, either mathematically in the com-
ponent conjugate beams (see Fig. 12.7.2), or by the use of numerical integration. Section
12.9 treats this subject further in a discussion of what value of the moment of inertia should
be used for reinforced concrete sections.
41
12.8 MODULUS OF ELASTICITY
As discussed in Section 1.10, variance in interpretation is encountered in references to the
modulus of elasticity of concrete. Ordinary beam theory presumes the modulus in tension to
be the same as that in compression for a homogeneous material. In a reinforced concrete sec-
tion, creep affects the apparent modulus primarily in the compression zone. Even the modulus
under short-duration loading, measured as the secant modulus, is considerably more variant
than the compressive strength, fc′. The secant modulus in tension is essentially the same as
that in compression when the stress magnitude is low, but the modulus reduces markedly as
the stress nears the cracking level. Also, in regions of high bending moment the concrete will
be cracked in the tension zone, while at low moment sections the concrete may be uncracked.
In other words, the actual modulus of elasticity in both tension and compression varies not
only with the magnitude of stress from top to bottom at a section, but also along the span.
Furthermore, creep and shrinkage over a period of time exert an effect equivalent to
reducing the modulus in compression and will generally magnify deflections by a factor of
2 to 3. Referring to Fig. 1.11.1, one may note that the true elastic strain decreases slightly
with time, indicating that the modulus of elasticity increases. The modulus increases
because it is dependent on fc′, which increases with age. In design practice, the apparent
elastic strain used is often computed with a code-specified Ec, presumably corresponding to
the 28-day age at loading. Beyond 28 days, however, the increase in modulus of elasticity
is relatively slight.
12.9 EFFECTIVE MOMENT OF INERTIA

In addition to the modulus of elasticity, the flexural rigidity includes the cross-sectional
property, moment of inertia. Even for sections whose external dimensions are constant, the
moment of inertia varies considerably along a span. Consider a span with a T-shaped sec-
tion in a continuous beam or frame, as shown in Fig. 12.9.1. At or near the support in such
a continuous structure, the concrete slab is cracked, so that the effective section is section
A-A; while in the positive moment zone it is the stem that is cracked, and the effective section
is section B-B. Furthermore, near points of inflection the entire section may be uncracked
and fully effective, as shown in section C-C. The problem is further compounded because
the amounts of reinforcement at the top and bottom of the beam vary along the span.
A C B A
w w
A C B A
Section A–A Section B–B
Section C–C
Figure 12.9.1 Effective moment of inertia for continuous T-shaped sections.

415
12.9 EFFECTIVE MOMENT OF INERTIA 415
In an elastic continuity analysis as discussed in Chapters 7 and 9, only relative stiffness

values are required; in deflection computations, however, the absolute magnitudes for Ec and
I must first be determined or assumed. Moreover, the amount of deflection under each incre-
ment of load is not constant. The flexural rigidity Ec I is greater at a low load level, since the
full uncracked section provides the greatest effective moment of inertia. As the load increases,
cross sections will gradually crack along the span, thereby reducing the flexural rigidity Ec I.
The subject of effective moment of inertia for computing deflections due to short-term
loading was studied by Yu and Winter [12.8], and by Branson [12.9, 12.10]. It is known that
the flexural rigidity Ec I varies with the magnitude of bending moment in the general manner
shown in Fig. 12.9.2. Of course, the moment of inertia Icr of the transformed cracked section
(see Section 12.5) increases roughly proportional with an increase in the percentage of rein-
forcement. Sections with higher percentages of reinforcement exhibit less change in rigidity
under increasing load than those with low percentages of reinforcement. Duan, Wang, and
Chen [12.11] reviewed and proposed an alternative expression for flexural rigidity.
For loads below the cracking load (see Fig. 12.9.3), deflections may be based on the
gross concrete section, with generally a small difference arising from whether the trans-
formed area of reinforcement is also included. However, as the load increases above the
cracking load, the moment of inertia approaches that of the cracked transformed section,
although it may be greater between cracks.
Generally, as shown by the typical load-deflection curve in Fig. 12.9.3, the use of gross
section underestimates the deflection, and the use of transformed cracked section overesti-
mates the deflection. However, the degree of accuracy is affected by the magnitude of the
service load compared to the load that causes cracking. Great accuracy in deflection com-
putations is rarely justified in design because (a) it is difficult to estimate the actual concrete
properties, in particular, its time-dependent properties associated with creep and shrinkage;
(b) deflection limits are rather arbitrary; and (c) except for dead loads, the magnitude and
duration of actual loads are very difficult to predict with accuracy.
EcI based on gross section plus

transformed area of reinforcement
EcI
EcI based on cracked

transformed section
0.2 Mu Mu
Figure 12.9.2 Typical variation of flexural rigidity with applied bending moment.
Service load
Computed deflection based on
transformed cracked section
Load
Computed deflections using gross I
Actual deflection
Cracking load
Deflection
Figure 12.9.3 Typical load-deflection curve for reinforced concrete beams under service loads.
416
ACI Effective Moment of Inertia Ie

To obtain an estimate of the flexural rigidity of the member, the ACI Code uses the expres-
sion developed by Branson [12.9] given as ACI Formula (24.2.3.5a),
 M cr 
3
  M 3
Ie =  I g + 1 −  cr   I cr ≤ I g (12.9.1)
 M max    M max  
where
Mcr = fr Ig /yt = cracking moment [ACI Formula (24.2.3.5b)]
Mmax = maximum service load moment acting at the condition under which deflection
is computed
Ig = moment of inertia of gross uncracked concrete section about the centroidal axis,
neglecting reinforcement
Icr = moment of inertia of transformed cracked section (see Section 12.5)
fr = modulus of rupture of concrete (see Section 1.8), taken by ACI Code as 7.5λ fc′;
in general, may be taken as 0.65 wc fc′, where wc is the unit weight of concrete in
lb/cu ft [12.1]
yt = distance from neutral axis to extreme fiber of concrete in tension
Equation (12.9.1) provides a smooth transition between the moment of inertia Icr of the
transformed cracked section and the moment of inertia Ig of the gross uncracked concrete
section. The effective moment of inertia, Ie, is intended to be calculated at the location of
maximum moment as a single value for the entire span in the case of simply supported
beams, or as a single value between points of inflection in continuous beams. If one wishes
to recognize the continuous variation of the moment of inertia along the span, Branson
proposed using the fourth power instead of the third power in Eq. (12.9.1). In this case, Mcr
and Mmax are the cracking moment and the applied moment, respectively, at each section
along the span. The span may be broken into segments, with each segment having a differ-
ent moment of inertia and numerical integration used to compute the deflection.
Equation (12.9.1) was developed from a statistical study of 54 test specimens that had
Mmax /Mcr values from 2.2 to about 4 and Ig /Icr values from 1.3 to 3.5. The study included
simple-span rectangular beams [12.12], T-beams [12.8], and two-span continuous rectan-
gular beams [12.13]. Branson has provided an excellent summary [12.10, 12.14, 12.15] of
background material relating to Eq. (12.9.1). Lutz [12.16] has provided an ingenious set of
diagrams to allow graphical evaluation of Eq. (12.9.1), and Shahrooz [12.17] has provided
a simplified method for computing Ie.
The recommendation of ACI Committee 435 to compute Ig neglecting reinforcement
was made for practical simplicity rather than for any improvement in accuracy. Al-Shaikh
and Al-Zaid [12.18] have studied the effect of the reinforcement ratio on Ie and suggested a
modification to the power of 3 in Eq. (12.9.1) based on the reinforcement ratio ρ.
The effective moment of inertia approach is also applicable to prestressed concrete (see
Chapter 20); Branson and Trost [12.19] have presented unified procedures using the Ie
method for partially cracked members, whether prestressed or nonprestressed.
Single Value of Effective Moment of Inertia for Practical Use

As an approximation, a single value of effective moment of inertia is suggested for practi-
cal use. Various methods have been suggested depending on the support conditions and the
expected cracking on the member along the span.
1. Midspan value alone. Recognized in ACI-24.2.3.7 for prismatic beams and one-way
slabs, this assumption is
I e = I m (12.9.2)
where Im is the effective moment of inertia at midspan for prismatic simply supported and
continuous spans, and at the support section for cantilevers. This is the simplest method, and
417
1 2 . 1 0 I N S TA N TA N E O U S D E F L E C T I O N S I N D E S I G N 417
when the inflection point is between about 0.2L and the support, the results are within ±20%
of those obtained using variable I, as long as 0.33 ≤ α ≤ 1.0 [where α = (Im at midspan) /(Ie at
end)]. When α lies between 0.50 and 1.0, the results are within ±5% of those obtained using
variable I. Zuraski, Salmon, and Shaikh [12.20] have suggested that this method is satisfac-
tory for ordinary design situations, and it is endorsed by ACI Committee 435 [12.1].
2. Simple average. For continuous members, ACI-24.2.3.6 permits the moment of iner-
tia to be taken as the average of the values for the critical positive and negative
moment sections. This average is to be taken as
1
( I e1 + I e 2 ) + I m
average I e = 2 (12.9.3)
2
where Ie1 and Ie2 are the effective moments of inertia at the two ends of the span. The use
of both Ie1 and Ie2 is appropriate only when there are end moments at both ends. For spans
having one end continuous, use
average I e = 0.75I m + 0.25I e1 (12.9.4)
3. Weighted average. In this method, an adjusted I is obtained by weighting the moments

of inertia in accordance with the magnitudes of the end moments [12.3]. The follow-
ing weighted average expression has been recommended by ACI Committee 435
[12.1, 12.3] as giving a somewhat better result than use of the simple average. For
spans with both ends continuous,
average I e = 0.70 I m + 0.15[ I e1 + I e 2 ] (12.9.5)
For spans with one end continuous,
average I e = 0.85I m + 0.15I e1 (12.9.6)
For uniform loading on continuous spans, Eq. (12.9.5) is slightly more accurate (say, on
the order of 5%) than using only the midspan value, but for concentrated loads it is less
accurate [12.20]. It may be emphasized that when an average value is used as permitted by
ACI-24.2.3.6, this should be done in accordance with Eq. (12.9.3), rather than taking the
sum of Im, Ie1, and Ie2 and dividing by 3 [12.3].
For a single heavy concentrated load, averaging reduces accuracy [12.20]; the midspan
value from Eq. (12.9.2) should be used in such cases.
12.10 INSTANTANEOUS DEFLECTIONS IN DESIGN

Throughout the history of reinforced concrete construction, computation of immediate
or short-term deflection has usually involved using either transformed cracked section or
gross uncracked section. In either case, Eq. (12.7.1) is suitable after it has been slightly
rewritten using Ec and Ie:
 ML2 
∆ = βa  (12.10.1)
 Ec I e 
where
βa = coefficient based on load and support conditions
Ie = effective moment of inertia
Ec = modulus of elasticity of concrete
418
For immediate deflection and also generally for long-term deflection under sustained
loads, ACI-24.2.3.4 permits the modulus of elasticity of concrete to be taken in accordance
with ACI-19.2.2 as
Ec = 33w1c,5 fc′
for concrete having a unit weight between 90 and 160 pcf. For normal-weight concrete,
Ec = 57, 000 fc′
The generally accepted effective moment of inertia for use in Eq. (12.10.1) is
Eq. (12.9.1), using a single value in accordance with Eqs. (12.9.2) through (12.9.6).
To compute deflection at different load levels, such as dead load or dead load plus live
load, the effective moment of inertia Ie should be computed by means of Eq. (12.9.1) for
the total load level in each case. The incremental deflection, such as for live load only,
is then computed as the difference between the deflections due to dead plus live load
and dead load only. It should be assumed that the live load cannot act in the absence of
dead load.
Computation of the live load deflection as ∆D+L –∆D gives the live load deflection occur-
ring during the first application of live load. Figure 12.10.1 shows the typical idealized
relationship between moment and deflection. For repeated loadings, the upper envelope
is nearly the same as the single-loading curve for both reinforced and prestressed con-
crete members, even though creep and cracking effects cause increasing residual deflec-
tions [12.2, 12.22]. Thus it seems reasonable to compute short-term deflections using Ie as
described above and the residual deflection separately, as discussed in Sections 12.13 and
12.14 on creep and shrinkage.
Figure 12.10.2 compares the measured short-term deflections with the deflections
computed by Eq. (12.9.1) [12.6]. A study by ACI Committee 435 [12.4] indicates that
by using the Code criteria for deflection for simply supported beams under controlled
laboratory conditions, “there is approximately a 90% chance that the deflections of a
particular beam will be within the range of 20% less-than to 30% more-than the calcu-
lated value.”
Examples 12.10.1 and 12.10.2 demonstrate the computation of immediate deflections.
Figure 12.10.1 Typical idealized moment deflection diagram for short-term loading. (Adapted
from Refs. 12.2, 12.14, and 12.21.)
419
%
20
2.0”
%
20
Measured deflections
1.0”
Branson
(ACI–since 1971)
Computed deflections
0 1.0” 2.0”
Simply supported rectangular beams

Simply supported T beams
2 span continuous rectangular beams
Figure 12.10.2 Comparison of computed and measured short-term deflections (From Reference 12.6.).
EXAMPLE 12.10.1
Investigate the immediate, short-term loading deflection for the simply supported beam
of Fig. 12.10.3 over a span of 40 ft. Assume that the member has been designed to sat-
isfy the strength requirements of the ACI Code using fc′ = 4000 psi (normal weight) and
fy = 60,000 psi.
SOLUTION
According to ACI Table 9.3.1.1 the minimum required depth for the beam is computed
as L /16 = 40(12)/16 = 30 in., which is greater than the beam depth of 24 in. Therefore,
the deflection limits of ACI Table 24.2.2 must be satisfied regardless of whether exces-
sive deflection is of concern.
(a) Dead load short-term deflection. The gross moment of inertia is
1
Ig = (18)(24)3 = 20, 700 in.4
12

0.45(40)2
M max = = 90 ft-kips
8
Using a modulus of elasticity ratio n = 8 (see Table 12.3.1), the neutral axis position for
the transformed cracked section shown in Fig. 12.10.3(b) is
18c 2
= 57.3(20.7 − c)
2
c 2 + 6.37c = 131.8

c = 8.73 in.
1
I cr = (18)(8.73)3 + 57.3(20.7 − 8.73)2 = 12, 200 in.4
3
(Continued)
420
The effective moment of inertia Ie is dependent on the bending moment Mcr that causes
cracking at the extreme tension fiber. For normal-weight concrete, the modulus of con-
crete is computed as (ACI-19.2.3)

fr = 7.5λ fc′ = 7.5(1.0) 4000 = 474 psi
The cracking moment is computed from ACI Formula (24.2.3.5b)

fr I g 0.474(20, 700)  1 
M cr = =  12  = 68 ft-kips
yt 12
Note that yt is the distance h/2 for the 24-in.-deep beam when the gross section is used
and the reinforcement is neglected.
3
M cr 68  M 
= (dead load only) = 0.756;  cr  = 0.431
M max 90  M max 
From ACI Formula (24.2.3.5a), Eq. (12.9.1), the effective moment of inertia is
 M 
3
  M 3
I e =  cr  I g + 1 −  cr   I cr
 M max    M max  
= 0.431(20, 7000) + 0.569(12, 200) = 15, 860 in.4

Ec = 33wc1.5 fc′ = 33(145)1.5 4000 = 3.64 × 106 psi
5wL4 5(0.45)(40)4 (12)3

(∆ i )D = = = 0.45 in.
384 EI 384(3.64)(103 )15, 860
c = 8.73”
Figure 12.10.3 Beam for Example 12.10.1

(Continued)
421
This deflection may not be harmful because it may be accommodated by creating an

upward deflection or camber during construction. Even if camber is not used, such dead
load instantaneous deflection will not affect plastered ceilings or other items that are put
into place after the immediate dead load deflection has taken place. Concern is primarily
with the instantaneous deflection from live load and the long-term creep and shrinkage
deflection from sustained loads.
(b) Dead load plus live load short-term deflection. The maximum service load moment
at this load level is
25(40)
M max = + 90 = 340 ft-kips (for dead load plus live load)
4
3
M cr 68  M cr 
= (dead load + live load) = 0.20;  M  = 0.008
M max 340 max
I e = (0.008)(20, 700) + (0.992)(12, 200) = 12, 270 in.4
At this higher load level, Ie is only slightly larger than Icr. Using Ie = Icr = 12,200 in.4,
5(0.45)(40)4(12)3
( ∆ i )beam weight = = 0.58 in.
384(3640)(12, 200)
25(40)3(12)3
( ∆ i )conc load = = 1.30 in.
48(3640)(12, 200)
( ∆ i ) D + L = 0.58 + 1.30 = 1.88 in.
(c) Live load short-term deflection. Consistent logic dictates that live load deflection be
obtained indirectly as
( ∆ i )L = ( ∆ i )D + L − ( ∆ i )D

= 1.88 − 0.45 = 1.43 in.
It is assumed that live load cannot act in the absence of dead load. Thus if the effective
moment of inertia when dead load alone is acting is considerably different from that
when dead load plus live load is acting, the live load deflection is properly obtained
only by subtracting ΔD from ΔD+L. Even if the service live load has been preceded by a
construction load of equal magnitude, the procedure is proper for repeated loadings, as
discussed in the literature [12.14, 12.19, 12.21].
From a practical viewpoint, Ie = Icr may be used whenever (Mcr /Mmax)3 is less than
about 0.1. Assuming this to be a floor beam that is not supporting partitions, the accept-
able immediate live load deflection from ACI Table 24.2.2 is
L 40(12)
allowable (∆ i )L = = = 1.33 in. < 1.43 in. NG
360 360
If plastered ceilings or frangible partitions are to be supported, the long-term deflec-

tion due to creep and shrinkage must be added to that due to live loads; the accept
able limit for such deflection is L /480 (ACI Table 24.2.2). One might expect that the
deflection of this beam would be excessive since the reinforcement ratio ρ is 0.0192.
This amount, corresponds to 1.06ρtc, which greatly exceeds the guideline value of about
0.6ρtc, or 0.375ρb, suggested in Chapter 3 as the maximum reinforcement ratio for
deflection control.
42
EXAMPLE 12.10.2
Compute the immediate deflection of the continuous beam of Fig. 12.10.4. Use
fc′ = 3000 psi (normal weight), fy = 60,000 psi, and the ACI Code.
SOLUTION
This continuous girder supports two smaller beams that frame to it and some uniform
loading that comes directly to it. ACI-24.2.3.6 indicates that Ie may be computed as
an average value for the critical positive and negative moment regions. For prismatic
members, the effective moment of inertia Ie at the positive moment region may be used
instead of the average value. In fact, for loading that is largely concentrated load, as in
the situation for this example, ACI Committee 435 recommends [12.1, 12.2] against
using an average value. The results will be compared using the various assumptions for
the single value of Ie, as discussed earlier (see Section 12.9).
(a) Section at the left support. For the gross section, neglecting reinforcement (see
Fig. 12.10.5),
36(18)18 + 90(4.5)38.25 27,160
c= = = 25.79 in.
36(18) + 90(4.5) 1053
1 1
I g = (18) (25.79)3 + (14.71)3  + (72)(4.5)3 + (72)(4.5)(12.46)2
3 12
= 173, 000 in.4

If the reinforcement is included, c would have been 25.29 in. and Ig 200,000 in.4 It is
generally accepted that Ig for use in ACI Formula (24.2.3.5a) should not include steel
reinforcement.
For the transformed cracked section [see Fig. 12.10.6(a)],
 c
18c   + 40.6(c − 2.6) = 40.9(36.7 − c)
 2
c 2 + 9.05c = 178.2
c = 9.56 in
n.
1
I cr = (18)(9.56)3 + 40.6(9.56 − 2.6)2 + 40.9(36.7 − 9.56)2
3
= 37, 200 in.4
It is noted that a modulus of elasticity ratio n = 9 (see Table 12.3.1) is used to transform
the area of steel into an equivalent area of concrete.
Compute the cracking moment for the beam with tension on the flange.
fr = 7.5λ fc′ = 7.5(1.0) 3000 = 411 psi
fr I g 0.411(173, 000)  1 
M cr = =  12  = 402 ft-kips
yt 14.71
M cr 402
= (dead load) >1; use I e = I g = 173, 000 in.4
M max 89.6
M cr 402
= (dead load +live load) >1;; use I e = I g = 173, 000 in.4
M max 200
(Continued)
423
(b) Midspan section [Fig. 12.10.6(b)]. Determine whether the neutral axis is located in
the flange by taking moments about the bottom of the flange,

90(4.5)(2.25) < 86.6(32.3)
Thus, the neutral axis is in the stem. Locate the neutral axis, including the effect of
compression in the stem.
 1
90(4.5)(c − 2.25) + 18(c − 4.5)2   = 86.6(36.8 − c)
 2
c 2 + 45.6c = 435
c = 8.09 in.
Figure 12.10.4 Data for Example 12.10.2. (Continued)

42
bE = 90”
4 12 ”
36”
c = 25.79”
18”
Figure 12.10.5 Cross section of beam for Example 12.10.2.
Figure 12.10.6 Transformed cracked sections for Example 12.10.2.
1 1
I cr = (72)(4.5)3 + 72(4.5)(5.84)2 + (18)(8.09)3 + 86.6(36.8 − 8.09)2
12 3
= 86, 200 in.4
The cracking moment for the beam with tension in the stem is
fr I g 0.411(173, 000)  1 
M cr = =  12  = 229.5 ft-kips
yt 25.79
M cr 229.5
M max 229
3
M cr 229.5  M 
= (dead load + live load) = 0.45;  cr  = 0.091
 M 
3
  M 3
I e =  cr  I g + 1 −  cr   I cr
 M max    M max  
= 0.091(173, 0000) + 0.909(86, 200) = 94, 000 in.4

(Continued)
425
(c) Section at the right support [Fig. 12.10.6(c)]. Use transformed cracked section and
locate the neutral axis,
 c
18c   + 40.6(c − 2.6) = 122.6(35.7 − c)
 2
c 2 + 18.13c = 498
c = 15.02 in.
1
I cr = (18)(15.02)3 + 122.6(35.7 − 15.02)2 + 40.6(15.02 − 2.60)2
3
= 79, 000 in.4
M cr = 402 ft-kips (same as left support)
M cr 402
M max 377
3
M cr 402  M 
= (dead load + live load) = 0.68;  cr  = 0.314
I e = 0.314(1773, 000) + 0.686(79, 000) = 109, 000 in.4
(d) Summary of values for Ie. The values of effective moment of inertia are
For DL For DL + LL
Left end Ie = 173,000 in.4 Ie = 173,000 in.4

Midspan Ie = 173,000 in.4 Ie = 94,000 in.4
Right end Ie = 173,000 in.4 Ie = 109,000 in.4
Having obtained the above values, usual practice is to use a single adjusted value of Ie
as discussed in Section 12.9.
(e) Compute a single adjusted value for Ie using the various procedures of Section 12.9.
1. Midspan value:
I e = 173, 000 in.4 (for DL only)

I e = 94, 000 in.4 (for DL + LL)
2. Simple average:
I e = 173, 000 in.4 (for DL only)

1  173, 000 + 109, 000  
I e =   + 94, 000 
2  2 
= 118, 000 in.4 (for DL + LL)
(Continued)
426
3. Weighted average:
I e = 0.70 I m + 0.15( I e1 + I e 2 )
I e = 173, 000 in.4 (for DL only)

I e = 0.70(94, 000) + 0.15(173, 000 + 109, 000)
= 108,100 in.4 (for DL + LL)
The task of determining the immediate deflection is actually one of analyzing a con-
tinuous beam with variable moment of inertia. Because the most “exact” computations
at best give deflections within probably ±20%, procedures more complex than those
illustrated here are not justified.
(f) Immediate dead load deflection. Ie for dead load only is 173,000 in.4 regardless of
whether midspan, simple average, or weighted average is used.
M a = 89.6 ft-kips M b = 377 ft-kips

w = 0.76 kip/ft P = 24.48 kips
Note that the span is taken as that measured between the centerlines of supports, and
the end moments are those computed for the same locations. Equally acceptable results
are obtained by using the clear span and the face-of-support moments. Referring to
Fig. 12.10.4, the total midspan deflection is
106
∆= (3.64 P + 1.333wL − 0.164 M a − 0.164 M b )
Ec I e
Ec = 33wc1.5 fc′ = 57, 000 3000 = 3.15 × 106 psi
1
(∆ i ) D =
3.15(173)
[3.64(24.48) + 1.333(0.76)(39) − 0.164(89.6 + 377.0)]
89.1 + 39.5 − 76.5 52.2
= = = 0.10 inn.
545 545
As previously mentioned, the immediate dead load deflection will usually cause no
difficulty, since most of it may be compensated for by the construction. However, it is
used as a basis for determining the long-term creep and shrinkage deflection, which is
discussed in the next section.
(g) Immediate live load deflection using the midspan value of Ie. The midspan value is
94,000 in.4 as computed in part (e).
( ∆ i ) D = 0.10 in. [see part (f)]

M a = 200 ft-kips M b = 591 ft-kips (dead load + live load)
w = 0.91 kip/ft P = 56.33 kips (dead load + live load)
1
( ∆ i )D + L =
3.15(94)
[3.64(56.33) + 1.333(0.91)(39) − 0.164(200 + 591))]
205.0 + 47.3 − 129.7 122.6
= = = 0.41 in.
296 296
(∆ i )L = 0.41 − 0.10 = 0.31 in.
(Continued)
427
(h) Immediate live load deflection using the simple average value of Ie.
( ∆ i ) D = 0.10 in. [see part(f)]

simple average value of I e = 118, 000 in.4 [see part (e )]
 94, 000 
( ∆ i ) D + L = 0.41  = 0.33 in.
 118, 000 
( ∆ i )L = ( ∆ i ) D + L − ( ∆ i ) D = 0.33 − 0.10 = 0.23 in.
(i) Immediate live load deflection using the weighted average value of Ie.
( ∆ i ) D = 0.10 in. [see part(f)]

weighted average value of I e = 108,100 in. [see part (e )]
4
 94, 000 
( ∆ i ) D + L = 0.41  = 0.36 in.
 108,100 
( ∆ i )L = 0.36 − 0.10 = 0.26 in.
(j) Conclusion. The immediate live load deflection is computed to be 0.31, 0.23,
and 0.26 in., respectively, depending on whether the midspan, simple average, or
weighted average value of Ie is used for dead and live load deflection. The use of
midspan Ie seems appropriate; for this example, one can conclude that the dead load
deflection is about 0.1 in. and the live load deflection is about 0.3 in., using one sig-
nificant figure.
If this is a floor that does not support frangible partitions, the limiting permissible
deflection is L /360, or
L 39(12)
= = 1.3 in. > 0.3 in (computed) OK
360 360
Because the beam of this example is continuous and has different reinforcement amounts
at the supports and at midspan, it is difficult to correlate the provided amounts of rein-
forcement with the calculated deflections. Furthermore, the beam behaves like a T-beam
in the positive moment region and as a rectangular cross section in the negative moment
regions. It is noted, however, that Ie is dominated largely by the midspan value. Since
the reinforcement ratio of about 0.2 ρtc or 0.14ρb for the positive moment region was
below the guideline value suggested in Chapter 3 for deflection control; thus excessive
deflection was not expected.
Recommended Values for Maximum Reinforcement Ratio ρ

for Deflection Control
For beams or one-way slabs, ACI Committee 435 [12.1, 12.2] has recommended the fol-
lowing values of the maximum reinforcement ratio ρ to be used in the positive moment
zone for deflection control:
1 For members of normal-weight concrete not supporting or not attached to nonstruc-

tural elements likely to be damaged by large deflections,
Rectangular cross section 0.35ρb
T-beams or box cross section 0.40ρb
428
2. For members of normal-weight concrete supporting or attached to nonstructural ele-

ments likely to be damaged by large deflections,
Rectangular cross section 0.25ρb
T-beams or box cross section 0.30ρb
3. For members of lightweight concrete, use 0.05ρb less than that indicated in items
1 and 2.
Additional guidance on deflection control may be found in Chapter 5 of the ACI Committee
435 report [12.1].
12.11 CREEP EFFECT ON DEFLECTIONS UNDER

SUSTAINED LOAD
The total long-term deflection consists of the instantaneous elastic deflection plus the contribu-
tions from creep and shrinkage. Creep is inelastic deformation with time under sustained loads
at unit stresses within the accepted elastic range (say, below 0.5 fc′ ), as shown in Fig. 12.11.1.
This inelastic deformation increases at a decreasing rate during the time of loading.
Factors that affect the magnitude of creep deformation [12.23, 12.24] are (1) the
constituents—such as the composition and fineness of the cement, the admixtures, and the
size, grading, and mineral content of the aggregates; (2) proportions, such as water content
and water-cement ratio; (3) curing temperature and humidity; (4) relative humidity during
storage; (5) size of the concrete member, particularly the thickness and the volume-to-
surface ratio; (6) age at loading; (7) duration of loading; and (8) magnitude of stress.
Since, as seen from Fig. 12.11.1, the result of creep is an increase in strain with constant
stress, one of the ways of accounting for it is by the use of a modified modulus of elastic-
ity Ect. An alternative, and generally preferred, procedure is to apply a multiplier Ct to the
short-term deflection ∆i.
To understand qualitatively the effect of creep on beam deformation, consider the strain
distributions before and after creep has occurred in the singly reinforced concrete cross
section of Fig. 12.11.2. Over time, the strains in the concrete will increase due to creep
and will cause the strain distribution to be as shown in Fig. 12.11.2(b). It is noted that the
strain at the tension steel is essentially unchanged because the concrete carries little if any
tension and because ordinary deformed steel reinforcement exhibits little creep. Since the
neutral axis increases, three observations may be made: (1) the concrete stress reduces at
the compression face (i.e., same compressive force acting and ccp exceeds ci); (2) the cur-
vature of the cross section can increase significantly; and (3) the increase in compressive
strain is much greater than the increase in curvature ϕ. This increase in curvature results in
increased deflection over time.
Figure 12.11.1 Typical concrete stress-strain curves for instantaneous and long-term loading.
429
12.11 CREEP EFFECT ON DEFLECTIONS 429
εi εi εcp
α
ci ci
ccp Creep effect
(shaded)
ϕi ϕi + ϕcp
As
Little change
εs εs due to creep
(a) Instantaneous strain (b) Strain after creep has occurred

due to initial loading
Figure 12.11.2 Creep effect on beam curvature.
For the purpose of computing deflections, it is frequently desirable to use a creep coef-
ficient Ct defined as the ratio of creep strain to elastic strain,
ε cp
Ct = (12.11.1)
εi
and the deflection resulting from creep is generally taken as
∆ cp = Ct ( ∆ i ) D (12.11.2)
where (∆i)D is the instantaneous, immediate deflection due to all sustained loads.

ACI Committee 209 has recommended [12.23] the hyperbolic-type equation of Branson
et al. [12.14, 12.23] for the creep coefficient, as follows:
 t 0.60 
Ct =  Cu (12.11.3)
 10 + t 0.60 
where
Ct = ratio of creep strain to elastic strain at any time t after a basic curing period
t = time in days after loading
Cu = ultimate creep coefficient; recommended average value is 2.35 for 40% humidity
The general relationship of Ct /Cu is shown in Fig. 1.11.2,

Equation (12.11.3) applies to the standard condition of 40% ambient relative humidity,
4 in. (100 mm) or less slump, average thickness of member of 6 in. (150 mm), and loading
age of 7 days for moist-cured concrete or 1 to 3 days for steam-cured concrete. For other
conditions, the standard condition value is to be multiplied by the following correction
factors (CF):
1. Age at loading.
For moist-cured concrete,
(CF )a = 1.25t a−0.118 (12.11.4a)
For steam-cured concrete,
(CF )a = 1.13t a−0.095 (12.11.4b)
In Eqs. (12.11.4), ta is the age at loading, in days, after the initial period of curing.
Several useful values of these equations appear in Table 12.11.1.
430
TABLE 12.11.1 CREEP CORRECTION FACTOR (CF)a FOR AGE AT LOADING,

EQS. (12.11.4)
Correction Factor, (CF)a
ta, Age in Days after Moist Cured for 7 Days Steam Cured for 1–3 Days
Initial Curing Period Initial Curing Period Initial Curing Period
10 0.95 0.90
20 0.87 0.85
30 0.83 0.82
60 0.77 0.76
90 0.74 0.74
2. Humidity. For H ≥ 40%,
(CF )h = 1.27 − 0.0067 H (12.11.5)
where H is the ambient relative humidity in percent. Values for this correction factor
appear in Table 12.11.2.
TABLE 12.11.2 CREEP CORRECTION FACTOR (CF)h

FOR HUMIDITY, EQ. (12.11.5)
Ambient Relative Correction Factor,

Humidity, H (%) (CF)h
40 or less 1.00
50 0.94
60 0.87
70 0.80
80 0.73
90 0.67
100 0.60
3. Average thickness of member. Where the average thickness of the member in inches
exceeds 6 in. (150 mm), a correction factor (reduction factor) may be applied.
However, for most design purposes such a correction may be neglected. For mem-
bers whose average thickness greatly exceeds 12 in. (300 mm), Meyers and Branson
[12.25] provide a chart that may be used to correct for the effect of average thickness.
4. Other correction factors. Additional correction factors are available [12.14, 12.25] to
account for slump greater than 4 in., cement content, percent of fine aggregate, and
air content. However, these tend either to be small or to offset one another and may
generally be neglected.
Compression Steel Effect on Creep

The presence of compression steel decreases the deformation due to creep (and shrinkage
as discussed in the next section). Evaluation of the effect of compression steel has been
reported by Washa and Fluck [12.13], Yu and Winter [12.8], and Hollington [12.26], and a
431
1 2 . 1 2 S H R I N K AG E E F F E C T O N D E F L E C T I O N S 431
multiplier factor has been given by Branson [12.27] and recommended by Committee 435
[12.1, 12.2], as follows:
0.85
kr =
1 + 50ρ′ (12.11.6)
where ρ′ is As′ / bd , the compression steel reinforcement ratio. Thus, the creep deflection ∆cp
would become
∆ cp = kr Ct ( ∆ i ) D (12.11.7)
instead of Eq. (12.11.2), when compression steel is present.

Paulson, Nilson, and Hover [12.28] have provided an evaluation of Eq. (12.11.6) as it
applies to high-strength concrete beams; these authors also recommended modifications.
12.12 SHRINKAGE EFFECT ON DEFLECTIONS UNDER

SUSTAINED LOAD
Shrinkage of concrete in beams may have a similar effect on the deflection as creep.
Shrinkage of an isolated plain concrete member would merely shorten it without causing
curvature. When steel reinforcement is added, however, bond between concrete and steel
restrains the shrinkage. Thus, a singly reinforced beam, having its shrinkage restrained at
the reinforced face and unrestrained at the unreinforced face, may have considerable cur-
vature. Generally, it is difficult to separate the effects of creep and shrinkage. Shrinkage is
more pronounced than creep during the first few months. Typically, 90% of the shrinkage
will have occurred by the end of 1 year, whereas 90% of the creep will not have occurred
until the end of 5 years. A number of investigators have studied shrinkage effects separately
from those of creep [12.14, 12.23–12.25, 12.29–12.31].
If the free shrinkage strain is known, shrinkage curvature ϕsh can be determined as a
function of shrinkage strain. Such curvature will be dependent on the relative amounts of
compression and tension steel just as creep is so affected. Finally, the shrinkage deflection
will involve the geometry of the support system. Shrinkage deflection ∆sh may be expressed
as [12.23]
∆ sh = α1ϕ sh L2 (12.12.1)
where α1 is a factor relating to the geometry of the support system and may be taken as the
following:
α1 = 0.50 for cantilever beams

= 0.125 for simply supported beamss
= 0.086 for beams continuous at one end only

= 0.063 for beams continuous at both ends
and L is the span length of the beam.
Shrinkage Strain, εsh
ACI Committee 209 has recommended [12.23] that the following expressions by Branson
et al. [12.14] be used for shrinkage strain εsh:
For any time t after age 7 days for moist-cured concrete,
t
ε sh = (ε sh )u (12.12.2a)
35 + t
432
For any time t after age 1 to 3 days for steam-cured concrete,
t
ε sh = (ε sh )u (12.12.2b)
55 + t
where
εsh = shrinkage strain at any time t after initial curing
t = time in days after initial curing
(εsh)u = ultimate shrinkage strain; average value suggested is 800 × 10–6 for 40% humidity
Equation (12.12.2a) is shown graphically in Fig. 1.11.4. For conditions other than 40%
ambient relative humidity, the standard condition value of Eqs. (12.12.2) is to be multiplied
by the following correction factor (CF):
(CF )h = 1.40 − 0.010 H , 40 ≤ H ≤ 80% (12.12.3)
(CF )h = 3.00 − 0.030 H , H ≥ 80% (12.12.4)
where H is the relative humidity in percent. Values for (CF)h appear in Table 12.12.1.
TABLE 12.12.1 SHRINKAGE CORRECTION FACTOR

(CF)h FOR HUMIDITY, EQS. (12.12.3) AND (12.12.4)
Ambient Relative Correction Factor,

Humidity, H (%) (CF)h
40 or less 1.00
50 0.90
60 0.80
70 0.70
80 0.60
90 0.30
100 0
Other correction factors may normally be neglected. Should such factors be desired,
corrections for average thickness other than 6 in., slump greater than 4 in., cement content,
percentage of fines, and air content are available [12.14, pp. 45–47].
Shrinkage Curvature, φsh
Several investigators [12.9, 12.23, 12.31] have developed expressions for curvature due to
warping that arises from nonuniform shrinkage. Reinforcement of different amounts in the
two faces of a beam is the principal cause of shrinkage warping.
Miller [12.31] established the following relationship for singly reinforced beams.
Referring to Fig. 12.12.1, by straight-line proportion,
ε sh − ε s ε sh  εs 
ϕ sh = =  1 − ε  (12.12.5)
d d sh
where εs is the compressive strain induced in the steel from shrinkage; εsh is the free shrink-
age strain at the unreinforced face. Miller empirically established values for εs /εsh as a func-
tion of the percentage of reinforcement ρ.
43
1 2 . 1 2 S H R I N K AG E E F F E C T O N D E F L E C T I O N S 433
εsh
εs
d
h
ϕsh
Figure 12.12.1 Shrinkage strain related to beam curvature for a singly reinforced beam. (After
Miller [12.31].)
Branson [12.9] modified Miller’s equation and empirically extended the results to give
equations including the effects of compression steel.
12
ε sh  ρ − ρ′ 
ϕ sh = 0.7 (ρ − ρ′ )1 3  for (ρ − ρ′ ) ≤ 3% (12.12.6)
h  ρ 
ε sh
ϕ sh = for (ρ − ρ′ ) > 3% (12.12.7)
h
Note that ρ and ρ′ are in percent, 100(As or As′ )/(bd). Equations (12.12.6) and (12.12.7) are
recommended [12.2, 12.14] as the most appropriate relationships.
Geometry of Warping from Shrinkage

The well-known moment area theorems for beam deflections may be used to establish the
factor α1 in Eq. (12.12.1) for the four typical cases. Since the quantity M/(EI) is in fact
the curvature due to bending moment, the φsh diagrams in Fig. 12.12.2 may be regarded as
the equivalent M/(EI) diagrams. In the derivations below, it is assumed that beam sections are
singly reinforced and that the same reinforcement is used for positive and negative bending.
For the cantilever beam [Fig. 12.12.2(a)],
∆ sh = BB ′ = moment of ϕ sh diagram between A and B about B
 L
= (ϕ sh L )   = 0.50ϕ sh L2
 2
For the simply supported beam [Fig. 12.12.2(b)],
θ A = area of ϕ sh diagram between A and C

 L
= ϕ sh  
 2
∆ sh = CC ′ = CC1 − C1C ′
 L
= θ A   − (moment of ϕ sh diagram between A and C about C)
 2
 L  L  L  L
= ϕ sh     − ϕ sh     = 0.125 ϕ sh L2
 2  2  2  4
43
Figure 12.12.2 Geometry of warping due to shrinkage.
For the beam fixed at one end only [Fig. 12.12.2(c)],
BB′ = moment of ϕ sh diagram between A and B about B

 L − x1   x12 
= −ϕ sh ( L − x1 )  x1 + + ϕ =0
2 
sh 
  2 
from which
2L
x1 =
2
For the tangent at C′ to be horizontal, the distances AD and DC must be equal; thus
 2
AD = DC = L − x1 =  1 − L
 2 
∆ sh = CC ′ = moment of ϕ sh diagram between A and C about C
= moment of a couple = ϕ sh ( L − x1 )2
2
 1   1
= ϕ sh L2  1 − 2  = ϕ sh L2  1 − 2 + 
 2   2
= 0.086ϕ sh L2
435
1 2 . 1 3 C R E E P A N D S H R I N K AG E D E F L E C T I O N 435
For the beam fixed at both ends [Fig. 12.12.2(d)], if the slope is to be horizontal at mid-
span for symmetry,
L
x1 =
4
Then
∆ sh = CC ′ = moment of ϕ sh diagram between A and C about C

2
 L
= moment of a couple =ϕ sh  
 4

= 0.063ϕ sh L2
Compression Steel Effect on Combined Shrinkage and Creep

Whenever shrinkage is to be included in combination with creep in a deflection computa-
tion, a multiplier similar to that of Eq. (12.11.6) may be applied to the short-term deflection
[12.23]. The following multiplier kr has been recommended by ACI Committee 435 [12.1,
12.2] as most appropriate to account for the compression steel effect on the long-term sus-
tained load deflection ∆cp+sh,
1
kr = (12.12.8)
1 + 50ρ′
where ρ′ = As′ / bd. Branson [12.27] and Shaikh [12.32] have discussed the use of this
expression to account for the compression steel effect. Equation (12.12.8) has been used in
the ACI Code since 1983 as a part of the combined multiplier λ∆ = kr ξ, which combines the
compression steel effect kr and the time-dependent effect ξ.
12.13 CREEP AND SHRINKAGE DEFLECTION—A CI

CODE METHOD
In the ACI Code method, the creep and shrinkage deflection due to sustained load is obtained
by multiplying the immediate, short- term deflection by a factor λ∆ (ACI-24.2.4.1.1).
Thus,
∆ cp + sh = kr ξ( ∆ i ) D = λ ∆ ( ∆ i ) D (12.13.1)
where
ξ
λ ∆ = kr ξ = (12.13.2)
1 + 50ρ′
and (∆i)D is the instantaneous deflection due to all sustained loads (usually dead load).
The value of ξ is permitted by ACI-24.2.4.1.3 to be taken in accordance with the dura-
tion of the sustained load as follows:
5 years or more 2.0

1 year 1.4
6 months 1.2
3 months 1.0
436
EXAMPLE 12.13.1
For the beam of Example 12.10.1 (see Fig. 12.10.3), determine the creep and shrinkage
deflection according to the ACI Code. Assume that only the dead load is sustained.
SOLUTION
First it is necessary to compute the immediate, short-term deflection due to all sustained
loads, in this case the dead load. From Example 12.10.1, part (a) of the solution,

( ∆ i ) D = 0.45 in.
Since no compression steel is used, Eq. (12.13.2) gives for 5 years or more load duration,
ξ
λ∆ = = 2.0
1 + 50ρ′
Then from Eq. (12.13.1),

∆ cp + sh = λ ∆ ( ∆ i ) D = 2.0(0.45) = 0.90 in.
If part of the live load were considered as sustained (as, e.g., in certain types of equip-
ment whose placement or installation is not expected to change for a period of 5 years or
more), it would be necessary to compute ∆i for the dead load plus the sustained live load.
Under the ACI Code, an additional effective moment of inertia Ie would be computed
using Mcr /Mmax, where Mmax is due to dead load plus sustained live load.
12.14 CREEP AND SHRINKAGE DEFLECTION—

ALTERNATIVE PROCEDURES
Separate Creep and Shrinkage Multiplier Procedure
A procedure for computing the deflections due to creep and shrinkage separately was rec-
ommended by ACI Committee 435 [12.1, 12.2], based on the work of Branson [12.9], as
modified by improvement in the prediction of creep and shrinkage [12.23].
Thus
∆ cp + sh = ∆ cp + ∆ sh (12.14.1)
where, using an equation given earlier
∆ sh = α1ϕ sh L2 [12.12.1]
and
∆ cp = kr Ct ( ∆ i ) D (12.14.2)
For evaluation of Eqs. (12.12.1) and (12.14.2),

α1 = constant [see (Eq. 12.12.1)]
φsh = shrinkage curvature [Eqs. (12.12.6) and (12.12.7)]
L = span length
kr = compression steel factor [Eq. (12.11.6)]
Ct = creep coefficient, using Eq. (12.11.3) with correction factors of Eqs. (12.11.4) and
(12.11.5), or per Ref. 12.23, “For average conditions, ultimate Ct = 1.6 may be used.”
(∆i)D = immediate deflection due to all sustained loads.
437
1 2 . 1 4 C R E E P A N D S H R I N K AG E D E F L E C T I O N 437
Combined Creep and Shrinkage Multiplier Procedure

This method is similar to the ACI Code method, except the time-dependent factor ξ can be
more accurately evaluated [12.14, 12.23, 12.26]. It may be stated as
∆ cp + sh = kr ξ( ∆ i ) D (12.14.3)
where
kr =1/ (1+ 50ρ′ )(same as ACI ) (12.14.4)

ξ = 
time-
dependent coefficient (creep plus shrinkage), which may be taken from
Table 12.14.1
TABLE 12.14.1 TIME-DEPENDENT COEFFICIENT ξ INCLUDING BOTH

CREEP AND SHRINKAGE EFFECTS, FOR BOTH NORMAL-WEIGHT AND
LIGHTWEIGHT CONCRETE MEMBERS OF COMMON TYPES, SIZES, AND
a
COMPOSITION (FROM BRANSON [12.14], P. 278).
Average Relative Humidity and Age (days) When Loaded
100% 70% 50%

Concrete Strength
fc′ at 28 Days ≤7d 14d ≥28d ≤7d 14d ≥ 28d ≤7d 14d ≥ 28d
2500–4000 psi 2.0 1.5 1.0 3.0 2.0 1.5 4.0 3.0 2.0
(17–28 MPa)
>4000 psi 1.5 1.0 0.7 2.5 1.8 1.2 3.5 2.5 1.5
(28 MPa)
aIt is suggested that the following percentages of the values in the table be used for sustained loads that are maintained for the
periods indicated:
25% for 1 month or less
50% for 3 months
75% for 1 year
100% for 5 years or more
The 50% values may normally be used for average relative humidities lower than 50%, which might be the case in heated
buildings, for example.
EXAMPLE 12.14.1
For the beam of Example 12.10.1 (see Fig. 12.10.3), determine the ultimate (i.e., 5-year
duration of load) creep and shrinkage deflection using the alternative methods discussed
in Section 12.14. Assume that only the dead load is sustained, the ambient relative
humidity is 70%, and the age at loading is 20 days after the initial moist-curing period.
SOLUTION
It is noted that ACI-24.2.4.1.1 permits computation of long-term deflection by a “more
comprehensive analysis,” which could include either of these alternative methods.
(a) Separate creep and shrinkage multiplier procedure.

∆ cp + sh = ∆ cp + ∆ sh
(Continued)
438
Using Eq. (12.14.2),

∆ cp = kr Ct ( ∆ i ) D
where,
0.85
kr = = 0.85 for ρ′ = 0
1 + 50ρ′
and from Eq. (12.11.3), for t = 5(365) days,
 t 0.60 
Ct =  Cu = 0.90Cu
 10 + t 0.60 
which for Cu = 2.35 as recommended for average conditions gives the basic value of Ct as
Ct = 2.12
Adjusting for 70% humidity, (CF)h = 0.80 from Eq. (12.11.5) or Table 12.11.2, and for
20-day age of loading after initial moist-curing period, (CF)a = 0.88 from Eq. (12.11.4a).
Thus the adjusted Ct is
Ct = 2.12(0.80)(0.88) = 1.49
From Example 12.10.1 using the ACI Code method,

( ∆ i ) D = 0.45 in.
Then
∆ cp = kr Ct ( ∆ i ) D = 0.85(1.49)0.45 = 0.57 in.
For shrinkage, from Eq. (12.12.1),

∆ sh = α1ϕ sh L2
where α1 = 0.125 for simply supported beams. For this beam, since ρ = 1.92% and
ρ′ = 0, Eq. (12.12.6) gives
ε 
ϕ sh = 0.7  sh  3 ρ
 h 
Using ε sh from Eq. (12.12.2a),
t
ε sh = (ε sh )u
35 + t
which for t = 5(365) days is, for average conditions,
ε sh ≈ ( ε sh )u = 800 × 10 −6
Adjusting for 70% humidity, (CF)h = 0.70, from Eq. (12.12.3) or Table 12.12.1, the
adjusted εsh is
ε sh = (800 × 10 −6 )0.70 = 560 × 10 −6
Then,
 560 × 10 −6  3
ϕ sh = 0.7   1.92 = 20.3 × 10 −6 rad/in.
 24 
∆ sh = α1ϕ sh L2 = 0.125(20.3 × 10 −6 )(480)2 = 0.58 in.
∆ cp + ∆ sh = 0.57 + 0.58 = 1.15 in.
(Continued)
439
12.15 ACI MINIMUM DEPTH OF FLEXURAL MEMBERS 439
(b) Combined creep and shrinkage multiplier procedure. Using Eq. (12.14.3),
∆ cp + sh = kr ξ( ∆ i ) D
1
kr = = 1.0
1 + 50ρ′
ξ = value from Table 12.14.1 ≈ 1.88
Note that age at loading is after initial curing period.
( ∆ i ) D = 0.45 in. (from Example 12.10.1)

∆ cp + sh = 1.0(1.8)0.45 = 0.81 in.
A comparison of computation methods for creep and shrinkage deflection may be

obtained from the following summary.
Method ∆sh+cp
1. ACI, using λ∆ = ξ/(1 + 50ρ′) = 2.0 0.90 in. (Example 12.13.1)

2. Separate creep and shrinkage per Eqs. 1.15 in. [Example 12.14.1, part (a)]
(12.12.1), (12.14.1) and (12.14.2).
3. Combined creep and shrinkage using Eq. 0.81 in. [Example 12.14.1, part (b)]
(12.14.3) with krξ = 1.8
12.15 ACI MINIMUM DEPTH OF FLEXURAL MEMBERS

The minimum depths specified in ACI-9.3.1.1 for beams and in ACI-7.3.1.1 for one-way
slabs must be satisfied unless computation of deflection indicates that a lesser thickness can
be used without adverse effects. The minimum depth (thickness) values apply to members
where large deflection is not likely to damage partitions, ceilings, or other frangible attach-
ments. When large deflection may cause such damage, calculated deflections must satisfy
the limits given in ACI-Table 24.2.2, regardless of whether the minimum thickness require-
ment is satisfied. The minimum depths prescribed by any such table are arbitrary and not
always conservative. For this reason, deflections should be computed even when the mini-
mum depth requirements of ACI-9.3.1.1 and ACI-7.3.1.1 are satisfied.
The logic behind the limitation on span-depth ratio given in the ACI Code as an attempt
to control deflection is explained in the following. The deflection at midspan of a simply
supported beam is
5wL4
∆= (12.15.1)
384 EI
where w is the service uniformly distributed load. The maximum bending moment is
wL2 fI
M= = (12.15.2)
8 y
where f is the service load stress and y is the distance from the neutral axis to the extreme
fiber, where f is computed.
40
Substituting Eq. (12.15.2) in Eq. (12.15.1) gives

5L2 f
∆= (12.15.3)
48 E y
Assuming a cracked section under service load conditions,
f fs f (12.15.4)
= = c
y n(d − c) c
where n is the modular ratio and d is the effective depth. Assuming that the tension steel is
stressed at, say, fs = 24,000 psi and that (d − c) ≈ 0.6h.
f 24, 000 40, 000 Ec E

≈ = = c (12.15.5)
y ( Es /Ec )(0.6h) hEs 725h
using Es = 29 × 106 psi. Substituting Eq. (12.15.5) in Eq. (12.15.3) and noting that E = Ec,
∆ 5  1 L
=  
L 48  725  h
1  L
min h =  L (12.15.6)
6960  ∆ 
Equation (12.15.6) represents an approximate relationship between depth, span, and

span-to-deflection ratio for a fully stressed section under short-term loading. If the mem-
ber is under a reduced stress, the depth may be decreased proportionally to give the same
short-term deflection. To account for the sustained load creep and shrinkage deflection, the
depth must be increased. Table 12.15.1 shows the evaluation of Eq. (12.15.6) to give the
minimum depth required for various deflection limitations under fully and partially stressed
conditions. The last two columns in Table 12.15.1 assume that total deflection including
creep and shrinkage effects is twice the immediate deflection.
Although the assumed stress under service loads in the tension steel is fs = 24,000 psi,
in practice, the tension steel stress under service loads is usually higher; in such a case, the
limits shown in Table 12.5.1 would be nonconservative.
Regardless of the assumed values, any selection of limiting values for minimum
depth from Table 12.15.1 can be considered only a crude attempt to control deflection.
Table 12.15.2 shows the minimum depth requirements for beams and one-way slabs given
in ACI Table 9.3.1.1 and 7.3.1.1, respectively. These values correspond to a compromise
between the relative conservative recommendations of ACI Committee 435 and the values
that practicing engineers believe to be suitable on the basis of experience.
When large deflections may cause cracking of partitions and other frangible attach-
ments, the total deflection that occurs after installation of such elements is limited to L /480
TABLE 12.15.1 MINIMUM DEPTH h FOR VARIOUS EQUIVALENT

IMMEDIATE DEFLECTIONS AND PERCENTAGE STRESSEDa
Percent ∆ = L /300 ∆ = L /360 ∆ = L /480 2∆ = L /300 2∆ = L /360

Stressed
100 L /23 L /19 L /15 L /12 L /9.7

67 L /35 L /29 L /22 L /17 L /15
60 L /39 L /32 L /24 L /19 L /16
50 L /46 L /39 L /29 L /23 L /19
a Assumed fs = 24,000 psi at 100% stressed.

41
1 2 . 1 6 S P A N -T O -D E P T H R A T I O 441
TABLE 12.15.2 MINIMUM DEPTH h FOR BEAMS AND ONE-WAY SLABS,

FOR MEMBERS NOT SUPPORTING OR ATTACHED TO PARTITIONS
OR OTHER CONSTRUCTION LIKELY TO BE DAMAGED BY LARGE
a
DEFLECTIONS
Type of Member Simple One End Both Ends Cantilever

Support Continuous Continuous
Beams fy = 60 ksi L /16 L /18.5 L /21 L /8

(ACI Table 9.3.1.1) fy = 40 ksi L /20 L /23 L /26 L /10
One-way slabs (solid) fy = 60 ksi L /20 L /24 L /28 L /10
(ACI Table 7.3.1.1) fy = 40 ksi L /25 L /30 L /35 L /12.5
a For structural lightweight concrete having weights wc from 90 to 115 pcf, multiply table values by 1.65 –0.005wc but not less
than 1.09. (60 ksi = 420 MPa; 40 ksi = 280 MPa, approximately.)
(ACI-24.2.2). This shows that the minimum depths of ACI Table 9.3.1.1 and 7.3.1.1 are
likely to be too small; hence those tables do not apply for such cases.
In general, minimum depth as a proportion of span is an inadequate criterion for con-
trolling deflection; computation of deflection should be made whenever deflection is of
concern.
12.16 SPAN-T O-D EPTH RATIO TO ACCOUNT FOR

CRACKING AND SUSTAINED LOAD EFFECTS
The general development, presented here, which is similar to that of Branson [12.33], illus-
trates the effects of the many variables on the span-to-depth ratio.
Grossman [12.34] has also provided a procedure for determining the minimum thick-
ness that would approximately satisfy any deflection limitation given by ACI Table 24.2.2.
The procedure proposed by Grossman uses an approximated effective moment of inertia
and eliminates the need to compute the moment of inertia of the cracked section, Icr.
The development presented in this section, which perhaps is less practical than the
method suggested by Grossman, is intended to illustrate how the variables interrelate and
to show why it is impossible to have a simple table of minimum thicknesses. Along with
Branson’s discussion [12.35] and Grossman’s closure [12.34], the work by Grossman
[12.34] provides further insight and a useful discourse on the subject of deflection control.
Rangan [12.36] has also presented a minimum thickness approach similar to that presented
in the following.
The short-term deflection ∆i may be expressed, according to Eq. (12.10.1), as
 M L2 
∆ i = β a  max  (12.16.1)
 Ec I e 
where
M max = maximum moment at the stage for which deflection is caalculated
I e = Eq.(12.9.1)[which is ACI code Formula (24.2.3.5a)]
= ( M cr /M max )3 I g + 1 − ( M cr /M max )3  I cr ≤ I g
M cr = cracking moment = fr I g /yt
fr = modulus of rupture = 0.65 wc fc′ (ACI uses 7.5λ fc′)
y t = distance from neutral axis to extreme fiber in tenssion
42
Multiplying Eq. (12.16.1) by fr Ig /(yt Mcr), which is equal to unity, gives
 M L2  fr I g
∆ i = β a  max  (12.16.2)
 Ec I e  yt M cr
Solving Eq. (12.16.2) for L /yt gives
L ∆ i  Ec   M cr  I e (12.16.3)
=
yt L  fr β a   M max  I g
Since both Ec and fr are proportional to fc′ , let
Ec 33w1c .5 fc′
βw = = = 50.8wc (12.16.4)
fr 0.65wc0.5 fc′
For normal-weight concrete,
β w = 50.8wc = 50.8(145) = 7370
Conversion from normal-weight concrete to lightweight concrete may be made by multi-

plying L /yt by the ratio of the unit weight of lightweight concrete to 145 pcf. Letting γ = Ie /Ig
and βw = Ec /fr in Eq. (12.16.3),
L ∆ i  β w   M cr 
= γ (12.16.5)
yt L  β a   M max 
where
I e  M cr    M cr   I cr
3 3
γ= = + 1 −  ≤1 (12.16.6)
I g  M max    M max   I g
 
The ratio of the moment of inertia of the cracked section to that of the gross section may be
computed for various shapes of beams using, for example, the transformed section method (see
Section 12.5). As an example, for a singly reinforced rectangular beam, assuming d = 0.9h,
I cr bc 3/ 3 + nAs (d − c)2
=
Ig bh3 /12
 ( c / d )3  c 
2
= 8.75  + nρ  1 −   (12.16.7)
 3  d  
where
c /d = (ρn)2 + 2ρn − ρn (12.16.8)
Charts are given by Lutz [12.16] to obtain Ig and Icr for T-sections.
For dead load deflection, Mmax = MD, which makes γ = γD; Eq. (12.16.5) then becomes
(∆ i )D βa  M D   L   1  (12.16.9)
=
L β w  M cr   yt   γ D 
For dead load plus live load,
(∆ i )D + L βa  M D + L   L   1  (12.16.10)
=
L β w  M cr   yt   γ D + L 
43
As discussed in Section 12.10, since live load cannot act in the absence of dead load, the
live load deflection must be obtained indirectly,
( ∆ i )L ( ∆ i )D + L ( ∆ i )D
= − (12.16.11)
L L L
which, using Eqs. (12.16.9) and (12.16.10), and letting CL = ML /MD, gives
( ∆ i )L β a  L   M D + L   1 1 
=  −  (12.16.12)
L β w  yt   M cr   γ D + L γ D (1 + CL ) 
When excessive deflection may cause damage to partitions and other nonstructural con-
struction, it is the sum of deflections due to live load plus creep and shrinkage that is
of concern. The instantaneous (short-term) dead load deflection will have occurred when
forms are removed and before any breakable attachments are put in place. Thus using
Eq. (12.13.1), and the ACI time-dependent multiplier λ∆,
∆ cp + sh = λ ∆ ( ∆ i )D

Finally, the deflection to be controlled to minimize possible damage is
( ∆ i )L ∆ cp + sh β a  L   M D + L   1 (λ ∆ − 1) 
+ =      + 
L L β w  yt   M cr   γ D + L γ D (1 + CL ) 
β a  L   M D + L  1 + CL + (λ ∆ − 1)( γ D + L / γ D )  1
=   (12.16.13)
β w  yt   M cr   1++ CL  γ D+L
Solving for L /yt gives
 L  ∆  β   M   CL + 1 
 y  =    w   cr  
   γ D+L (12.16.14)
t limit for L  β a   M D + L   CL + 1 + (λ ∆ − 1)( γ D + L / γ D ) 
L +cp +sh
For instantaneous dead load plus live load, Eq. (12.16.10) gives
 L  ∆  β   M 
 y  =    w   cr  γ D + L (12.16.15)
t limit for  L   βa   M D + L 
D+L
As shown below, if the span-to-depth ratio limit is available for short-term deflection under
dead load plus live load, such as from Table 12.15.1, the effect of creep and shrinkage can
be obtained by the use of a multiplier. Comparing Eqs. (12.16.14) and (12.16.15), in which
∆ /L may be taken as a stated limit,
 L  L  C L +1 
 y  =    (12.16.16)
 yt  D + L C
 L + 1 + ( λ ∆ − 1 )( γ D+L / γ )
D 
t L + cp + sh
For the situation in which the Ie under dead load only is approximately the same as Ie under
dead plus live load, γD+L ≈ γD, Eq. (12.16.16) becomes
 L  L  CL + 1 
 y  =  (12.16.17)
 yt  D + L  CL + λ ∆ 
t L + cp + sh
Charts are available [12.33] for the L /h ratio for short-term effects of dead load plus live
load, Eq. (12.16.10). The chart for singly reinforced rectangular beams ( yt = 0.5h) is given
in Fig. 12.16.1.
4
EXAMPLE 12.16.1
Determine the depth of beam required for the loading conditions of Example 12.10.1
(Figure 12.10.3) if the sum of the immediate live load plus creep and shrinkage deflec-
tion must not exceed L /480. Assume only the dead load is sustained with a sustained
load factor λ∆ = 2 as given by ACI-24.2.4.1.
SOLUTION
(a) Use ACI Table 9.3.1.1, or Table 12.15.2.
L
= 16
h
This considers average conditions and includes some effect of sustained load deflection.
480
min h = = 30 in.
16
130
120 Boundary conditions

Simple – use curve values directly
110 Cantilever – multiply curve values by
0.417
One end continuous (hinged–fixed)
100 – multiply curve values
by 2.41
h Both ends continuous (fixed–fixed)
90 – multiply curve values
by 3.33
Rectangular
80 beams
singly
reinforced
70
L
h
60
50
Curve No. fc’ (psi) ρ (%)
40 1 3000 3.0
2 4000 3.0
3 5000 3.0 1
30 4 3000 2.0 2
5 4000 2.0 3
6 5000 2.0 4
7 3000 1.0 65
20
8 4000 1.0 78
9 5000 1.0 9
10 3000 0.5 10
10 11 4000 0.5 11
12
12 5000 0.5
0
0 1 2 3
Mmax
Mcr
Figure 12.16.1 Curves of L /h versus Mmax /Mcr for the conditions Δ = L /360, normal-weight concrete and
uniformly distributed short-term loading, and for different boundary conditions, steel percentages, and concrete
strengths. (From Branson [12.33].)
(Continued)
45
(b) Use more accurate procedure with Fig. 12.16.1.

ρ = 0.0192
From Example 12.10.1, part (a)

fr I g
M cr = = 68 ft-kips
yt
and
M max 250 + 90
= = 5 > 3; use 3 for this example.
M cr 68
Compute CL
M 250
CL = L = = 2.78
M D 90
From Fig. 12.16.1 for Mmax /Mcr = 3, fc′ = 4000 psi, and ρ = 0.0192, find
 L ∆ 1
  = 23, for =
h D+L L 360
 L  360  ∆ 1
  = 23  = 17.3, for =
h D+L  480  L 480
 L  C +1   2.78 + 1 
  = 17.3  L  = 17.3   = 13.7
h L + cp + sh  CL + λ ∆   2.78 + 2 
480
min h = = 35 in.
13.7
which is more severe than the minimum depth required by ACI Table 9.3.1.1.
(c) Required depth and adequacy of beam in Example 12.10.1. For the given beam with
h = 24 in.,
( ∆ i )L = 1.43 in. (Example 12.10.1)

∆ cp + sh = 0.90 in. (Example 12.13.1)
where
L 480
allowable ∆ = = = 1 in.
480 480
calculated ∆ = 1.43 + 0.90 = 2.33 in.
which is clearly inadequate, as was expected based on the results obtained in parts (a)
and (b).
Increasing the beam depth will increase the moment of inertia, but it will also add
weight to the beam, which is sustained and will offset some of the reduction in deflec-
tion that would result due to an increase in depth. Because an important part of the
total deflection is due to the sustained load, a more effective strategy is to add some
compression steel to reduce the long-term deflection in addition to increasing beam
depth.
46
12.17 ACI CODE DEFLECTION PROVISIONS—B EAM

EXAMPLES
The ACI Code provisions (ACI-24.2) regarding deflection computations under service
loads may be summarized as follows:
1. For members not supporting or not attached to elements likely to be damaged by large
deflections that satisfy the minimum member depth requirements (ACI Table 9.3.1.1
for beams or ACI Table 7.3.1.1 for one-way slabs) are “… considered to satisfy the
requirements of the Code…” (ACI-R24.2). For members that do not meet the mini
mum depth requirements, both the immediate and the time-dependent deflections
must be computed and must satisfy the limits of ACI Table 24.2.2.
2. For members supporting or attached to elements likely to be damaged by large deflec-
tions, deflections must be computed and must satisfy the limits of ACI Table 24.2.2
regardless of whether they meet the minimum depth requirements.
As noted in Section 12.15, the use of a minimum depth as a proportion of span is an inade-
quate criterion for controlling deflections. The use of a minimum depth (ACI Table 9.3.1.1
for beams and 7.3.1.1 for one-way slabs) does not ensure that excessive deflections will not
occur. Therefore, computation of deflection should always be made whenever deflection
is of concern (as is often the case in practice), even for members under Category 1 above.
EXAMPLE 12.17.1
Investigate the deflection for the beam of Fig. 12.17.1 used on a simple span of 25 ft. The
maximum bending moments under service load are 158 ft-kips dead load and 105 ft-kips
live load. Assume that all loading is uniformly distributed and that none of the live load is
sustained. The beam supports partitions and other construction likely to be damaged by
large deflections. Use fc′ = 4000 psi (normal weight), fy = 60,000 psi, and the ACI Code.
b = 14”
c = 8.31”
21.5”
h = 25”
13.19”
2 – #8
3 – #9
Figure 12.17.1 Beam cross section for Example 12.17.1.
SOLUTION
(a) Because the beam supports elements likely to be damaged by large deflections, the
deflection limits of ACI Table 24.2.2 must be satisfied even if the minimum depth
requirement of ACI Table 9.3.1.1 is satisfied. Nonetheless, it is instructive to com-
pute the minimum required depth from ACI Table 9.3.1.1 (or from Table 12.15.2
above for fy = 60 ksi) and compare it with the provided beam depth. Thus,
L 25(12)
min h = = = 19 in.< 25 in. provided
16 16
(Continued)
47
1 2 . 1 7 AC I C O D E D E F L E C T I O N P R OV I S I O N S — B E A M E X A M P L E S 447
Since the minimum depth requirement of ACI Table 9.3.1.1 is satisfied, one might expect
that deflections for a member not supporting partitions likely to be damaged would be
acceptable.
(b) Examine reinforcement ratio ρ
As 4.58
ρ= = = 0.015
bd 14(21.5)
This exceeds the 0.25ρb = 0.007 limit recommended by ACI Committee 435 [12.1] for
members supporting nonstructural elements likely to be damaged by large deflections
(see Section 12.10). From this check, one might expect deflection to be a problem.
(c) Determine the moment of inertia Ig for the gross uncracked section without steel and
Icr for the cracked transformed section.
For the gross uncracked section,
1
Ig = (14)(25)3 = 18, 200 in.4
12
For the cracked section, locate the neutral axis under service loads. Using a modular
ratio n = 8,
14c 2
= 4.58(8)(21.5 − c)
2
c = 8.31 in.
1
I cr = (14)(8.31)3 + 4.58(8)(13.19)2 = 9050 in.4
3
(d) Determine the effective moment of inertia Ie (ACI Formula 24.2.3.5a). The cracking
moment is
fr I g 7.5 4000 (18, 200)
M cr = = = 57.6 ft-kips
yt 12.5(12, 000)
For dead load deflection,

3
M cr 57.6  M cr 
= = 0.365;  M  = 0.05
M max 158 max
 M 
3
  M 3
I e =  cr  I g + 1 −  cr   I cr
 M max    M max  
= 0.05(18, 200) + 0.95(9050) = 9500 in.4

For dead load plus live load deflection,
3
M cr 57.6  M cr 
= = 0.22  M  = 0.01
M max 263 max
I e ≈ I cr = 9050 in.4
(e) Compute immediate deflections.

Ec = 57, 000 fc′ = 57, 000 4000 = 3.6 × 106 psi
(Continued)
48
For dead load,
5wL4 5 ML2 5(158)(12)(300)2

(∆ i )D = = = = 0.52 in.
384 EI 48 EI 48(3.6)(103 )(9500)
For dead load plus live load,
5(263)(12)(300)2
(∆ i )D + L = = 0.91 in.
48(3.6)(103 )(9050)
Then the immediate live load deflection is

( ∆ i )L = ( ∆ i ) D + L − ( ∆ i ) D = 0.91 − 0.52 = 0.39 in.
(f) Compute creep and shrinkage deflection. From ACI-24.2.4.1, the multiplier is
2.0
λ ∆ = kr ξ = = 2.0
1 + 50ρ′
for sustained load at 5 years or more.

∆ cp + sh = λ ∆ ( ∆ i ) D = 2.0(0.52) = 1.04 in.
(g) Check limitation of ACI Table 24.2.2. For roof or floor construction supporting or
attached to nonstructural elements likely to be damaged by large deflection,
L
( ∆ i )L + ∆ cp + sh ≤
480
This limit is for the deflection that is estimated to occur after the nonstructural elements
have been put in place. Whatever portion, if any, of the live load or creep and shrinkage
deflection that has occurred prior to the placement of nonstructural elements may be
excluded from the L /480 limitation. Further, if adequate measures are taken to prevent
damage to supported or attached elements, the L /480 limit may be exceeded (footnote,
ACI Table 24.2.2).
For this example,
 L 300 
[0.39 + 1.04 = 1.43 or 1.4 in.] >  = = 0.63 in.
 480 480 
which is not acceptable.
EXAMPLE 12.17.2
Investigate the deflection for the one-way continuous slab shown in Fig. 12.17.2. The slab
is 5 in. thick. Assume that 60% of the 100 psf live load is sustained. Use fc′ = 3000 psi
(normal weight) and fy = 60,000 psi. Assume the slab supports non-structural elements
not likely to be damaged by large deflections.
SOLUTION
Equation (12.7.4) may be used to compute maximum deflection:
5L2  1 
∆m =  M s − 10 ( M a + M b ) [12.7.4]
48 EI  
(Continued)
49
1 2 . 1 7 AC I C O D E D E F L E C T I O N P R OV I S I O N S — B E A M E X A M P L E S 449
Center-to-center of supports will be used as the span L. However, when coefficients are
used to compute moments as discussed in Chapter 7 using clear span Ln, the Ln should be
used in computing deflections [12.2]. This mathematical model incorporates a conservative
assumption of zero moment at the exterior support. Though Eq. (12.7.4) provides midspan
deflection, maximum deflection occurs between the location of maximum moment and
midspan. It is reasonable and practical to use maximum moment for Ms in that equation.

Ec = 57, 000 fc′ = 57, 000 3000 = 3.12 × 106 psi
(a) Determine Ig and Icr using a 1-ft width of section.
1
(12)(5.0)3 = 125 in.4
Ig =
12
For the positive moment region, d = 3.94 in. Using n = 9,
12c 2
= 9(0.37)(3.94 − c)
2
c = 1.23 in.
1
I cr = (12)(1.23)3 + 9(0.37)(3.94 − 1.23)2 = 31.9 in.4
3
(b) Determine effective moment of inertia Ie (ACI-24.2.3.5). Note that for prismatic
one-way slabs, ACI-24.2.3.7 permits Ie to be computed at midspan for simple and
continuous spans.
fr = 7.5 fc′ = 7.5 3000 = 411 psi
fr I g 0.411(125)
M cr = = = 1.71 ft-kips
yt 2.50(12)
 M 
3
  M 3
I e =  cr  I g + 1 −  cr   I cr
 M max    M max  
Figure 12.17.2 End-span details for continuous slab of Example 12.17.2. (Continued)

450
In the positive moment region, for dead load

M cr M cr 1.71
= = > 1; I e = I g = 125 in.4
M D M max 0.83
and for the dead load plus live load,

3
M cr M cr 1.71  M cr 
= = = 0.8;  M  = 0.51
M D + L M max 2.14 max
I e = 0.51(125) + 0.49(31.9) = 79.4 in.4
(c) Immediate live load deflection.
( ∆ i )L = ( ∆ i )D + L − ( ∆ i )D
5(12)2 144  1.73 

(∆ i )D + L =  2.14 −  (12) = 0.21 in.
48(3.12)(103 )79.4  10 

5(12)2 144  0.67 
(∆ i )D =  0.83 −  (12) = 0.05 in.
3
48(3.12)(10 )125 10 
( ∆ i )L = 0.21 − 0.05 = 0.16 in.

Check short-term deflection limit from ACI Table 24.2.2 for immediate live load
deflection,
L 12(12)
allowable ∆ = = = 0.40 in. > 0.16 in. OK
360 360
(d) Consider the effect of sustained live load that contributes to the creep and shrinkage
deflection. The immediate deflection due to all sustained loads is required as the
base value on which to apply the time-dependent multiplier.
For the positive moment region, using sustained load moment of 0.83 + 0.6(1.31) =
1.62 ft-kips/ft,
M cr 1.71
= > 1.0
M max 1.62
This result suggests that one could use Ie = Ig for computing the immediate deflection
due to all sustained loads. However, because the presence at any point in time of the
design live load will cause cracking in the positive moment region, it would be noncon-
servative to use Ie = Ig, as upon removal of the transient live load, immediate deflections
would be those of a cracked slab. The actual stiffness of the cracked slab is unknown,
but it seems reasonable to assume that it would be similar to that corresponding to Ie
computed under dead load plus live load, i.e., Ie = 79.4 in4.
(e) Creep and shrinkage deflection. The immediate deflection due to sustained loads
may be computed as,
5(12)2 144  1.31
( ∆ i ) D + sustL =  1.62 −  (12) = 0.16 in.
3
48(3.12)(10 )(79.4) 10 
where the sustained moment at the support is 0.67 + 0.6(1.06) = 1.31 ft-kips/ft.
Considering 5 years or more load duration, and with no compression steel, λ∆= 2.0

∆ cp + sh = λ ∆ ( ∆ i ) D + sustL = 2.0(0.16) = 0.32 in.
(Continued)
451
1 2 . 1 8 C R A C K C O N T R O L F O R B E A M S A N D O N E - WAY S L A B S 451
(f) Check long-term deflection limit of L /240 in ACI Table 24.2.2,

L
( ∆ i )*L + ∆ cp + sh ≤
240
where ( ∆ i )*L represents the immediate deflection due to any additional (not sustained)
live load (see Section 12.6). At this stage, the slab is assumed to have been previously
cracked by the presence of live load and thus ( ∆ i )*L will be computed using Ie = 79.4 in4
calculated in part (b). Thus,
 0.4 L 
( ∆ i )*L = ( ∆ i ) D + L 
 D + L 
 0.4(0.10) 
= 0.21 
 0.163 

= 0.05 inn.
and
( ∆ i )*L + ∆ cp + sh = 0.05 + 0.32 = 0.37 in.
L 12(12)
allowable ∆ = = = 0.60 in. > 0.37 in. OK
240 240
The decision regarding whether part of the live load is sustained should be based on a
consideration of the actual loading and its duration. For instance, a floor system supporting
library stacks as the live load might well be considered to have a significant part of the live
load treated as being sustained. In most situations it is acceptable to consider that only the
dead load is sustained and thereby affects the magnitude of creep and shrinkage deflection.
Although the deflection calculations have been illustrated throughout this chapter without
permitting any shortcuts on steps in the formal procedure, deflection computations should
be made with practicality in mind. The effective moment of inertia Ie theoretically should
be computed at each total load level for which deflection is of concern—usually dead load,
dead load plus live load, and dead load plus sustained live load. To obtain an average, as is
encouraged by the ACI Code, Ie should be computed at each end and at the midspan region.
However, when the true accuracy obtainable from a deflection computation is recog-
nized, the designer should use the Ie equation only when the result will be significantly
different from using either Icr or Ig. For most situations, furthermore, only the midspan Ie
need be computed; an average does not measurably improve accuracy. The authors believe
these practical suggestions satisfy the spirit of the ACI Code.
For additional practical deflection examples based on the ACI Code and ACI Committee
435 recommendations, including composite beams and two-way systems, see the chapter
by Branson in Handbook of Concrete Engineering [12.2], the PCA Notes [12.37], and the
CRSI Design Handbook [2.21]. Also, see Example 21.3.1 in Chapter 21 for a complete
design example, including deflections, of a composite beam.
12.18 CRACK CONTROL FOR BEAMS

AND ONE-W AY SLABS
Cracking in concrete is generally the result of the following actions [12.38–12.40]: (1) vol-
umetric change, including that due to drying shrinkage, creep under sustained load, ther-
mal stresses, and chemical incompatibility of concrete components; (2) internal or external
direct stress due to continuity, reversible load, long-term deflection, camber in prestressed
concrete, or differential movement in structures; and (3) flexural stress due to bending.
452
Visible cracking is generally initiated by either internal microcracking (volumetric

change would usually include this type) or flexural microcracks. Flexural microcracks are
surface cracks that are not visible except by careful close investigation and are generally
initiated by flexural stress. Once flexural microcracks have formed, a slight increase in load
causes these cracks to open up suddenly to measurable widths. Under service loads, wide
cracks could form, which might be detrimental to steel reinforcement under a corrosive
environment. Factors such as humidity, salt air, and alternate wetting and drying or freezing
and thawing may accelerate corrosion of the steel reinforcement and contribute to concrete
deterioration in the vicinity of large-width cracks. Increased cover provides thicker protec-
tion but may result in wider cracks at the beam face, influencing corrosion [12.41, 12.42].
Today, epoxy-coated reinforcing bars are widely used to prevent and ameliorate corrosion
of the steel reinforcement in concrete structures. However, even when corrosion is not of
concern, wide cracks may be unsightly and contribute to doubt about structural safety.
Although cracking cannot be eliminated, it is generally more desirable to have many fine
hairline cracks than a few wide cracks. Thus, crack control is a matter of controlling the
distribution and size of cracks rather than eliminating them. Excellent sources of informa-
tion on control of cracking resulting from flexure and other causes are two ACI Committee
224 reports [12.38, 12.39].
To control cracking, it is better to use several smaller bars at moderate spacing than larger
bars at large spacing. The objective is, therefore, one of distributing the reinforcement in
the tension zone; hence ACI-24.3, entitled “Distribution of flexural reinforcement …,”
contains the crack control provisions for beams and one-way slabs. Control of cracking
is particularly important when reinforcement with a yield stress in excess of 40,000 psi is
used, as in most cases, or when tension reinforcement ratios exceed about 0.375ρb.
A good bar arrangement in the cross section will usually lead to adequate crack control.
Entirely satisfactory structures have been built, particularly in Europe, using design yield
stresses exceeding 80,000 psi, which is the current specified limit in ACI Table 2 0.2.2.4a
for the main reinforcement in non-seismic-force-resisting systems.
Many studies [12.43–12.52] have verified the generally accepted belief that crack width
is proportional to steel stress. Two other significant variables have been found to be the
thickness of concrete cover and the area of concrete surrounding each individual reinforc-
ing bar in the zone of maximum tension.
Since even in a controlled laboratory environment crack widths are highly variable and
difficult to predict accurately, they may be expected to vary widely within a given struc-
tural member. Rather than specifying limits on the maximum allowable crack width, the
ACI Code requires a proper arrangement and spacing of the reinforcing bars that will usu-
ally lead to adequate crack control. Accordingly, the maximum permitted center-to-center
spacing, s, of the nonprestressed reinforcement closest to the tension face is given in ACI
Table 24.3.2 as:
 40, 000  1
s = 15  − 2.5cc (12.18.1)
 fs 
but not greater than 12(40,000/fs), where fs in psi is the stress in the reinforcement closest to
the tension face at service loads computed based on the unfactored moment (ACI-24.3.2.1).
Recognizing that additional elastic analysis under service-level loads would be required to
determine fs for use in Eq. (12.18.1), ACI-24.3.2.1 permits taking fs as 2 3 of the specified
yield stress fy. For Grade 60 reinforcement, this approach is generally conservative because
the actual service load stress in the bar is usually less than ( 2 3 ) fy.
The clear cover, cc, to be used in Eq. (12.18.1) is measured from the nearest surface in
tension to the surface of the flexural tension reinforcement.
1 For SI, ACI 318-14M,

 280   280 
s = 380  − 2.5cc but not greater than 300  (12.18.1)
 fs   fs 
where s and cc are in mm and fs is in MPa.
453
1 2 . 1 8 C R A C K C O N T R O L F O R B E A M S A N D O N E - WAY S L A B S 453
For two-way slabs, the foregoing spacing limit formula does not apply. Other recom-
mendations are given in the ACI Committee 224 Report [12.38]. Chapter 16 of this text also
contains a brief treatment on the subject.
When structures are subject to very aggressive exposure or designed to be watertight,
the provisions of ACI-24.3.2 are not sufficient. Special investigations and precautions are
required for such structures.
For guidance in the design of sanitary structures, the reader is referred to the work of
ACI Committee 350 [12.51].
EXAMPLE 12.18.1
Check the crack control provisions of the ACI Code for the beam cross section of
Example 12.4.1 (Fig. 12.4.1). The selected beam has b = 18 in., h = 34 in., 2–#9 bars
and 4–#10 bars in one layer. Use #3 stirrups, 1.5-in. clear cover at bottom.
SOLUTION
Compute cc,

cc =1.5 in. ( clear cover to bottom ) + 0.375 (stirrup ) =1.875 in.
From Example 12.4.1, part (b), the calculated stress in the tension steel under service
loads is fs = 37,000 psi, which is somewhat less than ( 2 3 ) fy (i.e., fs = 40,000 psi, as per-
mitted by ACI-24.3.2.1.).
The maximum, center- to-
center spacing of the reinforcement, according to
Eq. (12.18.1), is
 40, 000   40, 000 

s = 15  − 2.5cc = 15  − 2.5(1.875) = 11.5 in.
 fs    37, 000 
but not greater than 12(40,000/fs) = 12(40,000/37,000) = 13 in.

The section of Fig. 12.4.1 clearly meets the 11.5-in. maximum spacing.
EXAMPLE 12.18.2
Investigate crack control at the maximum positive moment region for the beam of
Example 6.21.2.
SOLUTION
As shown in Fig. 6.21.2, the beam is reinforced with one #7 and 2–#9 bars in the positive
moment region. Using
2 2
fs = f y = 60, 000 = 40, 000 psi
3 3
as permitted by ACI-24.3.2.1, and assuming a 1.5-in. clear cover,

cc = 1.5(cover) + 0.375(stirrup) = 1.875 in.
then, according to Eq. (12.18.1)
 40, 000 
s = 15  − 2.5(1.875) = 10.3 in
 40, 000 
which is not greater than 12(40,000/40,000) or 12 in.

(Continued)
45
Using #3 stirrups and 1.5-in. clear side cover, the center-to-center spacing between
the bars is
14(beam width) − 2[1.5(cover ) + 0.375(stirrup)] 1.128(db# 9 )

sprovided = − = 4.6 in. OK
2 2
which is less than the limit of 10.3 in.
The above calculations are somewhat conservative. First, the calculation for sprovided
does not account for the #3 bend radius, which would result in an even smaller bar spac-
ing. Also, the actual service load stress in the bar is usually less than the 2 3 fy. A lesser
value could be computed if needed to satisfy the crack control limitation.
EXAMPLE 12.18.3
Investigate the crack control requirements for the floor girders 2G1-2G2-2G2-2G1 of
Example 9.9.1.
SOLUTION
Using Eq. (12.18.1), the maximum allowed spacing of the reinforcement closest to the
tension face is
 40, 000 
s = 15  − 2.5cc (12.18.1)
 fs 
but not greater than 12(40,000/fs).

With reference to Fig. 9.9.1, the most critical location will occur for the reinforcement
that is farthest from the concrete surface. In this case, this will occur for the negative
moment reinforcement, where cc = 3.25 in. Using
2 2
fs = f y = 60, 000 = 40, 000 psi
3 3
the maximum allowed spacing is
 40, 000 
s = 15  − 2.5(3.25) = 6.9 in.
 40, 000 
which is not greater than 12(40,000/40,000): that is, 12 in.

The provided center-to-center spacing between the bars at section A-A (largest spacing
between bars) is
18 − 2(1.5 + 0.375) 1.128(db# 9 )

sprovided = − = 6.6 in. OK
2 2
As noted before, this check is somewhat conservative because the bend radius of stirrups
has not been included in the computation of sprovided and because the actual service load
stress in the bar is usually less than ( 2 3 ) fy.
For flanges of T-beams in tension, the ACI Code requires that part of the flexural ten-
sion reinforcement be distributed over the effective flange width bE, but not wider than
Ln /10. If the effective flange width bE exceeds Ln /10, additional longitudinal reinforce-
ment must be provided in the outer portions of the flange (ACI-24.3.4). This requirement
(Continued)
45
12.19 SIDE FACE CRACK CONTROL FOR LARGE BEAMS 455
is intended to control the crack widths in the slab near the web and in the outer regions
of the flange when the negative moment beam reinforcement is concentrated within or
near the web. In this example, the flanges of girders are provided with the slab reinforce-
ment which, if spaced in accordance to Eq. (12.18.1), should provide adequate control
of cracking (see Fig. 9.9.3). In heavily reinforced beams, such as sections C-C, D-D,
and F-F of Fig. 9.9.1, the top reinforcement should be provided in a single layer across
the flange width.
12.19 SIDE FACE CRACK CONTROL FOR LARGE BEAMS

On deep beams, the maximum crack width may occur along the side faces between the neu-
tral axis and the main tension reinforcement [12.53–12.56]. In such cases, the main rein-
forcement provides a restraining effect on the opening of cracks near the extreme tension
face. Thus, the maximum width occurs somewhere between the neutral axis, where there
should be no flexural cracking, and the main tension reinforcement, where the crack open-
ing is restrained. Unacceptably wide side face cracks, sometimes three times as wide as
the crack width at the level of the main tension reinforcement, have been observed [12.54,
12.55]. Braam [12.56] has also treated crack width control in deep beams.
Therefore, ACI-9.7.2.3 has provisions requiring side face (skin) reinforcement when
the depth h of a beam or joist exceeds 36 in. Such skin reinforcement must be distributed
along both side faces of the member for a distance h/2 from the tension face (Fig. 12.19.1).
The spacing s of the skin reinforcement is computed using ACI Table 24.3.2, Eq. (12.18.1)
for deformed bars, where cc is the clear cover from the skin reinforcement to the side face.
According to ACI-R9.7.2.3, the area of the skin reinforcement is not specified because
“research has indicated that the spacing rather than the bar size is of primary importance.”
Bar sizes of #3 to #5 with a minimum area of 0.1 sq in. per foot of depth are typically pro-
vided. It is permitted to include the skin reinforcement in strength computations when a
strain compatibility analysis is made.
Ask
h > 36 in.
s h
2
As
Figure 12.19.1 Side face (skin) reinforcement for deep beams.

456
EXAMPLE 12.19.1
Design the “skin” reinforcement according to the ACI Code for a rectangular beam
18 in. wide by 48 in. deep, having 10–#9 as tension reinforcement (Fig. 12.19.2);
fc′ = 4000 psi and fy = 60,000 psi.
SOLUTION
(a) Establish whether skin reinforcement is needed. Since the overall depth exceeds
36 in., longitudinal skin reinforcement is needed according to ACI-9.7.2.3.
(b) Determine the spacing to be used for the skin reinforcement. Try #3 bars and con-
servatively assume fs as 2 3 of the specified yield strength. Using a side cover cc of
2 in., the maximum spacing of the skin reinforcement is
 40, 000   40, 000 

s = 15  − 2.5cc = 15  − 2.5(2) = 10.0 in.
 fs    40, 000 
but not greater than 12(40,000/fs) = 12(40,000/40,000) = 12 in. Considering that the
center of the top layer of longitudinal reinforcement is located at about 4.2 in. from the
bottom of the beam, use two layers of #3 bars spaced at 10 in. to satisfy the spacing and
distribution requirements of ACI-9.7.2.3. Details are shown in Fig. 12.19.2.
#4
Stirrup
48” d = 44.9”
10”
10 – #9
#3 @ 10”
18”
Figure 12.19.2 Side face (skin) longitudinal reinforcement, for Example 12.19.1.
12.20 CONTROL OF FLOOR VIBRATIONS—G ENERAL

The tendency to use smaller, more efficient structural members has resulted in floor systems
that are more flexible, and thus, susceptible to undesirable vibrations under the normal use of
the structure. Excessive vibrations may, for example, impair proper functioning of sensitive
equipment or of high-precision manufacturing machines. Even when equipment malfunction
is not of concern, floor vibrations may disrupt the tasks being performed and result in loss
of productivity, or may simply cause discomfort to the occupants. Human response to floor
vibrations is quite complex, and it depends on many factors (e.g., the source, type and mag-
nitude of the vibration, the damping of the floor system, and the activities being performed
by the occupants).
To date, there is no unique criterion for establishing an acceptable floor vibration level.
However, two key parameters in evaluating the serviceability of floor systems subject to
vibrations are the peak acceleration and the vibration frequency of the motion. In office and
457
residential construction, it has been found that peak accelerations for human comfort should
not exceed about 0.5% of the acceleration of gravity, g [12.57, 12.58]. On the other hand,
participants in some activities (e.g., weight lifting, dancing) can tolerate accelerations of about
5% of g [12.57]. The maximum accelerations that will develop on a floor system will depend
on the source and nature of the motion but, in particular, on the relationship between the fre-
quency of vibration of the source and the natural frequency of vibration of the floor system. For
rhythmic excitation (e.g., aerobic dancing), peak accelerations can be largely amplified owing
to the phenomenon referred to as resonance. Resonance occurs when the vibration frequency
of the source is close to or coincides with the natural frequency of the structural member. In a
floor system, the vibration frequency of the source, fsource, is dictated by the nature of the motion
(e.g., dancing). On the other hand, the natural frequency of the floor system, fn, depends on
the mass and stiffness of its members. Therefore, current design practice for controlling floor
vibrations is generally based on “fine-tuning” the stiffness of the floor to avoid resonance—
that is, so that the natural frequency of the floor, fn, does not coincide with the frequency of the
source, fsource. Increasing the amount of damping will help reduce floor vibrations. However,
the practice of adding damping to a floor system has not always been successful [12.57], and it
requires the installation of tuned dampers, which may be costly.
The current ACI Code does not contain specific provisions for the control of vibration of
reinforced concrete structures. However, a good summary of the background, acceptance
criteria, and design examples can be found in Refs. 12.59 and 12.60.
SELECTED REFERENCES
12.1. ACI Committee 435. “Control of Deflections in Concrete Structures,” Report ACI 435R-
95 (Reapproved 2000; Appendix B added 2003). Farmington Hills, MI: American Concrete
12.2. ACI Committee 435. “Proposed Revisions By Committee 435 to ACI Building Code and
Commentary Provisions on Deflections” ACI Journal, Proceedings, 75, June 1978, 229–238.
12.3. ACI Committee 435, Subcommittee 7. “Deflections of Continuous Beams,” ACI Journal,
12.4. ACI Committee 435. “Variability of Deflections of Simply Supported Reinforced Concrete
Beams,” ACI Journal, Proceedings, 69, January 1972, 29–35.
12.5. ACI Committee 435, Subcommittee 1. “Allowable Deflections,” ACI Journal, Proceedings, 65,
June 1968, 433–444. Disc., 65, 1037–1038.
12.6. ACI Committee 435. “Deflections of Reinforced Concrete Flexural Members,” ACI Journal,
Proceedings, 63, June 1966, 637–674.
12.7. Mark Fintel (Editor). Handbook of Concrete Engineering. New York: Van Nostrand Reinhold,
1985 (892 pp.) (see Chapter 2, “Deflections,” p. 53).
12.8. Wei-Wen Yu and George Winter. “Instantaneous and Long-Time Deflections of Reinforced
Concrete Beams under Working Loads,” ACI Journal, Proceedings, 57, July 1960, 29–50.
Disc., 57 1165–1171.
12.9. Dan E. Branson. “Instantaneous and Time-Dependent Deflections of Simple and Continuous
Reinforced Concrete Beams,” Part 1, Report No. 7. Alabama Highway Research Report, Bureau
of Public Roads, August 1963 (1965) (pp. 1–78).
12.10. Dan E. Branson. Discussion of “Variability of Deflections of Simply Supported Reinforced
Concrete Beams,” by ACI Committee 435, ACI Journal, Proceedings, 69, July 1972, 449–451.
12.11. Lian Duan, Fu-Ming Wang, and Wai-Fah Chen. “Flexural Rigidity of Reinforced Concrete
Members,” ACI Structural Journal, 86, July–August 1989, 419–427. Disc., 87, May–June
1990, 364–365.
12.12. G. W. Washa and P. G. Fluck. “Effect of Compressive Reinforcement on the Plastic Flow of
Reinforced Concrete Beams,” ACI Journal, Proceedings, 49, October 1952, 89–108.
12.13. G. W. Washa and P. G. Fluck. “Plastic Flow (Creep) of Reinforced Concrete Continuous
12.14. Dan E. Branson. Deformation of Concrete Structures. New York: McGraw-Hill, 1977.
12.15. Dan E. Branson. Discussion of “Proposed Revision of ACI 318-63 Building Code Requirements
for Reinforced Concrete,” ACI Journal, Proceedings, 67, September 1970, 692–693.
12.16. LeRoy A. Lutz. “Graphical Evaluation of the Effective Moment of Inertia for Deflection,” ACI
Journal, Proceedings, 70, March 1973, 207–213. Disc., 70, September 1973, 662–663.
458
12.17. Bahram M. Shahrooz. “A Simplified Method for Computing Effective Moment of Inertia,”
Concrete International, 14, January 1992, 38–40.
12.18. Abdulrahman H. Al-Shaikh and Rajeh Z. Al-Zaid. “Effect of Reinforcement Ratio on the
Effective Moment of Inertia of Reinforced Concrete Beams,” ACI Structural Journal, 90,
March–April 1993, 144–149.
12.19. Dan E. Branson and Heinrich Trost. “Unified Procedures for Predicting the Deflection and
Centroidal Axis Location of Partially Cracked Nonprestressed and Prestressed Concrete
Members,” ACI Journal, Proceedings, 79, March–April 1982, 119–130.
12.20. P. D. Zuraski, C. G. Salmon, and A. Fattah Shaikh. “Calculation of Instantaneous Deflections
for Continuous Reinforced Concrete Beams,” Deflections of Concrete Structures (SP-43).
12.21. N. H. Burns and C. P. Siess. “Repeated and Reverse Loading in Reinforced Concrete,” Journal
of the Structural Division, ASCE, 92, ST5 (October 1966), 65–78.
12.22. Koladi M. Kripanarayanan and Dan E. Branson. “Short-Time Deflections of Beams Under
Single and Repeated Load Cycles,” ACI Journal, Proceedings, 69, February 1972, 110–117.
12.23. ACI Committee 209. “Prediction of Creep, Shrinkage, and Temperature Effects in Concrete
Structures,” Designing for Effects of Creep, Shrinkage, and Temperature in Concrete Structures
(SP-27). Detroit: American Concrete Institute, 1971 (pp. 51–93).
12.24. ACI Committee 209. “Effects of Concrete Constituents, Environment, and Stress on the Creep
and Shrinkage of Concrete,” Designing for Effects of Creep, Shrinkage, and Temperature in
Concrete Structures (SP-27). Detroit: American Concrete Institute, 1971 (pp. 1–42).
12.25. B. L. Meyers and D. E. Branson. “Design Aid for Predicting Creep and Shrinkage Properties of
Concrete,” ACI Journal, Proceedings, 69, September 1972, 551–555.
12.26. M. R. Hollington. A Series of Long-Term Tests to Investigate the Deflection of a Representative
Precast Concrete Floor Component, Technical Report TRA 442. London: Cement and Concrete
Association, April 1970, 43 pp.
12.27. Dan E. Branson. “Compression Steel Effect on Long-
Time Deflection,” ACI Journal,
Proceedings, 68, August 1971, 555–559.
12.28. Kent A. Paulson, Arthur H. Nilson, and Kenneth C. Hover. “Long-Term Deflection of High-
Strength Concrete Beams,” ACI Materials Journal, 88, March–April 1991, 197–206.
12.29. T. C. Hansen and A. H. Mattock. “Influence of Size and Shape of Member on Shrinkage and
Creep of Concrete,” ACI Journal, Proceedings, 63, February 1966, 267–289.
12.30. Hans Gesund. “Shrinkage and Creep Influence on Deflections and Moments of Reinforced
Concrete Beams,” ACI Journal, Proceedings, 59, May 1962, 689–704.
12.31. Alfred L. Miller. “Warping of Reinforced Concrete Due to Shrinkage,” ACI Journal,

Proceedings, 54, May 1958, 939–950.
12.32. A. F. Shaikh. Discussion of “Proposed Revision of ACI 318–63 Building Code Requirements
for Reinforced Concrete,” ACI Journal, Proceedings, 67, September 1970, 722–723.
12.33. Dan E. Branson. “Design Procedures for Computing Deflections,” ACI Journal, Proceedings,
65, September 1968, 730–742.
12.34. Jacob S. Grossman. “Simplified Computations for Effective Moment of Inertia Ie and Minimum
Thickness to Avoid Deflection Computations,” ACI Journal, Proceedings, 78, November–
December 1981, 423–439. Disc., 79, September–October 1982, 413–419.
12.35. Dan E. Branson. Discussion of “Simplified Computations for Effective Moment of Inertia Ie
and Minimum Thickness to Avoid Deflection Computations,” ACI Journal, Proceedings, 78,
November–December 1981, 423–439, ACI Journal, Proceedings, 79, September–October
1982, 413–414.
12.36. B. Vijaya Rangan. “Control of Beam Deflections by Allowable Span–Depth Ratios,” ACI
Journal, Proceedings, 79, September–October 1982, 372–377.
12.37. PCA. Notes on ACI 318–05 Building Code Requirements for Structural Concrete and
Commentary. Skokie, IL: Portland Cement Association, 2005.
12.38. ACI Committee 224. “Control of Cracking in Concrete Structures,” Report ACI 224R-01.
12.39. ACI Committee 224. “Causes, Evaluation, and Repair of Cracks in Concrete Structures,”
Report ACI 224.1R-07. Farmington Hills, MI: American Concrete Institute, 2007, 26 pp.
12.40. Edward G. Nawy. “Crack Control in Reinforced Concrete Structures,” ACI Journal,

Proceedings, 65, October 1968, 825-836. Disc. 66, 308–311.
12.41. David Darwin, David G. Manning, Eivind Hognestad, Andrew W. Beeby, Paul F. Rice, and
Abdul Q. Ghowrwal. “Debate: Crack Width, Cover, and Corrosion,” Concrete International, 7,
May 1985, 20–35.
12.42. A. W. Beeby. “Cracking, Cover, and Corrosion of Reinforcement,” Concrete International, 5,
February 1983, 35–40.
459
PROBLEMS 459
12.43. Bengt B. Broms and LeRoy A. Lutz. “Effects of Arrangement of Reinforcement on Crack Width
and Spacing of Reinforced Concrete Members,” ACI Journal, Proceedings, 62, November
1965, 1395–1410. Disc. 62, 1807–1812.
12.44. Peter Gergely and LeRoy A. Lutz. “Maximum Crack Width in Reinforced Concrete

Flexural Members,” Causes, Mechanism, and Control of Cracking in Concrete (SP-20).
12.45. A. W. Beeby. “The Prediction of Crack Widths in Hardened Concrete,” The Structural Engineer,
57A, 1, January 1979, 9–17.
12.46. R. J. Frosch. “Another Look at Cracking and Crack Control in Reinforced Concrete,” ACI
Structural Journal, 96, May-June 1999, 437–442.
12.47. LeRoy A. Lutz, Nand K. Sharma, and Peter Gergely. “Increase in Crack Width in Reinforced
Concrete Beams Under Sustained Loading,” ACI Journal, Proceedings, 64, September 1967,
538–546.
12.48. A. W. Beeby. “The Prediction and Control of Flexural Cracking in Reinforced Concrete
Members,” Cracking, Deflection, and Ultimate Load of Concrete Slab Systems (SP-30).
12.49. John P. Lloyd, Hassen M. Rejali, and Clyde E. Kesler. “Crack Control in One-Way Slabs
Reinforced with Deformed Wire Fabric,” ACI Journal, Proceedings, 66, May 1969, 366–376.
12.50. A. W. Beeby. “An Investigation of Cracking in Slabs Spanning One Way,” Technical Report No.
TRA 433. London: Cement and Concrete Association, April 1970, 32 pp.
12.51. ACI Committee 350. “Concrete Sanitary Engineering Structures,” ACI Journal, Proceedings,
80, November–December 1983, 467–486.
12.52. LeRoy A. Lutz. “Crack Control Factor for Bundled Bars and for Bars of Different Sizes,” ACI
Journal, Proceedings, 71, January 1974, 9–10.
12.53. Gregory C. Frantz and John E. Breen. “Cracking on the Side Faces of Large Reinforced
Concrete Beams,” ACI Journal, Proceedings, 77, September–October 1980, 307–313.
12.54. Gregory C. Frantz and John E. Breen. “Design Proposal for Side Face Crack Control

Reinforcement for Large Reinforced Concrete Beams,” Concrete International, 2, October
1980, 29–34.
12.55. Perry Adebar and Joost van Leeuwen. “Side-Face Reinforcement for Flexural and Diagonal
Cracking in Large Concrete Beams,” ACI Structural Journal, 96, September–October 1999,
693–704.
12.56. C. R. Braam. “Control of Crack Width in Deep Reinforced Concrete Beams,” Heron, 35
(1990), 4 (published jointly by Delft University of Technology, Delft, The Netherlands, and
TNO—Institute for Building Materials and Structures, Rijswijk, The Netherlands).
12.57. David E. Allen. “Building Vibrations from Human Activities,” Concrete International, 12, June
1990, 66–73.
12.58. “Evaluation of Human Exposure to Whole-Body Vibration, Part 2: Vibration in Buildings (1 Hz
to 80 Hz),” (ISO 2631-2). International Standards Organization, Geneva, 2003.
12.59. David A. Fanella and Mike Mota. “Design Guide for Vibrations of Reinforced Concrete Floor
Systems,” 1 ed. Schaumburg, IL: Concrete Reinforcing Steel Institute, 2014.
12.60. Applied Technology Council (ATC). ATC Design Guide 1—Minimizing Floor Vibration. ATC,
Redwood City, CA, 1999.
PROBLEMS
All problems are to be done in accordance with the ACI Code unless otherwise indicated.
12.1 For the case assigned by the instructor, calcu-

late the stress fs at the centroid of the steel and Case Es /Ec Mw d (in.) b (in.) Tension Bars
fc(max) at the extreme compression fiber of the (ft-kips)
concrete due to application of the given service
1 10 85 19.5 12 3–#9
load bending moment Mw.
2 9 52 19.5 12 3–#7
(a) using internal equilibrium conditions (see 3 9 92 19.5 12 3–#10
Example 12.4.1) 4 8 400 32.7 16 7–#9 (2 layers;
(b) The flexure formula method with trans- 5 and 2 bars)
formed section (see Example 12.5.1) 5 8 400 27.5 22 7–#10
6 8 110 18.5 12 2–#8, 2–#10
460
12.2 Compute the immediate deflections due to dead 12.4 Compute the immediate deflections due to
load and live load on the beam of the figure dead load and live load on the beam of the fig-
for Problem 12.2. The span of the uniformly ure for Problem 12.4. The uniformly loaded,
loaded, simply supported beam is 30 ft, and the simply supported beam on a span of 30 ft
maximum service load moments are 80 ft-kips must resist maximum service load moments of
for dead load and 95 ft-kips for live load. Use 20 ft-kips dead load and 14 ft-kips live load. Use
fc′ = 3500 psi (n = 8.5) and fy = 60,000 psi. fc′ = 3000 psi (n = 9) and fy = 40,000 psi.
12.5 Repeat Problem 12.4, except the service load
12” moments are 29 ft-kips dead load and 19 ft-kips
live load, and fy = 60,000 psi.
10”
d = 19.5”
22”
3 – #10
13”
15”
Problems 12.2, 12.8, and 12.9

2 – #8
12.3 Compute the immediate deflections due to dead
load and live load on the beam of the figure for
Problem 12.3. The uniformly loaded, simply Problems 12.4, 12.5, 12.12, and 12.13
supported beam on a span of 29 ft must resist
maximum service load moments of 300 ft-kips 12.6 Investigate the acceptability of the beam of
dead load and 500 ft- kips live load. Use the figure for Problem 12.6 for immediate live
fc′ = 4000 psi (n = 8) and fy = 40,000 psi. load deflection if the limitation is L/360. Use
fc′ = 3500 psi (n = 8.5) and fy = 40,000 psi.
18”
12.7 Investigate the acceptability of the beam of
the figure for Problem 12.7 for immediate live
load deflection if the limitation is L/360. Use
fc′ = 3000 psi (n = 9) and fy = 60,000 psi.
Problem 12.2 if the limit for live load plus creep
and shrinkage deflection is L/360. Consider
d = 36.25” 40”
that none of the live load is sustained. Use data
computed in Problem 12.2 if that problem was
previously assigned.
2 – #9 12.9 Repeat Problem 12.8, except use the separate creep
8 – #11
and shrinkage multiplier procedure instead of the
ACI Code method. Assume that humidity is 80%,
age at loading is 28 days after initial curing period,
Problems 12.3, 12.10, and 12.11 and duration of sustained load is 5 years.
10” 10” B
A
w
M
6 – #7
d = 15.6”
19” A
d = 15.9” B
5 – #9 20’– 0”
38.4 ft-kips (DL)
+ 0 ft-kips (LL)
Section A–A Section B–B 42 ft-kips (DL)

75 ft-kips (LL) Service load
moments

461
PROBLEMS 461
A B C
w
7 – #9 each layer
B
40’–0”
A 630 ft-kips (DL) C

750 ft-kips (LL) 40”
d = 36.25”
405 ft-kips
382 ft-kips (DL) + (DL)
309 ft-kips (LL) 505 ft-kips
– (LL)
Service load 24”
moments Section A–A
45” 6 12 ” 6 12 ”
4 – #11 and
8 – #11 d = 35.1” 3 – #9 each
40”
9 – #10 40” layer d = 36”
4 – #9
24” 24”
Section B–B Section C–C
12.10 Investigate the acceptability of the beam of and so on, which limit the maximum deflec-
Problem 12.3 if the beam supports partitions tion due to live load plus creep and shrinkage
and other nonstructural construction likely to to L/480. Assume that none of the live load is
be damaged by large deflections. Consider that sustained, the relative humidity is 90%, age at
20% of the live load is sustained. loading is 30 days after initial curing period,
12.11 Repeat Problem 12.10, except use the sepa- and duration of sustained load is 5 years
rate creep and shrinkage multiplier procedure or more.
instead of the ACI Code method. The deflec-
(a) Use ACI Code method.
tion limit is the same as for ACI Code, however.
(b) Use separate creep and shrinkage multi-
Assume that humidity is 90%, age at loading is
plier procedure.
60 days after initial curing period, and duration
of sustained load is 5 years. 12.14 Investigate the acceptability of the beam
12.12 Investigate the acceptability of the beam of of Problem 12.6 if the live load plus creep
Problem 12.4 if the beam supports partitions, and shrinkage deflection is limited to L/250
and so on, which limit the maximum deflec- and 10% of the live load is considered sus-
tion due to live load plus creep and shrinkage tained. Assume that relative humidity is 50%,
to L/360. Assume that none of the live load is age at loading is 10 days after initial curing
sustained, the relative humidity is 50%, age at period, and duration of sustained loading is
loading is 20 days after initial curing, and sus- 1 year.
tained load will be in place for 1 year.
(a) Use ACI Code method.
(a) Use ACI Code method. (b) Use separate creep and shrinkage multi-
(b) Use separate creep and shrinkage multiplier procedure.
plier procedure.
(c) Use combined creep and shrinkage multi-
Problem 12.7 if the beam supports partitions
plier procedure.
and other nonstructural construction likely to
12.13 Investigate the acceptability of the beam of be damaged by large deflections. Assume that
Problem 12.5 if the beam supports partitions, 10% of the live load is sustained.
462
12.16 Investigate the acceptability of an interior 45-ft 12.17 Check the ACI crack control provisions for the
span (see the figure for Problem 12.16) of a con- tension steel detail of the figure for Problem
tinuous T-section with regard to deflection. The 12.17. Assume 1.5-in. clear cover and that #4
service load moments at midspan are 40 ft-kips stirrups are used.
dead load and 100 ft-kips live load; at both sup- 12.18 Check the ACI crack control provisions for the
ports these moments are 50 ft-kips dead load and tension steel detail of the figure for Problem
114 ft-kips live load. Assume that the L/480 limit 12.18. Assume 1.5-in. clear cover and that #4
of ACI Table 24.2.2 applies and that none of the stirrups are used.
live load should be considered as sustained. Use
fc′ = 3500 psi (n = 8.5) and fy = 60,000 psi.
27” 27”
6 – #8 2 12 ” 4”
4”
18”
5 – #8 2”
3 – #8 2 12 ” 2 12 ”
12” 12”
End sections Center section

Regions of negative moment Regions of positive moment
Problem 12.16
2 12 ” 2 12 ”
6 – #7
3 – #11
d = 36.1”
d = 27.3”
40”
30”
10 – #11
4 – #11
16” 18”
Problem 12.17 Problem 12.18
CHAPTER 13
SLENDERNESS EFFECTS
ON COLUMNS
13.1 GENERAL
In the basic treatment of compression members in Chapter 10, it was assumed that the
effects of buckling and lateral deflection on strength were small enough to be neglected.
Short compression members—that is, those having a low slenderness ratio L /r (L = column
height and r = radius of gyration = I / A )—experience a material failure (crushing of con-
crete) prior to reaching a buckling mode of failure. Further, the lateral deflections of short
compression members subjected to bending moments are small, thus contributing little
secondary bending moment PΔ as shown in Fig. 13.1.1. It is these buckling and deflec-
tion effects that reduce the strength of a compression member below the value computed
according to the principles of Chapter 10.
Over time, the use of higher-strength steel and concrete has led to the design of more
slender members. A stocky member having an L /r less than 20 will essentially exhibit a
material failure whose strength may be computed by using the procedures discussed in
Chapter 10. In contrast, a member having L /r greater than about 70 will have a considerable
reduction in strength, not only due to the increased likelihood of buckling but because of
an amplification of the bending moment due to second-order effects. To permit the greatest
flexibility in structural design, specifications must provide for adequate determination of
strength with any slenderness ratio. Thus the provisions of the ACI Code take into account
the slenderness effects on long compression members.
To compute the radius of gyration, r, ACI-6.2.5.1(a) permits the designer to use
Ig
r=
Ag
where Ig and Ag are the gross moment of inertia and cross-sectional area of the column,
respectively. Alternatively, r may be computed for a rectangular column having b as the
width and h as the depth as [ACI-6.2.5.1(b)]
1 3
Ig bh
r= = 12 = 0.288h ≈ 0.30h
Ag bh
46
464 C H A P T E R 1 3 S L E N D E R N E S S E F F E C T S O N C O L U M N S
Long columns in Denver. (Photo by C. G. Salmon.)
P P
Mi = primary
bending moment
Figure 13.1.1 Primary and secondary moment for beam-columns.
and for a circular column having h as the diameter as [ACI-6.2.5.1(c)]
Ig π h 4 ( 4)
r= = = 0.25h
Ag 64 π h 2
These values actually should be slightly larger owing to the effect of the reinforcement.
In the ACI Code, the evaluation of the effect of slenderness may be approximated by
using the moment magnifier method, whereby the sum of the primary and secondary
moments (Fig. 13.1.1) is treated as being equal to the product of the primary moment and
a magnification factor δ. The general idea relating to this approach is derivable from the
differential equation of a beam-column.
In the next several sections, the general concepts relating to the effect of slenderness on
the strength of compression members are presented.
465
1 3 . 2 B U C K L I N G O F C O N C E N T R I C A L LY L OA D E D C O L U M N S 465
13.2 BUCKLING OF CONCENTRICALLY LOADED

COLUMNS
Over 250 years ago Leonhard Euler derived the well-known Euler formula [13.1] for con-
centrically loaded columns stressed below the proportional limit. Engesser [13.2] in 1889
proposed the tangent modulus modification of the Euler formula. In 1910, von Kármán
[13.3] performed a series of careful tests verifying Engesser’s assumptions. The tangent
modulus formula has now been accepted as representing the lower bound for buckling
strength of concentrically loaded columns,
π 2 Et I
Pc = (13.2.1)
(kLu )2
where
Pc = buckling load
Et = tangent modulus of elasticity of concrete at the buckling load
I = effective moment of inertia of the section
kLu = equivalent pin-end length (Lu = actual unbraced length)
Although von Kármán was the first to use a rational analytical method for the inelastic
buckling of slender columns, his work did not consider reinforced concrete.
The fact that concentrically loaded columns rarely, if ever, exist in reinforced concrete
structures led investigators [13.4–13.22] to focus attention on the interaction of long col-
umns with beams in frame structures, resulting in the more rational provisions for slender-
ness effects on compression members beginning with the 1971 ACI Code.
The basic design limitation of a maximum axial strength equal to 80 or 85% of the
concentric capacity Po (see Fig. 10.11.1) means, of course, that from a practical viewpoint
the concentrically loaded column is considered not to exist. However, to help the reader
understand the effect of the slenderness ratio on the behavior of beam-columns over the
entire range from Pn = Po with Mn = 0 to Pn = 0 with Mn = Mo (see Fig. 10.6.1), the limiting
case of the concentrically loaded column is considered first.
To apply Eq. (13.2.1), a realistic expression for Et of concrete must be used. Since buck-
ling may occur at practically any value of concrete strain, it is necessary to know as accu-
rately as possible the stresses at all strain levels.
The idealized stress-strain diagram for steel was shown in Chapter 3 (Fig. 3.4.1), where
the modulus of elasticity is taken at 29,000,000 psi. A realistic stress-strain diagram for the
concrete in the compression zone is that of Hognestad [3.1], shown in Fig. 13.2.1, in which
the initial modulus of elasticity for concrete is taken as
Ec = 1, 800, 000 + 500 fc′′ psi (13.2.2)
where fc′′= 0.85 fc′. Such a stress-strain relationship may be used along with three assump-
tions: concrete resists no tensile stress, linear strain variation exists across the section, and
the deflected shape is part of a sine wave. On the basis of these assumptions, evaluation of
Eq. (13.2.1) gives the typical column strength curves for concentric loading, such as those
of Fig. 13.2.2.
A study of Fig. 13.2.2 shows that in the curves for fy = 40,000 psi there occurs a flat
leveling off (such as portion BC of Fig. 13.2.2) of the curve, indicating yielding of the
steel with a sudden drop in Es from 29 × 106 psi to zero. In such cases where εy < ε0,
the concrete may still increase the capacity (such as portion AB in Fig. 13.2.2) with an
increased strain up to ε0. For fy = 50,000 psi and fc′ = 4000 psi, the strains εy = 0.00173
and ε0 = 0.00194 are nearly equal, which means that little increase above the plateau of
yielding in the steel can occur. The effect of creep on long-time loading may be noted by
the cross-hatched region.
46
[ ( )]
ε ε 2
fc = fc′′ 2 εc – ε c
0 0
0.15fc′′
Stress fc
fc′′ = 0.85 fc’
Ec
ε0 = 2fc′′/Ec εcu = 0.0038

Strain εc
Figure 13.2.1 Hognestad’s stress-strain diagram for flexure [3.1].
EXAMPLE 13.2.1
Calculate the ordinate and abscissa of points A, B, and C of Fig. 13.2.2. Use fc′ = 4000 psi,
fy = 40,000 psi, and a steel area As = 0.02bh (0.01bh in each opposite face located at
0.45h from the center). Use the stress-strain diagrams of Figs. 13.2.1 and 3.4.1.
SOLUTION
The basic quantities of the concrete stress-strain diagram are computed first:
fc′′ = 0.85 fc′ = 0.85(4) = 3.4 ksi

Ec = 1800 + 500 fc′′ = 1800 + 500(3.4) = 3500 ksi
2 fc′′  3.4 
ε0 = = 2 = 0.00194
Ec  3500 
fy 40
εy = = = 0.00138
Es 29, 000
(a) Point A, maximum strength of section according to the principles of Chapter 10

(upper limit, εc = ε0 > εy),

Pn = 0.85 fc′ bh + As f y
without correcting for the displaced concrete.
Pn = 0.85 fc′ bh + 0.02bh(40)
 0.02(40) 
= 0.85 fc′ bh 1 +
 0.85(4) 
Pn
= 1.235 (ordinate of point A)
0.85 fc′ bh
(Continued)
467
1 3 . 2 B U C K L I N G O F C O N C E N T R I C A L LY L OA D E D C O L U M N S 467
(b) Point B, εc = εy = 0.00138 < ε0, Es = 0.
  ε   ε 2
fc = fc′′ 2  c  −  c  
  ε 0   ε 0  
  1.38   1.38  2 
= 3.4 2   −  
  1.94   1.94  
= 3.4[2(0.712) − 0.508] = 3.12 ksi
Pn = fc bh + As f y = 3.12bh + 0.02bh(40) = 3.92bh
Pn 3.92
= = 1.153 (ordinate of point B)
0.85 fc′ bh 3.4
0.45h 0.45h
1.6
fy = 50 ksi
ρ = 0.04
fy = 50 ksi
1.4 b
fy = 40 ksi
A
1.2
C
B
h
Pn /(0.85fc’bh)
1.0
fy = 50 ksi
ρ = 0.01
ρ = 0.02
0.5
fy = 50 ksi
ρ = 0.02
long-time loading
70 140 210
kL
Slenderness ratio, u
r
Figure 13.2.2 Strength curves for reinforced concrete ( fc′ = 4000 psi) concentrically loaded
pin-end columns. (Adapted from Ref. 13.23.)
(Continued)
468
Applying Eq. (13.2.1) with I of the uncracked transformed section,
dfc  2 2ε  Eε
Et = = fc′′  − 2c  = Ec − c c
dεc  ε0 ε0  ε0
 1.38 
= 3500  1 − = 1008 ksi
 1.94 
bh3 E 
I= + 0.02bh  s  (0.45h)2
12  Et 
1  0  2
= bh3  + 0.02   (0.45)  = 0.0833bh
3
 12  1008 

π 2 Et I
(kLu )2 =
Pc
π 2 (1008)(0.0833bh3 )
(kLu )2 = = 212h 2
3.92bh
kLu
= 14.55
h
Where the first equation for (kLu )2 is a restatement of Eq. (13.2.1).
For a rectangular section, r = h / 12,
kLu
= 14.55 12 = 50.3 (abscissa of point B)
r
(c) Point C, εc at an infinitesimal amount less than εy; E = 29,000 ksi.
1  29, 000  
I = bh3  + 0.02   (0.45)2 
12  1008  
= bh3 (0.0833 + 0.1167) = 0.20bh3
π 2 (1008)(0.20bh3 )
(kLu )2 = = 507h 2
3.92bh
kLu
= 22.5
h
kLu
= 22.5 12 = 78 (abscissa of point C )
r
13.3 EFFECTIVE LENGTH FACTOR
For conditions other than pin ends, where the factor k in Eq. (13.2.1) is 1.0, the equiva-
lent pin-end length (also called effective length) factor k must be determined for various
rotational and translational end restraint conditions. Where translation at both ends is ade-
quately prevented, the distance between points of inflection is shown in Fig. 13.3.1. For all
such cases the equivalent pin-end length is less than or equal to the actual unbraced length
(i.e., k ≤ 1.0).
469
13.3 EFFECTIVE LENGTH FACTOR 469
Figure 13.3.1 Equivalent pin-end (i.e., effective) lengths; no joint lateral translation.
P P
P
Lu Lu Lu
kLu > 2Lu

kLu = Lu kLu = 2Lu
Partial
restraint
P P
P
(a) End rotation (b) One end rotation fully (c) One end rotation partially restrained,
fully restrained restrained, other end other end unrestrained
unrestrained
Figure 13.3.2 Equivalent pin-end (i.e., effective) lengths; joint lateral translation possible.
If sidesway or joint lateral translation is possible, as in the case of an unbraced frame,

the equivalent pin-end length is greater than or equal to the actual unbraced length (i.e., k ≥
1.0), as shown in Fig. 13.3.2.
Reinforced concrete columns are in general part of a larger frame, and thus it is neces-
sary to understand the concepts of the braced frame (where joint lateral translation may be
assumed to be prevented by rigid bracing, shear walls, or attachment to an adjoining struc-
ture) and the unbraced frame (where stability is dependent on the stiffness of the beams and
columns that constitute the frame). As shown in Fig. 13.3.3(a) and 13.3.3(c), the effective
length kLu for cases where joint lateral translation is prevented may never exceed the actual
length Lu. In an unbraced frame [Fig. 13.3.3(b) and 13.3.3(d)] instability results in a side-
sway type of buckling with the effective length kLu always exceeding the actual length Lu.
Values for effective length factor k are discussed in Section 13.11.
470
P P P P
kLu kLu > 2Lu

Lu 2 Lu
0.7Lu < kLu < Lu
(a) Braced frame, hinged base (b) Unbraced frame, hinged base
0.5Lu < kLu < 0.7Lu

P P P P
Lu Lu
Lu < kLu < 2Lu
(c) Braced frame, fixed base
(d) Unbraced frame, fixed base
Figure 13.3.3 Equivalent pin-end (i.e., effective) lengths for columns in frames.
13.4 MOMENT MAGNIFICATION—M EMBERS WITH

TRANSVERSE LOADS—W ITHOUT JOINT LATERAL
TRANSLATION (i.e., NO SIDESWAY)
As stated previously, nearly all compression members are simultaneously subjected to
some bending moment that causes deflections. A deflected compression member with no
sidesway (i.e., with no joint lateral translation) will be further subjected to a secondary
bending moment PΔ, as shown earlier (see Fig. 13.1.1). One may consider this as a mag-
nification of the primary or first-order bending moment. An approximate determination
of the magnifying effect may be made by considering the member as finally achieving a
deflection Δmax, which is composed of the deflection ∆0 due to the primary bending moment
Mi and the additional deflection ∆1 due to the secondary moment from axial compression
(see Fig. 13.4.1). For a simply supported member, it may be assumed that the secondary
bending moment takes the shape of a sine curve (very nearly exact for members with no
end restraint and whose primary bending moment and deflection are both maximum at mid-
span). The midspan deflection ∆1 equals the moment of the M/(EI) diagram (for secondary
bending moment) between the support and midspan taken about the support. Thus
P  L 2  L PL2
∆1 = ( ∆ 0 + ∆1 )     = ( ∆ 0 + ∆1 ) 2 (13.4.1)
EI  2 π  π π EI
from which
 PL2 / (π 2 E I )   α 
∆1 = ∆ 0   = ∆ 0   (13.4.2)
1 − PL / (π EI ) 
2 2
1− α
where α = PL2/(π 2EI). Since Δmax is the sum of Δ0 and Δ1,
 α  ∆
∆ max = ∆ 0 + ∆1 = ∆ 0 + ∆ 0   = 0 (13.4.3)
1− α 1− α
471
1 3 . 4 M E M B E R S W I T H T R A N S V E R S E L O A D S — N O S I D E S WAY 471
Figure 13.4.1 Primary and secondary bending moments for a simply supported member with
transverse loads.
The maximum bending moment, including the effect of axial load, becomes
P∆0
M max = M m + P ∆ max = M m + (13.4.4)
1− α
On substituting the expression for Δmax of Eq. (13.4.3) and substituting P = απ 2EI/L2,
Eq. (13.4.4) becomes
 C 
M max = M m  m  = M m δ (13.4.5)
1− α
where
Cm
δ= = magnification factor (13.4.6)
1− α
and
 π 2 EI ∆ 0 
Cm = 1 +  − 1 α (13.4.7)
 Mm L
2

Thus, for common cases of single curvature deflection, the magnification factor to be applied
to the primary bending moment is Cm /(1 – α). For Cases 1 to 7 shown in Table 13.4.1, rigor-
ous solution of the differential equation may be obtained (see Section 13.5). The approximate
values of Cm for positive moment shown in this table for Cases 1 to 3, computed by using
Eq. (13.4.7), are in good agreement with those from theoretical solutions. Derivations for
approximate values of Cm for negative moment are presented by Iwankiw [13.24]. Positive
moments for the restrained end cases (4–7) are given to correlate with theoretical solutions
[13.24]. Inherent in the theoretical solutions for these restrained cases is the assumption that
the slope at a restrained support is zero, the practicality of which can be argued. One may
note that for all cases this Cm value will be close to 1.0, because in actual concrete structures
α (i.e., the ratio between the applied axial load and the column buckling strength) rarely
exceeds about 0.3. The approximate treatment of slenderness in ACI-6.6.4.5.3(b) conserva-
tively requires that Cm be taken as 1.0 for all cases with transverse loading between supports.
472
TABLE 13.4.1 SUGGESTED VALUES OF Cm FOR COMMON SITUATIONS WITH NO JOINT

TRANSLATION
Case Cm (Positive Cm (Negative Primary Bending Moment

Moment) Moment)
M M Mm
1 P P 1 + 0.2α – +
P w P Mm
2 1.0 – +
L/2 Q Mm
P P
3 1 – 0.2α – +
w Mm
P P +
4 1 – 0.3α 1 – 0.4α –
w Mm
P P
5 1 – 0.2α 1 – 0.4α –
+
–
L/2 Mm
Q +
6 P P 1.0 1 – 0.3α –
L/2 Mm
Q
7 P P 1 – 0.2α 1 – 0.2α +
– –
q Mm
P P +
8 Eq. (13.4.7) not available –
MA MB
MA = MB
13.5 MOMENT MAGNIFICATION—M EMBERS SUBJECT

TO END MOMENTS ONLY—W ITHOUT JOINT
LATERAL TRANSLATION (i.e., NO SIDESWAY)
Consider the general case shown in Fig. 13.5.1 wherein the end moments M1 and M2 consti-
tute the primary bending moment Mi, which is a function of z. The sum of primary and sec-
ondary moments causes the member to have a deflection y, and thus the secondary moment
is equal to P y. Stating the total moment Mz at the section z of Fig. 13.5.1 gives
d2 y
M z = Mi + Py = − EI (13.5.1)
dz 2
for members with constant EI, and dividing by EI gives

d2 y P M
2
+ y= − i (13.5.2)
dz EI EI
473
1 3 . 5 M E M B E R S W I T H E N D M O M E N T S O N LY — N O S I D E S WAY 473
M1 M2
P P z
y
z
y
Primary moment, Mi
M1 M2 M2 > M1
Py
Secondary moment, Py
Figure 13.5.1 Primary and secondary moments for a member subject to end moments only and no
joint lateral translation.
For design purposes, the general expression for moment Mz is of greater importance than
the deflection y. Differentiating Eq. (13.5.2) twice gives
d4 y P d2 y 1 d 2 Mi
+ =− (13.5.3)
dz 4
EI dz 2
EI dz 2
From Eq. (13.5.1),

2
d2 y M d4 y 1 d Mz
=− z and = −
dz 2
EI dz 4 EI dz 2
Substitution into Eq. (13.5.3) gives

2
1 d Mz P  Mz  1 d 2 Mi
− +  −  =−
EI dz 2
EI  EI  EI dz 2
Simplifying and letting λ2 = P/EI,
d 2 Mz d 2 Mi
+ λ2 Mz = (13.5.4)
dz 2
dz 2
which is of the same form as the deflection differential equation, Eq. (13.5.2).
The homogeneous solution for Eq. (13.5.4) is
M z = A sin λ z + B cos λ z (13.5.5)
To this must be added the particular solution that will satisfy the right-hand side of the dif-
ferential equation. In the special case of unequal end moments acting without transverse
loading,
 M − M1 
Mi = M1 +  2  z (13.5.6)
 L
Since
d 2 Mi
= 0
dz 2
47
Eq. (13.5.4) becomes a homogeneous equation, in which case Eq. (13.5.5) represents the
entire solution.
To determine the maximum moment,
dM z
= 0 = Aλ cos λ z − Bλ sin λ z (13.5.7)
dz
or
A
tan λ z = (13.5.8)
B
At maximum Mz,
A B
sin λ z = and cos λ z = (13.5.9)
A +B
2 2
A + B2
2
Substitution of Eq. (13.5.9) in Eq. (13.5.5) gives
A2 B2
M max = +
A2 + B 2 A2 + B 2

= A2 + B 2 (13.5.10)
Now the constants A and B are evaluated by applying the boundary conditions to
Eq. (13.5.5). The conditions are
(1) at z = 0, M z = M1
∴ B = M1
(2) at z = L, M z = M2
M 2 = A sin λ L + M1 cos λ L

M 2 − M1 cos λ L
∴A=
sin λ L
so that
 M − M1 cos λ L 

M2 =  2
 sin λ L  sin λ z + M1 cos λ z (13.5.11)
and
2
 M − M1 cos λ L 
M max =  2  + M1
2
 sin λ L
1 − 2( M1 /M 2 ) cos λ L + ( M1 /M 2 )2
= M2 (13.5.12)
sin 2 λ L
For the general case of a member subject to end moments, the maximum moment may be
either (1) the larger end moment M2 at the braced (supported) location, [Fig. 13.5.2(a)] or
(2) the magnified moment given by Eq. (13.5.12) that occurs within the span [Fig. 13.5.2(b)],
depending on the ratio M1/M2 and the value of α, where α is the ratio of P to the critical load
π 2EI/L2, making λ L = π α . Note that in Eq. (13.5.12) the bending moment ratio M1/M2 is
positive when the member is bent in single curvature, as shown in Fig. 13.5.1, and negative
when the member is bent in double curvature.
475
1 3 . 5 M E M B E R S W I T H E N D M O M E N T S O N LY — N O S I D E S WAY 475
To compute the required strength of a member, one needs to know whether the max-
imum moment occurs at a location between the supports, and if so, the correct location.
The concept of equivalent uniform moment [Fig. 13.5.2(c)] can eliminate the need for such
information. Thus, for the case with unequal end moments, use of the equivalent moment
assumes that Mmax is at midspan.
To establish the equivalent moment, let M1 = M2 = Mequiv in Eq. (13.5.12),
2(1 − cos λ L )
M max = M equiv (13.5.13)
sin 2 λ L
Equating Eq. (13.5.12) and Eq. (13.5.13) gives
( M1 / M 2 )2 − 2( M1 / M 2 ) cos λ L + 1
M equiv = M 2 (13.5.14)
2(1 − cos λ L )
According to the procedure of Section 13.4, the approximate expression for maximum
moment is
 C 
M max = M m δ = M m  m  (13.5.15)
1− α
For the case of uniform moment (M1 = M2 = Mequiv),
M equiv L2
∆0 =
8 EI
M m = M equiv
Mz < M2 for all values of α
M2
(a) Max moment at end
Zero
Mmax > M2 slope
M2
M1
(b) Max moment not at end

Mmax
Mequiv Mequiv
(c) Equivalent uniform moment with max

magnified moment at midspan
Figure 13.5.2 Combined primary and secondary bending moment diagrams for members having
end moments without transverse loading.
476
Replacing these expressions into Eq. (13.4.7),
 π 2 EI  M equiv L2 
Cm = 1 +   2  − 1 α ≈ 1
 L  8 EIM equiv

Thus
 1 
M max = M equiv  (13.5.16)
 1 − α 
Substitution of Eq. (13.5.14) into Eq. (13.5.16) gives
 C 
M max = M 2  m  (13.5.17)
1− α
in which
( M1 /M 2 )2 − 2( M1 /M 2 ) cos λ L + 1
Cm = (13.5.18)
2(1 − cos λ L )
Comparing Eq. (13.5.17) with Eq. (13.5.16), Cm M2 may be considered to be the equivalent
uniform moment along the span.
Equation (13.5.18) assumes that the strength of the member is limited by excessive
deflection in the plane of bending. Also, it does not fully cover the double-curvature cases
where M1/M2 lies between –0.5 and –1.0. The actual failure mode of members bent in dou-
ble curvature with such bending moment ratios is generally one of “unwinding” from dou-
ble to single curvature in a sudden type of buckling.
Massonnet [13.25] and the AISC specification for steel buildings1 have suggested
approximate expressions for Cm to be used for design in place of Eq. (13.5.18). The ACI
Code has adopted the long-used expression from AISC to compute Cm [ACI-6.6.4.5.3(a)]:
M1ns
Cm = 0.6 − 0.4 (13.5.19)
M 2 ns
where the additional subscript ns denotes that these moments are those acting on a non-
sway compression member. Note that the M1ns /M2ns ratio in Eq. (13.5.19) is positive when
the member is bent in double curvature, and negative when the member is bent is single
curvature.
Figure 13.5.3 compares Eq. (13.5.18) with the recommendations of Massonnet [13.25]
and the ACI expression. The reader should note that for a given value of α, the curves
plotted using Eq. (13.5.18) are terminated when the moment M2 at the end of the member
exceeds the magnified moment. The straight line adopted by ACI-6.6.4.5.3(a) falls near the
upper limit for Cm at any given bending moment ratio, and thus it provides a realistic and
simple approximation. Also, note that for members bent in double curvature, Cm will vary
between 0.2 and 0.6, while for members bent in single curvature, Cm will vary between
0.6 and 1.0.
1 See Manual of Steel Construction (14 Ed.). Chicago: American Institute of Steel Construction, 2011.
47
1 3 . 6 M O M E N T M AG N I F I C AT I O N — M E M B E R S W I T H S I D E S WAY 477
1.0 α= M1
0.1 M1ns [Eq. (13.5.19);
Cm = 0.6–0.4
M2ns ACl-6.6.4.5.3(a)]
M2 > M1
Max moment M2 at end

Equiv uniform moment
0.8 α = 0.2
α = 0.3 M2
P
0.6 α = 0.4
α = 0.5
Cm = ( )
M 2 M 2
( )
0.3 M1 + 0.4 M1 + 0.3
2 2
0.4 α = 0.6
(Massonnet–Ref. 13.25)
Cm =
α= 2P 2 α=
Curves plotted 0.7
π El/L α=
using Eq. (13.5.18) 0
0.2 α = .8
0.9
1.0 0.75 0.5 0.25 0 0.25 0.50 0.75 1.0

End moment ratio M1 / M2 or M1ns / M2ns
Single curvature Double curvature

Notes:
· M1 /M2 is positive for single curvature in Eq. (13.5.18) and in Massonnet’s equation [Ref. 13.25]
· M1ns /M2ns is positive for double curvature in Eq. (13.5.19) [ACI-6.6.4.5.3(a)]
Figure 13.5.3 Comparison of theoretical Cm with design recommendations for members subject to

end moments only, without joint lateral translation.
13.6 MOMENT MAGNIFICATION—MEMBERS WITH

SIDESWAY—U NBRACED (SWAY) FRAMES
In a braced (nonsway) frame, relatively stiff diagonal bracing or shear walls restrain lat-
eral movement, so that any end lateral displacement (or sidesway) of the member can be
assumed to be small. Moment magnification in columns of braced (nonsway) frames is
thus due to deflections between the ends of the member and may be computed using the
procedures described in Sections 13.4 and 13.5. In an unbraced (sway) frame, however, the
frame must rely on the stiffness of the beams and columns to limit the lateral translation
[see earlier: Figs. 13.3.2 and 13.3.3(b) and 13.3.3(d)] and thus, a relatively larger sidesway
deflection ∆ will be expected. Consequently, secondary moments will develop in columns
of unbraced frames as a result of both member end lateral translation and deflections occur-
ring between member ends.
To visualise the basic concepts of secondary moments that develop in an unbraced
(sway) frame, refer to Fig. 13.6.1(a), where the moments M1s and M2s (the subscript “s”
denotes “sway”) are the moments resulting from a first-order analysis. From equilibrium,
M1s + M 2 s = Vu Lc (13.6.1)
In the model of Fig. 13.6.1(a), acted on by Vu, the top of the member will exhibit a first-
order lateral translation ∆0 with respect of the bottom. This means that the gravity load, ΣPu,
is acting with an eccentricity ∆0, which, in turn, will increase the primary or first-order over-
turning moment Vu Lc by the amount ΣPu∆0. This additional moment will further increase
the relative lateral deflection, which upon equilibrium of the structure in the final displaced
478
∑Pu ∑Pu
M1s
Vu 1 1’
Lc
Vu
2 2’
M2s
∑Pu ∑Pu
(a) First-order analysis: For equilibrium, M1s + M2s = VuLc
∑Pu ∑Pu
δsM1s
Vu 1 1’
Lc
Vu
2 2’
δsM2s
∑Pu ∑Pu
Figure 13.6.1 First-and second-order end moments and sidesway deflection of columns in one
story of a multistory unbraced frame.
position will be ∆2u (larger than ∆0), as shown in Fig. 13.6.1(b). The total moments at the
column ends (primary plus secondary) then may be expressed as
δ s ( M1s + M 2 s ) = Vu Lc + ∑ Pu ∆ 2 u (13.6.2)
As shown by the right-hand side of Eq. (13.6.2), the total moment can be thought of as the
sum of the primary plus the secondary part, or as on the left, it can be thought of as the
primary moment magnified by the factor δs.
Second order effects due to lateral translation (sidesway) are usually more significant
under the action of lateral loads, but they can also occur under gravity loads alone. Indeed,
when the structural configuration or the loading condition is asymmetric, an unbraced
frame will exhibit sidesway under gravity loads. Furthermore, even if the frame and load-
ing condition were intended to be perfectly symmetric, randomly occurring construction
inaccuracies can cause columns to be “out of plumb” (i.e., the centroid of the top of the
column is not directly over the centroid of the bottom of the column) which will induce an
unintended eccentricity of the gravity loads (see study by MacGregor [13.28]).
As noted earlier, the lateral displacement in a braced system is usually small; thus the
secondary bending moments from any small amount of sidesway (the P-∆ effect) may
ordinarily be neglected. In unbraced frames, however, the relatively larger sidesway deflec-
tion usually cannot be ignored and must be accounted for in design. It is possible, however,
for a braced frame to undergo large sidesway deflection (e.g., if bracing is too flexible) or,
conversely, for an unbraced frame to develop very small sidesway deflection (e.g., if beams
and columns are very stiff). In other words, whether the P-∆ effects due to sidesway can or
479
1 3 . 7 I N T E R AC T I O N D I AG R A M S — E F F E C T O F S L E N D E R N E S S 479
cannot be neglected is a function of the amount of sidesway expected to occur under the
load combination being considered, regardless of whether the system is physically provided
with lateral bracing. Prior to 1995, the ACI Code Commentary indicated that the sidesway
effect could be neglected except in cases of asymmetrical gravity loading or structural con-
figuration where “appreciable” sidesway might occur. The ACI Code Commentary further
indicated that “appreciable” sway would be considered to exist when the calculated sway
∆ from a first-order frame analysis divided by the clear height Lu exceeded 1/1500. While
this recommendation no longer appears in the ACI Code Commentary, it does provide a
sense of the amount of sway under gravity load that could be permitted without incurring
significant P-∆ effects. The 2014 ACI Code requires that columns in frames be classified
as either nonsway or sway based on the expected amount lateral deflection or the percent
increase of first-order moments, as described later in Section 13.8.
For a more detailed treatment of braced and unbraced elastic frames, the reader is
referred to Refs. 13.26 and 13.27. Also, MacGregor and Hage [13.29] and MacGregor
[13.30] have provided an excellent summary of the various methods and the basic concepts
regarding the design of slender columns.
13.7 INTERACTION DIAGRAMS—E FFECT

OF SLENDERNESS
To understand the ACI Code procedure and its approximations as discussed in later sec-
tions of this chapter, the general approach is described for determining a point on the Pn–Mn
interaction diagram for kL/r not equal to zero. Chapter 10 treated the basic strength of a
section with zero kL/r, giving the interaction diagram shown in Fig. 10.6.1, where the curve
corresponds to the one designated kL/r = 0 in Fig. 13.7.1. Points A and B represent com-
binations of Pn and Mn with the neutral axes (N.A.) located at cA and cB, respectively, from
the extreme compression fiber whose crushing strain is taken as 0.003, as prescribed by the
ACI Code. The curve labeled kL /r ≈ 60 in Fig. 13.7.1 represents a typical strength interac-
tion curve that includes slenderness effects.
Pfrang and Siess [13.6], Pfrang [13.12], and MacGregor, Breen, and Pfrang [13.32] have
provided excellent discussions of the slenderness effects on interaction diagrams. A slender
column may fail in one of two ways: (1) by material failure, that is, by reaching a combined
Pn–Mn as computed by the methods of Chapter 10; or (2) by instability when an infinitesi
mal increase in axial load results in additional deflection such that equilibrium cannot be
achieved (i.e., buckling or instability failure).
Referring to Fig.13.7.2(a), consider a member with an increasing axial load P acting at
an initial eccentricity e. A “short” column (kL /r = 0) will follow the path of the straight line
to point C at P = P1 and can continue to carry an additional axial load and moment up to
point Pshort that lies on the interaction diagram for the member, that is, where material failure
cA
P0 kL ≈ 0
r εc = 0.003
kL ≈ 60
r
N.A. EIA
A
cB
Pn
εc = 0.003
EIB
B N.A.
M0
Mn
Figure 13.7.1 Slenderness effects on column strength interaction diagram.

480
Pn
Mm = P1e
P0
Pshort Primary eccentricity line

Material
kL = 0 failure D of Fig. 13.7.2(a)
Pn
r
C Plong C kL = 0
P1 P1 r
E D
Instability 70 35
kL ≈ 100 failure 100
r 130
M0 M0 Mr
Mn
(a) Slender column behavior (b) Interaction diagrams
Figure 13.7.2 Slender column interaction diagrams. (Adapted from Ref. 13.32.)
would occur. Consider a “long” column with large slenderness, say kL /r = 100. In such a
case, the member will follow the path to point D on the interaction diagram, where the mate-
rial failure is reached at an axial load Plong much less than Pshort. This behavior occurs because
lateral deflection of the column will in effect increase the axial load eccentricity and thereby
increase the applied moment to Mmax = Mm + P1Δmax as shown in Fig. 13.7.2(a). Note that
both the “short” and “long” columns would exhibit a material failure in this instance.
It is also possible, however, that the column could fail by instability. In such a case, the
column would follow the path (dashed) up to point E in figure 13.7.2(a). That is, it would
be unable to reach the material strength interaction diagram.
It is possible then to construct an interaction diagram where a material failure occur-
ring at D on the kL /r = 0 curve due to P1 plus a magnified moment Mmδ (equal to Mm +
P1Δmax) is plotted for the particular loading arrangement as point C on the primary moment
radial line for kL = 100, as shown in Fig. 13.7.2(b). Whatever the primary moment load-
ing arrangement—such as equal end eccentricities of axial load, unequal end eccentrici-
ties of axial load, or lateral transverse loading—the deflection of the member will differ,
and so the secondary bending moment will differ. This means that different types of pri-
mary moments will cause a member of kL /r = 100 to follow a curved path to intersect the
kL /r = 0 material strength interaction diagram at different locations such as point D. The
radial line through point C, however, is uniquely a function of the primary moment, which
is the same for all slenderness ratios.
The development of a correct interaction diagram for members of large slenderness,
such as Fig. 13.7.2(b), would require an elaborate analysis for each structure, taking into
account such factors as the following: (1) a realistic moment-curvature relationship, (2) the
time-dependent and cracking effects on deflections, and (3) influence of axial load on the
flexural stiffness of members. Mockry and Darwin [13.33] have made such an analysis and
prepared a practical set of design charts. Parme [13.34] has treated the subject in a similar
manner, and the Portland Cement Association has published design aids [13.35].
In practice, however, rather than constructing interaction diagrams for various kL /r val-
ues [Fig. 13.7.2(b)] the design of slender columns is done using the interaction diagram
for “short” columns (kL /r = 0) with the magnified moment. Instability or buckling failures
[point E in Fig. 13.7.2(a)] are not permitted.
13.8 ACI CODE—G ENERAL

Unless slenderness effects can be neglected (see Section 13.15), the design of columns in
both sway and nonsway frames in accordance with the ACI Code is to be done (a) by an
approximate moment magnifier method or (b) by conducting a second-order analysis from
which the magnified moments are obtained directly (see Section 13.12).
481
13.8 ACI CODE—GENERAL 481
Section Properties
In the moment magnifier method, a first-order analysis of the structure is used to compute
the first-order or primary bending moments. The moments thus computed are then mul-
tiplied by a moment amplification factor to obtain the total moments, including second
order effects. Ideally, the structure should be analyzed by using section properties that
take into account the influence of axial loads, the presence of cracked regions along the
length of the member, and the effects of the duration of the loads. However, the intrica-
cies and the computational effort involved in estimating the appropriate modeling param-
eters make such an approach often impractical for use in everyday design practice. For
this reason, alternative approaches based on simplified modeling assumptions are often
preferred. For a first-order, elastic analysis under factored loads, the ACI Code permits
the use of the properties listed in Table 13.8.1.
TABLE 13.8.1 MODULUS OF ELASTICITY, MOMENT OF INERTIA, AND

CROSS-SECTIONAL AREA FOR ELASTIC ANALYSIS AT FACTORED LOADS
(a) Modulus of elasticity Ec from ACI-19.2.2
(b) Moments of inertia [ACI-Table 6.6.3.1.1(a)]

Beams 0.35Ig
Columns 0.70Ig
Walls—uncracked 0.70Ig
Walls—cracked 0.35Ig
Flat plates and flat slabs 0.25Ig
(c) Area [ACI Table 6.6.3.1.1(a)] 1.0Ag
The moment- of-

inertia values of Table 13.8.1 are based on those suggested by
MacGregor and Hage [13.29] and include the effects of cracking and axial load on the
member stiffness. They also include a stiffness reduction factor, φK, as described next.
Consistent with the ultimate strength design approach, in which a strength reduction
factor is used, member stiffness should be also multiplied by a stiffness reduction factor,
φK, to account for the variability that exists in the member stiffness. Studies have shown
that the strength reduction factor φ and the stiffness reduction factor φK are not the same
[13.60]. For example, for an isolated column, φK was found to be about 0.75 whether the
column was tied or spirally reinforced. On the other hand, the column deflections in a
multistory frame will depend on the average concrete strength of all the columns in the
frame. The average concrete strength, and consequently, the average modulus of elasticity
of the columns in a frame, will be higher than those in a single understrength column (ACI-
R6.6.4.5.2). Based on these considerations, the moments of inertia given in Table 13.8.1
were obtained by using a stiffness reduction factor φK of 0.875 (larger than that of an iso-
lated column) to account for the variability of member stiffness.
Alternative moment of inertia expressions, more refined than those given in Table 13.8.1
[ACI Table 6.6.3.1.1(a)], are given in Table 13.8.2 [ACI Table 6.6.3.1.1.(b)]. These expres-
sions are based on the study by Khuntia and Ghosh [13.61], and require knowledge of the
axial force level, eccentricity, and reinforcing ratio. For this reason, these equations are
more appropriate for a more refined analysis rather than a preliminary or first estimate of
slenderness effects.
To account for the effects of load duration, the moments of inertia must be modified
when sustained loads act, as discussed later (see Sections 13.9 and 13.10).
Nonsway or Sway Frames
To compute the appropriate moment magnification by means of the moment magnifier
method, columns in frames must be classified as either nonsway or sway. According to
482
TABLE 13.8.2 ALTERNATIVE MOMENTS OF INERTIA FOR ELASTIC

ANALYSIS [ACI TABLE 6.6.3.1.1(b)]
Member Minimum I Maximum
Columns and  Ast   Mu P

walls
0.35Ig  0.80 + 25   1 − − 0.5 u  I g 0.875Ig
 Ag   Pu h Po 
Beams, flat plates,  b 

0.25Ig (0.10 + 25ρ)  1.2 − 0.2 w  I g 0.5Ig
and flat slabs  d
Po is the axial compressive strength of the member at zero eccentricity.
ACI-6.6.4.3, it is permitted to consider a frame as nonsway if one of the following condi-

tions is met.
1. The increase in column end moments due to second-order effects does not exceed
5% of the first-order end moments.
2. The story stability index, Q, is less than or equal to 0.05, where
∑ Pu ∆ 0
Q= (13.8.1)
Vu Lc
and
ΣPu = total vertical factored load in the story under consideration
∆0 = relative lateral deflection between the top and bottom of the story in question
due to Vu, computed using a first-order elastic frame analysis
Vu = factored shear in the story in question
Lc = length of the compression member under consideration, measured from center
to center of the joints in the frame
Alternatively, if bracing elements resisting lateral movement of a story have a stiffness of

at least 12 times the gross lateral stiffness of the columns in the direction considered, ACI-
6.2.5 permits the columns to be considered as braced against sidesway (i.e., nonsway).
Note that even if a column is part of a braced frame it may need to be treated as part of
a sway frame. For example, if the bracing elements (e.g., walls) are too flexible and relative
lateral end displacement cannot be ignored.
13.9 ACI CODE—M OMENT MAGNIFIER METHOD

FOR COLUMNS IN NONSWAY FRAMES
It was shown in Sections 13.4 and 13.5 that the maximum second-order moment in an
elastic member of a nonsway frame may be computed from Eq. (13.4.4), above, as follows:
P∆0
M max = M m + P ∆ max = M m + [13.4.4]
1− α
where α = P/Pc and Pc = π 2EI/L2. Furthermore, Eq. (13.4.5) for members with transverse
loading, and Eq. (13.5.15) for members with end moments only, indicate that the maximum
second-order moment may also be expressed as the maximum primary moment Mm times
a magnification factor,
 C 
M max =  m  M m [13.4.5]
1− α
483
1 3 . 9 A C I C O D E — C O L U M N S I N N O N S W A Y F R A M E S 483
or
M max = δ ns M m (13.9.1)
which is the same as ACI Formula (6.6.4.5.1)2, where

Cm
δ ns = (13.9.2)
1 − α
in which the subscript ns denotes nonsway. Several studies have shown [13.18, 13.32,
13.40] that this procedure is acceptable for reinforced concrete members. The design
strength of columns in nonsway frames is thus computed to satisfy the following equations:
φPn ≥ Pu and φ M n ≥ δ ns M m (13.9.3)
where design of the member is based on the factored axial load Pu combined with the
corresponding factored moment, Mm,2 computed from an elastic, first-order analysis, and
magnified by the factor δns given by ACI-6.6.4.5.2 as
Cm
δ ns = ≥ 1.0 (13.9.4)
Pu
1−
0.75Pc
The primary moment, Mm, in Eqs. (13.9.1) and (13.9.3) corresponds to the largest moment
occurring in the member and, in general, may occur at either end, or along the length of the
member. Equation (13.9.4) is basically the same as Eq. (13.9.2), where the critical buckling
load, Pc, has been multiplied by a stiffness reduction factor, φK, to account for the variability in
member stiffness. A single value φK = 0.75 is used for both tied and spirally reinforced isolated
columns as suggested by Mirza et al. [13.60]. The design strengths φPn, φMn are obtained from
the interaction diagram for short columns (kL /r = 0) as discussed in Section 13.7.
Factor Cm
For members with transverse loading, the quantity Cm may be computed from Eq. (13.4.7)
for simply supported members or from Table 13.4.1 for other support conditions. Since,
however, Cm will usually vary between 0.9 and 1.0 in practice, there is little advantage to
using the values given in Table 13.4.1. Accordingly, the ACI Code requires Cm to be taken
as 1.0 for members with transverse loading.
For members with end moments only, the ACI Code uses the approximate expression
given by Eq. (13.5.19) for Cm.
In summary, for columns in nonsway frames, Cm is computed under ACI-6.6.4.5.3 as
follows.
1. Members subject to transverse loading,
Cm = 1.0 (13.9.5)
2. Members subject to end moments only,
M1ns
Cm = 0.6 − 0.4 (13.9.6)
M 2 ns
where M1ns and M2ns are the first-order end moments. Note that M2ns is taken as the larger
of the end moments, and that the ratio M1ns/M2ns is negative when the member is bent in
single curvature.
2 Note that the ACI Code uses M2 instead of Mm in Eq. (13.9.1) [ACI Formula (6.6.4.5.1)]. Mm in Eq. (13.9.1) is the maxi-
mum moment occurring in the member and, in general, may occur at either end or, if there is transverse loading, within the span.
Here Mm is used instead to distinguish it from the larger end moment M2, in members with end moments only (see Section 13.5).
48
Stiffness Parameter EI
The other quantity required for evaluating the moment magnifier δns is the critical buckling
load, which in accordance with ACI-6.6.4.4.2 is computed as
π 2 ( EI )eff
Pc = (13.9.7)
(knonsway Lu )2
where Lu is the unsupported length of the member. The current ACI Code does not provide a
definition for Lu as in past editions, but it should be taken as the clear length of the column.
The parameter knonsway corresponds to the effective length factor k (see Section 13.3) and is
computed as described later (see Section 13.11) for nonsway (braced) frames.
To account for the effects of cracking, load duration, and nonlinearity of the con-
crete stress-strain curve, an effective stiffness parameter, (EI )eff, is used in Eq. (13.9.7).
MacGregor, Breen, and Pfrang [13.32] proposed the use of the larger of two simple expres-
sions. These appear in ACI-6.6.4.4.4 as Formulas (6.6.4.4.4a) and (6.6.4.4.4b), respectively,
0.4 Ec I g
( EI )eff = (13.9.8)
(1 + β dns )
or
0.2 Ec I g + Es I se
( EI )eff = (13.9.9)
(1 + β dns )
where
Ec = concrete modulus of elasticity = 57, 000 fc′ for normal-weight concrete (ACI-19.2.2.1)
Ig = gross moment of inertia of concrete section, neglecting reinforcement
Ise = moment of inertia of reinforcement about the concrete section centroid
βdns = ratio between the maximum factored axial load that is considered sustained and
the maximum factored axial load associated with the same load combination (usu-
ally the ratio between the factored axial dead load and the total factored axial load)
Alternatively, the stiffness parameter, (EI )eff, may be computed by using ACI Formula
(6.6.4.4.4c) as follows
Ec I
( EI )eff = (13.9.10)
(1 + β dns )
where I is computed in accordance with Table 13.8.2 [ACI Table 6.6.3.1.1(b)].

Under sustained loads, creep effects will increase the long-term deflections and, as a
result, they will also increase the second-order moments. To account for these effects, the
value of EI is divided by (1 + βdns) in Eqs. (13.9.8), (13.9.9), and (13.9.10).
The relative accuracy of Eqs. (13.9.8) and (13.9.9) as a function of the axial load ratio,
P/Pc, for the case of no sustained load (βdns = 0) is shown in Fig. 13.9.1 [13.32]. In the study
of Eqs. (13.9.8) and (13.9.9), EI values for about 100 cases were estimated by means of
theoretical load-moment-curvature diagrams for columns of various dimensions, strengths,
and steel percentages. Effective EI values were also computed for the University of Texas
frame tests [13.5, 13.8, 13.9], and for a series of computer simulations. It may be seen
that Eqs. (13.9.8) and (13.9.9) provide a lower-bound estimate of EI for most axial load
ratios, except at levels exceeding P/Pc ratios of about 0.8. Such high P/Pc ratios are rarely if
ever encountered in practice. Equation (13.9.8) is easier to use because it does not require
knowledge of the amount of reinforcement (which is not known in the initial stages of
design); however, it greatly underestimates the contribution of the reinforcement for heav-
ily reinforced columns.
485
1 3 . 1 0 A C I C O D E — C O L U M N S I N S W A Y F R A M E S 485
EcIg EcIg
EI = EI = + EsIse
2.5 5
5 2
EI from Eq. (13.9.9)

ρ = 0.08 ρ = 0.01
EI from (13.9.8)
Theoretical EI
Theoretical EI
4
3
1
2 ρ = 0.08
1
ρ = 0.01
0 0
0.6 0.7 0.8 0.9 0.6 0.7 0.8 0.9
P P
Pc Pc
(a) Eq. (13.9.8) (b) Eq. (13.9.9)
Figure 13.9.1 Comparison of Eqs. (13.9.8) and (13.9.9) for EI for short-duration loading (βdns = 0)
with EI values from moment-curvature diagrams. (Adapted from Ref. 13.32.)
Alternative expressions for EI have also been proposed by MacGregor, Oelhafen, and
Hage [13.41], Medland and Taylor [13.42], and Zeng, Duan, Wang, and Chen [13.43]. As a
result of these studies it seemed appropriate to divide only the concrete term in Eq. (13.9.9)
by the factor (1 + βdns). However, when this is done, the number of cases in which the ratio
of theoretical EI to the formula-computed EI is less than 1.0 becomes 3 times more than
otherwise, particularly when the reinforcement ratios are low. Thus, ACI Committee 318
concluded that the (1 + βdns) factor should be retained as the divisor for both the steel and
concrete terms. Studies of flexural stiffness expressions for EI include those of Diaz and
Roesset [13.44] and Mirza [13.45] for rectangular sections, and those of Sigmon and Ahmad
[13.46] and Ehsani and Alameddine [13.47] for circular sections. The sustained load effect
on slender columns has also been studied by Goyal and Jackson [13.48], Drysdale and
Huggins [13.49], and Rangan [13.50].
13.10 ACI CODE—M OMENT MAGNIFIER METHOD

FOR COLUMNS IN SWAY FRAMES
In this method, second-order effects are computed by combining the magnified moments
due to the deflections that occur along the length of the member (i.e., nonsway) with those
due to lateral translation at the member ends (i.e., sway). This is an approximate method
and can be used only in frames with loads causing deflections in the plane of the frame. If
significant torsion in plan exists, this procedure may significantly underestimate the mag-
nified moments. In such a case, a three-dimensional, second-order analysis must be used
(ACI-R6.6.4.6.1).
The moment magnifier method for sway frames consists of computing the first-order
nonsway moments, Mns, and the first-order sway moments, Ms, separately. The moments
thus computed are then amplified by a moment magnifier to find the second-order moments.
This approach of separating the nonsway and sway portions of the frame action is based, in
part, on the recommendations of Ford, Chang, and Breen [13.51]. While refinements to this
procedure have occurred over the years, the basic approach remains the same.
Computation of Nonsway Moments

According to the ACI Code, the nonsway moments, Mns, are to be computed from a first-
order, elastic frame analysis under loads that cause no appreciable sidesway. Traditionally,
this has been interpreted as an analysis of the frame under gravity loads alone. Clearly,
however, unless the layout of the frame, support conditions, and load distribution are per-
fectly symmetric, gravity loads will induce sidesway of the frame. The amount of appre-
ciable sidesway is not explicitly defined in the current ACI Code. Earlier editions of the
Code (ACI 318-92 Commentary) suggested that lateral deflections due to vertical loading
486
would be considered appreciable for a deflection ratio Δ /Lu greater than 1/1500. In other
words, if Δ /Lu is less than or equal to 1/1500, the effect of the sway component due to grav-
ity loads can be neglected and the corresponding moments can be considered as nonsway
moments. If that condition is not satisfied, the sway component due to vertical loads cannot
be ignored and must be considered. An alternative approach to computing the nonsway
moments, Mns, that does not depend on whether lateral deflections under gravity loads are
appreciable, is presented next.
(a) Frame subjected to factored vertical and lateral loads under the load
combination being considered
HR
H3
H2
H1
(b) Model for computing first-order nonsway moments
HR
H3
H2
H1
(c) Model for computing first-order sway moments
Figure 13.10.1 Computation of nonsway and sway moments

487
Consider the four-story, two-bay frame of Fig. 13.10.1(a) subjected to factored vertical

and lateral loads under a given load combination. In this approach, the nonsway moments
are computed by analyzing the frame under conditions that prevent lateral translation
in each floor (i.e., a nonsway condition is enforced). In this step, all the factored loads
(i.e., vertical and lateral loads) for the load combination being considered are applied, as
shown in Fig. 13.10.1(b). The bending moments thus obtained will then correspond to the
first-order, nonsway moments, Mns. Analysis of the frame under these load and support con-
ditions will generate horizontal reactions at the lateral supports from (a) the applied lateral
loads and (b) any sway component due to the applied vertical loads. These reactions are
then used to compute the sway moments, as described next. If no lateral loads exist under
the load combination being considered, any sway of the frame caused by gravity loads will
result in nonzero horizontal reactions. Thus, whether the amount of sway due to gravity
loads is appreciable is irrelevant because any sway effects due to gravity loads are included
in the horizontal reactions and the resulting sway moments when this approach is used.
Computation of Sway Moments

To obtain the first-order sway moments, Ms, a separate frame analysis is then conducted by
allowing lateral translation and by applying only the horizontal reactions obtained from the
nonsway analysis as lateral loads in the opposite direction, as shown in Fig. 13.10.1(c). The
actual loads acting on the frame are not applied in this analysis. The idea is that any tendency
to sway caused by the actual loads (vertical and lateral) is already captured by the horizontal
reactions computed from the nonsway analysis with lateral translations prevented.
Nonsway Moment Magnification Factor, δns

Consider the simple frame of Fig. 13.10.2(a) subjected to factored vertical and lateral loads.
The first-order moment diagrams obtained from the nonsway and sway analyses are also
shown in the figure. For the loading and support conditions shown, the column has no trans-
verse loading along its length and is bent in double curvature. This is the most common
situation found in practice (the special case of columns subjected to transverse loading or
in single curvature will be discussed later).
The nonsway moment magnification, δns, may be computed by using Eq. (13.9.4). Since
the column in Fig. 13.10.2 has no transverse loading, it can be treated as a member sub-
jected to end moments only, and thus, Cm may be computed from Eq. (13.9.6). For a mem-
ber bent in double curvature, the M1ns /M2ns ratio is always positive, and thus Cm will be at
most 0.6 when M1ns /M2ns = 0 (see Fig. 13.5.3). On the other hand, the ratio of Pu /(0.75Pc)
in Eq. (13.9.4) will rarely, if ever, exceed 0.4 in practice. As a result, the nonsway moment
magnification for a column bent in double curvature will generally be at most
0.6
δ ns <
∼ 1 − 0.4 = 1.0 (13.10.1)
This means that although the first-order moments are amplified along the length of the
member, they are smaller than at the ends, as shown in Fig. 13.10.2(b). In other words, the
largest nonsway moment for a column bent in double curvature, including second-order
effects, will most commonly occur at the member ends and will be equal to the first-order
moment M2ns [M2 in Fig. 13.5.2(a)]. This result is typical of most columns found in practice.
For this reason, for columns in sway frames, the ACI Code requires the nonsway moments,
including second order effects, to be taken as the first-order, factored moments at the mem-
ber ends, M1ns or M2ns. In other words, the ACI Code assumes δns = 1.0.
If the column has transverse loading along its length or if it is bent in single curvature
with end moments only, the maximum second-order moment can occur along the member
length and can be larger than that at the ends. This situation is discussed later when the
second-order nonsway and sway moments are combined.
48
P P
Mtop
ns
H H R
First
order Second order
bot
Mns
(a) (b)
top +
δs Mstop Mns δs Mstop
R
Mstop First
Moment
magnifier order
method
“True”
First second order Moment
order Msbot magnifier
method
bot +
δs Msbot Mns δs Msbot
(c) (d)
Figure 13.10.2 Computation of second-order moments using the moment magnifier method:

(a) simple frame subjected to gravity and lateral loads; (b) first-and second-order nonsway
moments; (c) first-and second-order sway moments; and (d) total first-and second-order moments.
(Second-order effects due to axial load in beam are neglected).
Sway Moment Magnification Factor δs

ACI-6.6.4.6.2 gives three approaches to computing the sway moment magnification factor,
δs, as follows:
1. Q Method: Use a magnification factor δs as follows (ACI Formula 6.6.4.6.2a):
1
δs = ≥ 1.0 (13.10.2)
1−Q
where Q is the story stability index given in ACI-6.6.4.4.1 as
∑ Pu ∆ 0
Q= (13.10.3)
Vu Lc
where
ΣPu = total vertical factored load in a story to be sway-resisted by the frame action
∆0 = relative lateral deflection between the top and bottom of the story in question
due to Vu, computed by using a first-order elastic frame analysis and stiffness
parameter values of ACI-6.6.3.1.1 (see Section 13.8)
Lc = length of the compression member in question, measured from center to center
of the joints in the frame
When δs computed according to Eq. (13.10.2) exceeds 1.5 [i.e., when Q from Eq. (13.10.3)
exceeds 1/3], either the more accurately computed sway magnifier of Eq. (13.10.4) (item 2
below) or a second-order elastic analysis (item 3 below) must be used.
489
2. Sum of P Method: The sway magnifier is computed according to ACI Formula

(6.6.4.6.1b), as follows:
1
δs = ≥ 1.0 (13.10.4)
∑ Pu
1−
0.75 ∑ Pc
where
π 2 ( EI )eff
Pc = (13.10.5)
(ksway Lu )2
and
ΣPc = summation of critical buckling loads for all sway-resisting columns
ksway = effective length factor k computed for unbraced (sway) frames (see Section 13.11)
Lu = unsupported length of the column taken as the clear length of the column
(EI)eff = Eqs. (13.9.8), (13.9.9), or (13.9.10) with βdns substituted for βds
βds = ratio of maximum factored sustained shear within a story to the maximum
factored shear in that story associated with the same load combination.
This factor is zero when sway moments are caused by loads of short dura-
tion, such as those induced by wind or earthquake. In the unusual case
of sway moments that are due to sustained loads, such as those caused
by lateral earth pressure or by appreciable sway due to sustained gravity
loads βds will not be zero.
3. Use a second-order elastic analysis based on the member stiffness parameters given
in ACI-6.6.3.1.1 (see Section 13.8). No magnification factor δs is used. The analysis
gives the magnified moments directly.
The typical resulting second-order sway moments are shown by the dashed lines in
Fig. 13.10.2(c). As shown, the moments are amplified along the length and at the ends,
with the maximum values occurring at the member ends in this case.
Combination of Nonsway and Sway Moments

In general, the total second-order moment in the column may be computed, albeit approx-
imately, by adding the nonsway and sway magnified bending moment diagrams along the
member length as follows:

2nd
M tot = δ ns M ns + δ s M s (13.10.6)
where
2nd
M tot = total factored moment including second-order effects
Mns, Ms = first-order factored moments computed from a nonsway and a sway frame
analysis, respectively
δns, δs = nonsway and sway moment magnifiers, respectively
Equation (13.10.6) assumes that the moment magnification factor is the same along the
length of the member, which is not true. Nonetheless, this approach provides in most cases
a reasonable estimate of the approximate second-order moment diagram for design pur-
poses. Because most columns in sway frames do not have transverse loading and because of
the presence of large double-curvature moments due to lateral loads, the nonsway moment
component including second-order effects is almost always maximum and equal to the
490
first-order moment at the column ends (i.e., δns = 1). As a result, the ACI Code requires the
total moment including second-order effects to be computed as the summation of the non-
sway and magnified sway moments at the column ends as follows (ACI-6.6.4.6.1):
M12 nd = M1ns + δ s M1s (13.10.7)
M 22 nd = M 2 ns + δ s M 2 s (13.10.8)
where
M12 nd, M 22 nd = smaller and larger factored end moment (including second-order effects),
respectively. [Note that the ACI Code refers to these moments as M1 and
M2 in ACI Formulas (6.6.4.6.1a) and 6.6.4.6.1b), respectively. Here, the
superscript “2nd” emphasizes that these moments include second-order
effects, and serves to avoid confusion with M1 and M2 in ACI Formula
(6.6.4.5.3a), and M2 in ACI Formula (6.6.4.5.1). The moments M1 and
M2 in these last two ACI Code equations refer to first-order moments.]
M1ns, M2ns = smaller and larger, first-order factored end moments, respectively, due to
loads that cause no appreciable sidesway (i.e., computed from a nonsway
frame analysis)
M1s, M2s = smaller and larger first-order factored end moments, respectively, due
to loads that cause appreciable sidesway (i.e., computed from a sway
frame analysis)
δs = sway moment magnifier
Equations (13.10.7) and (13.10.8) assume that the controlling combined effects for col-
umns in sway frames occur at the ends of the members, as for the column of the frame of
Fig. 13.10.2. This approach is reasonable when the column has no transverse loading along
its length and when it is bent in double curvature, as discussed earlier.
When the column is subjected to transverse loading along its length or, when the non-
sway moments cause single curvature, the maximum, nonsway moment, including second-
order effects, may occur along the length of the member rather than at the ends. Sway
moments, on the other hand, are always maximum at the ends. In other words, the locations
of the maximum magnified nonsway and sway moments do not coincide. In this case, a
reasonable estimate of the maximum second-order moment may be obtained by adding the
second-order nonsway and sway moments along the member length using Eq. (13.10.6),
as suggested earlier. This approach is not entirely correct because the true moment ampli-
fication factor is not constant along the member length. Rather, it leads to conservative
estimates of the maximum second-order moments. Otherwise, a second-order analysis (see
Section 13.12) should be used.
13.11 ALIGNMENT CHARTS FOR EFFECTIVE

LENGTH FACTOR k
The most commonly used procedure for obtaining the equivalent pin-end (effective) length
is to use the alignment charts from the Structural Stability Research Council Guide [13.27],
originally developed by O. J. Julian and L. S. Lawrence, and presented in detail by T. C.
Kavanagh [13.52]. The charts are shown in Fig. 13.11.1.
The effective length factor k is a function of the end restraint factors ψA and ψB, for the
top and bottom joints at the ends of the column, respectively, defined as
∑ EI /L for the columns in the plane of bending

ψ= (13.11.1)
∑ EI /L for thee beams in the plane of bending
491
13.11 ALIGNMENT CHARTS FOR EFFECTIVE LENGTH FACTOR k 491
which for a hinged end gives ψ = ∞ and for a fixed end, ψ = 0. Since a frictionless hinge
cannot exist in practical construction, ψ is to be taken equal to 10 for an end assumed as
hinged in the analysis. Similarly, perfect fixity rarely, if ever, can be achieved in reality.
Thus, it is recommended that ψ be taken equal to 1.0 for an assumed fixed end. The value
to be used for L in computing the end restraint factors ψA and ψB in Eq. (13.11.1) should be
the column or beam length measured from center to center of joints.
One nomogram (or alignment chart), Fig. 13.11.1(a), is for braced (nonsway)
frames where sidesway (joint translation) is prevented. The other nomogram, shown in
Fig. 13.11.1(b), is for unbraced (sway) frames where sidesway is possible, being restrained
only by the stiffness of interacting beams and columns.
The assumptions inherent in the development of the alignment chart for braced frames
[Fig. 13.11.1(a)] are as follows [13.52]:
1. All columns reach their respective critical loads simultaneously.

2. The structure is assumed to consist of symmetrical rectangular frames.
3. At any joint, the restraining moment provided by the beams is distributed among the
columns in proportion to their stiffnesses.
4. The beams are elastically restrained at their ends by the columns; at the onset of
buckling, the rotations of the beam at its ends are equal and opposite (i.e., the beams
are deflected in single curvature). See Fig. 13.3.3(a) and (c).
5. The beams carry no axial loads.
For the unbraced (sway) frame alignment chart [Fig. 13.11.1(b)], assumptions 1 through 3
and 5 are unchanged; however, the beams are assumed to be deflected in double curvature,
where the rotations of the ends are equal in magnitude and direction [see Fig. 13.3.3(b) and
(d)]. By means of the alignment charts, one may then determine the k factor for a column
of constant cross section in a multistory, multibay frame.
ψA knonsway ψB ψA ksway ψB
∞ ∞ ∞
50.0 1.0 50.0 ∞ 20.0 ∞
100.0 10.0 100.0
10.0 10.0 50.0 50.0
5.0 5.0 30.0 5.0 30.0
4.0 4.0
3.0 0.9 3.0 20.0 4.0 20.0
2.0 2.0
10.0 3.0 10.0
0.8 9.0 9.0
8.0 8.0
1.0 1.0 7.0 7.0
0.9 0.9 6.0 6.0
0.8 0.8 5.0 5.0
0.7 0.7
0.6 0.6 4.0 2.0 4.0
0.7
0.5 0.5
3.0 3.0
0.4 0.4
0.3 0.3 2.0 2.0

1.5
0.6
0.2 0.2
1.0 1.0
0.1 0.1
0 0.5 0 0 1.0 0
(a) Sidesway prevented (braced frame) (b) Sidesway not prevented (unbraced frame)
Σ EI /L,columns
Figure 13.11.1 Alignment charts for effective length factor for columns in continuous frames: ψ = . (From
Σ EI /L, beams
Ref. 13.52.)
492
The effective length procedure has been adopted by ACI-6.6.4.4.3 in the use of the
moment magnifier method to evaluate slenderness effects. ACI-6.6.4.4.3 says that for “non-
sway members k shall be permitted to be taken as 1.0, and for sway members, k shall be
at least 1.0.” Nonetheless, the use of the alignment charts for determining the k factor is
implicitly endorsed by their inclusion in the ACI Commentary (ACI-R6.6.4.4.3).
With steel frames where the material is considered homogeneous and isotropic, the
modulus of elasticity E is constant for all members, and the moment of inertia I is com-
puted for the gross cross section. In reinforced concrete, E varies with concrete strength
and magnitude of loading, while I also varies depending on the degree of cracking and the
reinforcement percentage. ACI-6.6.4.4.3 requires that the effective length factor k be deter-
mined using values of Ec in accordance with ACI-19.2.2 and I given in ACI-6.6.3.1.1 (see
Table 13.8.1).
As an alternative to using the nomograms of Figure 13.11.1, some approximate formulas
for the effective length factor k have been proposed [13.54–13.57].
For members in nonsway frames, the 1972 British Code of Standard Practice [13.54]
suggests that an upper bound for k is obtained by using the smaller of the following two
equations:
k = 0.7 + 0.05(ψ A + ψ B ) ≤ 1.0 (13.11.2)
k = 0.85 + 0.05ψ min ≤ 1.0 (13.11.3)
where ψA and ψB are the ψ values at the two ends of the member, and ψmin is the smaller of
the two values.
For members in sway frames, Furlong [13.55] proposed for members restrained at
both ends:
when ψavg < 2 (i.e., high-end restraint),
20 − ψ avg
k= 1 + ψ avg (13.11.4)
20
when ψavg ≥ 2 (i.e., moderate to low end restraint),
k = 0.9 1 + ψ avg (13.11.5)
where ψavg is the average of the ψ values at the two ends of the member.
Equations (13.11.4) and (13.11.5) give k values that are within 2% of those obtained by
the nomograms.
For members in sway frames, when hinged at one end, the British Code of Standard
Practice [13.54] proposes
k = 2.0 + 0.3ψ (13.11.6)
where ψ is the value at the restrained end.

Hu, Zhou, King, Duan, and Chen [13.56] provide additional discussion on the effective
length factor k, and Duan, King, and Chen [13.57] have proposed a partial-fraction equa-
tion for k, which, they assert, achieves “both accuracy and simplicity for design … .” Their
equation is as follows:
1 1 1
k = 1− − − (13.11.7)
5 + 9ψ A 5 + 9ψ B 10 + ψ A ψ B
493
13.13 MINIMUM ECCENTRICITY IN DESIGN 493
13.12 SECOND-O RDER ANALYSIS—A CI CODE

In a second-order analysis, equilibrium is established in the final displaced position of the
structure, ∆2u, shown by the dashed line in Fig. 13.6.1(b). Exact solutions of the problem are
often too cumbersome for practical applications in design, and thus approximate solutions
are used instead. The moment magnifier method presented in Sections 13.8 through 13.10
is one such approach.
Alternatively, an approximate solution may be obtained by conducting several iterations
of a first-order analysis wherein the moment Vu Lc is incremented by ΣPu ∆0 in each iter-
ation, or by using a bona fide second-order-analysis computer program. Today, a number
of commercial computer programs allow approximate second-order analyses of multistory
frames. The reader is cautioned, however, to study and fully understand the assumptions
and solution procedures used in these programs, which often apply to common, but not all,
situations encountered in practice. In particular, the designer is cautioned to ensure that
the software includes second-order effects from both joint lateral end translation and the
deflection between member ends.
The ACI Code allows both elastic (ACI-6.7) and inelastic (ACI-6.8) second-order analysis.
In accordance with ACI-6.7.1.1, an elastic second-order analysis must include “the influence
of axial loads, presence of cracked regions along the length of the member, and effects of load
duration,” which are considered to be satisfied when the section properties are calculated in
accordance with ACI-6.6.3.1 (see Section 13.8).
For an inelastic second-order analysis, the model must take into account material nonline-
arity, as well as the effects of “member curvature and lateral drift, duration of loads, shrinkage
and creep, and interaction with the supporting foundation” (ACI-6.8.1.1). Furthermore, the
member dimensions used in the analysis must be “… within 10% of the specified member
dimensions in construction documents…”; otherwise, the analysis must be repeated with the
updated dimensions (ACI-6.8.1.4). Also, the analysis procedure must “… have been shown
to result in prediction of strength in substantial agreement with the results of comprehensive
tests …” (ACI-6.8.1.2).
MacGregor and Hage [13.29], Wood, Beaulieu, and Adams [13.36, 13.37], Scholz [13.38,
13.39], and Furlong [13.58] have suggested various approaches to conducting second-order
analysis of reinforced concrete frames.
13.13 MINIMUM ECCENTRICITY IN DESIGN

Nonsway Frames
When computations indicate only a small moment to be acting on a member such that the
eccentricity Mu /Pu is less than (0.6 + 0.03h) in., the primary moment Mm in Eq. (13.9.1)3 is to
be calculated as (ACI-6.6.4.5.4)
M m ,min = Pu (0.6 + 0.03h) (13.13.1)4
In such cases, ACI-6.6.4.5.4 requires Cm to be taken as 1.0 or computed from Eq. (13.9.6)
based on the ratio of the computed first-order end moments. Note that the minimum
3 As noted earlier, the ACI Code uses M2 instead of Mm in Eq. (13.9.1) [ACI Formula Eq. (6.6.4.5.1)]. Mm in Eq. (13.9.1) is
the maximum moment occurring in the member in general, it may occur at either end or, if there is transverse loading, within the
span. Mm is used here to distinguish it from the larger end moment, M2, in members with end moments only (see Section 13.5).
4 For SI with 15 and h in mm, ACI 318M-14 gives Eq. (13.13.1) as
M m,min = Pu (15 + 0.03h)
49
moment, when it governs, is to be applied about each axis separately, not as a case of biax-
ial bending.
Sway Frames
Prior to the 1995 ACI Code, the minimum eccentricity requirement applied to members in
both nonsway (braced) and sway (unbraced) frames. However, ACI-6.6.4.5.4 applies only
to members in nonsway (braced) frames. The ACI Commentary makes no reference to this
change. In the design of sway (unbraced) frames, members of such frames only rarely have
end moments smaller than that given by Eq. (13.13.1). The writers believe, however, that
should an analysis give end moments less than that given by Eq. (13.13.1), the moment
obtained from that equation should be used as the largest primary moment.
When the minimum eccentricity requirements control, the provisions are intended to be
applied to bending about only one axis at a time, not as a case of biaxial bending.
13.14 BIAXIAL BENDING AND AXIAL COMPRESSION

With the advent of faster computers and the structural analysis software available today,
three-dimensional, elastic second-order analyses have become more common in practice.
Such analyses, if properly done, can adequately account for second-order effects in both
directions of bending.
Alternatively, an estimate of the magnified moments may be obtained by considering the
moment about each axis to be magnified by the factor δ computed from the restraint condi-
tions for that axis. Note that, in general, the effective length factor k, the stiffness factor EI,
and Cm will differ for each bending axis. Also, note that the member may be considered as
part of a braced system in one direction and unbraced in the other. Additional discussion
and design recommendations on biaxial bending of slender members may be found in Refs.
13.49 and 13.59, as well as Refs. 13.55 through 13.58, and 13.62.
13.15 ACI CODE—S LENDERNESS RATIO LIMITATIONS

Although the slenderness ratio is never zero in actual structures, there are certain limits
for kLu/r below which slenderness effects may reasonably be neglected. The ACI Code
provisions are based on the assumption that a moment increase due to slenderness effects
up to 5% can be tolerated without the designer having to consider these effects; thus a sig-
nificant number of ordinary columns can be designed considering only the provisions of
Chapter 10.
ACI-ASCE Committee 441 surveyed typical reinforced concrete buildings to determine
the normal range of variables found in columns of such buildings, as reported by Mac
Gregor, Breen, and Pfrang [13.32]. A great variety of buildings were studied, including
towers (braced frames) as high as 33 stories and an unbraced frame 20 stories high. The
total number of columns exceeded 20,000. The following results were reported [13.32].
For braced frames, 98% of the columns had L /h less than 12.5 (L /r ≈ 42) and e/h less than
0.64. For unbraced frames, 98% of the columns had L /h less than 18 (L /r ≈ 60) and e/h less
than 0.84. Furthermore, it was found that the practical upper limit on the slenderness ratio
kL /r is about 70 in building columns. In general, these limits provide a guide to the range
of variables that are used in any approximate method. Taking the idea that a 5% increase in
moment due to slenderness effects is acceptable, the effects of slenderness are permitted to
be neglected under the following conditions.
495
13.17 EXAMPLES 495
Braced (Nonsway) Frame Members

ACI-6.2.5 permits neglect of slenderness effects when
kLu M
≤ 34 + 12 1ns (13.15.1)
r M 2 ns
where M1ns is the smaller and M2ns the larger of end moments on the member, and [34 +
12M1ns /M2ns] ≤ 40. The subscript ns refers to nonsway (i.e., braced).
For single-curvature deformation, the ratio M1ns /M2ns is negative, while for double curva-
ture the ratio is positive. When the member is subject to large transverse loading (other than
end moments alone), the ratio M1ns /M2ns should probably be taken as –1.
Unbraced (Sway) Frame Members

ACI-6.2.5 permits neglect of slenderness when
kLu
< 22 (13.15.2)
r
Upper Limit on Second-Order Moments

When the limits given by Eqs. (13.15.1) or (13.15.2) are not satisfied, second-order
effects must be considered in accordance with the moment magnifier method of ACI-6.6.4
(Section 13.8) or by conducting second-order analyses in accordance with ACI-6.7 or ACI-
6.8 (Section 13.12). In no case, however, may the magnified moments exceed 1.4 times the
computed primary or first-order moments (ACI-6.2.6). In such a case, the probability of
stability failure is considered to be too high, so such a frame is not permitted and the struc-
ture must be redesigned.
13.16 AMPLIFICATION OF MOMENTS IN BEAMS

When a second-order analysis is conducted, the magnified column moments are read-
ily balanced by those in the beams to satisfy equilibrium. When the moment magnifier
method is used, however, the designer must recognize that the computed magnified
moments acting at the ends of the column must be carried by the beams in a manner
ensuring that equilibrium of the joint is maintained, as illustrated in Fig. 13.10.2(d).
For this reason, ACI-6.6.4.6.3 requires that flexural members (beams) “be designed for
the total magnified end moments of the columns at the joint.”
13.17 EXAMPLES
The preceding sections of this chapter discussed basic concepts underlying the ACI Code
provisions relating to slenderness effects on columns. To facilitate easy reference to the pro-
cedures and formulas, most of which appear in the ACI Code or Commentary, Table 13.17.1
provides a summary of the most common formulas needed for solving practical problems.
Examples 13.17.1 through 13.17.5 illustrate the use of these formulas.
496
TABLE 13.17.1 SLENDERNESS EFFECTS ON COLUMNS—SUMMARY

OF USEFUL FORMULAS
1. Section Properties to Be Used in First-Order Analysis

Ec = 57, 000 fc′ for normal-weight concrete
Ec = wc1.5 33 fc′ for wc between 90 and 155 lb/cu ft
Moment of inertia [to be divided by (1 + βds) for sustained lateral loads]
= 0.35Ig for beams and cracked walls
= 0.70Ig for columns and uncracked walls
= 0.25Ig for flat plates and flat slabs
Area = 1.0Ag
2. Radius of Gyration
r = 0.30h for rectangular sections
r = 0.25h for circular sections
3. Nonsway (Braced) Frames
(a) Slenderness effects may be ignored if
kLu M
≤ 34 + 12 1ns
r M 2 ns
M1ns /M2ns is negative for single curvature, and the kLu /r limit is not to be greater than 40.
ay be assumed as braced if second-order end moments ≤ [1.05 × (first-order end
(b) M
moments)], or if [Q = ∑ Pu ∆ 0 / (Vu Lc )] ≤ 0.05
(c) Moment magnifier δns
Mmax = δnsMm
Cm π 2 ( EI )eff
δ ns = ≥ 1.0; Pc =
Pu (knonsway Lu )2
1−
0.75Pc
M1ns M
Cm = 0.6 − 0.4 , for members with end moments only and 1ns is negative
M 2 ns M 2 ns
for single curvature
Cm = 1.0 for members with transverse loads
(EI)eff computed from Eqs. (13.9.8), (13.9.9), or (13.9.10)
Mm,min = Pu(0.6 + 0.03h)
4. Sway (Unbraced) Frames

kLu
(a) Slenderness effects may be ignored if ≤ 22
r
(b) Moment magnifier δs
M12 nd = M1ns + δ s M1s
M 22 nd = M1ns + δ s M1s
1 ΣP ∆ 1
δs = and Q = u 0 for Q ≤ or δ s ≤ 1.5 only
1−Q Vu Lc 3
or
1 π 2 ( EI )eff
δs = and Pc =
ΣPu (ksway Lu )2
1−
0.75ΣPc
( EI )eff computed from Eqs. (13.9.8), (13.9.9), or (13.9.10) with β ds substituted for β dns
497
13.17 EXAMPLES 497
EXAMPLE 13.17.1
Use the moment magnifier method to compute the required strength in flexure, Mu, and
axial load, Pu, for the columns and the beam of the frame shown in Fig. 13.17.1. The frame
is one of several identical frames in the structure and it supports uniformly distributed as
well as concentrated service-level dead and live loads, as shown. In addition, a strength-
level wind load is assumed to be acting at the top of the frame. Assume that only the dead
loads are sustained. Preliminary dimensions for the columns and the beam are as shown.
The beam is restrained against out-of-plane movement by prefabricated elements that are
simply supported on the top of the beam. There are no walls or any other bracing elements
in the building in the direction of the frame. Use fc′ = 5, 000 psi and Grade 60 steel.
PD = 160k
PD = 130k PD = 100k
PL = 70k
PL = 50k PL = 50k
6.25’
wD = 1 k/ft
wL = 0.8 k/ft
W = 25k
16” × 26” beam
20” × 20”
12’ columns
25’
Figure 13.17.1 Unbraced frame of Example 13.17.1.
SOLUTION
(a) Slenderness ratio limits. The first step is to determine whether second-order effects
need to be considered. Because the frame is permitted to sway (i.e., there are no
walls or other bracing elements), the columns are not braced against sidesway. In
such a case, ACI-6.2.5 permits neglect of slenderness effects when
kLu
≤ 22 (13.15.2)
r
The unsupported height, Lu, of both columns is
 1
Lu = 12 − 26   = 9.83 ft.
 12 
The effective length factor, k, may be computed using the alignment charts shown in
Fig. 13.11.1. This requires the computation of the end restraint factors, ψ, at the top and
bottom of the columns from Eq. (13.11.1). For the purpose of computing the end res-
traint factors, the moments of inertia given in ACI-6.6.3.1.1 (see Table 13.8.1) may be
used. Therefore,
16(26)3
( EI )beam = Ec (0.35I g ) = Ec (0.35) = 8202 Ec kip-in.2
12
20 4
( EI )col = Ec (0.70 I g ) = Ec (0.7) = 9333Ec kip-in.2
12
(Continued)
498
End restraint factors,
∑ EI /L for cols (9333)Ec /12

ψ A (top) = = = 2.4
∑ EI /L for beams (8202)Ec / 255
Since the concrete strength is assumed to be the same for the columns and the beam in
this example, the value of the modulus of elasticity is not needed for the computation
of the end restraint factor. Also note that the length L for the columns and the beam are
taken as the center-to-center distance for this calculation.
At the bottom, the columns are pinned. Thus, ψB (bottom), will be taken equal to the
upper practical value of 10.
Entering the alignment chart for an unbraced frame (i.e., sidesway not prevented) of
Fig. 13.11.1(b), the effective length factor is estimated as
ksway ≈ 2.2
Taking the radius of gyration, r, as 0.3 times the depth of the column per ACI-6.2.5.1(b),
the slenderness ratio for the columns is
ksway Lu 2.2(9.83)12
= = 43.3 > 22
r 0.3(20)
Therefore, the columns are slender and second-order effects must be considered.
(b) The frame is subjected to dead, live, and wind loads. Seven load combinations must
be considered in design (Section 2.7). In this example, however, only the following
three load combinations will be considered
1. U = 1.4D [2.7.1]
2. U = 1.2D + 1.6L [2.7.2]
3. U = 1.2D + 1.0W + 0.5L [2.7.4]
where the load factor on the live load L has been reduced to 0.5 in Eq. (2.7.4) as
permitted for areas that are not used as garages, places of public assembly, or areas
that do not sustain live loads greater than 100 lbs/sq ft.
1. Load combination: U = 1.4D.
(i) Nonsway or Sway Frame Analysis
For a first-order, factored analysis, the ACI Code requires that the section properties
given in ACI-6.6.3.1.1 and the modulus of elasticity, Ec, given by ACI-19.2.2, be used
(see Table 13.8.1). Thus, for fc′ = 5000 psi,
1
Ec = 57, 000 5000 = 4030 ksi
1000
and from part (a)
( EI )beam = 8202 Ec = 8202(4030) = 33, 050, 000 kip-in.2

( EI )col = 9333Ec = 9333(4030) = 37, 610, 000 kip-in.2
In ACI 318-14, columns are permitted to be analyzed as nonsway when (a) the increase
in column end moments due to second-order effects does not exceed 5% of the first-
order end moments or (b) the story stability index, Q, is less than or equal to 0.05 (ACI-
6.6.4.3). At this point, the second-order moments have not been computed; thus the story
stability index, Q, will be used to check this requirement.
(Continued)
49
13.17 EXAMPLES 499
The stability index, Q, is computed using Eq. (13.8.1)

∑ Pu ∆ 0
Q= [13.8.1]
Vu Lc
where
Δ0 = relative lateral deflection between the top and bottom of the story in question due
to Vu, computed using a first-order elastic frame analysis and section properties
of ACI-6.6.3.1.1.
Lc = length of the compression member in question, measured from center to center of
the joints in the frame
It is noted that the stability index, Q, is a function of the ratio Δ0 /Vu, in other words, the
lateral flexibility of the frame. Therefore, Q can be computed using any lateral load that
produces the shear, Vu, and the corresponding story deflection, Δ0. In general, to com-
pute Q in a multistory frame, an arbitrary lateral load is applied at the top of the story
being considered while the bottom of the story is restrained from lateral movement. No
other loads are applied for the computation of Q. In this example, a lateral load of 10
kips will be applied at the top of the columns.
The current ACI Code is unclear as to how to consider the effects of sustained loads
for computing the stability index, Q. Because a lateral load analysis is required to com-
pute the story deflection, Δ0, it may be assumed that ACI-6.6.3.1.1 (or ACI-6.6.3.1.2)
would apply. ACI-6.6.3.1.1 requires that the moment of inertia of compression members
be divided by (1 + βds) if sustained lateral loads, such as those induced by earth pres-
sure, are present. In this example, there are no sustained lateral loads according to this
definition, although the sustained dead load does induce sway (asymmetric loading) and
could potentially induce lateral instability of the frame. In past editions of the ACI Code
(2005 and earlier editions), it was required that the moments of inertia of all members
be reduced for stability checks involving the stability index, Q. In this example, the stiff-
ness parameter EI of all members will be divided by (1 + βds) to allow consideration of
the possible instability and moment magnification due to sway of the frame under the
sustained dead loads. This approach is conservative, but reasonable. For this load combi-
nation, all the load inducing sway is sustained (i.e., βds = βdns =1), and thus, the EI values
for the beam and the columns computed earlier will all be divided by (1 + βds) = 2. Thus,
33, 050, 000
( EI )beam = = 16, 525, 000 kips-in.2
2

37, 610, 000
( EI )col = = 18, 805, 000 kip-in.2
2
Analysis of the frame using these properties and a lateral load of 10 kips yielded a rela-
tive lateral deflection Δ0 of 0.58 in.
The total vertical factored load under this load combination is

∑ Pu = 1.4[130 + 100 + 160 + 1(25)] = 581 kips
Thus,
∑ Pu ∆ 0 581(0.58)
Q= = = 0.23 > 0.05
Vu Lc 10(12)12
Since the stability index, Q, exceeds 0.05, the story cannot be treated as nonsway and the
frame must be treated as a sway frame.
(Continued)
50
(ii) Computation of Nonsway Moments, Mns

The nonsway moments are computed by conducting a first-order analysis of the frame
with lateral end translation prevented [Fig. 13.10.1(b)] and subject to the factored loads
acting in this load combination (i.e., the factored dead load). The nonsway moment dia-
grams are shown in Fig. 13.17.2. Because of the asymmetry of the loading, a horizontal
reaction of 14.7 kips is calculated at the top.
401ft-k 224ft-k
14.7 k
382ft-k
left right
Pns = 312 k Pns = 269 k
Figure 13.17.2 Nonsway moments, Mns, for load combination U = 1.4D of Example 13.17.1.
(iii) Computation of Sway Moments, Ms

To compute the sway moments, a first-order analysis of the unbraced (sway) frame is
computed by applying a lateral load equal to the horizontal reaction [Fig. 13.10.1(c)] of
14.7 kips computed in (ii) above. No other loads are applied. The calculated sway bend-
ing moment diagrams are shown in Fig. 13.17.3.
88ft-k 88ft-k
left right
Ps = 7.1k Ps = 7.1k
Figure 13.17.3 Sway moments, Ms, for load combination U = 1.4D of Example 13.17.1.
(iv) Nonsway Magnifiers, δns

Because the columns do not have transverse loading and because the moment at the base
is zero, the maximum moment will be equal to the first-order moment at the top of the
columns. Thus, there is no need to compute the nonsway magnification factor. In this
(Continued)
501
13.17 EXAMPLES 501
example, however, the nonsway magnifier will be computed anyway to illustrate the
procedure. The nonsway magnification factor, δns, is given by ACI-6.6.4.5.2 as
Cm
δ ns = ≥ 1.0 [13.9.4]
Pu
1−
0.75Pc
This factor is a member factor and must be computed individually for each column.
Left Column
• Factor Cm
The column does not have transverse loading and is subjected to end moments only.
Accordingly, the factor Cm is computed from Eq. (13.9.6)
M1ns
Cm = 0.6 − 0.4 [13.9.6]
M 2 ns
where M1ns is the smaller of the nonsway, first-order end moments. Because the col-
umns are pinned at the base, M1ns is zero and Cm is computed as 0.6 in this case.
• Factored axial load, Pu
The factored axial load is the total factored axial load acting on the column for the
load combination being considered. Accordingly,

Puleft = Pnsleft + Psleft = 312 − 7.1 = 305 kips
• Critical buckling load, Pc
The critical buckling load is computed in accordance with ACI-6.6.4.4.2 as
π 2 ( EI )eff
Pc = [13.9.7]
(knonsway Lu )2
where the stiffness parameter, (EI )eff, may be computed using any of Eqs. (13.9.8),
(13.9.9), or (13.910). Since the amount of reinforcement has not been computed yet,
Eq. (13.9.9) cannot be used at this stage. Here, (EI )eff will be computed using Eq.
(13.9.8) to illustrate the procedure. Thus,
0.4 Ec I g 0.4(4030)13, 333
( EI )eff = = = 10, 750, 000 kip-in.2
(1 + β dns ) (1 + 1)
where βdns was taken equal to 1 since all of the load is sustained for the load combina
tion being considered.
Therefore, the critical buckling load Pc for the left column is
π 2 ( EI )eff π 2 (10, 750, 000)
Pcleft = 2
= = 9000 kips
(knonsway Lu ) [0.92(9.83)12]2
where the effective length factor, knonsway, was computed using the end restraint factors
ψA (top) and ψB (bottom) computed in part (a), and the alignment chart for braced
(sidesway prevented) frames of Fig. 13.11.1(a).
Thus,
0.6
δ left
ns = = 0.63 < 1
305
1−
0.75(9000) (Continued)
502
Therefore, δ left
ns = 1
and the maximum nonsway moment for the left column occurs at the
end and is equal to the first-order moment at the top.
Right Column
The procedure for the right column is the same as that used for the left column. Because
both columns have the same dimensions, have no transverse loading, have the same
shape of the bending moment diagram, and the same support conditions at the ends,
the parameters involved in the computation of the nonsway magnifier are all the same
except for the level of axial load level acting on the column. The following are the
parameter values for the right column.
• Factor Cm

Cm = 0.6 (column with end moments only and M1ns = 0)
• Factored axial load Pu:

Puright = Pnsright + Psright = 269 + 7.1 = 276 kips
• Critical buckling load Pc:

Pcright = Pcleft = 9000 kips
Thus,
0.6
δ right
ns = = 0.63 < 1
276
1−
0.75(9000)
Therefore, the maximum moment also occurs at the end (top) for the right column.
(v) Sway Moment Magnifier δs
The sway moment magnifier δs is computed in accordance to ACI-6.6.4.6.2 using Eq.
(13.10.2), (13.10.4), or a second-order elastic analysis. It is noted that the sway moment
magnifier is a story amplification factor and it is the same for all of the columns in a
given story.
• Q method [ACI-6.6.4.6.2(a)]
In this method, the sway moment magnifier is computed by using Eq. (13.10.2):
1
δs = ≥ 1.0 [13.10.2]
1−Q
Using the stability index Q of 0.23 computed above [see part 1(i)],
1
δs = = 1.30
1 − 0.23
This value is rather large, but less than 1.5 and thus, it is acceptable per ACI-6.6.4.6.2.
• Sum of P method [ACI-6.6.4.6.2(b)]
In this approach, the sway moment magnifier is computed using Eq. (13.10.4) as
1
δs = ≥ 1.0 [13.10.4]
∑ Pu
1−
0.75 ∑ Pc
where
π 2 ( EI )eff
Pc = [13.10.5]
(ksway Lu )2
(Continued)
503
13.17 EXAMPLES 503
where the stiffness parameter (EI )eff may be computed using Eq. (13.9.8), (13.9.9), or
(13.9.10) with βdns substituted for βds. Here, (EI )eff will be computed using Eqs. (13.9.8).
Since both columns have the same cross section, (EI )eff will be the same. Thus, using
Eq. (13.9.8),
0.4 Ec I g 0.4(4030)13, 333
( EI )eff = = = 10, 750, 000 kip-in.2
(1 + β ds ) (1 + 1)
where βds was taken equal to 1 because the sway moments are all produced by the sus-
tained dead load. Therefore, the critical buckling load Pc is for both columns
π 2 ( EI )eff π 2 (10, 750, 000)

Pc = = = 1575 kiips
(ksway Lu )2 [2.2(9.83)12]2
Therefore, ΣPc = 1575 + 1575 = 3150 kips.

The total vertical factored axial load was computed earlier in part 1(i) as ΣPu = 581 kips.
Thus,
1
δs = = 1.33
581
1−
0.75(3150)
It is seen that in this case, the Q method and the sum of P method yield very similar val-
ues for the sway moment magnifier δs.
(vi) Combination of Nonsway and Sway Moments
The total second-order moment is computed as the summation of the nonsway and sway
moments at the column ends using Eqs. (13.10.7) and (13.10.8). The moment at the
bottom of both columns is zero (pin). At the top, the maximum second-order moment
is computed as
M 22 nd = M 2 ns + δ s M 2 s [13.10.8]
The moment magnifier obtained with the Q method (i.e., δs = 1.30) will be used in the
following calculations. A comparison of the moments obtained with the two values for
the moment magnifier δs is presented later (see Table 13.17.2).
Left Column

M 22 nd = 401 + 1.3( −88) = 287 ft-kips
For this column, the nonsway and sway moments counteract each other. Therefore, the
second-order moment thus computed is smaller than the first-order moment. It is non-
conservative and inappropriate to use the value computed above. For design purposes,
the larger, first-order moment should be used, that is,

M 2 = 401 + ( −88) = 313 ft - kips
Thus, for this load combination, the column should be designed for
Pu = 305 kips

and Mu = 313 ft-kips
Right Column
For the right column, the nonsway and sway moments are additive. Thus,

M 22 nd = 224 + 1.3 (88) = 338 ft-kips
(Continued)
504
and the column should be designed for
Pu = 276 kips

2. Load combination: U = 1.2D +1.6L.
The approach to computing the second-order moments is the same as that for the preced-
ing load combination; thus only the differences in the calculations are described next.
(i) Nonsway or Sway Frame Analysis
For this load combination, the stability index Q is computed using Eq. (13.8.1) as above
∑ Pu ∆ 0
Q= [13.8.1]
Vu Lc
where that lateral deflection of the frame will be computed with the stiffness parame-
ter EI divided by (1 + β ds ) for all members to account for the effects of sustained loads.
Since this load combination includes gravity loading only, the amount of sustained load
inducing sway will be in direct proportion of the ratio between total sustained vertical
load and the total vertical load (i.e., βdns = βds). The total vertical factored load under this
load combination is
∑ Pu = 1.2[130 + 100 + 160 + 1(25)]
+1.6[50 + 50 + 70 + 0.8(25)]
= 498 + 304
∑ Pu = 802 kips
The total vertical factored sustained load (i.e., the dead load) under this load
combination is

∑ Psustained = 1.2[130 + 100 + 160 + 1(25)] = 498 kips
Thus,
498
βdns = = 0.62 [ = β ds in this case]
802
and (1 + β ds ) = 1.62. The revised EI values are
33, 050, 000

( EI )beam = = 20, 400, 000 kip-in.2
1.62

37, 610, 000
( EI )col = = 23, 220, 000 kip-in.2
1.62
Analysis of the frame using these properties and a lateral load of 10 kips yields a relative
lateral deflection, ∆0, of 0.47 in. [= 0.58(1.62)/2]. Thus,
∑ Pu ∆ 0 802(0.47)
Q= = = 0.26 > 0.05
Vu Lc 10(12)12
The stability index, Q, exceeds 0.05, and thus the story must be treated as a sway frame.
(Continued)
50
13.17 EXAMPLES 505

The nonsway moments computed from first-order analysis of the frame subject to all the
loads acting in this load combination and restrained against lateral movement are shown
in Fig. 13.17.4. Because of the asymmetry of the loading, a horizontal reaction of 20.9
kips is calculated at the top. The axial forces acting on each column are also shown.
591ft-k 340 ft-k

20.9 k
554ft-k
left right
Pns = 427k Pns = 375k
Figure 13.17.4 Nonsway moments, Mns, for load combination U = (1.2D + 1.6L) of Example 13.17.1.
(iii) Computation of Sway Moments, Ms

To compute the sway moments, the unbraced (sway) frame is subjected to a first-order
analysis by applying a lateral load equal to the horizontal reaction of 20.9 kips computed
in (ii) above. No other loads are applied. The calculated sway bending moment diagrams
are shown in Fig. 13.17.5.
126 ft-k 126 ft-k
left right
Ps = 10.1 k Ps = 10.1 k
Figure 13.17.5 Sway moments, Ms, for load combination U = (1.2D + 1.6L) of Example 13.17.1.

As shown for the preceding load combination, the nonsway magnifier will be less than
one because the columns do not have transverse loading and because the moment at the
base is zero. Thus, the nonsway moments will occur at the top of the columns and will
be equal to the first-order moments shown in Fig. 13.17.4.
(Continued)
506
(v) Sway Moment Magnifier δs

The sway moment magnifier δs is computed in the same manner as before. The main
difference in the computed values is due to the different β dns to account for the effects
of sustained loads.
• Q method [ACI-6.6.4.6.2(a)]
Using the stability index Q of 0.26 computed in part 2(i)
1
δs = = 1.35
1 − 0.26
This value is also rather large, but it is less than 1.5, thus it is acceptable per
ACI-6.6.4.6.2.
• Sum of P Method [ACI-6.6.4.6.2(b)]
In this approach, the sway moment magnifier is computed using Eq. (13.10.4) as
1
δs = ≥ 1.0 [13.10.4]
∑ Pu
1−
0.75 ∑ Pc
where
π 2 ( EI )eff
Pc = [13.10.5]
(ksway Lu )2
Using Eq. (13.9.8) for (EI )eff
0.4 Ec I g 0.4(4030)13, 333

( EI ) eff = = = 13, 270, 000 kip-in.2
(1 + β ds ) (1 + 0.62)
Therefore, the critical buckling load Pc is for both columns
π 2 ( EI )eff π 2 (13, 270, 000)

Pc = = = 1945 kiips
(ksway Lu )2 [2.2(9.83)12]2
and ΣPc = 2(1945) = 3890 kips.
The total vertical factored axial load was computed earlier in part 2(i) as ΣPu = 802 kips. Thus,
1
δs = = 1.38
802
1−
0.75(3890)
Again, both the Q method and the sum of P method yield similar values.
The moment magnifier obtained with the Q method (i.e., δs = 1.35) is used in the follow-
ing calculations. At the top, the maximum second-order moment is computed as
M 22 nd = M 2 ns + δ s M 2 s [13.10.8]
Left Column
For this column, the nonsway and sway moments counteract each other. For design pur-
poses, the larger, first-order moment without the sway magnifier should be used, that is,

M 2 = 591 + ( −126) = 465 ft-kips
(Continued)
507
13.17 EXAMPLES 507
Pu = 427 − 10.1 = 417 kips

Right Column

M 22 nd = 340 + 1.35(126) = 510 ft-kips

Pu = 375 + 10.1 = 385 kips


3. Load combination: U = 1.2D +1.0W + 0.5L.

For this load combination, two cases need to be considered: wind load acting to the right
and wind load acting to the left. The stability index, Q, and the moment magnifiers are
a function of the stiffness parameter EI and the applied loads for the load combination
being considered; both are the same and are independent of the wind direction. The
procedure will be illustrated in detail for wind acting to the right; the approach for wind
acting to the left is the same.
(i) Nonsway or Sway Frame Analysis.
Most of the load-inducing sway of the frame in this load combination is due to wind,
which is not sustained. Whether the member stiffness should be reduced to include the
sway component due the sustained dead load in this load combination is somewhat a
matter of opinion. In this example, the stiffness parameter EI is divided by (1 + βds) to
account for the sway component due to the sustained dead loads. This approach is con-
servative. The horizontal reaction at the top of the frame obtained from a nonsway anal-
ysis under 1.2D alone is computed as 12.6 kips. This will be considered as sustained. On
the other hand, the horizontal reaction obtained for the load combination being consid-
ered (1.2D +1.0W + 0.5L) is computed as 40 kips (see Fig. 13.17.6). Thus,
12.6
β ds = = 0.32
40
and (1 + βds) = 1.32. The revised EI values are
33, 050, 000

( EI )beam = = 25, 040, 000 kip-in.2
1.32
37, 610, 000
( EI )col = = 28, 500, 000 kip-in.2
1.32
Analysis of the frame using these properties and a lateral load of 40 kips yields a relative
lateral deflection, Δ0, of 1.52 in. [or (0.58(1.32)/2) × 4]
The total vertical factored load under this load combination is
∑ Pu = 1.2[130 + 100 + 160 + 1(25)]

+ 0.5[50 + 50 + 70 + 0.8(25)]
= 498 + 95
∑ Pu = 593 kips
(Continued)
508
Thus,
∑ Pu ∆ 0 593(1.52)
Q= = = 0.16 > 0.05
Vu Lc 40(12)12
The stability index, Q, exceeds 0.05, and thus the story must be treated as a sway frame,
as was expected for this load combination including wind load.
The nonsway moments computed from first-order analysis of the frame subject to all the
loads acting in this load combination and restrained against lateral translation are shown
in Fig. 13.17.6. The axial forces acting on each column, as well as the lateral restraining
force, are also shown.
418 ft-k 238 ft-k

40.0 k
401 ft-k
left right
Pns = 317 k Pns = 276 k
Figure 13.17.6 Nonsway moments, Mns, for load combination U = (1.2D + 1.0W + 0.5L) of
Example 13.17.1.
(iii) Computation of Sway Moments Ms

The sway moments obtained from a first-order analysis of the unbraced (sway) frame
under a lateral load equal to the horizontal reaction of 40.0 kips are shown in Fig. 13.17.7.
240 ft-k 240 ft-k
left right
Ps = 19.2 k Ps = 19.2 k
Figure 13.17.7 Sway moments, Ms, for load combination U = (1.2D + 1.0W +0.5L) of
Example 13.17.1.
(Continued)
509
13.17 EXAMPLES 509

As before, it can be shown that the nonsway magnifier will be less than one for this load
combination. Thus, the nonsway moments will occur at the top of the columns and will
be equal to the first-order moments shown in Fig. 13.17.6.
(v) Sway Moment Magnifier, δs
• Q method [ACI-6.6.4.6.2(a)]
Using the stability index Q of 0.16 computed in part 3(i)
1
δs = = 1.19
1 − 0.16
which is less than 1.5 and thus, acceptable per ACI-6.6.4.6.2.
• Sum of P method [ACI-6.6.4.6.2(b)]
Using Eq. (13.9.8) for (EI )eff
0.4 Ec I g 0.4(4030)13, 333
( EI )eff = = = 16, 280, 000 kip-in.2
(1 + β ds ) (1 + 0.32)
Therefore, the critical buckling load Pc is for both columns
π 2 ( EI )eff π 2 (16, 280, 000)

Pc = = = 2390 kiips
(ksway Lu )2 [2.2(9.83)12]2
Thus, ΣPc = 2(2390) = 4780 kips.
From part 3(i), the total vertical factored axial load under this load combination is
ΣPu = 593 kips.
Thus,
1
δs = = 1.20
593
1−
0.75(4780)

The moment magnifier obtained with the Q method (i.e., δs = 1.19) will be used in the
following calculations.
Left Column
The nonsway and sway moments counteract each other for this column. For design
purposes, the larger, first-order moment without the sway magnifier should be used,
that is,

M 2 = 418 + ( −240) = 178 ft-kips

Pu = 317 − 19.2 = 298 kips

Right Column

M 22 nd = 238 + 1.19(240) = 524 ft-kips
(Continued)
510
Pu = 276 + 19.2 = 295 kips

Mu = 524 ft-kips
(c) Summary of Results. Table 13.17.2 summarizes the first-order and second-order
moments computed at the top of the columns using the moment magnifier method
for this example problem. The second- to-
last column in the table shows the
moments obtained from an elastic second-order analysis of the frame that included
moment magnification due to deflections between member ends and due to lateral
end translation. The section properties used in these analyses were the same as those
used to compute the stability index Q for each load combination and accounted
for the effect of sustained loads using the same procedure to compute the moment
magnifiers.
It can be seen that the Q method and Sum of P method yield very similar moments,
as was expected since the calculated magnifiers were very close. It can also be seen that
the column moments computed by using the moment magnifier approach, as presented
here, are in very good agreement with those computed from an elastic second-order
analysis.
The last column in Table 13.17.2 shows the ratio between the second-order and first-
order moments computed for the columns in each load combination: the second-order
moments are at most 1.1 times the first-order moments, and thus the upper limit of 1.4
per the requirements of ACI-6.2.6 is satisfied (see Section 13.15).
(d) Required Strengths
Columns
For each factored load combination, the maximum moment, Mu, occurring concurrently
with the axial force, Pu, must be investigated. In no case, however, should Mu be taken
less than the first-order moment computed for that load combination.
The ACI Code does not require columns in sway frames to be designed for minimum
eccentricity (see Section 13.13), but it is instructive to compare the calculated moments
with the minimum moment of ACI-6.6.4.5.4:
M m ,min = Pu (0.6 + 0.03h) [13.13.1]
In this example, h = 20 in., and thus
M m ,min = Pu [0.6 + 0.03(20)] = 1.2 Pu [ kip-in.]
The maximum computed axial force is 417 kips on the left column under the load com-
bination U = 1.2D + 1.6L. In this case,
M m,min = 1.2(417) = 500 kip-in

= 42 ft-kips
which is much less than any of the first-order moments computed for either the left or
the right column and thus does not govern.
Though not required in this example, the design interaction diagram for a 20 by
20 in. column reinforced with 4–#11 corner bars and 8–#10 side bars (two in each
face) is shown in Fig. 13.17.8. It can be seen that the selected reinforcement satisfies
the required strengths (Pu, Mu) [boldface in Table 13.17.2] for the load combinations
considered.
(Continued)
51
a–c
TABLE 13.17.2 FIRST-AND SECOND-ORDER MOMENTS FOR THE COLUMNS OF EXAMPLE 13.17.1

Second-Order Moments (ft-kips)
Moment Magnifier 2nd Order Mu2 nd

First-Order Moments (ft-kips) Method, Mu2nd Elastic Mu1st
Nonsway Sway Mtot Sway Magnifiers, δs Mns + δs Ms

Load Combination Pu(kips) Mns Ms Mns + Ms (1) (2) (3) (4)
1.4D Left column 305 401 −88 313 1.30 1.33 287 284 285 0.91
13.17 EXAMPLES
Right column 276 224 88 312 338 341 339 1.09

ΣPu = 581
1.2D+1.6L Left column 417 591 −126 465 1.35 1.38 421 417 421 0.91
Right column 385 340 126 466 510 514 509 1.09
ΣPu = 802
1.2D+0.5L +1.0W Left column 298 418 −240 178 1.19 1.20 132 130 133 0.75
Right column 295 238 240 478 524 526 523 1.09
ΣPu = 593
1.2D+0.5L − 1.0W Left column 322 418 60 478 1.19 1.20 489 490 488 1.02
Right column 271 238 −60 178 167 166 167 0.94
ΣPu = 593
a Columns (1), (3), magnifier and corresponding second-order moment computed from Eq. (13.10.2) [Q method]
b Columns (2), (4), magnifier and corresponding second-order moment computed from Eq. (13.10.4) and (EI)eff from Eq. (13.9.8) [Sum of P method]
c Boldface type indicates required strengths (Pu and Mu) for the load combination considered based on second-order moments computed using the Q method
(Continued)
511
512
Beam
For the frame of this example, the moments at the beam ends must be in equilibrium
with the first-and second-order moments at the top of the columns. For this reason,
ACI-6.6.4.6.3 requires that flexural members in sway frames be designed for the total
magnified moment of the columns at the joint. Therefore, the beam would have to be
designed for a negative moment of 489 ft-kip at the left end and a negative moment of
524 ft-kip at the right end.
φPn, Pu (kips)
1800
Required strengths: 4 – #11 (corners)
left column 8 – #10
1600 right column ρg = 4.1%
1400
1200
fc’ = 5000 psi

1000 fy = 60,000 ksi
800
600
400
200
0 100 200 300 400 500 600

φMn, Mu (ft-kips)
Figure 13.17.8 Design interaction diagram for the columns of Example 13.17.1.
EXAMPLE 13.17.2
Determine the adequacy of the interior top floor column (column A) of the braced frame
of Fig. 13.17.9. The column is 10 × 10 with 4–#8 and 4–#9 bars ( fy = 60 ksi and fc′ = 3 ksi)
and is to carry a service axial compression of 106 kips live load and 39 kips dead load.
The bending moments that may act in combination with the axial load have been com-
puted and found to be negligible. If the member is not adequate, revise the design so that
it satisfies the moment magnifier method of the ACI Code.
Wall
12 × 20 12 × 20 A 10”
12 × 12 10 × 10
12 × 12 10 × 10
12 × 20 12 × 20
3 @ 12’–0”
1.5 in. clear

10”
cover to tie
#3 tie 4 – #9 (corner bars)

4 – #8
4 @ 24’–0”
(a) (b)
Figure 13.17.9 Frame of Example 13.17.2. (Continued)

513
13.17 EXAMPLES 513
SOLUTION
(a) Determine the slenderness ratio. Unless an evaluation of end restraint is made, ACI-
6.6.4.4.3 permits taking the effective length factor k for a braced frame equal to 1.0.
The radius of gyration may be taken as 0.3h according to ACI-6.2.5.1(b). The unsup-
ported height Lu is
 1
Lu = 12 − 20   = 10.33 ft.
 12 
Then
kLu 1.0(10.33)(12)
= = 41.3
r 0.3(10)
(b) Slenderness ratio limits. Since the end moments are negligible, the minimum eccen-
tricity provisions of ACI-6.6.4.5.4 govern the design. Accordingly, the deformation
should be considered as single curvature with M1ns /M2ns = –1.0. The slenderness
limit for a braced frame is [Eq. (13.15.1)]
 kLu   M 
 r  = 34 + 12 1ns  = [34 − 12] = 22
limit  M 2 ns 
which is less than 41.3. Thus slenderness effects must be considered.

(c) Braced frame moment magnifier, δns. For this example, only the load combination of
1.2D + 1.6L is used.
Cm
δ ns =
Pu
1−
0.75Pc
where
M1ns
Cm = 0.6 − 0.4 = 0.6 − 0.4( −1.0) = 1.0
M 2 ns
(for single-curvature meember in a braced frame )
Pu = 1.2(39) + 1.6(106) = 216 kips
π 2 ( EI )eff
Pc =
(knonsway Lu )2
For the stiffness parameter, (EI )eff, Eqs. (13.9.8) and (13.9.9) will be used to compare the
values and illustrate the procedure.
1 10(103 )
Ec = 57, 000 3000 = 3120 ksi, Ig = = 833 in.4
1000 12
Es = 29, 000 ksi, I se = 2(2.79)(2.58)2 = 37.1 in.4
The proportion of factored axial load that is sustained, βdns, is
39(1.2)
β dns = = 0.216
106(1.6) + 39(1.2)
(Continued)
514
Thus, using Eq. (13.9.8)

0.4 Ec I g 0.4(3120)833
( EI )eff = = = 855, 000 kip-in.2
(1 + βdns ) (1 + 0.216)
and using Eq. (13.9.9)

0.2 Ec I g + Es I se 0.2(3120)833 + 29, 000(37.1)
( EI )eff = =
(1 + βdns ) (1 + 0.216)
= 1, 310, 000 kip-in.2
Using the larger of the two values, the critical buckling load Pc for the column is
π 2 ( EI )eff π 2 (1, 310, 000)

Pc = = = 844 kips
(knonsway Lu )2 [1.0(10.33)12]2
Thus,
1
δ ns = = 1.52
216
1−
0.75(844)
Thus, the column must be designed for the minimum moment Mm,min magnified by 1.52
(see Section 13.13). In other words,
Mu = 1.52 M m ,min
However, because the magnified moment exceeds the 1.4 times the first-order moment
(ACI-6.2.6), the column would be considered inadequate.
(d) Alternative approach for effective length factor k. Using the moments of inertia
given in ACI-6.6.3.1.1,
( EI )beam = Ec (0.35I g ) = 3120(0.35)(8000) = 8, 740, 000 kip in.2
( EI )col = Ec (0.70 I g ) = 1, 820, 000 kip in.2 (10 in. × 10 in. upper column)
( EI )col = Ec (0.70 I g ) = 3, 770, 000 kip in.2 (12 in. ×12 in. lower column)
End restraint factors,
∑ EI /L for cols 1820 /12

ψ A (top) = = = 0.21
∑ EI /L for beams 2(8740) / 24

(1820 + 3770) / 12
ψ B (bottom) = = 0.64
2(8740) / 24
From Fig. 13.11.1(a), knonsway = k ≈ 0.65. In this case, the beams are very stiff compared
to the columns, which yields a lower value than 1.0. A more realistic effective slender-
ness ratio is
kLu 0.65(10.33)12
= = 26.9
r 0.3(10)
This value is still greater than the limit of 22 computed in part (a), and thus slenderness
effects must be considered.
(Continued)
51
13.17 EXAMPLES 515
The revised critical buckling load and corresponding magnification factor are
π 2 ( EI )eff π 2 (1, 310, 000)

Pc = = = 1995 kips
(knonsway Lu )2 [0.65(10.33)12]2
Thus,
1
δ ns = = 1.17
216
1−
0.75(1995)
The column should then be designed for

Pu = 216 kips
Mu = M max = 1.17 M m ,min

1
Mu = 1.17(216)[0.6 + 0.03(10)] = 19 ft-kips
12
In this case, the upper limit of 1.4 per the requirements of ACI-6.2.6 is satisfied (see
Section 13.15).
(e) Check strength. The strength of the section may be checked by the methods of
Chapter 10. The design interaction diagram for the column is shown in Fig. 13.17.10.
It is seen that the required strength is well below the design strength for the column.
In fact, the column appears to be overdesigned for this load combination and the
reinforcement could be reduced.
φPn, Pu (kips)
450
400
350
300
250
200
150
100
50
0 10 20 30 40 50 60
φMn, Mu (ft-kips)
Figure 13.17.10 Design interaction diagram for the column of Example 13.17.2.

516
EXAMPLE 13.17.3
Determine the adequacy of the square tied column (17 in. square, with 10–#9 bars,
fc′ = 3000 psi, fy = 60,000 psi) which is an exterior first-floor column in the frame of
Fig. 13.17.11. Assume that this frame is braced sufficiently to prevent relative transla-
tion of its joints. Also assume 40% of the factored axial load is sustained.
30’–0” 30’–0” 30’–0”
12’–0”
12’–0” B
14 × 24 14 × 24
17“ square
12’–0”
A columns
Assume hinged
(a) Frame dimensions
Pu = 525k
Mu = M2ns = 105’k
10 – #9
Mu
Clear height
e= = 2.4 in.
10’–0”
Pu
1 1 Min e = 0.6 + 0.03h

2 2” 2 2”
= 1.11 in.
17” square
(b) Cross section

Pu
(c) Member A
Figure 13.17.11 Braced frame of Example 13.17.3.
SOLUTION
(a) Effective length. In accordance with the more conservative procedure of ACI-
6.6.4.4.3, which allows the option of not doing an analysis to determine a k value
less than 1.0 for a braced frame, use

kLu = Lu = 12 − 2 = 10 ft
(b) Slenderness ratio limits. The column slenderness ratio is
kLu 120
= = 23.5
r 0.3(17)
Slenderness effects may be neglected when
kLu  M 
≤ 34 + 12 1ns  ≤ 40
r  M 2 ns 
(Continued)
517
13.17 EXAMPLES 517
In this case, M1ns /M2ns = 0,
 kLu 
 r  = 34 > 23.6
limit
Slenderness effects may be neglected. In the following, the moment magnifier method
will be illustrated even though it would not be required by the ACI Code.
(c) Braced frame moment magnifier δns.
Ec = 57 fc′ = 3120 ksi

1
Ig = (17)(17)3 = 6960 in.4
12
I se = 2(5)(6) = 360 in.
2 4
Using Eq. (13.9.8),

EI = 0.4 Ec I g = 8, 690, 000 kip in.2
or Eq. (13.9.9)
EI = 0.2 Ec I g + Es I se
= 0.2(3120)(6960) + 29, 000(360)

= 4, 340, 000 + 10, 440, 000 = 14, 800, 000 kip in. 2
Using the larger value of EI and applying the factor (1 + βdns) to account for sustained load,
14, 800, 000

( EI )eff = = 10, 500, 000 kip-in.2
(1 + 0.4)
π 2 ( EI )eff π 2 (10, 500, 000)
Pc = = = 7200 kips
(knonsway Lu )2 [1.0(10)12]2
Pu = 525 kips [from Fig.13.17.11(c)]

Pu 525
= = 0.097
0.75Pc 0.75(7200)
M1ns
Cm = 0.6 − 0.4 = 0.6
M 2 ns
Cm 0.6
δ ns = = = 0.66 < 1.0
1 − Pu / 0.75Pc 1 − 0.097
In this case, no moment magnification is required. The factored loads to be carried are
Pu = 525 kips and Mu = 105 ft-kips at the top of the column. The design strength φPn is
found to be 616 kips at e = 2.4 in. Thus, the column is adequate for the load combination
considered.
518
EXAMPLE 13.17.4
Determine the adequacy of the square tied column (20 in. square, with 12–#10 bars,
fc′ = 4000 psi, fy = 60,000 psi) which is an exterior first-story column in the unbraced
frame of Fig. 13.17.12. Assume that 20% of the factored axial load is sustained. The
ΣPu on the four columns in the lowest story is 3150 kips. Use the Sum of P method to
compute the moment magnifier.
12’–0”
12’–0”
14 × 24 14 × 24 14 × 24
20” square
12’–0” columns
A
Assume
bases fixed
30’–0” 30’–0” 30’–0”
(a) Frame dimensions
Pns = 488k; Ps = 0k
#3 ties Mu = M2ns + M2s

= 70 + 265 = 335’k
20”
12–#10
Clear height
10’–0”
5” 5” 5”
2.50” 2.50”
20” Mu = M1ns + M1s
= 35 + 145 = 180’k
(b) Cross section (c) Member A
Figure 13.17.12 Unbraced frame of Example 13.17.4.
SOLUTION
(a) Factored load combinations. Assume that for this example, the forces given in
Fig. 13.17.12 are from the factored load combination including wind, and that wind
causes insignificant axial compression (assumed zero here) on the member being
studied. At the top, the first-order nonsway and sway actions on the column are M2ns =
70 ft-kips with Pns = 488 kips, and M2s = 265 ft-kips with Ps = 0 kips (assumed),
respectively. At the bottom, the nonsway and sway moments are M1ns = 35 ft-kips
and M1s = 145 ft-kips. Only the check of strength required for U = 1.2D + 1.0W +
0.5L is illustrated in this example.
(b) Effective length and slenderness ratio. The end restraint factors ψ must be deter-
mined. Using the moment of inertia for the beam of 0.35Ig per ACI-6.6.3.1.1,
14(24)3
I beam = 0.35 = 5645 in.2
12
For the 20-in. column, use 0.70Ig per ACI-6.6.3.1.1,
(20)4
I col = 0.70 = 9333 in.4
12
(Continued)
519
13.17 EXAMPLES 519
For the unbraced frame, the magnifier δs involves the sum of the buckling loads Pc for
all columns participating in the sidesway resistance in the first story. Thus, the effective
length factor k is needed for each of these columns. In this example,
∑ EI /L for columns
ψ A (top of exterior column) =
∑ EI /L for beams
2 Ec (9333) /12
= = 8.26
Ec (5645) / 30
ψ A (top of interior column) = 4.13
For this calculation use of center-to-center span distances is recommended as being con-
sistent with the nominal frame analysis using those distances. Theoretically, at the fixed
end ψ equals zero; however, for practical purposes ψ should not be taken smaller than
1.0. (see section 13.11). Thus,

ψ B (bottom) = 1.0
Using Fig. 13.11.1(b), find
ksway of exterior column ≈ 1.85, ksway of interior column ≈ 1..65
The effective slenderness ratio for the exterior column being investigated is
kLu 1.85(10.0)(12)
= = 37
r 0.3(20)
which exceeds the limit of 22 given by ACI-6.2.5 for unbraced frames. Slenderness
effects must be considered.
(c) Magnification factors. For the unbraced (sway) frames,
M 22 nd = M 2 ns + δ s M 2 s [13.10.8]
For this example, the end moment M2ns under factored gravity load is given as
70 ft-kips and the end moment M2s caused by factored lateral load is given as 265 ft-kips.
Using the sum of P method, the unbraced frame magnifier, δs, is computed as
1
δs =
∑ Pu
1−
0.75 ∑ Pc
Compute (EI )eff from Eq. (13.9.9) as
0.2 Ec I g + Es I se
( EI )eff = [13.9.9]
(1 + β dns )
In this example, there is no sustained lateral load and it is assumed that the dead loads
cause no sway; thus βdns = βds = 0. Accordingly,
1
Ig = (20)(20)3 = 13, 333 in.4
12
I se = 1.27(2)[ 4(7.5)2 + 2(2.5)2 ] = 603 in.4
Ec = 57 fc′ = 57 4000 = 3605 ksi
( EI )eff = 0.2 Ec I g + Es I se

= 0.2(3605)(13, 333) + 29, 000(603)
= 9, 610, 000 + 17, 500, 000
= 27,110, 000 kiip in.2 (larger than 0.4Ec I g )
(Continued)
520
For the 20-in. square exterior column,
π 2 ( EI )eff π 2 (27,110, 000)

Pc = = = 5429 kipps
(ksway Lu )2 [1.85(10)12]2
For the 20-in. square interior column,
π 2 ( EI )eff π 2 (27,110, 000)

Pc = = = 6825 kipps
(ksway Lu )2 [1.65(10)12]2
Then, ΣPc is in this case,

∑ Pc = 2(5429 + 6825) = 24, 508 kips
The total factored load ΣPu acting on the four columns in the first story is given as 3150 kips.
Thus, the magnifier δs for the P-Δ effect is
1
δs = = 1.21
3150
1−
0.75(24, 508)
Using Eq. (13.10.8),
M 22 nd = M 2 ns + δ s M 2 s

= 70 + 1.21(265) = 391 ft-kips
The ratio of second-order to first-order moment is 391/335 = 1.17, which is less than the
upper limit of 1.4 of ACI-6.2.6. Thus, the column must be designed for Pu = 488 kips
and Mu = 391 ft-kips, or for Pu = 488 kips and an eccentricity e = 391(12)/488 = 9.61 in.,
for this load combination.
(d) Compute the design column strength φPn and φMn for the member.
Using the principles of Chapter 10, at an eccentricity e = 9.61 in., the design strength is
φPn = 506 kips, which is greater than Pu = 488 kips. OK
EXAMPLE 13.17.5
Determine the adequacy of the 14 × 20 in. compression member reinforced with 6–#11
bars, fc′ = 4500 psi, and fy = 60,000 psi, as shown in Fig. 13.17.13. The member serves
as an exterior column in a braced frame; the loading and factored moment diagram are
as shown in the figure, and the clear height is 22 ft 6 in.
SOLUTION
(a) For possible instability in the plane of the frame,
kLu 1.0(22.5)12
= = 45
rx 0.3(20)
where k for this braced column has been conservatively assumed as 1.0.
(Continued)
521
13.17 EXAMPLES 521
Pu1
x Bending axis
#4 ties
2.71”
6 – #11
14” 7’–6”
2.71” ePu2
Pu2 Mm = 279’k
2.71”
2.71” x
Moment
20”
15’–0”
Pu1 + Pu2 = Pu = 115k
Figure 13.17.13 Column cross section, loading, and factored, first-order bending moment
diagram of Example 13.17.5.
For a braced column without transverse loading having this irregular moment dia-
gram, when the bending moment diagram has the largest moment at a location other
than at an end, M1ns /M2ns should be taken conservatively as –1.0. Thus, the slenderness
limit is [Eq. (13.15.1)]
 kLu   M 
 r  = 34 + 12 1ns  = [34 − 12] = 22
limit  M 2 ns 
which is less than 45. Thus slenderness effects must be considered.

For this column,
1
Ig = (14)(20)3 = 9330 in.4
12
I se = 2(3)(1.56)(7.29)2 = 497 in.4

Ec = 57 4500 = 3820 ksi
Assuming no sustained load,
β dns = 0

EI = 0.4 Ec I g = 0.4(3820)(9330) = 14, 300, 000 kip in.2
or

= 0.2(3820)(9330) + 29, 000(497) = 21, 500, 000 kipps in.2
(Continued)
52
The column is not subjected to end moments only and it does not have transverse loading.
Conservatively, assume Cm =1.0. Using the larger value of EI,
π 2 ( EI )eff π 2 (21, 500, 000)

Pc = = = 2910 kips
(knonsway Lu )2 [1.0(22.5)12]2
and
Cm 1.0
δ ns = = = 1.06
1 − Pu / 0.75Pc 1 − 115 / [0.75(2910)]
The required eccentricity for the column is

Mu δ ns M m
e= =
Pu Pu

1.06(279)12
e= = 30.9 in.
115
Assuming φ = 0.90 for this rather large eccentricity (e /h = 1.55), the required nomi-
nal strength of the member in the plane of the frame (i.e., the strong orientation of the
member) is
Pu 115
φ 0.90

δ ns M n 1.06(279)
φ 0.90
A statics analysis of this section using an eccentricity of 30.9 in. gives

Pn = 194 kips > [ Pu /φ = 128 kips] OK
(b) Instability transverse to the plane of the frame. The slenderness ratio is
kLu 1.0(22.5)12
= = 64.3
ry 0.3(14)
which exceeds the limiting value of 22 for which the effect of slenderness may be
neglected. Again, as in part (a), M1ns /M2ns may be conservatively taken as –1.0. There
are no moments in this direction, thus the minimum moment Mm,min must be considered
and magnified by the factor δns (see Section 13.13).
1
Ig = (20)(14)3 = 4570 in.4
12
I se = 2(2)(1.56)(4.29)2 = 115 in.4

EI = 0.4 Ec I g = 0.4(3820)(4570) = 6, 980, 000 kip in. 2
or

= 0.2(3820)(4570) + 29, 000(115) = 6, 830, 000 kip in.2
(Continued)
523
Using the larger of these two values and Cm = 1.0,
π 2 ( EI )eff π 2 (6, 980, 000)

Pc = 2
= = 945 kips
(knonsway Lu ) [1.0(22.5)12]2
and
Cm 1.0
δ ns = = = 1.19
1 − Pu / 0.75Pc 1 − 115 / [0.75(945)]
Thus, in the weaker direction

e = 1.19(0.6 + 0.03h) = 0.71 + 0.036h
= 0.71 + 0.036(14) = 1.21 in.
and
1
M max = Pu e = 115(1.21) = 12 ft-kips.
12
Assuming φ = 0.65 for this low eccentricity (e/h = 0.09), the member must have a strength
Pn = Pu /0.65 = 177 kips for an eccentricity of 1.21 in. A statics analysis indicates that
the nominal strength Pn is 1390 kips at this eccentricity with respect to the weaker axis.
However, the strength Pn may not be taken greater than 0.80Po,
Pn(max) = 0.80[0.85(4.5)(280 − 9.36) + 60(9.36)] = 1277 kips
which is less than the capacity at the minimum eccentricity of 1.21 in. Therefore,
Pn = 1277 kips > [ Pu /φ = 177 kips] OK
SELECTED REFERENCES
13.1. L. Euler. De Curvis Elasticis, Additamentum I, Methodus In-veniendi Lineas Curvas Maximi
Minimive Proprietate Gaudentes. Lausanne and Geneva: 1744 (pp. 267–268); and “Sur la Force
des Colonnes,” Mémoires de l’Académie de Berlin, Vol. 13. Berlin: 1759 (pp. 252–282).
13.2. F. Engesser. “Über die Knickfestigkeit gerader Stäbe,” Zeitschrift des Architektenund Ingenieur-
Vereins zu Hannover, 35 (1889), 445–462. Also “Die Knickfestigkeit gerader Stäbe,” Zentralblatt
der Bauverwaltung, Berlin, Dec. 5, 1891, p. 483.
13.3. T. von Kármán. “Die Knickfestigkeit gerader Stäbe,” Physikalische Zeitschrift, Vol. 9, 1908
(p. 136). Also, “Untersuchungen über knickfestigkeit,” Mitteilungen über Forschungsarbeiten
auf dem Gebiete des Ingenieurwesens, No. 81. Berlin: 1910.
13.4. Luis P. Sáenz and Ignacio Martín. “Test of Reinforced Concrete Columns with High Slenderness
Ratios,” ACI Journal, Proceedings, 60, May 1963, 589–616.
13.5. John E. Breen and Phil M. Ferguson. “The Restrained Long Concrete Column As a Part of a
Rectangular Frame,” ACI Journal, Proceedings, 61, May 1964, 563–587.
13.6. Edward O. Pfrang and Chester P. Siess. “Behavior of Restrained Reinforced Concrete Columns,”
Journal of the Structural Division, ASCE, 90, ST5 (October 1964), 113–135; Disc., 91, ST3
(June 1965), 280–287.
13.7. J. G. MacGregor. Discussion of “Behavior of Restrained Reinforced Concrete Columns,” by
E. O. Pfrang and C. P. Siess, Journal of the Structural Division, ASCE, 91, ST3 (June 1965),
280–287.
524
13.8. Richard W. Furlong and Phil M. Ferguson. “Tests of Frames with Columns in Single Curvature,”
Symposium on Reinforced Concrete Columns (SP-13). Detroit: American Concrete Institute,
1966 (pp. 55–73).
13.9. Phil M. Ferguson and John E. Breen. “Investigation of the Long Column in a Frame Subject
to Lateral Loads,” Symposium on Reinforced Concrete Columns (SP-13). Detroit: American
Concrete Institute, 1966 (pp. 75–119).
13.10. Ignacio Martín and Elmer Olivieri. “Test of Slender Reinforced Concrete Columns Bent in
Double Curvature,” Symposium on Reinforced Concrete Columns (SP-13). Detroit: American
13.11. J. G. MacGregor and S. L. Barter. “Long Eccentrically Loaded Concrete Columns Bent in
Double Curvature,” Symposium on Reinforced Concrete Columns (SP-13). Detroit: American
Concrete Institute, 1966 (pp. 139–156.)
13.12. Edward O. Pfrang. “Behavior of Reinforced Concrete Columns with Sidesway,” Journal of the
Structural Division, ASCE, 92, ST3 (June 1966), 225–252.
13.13. R. Green and John E. Breen. “Eccentrically Loaded Concrete Columns under Sustained Load,”
13.14. John E. Breen and Phil M. Ferguson. “Long Cantilever Columns Subject to Lateral Forces,”
13.15. Shriniwas N. Pagay, Phil M. Ferguson, and John E. Breen. “Importance of Beam Properties on
Concrete Column Behavior,” ACI Journal, Proceedings, 67, October 1970, 808–815.
13.16. Hajime Okamura, Shriniwas N. Pagay, John E. Breen, and Phil M. Ferguson. “Elastic Frame
Analysis—Corrections Necessary for Design of Short Concrete Columns in Braced Frames,”
13.17. Phil M. Ferguson, Hajime Okamura, and Shriniwas N. Pagay. “Computer Study of Long
Columns in Frames,” ACI Journal, Proceedings, 67, December 1970, 955–958.
13.18. G. A. Blomier and J. E. Breen. “Effect of Yielding of Restraints on Slender Concrete Columns
with Sidesway Prevented,” Reinforced Concrete Columns (SP-50), Detroit: American Concrete
Institute, 1975 (pp. 41–65).
13.19. F. N. Rad and R. W. Furlong. “Behavior of Unbraced Reinforced Concrete Frames,” ACI
Journal, Proceedings, 77, July–August 1980, 269–278.
13.20. J. S. Ford, D. C. Chang, and J. E. Breen. “Behavior of Concrete Columns under Controlled
Lateral Deformation,” ACI Journal, Proceedings, 78, January–February 1981, 3–20.
13.21. J. S. Ford, D. C. Chang, and J. E. Breen. “Experimental and Analytical Modeling of Unbraced
Multipanel Concrete Frames,” ACI Journal, Proceedings, 78, January–February 1981, 21–35.
13.22. J. S. Ford, D. C. Chang, and J. E. Breen. “Behavior of Unbraced Multipanel Concrete Frames,”
ACI Journal, Proceedings, 78, March–April 1981, 99–115.
13.23. Bengt Broms and I. M. Viest. “Ultimate Strength Analysis of Long Hinged Reinforced

Concrete Columns,” Journal of the Structural Division, ASCE, 84, ST1 (January 1958) (Paper
No. 1510).
13.24. Nestor R. Iwankiw. “Note on Beam–Column Moment Amplification Factor,” Engineering
Journal, AISC, 21, 1(1st quarter 1984), 21–23.
13.25. Charles Massonnet. “Stability Considerations in the Design of Steel Columns,” Journal of the
Structural Division, ASCE, 85, ST7 (September 1959), 75–111.
13.26. Charles G. Salmon, John E. Johnson, and Faris A. Malhas. Steel Structures—Design and
Behavior: Emphasizing Load and Resistance Factor Design (5th ed.) New Jersey: Pearson-
Prentice Hall Publishers, 2009.
13.27. Ronald D. Ziemian Ed. Guide to Stability Design Criteria for Metal Structures (6th ed.)
New York: John Wiley & Sons, 2010.
13.28. J. G. MacGregor. “Out-of-Plumb Columns in Concrete Structures,” Concrete International, 1,
June 1979, 26–31.
13.29. James G. MacGregor and Sven E. Hage. “Stability Analysis and Design of Concrete Frames,”
Journal of the Structural Division, ASCE, 103, ST10 (October 1977), 1953–1970.
13.30. J. G. MacGregor. “Design of Slender Concrete Columns—Revisited,” ACI Structural Journal,
90, May–June 1993, 302–309.
13.31. Chu-Kia Wang and Carlito P. Talaboc. “Second-Order Analysis of Multistory Building Frames
by Marching Technique,” Microcomputers in Structural Engineering, 1, 3 (December 1986),
209–232.
13.32. James G. MacGregor, John E. Breen, and Edward O. Pfrang. “Design of Slender Concrete
Columns,” ACI Journal, Proceedings, 67, January 1970, 6–28.
13.33. Eldon F. Mockry and David Darwin. “Slender Column Interaction Diagrams,” Concrete

International, 4, June 1982, 44–50.
13.34. Alfred L. Parme. “Capacity of Restrained Eccentrically Loaded Long Columns,” Symposium
on Reinforced Concrete Columns (SP- 13). Detroit: American Concrete Institute, 1966
(pp. 325–367).
52
13.35. PCA. Design Constants for Rectangular Long Columns. Advanced Engineering Bulletin No.
12. Skokie, IL: Portland Cement Association, 1964, 36 pp.
13.36. Brian R. Wood, Denis Beaulieu, and Peter F. Adams. “Column Design by P–Delta Method,”
Journal of the Structural Division, ASCE, 102, ST2 (February 1976), 411–427.
13.37. Brian R. Wood, Denis Beaulieu, and Peter F. Adams, “Further Aspects of Design by P–Delta
Method,” Journal of the Structural Division, ASCE, 102, ST3 (March 1976), 487–500.
13.38. H. Scholz. “A Novel P–Delta Method for Reinforced Concrete Frames Designed by LRFD
Approaches,” ACI Journal, Proceedings, 83, July–August 1986, 633–641.
13.39. H. Scholz. “Alternative P–Delta Method for Prestressed Members Subjected to Axial Load and
Bending,” ACI Structural Journal, 87, July–August 1990, 464–472.
13.40. James Colville. “Slenderness Effects in Reinforced Concrete Square Columns,” Reinforced
Concrete Columns (SP-50). Detroit: American Concrete Institute, 1975 (pp. 165–191).
13.41. J. G. MacGregor, U. H. Oelhafen, and S. E. Hage. “A Reexamination of the EI Value for Slender
Columns,” Reinforced Concrete Columns (SP-50). Detroit: American Concrete Institute, 1975
(pp. 1–40).
13.42. Ian C. Medland and Donald A. Taylor. “Flexural Rigidity of Concrete Column Sections,”
Journal of the Structural Division, ASCE, 97, ST2 (February 1971), 573–586.
13.43. Jian-
Min Zeng, Lian Duan, Fu- Ming Wang, and Wai- Fah Chen. “Flexural Rigidity of
Reinforced Concrete Columns,” ACI Structural Journal, 89, March–April 1992, 150–158.
13.44. M. A. Diaz and J. M. Roesset. “Evaluation of Approximate Slenderness Procedures for
Nonlinear Analysis of Concrete Frames,” ACI Structural Journal, 84, March–April 1987,
139–148.
13.45. S. A. Mirza. “Flexural Stiffness of Rectangular Reinforced Concrete Columns,” ACI Structural
13.46. Gregory R. Sigmon and Shuaib H. Ahmad. “Flexural Rigidity of Circular Reinforced Concrete
Sections,” ACI Structural Journal, 87, September–October 1990, 548–556.
13.47. M. R. Ehsani and F. Alameddine. “Refined Stiffness of Slender Circular Reinforced Concrete
Columns,” ACI Structural Journal, 84, September–October 1987, 419–427.
13.48. Brij B. Goyal and Neil Jackson. “Slender Concrete Columns under Sustained Load,” Journal of
the Structural Division, ASCE, 97, ST11 (November 1971), 2729–2750.
13.49. Robert G. Drysdale and Mark W. Huggins. “Sustained Biaxial Load on Slender Concrete
Columns,” Journal of the Structural Division, ASCE, 97, ST5 (May 1971), 1423–1443.
13.50. B. Vijaya Rangan. “Lateral Deflection of Slender Reinforced Concrete Columns under

Sustained Load,” ACI Structural Journal, 86, November–December 1989, 660–663.
13.51. J. S. Ford, D. C. Chang, and J. E. Breen. “Design Indications from Tests of Unbraced Multipanel
Concrete Frames,” Concrete International, 3, March 1981, 37–47.
13.52. Thomas C. Kavanagh. “Effective Length of Framed Columns,” Transactions ASCE, 127, Part
II, 1962, 81–101.
13.53. John E. Breen, James G. MacGregor, and Edward O. Pfrang. “Determination of Effective
Length Factors for Slender Concrete Columns,” ACI Journal, Proceedings, 69, November
1972, 669–672.
13.54. BSI. Code of Practice for the Structural Use of Concrete, Part I. Design, Materials and
Workmanship. London: British Standards Institution, 1972.
13.55. Richard W. Furlong. “Column Slenderness and Charts for Design,” ACI Journal, Proceedings,
68, January 1971, 9–18.
13.56. Yu-Xia Hu, Ren-Gen Zhou, Won-Sun King, Lian Duan, and Wai-Fah Chen. “On Effective
Length Factor of Framed Columns in the ACI Building Code,” ACI Structural Journal, 90,
March–April 1993, 135–143.
13.57. Lian Duan, Won-Sun King, and Wai-Fah Chen. “K-Factor Equation to Alignment Charts for
Column Design,” ACI Structural Journal, 90, May–June 1993, 242–248.
13.58. Richard W. Furlong, “Rational Analysis of Multistory Concrete Structures,” Concrete

International, 3, June 1981, 29–35.
13.59. S. I. Abdel-Sayed and N. J. Gardner. “Design of Symmetric Square Slender Reinforced
Concrete Columns under Biaxially Eccentric Loads,” Reinforced Concrete Columns (SP-50)
13.60. S. A. Mirza, P. M. Lee, and D. L. Morgan. “ACI Stability Resistance Factor for RC Columns,”
Journal of Structural Engineering, ASCE, 113, No. 9, September, 1987, 1963–1976.
13.61. Madhu Khuntia and S. K. Ghosh. ”Flexural Stiffness of Reinforced Concrete Columns and
Beams: Analytical Approach,” ACI Structural Journal, 101, No. 3, May-June 2004, 351–363.
13.62. Richard W. Furlong, Cheng-Tzu Thomas Hsu, and S. Ali Mirza. “Analysis and Design of
Concrete Columns for Biaxial Bending—Overview,” ACI Structural Journal, 101, No. 3, May–
June 204, 413–423.
526
PROBLEMS
All problems are to be done in accordance with the ACI Code unless otherwise indicated.
All dead and live loads given are service-level loads, while wind loads are strength-level
loads computed per ASCE 7 unless otherwise indicated. Even though the slenderness
effects might be neglected according to the ACI Code in a given problem, consider slender-
ness effects anyway. For design problems, use whole inches for member size.
13.1 Determine the adequacy of the section, includ- 13.3 Assume that the frame of Problem 13.2 is
ing slenderness effects, for a 16-in. square tied unbraced and consists of 4 bays symmetrical
column that has a clear height of 18 ft and serves about the middle. The gravity axial loads are 44
as an interior member of a braced frame. The kips dead load and 66 kips live load. The largest
member is designed as axially loaded with the moments in column A are 15 ft-kips dead load,
following service loads: live load, 130 kips; 12 ft-kips live load, and 58 ft-kips wind load.
dead load, 200 kips. Use fc′ = 4000 psi and fy An exterior column carries one-half the factored
= 60,000 psi. Without actually selecting bars, axial load of an interior column. Investigate the
assume about 2 1 2 % total reinforcement equally adequacy of the interior column A. If not ade-
divided in the opposite faces of the member. Use quate, redesign it to make it satisfactory accord-
the ACI moment magnifier method. ing to the moment magnifier method of the
13.2 Determine the adequacy, including slenderness ACI Code.
effects, for a 14-in.-diameter spirally reinforced 13.4 Determine the adequacy of the exterior square
1
column (assume about 2 2 % reinforcement) column (column B) of the figure for Problem
which is an interior second-floor column (column 13.2 if the member is carrying a gravity axial
A) in the braced frame of the figure for Problems compression of 217 kips dead load and 145
13.2 through 13.8. The member is to carry an axial kips live load and primary bending moments
service load of 87 kips dead load and 133 kips of 27 ft-kips dead load and 18 ft-kips live load.
live load, with only the dead load considered sus- Assume that the primary bending moment var-
tained. Use fc′ = 5000 psi and fy = 60,000 psi. Use ies from +M at the top of the member linearly
the ACI moment magnifier method. to –M/2 at the bottom, with joint translation
14” square 14” diameter 14”

14 × 22 14 × 22
14” square 14” diameter

15’ – 0” Column B Column A
4 – #9
14 × 22 14 × 22
16” square 16” diameter

15’ – 0” Column C
1
22”
Column B
40’ – 0” 40’ – 0”
16”
Symmetrical
about this
line
6 – #9
1
22”
Column C
Problems 13.2 to 13.8

527
PROBLEMS 527
adequately prevented (i.e., a braced frame). Use the adequacy according to the moment magnifier
fc′ = 5000 psi and fy = 60,000 psi. Use the ACI method of the ACI Code. If not adequate, redesign
moment magnifier method. assuming the loadings are not affected by changes
13.5 Repeat Problem 13.4, except consider the pri- in member stiffnesses. The total factored load ΣPu
mary bending moment constant over the height to all columns in the lower story is 885 kips for
of the column. load combination 1.2D + 1.0W + 0.5L.
13.6 Assume that the member (Column B) in Problem 13.9 Design column A for the unbraced frame of the
13.4 is part of an unbraced frame and that max- figure for Problem 13.9 for the load combination
imum nonsway moments M2ns are 28 and 20 of 1.2D + 1.0W + 1.0L (Assume that the columns
ft-kips due to dead and live loads, respectively. will support areas having live loads greater than
Also, a maximum M2s sway moment of 85 ft- 100 psf.) The axial compression is 112 kips dead
kips due to strength-level wind load must be load and 112 kips live load; bending moment is
carried. The axial compressive loads are 30 kips 35 ft-kips dead load, 35 ft-kips live load, and
service dead load and 40 kips service live load. 118 ft-kips wind load. Consider only the dead
The total factored load ΣPu to all five columns load to be sustained. The factored axial load Pu
in the story is 590 kips for the load combination on the interior columns is 1.8 times that on the
1.2D + 1.0W + 0.5L. Investigate the adequacy of exterior columns. Select a square member to
the 14-in. square exterior column (column B). contain approximately 2 1 2 % reinforcement. Use
If found to be inadequate, redesign the member fc′ = 5000 psi, fy = 60,000 psi, and the moment
assuming that the loadings are not affected by a magnifier method of the ACI Code.
change in member stiffness. 13.10 (a) Redesign the column used in Problem 13.9
13.7 Determine the adequacy of the 16-in. square except for service loads, use an axial com-
first-story column (column C) of the figure for pression of 132 kips dead load and 132 kips
Problem 13.2, which is carrying a service axial live load, and bending moments M2 of
load of 370 kips and a primary bending moment 26 ft-kips dead load, 26 ft-kips live load,
of 121 ft-kips (assume loads 70% dead load). and 123 ft-kips wind load. The length for
The primary bending moment varies from a column A is 20 ft instead of 21 ft.
maximum at the top to zero at the bottom and (b) How much smaller could the member have
joint translation is adequately prevented. Use been if the frame had been adequately
fc′ = 5000 psi and fy = 60,000 psi. Use ACI braced to prevent joint translation?
moment magnifier method. 13.11 (a) Redesign the columns used in Problem
13.8 Assume that the member (column C) of Problem 13.9 except use 18 ft instead of 21 ft for the
13.7 is part of an unbraced frame. The axial loads length of column A.
are 45 kips dead load and 60 kips live load. The (b) How much smaller could the member have
bending moments are 30 ft-kips dead load, 22 ft- been if the frame had been adequately
kips live load, and 110 ft-kips wind load. Investigate braced to prevent joint translation?
Unbraced frame
14’–0”
h
16 × 28 beam
Assume same size 1

2 2”
1
22”
for all columns h
Place reinforcement
21’–0” symmetrically in
A two faces
There are three
interior columns
ψ = 1.0
30’– 0”

CHAPTER 14
STRUT-A ND-T IE MODELS—
DEEP BEAMS, BRACKETS,
AND CORBELS
14.1 INTRODUCTION
As discussed in Section 5.7, the zone in the vicinity of a concentrated load, or at an abrupt
discontinuity in the member (such as an abrupt change in member depth), is called a
“D-region” to refer to a discontinuity or a disturbance [5.9]. A typical example of a
D-region is that of a deep beam where the shear span-to-depth ratio, a /d, is less than about
2. In such cases, plane sections cannot be assumed to remain plane after bending (Bernoulli
hypothesis), and thus the distribution of stresses and strains cannot be obtained by using
ordinary beam theory, as presented in Chapters 3, 4, and 5. For many years, the design of
D-regions relied on experimental evidence, past experience, and good practice. Today, how-
ever, truss models or strut-and-tie models are accepted tools for the design of D-regions.
Intended to represent the flow of stresses at failure, a strut-and-tie model is an idealized
truss that is in equilibrium with the externally applied loads and that satisfies the require-
ments of the lower-bound theorem of plasticity theory. This means that the failure load cal-
culated by using a strut-and-tie model will be less than or equal to the actual failure load,
and thus it offers a safe value of the load-carrying capacity of the structure. In design, either
two-or three-dimensional models, or both, are used to represent the D-regions in various
structures or elements. Two-dimensional models are used for planar structures such as deep
beams, brackets or corbels, dapped-end beams, walls, and beam-column joints. Elements
for which three-dimensional models may be required include bridge piers and pile caps.
Consider the simply supported deep beam with concentrated loads and the correspond-
ing truss or strut-and-tie model shown in Fig. 14.1.1. The flow of stresses may be idealized
by two inclined members in compression, or struts in the concrete that carry the load to the
supports and a tension member or tie at the bottom of the beam represented by the longi-
tudinal reinforcement. Also, a horizontal strut is used to represent the stresses that develop
in the compression zone of the beam at midspan. In the model, struts and ties have in-and
out-of-plane finite dimensions (width, length, and thickness) that will depend on the truss
geometry and the forces developed in each truss element. The intersection of the axes of
struts and ties defines the nodes (or joints) of the model, and the region surrounding the
node defines what is referred to as the nodal zone. In general, failure of a truss will occur
when one or more truss elements, including the nodes, reach the design strength. In the
strut-and-tie model of Fig. 14.1.1, failure is assumed to occur when the struts reach their
crushing strength or when the tie reaches yielding. Failure may also occur at the joints by
failure of the nodal zone.
529
14.1 INTRODUCTION 529
Nodal zones
Bottle-shaped
strut Idealized
tapered strut
Idealized
prismatic,
parallel stress
field
Nodal zone
Nodal zone
Tie
Figure 14.1.1 Idealized flow of stresses and corresponding strut-and-tie model for a simply
supported deep beam.
Struts
Strut elements represent the compression stress field regions that develop in the member
under the applied loads. In practice, the shape of the compression field may vary depend-
ing on the geometry, the load, and the support conditions of the member. Jörg Schlaich,
Schäfer, and Jennewein [14.1] have proposed three basic types or shapes of compression
stress fields commonly encountered in practice, as shown in Fig. 14.1.2. The “basic” type
is a prism of uniform cross-sectional area, where the compression stress field is constant
over the length [Fig. 14.1.2(a)]. The prismatic, parallel stress field is commonly used to
represent the compression zone of a beam, such as that at midspan of the beam shown in
Fig. 14.1.1. The second type of strut, referred to as a bottle-shaped strut, is meant to rep-
resent regions where the concrete compressive stresses spread out at some distance away
from the loading or reaction point [Fig. 14.1.2(b)]. These stress fields may be found, for
example, between the applied load and the reaction of the deep beam of Fig. 14.1.1, which
may be modeled as a bottle-shaped strut. For simplicity in design, however, bottle-shaped
struts are often idealized as tapered struts instead, as shown in Fig. 14.1.1. In the third type
of stress field, the stresses “fan out,” as shown in Fig. 14.1.2(c). Fan-shaped action will
develop, for example, in a simply supported deep beam with a uniformly distributed load
as shown in Fig. 14.1.3(a). In practice, such regions will often be modeled as a series of
prismatic struts as illustrated in Fig. 14.1.3(b).
The strength (stress capacity) of the strut will depend on its shape and on the presence of
reinforcement (if any) crossing the strut axis. In a bottle-shaped strut, the spread of stresses
from the end to the middle of the strut will tend to split the strut near the end, which will
weaken the strut [see Fig. 14.1.4(a)]. In fact, the flow of forces within a bottle-shaped
strut may be idealized as a strut-and-tie model as shown in Fig. 14.1.4(b). A number of
researchers have suggested values for the strength of struts of different shapes (J. Schlaich
et al. [14.1], Marti [14.2], Rogowsky and MacGregor [14.3], Bergmeister, Breen, and Jirsa
[14.4], and Ramirez and Breen [14.5]). While the range of suggested values is diverse,
researchers agree that the strength of compressive struts is less than the compressive cylin-
der strength fc′ with values ranging from approximately 0.5 to 1 times fc′.
530
530 C H A P T E R 1 4 S trut- and - T ie M odels
b b
a a α
(a) (b)
a
(c)
Figure 14.1.2 Basic compression fields: (a) parallel field; (b) bottle-shaped field; (c) fan-shaped

field. (Adapted from J. Schlaich et al. [14.1].)
CL CL
a a/4 a/2 a/4
w wa/2 wa/2
C C
T T
(a) (b)
Figure 14.1.3 Fan action in a deep beam: (a) fan-shaped stress field; (b) strut-and-tie model.

(Adapted from Marti [14.2].)
Crack Tie
2
Strut
1
2
1
(a) (b)
Figure 14.1.4 Bottle-shaped strut: (a) splitting cracks near the strut ends; (b) idealized strut-and-
tie model of the strut. (Adapted from MacGregor [14.6].)
531
Ties
Ties represent the regions where tensile stresses develop in the member and are the tension
members in the model. They consist of the reinforcement (nonprestressed or prestressed)
plus a portion of the concrete around it. A tie may consist of one or several layers of rein-
forcement over the width of the tie, as shown in Fig. 14.1.5. For better crack distribution, it
is recommended that the reinforcement be distributed uniformly within the cross section of
the tie. It is noted that the centroid of the reinforcement must coincide with the axis of the
tie in the model. In practice, it is assumed that the concrete surrounding the reinforcement
does not carry tension; the strength of the tie is given by the strength of the reinforcement
alone, which is often assumed to be at yield.
Nodal Zones
The intersection between the axes of two or more struts and ties defines the nodes in the
model. The region surrounding the nodes is called a nodal zone. To satisfy equilibrium in
a planar structural model, at least three forces must intersect at a node. Depending on the
type of forces acting at a node intersection (compression or tension), nodes are often clas-
sified as C-C-C nodes when they resist forces from three or more struts or as C-C-T nodes
when two or more struts and a tie intersect at the node. A third type of node is a C-T-T node,
where two ties intersect a strut.
Preliminary dimensions of a nodal zone may be determined from the estimated size of
the struts, ties, and the bearing area under the applied loads or reactions, such as bearing
plates or column base dimensions. These dimensions, however, may need to be revised if
the strength of the nodal zone is exceeded at one or more of its faces.
If the faces of the nodal zone are perpendicular to the axes of the struts, the stresses
will be the same on all faces and equal to the in-plane principal stress. In such a case, the
dimensions of the nodal zone are in the same proportion as the acting forces, as shown in
Fig. 14.1.6(a). Such nodal zones are often called hydrostatic nodal zones, though in real-
ity the state of stresses is not truly hydrostatic because the out-of-plane stress in a planar
model is zero. A node anchoring one or more ties may also be considered as “hydrostatic”
by treating the tie force as a compressive force acting on the far side of the nodal zone, as
shown in Fig. 14.1.6(b).
Laying out a model with hydrostatic nodal zones to meet the strength requirements in all
faces can be cumbersome in practice. In a simplified approach to sizing the nodal region,
Schlaich and Schäfer [14.7] have suggested that the faces of the nodal zone need not be
perpendicular to the axis of the strut (or tie), as shown in Fig.14.1.7. In such a case, the node
region is simply defined by the width of the struts and ties intersecting at the node. Unlike
hydrostatic nodes, both normal and shear stresses will act on the faces of the nodal zone.
Schlaich and Schäfer [14.7] have also suggested that the nodal zone be considered safe if
the stresses acting on the cross-sectional area taken perpendicular to the strut (or tie) axis
are below the nodal zone stress limit. Note that the normal stresses acting on the faces of the
Tie
T width
Figure 14.1.5 Tie consisting of several layers of reinforcement plus a portion of surrounding

concrete.
532
C1 C1
d1
C1 C2 C3 T C2
= =
d1 d2 d3
C3 C1
d3
Nodal Nodal
zone zone
C3
d2
C1 C2 C2
C2
T
(a) C–C–C node (b) C–C–T node
Figure 14.1.6 Example of hydrostatic nodal zones: (a) C-C-C node; (b) C-C-T node.
d3
C3
fc3
fc2
d2
C2
fc1
C1
d1
Figure 14.1.7 Simplified arrangement of a C-C-C node. (After M. Schlaich and Schäfer [14.7].)
nodal zone need not be equal in this case. This approach can be particularly useful for nodal
zones with four or more struts (or ties) intersecting at a node, as shown in Fig. 14.1.8(a).
In another approach, though somewhat more laborious, M. Schlaich and Anagnostou
[14.8] have shown that the nodal zone can be treated by using a combination of hydrostatic
“subnodal” zones and “transition stress fields.” Figure 14.1.8(b), which exemplifies this
approach, shows the same nodal zone of Fig. 14.1.8(a), where the entire nodal zone has
been subdivided into two hydrostatic nodal zones and one short strut (or transition stress
field). In this situation, the stresses on all faces of the two “subnodal” zones and within the
short strut are the same.
If a nodal zone anchoring a tie is too small to develop the tie reinforcement within the
nodal zone, an anchor plate may be required [see Fig. 14.1.9(a)]. The tie reinforcement may
be developed, however, within the region defined by the intersection of the strut and tie
widths outside of the nodal zone, as shown by the dotted area in Fig. 14.1.9(b). This region,
referred to as the extended nodal zone in the ACI Code, may be used to compute the avail-
able development length for the tie reinforcement (ACI-R23.2.6). It is assumed that within
the extended nodal zone the compression stresses due to the reactions and the struts help
53
transfer the forces from the tie to the node. In some situations, however, the development
length computed with the inclusion of the extended nodal zone may still be insufficient to
anchor the tie. Hooks, mechanical anchorage, and/or a reduction in bar diameter will be
needed in such cases.
The strength of a nodal zone will depend on its shape and the types of elements fram-
ing into the node. In a “hydrostatic” C-C-C node [see Fig. 14.1.8(b)], the nodal zone is
subjected to a biaxial or triaxial state of compressive stresses, a rather favorable condition
for the node. On the other hand, if a node anchors one or more ties, the tension stresses in
the tie reinforcement will weaken the node, and so a lower strength of the nodal zone can
be expected. Very limited experimental data on the strength of nodal zones exist (Jirsa et al.
[14.9]). Suggested values range from about 0.7 to 0.85 fc′ for C-C-C nodes and from 0.52
to 0.6 fc′ for C-T-T nodes (MacGregor [14.6]).
Figure 14.1.8 Example of nodal zone with four struts: (a) simplified arrangement;

(b) “hydrostatic” nodal zones and a “transition stress field.”
534
C
(a)
Extended
nodal zone
Nodal zone
hn
2
T
hn
2
La
C
(b)
Figure 14.1.9 Anchorage of tie reinforcement: (a) anchored by a plate; (b) anchored by bond.
ACI Code Provisions

Chapter 23 of the 2014 ACI Code contains the provisions for the design of D-regions using
strut-and-tie models. The basic design approach consists of ensuring that the selected strut-
and-tie model (i.e., the idealized truss) is capable of transferring the loads to the supports
and to the adjacent B-regions. Accordingly, the strut-and-tie model must satisfy the follow-
ing main requirements.
1. The strut-and-tie model must be in equilibrium with the applied factored loads and
the reactions (ACI-23.2.4).
2. Struts shall not cross or overlap each other except at nodal zones (ACI-23.2.6). If
struts overlap each other, then a portion of the struts would be overstressed. The
strength of the strut is calculated based on its geometry (e.g., prismatic or bottle-
shaped), the concrete strength, and the reinforcement crossing the strut.
3. Ties shall be permitted to cross struts or other ties (ACI-23.2.5). A tie crossing a
strut will induce tensile strains in the transverse direction, which will reduce the strut
strength. Thus, the ACI Code reduces the strength of struts crossed by ties or when-
ever the compressive stresses are transferred across cracks in a tension zone.
4. The angle between the axis of a strut and that of a tie at a node shall not be taken less
than 25° (ACI-23.2.7). This requirement is meant to avoid incompatibility resulting
from shortening of a strut and lengthening of a tie in nearly the same direction.
53
5. The design strength of each strut, tie, and nodal zone in the strut-and-tie model must
satisfy the following (ACI-23.3.1):
φFns ≥ Fus (Struts) (14.1.1)
φFnt ≥ Fut (Ties) (14.1.2)
φFnn ≥ Fun (Nodal zones) (14.1.3)
where
Fns = nominal compressive strength of a strut
Fnt = nominal tenssile strength of a tie
Fnn = nominal strength on the face of a nodal zone
Fus = factored compressive force in a strut
Fut = factored tensile force in a tie
Fun = factored force on the face of a node
φ = strength reduction factor, taken equal to 0.75 for struts, ties, nodal zones, and
bearing areas designeed in accordance with the strut-and-tie method (ACI-21.2)
Strength of Struts, Fns
The nominal strength Fns of a strut is taken as the smaller of the strengths computed at the
two ends of the strut as follows [ACI-23.4.1(a)]:
Fns = Acs fce (14.1.4)
where Acs is the area of the strut at one end and fce is the effective compressive strength of
the strut. As mentioned earlier, the compressive strength of a strut will depend on its shape
and the presence of reinforcement (if any) perpendicular to the strut axis. In the ACI Code,
the effective compressive strength of a strut is computed as (ACI-23.4.3)
fce = 0.85β s fc′ (14.1.5)
where βs is a factor that accounts for the effect of cracking and crack-control reinforcement,
as shown in Table14.1.1. For example, βs is taken as 1.0 for a prism of uniform cross sec-
tion, where the compression stress field can be considered parallel and constant over the
length. This results in fce = 0.85 fc′, which is consistent with the strength of the rectangular
stress block in the compression zone of a beam according to the ACI Code.
TABLE 14.1.1 VALUES OF βS FOR COMPUTING STRUT STRENGTH

ACCORDING TO ACI-23.4.3
Strut Type βs
Prism of uniform cross section over its length (parallel stress field) 1.0
Bottle-shaped struts:
with reinforcement satisfying ACI-23.5 0.75
without reinforcement satisfying ACI-23.5 0.60λa
Struts in tension members or in the tension flanges of members 0.40
All other cases 0.60λa
a λ = modification factor to account for the unit weight of concrete (see Section 1.8).
536
For bottle-shaped struts, βs is reduced to 0.6λ because of the tendency to split the
strut as the stresses spread out from the strut end toward the midlength of the strut [see
Fig. 14.1.4(a)]. The parameter λ is the modification factor that accounts for the unit weight
of the concrete as shown in the table.
If a bottle-shaped strut is crossed by reinforcement that resists the transverse tensile
stresses, a value of 0.75 may be used for βs (see Table 14.1.1) However, such an increase
in strut strength is permitted only when the reinforcement is provided in accordance with
ACI-23.5. This section allows designers to use a strut-and-tie model for the bottle-shaped
strut to compute the required amount of reinforcement crossing the strut by assuming that
the compressive forces spread out at a 2:1 slope, as shown earlier [see Fig. 14.1.4(b)].
Alternatively, for fc′ ≤ 6000 psi, the required amount of reinforcement crossing the strut
may be computed to satisfy Eq. (14.1.6) [ACI Formula (25.5.3)] as follows:
Asi
∑bs sin α i ≥ 0.003

(14.1.6)
s i
where Asi is the total area of reinforcement at a spacing si of a layer of reinforcement at an

angle αi with the axis of the strut, and bs is the strut thickness. ACI-23.5.3.1 allows the rein-
forcement to be provided in two orthogonal directions, as shown in Fig. 14.1.10, or in one
direction. In the latter case, the angle between the reinforcement layer and the strut axis,
α1, shall be at least 40°
For struts crossing the tension region of a member, such as the flanges of inverted
T-sections, βs is reduced to 0.4 to account for the reduced strenth of struts carrying com-
pressive stresses across a cracked tension zone. For all other cases, the ACI Code specifies
a βs value of 0.6λ. Such cases include, for example, fan-shaped struts [Fig. 14.1.3(a)] and
inclined struts either parallel to cracks [Fig. 14.1.11(a)] or crossed by cracks [Fig. 14.1.11(b)]
in the compression field of a beam web.
The strength of a strut may be increased by using compression reinforcement along the
length of the strut (ACI-23.4.1). Such a reinforcement must be parallel to the strut axis and
located within the strut width. In addition, the reinforcement must be properly anchored and
enclosed by ties, hoops, or spirals in accordance with the detailing requirements in ACI-
23.6.3. The strength of a strut with compression reinforcement is computed as [ACI-23.4.1(b)]
Fns = Acs fce + As′ fs′ (14.1.7)
where As′ is the area of compression reinforcement along the length of the strut and fs′ is
the stress in the compression reinforcement at the nominal axial strength of the strut. For
Grades 40 and 60 reinforcement, fs′ may be taken equal to fy.
Axis of
strut
Strut
boundary Strut boundary
α1 As1
s2
α2
As2
s1
Figure 14.1.10 Details of reinforcement crossing a strut according to ACI-23.5.3.1. (From ACI-

318 Code and Commentary, 2014.)
537
Strut Cracks
Tie
Strut
(a) Struts in a beam web with inclined
cracks parallel to struts
Strut Cracks
Tie
(b) Struts crossed by skew cracks
Figure 14.1.11 Struts in the compression field of a beam web: (a) struts parallel to cracks;

(b) struts crossed by cracks. (From ACI 318-05 Code and Commentary, 2005.)
Strength of Nodal Zones, Fnn

The nominal strength of a nodal zone is taken as (ACI-23.9.1)
Fnn = Anz fce (14.1.8)
where fce is the effective compressive strength of concrete at a face of a nodal zone and Anz
is the smaller of
a. the area of the face of the nodal zone taken perpendicular to the line of action of the
strut or tie force, or
b. the area of a section through the nodal zone taken perpendicular to the line of action
of the resultant force. This case may be encountered when more than three forces
meet at a node, as shown in Fig. 14.1.12. In the figure, the struts acting on faces AD
and DC may be replaced by the resultant acting on face AC.
The effective compressive strength fce in the nodal zone is computed as (ACI-23.9.2)
fce = 0.85β n fc′ (14.1.9)
where βn is a factor that accounts for the type of node as shown in Table 14.1.2. A node
bounded only by struts (a C-C-C node) represents the most favorable stress condition for
the node and thus is assigned βn = 1.0 and fce = 0.85 fc′. On the other hand, nodes anchoring
one or more ties are assigned a lower strength to reflect the reduction in the capacity of the
nodal zone caused by the tensile stresses induced by the ties.
TABLE 14.1.2 VALUES OF βn FOR CALCULATING THE STRENGTH OF THE

NODAL ZONE ACCORDING TO ACI-23.9.2
Nodal Zone Type βn
Bounded by struts, bearing areas, or both (C-C-C nodes) 1.0

Anchoring one tie (C-C-T nodes) 0.80
Anchoring two or more ties (C-T-T nodes or bounded only by ties) 0.60
538
B C
Nodal
zone
A
Node
B C
Figure 14.1.12 Example of nodal zone with four struts. Struts acting on faces AD and DC may be
resolved into a single strut acting on face AC to check the nodal zone.
Strength of Ties, Fnt
Ties may consist of nonprestressed or prestressed reinforcement, or both. Accordingly, the
nominal strength of a tie, Fnt, is computed as follows (ACI-23.7.2):
Fnt = Ats f y + Atp ( fse + ∆f p ) (14.1.10a)
where
Ats = area of nonprestressed reinforcement in a tie
fy = yield strength of nonprestressed reinforcement
Atp = area of prestressing reinforcement in a tie
fse = effective stress in the prestressing reinforcement after losses1
Δfp = increase in stress in the prestressing steel (see Chapter 20) due to factored loads.
Unless justified by analyses, the ACI Code permits Δfp to be taken as 60,000 psi
for bonded tendons and 10,000 psi for unbonded tendons. In no case, however, can
the sum (fse + Δfp) exceed the yield strength fpy of the prestressing reinforcement.
When tie reinforcement consists of only nonprestressed bars, Eq. (14.1.10a) reduces to
Fnt = Ats f y (14.1.10b)
Tie reinforcement must be anchored by mechanical devices, post-tensioning anchoring

devices, standard hooks, or straight bar development (ACI-23.8.2). When straight bars are
used, the available development length La (i.e., the length available to anchor the bar) is
measured from the point of intersection of the centroid of the bars in the tie and the exten-
sion of the outlines of either the strut or the bearing area [ACI-23.8.3(b)], as shown in
Fig. 14.1.9(b)]. Also, note that the bar may be developed by extending the reinforcement
beyond the nodal zone if enough room is available. If La is insufficient to anchor the bar, the
reinforcement may be anchored using 90° hooks or mechanical anchors, such as an anchor
plate as shown in Fig. 14.1.9(a). If 90° hooks are provided, the hooks should be confined by
reinforcement to avoid splitting of the concrete within the anchorage region.
1 Prestressed concrete requirements are treated in Chapter 20.

539
For nodes with ties on opposite faces, tie reinforcement must be developed within the
nodal zone for the difference between the tie force on one face and the tie force on the
opposite face [ACI-23.8.3(a)].
Selecting a Strut-and-Tie Model
The first and most important task is the selection of a statically admissible truss. Since
a strut-and-tie model represents a lower-bound solution of the theory of plasticity, more
than one admissible truss (solution) may exist, provided equilibrium and strength require-
ments are satisfied. Although all admissible solutions will be safe if the above conditions
are met, some are preferable to others. Since ties are generally more flexible than struts, a
model with the minimum number and shortest ties is preferred [14.7]. Figure 14.1.13 shows
examples of preferred and undesirable models for a simply supported beam with a uniform-
ily distributed load. The model in Fig. 14.1.13(b) is not only more complex, but also has
longer ties than the simpler model shown in Fig. 14.1.13(a). Furthermore, the model of
Fig. 14.1.13(b) has struts crossing each other, a condition that is not permitted by the ACI
Code (ACI-23.2.6).
The truss geometry may be selected by visualizing the stress fields that develop within
the structure under the load combination being considered. For isolated elements or
simple structures, such as simple supported beams, the flow of stresses may be easily
visualized, as illustrated earlier (see Fig. 14.1.1). For more complex structural systems,
however, the selection of a suitable truss may be considerably more difficult, even for
an experienced design engineer. In such cases, Schlaich et al. [14.1] recommend that
the model be based on the principal stress trajectories obtained from linear elastic finite
element analyses. This approach can be particularly useful in design because the need to
consider multiple load factors and load combinations can result in a number of different
strut-and-tie models. Figures 14.1.14 through 14.1.17 show various strut-and-tie models
for elements of different types. In addition to finite element analyses, several computer-
based tools have been developed to assist the engineer in the selection of an appropri-
ate and optimal layout for the strut-and-tie model (Ali and White [14.10], Alshegeir
and Ramirez [14.11], Liang, Uy, and Steven [14.12], Tjhin and Kuchma [14.13], and
Herranz et al. [14.14]).
h=L h=L
L L
(a) Preferred (b) Undesirable

(not permitted by the ACI Code)
Figure 14.1.13 Strut-and-tie models for a deep beam with uniformly distributed load. (After
Schlaich and Schäfer [14.7].)
540
Uniform
field Strut
Fan
Tie
(a) Simply supported beam (b) Dapped end beam on corbel
(c) Deep beam (d) Wall with concentrated loads
Figure 14.1.14 Examples of strut-and-tie models. (Adapted from Cook and Mitchell [14.16].)
(b)
(a)
(c)
(d)
Figure 14.1.15 Strut-and-tie action: (a) corbel; (b) knee joint under closing moment; (c) knee

joint under opening moment; (d) interior beam-column joint. (Adapted from Marti [14.2].)
541
Figure 14.1.16 Strut-and-tie action in shear wall with openings. (Adapted from Marti [14.2].)
Strut
Tie
(a) Pile cap supported on four piles (b) Simple three-dimensional truss model
for a four-pile cap
Figure 14.1.17 Strut-and-tie model for pile cap. (From Adebar, Kuchma, and Collins [14.15].)
Dimensioning Struts, Ties, and Nodal Zones

Once a suitable model (i.e., one that satisfies equilibrium) has been selected, the strength of
its elements (struts, ties, and nodal zones) must be checked. The size (width and thickness)
of the struts, ties, and nodal zones will depend on the forces acting on the member and the
geometry of members meeting at a node. The designer should realize that dimensioning of
the struts, ties, and nodal zones is an iterative procedure and, in many cases, a revision of
the selected strut-and-tie model or the truss geometry will be required.
In a planar model, the out-of-plane dimension of the truss elements (strut, tie, and nodal
zone) is often taken as the thickness of the member, which reduces the task to simply deter-
mining the width of the element. The size of the struts can be initially defined in proportion
to the magnitude of the force in the element using the strength for each strut type [i.e., struts
of uniform cross sections (prisms) or bottle-shaped struts]. In many instances, however,
the size of the strut may have to be increased at one or both ends to reduce the stresses and
542
prevent failure of the adjoining nodal zones. Under externally applied loads or the reactions
at supports, bearing strength requirements are often used to determine the width of one
or more faces of the nodal zone at those locations. Additionally, the force resultants from
adjacent B-regions, for example, the depth of the compression zone within the B-region of
a beam, can be used (see Fig. 14.1.5).
Ties may also be initially sized in proportion to the force in the element. Calulations
of the strength of the tie require only the total area of reinforcement and its yield strength
[assuming that only nonprestressed reinforcement is used; see Eq. (14.1.10b)].
To size and check the strength of the nodal zones at the ends of a tie, the cross-
sectional area of the tie, including a portion of the surrounding concrete, is needed (see
Fig. 14.1.5). When the tie reinforcement is anchored with a bearing plate behind the joint
[see Fig. 14.1.9(a)], the area of the tie is often taken as the size of the plate required to
meet the bearing strength requirement for the nodal zone on that face. On the other hand,
when the tie reinforcement is anchored by hooks or straight bar development, a hypo-
thetical bearing area behind the joint must be defined [see Fig. 14.1.9(b)]. When a single
layer of reinforcement is provided, the ACI Code Commentary recommends that the width
of the tie be taken as the diameter of the bars plus twice the clear cover to the surface of
the bars [ACI-R23.8.1(a)]. If more than one layer of reinforcement is required, the ACI
Code Commentary recommends that the tie reinforcement be uniformly distributed over
the width and thickness of the tie. The writers recommend that bar sizes be selected so
that the maximum spacing between bars within the tie does not exceed more than two bar
diameters and that the tie area be taken as the perimeter defined by the outermost layer of
reinforcement plus the cover to the surface of the bar.
14.2 DEEP BEAMS
When the shear span-to-depth ratio (a/d) is lower than about 2, a simply supported beam
tends to behave like a tied arch as shown in Fig. 5.4.5(a). In such beams, often referred to
as deep beams, the region between the support and the concentrated load must be treated as
a D-region (“disturbed region”) [14.1], since the assumption of ordinary beam theory that
plane sections remain plane does not apply.
The ACI Code defines a deep beam as a member with a ratio of clear span, Ln, to overall
member depth, h, equal to or less than 4 [ACI-9.9.1.1(a)] or as the region of a beam where
a concentrated load exists within a distance equal to twice the member depth, h, from the
face of the support [ACI-9.9.1.1(b)]. Note that a concentrated load may be located near a
support, creating a deep beam (D-region) between the load and the nearer support, but an
ordinary beam (B-region) between the load and the more distant support. In this case, the
provisions for deep beams would apply to the beam region between the nearer support and
the concentrated load if the concentrated load is located within a distance of 2 times the
member depth, h [ACI-9.9.1.1(b)].
It is noted that the provisions for deep beams in ACI-9.9 apply only to “members that
are loaded on one face and supported on the opposite face such that strut-like compression
elements can develop between the loads and the supports.” If the loads, whether uniformly
distributed or concentrated, are applied through the bottom or the sides of the member, the
reinforcement should be designed by using strut-and-tie models to transfer the loads to the
top of the member and distribute them to the supports.
The design of deep beams may be done by (1) conducting nonlinear analyses that account
for the nonlinear distribution of the longitudinal strains over the depth of the member (ACI-
9.9.1.2) or (2) using strut-and-tie models in accordance with Chapter 23 of the ACI Code
(ACI-9.9.1.3). The latter are based on the strut-and-tie approach, presented in general
by J. Schlaich et al. [14.1] and in detail for deep beams by Rogowsky and MacGregor
[14.3], Adebar, Kuchma, and Collins [14.15], Cook and Mitchell [14.16], M. Schlaich and
Anagnostou [14.8], Siao [14.17, 14.18, 14.20], Walraven and Lehwalter [14.19], Foster and
Gilbert [14.21], Wight and Parra-Montesinos [14.22].
The ACI Code also requires minimum beam cross-sectional dimensions to “control
cracking under service loads and to guard against diagonal compression failures …”
543
14.2 DEEP BEAMS 543
(ACI-R9.9.2.1) by limiting the maximum design shear stress to φ10 fc′(psi). Accordingly,
minimum beam dimensions are computed so that (ACI-9.9.2.1)
Vu ≤ φ10 fc′ bw d (14.2.1)

with φ = 0.75 for shear.
Minimum Reinforcement
The minimum amount of flexural tension reinforcement for deep beams is the same as
that required for ordinary beams (see Section 3.7). In addition, a minimum amount of both
vertical and horizontal reinforcement distributed along the side faces of deep beams is
required. This minimum reinforcement is intended to control the growth of inclined cracks
and is to be provided irrespective of the method used for the design of the member.
The minimum amount of vertical (i.e., perpendicular to the beam axis) reinforcement,
Av, required by ACI-9.9.3.1(a) is 0.0025bws, where bw is the beam width and s is the spacing
of the vertical reinforcement. Similarly, the minimum required amount of horizontal rein-
forcement, Avh, is 0.0025bws2 where s2 is the spacing of the horizontal reinforcement [ACI-
9.9.3.1(b)]. Neither the vertical nor the horizontal spacing of the reinforcement should
exceed the lesser of d/5 and 12 in. (ACI-9.9.4.3).
EXAMPLE 14.2.1
Design the flexural and shear reinforcement for a simply supported beam that carries
two concentrated service live loads of 126 kips each on a clear span of 12 ft, as shown in
Fig. 14.2.1. The beam has a width of 14 in. and an overall depth of 42 in.; it is supported
on 14 by 16 in. columns. Use fc′ = 3500 psi and fy = 60,000 psi.
Service loads
k
126 126k
4’–0” 4’–0” #3 horizontal bars
42”
Tie width
3” 23 @ 6” = 11–6” 3” 14”
#3 stirrups
16” 12’–0” 16”
Figure 14.2.1 Deep beam of Example 14.2.1.
SOLUTION
(a) Determine whether the provisions for deep beams of ACI-9.9.1 are applicable.
The beam is loaded on one face and supported on the other with a concentrated
load at 4(12)/42 = 1.14 times the member depth. Therefore, the provisions of
ACI-9.9.1 apply.
(b) Verify minimum beam cross-sectional dimensions per ACI-9.9.2.1. Neglecting the
self-weight of the beam, which is small in comparison to the applied concentrated
loads, the maximum factored shear is

Vu = 1.6(126) = 201.6 kips
(Continued)
54
Assuming an effective depth d = 0.8h, the maximum permitted factored shear force from
Eq. (14.2.1) is
φ10 fc′ bw d = 0.75(10) ( )

3500 (14)(0.8)42
1
1000
= 208.7 kips [> Vu = 201.6 kips] OK
(c) Select strut-and-tie model and geometry. Assume a strut-and-tie model consisting of
a single truss as shown in Figure 14.2.2. The horizontal location of the truss nodes
(joints) may be assumed to coincide with the line of action of the concentrated loads
(nodes B and C) and the centerline at the supports (nodes A and D). The width and
location of strut BC and the tie AD are unknown at this stage, but they can be ini-
tially estimated from flexural considerations at midspan. Neglecting the self-weight,
the factored maximum moment at midspan is
 1
Mu = 1.6(126)  4 + 8  = 941 ft-kips
 12 
18” 18”
d e
13.5”
hBC
s
2
B C
c f
A α Tie α D
htAD 16.8”
2
a b
8” 8”
Figure 14.2.2 Strut-and-tie model for the beam of Example 14.2.1.
Strength of Strut BC

The strength of a strut will be given by the smaller of the strengths computed at the
two ends of the strut. Because of symmetry, the strength of strut BC at ends B and C
will be the same. The effective compressive stress of the concrete in a strut is given by
Eq. (14.1.5)
fce = 0.85β s fc′
Since strut BC is located in the compression zone of the beam, it is considered as a

prism of uniform cross section over its length (parallel stress field); thus βs = 1 (see
Table 14.1.1).
Therefore, the effective compressive stress of strut BC is

fce = 0.85(1) fc′ = 0.85 fc′
Assuming a strut with no compression reinforcement, the design strength of strut BC is

φFnsBC = φ fce Acs = 0.75(0.85) fc′ bhsBC
where b is the strut thickness (equal to the beam width), hsBC is the width of strut BC (see
Fig. 14.2.2), and φ is taken as 0.75 per ACI-21.2.
(Continued)
54
14.2 DEEP BEAMS 545
Strength of Tie AD

The strength of the tie will be governed by the strength of the nodes at its ends or by
the yield strength of the tie reinforcement. Usually, however, the strength of the nodal
zones anchoring the tie controls, and thus it will be used here to size the width of the tie.
Nodes A and D anchor one tie, and thus, they are classified as C-C-T nodes. Their nom-
inal strength is given by Eq. (14.1.8) as

FnnA = FnnD = fce Anz = 0.85β n fc′Anz
where βn = 0.80 for C-C-T nodes (see Table 14.1.2), and Anz is the area of the face of the
nodal zone perpendicular to the line of action of the tie. Thus, the design strength of node A is

φFnnA = 0.75(0.85)(0.80) fc′ bhtAD
where htAD is the tie width (see Fig. 14.2.2).
Using the expressions for the strengths of the strut and the tie obtained above, compute
the minimum required widths of strut BC and tie AD based on flexural strength require-
ments. Since C = T, then

C = φFnsBC = 0.75(0.85) fc′ bhsBC = 0.75(0.85)(0.80) fc′ bhtAD = T
and

hsBC = 0.8htAD
Since

φ M n = (C or T )(arm )
then
 h BC h AD 
φ M n = 0.75(0.85)(3.5)(14)hsBC  42 − s − t 
 2 2 
or
 h BC h BC 
φ M n = 31.24hsBC  42 − s − s 
 2 2(0.8) 
Then, equating the design strength and the factored moment gives

φ M n = 31.2 4 hsBC (42 − 1.125hsBC ) = 941 (12) = Mu
therefore

hsBC = 13.45 in.
and
hsBC 13.45
hsAD = = = 16.81 in.
0.8 0.8
Therefore, the required force in strut BC and tie AD is
Mu 941(12)
FusBC = FutAD = = = 421 kips
arm  13.5 16.81
 42 − − 
2 2 
(Continued)
546
(d) Check angle between diagonal struts AB and CD and tie axes, and compute forces
in diagonal struts AB and CD. Knowing the widths of strut BC and tie AD, and the
line of action of the applied loads and the reactions, the locations of the nodes can
be established. The angle α between the diagonal struts and the tie is
 13.45 16.81 
42 − −
 2 2  = 25.6° > 25°(min, per ACI-23.2.7)
α = arctan   OK
56
 
 
Therefore, the force in diagonal struts AB and CD is

421
FusAB = FusCD = = 467 kips
cos(25.6)
(e) Check strength of nodal zones. Since the strut-and-tie model was not built with
hydrostatic nodal zones (i.e., zones with equal stresses on each face), the strength of
each face of the nodal zone must be checked separately.
Nodal Zone A (same as Nodal Zone D)

This node anchors one tie and thus is classified as a C-C-T node, as noted earlier. The
nominal strength of the nodal zone is

FnnA = 0.85β n fc′Anz
with βn = 0.80 (see Table 14.1.2).
Face a-c. The strength of the nodal zone at the far end of the tie (face a-b) is already sat-
isfied, since its width htAD was sized to satisfy the strength of the nodal zone.
Face a-b. On this face,

Anz = b × hcol
where b is the node thickness (equal to the beam width) and hcol is the width of the sup-
port (i.e., equal to the column depth). Thus,

Anz = 14(16) = 224 sq in.
and

φ FnnA = 0.75(0.85)(0.80)(3.5)224 = 400 kips
The reaction at the supports is

RA = 1.6(126) = 201.6 kips < [φ FnnA = 400 kips] OK
Face b-c. On this face,

Anz = b × ds
where b is the node thickness (equal to the beam width) and ds is the width at end A
of strut AB taken perpendicular to the line of action of the strut force (see Fig. 14.2.3).
From the geometry of the truss and dimensions of the nodal zone,

ds = hn cos α + hcol sin α
(Continued)
547
14.2 DEEP BEAMS 547
where hn = htAD = 16.8 in. and hcol = 16 in.; thus

ds = 16.8 cos (25.6) + 16 sin (25.6) = 22.1 in.
and
A
φFnn = 0.75(0.85)(0.80)(3.5)(14)22.1 = 552 kips > [ 467 kips = FusAB ] OK
Therefore, nodal zones A and D are adequate.
ds
16.8”
2 α
hn Centroid of tie
16.8”
2
a b
hcol = 16”
1.5”
La
Figure 14.2.3 Nodal zone at end A of strut AB of beam of Example 14.2.1.
Nodal Zone B (same as Nodal Zone C)

This nodal zone is bounded by struts AD and BC, and the bearing area under the con-
centrated loads. Thus, it is classified as a C-C-C node. The nominal strength of the nodal
zone is

FnnB = 0.85β n fc′Auz
with βn = 1.0 (see Table 14.1.2).
Face e-f. On this face, the strength should be adequate since the design strength of the
nodal zone and strut BC are the same. Check
B
φFnn = 0.75(0.85)(1.0)(3.5)(14)13.5 = 422 kips ≈ [ 421 kips = FusBC ] OK
Face d-e. Underneath the concentrated load, the nodal zone strength requirements will
be satisfied by providing a bearing plate of length  B, such that

φFnnB = 0.75(0.85)(1.0)(3.5)(14) B > 201.6 kips
or

 B ≥ 6.5 in
Face d-f. The width taken perpendicular to the line of action of the force in strut AB is

ds = 13.5 cos (25.6) + 6.5 sin (25.6) = 15 in.
(Continued)
548
and
φFnnB = 0.75(0.85)(1.0)(3.5)(14)15 = 469 kips > [ 467 kips = FusAB ]
OK
Therefore, nodal zones B and C are adequate, provided a bearing plate 6.5 in. long or
longer is used underneath the concentrated loads.
(f) Check strength of diagonal strut AB (same as strut CD). The nominal compressive
strength of the struts is

FnsAB = fce Acs
where Acs is the smaller of the areas at the ends of the strut and

fce = 0.85β s fc′
where the value βs depends on the strut type according to Table 14.1.1. Strut AB can be
considered as a bottle-shaped strut because there is room for the stresses to spread out
from the nodal zones toward midlength of the strut. For simplicity, however, the strut is
idealized as a tapered strut in Fig. 14.2.2. Assuming that the web will be provided with
reinforcement as required by ACI-23.5 (this will be checked later), βs may be taken as
0.75 (see Table 14.1.1). Therefore, the effective compressive strength of the strut is

fce = 0.85 (0.75) 3.5 = 2.23 ksi
At end A, the width ds taken perpendicular to the line of action of the strut was computed
earlier in part (e) as 22.1 in. Thus, the design strength of the strut at end A is
AB
φFns = 0.75(2.23)(14)22.1 = 517 kips > [ 467 kips = FusAB ] OK
At end B, the strut width ds taken perpendicular to the line of action of the strut is 15 in.
when a 6.5-in. long bearing plate is provided under the concentrated loads. Thus, the
design strength of the strut at end B is
AB
φFns = 0.75(2.23)(14)15 = 351 kips < [ 467 kips = FusAB ] NOT OK
This means that the strut width is too small and that the stresses will be too high at end
B. To reduce those stresses, the design strength of the strut must be increased by (a) add-
ing compression reinforcement parallel to the strut axis (impractical in this case) or
(b) increasing the bearing area under the concentrated load, if possible, to increase the
strut area. The latter approach will be followed in this example. The required width of
the strut at end B is
467
ds = (15) = 19.9 in.
351
which results in a required bearing length at nodes B and C of
19.9 − (13.5) cos(25.6)

B =  C = = 17.9 in.
sin (25.6)
Earlier it was shown that the nodal zones at ends B and C satisfied the strength require-
ments with a smaller bearing plate. Clearly, a larger plate will be satisfactory. Therefore,
use a bearing plate 14 in. wide by 18 in. long below each concentrated load.
(Continued)
549
14.2 DEEP BEAMS 549
(g) Compute the required area of steel and select bars for the tie. The required area for
tie AD is
FutAD 421
Ats ≥ = = 9.36 sq in
φ f y 0.75(60)
Select 12–#8 bars, Ats = 9.48 sq in.

Since fc′ = 3500 psi < 4444 psi, then the minimum required flexural reinforcement
ratio will be given by
200 200
min ρ = = = 0.0033
fy 60, 000
or
min As = 0.0033 (14 ) 33.6 = 1.55 sq in. < [ 9.48 sq in. = Ats provided ] OK
Use 12–#8 bars. These bars are to be uniformly distributed over the tie width 16.8 in.,
as shown in Fig. 14.2.1.
(h) Check anchorage of tie AD. According to ACI-R23.8.2, the available development
length is defined as the extension of the bearing area or the assumed outlines of the
struts, whichever is larger. In this example, the larger of these two values is given by
the outlines of the strut (see Fig. 14.2.3). Assuming a 1.5-in. cover to the reinforce-
ment at the far end of the tie, the available length La to anchor the tie reinforcement is
16.8 / 2
La = + 16 − 1.5 = 32 in.
tan(25.6)
It may be shown that the available length, La, is insufficient to develop a straight #8 bar,
and thus hooked bars are required. Using bars terminating in a 90° hook, the required
anchorage length is [see Eq. (6.10.1) in Chapter 6]
Ldh =   db [6.10.1]
 50 λ fc′ 
but not less than 8db nor less than 6 in. For uncoated bars, ψe is equal to 1.0. Similarly,
λ is equal to 1.0 for normal-weight concrete. Since the clear cover is less than 2 in. and
assuming that confinement reinforcement is not provided, the modification factors ψc
and ψr are both equal to 1.0. Thus, for a #8 bar
 60, 000(1.0)(1.0)1.0 
Ldh =  1.0 = 20.3 in. < [32 in. = La ] OK
 50(1.0) 3500 
Should the available length La be insufficient, the bars could be anchored by using addi-
tional layers of smaller bars. While an end plate at the exterior face of the beam could be
provided, it may not be practical to do so for the12–#8 bars in this case.
(i) Minimum distributed reinforcement. In accordance with ACI-9.9.3.1, a minimum
amount of vertical and horizontal reinforcement must be provided along the side
faces of deep beams. The minimum required amount of horizontal reinforcement,
Avh, per ACI-9.9.3.1(a) is
Avh
≥ 0.0025(14) = 0.035 in.
s2
(Continued)
50
Using 2–#3 bars (one in each face), the required spacing is
2(0.11)
s2 ≤ = 6.3 in.
0.035
This reinforcement will be uniformly distributed above the tie reinforcement.
Check maximum spacing in accordance to ACI-9.9.4.3:
d 33.6
max s2 ≤ = = 6.72 in. Controls
5 5
≤ 12 in.
To meet the required spacing requirement of 6.3 in. over a depth of (42 –16.8) =
25.2 in., use four layers of #3 @ 6 in., as shown in Fig. 14.2.1.
Similarly, the minimum required amount of vertical reinforcement,  Av, per
ACI-9.9.3.2(b) is
Av
≥ 0.0025(14) = 0.035 in.
s
Using #3 U stirrups, the required spacing is
2(0.11)
s≤ = 6.3 in.
0.035
which is less than the maximum permitted spacing of 6.72 in. (maximum spacing
requirements are the same for both the vertical and the horizontal reinforcement).
Use #3 U stirrups @ 6 in. for both the horizontal and vertical reinforcement, as shown
in Fig. 14.2.1.
(j) Check requirements for reinforcement crossing bottle-shaped struts. The diagonal
struts AB and CD were sized as bottle-shaped struts assuming that they would be
provided with reinforcement crossing the struts in accordance with the require-
ments of ACI-23.5 [see item (f)]. The required amount of this reinforcement may
be computed by using the strut-and-tie model for bottle-shaped struts shown in
Fig. 14.1.4(b) [ACI-23.5.1] or, when fc′ ≤ 6000 psi, by satisfying Eq. (14.1.6) [ACI-
23.5.3]. The latter approach is used in this example.
In accordance with ACI-23.5.3.1, the reinforcement crossing the strut may be provided
in two orthogonal directions. Thus, use the grid of horizontal and vertical reinforcement
computed in item (i) to verify whether such reinforcement is sufficient to satisfy the
requirements of ACI-23.5. If not, additional horizontal and vertical reinforcement would
have to be provided. Using Eq. (14.1.6) with reference to Fig. 14.1.10 and noting that
As1 = Av and s1 = s (i.e., stirrup spacing )

As 2 = Avh and bs = strut thickness (i.e., beam width)
Then
Asi 0.22 0.22

∑ sin α i = sin(90 − 25.6) + sin(25.6) = 0.00035 > 0.003 OK
bs si 14(6) 14(6)
Therefore, the provided reinforcement along the side faces of the beam satisfies the
requirements of ACI- 23.5 for bottle-
shaped struts; no additional reinforcement is
required. The proposed reinforcement layout for the beam is shown in Fig. 14.2.1.
51
14.2 DEEP BEAMS 551
EXAMPLE 14.2.2
The continuous beam of Fig. 14.2.4 is to support a uniformly distributed, factored load of

23 k/ft from heavy machinery. Design the flexural and shear reinforcement for the beam.
The factored moment and shear diagrams are given in Fig. 14.2.4. Use fc′ = 3500 psi and
fy = 60,000 psi.
23 k/ft
#4 stirrups #3 stirrups #4 stirrups

3–#7
2.5”
26”
#3 horizontal bars
3” 3” 2.5”
1 2–#7 14”
4 2 in. spacing
15” Ln = 8’–0” 15”

(a)
92k
+
Vu
(b) –
Critical section
face of support
39 ft-kips
+
Mu
– –
145 ft-kips
(c)
Figure 14.2.4 Continuous deep beam of Example 14.2.2.
SOLUTION
(a) Compute the ratio of beam span to overall depth.
Ln 8(12)
= = 3.69 < 4
h 26
Therefore, the provisions of ACI-9.9 for deep beams apply.
(b) Check minimum beam cross- sectional dimensions (ACI-
9.9.2.1). Assuming
an effective depth d = 0.8h, the maximum permitted factored shear force from
Eq. (14.2.1) is
1

φ10 fc′ bw d = 0.75(10)( 3500 )(14)(0.8)26 = 129.2 kips [ > Vu = 92 kips] OK
1000
(c) Select top and bottom reinforcement based on flexural considerations. Assuming a
cover dimension to the center of the reinforcement of 2.5 in., the required Rn at the
beam ends is
145(12)
Rn = 1000 = 300 psi
0.75(14)(23.5)2
(Continued)
52
where a φ factor of 0.75 has been used per ACI-21.2.1 for strut-and-tie models. Then
Equation (3.8.5) is used to estimate ρ as 0.0053. Since fc′ = 3500 psi < 4444 psi, then
200 200
min ρ = = = 0.0033 < 0.0053 OK
fy 60, 000
Thus,

required As = 0.0053(14)23.5 = 1.7 sq.in.
Choose 3–#7 bars, As = 1.8 sq in.

Check flexural strength
1.8(60)
a= = 2.59 in.
0.85(3.5)14

 2.59  1
φ M n = 0.75(1.8)(60)  23.5 −  = 150 ft-kips > 145 ft-kips OK
 2  12
At midspan, the required Rn is
39(12)
Rn = 1000 = 81 psi
0.75(14)(23.5)2
From Fig. 3.8.1, minimum reinforcement controls. Thus,

As = 0.0033(14)23.5 = 1.09 sq in.
Choose 2–#7 bars, As = 1.20 sq in.
Check flexural strength,
1.2(60)
a= = 1.73 in.
0.85(3.5)14

 1.73  1
φ M n = 0.75(1.2)(60)  23.5 −  = 102 ft-kips > 39 ft-kips OK
 2  12
(d) Select strut-and-tie model and geometry. Figure 14.2.5 shows a truss model for the
beam. Parallel chords through the centroid of the top and bottom longitudinal reinforce-
ment at a distance of (26 –2.5 –2.5) = 21 in. are assumed.2 The uniformly distributed
load is applied as a nodal force using the tributary width to each side of the node. The
end moment (145 ft-kips) is applied as an internal couple of 145(12)/21 = 82.9 kips.
(e) Check angle between the strut and tie axes. Table 14.2.1 shows the values for the
angle between the different struts and ties of the selected model. All values are
greater than 25°, thus satisfying ACI-23.2.7.
(f) Compute truss member forces and check strengths of struts, ties, and nodal zones.
With the geometry of the truss defined, all truss member forces are determined.
These forces are then used to size the truss elements and to check the design
strengths of the struts, ties, and the nodal zones. Table 14.2.2 shows the forces in all
truss members, as well as the effective compressive strength φ fce of the struts and
nodes, and the tie strength φ f y . For the horizontal struts in the compression zone
of the beam, a value of βs = 1.0 was used (see Table 14.1.1). Bottle-shaped struts
with reinforcement satisfying ACI-23.5 (βs = 0.75) were used for all diagonal struts.
(Continued)
2 This distance is less than the internal lever arms based on external strength calculations at the ends and at midspan. The
implications of this assumption are discussed later in this example.
53
14.2 DEEP BEAMS 553
23.0 k 25.9 k 28.8 k 28.8 k 28.8 k 25.9 k 23.0 k
0’–6” 1’–0” 1’–3” 1’–3” 1’–3” 1’–3” 1’–0” 0’–6”
82.9 k 82.9 k
A B C E F G
D
26” 21” 21”
82.9 k H I J K L M 82.9 k
92.0 k 92.0 k
1.5” 1.5”
Figure 14.2.5 Strut-and-tie model for the continuous beam of Example 14.2.2.
Except for nodal zones D, H, and M, which are bounded by struts or a bearing area
(βn = 1.0), all other nodal zones anchor two or more ties and, thus, βn = 0.6. The
last column of Table 14.2.2 shows the minimum required widths of the truss ele-
ments. The strut widths are based on the smaller of the effective strengths (shown
in boldface) computed for the strut itself or the nodal zones at the ends of the strut.
Note that all struts have at least one end connected to a node with two or more ties
(Fig. 14.2.5). Such nodes have βn = 0.6, whereas for the struts βs is either 1.0 or
0.75. Thus, the width of all struts is controlled by the strength of the nodal zones.
The minimum required width for the ties is thus based on the effective compressive
strength of the nodal zone. The strength φ f y of the tie itself is given by the tie rein-
forcement and has no bearing in the calculation of the tie width.
TABLE 14.2.1 ANGLES BETWEEN STRUTS

AND TIES FOR MODEL OF FIG. 14.2.5
Strut Tie Angle (degrees)
BH and FM AB and FG 47.1

BI and FL 42.9
CI and EL BC and EF 54.5
CJ and EK 35.5
DJ and DK JK 54.5
CJ and EK 35.5
To illustrate the steps involved in selecting the dimensions of the truss elements,
detailed calculations are shown below for nodal zone I (same as nodal zone L).
Nodal Zone I (same as nodal zone L)

Figure 14.2.6 shows a detailed view of nodal zone I where the minimum required widths of
the truss elements obtained from Table 14.2.2 have been increased to the nearest half-inch.
(Continued)
54
TABLE 14.2.2 FORCES, DESIGN STRENGTHS, AND MINIMUM REQUIRED WIDTHS

FOR THE STRUT-AND-TIE ELEMENTS OF EXAMPLE 14.2.2
Strut or Tie
Strength Strength of Nodal Zones
Member Force βs φ fcea or φfy Node βn φfcea Node βn φ fcea dmin

(kips) (ksi) (ksi) (ksi) (in.)
Tie AB (FG) 74.6 — 45.00 A (G) 0.6 1.34 B (F) 0.6 1.34 4.0
Tie BC (EF) 10.6 — 45.00 B (F) 0.6 1.34 C (E) 0.6 1.34 0.6
Strut CD (DE) 20.2 1.00 2.23 C (E) 0.6 1.34 D 1.0 2.23 1.1
Strut HI (LM) 10.6 1.00 2.23 H (M) 1.0 2.23 I (L) 0.6 1.34 0.6
Tie IJ (KL) 20.2 — 45.00 I (L) 0.6 1.34 J (K) 0.6 1.34 1.1
Tie JK 30.5 — 45.00 J 0.6 1.34 K 0.6 1.34 1.6
Strut AH (GM) 24.0 0.75 1.67 A (G) 0.6 1.34 H (M) 1.0 2.23 1.3
Strut BH (FM) 94.2 0.75 1.67 B (F) 0.6 1.34 H (M) 1.0 2.23 5.0
Strut CI (EL) 53.0 0.75 1.67 C (E) 0.6 1.34 I (L) 0.6 1.34 2.8
Strut DJ (DK) 17.7 0.75 1.67 D 1.0 2.23 J (K) 0.6 1.34 0.9
Tie BI (FL) 43.1 — 45.00 B (F) 0.6 1.34 I (L) 0.6 1.34 2.3
Tie CJ (EK) 14.4 — 45.00 C (E) 0.6 1.34 J (K) 0.6 1.34 0.8
a
φ fce = φ (0.85)β s f ′c for struts or φ (0.85)β n fc′ for nodal zones. φ = 0.75.
This node is intersected by two ties, and thus it is classified as a C-T-T node (βn = 0.6).
Its strength is given by
φFnnI = φ (0.85)β n fc′Anz

= 0.75(0.85)(0.6)3.5 Anz
= 1.34 Anz
It is noted that more than three forces act on this nodal zone. In the following, the
simplified approach proposed by Schlaich and Schäfer [14.7] for nodal zones (Section
14.1), where the stresses acting on the cross-sectional area taken perpendicular to the
strut or (tie) axis on each face, is used to verify the adequacy of the node. An alter-
native approach where the four forces are resolved into three intersecting forces is
presented later.
Face a-b. On this face, the nodal zone area taken perpendicular to the line of action of
the strut HI is

Anz = 1(14) = 14 sq in.
thus

φFnnI = 1.34 Anz = 1.34(14) = 18.8 kips > [10.6 kips = FusHI ] OK
(Continued)
5
14.2 DEEP BEAMS 555
43.1k
Tie BI Strut CI
53.0k
1
22”
e 3”
a d
Strut Hl Tie IJ
1
1” 12 ” 20.2k
10.6k
b
c 2.5”
Bottom of beam
Figure 14.2.6 Details of nodal zone I with minimum required widths from Table 14.2.2.
Face c-d. The nodal zone area taken perpendicular to the line of action of the tie IJ is

Anz = 1.5(14) = 21 sq in.
and
φFnnI = 1.34(21) = 28.1 kips > [20.2 kips = FutIJ ] OK
Face e-d. Nodal zone area taken perpendicular to the line of action of the strut CI is

Anz = 3(14) = 42 sq in.
and
φFnnI = 1.34(42) = 56.3 kips > [53.0 kips = FusCI ] OK
Face a-e. Nodal zone area taken perpendicular to the line of action of the tie strut BI is

Anz = 2.5(14) = 35 sq in.
and
φFnnI = 1.34(35) = 46.9 kips > [ 43.1 kips = FutBI ] OK
Similar calculations can be done for the rest of the nodes.

Note that the maximum horizontal force at nodes A, G, H, and M is due to the inter-
nal force of 82.9 kips (see Fig. 14.2.5). At A and G, the required depth of the nodal zone
is 82.9/[14(1.34)] = 4.2 in., which is less than the available width of 5 in. (i.e., twice the
assumed cover dimension to the tie axis). Similarly, at H and M, the required width is 82.9/
[14(2.23)] = 2.7 in., which can be easily accommodated within the available 5 in. The verti-
cal reaction at nodes H and M is assumed to be 1.5 in. from the inside column face to satisfy
the minimum width of 92/[14(2.23)] = 2.9 in. on the side of the node facing the column.
(Continued)
56
The dimensions of the struts, ties, and nodal zones based on the values computed in
Table 14.2.2 (increased to the nearest half-inch.) are shown as light gray shading in
Fig. 14.2.5. Note that all elements can be accommodated within the available dimensions
and geometry of the beam. Thus, the selected truss and elements can be considered adequate.
As mentioned earlier (Fig. 14.1.12), when four or more forces act on a node in a two-
dimensional model, it is possible to resolve some of the forces into three intersecting
forces. With reference to Fig. 14.2.7(a), the four forces acting on node I may be con-
veniently resolved into a vertical force from tie BI, a diagonal force from strut CI, and a
net horizontal force equal to the summation of the forces from strut HI and from tie IJ,
as shown in Fig. 14.2.7(b). In this case, the required node area perpendicular to the net
horizontal force is computed from

[ FunI = 30.8 kips] ≤ [φFnnI = 1.34 Anz ]
thus
30.8
Anz = = 23.0 in.
1.34
which results in a required width of 23.0/14 = 1.6 in. This value is slightly larger than that
computed earlier (i.e., somewhat more conservative), but it can be easily accommodated
within the available dimensions and assumed geometry, as before. Note that the required
minimum widths for the node faces perpendicular to tie BI and strut CI remain the same.
A similar analysis could be done for other nodes with more than three forces. In node
B, for example, the horizontal force acting on the node would be taken as the difference
between the forces in ties AB and BC or (74.6 –10.6) = 64.0 kips. This would result in
a minimum required width of 64.0/[1.34(14)] = 3.4 in., which is smaller than that given
in Table 14.2.2 for tie AB.
43.1k 43.1k
Strut CI Strut CI
Tie BI Tie BI
53.0 k 53.0 k
Strut HI Tie IJ (Tie IJ + Strut HI)
10.6 k 20.2 k 30.8 k
(a) (b)
Figure 14.2.7 Resolution of nodal forces into three intersecting forces for nodal zone I.
(g) Design the web reinforcement. In Table 14.2.2, the effective compressive strength
of the diagonal struts was computed using βs = 0.75, which requires that reinforce-
ment in accordance with ACI-23.5 be provided. Note that since the depth of all struts
was governed by the strength of the nodal zones with βn = 0.6, the strength and size
of the diagonal struts would conform to the requirements for bottle-shaped struts
without web reinforcement according to ACI-23.4. Thus, reinforcement according
to ACI-23.5 need not be strictly satisfied in this case. Vertical reinforcement must
be provided, however, to carry the forces in the vertical ties BI, CJ, EK, and FL. In
addition, ACI-9.9.3.1 requires that minimum vertical and horizontal reinforcement
be provided along the side faces of this deep beam.
Minimum Vertical Reinforcement.

According to ACI-9.9.3.1(a), the minimum required area of vertical reinforcement is

min Av ≥ 0.0025bw s
(Continued)
57
14.2 DEEP BEAMS 557
where
 d 23.5 
max s ≤  = = 4.7 in. Controls
5 5 
≤ 12 in.
Choose s = 4.5 in., then

min Av ≥ 0.0025(14)4.5 = 0.16 sq in.
Thus, #3 U stirrups @ 4 1 2 in. Av = 2(0.11) = 0.22 sq in., satisfy the minimum amount
required by ACI-9.9.3.1(a).
Forces in Vertical Ties

Ties CJ and EK. These ties must carry a force of 14.4 kips (see Table 14.2.2). Thus the
total required area of steel is
F 14.4 
Ats ≥  ut = = 0.32 sq in.
 φ f y 0.75(60) 
Assuming that this reinforcement is distributed over the adjacent panels and spaced at
4 1 2 in., the required amount of vertical steel is
4.5
Av = 0.32 = 0.096 sq in.< min Av
15
Use #3 U stirrups @ 4 1 2 in. in the center portion of the beam, about 2 ft 2 in. from the
face of the column.
Ties BI and FL. The total required area of steel for these ties is
43.1
Ast ≥ = 0.96 sq in.
0.75(60)
Assuming the reinforcement is also distributed at 4 1 2 in., the required vertical steel is
4.5
Av = 0.96 = 0.32 sq in. > min Av
13.5
Use #4 U stirrups @ 4 1 2 in., Av = 2(0.20) = 0.40 sq in., at the end portions of the beam
to about 2 ft 2 in. from the face of each column.
Minimum Horizontal Reinforcement

The minimum required area of horizontal reinforcement per ACI-9.9.3.1(b) is

min Avh ≥ 0.0025bw s2
where
max s2 ≤ 4.7 in. Controls

but not greater than 12 in.
(Continued)
58
Using #3 bars (in each face), the required spacing is
2(0.11)
s2 = = 6.3 in. > max s2
0.0025(14)
Thus, use #3 bars horizontally in each face spaced at 4 1 2 in. to satisfy max s2.
Minimum Reinforcement for Bottle-Shaped Struts According

to ACI-23.5.
Although not required in this example, it is instructive to compare the minimum rein-
forcement requirements of ACI-9.9.3.1(a) and 9.9.3.1(b) with those of ACI-23.5.3. For
struts CI, EL, DJ, and DK, ACI-23.5.3 would require
Asi
∑bs sin α i ≥ 0.003 [14.1.6]
s i
Using the minimum reinforcement provided in the center portion of the beam per
ACI-9.9.3.1,
2(0.11)
(sin 35.5 + sin 54.5) = 0.0049 ≥ 0.003 OK
14(4.5)
Therefore, the provisions of ACI-23.5.3 are already satisfied. It can be shown that the
reinforcement provided in the end regions of the beam also satisfies ACI-23.5.3.
(h) Check strength of horizontal ties.
The reinforcement for the horizontal ties was computed in part (c) based on flexural
considerations and should be adequate to carry the tie forces. It is assumed that the top
and bottom reinforcement will be as follows.
Bottom Longitudinal Reinforcement. The required area of steel for tie JK (the tie with
the largest force in the bottom of the beam) is
30.5
Ast ≥ = 0.68 sq in. < [1.20 sq in. = As for 2 − # 7] OK
0.75(60)
Although not required near the supports, this reinforcement will be extended into the
joints.
Top Longitudinal Reinforcement. Ties AB and FG carry the largest force in the top of
the beam. The required area of steel for these ties is
74.6
Ast ≥ = 1.66 sq in. < [1.80 sq in. = As for 3 − # 7] OK
0.75(60)
Note that this reinforcement will be extended into the joint; according to the strut-and-
tie model of Figure 14.2.5, it should be able to carry a force of 82.9 kips computed in
part (d). The required area of steel to carry this force is
82.9
Ast ≥ = 1.84 sq in. ≈ [1.80 sq in. = As for 3 − # 7]
0.75(60)
(Continued)
59
14.3 BRACKETS AND CORBELS 559
The required area of steel is somewhat larger than that provided because the internal
lever arm assumed in the model is smaller than that based on flexural strength calcula-
tions. It is possible to construct a model where the location of the tie and the internal
lever arm are consistent with those obtained from flexural considerations, as shown in
Fig. 14.2.8. Such a model requires a variable depth and location for the top and bottom
chords, a refinement that is hardly justified in practice. It can be shown that the forces in
the truss elements in the more refined model of Fig. 14.2.8 are nearly identical to those
of the parallel chord truss of Fig. 14.2.5, and thus the design based on the parallel chord
truss is adequate. The final design is shown in Fig. 14.2.4.
23.0 k 25.9 k 28.8 k 28.8 k 28.8 k 25.9 k 23.0 k
0’–6” 1’–0” 1’–3” 1’–3” 1’–3” 1’–3” 1’–0” 0’–6”

2.5”
78.4 k 78.4 k
B C D E F
A G
26” 22.6”
22.2” 22.2”
78.4 k H I J K L M 78.4 k
2.5”
92.0 k 92.0 k
1.5” 1.5”
Figure 14.2.8 Strut-and-tie model of variable depth for the continuous beam of Example 14.2.2.
14.3 BRACKETS AND CORBELS

Design of brackets and corbels in accordance to the provisions of ACI-16.5 was discussed
in Section 5.16. As mentioned in Chapter 5, the provisions of ACI-16.5 are primarily based
on test data and thus are limited to brackets and corbels with shear span-to-depth ratios
a/d ≤ 1.0 and with a factored horizontal tensile force Nu ≤ Vu. Alternatively, the ACI
Code allows brackets and corbels to be designed using the provisions for strut-and-tie
models (ACI-R16.5.1.1). Unlike the provisions of ACI-16.5, the strut-and-tie provisions
of Chapter 23 of the ACI Code can be used for brackets and corbels with any a/d and
Nuc /Vu ratios. This section presents the use of strut-and-tie models as a tool for the design
of brackets and corbels.
The flow of forces in a corbel subjected to a factored shear Vu and a horizontal force Nuc
may be visualized with the aid of the strut-and-tie model shown in Fig. 14.3.1. The shear
force Vu must be transferred into the column by a diagonal strut (AB). The vertical compo-
nent of strut AB must be balanced by the shear force Vu at node A and by the compression
resultant C in the column at node B. A horizontal tie (AC) is required to equilibrate the hor-
izontal force Nuc and the horizontal component of the strut AB. The moment generated by
the applied forces (Vu and Nuc) must be balanced by the internal couple (C-T) in the column.
Thus, a tie will be required on the opposite face of the column to carry the tension resultant
T. The force imbalance generated by the ties meeting at node C must be balanced by strut
CB. Finally, the horizontal shear caused by Nuc, resisted by shear reinforcement in the col-
umn, is represented by tie BD in the model.
560
Vu
Nuc A C
D
B
C T
Figure 14.3.1 Strut-and-tie model for a reinforced concrete corbel.
EXAMPLE 14.3.1
Redesign the bracket of Example 5.16.1 using the ACI Code provisions for strut-and-tie
models. Compare the final design with that obtained previously using the provisions of
ACI-16.5.
SOLUTION
(a) Factored loads.

Vu = 113 kips (from Example 5.16.1)
(b) Select a strut-and-tie model. Assume a model as shown in Fig. 14.3.2.
(c) Determine preliminary bracket size. The nodal zone beneath the bearing plate is a
C-C-T node; thus, its effective compressive stress is

fce = 0.85β n fc′
where βn = 0.8 (see Table 14.1.2).
Therefore,
113, 000
bearing plate width = = 3.17 in.
0.75(0.85)(0.8)(5000)14
where φ = 0.75 for strut-and-tie models.3
Use 3 1 2 in. for bearing plate width.
Allowing a tolerance gap of 1-in. clear between the face of the column and the beam for
possible overrun in beam length and also because the beam might be 1 in. too short, then
1
a = 2 + (bearing plate width ) = 2 + 1.75 = 3.75 in.
2
(Continued)
3 Note that although a larger φ factor is used here (compared to φ = 0.65 in Example 5.16.1), the strength of the nodal zone
is only 80% of that required by the bearing provisions of ACI-22.8.3.2. Thus, a slightly larger bearing plate width is required in
this example.
561
a = 3.75”
14” square column

113k
3½”
48k C 1.5”
A
8”
123
15”
k
81
k
65k
178k B
D E
2”
dsBD
C
T
Figure 14.3.2 Strut-and-tie model for the corbel of Example 14.3.1.
(d) Determine the depth of the bracket. Choose, say h = 15 in., so that d ≈ 13.5 in. (as before
in Example 5.16.1). Also, ACI-16.5.2.2 requires that the depth at the outer edge of the
bearing area should not be less than 0.5d. Select a depth of 8 in. for the outer face.
(e) Define the geometry of the strut-and-tie model. Using a 1-in. clear cover and one
layer of bars for tie AC, assume the center of the tie to be at 1.5 in. from the top.
Similarly, assuming a 1.5-in. clear cover for tie CE and one layer of bars, assume the
center of the tie is 2 in. from the face of the column.
The width dsBD of strut BD will be governed by the strength of the strut itself or by the
strength of the nodal zone B; therefore, the design strength is the smaller of

φFnsBD = φ (0.85)β s fc′Acs
or

φFnnB = φ (0.85)β n fc′Anz
where
φ = 0.75
Acs = Anz = area of the strut at end B taken perpendicular to the line of action of the strut
(Continued)
562
Strut BD is located in the compression zone of the column and may be considered
a prism of uniform cross s ection (parallel stress field). Thus, βs = 1 (Table 14.1.1). On
the other hand, nodal zone B is a C-C-C node and βn = 1 (Table 14.1.2). Therefore,
the strength at the end of the strut is the same as that on that face of the nodal zone.
Thus,
φFnsBD = 0.75(0.85(1)5 Acs = 3.19 Acs

= 3.19(14)dsBD = 44.6 dsBD
By taking moments about the axis through tie CE,
 d BD 
Mu = 113(3.75 + 14 − 2) = C  14 − 2 − s 
 2 

 d BD 
1780 = 44.6 dsBD  12 − s 
 2 
which leads to a quadratic equation for dsBD . Solving this equation gives

dsBD = 3.99 ≈ 4 in.
This defines the location of the resultant C and node B. The geometry of the truss is now
completely defined, and it can be analyzed.
(f) Check the angle between the strut and tie axes at nodes A and C.
Angle between Strut AB and Tie AC
 15 − 1.5 
α = arctan  = 67° > 25° (ACI − 23.2.7 minimum ) OK
 3.75 + 2 
Angle between Strut CE and Tie BC
 14 − 2 − 2 
α = arctan  = 36.5° > 25°
 15 − 1.5  OK
and
Angle between Strut AC and Tie BC

α = 90 − 36.5 = 53.5° > 25° OK
(g) Compute truss member forces and check strength of struts, ties, and nodal zones.
With the geometry of the truss defined, the truss member forces are computed
from equilibrium. The calculated forces in each member of the truss are shown in
Fig. 14.3.2. These forces must then be checked against the design strength of the
struts, ties, and the nodal zones, as follows.
Strut AB
Strut AB needs to carry a compression force FusAB of 123 kips. Its strength will be
governed by the strength of the strut itself or by the strength of the nodes at its ends.
(Continued)
563
Therefore, the minimum required width of the strut will be given by the condition that
FusAB does not exceed the smallest of the following:
φFnsAB = strength of the strut
φFnnA = strength of node A at thee end of the strut
φFnnB = strength of node B at the end of the strut
This strut may be assumed to be a bottle-shaped strut because there is room for the
stresses to spread out from the nodal zones toward midlength of the strut. Assuming that
the corbel will be provided with reinforcement according to ACI-23.5, βs may be taken
as 0.75 (Table 14.1.1). Thus, the design compressive strength of the strut is

φ fce = 0.75(0.85)(0.75)5 = 2.39 ksi
Nodal zone A is a C-C-T node and βn = 0.8, while nodal zone B is a C-C-C node and βn = 1.0
(Table 14.1.2). As a result, design compressive strengths in the nodal zones A and B are
φ fce = 0.75(0.85)(0.8)5 = 2.55 ksi at node A

φ fce = 0.75(0.85)(1.0)5 = 3.19 ksi at node B
AB
The minimum required width dmin for the strut is governed by the smallest of these val-
ues, which in this example is the strength of the strut itself, φfce = 2.39 ksi. The strut
AB
width dmin (taken perpendicular to the line of action of the strut) is then computed from
[ FusAB = 123 kips] ≤ [φ FnsAB = Acs φ fce = dmin

AB
(14)2.39]

AB
dmin ≥ 3.7 in.
From the geometry of the bracket and the chosen strut-and-tie model, one can verify that
the available space to accommodate strut AB at node B is

davailable = 2[2 cos(90 − 67)] = 3.7 in. = dmin
AB
OK
Similar calculations can be made for struts BC and BD, and for ties AC and CE. These
results are summarized in Table 14.3.1, which lists the design compressive strength
φ fce of the struts and nodes, and the design strength φ fy of the ties. The last column of
the table shows the minimum required depths for the members based on the smaller of
the design strengths (shown in boldface in the table) computed for the strut itself or the
nodal zones.4
Figure 14.3.3, which shows the minimum dimensions from Table 14.3.1 for the struts,
ties, and nodal zones for the model, illustrates that all can be accommodated within the
dimensions and geometry selected for the corbel.
(h) Compute the required area of steel and select bars for the ties.
Tie AC The required area for this tie is

FutAC 48
Ats ≥ = = 1.07 sq in.
φ f y 0.75(60)
Use 4–#5 (Ats = 1.24 sq in.)
(Continued)
4 The minimum depth shown for the ties ensures that the tie force can be carried safely in the nodal zone, and it is based on the
design compressive strength of the nodal zone. The design strength φ fy of the tie itself depends only on the tie reinforcement and is unre-
lated to this calculation. It is shown in Table 14.3.1 for completeness and will be used later to compute the required tie reinforcement.
564
This horizontal reinforcement must be anchored and extended as far as possible into
the column so that strut BC can be developed as assumed in the model. In practice, these
bars can often be developed using 90° hooks. Also note that these bars will fit all in one
layer at the assumed location of the tie axis in the model. If, for example, more than
one layer or larger diameter bars were required, resulting in a change in the tie axis, the
model would have to be revised.
Tie CE This tie force will be resisted by longitudinal reinforcement provided at the
far end of the column. As a minimum, however, this reinforcement amount should be
FutCE 65
Ats ≥ = = 1.44 sq in.
φ f y 0.75(60)
TABLE 14.3.1 FORCES, DESIGN STRENGTHS, AND MINIMUM REQUIRED WIDTHS

FOR THE STRUT-AND-TIE ELEMENTS IN EXAMPLE 14.3.1
Strut or Tie Strength Node Strength
Member Force βs φ fce or Node βn φ fce Node βn φ fce dmin

(kips) φ fy (ksi) (ksi) (ksi) (in.)
Strut AB 123 0.75 2.39 A 0.8 2.55 B 1.0 3.19 3.7
Strut BC 81 0.75 2.39 B 1.0 3.19 C 0.6 1.91 3.0
Strut BD 178 1.00 3.19 B 1.0 3.19 — — — 4.0
Tie AC 48 — 45.00 A 0.8 2.55 C 0.6 1.91 1.8
Tie CE 65 — 45.00 C 0.6 1.91 — — — 2.4
C
A 1.8”
3.7” 3”
2.4”
B
4”
D E
Figure 14.3.3 Minimum required widths for struts and ties of corbel of Example 14.3.1.
(Continued)
56
14.4 ADDITIONAL REMARKS 565
(i) Design closed stirrups or ties. In accordance with ACI-16.5.5.2, closed stirrups or ties
must be provided parallel to the main reinforcement (tie AC). From Eq. (5.16.20),

required Ah ≥ 0.5( Asc − An ) = 0.5(1.24 − 0) = 0.62 sq in.
Try 3–#3 closed hoops [Ah = 2(3)0.11 = 0.66 sq in.]. This reinforcement must be dis-
tributed within two-thirds of the effective depth [i.e., within 2(15 –1.5)/3 = 9 in. spaced
at 3 in.].
Since struts AB and BC were sized using a βs value of 0.75, the horizontal rein-
forcement must also satisfy ACI-23.5. The ACI Code allows designers to compute this
reinforcement amount using the strut-and-tie model for bottle-shaped struts shown in
Fig. 14.1.4. In this example, however, the required amount of reinforcement is computed
using Eq. (14.1.6),
Asi
∑bs sin α i ≥ 0.003 [14.1.6]
s i
where Asi is the total area of reinforcement at a spacing si of a layer of reinforcement at

an angle αi with the axis of the strut, and bs is the strut thickness. Since only horizontal
reinforcement is provided, ACI-23.5.3.1 requires that the angle αi be greater than 40°.
The angles between the struts AB and BC and the horizontal were computed earlier as
67° and 53.5°, respectively. Thus ACI-23.5.3.1 is satisfied. Using the horizontal rein-
forcement amount calculated above in accordance with ACI-16.5.5.2 and the smallest
angle of 53.5° for strut BC
Asi 2(0.11)
∑bs sin α i =
14(3)
sin 53.5 = 0.0042 > 0.003 OK
s i
Use 3–#3 @ 3 in. closed hoops.

( j) Final design. In this case, the dimensions and detailing of the bracket are essentially
the same as those designed using the traditional provisions of ACI-16.5 shown in
Fig. 5.16.7. The only difference is in the bearing plate size (3.5 in.) and location
(3.75 in. from the face of the column). This discrepancy is not important within the
accuracy and details of the methods. However, the strut-and-tie provisions are much
more versatile and can be applied to corbels of any size and with any Nuc / Vu ratio.
14.4 ADDITIONAL REMARKS
The strut-and-tie approach is a powerful tool that can be used for the design of reinforced
concrete members of virtually any shape under a myriad of loading conditions. The approach
is particularly useful, for example, for members with unusual shapes or with openings (see
Fig. 14.1.16). Even when alternative methods are used in design, strut-and-tie models are
very useful aids in visualizing the load paths within the member from the point of appli-
cation of the loads to the supports. As noted earlier, determining an appropriate geometry
for the model can be a time-consuming task, even for an experienced engineer. While sev-
eral computer-based tools have been developed [14.10–14.14], the reader is referred to
Refs. 14.6 and 14.23, which contain several examples for a variety of members using strut-
and-tie models. Additional examples may be found in Refs. 14.2 and 14.22.
56
SELECTED REFERENCES
14.1. Jörg Schlaich, Kurt Schäfer, and Mattias Jennewein. “Toward a Consistent Design of Structural
Concrete,” PCI Journal, 32, May–June 1987, 74–150.
14.2. Peter Marti. “Basic Tools of Reinforced Concrete Beam Design,” ACI Journal, Proceedings,
82, January–February 1985, 46–56.
14.3. D. M. Rogowsky and J. G. MacGregor. “Design of Reinforced Concrete Deep Beams,”
Concrete International, 8, August 1986, 49–58.
14.4. K. Bergmeister, J. E. Breen, and J. O. Jirsa. “Dimensioning of Nodal Zones and Anchorage
of Reinforcement,” Structural Concrete, International Association for Bridge and Structural
Engineering Colloquium on Structural Concrete, Stuttgart, Germany, 1991, 551–564.
14.5. Julio A. Ramirez and John E. Breen. “Evaluation of a Modified Truss-Model Approach for
Beams in Shear,” ACI Structural Journal, 88, September–October 1991, 562–571.
14.6. James G. MacGregor. “Derivation of Strut-and-Tie Models for the 2002 ACI Code,” Examples
for the Design of Structural Concrete with Strut-and-Tie Models, Karl-Heinz Reineck, Editor.
(SP-208). Farmington Hills, MI: American Concrete Institute, 2002 (pp. 7–40).
14.7. J. Schlaich and K. Schäfer. “Design and Detailing of Structural Concrete Using Strut-and-Tie
Models,” The Structural Engineer, 69, 6 (March 1991), 113–125.
14.8. Michael Schlaich and Georg Anagnostou. “Stress Fields for Nodes of Strut-and-Tie Models,”
Journal of Structural Engineer, ASCE, 116, 1 (January 1990), 13–23.
14.9. J. O. Jirsa, K. Bergmeister, R. Anderson, J. E. Breen, D. Barton, and H. Bouadi. “Experimental
Studies of Nodes in Strut-and-Tie Models,” Structural Concrete, International Association for
Bridge and Structural Engineering Colloquium on Structural Concrete, Stuttgart, Germany,
1991, 525–532.
14.10. Mohamed A. Ali and Richard N. White. “Automatic Generation of Truss Model for Optimal
Design of Reinforced Concrete Structures,” ACI Structural Journal, 98, July–August 2001,
431–442.
14.11. A. Alshegeir and J. A. Ramirez. “Computer Graphics in Detailing Strut-Tie Models,” Journal
of Computing in Civil Engineering, 6, 2 (April 1992), 220–232.
14.12. Qing Quan Liang, Brian Uy, and Grant P. Steven. “Performance-Based Optimization for Strut-
Tie Modeling of Structural Concrete,” Journal of Structural Engineering, ASCE, 128, 6 (June
2002), 815–823.
14.13. Tjen N. Tjhin and Daniel A. Kuchma. “Computer-Based Tools for Design by Strut-and-Tie
Method: Advances and Challenges,” ACI Structural Journal, 99, September–October 2002,
586–594.
14.14. Juan Pablo Herranz, Hernan Santa Maria, Sergio Gutierrez, and Rafael Riddell. “Optimal Strut-
and-Tie Models Using Full Homogenization of Optimization Method,” ACI Structural Journal,
109, September-October, 2012, 605–614.
14.15. Perry Adebar, Daniel Kuchma, and Michael P. Collins. “Strut-and-Tie Models for the Design of
Pile Caps: An Experimental Study,” ACI Structural Journal, 87, January–February 1990, 81–92.
14.16. William D. Cook and Denis Mitchell. “Studies of Disturbed Regions near Discontinuities in
Reinforced Concrete Members,” ACI Structural Journal, 85, March–April 1988, 206–216.
14.17. Wen Bin Siao. “Strut-and-Tie Model for Shear Behavior in Deep Beams and Pile Caps Failing
in Diagonal Tension,” ACI Structural Journal, 90, July–August 1993, 356–363.
14.18. Wen Bin Siao. “Shear Strength of Short Reinforced Concrete Walls, Corbels, and Deep Beams,”
ACI Structural Journal, 91, March–April 1994, 123–132.
14.19. Joost Walraven and Norbert Lehwalter, “Size Effects in Short Beams Loaded in Shear,” ACI
14.20. Wen Bin Siao, “Deep Beams Revisited,” ACI Structural Journal, 92, January–February 1995,
95–102.
14.21. Stephen J. Foster and R. Ian Gilbert. “The Design of Nonflexural Members with Normal and
High-Strength Concretes,” ACI Structural Journal, 93, January–February 1996, 3–10.
14.22. James K. Wight and Gustavo J. Parra-Montesinos. “Strut-and-Tie Model for Deep Beam
Design,” Concrete International, 25, May 2003, 63–70.
14.23. Further Examples for the Design of Structural Concrete with Strut-and-Tie Models (SP-
273). Karl-Heinz Reineck and Lawrence C. Novak, Editors. Farmington Hills, MI: American
Concrete Institute, 2010, 288 pp.
567
PROBLEMS 567
PROBLEMS
All problems are to be worked in accordance with the provisions of the ACI Code unless
otherwise indicated.
14.1. Redesign the beam of Example 14.2.1, but (c) Compute the minimum required distrib-
instead of using a single diagonal strut, use a par- uted vertical and horizontal reinforcement
allel chord, with two panel trusses between the between the concentrated load and the
concentrated loads and the supports, as shown in support.
the figure for Problem 14.1. Compare your solu- 14.3. Design the reinforcement for the beam shown
tion with that presented in Example 14.2.1 and in the figure for Problem 14.3. The rectangular
discuss. continuous beam is to support heavy machin-
14.2. The beam shown in the figure for Problem ery; it carries a uniformly distributed, factored
14.2 is to carry a uniform live load of 0.6 kip/ft load of 65 kips/ft. The factored moment Mu dia-
in addition to the concentrated load shown gram is given in the accompanying figure. Use
and the beam weight. Use fc′ = 4000 psi and fc′ = 4000 psi and fy = 60,000 psi.
fy = 50,000 psi. 14.4. Redesign the bracket of Problem 5.21 in
(a) Construct an acceptable strut-and-tie model Chapter 5 using the provisions of Chapter 23
for the beam in accordance with Chapter 23 in the ACI Code for strut-and-tie models. The
of the ACI Code. bracket projects from one side of a 16 × 16 col-
(b) Using the model constructed in part (a), umn to support a vertical load of 35 kips dead
verify the adequacy of the longitudinal rein- load and 65 kips live load. Assume that suita-
forcement provided in the bottom of the ble bearings are provided so that horizontal res-
beam. If inadequate, prescribe the number traint is eliminated. The reaction is located 5
and location of the required longitudinal in. from the column face. Use = 5000 psi and
bars. Also, check the bar anchorage. fy = 60,000 psi.
2’– 4” 2’– 4” 4’ 2’– 4” 2’– 4”

Center Center
of of
column column
Problem 14.1
Column load: 20k (LL) 30k (DL)

5’–0” (12 in. sq column)
d = 26”
30” 9 –#9
12” support Symmetrical

about CL
of span 16”
10’–0
Problem 14.2
568
50”
24”
18” 18”
12’–0”
symmetric
+
Mu
– –
900 ft-kips 900 ft-kips
Problem 14.3
14.5. Repeat Problem 14.4 if the reaction is from a is welded to the bracket. Creep, shrinkage, and
restrained beam that induces a horizontal ten- temperature effects on the restrained girder
sion equal to 50% of the total gravity reaction. induce a horizontal force of 50 kips (unfactored)
14.6. Redesign the bracket (corbel) of Example 14.3.1 on the bracket.
considering that the supported prestressed girder
CHAPTER 15
STRUCTURAL WALLS
15.1 GENERAL
Reinforced concrete walls are common structural elements used in buildings and non-
building structures to carry both vertical (gravity) and lateral loads. When properly designed
and judiciously situated within the structure, walls can be very effective to resist a large
portion of the loads (gravity or lateral load) and to augment the lateral stiffness of the struc-
ture. In practice, walls may be classified according to their intended function within the
structure as follows (see Fig. 15.1.1):
1. Nonbearing walls. These walls are designed to carry no loads other than their own
weight and are usually supported by other elements within the structure. Nonbearing
walls, also called curtain walls, are often used as partitions to divide and separate
space in the interior of the structure, as well as for the building façade.
2. Bearing walls. A bearing wall is used primarily to bear, or support, gravity loads,
such as the weight of floors and other superimposed dead loads, as well as live loads.
3. Shear walls. The primary function of shear walls in a structure is to resist a large por-
tion of the total horizontal lateral load (e.g., due to wind or earthquake) by means of
in-plane shear and flexure.
4. Retaining walls. These walls are designed to serve as a barrier and to hold back
soil or other loose material. As such, they are designed to resist the lateral pressure
exerted on the wall through shear and flexure about their minor or weak bending axis.
15.2 MINIMUM WALL DIMENSIONS AND

REINFORCEMENT REQUIREMENTS—A CI CODE
Wall Thickness
Minimum design and detailing requirements for reinforced concrete walls are given in
Chapter 11 of the ACI Code. The minimum required wall thickness, h, is the smaller
between the unsupported length, ℓw, and the unsupported height ℓc, multiplied by either 1 30
for nonbearing walls or 1 25 for bearing walls, but not less than 4 in. (see Fig. 15.2.1). For
exterior basement and foundation walls, however, the minimum thickness is 7.5 in. These
minimum thickness requirements only apply to bearing walls and exterior basement and
foundation walls designed in accordance with the simplified method presented later (see
Section 15.4.) Thinner walls may be permitted if adequate strength and stability can be
shown by structural analysis (ACI-11.3.1.1).
570
570 C H A P T E R 1 5 S T R U C T U R A L W A L L S
Erection of core wall in Lincoln Square Expansion Project, Bellevue, WA. (Photo courtesy of Cary
Kopczynski & Co.)
Wall Reinforcement
Walls are required to have a minimum amount of uniformly distributed reinforcement in
both the longitudinal (vertical) and the transverse (horizontal) directions. Reinforcement
may consist of deformed bars or welded wire mesh placed parallel to the wall faces. The
minimum required amounts of reinforcement for cast-in-place walls when the in-plane shear
Vu ≤ 0.5φVc are shown for common grades of steel in Table 15.2.1. Minimum reinforcement
requirements when the in-plane shear Vu > 0.5φ Vc are presented later (see Section 15.5).
It is noted that the reinforcement ratios ρℓ and ρt in Table 15.2.1 are defined as the area
of steel reinforcement per gross area of concrete perpendicular to the direction of the rein-
forcement being considered.
Wall reinforcement may be provided in one or two layers depending on wall thickness.
A single layer of reinforcement is permitted for walls 10 in. thick or less. Walls with thick-
ness greater than 10 in. are required to have two layers of reinforcement placed parallel
with the wall faces, except basement walls and cantilever retaining walls.
571
y y
x x Gravity loads
(e.g., weight of above floors)
x x

y y
Wall self-weight Wall self-weight
Footing/foundation Footing/foundation
(a) Nonbearing wall (b) Bearing wall
y
y Lateral pressure
(e.g., from soil)
Lateral load x
x Axial loads
(e.g., from wind or x
(e.g., weight of
earthquake) x
above floors) My-y
y
Mx-x Wall self-weight
y
Vy-y Vx-x
Wall self-weight
Footing/foundation
Footing/foundation
15.2 DIMENSION AND REINFORCEMENT REQUIREMENTS
(c) Shear wall (d) Retaining wall
Figure 15.1.1 Forces and main actions on different types of walls: (a) nonbearing wall, (b) bearing wall, (c) shear wall, and (d) retaining wall.
571
572
Minimum Wall Thickness, h
Nonbearing walls:
ℓc Largest of:
h a) 4 in.
hw
1 ℓ
b) x smaller of ℓw
30 c
Bearing walls*:
Largest of:
ℓc
a) 4 in.
ℓw 1 ℓ
b) 25 x smaller of ℓw
c
Exterior basement and foundation walls*: 7.5 in.
* Applies only to walls designed according to the

simplified method.
Figure 15.2.1 Minimum wall thickness requirements (ACI 11.3.1.1).
TABLE 15.2.1 MINIMUM REINFORCEMENT FOR WALLS WHEN IN-PLANE

Vu IS 0.5φVc OR LESS (ACI-11.6.1)
Reinforcement
Steel Reinforcement Bar Size Minimum Minimum

Type and Grade Longitudinal, ρℓ Transverse, ρt
Deformed bars of #5 or smaller 0.0012 0.0020

Grades 60, 75 or 80 #6 or larger 0.0015 0.0025
Deformed bars of Any size 0.0015 0.0025
Grades 40 or 50
Welded-wire W31, D31 0.0012 0.0020
reinforcement of or smaller
any Grade
When two layers of reinforcement are used, wall reinforcement is commonly distributed
into equal amounts between the two layers. The ACI Code, however, permits up to two-
thirds of the total required amount in one layer, with the balance of the required amount in
that direction placed in the other layer (ACI-11.7.2.3). The layer with the larger amount of
reinforcement must be placed at least 2 in., but not more than h/3, from the exterior sur-
face. The other layer must be placed at least 3 4 in., but no more than h/3, from the interior
surface.
Additionally, minimum requirements of concrete cover for the reinforcement must be
satisfied. The minimum clear cover for cast-in-place walls not exposed to weather or in
contact with ground is 3 4 in. when #11 or smaller bars are used (ACI-20.6.1.3.1). This is
the most common situation for walls built above ground. For walls cast against and perma-
nently in contact with ground, such as retaining walls, the minimum concrete cover for any
bar size is 3 in. Concrete cover requirements for other bar sizes or exposure conditions are
given in ACI-20.6.1.3.1.
The spacing of the longitudinal (vertical) and transverse (horizontal) reinforcement in
cast-in-place walls is limited to 3 times the wall thickness or 18 in., unless shear reinforce-
ment is required for in-plane shear strength (ACI-11.7.2.1 and ACI-11.7.3.1). Spacing that
is required for shear strength is further limited to ℓw /3 for the longitudinal reinforcement
and to ℓw /5 for the transverse reinforcement (see Fig. 15.2.2).
573
1 5 . 4 D E S I G N O F B E A R I N G WA L L S 573
0.0012 (#5 or smaller bars)

min ρℓ =
0.0015 (#6 or larger bars)
shoriz ≤ 3h or 18 in.
svert ≤ 3h or 18 in.
0.0020 (#5 or smaller bars) ℓw

min ρt =
0.0025 (#6 or larger bars)
Notes:
i. A single layer may be provided when h < 10 in.
ii. Shear walls need also satisfy shoriz ≤ ℓw/3 and svert ≤ ℓw/5
when in-plane Vu > 0.5φVc.
Figure 15.2.2 Minimum requirements for Grade 60 reinforcement in walls.
15.3 DESIGN OF NONBEARING WALLS
Nonbearing walls are designed to carry no gravity loads other than their own weight. They
may be designed to carry some lateral loads (e.g., from soil pressure), in which case they
are designed as a retaining wall (treated in detail later: see Section 15.7). When used to
divide space or separate areas, as façades in buildings, nonbearing walls are supported
laterally by other members in the structure. Design of these walls is based primarily on
experience and good practice; usually it involves satisfying the minimum requirements for
reinforcement and wall dimensions.
15.4 DESIGN OF BEARING WALLS
Walls used to carry, in addition to their own weight, gravity loads such as the weight of
floors and other superimposed dead loads, as well as live loads, are referred to as bearing
walls. The strength of bearing walls may be calculated by using the same principles (i.e.,
strain compatibility and equilibrium) presented in Chapter 10 for members subjected to
combined bending and axial forces. Alternatively, for solid rectangular walls with no or
small out-of-plane bending, the ACI Code allows the use of a simplified design procedure
(ACI-11.5.3).
Simplified Design Method—ACI Code

The simplified method can be used only for solid walls of rectangular cross section and
only if the resultant of all factored axial loads is located within the middle third of the
wall thickness in the out-of-plane direction (ACI-11.5.3.1); that is, the eccentricity may
not be greater than h /6. Walls with nonrectangular cross section, with larger out-of-plane
eccentricity, or subjected to in-plane bending must be analyzed by using the procedures for
members subjected to combined bending and axial forces (Chapter 10). Furthermore, walls
574
subjected to significant in-plane shear forces must be designed as shear walls, as discussed
next, in Section 15.5.
Wall Design Strength

Since the eccentricity of the factored axial load acting on the wall must be within h /6 (i.e.,
the kern point), no tension will be induced in the wall under the design actions. The design
strength of the wall may then be simplified by computing a nominal axial compressive
strength, Pn, so that
φPn ≥ Pu [2.6.3]
where φ is the strength reduction factor for a compression-controlled section and Pu is the
factored axial force.
The nominal axial compressive strength, Pn, of the wall is computed by ACI Formula
(11.5.3.1) as
  kℓ  2

Pn = 0.55 fc′Ag 1 −  c   [15.4.1]
  32h  
where Ag is the gross concrete area of the wall, ℓc is the unsupported length of the wall (see
Fig. 15.2.1), and k is the effective length factor for the wall given in Table 15.4.1. The first
term in Eq. (15.4.1) is obtained by assuming a rectangular stress block over a depth equal
to two-thirds of the wall thickness, h, which results in a net axial compression resultant of
0.85 fc′(2 / 3h)ℓ w or 0.57fc′ hℓ w , where the coefficient is taken as 0.55. The term in brackets
accounts for slenderness effects on the axial load capacity. It is noted that Pn is independent
of the amount of steel reinforcement. This result is based on analyses of walls with no or
low eccentricity, which showed that the minimum amount reinforcement as required by
the ACI Code (Table 15.2.1) does not substantially contribute to wall strength [15.1]. Test
results as well as a comparison of test data with the simplified method of the ACI Code may
be found elsewhere [15.2].
TABLE 15.4.1 EFFECTIVE LENGTH FACTORS, k, FOR WALLS (ACI-11.5.3.2)
End Restraint Conditions k
Braced against lateral translation at top and bottom:

Restrained against rotation at one or both ends 0.8
Unrestrained against rotation at both ends 1.0
Not braced against lateral translation 2.0
EXAMPLE 15.4.1
A solid, rectangular, bearing wall with a length of 30 ft. is to be designed to carry a fac-
tored axial load, Pu, of 1350 kips with an eccentricity of 1.1 in. The wall supports a floor
system made of hollow-core slabs that may be assumed to restrain lateral movement at
the top of the wall. At the base, the wall is supported on a concrete footing that prevents
lateral translation. Wall height is 15 ft. Find the required wall thickness and the amount
of vertical and horizontal reinforcement. Use fc′ = 4000 psi and fy = 60,000 psi.
(Continued)
57
1 5 . 4 D E S I G N O F B E A R I N G WA L L S 575
SOLUTION
(a) Because this is a solid wall with a rectangular cross section subjected to an axial
force with a small eccentricity, it may be designed in accordance with the simplified
method of the ACI Code. In such a case, the minimum required thickness, h, of the
bearing wall is to be computed as the larger of:
a) h = 4 in.
1
b) h = × smaller of
25
the unsupported wall length, ℓ w = 30 ft = 360 in.
 the unsupported wall height, ℓ = 15 ft = 180 in.  Governs!
 c 
or
1
h= (180) = 7.2 in.
25
Therefore, a minimum thickness of 7.2 in. is required. Use h = 8 in.
(b) Compute the design strength, φ Pn
For the simplified method of the ACI Code to be applicable, the eccentricity of the fac-
tored axial load must be within h/6 = 1.3 in. which is greater than the eccentricity of 1.1
in. of the applied axial load. Therefore, the simplified method may be used.
From Eq. (15.4.1), Pn is computed as
  kℓ c  2 
Pn = 0.55 fc′Ag 1 −    [15.4.1]
  32h  
where Ag = hℓ w = 8(30 × 12) = 2880 sq in.

The unsupported wall height or the vertical distance between lateral supports, ℓc, is
15 ft. The wall is braced against lateral translation at the top and at the bottom, but it will
be assumed to be unrestrained against rotation. From Table 15.4.1, the effective length
factor, k, is taken as 1.0. Thus,
  1.0(15)12  2 
Pn = 0.55(4)2880 1 −    = 3204 kips
  32(8)  
Assume that the wall will be provided with rectilinear ties (no spirals), as is the case in
common practice. Thus, φ = 0.65 and

φPn = 0.65(3204) = 2083 kips > Pu = 1350 kips OK
(c) Select the vertical and horizontal reinforcement

Since the wall thickness is 8 in. < 10 in., the wall may be reinforced with a single layer
of reinforcement. In this example, a single layer of uniformly distributed reinforcement
will be provided. For this bearing wall, the maximum spacing of the vertical and the
horizontal reinforcement is the smaller of
• 3h = 3(8) = 24 in. and
• 18 in. Governs!
Using Grade 60, #5 or smaller bars, a minimum ratio of vertical reinforcement, ρℓ , of
0.0012 is required (Table 15.2.1). Also, a minimum amount of 0.0020 is required for the
horizontal reinforcement ratio ρt (Table 15.2.1).
(Continued)
576
Thus, the total required area of vertical reinforcement per foot of wall length is
As vert = ρℓ h × (12 in.)

As vert = 0.0012(8)12 = 0.12 sq in./ft
From Table 8.3.1, choose #4 bars at the maximum permitted spacing of 18 in. to provide
a total area of 0.13 sq in./ft.
Similarly, the total required area of horizontal reinforcement per foot of wall length is:
As horiz = ρt h × (12 in.)

As horiz = 0.0020(8)12 = 0.19 sq in./ft
From Table 8.3.1, choose #4 bars spaced at 12 in., which provide an average area of
0.20 sq in./ft.
(d) Check concrete cover requirements
The clear cover for a single layer of #4 bars for the horizontal and vertical reinforce-
ment centered in the 8 in. thick wall is
8 − [0.5 + 0.5]
cover = = 3.5 in. >> 3/4 in. required OK
2
Use an 8-in.-thick wall with one layer of reinforcement at the center of the wall made
with #4 vertical bars @ 18 in. and #4 horizontal bars @ 12 in.
15.5 DESIGN OF SHEAR WALLS
In addition to carrying gravity loads, walls may be used to carry a large portion of the lat-
eral load by in-plane shear and bending [Fig. 15.1.1(c)]; consequently, these are commonly
referred to as shear walls. This terminology, however, has often led to the misconception
that all shear walls exhibit a brittle failure in shear. This is not so. Shear walls can be
designed to be an excellent source of lateral stiffness and strength, and they can be effec-
tively used to control lateral drift and to resist lateral loads in multistory buildings.
Common wall configurations used in commercial and residential construction are shown
in Fig. 15.5.1. In many tall buildings, shear walls are designed to carry most if not all of the
lateral loads from wind and earthquakes. Gravity loads are often carried by a slab-column
frame system. These shear walls are commonly concentrated around elevator shafts and
stairwells, and thus they are often referred to as core walls [see Fig. 15.5.1(a)]. A combi-
nation of shear walls and moment-resisting frames, often referred to as a dual system, is
shown in Fig. 15.5.1(b). In these dual systems, which are commonly used in office build-
ings, lateral loads are shared by the shear walls and the frames. A common configuration
used in apartment buildings is shown in Fig. 15.5.1(c). In such cases, the walls are designed
to carry most if not all of the gravity loads in addition to resisting the lateral loads as
shear walls.
The amount of wall cross-sectional area with respect to the floor area, or wall-to-floor
area ratio, may vary significantly depending on the configuration of the floor plan [15.3,
15.4]. When shear walls are provided as part of a core wall system [Fig. 15.5.1(a)] or as
part of a dual system [Fig. 15.5.1(b)], the ratio of wall to floor area can be much lower
57
1 5 . 5 D E S I G N O F S H E A R WA L L S 577
(a) Shear wall system (reinforced concrete core)
(b) Shear wall and moment-resisting frame (dual system)
(c) Bearing and shear wall system
Figure 15.5.1 Common wall configurations in residential and commercial buildings.
than in buildings where walls form the primary gravity and lateral load resisting system
[Fig. 15.5.1(c)]. In general, walls in buildings with low ratios of wall to floor area (less
than about 0.5% of wall area in each direction) can be expected to be more heavily rein-
forced than those in buildings with high ratios (greater than 1.5% of wall area in each
direction).
Shear wall cross sections commonly found in practice are shown in Fig. 15.5.2. Walls of
rectangular cross section are the most common shape. Other shapes such as C-, I-, L-, and
T-shaped cross sections are also commonly used in building construction.
In walls with high flexural strength demands, such as in the lower stories of a high-rise
building, vertical reinforcement in addition to that required to be uniformly distributed is
often concentrated near the ends of the walls. These end regions are commonly referred
to as wall boundary elements (see Fig. 15.5.3). In some cases, the wall thickness may be
increased over the length of the boundary element region to accommodate the required
amount of vertical reinforcement at the ends. Another reason to provide an enlarged sec-
tion at the wall ends is to stabilize and to prevent buckling of the boundary element in
thin walls (Fig. 15.5.4). Walls with an enlarged boundary element region are referred to as
“barbell” walls.
In low-to medium-rise buildings, walls are often designed with a constant cross section
over the height; that is, they are prismatic. In such cases, variations in the required strength
with wall height are accommodated by varying the amount of vertical and horizontal rein-
forcement, and by varying the concrete strength. In taller structures, wall cross section may
578
Figure 15.5.2 Typical shear wall cross sections.
Boundary Boundary
element element
region region
Boundary Boundary
element element
Figure 15.5.3 Boundary element regions. (Note: Transverse reinforcement not shown

for clarity.)
be reduced in the upper stories by varying the wall thickness or the wall length. Changes
in cross section must be done with judgment and caution, without causing drastic changes
in stiffness or strength over the wall height, which might otherwise result in an undesired
failure at the transition level.
In practice, walls may be found as solid walls or with openings, usually to allow for
windows or doorways, or both. A detailed treatment of walls with openings is beyond the
scope of this chapter, but some basic guidelines and terminology are provided. The analy-
sis and design of walls with openings depend primarily on the size and arrangement of the
openings. In general, the influence of the openings may be neglected when the total area
of opening is relatively small in comparison to the wall surface area [see Fig. 15.5.5(a)]. In
such cases, the wall may be treated as a solid wall with special provisions for reinforcement
around the openings (ACI-11.7.5.1). In contrast, when the openings are relatively large
or are located near a critical region [see Fig. 15.5.5(b)], such as the boundary elements
at the wall ends, their effect on the wall flexural and shear strength must be considered.
The ACI Code offers little or no specific guidance for the design of walls with openings.
In most cases, however, walls with openings may be designed by using the strut-and-tie
method presented in Chapter 14. Tests on walls with openings are scarce [15.5–15.8].
While the studies cited focused primarily on the response of walls subjected to reversed
cyclic loads, they provide good insight into the general behavior and design of walls with
openings.
579
Figure 15.5.4 Wall buckling during the Maule, Chile, earthquake in 2010. Maule, Chile, 2010.
(Photo courtesy of Fernando Yañez, University of Chile.)
A common and efficient structural system for resisting lateral loads may be obtained
when openings are arranged in a regular pattern, by connecting two or more solid
walls with beams in every floor level, as shown in Fig. 15.5.6. Under lateral loads, the
beams will interact with and affect the behavior of the connected walls. Because the
behavior of the connected walls and that of the beams is coupled, the walls are com-
monly referred to as coupled walls. Accordingly, the connecting beams are referred to
as coupling beams.
The level of interaction or level of coupling depends on the relative flexibility and
strength of the beams. In a structure with stiff and strong coupling beams, the level of cou-
pling will be high. In such cases, the shear forces in the coupling beams will induce high
axial forces in the walls that must be considered in addition to those resulting from gravity
loads. For the direction of the lateral forces shown in Figure 15.5.6, summation of the cou-
pling beam shear forces will induce axial tension on the left wall and axial compression
on the right wall. Clearly, the reverse is true for lateral loads acting in the opposite direc-
tion. If the walls carry little or no gravity loads, each wall will have to be designed under
combined bending and axial tension and under combined bending and axial compression.
580
(a) (b)
Figure 15.5.5 Walls with (a) small and (b) large openings.
Figure 15.5.6 Coupled wall geometry and actions under lateral loads.
If the beams are relatively flexible and weak, the level of coupling will be low and may
be ignored in design. In those cases, each wall may be designed as an isolated cantilever,
solid wall.
Behavior and Design of Cantilever Walls

The basic failure mechanisms of cantilever shear walls as described by Paulay [15.9] are
shown in Figure 15.5.7. In general, the behavior and load transfer mechanism of cantilever
shear walls may be divided into two categories in accordance to their height-to-length, or
aspect, ratio (hw /ℓw). Walls with an aspect ratio hw /ℓw greater than 2 are often considered
tall walls. The behavior of cantilever tall walls has been under investigation for decades
[15.5, 15.10–15.11]. These walls exhibit a flexural-dominated behavior [Fig. 15.5.7(b)],
and thus they can be analyzed and designed by means of the same principles described for
members subjected to combined flexure and axial loads in Chapter 10. Shear failures of tall
walls involving diagonal tension or diagonal compression [Fig. 15.5.7(c)] or sliding shear
[Fig. 15.5.7(d)] should and can be avoided by means of adequate proportion and distribu-
tion of the transverse and longitudinal reinforcement.
581
V H
Nf
hv Vf
H
V Vf
N M V
T C Nf
(a) Wall actions (b) Flexure (c) Diagonal tension (d) Sliding shear
or compression
Figure 15.5.7 Basic failure mechanisms of cantilever walls. (Adapted from Paulay [15.9].)
Walls with aspect ratios (hw /ℓw) less than 2, often called squat walls, exhibit a different
load transfer and failure mechanism. Such walls, commonly found in low-rise construction,
can develop large flexural strength even when minimum amounts of vertical reinforcement
are provided. As a result, relatively large shear forces will be expected to act concurrently
with the development of their flexural strength [15.12]. Because of the low aspect ratio,
the behavior of squat walls is akin to that of deep beams, where the horizontal shear is
transferred primarily by diagonal struts in equilibrium with both the vertical and horizontal
reinforcement. Wood [15.13] has provided an excellent review of experimental results in
light of ACI Code design provisions for low-rise shear walls. Experimental results on the
behavior of low-rise squat walls are reported elsewhere [15.14–15.19].
Flexural Strength—Solid Rectangular Walls

In general, cantilever shear walls are subjected to shear forces and bending moments from
lateral forces and to axial forces from gravity loads. The flexural strength of shear walls
with aspect ratios hw /ℓw > 2 can be evaluated by using the same principles (i.e., strain com-
patibility and equilibrium) presented in Chapter 10 for members subjected to combined
bending and axial forces.
For walls of rectangular cross section with uniformly distributed vertical reinforcement,
Cardenas and Magura [15.20] have developed an approximate procedure to estimate flex-
ural strength. In their approach, it is assumed that the total area of vertical reinforcement is
continuous over the entire wall length, and that the strain distribution is such that the axial
load at ultimate is less than that at the balanced strain condition (see Fig. 15.5.8). By ignor-
ing the contribution of the reinforcement that remains elastic near the neutral axis (i.e.,
within βc in Fig. 15.5.8), Cardenas and Magura [15.20] proposed the following approxi-
mate equation:
 Pn   c
M n ≈ 0.5 As vert f y ℓ w  1 +  . 1 +  (15.5.1)
 As vert f y   ℓ w 
where
Mn = nominal flexural strength
As vert = total area of distributed vertical reinforcement
Pn = nominal axial force (taken as positive for compression)
c q+α
=
ℓ w 2q + 0.85β1 (15.5.2)
582
As vert f y ρℓ f y
q= =
hℓ w fc′ fc′ (15.5.3)
Pn
α = axial force ratio =
hℓ w fc′ (15.5.4)
Dividing Eq. (15.5.1) by hℓw 2 yields the following strength coefficient:
 
Mn  α   c
= 0.5ρℓ f y  1 +  . 1 −  (15.5.1a)
hℓw2  fy   ℓw 
 ρℓ f ′ 
c
Mn
For design purposes, the strength coefficient has been plotted in Fig. 15.5.9 as a
hℓw2
function of the reinforcement ratio, ρℓ, for Grade 60 steel and for fc′ equal to 4000 psi and
5000 psi, respectively. The curves are shown for axial load ratios, α, of 0, 0.05, 0.1, 0.15,
and 0.2. These charts can be used to quickly estimate the wall flexural strength for a given
reinforcing ratio, ρℓ, and axial load ratio (see Example 15.5.1), or to obtain a preliminary
estimate of the required amount of reinforcement for a required moment and axial load
(see Example 15.5.2). It must be emphasized that Eqs. (15.5.1) and (15.5.1a), as well as the
charts shown in Fig. 15.5.9, are limited to walls with an axial load smaller than that at the
balanced strain condition.
0.003 0.85fc’ ƒy
εy β1c 0.5β1c
c Cc
ϕu βc βc
As
βc βc
εy
ℓw
Pn
ℓw-c
ℓw/2
h ƒy
εy (c) Concrete stress (d) Steel stress
(a) Cross section (b) Strain distribution; β = distribution and distribution
0.003
axial load
Figure 15.5.8 Assumed strain and stress distributions at nominal strength in a rectangular wall
with uniformly distributed vertical reinforcement. (After Cardenas and Magura [15.20]).
583
0.8
0.75 Pn
α= = 0.20
0.7 hℓw f’c
0.15
0.65 0.10
0.05
0.6
0
0.55
0.5
Mn 0.45
hℓw2 0.4
(ksi) 0.35
0.3
0.25
0.2
0.15
f’c = 4000 psi
0.1 fy = 60,000 psi
0.05
0
0 0.005 0.01 0.015 0.02 0.025
0.8
Pn
0.75 α= = 0.20
hℓw f’c
0.15
0.7 0.10
0.65 0.05
0.6 0
0.55
0.5
Mn 0.45
hℓw2 0.4
(ksi) 0.35
0.3
0.25
0.2
0.15 f’c = 5000 psi
fy = 60,000 psi
0.1
0.05
0
0 0.005 0.01 0.015 0.02 0.025
ρℓ
Mn
Figure 15.5.9 Strength coefficient, , for rectangular walls with uniformly distributed
hℓw 2
vertical reinforcement as a function of the longitudinal reinforcement ratio, ρℓ , for axial
load ratios, α, of 0, 0.05, 0.1, 0.15 and 0.2. Charts are shown for fc′ = 4,000 psi (top chart)
and fc′ = 5,000 psi (bottom chart) and Grade 60 reinforcing bars.
584
EXAMPLE 15.5.1
A shear wall in a multistory building has cross-sectional dimensions and reinforce-

ment layout as shown in Fig. 15.5.10. The wall is 12 ft. long and 12 in. thick. The
vertical reinforcement consists of uniformly distributed #5 bars spaced at 15 1 2 in.,
while the horizontal reinforcement consists of #5 bars spaced at 18 in. Assume that the
required in-plane shear strength Vu < 0.5φVc . Compute the maximum factored moment
Mu that can be applied when the concurrent axial compressive force Pu is 500 kips. Use
fc′ = 4, 000 psi and fy = 60,000 psi.
#5 @ 18” (horiz.) #5 @ 15½” (vert.) 15½” ¾”(min)
12”
¾”(min)
12’ – 0”
Figure 15.5.10 Dimensions and reinforcement layout for wall cross section of Example 15.5.1.
SOLUTION
1. First, verify that the requirements for maximum spacing and minimum amount of
vertical and horizontal reinforcement are satisfied:
a) Since wall thickness of 12 in. is greater than 10 in., two layers of reinforcement
placed parallel with the wall faces are required as provided. OK
b) Clear cover of reinforcement is 3 4 in. This satisfies the minimum cover required
by ACI-20.6.1.3.1 for walls not exposed to weather or in contact with ground
when #11 bars or smaller are used (see Section 15.2). OK
c) Vertical (Longitudinal) Reinforcement:
i. Since #5, Grade 60 deformed bars are used, and Vu < 0.5φVc , the minimum
required amount of vertical (longitudinal) reinforcement, ρℓ min, is 0.0012 or
0.12% (see Table 15.2.1). For two layers of #5 bars spaced at 15 1 2 in., the
provided reinforcement ratio is
As vert 2(0.31)
provided ρℓ = = = 0.0033 > 0.0012 = ρℓ min OK
h shoriz 12(15.5)
ii. Per ACI 11.7.2.1, maximum spacing of vertical reinforcement is the smaller

of (see Fig. 15.2.2)
• 18 in. Governs!
or
• three times the wall thickness h, or 3(12) = 36 in.
The provided spacing of the reinforcement is 15 1 2 in. < 18 in. OK
Since Vu < 0.5φVc the limit based on ℓw / 3 need not be satisfied.
d) Horizontal (Transverse) Reinforcement:

i. Since #5, Grade 60 deformed bars are used, and Vu < 0.5φVc , the minimum
amount of horizontal (transverse) reinforcement, ρt min, is 0.0020 or 0.20%
(see Table 15.2.1). For two layers of #5 bars spaced at 18 in., the provided
reinforcement ratio is:
As horiz 2(0.31)
provided ρt = = = 0.0029 > 0.0020 = ρt min OK
h svert 12(18)
(Continued)
58
ii. Per ACI 11.7.3.1, maximum spacing of horizontal reinforcement is the smaller of:

• 18 in. Governs!
or,
• three times the wall thickness h, or 3(12) = 36 in.
The provided spacing of the reinforcement is 18 in., equal to the maximum
permitted, 18 in. OK
Since Vu < 0.5φ Vc , the limit based on ℓw /5 need not be satisfied.
2. Compute the nominal strength, Mn
Use the Cardenas and Magura approach [Eq. (15.5.1a)] to estimate the flexural strength
of the wall. Assuming a tension-controlled section (φ = 0.9), use Eq. (15.5.4) to compute
the axial load ratio α:
Pn P /φ 500 / 0.9
α= = u = = 0.080
hℓ w fc′ hℓ w fc′ 12(144)4
with ρℓ = 0.0033 and α = 0.080, enter the top chart ( fc′ = 4000 psi; fy = 60,000 psi) of
Fig. 15.5.9 to obtain by interpolation:
Mn
≈ 0.22 ksi
hℓw 2

 1
M n = 0.22(12)144 2   = 4562 ft-kips
 12 
Therefore,

Mu ≤ φ M n = 0.9(4562) = 4100 ft-kips
3. The maximum factored moment that can be applied on the cross section can also
be computed from the design axial load–moment interaction diagram. Analysis of
the wall cross section by using equilibrium and strain compatibility results in the
interaction diagram shown in Fig. 15.5.11. As shown in the figure, for an axial load
Pu of 500 kips, the design flexural strength, φ M n, is 4300 ft-kips. This value is about
5% larger than the factored moment obtained with the approximate method. In other
words, the approximate approach is somewhat conservative in this case, but it pro-
vides a very good estimate of the flexural strength of the wall. Note that the wall is
tension controlled, as assumed above in item 2.
Figure 15.5.11 Nominal and design strength interaction diagram for wall of Example 15.5.1.
586
EXAMPLE 15.5.2
A shear wall 10 in. thick and 10 ft long is to carry a factored moment Mu of 4500 ft-kips
with a concurrent axial force Pu of 900 kips. Find the required amount of uniformly
distributed longitudinal reinforcement to carry the required moment and axial force. Use
fc′ = 5000 psi and f y = 60,000 psi. Assume that Vu < 0.5φ Vc (i.e., shear reinforcement is
not required for in-plane strength).
SOLUTION
1. Use the charts of Fig. 15.5.9 to compute a first estimate of the required longitudinal
reinforcing ratio, ρℓ.
a. Compute the axial load ratio α [Eq. (15.5.4)] assuming the wall is tension controlled:
Pn P /φ 900
α= = u = = 0.17
hℓ w fc′ hℓ w fc′ 0.9(10)(120)5
Mn
b. Compute the required strength coefficient :
hℓw2
M n M u /φ 4500(12)
required = = = 0.42
hℓw2 hℓw 2 0.9(10)120 2
c. From Fig. 15.5.9 (bottom chart), estimate ρℓ ≈ 0.0055 or 0.55%.
2. Select reinforcing bars and spacing.

Since Vu < 0.5φ Vc, the minimum required amount of vertical reinforcement from
Table 15.2.1 is ρℓ = 0.0012 (assuming that #5 bars are to be used), which is less than
the required amount for strength of 0.0055.
The total required area of reinforcement per foot of wall length is:
As vert = ρℓ h × (12 in.)

As vert = 0.0055(10)12 = 0.66 sq in./ft
Since wall thickness is 10 in., two layers of reinforcement are required. Therefore, the
required amount of reinforcement per layer is
As vert = 0.66 / 2 = 0.33 sq in./ft
From Table 8.3.1, choose #5 bars spaced at 12 in. This amount, which will provide an
average area of 0.31 sq in./ft, is somewhat less than that required above. Given the con-
servative nature of Eq. (15.5.1), however, it should be adequate.
Maximum spacing requirements are
• 18 in. Governs!
• three times the wall thickness, h, or 3(10) = 30 in.
Since Vu < 0.5 φVc , the limit based on ℓw / 3 need not be satisfied. Maximum permitted
spacing is 18 in., which is greater than 12 in. OK
Try two layers of #5 bars spaced at 12 in.
(Continued)
587
3. Compute the design interaction diagram

Assuming a minimum clear cover of 3 4 in., and #5 bars for the horizontal, transverse
reinforcement, the total required number of bars spaced at 12 in. per layer is
ℓ w − 2 × (clear cover ) − 2 × dbhoriz − dbvert

No. of bars ≈ +1
spacing (in.)
120 − 2 × 3 / 4 − 2 × 0.625 − 0.625
≈ +1
12
≈ 10.7
Try 11 bars per layer for a total of 22 equally spaced #5 bars at 11 1 2 in.
The design interaction diagram for the wall cross section with 22–#5 bars at 11 1 2 in.,
equally distributed in two layers, is shown in Fig. 15.5.12: the design strength (φ Mn, φ Pn) is
greater than the required strength (Mu, Pu). Note that the wall is tension controlled, as pre-
viously assumed. Therefore, the provided reinforcement is adequate, although somewhat in
excess of that required. For this example, an attempt will be made to optimize the amount
of reinforcement by increasing the bar spacing without exceeding the maximum permitted
of 18 in. A revised analysis of the cross section with #5 bars spaced at 16 1 2 in. in two lay-
ers (ρℓ ≈ 0.0038 > ρℓ min ) shows that the strength requirement is satisfied but with a design
strength that is much closer to the required strength. In practice, however, members must be
designed for several load combinations. Therefore, trying to optimize wall reinforcement or
size for a single load combination is not warranted. Use two layers of #5 spaced at 16 1 2 in.
A design sketch of the final reinforcement layout is shown in Fig. 15.5.13.
Figure 15.5.12 Design strength interaction diagram for wall of Example 15.5.2.
#5 @ 16½” (vertical)
¾”(min)
10”
¾”(min)
10’–0”
Figure 15.5.13 Design sketch for the shear wall of Example 15.5.2.

58
Shear Strength—Solid Rectangular Walls

The ACI Code shear strength provisions for walls are largely based on the results of a ser-
ies of tests conducted by the Portland Cement Association (PCA) on large rectangular shear
wall specimens representative of both high-and low- rise building construction [15.21].
Experimental results from Muto and Kokusho [15.22], Ogura et al. [15.23], Williams and
Benjamin [15.15], Benjamin and Williams [15.16], and Antebi et al. [15.17] were also consid-
ered in the development of the provisions. Based on these studies, the in-plane shear strength
of cantilever shear walls may be computed by using the same approach for members subjected
to combined bending and axial load, where the nominal shear strength Vn is computed as
Vn = Vc + Vs [5.8.1]
where Vc and Vs are the contributions of the concrete and shear reinforcement to shear
strength. This approach may be used to compute the shear strength of both squat (hw /ℓw ≤ 2)
and tall (hw /ℓw > 2) walls. Alternatively, ACI-11.5.4.1 permits squat walls (hw /ℓw ≤ 2) to be
designed with the strut-and-tie method described in Chapter 14.
Concrete Contribution to Shear Strength, Vc

The current ACI Code provisions provide both a simplified and a detailed method for
computing Vc.
Simplified Method (ACI-11.5.4.5)

In this approach, Vc is approximately taken as the shear corresponding to the onset of diag-
onal cracking for beams, and may be used when the wall is subjected to flexure with axial
compression or with no axial load, as follows
Vc = 2 λ fc′hd (15.5.5)1
where λ is the modification factor for lightweight concrete in accordance with ACI-19.2.4.2
(see Section 1.8), h is the wall thickness, and d is taken as 0.8ℓw. A larger value of d is per-
mitted (ACI-11.5.4.2) if the centroid of the tension force resultant is computed on the basis
of strain compatibility.
If the wall is subjected to axial tension, the concrete contribution to shear can be either
neglected (i.e., Vc = 0) or computed as
 N u 

Vc = 2 1 +  λ fc′hd (15.5.6)2
 500 Ag 

Vc = 0.17 λ fc′hd (15.5.5)
2 For SI, ACI 318-14M, with fc′ , N u /Ag , N u /(hℓ w ) in MPa gives
 N u 
Vc = 0.17 1 +  λ fc′ hd (15.5.6)
 3.5 Ag 
 Nu 
Vc =  0.27λ fc′ +
4hℓ w 
hd (15.5.7)

  0.2 N u  
  0.1λ fc′ + hℓ  ℓ w 

Vc = 0.05λ fc′ + w  hd (15.5.8)

Mu ℓ w
−



 Vu 2 
589
where Nu is the factored axial tensile force in pounds acting concurrently with the factored
shear force, Vu, and Ag is the gross area of concrete of the wall in sq in. Note that Nu in Eq.
(15.5.6) must be entered as a negative quantity.
Detailed Method
In this method, Vc is computed as the lesser of the following (ACI-11.5.4.6)
 N 
Vc =  3.3λ fc′ + u  hd (15.5.7)2
 4hℓ w 
or
  0.2 N u  
  1.25λ fc′ + hℓ  ℓ w 
Vc = 0.6 λ fc′ + w  hd (15.5.8)2
 Mu ℓ w 
 − 
 Vu 2 
where Nu is the factored axial force (positive for compression and negative for tension)
acting concurrently with the factored shear force, Vu. In these equations, the term N u /(hℓ w )
must be in psi. Note that Eq. (15.5.8) does not apply when (Mu / Vu – ℓw /2) is negative.
Equation (15.5.7) approximately corresponds to the development of web-shear cracking
when the principal tensile stress reaches 4 fc′ at the centroid of a rectangular cross section
subjected to combined axial load and shear [15.21]. This equation may govern for thin-
web flanged, or barbell walls, and will usually control for low-rise squat rectangular walls.
Alternatively, the ACI Code permits squat walls (hw /ℓw ≤ 2) to be designed for in-plane
shear in accordance with the strut-and-tie method (ACI-11.5.4.1), as noted earlier.
Equation (15.5.8) corresponds to the onset of flexure-shear cracking and will usually
govern for walls with hw /ℓw > 2. In developing Eq. (15.5.8) for computing the concrete
contribution Vc, it was assumed that flexural diagonal cracking would initiate at a sec-
tion located at ℓw /2 from the base of the wall [15.21]. In general, ACI-11.5.4.7 permits
sections located between the wall base and a distance ℓw /2 or hw /2 from the base of the
wall, whichever is less, to be designed for Vc computed by using the Mu / Vu ratio at ℓw /2
or hw /2, whichever is less from the wall base, as shown in Fig. 15.5.14. Note that the term
(Mu / Vu – ℓw /2) in Eq. (15.5.8) may become negative for walls with low Mu / Vu ratios. In
such a case, Eq. (15.5.8) does not apply, and Eq. (15.5.7) must be used to compute Vc. In
other words, web-shear cracking instead of flexure-shear cracking will be expected to occur.
Figure 15.5.15 shows the variation in the shear strength attributed to the concrete, Vc,
for walls computed with the simplified [Eq. (15.5.5)] and the detailed [Eqs. (15.5.7) and
ℓw
h
hw
Smaller of
ℓw or hw
2 2
Critical section for M
Critical section for V
Figure 15.5.14 Definition of the critical section when the detailed method of ACI-11.5.4.6 [Eqs.
(15.5.7) and (15.5.8)] is used for computing the concrete contribution to shear strength, Vc.
590
10
fc’ = 5000 psi
9 Web-shear cracking λ = 1.0
[Eq. (15.5.7)]
8
Nu
= 1000 psi
7 hℓw
6 Nu
= 500 psi Flexure-shear cracking
Vc hℓw
5 [Eq. (15.5.8)]
fc’ hd
Nu
4 = 0 psi
hℓw
3
1 Simplified Eq. (15.5.5)

0
0 1 2 3 4 5 6
hw
ℓw
Figure 15.5.15 Variation in the shear strength attributed to concrete, Vc, as a function of the wall
aspect ratio, hw /ℓw, in accordance with Eqs. (15.5.5), (15.5.7), and (15.5.8), for a cantilever wall
subjected to an inverted-triangle load distribution and axial compressive stresses of 0, 500, and
1000 psi.
(15.5.8)] methods (ACI-11.5.4.5 and ACI-11.5.4.6). In the figure, Vc is shown as an average

shear stress Vc /(hd) divided by fc′ , as function of the wall aspect ratio, hw /ℓw. The Mu / Vu
ratio in Eq. (15.5.8) was computed based on an inverted triangular force distribution over
the wall height. The curves are shown for three values of axial compressive stress in the
wall: 0, 500, and 1000 psi, and for normal-weight concrete with a compressive strength fc′
of 5000 psi.
As shown in the figure, web-shear cracking [i.e., Eq. (15.5.7)] provides an upper limit to
the concrete contribution to shear strength and will govern for walls with low hw /ℓw ratios.
At larger aspect ratios, flexure-shear cracking, Eq. (15.5.8), will govern and will result in
a reduced contribution from the concrete to shear strength. However, Vc need not be taken
less than 2 fc′hd [15.21], and thus, this value should be taken as the lower limit for Vc in
walls with compressive or zero axial force.
Shear walls in bending combined with axial tensile forces may be found in tall walls
coupled with stiff and strong coupling beams (see Fig. 15.5.6). In such cases, it is conserv-
ative to assume no contribution from the concrete to shear strength under axial tension (i.e.,
Vc = 0). Alternatively, Vc may be computed by using Eq. (15.5.6) of the simplified method
or, Eqs. (15.5.7) and (15.5.8) of the detailed method.
The variation in shear strength attributed to the concrete as a function of the wall aspect
ratio, hw /ℓw, is shown in Fig. 15.5.16 for two levels of axial tension (–100 and –400 psi). As
before, the curves were obtained by assuming an inverted triangular force distribution for
computing the Mu / Vu ratio in Eq. (15.5.8). At low levels of axial tension (e.g., –100 psi),
the simplified method [Eq. (15.5.6)] provides a lower bound to Vc for walls with rela-
tively low aspect ratios (about 3 or less). For walls with higher aspect ratios, however,
the flexure-shear cracking strength computed with Eq. (15.5.8) provides a lower estimate
of Vc,—perhaps as much as 60% lower for walls with aspect ratios of 6. At higher levels
of axial tension (e.g., –400 psi), the simplified method [Eq. (15.5.6)] provides a lower
bound over the entire range of aspect ratios shown in Fig. 15.5.16. Test data on the shear
strength of walls with axial tensile forces are scarce; thus, the ACI Code equations provide
only estimates of wall shear strength under axial tension, and these should be used with
caution.
591
4
f’c = 5000 psi
Web-shear cracking λ = 1.0
[Eq. (15.5.7)]
3 Flexure-shear cracking
[Eq. (15.5.8)]
Vc Nu
= –100 psi
2 hℓw
fc’ hd
Nu
= –400 psi
Simplified hℓw
1
Eq. (15.5.6)
0
0 1 2 3 4 5 6
hw
ℓw
Figure 15.5.16 Variation in the shear strength attributed to concrete, Vc, as a function of the wall
aspect ratio, hw /ℓw, in accordance with Eqs. (15.5.6), (15.5.7), and (15.5.8) for a cantilever wall
subjected to an inverted-triangle load distribution and axial tensile stresses of –100 and –400 psi.
Shear Reinforcement Contribution to Shear Strength, VS

As noted earlier, all walls are required to have at least a minimum amount of uniformly dis-
tributed vertical and horizontal reinforcement (see Section 15.2). In shear walls, this rein-
forcement will help restrain the growth (width and length) of diagonal, shear cracks; thus,
both the horizontal and vertical bars contribute to shear strength. In design, however, only
the horizontal reinforcement is used to compute the contribution of the steel reinforcement
to shear strength, Vs. Accordingly, the contribution of the reinforcement to shear strength is
computed, per ACI-11.5.4.8, as
Av f yt d
Vs = (15.5.9)
s
where Av is the area of horizontal reinforcement, As horiz spaced at a distance s over the wall
height (svert in Figure 15.2.2), and f yt is the yield strength of the horizontal reinforcement.
In other words, the steel reinforcement contribution to shear strength may be computed as
Vs = ρt f yt hd (15.5.9a)
where ρt is the transverse (horizontal) reinforcement ratio and h is the wall thickness.
The effective depth, d, is taken as 0.8ℓw unless a larger value is used, per ACI-11.5.4.2
as discussed earlier.
Minimum Shear Reinforcement Requirements

Minimum requirements for the horizontal shear reinforcement in walls were presented in
Section 15.2. These requirements apply to shear walls when the in-plane required shear
strength Vu ≤ 0.5φVc (ACI-11.6.1).
If Vu > 0.5φVc, however, both the horizontal shear reinforcement ratio, ρt , and the longi-
tudinal reinforcement ratio, ρℓ , must be at least 0.0025 (ACI-11.6.2), irrespective of the size
and Grade of reinforcement. In squat walls, however, tests have shown that the horizontal
592
shear reinforcement becomes less effective than the longitudinal (vertical) reinforcement
[15.14]. Accordingly, when Vu > 0.5φVc and hw /ℓw ≤ 2.5, the required longitudinal reinforce-
ment ratio, ρℓ , is computed as follows:
 h 
ρℓ ≥ 0.0025 + 0.5  2.5 − w  (ρt − 0.0025) (15.5.10)
 ℓw 
However, ρℓ need not exceed the horizontal reinforcement ratio, ρt , required for shear
strength, as computed from Eq. (15.5.9a). Also, note that when hw /ℓw = 0.5, the amount of
longitudinal reinforcement, ρℓ , is equal to the amount of horizontal shear reinforcement, ρt .
In addition to the maximum spacing requirements for the horizontal and vertical rein-
forcement for walls with Vu ≤ 0.5φ Vc (smaller of 3h or 18 in.), the maximum spacing
is further limited to ℓw /3 and to ℓw /5 for the longitudinal and transverse reinforcement,
respectively, when Vu > 0.5φVc (see Fig. 15.2.2).
Maximum Amount of Shear Reinforcement

Similar to the behavior of reinforced concrete beams, a heavily reinforced wall in shear
may reach failure by crushing of the concrete diagonal struts prior to yielding of the trans-
verse reinforcement (see Sections 5.7 and 5.10). To prevent such a failure mode in shear
walls, ACI-11.5.4.3 requires that at any horizontal section, the nominal strength, Vn, be
limited to 10 fc′ hd .
EXAMPLE 15.5.3
The factored axial force, shear force, and bending moment diagrams in a wall of a
9-story building are shown in Fig. 15.5.17. The wall is 20 ft long and 10 in. thick and
is one of several walls resisting gravity and wind loads. Compute the amount of verti-
cal and horizontal reinforcement required in the first story. Use normal-weight concrete
with fc′ = 5000 psi and Grade 60 steel.
Figure 15.5.17 Factored axial force, shear force and bending moment diagrams due to gravity
and wind loads in the wall of Example 15.5.3.
(Continued)
593
SOLUTION
1. Compute the required amount of vertical reinforcement, ρℓ , based on flexural

strength requirements. From Fig. 15.5.17, the required moment strength at the base
of the wall is Mu = 15,162 ft-kips, and the factored axial force acting currently with
this moment is Pu = 1036 kips.
a. Assuming a tension-controlled section (φ = 0.9), compute the axial load ratio, α,
from Eq. (15.5.4)
Pn P /φ 1036
α= = u = = 0.096
hℓ w fc′ hℓ w fc′ 0.9(10)(240)5
Mn
b. Compute the required strength coefficient :
hℓw2
M n M u /φ 15,162(12)
required = = = 0.35 ksi
hℓw2 hℓw2 0.9(10)240 2
c. From Fig. 15.5.9 (bottom chart), estimate ρℓ ≈ 0.0065 or 0.65% .

d. Assuming #5 bars, the minimum amount of longitudinal reinforcement when
Vu ≤ 0.5φ Vc is 0.0012 (see Table 15.2.1), which is less than the required amount
computed above based on flexural strength requirements. If, however, Vu > 0.5φVc,
a larger minimum amount may be required. Thus, minimum requirements for
the amount of vertical reinforcement will be checked later after shear strength
requirements have been computed.
2. Compute the amount of horizontal reinforcement, ρt , based on shear strength
requirements.
a. Concrete contribution Vc—According to ACI-11.5.4.5 and 11.5.4.6, Vc may be
calculated by using the simplified or the detailed method. Here, both methods
are used and then compared.
i. Simplified Method—Since the wall is subjected to axial compression under the
given factored loads, use Eq. (15.5.5). Thus,
Vc = 2 λ fc′hd
1

= 2(1) 5000 (10)0.8(240) = 272 kips
1000
where d was taken as 0.8ℓw and λ was taken as 1.0 for normal-weight concrete.
ii. Detailed Method—According to ACI-11.5.4.6, Vc is to be computed as the
lesser of the values resulting from Eqs. (15.5.7) and (15.5.8.)
The axial compressive stress in the wall is
N u 1036(1000 )
= = 432 psi
hℓ w 10(240)
and the wall-to-height ratio, hw /ℓw, is 114/20 = 5.7.
• Vc based on web-shear cracking [Eq. (15.5.7)]:
 N 
Vc =  3.3λ fc′ + u  hd
 4hℓ w 

 1036(1000)   
=  3.3(1) 5000 + 10(0.8)240  1  = 655 kips
 4(10)240   1000 
(Continued)
594
•
Vc based on flexure-shear cracking [Eq. (15.5.8)]: In accordance with
ACI-11.5.4.7, for sections located between the wall base and the smaller
of ℓw /2 or hw /2, Vc is permitted to be computed for the Mu / Vu ratio at a
distance equal to ℓw /2 or hw /2 from the base of the wall, whichever is
less. For this wall, ℓw /2 = 10 ft (< hw /2 = 57 ft) governs. At this location,
the factored moment Mu is 13,175 ft-kips and the corresponding factored
shear is 198 kips, resulting in a Mu / Vu ratio of 66.5 ft. This value is
greater than ℓw /2 and Eq. (15.5.8) applies. Thus,
  0.2 N u  
  1.25λ fc′ + hℓ  ℓ w 
Vc = 0.6 λ fc′ + w  hd
 Mu ℓ w 
 − 
 Vu 2 
  0.2(1036)(1000)  
  1.25(1) 5000 +  240 
= 0.6(1) 5000 +
10 ( 240 )  10(0.8)240  1 
 (66.5 − 10)(12)   1000 
 
 
= 200 kips
Therefore, Vc = 200 kips based on flexure-shear cracking governs.

Since Vc computed with the detailed method (200 kips) is less than that
computed with the simplified method (lower limit), Vc will be taken as the
lower limit of 272 kips. In other words, for this wall with an aspect ratio
hw /ℓw of 5.7, there would be no need to compute Vc with the detailed method.
b. Shear reinforcement contribution, Vs—Because the simplified method was used to
compute Vc, the critical section for shear is taken at the base of the wall where the
factored shear force is maximum in the first story. At this location, Vu = 200 kips.
Since

0.5φVc = 102 kips < Vu = 200 kips < φVc = 204 kips
only minimum transverse, horizontal reinforcement is required. Also, since

Vu > 0.5φ Vc, a minimum of horizontal reinforcement ratio, ρt , of 0.0025 is
required (ACI-11.6.2). Because this wall has an aspect ratio hw /ℓw= 5.7 > 2.5,
Eq. (15.5.10) does not apply; thus the minimum required amount of longitudinal,
vertical reinforcement, ρℓ , is equal to ρt , or 0.0025. This amount is less than that
required from flexural strength requirements (0.0065) and thus, it is satisfied.
Since this wall requires only the minimum amount of horizontal reinforcement, a check
on the maximum permitted amount to satisfy the requirement that Vn ≤ 10 fc′ hd is
clearly not necessary. The procedure, however, is illustrated below:
Vn = Vc + Vs = 272 + 0.0025(60)(10)(0.8)240
Vn = 560 kips

1
10 fc′ hd = 10 5000 (10)(0.8)240 1000 = 1358 kips
Thus, Vn < 10 fc′ hd , as was expected. OK

(Continued)
59
3. Select reinforcement and bar spacing.

a. Vertical reinforcement—Since Vu > 0.5φVc, spacing of the vertical reinforcement
is the smaller of
• 3h = 3(10) = 30 in.
• ℓw /3 = 240/3 = 80 in., or
• 18 in. Governs!
The required amount of vertical reinforcement per flexural strength requirements is
ρℓ ≈ 0.0065. Thus, the total required area of reinforcement per foot of wall length is
required As vert = ρℓ h × (12 in.)
= 0.0065(10)12

= 0.78 sq in./ft.
Since the wall thickness is 10 in., two layers of reinforcement are required. Therefore,
the required amount of reinforcement per layer is

required As vert = 0.78 / 2 = 0.39 sq in./ft / layer
From Table 8.3.1, choose #5 bars spaced horizontally at 10 in. to provide a total
amount of 0.37 sq in./ft per layer. This amount is slightly less than that computed
above. Given the approximations involved in the calculations, however, it can be
considered adequate as a first trial.
Try 24 bars per layer for a total of 48 #5 bars equally spaced at 10 in.
The design interaction diagram of the wall with this amount of reinforcement shows
that the section is tension controlled and yields a design flexural strength, φ M n , of
15,170 ft-kips at a factored axial force Pu = 1036 kips. Therefore,

Mu = 15,162 ft-kips < φ M n = 15,170 ft-kips OK
Use two layers of #5 bar spaced at 10 in.
b. Horizontal reinforcement—Since Vu > 0.5φ Vc spacing of the horizontal rein-
forcement is the smaller of
• 3h = 3(10) = 30 in.
• ℓw /5 = 240/5 = 48 in., or
• 18 in. Governs!
The required amount of horizontal reinforcement is ρt = 0.0025, and thus the total
required area of reinforcement per foot of wall height is
required As horiz = ρt h × (12 in.)
= 0.0025(10)12

= 0.30 sq in./ft
Since two layers of reinforcement are required, the required amount of reinforce-
ment per layer is

required As horiz = 0.30 / 2 = 0.15 sq in./ft/layer
This requirement is satisfied with two layers of #4 bars spaced vertically at 16 in.
(see Table 8.3.1).
Use two layers of #4 bars spaced at 16 in. A final design sketch is shown in
Fig. 15.5.18.
(Continued)
596
#4 @ 16” (horizontal) #5 @ 10” (vertical)

¾”(min)
10”
¾”(min)
20’ – 0”
Figure 15.5.18 Design sketch for the shear wall of Example 15.5.3.
15.6 LATERAL SUPPORT OF LONGITUDINAL

REINFORCEMENT
As described in Section 10.8 for columns, longitudinal reinforcement may tend to buckle
if subjected to high compressive stresses at or shortly after spalling of the concrete cover.
In walls not subjected to seismic actions, ACI-11.7.4.1 requires that the longitudinal rein-
forcement be laterally supported by transverse ties whenever the total area of longitudinal
steel exceeds 1% of the gross cross-sectional area of the wall, or when longitudinal rein-
forcement is required for axial strength. The condition involving axial strength is taken to
apply when vertical reinforcement is required as compression reinforcement to resist wall
actions. In practice, vertical reinforcement in walls with low amounts of reinforcement (i.e.,
less than 1% of the gross area of the wall) will rarely be needed as compression reinforce-
ment and thus would seldom require lateral support. On the other hand, walls with larger
reinforcement amounts may have some of their longitudinal reinforcement concentrated at
the wall ends (boundary elements), as shown in Fig. 15.5.3. In such cases, the longitudinal
reinforcement in the boundary elements is designed to act in both tension and compression
and thus must be laterally supported by transverse ties (see Fig. 15.6.1). The remaining
vertical reinforcement in the web is commonly designed to satisfy shear strength require-
ments only and would not need to be laterally supported unless web reinforcement is used
as compression reinforcement. Transverse ties must satisfy the requirements of ACI-25.7.2
and may be arranged as shown in Fig. 10.8.1. It must be emphasized that these require-
ments apply to walls not subjected to seismic actions; additional requirements must be met
for walls subjected to earthquake-induced actions.
Figure 15.6.1 Common transverse reinforcement details to provide lateral support of longitudinal

reinforcement in wall boundary elements.
597
1 5 . 7 . 1 F O R C E S O N R E TA I N I N G WA L L S 597
15.7 RETAINING STRUCTURES
Retaining structures hold back soil or other loose material and prevent assumption of the
natural angle of repose at locations where an abrupt change in ground elevation occurs. The
retained material exerts a push on a structure and thus tends to overturn or slide it or both.
Several types of these structures are described below (see Fig. 15.7.1).
1. Gravity wall. A gravity wall is usually of plain concrete and depends entirely on its
weight for stability. It is used for walls up to about 10 ft high.
2. Cantilever retaining wall. Cantilever walls, the most common type of retaining struc-
ture, are used in the height range of 10 to 25 ft. The stem, heel, and toe of such a wall
each acts as a cantilever beam.
3. Counterfort wall. In a counterfort wall the stem and slab are tied together by trans-
verse walls (counterforts) spaced at intervals, which act as tension ties to support the
stem wall. Counterfort walls are often economical for heights over about 25 ft.
4. Buttress wall. A buttress wall is similar to a counterfort wall except that the trans-
verse support walls are located on the side of the stem opposite the retained material,
where they act as struts. Buttresses, as compression elements, are more efficient than
the tension counterforts and are economical in the same height range. The counter-
fort is more widely used than the buttress, however, because the counterfort is hidden
beneath the retained material, whereas the buttress occupies what could otherwise be
usable space in front of the wall.
5. Bridge abutment. A wall-type bridge abutment acts like a cantilever retaining wall
except that the bridge deck provides an additional horizontal restraint at the top of
the stem. Thus this abutment is designed as a beam fixed at the bottom and simply
supported or partially restrained at the top.
6. Box culvert. The box culvert, with either single or multiple cells, acts as a closed
rigid frame that must resist not only lateral earth pressure but also vertical load, either
from the soil that it supports or from both the soil and highway vehicle loads.
15.7.1 FORCES ON RETAINING WALLS
The magnitude and direction of the earth pressure that tends to overturn and slide a retain-
ing wall may be determined by applying the principles of soil mechanics. Excellent texts
such as those reported in Terzaghi, Peck, and Mesri [15.24] and Huntington [15.25] will be
useful for any extensive study of how to determine the soil pressure to be used in a given
situation.
The pressure exerted by the retained material is proportional to the distance below the
earth surface and to its unit weight. Analogous to the action of a fluid, the unit pressure p at
a distance h below the earth surface may be expressed as
p = Cwh (15.7.1.1)
where w is the unit weight of the retained material and C is a coefficient that depends on the
physical properties of the material.
There are two categories of earth pressure: (1) the pressure exerted as the earth moves in
the same direction as the retaining structure deflects, known as active pressure and (2) the
resistance developed as a structure moves against the earth, known as passive pressure.
Passive pressure is several times larger than active pressure. Both active and passive pres-
sures may be expressed in the form of Eq. (15.7.1.1), but using Ca and Cp as the active and
passive pressure coefficient C, respectively.
The force Pa caused by active pressure on a wall of height h may be expressed as
h2
Pa = Ca w (15.7.1.2)
2
598
Retained Retained
material Stem material
Toe
Heel
(a) Gravity wall (b) Cantilever retaining wall
Retained Retained
material material
Buttress
Stem
Stem Counterfort
Slab Slab
(c) Counterfort wall (d) Buttress wall
Bridge deck Approach pavement Top
Retained
material Wall
Floor
Stem
Toe
Heel
(e) Bridge abutment (f) Box culvert
Figure 15.7.1 Types of retaining structures.
and the force Pp developed in passive pressure may be expressed as

h2
Pp = C p w (15.7.1.3)
2
where Caw and Cpw in Eqs. (15.7.1.2) and (15.7.1.3) may be considered equivalent fluid
pressures. Typical values for Ca and Cp are 0.3 and 3.3, respectively, for granular material
such as sand. Roughly, Cp = 1/Ca.
The proper evaluation of earth pressures on retaining structures is outside the scope of
this book; the preceding brief comments are offered to provide some logic for the earth
pressures that are used in the design example of this chapter.
The following factors may affect the pressure (active pressure) on a wall:
1. Type of backfill used.
2. Seasonal condition of the backfill material, such as wet, dry, or frozen.
3. Drainage of backfill material.
4. Possibility of backfill overload, such as trucks and equipment near the wall.
5. Degree of care exercised in backfilling.
6. Degree of rotational restraint between various components of the retaining structure.
7. Possibility of vibration in the vicinity of the wall (especially in the case of
granular soil).
8. Type of material beneath the footing of the retaining structure.
9. Level of the water table.
59
1 5 . 7 . 2 S TA B I L I T Y R E QU I R E M E N T S 599
Probably the most important single factor is the need to prevent water from accumulat-
ing in the backfill material. Walls are rarely designed to retain saturated material, which
means that proper drainage must be provided.
When vehicles may travel near by, exerting their loads, or when buildings are constructed
near the top of a retaining wall, the lateral pressure against that wall is increased. In the case of
a fixed static load such as a building, the weight of the building can be converted into an addi-
tional height (surcharge) of backfill soil material. The effect of a highway or railroad passing
over the retained material near the wall causes a dynamic reaction that cannot accurately be
converted into a static effect. However, some specifications have traditionally prescribed an
equivalent static surcharge corresponding to a number of additional feet of backfill material.
15.7.2 STABILITY REQUIREMENTS
The first step in retaining wall design is to establish the proportions such that the stability of
the structure under active pressure is assured (see Fig. 15.7.2.1). Three requirements must
be satisfied: (a) the overturning moment Pah(h′/3) must be more than balanced by the resist-
ing moment Wx1 + Pav L, so that an adequate factor of safety against overturning is provided,
usually about 2.0; (b) sufficient frictional resistance F in combination with any reliable
passive resistance Pp against the toe must provide an adequate factor of safety (usually 1.5)
against sliding caused by Pah; and (c) the base width L must be adequate to distribute the
load R to the foundation soil without causing excessive settlement or rotation.
Typically, referring to the pressure distribution of Fig. 15.7.2.1, the overturning factor of
safety (FS) would be computed as follows:
resisting moment Wx1 + Pav L

FS = = (15.7.2.1)
overturing moment Pah (h ′ / 3)
or, perhaps more frequently, the vertical component of Pa would be neglected:
Wx1
FS = (15.7.2.2)
Pah (h ′ / 3)
where W represents the weight of the concrete wall and footing and of the soil resting on
the footing.
δ Commonly
accepted
pressure
distribution
h’
h x1 W
Pa
δ
Pav
Pah
h’/3
Pp
Toe F = µR Heel
x2
R
L
Figure 15.7.2.1 Forces on retaining wall.

60
Using the notation of Fig. 15.7.2.1, the factor of safety against sliding may be
computed as,
µ R + Pp
FS = (15.7.2.3)
Pah
where µ is the coefficient of friction between the soil and footing. Table 15.7.2.1 gives a
range of coefficients of friction that may be used as a guide for typical values in lieu of
more accurate ones.
The inclusion of some passive resistance Pp on the toe of the footing may or may not be
justified. Certainly, to actually develop passive pressure in the soil in front of the wall, the
design should specify that the concrete be placed without using forms for the toe and with-
out disturbing the soil against which the concrete is placed.
Referring to Fig. 15.7.2.2, the ordinary passive resistance against the toe is
1
Pp1 = C p wh12 (15.7.2.4)
2
TABLE 15.7.2.1 VALUES OF COEFFICIENT OF

FRICTION μ BETWEEN SOIL AND CONCRETE
Soil µ
Coarse-grained soils (without silt) 0.5–0.6

Coarse-grained soils (with silt) 0.4–0.5
Silt 0.3–0.4
Sound rock (with rough surface) 0.60
Frequently, however, Pp1 is neglected, and in nearly all cases at least the value of h1
is reduced under the assumption that during construction, or after, the earth surface can-
not be expected to remain undisturbed at its final designated elevation. If, when all reli-
able resistances have been included, the factor of safety remains inadequate, a base key
(Fig. 15.7.2.2) may be used. Essentially, the base key, when placed in an unformed excava-
tion against undisturbed material, may be expected to develop an additional passive force
Pp2, shifting the possible failure plane from line 1 to line 2.
The base key develops the additional resistance
1
Pp 2 = C p w(h22 − h12 ) (15.7.2.5)
2
and also, an inert region, bced of Fig. 15.7.2.2, is created, which moves the friction plane
from bd to ce. Thus the frictional force developed along ce is based on the angle α, the
angle of internal friction of the soil, rather than on the friction angle between soil and
a Earth surface Stem
1
h1
2 Pp1
Toe
h2
Pp2 b d Base key
c e f
Figure 15.7.2.2 Passive resistance and effect of base key.

601
1 5 . 7 . 3 P R E L I M I NA RY P R O P O RT I O N I N G O F CA N T I L E V E R WA L L S 601
L L
L/3 L/3 L/3 L/3 L/3 L/3
R 3x2
R
ph
pt
pt CL of footing
e L/2
CL of footing
x2 e
(a) Resultant within middle third (b) Resultant outside middle third
Figure 15.7.2.3 Soil pressure distribution.
concrete. Normally, tan α > µ for granular material, so that, by making the base key deep
enough to create an inert block, additional frictional resistance is developed.
Finally, the magnitude and distribution of the soil pressure requires investigation. Usual
practice is to require the resultant vertical force R to be inside the middle third of the foot-
ing for sand and gravel subbases and within the middle half for rock subbase. In addition,
the maximum pressure may not exceed the allowable value. Comments regarding allowable
bearing capacity as well as some typical safe bearing values are to be found in Chapter 19.
Referring to Fig. 15.7.2.3(a), when the entire footing is under compression, the basic
equation for combined bending and axial compression acting on a 1-ft strip along the wall is
R Re( L / 2) R  6e 
p= ± 3 = 1 ±  (15.7.2.6)
L L /12 L L
For the limiting condition of zero stress at the heel, e = L / 6; thus Eq. (15.7.2.6) is valid for
all positions of R within the middle third.
When the resultant R is outside the middle third [Fig. 15.7.2.3(b)], vertical force equi-
librium requires
1
R= pt (3 x2 )
2
2R (15.7.2.7)
pt =
3 x2
for 0 < 3x2 < L .
15.7.3 PRELIMINARY PROPORTIONING
OF CANTILEVER WALLS
The pressure exerted by the retained material must be resisted by out-of-plane bending
and shear along the length of the wall. Design of retaining walls entails considering a
12-in.-wide imaginary strip. The wall is then analyzed as a cantilever beam having a width
of 12 in. where the wall thickness and reinforcement is to be determined.
According to ACI-11.1.4, cantilever retaining walls must be designed using the provi-
sions for sectional strength in ACI-22.2 through 22.4, with minimum horizontal reinforce-
ment (in the plane of the wall) per Table 15.2.1 (ACI-11.6). Cantilever retaining walls are
exempted from the requirement of providing two layers of reinforcement in walls 10 in.
thick or greater (ACI-11.7.2.3). Good practice, however, may dictate the use of two layers
of reinforcement for walls thicker than 1 ft.
602
To begin the design, it is necessary to apply certain “rules of thumb” along with the use
of statics to estimate dimensions such as the length and thickness of the base footing and
the relative position of the wall with respect to the footing.
Height of Wall
Since the bottom of the footing must be below the frost penetration depth—say, 3 to 4 ft
in the northern United States—the overall height equals the desired difference in elevation
plus the frost penetration depth.
Position of Stem on the Base Footing

The following demonstration using statics shows that the front face of the wall should
coincide with the desired position of the resultant soil pressure beneath the base. Consider
the most typical situation of a vertical wall with level backfill, as shown in Fig. 15.7.3.1.
Assume an average unit weight w for all material (concrete and earth) enclosed within abcd
and neglect entirely the concrete in the toe. Thus
R = W = whγ L
Satisfying rotational equilibrium about the heel,
h γL
Pa +W = Rξ L
3 2
wh 2  h  γL
Ca   + whγ L = whγ L (ξ L )
2  3 2
Solving for L /h gives
L Ca (15.7.3.1)
=
h 3 γ (2 ξ − γ )
The variable ξ must be selected by the designer, based on the type of soil and desired pres-
sure distribution. For good granular soil and acceptance of a triangular pressure distribution
with the resultant at the outer edge of the middle third, ξ = 2 3 . For clay where a uniform
distribution might be desired, ξ would be 1 2 .
Earth surface
a d
γ L/2
h2 h
Pa = Caw
2
(1 – γ )L γL
Neglect h
for 3
derivation
b c
R X = ξL
Figure 15.7.3.1 Data for economical proportioning of wall.

603
1 5 . 7 . 4 D E S I G N of C A N T I L E V E R R E T A I N I N G W A L L s 603
Minimizing the L /h in Eq. (15.7.3.1) such that the base width is a minimum,
γ = ξ (15.7.3.2)
Thus the front face of the wall should line up with the desired position of the soil pressure
resultant.
Length of Base
A preliminary base length may be obtained by the application of static moments with respect
to point b in Fig. 15.7.3.1, employing the same assumptions that underlie Eq. (15.7.3.1).
This process makes it possible to obtain a more accurate result, since one can easily esti-
mate the stem thickness and footing thickness as fractions of L or of h.
Thickness of Footing
The footing thickness is usually 7 to 10% of the total height h, with a minimum of about 1
ft. It should be about equal to the base thickness of the stem.
Thickness of Stem
The thickness at the top of the wall is arbitrary; however, cover requirements and construc-
tion constraints will dictate how thin it may be. Generally a 10-or 12-in. minimum is pre-
ferred, though no minimum is prescribed by the ACI Code.
The base thickness of the stem is determined as required for bending moment and shear,
though it may be estimated as 12 to 16% of the base width or 10 to 12% of the wall height.
The stem thickness should not be too skimpy, since a thin wall deflects considerably at the
top and savings in reinforcement due to larger effective depth will tend to offset the cost of
any extra concrete used. It is recommended that a batter of 1 4 in./ft of height be provided on
the front face to offset deflection or forward tilting of the structure.
15.7.4 DESIGN OF C ANTILEVER RETAINING WALLS

The design of cantilever retaining walls will be illustrated through an example.
EXAMPLE 15.7.4.1
Design a cantilever retaining wall to support a bank of earth 16 ft high above the final
level of earth at the toe of the wall. The backfill is to be level, but a building is to be built
on the fill. Assume that an 8-ft surcharge will approximate the lateral earth pressure effect.
Data: weight of retained material = 120 pcf; equivalent fluid pressure = 32 pcf; angle of
internal friction = 35°; coefficient of friction between concrete and soil = 0.40; for pas-
sive pressure, use equivalent fluid pressure of 400 pcf; fc′ = 3000 psi (normal weight);
fy = 60,000 psi; maximum soil pressure = 5 ksf (kips per square foot). Use the ACI Code.
SOLUTION
(a) Basic design data. For adequate deflection control on the cantilever wall, choose to
use a reinforcement ratio ρ about 25–35% of ρb (as suggested in Section 12.10).
Choose
ρ ≈ 0.35ρb = 0.35(0.0214) = 0.0075 (see Table 3.6.1)
This corresponds to Rn ≈ 410 psi. [see Eq. (3.8.4b)]

(b) Height of wall. Allowing 4 ft for frost penetration to the bottom of the footing in
front of the wall, the total height becomes
h = 16 ft + 4 ft = 20 ft
(Continued)
604
Example 15.7.4.1 (Continued)
(c) Thickness of footing. The thickness may be estimated at this stage of design to be 7 to
10% of the overall height h. Assume a uniform footing thickness, t = 2 ft (10% of h).
(d) Base length. One could determine the base length by considering the equilibrium of
factored loads (using the factor 1.6 for the horizontal earth pressure and using 0.9 for
the earth and concrete weight); this practice will give a conservative result. However,
it seems to be within the intent of the ACI Code to use actual unfactored service loads
in establishing base dimensions for foundation structures (see ACI-13.3.1.1).
Using a 1-ft length of wall and letting the unit weight of material bounded by points
a, b, c, and d in Fig. 15.7.4.1 equal 120 pcf,
P1 = 0.256(20)(1) = 5.12 kips
1
P2 = (0.640)(20)(1) = 6.40 kips
2
W = 0.120(20 + 8) x = 3.36 x kips
Summation of moments about point b gives

 x
W   = P1 (10.0) + P2 (6.67)
 2
3.36 x 2
= 5.12(10.0) + 6.40(6.67)
2
1.668 x 2 = 93.9
x = 7.47 ft
Since for this granular material the resultant soil pressure is desired to be at the outer
edge of the middle third of the footing,
base length = 1.5 x = 1.5(7.47) = 11.2 ft
Try 11 ft 3 in.
Figure 15.7.4.1 Preliminary proportioning of cantilever wall in the design example 15.7.4.1.

(Continued)
605
(e) Stem thickness. Prior to computing soil pressures and stability factors of safety, a
more accurate knowledge of the concrete dimensions is necessary. The thickness at
the base of the stem is selected with due regard for the bending moment and shear
requirements.
Bending moment will normally provide the governing criterion, for which the gen-
eral expression (y measured from the top of the wall) is
0.256 y 2 0.032 y 3
My = + = 0.128 y 2 + 0.00533 y 3 (15. 7.4.1)
2 6
At the bottom of the assumed 18-ft high stem wall, using a 1.6 overload factor,
Mu = 1.6 0.128(18)2 + 0.00533(18)3  = 116 ft-kips
For a desired Rn = 410 psi,

Mu 116 (12, 000)
d= = = 17.7 in.
φ Rn b 0.90(410)12
Total thickness = 17.7 + 2 (cover) + 0.5 (estimated bar radius) = 20.2 in. Try 21 in. for
the stem base thickness and select 12 in. as the practical minimum for the top of the wall.
The selected Rn of 410 psi is a chosen guideline value and need not be rigidly adhered
to. In this case, the 3-in. multiple of 21 in. is preferred, and the somewhat thicker stem
will reduce deflection.
The critical section for shear strength may be permitted to be taken at a distance d
from the bottom of the stem. Conservatively, the shear at the base of the wall can be
used. The general expression for shear is
1
Vy = 0.256 y + (0.032) y 2 = 0.256 y + 0.016 y 2 (15.7.4.2)
2
At the base of the stem (y = 18 ft), using a 1.6 overload factor,
Vu = 1.6 0.256(18) + 0.016(18)2  = 15.7 kips
Using the provisions for one-

way shear strength of nonpresstressed members
(ACI-22.5.5.1),
(
φ Vc = φ 2 λ fc′ bd )
1
= 0.75[2(1) 3000 ](12)(≈ 18) = 17.7 kips > 15.7 kips OK
1000
The ACI Code is unclear concerning the out-of-plane shear reinforcement requirements
for cantilever retaining walls. However, cantilever walls have been traditionally treated
as a one-way slab in applying any ACI Code limitations. Accordingly, shear reinforce-
ment is not required for one-way slabs when Vu < φVc (ACI-7.6.3.1).
Where appearance on the front face is important, a batter of that face should be pro-
vided to counteract the effect of deflection. The usual batter is about 1 4 in./ft of wall
height. Thus the minimum batter here is
1 1
(18) = 4 in.
4 2
In this case, the thickness increases by 9 in. from the top to the bottom of the wall, so
batter the front face 5 in. and the rear face 4 in. When the wall is in place, it will deflect
(Continued)
60
Figure 15.7.4.2 Dimensions for computing resultant soil pressure in the design example.
forward and become nearly vertical; therefore, the analysis from this point will consider
the front face to be vertical. Economics may dictate having only one sloping face, or in
some cases no sloping face.
(f) Factor of safety against overturning. Using the dimensions given in Fig. 15.7.4.2,
locate the resultant of vertical forces (without overload factors) with respect to the
heel, as in Table 15.7.4.1:
105
resultant, from heel = = 3.95 ft
26.6
resisting moment = 26.6(11.25 − 3.95) = 194 ft-kips
overturning moment = P1 (10) + P2 (6.67)
= 5.12(10) + 6.40(6.67)
= 93.9 ft-kips

194
FS against overturning = = 2.07 > 2.0 OK
93.9
TABLE 15.7.4.1
Force Arm Moment
W1 = 0.120(18 +8)(6.5) = 20.3 3.25 66

1 1
W2 = 0.030   (18)(0.75) = 0.2 6.25 1
 2
W3 = 0.150(11.25)(2.0) = 3.4 5.63 19
W4 = 0.150(1.0)(18) = 2.7 7.00 19
Totals
26.6 105
1 This value corresponds to the difference between the concrete and retained material unit weights.
(Continued)
607
Alternatively, if a stability check is to be made using factored loads, the load com-
bination with 0.9D should be used with the horizontal earth pressure H (P1 and P2 in
Fig. 15.7.4.2) multiplied by a load factor of 1.6 in accordance with ACI-5.3.8(a), that is,
U = 0.9 D + 1.6 H
In such a case, it would be required that
resisting moment ≥ overturning moment
0.9(194) ≥ 1.6(93.9)
175 ≥ 150 OK
In effect, the required factor of safety against overturning under service load conditions
is 1.6 / 0.9 = 1.8, which is not much less than the traditional value of 2.0.
(g) Location of resultant and footing soil pressures. Referring to Fig. 15.7.4.3, and using
service loads because the maximum soil pressure limitation is given for that condition,
R = 26.6 kips
105 + 93.9
x= = 7.48 ft
26.6
11.25
e = 7.48 − = 1.85 ft
2
6e 6(1.85)
= = 0.99 < 1.0
base length 11.25
The resultant lies just inside of the middle third.

The service load soil pressure diagram is essentially a triangle; thus
1
R= ( pmax )(base length )
2
1
26.6 = ( pmax )(11.25)
2

pmax = 4.73 ksf < 5 ksf OK
Figure 15.7.4.3 Soil pressure and location of resultant under service load in the design example.
(h) Factor of safety against sliding. Neglecting passive pressure against the toe of the
footing and using service loads,
force causing sliding = P1 + P2 = 5.12 + 6.40 = 11.52 kips
frictional force = µ R = 0.40(26.6) = 10.64 kips
10.64
factor of safety = = 0.92 < 1.5 NG
11.52
(Continued)
608
Sliding resistance may also be checked using factored loads, U = 0.9D + 1.6H, in
which case it is required that
resisting force ≥ sliding force
0.9(10.64) ≥ 1.6(11.52)
9.58 kips < 18.4 kips
NG
Thus, ACI-5.3.8 would require the same factor of safety against sliding as against over-
turning, that is, 1.6 / 0.9 = 1.78. This exceeds the traditional factor to resist sliding of 1.5.
Since the resistance provided does not give an adequate safety factor, a key
(Figs. 15.7.4.4 and 15.7.4.5) against sliding is required. Such a key is intended to develop
passive pressure in the region in front of and below the bottom of the footing. The pro-
cedure for determining the size of a key is one that is debated by designers. Generally it
seems desirable to place the front face of the key about 5 in. in front of the back face of
the stem. This will permit anchoring the stem reinforcement in the key.
Various procedures may reasonably be used for granular cohesionless materials to
estimate the effect of the key. The maximum effect of a key would be to develop the pas-
sive resistance over the depth BC of Fig. 15.7.4.4, with any resulting failure occurring
along a curved path such as C’C. Fisher and Mains [15.26] have advocated the inert-
block concept of frictional resistance. In this method, any key used must extend deep
enough below the footing to develop an inert block of soil, BCDE of Fig. 15.7.4.5; it will
then have a failure plane approximately as C’CDGFH, so that the passive resistance is
developed over only the depth BC. In determining the passive resistance, the top 1 ft of
overburden is usually neglected in the height h1, of Fig. 15.7.4.4 or Fig. 15.7.4.5.
For this design, a factor of safety of 1.5 against sliding under service loads has
been considered proper. When passive resistance against the toe is also included
(Fig. 15.7.4.4), a higher factor of safety, say 2.0, should be used.
Figure 15.7.4.4 The passive resistance concept of frictional resistance.
Figure 15.7.4.5 The inert-block concept of frictional resistance.

(Continued)
609
Using the inert-block concept, the passive force Pp developed over the distance BC
of Fig. 15.7.4.5 is
equivalent fluid pressure = 400 pcf (given)
0.400(h1 + a )2 0.400(h1 )2
Pp = −
2 2
which for h1 = 3 ft gives
Pp = 0.200(6 a + a 2 )
When an inert block is induced, the frictional coefficient over the region CD becomes
tan α = tan 35° = 0.7, while the coefficient of 0.40 applies on DG and FH. Thus the fric-
tional resistance
F = µ1 R1 + µ 2 R2
 1  1
= 0.7   (4.73 + 2.59)(5.08) + 0.4   (2.59)((6.17)
 2  2

= 13.0 + 3.2 = 16.2 kips
Force equilibrium, incorporating a 1.5 factor of safety, gives
( P1 + P2 )1.5 = Pp + F
11.52(1.5) = 0.200(6 a + a 2 ) + 16.2
a 2 + 6 a − 5.40 = 0

a = 0.8 ft
Neglecting the frictional force in front of the key and considering the passive resist-
ance developed below the toe as shown in Fig. 15.7.4.4, the depth of key required may
also be computed as
b = 5.08 tan α = 5.08(0.7) = 3.55 ft
(h1 + a + b)2 0.400(h1 )2
PP = 0.400 −
2 2

= 0.200(a 2 + 13.1a + 33.9)
Force equilibrium, using a 1.5 safety factor, gives
( P1 + P2 )1.5 = Pp + µ 2 R2
 1
11.52(1.5) = 0.200(a 2 + 13 + 1a + 33.9) + 0.4   (2.59)(6.17)
 2
a 2 + 13.1a − 36.5 = 0

a = 2.4 ft
This would indicate that the latter method is more conservative. With this discussion
and these computations to serve as a guide, make the key depth 18 in. The key may be
made square; use 1 ft 6 in. × 1 ft 6 in. Reinforcement will rarely be required, but it is
common practice to extend some of the stem steel down into the key.
(Continued)
610
(i) Design of heel cantilever. The loads considered are included in Fig. 15.7.4.6, where
the effect of the base key is neglected.
For bending moment, the critical section is taken at the center of the stem steel
(2.5 in. from face of wall). It seems that any possible plane of weakness will occur at
the stem steel rather than at the face of support. Thus the downward uniform loading
due to earth overburden, surcharge, and footing concrete using an overload factor of
1.2 for the earth and footing concrete and 1.6 for the surcharge is, from Fig. 15.7.4.6,
wu = 1.2(2.16 + 0.30) + 1.6(0.96) = 4.49 kips/ft
1
Mu (downward) = (4.49)(5.96)2 = 79.7 ft-kips
2
w = 0.120 (18) = 2.16 kips/ft

2” cover (earth overburden)
= 0.120 (8) = 0.96 kips/ft
Stem reinforcement (surcharge)
w = 0.15(2) = 0.30 kips/ft

(footing)
Critical section
2’–0” for inclined 2’–0” Heel
cracking due (revised to 2’– 8”)
to shear
Base key
Neglect upward
pressure for
5.75 heel design
4.73 = 2.42 ksf
11.25
(under service loads)
5’–9”
Figure 15.7.4.6 Design of heel cantilever in the design example.
The upward soil pressure would reduce this value. However, the upward soil pressure
may not actually act in the linear fashion assumed; in fact, it might not be there at all.
Applying overload conditions under ACI-5.3.8, the most critical situation results under
the condition of 0.9D (i.e., gravity dead load) and 1.6H (i.e., horizontal earth pressure).
This would eliminate pressure under the heel entirely.
The critical section for shear may be taken at the distance d from the face when sup-
port reaction, in direction of applied shear, introduces compression into the end regions
of the member. In this case, tension is induced in the concrete where the heel joins the
stem, and an inclined crack could extend into the region ahead of the back face of the
stem. Thus, the critical section for shear must be taken at the face of the stem, as shown
in Fig. 15.7.4.6. The factored shear at the face without including upward soil pressure is
Vu = 4.49(5.75) = 25.8 kips
The design shear strength, unless shear reinforcement is used, is
φ Vc = φ (2 λ fc′)bd
1
= 0.75[2(1) 3000 ](12)(21.5) = 21.2 kips < 25.8 kips NG
1000
(Continued)
61
Shear appears to control. The effective depth must be increased in the ratio 25.8 / 21.2
unless shear reinforcement is to be used.
21.5(25.8)
21.2
Although slightly larger than required, use a heel thickness of 32 in. (d ≈ 29.5 in.).
For flexural strength, using Mu = 82 ft-kips (corrected for the weight of the 32-in.
footing),
Mu 82(12, 000 )
φ bd 2 0.90 (12)(29.5)2
This is less than the Rn corresponding to the minimum percentage of reinforcement

required for beams (see Fig. 3.8.1). The heel cantilever may be considered as a shal-
low foundation that must be designed according to the applicable ACI provisions
for beams and slabs (ACI-13.3.2.1). If treated as a one-way slab, the minimum rein-
forcement per ACI-7.6.1.1 should be provided. However, the retaining wall is a major
beamlike structure and use of the minimum reinforcement required for beams is recom-
mended. Increasing the heel thickness from the preliminary 24 in. to 32 in. reduces the
stem height and would permit reducing its thickness. The extra heel thickness will also
improve stability against overturning.
From Eq. (3.8.5), the required reinforcement ratio ρ to satisfy the factored loading is
about 0.0018 for Rn = 105 psi.
The minimum reinforcement requirement is, according to ACI-9.6.1.2,
3 fc′
As ,min = bw d [3.7.9]
fy
but not less than 200bw d /fy. Thus,
As ,min 3 fc′ 200   3 3000 200 

ρmin = = or = or 
bw d  f y f y   60, 000 60, 000 

= [0.0027 or 0.0033] = 0.0033
Less than 0.0033 may be used, provided the amount is at least one-third more than
required for strength (ACI-9.6.1.3); that is, 1.33(0.0018) = 0.0024.
required As = 0.0024(12)29.5 = 0.85 sq in./ft
Use #6@6 (As = 0.88 sq in./ft).

Regarding the development length for the #6 bars, the lateral spacing of 6 in. on centers
means clear spacing exceeding 2db. Also, the 2-in. clear cover exceeds db. Therefore,
condition 2 of Category A of the simplified equations applies (see Section 6.6). From
Table 6.6.1, Ld = 32.9 in. Because these bars are cast with more than 12 in. of concrete
beneath them, a casting position factor ψt = 1.3 must be used. Thus,
Ld (for #6) = 1.3(32.9) = 42.8 in. (simplified method)
(Continued)
612
This is a very long development length. Use the general equation, Eq. (6.6.1), to
calculate Ld :
 

3 f ψ t ψ e ψ s 
Ld = 
y
d [6.6.1]
 40 λ fc′  cb + K tr   b

 d  
 b 
where
Ld = development length
db = nominal diameter of bar or wire
cb = cover or spacing dimension
= the smaller of (1) distance from center of bar being developed to the nearest
concrete surrface, and (2) one-half the center-to-center spacing of baars being
developed
In this situation, ψt = 1.3, ψe = 1.0, and λ = 1.0. The ACI Code permits a size factor ψs
of 0.8 for #6 and smaller bars, though ACI Committee 408 does not support the use of
this factor (see Section 6.7). Here, the authors will adopt the latter, more conservative
approach, and use ψs = 1.0. Since there is no shear reinforcement, Ktr = 0. The dimension
cb for cover or spacing is controlled by the cover,
cb = 2 (i.e.,clear cover ) + 0.375 ( i.e., bar radius) = 2.375 in.
thus, (cb + Ktr)/db = 2.375/0.75 = 3.2 > 2.5 max. Using Eq. (6.6.1) with λ = ψe = ψs = 1.0
and ψ t = 1.3,
 

3 fy ψ t ψ e ψ s   3 60, 000 1.3 
Ld =  db =  0.75 = 32 in. (2.7 ft)
 40 λ fc′  cb + K tr    40 1.00 3000 2.5 

 d  
 b 
Thus, the #6 bars in the heel must be embedded 32 in. to develop their full strength. This is
measured as the distance beyond the main vertical reinforcement in the back face of the wall.
Use an embedment of 2.7 ft from back face of wall.
( j) Design of toe cantilever. The toe of the footing is also treated as a cantilever beam,
with the critical section for moment at the front face of the wall and the critical
section for shear (inclined cracking) at a distance d (approximately 20.5 in.) from
the front face of the wall. Shear will usually control the required toe thickness. The
thickness need not be the same as the heel, though many engineers would make
them the same. In this example the heel is unusually thick because of the heavy sur-
charge. Try a toe thickness somewhat less than the heel, say, 2 ft.
Referring to Fig. 15.7.4.7 and neglecting the earth on the toe, the factored shear and
bending moment are
 4.73 + 3.87 
Vu = 1.6  − 0.300 (2.04) = 13.1 kips
 2 
1  2 1  1 1 
Mu = 1.6  (4.73)(33.75)2   + (3.15)(3.75)2   − (0.300)(3.75)2 
2  3 2  3 2 
 (3.75)2 
= 1.6  (9.46 + 3.15 − 0.90) = 43.9 ft-kips
 6 
(Continued)
613
Critical section for inclined

cracking due to shear
2.04’
3’–9”
w = 0.30 kips/ft for

concrete footing only
2’–0” 3” cover
3.15 ksf
3.87 ksf
4.73 ksf Service load stresses
Figure 15.7.4.7 Design of toe cantilever in the design example.
where the toe soil pressure is considered primarily the result of horizontal earth pres-
sure; hence the use of the 1.6 factor [ACI-5.3.8(a)].
43.9(12, 000)
0.90(12)(20.5)2
From Eq. (3.8.5), the required ρ ≈ 0.002 is less than the minimum reinforcement ratio
[computed in part (i)] of 0.0033. Applying ACI-9.6.1.3,
minimum ρ = 1.33(0.002) = 0.0027
required As = 0.0027(12)(20.5) = 0.65 sq in./ft
Use #6 @ 8 (As = 0.66 sq in./ft).

The shear strength is
φ Vc = φ (2 λ fc′)bd
1
= 0.75[2(1) 3000 ](12)(20.5) = 20.2 kips > 13.1 kips OK
1000
The #6 bars in the heel have wide lateral spacing and 3-in. cover; clearly the maximum
(cb + Ktr)/db = 2.5 applies. Using Eq. (6.6.1) with ψt = ψe = ψs = λ = 1.0,
 

3 fy ψ t ψ e ψ s   3 60, 000 1.0 
Ld =  db =   0.75 = 24.6 in. (2.1ft)
 40 λ fc′  cb + K tr    40 1.00 3000 2.5 

 d  
 b 
Use embedment of 2.1 ft from front face of wall.
(Continued)
614
(k) Reinforcement for wall. The wall height (17.33 ft) is now 8 in. less than that used
in the preliminary design calculations. Retain the 21-in. thickness at the base of the
stem. The factored moment diagram, Mu, is shown in Fig. 15.7.4.8, calculated using
Eq. (15.7.4.1) with maximum y = 17.33 ft along with an overload factor of 1.6. Also
shown in the figure is the design strength diagram, φ Mn, of the selected wall steel.
The steel area required at the base of stem is
106(12, 000)
0.90(12)(18.5)2
required As = 0.006(12)(18.5) = 1.33 sq in./ft
Use #6 and #7 bars alternated @4 1 2 in. on centers (As = 1.39 sq in./ft). The shear has
been checked and found satisfactory.
The embedment of the bars into the footing must equal or exceed the development
length Ld. If all of the wall steel is run the entire height of the wall, vertical dowels
extending above the footing must be used, since the footing is placed and cast first and
the bars cannot extend 18 ft or so out of it. When bars extend out too far, they are often
bent or broken off during construction. Also, in an endeavor to economize, the quantity
of steel per foot of wall should be reduced in the upper parts of the wall so that the design
strength approximately equals that required over the height of the wall. In this design, it
is proposed that the #7 bars be extended up to 8 to 10 ft out of the footing (dowel bars).
These bars may be terminated at a shorter distance where they are no longer required to
satisfy flexural strength requirements, provided they can be developed. The #6 bars, on
the other hand, will be lap spliced above the footing into the wall.
(k.1) Embedment length into the footing, Ld (all bars). Ld for embedment into the
footing can be based on the clear spacing because cover is not involved.
For the #7 bars having the adjacent #6 bars centered 4 1 2 in. away,
 0.875 0.75 
4.5 − −
 2 2  d = 4.21d
clear spacing =   b b
0.875
 
 
Since no transverse reinforcement is provided, Ktr = 0, and thus (cb + Ktr)/db = 4.21 >

2.5 max. Using Eq. (6.6.1) with λ = ψ t = ψ e = ψ s = 1.0 ,
 

3 fy ψ t ψ e ψ s   3 60, 000 1.0 
Ld =  d = 0.875 = 28.8 in. (2.4 ft)
 40 λ fc′  cb + K tr   b  40 1.00 3000 2.5 

 d  
 b 
Use an embedment of 2 ft 6 in. into the footing for both the #6 and #7 bars. The base key
is available for embedment of stem bars if it might be necessary.
(k.2) Embedment length into the wall (#7 bars). When dowels are used, the clear
cover of 2 in. over the portion of the dowels extending above the footing into the wall
will control the development length Ld. Thus, for the general equation, Eq. (6.6.1),
cb = 2 (i.e., clear cover) + 0.438 (i.e., bar radius) = 2.44 in.
Thus (cb + Ktr)/db = 2.79 > 2.5 max, and Ld = 2.4 ft (same length as into the footing).
(Continued)
615
(k.3) Splice length above the footing (#6 bars). Since it is proposed to lap the #6 bars
at the base of the wall, ACI-25.5 must be applied to determine the required lap distance.
When no more than half of the bar area is to be lap spliced within the lap length (less
than half of the total bar area is to be spliced in this case), the tension splice must meet
the requirements for a Class A splice. Referring to Table 6.16.1 (which summarizes ACI-
25.5.2.1), the lap required is equal to Ld (or 12 in. min), where Ld is the development
length required for unspliced bars. The #6 bars to be spliced are located 9 in. on centers.
The #7 bars 4.5 in. away are not being developed within the splice region; thus they do
not influence the development length for the bars being spliced. The clear spacing is
9 − 2(0.75)
clear spacing = db = 10 db
0.75
allowing an extra bar diameter for the lap splice. Clearly, (cb + Ktr)/db is controlled by
the 2-in. clear cover to the tension steel at the back face of the wall. Thus, for the #6 bars
cb = 2 (i.e., clear cover) + 0.375 (i.e., bar radius) = 2.38 in.
Thus, (cb + Ktr)/db = 2.38 < 2.5 max., and Ld = 25.9 in.

Use a splice length of 2 ft 3 in. for the #6 bars into the wall terminating at point A in
Fig. 15.7.4.8 above the footing.
(k.4) Bar cutoff locations. Proceeding from the stem base, the #7 bars embedded in
the footing should not be extended more than about 8 to 10 ft out of the footing, as noted
earlier. Bar spacing and cutoff are done in accordance with the ACI Code, by drawing
the φ Mn diagram as described in Chapter 6.
For #6 and #7 bars alternated at 4 1 2 in., the design strength is
C = 0.85(3)(12)a = 30.6 a
T = 1.39(60) = 83.4 kips

a = 2.72 in.
At top of wall,
φ M n = 0.90(83.4) [ 9.5 − 0.5(2.72)] 1

= 50.9 ft-kips
12
At bottom of wall,
φ M n = 0.90(83.4) [18.5 − 0.5(2.72)] 1

= 107 ft-kips
12
(Note: It may be readily verified that both sections are tension controlled, and thus φ = 0.9.)
The design strengths computed above are used to locate the outer dashed line in
Fig. 15.7.4.8.
Next, use the moment diagram (Fig. 15.7.4.8) to locate the point where the #7 bars
may be terminated, leaving the remaining design strength, φ Mn, based on #6 @ 9 in.
This design strength is shown in Fig. 15.7.4.8 by an inclined dashed line that is plotted
by using the following φ Mn values:
C = 30.6 a, T = 0.59(60) = 35.4 kips, a = 1.16 in.
φ M n (at top) = 0.9(35.4) [ 9.5 − 0.5(1.16)] 1

= 23.7 ft-kips
12
φ M n (at bottom ) = 0.9(35.4) [18.5 − 0.5(1.16)] 1

= 47.6 ft-kips
12
(Continued)
61
23.7 50.9 d = 9.5”

Ld = 25.9”
for #6 bars
16
14 #6 @ 18”
#6 and #7
1
alternate @ 4 2 ”
12 #6 @ 9”
Distance from base of stem, ft
C d = 1.14’ > 12 diam

10 Ld = 25.9” for #6 d = 13.6”
Terminate every other #6 bar
8
d = 1.28’ > 12 diam
B
Ld = 28.8” for #7 d = 15.4”

6
10’–6”
φMn diagram
Terminate #7 bars
Factored
7’–3”
4
moment
diagram
A
2 2’–3”
End of
lap
splice 107
0
47.6 106 d = 18.5”
Factored bending moment Mu (ft-kips)
Figure 15.7.4.8 Determination of stem reinforcement for the design example.
The termination point B is found by extending beyond the intersection of the remain-
ing φ Mn line (#6 @ 9 in.) with the factored moment diagram a distance of either the
effective depth d or 12 bar diameters, whichever is greater. In this case, the calculated
termination point and the practical location in 3-in. increments approximately coincide
(they differ by less than 1 2 in.)
Whenever tension bars are terminated in a tension zone, reduction in shear strength
and ductility may occur. For this reason, the ACI Code does not permit flexural rein-
forcement in beams or one-way slabs to be terminated in a tension zone unless addi-
tional conditions are satisfied. Chapter 11 of the ACI Code is silent with respect to this
requirement in cantilever retaining walls. Since the cantilever wall is assumed to behave
as a beamlike structure, the provisions for cutting bars in the tension zone of beams will
be used (see Section 6.11).
Since stirrups are not used in retaining walls, the conditions, one of which must be
satisfied in this case, are (1) that the factored shear Vu at the cutoff point not exceed two-
thirds of the shear strength φVn and (2) that the continuing bars provide at least twice the
area required for the bending moment at the cutoff point and that the factored shear Vu
at the cutoff point not exceed three-fourths of the shear strength φVn.
From inspection of Fig. 15.7.4.8, it is clear that since the continuing bars provide
less than twice the area required at the cutoff point, condition (2) above is not satisfied.
(Continued)
617
Check to see if item (1) above is satisfied. Using a 1.6 overload factor and Eq. (15.7.4.2),
the shear at y = 10.12 ft (i.e., 17.33 ft –7.21 ft) from the top is
Vu = 1.6 0.256(10.12) + 0.016(10.12)2  = 6.8 kips
(
φVc = φ 2λ fc′bd )
1
= 0.75[2(1) 3000 ](12)(14.8) = 14.6 kips
1000
2
(14.6) > 6.8 OK
3
Thus, terminate #7 bars at 7 ft 3 in. from the top of heel.

Additional economy may be achieved by cutting every other #6 bar, leaving #6 @
18 in. to extend to the top of the wall (According to ACI-11.7.2.1, the longitudinal rein-
forcement in cast-in-place walls shall be spaced not farther apart than 3 times the wall
thickness nor more than 18 in.). Figure 15.7.4.8 shows the actual cutoff point C for every
other #6. In this case, the extension d gives a point slightly below point C, since the
actual cut point is located to produce bar lengths in the usual 3-in. increments. Terminate
every other #6 at 10 ft 6 in. from the top of heel. A tension zone cutoff check at point C
(y = 6.83 ft from top; i.e., 17.33 –10.5) shows
Vu = 1.6 0.256(6.83) + 0.016(6.83)2  = 4.0 kips
(
φVc = φ 2 λ fc′bd )
1
= 0.75[2(1) 3000 ](12)(13.0) = 12.8 kips
1000
2
(12.8) > 4.0 OK
3
The complete bar arrangement for the stem is shown later (see Fig. 15.7.4.9).
The reader may note that the moment decreases rapidly from the base of the stem
toward the top of the wall. At about 5 ft from the base, the factored moment has decreased
about 50%. Somewhere in this vicinity, the reinforcement ratio ρ that is required for
strength equals the minimum of ACI-9.6.1.2 [ρmin = 0.0033 from part (i)]. For the upper
portion of the wall, ACI-9.6.1 would require the minimum ρ to be either 0.0033 or four-
thirds of the required ρ based on the factored moment. For example, the design strength
φ Mn at point C is indeed approximately equal to four-thirds of the factored moment at C.
Thus, in general, wherever actual ρ is less than ρmin, the requirement of ACI-9.6.1 may
be checked by observing whether the design strength diagram is offset at one-third or
more from the factored moment curve.
It is further noted that the stem has been designed for bending moment and shear
only. However, the weight of the stem causes compression in itself, so that, strictly
speaking, it might be treated as a compression member (under large moment) in accord-
ance with the concepts of Chapter 10. In this design problem the compressive force Pu
in the wall under factored loads would be
 12 + 21
Pu = 1.2  (0.5)(17.33)(0.15) = 4.3 kips
 12 
(Continued)
618
The combination of this small Pu with Mu = 106 ft-kips will give a design near the bot-
tom of the tension-controlled region (see Chapter 10). In general, treatment of such
walls as compression members is rarely justified but, if done, it might permit some slight
reduction in reinforcement or thickness of the section.
(l) Minimum Horizontal Reinforcement. Horizontal bars along the length of the wall
must be provided in accordance with ACI-11.1.4 and 11.6. Accordingly, the mini-
mum required amount if #5 or smaller bars are used is (see Table 15.2.1)
As = 0.0020bh = 0.0020(12)(16.5) = 0.40 sq in./ft
where the average wall thickness is used for h.

Since it is primarily the front face that is exposed to temperature changes, more of
the horizontal reinforcement should be placed there. Thus, it is suggested that about
two-thirds of this reinforcement be placed in the front face and one-third in the rear face.
Accordingly,
2 2
As = (0.40) = 0.27 sq in./ft
3 3
1 1
As = (0.40) = 0.13 sq in./ft
3 3
Use on the front face #5 @ 12 (As = 0.31 sq in./ft).

Use on the rear face #4 @ 14 (As = 0.17 sq in./ft).
For vertical reinforcement on the front face, use any nominal amount that is adequate
for supporting the horizontal temperature and shrinkage steel in that face; the minimum
amount specified in ACI-11.6.1 is a good guide. Using an average wall thickness of 16.5
in. and assuming #6 bars (see Table 12.5.1),
min As = 0.0015(16.5)(12) = 0.30 sq in./ft
Use #6 @ 1 ft 6 in. spacing.

(m) Drainage and other details. Adequate drainage of backfill must be provided because
the pressures used are for drained material. A common minimum provision is for
weep holes (say, 4-in. diameter tile) every 10 to 15 ft along the wall.
Construction of a retaining wall is accomplished in at least two stages; the footing is
placed first and then the wall. A shear key is usually desirable for positive shear transfer
between wall and footing (see Fig. 15.7.4.9). Such a key is made by embedding a bev-
eled 2 × 4 or 2 × 6 timber in the top of the footing. This key may be designed using the
shear-friction provisions of ACI-22.9 (discussed in Section 5.15). In this situation, it is
not appropriate to consider shear as a measure of diagonal tension. The nominal shear
strength of the shear key is based on a nominal stress of 0.2 fc′ but not to exceed 800 psi,
as given by ACI-22.9.4.4.
This design uses no bent bars. Frequently, the bar arrangement in the toe can be con-
veniently bent up into the stem and meshed with additional stem reinforcement to give
an economical system.
(n) Design sketch. The final details of this design are presented in a design sketch
(Fig. 15.7.4.9) which must accompany any set of design computations.
(Continued)
619
1’–0”
5” 4”
1 12 ” clear cover 2”clear cover

#6 @ 18” × 17’–3”
12”
14”
#5 (typical)
#4 (typical)
#6 @ 1’–6”
For main #6 @ 18” × 10’–6”
reinforcement,
18’–0” see typical
pattern in 10’–6”
projection
4” Ø weep holes
#7 @ 9 × 9’– 9”
@ 10’–0” spacing
along wall
7’–3”
1’–9” #6 @ 6 × 8’– 9”
Ground surface
2’–3”
2” × 6” keyway
3’–9” 5’–9”
4’–0”
2’–0” 2’–8”
2” clear cover
#6 @ 8 × 5’–9” 3” clear cover #4 as shown #6 @ 9”× 4’–9”
1’–6” 1’–6”
5’–1” 4’–8”
11’–3”
Figure 15.7.4.9 Design sketch for the cantilever retaining wall.
SELECTED REFERENCES
15.1. K. M. Kriparanayan. “Interesting Aspects of the Empirical Wall Design Equation,” ACI Journal,
74, May 1977, 204–207.
15.2. Garold D. Oberlender and Noel J. Everard. “Investigation of Reinforced Concrete Walls,” ACI
Journal, 74, June 1977, 256–263.
15.3. Sharon L. Wood. “Performance of Reinforced Concrete Buildings During the 1985 Chile
Earthquake: Implications for the Design of Structural Walls,” EERI Earthquake Spectra, 7, No.
4, 1991, 607–638.
15.4. Burcu Burak and Hakki G. Comlekoglu. “Effect of Shear Wall Area to Floor Area Ratio on the
Seismic Behavior of Reinforced Concrete Buildings,” ASCE Journal of Structural Engineering,
139, No. 11, November 2013, 1928–1937.
15.5. K. N. Shiu, J. I. Daniel, J. D. Aristizabal-Ochoa, A. E. Fiorato, and W. G. Corley. “Earthquake
Resistant Structural Walls—Tests of Walls with and without Openings.” Skokie, IL. Construction
Technology Laboratories, Portland Cement Association (PCA), July 1981, 123 pp.
15.6. Aejaz Ali and James K. Wight, “RC Structural Walls with Staggered Door Openings,” ASCE
Journal of Structural Engineering, 117, No. 5, May 1991, 1514–1531.
15.7. Christopher P. Taylor, Paul A. Cote, and John Wallace. “Design of Slender Reinforced Concrete
Walls with Openings,” ACI Structural Journal, 95, No. 4, July–August 1998, 420–433.
15.8. F. V. Yanez, R. Park, and T. Paulay. “Seismic Behavior of Reinforced Concrete Structural Walls
With Regular and Irregular Openings,” Pacific Conference on Earthquake Engineering, New
Zealand, November 1991.
15.9. Thomas Paulay. “The Design of Ductile Reinforced Concrete Structural Walls for Earthquake
Resistance,” Earthquake Spectra, 2, No. 4, 1986, 783–823.
620
15.10. R. G. Oesterle, A. E. Fiorato, L. S. Johal, J. E. Carpenter, H. G. Russel, and W. G. Corley.
“Earthquake Resistant Structural Walls—Tests of Isolated Walls.” Skokie, IL: Construction
Technology Laboratories, Portland Cement Association (PCA), November 1976, 328 pp.
15.11. John H. Thomsen and John W. Wallace, “Displacement-Based Design of Slender Reinforced
Concrete Structural Walls—Experimental Verification,” ASCE Journal of Structural
Engineering, 130, No. 4, April 2004, 618–630.
15.12. Thomas Paulay. “The Shear Strength of Shear Walls,” Bulletin of the New Zealand Society of
Earthquake Engineering, 3, No. 4, December 1970, 148–162.
15.13. Sharon L. Wood. “Shear Strength of Low-Rise Reinforced Concrete Walls,” ACI Structural
Journal, 87, January–February, 1990, 99–107.
15.14. Felix Barda, John M. Hanson, and W. Gene Corley, “Shear Strength of Low-Rise Walls
with Boundary Elements,” Reinforced Concrete Structures in Seismic Zones, ACI SP53-8.
15.15. H. A. Williams and J. R. Benjamin. “Investigation of Shear Walls, Part 3, Experimental and
Mathematical Studies of the Behavior of Plain and Reinforced Concrete Walled Bents Under
Static Shear Loading.” Department of Civil Engineering, Stanford University, Stanford, CA,
June 1953, 142 pp.
15.16. J. R. Benjamin and H. A. Williams. “Investigation of Shear Walls, Part 6, Continued Experimental
and Mathematical Studies of Reinforced Concrete Walled Bents Under Static Shear Loading.”
Department of Civil Engineering, Stanford University, Stanford, CA, August 1954, 59 pp.
15.17. J. Antebi, S. Utku, and R. J. Hansen, “The Response of Shear Walls to Dynamic Loads,” MIT
Department of Civil and Sanitary Engineering, (DASA-1160). Cambridge, MA, August 1960,
290 pp.
15.18. Thomas Paulay, M. J. Nigel Priestley, and A. J. Synge, “Ductility in Earthquake Resisting
Squat Shear Walls,” ACI Journal, 79, July–August 1982, 257–269.
15.19. Pedro A. Hidalgo, Christian A. Ledezma, and Rodrigo M. Jordan, “Seismic Behavior of Squat
Reinforced Concrete Shear Walls,” EERI Earthquake Spectra, 18, No. 2, May 2002, 287–308.
15.20. A. E.Cardenas, and D. D. Magura. “Strength of High-
Rise Shear Walls— Rectangular
Cross Sections,” Response of Multistory Concrete Structures to Lateral Forces, ACI SP-36.
Detroit: American Concrete Institute, 1973, (pp. 119–150).
15.21. A. E., Cardenas, J. M., Hanson, W. G., Corley, and E. Hognestad. “Design Provisions for Shear
Walls,” ACI Journal, Proceedings, 70, March 1973, 221–230.
15.22. K. Muto, and K. Kokusho. “Experimental Study on Two-Story Reinforced Concrete Shear
Walls,” Muto Laboratory, University of Tokyo, Japan. Translated by T. Akagi, University of
Illinois, Urbana, August 1959.
15.23. K. Ogura, K. Kokusho, and N. Matsoura. “Tests to Failure of Two-Story Rigid Frames with
Walls, Part 24, Experimental Study No. 6,” Japan Society of Architects Report No. 18, February
1952. Translated by T. Akagi, University of Illinois, Urbana, August 1959.
15.24. Karl Terzaghi, Ralph B. Peck and Gholamreza Mesri. Soil Mechanics in Engineering Practice
(3rd ed.). New York: John Wiley, & Sons, 1996.
15.25. Whitney Clark Huntington. Earth Pressures and Retaining Walls. New York: John Wiley &
Sons, 1957.
15.26. G. P. Fisher and R. M. Mains. “Sliding Stability of Retaining Walls,” Civil Engineering, July
1952, 490.
PROBLEMS
Shear Walls
15.1 For the wall of Example 15.5.3, use the ACI height, determine the level at which the concrete
Code provisions to compute the amount of ver- strength may be reduced from 5,000 psi to 4,000
tical and horizontal reinforcement required in psi for the wall of Example 15.5.3.
the first story if normal-weight concrete with 15.3 Assuming that the wall thickness and the amount
fc′ = 8, 000 psi and Grade 60 steel is used. of vertical reinforcement will remain the same
15.2 Since the required strength of shear walls varies over the height, determine the level at which the
significantly with height, it is common practice concrete strength may be reduced from 8,000 psi
to vary the concrete strength or the amount of to 6,000 psi for the wall of Problem 15.1. If the
reinforcement, or both, while maintaining wall concrete strength were to be further reduced in
thickness constant over the height. Assuming the upper levels of the wall, determine the level
that the wall thickness and the amount of verti- at which concrete strength could be reduced to
cal reinforcement will remain the same over the 4,000 psi.
621
PROBLEMS 621
15.4 Use the ACI Code provisions to determine the determined probably be permissible if the
adequacy of the barbell wall of the figure for foundation material is rock?
Problem 15.4. The wall is subjected to an axial 15.6 Determine the adequacy of the retaining wall of
compressive, factored load, Pu, of 1350 kips act- the figure for Problem 15.6 with regard to stability
ing concurrently with a factored moment, Mu, of (overturning, sliding, and soil pressure magnitude
14,900 ft-kips and an in-plane, factored shear and distribution). Assume that the wall is on good
force, Vu, of 120 kips. Vertical reinforcement in granular soil with a maximum safe soil pressure
the boundary elements consist of 8–#8 bars. In of 5000 psf under service load conditions.
the web, both the vertical and horizontal rein- 15.7 Reconsider the retaining wall conditions of
forcement consist of #5 bars spaced at 18 in. on Section 15.7.4 if the surcharge is changed to 10 ft
centers. Use fc′ = 6, 000 psi and Grade 60 steel. to approximate the effect of a railroad located par-
allel to and near the top of the wall. Sometimes
Cantilever Retaining Walls under such conditions the lateral effect of a sur-
Note: For all problems assume that frost pen charge is included but the beneficial stabilizing
etration depth is 4 ft. effect of the vertical surcharge weight is omitted.
15.5 Given the cantilever retaining wall of the figure Compare the wall proportions for both condi-
for Problem 15.5 with the active pressure coef- tions, using a minimum length of base and keep-
ficient Ca = 0.27, the weight of retained soil = ing the soil pressure resultant under service load
100 pcf, and the coefficient of friction between conditions within the middle third of the base.
concrete and earth = 0.40, use service load con- Verify the adequacy of the dimensions, but do not
ditions to determine the following. design the reinforcement. Use fc′ = 4000 psi, fy =
(a) The factor of safety provided against overturn- 60,000 psi, and the ACI Code.
ing. Would you consider this to be adequate? 15.8 Assuming that the overall proportioning of the
(b) The factor of safety provided against slid- retaining wall of Problem 15.6 is adequate for
ing. Neglect any passive pressure resistance earth stability, design the reinforcement for the
on the toe. Would you consider this factor of wall cantilever. Besides dowels at the base, use
safety adequate? three changes of reinforcement over the 20-ft wall
(c) The location of the resultant bearing pres- height. Draw the resulting design strength φMn
sure under footing. Is it within the middle diagram superimposed on the factored moment
third of the base? Would the position you Mu diagram. Use fc′ = 4000 psi, fy = 60,000 psi.
8 #8 8 #8
#5 @ 18
Problem 15.4
12”
Level
Equivalent
fluid active
pressure = 29 pcf
Weight of
Retained 20’–0’’ retained
material material = 100 pcf
22’–0’’
Equivalent
Assumption 1’–6” fluid passive
for 15’–0” pressure = 300 pcf
computing
stability
2’–6”
2’–0”
2’–0’’
1’–4’’ 5’–5’’ 1’–4’’
2’– 0’’
2’– 8’’ 6’–4’’
6’–0” 2’–0”
11’–0’’
CHAPTER 16
DESIGN OF TWO-WAY
FLOOR SYSTEMS
16.1 GENERAL DESCRIPTION
In reinforced concrete buildings, a basic and common type of floor is the slab-beam-girder
construction, which was treated in Chapters 8 and 9. As shown in Fig. 16.1.1(a), the shaded
slab area is bounded by the two adjacent beams on the sides and portions of the two girders
at the ends. When the length of this area is two or more times its width, almost all of the
floor load is transferred to the beams, and very little, except some near the edge of the gird-
ers, is transferred directly to the girders. Thus the slab may be designed as a one-way slab
as treated in Chapter 8, with the main reinforcement parallel to the girder and the shrinkage
and temperature reinforcement parallel to the beams. The deflected surface of a one-way
slab is primarily one of curvature in its short direction.
When the ratio of the long span L to the short span S as shown in Fig. 16.1.1(b) is less
than about 2, the deflected surface of the shaded area becomes one of curvature in both
directions. The floor load is carried in both directions to the four supporting beams around
the panel; hence the panel is a two-way slab. Obviously, when S is equal to L, the four
beams around a typical interior panel should be identical; for other cases, the longer beams
carry more load than the shorter beams.
Flat slab with drop panels and capitals. Parking garage of Helen C. White Hall, University of
Wisconsin–Madison. (Photo by José A. Pincheira.)
623
16.1 GENERAL DESCRIPTION 623
Two-way floor systems may also take other forms in practice. Figure 16.1.2 shows flat
slab and flat plate floor construction, which are characterized by the absence of beams
along the interior column lines; edge beams may or may not be used at the exterior edges
of the floor.
Flat slab floors differ from flat plate floors in that slab floors provide adequate shear
strength by having either or both of the following: (a) drop panels (i.e., increased thickness
of slab) in the region of the columns or (b) column capitals (i.e., tapered enlargement of the
upper ends of columns). In flat plate floors a uniform slab thickness is used and the shear
strength is obtained by the embedment of multiple-U stirrups [Fig. 16.16.1(a)], structural
steel devices known as shearhead reinforcement [Fig. 16.16.1(b)] or, more commonly used
today, headed shear stud reinforcement [Fig. 16.16.2] within the slab of uniform thickness.
Relatively speaking, flat slabs are more suitable for larger panel size or heavier loading than
flat plates.
Historically [16.2], flat slabs predate both two-way slabs on beams and flat plates. Flat
slab floors were originally patented by O. W. Norcross in the United States on April 29,
1902. Several systems of placing reinforcement have been developed and patented since
then—the four-way system, the two-way system, the three-way system, and the circumfer-
ential system. C. A. P. Turner was one of the early advocates of a flat slab system known
as the “mushroom” system. About 1908, the flat slab began being recognized as an accept-
able floor system, but for many years designers were confronted with difficulties of patent
infringements.
G B
B B B B S
G B
(a) (b)
Figure 16.1.1 One-way and two-way slabs.
Drop panel
Column capital
(b) Flat plate floor
Column Drop panel

capital
(a) Flat slab floor
Figure 16.1.2 Flat slab and flat plate floor construction.

624
624 C hapter 1 6 D esign of T wo - W ay F loor S ystems
The terms two-way slab [Fig. 16.1.1(b)], flat slab [Fig. 16.1.2(a)], and flat plate
[Fig. 16.1.2(b)] are arbitrary. Following tradition, the implication is that in two-way slabs
there are beams between columns, but no such beams, except perhaps for edge beams
along the exterior sides of the entire floor area in flat plates. In flat slabs, on the other
hand, the slab depth near the columns is increased to provide additional shear and bending
resistance.
If methods of structural analysis and design are developed for two-way slabs with
beams, many of these general provisions should apply equally well to flat slabs or flat
plates. Historically, until 1971, the design of two-way slabs supported on beams was treated
separately from that of flat slabs or flat plates without beams. Various empirical procedures
have been proposed and used [16.6–16.8]. Chapter 8 of the 2014 ACI Code takes an inte-
grated view and refers to two-way slab systems with or without beams. In addition to solid
slabs, hollow slabs with interior voids to reduce dead weight, slabs (such as waffle slabs)
with recesses made by permanent or removable fillers between joists in two directions, and
slabs with paneled ceilings near the central portion of the panel are also included in this
category (ACI-8.1.3).
Thus the term two-way floor systems (rather than the term two-way slab systems as in
the ACI Code) is used in this book to include all three types of floor systems: the two-way
slab with beams, the flat slab, and the flat plate.
16.2 GENERAL DESIGN CONCEPT OF THE ACI CODE

The basic approach to the design of two-way floor systems involves imagining that verti
cal cuts are made through the entire building along lines midway between the columns.
The cutting creates a series of frames whose width lies between the centerlines of the
two adjacent panels as shown in Fig. 16.2.1. The resulting series of rigid frames, taken
separately in the longitudinal and transverse directions of the building, may be treated for
gravity loading floor by floor, as would generally be acceptable for a rigid frame structure
consisting of beams and columns, in accordance with ACI-6.3.1.2. A typical rigid frame
would consist of (1) the columns above and below the floor and (2) the floor system, with
or without beams, bounded laterally between the centerlines of the two panels (one panel
for an exterior line of columns) adjacent to the line of columns. Thus, the design of a two-
way floor system (including two-way slab, flat slab, and flat plate) is reduced to that of an
equivalent frame.
As in the case of design of actual rigid frames consisting of beams and columns, approx-
imate methods of analysis may be suitable for many usual floor systems, spans, and story
heights. As treated in Chapter 7, the analysis for actual frames could be (a) approximate,
using the moment and shear coefficients of ACI-6.5, or (b) more accurate, using structural
analysis after assuming the relative stiffnesses of the members.
For gravity load only and for floor systems within the specified limitations, the moments
and shears on these equivalent frames may be determined (a) by approximately using
moment and shear coefficients prescribed by the Direct Design Method of ACI-8.10 or
(b) by structural analysis in a manner similar to that for actual frames, using the special
provisions of the Equivalent Frame Method of ACI-8.11. It is noted that an elastic analysis
(such as by the equivalent frame method) must be used for lateral loads even if the floor
system meets the limitations of the direct design method for gravity load.
The equivalent rigid frame is the structure being dealt with whether the moments are
determined by the Direct Design Method (DDM) or by the Equivalent Frame Method
(EFM). These two ACI Code terms describe two ways of obtaining the longitudinal varia-
tion of bending moments and shears.
When the EFM is used for obtaining the longitudinal variation of moments and shears,
the relative stiffness of the columns, as well as that of the floor system, can be assumed in
the preliminary analysis and then reviewed, as is the case for the design of any statically
indeterminate structure. Design moment envelopes may be obtained for dead load in com-
bination with various patterns of live load, as described in Chapter 7 (Section 7.3). In lateral
625
1 6 . 3 TOTA L FAC TO R E D S TAT I C M O M E N T 625
rigid frame
Interior equivalent frame floor area
equivalent
Width of
Figure 16.2.1 Tributary floor area for an interior equivalent rigid frame of a two-way floor system.
load analyses, moment magnification in columns due to sidesway of vertical loads must be
taken into account as prescribed in ACI-6.6.4, 6.7, or 6.8 (see Chapter 13).
Once the longitudinal variation in factored moments and shears has been obtained,
whether by the DDM or EFM, the moment across the entire width of the floor system being
considered is distributed laterally to the beam, if used, and to the slab. The lateral distribu-
tion procedure and the remainder of the design are essentially the same whether the DDM
or EFM has been used.
The accuracy of analysis methods utilizing the concept of dividing the structure into
equivalent frames has been verified for gravity load analysis by tests [16.12–16.25] and
analytical studies [16.26–16.35]. For lateral load analysis, where there is less agreement
on procedure, various studies have been made, including those of Pecknold [16.38], Allen
and Darvall [16.39,16.47], Vanderbilt [16.32, 16.40], Elias [16.41–16.43], Fraser [16.44],
Vanderbilt and Corley [16.45], Lew and Narov [16.46], Pavlovic and Poulton [16.48],
Moehle and Diebold [16.49], Hsu [16.50], Cano and Klingner [16.51], Hwang and Moehle
[16.145], and Dovich and Wight [16.146].
16.3 TOTAL FACTORED STATIC MOMENT

Consider ABCD and CDEF, two typical interior panels in a two-way floor system, as shown
in Fig. 16.3.1(a). Let L1 and L2 be the panel size in the longitudinal and transverse direc-
tions, respectively. Let lines 1-2 and 3-4 be centerlines of panels ABCD and CDEF, both
parallel to the longitudinal direction. Isolate as a free body [see Fig. 16.3.1(b)] the floor slab
and the included beam bounded by the lines 1-2 and 3-4 in the longitudinal direction and
the transverse lines 1′-3′ and 2′-4′ at the faces of the columns in the transverse direction.
The load acting on this free body [see Fig. 16.3.1(c)] is wu L2 per unit distance in the lon-
gitudinal direction. The total upward force acting on lines 1′-3′ or 2′-4′ is wu L2 Ln /2, where
wu is the factored load per unit area and Ln is the clear span in the longitudinal direction
between faces of supports (as defined by ACI-8.10.3.2.1).
If Mneg and Mpos are the numerical values of the total negative and positive bending
moments along lines 1′-3′ and 5-6, respectively, then equilibrium of the free body of
Fig. 16.3.1(d) requires
wu L2 L2n
M neg + M pos = (16.3.1)
8
For a typical exterior panel, the negative moment at the interior support would be larger
than that at the exterior support, as has been shown in Section 7.5. The maximum positive
62
moment would occur at a section to the left of the midspan, as shown in Fig. 16.3.2(c). In
design, it is customary to use Mpos at midspan for determining the required positive moment
reinforcement. For this case,
M neg (left) + M neg (right) wn L2 L2n

+ M pos = (16.3.2)
2 8
A proof for Eq. (16.3.2) can be obtained by writing the moment equilibrium equation about
the left end of the free body shown in Fig. 16.3.2(a),
wu L2 Ln  Ln   Ln 
M neg (left) + M pos =  4  − Vmidspan  2 
2
and by writing the moment equilibrium equation about the right end of the free body shown
in Fig. 16.3.2(b),
wu L2 Ln  Ln  L 
M neg (right) + M pos =   + Vmidspan  n 
2  4  2 
Equation (16.3.2) is arrived at by adding the two preceding equations and dividing by 2
on each side. Note that Eq. (16.3.2) may also be obtained, as shown in Fig. 16.3.2(c), by the
superposition of the simple span uniform loading parabolic positive moment diagram over
the trapezoidal negative moment diagram due to end moments.
L1
A B
wu L2 per unit distance
Mneg Mneg
1 2
L2
C D Clear span = Ln
wu L2 Ln wu L2 Ln
2 2
L2
3 4
(c)
E F
(a)
Mneg Mpos
Ln
1’ 5 2’ 4
1 2
0
Centerline
wu L2 Ln
wu L2 Ln 2
2
Factored (d)
floor load Mpos
= wu per unit area
3 4 M0
3’ 6 4’
Mneg Mneg
(b) (e)
Figure 16.3.1 Statics of a typical interior panel in a two-way floor system.

627
wu L2 Ln
2
Mneg(left) Mpos
Ln /2
Ri VCL wu L2 Ln
2
(a)
Mpos Mneg(right)
Ln /2
VCL RJ
(b)
Mmax Mpos
+
– –
Mneg(left)
Mneg(right)
(c)
Figure 16.3.2 Statics of typical exterior panel in a two-way floor system.
ACI-8.10.3 uses the symbol M0 to mean wu L2 L2u /8 and calls M0 the total factored static
moment. Further, it requires that the “absolute sum” of positive and average negative fac-
tored moments in each direction be not less than M0, or
M neg (left) + M neg (right)  w L L2 

+ M pos ≥  M 0 = u 2 n  (16.3.3)
2  8 
in which
wu = factored load per unit area
Ln = clear span in the direction moments are being determined, measured face-to-
face of supports (ACI-8.10.3.2.1), but not less than 0.65L1
L1 = span length in the direction that moments are being determined, measured
center-to-center of supports
L2 = span length in direction perpendicular to L1, measured center-to-center of supports
Equations (16.3.1) and (16.3.2) are theoretically derived on the basis that Mneg(left),
Mpos, and Mneg(right) occur simultaneously for the same live load pattern on the adjacent
panels of the equivalent rigid frame defined in Fig. 16.2.1. If the live load is relatively
heavy compared with dead load, then different live load patterns should be used to obtain
the critical positive moment at midspan and the critical negative moments at the supports.
In such a case, the “equal” sign in Eqs. (16.3.1) and (16.3.2) becomes the “greater than”
sign. This is why ACI-8.10.3 states that “absolute sum …” shall be at least M0 as the design
requirement. The designer should keep this in mind when steel reinforcement is selected for
positive and negative bending moment in two-way floors when the direct design method is
used for gravity load. To avoid the use of excessively small values of M0 in the case of short
spans and large columns or column capitals, the clear span Ln to be used in Eq. (16.3.3) is
not to be less than 0.65L1 (ACI-8.10.3.2.1).
When the limitations for using the direct design method are met, it is customary to
divide the value of M0 into Mneg and Mpos if the restraints at each end of the span are
628
identical (Fig. 16.3.1), or into [Mneg (left) + Mneg(right)] /2 and Mpos if the span end restraints
are different (Fig. 16.3.2). Then the moments Mneg(left), Mneg (right), and Mpos must be dis-
tributed transversely along the lines 1′-3′, 2′-4′, and 5-6, respectively. This last distribution
is a function of the relative flexural stiffness between the slab and the included beam.
Total Factored Static Moment in Flat Slabs

The ability of flat slab floor systems to carry load has been substantiated by numerous tests
of actual structures [16.2]. However, the amount of reinforcement used, say, in a typical
interior panel, was less than what it should be to satisfy an analysis by statics, as is demon-
strated in this section. This led to some controversy [16.1]. After studies by Westergaard and
Slater [16.3], however, a provision was adopted (about 1921) into the Code that allowed a
reduction of the moment coefficient from the statically required value of 0.125 to 0.09. This
reduction was not regarded as a violation of statics but was used as a way of permitting an
increase in the usable strength. The reduction, moreover, was applicable only to flat slabs
that satisfied the limitations then specified in the code. Over the years these limitations had
been liberalized, but at the same time the moment coefficient was raised to values closer
to 0.125. The present ACI Code logically stipulates the use of the full statically required
coefficient of 0.125.
The statical analysis of a typical interior panel was first made in 1914 by Nichols [16.1]
and further developed later by Westergaard and others [16.3–16.5]. Consider the typical
interior panel of a flat slab floor subjected to a factored load of wu per unit area, as shown
in Fig. 16.3.3(a). The total load on the panel area (rectangle minus four quadrantal areas)
is supported by the vertical shears at the four quadrantal arcs. Let Mneg and Mpos be the total
negative and positive moments about a horizontal axis in the L2 direction along the edges of
ABCD and EF, respectively. Then,
load on area ABCDEF = sum of reactions at arcs AB and CD

LL π c2 
= wu  1 2 −
 2 8 
Considering the half-panel ABCDEF as a free body, recognizing that there is no shear at the
edges BC, DE, EF, and FA, and taking moments about axis 1-1,
LL π c 2   c  wu L1 L2  L1  wu π c 2  2c 
M neg + M pos + wu  1 2 −  − +   = 0
 2 8   π  2  4  8  3π 
c /2
1 A F wu L2 per linear ft
B
2
(L1 – c)
3
L2
Mpos
C
M0
1
D E
Mneg
L1/2 L1/2
L1
Mneg 1
2 panel load
Mpos
c
π
exact, c3 approximate
1
2 panel load
(a) (b)
Figure 16.3.3 Statics of a typical interior panel in a flat slab floor system.

629
Letting M0 = Mneg + Mpos,
2
1  4c c3  1  2c 
M 0 = wu L2 L21  1 − + 2
≈ wu L2 L21  1 − (16.3.4)
8  π L1 3L2 L1  8  3L1 
Actually, Eq. (16.3.4) may be more easily visualized by inspecting the equivalent interior
span as shown in Fig. 16.3.3(b).
ACI-8.10.1.3 states that circular or regular polygon-shaped supports shall be treated as
square supports having the same area. For flat slabs, particularly with column capitals, the
clear span Ln, computed from using equivalent square supports, should be compared with
that indicated by Eq. (16.3.4), which is L1 minus 2c/3. In some cases the latter value is
larger and should be used, consistent with the fact that ACI-8.10.3.2 does express its intent
in an inequality.
Design Examples
In an effort to present, explain, and illustrate the design procedure for the three types of
two-way floor systems, identified in this chapter as two-way slabs (with beams), flat slabs,
and flat plates, it will be necessary to assume that preliminary dimensions and sizes of the
slab (and drop, if any), beams, and columns (and column capitals, if any) are available. In
the usual design processes, not only the preliminary sizes may need to be revised as they
are found unsuitable, but also designs based on two or three different relative beam sizes
(when used) to slab thickness should be made and compared. Preliminary data for the three
types of two-way floor systems to be illustrated are as follows.
Data for Two-Way Slab (with Beams) Design Example

Figure 16.3.4 shows a two-way slab floor with a total area of 12,500 sq ft. It is divided
into 25 panels with a panel size of 25 ft × 20 ft. Concrete strength is fc′ = 3000 psi and
steel yield strength is fy = 60,000 psi. Consider a superimposed service dead load of
24.5 psf and a service live load of 120 psf. Story height is 12 ft. The preliminary sizes are
as follows: slab thickness is 6 1 2 in., long beams are 14 × 28 in. overall; short beams are
12 × 24 in. overall; upper and lower columns are 15 × 15 in. The four kinds of panels
(corner, long-sided edge, short-sided edge, and interior) are numbered 1, 2, 3, and 4 in
Fig. 16.3.4.
1 2 f’c = 3 ksi
fy = 60 ksi
Service LL = 120 psf
3 4
Story height = 12 ft
5 @ 20’ = 100’
N
Assume:
W E 1
Slab thickness = 6 2 ”
S Long beams 14 × 28
3 4 overall
Short beams 12 × 24
overall
1 2 Columns 15” × 15”
5 @ 25’ = 125’
Figure 16.3.4 Plan view and data for two-way slab (with beams) design examples.
630
EXAMPLE 16.3.1
For the two-way slab (with beams) (see Fig. 16.3.4), determine the total factored static
moment in a loaded span in each of four equivalent frames whose widths are designated
A, B, C, and D in Fig. 16.3.5.
SOLUTION
The total service dead load per unit floor area is
Slab (6.5)(150) 1 = 81.25 psf

12
Superimposed = 24.50 psf
Total service dead load: = 105.75 psf

The factored load wu per unit floor area is then
wu = 1.2 wD + 1.6 wL = 1.2(105.75) + 1.6(120)

= 127 + 192 = 319 psf
5 @ 20’ = 100’
5 @ 20’ = 100’
B D C
5 @ 25’ = 125’ 5 @ 25’ = 125’
Figure 16.3.5 Equivalent frame notations in the two-way slab (with beams) of Example 16.3.5.
1 1
for frame A, M0 = wu L2 L2n = (0.319)(20)(25 − 1.25)2 = 450 ft-kips
8 8
for frame B, M 0 = 225 ft-kips
1 1
for frame C, M0 = wu L2 L2n = (0.319)(25)(20 − 1.25)2 = 350 ft-kips
8 8
for frame D, M 0 = 175 ft-kips
Data for Flat Slab Design Example

Figure 16.3.6 shows a flat slab floor with a total area of 12,500 sq ft. It is divided into 25
panels with a panel size of 25 × 20 ft. Concrete strength is fc′ = 3000 psi and steel yield
strength is fy = 60,000 psi. Service live load is 140 psf. Story height is 10 ft. Exterior col-
umns are 16 in. square and interior columns are 18 in. round. Edge beams are 14 × 24 in.
overall. Thickness of slab is 7 1 2 in. outside of drop panel and 10 1 2 in. through the drop
panel. Sizes of column capitals and drop panels are as shown in Fig. 16.3.6.
631
1
f’c = 3 ksi Slab thickness outside of drop = 7 2 ”
fy = 60 ksi 1
Slab thickness within drop = 10 2 ”
Service LL = 140 psf
Story height = 10 ft L = 25’
S = 20’
3
4
18” diam 5’ – 0”
column
(Revised to 7’ – 0”)
5 @ 20’ = 100’
4’ – 6”
6’ – 8”
16” sq column 8’ – 4”
1 2
14” × 24” edge beams
5 @ 25’ = 125’
Figure 16.3.6 Plan view and data for flat slab design examples.
EXAMPLE 16.3.2
Compute the total factored static moment in the long and short directions of an inte-
rior panel in the flat slab design example as shown in Fig. 16.3.6. Compare the results
obtained by using Eqs. (16.3.3) and (16.3.4).
SOLUTION
Neglecting the weight of the drop panel, the service dead load is 150 (7.5/12) =
94 psf; thus

wu = 1.2 wD + 1.6 wL = 1.2(94) + 1.6(140) = 113 + 224 = 337 psf
Using Eq. (16.3.4),

2 2
1  2c  1  2(5) 
M0 = wu L2 L21  1 − = (0.337)(20)(25)2 1 − = 396 ft-kips
8  3L1  8  3(25) 
(in long direction)
2 2

1 2  2c  1 2  2(5) 
M 0 = wu L2 L1  1 − = (0.337)(25)(20) 1 −  = 293 ft-kips
8  3L1  8  3(20) 
(in short direction)
(Continued)
632
The equivalent square area for the column capital (ACI-8.10.1.3) has its side equal to
4.43 ft; then, using Eq. (16.3.3), with Ln measured to the face of capital (i.e., equivalent
square),
1 1
M0 = wu L2 L2n = (0.337)(20)(25 − 4.43)2 = 356 ft-kips
8 8
(in long diirection )
1 1
M0 = wu L2 L2n = (0.337)(25)(20 − 4.43)2 = 255 ft-kips
8 8
(in short direction)
As to flat slabs with column capitals, it appears that the larger values of 396 and
293 ft-kips should be used because Eq. (16.3.4) is especially suitable; in particular,
ACI-8.10.3.2 states that the total factored static moment shall be at least that given by
Eq. (16.3.3).
Data for Flat Plate Design Example

Figure 16.3.7 shows a flat plate floor with a total area of 4500 sq ft. It is divided into
25 panels with a panel size of 15 × 12 ft. Concrete strength is fc′ = 4000 psi and steel yield
strength is fy = 60,000 psi. Service live load is 72 psf. Story height is 9 ft. All columns are
rectangular, 12 in. in the long direction and 10 in. in the short direction. Preliminary slab
thickness is set at 5 1 2 in. No edge beams are used along the exterior edges of the floor.
No edge beams Live load = 72 psf
3 4
5 @ 12’ = 60’
10” f’c = 4 ksi

fy = 60 ksi
Service LL = 72 psf
12”
Story height = 9 ft
1
Slab thickness = 5 2 ”
1 2 (preliminary)
5 @ 15’ = 75’
Figure 16.3.7 Plan view and data for flat plate design examples.
EXAMPLE 16.3.3
Compute the total factored static moment in the long and short directions of a typical
panel in the flat plate design example as shown in Fig. 16.3.7.
SOLUTION
The dead load for a 5 1 2 -in. slab is

wD = (5.5 /12)(150) = 69 psf
(Continued)
63
1 6 . 4 R AT I O O F F L E X U R A L S T I F F N E S S E S O F B E A M T O S L A B 633
The factored load per unit area is

wu = 1.2 wD + 1.6 wL = 1.2(69) + 1.6(72) = 83 + 115 = 198 psf
Using Eq. (16.3.3), with clear span Ln measured face-to face of columns,
1
M 0 = (0.198)(12)(15 − 1)2 = 58.2 ft-kips (in long direction)
8
1
M 0 = (0.198)(15)(12 − 0.83)2 = 46.3 ft-kips (in short directtion)
8
16.4 RATIO OF FLEXURAL STIFFNESSES

OF LONGITUDINAL BEAM TO SLAB
When beams are used along the column lines in a two-way floor system, an important param-
eter affecting the design is the relative size of the beam to the thickness of the slab. This
parameter can best be measured by the ratio αf of the flexural rigidity (called flexural stiffness
by the ACI Code) Ecb Ib of the beam to the flexural rigidity Ecs Is of the slab in the transverse
cross section of the equivalent frame shown in Fig. 16.4.1. The separate moduli of elasticity
Ecb and Ecs, referring to the beam and slab, provide for different concrete strength (and thus
different Ec values) for the beam and slab. The moments of inertia Ib and Is refer to the gross
sections of the beam and slab, respectively, within the cross section of Fig. 16.4.1(c). ACI-
8.4.1.8 permits the slab on each side of the beam web to act as a part of the beam. This slab
portion is limited to a distance equal to the projection of the beam above or below the slab,
whichever is greater, but not greater than 4 times the slab thickness, as shown in Fig. 16.4.2.
More accurately, the small portion of the slab already counted in the beam should not be used
in Is, but ACI permits the use of the total width of the equivalent frame in computing Is. Thus,
Ecb I b
αf = (16.4.1)
Ecs I s
The moment of inertia of a flanged beam section about its own centroidal axis
(Fig. 16.4.2) may be shown to be
bw h3
Ib = k (16.4.2a)
12
in which
 t  t 
2 3
b  t  t b
1 +  E − 1    4 − 6   + 4   +  E − 1   
 bw   h    h  h   bw   h  
k= (16.4.2b)
b  t
1 +  E − 1  
 bw   h 
where
h = overall beam depth
t = overall slab thickness
bE = effective width flange computed as shown in Fig. 16.4.2. (Note that th his
parameter is used here to compute the beam-to-slab stiffnness ratio α f and is
different from the effective flange width used for computing the flexural
strength of T-sections in Chapter 4.)
bw = width of web
634
L2
Equivalent frame
L2
L1 L1
(a) Plan
L2
t
A
h
A
L1 L1 (c) Cross section A–A
(b) Elevation
Figure 16.4.1 Plan, elevation, and beam and slab cross section of equivalent frame in a two-way
floor system.
bE bE
t t
(h–t) ≤ 4t h (h–t) ≤ 4t (h–t) ≤ 4t h

ACI–8.4.1.8 ACI–8.4.1.8 ACI–8.4.1.8
bw bw
Figure 16.4.2 Cross s ections for moment of inertia of a flanged beam section.
Equation (16.4.2b) expresses the nondimensional constant k in terms of (bE /bw) and (t /h).
Typical values of k are tabulated in Table 16.4.1 and three curves are plotted in Fig. 16.4.3.
The values of k are about 1.4, 1.6, and 1.8, respectively, for bE /bw values of 2, 3, and 4, when
t /h values are between 0.2 and 0.5. Thus
b  bE t
k ≈ 1.0 + 0.2  E  for 2 < <4 and 0.2 < < 0.5 (16.4.2c)
 bw  bw h
TABLE 16.4.1 VALUES OF k IN TERMS OF (bE /bW) AND (t/h) IN EQ. (16.4.2b)
t/h
bE /bw 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
2 1.222 1.328 1.366 1.372 1.375 1.396 1.454 1.565 1.743 2.000
3 1.407 1.564 1.605 1.608 1.625 1.694 1.844 2.098 2.477 3.000
4 1.564 1.744 1.777 1.781 1.825 1.956 2.212 2.621 3.209 4.000
635
1 6 . 4 R AT I O O F F L E X U R A L S T I F F N E S S E S O F B E A M T O S L A B 635
k=4
bE
t
CG h
axis k=3
bE
=4
bw bw
bw h3
Ig = k
12
k=2 k=2
bE
=3
bw
bE
=2
bw
k=1 k=1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
t/h
Figure 16.4.3 Values of k in terms of bE /bw and t/h.
EXAMPLE 16.4.1
For the two-way slab (with beams) design example described in Section 16.3 (see Fig
16.3.4), compute the ratio α f of the flexural stiffness of the longitudinal beam to that of the
slab in the equivalent frame, for all the beams around panels 1, 2, 3, and 4 in Fig. 16.4.4.
SOLUTION
(a) B1–B2. Referring to Fig. 16.4.4, the effective width bE for these interior beams is
the smaller of 14 + 2(21.5) = 57 in. and 14 + 8(6.5) = 66 in.; thus bE = 57 in. Using
Eq. (16.4.2b),
bE 57 t 6.5
= = 4.07, = = 0.232
bw 14 h 28

14(28)3
k = 1.774, I b = 1.774 = 45, 400 in.4
12
A slightly higher value of k would have been obtained by using Eq. (16.4.2c). Using
Eq. (16.4.1), where Ecb = Ecs,
1 Ecb I b 45, 400
Is = (240)(6.5)3 = 5490 in.4 , αf = = = 8.27
12 Ecs I s 5490
(b) B3–B4. Referring to Fig. 16.4.4, the effective width bE for these exterior beams is
the smaller of 14 + 21.5 = 35.5 in. and 14 + 4(6.5) = 40 in.; thus bE = 35.5 in. Using
Eq. (16.4.2b),
bE 35.5 t 6.5
= = 2.54, = = 0.232
bw 14 h 28

14(28)3
k = 1.484, I b = 1.484 = 38, 000 in.4
12
(Continued)
63
t = 6.5”
25’– 0” 25’– 0” bE = 57” bE = 35.5”
B1 B2
(h-t) = 21.5”
h = 28”
20’–0”
3 4
B8
B6
B6
B1 B2 bw = 14” bW = 14”
(h-t) ≤ 4t
(typical for
B1–B2 B3 – B4
flange
20’–0”
1 2
B7
B5
B5
projection)
B3 B4
t = 6.5”
Computed αf values
bE = 47” bE = 29.5”
8.27 8.27
h = 24”
17.5”
5.96
3.50
3.50
3 4
8.27 8.27 bw = 12” bw = 12”

B5–B6 B7–B8
5.96
3.50
3.50
1 2
13.83 13.83
Figure 16.4.4 Beam cross sections and computed αf values in Example 16.4.1.
Using Eq. (16.4.1),
1 Ecb I b 38, 000

Is = (120)(6.5)3 = 2745 in.4 , αf = = = 13.83
12 Ecs I s 2745
(c) B5–B6. Referring to Fig. 16.4.4, the effective width bE for B5–B6 is the smaller of
12 + 2(17.5) = 47 in. and 12 + 8(6.5) = 64 in.; thus bE = 47 in. Using Eq. (16.4.2b),
bE 47 t 6.5
= = 3.92, = = 0.271
bw 12 h 24

12(24)3
k = 1.762, I b = 1.762 = 24, 400 in.4
12
Using Eq. (16.4.1),
1 Ecb I b 24, 000

Is = (300)(6.5)3 = 6870 in.4 , αf = = = 3.50
12 Ecs I s 6870
(d) B7–B8. Referring to Fig. 16.4.4, the effective width bE for B7–B8 is the smaller of
12 + 17.5 = 29.5 in. and 12 + 4(6.5) = 38 in.; thus bE = 29.5 in. Using Eq. (16.4.2b),
bE 29.5 t
= = 2.46, = 0.271
bw 12 h

12(24)3
k = 1.480, I b = 1.480 = 20, 500 in.4
12
(Continued)
637
16.5 MINIMUM SLAB THICKNESS FOR DEFLECTION CONTROL 637
Using Eq. (16.4.1),
1 Ecb I b 20, 500

Is = (150)(6.5)3 = 3435 in.4 , αf = = = 5.96
12 Ecs I s 3435
For the resulting αf values for B1 through B8 around panels 1, 2, 3, and 4, see
Fig. 16.4.4. For this design, the αf values vary between 3.50 and 13.83; thus the rigidity
of the equivalent frames is provided primarily by the beams and the slab portion close to
the column lines, even though the widths of the frames vary from 10 to 25 ft.
16.5 MINIMUM SLAB THICKNESS FOR

DEFLECTION CONTROL
Control of deflections in two-way floor systems is dealt with in ACI-8.3.2. Deflections
are to be computed according to ACI-24.2 and must satisfy the limits of ACI Table 24.2.2
(see Chapter 12). To compute deflections, the use of the effective moment of inertia Ie, Eq.
(12.9.1), is endorsed unless computed deflections using other procedures are “in reasonable
agreement with the results of comprehensive tests” (ACI-R24.2.3.3). Various methods for
obtaining deflections of two-way floor systems have been proposed [16.52–16.71]; how-
ever, no specific procedure is given by the ACI Code or Commentary. Computation of two-
way floor system deflections is outside the scope of this book.
To aid the designer, ACI-8.3.1.1 provides a minimum thickness table [ACI Table 8.3.1.1]
for slabs without interior beams, though there can be exterior boundary beams. For slabs
with beams spanning between the supports on all sides, ACI-8.3.1.2 provides minimum
thickness equations. A thickness less than those indicated by ACI-8.3.1.1 or 8.3.1.2 may be
used if computations show that the deflection will not exceed the limits stipulated in ACI
Table 24.2.2.2. Computation of deflections must take into account “size and shape of the
panel, conditions of support, and nature of restraints at the panel edges” (ACI-24.2.3.3).
The minimum thicknesses given in ACI-8.3.1.1 or 8.3.1.2 are based on experience and are
considered satisfactory.
Slabs without Interior Beams Spanning between Supports

The minimum thickness, with the requirement that the ratio of long to short span be not
greater than 2, shall be that given by Table 16.5.1 [ACI Table 8.3.1.1], but not less than:
For slabs without drop panels 5 in.
For slabs with drop panels 4 in.
Note that in the flat slab and flat plate two-way systems, there may or may not be edge
beams but there are no interior beams.
Slabs Supported on Beams

Four parameters affect the minimum thickness requirements of ACI-8.3.1.2 for slabs sup-
ported on beams on all sides; they are (1) the longer clear span Ln of the slab panel; (2) the
ratio β of the longer clear span Ln to the shorter clear span Sn; (3) the yield strength fy of the
steel reinforcement; (4) the average αf m for the four αf values of relative stiffness for a panel
perimeter beam compared to the slab, as described in Section 16.4.
In terms of these parameters, ACI-8.3.1.2 requires the following for “slabs with beams
spanning between the supports on all sides.”
638
TABLE 16.5.1 MINIMUM THICKNESS OF SLAB WITHOUT INTERIOR

a
BEAMS [ADAPTED FROM ACI TABLE 8.3.1.1]
Without Drop Panelsc With Drop Panelsc
Exterior Panels Interior Exterior Panels Interior

Panels Panels
Without Edge With Edge Without Edge With Edge
fy Beams Beams Beams Beams
(ksi)b αf < 0.8 αf ≥ 0.8 αf < 0.8 αf ≥ 0.8
Ln Ln Ln Ln Ln Ln
40
33 36 36 36 40 40
Ln Ln Ln Ln Ln Ln
60 30 33 33 33 36 36
Ln Ln Ln Ln Ln Ln
75
28 31 31 31 34 34
a Ln is the clear span in the long direction.
b For fy between the values given in the table, min t is to be obtained by linear interpolation.
c Drop panels are defined in ACI-8.2.4.
Slabs Supported on Shallow Beams Where αfm ≤ 0.2

The minimum slab thickness requirements are the same as those for slabs without inte-
rior beams, i.e., those given in Table 16.5.1.
Slabs Supported on Medium Stiff Beams Where 0.2 < αfm ≤ 2.0

For this case,
Ln (0.8 + f y / 200, 000)

min t = (16.5.1)
36 + 5β (α fm − 0.2)
but not less than 5 in.
Slabs Supported on Very Stiff Beams Where αfm > 2.0

For this case,
Ln (0.8 + f y / 200, 000)

Min t = (16.5.2)
36 + 9β
but not less than 3.5 in.

The effect of Eqs. (16.5.1) and (16.5.2) may be observed from Fig. 16.5.1 for Grade 60
steel reinforcement, where the vertical axis is the ratio of the long direction clear span Ln
to the minimum thickness t. This approach is similar to the span-to-depth ratio limitations
used for beams in Sections 12.15 and 12.16. Figure 16.5.1 includes the full feasible range
of parameters: (1) the panel proportions Ln /Sn ranging from square to two-to-one rectangu-
lar, and (2) the average edge stiffness parameter αfm ranging from zero with no edge beams
to 2.5 or so for very stiff edge beams.
639
16.6 NOMINAL REQUIREMENTS FOR SLAB THICKNESS 639
0.2 0.5 1.0 1.5 2.0 2.5
50 Flexible
49
edge beams
Min t = 5 in.
(without drop panel)
Min t = 4 in.
45
0
(with drop
2.
=
panel)
n
/S
Ln
Ln
t 41
=
.0
β
40 =1
Sn
L n/
β=
35 Very stiff
edge beams
33
No edge beams
30
0 0.2 0.5 1.0 1.5 2.0 2.5
αfm
Figure 16.5.1 ACI minimum slab thickness formulas for Grade 60 steel.
Edge Beams at Discontinuous Edges

For all slabs supported on beams, there must be an edge beam at discontinuous edges hav-
ing a stiffness ratio αf not less than 0.80; otherwise, the minimum thickness required by
Eqs. (16.5.1) or (16.5.2) “shall be increased by at least 10 percent in the panel with the dis-
continuous edge” (ACI-8.3.1.2.1).
16.6 NOMINAL REQUIREMENTS FOR SLAB THICKNESS

AND SIZE OF EDGE BEAMS, COLUMN CAPITAL,
AND DROP PANEL
Whether the ACI Direct Design Method or the Equivalent Frame Method is used for deter-
mining the longitudinal distribution of bending moments, certain nominal requirements for
slab thickness and size of edge beams, column capital, and drop panel must be fulfilled.
These requirements are termed “nominal” because they are prescribed by the ACI Code. It
should be realized, of course, that those provisions are based on a combination of experi-
ence, judgment, tests, and theoretical analyses.
Slab Thickness
As discussed in Section 16.5, ACI-8.3.1.1 and ACI-8.3.1.2 [Eqs. (16.5.1) and (16.5.2) along
with Table 16.5.1] set minimum slab thickness for two-way floor systems. In addition,
the ACI Code sets lower limits for the minimum value based on experience and practical
requirements. These lower limits are summarized below:
Flat plates and flat slabs without drop panels 5 in.
Slabs on shallow interior beams having αfm < 0.2 5 in.
Slabs without interior beams but having drop panels 4 in.
Slabs with stiff interior beams having αfm ≥ 2.0 3.5 in.
640
Edge Beams
For slabs supported by interior beams, the minimum thickness requirements assume an
edge beam having a stiffness ratio αf not less than 0.80. If such an edge beam is not pro-
vided, the minimum thickness as required by Eqs. (16.5.1) and (16.5.2) must be increased
by 10% in the panel having the discontinuous edge. For slabs not having interior support
beams, the increased minimum thickness in the exterior panel having the discontinuous
edge is given by Table 16.5.1.
Column Capital
Used in flat slab construction, the column capital is an enlargement of the top of the col-
umn as it meets the floor slab or drop panel (Fig. 16.6.1). Since no beams are used, the
purpose of the capital is to enlarge the perimeter around the column to transmit shear from
the floor loading and to provide increasing thickness as the perimeter decreases near the
column. Assuming a maximum 45° line for distribution of the shear into the column, ACI-
8.4.1.4 requires that the effective column capital for strength considerations be within
the largest circular cone, right pyramid, or tapered wedge with a 90° vertex that can be
included within the outlines of the actual supporting element (see Fig. 16.6.1). The diam-
eter of the column capital is usually about 20 to 25% of the average span length between
columns.
Drop Panel
The drop panel is often used in flat slab and flat plate construction as a means of increasing
the shear strength around a column or reducing the negative moment reinforcement over a
column. It is an increased slab thickness in the region surrounding a column (Fig. 16.6.1).
A drop panel must comply with the dimensional limitations of ACI-8.2.4. The panel must
extend in each direction from the centerline of the support a minimum distance of one-
sixth of the span length measured from center-to-center in that direction. In addition, the
projection of the panel below the slab must be at least one-fourth of the slab thickness
outside of the drop (ACI-8.2.4). Note that when a qualifying drop is used, the minimum
thickness given by ACI-Table 8.3.1.1 has been reduced by 10% from the minimum when
a drop is not used.
For determining the slab flexural strength, ACI-8.5.2.2 stipulates that the thickness of
the drop below the slab be assumed no larger than one-quarter of the distance between the
edge of the drop panel and the edge of the column or column capital. Therefore, there is
little reason to use a drop panel of greater plan dimensions or thickness than that defined
by ACI-8.5.2.2.
Slab
CL of slab Drop panel

90°
Effective dimension
Actual column of column capital
capital
Center of column
Column
Figure 16.6.1 Effective dimension of column capital.

641
EXAMPLE 16.6.1
For the two-way slab (with beams) described in Section 16.3 (see Fig. 16.3.4), deter-
mine the minimum slab thickness required for deflection control and compare it with the
preliminary thickness of 6 1 2 in.
SOLUTION
The average ratios αfm for panels 1, 2, 3, and 4 may be computed from the αf values
shown in Fig. 16.4.4; thus
1
α fm for panel 1 = (5.96 + 8.27 + 3.50 + 13.83) = 7.89
4
1
α fm for panel 2 = (3.50 + 8.27 + 3.50 + 13.83) = 7.28
4
1
α fm for panel 3 = (5.96 + 8.27 + 3.50 + 8.27) = 6.50
4
1
α fm for panel 4 = (3.50 + 8.27 + 3.50 + 8.27) = 5.89
4
Since the αfm values for all four panels are well above 2, Eq. (16.5.2) applies. The minimum
thickness for all panels, using Ln = 23.75 ft, Sn = 18.75 ft, and fy = 60,000 psi, becomes
Ln (0.8 + 0.2 f y / 40, 000) 23.75(12)1.1
min t = = = 6.61 in.
36 + 9 β 36 + 9(23.75) /18.75
If a uniform slab thickness for the entire floor area is to be used, the minimum for
deflection control is 6.61 in. This value compares well, but it is slightly larger than the
6 1 2 -in. preliminary thickness. Given the approximations involved in the calculations
and the limits used for deflections, as well as the small difference between the required
and the provided thickness (0.1 in.), a slab thickness of 6 1 2 -in. will be considered
adequate. If deflections are of concern, they should be computed in accordance with
ACI-24.2 and must satisfy the limits of ACI Table 24.2.2.
► EXAMPLE 16.6.2
Review the slab thickness and other nominal requirements for the dimensions of the flat
slab described in Section 16.3 (see Fig. 16.3.6).
SOLUTION
(a) Stiffness of edge beams. Before using Table 16.5.1 or ACI Table 8.3.1.1, the αf val-
ues for the edge beams are needed. The moment of inertia of the edge beam section
shown in Fig. 16.6.2(b) is 22,900 in.4 Thus the αf value for the long edge beam is
Ib 22, 900 22, 900

αf = = = = 5.42
I s 120(7.5)3 /12 4220
and for the short edge beam, it is
Ib 22, 900 22, 900

αf = = = = 4.34
I s 150(7.5)3 /12 5270
These αf values are shown on Fig. 16.6.2(a).

(Continued)
642
(b) Minimum slab thickness using Table 16.5.1 or ACI Table 8.3.1.1. The long and short
clear spans for deflection control are

Ln = 25 − 4.43 = 20.57 ft; Sn = 20 − 4.43 = 15.57 ft
from which
Ln 20.57
β= = = 1.32
Sn 15.57
αf = 0 αf = 0
αf = 0
αf = 4.34
αf = 0
3 4
5’– 0”
5 @ 20’ = 100’
(Revised to 7’– 0”)

αf = 0 αf = 0
6’–8”
8’– 4”
αf = 4.34
1 2
αf = 0
αf = 0
αf = 5.42 αf = 5.42
5 @ 25’ = 125’
(a) Plan view
bE = smaller of 30.5” and 44”
1”
72
24”
14”
(b) Edge beam section
Figure 16.6.2 Plan view, edge beam section, and αf values for the flat slab of Example 16.6.2.
(Continued)
643
For fy = 60 ksi, a flat slab with drop panels, and αf = smaller of 4.34 and 5.42,
Table 16.5.1 gives
Ln 20.57(12)
min t = = = 6.86 in.
36 36
for both, the exterior and interior panels.
(c) Nominal requirement for slab thickness. The minimum thickness required is, from
part (b), 6.86 in. The 7 1 2 in. slab thickness used is more than ample; 7 in. probably
should have been used.
(d) Thickness of drop panel. Flexural strength within the drop panel must be computed
on the basis of the 10 1 2 -in. slab thickness within the drop (Fig. 16.3.6) or 7 1 2 in.
plus one-fourth of the projection of the drop beyond the column capital, whichever
is smaller. For the full 3-in. projection of the drop below the 7 1 2 -in. slab to be usable
in computing the flexural strength, the side of the drop measuring 6 ft 8 in. is revised
to 7 ft so that one-fourth of the distance between the edges of the 5-ft column capital
and the 7-ft drop is just equal to (10.5 –7.5) = 3 in.
EXAMPLE 16.6.3
Review the slab thickness and other nominal requirements for the dimensions of the flat
plate described in Section 16.3 (see Fig. 16.3.7).
SOLUTION
(a) Minimum slab thickness from ACI Table 8.3.1.1. For fy = 60 ksi, a flat plate that
inherently has αf = 0, and Ln = 15 –1 = 14 ft, Table 16.5.1 gives
Ln 14(12)
min t = = = 5.6 in. (exterior panel)
30 30
and
Ln 14(12)
min t = = = 5.1 in. (interior panel)
33 33
(b) The 5 1 2 -in. slab thickness used for all panels satisfies the nominal minimum of
5 in. for slabs without drop panels and without interior beams. It is also greater than
that required for an interior panel (5.1 in.), but slightly smaller than that required
for an exterior panel (5.6 in.). Again, given the approximations involved in the cal-
culation of deflections and the small difference between the required and provided
values (0.1 in.), a slab thickness of 5 1 2 in. will be considered adequate. If deflections
are of concern, they should be computed in accordance with ACI-24.2 and must sat-
isfy the limits of ACI Table 24.2.2.
64
16.7 DIRECT DESIGN METHOD—L IMITATIONS

Over the years, the use of two-way floor systems has been extended from one-story or
low-rise to medium-or high-rise buildings. For the common cases of one-story or low-rise
buildings, lateral load (wind or earthquake) is of lesser concern. In particular, when the
dimensions of the floor system are quite regular and when the live load is not excessively
large in comparison to the dead load, the use of a set of prescribed coefficients to distribute
longitudinally the total factored static moment M0 seems reasonable. As shown in Figs.
16.3.1 and 16.3.2, for each clear span in the equivalent frame, the equation
M neg (left) + M neg (right )  w L L2 

+ M pos ≥  M 0 = u 2 n  [16.3.3]
2  8 
is to be satisfied.
To use the Direct Design Method (DDM), in which a set of prescribed coefficients gives
the negative end moments and the positive moment within the span, ACI-8.10.2 imposes
the following limitations.
1. There is a minimum of three continuous spans in each direction.

2. Panels must be rectangular, with the ratio of longer to shorter span, measured center-
to-center of supports, within a panel not greater than 2.
3. The successive span lengths, measured center-to-center of supports, in each direction
do not differ by more than one-third of the longer span.
4. Columns are not offset by more than 10% of the span in the direction of the offset.
5. The load is due to gravity only and is uniformly distributed over an entire panel, and
the service live load does not exceed two times the service dead load.
6. For a panel with beams on all sides, the relative stiffness ratio of L22 /α f 2 to L21 /α f 1
must lie between 0.2 and 5.0, where αf is the ratio of the flexural stiffness of the
included beam to that of the slab.
Though the design of two-way floor systems is to a large extent empirical, the ACI limi
tations conform to the experimental results that are available [16.15–16.22] and to many
years of experience with slabs in actual structures. The Direct Design Method (DDM) can
also be used when it can be demonstrated that variations from any of the six limitations will
still produce a slab system that satisfies the conditions of equilibrium and geometric com-
patibility, and that all the strength and serviceability conditions are met, including speci-
fied limits on deflection. Van Buren [16.28] has provided such an analysis for staggered
columns in flat plates.
EXAMPLE 16.7.1
Show that for the two-way slab (with beams) described in Section 16.3 (see Fig. 16.3.4)
the six limitations of the Direct Design Method are satisfied.
SOLUTION
The first four limitations are satisfied by inspection. For the fifth limitation, from
Example 16.3.1
service dead load wD = 106 psf
service live load wL = 120 psf

wL 120
= <2 OK
wD 106
(Continued)
645
16.8 DIRECT DESIGN METHOD 645
For the sixth limitation, referring to Fig. 16.4.4 and taking L1 and L2 in the long and short
directions, respectively,
L21 625
Panel 1, = = 56.6
α f 1 0.5(13.83 + 8.27)

L22 400
= = 84.66
α f 2 0.5(5.96 + 3.50)
L21 625
Panel 2, = = 56.6
α f 1 0.5(13.83 + 8.27)

L22 400
= = 114.3
α f 2 3.50
L21 625
Panel 3, = = 75.6
α f 1 8.27

L22 400
= = 84.6
α f 2 0.5(5.96 + 3.50)
L21 625
Panel 4, = = 75.6
α f 1 8.27

L22 400
= = 114.3
α f 2 3.50
It may readily be verified that all ratios of L22 /α f 2 to L21 /α f 1 lie between 0.2 and 5.
16.8 DIRECT DESIGN METHOD—L ONGITUDINAL

DISTRIBUTION OF MOMENTS
In the Direct Design Method, moment diagrams in the direction of span length need not be
computed by an elastic analysis of the equivalent frame subjected to various pattern load-
ings; instead, they are nominally defined for common situations.
Figure 16.8.1 shows the longitudinal moment diagram for a typical interior span of
the equivalent frame in a two-way floor system as prescribed by ACI-8.10.4.1. Later (see
Section 16.12), the positive moment 0.35M0 or the negative moment 0.65M0 is to be dis-
tributed transversely to the slab having width L2 and to the included beam (if any) having
clear span Ln. Recall that
1
M0 = wn L2 L2n [16.3.3]
8
For a span that is completely fixed at both ends, the negative moment at the fixed end is
twice as large as the positive moment at midspan. For a typical interior span satisfying the
limitations of the DDM, the specified negative moment of 0.65M0 is a little less than twice the
specified positive moment of 0.35M0, which is fairly reasonable because the restraining effect
of the columns and adjacent panels is less than that of a beam with fixed ends.
For the exterior span, ACI-8.10.4.2 provides the longitudinal moment diagram for each
of the five categories as described in Fig. 16.8.2. On examination of these diagrams, one
64
Ln /2 Ln /2
0.35M0
+
i j
– at least –
M0
0.65M0 0.65M0
Figure 16.8.1 Longitudinal moment diagram for interior span.
sees that the negative moment at the exterior support increases from 0 to 0.65M0, the pos-
itive moment within the span decreases from 0.63M0 to 0.35M0, and the negative moment
at the interior support decreases from 0.75M0 to 0.65M0, all gradually as the restraint at the
exterior support increases from the case of a slab simply supported on a masonry or con-
crete wall (unrestrained) to that of a reinforced concrete wall built monolithically with the
slab (fully restrained). ACI Commentary R8.10.4.2 states that high positive moments are
purposely assigned into the span, since design for exterior negative moment will be gov-
erned by minimum reinforcement to control cracking in most slab systems.
Regarding the ACI Code–suggested moment diagrams of Figs. 16.8.1 and 16.8.2, ACI-
8.10.4.3 permits these moments to be modified by 10%, provided the total factored static
Ln /2 Ln /2
0.63 M0
+ At least M0 Case 1
Exterior edge
–
unrestrained
0.75 M0
Ln /2 Ln /2
0.57M0
At least M0
+ Case 2
Beams on all
0.16 M0 –
column lines
0.70 M0
Ln /2 Ln /2
0.52M0
At least M0
+ Case 3
– No beams on all
0.26 M0 column lines
0.70 M0
Ln /2 Ln /2
0.50M0
At least M0
+ Case 4
– Edge beams only
0.30 M0
0.70 M0
Ln /2 Ln /2
0.35M0 Case 5
At least M0
+ Exterior edge
– – fully restrained
(same as interior span)
0.65 M0 0.65 M0
Figure 16.8.2 Longitudinal moment diagram for exterior span.

647
moment M0 for the panel is statically accommodated. Note that moment redistribution as
described in Section 19.12 is not permitted when the approximate moments given in ACI-
8.10.4.3 (Figs. 16.8.1 and 16.8.2) are used.
16.9 DIRECT DESIGN METHOD—E FFECT OF PATTERN

LOADINGS ON POSITIVE MOMENT
To understand the effect of pattern loading on the longitudinal moment values in multiple-
panel, two-way floor systems, it is convenient to review some aspects of the continuity
analysis of the usual column-beam frames discussed in Chapter 7. Some of the findings,
which might be visualized from knowledge of influence lines and maximum moment enve-
lopes due to dead and live load combinations, are as follows:
1. The higher the ratio of column stiffness to beam stiffness, the smaller the effect of
pattern loadings. This is because the ends of the span are closer to the fixed condition
and less effect is exerted on the span by loading patterns on adjacent spans.
2. The higher the ratio of live load to dead load, the larger the effect of pattern loadings.
This is because dead load exists constantly on all spans and the pattern is related to
live load only.
3. Maximum negative moments at supports are less affected by pattern loadings than
maximum positive moments within the span.
Prior to the 1995 ACI Code, the adjustment of positive moment to account for pattern
loading had to be considered. Since 1995, the ACI Code restricts the uses of the DDM to
cases where the service live load does not exceed 2 (instead of 3 used previously) times
the service dead load. With this lower maximum ratio for live load to dead load, the ACI
Code committee concluded the number of cases where pattern loading would have a sig
nificant effect would be small; thus, adjustment for pattern loading no longer appears in the
ACI Code.
16.10 DIRECT DESIGN METHOD—P ROCEDURE FOR

COMPUTATION OF LONGITUDINAL MOMENTS
The background explanation for the distribution of the total static moment M0 in the longitu-
dinal direction, and the effects of pattern loading, have been discussed in the two preceding
sections. Using this information, the procedure for computing the longitudinal moments by
the direct design method may be summarized as follows:
1. Check limitations 1 through 5 for the DDM listed in Section 16.7. If they comply,
and the slab is supported on beams, follow Steps 2 through 6 given below. For slabs
not supported on beams, proceed to Step 6.
2. Compute the slab moment of inertia Is
 t3 
I s = ∑ L2  
 12 
3. Compute the longitudinal beam (if any) moment of inertia Ib (ACI-8.4.1.8).

4. Compute the ratio αf of the flexural stiffness of beam section to flexural stiffness of
a width of slab bounded laterally by centerlines of adjacent panels (if any) on each
side of the beam
Ecb I b
αf =
Ecs I s
5. Check that the ratio L22 /α f 2 to L21 /α f 1 lies between 0.2 and 5.0 for the cases where the
slab is supported by beams.
648
6. Compute the total static moment M 0 = wu L2 L2n /8 as stated by Eq. (16.3.3). Note that
Ln is not to be taken less than 0.65L1. For flat slabs, Eq. (16.3.4) should preferably be
used for computing M0.
7. Obtain the three critical ordinates on the longitudinal moment diagrams for the exte-
rior and interior spans using Figs. 16.8.1 and 16.8.2.
EXAMPLE 16.10.1
For the two-way slab (with beams) described in Section 16.3 (see Fig. 16.3.4), deter-
mine the longitudinal moments in frames A, B, C, and D, as shown in Fig. 16.3.5.
SOLUTION
(a) Check the six limitations for the direct design method. (These limitations were
checked in Example 16.7.1.)
(b) Total factored static moment M0. The total factored static moments M0 for the equiv-
alent rigid frames A, B, C, and D (computed in Example 16.3.1)
M0 (frame A) = 450 ft-kips

M0 (frame B) = 225 ft-kips
M0 (frame C) = 350 ft-kips
M0 (frame D) = 175 ft-kips
(c) Longitudinal moments in the frames. The longitudinal moments in frames A, B, C,

and D are computed using Case 2 of Fig. 16.8.2 for the exterior span and Fig. 16.8.1
for the interior span. The computations are as shown below, and the results are sum-
marized in Fig. 16.10.1.
For Frame A, M0 = 450 ft-kips
Mneg at exterior support = 0.16(450) = 72 ft-kips
Mpos in exterior span = 0.57(450) = 257 ft-kips
Mneg at first interior support = 0.70(450) = 315 ft-kips
Mneg at typical interior support = 0.65(450) = 293 ft-kips
Mpos in typical interior span = 0.35(450) = 158 ft-kips
For Frame B, M0 = 225 ft-kips
For Frame C, M0 = 350 ft-kips
For Frame D, M0 = 175 ft-kips
(Continued)
649
Equivalent frame D
Equivalent frame C
5 @ 20’ = 100’
Equivalent frame A
Equivalent frame B
5 @25’ = 125’
–72 –315 –293 –293 –56 –245 –228 –228

+257 +158 +200 +123
Moments in A Moments in C
–36 –158 –146 –146 –28 –123 –114 –114

+128 +79 +100 +61
Moments in B Moments in D
Figure 16.10.1 Longitudinal moments in ft-kips for the two-way slab (with beams) of Example 16.10.1.
EXAMPLE 16.10.2
For the flat slab design example described in Section 16.3 (see Fig. 16.3.6), compute the
longitudinal moments in frames A, B, C, and D as shown in Fig. 16.10.2.
SOLUTION
(a) Check the five limitations (the sixth limitation does not apply here) for the direct
design method. These five limitations are all satisfied.
5 @20’ = 100’
Equivalent frame D
Equivalent frame C
Equivalent frame A
Equivalent frame B
5 @25’ = 125’
–119 –277 –257 –257 –88 –205 –190 –190

+198 +139 +147 +103
–59 –139 –129 –129 –44 –103 –96 –96

+99 +69 +74 +51
Figure 16.10.2 Longitudinal moments in ft-kips for the flat slab of Example 16.10.2.
(Continued)
650
(b) Total factored static moment M0. Referring to the equivalent frames A, B, C, and D
in Fig. 16.10.2, the total factored static moment may be taken from the results of
Example 16.3.2; thus
M 0 for A = 396 ft-kips
1
M 0 for B = (396) = 198 ft-kips
2
M 0 for C = 293 ft-kips

1
M 0 for D = (293) = 147 ft-kips
2
(c) Longitudinal moments in the frames. The longitudinal moments in Frames A, B, C,
for the interior span. The computations are shown in Table 16.10.1 and the results
are summarized in Fig. 16.10.2.
TABLE 16.10.1 LONGITUDINAL MOMENTS (FT-KIPS) FOR THE FLAT

SLAB DESIGN EXAMPLE
Frame A B C D
M0 396 198 293 147

Mneg at exterior support, 0.30M0 119 59 88 44
Mpos in exterior span, 0.50M0 198 99 147 74
Mneg at first interior support, 0.70M0 277 139 205 103
Mneg at typical interior support, 0.65M0 257 129 190 96
Mpos in typical interior span, 0.35M0 139 69 103 51
EXAMPLE 16.10.3
For the flat plate described in Section 16.3 (see Fig. 16.3.7) compute the longitudinal
moments in frames A, B, C, and D as shown in Fig 16.10.3.
SOLUTION
(a) Check the five limitations (the sixth limitation does not apply here) for the direct
design method. These limitations are all satisfied.
(b) Total factored static moment M0 from the results of Example 16.3.3.
M 0 for A = 58.2 ft-kips

1
M 0 for B = (58.2) = 29.1 kips
2
M 0 for C = 46.3 ft-kips
M 0 for D = 23.2 ft-kips
(Continued)
651
16.11 TORSION STIFFNESS OF THE TRANSVERSE ELEMENTS 651
(c) Longitudinal moments in the frames. The longitudinal moments in frames A, B, C,

for the interior span. The computations are as shown in Table 16.10.2 and the results
are summarized in Fig. 16.10.3.
TABLE 16.10.2 LONGITUDINAL MOMENT (FT-KIPS) FOR THE FLAT

PLATE DESIGN EXAMPLE
Frame A B C D
M0 58.2 29.1 46.3 23.2

Mneg at exterior support, 0.26M0 15.1 7.6 12.0 6.0
Mpos in exterior span, 0.52M0 30.3 15.1 24.1 12.1
Mneg at first interior support, 0.70M0 40.7 20.4 32.4 16.2
Mneg at typical interior support, 0.65M0 37.8 18.9 30.1 15.1
Mpos in typical interior span, 0.35M0 20.4 10.2 16.2 8.1
10”
5 @ 12’ = 60’
Equivalent frame D
Equivalent frame C
Equivalent frame A
12”
Equivalent frame B
5 @ 15’ = 75’
Exterior edge of floor
Exterior edge of floor No edge beams used
Typical column 10 × 12”
–15.1 –40.7 –37.8 –37.8 –12.0 –32.4 –30.1 –30.1

+30.3 +20.4 +24.1 +16.2
–7.6 –20.4 –18.9 –18.9 –6.0 –16.2 –15.1 –15.1

+15.1 +10.2 +12.1 +8.1
Figure 16.10.3 Longitudinal moments in ft-kips for the flat plate of Example 16.10.3.
16.11 TORSION STIFFNESS OF THE TRANSVERSE

ELEMENTS
Up to this point, the stiffness of the equivalent frame has been considered with regard to
the members in the plane of the frame only. The transverse members, however, will also
contribute to the stiffness of the frame by resisting the in-plane bending through torsion. In
the ACI Code, this contribution is considered by the torsional constant C of the transverse
652
c1 = width of column, bracket, or capital (bw+ larger of h1 or h2

but no larger than bw + 4t)
c1
c1
h1
t or t or t
h2
For torsional member For torsional bw For flexural and
(no actual beam) member torsional member
(ACI–8.11.5.1a) (ACI–8.11.5.1b) (ACI–8.11.5.1c)
(a) (b) (c)
Figure 16.11.1 Definition of cross sections for transverse beams in torsion. [Projection of slab
beyond beam in Case (c) is allowed on each side for interior beam.]
beams spanning from column to column. Even if no such beam (as defined by projection
above or below the slab) actually exists, one still should imagine that there is a “beam”
made of a portion of the slab having a width equal to that of the column, bracket, or capi-
tal in the direction of the span for which moments are being determined (ACI-8.11.5.1a).
When there is actually a transverse beam web above or below the slab, the cross section of
the transverse beam should include the portion of slab within the width of column, bracket,
or capital described above plus the projection of beam web above or below the slab (ACI-
8.11.5.1b). As a third possibility, the transverse beam may include that portion of slab on
each side of the beam web equal to its projection above or below the slab, whichever is
greater, but not greater than 4 times the slab thickness (ACI-8.11.5.1c). The largest of the
three definitions as shown in Fig. 16.11.1 may be used.
The torsional constant C of the transverse beam equals
 x   x3 y 
C = ∑  1 − 0.63   (16.11.1)
 y   3 
which is given by ACI Formula (8.10.5.2b), where
x = shorter dimension of a component rectangle

y = longer dimenssion of a component rectangle
and the component rectangle should be taken in such a way that the largest value of C is
obtained. Equation (16.11.1) is identical to Eq. (18.3.5), for which there is additional dis-
cussion in Chapter 18.
EXAMPLE 16.11.1
Compute the torsional constant C for the edge and interior beams in the short and long
directions for the two-way slab (with beams) (see Fig.16.3.4).
SOLUTION
Each cross section shown in Fig. 16.11.2 may be divided into component rectangles in
two different ways, and the larger value of C is to be used.
Edge beam in short direction
3 3
 edge   0.63(6.5)  29.5(6.5)  0.63(12)  17.5(12)
C = 1 − + 1 −
 beam   29.5  3 
 17.5  3

= 2325 + 5725 = 8050 in.4
(Continued)
653
29.5” 29.5” 47” 47”

6.5”
17.5” 17.5” 17.5” 17.5”
12” 12” 12” 12”
(a) Short direction
35.5” 35.5” 57” 57”

6.5” 6.5”
21.5” 21.5” 21.5” 21.5”
14” 14” 14” 14”
(b) Long direction
Figure 16.11.2 Effective cross sections of transverse beams resisting torsion in the two-way slab
(with beams) of Example 16.11.1.
or
3 3
 edge   0.63(6.5)  17.5(6.5)  0.63(12)  24(12)
C = 1− + 1 −
 beam   17.5  3  24  3

= 1230 + 9470 = 10, 700 in.4
Use 10,700 in.4
Interior beam in short direction
3 3
 interior   0.63(6.5)  47(6.5)  0.63(12)  17.5(12)
C = 1− + 1 −
 beam   47  3  17.5  3

= 3925 + 5725 = 9650 in.4
or
 interior 
C = 2(1230) + 9470 = 11, 930 in.4
 beam 
Use 11,930 in.4
Edge beam in long direction
3 3
 edge   0.63(6.5)  35.5(6.5)  0.63(14)  21.5(14)
C = 1− + 1 −
 beam   35.5  3  21.5  3

= 2900 + 11, 600 = 14, 500 in. 4
or
3 3
 edge   0.63(6.5)  21.5(6.5)  0.63(14)  28(14)
C = 1− + 1 −
 beam   21.5  3  28  3
= 1600 + 17, 500 = 19,100 in.4
Use 19,100 in.4
(Continued)
654
Interior beam in long direction
3 3
 interior   0.63(6.5)  57(6.5)  0.63(14)  21.5(14)
C = 1− + 1 −
 beam   57  3  21.5  3

= 4800 + 11, 600 = 16, 400 in.4
or
 interior 
C = 2(1600) + 17, 500 = 20, 700 in.4
 beam 
Use 20,700 in.4
EXAMPLE 16.11.2
Compute the torsional constant C for the edge beam and the interior beam in the short
and long directions for the flat slab design example of Fig. 16.3.6.
SOLUTION
For the short or long edge beam [Fig. 16.11.3(a)], the torsional constant C is computed
on the basis of the cross section shown in Fig. 16.11.3(a).
3 3
 0.63(7.5)  (7.5) (51.5)  0.63(14)  (114) (16.5)
C = 1 − + 1 −
 51.5  3 
 16.5  3
= 6575 + 7025 = 13, 600 in.4
or
3 3
 0.63(14)  (14) (24)  0.63(7.5)  (7.5) (37.5)
C = 1 − + 1 −
 24  3 
 37.5  3
= 13, 890 + 4610 = 18, 500 in.4
Use 18,500 in.4
For the short or long interior beam [Fig. 16.11.3(b)], a weighted slab thickness of
8.5 in. is used, on the assumption that one-third of the span has a 10 1 2 -in. thickness
and the remainder has a 7 1 2 -in. thickness. (Actually, the ratio is not exactly so because
the drop width has been revised from 6 ft 8 in. to 7 ft.)
 x  x 3 y  0.63(8.5)   (8.5)3 (53.2) 

C =  1 − 0.63  = 1 −   = 9800 in.
4
 y 3  53.2   3 
(Continued)
65
35 + 16.5 = 51.5” 51.5”
7.5”
16.5” 16.5”
14” 14”
35” = distance from outer edge of exterior column to inner

edge of square capital (i.e., 2’–3” + half the 16” column)
(a) Short or long edge beam
8.5” = weighted slab thickness
12 (4.43’) = 53.2”
(b) Short or long interior beam
Figure 16.11.3 Cross s ections of torsional transverse beams in the flat slab of Example 16.11.2.
EXAMPLE 16.11.3
For the flat plate design example (see Fig. 16.3.7), compute the torsional constant C for
the short and long beams.
SOLUTION
Since no actual edge beams are used, the torsional member is, according to Fig. 16.11.4,
equal to slab thickness t by the column width c1.
3
 0.63(5.5)  (5.5) (12)
C for short beams = 1 −  = 474 in.4
 12  3
3

 0.63(5.5)  (5.5) (10)
C for long beams = 1 − = 362 in. 4
 10  3
5.5”
10”
Imaginary beam Long beam
5.5”
12”
10”
12”
Short beam
Figure 16.11.4 Cross s ections of torsional transverse beams for the flat plate of Example 16.11.3.
65
16.12 TRANSVERSE DISTRIBUTION
OF LONGITUDINAL MOMENT
The longitudinal moment values, whether those of the Direct Design Method shown
in Figs. 16.8.1 and 16.8.2 or those obtained by structural analysis using the Equivalent
Frame Method (see Section 16.20), are for the entire width (sum of the two half-panel
widths in the transverse direction, for an interior column line) of the equivalent frame.
Each of these moments is to be divided, on the basis of studies by Gamble, Sozen, and
Siess [16.12], between the column strip and the two half middle strips as defined in
Fig. 16.12.1. If the two adjacent transverse spans are each equal to L2, the width of
the column strip is then equal to one-half of L2, or one-half of the longitudinal span
L1, whichever is smaller (ACI-8.4.1.5). This is reasonable, since when the longitu-
dinal span is shorter than the transverse span, a larger portion of the moment across
the width of the equivalent frame might be expected to concentrate near the column
centerline.
The transverse distribution of the longitudinal moment to column and middle strips is
a function of three parameters, using L1 and L2 for the longitudinal and transverse spans,
respectively: the aspect ratio L2 /L1; the ratio αf 1 = Ecb Ib /(Ecs Is) of the longitudinal beam
stiffness to slab stiffness; and the ratio βt = EcbC/(2Ecs Is) of the torsional rigidity of the edge
beam section to the flexural rigidity of a width of slab equal to the span length of the edge
beam. According to ACI-8.10.5, the column strip is to take the percentage of the longitudinal
moment as shown in Table 16.12.1. As may be seen from Table 16.12.1, only the first two
parameters affect the transverse distribution of the negative moments at the first and typical
interior supports as well as the positive moments in exterior and interior spans, but all three
parameters are involved in the transverse distribution of the negative moment at the exterior
support.
Regarding the distributing percentages shown in Table 16.12.1, the following observa-
tions may be made:
1. In general, the column strip takes more than 50% of the longitudinal moment.
2. The column strip takes a larger share of the negative longitudinal moment than the
positive longitudinal moment.
3. When no longitudinal beams are present (αf 1 = 0), the column strip takes the same
share of the longitudinal moment, irrespective of the aspect ratio. The reader may
note, however, that the column strip width is a fraction of L1 or L2 (0.25L1 or 0.25L2
on each side of the column line), whichever is smaller.
Centerline of panel
Width of equivalent
Half middle strip frame
whichever is smaller
Exterior wall or beam, if any
0.25L1 or 0.25L2
Column strip
Half middle strip
Centerline of panel
Figure 16.12.1 Definition of column and middle strips.

657
16.12 TRANSVERSE DISTRIBUTION OF LONGITUDINAL MOMENT 657
TABLE 16.12.1 PERCENTAGE OF LONGITUDINAL MOMENT IN COLUMN

a
STRIP (ACI-8.10.5)
Aspect Ratio L2 /L1

0.5 1.0 2.0
Negative moment at αf1L2 /L1 = 0 βt = 0 100 100 100

exterior support βt ≥ 2.5 75 75 75
αf1L2 /L1 ≥ 1.0 βt = 0 100 100 100
βt ≥ 2.5 90 75 45
Positive moment αf1L2/L1 = 0 60 60 60
αf1L2/L1 ≥ 1.0 90 75 45
Negative moment at αf1L2/L1 = 0 75 75 75
interior support αf1L2/L1 ≥ 1.0 90 75 45
a Linear interpolation between values shown is permitted.
4. In the presence of longitudinal beams (αf 1 > 0), the larger the aspect ratio, the smaller
the distribution to the column strip. This seems consistent because the same reduc-
tion in the portion of moment going into the slab is achieved by restricting the col-
umn strip width to a fraction of L1 when L2/L1 is greater than one.
5. The column strip takes a smaller share of the exterior moment as the torsional rigid-
ity of the edge beam section increases.
When the exterior support consists of a column or wall extending for a distance equal to
or greater than three-fourths of the transverse width, L2, the exterior negative moment is to
be uniformly distributed over the transverse width (ACI-8.10.5.4).
The procedure for distributing the longitudinal moment across a transverse width to the
column and middle strips may be summarized as follows.
1. Divide the total transverse width applicable to the longitudinal moment into a
column strip width and two half middle strip widths, one adjacent to each side of
the column strip. For an exterior column line, the column strip width is 0.25 L1, or
0.25 L2, whichever is smaller; for an interior column line, the column strip width is
∑(0.25 L1 or 0.25 L2 ), whichever is smaller, of the panels on both sides.
2. Determine the ratio βt = EcbC/(2Ecs Is) of edge beam torsional rigidity to slab flexural
rigidity. (Note: The coefficient 2 in the denominator arises from approximating the
shear modulus of elasticity in the numerator as Ecb /2.)
3. Determine the ratio αf 1 = Ecb Ib /(Ecs Is) of longitudinal beam flexural stiffness to slab
flexural stiffness.
4. Divide the longitudinal moment at each critical section into two parts according to
the percentage shown in Table 16.12.1: one part to the column strip width; and the
remainder to the half middle strip for an exterior column line, or to the half middle
strips on each side of an interior column line.
5. If there is an exterior wall instead of an exterior column line, the strip ordinarily
called the exterior column strip will not deflect and therefore no moments act. In this
case there can be no longitudinal distribution of moments; thus there is no computed
moment to distribute laterally to the half middle strip adjacent to the wall. This half
middle strip should be combined with the next adjacent half middle strip, which itself
receives a lateral distribution in the frame of the first interior column line. The total
middle strip in this situation is designed for twice the moment in the half middle strip
from the first interior column line (ACI-8.10.6.3).
658
Distribution of Moment in Column Strip to Beam and Slab

When a longitudinal beam exists in the column strip along the column centerline, the col-
umn strip moment as determined by the percentages in Table 16.12.1 (ACI-8.10.5) should
be divided between the beam and the slab. ACI-8.10.5.7.1 states that 85% of the column
strip moment be taken by the beam if αf 1L2/L1 is equal to or greater than 1.0; for values
of αf 1L2/L1 between 0 and 1.0, the proportion of moment to be resisted by the beam is to
be obtained by linear interpolation between 0 and 85%. In addition, any beam must be
designed to carry its own weight (projection above and below the slab), and any concen-
trated or linear load applied directly on it (ACI-8.10.5.7.2).
EXAMPLE 16.12.1
For the two-way slab (with beams) of Example 16.10.1, distribute the longitudinal
moments computed for Frames A, B, C, and D (see Fig. 16.10.1) into three parts—namely,
for the longitudinal beams, for the column strip slab, and for the middle strip slab.
SOLUTION
The values for the total longitudinal moments in frames A, B, C, and D at the five critical
sections are taken from Example 16.10.1 (see Fig. 16.10.1) and shown again at the end
of this example (see Table 16.12.4).
(a) Negative moment at face of exterior support.
Interior frame A, L2 /L1 = 0.80; αf 1 = 8.27 (Fig. 16.4.4); αf 1L2 /L1 = 6.61; C = 10,700
in.4 (Example 16.11.1); Is = 240(6.5)3/12 = 5490 in.4; and βt = C/(2Is) = 10,700/[2(5490)]
= 0.98. Table 16.12.2 shows the linear interpolation for obtaining the column strip per-
centage from the prescribed limits of Table 16.12.1. The total moment of 72 ft-kips is
divided into three parts, 92.6% to column strip (of which 85% is carried by the beam
and 15% by the slab, since αf 1L2/L1 = 6.61 ≥ 1.0) and 7.4% to the middle strip slab. The
results are shown in Table 16.12.4.
Exterior frame B, L2/L1 = 0.80; αf 1 = 13.83 (Fig. 16.4.4); αf 1L2/L1 = 11.1; βt = 0.98,
the same as for frame A, and column strip moment percentage = 92.6%, the same as for
frame A.
Interior frame C, L2/L1 = 1.25; αf 1 = 3.50 (Fig. 16.4.4); αf 1L2/L1 = 4.38; C = 19,100 in.4
(Example 16.11.1); Is = 300(6.5)3/12 = 6870 in.4; and βt = C/(2Is) = 19,100/
[2(6870)] = 1.39. Table 16.12.3 shows the linear interpolation for obtaining the column
strip percentage from the prescribed limits of Table 16.12.1. The total moment of 56 ft-kips
is divided into three parts, 81.9% to column strip (of which 85% is carried by the beam
and 15% by the slab since αf 1L2/L1 = 4.38 ≥ 1.0) and 18.1% to the middle strip slab.
Exterior frame D, L2/L1 = 1.25; αf 1 = 5.96 (Fig. 16.4.4), αf 1L2/L1 = 7.45; βt = 1.39, the
same as for frame C; and column strip moment percentage = 81.9%, the same as for
frame C.
TABLE 16.12.2 LINEAR INTERPOLATION FOR COLUMN STRIP

PERCENTAGE OF EXTERIOR NEGATIVE MOMENT—FRAME A
L2/L1 0.5 0.8 1.0
βt = 0 100% 100% 100%

αf1L2 /L1 = 6.61 βt = 0.98 96.1% 92.6% 90.2%
βt ≥ 2.50 90% 81% 75%
(Continued)
659
TABLE 16.12.3 LINEAR INTERPOLATION FOR COLUMN STRIP

PERCENTAGE OF EXTERIOR NEGATIVE MOMENT—FRAME C
L2/L1 1.0 1.25 2.0
βt = 0 100% 100% 100%

αf 1L2/L1 = 4.38 βt = 1.39 86.1% 81.9% 69.4%
βt ≥ 2.50 75% 67.5% 45%
(b) Negative moments at exterior face of first interior support and at face of typical inte-
rior support.
Interior frame A, L2/L1 = 0.80 and αf 1L2/L1 = 6.61 > 1.0. Using the prescribed values
in Table 16.12.1, the proportion of moment going to the column strip is determined to
be 81% by linear interpolation.
L2/L1 0.5 0.8 1.0

αf 1L2/L1 = 6.61 90% 81% 75%
Exterior frame B, L2 /L1 = 0.80 and αf1L2 /L1 = 11.1. The proportion of moment is
again 81% for the column strip, the same as for Frame A.
Interior frame C, L2 /L1 = 1.25 and αf1L2 /L1 = 4.38. Using the prescribed values in
Table 16.12.1, the proportion of moment going to the column strip is determined to be
67.5% by linear interpolation:
L2/L1 1.0 1.25 2.0

αf 1L2/L1 = 4.38 75% 67.5% 45%
Exterior frame D, L2 /L1 = 1.25 and αf1L2 /L1 = 7.45. The proportion of moment is
again 67.5% for the column strip, the same as for Frame C.
(c) Positive moments in exterior and interior spans. Since the percentages for αf 1L2/L1
≥ 1.0 are the same for positive moment and for negative moment at interior support,
the percentages of column strip moment for positive moments in exterior and inte-
rior spans are identical to those for interior negative moments (see Table 16.12.1) as
determined in part (b) of this example.
(d) The computed moments in the beams, the column strip slabs, and the middle strip
slabs for frames A, B, C and D are summarized in Table 16.12.4.
60
TABLE 16.12.4 TRANSVERSE DISTRIBUTION OF LONGITUDINAL

MOMENTS (FT-KIPS) IN TWO-WAY SLAB (WITH BEAMS) DESIGN EXAMPLE
(ALSO SEE FIG. 16.10.1)
Interior frame A
Total Width = 20 ft, Column Strip Width = 10 ft, Middle Strip Width = 10 ft
Exterior Span Interior Span
Exterior Positive Interior Negative Positive

Negative Negative
Total moment –72 +257 –315 –291 +158
Moment in beam –57 +177 –217 –200 +109
Moment in column strip slab –10 +31 –38 –36 +19
Moment in middle strip slab –5 +49 –60 –55 +30
Exterior frame B
Total Width = 10 ft, Column Strip Width = 5 ft, Half Middle Strip Width = 5 ft

Negative Negative
Total moment –36 +128 –158 –146 +79
Moment in beam –28 +88 –109 –101 +54
Interior frame C
Total Width = 25 ft, Column Strip Width = 10 ft, Middle Strip Width = 15 ft

Negative Negative
Total moment –56 +200 –245 –228 +123
Moment in beam –39 +115 –140 –131 +71
Exterior frame D
Total Width = 12.5 ft, Column Strip Width = 5 ft, Half Middle Strip Width = 7.5 ft

Negative Negative
Total moment –28 +100 –123 –114 +61
Moment in beam –20 +57 –71 –65 +35
61
EXAMPLE 16.12.2
Divide the five critical moments in each of the equivalent frames A, B, C, and D in
the flat slab design example (see Example 16.10.2), as shown in Fig. 16.10.2, into two
parts: one for the half column strip (for frames B and D) or the full column strip (for
frames A and C), and the other for the half middle strip (for frames B and D) or the two
half middle strips on each side of the column line (for frames A and C).
SOLUTION
The percentages of the longitudinal moments carried by the column strip width are
shown in lines 10 to 12 of Table 16.12.5. Note that the column strip width shown in line
2 is one-half of the shorter panel dimension for frames A and C, and one-fourth of this
value for frames B and D. Note also that the sum of the values on lines 2 and 3 should
be equal to that on line 1, for each respective frame.
The moment of inertia of the slab equal in width to the transverse span of the edge
beam is
240(7.5)3
I s in βt for frames A and B = = 8440 in.4
12
and
300(7.5)3
I s in βt for frames C and D = = 10, 600 in.4
12
These values are shown in line 5 of Table 16.12.5.
The percentages shown in lines 10 to 12 are obtained from Table 16.12.1, by interpo-
lation (as illustrated in Tables 16.12.2 and 16.12.3) if necessary. Having these percent-
ages, the separation of each of the longitudinal moment values shown in Fig. 16.10.2
into two parts is a simple matter and thus is not shown further.

MOMENT FOR FLAT SLAB DESIGN EXAMPLE
Line Equivalent Frame A B C D

Number
1 Total transverse width (in.) 240 120 300 150

2 Column strip width (in.) (see 120 60 120 60
Fig. 16.12.1)
3 Half middle strip width (in.) 2 @ 60 60 2 @ 90 90
4 C (in.4) from Example 16.11.2 18,500 18,500 18,500 18,500
5 Is (in.4) in βt 8440 8440 10,600 10,600
6 βt = EcbC/(2Ecs Is) 1.10 1.10 0.87 0.87
7 αf 1 from Example 16.6.2 0 5.42 0 4.34
8 L2/L1 0.80 0.80 1.25 1.25
9 αf 1L2/L1 0 4.34 0 5.43
10 Exterior negative moment, percent
to column strip (Table 16.12.1) 89.0% 91.6% 91.3% 88.7%
11 Positive moment, percent to
column strip (Table 16.12.1) 60.0% 81.0% 60.0% 67.5%
12 Interior negative moment, percent
to column strip (Table 16.12.1) 75.0% 81.0% 75.0% 67.5%
62
EXAMPLE 16.12.3
Divide the five critical moments in each of the equivalent frames A, B, C, and D in the
flat plate of Example 16.10.3, as shown in Fig. 16.10.3, into two parts: one for the half
column strip (for frames B and D) or the full column strip (for frames A and C), and the
other for the half middle strip (for frames B and D) or the two half middle strips on each
side of the column line (for frames A and C).
SOLUTION
The percentages of the longitudinal moments carried by the column strip width are
shown in lines 10 to 12 of Table 16.12.6. Explanations are identical to those for the
preceding example.

MOMENT FOR FLAT PLATE DESIGN EXAMPLE
Line Equivalent Frame A B C D

Number
1 Total transverse width (in.) 144 72 180 90

2 Column strip width (in.) 72 36 72 36
3 Half middle strip width (in.) 2@36 36 2@54 54
4 C (in.4) from Example 16.11.3 474 474 362 362
5 Is (in.4) in βt 2000 2000 2500 2500
6 βt = EcbC/(2Ecs Is) 0.119 0.119 0.072 0.072
7 αf1 0 0 0 0
8 L2 /L1 0.80 0.80 1.25 1.25
9 αf1L2/L1 0 0 0 0
10 Exterior negative moment, percent to
98.8% 98.8% 99.3% 99.3%
column strip
11 Positive moment, percent to column
60% 60% 60% 60%
strip
12 Interior negative moment, percent to
75% 75% 75% 75%
column strip
16.13 DESIGN OF SLAB THICKNESS

AND REINFORCEMENT
Slab Thickness
Ordinarily the minimum thickness specified in ACI-8.3.1 to control deflections governs the
thickness for design. Of course, reinforcement for bending moment must be provided, but
the reinforcement ratio ρ required is usually well below 0.5ρtc; thus, it does not dictate slab
thickness.
Slab thickness based on shear strength requirements must also be investigated. The
shear requirement for two-way slabs (with beams) may be investigated by observing strips
1-1 and 2-2 in Fig. 16.13.1. Beams with αf1L2/L1 values larger than 1.0 are assumed to
carry the loads acting on the tributary floor areas bounded by 45° lines drawn from the
63
16.13 DESIGN OF SLAB THICKNESS AND REINFORCEMENT 663
45° 1
2 2
S
G H
Figure 16.13.1 Load transfer from floor area to beams.
corners of the panel and the centerline of the panel parallel to the long side (ACI-8.10.8.1).
If this is the case, the loads on the trapezoidal areas E and F of Fig. 16.13.1 are carried by
the long beams, and those on the triangular areas G and H are carried by the short beams.
The shear per unit width of slab along the beam is highest at the ends of slab strips 1-1 and
2-2, which, considering the increased shear at the exterior face of the first interior support,
is approximately equal to
 w S
Vu = 1.15  u  (16.13.1)
 2 
where S is the span length in the short direction.

If αf1L2 /L1 is equal to zero, there is, of course, no load on the beams (because there are
no beams). When the value of αf1L2 /L1 is between 0 and 1.0, the percentage of the floor load
carried by the beams should be obtained by linear interpolation. In such a case, the shear
expressed by Eq. (16.13.1) would be reduced, but the shear around the column due to the
portion of the floor load transferred directly to the columns by two-way action should be
investigated as for flat plate floors.
The shear strength requirement for flat slab and flat plate systems (without beams) is
treated separately in Sections 16.15, 16.16, and 16.18.
Reinforcement
When the nominal requirements for slab thickness as discussed in Section 16.6 are satis-
fied, no compression reinforcement will likely be required. The tension steel area required
within the strip being considered can then be obtained by the following steps:
factored moment Mu in the strip

1. required M n =
(φ = 0.90 — assumed, but reasonable for slabs)
fy Mn
2. m = , Rn = ,
0.85 fc′ bd 2
1 2 mRn 
ρ= 1 − 1 − , As = ρ bd
m fy 
64
Instead of using the equation for ρ in Step 2, the curves in Fig. 3.8.1 may be used.
Note also that the values of b and d to be used in Step 2 for negative moment in a column
strip with drop are the drop width for b and, for d, the smaller of the actual effective depth
through the drop and that provided by a drop thickness below the slab at no more than
one-fourth the distance between the edges of the column capital and the drop. For positive
moment computation, the full column strip width should be used for b, and the effective
depth in the slab for d. After obtaining the steel area As required within the strip, a number
of bars may be chosen so that they provide either the area required for strength, but not less
than the minimum amount, As,min, equal to 0.0018bt for Grade 60 steel (see ACI-8.6.1.1).
The spacing of reinforcing bars must not exceed the smaller of 2 times the slab thickness or
18 in. at critical sections, and the smaller of 3 times the slab thickness or 18 in. at other sec-
tions (ACI-8.7.2.2). In slabs of cellular or ribbed construction (i.e., two-way joist systems),
where the requirement for shrinkage and temperature reinforcement governs, maximum
spacing must not exceed the smaller of 5 times the slab thickness or 18 in. (ACI-24.4.3.3).
Reinforcement Details in Slabs without Beams

ACI-8.7, in particular ACI Fig. 8.7.4.1.3a, provides detailed dimensions for minimum
extensions required for each portion of the total number of bars in the column and middle
strips. Since the forces acting in the bars were empirically determined, there is no practical
means to evaluate the distances required to develop the reinforcement. Thus, past practice
and engineering judgment were used in preparing ACI Fig. 8.7.4.3.1a. In that figure, no
details for bent bars are shown because they are seldom used, although their use is still
permitted by ACI-8.7.4.1.3c when the depth-to-span ratio allows bends at 45° or less.
For unbraced frames, reinforcement lengths must be determined by analysis but may not
be less than those prescribed in ACI Fig. 8.7.4.1.3a. Also, ACI-8.7.4.2 requires the use of
“integrity steel,” which consists of a minimum of two of the column strip bottom bars pass-
ing continuously (or spliced with Class B splices or anchored within support) through the
column core in each direction. The purpose of this integrity steel is to provide some residual
strength following a single punching shear failure. The ACI Code also allows the use of full
mechanical and full welded splices as alternative methods of splicing the reinforcement.
Corner Reinforcement for Two-Way Slab (with Beams)

It is well known from plate bending theory that a transversely loaded slab simply supported
along four edges will tend to develop corner reactions as shown in Fig. 16.13.2, for which
reinforcement must be provided. Thus, in slabs supported by edge walls or on one or more
edge beams having a value of αf greater than 1.0, reinforcement shall be provided at exte-
rior corners in both the bottom and top of the slab (see Fig. 16.13.3). This reinforcement in
both the top and bottom of the slab must be sufficient to resist a moment equal to the max-
imum positive moment per foot of width in the slab (ACI-8.7.3.1). The reinforcement is to
be provided for a distance in each direction from the corner equal to one-fifth the longer
span (ACI-8.7.3.1.2), and it may be placed in a single layer parallel to the diagonal from the
corner in the top of the slab and perpendicular to that diagonal in the bottom of the slab, or
in two layers parallel to the sides of the slab (ACI-8.7.3.1.3), as shown in Fig. 16.13.3. The
latter is the preferred option in practice, as providing reinforcement along the diagonal can
be confusing in the field and would require bars of varying lengths. Reinforcement provided
for flexure in the primary directions can be used to satisfy this requirement (ACI-R8.7.3.1).
Since the maximum positive moment in the slab will exceed the negative moment at
the exterior slab support, the corner reinforcement amount computed per ACI-8.7.3.1
will govern whenever it is required. It is also noted that the moment per foot of width of
slab, whether computed using the Direct Design Method or the Equivalent Frame Method
(Section 16.20), must be distributed within the column strip width, which for a corner panel
is one-quarter of the shorter span. The corner reinforcement, however, is to be distributed
over one-fifth of the longer span. It is possible that the computed column strip width is
larger than the distance where the corner reinforcement is to be distributed (e.g., for square
panels). In such cases, it is recommended that the remaining portion of the column strip be
65
S
R
Edge shear distribution
R Edge shear distribution R
Figure 16.13.2 Edge reactions for a simply supported two-way slab.
Llong
(Llong)/5
(Llong)/5
As top per 8.7.3
As bottom per 8.7.3
Lshort
OPTION 1
Llong
(Llong)/5
(Llong)/5
Lshort
As in two layers
per 8.7.3
top and bottom
OPTION 2
Figure 16.13.3 Corner reinforcement in two-way slabs supported by edge walls or where one or
more edge beams have a value αf greater than 1.0. (Adapted from ACI 318-14 Commentary.)
reinforced with the same amount computed for the corner reinforcement. It is also recom-
mended that 50% of the top corner reinforcement be extended to one-fifth of the longer
span or 0.3Ln, whichever is larger, to be consistent with the required minimum bar exten-
sions at the exterior supports (ACI Table 8.7.4.1.3a). The remainder must be extended at
least to one-fifth of the longer span in accordance with ACI-8.7.3.1.2.
6
Crack Control
In addition to deflection control, crack control is the other major serviceability requirement
usually considered in the design of flexural members. ACI-24.3 gives criteria for distribut-
ing the flexural tension reinforcement in beams and one-way slabs to minimize crack width
under service loads. No ACI Code provisions are given for two-way floor (or roof) systems;
however, ACI Committee 224, Cracking, has suggested a formula to predict the possible
crack width in two-way acting slabs, flat slabs, and flat plates. The recommendations are
based on the work of Nawy et al. [16.72–16.75]. When the calculated crack width is con-
sidered excessive (there are no ACI Code limits for slabs), the distribution (size and spac-
ing) of flexural reinforcement may be adjusted [16.75] to reduce the width of the expected
cracks. Ordinarily, crack width is not a problem on two-way acting slabs designed with fy
less than or equal to 60,000 psi.
EXAMPLE 16.13.1
Investigate if the preliminary slab thickness of 6 1 2 in. in the two-way slab (with beams)
(see Fig. 16.3.4) is sufficient for resisting flexure and shear.
SOLUTION
For each of the equivalent frames A, B, C, and D, the largest bending moment in the
slab occurs at the exterior face of the first interior support in the middle strip slab. From
Table 16.12.4, this moment is observed to be 60/10, 30/5, 80/15, or 40/7.5 ft-kips per ft
of width in frames A, B, C, and D, respectively. Taking the effective depth to the con-
tact level between the reinforcing bars in the two directions, and assuming #5 bars with
3 -in. clear cover,
4

average d = 6.50 − 0.75 − 0.63 = 5.12 in.
Assuming φ = 0.90, the largest Rn required is
Mu 6000(12)
Rn = = = 254 psi
φ bd 2 0.90(12)(5.12)2
From Eq. (3.8.5), the reinforcement ratio ρ for this value of Rn is about 0.0045, which is
well below 0.35ρb = 0.0075, recommended for beams and one-way slabs not supporting
elements likely to be damaged by large deflections (Section 12.10). For two-way slabs,
deflections should be smaller. Hence excessive deflection should not be expected; this is
further verification of the minimum thickness formulas given in ACI-8.3.1.
The factored floor load wu is

wn = 1.2 wD + 1.6 wL = 319 psf
Since all af 1L2 /L1 values are well over 1.0, take V from Eq. (16.13.1) as
1.15wu S 1.15(0.319)(20)
Vu = = = 3.67 kips
2 2
1
Vc = 2 λ fc′bw d = 2(1.0) 3000 (12)(5.12) = 6.73 kips
1000

φVc = 0.75(6.73) = 5.05 kips > [Vu = 3.67 kips] OK
Note that the factored shear 3.67 kips is maximum at strip 1-1 of Fig. 16.13.1; actually,
the average for all such strips will be lower.
67
EXAMPLE 16.13.2
Design the reinforcement in the exterior and interior spans of a typical column strip and
a typical middle strip in the short direction of the flat slab (see Fig. 16.3.6). As described
earlier, in Section 16.3, fc′ = 3000 psi and fy = 60,000 psi.
SOLUTION
(a) Moments in column and middle strips. The typical column strip is the column
strip of equivalent frame C of Fig. 16.10.2; but the typical middle strip is the sum
of two half middle strips, taken from each of the two adjacent equivalent frames
C. The factored moments in the typical column and middle strips are shown in
Table 16.13.1.
(b) Design of reinforcement. The design of reinforcement for the typical column
strip is shown in Table 16.13.2; for the typical middle strip, it is shown in
Table 16.13.3. Because the moments in the long direction are larger than those
in the short direction, the larger effective depth is assigned to the long direction
wherever two layers of steel (one in each direction) are required. This occurs in
the top steel at the intersection of column strips and in the bottom steel at the
intersection of middle strips. Assuming #5 bars and 3 4 -in. clear cover, the effec-
tive depths provided at various critical sections of the long and short directions
are shown in Fig. 16.13.4.
(c) Slab thickness for flexure. For fc′ = 3000 psi and fy = 60,000 psi, the reinforcement
ratio corresponding to ρ(εt = 0.005) = 0.0135 (Table 3.6.1). The actual percentages
used (line 6 of Tables 16.13.2 and 16.13.3) are nowhere near this maximum. This
result was expected since the slab thickness was governed by the minimum thick-
ness required for deflection control.
TABLE 16.13.1 FACTORED MOMENTS IN A TYPICAL COLUMN STRIP AND MIDDLE

STRIP, EXAMPLE 16.13.2 (FLAT SLAB)
Line Moments at Critical Negative Positive Negative Negative Positive Negative

Number Section (ft-kips) Moment Moment Moment Moment Moment Moment
1 Total M in column
and middle strips
(Fig. 16.10.2, frame C) –88 +147 –205 –190 +103 –190
2 Percentage to column
strip (Table 16.12.5) 91.3% 60% 75% 75% 60% 75%
3 Moment in column –80 +88 –154 –143 +62 –143
strip
4 Moment in middle –8 +59 –51 –47 +41 –47
strip
(Continued)
68
TABLE 16.13.2 DESIGN OF REINFORCEMENT IN COLUMN STRIP OF FRAME C,

EXAMPLE 16.13.2 (FLAT SLAB) (fy = 60,000 psi, f c′ = 3000 psi)
Line Item Negative Positive Negative Negative Positive Negative

No. Moment Moment Moment Moment Moment Moment
1 Moment, Table 16.13.1,
line 3 (ft-kips) –80 +88 –154 –143 +62 –143
2 Width b of drop or strip (in.) 100 120 100 100 120 100
3 Effective depth d (in.) 8.81 6.44 8.81 8.81 6.44 8.81
4 Mu /φ (ft-kips) –89 +98 –171 –159 +69 –159
5 Rn (psi) = Mu /(φbd2) 138 236 264 246 166 246
6 ρ, Eq. (3.8.5) 0.24% 0.41% 0.47% 0.43% 0.29% 0.43%
7 As = ρbd 2.11 3.17 4.14 3.78 2.24 3.78
a
8 As,min = 0.0018bt 2.16 1.62 2.16 2.16 1.62 2.16
9 N = larger of (7) or (8)/0.31 7 11 14 13 8 13
10 N = width of strip/(smaller 5 8 5 5 8 5
of 2t or 18 in.)
11 N required, larger of (9) 7 11 14 13 8 13
or (10)
a
bt = 100(10.5) + 20(7.5) = 1200 sq in. for negative moment region.
TABLE 16.13.3 DESIGN OF REINFORCEMENT IN MIDDLE STRIP OF FRAME C,

EXAMPLE 16.13.2 (FLAT SLAB) (fy = 60,000 psi, fc′ = 3000 psi)
Line Item Negative Positive Negative Negative Positive Negative

No. Moment Moment Moment Moment Moment Moment
1 Moment, Table 16.13.1, line 4

(ft-kips) –8 +59 –51 –47 +41 –47
2 Width of strip, b (in.) 180 180 180 180 180 180
3 Effective depth d (in.) 6.44 5.81 6.44 6.44 5.81 6.44
4 Mu/φ (ft-kips) –9 +66 –57 –52 +46 –52
5 Rn (psi) = Mu/(φbd2) 14 130 92 84 91 84
6 ρ, Eq. (3.8.5) 0.02% 0.22% 0.16% 0.14% 0.15% 0.14%
7 As = ρbd 0.23 2.30 1.85 1.62 1.57 1.62
8 As,min = 0.0018bt 2.43 2.43 2.43 2.43 2.43 2.43
9 N = larger of (7) or (8)/0.31a 8 8 8 8 8 8
10 N = width of strip/(smaller 12 12 12 12 12 12
of 2t or 18 in.)
11 N required, larger of line 9 12 12 12 12 12 12
or 10
a
A mixture of #5 and #4 bars could have been selected.
(Continued)
69
16.14 SIZE REQUIREMENT FOR BEAM 669
Long direction Short direction
d = 8.81”
d = 6.44”
d = 8.81”
d = 9.44” d = 6.44” d = 9.44”
d = 6.44”
d = 5.81”
d = 6.44”
d = 6.44” d = 6.44” d = 6.44”
d = 8.81”
d = 6.44”
d = 8.81”
d = 9.44” d = 6.44” d = 9.44”
Figure 16.13.4 Effective depths provided at critical sections in the flat slab of Example 16.13.2.
16.14 SIZE REQUIREMENT FOR BEAM (IF USED)

IN FLEXURE AND SHEAR
The size of the beams along the column centerlines in a two-way slab (with beams) should
be sufficient to provide the bending moment and shear strengths at the critical sections.
For approximately equal spans, the largest bending moment should occur at the exterior face
of the first interior column where the available section for strength computation is rectangular
because the effective slab projection is on the tension side. Then, with the preliminary beam
size, the required reinforcement ratio ρ may be determined. Although unlikely to be a problem
with T-sections, deflection must be investigated if excessive deflections are of concern.
The maximum shear in the beam should also occur at the exterior face of the first interior
column. The shear diagram for the exterior span may be obtained by placing the negative
moments already computed for the beam at the face of the column at each end and loading
the span with the percentage of floor load interpolated (ACI-8.10.8.1) between αf 1L2/L1 = 0
and αf 1L2 /L1 ≥ 1.0. As discussed in Section 9.2, the stem (web) bw d should for practicality
be sized such that nominal shear stress vn = Vu /(φ bwd) does not exceed about 6 fc′ at the
critical section d from the face of support.
EXAMPLE 16.14.1
Investigate if the preliminary overall sizes of 14 × 28 in. for the long beam and 12 × 24 in.
for the short beam are suitable for the two-way slab (with beams) design example.
SOLUTION
Since the values of αf , or of αf 1L2 /L1, are considerably larger than 1.0 for all beam spans,
there is to be no reduction of the floor load going into the beams from the tributary areas
(ACI-8.10.8.1). As shown in Fig. 16.14.1, the most critical span is B1 for the long direc-
tion and B5 for the short direction. Actually, the load acting on the clear span of the beam
should include the floor load (including the weight of the beam stem itself or any other
load) directly over the beam stem width plus the floor load on the tributary areas bounded
by the 45° lines from the corner of the panel. Also, for practical purposes, it is acceptable
to consider the shear due to floor load at the face of column equal to one-half of the floor
load on the tributary areas between column centerlines, as shown in Fig. 16.14.1.
(Continued)
670
B1 B2
10’ 5’ 10’
Factored floor
6.38k/ft 6.38k/ft
load = 319 psf
57 ft-fk 216 ft-fk
B8
B6
B6
23’– 9”
25’– 0”
20’–0”
B1 B2
Beam B1
6.38k/ft
39 ft-k 140 ft-k
B7
B5
B5
18’–9”
20’–0”
B3 B4
Beam B5
Figure 16.14.1 Tributary areas for the beams around the two-way slab panel of Example 16.14.1.
(a) Size of long beam B1. The negative moments at the face of supports, 57 and
216 ft-kips, are taken from Table 16.12.4, frame A.
1
weight of beam stem = 14(21.5) (150) = 314 lb/ft
144
1
maximum negative moment = (1.2)(0.314)(23.75)2 + 216
10
= 21 + 216 = 237 ft-kips

b = 14 in, d = 28 − 2.5 (assume one layer of steel) = 25.5 in.
Mu 237(12, 000 )
Rn = = = 347 psi
φ bw d 2
0.90(14)(25.5)2
From Eq. (3.8.5), ρ = 0.006, which is well below ρ (εt = 0.005) = 0.0135. Perhaps the

beam size should be reduced. From Fig. 16.14.1,
total factored floor load on B1 = 6.38(15) = 95.7 kips
23.75 1 216 − 57
max Vu = 1.115(1.2)(0.314) + (95.7) +
2 2 23.75
= 5.1 + 47.9 + 6.7 = 59.7 kips
Vu 59, 700
vn = = = 223 psi = 4.1 fc′ < 6 fc′ OK
φ bw d 0.75(14)(25.5)
(b) Size of short beam B5. The negative moments at the face of supports, 39 and 140
ft-kips, are taken from Table 16.12.4, frame C.
1
weight of beam stem = 12(17.5) (150) = 219 Ib/ft
144
1
maximum negative moment = (1.2)(0.219)(18.75)2 + 140
10
= 9 + 140 = 149 ft-kips

b = 12 in. d = 24 − 2.5(assume one layer of steel) = 211.5 in.
Mu 149(12, 000)
Rn = = = 358 psi
φ bw d 2
0.90(12)(21.5)2
(Continued)
671
1 6 . 1 5 S H E A R S T R E N G T H I N T W O - WAY F L O O R S Y S T E M S 671
From Eq. (3.8.5), ρ = 0.007, which is well below ρ(εt = 0.005) = 0.0135. From Fig. 16.14.1,

total factored floor load on B5 = 6.38(10) = 63.8 kips
18.75 1 140 − 39
max Vu = 1.15(1.2)(0.219) + (63.8) +
2 2 18.75
= 2.8 + 31.9 + 5.4 = 40.1 kips

V 40,100
vn = u = = 207 psi = 3.8 fc′ < 6 fc′ OK
φ bw d 0.75(12)(21.5)
As mentioned above, the size of beams in both the long and short directions should prob-
ably be reduced; the nominal stress vn is well below 6 fc′ at the face of support and is
even lower at d therefrom.
16.15 SHEAR STRENGTH IN TWO-W AY FLOOR

SYSTEMS
The shear strength of a flat slab or flat plate floor around a typical interior column under
dead and full live loads is analogous to that of a square or rectangular spread footing sub-
jected to a concentrated column load, except each is an inverted situation of the other. The
area enclosed between the parallel pairs of centerlines of the adjacent panels of the floor
is like the area of the footing, because there is no shear force along the panel centerline of
a typical interior panel in a floor system. Consequently the discussion here is essentially
identical to what is included in Chapter 19 on footings.
The shear strength of two-way slab systems without shear reinforcement has been stud-
ied by many investigators [16.76–16.92, 16.142]. An excellent summary is provided by
ASCE-ACI Task Committee 426 [16.83].
One-Way Action
The shear strength of a flat slab or flat plate should be computed for one-way action (ACI-
8.5.3.1.1) and for two-way action (ACI-8.5.3.1.2). To compute the one-way shear strength
(one-way action), the slab is treated as a wide beam, where the critical section is parallel to
the panel centerline in the transverse direction and extends across the full distance between
two adjacent longitudinal panel centerlines. As in beams, this critical section of width bw
times the effective depth d is located at a distance d from the face of the equivalent square
column capital or from the face of the drop panel, if any. The nominal strength in usual
cases where no shear reinforcement is used may be computed as
Vn = Vc = 2 λ fc′bw d (16.15.1)1
according to the simplified method of ACI-22.5.5.1.
1 For SI, with fc′ in MPa and bw and d in mm, ACI-318-14M gives

Vc = 0.17λ fc′bw d (16.15.1)
672
Figure 16.15.1 Punching shear failure along a truncated pyramid around the column. (Photo
courtesy of James O. Jirsa.)
Two-Way Action
A second failure mode may occur by diagonal cracking along a truncated cone or pyramid
around columns, concentrated loads, or reactions (see Fig. 16.15.1). This failure mode is
commonly called “punching” shear. The critical section is assumed to be located so that its
periphery b0 is at a distance d/2 (i.e., half the effective depth) from the edges or corners of
a column, concentrated load, or reaction areas and from changes in slab thickness, such as
edges of capitals or drop panels.
The ACI Code states that b0 is a “minimum but need not approach closer than d/2 …”
(ACI-22.6.4.1). Some confusion may arise as to whether the “minimum” at d/2 would
require using a curved-corner perimeter around a square or rectangular column. Since exact-
ness is neither improved nor reduced by calculating the critical section b0 by such elaborate
procedures, ACI-22.6.4.1.1 permits the critical section for square or rectangular loaded areas
to have four “straight sides.” For slabs with changes in thickness, such as slabs with capitals
or drop panels, shear must be checked at several sections to determine the critical section.
When shear reinforcement is not used, the nominal shear strength Vn = Vc, which is given
by ACI-22.6.5.2 as the smallest of
Vc = 4 λ fc′b0 d (16.15.2a)2
 4 2
Vc =  2 +  λ fc′b0 d (16.15.2b)
 β
 α d
Vc =  2 + s  λ fc′b0 d (16.15.2c)2
 b0 
where
b0 = perimeter of critical section
β = ratio of long side to short side of the column
αs = 40 for interior columns, 30 for edge columns, and 20 for corner columns
d = average of effective depths in the two orthogonal directions
2 For SI, with fc′ in MPa and b0 and d in mm, ACI-318-14M gives
Vc = 0.33λ fc′b0 d (16.15.2a)
 2
Vc = 0.17  1 +  λ fc′b0 d (16.15.2b)
 β
 α d
Vc = 0.083  2 + s  λ fc′b0 d (16.15.2c)
 b0 
673
Equation (16.15.2b) recognizes that there should be a transition between, say, a square
column (β = 1) where Vc might be based on 4 λ fc′ for two-way action, and a wall (β = ∞),
where Vc should be based on the 2 λ fc′ used for one-way action, as for beams. Note that
unless β is larger than 2.0, Eq. (16.15.2b) does not control. For shapes other than rectangu-
lar (e.g., L-shaped), β is taken as the ratio between the longest and the largest perpendicular
dimensions of the effective loaded area (ACI-R22.6.5.2).
The two-way shear strength may be reduced below 4 λ fc′, even for a square concen-
trated load area, both (1) as the distance to the critical section from the concentrated load
increases, such as for drop panels, and (2) as the perimeter becomes large in comparison
to the slab thickness [e.g., a 6-in. slab supported by a 30-in square column (b0 /d ≈ 28)].
Equation (16.15.2c) accounts for this reduced strength.
In the application of Eqs. (16.15.2), b0 is the perimeter of the critical section at a distance
d/2 from the edge of column capital or drop panel. For Eq. (16.15.2c), αs for an “interior
column” applies when the perimeter is four-sided, for an “edge column” when the perim-
eter is three-sided, and for a “corner column” when the perimeter is two-sided. As shown
by Fig. 16.15.2, Eq. (16.15.2c) will give a Vc smaller than 4 λ fc′b0d for large columns
(or very thin slabs), such as a square interior column having side larger than 4d, a square
edge column having side larger than 4.33d, and a square corner column having side larger
than 4.5d. Thus, the nominal shear strength Vc in a two-way system is generally set by
Eq. (16.15.2a): that is, Vc = 4 λ fc′b0 d , unless either of Eq. (16.15.2b) or Eq. (16.15.2c)
gives a lesser value.
Unlike the design of beams, a minimum amount of shear reinforcement is not required
for two-way slabs because there is the possibility of load sharing between the weak and
strong areas. For deep, lightly reinforced one-way slabs, however, shear failure may occur
at loads less than Vc, especially if made of high-strength concrete (ACI-R7.6.3.1). Thus it
would be prudent to provide a minimum amount of shear reinforcement even if not required
by the code in these cases.
The investigation for concentric shear (without moment) transfer from slab to column
is shown in Examples 16.15.1 and 16.15.2, for the flat slab and flat plate design examples,
respectively. When both shear and moment must be transferred from the slab to the column,
ACI-8.4.4.2 applies, as will be discussed in Section 16.18.
d d d
2 2 2
s
b0 = 4(s + d) b0 = 3s + 2d b0 = 2s + d
40 30 20
If s = 4d, =2 If s = 4.33d, =2 If s = 4.5d, =2
b0 /d b0 /d b0 /d
(a) Interior column (b) Edge column (c) Corner column
Figure 16.15.2 Minimum size of square columns for Vc = 4 λ fc′b0 d .
674
EXAMPLE 16.15.1
Compute the shear strength for one-way and two-way actions in the flat slab design of
Example 16.3.2 for an interior column with no bending moment to be transferred. Note
that slab thickness is 7.5 in. and that normal-weight concrete with fc′ = 3000 psi is used
(i.e., λ = 1.0)
SOLUTION
(a) One-way action. Investigation for one-way action is made for sections 1-1 and 2-2
in the long direction, as shown in Fig. 16.15.3(a). The short direction has a wider
critical section and a shorter span; thus it does not control. For section 1-1, if the
entire width of 20 ft is conservatively assumed to have an average effective depth of
(6.44/2 + 5.81/2) = 6.12 in. (see Fig. 16.3.4),
Vu = 0.337(20)(9.52) = 64 kips (section 1-1)
1
Vn = Vc = 2 λ fc′(240)(6.12) = 161 kips
1000
φVn = 0.75(161) = 121 kips > Vu OK
If, however, bw is taken as 84 in. and d as 9.12 in. on the contention that the increased
depth d is only over a width of 84 in.,
1
Vn = Vc = 2 λ fc′(84)(9.12) = 84 kips
1000
1 2
avg d = 6.12”
7.82’
avg d = 9.12” 9.52’ Drop
Equivalent
20’– 0”
20’– 0”
7’–0”
square
column
4.43’
8’– 4”
1 2
25’–0” 25’–0”
(a) One-way action (long direction)
= 3.06”
avg d
Section 2–2
2
Section 1–1 Drop

Capital
20’– 0”
20’–0”
7’–0”
7.51’
avg d 5’– 0” avg d

= 4.56” = 4.56” avg d avg d
2 2 = 3.06” 8’– 4” = 3.06”
5.76’ 2 2
8.84”
25’– 0” 25’– 0”
(b) Two-way action
Figure 16.15.3 Critical sections for shear in the flat slab of Example 16.15.1. (Continued)
675
This latter value is probably unrealistically low.

For section 2-2, the shear resisting section has a constant d of 6.12 in.; thus
Vu = 0.337(20)(7.82) = 53 kips (section 2-2)

φVn = 121 kips > Vu at sections 1-1 and 2-2 OK
In general, it will be rare that one-way action will govern.

(b) Two-way action. The critical sections for two-way action are the circular section
1-1 at d/2 = 4.56 in. from the edge of the column capital and the rectangular section
2-2 at d/2 = 3.06 in. from the edge of the drop, as shown in Fig. 16.15.3(b). Since
there are no shearing forces at the centerlines of the four adjacent panels, the shear
forces around the critical sections 1-1 and 2-2 in Fig. 16.15.3(b) are
 π (5.76)2   π (5.76)2 
Vu = 0.337 500 −  + 1.2(0.038) 7(8.33) − 
 4   4 
= 159.2 + 1.5 = 161 kips (section 1-1)
In the second term, the 0.038 is the weight of the 3-in. drop in ksf

Vu = 0.337[500 − 8.84(7.51)] = 146 kips (section 2-2)
Compute the shear strength at section 1-

1 around the perimeter of the capital
[Fig. 16.15.3(b)],
b0 217
b0 = π (5.76)12 = 217 in., = = 23.8
d 9.12
Since b0 /d > 20, and β = 1, Eq. (16.15.2c) controls. Thus,
 40(9.12) 
φVn = φVc = φ  2 +  λ fc′b0 d = φ (3.68λ fc′b0 d )
 217 

= 0.75 3.68λ fc′(217)(9.12)
1
= 299 kips (section 1-1)
1000
At section 2-2, Fig. 16.15.3(b),
b0 392
b0 = [(2(8.84) + 2(7.51)]12 = 392 in.: = = 64.1
d 6.12
and since b0/d > 20, Eq. (16.15.2c) controls. Thus,
 40(6.12) 
φVn = φVc = φ  2 +  λ fc′b0 d = φ (2.62 λ fc′b0 d )
 392 

= 0.75  2.62 λ fc′(392)(6.12)
1
= 258 kips (section 2-2)
1000
Though both sections 1-1 and 2-2 have φVn significantly greater than Vu, the section
around the drop panel is loaded to a slightly higher percentage of its strength (57% for
section 2-2 vs. 54% for section 1-1). In any case, shear reinforcement is not required at
this interior location.
67
EXAMPLE 16.15.2
Compute the one-way and two-way shear strengths in the flat plate of Example 16.3.3
for an interior column with no bending moment to be transferred. Note that slab thick-
ness is 5.5 in. and that normal-weight concrete of fc′ = 4000 psi is used (i.e., λ = 1.0).
SOLUTION
(a) One-way action. Assuming 3 4 -in. clear cover and #4 bars, the average effective
depth when bars in two directions are in contact is

avg d = 5.50 − 0.75 − 0.50 = 4.25 in.
Referring to Fig. 16.15.4(a),
Vu = 0.198(12)6.65 = 15.8 kips

12
φVn = φVc = 0.75 2(1.0) 4000  (12) 4.25 1000 = 58.1 kips > Vu OK
Avg d = 4.25” 6.65’

2.125” Section 2–2
10”
12’–0” 10” 14.25” 12’– 0”
2.125” 12”
12” 2.125” 2.125”
16.25”
1
15’–0” 15’–0”
(a) One-way action (long direction) (b) Two-way action
Figure 16.15.4 Critical sections for shear in the flat plate of Example 16.15.2.
(b) Two-way action. Referring to Fig. 16.15.4(b),

Vu = 0.198 [15(12) − 1.35(1.19)] = 35.3 kips
The perimeter of the critical section at d/2 around the column is
b0 61.0
b0 = 2(16.25) + 2(14.25) = 61.0 in., = = 14.4 < 20
d 4.25
Since b0 /d < 20, and β = 1.2, Eq. (16.15.2a) controls. Thus,
φVn = φVc = 0.75  4(1.0) 4000  (61.0)4.25 1000

1
= 49.2 kips > V u OK
Shear reinforcement is not required at this interior location.
16.16 SHEAR REINFORCEMENT IN FLAT

PLATE FLOORS
In flat plate floors, where neither column capitals nor drop panels are used, shear reinforce-
ment is frequently necessary. In such cases, two-way action usually controls. The shear
reinforcement may take the form of single-or multiple-leg stirrups placed in vertical sec-
tions around the column [Fig. 16.16.1(a)]. Alternatively, they may consist of shearheads,
67
1 6 . 1 6 S H E A R R E I N F O R C E M E N T I N F L AT P L AT E F L O O R S 677
which are steel I-or channel-shaped sections welded to form four (or three for an exte-
rior column) identical arms at right angles and uninterrupted within the column section
[Fig. 16.16.1(b)].
Stirrups can be difficult to install in the slab around the column because the region is
commonly congested with the column and slab reinforcement. Shearheads can be used
instead, but sometimes they are more expensive to fabricate and install. Alternatively,
headed shear studs have been used in Canada and Europe as shear reinforcement in slabs
for many years [16.101, 16.144], and today are also widely used in the United States. This
type of reinforcement consists of headed steel studs welded to a steel strip as shown in
Fig. 16.16.2. The strips may be arranged in orthogonal directions for rectangular and square
columns or in radial directions for both rectangular (Fig. 16.16.2) and circular columns.
Critical section
outside slab shear Critical
reinforcement c1
Critical section section
through slab shear d d
Critical 2
reinforcement at d/2 2 d d
section 2
from face of column 2
c2
d d
2 2
Stirrups 3
(L – c1 )
4 v 2
Lv
Double U-stirrup
(a) Bar reinforcement (b) Shearhead reinforcement
Figure 16.16.1 Bar and shearhead reinforcement in flat plate floors.
Figure 16.16.2 Radial-type arrangement of headed shear stud reinforcement. (Photo courtesy of Cary
Kopczynski & Co.)
678
The strength of two-way slab systems with shear reinforcement has been summarized
by Hawkins [16.94]. Corley and Hawkins [16.93, 16.110] have studied shearhead rein-
forcement. Other studies of shear reinforcement in flat plates can be found in Refs. [16.95–
16.102, 16.115, and 16.148], Yamada, Nanni, and Endo [16.102], and Pillai, Kirk, and
Scavuzzo [16.115].
Shear Strength Provided by Stirrups

When stirrups are used in accordance with ACI-8.7.6, their nominal shear strength is
computed as
 Av f yt d 
Vn = Vc + Vs = 2 λ fc′b0 d +  ≤ 6 fc′b0 d (16.16.1)
 s 
where b0 is the perimeter of the critical section, Av is the total stirrup bar area around b0, s
is the stirrup spacing in the direction perpendicular to the column face, and fyt is the yield
strength of the bars used for stirrups. Note that shear reinforcement is required wher-
ever Vu exceeds φVc based on Vc of Eqs. (16.15.2), which can be greater than 2 λ fc′b0 d .
However, in the design of stirrups, Vc for Eq. (16.16.1) may not be taken greater than
2 λ fc′b0 d (ACI-22.6.6.1). Furthermore, according to ACI-22.6.6.2, the nominal strength
Vn (i.e., Vc + Vs) must not exceed 6 fc′b0 d .
When stirrups are provided, shear stresses must also be checked at a critical section
located at a distance d/2 beyond the point at which shear reinforcement is to be terminated.
The shape of this critical section is a polygon that minimizes b0, as shown by the outermost
critical section in Fig. 16.16.1(a). At this critical section, the factored shear stress, vu, can-
not exceed φ 2 λ fc′ .
The requirements for the geometry and anchorage of stirrups in slabs (ACI-8.7.6.2) may
be difficult to meet in slabs thinner than 10 in. Furthermore, single-or multiple-leg stirrups
are permitted as shear reinforcement in slabs (and footings) only when the effective depth
d is greater than or equal to 6 in., but not less than 16 times the diameter of the stirrups
(ACI-22.6.7.1).
In accordance with ACI-8.7.6.3, the first stirrup must be located at a distance d/2 from
the face of the column, and the spacing between stirrups cannot exceed d/2. In addition, the
spacing between the vertical legs of stirrups (measured parallel to the face of the column)
must not exceed 2d. Details of typical bar arrangements for stirrups in slabs are given in
ACI-R8.7.6.
Shear Strength Provided by Shearheads

Shear strength may be provided by shearheads under ACI-22.6.9 whenever Vu /φ at the
critical section is between that permitted by Eqs. (16.15.2) and 7 fc′b0 d . These provisions,
based on the tests of Corley and Hawkins [16.93], apply only where shear alone (i.e., no
bending moment) is transferred at an interior column. When there is moment transfer to
columns, ACI-22.6.9.12 applies, as is discussed in Section 16.18.
With regard to the size of the shearhead, it must furnish a ratio αv of 0.15 or larger
(ACI-22.6.9.5) between the flexural stiffness for each shearhead arm (Es Ix) and that for the
surrounding composite cracked slab section of width (c2 + d), or
Es I x
min α v = = 0.15 (16.16.2)
Ec ( composite I s )
where c2 is the dimension of the column measured perpendicular to the span for which the
moments are being calculated. The steel shape used must not be deeper than 70 times its
web thickness, and the compression flange must be located within 0.3d of the compression
679
surface of the slab (ACI-22.6.9.2 and 22.6.9.4). In addition, the plastic moment capacity Mp
of the shearhead arm must be at least (ACI-22.6.9.6).
Vu   c1  

min M p = hv + α v  Lv −   (16.16.3)
2 ηφ  2 
where
η = number (usually 4) of identical shearhead arms
Vu = factored shear around the periphery of column face
hv = depth of shearhead
Lv = length of shearhead measured from column centerline
c1 = dimension of the column measured in the direction of the span for which the
moments are being calculated
φ = 0.90, strength reduction factor for tension-controlled members
The purpose of Eq. (16.16.3) is to ensure that the required shear strength of the slab is
reached before the flexural strength of the shearhead is exceeded.
The length of the shearhead, Lv, should be such that the nominal shear strength Vn
will not exceed 4 fc′b0 d computed at a peripheral section located at 3 4 ( Lv − c1 / 2) along
the shearhead, but no closer elsewhere than d/2 from the column face (ACI-22.6.9.8 and
22.6.9.10). This length requirement is shown in Fig. 16.16.1(b).
When a shearhead is used, it may be considered to contribute a resisting moment
(ACI-22.6.9.7)
φ α vVu  c1 
Mv =  Lv − 2  (16.16.4)
2η
to each column strip, but not more than 30% of the total moment resistance required in the
column strip, nor more than the change in column strip moment over the length Lv , and nor
more than the required Mp given by Eq. (16.16.3).
Shear Strength Provided by Headed Shear Studs

Shear reinforcement in the form of headed steel studs welded to a steel strip (Fig. 16.16.2),
often called headed shear studs, are a common alternative to stirrups or shearheads. The
steel studs provide the same function as the vertical legs of stirrups, while the head of the
stud and the bottom strip provide anchorage to the stud.
Design requirements for headed shear stud reinforcement are given in ACI-8.7.7 and
ACI-22.6.8, The nominal shear strength is computed as
 Av f yt d 
Vn = Vc + Vs = 3λ fc′b0 d +  ≤ 8 fc′b0 d (16.16.5)
 s 
where b0 is the periphery around the critical section, Av is the total area of shear studs
around b0, s is the stud spacing in the direction perpendicular to the column face, and fyt is
the yield strength of the studs.
Shear reinforcement is required wherever Vu exceeds φVc based on Vc of Eqs. (16.15.2).
However, in the design of headed shear studs, Vc for Eq. (16.16.5) may not be taken
greater than 3λ fc′b0 d (ACI-22.6.6.1), and the nominal strength Vn (i.e., Vc + Vs) must
not exceed 8 fc′b0 d (ACI-22.6.6.2). Furthermore, a minimum amount of shear stud rein-
forcement must be provided, such that
Av b
≥ 2 fc′ 0 (16.16.6)
s f yt
680
Outermost
Column peripheral
line of studs
S
d/2
≤ d/2
(typ.) ≤ 2d
(typ.)
d/2
Shear
critical
sections
Figure 16.16.3 Layout requirements (plan view) for headed shear studs at an interior column.
(Adapted from ACI 318-14 Commentary.)
Although the ACI code allows higher shear stress limits for the concrete contribution (3 fc′ )
and for the maximum shear stress (8 fc′ ), the authors recommend that the same limits that
apply to stirrups be used (i.e., 2 fc′ and 6 fc′ , respectively).
As for slabs with stirrups, the shear studs must be extended over a distance such that
the factored shear stresses, vu, at the outermost critical section do not exceed φ 2 λ fc′. The
shape of this critical section is defined as a polygon located at a distance d/2 beyond the
outermost peripheral line of shear studs, as shown by in Fig. 16.16.3.
Shear stud location and spacing limits are given in ACI Table 8.7.7.1.2. The first line
of shear studs is not to be located farther than d/2 from the face of the column. Shear stud
spacing in the direction perpendicular to the column face should not exceed d/2, as with
stirrups. However, stud spacing may be increased to 0.75d for subsequent lines of shear
studs if the factored shear stress, vu, at the critical section located at d/2 from the face of
the column does not exceed φ 6 fc′b0 d. Stud spacing in the direction parallel to the column
face cannot exceed 2d. A typical radial-type arrangement of headed shear studs at an inte-
rior column is shown in Fig. 16.16.2. Research has shown that a radial-type arrangement of
headed studs leads to better ductility in slabs compared to a cruciform-type layout [16.148].
EXAMPLE 16.16.1
For the flat plate shown in Fig. 16.16.4, compute the shear strength for one-way and two-
way actions around an interior column. Assume a slab thickness of 8 in. and that the serv-
ice live load is 200 psf. If the required nominal shear strength Vn for two-way action is
between that permitted by Eqs. (16.15.2) and 6 fc′b0 d , determine the Av /s requirement for
shear reinforcement at the peripheral critical section and show the variation in the nominal
shear stress (i.e., factored shear Vu divided by φ b0 d) from the critical section to the panel
centerline. Use normal-weight concrete with fc′ = 4000 psi and fy = 60,000 psi; assume #5
bars for the slab longitudinal reinforcement and #3 bars for the stirrups, if needed.
SOLUTION
(a) One-way action.
1
wu = 1.2 wD + 1.6 wL = 1.2(150)(8) + 1.6(200)
12

= 120 + 320 = 440 psf
(Continued)
681
The average effective depth in a column strip is

avg d in column strip = 8.00 –0.75 –0.375 –0.63 = 6.25 in.
which is greater than 6 in. and 16db of stirrup (ACI-22.6.71). OK
For a 12-in.-wide strip along section 1-1 of Fig. 16.16.4,
vn =
Vu
=
440(7.98)
φ bw d 0.75(15)(6.25)
= 50 psi < 2 λ fc′ = 126 psi ( ) OK
(b) Two-way action. Referring to Section 2-2 of Fig. 16.16.4,

b0 = 2(18.25) + 2(16.25) = 69 in.
b0 69
= = 11.0 < 20
d 6.25
With a rectangular perimeter b0 having a ratio of long to short side of less than 2
(i.e., β less than 2), and b0 /d less than 20 for an interior column, Eq. (16.15.2a) controls;
thus, the strength without shear reinforcement is Vc = 4 λ fc′b0 d . Using nominal stress
vn = Vu /φ b0d,
Vu 440 [ 270 − 1.52(1.35)]
vn = = = 365 psi > (4 λ fc′ = 253 psi)
φ b0 d 0.75(69)(6.25)
7.98’
avg d = 6.25”
10”×12” column
16.25”
15’–0”
Section 2–2
18.25” 8” slab
1
18’– 0”
Figure 16.16.4 Critical sections for shear Example 16.16.1.

2 3 45 6
81.88” 16.25” 81.88”
Excess (vn –2λ fc’) diagram

15’–0”
365 psi
2 3 4 5 6 2λ fc’ = 126 psi

91 psi
41 psi
98.88” 98.88” 17 psi
2 3 4 5 6
18.25”
18’–0” 98.88” in long direction
81.88” in short direction
Figure 16.16.5 Variation of two-way nominal shear stress (Vu /φ b0d), Example 16.16.1.
(Continued)
682
Since the maximum nominal shear stress of 365 psi exceeds 4 λ fc′ psi, but not the
maximum 6 fc′ = 380 psi permitted when bar or wire shear reinforcement is used, shear
reinforcement is required to take the excess stress vn that exceeds 2 λ fc′ = 126 psi. The
shear reinforcement in this case may consist of properly anchored bars or wires and need
not be a shearhead. The Av /s requirement around the critical section of 69-in. periphery,
from applying Eq. (16.16.1) with required Vn = Vu /φ, is
Av
Vu
φ
(
− 2 λ fc′ b0 d) (
vn − 2 λ fc′ b0 )
= =
s f yt d f yt

(365 − 126)(69)
= = 0.275 in.
60, 000
Assuming s = d/2 ≈ 3-in. spacing (ACI-8.7.6.3),

required Av = 0.83 sq in.
If a double #3 U stirrup is used at each of the four sides,

provided Av = 4(2)(0.11) = 0.88 sq in.
which is greater than the required Av of 0.83 sq in. OK

The variation of the nominal shear stress vn from the maximum value of 365 psi to zero
at the panel centerline over the equally spaced points 2 to 6 is shown in Fig. 16.16.5.
The nominal shear stress (Vu /φb0d) drops to 126 psi in a rather steep manner. The num-
ber and spacing of the U stirrups may be laid out with the aid of the excess (vn − 2 λ fc′)
diagram.
EXAMPLE 16.16.2
Redesign the connection using shearhead reinforcement for the two-way shear action of
Example 16.16.1.
SOLUTION
(a) Two-way action. Since the maximum nominal shear stress vn = Vu /φ b0 d of 365 psi
is between 4 λ fc′ = 253 psi and the maximum of 7 fc′ = 443 psi when a shearhead
is used, shearhead reinforcement for the interior column (having no moment transfer
to column) is permitted to be designed according to ACI-22.6.9.
(b) Length of shearhead. The length of shearhead should be such that the nominal shear
stress is less than 4 fc′, computed around a periphery passing through points at
3 ( L − c / 2) from, but no closer than d/2 to the column faces. Assuming a square
4 v 1
as the critical periphery, since the shearhead is to have four identical arms (ACI-
22.6.9.1), the required b0 (ft) may be computed from Fig. 16.16.6 so that
440 270 − (b0 / 4)2   1 

4 4000 ≤  12 
0.75(b0 ) 6.25
(Continued)
683
b 3 L – c1
(
4 v 2
)
/4
0
b
3.13”
Lv
3.13”
a
c o
45°
c2 = 10”
c1 = 12”
Lv
Figure 16.16.6 Required length of shearhead in Example 16.16.2.
Neglecting the (b0 /4)2 in the numerator,

440(270)  1 
b0 =   = 8.3 ft (100 in.)
253(0.75)6.25  12 
Note that because shearhead reinforcement is used instead of stirrups, slab reinforcement
could have been placed closer to the top and bottom of the slab thus resulting in a slightly
larger average effective depth d and a somewhat smaller b0. Nonetheless, the same d com-
puted for the slab with stirrups in Example 16.16.1 will be used in this example.
The required distance Lv may be computed from the following geometric consider-
ations. From right triangle oab,
3  c c b0
 4  Lv − 2  + 2  2 = 4
 
which gives, based on leg ob, c = c2 = 10 in.,
 100 4
Lv =  − 5 + 5 = 21.9 in.
4 2 3
and, based on leg oa, c = c1 = 12 in.,
 100 4
Lv =  − 6 + 6 = 21.5 in.
4 2 3
For the periphery abcd not to approach closer than d/2 (3.13 in.) to the periphery of the
column section,

oa = ob = (6 + 3.13) + (5 + 3.13) = 17.26 in.
But
3 3
oa = 6 + ( Lv − 6) and ob = 5 + ( Lv − 5)
4 4
(Continued)
684
which gives
4
Lv = (17.26 − 6) + 6 = 21.0 in.
3
4
Lv = (17.26 − 5) + 5 = 21.3 in.
3
Use Lv = 22 in.
nAs = 8(0.84) = 6.7 sq in. #5 bars
Column strip width = 90”
c = 2.23″
12 + d
avg d = 6.25”
A
10 + d nAs = 8(2.79)
= 22.3 sq in. 4”
A 0.75″
c2 + d = 16.25”
Section A–A
Column strip width = 90”
Figure 16.16.7 Cracked slab section of width (c2 + d) in Example 16.16.2.
(c) Size of shearhead. The shearhead stiffness must be at least 0.15 that of the composite
cracked slab section of width (c2 + d). It can be shown that 15–#5 bars and 13–#5 bars
are required for negative slab reinforcement in the 90-in.-wide column strips of the
long and short directions, respectively. The composite cracked section across width
A-A in Fig. 16.16.7 should be used because there is more steel in the slab in the long
direction. The steel area As at the top of the slab in section A-A, of width c2 + d is
10 + d 16.25
As = (15)(0.31) = (15)(0.31) = 0.84 sq in.
90 90
with 3 4 -in. cover at the top and bottom face of slab, and assuming #5 bars for top and
bottom reinforcement in the two orthogonal directions, the clear distance between the
top and bottom layers of reinforcement is 4 in. Assume an S4 × 9.5 section for the shear-
head placed as shown in Fig. 16.16.7. Note that at least two bottom bars are required
to pass within the region bounded by the longitudinal reinforcement of the column to
satisfy the structural integrity requirements of ACI-8.7.4.2. In addition, ACI-22.6.9.4
requires that the compression flange of the steel shape be within 0.3d of the com-
pression surface (i.e., the bottom of the slab in this case). For this reason, the S4 × 9.5
section is placed closer to the bottom of the slab with the same cover of the reinforcing
bars. Because clearance under the shearhead is insufficient for the bottom bars to pass
in the transverse direction, they will be made to pass through holes in the shearhead
arms, as permitted by ACI-8.7.4.2.3.
The ratio between the depth and the web thickness for this section is 4.00/0.326 = 12.3,
which is much less than 70 in accordance with ACI-22.6.9.2.
(Continued)
685
The centroidal axis of the composite cracked section may be obtained by equating the
static moments of the compression-and tension-transformed areas (ignoring the bottom
steel bars),
16.25c 2
= 22.3 (2.75 − c) + 6.7 (6.56 − c)
2
c = 2.23 in.
16.25(2.23)3
composite I s = + n ( I x of steel section ) + 22.3(0.52)2 + 6.7(4.71)2
3 OK
= 60.1 + 8(6.76) + 6.03 + 148.6 = 269 in.4
Es (6.76) 8(6.76)
provided α v = = = 0.20 > 0.15
Ec (composite I s ) 269
The plastic section modulus of the S4 × 9.5 is given by the AISC Manual [16.147] as
4.04 cu in. Using A992 Grade 50 steel, the provided Mp is

provided Mp = 50 ( 4.04 ) = 202 kip-in.
The required Mp is computed from Eq. (16.16.3) as
Vu   c1  
required Mp = hv + α v  required Lv −  
8φ  2 
0.440(270)

=
8(0.90)
[ 4 + 0.20(21.9 − 5)]
= 122 kip-in. < 202 kip-in. OK
The section has about 66% more strength than required. Perhaps a smaller section, such
as an S4 × 7.7, could be used.
(d) Shearhead contribution to resisting negative moment in slab. It can be shown that the
negative moments at the face of the column in the 90-in. column strip width in the
long and short directions are
column strip moment in the long direction = 125 ft-kips
column strip moment in the short direction = 102 ft-kips
The resisting moment of the shearhead may be computed from Eq. (16.16.4),
φ α vVu  c1 
Mc =  Lv − 2 
2η
0.90(0.20)(0.440)(270)
=
8
[(22 − 6) or (22 − 5) ] 121

= 3.6 or 3.8 ft-kips
Thus the contribution is rather small, and a revision of slab reinforcement is unnecessary.
68
16.17 DIRECT DESIGN METHOD—M OMENTS

IN COLUMNS
The moments in columns due to unequal loads on adjacent panels are readily available
when an elastic analysis is performed on the equivalent frame for the various pattern load-
ings. In the Direct Design Method, wherein the limitations listed in Section 16.7 are satis-
fied, the longitudinal moments in the slab are prescribed by the provisions of ACI-8.10.4.
In a similar manner, the code prescribes the moment at an interior column as follows [ACI
Formula (8.10.7.2)]:
 1  
M sc = 0.07  wDu + wLu  L2 L2n − wDu
′ L2′ ( Ln′ )2  (16.17.1)
 2  
where
wDu = factored dead load per unit area
wLu = factored live load per unit area
′ , L2′ , Ln′ = quantities referring to shorter span
wDu
The moment is yet to be distributed between the two ends of the upper and lower columns
meeting at the joint.
The rationale for Eq. (16.17.1) may be observed from the stiffness ratios at a typical
interior joint shown in Fig. 16.17.1(a), wherein the distribution factor for the sum of the
column end moments is taken as 7 8 and the unbalanced moment in the column strip is taken
to be 0.080/0.125 times the difference in the total static moments due to the dead load plus
half the live load on the longer span and the dead load only on the shorter span.
For an edge column, ACI-8.10.7.3 requires using 0.3M0 as the moment to be transferred
between the slab and an edge column.
Col. 1
Col. 1
(wDu + 1 wLu) L2 (wDu + 1 wLu) L2

w’DuL’2 2 2
Span 2 Span 1 Span 1

Col. 2
Col. 2
(a) Interior joint (b) Exterior joint
Figure 16.17.1 Direct design method—moments in columns.
EXAMPLE 16.17.1
Obtain the factored moments in the interior and exterior columns in each direction for
the flat plate of Example 16.3.3.
SOLUTION
(a) Exterior column, long direction (Frame A). The factored moment Msc to be trans-
ferred to the exterior column is (ACI-8.10.7.3) 0.3M0,

M sc = 0.3 M 0 = 0.3(58.2) = 17.5 ft-kips
(Continued)
687
16.18 TRANSFER OF MOMENT AND SHEAR 687
where the 58.2 ft-kips was obtained from Table 16.10.2. The moment Msc is to be
divided between upper and lower columns in proportion to their stiffnesses (in this case,
equally).
In this example, nearly all (98.8% for Frame A and 99.3% for Frame C–see Table
16.12.6) of the exterior frame moment is taken by the column strip; the arbitrary use of
0.3M0 to be taken by the column strip seems appropriate.
(b) Interior column, long direction. The factored moment to be transferred to the col-
umn is empirically the amount obtained from ACI Formula (8.10.7.2) [Eq. 16.17.1],
M sc = 0.07(0.083 + 0.058)(12)(15 − 1)2 − 0.083(12)(15 − 1)2

= 0.07(0.058)(12)(14)2 = 9.5 ft-kips
The moment Msc is to be divided between upper and lower columns.

(c) Exterior column, short direction (Frame C). The factored moment to be transferred is

M sc = 0.3 M 0 = 0.3(46.3) = 13.9 ft-kips
where the 46.3 ft-kips was obtained from Table 16.10.2. The moment Msc is to be divided
between upper and lower columns.
(d) Interior column, short direction. The factored moment to be transferred is

M sc = 0.07(0.058)(15)(12 − 0.83)2 = 7.6 ft-kips
The moment Msc is to be divided between upper and lower columns.
16.18 TRANSFER OF MOMENT AND SHEAR AT

JUNCTION OF SLAB AND COLUMN
Inasmuch as the columns meet the slab at monolithic joints, there must be moment as
well as shear transfer between the slab and the column ends. The moments may arise
out of lateral loads due to wind or earthquake effects acting on the multistory frame, or
they may be due to unequal gravity loads, as considered in Section 16.17. The trans-
fer of moment and shear at the slab-column interface is extremely important in the
design of flat plates and has been the subject of numerous research studies [16.103–
16.138, 16.143]. The current provisions are largely based on the ACI-ASCE Committee
352 report, “Recommendations for Design of Slab-Column Connections in Monolithic
Reinforced Concrete Structures” [16.128], and the background explanation by Moehle,
Kreger, and Leon [16.127].
Let Msc be the total slab factored moment that is to be transferred to the ends of the
upper and lower columns meeting at an exterior or an interior joint. Test results by Hanson
and Hanson [16.104] have shown that about 60% of the moment is transferred by flexure
and the remainder by eccentricity of the shear around the critical periphery located at d/2
from the column faces. The ACI Code requires that the fraction of the total slab factored
68
moment resisted by the column at a joint, Msc, be transferred by flexure and computed as
(ACI-8.4.2.3.2)
 
 
1
Mu f = γ f M sc =  M (16.18.1)
 2 b1 
sc
1+ 
 3 b2 
where
Mu f = fraction of total slab factored moment transferred by flexure
b1 = critical section dimension in the direction of the span for which moments are
determined
= c1 + d/2 for exterior columns [Fig. 16.18.1(a)]
= c1 + d for interior columns [Fig. 16.18.1(b)]
b2 = critical section dimension in the direction perpendicular to b1
= c2 + d (Fig. 16.18.1)
The remaining portion of the total slab factored moment is assumed to be transferred by
eccentricity of shear and is computed as (ACI-8.4.4.2.2)
Muv = γ v M sc = (1 − γ f ) M sc (16.18.2)
The moment Mu f is considered to be resisted within an effective slab width equal to
(c2 + 3t) at the column (ACI-8.4.2.3.3), where t is the slab or drop panel thickness. The
moment strength for Mu f is achieved by using additional reinforcement and closer spacing
within the width (c2 + 3t) [ACI-8.4.2.3.5].
b1 = c1 + d b1 = c1 + d
2
z z
Column
c1
b2 = c2 + d
b2 = c2 + d
c2
z z
b1 Vu b1
Ac
Stresses
due to Vu
b2
b2
Muv x1 x2 x1
Muv x1 x2
Jc
x1 Jc
Stresses
Muv x2 due to Muv
Jc Muv x2
Jc
vu1
vu1 vu2
vu2
Total
stresses
(a) Exterior column (b) Interior column
Figure 16.18.1 Transfer of moments to columns through shear stresses.

689
Note that if b2 = b1, Eq. (16.18.1) becomes
Mu f = 0.60 M sc
and that if b2 = 1.5b1, Eq. (16.18.1) becomes

Mu f = 0.648 M sc
It appears reasonable that when b2 is larger than b1, the moment transferred by flexure is
greater because the effective slab width (c2 + 3t) resisting the moment is larger.
Analysis of the available test data [16.126, 16.127] has shown that the interaction
between flexure and shear is reduced when the ratio between the shear caused by gravity
loads and the slab shear strength (in the absence of moment transfer) is relatively low. In
other words, a greater proportion of the total slab factored moment Msc may be assumed to
be transferred by flexure; i.e., a larger value of γf may be used in Eq. (16.18.1). Accordingly,
ACI-8.4.2.3.4 allows γf to be increased when the applied-to-strength shear ratio does not
exceed a threshold value that depends on the connection type (corner, edge, or interior), and
when the slab flexural reinforcement provided within (c2 + 3t) satisfies a minimum value
for the net tensile strain. When the conditions set forth in ACI-8.4.2.3.4 are satisfied, flex-
ure and shear interaction may be neglected in some cases, which can greatly simplify slab
design. These conditions and the allowed increase in the values of γf are described next.
Modifications to γf (ACI-8.4.2.3.4)
Corner Columns
At a corner support, ACI-8.4.2.3.4 permits neglect of the interaction between shear and
moment when the factored shear due to gravity loads, Vug, acting on the slab critical section
does not exceed 0.5φVc. In other words, for such situations, the full slab moment can be
considered to be transferred by flexure (i.e., γf = 1.0), and the factored shear Vu can be con-
sidered independently. This provision applies to moments being transferred in either direc-
tion, and is permitted only when the slab flexural reinforcement required to resist Msc within
the effective slab width (c2 + 3t) has a net tensile strain, εt, greater than or equal to 0.004.
Edge Columns
For moments about an axis perpendicular to the edge, the full slab moment can also be
considered to be transferred by flexure alone (i.e., γf = 1.0) when the factored shear due to
gravity loads, Vug, acting on the slab critical section does not exceed 0.75φVc. As for corner
columns, this is permitted only when the net tensile strain, εt, in the slab flexural reinforce-
ment provided within the effective slab width (c2 + 3t) is greater than or equal to 0.004.
On the other hand, for moments about an axis parallel to the edge, ACI-8.4.2.3.4 per-
mits increasing by as much as 25% the proportion γf of the full exterior moment transferred
by flexure when the factored shear due to gravity loads, Vug, acting on the slab critical sec-
tion does not exceed 0.4φVc. The 25% increase in γf , however, is permitted only when the
net tensile strain, εt, in the slab flexural reinforcement provided within the effective slab
width (c2 + 3t) is greater or equal to 0.01.
Interior Columns
For interior columns, a 25% increase in γf is permitted for moments in either direction when
factored shear due to gravity loads, Vug, acting on the slab critical section does not exceed
0.4 φVc, provided the net tensile strain, εt, in the slab flexural reinforcement within the effec-
tive slab width (c2 + 3t) is greater than or equal to 0.01.
The modifications to γf described above are not permitted for prestressed concrete systems.
Computation of Shear Stresses due to Flexure and Shear Interaction (γv > 0)—
Eccentric Shear Stress Model
When part of the slab factored moment is assumed to be transferred by eccentricity of
shear; i.e., when γv > 0, the shear stresses due to Muv must be combined with those induced
690
by the factored shear force Vu. The procedure to compute shear stresses has been tradition-
ally based on the eccentric shear stress model [16.103, 16.104], where shear stresses are
assumed to vary linearly about the centroid of the critical section (located at d/2 from the
column faces), as shown in Fig. 16.18.1. Based on these assumptions, the minimum and
maximum values of the factored shear stresses acting along the faces of the critical section
may be computed as (see Fig. 16.18.1)
Vu Muv x1
vu1 = − (16.18.3)
Ac Jc
Vu Muv x2
vu 2 = + (16.18.4)
Ac Jc
By using a section property Jc analogous to the polar moment of inertia of the shear area
along the critical periphery taken about the z-z axis, it is assumed that there are both hori-
zontal and vertical shear stresses on the shear areas having dimensions b1 by d as shown in
Fig. 16.18.2. The z-z axis is perpendicular to the longitudinal axis of the equivalent frame:
that is, in the transverse direction, and located at the centroid of the shear area.
For an exterior column, x1 and x2 are obtained by locating the centroid of the channel-
shaped vertical shear area represented by the dashed line (b1 + b2 + b1) shown in
Fig. 16.18.1(a), and
Ac = (2b1 + b2 )d (16.18.5)
b12 d
x2 = (16.18.6)
Ac
 b3  b d3
J c = d 2 1 − (2b1 + b2 ) x22  + 1 (16.18.7)
 3  6
For an interior column, referring to Fig. 16.18.1(b),
Ac = (2b1 + b2 )d (16.18.8)
 b3 b b 2  b d 3
Jc = d  1 + 2 1  + 1 (16.18.9)
6 2  6
Equations (16.18.5) to (16.18.9) may be derived by letting the shear stress at any loca-
tion resulting from Muv alone be proportional to the distance from the centroidal axis z-z to
x1 z x2 z
Muv x1
Jc
Muv
Muv x2 vuh
Jc
d
Muv ( 2 )
Jc
z z
(a) Shear areas (b) Vertical resisting (c) Horizontal resisting
to resist Muv shear stresses shear stresses
Figure 16.18.2 Resisting shear stresses due to Muv acting on an exterior column.

691
the shear areas b1 by d, and either (a) to the one shear area b2 by d for an exterior column as
shown in Fig. 16.18.2, or (b) to the two shear areas b2 by d for the interior column.
For slabs without shear reinforcement, the larger factored shear stress vu2 shown in
Fig. 16.18.1 must not exceed the stress φ vn = φVc /b0d obtained from Eqs. (16.15.2); other-
wise shear reinforcement is required.
EXAMPLE 16.18.1
For the flat plate design of Example 16.3.3, investigate the transfer of gravity load
moments in the long direction, as already computed in Example 16.17.1, to the exterior
and interior columns, respectively.
SOLUTION
(a) Exterior column (long direction). Investigate whether the full slab moment may be
considered to be transferred by flexure alone; i.e., whether γf is permitted to be taken
as 1.0 in accordance with ACI-8.4.2.3.4. From Example 16.17.1, the moment to be
transferred is

Mu = M sc = 17.5 ft-kips
The factored shear due to gravity loads, Vug, is taken as wu times the floor area, 12 ft ×
7.5 ft, tributary to the exterior column.

Vug = 0.198(12)7.5 = 17.8 kips
Using the dimensions shown in Fig. 16.15.4 as a reference, b0 for an exterior (edge)
column is
b0 = 2(12 + 2.125) + 14.25 = 42.5 in.
Using the average effective depth d = 4.25 in. for #4 slab reinforcement, the nominal
shear strength Vc in accordance with ACI-22.6.5.2 is the smallest of
Vc = 4 λ fc′b0 d Controls
 4  4  
V = 2 +  λ fc′b0 d =  2 + λ fc′b0 d  = 5.3λ fc′b0 d
c  β  
12 /10  
 α d  30(4.25)  
Vc =  2 + s  λ fc′b0 d =  2 +  λ fc′b0 d  = 5.0 λ fc′b0 d
 b0   42.5  
Thus,
 1 
Vc = 4 λ fc′b0 d = 4(1.0) 4000 (42.5)4.25  = 45.7 kips
 1000 
According to ACI-8.4.2.3.4, for an interior (or edge) column the full slab moment may
be considered to be transferred by flexure alone when

[0.75φ Vc = 0.75 ( 0.75)( 45.7) = 0.75 (34.3) = 25.7 kips ] > [ Vug = 17.8 kips]
Thus, the interaction between shear and flexure may be neglected (i.e., γf = 1.0), inso-
far as the reinforcement required to resist γf Msc within the effective slab width has a net
tensile strain, εt, greater than or equal to 0.004. The effective slab width is equal to the
column width plus three times the slab thickness (ACI-8.4.2.3.3); that is, 26.5 in. Using
the average effective depth d = 4.25 in. for #4 slab reinforcement,
M sc 17.5(12, 000)
Rn = = = 487 psi
φ bd 2 0.9(26.5)(4.25)2
(Continued)
692
From Eq. (3.8.5), with fc′ = 4000 psi and fy = 60,000 psi, the required ρ ≈ 0.008, which is
much less than the maximum reinforcement ratio corresponding to a net tensile strain εt of
0.004. Thus, the net tensile strains in the tension steel will be larger than 0.004. Therefore,
the full slab factored moment can be considered to be transferred by flexure alone and

Mu f = γ f M sc = 1.0(17.5) = 17.5 ft-kips
The shear can be considered independently.
(b) Exterior column (long direction). Check moment transfer using the eccentric shear
stress model [Eqs. (16.18.3) and (16.18.4)]. This procedure involves more calcula-
tions and is more conservative.
(i) Transfer by flexure. From Eq. (16.18.1), using the average effective depth d =
4.25 in. for #4 slab reinforcement,
 
 
1
Mu f = γ f M sc =  M
 2 b1 
sc
1+ 
 3 b2 

 
 1 
=  17.5 = 0.601(17.5) = 10.5 ft-kips
 2 12 + 2.125 
 1 + 
3 10 + 4.25 
As shown by Fig. 16.18.3, this moment is to be carried over a slab width equal to the
column width plus 3 times the slab thickness, that is, 26.5 in. From Table 16.12.6 and
Fig. 16.10.3 (Frame A), the total moment in the 72-in.-wide column strip is

M in column strip = 0.988 (15.1) = 15 ft-kips
If the slab reinforcement is placed at equal spacing in the column strip, additional rein-
forcement is needed in the 26.5-in. width for a bending moment of
 26.5 
Mu f − 15  = 10.5 − 5.5 = 5 ft-kips
 72 
(ii) Transfer by eccentricity of shear. From part (a), the factored shear force is
Vu = Vug = 0.198(12)7.5 = 17.8 kips
and from Eq. (16.18.2)

Muv = M sc − Muf = 17.5 − 10.5 = 7 ft-kip
ps
From Fig. 16.18.3,
2(14.12)7.06
x2 = = 4.7 in.
28.24 + 14.25
Ac = 4.25(28.24 + 14.25) = 181 sq in.
 2(14.12)3  14.12(4.25)3
J c = 4.25  − 42.49(4.7)2  +
 3  6

= 3990 + 181 = 4171 in.4
17, 800 7000(12)9.42
vu1 = − = 98 − 190 = −92 psi
181 4171
17, 800 7000(12)4.70
vu 2 = + = 98 + 95 = +193 psi
181 4171 (Continued)
693
d
12 + = 14.12”
2
10 + 3t = 26.5” 10 + d = 14.25”
x2
x1
t = slab or drop panel thickness
Vu
Muv
Muf
Figure 16.18.3 Transfer of moments at exterior column in Example 16.18.1.
The nominal stress limit based on strength in shear was determined in part (a) to be that
based on 4 λ fc′. Thus, the limit to the above stresses is

limit vu = φ vc = φ (4 λ fc′) = 0.75(253) = 190 psi ≈ [vu2 = 192 psi]
The required and design shear strengths are very close; the slab may be considered ade-
quate based on the shear-flexure interaction procedure. Otherwise, shear reinforcement
could be provided.
The horizontal shear stress vuh at the upper or lower edge of the two shear areas b1
by d is
7000(12)(4.25/ 2)
vuh = = 43 psi
4171
The vuh of 43 psi, vu1 of –91 psi, and vu2 of +192 psi may be drawn on a sketch like that of
Fig. 16.18.2, and by basic statics computation of the resultant upward force should equal
Vu of 17.8 kips and the resultant moment about the z-z axis should equal Muv of 7 ft-kips.
(c) Interior column (long direction). Investigate whether an increase in γf would be per-
mitted in accordance with ACI-8.4.2.3.4. The factored shear force due to gravity
loads Vug is computed as wu times the tributary floor area of 12 × 15 ft,

Vug = 0.198(12)15 = 35.6 kips
Using Eqs. (16.15.2), as shown in part (a) for the exterior column, will indicate that the
shear strength based on 4 λ fc′ controls. Since the interior and exterior columns have the
same size, Eq. (16.15.2b), involving the aspect ratio β, gives the same value as in part (a),
that is, 5.3λ fc′. Regarding Eq. (16.15.2c), αs is taken as 40 for interior columns, and

b0 = 2(16.25) + 2(14.25) = 61.0 in.
Thus, Eq. (16.15.2c) gives
 α d  40(4.25)  
Vc =  2 + s  λ fc′b0 d =  2 +  λ fc′b0 d  = 4.8λ fc′b0 d
 b0   61.0  
(Continued)
694
The strength Vc cannot exceed that based on 4 λ fc′ from Eq. (16.15.2a). Thus,
1
Vc = 4 λ fc′b0 d = 4(1.0) 4000 (61.0)(4.25) = 65.6 kips
1000
For interior columns, a 25% increase in the value of γf may be permitted when

[0.4φVc = 0.4(0.75) (65.6) = 0.4(49.2) = 19.7 kips] < [Vug = 35.6 kipss]
Since the factored shear Vug is not less than 0.4φVc, the increase in γf is not permitted!
(Note that for interior columns, γf may be increased only if the net tensile strain in the
slab reinforcement provided within (c2 + 3t) is greater than or equal to 0.01, even if 0.4
φVc were to be greater than Vug).
(i) Transfer by flexure. From part (b) of Example 16.17.1, the factored moment to
be resisted by the column is
Mu = M sc = 9.5 ft-kips
and the fraction to be transferred by fleexure is
 
 
1

 2 b1 
sc
1+ 
 3 b2 
 
 1 
=  9.5 = 0.584(9.5) = 5.5 ft-kips
 2 12 + 4.25 
 1 + 
3 10 + 4.25 
From Table 16.12.6 and Fig. 16.10.3, the total moment in the 72-in.-wide column strip is

M in column strip = 0.75(37.8) = 28.4 ft-kips
Assuming that the slab reinforcement is placed at equal spacing within the column strip,
the moment resisted within the 26.5-in. width is 26.5(28.4)/ 72 = 10.5 ft-kips, which is
larger than 5.5 ft-kips; no additional reinforcement is needed.
(ii) Transfer by eccentricity of shear. The factored shear Vu is 35.6 kips. From
Eq. (16.18.2)

Muv = M sc − Muf = 9.5 − 5.5 = 4.0 ft-kips
From Fig. 16.18.1(b),
Ac = 4.25(32.50 + 28.50) = 259 sq in.
 (16.25)3 14.25(16.25)2  16.25(4.25)3

J c = 4.25  + +
 6 2  6
= 11, 040 + 210 = 11, 250 in.4

35, 600 4000(12)8.12
vu1 = − = 137 − 35 = −102 psi
259 11, 250
35, 600 4000(12)8.12
vu 2 = + = 137 + 35 = +172 psi
259 11, 250
( )
both of which are less than the design strength φ vn = φ vc = φ 4 λ fc′ = 0.75(253) =
190 psi when no shear reinforcement is provided.
(Continued)
695
Again, by basic statics the sum of the factored load vertical shear stresses on the two
areas b1 × d plus that on the two areas b2 × d should total Vu of 35.6 kips. Likewise, the
moment of these shear stress resultants along with that of the horizontal shear stresses
on the two faces b1 × d should add up to 4.0 ft-kips. The horizontal shear stress at the
upper or lower edge of the faces b1 × d is
4000(12)(4.25 / 2)
vuh = = 9 psi
11, 250
Moment Transfer from Flat Plate to Column When Shearheads Are Used

Tests [16.93, 16.110] have indicated that shear stresses computed for factored loads at
the critical section distance d/2 from the column face are appropriate for transfer of Muv =
Mu – Muf as described above, even when shearheads are used. However, the critical section
for Vu is at a periphery passing through points at 3 4 ( Lv − c1 / 2) from, but no closer than,
d/2 to the column faces. When both Vu and Mu are to be transferred, ACI-22.6.9.12
requires that the sum of the shear stresses due to Muv and due to vertical load Vu com-
puted at their respective critical sections not exceed φ (4 λ fc′). The reason for this apparent
inconsistency (ACI Commentary R22.6.9.11) is that these two critical sections are in close
proximity at the column corners, where the failures initiate.
EXAMPLE 16.18.2
Compute the moment at an interior column of the flat plate with shearheads of
Example 16.16.2 using Eq. (16.17.1), ACI Formula (8.10.7.2). Also, compute the shear
stresses due to moment transferred by eccentricity of shear.
SOLUTION
(a) Referring to Example 16.16.1, part (a) of solution,

wu = 440 psf
Referring to Example 16.16.2, part (d) of solution,
column strip moment in the long direction = 125 ft-kips
Using Eq. (16.17.1), the moment to be transferred to the column is,
M sc = 0.07 (0.120 + 0.320)(15)(18 − 1)2 − 0.120(15)(18 − 1)2 

= 0.07(0.320)(15)(17)2 = 97.1 ft-kips
The effective slab width is

c2 + 3t = 10 + 3(8) = 34 in.
Column strip moment in 34-in. width of Fig. 16.18.4 is
 34 
125   = 47.2 ft-kips
 90 
(Continued)
69
Additional reinforcement is needed to carry (97.1 –47.2) = 49.9 ft-kips—say, 50 ft-kips

within the 34-in. width unless 97.1/125 = 78% of the total column strip reinforcement is
concentrated in the 34-in. width.
(b) Compute factored load shear stress at critical section of Fig. 16.18.4 due to Mu v only.
The moment to be transferred by flexure is
 
 
1
 2 b1 
sc
1+ 
 3 b2 
 
 1 
=  97.1
 2 12 + 6.25 
 1 + 
3 10 + 6.25 
Mu f = 0.586 M sc = 0.586 ( 97.1) = 56.9 ft-kips
and

Mu v = 97.1 − 56.9 = 40.2 ft-kips
The maximum shear stress due to Muv (see Fig. 16.8.1) computed about the centroid of
the critical section at d/2 (ACI-22.6.4.1) is given by (see Fig. 16.8.1)
Muv x1
vu =
Jc
The critical section properties for an interior column are

Ac = 2(b1 + b2 )d [16.18.8]
Thus,
Ac = 2(18.25 + 16.25)6.25

= 431.3 sq in.
and
 b3 b b 2  b d 3
Jc = d  1 + 2 1  + 1 [16.18.9]
6 2  6
 (18.25)3 16.25(18.25)2  18.25(6.25)3

J c = 6.25  + +
 6 2  6
= 24, 000 in.4
Thus, the maximum shear stress due to Muv is
40.2(9.13)(12, 000)
vu = = 184 psi
24, 000
(Continued)
697
1 6 . 1 9 O P E N I N G S A N D C O R N E R C O N N E C T I O N S I N F L AT S L A B S 697
These stresses alone are very high and by adding the Vu /Ac component they will exceed
the maximum permitted of φ (4 λ fc′) = 190 psi. Therefore, the slab thickness should be
increased around the column: that is, a drop panel would be required.
12 + d = 18.25”
10 + 3t = 34” 10 + d = 16.25”
Vu
Muv
Muf
Figure 16.18.4 Transfer of moments in the long direction at interior column in Example 16.18.2.
16.19 OPENINGS AND CORNER CONNECTIONS

IN FLAT SLABS
When openings and corner connections are present in floor slabs, designers must make sure
that adequate provisions are made for them. ASCE-ACI Joint Task Committee 426 [16.83]
has summarized available information. Tests by Roll, Zaidi, Sabnis, and Chuang [16.79]
have provided additional data for treating openings, while Zaghlool and de Paiva [16.107,
16.108] have provided data for corner connections.
ACI-8.5.4.1 first prescribes in general that openings of any size may be provided if it
can be shown by analysis that all strength and serviceability conditions, including the limits
on the deflections, are satisfied. In common situations (ACI-8.5.4.2), however, no special
analysis need be made for slab systems not having beams when (1) openings are within
the middle half of the span in each direction, provided the total amount of reinforcement
required for the panel without the opening is maintained; (2) openings in the area com-
mon to two column strips do not interrupt more than one-eighth of the column strip width
in either span, and the equivalent of reinforcement interrupted is added on all sides of the
openings; and (3) openings in the area common to one column strip and one middle strip do
not interrupt more than one-fourth of the reinforcement in either strip, and the equivalent of
reinforcement interrupted is added on all sides of the openings.
In regard to shear strength in two-way action, when the opening is within a column strip
or closer than 10 times the slab thickness from a concentrated load or reaction area [ACI-
8.5.4.2(d)], the critical section for slabs without shearheads must not include the part of the
periphery that is enclosed by radial projections of the openings to the center of the column (ACI-
22.6.4.3). For slabs with shearheads, the critical periphery is to be reduced only by one-half of
what is cut away by the radial lines from the center of the column to the edges of the opening
(ACI-22.6.9.9). Some critical sections with cutaways by openings are shown in Fig. 16.19.1.
698
Opening
Ineffective
d (typical)
2
Critical section
(a) (b)
Free corner
Regard as
free edge
(c) (d)
Figure 16.19.1 Effect of openings and free edges on critical perimeter of two-way shear action.
(From ACI Commentary R 22.6.4.3.)
16.20 EQUIVALENT FRAME METHOD FOR GRAVITY

LOAD ANALYSIS
The design of two-way floor systems under gravity loads has been presented in earlier
sections of this chapter, wherein the longitudinal distribution of moments in the equivalent
frame3 follows the coefficients of the ACI Code, within the limitations set forth in the Direct
Design Method (DDM). When the limitations of the DDM are not met, an elastic analysis
of the equivalent frame must be made for various gravity load patterns to obtain the longi-
tudinal moment and shear envelopes. For lateral loads, however, an elastic analysis of the
equivalent frame must be made regardless of whether the system is within the limitations
of the DDM for gravity loads. According to ACI-8.2.1, “A slab system shall be designed by
any procedure satisfying equilibrium and geometric compatibility, provided that the design
strength at every section is at least equal to the required strength, and all serviceability
requirements are met.”
For gravity load analysis, ACI-8.11.2.5 permits a separate analysis of each floor, with
the far ends of columns considered fixed. The elevation and plan of a subassembly contain-
ing a floor with the attached upper and lower columns in a two-way floor system are shown
in Fig. 16.20.1(a) and 16.20.1(b). The subassembly enclosed between the two parallel cen-
terlines of two adjacent panels in a multistory two-way floor system is a three-dimensional
structure that includes the slab, beams (longitudinal and transverse), columns, drop panels,
and column capitals, if present.
A difficult problem in the analysis of the subassembly shown in Fig. 16.20.1 is modeling
of the transfer of the slab moment to the columns. In the Equivalent Frame Method (EFM),
the three-dimensional subassembly is idealized by a series of two-dimensional frames
using equivalent columns and slab-beams.
For gravity load analysis the EFM prescribed by the ACI Code differs from the DDM
only in the way by which the longitudinal moments along the spans of the equivalent frame
(as defined in Section 16.2) are obtained. In either method of analysis, the transverse dis-
tribution of the longitudinal moments may be carried out as described in Section 16.12 as
3 Note that the concept of “equivalent frame” is applicable in both the direct design method and the equivalent frame method.
69
1 6 . 2 0 E QU I VA L E N T F R A M E M E T H O D 699
long as, in two-way slabs supported on beams, the beams are sufficiently stiff and satisfy
limitation No. 6 of the direct design method discussed in Section 16.7 per ACI-8.11.6.6.
It is noted that when the EFM is used for gravity load analysis of two-way floor systems
that meet the limitations of the DDM, the computed moments may be reduced such that
the absolute sum of the positive and average negative moments is at least equal to the total
static moment wu L2 L2n /8 (ACI-8.11.6.5). This provision exists because the “Code should
not require a value greater than the least acceptable value” (ACI-R8.11.6.5); in other words,
if the values obtained from the DDM are acceptable, there is no reason to design the slab
for larger moments irrespective of the method used in the analyses.
At this point, axial deformation of the columns, reduction of column stiffness due to
heavy gravity load, and increase of column moments and shears due to P-Δ effects (see
Chapter 13) are neglected to focus attention on the characteristics of the equivalent rigid
frame in a two-way floor system.
G H K
A B C
D E F
(a) Elevation
CL of panel
1
Middle strip
Width of equivalent frame
2
Column strip
1
Middle strip
2
CL of panel
(b) Plan
G H K
columns columns
torsional torsional
A elements B element C
A’ slab-beam B’ slab-beam C’
D E F
(c) Frame model
Figure 16.20.1 Equivalent frame of two-way slab subassembly for gravity load analysis.
70
Section Properties
The subsections that follow describe the procedure for computing the section properties
of the equivalent columns and beams. These properties are computed based on extensive
analytical and experimental test data on two-way slabs under gravity loads. The EFM and
thus computation of section properties of the equivalent columns and beams, were origi-
nally developed to be used with the method of moment distribution. The equivalent frame
method, however, may also be used with virtually any standard analysis software, as dis-
cussed in Section 16.21.
Flexural Stiffness of Slab-Beams, Ksb

Consideration of the flexural stiffness of the slab-beams, Ksb, must incorporate the stiffness
of the slab, including any drop panels and capitals around columns, as well as the stiffness
provided by any beams spanning in the direction of the frame. As a result, the equivalent
beams will have a variable moment of inertia along the beam axis; that is, they will be
nonprismatic.
Figure 16.20.2 shows cross sections for computing the flexural stiffness of the slab-
beams for various slab systems. As an example, consider the slab system with column cap-
itals shown in Fig. 16.20.2(c). The moment of inertia of the slab-beam between the edges
of the drop panels, I1, is computed assuming a rectangular section of width, L2, and depth
equal to the slab thickness, h1, as shown by Section A-A in Fig. 16.20.2(c). Where a drop
panel exists, the moment of inertia, I2, is computed for the T-shaped section (Section B-B)
shown in Fig. 16.20.2(c). The moment of inertia from the center of the column to the face
of the capital is assumed to be equal to that of the slab-beam at the face of the capital, I2,
divided by the quantity (1 –c2 /L2)2, where c2 is the width of the equivalent square column
capital (ACI-8.11.3.1). For other slab systems, c2 is the width of the support measured
transverse to the direction of the span in which the moments are being computed. The coef-
ficient 1/(1 – c2 /L2)2 accounts for the increased stiffness of the slab-beam over the support
in a manner that is consistent with the results of three-dimensional slab analyses as well as
with test results [16.27].
The variation of moment of inertia along the axis of the slab-beams is taken into account
(as required by ACI-8.11.3.2) as shown by the equivalent slab-beam stiffness diagram
shown for each slab system in Fig. 16.20.2. Note that ACI-8.11.3.3 permits the use of the
gross concrete area for computation of the moment of inertia of slab-beams in gravity load
analysis. For lateral load analyses, the effects of cracking and reinforcement must be taken
into account, as discussed later in Section 16.22.
Stiffness of Equivalent Columns, Kec

Consider the moment transfer mechanism from the slab to the edge column shown in
Fig. 16.20.3. It may be easily visualized that some of the moment will be transferred by
flexure from the slab directly to the column, while the remainder will be transferred first to
the transverse beams and then to the column by twisting of the transverse beams framing
into the column. Note that a similar transfer mechanism would occur for slab systems with-
out beams: that is, some of the moment will be transferred to the column by twisting of the
slab in the transverse direction.
Stiffness of Torsional Members

To account for the moment transferred by twisting of the transverse elements, Corley,
Sozen, and Siess [16.26] developed the idea of using a torsional member attached to the
column (a cutaway from the three-dimensional structure) in the transverse direction at one
end and connected to the slab-beam elements at the other end. The conceptual model is
illustrated in Fig. 16.20.1(c). In this model, beams A′-B′ and B′-C′ represent the slab-beams
while elements AA′, BB′, and CC′ represent the torsional elements.
701
Figure 16.20.2 Sections for computing the flexural stiffness of equivalent beams (slab-beams), Ksb, and variation of
moment of inertia for two-way slab systems. (Adapted from ACI 318-77 Commentary.)
702
Based on tests results, Corley and Jirsa [16.27] developed a formula for the torsional
stiffness Kt of the torsional member as follows [ACI Commentary R8.11.5]:
9 Ecs C  I sb 
Kt = ∑ 3  I  (16.20.1)
 c  s

L2  1 − 2 
 L  2
in which
C = torsional constant of the transverse beam (see Section 16.11)
Ecs = modulus of elasticity of slab concrete
Is = moment of inertia of slab over width of equivalent frame
Isb = moment of inertia of entire T-section (if so) within the width of the equivalent
frame
L2 = span of member subject to torsion
c1 = width of rectangular or equivalent rectangular column, capital, or bracket meas-
ured in the direction for which moments are being computed
c2 = width of rectangular or equivalent rectangular column, capital, or bracket meas-
ured in the direction perpendicular to c1 for which moments are being computed
The summation sign is for the transverse spans (denoted by L2) on each side of the column.
Flexural Stiffness of Column

The flexural stiffness of the columns, Kc, must account for the variation of moment of iner-
tia along the column axis (ACI-8.11.4.2). The height of the column is to be measured from
c1
CL of panel
c2
Rotation equals
column
e
rotation
m
ra
tf
len
va
Rotation
ui
more than
eq
of
column
th
rotation
id
W
A
A
x1
CL of panel
Slab thickness, t
Cross section
z c1 of torsional
Axis of
rotation member
t
Section A–A
Figure 16.20.3 Torsional members for equivalent columns.

703
I= ∞
I = linear variation
H
Constant Ic
I= ∞
Figure 16.20.4 Basis of calculation of column stiffness.
middepth of slab above to middepth of slab below (ACI-R8.11.4), as shown in Fig. 16.20.4.

From the top to the bottom of the slab-beam at the joint (ACI-8.11.4.1), the moment of
inertia is to be taken as infinite. In flat slabs having column capitals, the moment of inertia
may be assumed to vary linearly from infinite at the top to its value based on gross cross
section of the column at the bottom of the capital, as shown in Fig. 16.20.4. Outside of
joints or column capitals, the moment of inertia may be based on the column gross cross
section (ACI-8.11.4.3).
The model shown in Fig. 16.20.1(c) consists of three element types: the equivalent
beams (slab-beams), the torsional elements, and the columns. In the traditional equivalent
frame method, the columns and the torsional elements are combined into and replaced by
equivalent columns in order to allow analysis of the slab system as a plane frame. The flex-
ibility (inverse of stiffness) of this equivalent column is taken as the sum of the flexibilities
of the actual columns above and below the slab-beam and the flexibility of the attached
torsion member, such that
1 1 1 (16.20.2)
= +
K ec ∑ K c K t
from which the stiffness of the equivalent column, Kec, is
∑ Kc
K ec = (16.20.3)
∑ Kc
1+
Kt
where the summation sign is for the columns above and below the slab.
With stiffness properties computed for the slab-beams and the equivalent columns, the
three-dimensional subassembly shown in Fig. 16.20.1 may then be analyzed as a plane
frame as shown in Fig. 16.20.5.
Ksb Ksb
Kec Kec Kec
Figure 16.20.5 Two-dimensional equivalent frame of a two-way slab system for gravity load
analysis.
704
Arrangement of Live Load
When the load pattern is known, analysis should be made for it (ACI-6.4.3.1). When service
live load does not exceed 3 4 of the service dead load, analysis is permitted to be made only
for full factored dead and live load on all spans (ACI-6.4.3.2). When load patterns in accord-
ance with influence line concepts are used, only 3 4 of the full factored live load may be used
in accordance with ACI-6.4.3.3; however, factored moments used in design should not be less
than those due to full factored dead and live loads on all panels (ACI-6.4.3).
Reduction of Negative Moments Obtained at Column Centerlines

from Structural Analysis
Negative moments obtained at interior column centerlines may be reduced to the face of
rectilinear or equivalent square (for circular or polygonal) supports, but not greater than
0.175L1 from the column centerline (ACI-8.11.6.1). For exterior columns having capi-
tals or brackets, reduction of negative moments can be made only to a section no greater
than halfway between the face of the column and the edge of the capital or bracket (ACI-
8.11.6.3). For exterior columns without brackets or capitals, the critical section for negative
moments is taken at the face of the column (ACI-8.11.6.2).
Deflections
When the deflection must be calculated for a two-way slab system, the ACI Code (ACI-
24.2.3.3) provides little guidance other than the requirement to take into account “size and
shape of the panel, conditions of support, and nature of restraints at panel edges.” The effec-
tive moment of inertia Ie [Eq. (12.9.1)] is required to be used in such calculations. Although
a number of techniques have been proposed [16.52–16.71], adoption of the equivalent
frame concept seems to have the most promise of being relatively simple to apply and giv-
ing reasonable results. This equivalent frame application has been developed by Nilson and
Walters [16.53] for essentially uncracked systems and extended by Kripanarayanan and
Branson [16.55] for partially cracked load ranges. Scanlon and others [16.58, 16.63, 16.65,
16.66, 16.68, and 16.70] have also treated the subject in detail.
EXAMPLE 16.20.1
Assuming that the equivalent frame method is to be applied to the two-way slab (with
beams) design of Example 16.3.1, obtain the properties necessary for the analysis of the
equivalent rigid frames A, B, C, and D as shown by the notations in Fig. 16.3.5.
SOLUTION
(a) The cross sections and the moment of inertia values for the slab-beams for frames A,
B, C, and D are shown in Fig. 16.20.6. Note that moments of inertia are based on the
gross sections as permitted by ACI-8.11.3.3 for gravity load analysis. The variations
in the moment of inertia of the slab-beam in the long and short directions are shown in
Fig. 16.20.7. For the long slab-beam, the ratio of moment of inertia between the center
and the face of the column to the moments of inertia of the rest of the span is 1.0/(1 –
15/240)2 = 1.14; it is 1.0/(1 –15/300)2 = 1.11 for the short slab-beam (ACI-8.11.3.1).
(b) Compute flexure properties of columns. The values and the variations in the moment
of inertia of the column section in the long and short directions are shown in
Fig. 16.20.8.
(c) Compute torsional stiffness of transverse torsional members. The torsional constants
C for the transverse members are taken from Example 16.11.1. These values, needed
to compute the torsional stiffness Kt [Eq. (16.20.1)] in each direction, are shown in
Fig. 16.20.9.
(Continued)
705
240” 7” 120”
0.98”
0.49”
21.5” 1
21.5”
6 2 in. slab
Isb = 57,730 in.4
Isb = 66,540 in.4
Is = 2910 in.4
14” Is = 5490 in.4 14”
Isb /Is = 19.84
Isb /Is = 12.12
(a) For frames A and B
300” 6” 150”
2.09” 1.19”
17.5” 1 17.5”
6 2 in. slab
Isb = 39,530 in.4 Isb = 33,980 in.4
12” Is = 6870 in.4 12” Is = 3570 in.4

Isb /Is = 5.75 Isb /Is = 9.52
(b) For frames C and D
Figure 16.20.6 Slab-beam cross s ections in the two-way slab (with beams) of Example 16.20.1.
15” 15” 15” 15”
6.5” 28” 6.5” 24”
25’ 20’
Isb Isb
= 1.14 Isb = 1.11 Isb
(1 – c2/L2)2 (1 – c2/L2)2
66,540 in 4 (frame A) 39,530 in 4 (frame C)

Isb = Isb =
57,730 in 4 (frame B) 33,980 in 4 (frame D)
23.75’ 0.625’ 18.75’ 0.625’
0.625’ 25.00’ 0.625’ 20.00’
(a) Long slab-beam (b) Short slab-beam
Figure 16.20.7 Flexure properties of slab-beam in the two-way slab (with beams) of Example 16.20.1.
(Continued)
706
I=∞
6.5”
(ACI-8.11.4.1)
21.5” 24.75” 2.06’
Bottom of
slab-beam
12’ – 0” 12’ –0”

9.67’ Ic = 4220 in.4
3.25” I=∞
0.27’
(a) Column section in long direction (for frames A and B)
6.5”
17.5” 20.75” 1.73’ I=∞
12’–0”
12’ – 0”
10.00’ Ic = 4220 in.4
3.25” I=∞
0.27’
(b) Column section in short direction (for frames C and D)
Figure 16.20.8 Flexure properties of columns in the two-way slab (with beams) of Example 16.20.1.
D C
C = 10,700 in.4
C = 11,930 in.4
20’
C = 20,700 in.4 C = 20,700 in.4

A
C = 10,700 in.4
C = 11,930 in.4
20’
C = 19,100 in.4 C = 19,100 in.4
25’ 25’
Figure 16.20.9 Torsional constants in the two-way slab (with beams) in Example 16.11.1.
(Continued)
70
For Frame A, using Isb/Is = 12.12 for 14 × 21.5 projection below 240 × 6.5 slab, and
noting that Ecs = E,
18 E (10, 700)
exterior K t = (12.12) = 974 E (12.12) = 11, 800 E
240 (1 − 15 / 240)3

18 E (11, 930)
interior K t = (12.12) = 1086 E (12.12) = 13, 200 E
240 (1 − 15 / 240)3
For Frame B, using Isb/Is = 19.84 for 14 × 21.5 projection below 127 × 6.5 slab,
exterior K t = 487 E (19.84) = 9660 E

interior K t = 543E (19.84) = 10, 800 E
For Frame C, using Isb/Is = 5.75 for 12 × 17.5 projection below 300 × 6.5 slab,
18 E (19,100)
exterior K t = (5.75) = 1340 E (5.75) = 7700 E
300 (1 − 15 / 300)3

18 E (20, 700)
interior K t = (5.75) = 1450 E (5.75) = 8340E
300 (1 − 15 / 300)3
For Frame D, using Isb/Is = 9.52 for 12 × 17.5 projection below 156 × 6.5 slab,
exterior K t = 670 E (9.52) = 6380 E

interior K t = 725E (9.52) = 6900 E
EXAMPLE 16.20.2
Assuming the equivalent frame method is to be applied to the flat slab of Example 16.3.2,
obtain the properties necessary for the analysis of equivalent rigid Frame A in the long
direction.
SOLUTION
(a) Compute flexure properties of slab strip. The moment of inertia through the 7 1 2 -in.
slab is computed as that of a rectangular section of width L2 = 240 in. Thus,
240(7.5)3
I sb = = 8440 in.4
12
The moment of inertia through the drop, where there is a T-section of a 240 × 7.5 in.
flange and an 84 × 3 in. web, is 14,720 in.4 or 1.75Isb. The moment of inertia between
the column centerline and the face of the equivalent square column capital is 1.75Isb /
(1 –4.43/20)2 = 2.88 Isb.
The variation in the moment of inertia along the interior span of the slab strip is
(b) Compute flexure properties of columns. The column length is measured between the
centerlines of slab thickness, as shown in Fig. 16.20.11. The moment of inertia is
assumed to be infinite from the top of the slab to the bottom of the drop panel, and
then is taken to vary linearly to the base of the column capital.
(Continued)
708
(c) Compute torsional stiffness of transverse torsional members. From Example 16.11.2,
C (edge beam ) = 18, 500 in.4

C (interior beam ) = 9800 in.4
For the two members, one framing in from each side,
2(9 E )C 2(9 E )(18, 500)

K t (edge) = 3
= 3
= 2980 E
 c   4.5 
L2  1 − 2  240  1 − 
 L  2
 20 

2(9 E )9800
K t (interior ) = 3
= 1560 E
 4.43 
240  1 − 
 20 
18” 18”
7.5” 10.5”
1.5”
21”
1/2 width of equivalent

square column capital
Isb = 8440 in.4

Idrop
= 2.88 Isb Idrop = 1.75Isb Idrop = 1.75Isb 2.88Isb
(1– c2/L2)2
2.21’ 16.66” 2.21’
1.96’ 25.00” 1.96’
Figure 16.20.10 Flexure properties of frame A in the flat slab in Example 16.20.2.

(Continued)
709
11 “ 0.056 L
2 I= ∞
29.25”
0.188L
21”
21”
Linear variation
10’–0”
10’–0”
18”
Ic = 5150 in.4
0.725 L
3.75” I= ∞
0.031 L
(a) Interior column
11 “ 0.056 L
2 I= ∞
27.25”
0.171 L
19”
19”
Linear variation
10’– 0”
16”
10’– 0”
Ic = 5460 in.4
0.742 L
3.75” I= ∞
0.031 L
(b) Exterior column
Figure 16.20.11 Flexure properties of columns in the flat slab of Example 16.20.2.

710
16.21 EQUIVALENT FRAME MODELS

As noted in the preceding section, the equivalent frame method was originally devel-
oped to be used with the moment distribution method—that is, the usual method of
structural analysis for plane frames at the time. The method, however, can be used with
any standard analysis software by specifying appropriate values of the stiffness proper-
ties for the slab-beams, columns, and torsional elements. This section briefly discusses
several alternative models for implementing the EFM into virtually any structural anal-
ysis software.
Perhaps the most sophisticated model would be to divide the slab into a mesh of finite
elements supported on columns as shown in Fig. 7.2.1. This technique has been used in
some investigations [16.32] and is sometimes used in design practice as well. One of the
main challenges in using this technique is arriving at a definition of the appropriate stiff-
ness of the slab elements (and any beams) that will adequately capture the moment transfer
between the slab and the column observed in experiments.
When a finite element analysis may not be efficient or is too costly for a given project,
a two-dimensional model, where the slab-to-column moment transfer is modeled via tor-
sional or rotational springs attached to the columns or the slab-beams, could be used. These
concepts are presented in the following sections.
Model with Torsional Springs

This model is essentially that shown in Fig. 16.20.1(c) and corresponds to the original
model developed for the EFM: the ends of the slab-beam are connected to the columns with
zero length elements that are rigid except in torsion. This approach may require placing a
special joint at the same location where the slab-beams are connected to the columns to
insert the torsional springs. The properties of the slab-beams, Ksb, columns, Kc, and tor-
sional elements, Kt, are those presented earlier (see Section 16.20).
Model with Rotational Springs

An alternative to using torsional springs is to use equivalent rotational springs at the ends
of columns or the slab-beams. In Fig. 16.21.1, the torsion element is replaced by rota-
tional springs at the column ends, one joined to the lower end of the upper column and
the other joined to the upper end of the lower column. In such a case, the stiffness of the
original torsional element is divided proportionally according to the column stiffness at
that end.
Alternatively, the torsion element may be replaced by rotational springs at the slab-
beam ends, as shown in Fig. 16.21.2. At each interior column, the torsional element may
be divided into two such elements in parallel, one joined to the right end of the left beam
and the other to the left end of the right beam, with the total torsional stiffness proportioned
according to the beam stiffness at that end.
Figure 16.21.1 Torsion elements replaced by rotational springs at column ends, distributed to

upper and lower columns in proportion to the column flexural stiffnesses (same as Kec model of
Vanderbilt [16.32]).
71
Figure 16.21.2 Torsion elements replaced by rotational springs at slab-beam ends (same as Keb
model of Vanderbilt [16.32]).
Other Models
As yet another alternative approach to using torsional or rotational springs, some research-
ers have proposed the use of an effective slab width to define an equivalent beam. In this
method, a slab width factor is computed based on the assumed stiffness of the columns
and beam-column connection regions and the expected crack regions in the slab. Studies
concerning the use of an effective slab width include those of Pecknold [16.38], Allen and
Darvall [16.39, 16.47], Elias [16.41–16.43], Luo and Durrani [16.140, 16.141], Hwang and
Moehle [16.145], and Dovich and Wight [16.146].
16.22 EQUIVALENT FRAME METHOD FOR LATERAL

LOAD ANALYSIS
As noted in Section 16.20, the equivalent frame method was developed and validated using
data from tests and analyses of two-way slab systems under gravity loads. The method,
however, can also be used for lateral load analysis with few modifications. First, the anal-
ysis must be conducted for the entire frame rather than just the subassembly used for grav-
ity loads. In addition, rather than gross section properties, the moments of inertia must be
computed in accordance with ACI-6.6.3.1 (see Section 13.8).
Note that because the stiffness properties are different, two separate models are
required: one for gravity load analysis and another for lateral load analysis. The results of
these analyses are permitted to be combined in accordance with ACI-8.4.1.9.
Clearly, a more convenient approach would be to construct a single model for the anal-
ysis of both gravity and lateral loads. While this is possible with the available structural
analysis software, the agreement between the results of the analyses and the test data can be
quite sensitive to the model type and to the stiffness properties assumed in the model. Cano
and Klingner [16.51] have provided a good review and comparison of various analysis pro-
cedures for two-way slab systems under combined gravity and lateral loads. Additionally,
Hwang and Moehle [16.145] have provided good insight and guidance for the analysis of
slab-column frames under lateral loads.
SELECTED REFERENCES
General and Historical
16.1. J. R. Nichols. “Statical Limitations Upon the Steel Requirement in Reinforced Concrete Flat
Slab Floors,” Transactions ASCE, 77, 1914, 1670–1736.
16.2. George A. Hool and Nathan C. Johnson. Concrete Engineers Handbook. New York: McGraw-
Hill, 1918 (pp. 457–486).
16.3. H. M. Westergaard and W. A. Slater. “Moments and Stresses in Slabs,” ACI Proceedings, 17,
1921, 415.
712
16.4. H. M. Westergaard. “Formulas for the Design of Rectangular Floor Slabs and Supporting
Girders,” ACI Proceedings, 22, 1926, 26.
16.5. Joseph A. Wise, “Design of Reinforced Concrete Slabs,” ACI Proceedings, 25, 1929, 712.
16.6. Joseph DiStasio and M. P. Van Buren. “Slabs Supported on Four Sides,” ACI Journal,
16.7. R. L. Bertin, Joseph DiStasio, and M. P. Van Buren. “Slabs Supported on Four Sides,” ACI
Journal, Proceedings, 41, June 1945, 537–556.
16.8. C. P. Siess and N. M. Newmark. “Rational Analysis and Design of Two-Way Concrete Slabs.”
ACI Journal, Proceedings, 45, December 1948, 273–316.
16.9. Sidney H. Simmonds and Janko Misic. “Design Factors for the Equivalent Frame Method,” ACI
Journal, Proceedings, 68, November 1971, 825–831.
16.10. Alex E. Cardenas, Rolf J. Lenschow, and Mete A. Sozen. “Stiffness of Reinforced Concrete
Plates,” Journal of the Structural Division, ASCE, 98, ST11 (November 1972), 2587–2603.
16.11. C. K. Wang. Intermediate Structural Analysis. New York: McGraw-Hill, 1983.
Gravity Load—Tests
16.12. W. L. Gamble, M. A. Sozen, and C. P. Siess. “Measured and Theoretical Bending Moments
in Reinforced Concrete Floor Slabs,” Civil Engineering Structural Research Series No. 246.
Urbana: University of Illinois, June 1962.
16.13. M. A. Sozen and C. P. Siess. “Investigation of Multi-Panel Reinforced Concrete Floor Slabs,”
ACI Journal, Proceedings, 60, August 1963, 999–1028.
16.14. S. A. Guralnick and R. W. LaFraugh. “Laboratory Study of a 45-Foot Square Flat Plate
Structure,” ACI Journal, Proceedings, 60, September 1963, 1107–1185.
16.15. David S. Hatcher, Mete A. Sozen, and Chester P. Siess. “Test of a Reinforced Concrete Flat
Plate,” Journal of the Structural Division, ASCE, 91, ST5 (October 1965), 205–231.
16.16. James O. Jirsa, Mete A. Sozen, and Chester P. Siess. “Test of a Flat Slab Reinforced with
Welded Wire Fabric,” Journal of the Structural Division, ASCE, 92, ST3 (June 1966), 199–224.
16.17. W. L. Gamble, M. A. Sozen, and C. P. Siess. “Tests of a Two-Way Reinforced Concrete Floor
Slab,” Journal of the Structural Division, ASCE, 95, ST6 (June 1969), 1073–1096.
16.18 M. Daniel Vanderbilt, Mete A. Sozen, and Chester P. Siess. “Tests of a Modified Reinforced
Concrete Two-Way Slab,” Journal of the Structural Division, ASCE, 95, ST6 (June 1969),
1097–1116.
16.19. D. S. Hatcher, Mete A. Sozen, and Chester P. Siess. “Test of a Reinforced Concrete Flat Slab,”
Journal of the Structural Division, ASCE, 95, ST6 (June 1969), 1051–1072.
16.20. E. Ramzy, F. Zaghlool, H. A. Rawdon de Paiva, and Peter G. Glockner. “Tests of Reinforced
Concrete Flat Plate Floors,” Journal of the Structural Division, ASCE, 96, ST3 (March 1970),
487–507.
16.21. Alex E. Cardenas and Paul H. Kaar. “Field Test of a Flat Plate Structure,” ACI Journal,
Proceedings, 68, January 1971, 50– 58.
16.22. Donald D. Magura and W. Gene Corley. “Tests to Destruction of a Multipanel Waffle Slab
Structure—1964–1965 New York World’s Fair,” ACI Journal, Proceedings, 68, September
1971, 699–703.
16.23. A. S. Hall and B. V. Rangan. “Moments in Edge Panels of Flat Plate Floors,” Journal of
Structural Engineering, ASCE, 109, 11 (November 1983), 2638–2650.
16.24. Abdel Wahid Hago and Prabhakara Bhatt. “Tests on Reinforced Concrete Slabs Designed
by Direct Design Procedure,” ACI Journal, Proceedings, 83, November–December 1986,
916–924.
16.25. B. Vijaya Rangan and A. S. Hall. “Moment Redistribution in Flat Plate Floors,” ACI Journal,
Proceedings, 81, November–December 1984, 601–608.
Gravity Load—Analytical Methods

16.26. W. G. Corley, M. A. Sozen, and C. P. Siess. “The Equivalent Frame Analysis for Reinforced
Concrete Slabs,” Structural Research Series No. 218. Urbana: University of Illinois, Civil
Engineering Department, June 1961.
16.27. W. G. Corley and J. O. Jirsa. “Equivalent Frame Analysis for Slab Design,” ACI Journal,
713
16.28. Maurice P. Van Buren. “Staggered Columns in Flat Plates,” Journal of the Structural Division,
ASCE, 97, ST6 (June 1971), 1791–1797.
16.29. William L. Gamble. “Moments in Beam Supported Slabs,” ACI Journal, Proceedings, 69,
March 1972, 149–157.
16.30. Donald J. Fraser. “Equivalent Frame Method of Beam-
Slab Structures,” ACI Journal,
Proceedings, 74, May 1977, 223– 228.
16.31. S. K. Sharan, D. Clyde, and D. Turcke. “Equivalent Frame Analysis Improvements for Slab
Design,” ACI Journal, Proceedings, 75, February 1978, 55–59.
16.32. M. Daniel Vanderbilt. Equivalent Frame Analysis of Unbraced Reinforced Concrete Buildings
for Static Lateral Loads. Structural Research Report No. 36, Civil Engineering Department,
Colorado State University, Fort Collins, June 1981.
16.33. D. J. Fraser. “The Equivalent Frame Method Simplified for Beam and Slab Construction,”
Concrete International, 4, April 1982, 66–73.
16.34. J. F. Mulcahy and J. M. Rotter. “Moment Rotation Characteristics of Flat Plate and Column
Systems,” ACI Journal, Proceedings, 80, March–April 1983, 85–92.
16.35. D. J. Fraser. “Simplified Frame Analysis for Flat Plate Construction,” Concrete International,
6, September 1984, 32–41.
Pattern Gravity Loading

16.36. J. O. Jirsa, M. A. Sozen, and C. P. Siess. “Pattern Loadings on Reinforced Concrete Floor
Slabs,” Journal of the Structural Division, ASCE, 95, ST6 (June 1969), 1117–1137.
16.37. Jan C. Jofriet and Gregory M. McNeice. “Pattern Loading on Reinforced Concrete Flat Plates,”
ACI Journal, Proceedings, 68, December 1971, 968–972.
Lateral Load Analysis on Slab Frames

16.38. David A. Pecknold. “Slab Effective Width for Equivalent Frame Analysis,” ACI Journal,
16.39. Fred Allen and Peter Darvall. “Lateral Load Equivalent Frame,” ACI Journal, Proceedings, 74,
July 1977, 294– 299.
16.40. M. Daniel Vanderbilt. “Equivalent Frame Analysis for Lateral Loads.” Journal of the Structural
Division, ASCE, 105, ST10 (October 1979), 1981–1998. Disc., 106, ST7 (July 1980), 1671–
1672; 107, ST1 (January 1981), 245.
16.41. Ziad M. Elias. “Sidesway Analysis of Flat Plate Structures,” ACI Journal, Proceedings, 76,
March 1979, 421–442.
16.42. Ziad M. Elias and Constantinos Georgiadis. “Flat Slab Analysis Using Equivalent Beams,” ACI
Journal, Proceedings, 76, October 1979, 1063–1078.
16.43. Ziad M. Elias. “Lateral Stiffness of Flat Plate Structures,” ACI Journal, Proceedings, 80,
16.44. Donald J. Fraser. “Elastic Analysis of Laterally Loaded Frames,” Journal of Structural

Engineering, ASCE, 109, 6 (June 1983), 1479–1489.
16.45. M. Daniel Vanderbilt and W. Gene Corley. “Frame Analysis of Concrete Buildings,” Concrete
International, 5, December 1983, 33–43.
16.46. I. Paul Lew and Fruma Narov. “Three-Dimensional Equivalent Frame Analysis of Shearwalls,”
Concrete International, 5, October 1983, 25–30.
16.47. Peter Darvall and Fred Allen. “Lateral Load Effective Width of Flat Plates with Drop Panels,”
ACI Journal, Proceedings, 81, November–December 1984, 613–617.
16.48. Milija N. Pavlovic and Steven M. Poulton. “On the Computation of Slab Effective Widths,”
Journal of Structural Engineering, ASCE, 111, 2 (February 1985), 363–377.
16.49. Jack P. Moehle and John W. Diebold. “Lateral Load Response of Flat-Plate Frame,” Journal of
Structural Engineering, ASCE, 111, 10 (October 1985), 2149–2164.
16.50. Cheng-Tzu Thomas Hsu. “Lateral Displacement for Unbraced Concrete Frame Buildings,” ACI
16.51. Mary Theresa Cano and Richard E. Klingner. “Comparison of Analysis Procedures for Two-
Way Slabs,” ACI Structural Journal, 85, November–December 1988, 597–608. Disc. 86,
714
Two-Way Slab Deflections

16.52. Mortimer D. Vanderbilt, Mete A. Sozen, and Chester P. Siess. “Deflections of Multiple-Panel
Reinforced Concrete Floor Slabs,” Journal of the Structural Division, ASCE, 91, ST4 (August
1965), 77–101.
16.53. Arthur H. Nilson and Donald B. Walters, Jr. “Deflection of Two-Way Floor Systems by the
Equivalent Frame Method,” ACI Journal, Proceedings, 72, May 1975, 210–218.
16.54. B. Vijaya Rangan. “Prediction of Long-Term Deflections of Flat Plates and Slabs,” ACI Journal,
16.55. K. M. Kripanarayanan and D. E. Branson. “Short Time Deflections of Flat Plates, Flat Slabs,
and Two-Way Slabs,” ACI Journal, Proceedings, 73, December 1976, 686–690.
16.56. P. J. Taylor and J. L. Heiman. “Long-Term Deflection of Reinforced Concrete Flat Slabs and
Plates,” ACI Journal, Proceedings, 74, November 1977, 556–561.
16.57. B. Vijaya Rangan and Arthur E. McMullen. “A Rational Approach to Control of Slab

Deflections,” ACI Journal, Proceedings, 75, June 1978, 256–262.
16.58. Andrew Scanlon and David W. Murray. “Practical Calculation of Two-Way Slab Deflections,”
Concrete International, 4, November 1982, 43–50.
16.59. John A. Sbarounis. “Multistory Flat Plate Buildings: Construction Loads and Immediate
Deflections,” Concrete International, 6, February 1984, 70–77.
16.60. John A. Sbarounis. “Multistory Flat Plate Buildings: Effect of Construction Loads on Long-
Term Deflections,” Concrete International, 6, April 1984, 62–70. Disc., 7, February 1985, 62.
16.61. John A. Sbarounis. “Multistory Flat Plate Buildings: Measured and Computed One-Year
Deflections,” Concrete International, 6, August 1984, 31–35.
16.62. R. I. Gilbert. “Deflection Control of Slabs Using Allowable Span to Depth Ratios,” ACI Journal,
16.63. Cameron J. Graham and Andrew Scanlon. “Deflections of Concrete Slabs Under Construction
Loading,” Deflections of Concrete Structures (SP- 86). Detroit: American Concrete
Institute, 1985.
16.64. B.Vijaya Rangan. “Estimation of Slab Deflections in Flat Plate Buildings,” ACI Journal,
Proceedings, 83, March–April 1986, 269–273.
16.65. K. S. Stephen Tam and Andrew Scanlon. “Deflection of Two-Way Slabs Subjected to Restrained
Volume Change and Transverse Loads,” ACI Journal, Proceedings, 83, September–October
1986, 737–744.
16.66. Cameron J. Graham and Andrew Scanlon. “Long-Time Multipliers for Estimating Two-Way
Slab Deflections,” ACI Journal, Proceedings, 83, November–December 1986, 899– 908.
16.67. N. J. Gardner and H. C. Fu. “Effects of High Construction Loads on the Long-Term Deflections
of Flat Slabs,” ACI Structural Journal, 84, July–August 1987, 349–360. Disc., 85, May–June
1988, 359.
16.68. David P. Thompson and Andrew Scanlon. “Minimum Thickness Requirements for Control of
Two-Way Slab Deflections,” ACI Structural Journal, 85, January–February, 1988, 12–22.
16.69. Amin Ghali. “Prediction of Deflections of Two-Way Floor Systems,” ACI Structural Journal,
86, September–October 1989, 551–562.
16.70. Andrew Scanlon and David P. Thompson. “Evaluation of ACI 318 Requirements for Control
of Two-Way Slab Deflections,” ACI Structural Journal, 87, November–December 1990,
657–661.
16.71. ACI Committee 435. “State-of-the-Art Report on Two-Way Slab Deflections,” ACI Structural
Crack Control in Two-Way Slab Systems

16.72. Edward G. Nawy. “Crack Width Control in Welded Fabric Reinforced Centrally Loaded Two-
Way Concrete Slabs,” Causes, Mechanism, and Control of Cracking in Concrete (SP-20).
Detroit: American Concrete Institute, 1968 (pp. 211– 235).
16.73. Edward G. Nawy and G. S. Orenstein. “Crack Width Control in Reinforced Concrete Two-Way
Slabs,” Journal of the Structural Division, ASCE, 96, ST3 (March 1970), 701– 721.
16.74. Edward G. Nawy and Kenneth W. Blair. “Further Studies on Flexural Crack Control in
Structural Slab Systems,” Cracking, Deflection, and Ultimate Load of Concrete Slab Systems
16.75. Edward G. Nawy. “Crack Control Through Reinforcement Distribution in Two-Way Acting
Slabs and Plates,” ACI Journal, Proceedings, 69, April 1972, 217–219.
715
Shear Strength
16.76. Johannes Moe. Shearing Strength of Reinforced Concrete Slabs and Footings Under
Concentrated Loads, Development Department Bulletin D47. Chicago: Portland Cement
Association, April 1961, 130 pp.
16.77. ACI-
ASCE Committee 326. “Report on Shear and Diagonal Tension,” ACI Journal,
Proceedings, 59, January, February, and March, 1962, 1–30, 277–344, and 352–396.
16.78. Neil M. Hawkins, H. B. Fallsen, and R. C. Hinojosa. “Influence of Column Rectangularity on
the Behavior of Flat Plate Structures,” Cracking, Deflection, and Ultimate Load of Concrete
Slab Systems (SP-30). Detroit: American Concrete Institute, 1971 (pp. 127–146).
16.79. Frederic Roll, S. T. H. Zaidi, Gajanan Sabnis, and Kuang Chuang. “Shear Resistance of
Perforated Reinforced Concrete Slabs,” Cracking, Deflection, and Ultimate Load of Concrete
Slab Systems (SP-30). Detroit: American Concrete Institute, 1971 (pp. 77–100).
16.80. M. Daniel Vanderbilt. “Shear Strength of Continuous Plates,” Journal of the Structural Division,
ASCE, 98, ST5 (May 1972), 961–973.
16.81. M. E. Criswell and N. M. Hawkins. “Shear Strength of Slabs: Basic Principles and Their
Relation to Current Methods of Analysis,” Shear in Reinforced Concrete, Vol. 2 (SP-42).
16.82. N. M. Hawkins, M. E. Criswell, and F. Roll. “Shear Strength of Slabs Without Shear

Reinforcement,” Shear in Reinforced Concrete, Vol. 2 (SP-42). Detroit: American Concrete
Institute, 1974 (pp. 677–720).
16.83. Neil M. Hawkins, (Chairman) “The Shear Strength of Reinforced Concrete Members—Slabs,
by the Joint ASCE-ACI Task Committee 426 on Shear and Diagonal Tension of the Committee
on Masonry and Reinforced Concrete of the Structural Division,” Journal of the Structural
Division, ASCE, 100, ST8 (August 1974), 1543–1591.
16.84. Brian E. Hewitt and Barrington de V. Batchelor. “Punching Shear Strength of Restrained
Slabs,” Journal of the Structural Division, ASCE, 101, ST9 (September 1975), 1837–1853.
16.85. Timothy J. Ross and Helmut Krawinkler. “Impulsive Direct Shear Failure in RC Slabs,” Journal
of Structural Engineering, ASCE, 111, 8 (August 1985), 1661–1677.
16.86. Fernando Gonzalez-Vidosa, Michael D. Kotsovos, and Milija N. Pavlovic. “Symmetrical
Punching of Reinforced Concrete Slabs: An Analytical Investigation Based on Nonlinear Finite
Element Modeling,” ACI Structural Journal, 85, May–June 1988, 241–250.
16.87. Ibrahim A. E. M. Shehata and Paul E. Regan. “Punching in RC Slabs,” Journal of Structural
Engineering, ASCE, 115, 7 (July 1989), 1726–1740.
16.88. Lionello Bortolotti. “Punching Shear Strength in Concrete Slabs,” ACI Structural Journal, 87,
March–April 1990, 208– 219.
16.89. Carl Erik Broms. “Punching of Flat Plates—A Question of Concrete Properties in Biaxial
Compression and Size Effect,” ACI Structural Journal, 87, May–June 1990, 292–304.
16.90. David I. McLean, Long T. Phan, H. S. Lew, and Richard N. White. “Punching Shear Behavior
of Lightweight Concrete Slabs and Shells,” ACI Structural Journal, 87, July–August 1990,
386–392. Disc., 88, May–June 1991, 382–383.
16.91. John S. Lovrovich and David I. McLean. “Punching Shear Behavior of Slabs with Varying
Span–Depth Ratios,” ACI Structural Journal, 87, September–October 1990, 507–511. Disc.,
87, July–August 1991, 515–516.
16.92. J. S. Kuang and C. T. Morley. “Punching Shear Behavior of Restrained Reinforced Concrete
Slabs,” ACI Structural Journal, 89, January–February 1992, 13–19.
Shear Reinforcement in Slabs

16.93. W. G. Corley and N. M. Hawkins. “Shearhead Reinforcement for Slabs,” ACI Journal,

16.94. N. M. Hawkins. “Shear Strength of Slabs with Shear Reinforcement,” Shear in Reinforced
Concrete, Vol. 2 (SP-42). Detroit: American Concrete Institute, 1974 (pp. 785–816).
16.95. Paul H. Langohr, Amin Ghali, and Walter H. Dilger. “Special Shear Reinforcement for Concrete
Flat Plates,” ACI Journal, Proceedings, 73, March 1976, 141–146.
16.96. Amin Ghali, Mahmoud Z. Elmasri, and Walter Dilger. “Punching of Flat Plates under Static
and Dynamic Horizontal Forces,” ACI Journal, Proceedings, 73, October 1976, 566–576.
16.97. Walter Dilger, Mahmoud Z. Elmasri, and Amin Ghali. “Flat Plates with Special Shear

Reinforcement Subjected to Static Dynamic Moment Transfer,” ACI Journal, Proceedings, 75,
October 1978, 543–549.
716
16.98. Frieder Seible, Amin Ghali, and Walter H. Dilger. “Preassembled Shear Reinforcing Units for
Flat Plates,” ACI Journal, Proceedings, 77, January–February 1980, 28–35.
16.99. Walter H. Dilger and Amin Ghali. “Shear Reinforcement for Concrete Slabs.” Journal of the
Structural Division, ASCE, 107, ST12 (December 1981), 2403–2420.
16.100. Adel A. Elgabry and Amin Ghali. “Design of Stud-Shear Reinforcement for Slabs,” ACI
16.101. ACI-ASCE Committee 421. “Guide to Shear Reinforcement for Slabs,” ACI Manual of
Concrete Practice, ACI 421.1R-08. Farmington Hills, MI: American Concrete Institute, 2008
(27 pp.).
16.102. Tetsuya, Yamada, Antonio Nanni, and Katsushiko Endo. “Punching Shear Resistance of Flat
Slabs: Influence of Reinforcement Type and Ratio,” ACI Structural Journal, 89, September–
October 1992, 555–563.
Shear-Moment Transfer
16.103. Joseph DiStasio and M. P. Van Buren. “Transfer of Bending Moment between Flat Plate Floor
and Column,” ACI Journal, Proceedings, 57, September 1960, 299–314.
16.104. N. W. Hanson and J. M. Hanson. “Shear and Moment Transfer between Concrete Slabs and
Columns,” Journal of the PCA Research and Development Laboratories, 10, (1) January
1968, 2–16.
16.105. Neil M. Hawkins and W. Gene Corley. “Transfer of Unbalanced Moment and Shear from Flat
Plates to Columns,” Cracking, Deflection, and Ultimate Load of Concrete Slab Systems (SP-
30). Detroit: American Institute, 1971 (pp. 147–176).
16.106. Adrian E. Long. “Punching Failure of Slabs—Transfer of Moment and Shear,” Journal of the
16.107. E. Ramzy F. Zaghlool and H. A. Rawdon de Paiva. “Strength Analysis of Corner Column-
Slab Connections,” Journal of the Structural Division, ASCE, 99, ST1 (January 1973), 53–70.
16.108. E. Ramzy F. Zaghlool and H. A. Rawdon De Paiva. “Tests of Flat-Plate Corner Column-Slab
Connections,” Journal of the Structural Division, ASCE, 99, ST3 (March 1973), 551–572.
16.109. N. M. Hawkins. “Shear Strength of Slabs with Moments Transferred to Columns,” Shear
in Reinforced Concrete, Vol. 2 (SP- 42). Detroit: American Concrete Institute, 1974
(pp. 817–846).
16.110. N. M. Hawkins and W. G. Corley. “Moment Transfer to Columns in Slabs with Shearhead
Reinforcement,” Shear in Reinforced Concrete, Vol. 2 (SP-42). Detroit: American Concrete
Institute, 1974 (pp. 847–880).
16.111. Shafiqul Islam and Robert Park. “Tests on Slab–
Column Connections with Shear and
Unbalanced Flexure,” Journal of the Structural Division, ASCE, 102, ST3 (March 1976),
549–568.
16.112. Robert Park and Shafiqul Islam. “Strength of Slab–Column Connections with Shear and
Unbalanced Flexure,” Journal of the Structural Division, ASCE, 102, ST9 (September 1976),
1879–1901.
16.113. Hans Gesund and Harinatha B. Goli. “Local Flexural Strength of Slabs at Interior Columns,”
16.114. V. W. Neth, H. A. R. de Paiva, and A. E. Long. “Behavior of Models of a Reinforced Concrete
Flat Plate Edge-Column Connection,” ACI Journal, Proceedings, 78, July–August 1981,
269–275.
16.115. S. Unnikrishna Pillai, Wayne Kirk, and Leonard Scavuzzo. “Shear Reinforcement at

Slab–Column Connections in a Reinforced Concrete Flat Plate Structure,” ACI Journal,
16.116. Harinatha B. Goli and Hans Gesund. “Flexural Strength of Flat Slabs at Exterior Columns,”
Journal of the Structural Division, ASCE, 108, ST11 (November 1982), 2479– 2495.
16.117. B. Vijaya Rangan and A. S. Hall, “Moment and Shear Transfer Between Slab and Edge
Column,” ACI Journal, Proceedings, 80, May–June 1983, 183–191.
16.118. Denby G. Morrison, Ikuo Hirasawa, and Mete A. Sozen. “Lateral-Load Tests of R/C Slab–
Column Connections,” Journal of Structural Engineering, ASCE, 109, 11 (November 1983),
2698–2714.
16.119. Denby G. Morrison. “Dynamic Lateral-
Load Tests of R/C Column– Slabs,” Journal of
Structural Engineering, ASCE, 111, 3 (March 1985), 685–698.
16.120. Paul F. Rice and Edward S. Hoffman. “Shear-Moment Transfer,” Concrete International, 9,
January 1987, 31–35.
71
16.121. P. R. Walker and P. E. Regan. “Corner Column–Slab Connections in Concrete Flat Plates,”
Journal of Structural Engineering, ASCE, 113, 4 (April 1987), 704–720.
16.122. Scott D. B. Alexander and Sidney H. Simmonds. “Ultimate Strength of Slab–
Column
Connections,” ACI Structural Journal, 84, May–June 1987, 255–261. Disc., 85, March–April
1988, 226–232.
16.123. Sidney H. Simmonds and Scott D. B. Alexander. “Truss Model for Edge Column–Slab
Connections,” ACI Structural Journal, 84, July–August 1987, 296–303. Disc., 85, May–June
1988, 352–353.
16.124. Ahmad J. Durrani and Hikmat E. Zerbe. “Seismic Resistance of R/C Exterior Connections
with Floor Slab,” Journal of Structural Engineering, ASCE, 113, 8 (August 1987), 1850–1864.
16.125. Adel A. Elgabry and Amin Ghali. “Tests on Concrete Slab–Column Connections with Stud-
Shear Reinforcement Subjected to Shear-Moment Transfer,” ACI Structural Journal, 84,
16.126. Jack P. Moehle. “Strength of Slab–Column Edge Connections,” ACI Structural Journal, 85,
January–February 1988, 89–98. Disc., 85, November–December 1988, 703–709.
16.127. Jack P. Moehle, Michael E. Kreger, and Roberto Leon. “Background to Recommendations
for Design of Reinforced Concrete Slab–Column Connections,” ACI Structural Journal, 85,
16.128. ACI-ASCE Committee 352. “Recommendations for Design of Slab–Column Connections
in Monolithic Reinforced Concrete Structures,” ACI Structural Journal, 85, November–
December 1988, 675–696. Disc., 86, July–August 1989, 496–499.
16.129. Amin Ghali. Discussion of “Recommendations for Design of Slab–Column Connections
in Monolithic Reinforced Concrete Structures,” ACI-ASCE Committee 352 Report (ACI
Structural Journal, 85, November–December 1988, 675–696), ACI Structural Journal, 86,
July–August 1989, 496–499.
16.130. Austin Pan and Jack P. Moehle. “Lateral Displacement Ductility of Reinforced Concrete Flat
Plates,” ACI Structural Journal, 86, May–June 1989, 250–258.
16.131. Neil M. Hawkins, Aibin Bao, and Jun Yamazaki. “Moment Transfer from Concrete Slabs to
Columns,” ACI Structural Journal, 86, November–December 1989, 705–716.
16.132. B. Vijaya Rangan. “Punching Shear Design in the New Australian Standard for Concrete
Structures,” ACI Structural Journal, 87, March–April 1990, 140–144. Disc., 88, January–
February 1991, 115.
16.133. Douglas A. Foutch, William L. Gamble, and Harianto Sunidja. “Tests of Post-Tensioned
Concrete Slab–Edge Column Connections,” ACI Structural Journal, 87, March–April 1990,
167–179. Disc., 88, January–February 1991, 116–117.
16.134. Peter Marti. “Design of Concrete Slabs for Transverse Shear,” ACI Structural Journal, 87,
March–April 1990, 180–190. Disc., 88, January–February 1991, 117–118.
16.135. Ian N. Robertson and Ahmad J. Durrani. “Gravity Load Effect on Seismic Behavior of Interior
Slab–Column Connections,” ACI Structural Journal, 89, January–February 1992, 37–45.
16.136. Scott D. B. Alexander and Sidney H. Simmonds. “Tests of Column—Flat Plate Connections,”
16.137. Adel A. Elgabry and Amin Ghali. “Transfer of Moments between Columns and Slabs: Proposed
Code Revisions,” ACI Structural Journal, 93, January–February 1996, 56–61.
16.138. Homayoun H. Abrishami, William D. Cook, and Denis Mitchell. “The Effect of Epoxy-Coated
Reinforcement and Concrete Quality on Cracking of Flat Plate Slab–Column Connections,”
ACI Materials Journal, 93, March–April 1996, 121–128.
16.139. ACI Committee 340. Supplement to Design Handbook Volume 1, Design of Two-
Way
Slabs, in Accordance with the Strength Design Method of ACI 318–83. [SP-17(84)(S)].
Detroit: American Concrete Institute, 1985.
Other References
16.140. Y. H. Luo and A. J. Durrani. “Equivalent Beam Model for Flat-Slab Buildings—Part I,” ACI
16.141. Y. H. Luo and A. J. Durrani. “Equivalent Beam Model for Flat-Slab Buildings—Part II,” ACI
Structural Journal, 92, March–April 1995, 250–257.
16.142. N. J. Gardner and Xiao-yun Shao. “Punching Shear of Continuous Flat Reinforced Concrete
Slabs,” ACI Structural Journal, 93, March–April 1996, 218–228.
16.143. Adel A. Elgabry and Amin Ghali. “Moment Transfer by Shear in Slab–Column Connections,”
16.144. Amin Ghali. “An Efficient Solution to Punching of Slabs,” Concrete International, 11, June
1989, 50–54.
718
16.145. Shyh-Jiann Hwang and Jack P. Moehle. “Models for Laterally Loaded Slab-Column Frames,”
16.146. Laurel M. Dovich and James K. Wight. “Effective Slab Width Model for Seismic Analysis of
Flat Slab Frames,” ACI Structural Journal, 102, November–December 2005, 868–875.
16.147. AISC. Manual of Steel Construction (14th ed.). Chicago: American Institute of Steel
Construction, 2016.
16.148. T. X. Dam, J. K. Wight, and G. J. Parra-Montesinos, “Behavior of Monotonically Loaded
Slab-Column Connections Reinforced with Shear Studs,” ACI Structural Journal, Vol. 114,
No. 1, 2017, pp. 221–232.
PROBLEMS
Two-Way Slabs (With Beams)
16.1 Design the typical interior frame along columns 16.2 Compute the moments for the column and mid-
2-5-7 for the two-way slab system shown in the dle strips, as well as the moments carried by the
figure for Problem 16.1. The 13-ft-long columns beams for the interior frame of Problem 16.1,
are connected by beams, and no column capi- except that a 12-in. wall exists at the lower-story
tals or drop panels are used. As an initial trial, level and contains the 24-in.-square columns at
assume that all beams (interior) are 12 × 24 in. locations 1, 2, 3, 4, 6, and 10.
overall. Revise beam size as necessary during 16.3 Compute the moments for the column and mid-
the design. Determine slab thickness based on dle strips, as well as the moments carried by the
ACI-8.3.1; then use the Direct Design Method beams for the typical interior frame along col-
for longitudinal distribution of moments. Show umn lines 4-5-8 for the two-way slab system of
a design sketch giving all your decisions, Problem 16.1.
including dimensions, bar sizes, and stirrups 16.4 Compute the moments for the column and middle
for the two spans from column 2 to column 7. strips, as well as the moments carried by the beams
The live load is 150 psf, fc′ = 4000 psi, and fy = for the exterior half- frame along column lines
60,000 psi. 1-2-3 for the two-way slab system of Problem 16.1.
Parapet extends
from columns
1 to 10 and from
columns 1 to 3
10”
10
All columns 2’– 0

27’– 0”
24 × 24
6 7 9
A A Columns
27’– 0”
Section A–A
B D above
and below
4 5 8
Crosshatch indicates
27’–0”
A C column only
below the slab
1 2 3
23’– 0” 23’– 0” 23’– 0”

719
PROBLEMS 719
Flat Slabs on a 24-in. square column, along with a 7 ft 8 in.

width of drop panel that is 11 in. thick. The slab
16.5 In the flat slab shown in the figure for Problem is 8 1 2 in. thick. Assume that there is no edge
16.5, the columns are 24 in. square with col- beam or wall at the exterior support location.
umns 1 through 6 existing only below the floor The factored moment Mu to be transferred is
slab, while columns 7 through 9 exist both above 340 ft-kips and the factored shear Vu is 115 kips.
and below the floor slab. All columns are 13 ft The negative moment reinforcement provided
long center-to-center of floor slabs. The live load in the column strip is #5 at 10-in. spacing. Use
is 150 psf, fc′ = 4000 psi , and fy = 60,000 psi. fc′ = 4000 psi and fy = 60,000 psi.
Use rectangular column capitals and drop panels 16.7 Rework Example 16.10.2, assuming that the
to design the flat slab. service live load is 200 psf instead of 140 psf.
(a) Determine the slab thickness based on 16.8 Rework Example 16.13.2, assuming that the
ACI-8.3.1. service live load is 200 psf instead of 140 psf.
(b) Use the Direct Design Method for longi-
tudinal distribution of moments in interior Flat Plates
equivalent frame defined by columns 2, 5,
and 7 along its centerline. 16.9 Design a typical interior rigid frame (long direc-
(c) Determine transverse distribution and select tion) of a flat plate floor using the data for the
reinforcement for the column strip (defined design example of Section 16.3, assuming that
by columns 2, 5, and 7) and adjacent half the service live load is 120 psf and the columns
middle strips. are 16 × 14 in.
16.10 Assuming that the slab thickness is 6 1 2 in.,
16.6 Investigate the moment and shear transfer at investigate the moment and shear transfer in the
the exterior support of a flat slab structure as flat plate design example (see Examples 16.17.1
detailed in the figure for Problem 16.6. The exte- and 16.18.1) for a service live load of 150 psf
rior support has a 5-ft flat-sided column capital instead of 72 psf.
Columns above
and below
4’– 10”
6 7 9 Live load = 150 psf
3’–6”
fc’ = 4000 psi
fy = 60,000 psi 2’–0”
27’–0”
8 2”
1
All columns: 11”
24” square
13’– 0” long
4 5 8
7’–8”
5’–0”
Crosshatch indicates
column below slab only
27’–0”
A A
Section A–A Section A–A
1 2 3
23’– 0” 23’–0” A A
CHAPTER 17
YIELD LINE THEORY
OF SLABS
17.1 INTRODUCTION
Reinforced concrete design methods under the present ACI Code are based on the results
of an elastic analysis of the structure as a whole, when subjected to the action of factored
loads, such as 1.2D + 1.6L, where D and L refer to service dead and live loads. Actually,
the behavior of a statically indeterminate structure is such that inelastic behavior develops
at one or more regions along the structural members and, consequently, the results of an
elastic analysis are no longer valid. These regions of concentrated inelastic behavior are
commonly called “plastic hinges.” If sufficient ductility exists, redistribution of bending
moments will occur until a sufficient number of plastic hinges has formed to change the
structure into a mechanism, at which time the structure collapses or fails. It is then said that
the structure has reached a “collapse mechanism.” The term “ultimate load analysis,” as
opposed to “elastic analysis,” relates to the use of the bending moment diagram at the verge
of collapse as the basis for design. Other than the provisions for redistribution of moments
at the supports of continuous flexural members (ACI-6.6.5), the present ACI Code has no
explicit allowance for ultimate load analysis. The redistribution as described in ACI-6.6.5
has been presented and illustrated in Section 9.12.
The design and analysis of two-way slab systems was treated in Chapter 16. The fac-
tored moments are based on the elastic analysis of an equivalent frame, which has been
devised as a simple substitute for the elastic analysis of a three-dimensional system.
The chief concern of this chapter is to develop the yield line theory for two-way slabs.
Although not included in the ACI Code, slab analysis by yield line theory may be useful
in providing the needed information for understanding the behavior of irregular or single-
panel slabs with various boundary conditions.
17.2 GENERAL CONCEPT
Although the study of flexural behavior of plates up to the ultimate load may date back
to the 1920s [17.1], the fundamental concept of the yield line theory for the ultimate load
design of slabs was expanded considerably by K. W. Johansen [17.2, 17.3]. In this theory
the strength of a slab is assumed to be governed by flexure alone; other effects such as
shear and deflection are to be separately considered. At collapse, it is assumed that the steel
reinforcement has fully yielded and that the bending and twisting moments are uniformly
distributed along the yield lines.
721
17.2 GENERAL CONCEPT 721
Flat plate building, Chicago. (Photo by Gustavo J. Parra-Montesinos.)
Yield line theory for one-way slabs is not much different from the limit analysis of con-
tinuous beams. On a continuous beam, achievement of flexural strength at one location—
say, in the negative moment region over a support—does not necessarily mean that the
ultimate load on the beam has been reached. If the section, having reached its flexural
strength, can continue to provide a constant resistance while undergoing further rotation,
then the flexural strength may be reached at additional locations. Complete failure the-
oretically cannot occur until yielding has occurred at several locations (or along several
parallel lines in the case of one-way slabs), so that a mechanism forms, giving a condition
of unstable equilibrium.
Consider, for example, the one-way slab of finite width shown in Fig. 17.2.1. A uniform
loading on the slab will cause uniform maximum negative bending moment along AB and
EF and uniform positive bending moment along CD, which is parallel to the supports.
When the uniform load is increased until the moments along AB, CD, and EF reach
their respective ultimate moment capacities, rotation of the slab segments will occur, with
A C Free edge E
Yield lines
B D Free edge F
AB EF
CD
Figure 17.2.1 Collapse mechanism of a one-way slab.

72
722 C H A P T E R 1 7 Y ield L ine T heory of S labs
yield lines AB, CD, and EF acting as axes of rotation. Assuming an elastic–plastic moment–
curvature relationship (see Fig. 9.12.1), angle change can occur without an increase in the
resisting moment. Thus, under the limiting condition with the slab segments able to rotate
with no change in resisting moment, the slab system is geometrically unstable. This condi-
tion is known as a collapse mechanism.
Yield line theory for two-way slabs requires a different treatment from limit analysis of
continuous beams, because in this case the yield lines will not in general be parallel to each
other but instead will form a yield line pattern. The entire slab area will be divided into
several segments, which can rotate along the yield lines as rigid bodies at the condition of
collapse or unstable equilibrium. Some yield line patterns for typical situations are shown
in Fig. 17.2.2.
The slab of Fig. 17.2.2(a) has nonparallel supports. At the collapse condition, this slab
will break into two segments: one will have an edge rotating about line I and the other will
have an edge rotating about line II. The positive moment yield line must then intersect lines
I and II at their intersection, point 0. The exact position of yield line III will depend on the
reinforcement amount and direction, in both the positive and negative moment regions.
For the case of Fig. 17.2.2(b) where a rectangular panel is either simply supported or
continuous over four linear supports, the collapse mechanism consists of four slab seg-
ments. The exact locations of points a and b will depend on the moment strengths at the
supports and the positive moment reinforcement in each direction.
The slab in Fig. 17.2.2(c) is supported along two edges and, in addition, is supported
by two isolated columns. The rotational axes for the slab segments at collapse must occur
along the supports (lines I and II), and additional rotational axes must pass through the
isolated columns. The critical position of the positive moment yield lines a, b, c, d, and e
is a function of the reinforcement amount and direction; in the meantime, compatibility of
deflection along the yield lines must be maintained during the rigid body rotations of the
slab segments.
For a concentrated load at a significant distance from a supported edge, the yield line
pattern will be circular, as shown in Fig. 17.2.2(d). The circle pattern will be a yield line of
negative bending moment, while the radial yield lines are due to positive bending moment.
For concentrated loads near a free edge, a fan or partial circular pattern is typical.
Supported edges
Free edge II
a b
Supported
edges e IV
III
c
d
Columns
Free edge
I II
I III
0 (a) (c)
Four supported edges

Negative
P moment
a b yield line
Positive
moment
yield line
(b) (d)
Figure 17.2.2 Typical yield line patterns.

723
1 7 . 3 F U N DA M E N TA L A S S U M P T I O N S 723
17.3 FUNDAMENTAL ASSUMPTIONS
In applying the yield line theory to the ultimate load analysis of reinforced concrete slabs,
the following fundamental assumptions are made.
1. The steel reinforcement is fully yielded along the yield lines at failure. In the usual
case, when the slab reinforcement is well below that in the balanced condition, the
moment-curvature relationship [17.4] is as shown in Fig. 17.3.1.
2. The slab deforms plastically at failure and is separated into segments by the
yield lines.
3. The bending and twisting moments are uniformly distributed along the yield line and
correspond to the maximum values provided by the flexural strengths in two orthog-
onal directions (for two-way slabs).
4. The elastic deformations are negligible compared with the plastic deformations; thus
the slab parts rotate as plane segments in the collapse condition.
Assumption No. 3 may be considered to be the yield criterion of orthotropic reinforced

concrete slabs. It means that along a yield line as shown in Fig. 17.3.2, the bending moment
strength Mnb and twisting moment strength Mnt, each per unit distance along the yield line,
are exactly equal to the sum of the projected components along the yield line of the moment
strengths Mnx and Mny per unit distance in the y- and x-directions, respectively. It may be
noted that Mnx is the strength contributed by the reinforcement in the x-direction, and Mny is
the strength contributed by the reinforcement in the y-direction. Also, the sign convention
is that the bending moments Mnx, Mny, and Mnb are positive when they induce tension in the
lower portion of the slab, and the twisting moment Mnt is positive if its vector is directed
away from the free body on which it acts.
Typical
Idealized
Moment, M
Elastic deformation
Plastic deformation
Curvature, ϕ
Figure 17.3.1 Typical and idealized M-ϕ relationship for reinforced concrete slab.
Steel for
y Mny per unit width
L M nb
Steel for
Mnx per unit width M nt
line
L sin θ
ld
Mnx
Yie
x Mny
L cos θ
Figure 17.3.2 Bending and twisting moments on yield line.

724
The bending moment strength Mnb and twisting moment strength Mnt along the yield line
in Fig. 17.3.2 may be expressed in terms of Mnx and Mny. Taking equilibrium of moment
vectors parallel to the yield line,
M nb ( L ) = M nx ( L sinθ)sin θ + M ny ( L cos θ) cos θ

M nb = M nx sin 2 θ + M ny cos2 θ
M nx + M ny M nx − M ny
M nb = − cos 2θ (17.3.1)
2 2
and, taking equilibrium of moment vectors perpendicular to the yield line,
M nt ( L ) = M nx ( L sin θ) cos θ − M ny ( L cos θ)sin θ

M nt = ( M nx − M ny )sin θ cos θ
( M nx − M ny )
= sin 2θ (17.3.2)
2
In using Eqs. (17.3.1) and (17.3.2), it is important to note that θ is the counterclockwise
angle measured from the positive x-axis to the yield line.
17.4 METHODS OF ANALYSIS
There are two methods of yield line analysis of slabs: the virtual work method and the equi-
librium method. Based on the same fundamental assumptions, the two methods should give
exactly the same results. In either method, a yield line pattern must be first assumed so that
a collapse mechanism is produced. For a collapse mechanism, rigid body movements of the
slab segments are possible by rotation along the yield lines while deflection compatibility is
maintained at the yield lines between slab segments. There may be more than one possible
yield line pattern, in which case solutions to all possible yield line patterns must be sought;
the one giving the smallest ultimate load would actually happen and thus should be used in
design. For instance, the failure pattern of the simply supported rectangular slab subjected
to uniform load may be that shown either in Fig. 17.4.1(a) or in 17.4.1(b), depending on the
aspect ratio of the rectangular panel and the moment strengths Mnx and Mny.
After a yield line pattern has been assumed, the next step is to determine the position of
the yield lines, such as that defined by the unknown x in Fig. 17.4.1(a) or 17.4.1(b). It is at
this point that one may choose to use the virtual work method or the equilibrium method.
In the virtual work method, the total external work done by the applied loads during simul-
taneous rigid body rotations of the slab segments (while maintaining deflection compatibil-
ity) must balance the total internal work done by the bending and twisting moments on all
the yield lines. The value of x that gives the smallest ultimate load is then found by means
of differential calculus. In the equilibrium method, the value of x is obtained by applying
a b c
d
x
(a) (b)
Figure 17.4.1 Yield line patterns of a simply supported rectangular slab.

725
1 7 . 5 Y I E L D L I N E A N A LY S I S O F O N E - WAY S L A B S 725
the usual equations of statical equilibrium to the slab segments, but the optimal position x is
defined by predetermined nodal forces placed at the intersection of yield lines. Expressions
for the nodal forces in typical situations, once derived, can be conveniently used to avoid
the necessity of mathematical differentiation, as required in the virtual work method.
In the following sections, yield line analysis for one-way slabs is dealt with first in a
manner similar to limit analysis of continuous beams. Then both the virtual work method
and the equilibrium method are illustrated for two-way slabs.
17.5 YIELD LINE ANALYSIS OF ONE-W AY SLABS

The continuous slab span shown in Fig. 17.5.1(a) has nominal moment strengths of Mni
and Mnj provided by top reinforcement at the supports and a nominal strength Mnp pro-
vided by bottom reinforcement within the span. The nominal strengths Mni, Mnj, and Mnp
are absolute values of the moment strength per unit slab width. For uniform loading, the
only possible yield pattern consists of three parallel yield lines as shown in Fig. 17.5.1(a),
one each along the left and right supports and one at a distance x from the left support.
The moment strength per unit slab width when collapse is imminent should be as shown
in Fig. 17.5.1(b). The problem is to determine the collapse load wu /φ per unit slab area in
terms of Mni, Mnj, and Mnp, and the span length L. It will be shown that the virtual work and
equilibrium methods will give exactly the same results.
Referring to Fig. 17.5.1(c), the rigid body rotations of the slab segments at collapse
are measured from the original horizontal positions to those indicated by the dashed lines,
where the rotation of the left segment is θ1 in the clockwise direction and that of the right
segment is θ2 in the counterclockwise direction, all the while maintaining the compatible
deflection Δ at the positive yield line. The total external work done by the uniform load on
the left and right segments of unit slab width is
∆ ∆
total external work = (wu /φ) x + (wu /φ)( L − x )
2 2
The total internal work done by Mni and Mnp on the left segment is (Mni + Mnp) θ1 and that
done by Mnj and Mnp on the right segment is (Mnj + Mnp) θ2, both in absolute quantities. Thus
total internal work = ( M ni + M np )θ1 + ( M nj + M np )θ2
in which, for small deformations

∆ ∆
θ1 = , θ2 =
x L − x
The principle of virtual work states that the total work done by a force system in equilib-
rium in going through a virtual rigid body displacement is zero. By means of this principle,
one can equate the total external work to the absolute value of the total internal work; or
∆ ∆ ∆ ∆
(wu / φ) x + (wu / φ)( L − x ) = ( M ni + M np ) + ( M nj + M np )
2 2 x ( L − x )
Dividing out the compatible deflection Δ from the above equation and solving for wu /φ,
wu 2 M ni 2 M np 2 M nj
= + + (17.5.1)
φ L x x( L − x ) L ( L − x )
Differentiating Eq. (17.5.1) for wu with respect to x and setting the derivative to zero, one
obtains the quadratic equation
( M nj − M ni ) x 2 + 2( M ni + M np ) L x − ( M ni + M np )L2 = 0 (17.5.2)
726
x
Positive moment yield line Free edge
Top steel, Mni Top steel, Mnj
Negative Bottom steel, Mnp Negative

moment moment
yield line Free edge yield line
L
(a) Plan of continuous one-way slab
Mnp
Mni
Mnj
(b) Moment strength per unit slab width when collapse condition is imminent
wu /φ per unit distance
Mni Mnj
θ1 θ2
θ1 θ2 wu
(L – x)
wu φ
x
φ Mnp Mnp
V V
(c) Virtual work method
wu/φ Mnp Mnp wu/φ

Mni Mnj
V=0 V=0
1 wu Mnj – Mni 1 wu Mnj – Mni
Ri = L– Nodal forces Rj = L+
2 φ L 2 φ L
(d) Equilibrium method
Figure 17.5.1 Yield line analysis of one-way slabs.
In the virtual work method, then, first the value of x is found by solving the quadratic equa-
tion (17.5.2) and the collapse load wu /φ is determined from Eq. (17.5.1).
In the equilibrium method, the value of x is not to be obtained by differential calculus.
Instead, it is defined by the magnitude of the nodal forces [V and –V shown in Fig. 17.5.1(d)]
acting on the slab segments on either side of the positive yield line. In this particular instance,
it is known from elementary beam theory that the shear at a section of maximum positive
bending moment should be zero. From this point on, the unknown values of Ri, Rj, wu, and x
in Fig. 17.5.1(d) are obtained from the four independent equilibrium equations for the entire
slab and either of the two slab segments. The left and right reactions on the free bodies of
Fig. 17.5.1(d) are found by applying the equilibrium equations to the entire slab; thus
1  w  M nj − M ni  1  w  M nj − M ni 
Ri =   u  L − , Rj =   u  L + 
2  φ  L  2  φ  L 
72
1 7 . 5 Y I E L D L I N E A N A LY S I S O F O N E - WAY S L A B S 727
Then, summing the vertical forces on the left slab segment,
1  wu  M nj − M ni wu
Ri =   L− = x
2 φ  L φ
from which
wu ( M nj − M ni ) (17.5.3)
=
φ L ( L /2 − x )
and, taking moments about the positive moment yield line on the left segment,
1  w  M nj − M ni  1  wu  2
M np = − M ni +   u  L − x−   x (17.5.4)
2  φ  L  2 φ 
Substitution of Eq. (17.5.3) into Eq. (17.5.4) results in the same quadratic equation as
Eq. (17.5.2). This shows that the two methods, virtual work and equilibrium, give exactly
the same results.
In the solution of a numerical problem, the quadratic equation (17.5.2) is first solved
for x. Although wu can then be computed from either Eq. (17.5.1) or Eq. (17.5.3), using
Eq. (17.5.1) is far superior to Eq. (17.5.3), because Eq. (17.5.1) contains the sum of
three absolute quantities but Eq. (17.5.3) involves the division by a sensitive quantity
(L/2 – x).
EXAMPLE 17.5.1
Given the nominal moment strengths Mni = 14 ft-kips, Mnp = 16 ft-kips, and Mnj = 22 ft-
kips per ft width of a 20-ft continuous slab span as shown in Fig. 17.5.2, determine the
location x of the positive moment yield line and the collapse load wu /φ in kips per ft per
foot width of slab.
SOLUTION
Using the quadratic equation (17.5.2),
( M nj − M ni ) x 2 + 2( M ni + M np )L x − ( M ni + M np )L2 = 0
(22 − 14) x 2 + 2(14 + 16)(200 x ) − (14 + 16)(20)2 = 0
8 x 2 + 1200 x − 12, 000 = 0

x 2 + 150 x + 752 = 1500 + 5625
x = 7125 − 75 = 84.41 − 75 = 9.41 ft
Substituting the value of x = 9.41 ft in Eq. (17.5.1),
wu 2 M ni 2 M np 2 M nj
= + +
φ L x x( L − x ) L ( L − x )
2(14) 2(16) 2(22)
= + +
20(9.41) 9.41(10.59) 20(10.59)
= 0.149 + 0.321 + 0.208 = 0.678 kip/ft
(Continued)
728
Using wu /φ = 0.678 kip/ft and the end moments of 14 ft-kips and 22 ft-kips, the end
reactions, the shear diagram, and the moment diagram are computed and shown in
Fig. 17.5.2.
Mnp = 16 ft-k/ft
Mni = 14 ft-k/ft Mnj = 22 ft-k/ft

1.0’
20’
14 ft-k 22 ft-k
wu/φ = 0.678 k/ft
+6.78 +6.78
–0.40 +0.40
+6.38 +7.18
+6.38k
Shear
diagram
at collapse
condition
x = 9.41’
–7.18k
16 ft-k Moment
diagram
+ at collapse
– – condition
14 ft-k
22 ft-k
Figure 17.5.2 One-way slab for Example 17.5.1.
17.6 WORK DONE BY YIELD LINE MOMENTS IN RIGID

BODY ROTATION OF SLAB SEGMENT
Before taking up the virtual work method of yield line analysis of two-way slabs, it is
desirable to derive a general procedure for obtaining the absolute value of the internal work
done by the bending and twisting moments acting on the yield line through rigid body rota-
tion of the slab segment. Figure 17.6.1(a) shows the moments acting on the edges of a slab
segment having yield line of length L with horizontal and vertical projections of Lx and Ly,
respectively. Let this slab undergo a rigid body rotation, whose components are θx and θy,
shown in vector notations in Fig. 17.6.1(b). It can be shown algebraically that the absolute
729
1 7 . 7 N O DA L F O R C E S AT I N T E R S E C T I O N W I T H F R E E E D G E 729
L
Mnt (L)
θx
Mnx (Ly )
Mnb (L) Ly
Yield line
θ
θy
Mny (Lx)
Lx
(a) Total moments on edges (b) Right body rotation
Figure 17.6.1 Work done by yield line moments in rigid body rotation of a slab segment.
value of the work done by the moments (Mnb L) and (Mnt L) acting on the yield line is equal
to the work done by the moments (Mnx Ly) and (Mny Lx) acting on the horizontal and vertical
projections of the yield line—for the same rigid body rotation, of course. This is obviously
correct because the moments (Mnx Ly) and (Mny Lx) on Fig. 17.6.1(a) are the equilibrants of
the moments (Mnb L) and (Mnt L) on the same figure. Certainly then, the work done by one
set of generalized forces should be equal, numerically, to the work done by the alternative
set of equilibrating generalized forces.
17.7 NODAL FORCES AT INTERSECTION OF YIELD

LINE WITH FREE EDGE
In the equilibrium method of yield line analysis, the position of a yield line is characterized
by the insertion of nodal forces at the intersection of a yield line with another yield line,
or of a yield line with a free edge. Section 17.5 demonstrated that for one-way slabs, on
the basis of elementary beam theory, the nodal force on either side of a positive moment
yield line is zero. This section derives the expression for the pair of equal and opposite
nodal forces V acting on each side of the intersection of a yield line with a free edge in a
two-way slab.
Shown in Fig. 17.7.1(a) is a two-way slab with nominal moment strengths of Mnx and
Mny provided by the bottom reinforcement in the x- and y-directions, respectively. A posi-
tive moment yield line is assumed to intersect the free edge at an angle α. The upward nodal
force V acts on the left segment and the downward nodal force V acts on the right segment,
shown by a dot and a cross for the upward and downward forces in Fig. 17.7.1(a) according
to the convention used by Johansen [17.2, 17.3]. For equilibrium the upward and downward
nodal forces must be equal in magnitude.
Consider the equilibrium of an infinitesimal slab element shown in either Fig. 17.7.1(b)
or 17.1.1(c). By its own definition, a yield line is always at the optimal position. Inasmuch
as the edge ac in Fig. 17.7.1(b), or the edge ad in Fig. 17.7.1(c), is at an infinitesimal
angle ∆α from the yield line, the same yield line moments as those on the yield line ab
act on edges ac or ad. By applying the principle of equivalent force systems as indi-
cated by Fig. 17.6.1(a), the free-body diagram shown on the left side of the equal sign in
Fig. 17.7.1(b) or 17.7.1(c) is transformed to the equivalent free-body diagram on the right
side of the equal sign. Using the equivalent free body and summing the moments about the
line ac in Fig. 17.7.1(b),
M ny ( ∆L x ) cos(α − ∆α ) = V ( ∆L x ) sin(α − ∆α ) (17.7.1)
and summing the moments about the line ad in Fig. 17.7.1(c),
M ny ( ∆L x ) cos(α + ∆α ) = V ( ∆L x ) sin(α + ∆α ) (17.7.2)

730
Bottom reinforcement for Mny

Bottom reinforcement
for Mnx
•Up α
+Down V +V V = Mny cot α
Free edge
(a) Two-way slab with free edge
a a
Mnt
Mnx Ly
Mnx Ly
Ly
Mnb
α Mny Lx
V Mnb Mnt V
c c
b b Lx
(b) Free-body diagram of inﬁnitesimal piece to left of yield line
Mny Lx a
a
Mnt Mnb
Mnx Ly
Ly
Mnx Ly
+V Mnt +V
Mnb b
b d d
Lx
(c) Free-body diagram of inﬁnitesimal piece to right of yield line
Figure 17.7.1 Nodal force at intersection of yield line with free edge.
Solving for V by using either Eq. (17.7.1) or Eq. (17.7.2),
V = M ny cot α (17.7.3)
Note that on the left side of Eq. (17.7.1) or Eq. (17.7.2), Mny(∆Lx) represents the net moment
along the edges ca and ab, or the edges ba and ad, respectively; while on the right side, the
moment of the nodal force V about an axis coincident with ac or ad is involved.
Equation (17.7.3) is used when applying the equilibrium method to a situation where a
positive moment yield line intersects a free edge of a slab at an angle other than 90°. This
is illustrated by the following example.
EXAMPLE 17.7.1
The triangular slab ABC shown in Fig. 17.7.2(a) is uniformly loaded and is simply sup-
ported along edges AC and BC but has a free edge along AB. The reinforcement in the
x- and y-directions in the lower face of the slab provides nominal moment strengths
Mnx = 8 ft-kips and Mny = 10 ft-kips, each per foot width of slab. Determine the yield line
pattern and the collapse uniform load wu /φ.
(Continued)
731
1 7 . 7 N O DA L F O R C E S AT I N T E R S E C T I O N W I T H F R E E E D G E 731
C
Simply Bottom reinforcement
supported for Mnx = 8 ft-k/ft
edges
12’
A B Bottom reinforcement
Free edge for Mny = 10 ft-k/ft
16’
(a) Uniformly loaded triangular slab

C 10x ft-k C
96 ft-k
96 ft-k
E
V α
θ • +V
D
A D 10x ft-k x B
(b) Free-body diagrams
Figure 17.7.2 Yield line analysis of the triangular slab of Example 17.7.1.
SOLUTION
(a) Yield line pattern. It was shown in Fig. 17.2.2(a) that a yield line should pass through
the point of intersection of two nonparallel supported edges. In this example, a posi
tive moment yield line CD will divide the slab into two segments ACD and BCD,
wherein a common compatible deflection Δ exists at point D due to the rigid body
rotations of the slab segment ACD about the supported edge AC and of the slab seg-
ment BCD about the supported edge BC.
(b) Equilibrium method of finding BD = x. The nodal forces V and –V acting on the left
and right slab segments of Fig. 17.7.2(b) are computed from Eq. (17.7.3); thus
 x 5
V = M ny cot α = (+10)   = x kips
 12  6
Note that the nodal force acts upward in the obtuse angle and it acts downward in
the acute angle. Equilibrium of moments about the edge AC of the left segment
requires
1 1
(wu /φ)(16 − x )(12) ( DE ) = V ( DE ) + 8(12)sin θ + 10 x cos θ
2 3
1  1 5
(wu /φ)(16 − x )(12)   (16 − x )(0.6) = x(16 − x )(0.6) + 96(0.6) + 10 x(0.8
8)
2  3 6
from which
wu 576 + 160 x − 5 x 2
=
φ 12(16 − x )2
(Continued)
732
Similarly, the equilibrium of moments about the edge BC of the right segment requires
1  BD 
(wu / φ)( x )(12)  + V ( BD) = 96
2  3 

1  wu   x  5x
2
  ( x )(12)   + = 96
2 φ   3 6
from which
wu 576 − 5 x 2
=
φ 12 x 2
Equating the two expressions for w u /φ,
576 + 160 x − 5 x 2 576 − 5 x 2

=
12(16 − x )2 12 x 2
x 2 + 14.4 x − 115.2 = 0

x = 5.72 ft
With the known value of x, the value of wu /φ may be computed,
wu
= 1.05 kips/sq ft
φ
It can be shown that the same quadratic equation in x may be obtained via the virtual
work method by following the procedure illustrated in Section 17.5.
17.8 NODAL FORCES AT INTERSECTION OF THREE

YIELD LINES
Figure 17.8.1 shows three possible yield line patterns for an irregular quadrilateral slab
with four simply supported edges. The optimum positions of the yield lines in each pat-
tern should be such as to give the lowest collapse load. These positions are defined by the
locations of points a and b in Fig. 17.8.1(a), of points c and d in Fig. 17.8.1(b), and of
point e in Fig. 17.8.1(c). In the equilibrium method, the yield lines are characterized by the
insertion of predetermined nodal forces. This section derives the expressions for the nodal
forces at the intersection of three yield lines, such as at points a, b, c, or d in Fig. 17.8.1.
A B A B A B
b c d
a
e
D D
D
C C C
(a) (b) (c)
Figure 17.8.1 Yield line patterns of a simply supported quadrilateral slab.

73
1 7 . 8 N O D A L F O R C E S AT I N T E R S E C T I O N O F T H R E E Y I E L D L I N E S 733
2
φ3
1
V23 φ2
Bottom reinforcement Bottom reinforcement
V12
φ1 for Mnx1 – Mnx2 – Mnx3 for Mny1 – Mny2 – Mny3
0 x under yield lines 1-2-3 under yield lines 1-2-3
V31
(a) (b) (c)
Figure 17.8.2 Nodal forces at the intersection of three yield lines.
The derivation shown below follows the works of Johansen [17.2, 17.3], Jones and Wood
[17.4], and Wood [17.5].
Assume that three yield lines 1–2–3, intersecting at a common point 0, are situated
at angles φ1–φ2–φ3 measured counterclockwise from the positive x-axis, as shown in
Fig. 17.8.2(a). The nominal moment strengths under the yield lines 1–2–3 are, as shown in
Fig. 17.8.2(b) and 17.8.2(c), Mnx1–Mnx2–Mnx3 provided by the reinforcement shown horizon-
tally and Mny1–Mny2–Mny3 provided by the reinforcement shown vertically, all reinforcement
being near the lower face of the slab. The nodal forces V12, V23, and V31 are shown by the
dots (which mean upward nodal forces) in Fig. 17.8.2(a). Note that for vertical equilibrium
at the point of intersection, the sum of V12, V23, and V31 must be zero.
It is important to emphasize at the outset that, by definition, the yield lines 1–2–3 are all
situated at their respective optimal positions. The bending and twisting moments on any
line passing through the point of intersection and deviating by an infinitesimal angle from
a particular yield line should be equal to the bending and twisting moments on that line as
provided by the orthogonal moment strengths.
Consider the equilibrium of an infinitesimal slab segment 0AB [Fig. 17.8.3(a)], bounded
by a differential length 0A on yield line 3 and an arbitrary length 0B on yield line 1. Since
BA is at a differential angle ∆α from B0, the bending and twisting moments on both B0 and
BA are those provided by Mny1 and Mnx1 on their respective horizontal and vertical projec-
tions. Likewise the bending and twisting moments on the differential length 0A are those
provided by Mny3 and Mnx3 on its horizontal and vertical projections. Call the upward nodal
force at point 0 inside triangle B0A and the force bounded by yield line 1 and a different
length on yield line 3 by the name V1–∆3.
Next, write the equation of equilibrium for moments about AB as the axis of rotation for
the slab segment 0AB in Fig. 17.8.3(a), noting that the moment of the uniform load on 0AB
about any axis is a differential of the second order and may be neglected.
− V1− ∆ 3 0 A sin(φ3 − φ1 + ∆α )
+ ( M ny 3 − M ny1 ) 0 A cos φ3 cos(φ1 − ∆α )

+( M nx 3 − M nx1 ) 0 A sin φ3 sin(φ1 − ∆α ) = 0
Solving the above equation for V1–∆3 and letting ∆α approach zero at the limit,
( M nx 3 − M nx1 )sin φ3 sin φ1 + ( M ny 3 − M ny1 ) cos φ3 cos φ1

V1− ∆ 3 = (17.8.1)
sin(φ3 − φ1 )
Making a similar analysis for the infinitesimal slab segment 0AC [Fig. 17.8.3(b)] bounded
by a differential length 0A on yield line 3 and an arbitrary length 0C on yield line 2 gives
the expression for the upward nodal force V2–∆3 in Fig. 17.8.3(b) as
( M nx 3 − M nx 2 )sin φ3 sin φ2 + ( M ny 3 − M ny 2 ) cos φ3 cos φ2

V2 − ∆ 3 = (17.8.2)
sin(φ3 − φ2 )
734
Figure 17.8.3 Determination of nodal force V12 between yield lines 1 and 2.
For vertical equilibrium at the point of intersection, the upward nodal force V∆3–1 in
Fig. 17.8.3(a) in the zone going counterclockwise from the differential length on yield line
3 to yield line 1 is
V∆ 3 −1 = −V1− ∆ 3 (17.8.3)
and, for the same reason, the upward nodal force V12 in Fig. 17.8.3(c) is
V12 = −V2 − ∆ 3 − V∆ 3 −1 (17.8.4)

735
1 7 . 8 N O D A L F O R C E S AT I N T E R S E C T I O N O F T H R E E Y I E L D L I N E S 735
Substitution of Eqs. (17.8.1), (17.8.2), and (17.8.3) into Eq. (17.8.4) gives

V12 =
sin(φ3 − φ1 )
− (17.8.5)
sin(φ3 − φ2 )
Replacing each numerical subscript in Eq. (17.8.5) by its successor in the cyclic order of
1–2–3–1 (counterclockwise around the point of intersection) and then once more in the
same manner, the following expressions for the upward nodal forces V23 and V31 as shown
in Fig. 17.8.2(a) are obtained.
( M nx1 − M nx 2 )sin φ1 sin φ2 + ( M ny1 − M ny 2 ) cos φ1 cos φ2

V23 =
sin(φ1 − φ2 )
( M nx1 − M nx 3 )sin φ1 sin φ3 + ( M ny1 − M ny 3 ) cos φ1 cos φ3
− (17.8.6)
sin(φ1 − φ3 )

V31 =
sin(φ2 − φ3 )
− (17.8.7)
sin(φ2 − φ1 )
Equations (17.8.5), (17.8.6), and (17.8.7) are expressions for the upward nodal forces at the
intersection of three yield lines.
Nodal Force at Intersection of Yield Line with Free Edge

The nodal forces at the intersection of a yield line with a free edge, as shown by Fig. 17.8.4,
may be obtained by substituting φ1 = 0, φ2 = α, φ3 = π, Mnx1 = Mnx3 = Mny1 = Mny3 = 0,
Mnx2 = Mnx, and Mny2 = Mny in Eqs. (17.8.5) to (17.8.7); thus
( − M ny 2 )( −1) cos α
V12 = − = − M ny cot α
sin(π − α )
( − M ny 2 )(+1) cos α
V23 = = + M ny cot α
sin(0 − α )
( M ny 2 ) cos α( −1) ( M ny 2 ) cos α(+1)
V31 = − =0
sin(α − π) sin(α − 0)
The above results check with the findings in Section 17.7.
Bottom reinforcement for Mnx
e
lin
i eld Bottom reinforcement for Mny
Y α
V23
V12
3 1
Free edge V12 = –Mny cot α
V31
V23 = +Mny cot α
Figure 17.8.4 Nodal forces at the intersection of a yield line with a free edge.

736
Nodal Forces at Intersection of Three Yield Lines Having Identical

Mnx and Mny Nominal Moment Strengths
From Eqs. (17.8.5) to (17.8.7) it can be observed that wherever the nominal moment
strengths under three intersecting yield lines are identical (i.e., Mnx1 = Mnx2 = Mnx3 = Mnx and
Mny1 = Mny2 = Mny3 = Mny), the nodal forces at the intersection are zero. This fact is of great
convenience when using the equilibrium method for the yield line analysis of two-way
rectangular slabs.
17.9 YIELD LINE ANALYSIS OF RECTANGULAR

TWO-W AY SLABS
A typical rectangular two-way slab panel, as shown in Fig. 17.9.1, has two-way reinforce-
ment within the panel near the bottom face providing positive moment nominal strengths
Mnpx and Mnpy, as well as two-way reinforcement along the edges near the top face provid-
ing negative moment nominal strengths Mnnx and Mnny; these strengths are per unit width
of slab. The uniform load to give the collapse condition based on the yield line theory may
be determined in terms of the sides a and b, and the absolute values of Mnpx, Mnpy, Mnnx,
and Mnny.
Yield Line Pattern

Three possible yield line patterns are shown in Fig. 17.9.2. There is no unknown position
in yield line pattern No. 1 of Fig. 17.9.2(a); consequently the nodal forces V need not be
predetermined, and their value is dictated by statics alone. The unknowns x and y in yield
line patterns Nos. 2 and 3 of Fig. 17.9.2(b) and 17.9.2(c) must be determined by means of
differential calculus in the virtual work method. For the equilibrium method, in this particu-
lar case, however, the nodal forces are all zero because the moment strengths under a set of
three intersecting yield lines are identical.
Capacity = Mnnx
Capacity = Mnpy
Edges supported
and restrained
b
Capacity = Mnny Capacity = Mnpx

a
(a) Dimensions (b) Top reinforcement (c) Bottom reinforcement
Figure 17.9.1 A rectangular two-way slab panel.
V
+
V V
+
V y
(a) Yield pattern No. 1 (b) Yield pattern No. 2 (c) Yield pattern No. 3
Figure 17.9.2 Yield line patterns for a rectangular two-way slab panel.

73
1 7 . 9 Y I E L D L I N E A N A LY S I S O F R E C TA N G U L A R T W O - WAY S L A B S 737
a
Mnpy
θ=
b
Mnpx
C
+V
Mnnx
Mnpy Mnpy
b D V V B D V
+ + θ=
V a
Mnpx
Mnpx
V
Mnpx
A A
Mnpy
Mnny
Figure 17.9.3 Analysis of yield line pattern No. 1.
Analysis for Yield Pattern No. 1

Assuming a vertical deflection Δ at the intersection of the diagonal yield lines in Fig. 17.9.3,
the deflection at the centroids of the four triangles A–B–C–D is ∆ /3. The work done at the
collapse condition by the uniform load is the product of the total load on the entire panel
and Δ /3; thus
wu  ∆ 
W= ab   (17.9.1)
φ  3
The work done by the yield moments on the boundaries of all four slab segments, referring
to Fig. 17.9.3, is
 2∆   2∆ 
W = 2( M nny + M npy )(a )   + 2( M nnx + M npx )(b)   (17.9.2)
 b   a 
Equating Eq. (17.9.1) to Eq. (17.9.2) and solving for wu,
wu  M nnx + M npx M nny + M npy 

= 12  +  (17.9.3)
φ  a2 b2
Alternatively, the same solution is obtained using the equilibrium method. Taking moments
about the lower edge of slab segment A in Fig. 17.9.3,
1  wu   b   b   b
a     + V   = ( M nny + M npy )(a ) (17.9.4)
2  φ   2   6   2
Taking moments about the left edge of slab segment D in Fig. 17.9.3,
1  wu   a   a   a
b     = ( M nnx + M npx )(b) + V   (17.9.5)
2  φ   2   6   2
Eliminating V between Eqs. (17.9.4) and (17.9.5) and solving for wu /φ yields the same
expression for wu /φ as Eq. (17.9.3).

Assuming a vertical deflection Δ at the two points of intersection of the yield lines in
Fig. 17.9.4, the work done at the collapse condition by the uniform load on the entire panel is
W = 2WD + 2WA1 + 4WA 2

1  w    ∆ w   b  ∆  1  w  b  ∆
= 2   u  bx    + 2  u  (a − 2 x )     + 4   u  x   
2  φ    3   φ  2  2  2  φ  2  3 
wu  ∆
= (3ab − 2bx ) (17.9.6)
φ  6 
738
The work done by the yield moments on the boundaries of all four slab segments is, refer-
ring to Fig. 17.9.4,
 2∆   ∆
W = 2( M nny + M npy )(a )   + 2( M nnx + M npx )(b)   (17.9.7)
 b   x
Equating Eq. (17.9.6) to Eq. (17.9.7) and solving for wu /φ,
wu 12[b ( M nnx + M npx ) + 2 ax ( M nny + M npy )]

2
= (17.9.8)
φ b 2 (3ax − 2 x 2 )
Setting to zero the derivative of Eq. (17.9.8) with respect to x gives the quadratic
equation in x,
4 a( M nny + M npy ) x 2 + 4b 2 ( M nnx + M npx ) x − 3ab 2 ( M nnx + M npx ) = 0 (17.9.9)
Using the equilibrium method with V = 0 because there are three intersecting yield lines,
and taking moments about the lower edge of slab segment A in Fig. 17.9.4,
1  w  b   b w  b  b
2   u  x    + u (a − 2 x )     = ( M nny + M npy )(a )
2  φ  2  6 φ  2  4
wu 24 a( M nny + M npy )
= 2 (17.9.10)
φ 2b x + 3b2 (a − 2 x )
Taking moments about the left edge of slab segment D in Fig. 17.9.4,
1  wu   x 
bx   = ( M nnx + M npx )(b)
2  φ   3 
wu 6( M nnx + M npx ) (17.9.11)
=
φ x2
Equating Eq. (17.9.10) to Eq. (17.9.11) gives the same quadratic equation in x as
Eq. (17.9.9).
The condition for x = a/2 in Eq. (17.9.9) can be shown to be
M nnx + M npx a2 a
= for x = (17.9.12)
M nny + M npy b2 2
which means that if the sum of the positive and negative moments contributed by the rein-
forcement in the a direction, each per unit width of slab, is equal to (a2/b2) times the sum of
the positive and negative moments contributed by the reinforcement in the b direction, each
per unit width of slab, yield pattern No. 1 prevails.
a Mnpy
θ=
b θ=
x
Mnpx
C Mnpy
Mnnx
Mnpy Mnpy
b D B D
Mnpx
Mnpx
Mnpx
A1 A1
A2 A2 A2 A2
x x Mnpy
Mnny
Figure 17.9.4 Analysis of yield line pattern No. 2.

739
The condition for x < a/2 in Eq. (17.9.9) can be shown to be

M nnx + M npx a2 a
< for x < (17.9.13)
M nny + M npy b2 2
which means that for yield pattern No. 2 to prevail, the sum of the positive and negative
moments contributed by the reinforcement in the a direction must be less than (a2 /b2) times
the sum of the positive and negative moments contributed by the reinforcement in the b
direction.

By interchanging the subscripts x and y as well as the quantities a and b in Eqs. (17.9.8),
(17.9.9), (17.9.10), and (17.9.11), the following equations applicable to yield line pattern
No. 3 are obtained. The quadratic equation in y (Fig. 17.9.2) is
4b( M nnx + M npx ) y 2 + 4 a 2 ( M nny + M npy ) y − 3ba 2 ( M nny + M npy ) = 0 (17.9.14)
Similarly, the expressions analogous to Eqs. (17.9.8), (17.9.10), and (17.9.11) for wu /φ in
terms of y are
wu 12  a ( M nny + M npy ) + 2by( M nnx + M npx )

2
= (17.9.15)
φ a 2 (3by − 2 y 2 )
wu 24b( M nnx + M npx )

= 2 (17.9.16)
φ 2 a y + 3a 2 ( b − 2 y )
wu 6( M nny + M npy ) (17.9.17)

=
φ y2
The condition for y < b/2 in Eq. (17.9.14) can be shown to be

M nnx + M npx a2 b
> for y < (17.9.18)
M nny + M npy b2 2
in which case yield pattern No. 3 will prevail.
EXAMPLE 17.9.1
Determine the controlling yield line pattern and the corresponding collapse uniform
load for a rectangular two-way slab panel with dimensions as shown in Fig. 17.9.5(a).
The slab has reinforcement in the top near the edges and in the bottom within the panel.
Obtain solutions for the following three cases:
1. M nnx + M npx = 6.25 ft-kips/ft M nny + M npy = 4 ft-kips/ft

2. M nnx + M npx = 2 ft-kips/ft M nny + M npy = 4 ft-kips/ft
3. M nnx + M npx = 8 ft-kips/ft M nny + M npy = 4 ft-kips/ft
SOLUTION
(a) Case 1. The controlling yield line pattern may be determined by comparing the ratio
of (Mnnx + Mnpx) to (Mnny + Mnpy) with the ratio of a2 to b2. In this case,
( M nnx + M npx ) 6.25 a 2 625
= = 1.5625, = = 1.5625
( M nny + M npy ) 4 b 2 400
(Continued)
740
Since the moment ratio is equal to the ratio of a2 to b2, the yield pattern is as shown in
Fig. 17.9.5(d). Then from Eq. (17.9.3),
wu  M nnx + M npx M nny + M npy 

= 12  + 
φ  a2 b2

 6.25 4 
= 12  + = 0.240 ksf
 625 400 
Figure 17.9.5 Rectangular two-way slab of Example 17.9.1.
(b) Case 2. The ratio of (Mnnx + Mnpx) to (Mnny + Mnpy) is, in this case,
( M nnx + M npx ) 2 a2
= = 0.5 < 2 = 1.5625
( M nny + M npy ) 4 b
The yield line pattern is as shown in Fig. 17.9.5(e). The quadratic equation (17.9.9) is
used to solve for x.
4 a( M nny + M npy ) x 2 + 4b 2 ( M nnx + M npx ) x − 3ab 2 ( M nnx + M npx ) = 0
4(25)(4) x 2 + 4(400)(2) x − 3(25)(400)(2) = 0

x 2 + 8 x − 150 = 0
a 
x = 166 − 4 = 8.884 ft <  = 12.5 ft  OK
2 
(Continued)
741
The same uniform load wu /φ is obtained from Eqs. (17.9.8), (17.9.10), or (17.9.11); the
fact that it is so serves as a check on the numerical computation.
wu 12[b ( M nnx + M npx ) + 2 ax( M nny + M npy )]

2
=
φ (
b 2 3ax − 2 x 2 )
12[ 400(2) + 2(25)(8.884)(4)]
= = 0.152 ksf
400[3(25)(8.884) − 2(8.884)2 ]
wu 24 a( M nny + M npy )
=
φ 2b2 x + 3b2 (a − 2 x )
24(25)(4)
= = 0.152 ksf
2(400)(8.884) + 3(400)[25 − 2(8.884)]
wu 6( M nnx + M npx ) 6( 2 )
= = = 0.152 ksf
φ x2 (8.884)2
(c) Case 3. The ratio of (Mnnx + Mnpx) to (Mnny + Mnpy) is, in this case,
( M nnx + M npx ) 8 a2
= = 2 > 2 = 1.5625
( M nny + M npy ) 4 b
The yield line pattern is as shown in Fig. 17.9.5(f). The quadratic equation (17.9.14) is
used to solve for y.
4b( M nnx + M npx ) y 2 + 4 a 2 ( M nny + M npy ) y − 3ba 2 ( M nny + M npy ) = 0
4(20)(8) y 2 + 4(625)(4) y − 3(20)(625)(4) = 0

8 y 2 + 125 y − 1875 = 0
b 
y = 9.375 ft <  = 10 ft  OK
 2 
The same uniform load wu /φ is obtained from Eqs. (17.9.15), (17.9.16), or (17.9.17); the
fact that it is so serves as a check on the numerical computation.
wu 12[ a ( M nny + M npy ) + 2by( M nnx + M npx )]

2
=
φ a 2 (3by − 2 y 2 )
12[625(4) + 2(20)(9.375)(8)]
= = 0.273 ksf
625[3(20)(9.375) − 2(9.375)2 ]
wu 24b( M nnx + M npx )
= 2
φ 2 a y + 3a 2 ( b − 2 y )
24(20)(8)
= = 0.273 ksf
2(625)(9.375) + 3(625)[20 − 2(9.375)]
wu 6( M nny + M npy ) 6(4)
= = = 0.273 ksf
φ y 2
(9.375)2
742
17.10 CORNER EFFECTS IN RECTANGULAR SLABS

In Section 17.4, it was stated that there may be more than one possible yield line pattern, in
which case solutions to all possible yield line patterns must be sought; the one giving the
smallest collapse load would actually happen and thus should be used in design. Although
the three typical yield line patterns for a rectangular two-way slab panel have been shown
in Fig. 17.9.2 and their analysis has been completely treated in Section 17.9, it can be
demonstrated that the corner yield patterns 4–5–6 shown in Fig. 17.10.1—in one-to-one
correspondence to yield patterns 1–2–3 of Fig. 17.9.2—may indeed give a smaller collapse
load and therefore control. These corner patterns are complicated to analyze, either by the
virtual work method or by the equilibrium method. For instance, there are three unknowns
E–F–G for the yield line positions; then, once the expression for wu /φ has been obtained
from the virtual work equation as a function of three independent variables, the partial
derivative of wu /φ with respect to each of the three unknown variables can be equated
to zero. In the equilibrium method, the same set of equations for the positions of points
E–F–G may be obtained by inserting the predetermined zero or nonzero nodal forces and
applying the moment equation of equilibrium to each of the slab segments.
An analysis [17.6] of a square slab with equal reinforcement in the x- and y-directions
will show that corner yield pattern No. 4 of Fig. 17.10.2(b) [see also Fig. 17.10.1(a)] results
in wu /φ = 22(Mnn + Mnp) /a2, whereas the regular yield pattern of Fig. 17.10.2(a) indicates
wu /φ = 24(Mnn + Mnp)/a2. Mnn and Mnp are the nominal moment strengths per unit slab width
for the negative moment and positive moment regions, respectively, in each direction, and a
is the side of the square. Thus the corner pattern is more critical by approximately (24 –22)/
24 = 8.3%. It may be proper then to discount the results of a regular yield pattern analysis
as made in Section 17.9 for most rectangular slabs by 8 to 10% for reason of corner effects.
It may be pointed out that yield line EF in Fig. 17.10.2(b) is a negative moment yield
line; thus when there is no negative moment reinforcement, the moment strength along EF
is zero. In this case the crack or yield line EF will not form if the corner A is not held down
G G G
E E E
F F F
(a) Yield pattern No. 4 (b) Yield pattern No. 5 (c) Yield pattern No. 6
Figure 17.10.1 Corner yield patterns for a rectangular two-way slab panel.
wu 22 (Mnn + Mnp)
=
D φ a2 C
wu 24 (Mnn + Mnp)
=
φ a2
βa
0
a
G H
A B
F P Q K
a αa
(a) Regular yield pattern (b) Corner yield pattern
Figure 17.10.2 Square slab panel with equal reinforcement in two directions.

743
1 7 . 1 1 A P P L I CAT I O N TO S P E C I A L CA S E S 743
because the corner would simply lift up. ACI-8.7.3 requires that special reinforcement be
provided at exterior corners in both top and bottom of the slab, for a distance in each
direction from the corner equal to one-fifth the longer span. The use of negative moment
reinforcement near the corner tends to move point G farther away from the corner, and thus
such reinforcement helps increase the load that will cause a collapse mechanism.
17.11 APPLICATION OF YIELD LINE ANALYSIS

TO SPECIAL CASES
The yield line theory of slabs, as developed and illustrated in the preceding sections, is
particularly suitable for special cases involving irregular shapes or irregular boundary con-
ditions. Prerequisite to the analysis of these cases is the picturing of an applicable yield
line pattern. The governing concept here is that rigid body plane rotations of slab segments
separated at yield lines are possible under compatible deflection conditions. To this end, the
following guides may be provided.
1. Yield lines end at a slab boundary.

2. A yield line (or its prolongation) between two slabs segments passes through the
intersection of the axes of rotation of the two adjacent slab segments.
3. The axes of rotation lie along lines of supports or pass over column supports.
In addition to those already described, two other yield line patterns to further illustrate the
use of the above guides are shown in Fig. 17.11.1.
Special Case
Shown in Fig. 17.11.2 is a rectangular slab simply supported at three edges and free at the
upper edge. The positive moment contributed by the reinforcement parallel to the a dimen-
sion provides a nominal moment strength of Mnx per unit of the b distance; the positive
moment contributed by the reinforcement parallel to the b dimension provides strength
Mny per unit of the a distance. Two possible yield patterns are shown in Fig. 17.11.2(c) and
17.11.2(d); the unknown is x in yield pattern No. 1 and y in yield pattern No. 2.
For yield pattern No. 1, referring to Fig. 17.11.2(c) and letting Δ be the deflection where
the yield line meets the free edge, the virtual work done by the uniform load is
1  wu   ∆  1  w   ∆ w  ∆
W=   bx   2 +  u  xb   2 + u (a − 2 x )b  
2 φ   3 2 φ   3 φ  2
wu  b
W= ( ∆ )   (3a − 2 x )
φ  6
Supported edge Free edges

Column
Supported
edge
Supported edge Supported edge

(a) (b)
Figure 17.11.1 Yield line patterns in special cases.

74
Free edge Bottom reinforcement for Mny
b Simply
supported edges
a Bottom reinforcement for Mnx
(a) Dimensions (b) Reinforcement

Rotation =
b
x x Mny
Mny Mny
+ +
V α V V V
x
Rotation =
b
Mnx
Mnx
Mnx
a Mny
(c) Yield pattern No. 1

a/2 a/2
Rotation =
y
Mnx
Mny Mny
a
Rotation =
Mnx
y
Mnx
Mny Mnx
(d) Yield pattern No. 2
Figure 17.11.2 Rectangular slab simply supported at three edges and free at one edge.
The virtual work done by the yield line moments is
 ∆  ∆
W = 2 M nx b   + 2 M ny x  
 x  b
Equating the two expressions for W and solving for wu /φ,
wu 12(b M nx + x M ny )
2 2
= (17.11.1)
φ b 2 x(3a − 2 x )
Setting to zero the derivative of Eq. (17.11.1) with respect to x,
 M ny  2
3a  x + 4b 2 x − 3ab 2 = 0 (17.11.2a)
 M nx 
In order that the root of Eq. (17.11.2a) be less than a/2,
 M ny  4b 2
 M  > 3a 2 (17.11.2b)
nx
For the equilibrium method, the nodal force V in Fig. 17.11.2(c) may be predetermined
by use of Eq. (17.7.3); or
 x
V = M ny cot α = M ny  
 b
745
1 7 . 1 1 A P P L I CAT I O N TO S P E C I A L CA S E S 745
For equilibrium of the triangular segment in Fig. 17.11.2(c),
1  wu   x   x
  bx   + M ny   x = M nx b
2 φ   3   b
from which
wu 6(b M nx − x M ny )
2 2
= (17.11.3)
φ b2 x 2
For equilibrium of the trapezoidal segment in Fig. 17.11.2(c),
1  wu   b  w  b  x
xb   2 + u (a − 2 x )b   = 2 M ny x + 2 M ny   b
2  φ   3  φ  2  b
from which
wu 24 x M ny
= 2 (17.11.4)
φ b (3a − 4 x )
The same quadratic equation in x as Eq. (17.11.2) is obtained from equating Eq. (17.11.3)
to Eq. (17.11.4).
For yield pattern No. 2, referring to Fig. 17.11.2(d) and letting Δ be the deflection at the
yield line perpendicular to the free edge, the virtual work done by the uniform load is
1  wu   a   ∆  w  a  ∆ 1  w   ∆
W=   y     2 + u (b − y)     2 +  u  ay  
2 φ  2 3     φ  
2 2  2 φ   3
wu  a
W= ( ∆ )   (3b − y)
φ  6
The virtual work done by the yield line moments is
 2∆   ∆
W = 2 M nx b   + M ny a  
 a   y
Equating the two expressions for W and solving for wu /φ,
wu 6(4byM nx + a M ny )
2
= (17.11.5)
φ a 2 y(3b − y)
Setting to zero the derivative of Eq. (17.11.5) with respect to y,
 M ny   M ny 
4by 2 + 2 a 2   y − 3a 2 b  =0 (17.11.6a)
 M nx   M nx 
For the root of Eq. (17.11.6a) to be less than b,
 M ny   4b 2 
 M  <  a 2  (17.11.6b)
nx
Since the moment strengths at the three intersecting yield lines in Fig. 17.11.2(d) are
identical, the nodal forces are all zero, based on the treatment presented in Section 17.8. For
equilibrium of the trapezoidal segment in Fig. 17.11.2(d),
1  wu   a a w  a a
y   + u (b − y)   = M nx b
2  φ   2 6 φ  2 4
746
from which
wu 24bM nx
= 2 (17.11.7)
φ a (3b − 2 y)
For equilibrium of the triangular segment in Fig. 17.11.2(d),
1  wu  y
ay = M ny a
2  φ  3
from which
wu 6 M ny
= 2 (17.11.8)
φ y
The same quadratic equation in y as Eq. (17.11.6a) is obtained from equating Eq. (17.11.7)
to Eq. (17.11.8).
Thus Eqs. (17.11.1) to (17.11.4) apply to yield pattern No. 1, which cannot happen if
(Mny /Mnx) is smaller than 4b2/(3a2), and Eqs. (17.11.5) to (17.11.8) apply to yield pattern
No. 2, which cannot happen if (Mny /Mnx) is larger than 4b2/a2. When (Mny /Mnx) is between
4b2/(3a2) and 4b2/a2, both yield patterns should be analyzed; the one giving the smaller col-
lapse load controls. Although an exact value of (Mny /Mnx) in terms of b2/a2 (between 1.33
and 4.00) at the transition point where yield pattern No. 1 begins to control over yield pat-
tern No. 2 may be determined, the complication in the algebra does not seem to warrant the
effort. Table 17.11.1 shows the results of analysis for six different cases in which the values
of (Mny /Mnx) become progressively larger while Mny remains constant.
TABLE 17.11.1 YIELD LINE ANALYSIS OF A RECTANGULAR SLAB

WITH THREE SUPPORTED EDGES AND ONE FREE EDGE (FIG. 17.11.2),
a = 25 ft, b = 20 ft
Case 1 2 3 4 5 6
Mny (ft-kips/ft) 16 16 16 16 16 16
Mnx (ft-kips/ft) 24 18.75 16 8 6.25 4
Yield pattern No. 1
x (ft) — 12.5 12 9.78 9.01 7.68
[Eq. (17.11.2)]
w (ksf ) — 0.480 0.427 0.262 0.222 0.167
[Eq. (17.11.1),
(17.11.3), or (17.11.4)]
Yield pattern No. 2
y (ft) 13.22 14.41 15.20 18.75 20 —
[Eq. (17.11.6)]
w (ksf ) 0.549 0.462 0.415 0.273 0.240 —
[Eq. (17.11.5),
(17.11.7), or (17.11.8)]
Controlling yield pattern No. 2 No. 2 No. 2 No. 1 No. 1 No. 1
Collapse load wu /φ (ksf ) 0.549 0.462 0.415 0.262 0.222 0.167
74
PROBLEMS 747
SELECTED REFERENCES
17.1. A. Ingerslev. “The Strength of Rectangular Slabs,” Journal of the Institute of Structural

Engineers, London, 1 (1), January 1923, 3–14. Disc., 14–19.
17.2. K. W. Johansen. “The Ultimate Strength of Reinforced Concrete Slabs,” Final Report. Third
Congress, International Association for Bridge and Structural Engineering, Liège, Belgium.
September 1948 (pp. 565–570).
17.3. K. W. Johansen. Yield Line Theory. London: Cement and Concrete Association, 1962.
17.4. L. L. Jones and R. H. Wood. Yield Line Analysis of Slabs. New York: Elsevier, 1967.
17.5. R. H. Wood. “Plastic Design of Slabs Using Equilibrium Methods,” Flexural Mechanics of
Reinforced Concrete, Proceedings of the International Symposium, Miami, 1964 (pp. 319–336).
17.6. E. Hognestad. “Yield Line Theory for the Ultimate Flexural Strength of Reinforced Concrete
Slabs,” ACI Journal, Proceedings, 49, March 1953, 637–656.
PROBLEMS
17.1 Assuming a 6-in. slab thickness for the continu- moment yield line intersects the edges at 5 and
ous slab span described in Example 17.5.1, the 4 ft from the corner along the 25-and 20-ft
uniform load wu due to the weight of the slab edges, respectively.
itself may be considered to be 1.2 wD = 1.2(75) 17.4 Same as Problem 17.3, except for Case 2 of
= 90 psf. If the slab supports a transverse wall Example 17.9.1.
at 7 ft from the left support line, determine the 17.5 Same as Problem 17.3, except for Case 3 of
maximum wall load per transverse foot that will Example 17.9.1.
cause a collapse mechanism to occur. 17.6 Verify the solution for Case 1 in Table 17.11.1.
17.2 Solve Example 17.7.1, but use Mnx = 10 ft-kips 17.7 Verify the solution for Case 2 in Table 17.11.1.
and Mny = 8 ft-kips, each per foot width of slab. 17.8 Verify the solution for Case 3 in Table 17.11.1.
17.3 For the regular yield pattern solution to Case 17.9 Verify the solution for Case 4 in Table 17.11.1.
1 of Example 17.9.1, investigate the effect of 17.10 Verify the solution for Case 5 in Table 17.11.1.
a corner yield pattern in which the negative 17.11 Verify the solution for Case 6 in Table 17.11.1.
CHAPTER 18
TORSION
18.1 GENERAL
Reinforced concrete members may be subjected to torsion, frequently in combination with
bending and shear. The cantilever member in Fig. 18.1.1(a) is largely subjected to torsion,
although some bending and shear also exist due to its own weight. The fixed-ended beam of
Fig. 18.1.1(b) is subjected to substantial amounts of bending, shear, and torsion.
Spandrel beams at the edge of a building built integrally with the floor slab are subjected
not only to transverse loads but also to a torsional moment per unit length equal to the
restraining moment at the exterior end of the slab. Similarly, spandrel girders receive tor-
sional moments from the exterior ends of the floor beams that frame into them.
Torsion on structural systems may be classified into two types: statically determinate
torsion (sometimes called “equilibrium torsion”), for which the torsion can be determined
from statics alone, and statically indeterminate torsion (sometimes called “compatibility
torsion”), for which the torsion cannot be determined from statics alone and a rotation
(twist) is required for deformational compatibility between interconnecting elements, such
as a spandrel beam, slab, or column. Both examples in Fig. 18.1.1 are cases of statically
determinate torsion.
In cases of statically determinate torsion, as in Fig. 18.1.2(a) and 18.1.2(b), the amount
of torsion the member is required to resist is based on the requirement of statics and is
independent of the stiffness of the member. Statically indeterminate torsion, as shown in
Fig. 18.1.2(c) and 18.1.2(d), sometimes exists where there would be no torsion if the sta-
tistical indeterminacy were eliminated. For instance, if the support at A is eliminated in
Fig. 18.1.2(c), the torsion is eliminated. Similarly, in Fig. 18.1.2(d), if a flexural hinge is
put at B, the torsion is eliminated. For such statically indeterminate torsion situations, the
amount of torsion in a member depends on the magnitude of the torsional stiffness of the
member itself in relation to the stiffnesses of the interconnecting members.
For an excellent overview of the entire subject of torsion as it affects structures, with
specific reference to reinforced concrete design, the reader is referred to Tamberg and
Mikluchin [18.3]. An extensive bibliography is also included. An extensive treatment of
torsion for reinforced and prestressed concrete can be found in the 1983 book by Hsu
[18.12]. Collins and Mitchell [18.9] have provided a unified rational treatment of the func-
tion of reinforcement to resist shear and torsion, and have made design recommendations
for reinforced and prestressed concrete members. A number of general references are avail-
able [18.1–18.18]. Joint ACI-ASCE Committee 445 on Shear and Torsion also recently
published a Report on Structural Concrete [18.19].
In this chapter, brief treatment is given to the computation of torsional stress and torsional
rigidity of homogeneous sections, the torsional strength of reinforced concrete sections, the
749
18.2 TORSIONAL STRESS IN HOMOGENEOUS SECTIONS 749
Inclined cracks due to torsion; test by J. P. Klus at the University of Wisconsin–Madison.
(a) (b)
Figure 18.1.1 Reinforced concrete members subjected to torsion.
P P
B
P
P
A
(a) (b) (c) (d)
Figure 18.1.2 Comparison of statically determinate torsion (Cases a and b) and statically

indeterminate torsion (Cases c and d). (Structures shown in plan view with load P perpendicular to
plane of frame.)
development and background for the ACI Code requirements, design examples, and the
effect of torsional stiffness in a continuity analysis.
18.2 TORSIONAL STRESS IN HOMOGENEOUS

SECTIONS
A torsional moment T acting on a shaft of homogeneous material as shown in Fig. 18.2.1
causes shear stresses υ over the cross section.
Circular Sections
For a circular section, a plane transverse section before twisting remains plane after twist-
ing. Consequently, the resultant shear stress υ at any point is proportional to its distance
from the center and is in a direction perpendicular to the radius. Calling h the diameter of
the circle [Fig. 18.2.1(b)] and υt the maximum torsional shear stress at the perimeter,
 2υ  2υ 2C
T = ∫ r υ dA = ∫ r  t r  dA = t ∫ r 2 dA = υt
A A  h  h A h
750
750 C hapter 1 8 T orsion
dV = v dA
h/2
Shear stress
due to torsion vt
(a) (b)
x
(c)
Figure 18.2.1 Torsional stress in homogeneous sections.
in which C, the polar moment of inertia, is

h/2
h/2  2 πr 4  πh 4
C = ∫ r 2 dA = ∫ r 2 2 πr dr =  =
A 0  4  0 32
The torsional shear stress υt becomes
16T
υt = (18.2.1)
πh 3
Rectangular Sections
The torsional shear stress distribution over a rectangular section of dimension x by y can-
not be as easily derived as that for a circular section. Unlike the circular section, where
plane transverse sections remain plane after twisting, the noncircular cross section warps
under torsion. If plane sections were maintained after twisting, the maximum shear stress
would exist at a point farthest from the axis of twist. Such is not the case for rectangu-
lar sections. From the mathematical theory of elasticity [18.116], it has been found that
the maximum torsional shear stress υt occurs at the midpoint of the long side and paral-
lel to it. The magnitude of υt is a function of α, the ratio of y to x (long to short sides)
[Fig. 18.2.1(c)]; and
T
υt = (18.2.2)
α x2 y
The values for α are given in Table 18.2.1.
TABLE 18.2.1 VALUES FOR α
y/x 1.0 1.2 1.5 2.0 2.5 3 5 ∞
α 0.208 0.219 0.231 0.246 0.256 0.267 0.290 0.333

751
18.3 TORSIONAL STIFFNESS OF HOMOGENEOUS SECTIONS 751
T-, L-, and I-Sections

The torsional shear stress distribution in T-, L-, or I-sections may be approximated by divid-
ing the section into several component rectangles assuming that each component rectangle
has a large ratio y/x so that the value for α may be assumed to be 1 3 [18.116]. The maximum
shear stress υt occurs at the midpoint of the long side y of the rectangle having the greatest
thickness xm and
Txm
υt = (18.2.3)
1
∑ 3 x3 y
in which x and y are the thickness and side dimension, respectively, of each component rec-
tangle. Since the web of the sections considered is usually thicker than the flange, xm will
usually be the web thickness.
18.3 TORSIONAL STIFFNESS OF HOMOGENEOUS

SECTIONS
The torsional stiffness Kt of a member is defined as the ratio of torsional moment T to the
angle of twist φ over a length L. The torsional rigidity is usually represented by the symbol
GC in which G is the modulus of elasticity in shear and C is the torsion constant. Thus if θ
is the total angle of twist over a length L,
T GC (18.3.1)
Kt = =
θ L
Circular Sections
It was just shown (see Section 18.2, subsection on Circular Sections) that the torsion con-
stant C of a circular section of diameter h is the polar moment of inertia
πh 4
C= (18.3.2)
32
Rectangular Sections
The torsion constant C of a rectangular section of height y and width x may be expressed,
per Timoshenko and Goodier [18.116] as
C = βx3 y (18.3.3)
in which β is a function of y to x. The values for β are given in Table 18.3.1.
TABLE 18.3.1 VALUES OF β
y/x 1.0 1.2 1.5 2.0 2.5 3 4 5
β 0.141 0.166 0.196 0.229 0.249 0.263 0.281 0.291

752
T-, L-, and I-Sections

The torsion constant C of a T-, L-, or I-section may be approximated [18.116] by the
expression
1
C = ∑ x3 y (18.3.4)
3
in which y and x are, respectively, the side length and thickness of each of the component
rectangles into which the section may be divided.
The following is a more exact expression [18.116, p. 313] for the torsion constant, giving
values closer to those in Table 18.3.1 for sections composed of rectangular elements having
y/x less than about 10,
1  x
C = ∑ x 3 y  1 − 0.63  (18.3.5)
3  y
In the ACI Code, torsion provisions are based on the thin-walled tube, space truss anal-
ogy (see Section 18.7); thus, no explicit use is made of the torsion constant C. However,
in the provisions for two-way slab systems as discussed in Chapter 16 (the definition for C
in ACI-8.10.5.2), Eq. (18.3.5) is used (see Section 16.11). The lower estimate of stiffness
using Eq. (18.3.5) is desirable in structural analysis when one is determining the restraining
effect of spandrel members in two-way floor systems.
18.4 EFFECTS OF TORSIONAL STIFFNESS

ON COMPATIBILITY TORSION
General Treatment of Torsion on Statically Indeterminate Systems
To perform a statically indeterminate structural analysis, it is necessary first to be able to deter-
mine the relative stiffnesses of interacting members. The “compatibility torsion” discussed
in Section 18.1 is involved. For example, if a spandrel member is uncracked and its torsional
stiffness GC/L is computed as shown in Section 18.3, the torsional moment that the member
will attempt to carry may be very large. As the member cracks, its torsional stiffness reduces
drastically, the member will twist, and the torsional moment carried is likewise reduced.
Postcracking stiffness and torsional moment have been studied by Lampert [18.4] and
Collins and Lampert [18.5], who proposed an expression for the torsional rigidity of a
cracked section. Since the stiffness is needed before the torsional moment can be deter-
mined, the cracked section stiffness is not available because it requires knowledge of the
steel reinforcement.
Alternatively, Collins and Lampert [18.5] have indicated that in cases of compatibility
torsion (where torsional moment depends on torsional stiffness), analysis on the basis of
zero torsional stiffness resulted in a design as satisfactory as an analysis using uncracked
stiffness. In fact, added steel may increase the torsional moment in the member but have
little effect on the twist (rotation). Consequently, it may be more effective to design for a
twist (rotation) than for a torque (torsional moment). The purpose of the torsional rein-
forcement then is to provide ductility and distribute cracks caused by the twist. Such a
procedure would be within the spirit of the ACI Code, where ACI-6.3.1.1 states, “Relative
stiffnesses of members within structural systems shall be based on reasonable and con-
sistent assumptions.” Further, ACI Commentary R6.3.1.1 indicates that “torsional stiffness
may be neglected” when one is dealing with compatibility torsion.
Using the zero torsional stiffness assumption, the member resisting torsion (say, a span-
drel beam) would be designed for flexure and shear, neglecting torsion; then the torsional
stiffness based on the cracked section could be computed. The structure may then be ana-
lyzed to determine the torsional moments, and the section may be checked according to the
ACI Code rules for torsion design.
753
18.4 EFFECTS OF TORSIONAL STIFFNESS 753
ACI Code Procedure

The ACI Code (ACI-22.7.3.3) provides an optional simple procedure to reduce the
design complexity for cases involving compatibility torsion. When a statically inde-
terminate situation involves torsion, and an internal redistribution of forces can occur
as a result of cracking, the factored torsional moment to be used in design is reduced
to a minimum value sufficient to provide the necessary rotation capacity (ductility). In
other words, if the torsional restraint is omitted in determining the bending moments
and shears on the structural elements, the design of those elements may be more con-
servative than otherwise, but the difficulty of determining the torsional moments has
been eliminated. The torsional members must, however, have the ductility to twist the
necessary amount.
To summarize, the ACI Code provides two options for the design of torsional members
when the torsional moment is dependent on the relative stiffness of the interacting members
(compatibility torsion).
1. Estimate the torsional and flexural stiffness of all interacting members by making
“reasonable and consistent assumptions.” (ACI-6.3.1.1). Determine the moments,
shears, and torsional moments by statically indeterminate structural analysis using
factored loads. Then apply the ACI Code provisions for torsion design.
2. Neglect torsional stiffness in the statically indeterminate structural analysis. This
assumes that the torsional member will crack and that “a large twist occurs under an
essentially constant torsional moment, resulting in a large redistribution of forces in
the structure” (ACI-R22.7.3). Since no torsional moment will then be available from
the computation, the torsional members must be designed for an ACI Code–specified
minimum torsional moment intended to control the width of torsional cracks. Under
ACI-22.7.3.2, the factored torsional moment Tu is permitted to be reduced to a value
approximating the torsional cracking moment.
For nonprestressed members, the design factored torsional moment φTn is permitted to be
taken at the critical sections of ACI-9.4.4.3 as [ACI-Table 22.7.5.1(a)]:
 Acp 2 
φ Tn = φ 4 λ fc′   (18.4.1)
 pcp 
For prestressed members (see Chapter 20), the design factored torsional moment φTn is
permitted to be taken at the critical sections of ACI-9.4.4.3 as [ACI Table 22.7.5.1(b)]:
 Acp 2  f pc
φ Tn = φ 4 λ fc′   1+ (18.4.2)
p
 cp  4 λ fc′
For nonprestressed members subjected to an axial tensile or compressive force, the design
factored torsional moment is permitted to be taken as [ACI Table 22.7.5.1(c)]:
 Acp 2  Nu
φ Tn = φ 4 λ fc′   1+ (18.4.3)
 pcp  4 Ag λ fc′
where
Acp = area enclosed by outside perimeter of concrete cross section
Ag = gross area of concrete
pcp = outside perimeter of the concrete cross section
fpc = compressive stress in concrete at the centroid of the cross section resisting exter-
nally applied loads
Nu = factored axial force
754
2G3 2G4 2G4 2G3
2B3
2B1
2B1
2B1
2G1 2G2 2G2 2G1
2S3
3 @26’ = 78’
2B4
2B2
282
282
2S1 2S2 2S3
2B3
2B1 4 @39’ = 156’
Figure 18.4.1 Floor plan of typical slab-beam-girder construction.
The compressive stress fpc used in prestressed concrete is discussed in Section 20.10 [see
Eq. (20.10.9)]. Equations (18.4.1), (18.4.2), and (18.4.3) can be viewed as representing the
cracking torsional moment at a principal tensile stress of 4 λ fc′ psi.
Spandrel Beams and Girders

Consider the spandrel beam 2B4 (Fig. 18.4.1) in the typical slab-beam-girder floor of
Example 8.3.1. This beam receives a vertical load and a torsional moment per unit length
from the slab 2S1, which are equal, respectively, to the reaction and restraining moment at
the exterior end of slab 2S1. In addition, the beam supports the weight of whatever walls
or windows may rest directly on it. Thus the spandrel beam 2B4 is subjected to a torsional
moment per unit length in addition to bending and shear. A similar condition exists in the
spandrel girder 2G4; in that case, however, the torsional moments are applied only at the
junction points with the beams.
The torsional moments in the spandrel beams or girders cause torsional shear stresses,
which are additive to the bending shear stresses at the inside face of the member. The
usual approach in design is to provide for the sum of the torsional shear and flexural shear
requirements. Since torsional shear stress goes around the member, closed stirrups or hoops
are necessary.
The magnitude of the torsional moment acting uniformly along a spandrel beam (ACI-
9.4.4.1) such as 2B4 of Fig. 18.4.1 might be roughly approximated as equal to the restrain-
ing moment along the exterior edge of the slab, using a value such as 1 24 wL2, as given
by ACI Table 6.5.2. Alternatively, the torsional moment may be neglected if the slab is
designed assuming the absence of a restraining moment along the spandrel beam. In such a
case, the spandrel beam must be designed for a minimum torsional strength corresponding
to that which will provide adequate ductility to twist (see Section 18.11).
The design and behavior of spandrel beams has been the subject of many studies [18.87–
18.101], the most extensive of which are those of Raths [18.99] and Klein [18.100].
EXAMPLE 18.4.1
Estimate the maximum factored torsional moment Tu in the spandrel beam 2B4 of
Fig. 18.4.1 if the restraining moment at the exterior end of slab panel (5.5-in. slab and
a clear span of 11.92 ft) 2S1 is M = wL2 /24. The service live and dead loads are 100
1
and 69 psf, respectively. Assume an 18 × 18 in. column and a 13 × 22 in. overall size
beam. Use f ′ = 3000 psi (normal weight). 2
c
(Continued)
75
1 8 . 5 T O R S I O N A L M O M E N T S T R E N G T H T cr AT C R A C K I N G 755
SOLUTION
(a) Compute the factored torsional loading. The factored restraining moment Mu along
the edge of the slab is approximately
wu = 1.2(69) + 1.6(100) = 243 psf

1
Mu = (0.243)(11.92)2 = 1.44 ft-kiips/ft of width
24
The torsional moment is largest at the face of the column and decreases nearly line-
arly to zero at midspan. The factored torsional moment Tu at the face of the column is
approximately
1 
Tu =  clear span of spandrel Mu = 12.25(1.44) = 17.6 ft-kips
2 
(b) Compute the ACI Code–specified design torsional moment φTn permitted to be used
in lieu of a structural analysis for statically indeterminate torsion. The spandrel
member would then be designed arbitrarily to have a design strength
 Acp 2 
φ Tn = φ 4 λ fc′   (18.4.1)
 pcp 
From the thin-walled space truss concept, the section resisting torsion is the primary
rectangular portion bwh without any of the slab included. Thus

Acp = bw h = 13(22.5) = 292.5 sq in.
and the outside perimeter ppc of the resisting section is

p pc = 2(13 + 22.5) = 71 in.
Thus
 Acp 2   292.52  1
φ Tn = φ 4 λ fc′   = 0.75(4)(1) 3000  = 16.4 ft-kips
 pcp   71  12,000
In this case, the design for ductility using Eq. (18.4.1) would be slightly less conserva-
tive than using the Tu = 17.6 ft-kips computed in part (a).
18.5 TORSIONAL MOMENT STRENGTH T cr

AT CRACKING
From Eq. (18.2.2) it follows that the nominal torsional moment strength Tn of a plain con-
crete rectangular section may be expressed
Tn = α x 2 y [ ft (max)] (18.5.1)
because torsional stress υt equals maximum principal tensile stress ft (max) in the situation
of pure torsion (same as pure shear). In Eq. (18.5.1), α is an elastic theory coefficient as
given in Table 18.2.1, ranging from 0.208 to 1 3 as y/x varies from 1.0 to ∞, and up to 0.50
for plastic theory.
756
x = dimension of shorter side

90°
Compression
zone
Outline
of actual
failure
surface y = dimension of longer side
from test
Failure plane in
tension zone
φ1 = 45°
NA for skew bending
Figure 18.5.1 Skew bending of plain concrete rectangular section. (According to Hsu [18.21].)
Hsu [18.21] has shown that when ft (max) is taken at about 5 to 6 fc′ (i.e., representing
the stress at which normal-weight concrete cracks in tension), and α at 1 3 , Eq. (18.5.1)
gives the torsional cracking moment strength Tcr,
Tcr =
1 2
3
(
x y 5 to 6 fc′ ) (18.5.2)
Hsu [18.21] has also shown that a torsion failure of a rectangular section does not occur in
a spiral form, as might be expected from a circular shaft. Instead, a rectangular section in
torsion cracks by bending about an axis parallel to the wider face of the section and inclined
at about 45° to the axis of the beam, as shown in Fig. 18.5.1. This is called the skew bending
theory (see Section 18.6).
An alternative line of thinking makes the analogy to torsion in thin-walled sections. It
is well known from mechanics of materials1 that for closed thin-walled sections, the shear
flow υt t resulting from torsion is
T
υt t = (18.5.3)
2 Ao
where υt is the shear stress, t is the thickness of the tube wall, T is the torsional moment,
and Ao is the area enclosed by the tube (measured at mid-thickness of the tube wall). When
the section is actually solid rather than a tube or box, the wall thickness must be defined.
According to MacGregor and Ghoneim [18.86], the Euro-International Committee (CEB)
approximates t as Acp /pcp, where pcp is the perimeter of the concrete section and Acp is
the area enclosed by that perimeter. The Canadian Concrete Code (see Chapter 5, Ref.
5.57) assumes that, prior to cracking, the equivalent wall thickness is 0.75Acp /pcp and the
area Ao enclosed by the tube centerline is 2Acp /3. Substituting the Canadian values into
Eq. (18.5.3) gives
T T 3 pcp T pcp
υt = = = 2 (18.5.4)
2 Ao t 2 2 Acp 0.75 Acp Acp
The principal tensile stress ft (max) resulting from pure torsion equals the shear stress
in a thin-walled tube, that is, Eq. (18.5.4). Then, setting maximum υt equal to ft (max), the
1 See, for example, Charles G. Salmon and John E. Johnson, Steel Structures, Design and Behavior Emphasizing Load and
Resistance Factor Design (4th ed). New York: Harper Collins College Publishers, 1996 (pp. 458–462).
75
1 8 . 6 S K E W B E N D I N G T H E O R Y 757
cracking tensile stress lower bound for concrete in biaxial tension-compression, assumed
to be 4 λ fc′, gives the torsional moment Tcr to cause cracking as
Tcr pcp
υt = ft (max) = = 4 λ fc′ (18.5.5)
Acp 2
 Acp 2 
Tcr = 4 λ fc′  
 pcp  (18.5.6)
The cracking stress 5 to 6 fc′ used in the skew bending model for normal-weight concrete
is somewhat higher than the 4 fc′ used in the thin-walled tube model because the modulus
of rupture for tensile strength in bending (i.e., skew bending model) is larger than the prin-
cipal tensile stress causing cracking for the biaxial tension-compression state of stress (i.e.,
pure torsion on a thin-walled tube).
18.6 STRENGTH OF RECTANGULAR SECTIONS

IN TORSION—S KEW BENDING THEORY
Once steel reinforcement, both longitudinal and transverse, has been placed in a rectangu-
lar section, the behavioral mechanism changes from that of plain concrete. The resisting
action of transverse reinforcement in the form of closed stirrups or hoops is similar to that
of stirrups resisting flexural shear. Prior to cracking, the reinforcement participates little
if at all; but after cracking, the reinforcement carries a large portion of the total torsional
moment. The contribution of concrete is only about 40% of the torsional strength of an
unreinforced section. The failure mode according to this theory, however, does continue to
be one of skew bending.
The skew bending concept was proposed by Lessig [18.20] and extended by Goode and
Helmy [18.25], Collins, Walsh, Archer, and Hall [18.26], and Below, Rangan, and Hall
[18.27], all of whom applied it to the case of combined bending and torsion. Hsu [18.23]
has applied the concept to the case of torsion alone and has developed the expression that
formed the basis for the 1971–1989 ACI Code procedures. Hsu [18.22, 18.24], Zia [18.1,
18.2], and Warwaruk [18.10] have provided summaries of the theories relating to rectangu-
lar sections in torsion. McMullen and Rangan [18.7] have reviewed the research to clarify
the contradictions between the skew bending and space truss theories (see Section 18.7
for the space truss theory). The following development presents some of the ideas relating
to the strength expression developed by Hsu [18.23].
Referring to Fig. 18.6.1, the failure section is assumed to be a plane that is perpendicular
to the wider face of the member and inclined at 45° to the axis of the member. The failure
plane may be as shown in Fig. 18.6.1, due to twisting moment in the direction indicated.
Since a bending mode of failure is assumed, the compression zone is treated as in any
beam analysis; that is, it has a depth a over which the compressive stress may be assumed
uniform. On the tension side, where the concrete cracks and no resistance to tension is
assumed, the reinforcing hoops carry tensile forces Pυ and the longitudinal bars resist shear
across the cracked concrete via dowel action (see Section 5.4). The horizontal and verti-
cal components of this dowel action are designated Qx and Qy. As long as the concrete is
uncracked and the concrete itself transmits shear, no dowel action exists.
On the compression side [Fig. 18.6.2(a)], the longitudinal bars contribute tensile force
Pℓ; the concrete contributes shear resistance Ps in the failure plane and also compressive
resistance Pc normal to the failure plane. The components of the resultant force P are shown
in Fig. 18.6.2(b). The hoop reinforcement in compression is neglected because, as has been
shown in Section 3.10, the nominal moment strength Mn is not significantly affected by the
inclusion of compression reinforcement.
758
a
y
x ent
mom
sting
90° Twi
Qy
P
Compression
zone under Qx y
skew bending Pv
Pv
2
y√
ist Pv
so f tw
Axi Qy 45° x
nt
me
mo
sting
Twi
Qx
z1 s
s
Figure 18.6.1 Forces acting on skew bending failure section.
On the tension side, it is noted that no longitudinal force can exist. If such a force were
to act, it would have to be counterbalanced by a component of resistance acting oppositely.
Since only Pυ, Qx, and Qy are assumed to be acting, and the resultant force must be directed
upward (opposite P on the compression side), no resultant tension or compression can exist
in the longitudinal direction of the tension zone under skew bending.
The reader is reminded that unless the direction of the twisting moment is definitely
established by analysis, a potential failure plane can exist opposite to that in Fig. 18.6.1,
having the compression and tension sides interchanged. Thus, the longitudinal forces, stir-
rup forces, and dowel forces must be resisted on each side of the section.
Strength Attributable to Concrete

The shear resistance Ps (Fig. 18.6.2) may be expressed as
( )
Ps = υavg y 2 a (18.6.1)
where υavg is the average shear stress acting over the compression zone, y 2 is the width of
the compression zone, and a is the depth of the compression zone.
Alternatively, the shear strength Ps may be considered proportional to the effective
area xy 2 and to fc′ (or λ fc′ in the case of lightweight concrete) [see Eq. (5.5.6) omit-
ting effect of reinforcement]. Thus
Ps = k1 xyλ fc′ (18.6.2)
where k1 is a proportionality constant.

From Fig. 18.6.2(b), one may note that
P = 2 Ps + P (18.6.3)
The first term of Eq. (18.6.3) represents the portion attributable to concrete. Thus the tor-
sional strength Tc attributable to concrete equals the force 2Ps times the moment arm (say,
0.80x),
Tc = 2 Ps (arm)
= 2 k1 xyλ fc′(0.80 x )

= k2 λ fc′x 2 y (18.6.4)
759
1 8 . 6 S K E W B E N D I N G T H E O R Y 759
Pl
Pc Ps
Pl
Longitudinal bar
P 45°
Ps 45°
45° P
Pc
45°
(a) (b)
Figure 18.6.2 Components of resultant force P acting on compression zone of failure plane.
Equation (18.6.4) is of the same form as Eq. (18.5.2) but includes the lightweight concrete
factor λ. If 6 λ fc′ is used in Eq. (18.5.2), and 40% of that equation is used in recognition
that the concrete strength contribution to the total torsional strength of a reinforced section
represents only about 40% of the torsional cracking strength of a plain concrete section,
then Tc of Eq. (18.6.4) becomes
1 
Tc = 0.4  x 2 y(6 λ fc′)
3 
= 0.8λ fc′x 2 y (18.6.5)
Strength Attributable to Hoops and Longitudinal Reinforcement

Referring to Figs. 18.6.1 and 18.6.2, the forces Pυ, Qx, and Qy on the tension side, and Pℓ on
the compression side, have yet to be considered.
Assuming both the transverse and longitudinal steel have yield strength fy, the contribu-
tion of the closed vertical stirrups (or hoops) is
y 
Pυ = At f y  1  (18.6.6)
 s 
where y1/s (see Fig. 18.6.3) is the number of hoops intercepted by the 45° failure plane.
The tensile force Pℓ in the longitudinal bars intercepting the compression concrete
zone is
A 
P = ξ    f y
 2  (18.6.7)
where ξ is an efficiency factor to account for location of the longitudinal bars at two or
more points within the compression zone, and Aℓ is the total area of all longitudinal bars
(assumed to be Aℓ /2 in the compression zone). Pℓ contributes to the torsional strength, since
from Eq. (18.6.3) it is part of P.
y2 y1
x2
x1
Figure 18.6.3 Cross s ection dimensions.

760
The dowel forces Qx and Qy act after the concrete has cracked and these forces may be
assumed proportional to the bar cross-sectional area and to the bar lateral displacement,
which is proportional to the distance (either 0.5x2 or 0.5y2) from the center of twist to the
bar. Thus
Qx = k3 A y2 , Qy = k3 A x2 (18.6.8)
where k3 is a proportionality constant. Dowel action in beams subject to torsion has been
reviewed by Youssef and Bishara [18.84].
Next let m equal the ratio of the volume of longitudinal bars to the volume of closed
stirrups or hoops such that
A s
m= (18.6.9)
2 At ( x1 + y1 )
or
 2 m( x1 + y1 ) 
A = At   (18.6.10)
 s 
 y   x1 At f y 
P = ξm  1 + 1   (18.6.11)
 x1   s 
Substituting Eq. (18.6.10) into Eq. (18.6.8) gives
 2 mAt ( x1 + y1 ) 
Qx = k3 y2  
 s
k3  y2   y   x1 y1 At f y 
=2 m 1 + 1    (18.6.12)
 
f y  y1   x1   s
Similarly,
k3  x2   y   x1 y1 At f y 
Qy = 2   m 1 + 1    (18.6.13)
f y  y1   x1   s
The torsional resistance from reinforcement then is
x  x  y  x 
Ts = Pυ  1  + Pt  2  + 2Qx  2  + 2Qy  2  (18.6.14)
 2  2  2  2 
Substitution of Eqs. (18.6.6), (18.6.11), (18.6.12), and (18.6.13) into Eq. (18.6.14) gives
 x1 y1 At f y 
Ts = α t   (18.6.15)
 s
where
1  y  x  k  y   1
αt = + ξm  1 + 1   2  + 2 3 m  1 + 1  ( x22 + y22 )   (18.6.16)
2  x1   2 y1  fy  x1   y1 
761
1 8 . 7 S P A C E T R U S S A N A L O G Y 761
Assuming x2 ≈ x1 and y2 ≈ y1, the quantity αt is essentially a function of two parameters, m

and y1/x1 and might be written as
y 
α t = C1 + C2 m + C3  1  (18.6.17)
 x1 
where the constants C1, C2, and C3 may be experimentally determined.

Hsu [18.23] has shown that for equal volume of longitudinal bars to closed hoops (i.e.,
m = 1), αt may be expressed as
y 
α t = 0.66 + 0.33  1  (18.6.18)
 x1 
which was used in the 1989 ACI Code.

Thus the full nominal strength Tn of rectangular reinforced concrete sections may be
written by combining Eqs. (18.6.5) and (18.6.15),
 x1 y1 At f y 
Tn = Tc + Ts = 0.8λ fc′x 2 y + α t   (18.6.19)
 s
The first term represents the strength attributable to one rectangle. When the cross section
consists of several rectangular segments, Eq. (18.6.19) becomes
 x1 y1 At f y 
Tn = Tc + Ts = 0.8λ fc′(∑ x 2 y) + α t   (18.6.20)
 s
where x2y is computed for each segment having its short dimension x and long dimension y.
18.7 STRENGTH OF RECTANGULAR SECTIONS IN

TORSION—S PACE TRUSS ANALOGY
Evaluation of torsional strength using the space truss analogy was first suggested by
Rausch [18.28] and developed by Lampert and Thürlimann [18.4, 18.29]. The concept was
further explained by Lampert and Collins [18.30], Müller [18.33], and Rabbat and Collins
[18.34]. Advances in knowledge of reinforced concrete in torsion led to the conclusion
that the space truss analogy is likely the simplest way to give unified treatment to flexure
and torsion. Collins and Mitchell [18.9] suggested a rational design approach based on the
space truss analogy, which led to changes in the ACI Code provisions in 1995 for design
to include torsion. The European Concrete Committee (CEB) [18.31] has since about 1978
used the space truss analogy as its basis for design.
The torsional strength of reinforced concrete rectangular sections is almost entirely pro-
vided by the reinforcement and the concrete that immediately surrounds the steel. Then
one may consider a rectangular section as a thin-walled tube [Fig. 18.7.1(a)] or as a space
truss [Fig. 18.7.1(b)]. The longitudinal bars in the corners contribute tensile forces, while
the concrete strips between cracks provide compressive strength. The inclined compressive
forces act in a spiral fashion around the box section, giving compressive forces on both
vertical and horizontal sides.
The current provisions for torsion design are well explained by MacGregor and Ghoneim
[18.86] and the following presentation is largely from that source.
After a solid rectangular concrete member has cracked in torsion, the concrete interior
contributes little strength. Thus, such sections can be treated as equivalent tubular mem-
bers. Before cracking, the section acts as a tube as in Fig. 18.7.1(a), and after cracking
762
Diagonal compressive T x0 Stirrups

stress at angle θ Cracks
T
Shear flow y0
path
V1
θ V4
θ V2
A0
Longitudinal
Longitudinal Concrete V3
bar
tensile force compression
Shear diagonals
flow, q
(a) Thin-walled tube analogy (b) Space truss analogy
Figure 18.7.1 Thin-walled tube and space truss analogies. (From MacGregor and Ghoneim [18.86].)
it acts as a space truss as in Fig. 18.7.1(b). Tests reported by Hsu [18.24] comparing the
strengths of solid and hollow sections show that once cracking has occurred, the concrete
in the center of the solid member has little effect on torsional strength.
Thin-Walled Tube Shear Stress

As discussed in Section 18.5, the shear flow υt t resulting from torsion in closed thin-wall
sections is
T
υt t = [18.5.3]
2 Ao
where υt is the shear stress, t is the thickness of the tube wall, T is the torsional moment,
and Ao is the area enclosed by the tube (measured at mid-thickness of the wall). When the
section is actually solid rather than a tube or box, the wall thickness must be defined. After
cracking has occurred, the torsional strength is provided by the closed stirrups, longitudinal
bars, and outer concrete skin. According to MacGregor and Ghoneim [18.86], Ao is empir-
ically taken as 0.85Aoh, where Aoh is the area enclosed by the closed stirrups, measured to
the centerline of the outermost hoops. The thickness t is taken as Aoh /ph, where ph is the
perimeter of the centerline of the closed stirrups.
Space Truss Idealization

After cracking, the reinforced concrete section becomes essentially a space truss consisting
of longitudinal bars in the corners, closed stirrups, and concrete compression diagonals that
spiral around the member between torsional cracks [see Fig. 18.7.1(b)]. The width x0 and
height y0 of the truss are measured center-to-center of the sides of the closed stirrups. The
angle θ of the crack to the horizontal is initially about 45° on formation of the crack for non-
prestressed beams, but may vary from approximately 30° to 60° as the cracking increases.
Area of Closed Stirrups

Equation (18.5.3) gives the shear force per unit length of perimeter, known as shear flow
υt t. The total shear force, say V2 in Fig. 18.7.1(b), for a given side of the tube is the shear
flow υt t times the length of the side, say y0, to obtain V2. Thus, the shear force in one vertical
side and in one horizontal side is
T T
V2 = y0 and V1 = x0 (18.7.1)
2 Ao 2 Ao
Similar forces act on all four sides, as shown in Fig. 18.7.1(b), oriented to cause a torsional
moment about the axis of the member. Figure 18.7.2 shows a portion of one of the vertical
763
1 8 . 7 S P A C E T R U S S A N A L O G Y 763
At fyt Longitudinal bar
Vertical
hoops
At fyt
V2 y0
s θ
y0 cot θ
Longitudinal bar
Figure 18.7.2 Vertical forces on one side of space truss of Fig. 18.7.1(b). (From MacGregor and
Ghoneim [18.86].)
sides, where s is the spacing of closed stirrups, θ is the angle of cracks with the member
longitudinal axis, y0 is the vertical height to the center of stirrups, and y0 cot θ is the hori-
zontal projection of the crack intercepted by m stirrups. Thus,
ms = y0 cot θ (18.7.2)
The force in the stirrups crossing the diagonal crack must equal V2. Assume the stirrups
yield at nominal torsional strength, then
V2 = mAt f yt (18.7.3)
where At is the cross-sectional area of one leg of a closed stirrup, and fyt is the yield stress
of the stirrup steel. Substitution of Eq. (18.7.2) into Eq. (18.7.3) gives
At f yt y0
V2 = cot θ (18.7.4)
s
Similarly, the shear force V1 will involve the x0 dimension, as follows:
At f yt x0
V1 = cot θ (18.7.5)
s
Substituting Eq. (18.7.1) into Eq. (18.7.4) for V2 [or Eq. (18.7.1) into Eq. (18.7.5) for V1],
and letting T be the nominal torsional moment strength, Tn , gives
2 Ao At f yt
Tn = cot θ (18.7.6)
s
Equation (18.7.6) is ACI Formula (22.7.6.1a).

Hsu [18.35] has shown how to determine θ by analysis. The value can vary from about
30 to 60°; the angle will be smaller for prestressed concrete (see Chapter 20) than for non-
prestressed concrete.
From Eq. (18.7.6) the required stirrup cross-sectional area At per unit spacing s at each
side of the section becomes
At Tn T /φ
= tan θ = u tan θ (18.7.7)
s 2 Ao f yt 2 Ao f yt
where Tu /φ is the factored torsional moment Tu divided by the strength reduction factor φ;
the quantity Tu /φ represents the required nominal strength Tn.
764
Longitudinal Reinforcement
Figure 18.7.3 shows the forces on the side of the space truss of Fig. 18.7.1(b) designated
by the shear force V2. The torsional cracks have created inclined concrete struts crossed by
one leg of a closed stirrup (or hoop). The shear force V2 can be represented by a diagonal
compressive force D2 parallel to the concrete compressive struts along with an axial tensile
force N2 (which is divided equally between the top and bottom longitudinal steel). The
forces D2 and T2 are
V2
D2 = (18.7.8)
sin θ
N 2 = V2 cot θ (18.7.9)
Similarly, for the force N1 on the face of width dimension x0 where V1 acts,
N1 = V1 cot θ (18.7.10)
Assuming that the shear flow is constant along the side of the thin-walled tube, it is reason-
able to assume that the resultant forces D2 and N2 act at midheight of a side. Since longitu-
dinal bars are expected to be located in all four corners, the total longitudinal force N for
all four sides would be
N = 2( N1 + N 2 ) (18.7.11)
Substituting Eqs. (18.7.1) into Eqs. (18.7.9) and (18.7.10), and then into Eq. (18.7.11), gives
 T T 
N = 2 y0 cot θ + x0 cot θ (18.7.12)
 2 Ao 2 Ao 
Then replacing T by the nominal strength Tn and N by the nominal strength Nn gives
Tn
Nn = 2( x0 + y0 ) cot θ (18.7.13)
2 Ao
Noting that 2(x0 + y0) is the perimeter ph of the closed stirrups, Eq. (18.7.13) may be written
Tn ph
Nn = cot θ (18.7.14)
2 Ao
The total longitudinal steel force Nn at nominal strength is
N n = Al f y (18.7.15)
N2/2
D2
fcd V2
y0 θ
θ N2
y0 cos θ
N2/2
Figure 18.7.3 Resolution of shear force on the side of space truss shown in Fig. 18.7.1(b). (From
MacGregor and Ghoneim [18.86].)
765
1 8 . 8 C O M B I N E D B E N D I N G A N D T O R S I O N 765
Thus, combining Eqs. (18.7.14) and (18.7.15) gives the total required longitudinal steel
area Aℓ as
Tn ph (T /φ) ph
Al = cot θ = u cot θ (18.7.16)
2 Ao f y 2 Ao f y
Solving for the nominal torsional moment strength Tn,
2 Al Ao f y
Tn = tan θ
ph (18.7.17)
which is ACI Formula (22.7.6.1b).
18.8 STRENGTH OF SECTIONS IN COMBINED

BENDING AND TORSION 2
In most practical situations, torsion will occur simultaneously with flexure. There have
been many studies of the interaction between bending and torsion [18.4, 18.9, 18.27, 18.29,
18.36–18.46]. Both the skew bending theory [18.20, 18.26, 18.38] and the space truss anal-
ogy [18.4, 18.29] are in general agreement on the interaction behavior. Hsu [18.18] has
provided a summary of the interaction relationship.
Assuming that the concrete does not participate in carrying flexure, the bending moment
M gives rise to a tensile force M/y0 in the bottom steel and an equal compressive force in the
top steel (see Fig. 18.8.1). In addition, the torsional moment T causes a total tensile force N
in the longitudinal steel, given by Eq. (18.7.14):
Tn ph
Nn = cot θ (18.7.14)
2 Ao
Assuming that the force N from torsion alone is divided equally between the top and bottom
longitudinal steel,
Tn ph
N top = N bottom = cot θ (18.8.1)
4 Ao
According to Collins and Lampert [18.30, 18.81] and Hsu [18.18], two failure modes
are possible depending on relative amounts of longitudinal reinforcement in the top (flex-
ural compression zone) and in the bottom (flexural tension zone). The first mode involves
yielding of both the bottom longitudinal steel and the transverse steel (i.e., closed stirrups
or hoops); the second involves yielding of both the top longitudinal steel and the trans-
verse steel. In general, the strength interaction relationship is given by Collins and Lampert
[18.81] and explained further by Hsu [18.18].
Symmetrically Reinforced Sections, As′ = As

When equal amounts of top and bottom longitudinal reinforcement are used and a positive
bending moment acts with torsion, the bottom steel will always yield before the top steel
because the compressive force resulting from flexure counteracts the tensile forces aris-
ing from torsion. In this case, the addition of a bending moment will reduce the torsional
2 The discussion in this section is for combined action of torsional moment and positive bending moment, which if acting
alone would cause tension in the bottom steel As and compression in the top steel As′ . If the bending moment is negative, the reader
should interpret As′ as being the bottom steel and As as the top steel.
76
M Tph
+ = – + cot θ = Ntop
y0 4Ao
y0 Tph
M
+ + cot θ = Nbottom
Tph M y0 4Ao
cot θ
4Ao y0
(a) Torsion (b) Bending moment (c) Torsion + bending moment
Figure 18.8.1 Superposition of longitudinal forces resulting from torsion and bending.
1.2
Closed
1.0 stirrup
2
Tn Mn
0.8 + =1
Tn0 Mn0 As’
Tn
0.6
Tn0
As
0.4
As = As’
0.2
Mn
0 0.2 0.4 0.6 0.8 1.0 Mn0
Figure 18.8.2 Bending-torsion interaction diagram for equal tension and compression longitudinal steel.
strength. The interaction relationship between bending and torsion for this case may be
approximated as follows (see Fig. 18.8.2):
2
 Tn  Mn
 T  + M = 1 (18.8.2)
n0 n0
where
Tn = nominal torsional moment strength in presence of flexure
Tn0 = nominal torsional strength when member is subject to torsion alone
Mn = nominal flexural strength in the presence of torsion
Mn0 = nominal flexural strength when the member is subject to flexure alone
Unsymmetrically Reinforced Sections, As′ < As

When pure torsion is applied to an unsymmetrically reinforced section with the top com-
pression steel area less than the bottom tensile reinforcement area, the top steel will yield
first and will govern the failure mode. Addition of a positive bending moment will induce
compression in the top steel and allow it to carry larger tensile forces arising from tor-
sion. As a result, the torsional strength of the cross section will increase when a bending
moment is applied. One of two failure modes is then possible: (a) if the bending moment
is small (i.e., low compressive force in the top steel), then the top steel will yield before
the bottom steel or (b) if the bending moment is large, then the bottom steel will yield first.
Regardless of the failure mode, the torsional strength of an unsymmetrically reinforced sec-
tion will increase in comparison to the section under pure torsion. The strength interaction
for unsymmetrically reinforced sections can be approximated by the following expressions.
For yielding of the bottom (flexural tensile zone) and transverse steel:
2
T  M
r n  + n =1 (18.8.3)
 n0 
T M n0

( )( )
where r = As′ f y′ / As f y and f y′ and fy are the yield stresses of the top and bottom steel,
respectively.
76
1 8 . 9 C O M B I N E D S H E A R A N D T O R S I O N 767
2
Tn 1 Mn
= 1+
Tn0 r Mn0
3
1/
ld l
=
e
yie ste
r
Bo
p
To
tto
r
m
=
1.0
ste
1/3 elds
el
yi
r = 2
Bo Tn 1 Mn
tto 1 = 1–
Tn m Tn0 r Mn0
ste
Tn0 el
yie A’s f’y
lds r=
Asfy
1.0
Mn
Mn0
Figure 18.8.3 Bending-torsion interaction relationships. (From Collins and Lampert [18.81].)
For the case of the top steel yielding first, the expression is
2
 Tn  Mn 1
 T  − M r = 1 (18.8.4)
n0 n0
Equations (18.8.3) and (18.8.4) are shown graphically in Fig. 18.8.3. Note that the ben-
eficial effect of a bending moment reaches its peak when top and bottom steel yield
simultaneously.
Since the ratio r is normally less than 1.0—that is, the tension steel area As required for
bending moment strength is significantly larger than the compression steel area As′ used for
ductility—the torsional strength Tn in the presence of bending moment Mn is not reduced by
the interaction effect; in fact, for r in the range of 1 3 to 1 2 , the torsional strength increases.
For practical purposes, it is reasonable to design for the sum of the requirements for torsion
and bending moment.
Accordingly, ACI-9.5.4.5 permits reducing the area of longitudinal torsion reinforce-
ment Aℓ in the flexural compression zone by an amount equal to Mu /(0.9dfy), where fy is
the yield stress of the longitudinal steel. This procedure would be in agreement with the
requirement for the top steel in Fig. 18.8.1(c). That reduced amount, however, is not permit-
ted to be less than that required (ACI-9.6.4.3) as a minimum amount of torsion reinforce-
ment, or the minimum torsion reinforcement required (ACI-9.7.5.1) to be distributed at a
maximum spacing of 12 in. around the perimeter of the closed stirrups.
18.9 STRENGTH OF SECTIONS IN COMBINED SHEAR

AND TORSION
Rectangular and L-shaped sections have been studied under combined shear and torsion
by a few investigators [18.47–18.50]. However, since shear usually accompanies flexure,
the combination of flexure, shear, and torsion has received more attention [18.51–18.74].
Generally, though, flexural shear and torsional shear are of significance in regions where
bending moment is low. Thus, for design purposes it is most necessary to know the strength
interaction relationship between shear and torsion.
Test data have provided a wide range of points on the interaction relationship using torsion
and shear coordinates. Because of the unknowns involved, some investigators have proposed
768
a linear interaction equation [18.1, 18.2, 18.56, 18.65] for design purposes. However, a num-
ber of studies at the University of Texas [18.47, 18.49, 18.51, 18.55, 18.65] on rectangular,
L-shaped, and T-shaped beams have indicated that a quarter-circle interaction relationship is
acceptable for members without web reinforcement. For members with web reinforcement,
the interaction is curved but flatter than the quarter-circle. The quarter-circle expression is
2 2
 Tn   Vn 
 T  +  V  = 1 (18.9.1)
n0 n0
where Tn and Vn are the nominal strengths in torsion and shear, respectively, acting simul-
taneously; Tn0 is the nominal strength under torsion alone; and Vn0 is the nominal strength
under shear alone. Equation (18.9.1) was used for the 1989 ACI Code expression for the
strength in combined shear and torsion on sections without web reinforcement. However,
for web reinforcement the separate requirements for shear and torsion were added together,
rather than arrival at by following Eq. (18.9.1).
18.10 STRENGTH INTERACTION SURFACE FOR

COMBINED BENDING, SHEAR, AND TORSION
The effect of the simultaneous application of bending, shear, and torsion may be most eas-
ily examined by means of an interaction surface. Such a concept was used in Section 10.20
for biaxial bending of columns. Various investigators have proposed interaction surfaces.
Some of these surfaces are shown in Figs. 18.10.1 and 18.10.2. The work of Collins, Walsh,
and Hall [18.56] seems to provide the most complete rational approach by correlating the
regions of the failure surface with modes of failure, as shown in Fig. 18.10.2.
Tn Tn
Tn0 Tn0
1.0 1.0
Discontinuity
1.0 1.0
Vn Vn
0.5 Vn0 Vn0
Mn 1.0
1.0 Mn0
Mn
Mn0
(a) Hsu interaction surface (Ref. 18.53) (b) Mirza-McCutcheon interaction surface (Ref. 18.54)
Tn
Tn0
1.0
Vn
1.0 Vn0
Mn
Mn0
1.0
(c) Victor-Ferguson interaction surface (Ref. 18.55)
Figure 18.10.1 Interaction surfaces for combined bending, shear, and torsion (for members
without transverse reinforcement).
769
18.10 COMBINED BENDING, SHEAR, AND TORSION 769
Since most members that are subjected to bending, shear, and torsion will have less
longitudinal steel in the compression zone due to flexure alone (top steel in the positive
moment zone) than in the tension zone (bottom steel in the positive moment zone), only
that case is given in Fig. 18.10.2(a). When equal amounts of top and bottom longitudinal
steel are used, the Mode 3 (“negative” yield or yield in top steel when a positive moment is
applied) cannot occur; in such a case the surfaces of Modes 1 and 2 should extend upward
until they intersect the Tn /Tn0 axis.
The following conclusions may be drawn regarding the interaction of bending, shear,
and torsion.
1. The interaction between torsion and shear may be represented for most situations
(with As′ < As) by a quarter-circle and is relatively unaffected by the simultaneous
application of bending moment of a magnitude equal to one-third to one-half of the
nominal bending strength when shear and torsion are absent.
2. When equal amounts of top and bottom longitudinal steel are used, the quarter-circle
shear–torsion interaction still seems acceptable, but there is a reduction in strength
when bending moment is also applied.
3. Based on most test results, the straight-line shear-torsion interaction, while simple to
use, appears overly conservative.
Tn
Tn0 Negative yield (Mode 3)
1.0
Side yield (Mode 2)
Shear failure (Mode 4)

Positive yield
(Mode 1)
1.0
Vn
Vn0
Mn 1.0
Mn0
(a) Interaction surface for beam with weaker top
steel than bottom steel (As’ < As)
Compression zone
(b) Mode 1 (bending and torsion) (c) Mode 2 (low shear–high torsion)
Top steel
(d) Mode 3 (low bending–high (e) Mode 4 (high shear–low torsion)

torsion; weaker top steel)
Figure 18.10.2 Interaction surface and failure modes (for members with transverse reinforcement)
according to Collins, Walsh, and Hall. (From Ref. 18.56.)
70
18.11 TORSIONAL STRENGTH OF
CONCRETE AND CLOSED TRANSVERSE
The traditional ACI Code design procedure for torsion prior to the 1995 ACI Code was
based largely on the work of Hsu along with recommendations of ACI Committee 438
[18.75–18.78, 18.80]. Since 1995, however, the design for torsion has been based on the
space truss analogy presented in Section 18.7.
References are available for special topics such as design of inverted T-girders [18.102,
18.103], ledger beams [18.104], channel-shaped sections [18.105–18.107], and beams hav-
ing openings [18.108–18.113].
Threshold Torsion
The ACI Code permits (ACI-9.5.4.1) neglect of torsion in solid sections when the factored
torsional moment is less than the threshold torsion, Tth , defined in ACI-22.7.4.1, multiplied
by the strength reduction factor φ.
For nonprestressed members,
 Acp 2 
Tu < φ Tth = φ λ fc′   (18.11.1)
 pcp 
For prestressed members,
 Acp 2  f pc
Tu < φ Tth = φ λ fc′   1+ (18.11.2)
 pcp  4 λ fc′
For nonprestressed members subjected to an axial tensile or compressive force,
 Acp 2  Nu
Tu < φ Tth = φ λ fc′   1+ (18.11.3)
where
Acp = area enclosed by outside perimeter of concrete cross section
fpc = compressive stress in concrete at the centroid of the cross section resisting exter-
nally applied loads
Nu = factored axial force (taken positive for compression and negative for tension)
Comparing Eqs. (18.11.1) through (18.11.3) with Eqs. (18.4.1) through (18.4.3), it is
clear that the threshold torsion is equal to one-fourth of the torsional cracking moment
strength, assuming a concrete tensile strength of 4 λ fc′.
Equations (18.11.1), (18.11.2), and (18.11.3) apply to hollow sections as well, which are
defined in ACI-R22.7.4 as those “having one or more longitudinal voids, such as a single-
cell or multiple-cell box girder.” For hollow sections, Acp must be replaced by the gross area
Ag of concrete in the above equations (ACI Table 22.7.4.1b). The outer boundaries of the
cross section must meet ACI-9.2.4.4 (see Section 16.4).
For isolated members with flanges and for members cast monolithically with a slab,
ACI-9.2.4.4a indicates that the overhanging flange width on each side of the beam web
used to compute Acp and pcp shall be taken equal to the greatest projection of the beam
above or below the slab, but not greater than four times the flange thickness. ACI-9.2.4.4b
indicates, however, that the “overhanging flanges shall be neglected in cases where the
Ag 2
parameter Acp2 / pcp for solid sections and p for hollow sections calculated for a beam with
cp
flanges is less than that calculated for the same beam ignoring the flanges.”
71
18.11 TORSIONAL STRENGTH—ACI CODE 771
Nominal Torsional Moment Strength Tn

The nominal torsional moment strength Tn is given by ACI-22.7.6.1 as the lesser of
2 Ao At f yt
Tn = cot θ [18.7.6]
s
which is ACI Formula (22.7.6.1a), and
2 Al Ao f y
Tn = tan θ (18.7.17)
ph
which is ACI Formula (22.7.6.1b).

In Eq. (18.7.6), Ao “shall be determined by analysis” (ACI-22.7.6.1) except that it is
permitted to be taken equal to 0.85Aoh (ACI-22.7.6.1.1), where Aoh is the area enclosed
by the closed stirrups, measured to the centerline of the outermost hoops. The angle θ of
potential cracking “shall not be taken less than 30 degrees nor greater than 60 degrees”
(ACI-22.7.6.1).
ACI-22.7.6.1.2 permits θ to be taken as 45° for nonprestressed concrete and 37.5° for
prestressed members having an effective prestress force not less than 40% of the tensile
strength of the longitudinal reinforcement. The value of 37.5° has been chosen arbitrarily at
halfway between the minimum 30° angle and the 45° angle traditionally used in the design
of stirrups.
According to ACI-9.7.5.1, torsion reinforcement must consist of longitudinal bars
distributed around the perimeter of transverse reinforcement. Transverse torsional rein-
forcement must consist of closed stirrups or hoops (ACI-9.7.6.3.1). Details of torsional
reinforcement are given in ACI-9.7.5 and ACI-9.7.6.3.
Area of Closed Stirrups (Hoops) Required for Torsion

As developed in Section 18.7, Eq. (18.7.7) gives the area At of closed stirrup (hoop) rein-
forcement per unit spacing s as
At Tn T /φ
= tan θ = u tan θ [18.7.7]
s 2 Ao f yt 2 Ao f yt
The quantity Tu /φ represents the required nominal strength Tn. Equation (18.7.7) is essen-
tially ACI Formula (22.7.6.1a). As used traditionally in design, At is the area of one leg of a
closed stirrup resisting torsion within a distance s.
Area of Longitudinal Reinforcement Required for Torsion

From Section 18.7, the longitudinal reinforcement Aℓ required in addition to that needed for
flexure is obtained from Eq. (18.7.16) as
Tn ph (T /φ) ph
Al = cot θ = u cot θ (18.7.16)
2 Ao f y 2 Ao f y
The perimeter distance ph of the closed stirrups (hoops) is 2(x0 + y0) for a rectangular sec-
tion, where x0 and y0 are the horizontal and vertical dimensions measured to the center of
the closed stirrups (hoops). In general, ph is the perimeter of the centerline of the outermost
closed transverse torsional reinforcement. The angle θ is the same as that used to compute
the area of closed stirrups in Eq. (18.7.7).
Note that to provide control of the width of diagonal cracks, the ACI Code limits the
values of fy and fyt to 60,000 psi (ACI-20.2.2.4).
72
Critical Section for Torsion

The critical section for torsion is taken essentially the same as for shear. ACI-9.4.4.3 pro-
vides the location of the first critical section for both nonprestressed and prestressed con-
crete beams. For nonprestressed members, the critical section is to be taken at the effective
depth d from the face of support unless a concentrated torsional moment occurs between
the face of support and the critical section, in which case the value at the face of support is
to be used. The portion of a member between the critical section and the face of support is
to be designed for the same (or greater if a concentrated torsion occurs in that region) tor-
sional moment as that at the critical section. The explanation for using the distance d from
the face of support is given in Section 5.11 on shear.
For prestressed members, the critical section is to be taken at h/2 from the face of sup-
port. The same conditions of the above paragraph apply as for nonprestressed members.
18.12 COMBINED TORSION WITH SHEAR

OR BENDING—A CI CODE
Combined Shear and Torsion
When torsion and shear act on a member, the nominal strength is reached either when the
closed stirrups reach their yield stress or when the longitudinal steel reaches its yield stress.
The member may not be considered “serviceable,” however, if inclined cracks are large at
service load. The interaction relationship used in the ACI Code relating shear V and torsion
T has been traditionally based on limiting crack width.
The average shear stress υu resulting from the factored shear Vu is
Vu
υu = (18.12.1)
bw d
where bw is the web width and d is the effective depth. The shear stress υtu resulting from
factored torsion Tu using Eq. (18.5.4) taking Ao = 0.85Aoh and t = Aoh /ph is
T T  1  ph
υtu = τ = = u
2 Ao t 2  0.85 Aoh  Aoh
Tu ph
υtu = (18.12.2)
2
1.7 Aoh
In the hollow section of Fig. 18.12.1(a), the shear stresses resulting from shear and torsion
are additive on the left wall. Thus, algebraic summation of the stresses υu and υtu seems
appropriate to compare with a limit stress; thus,

Vu T p
(
+ u h2 ≤ φ υc + 8 fc′
bw d 1.7 Aoh ) (18.12.3)
The limit stress on the right-hand side of Eq. (18.12.3) is the same as the upper limit of ACI-
22.5.1.2 of 8 fc′ for the shear stress resisted by the transverse reinforcement. The stress
υc is the nominal strength Vc from ACI-22.5.5 and ACI-22.5.6 divided by bw d to get it into
stress units. Thus, for hollow sections the limit of ACI-22.7.7.1(b) is given as
Vu T p  V 
+ u h2 ≤ φ  c + 8 fc′ (18.12.4)
bw d 1.7 Aoh  bw d 
which is ACI Formula (22.7.7.1b). For cases in which the wall thickness is less than Aoh /ph,
the second term on the left-hand side of Eq. (18.12.4) shall be taken as Tu / (1.7 Aoh t )
(ACI-22.7.7.2).
73
1 8 . 1 3 M I N I M U M TO R S I O NA L R E I N F O R C E M E N T — AC I C O D E 773
Shear Shear Shear Shear

stresses stresses stresses stresses
due to torsion due to shear due to torsion due to shear
(a) Hollow section (b) Solid section
Figure 18.12.1 Shear stresses due to shear and torsion.
In the solid section of Fig. 18.12.1(b), the shear stresses resulting from direct shear are
assumed to be distributed uniformly across the section width, while the torsional shear
stresses exist only in the walls of the assumed thin-walled tube. Thus, since the shear stress
can redistribute through the core region, the solid section would be more conservatively
treated than the thin-walled section by using Eq. (18.12.4). Thus, the following square-root
summation is used for the limiting stress on solid sections,
2 2
 Vu   Tu ph   Vc 
 b d  +  1.7 A2  ≤ φ  b d + 8 fc′ (18.12.5)
w oh w
which is ACI Formula (22.7.7.1a).
Combined Bending and Torsion

The ACI Code does not explicitly consider this combination of loadings. Torsion induces an
axial force in the longitudinal reinforcement. In regions where bending moment also exists,
the flexural reinforcement requirement is added to the longitudinal torsion reinforcement
requirement. This is a conservative procedure for the flexural compression zone. In that
region the torsion improves the resistance, as shown in Fig. 18.8.1(c).
For this reason, ACI-9.5.4.5 permits reducing the area of longitudinal torsion reinforce-
ment Aℓ in the flexural compression zone by an amount equal to Mu /(0.9dfy), where fy is
the yield stress of the longitudinal steel. That reduced amount is not permitted to be less
than the minimum (ACI-9.6.4.3) nor less than the minimum (ACI-9.7.5.1) based on a bar
in each closed stirrup corner and a maximum spacing of 12 in. around the perimeter of the
closed stirrup.
In prestressed concrete beams (ACI- 9.5.4.4), the total longitudinal reinforcement
(including tendons and nonprestressed steel) shall resist the factored bending moment Mu
at that section plus an additional concentric longitudinal tensile force equal to Aℓ fy, based
on the factored torsion Tu at that section; ACI-9.5.4.5, however, permits reducing the area
of longitudinal torsion reinforcement in the compression zone below that required by ACI-
9.5.4.4 by Mu /(0.9dfy).
18.13 MINIMUM REQUIREMENTS FOR TORSIONAL

The minimum area requirements for the transverse reinforcement At and longitudinal rein-
forcement Aℓ are to ensure that there is ductile behavior prior to failure. Hsu [18.24] and
Collins [18.78] have found that the strength attributable to concrete when closed stirrups
are present is only about 40% of the torsional strength of a plain concrete member. When
74
torsional reinforcement is required, the area At of transverse torsion reinforcement in com-

bination with area Aυ for shear reinforcement must satisfy (ACI-9.6.4.2),
bw s
( Av + 2 At ) ≥ 0.75 fc′ (18.13.1)
f yt
but not less than 50bw s/fyt .

For torsion alone, Eq. (18.13.1) becomes
bw s
At ≥ 0.375 fc′ (18.13.2)
f yt
but not less than 25bw s/fyt .

The background for the equation for minimum area Aℓ of longitudinal reinforcement is
provided by MacGregor and Ghoneim [18.86], whose work is the basis for much of what
follows here.
Two rectangular reinforced concrete beams in the tests by Hsu [18.24] failed at the tor-
sional cracking load. In those beams, the total volumetric ratio of the sum of closed stirrups
and longitudinal reinforcement to the volume of the concrete, respectively, was 0.80 and
0.88%. A beam having volumetric ratio of 1.07% failed at 1.08 times the cracking torsional
moment Tcr. All other beams had volumetric ratios exceeding 1.07% and achieved strengths
at least 1.2Tcr. Those tests suggest that beams loaded in pure torsion should have a mini-
mum volumetric ratio of 0.9 to 1%, which can be expressed [18.86]
A,min s At ph
+ ≥ 0.01 (18.13.3)
Acp s Acp s
or, solving for Aℓ, min, and recognizing that the yield stress of the stirrups and longitudinal
steel may be different, gives
f yt At ph
f y A, min = 0.01kAcp − (18.13.4)
s
where k is a constant assumed to be a function of the concrete strength. Dividing by fy gives
0.01k  A   f yt 
A, min = Acp −  t  ph   (18.13.5)
fy  s   fy 
MacGregor and Ghoneim [18.86] indicate that the constant 0.01k, which now has stress
(psi) units, can be taken as 7.5 fc′, which then makes Eq. (18.13.5) become
7.5 fc′  A   f yt 
A, min = Acp −  t  ph   (18.13.6)
fy  s   fy 
During the Committee 318 (ACI Code Committee) balloting process for the 1995 ACI
code, various forms of Aℓ,min were discussed. An objective was greater simplicity than the
equation of the 1989 ACI Code, while still providing the performance objective. Professors
Alan Mattock of the University of Washington and Thomas T. C. Hsu of the University of
Houston developed [18.86] the final simplified version of ACI 9.6.4 3(a) as follows,
5 fc′Acp A f yt
A, min = −  t  ph (18.13.7)
fy  s fy
75
18.14 EXAMPLES 775
In the above equation [ACI-9.6.4.3(b)],
At 25bw
≥ (18.13.8)
s f yt
where fyt refers to closed transverse torsional reinforcement, and fy refers to longitudinal
torsional reinforcement.
Equation (18.13.7) appears to give results less than Eq. (18.13.6); however, Hsu in
developing Eq. (18.13.6) applied the stress ratio υtu /(υtu + υu) to the first term, taking the
ratio as 2 3 . That would make Eq. (18.13.6) become the same as the ACI Code equation,
using the coefficient 5.
Spacing Limitations
Transverse torsion reinforcement (closed stirrups) must be spaced not farther apart than
ph /8 or 12 in. (ACI-9.7.6.3.3). This reinforcement must be provided over a distance at least
bt + d beyond the section where it is no longer required by analysis (ACI-9.7.6.3.2), where
bt is the width of the section containing the torsional transverse reinforcement.
Longitudinal bars must have a diameter of at least 0.042 times the stirrup spacing, but
not less than #3 in size (ACI-9.7.5.2) and must be distributed around the perimeter of the
closed stirrups with a maximum spacing of 12 in. (ACI-9.7.5.1). The longitudinal bars or
tendons must be inside the closed stirrups, and there must be at least one longitudinal bar
or tendon in each corner of the closed stirrups. As in the case of transverse reinforcement,
longitudinal torsional reinforcement must be extended at least bt + d beyond the section
where it is no longer required by analysis (ACI-9.7.5.3).
Termination of Torsion Reinforcement

Closed stirrups are permitted to be terminated (ACI-9.7.6.3.2) at a location (bt + d) beyond
the section where it is no longer required by analysis (i.e., Tu < φTth; ACI-9.5.4.1). For solid,
nonprestressed beams, this section is located where Eq. (18.11.1) is satisfied.
 Acp 2 
Tu = φ Tth = φλ fc′   [18.11.1]
 pcp 
where bt is the width of the part of the cross section containing the closed stirrups that are
resisting torsion.
Similar equations are used for prestressed members [Eq. (18.11.2)] and for nonpre-
stressed members having axial load [Eq. (18.11.3)]. Where the torsional moment exceeds
the threshold torsion, torsion reinforcement must be provided.
18.14 EXAMPLES
Several examples are presented to illustrate use of the ACI Code procedures. Table 18.14.1,
which summarizes most of the ACI provisions on torsion, can help the reader in reviewing
the computations in the examples.
76
TABLE 18.14.1 ACI CODE PROVISIONS FOR TORSION
1. For nonprestressed concrete:

 Acp 2  9.5.4.1;
Tu = 0 if computed Tu < φλ fc′  . 22.7.4.1
 pcp 
 Acp 2  9.4.4.4;
Tu = φ 4 λ fc′   , with redistribution 22.7.3.2;
 pcp  22.7.5.1
For prestressed concrete:
 Acp 2  f pc 9.5.4.1;
Tu = 0 if computed Tu < φλ fc′   1+ 22.7.4.1
 pcp  4 λ fc′
 Acp 2  f pc 9.4.4.4;
Tu = φ 4 λ fc′   1+ , with redistribution 22.7.3.2;
 pcp  4 λ fc′ 22.7.5.1
For nonprestressed members subjected to an axial tensile or compressive force
 Acp 2  Nu 9.5.4.1;
Tu = 0 if computed Tu < φλ fc′   1+ 22.7.4.1
 Acp 2  Nu 9.4.4.4;
Tu = φ 4 λ fc′   1+ , with redistribution 22.7.3.2;
22.7.5.1
where Acp = area enclosed by outside perimeter of concrete cross section 2.2
Nu = factored axial force
2. Limitations on cross-sectional dimensions:
For solid sections:

2 2
 Vu   Tu ph   Vc 
 b d  +  1.7 A2  ≤ φ  b d + 8 fc′ Formula (22.7.7.1a) 22.7.7.1(a)
w oh w
where ph = perimeter of centerline of outermost closed transverse torsional

reinforcement
For hollow sections:
Vu T p  V 
+ u h2 ≤ φ  c + 8 fc′
Formula (22.7.7.1b) 22.7.7.1(b)
bw d 1.7 Aoh  bw d 
If wall thickness < Aoh /ph, second term on left side above becomes
 Tu  22.7.7.2
 1.7 A t 
oh
3. φTn ≥ Tu 9.5.1.1(c)
(Continued)
7
18.14 EXAMPLES 777
TABLE 18.14.1 (Continued)
4. Nominal torsional strength:

2 Ao At f yt 22.7.6.1
Tn = cot θ
Formula (22.7.6.1a)
s
2 Al Ao f y
Tn = tan θ Formula (22.7.6.1b)
ph
Ao = 0.85Aoh unless determined by analysis

Aoh = area enclosed by centerline of outermost closed transverse torsional reinforcement
30° ≤ θ ≤ 60°
Permitted: θ = 45° for nonprestressed members
Permitted: θ = 37.5° for prestressed members having an effective prestress force not
less than 40% of the tensile strength of the longitudinal reinforcement
5. Transverse and longitudinal reinforcement:
At T /φ 9.5.1.1(c);
≥ u tan θ
s 2 Ao f yt 22.7.6.1
(Tu /φ) ph 9.5.1.1(c);

Al ≥ cot θ
2 Ao f y 22.7.6.1
6. Minimum reinforcement:
bw s 9.6.4.2
( Aν + 2 At ) ≥ 0.75 fc′
f yt
but not less than 50 bw s /fyt
5 fc′Acp A f yt 9.6.4.3(a)
A,min = −  t  ph
fy  s fy
At 25bw 9.6.4.3(b)
where ≥
s f yt
7. Spacing of reinforcement:
ph 9.7.6.3.3
max s (closed stirrups) ≤ ≤ 12 in.
8
torsion reinforcement to terminate at (bt + d) beyond theoretical point 9.7.6.3.2;
9.7.5.3
 longitudinal bars, diameter at least 0.042 times  9.7.5.2;
max s  ≤ 12 in.
 stirrup spacing, #3 or larger, one in each corner  9.7.5.1
78
EXAMPLE 18.14.1
A reinforced concrete spandrel beam has overall dimensions of 10 × 18 in. and is joined
integrally with a 6-in. slab (based on Example in Ref. 18.53), as shown in Fig. 18.14.1(a).
The section shown is that at the critical location a distance d from the face of support. At
this section, the factored loads are negative bending moment Mu = 75 ft-kips, shear force
Vu = 18 kips, and torsional moment Tu = 7 ft-kips. Assume that the torsional stiffness was
estimated and used in a structural analysis to obtain these design loadings. Check the
adequacy of this section and select the transverse reinforcement steel required, if any,
according to the ACI Code using fc′ = 4000 psi (normal weight) and fy = fyt = 40,000 psi.
SOLUTION
(a) Flexural strength. The required coefficient of resistance Rn is
required M n Mu
required Rn = =
bd 2 φ bd 2
Assuming φ = 0.90,
75(12, 000)
0.90(10)(15.6)2
From Eq. (3.8.5), the required reinforcement ratio ρ is

required ρ = 0.011
required As = 0.011(10)15.6 = 1.72 sq in.

provided As = 3(0.79) = 2.37 sq in. > [ required As = 1.72 sq in.] OK
Available steel area for longitudinal torsion reinforcement,
available As = 2.37 − 1.72 = 0.65 sq in.

available As′ = 2(0.60) = 1.220 sq in.
hf = 6”
3 – #8
d = 15.6”
18”
fc’ = 4000 psi

fy = 40,000 psi Beam
defined
2 – #7
by Hsu
10” 3hf = 18”
(a) Given data (from Hsu [18.53])
3 – #8 #3 hoop @ 5”
#3 hoop @ 5”
#3 minimum size, slab

#5
design will determine
2 – #7 this reinforcement
(b) Details of torsion reinforcement
Figure 18.14.1 Spandrel beam of Example 18.14.1. (Continued)

79
18.14 EXAMPLES 779
(b) Compute maximum Tu to neglect torsion. The maximum factored torsional moment
Tu that may be neglected (ACI-9.5.4.1) is
 Acp 2 
Tu = φλ fc′  
 pcp 
Acp = 10(18) + 18(6) = 288 sq in.
pcp = 2(10 + 18) + 18 + 18 = 92 in.
Note that Acp, the area enclosed by outside perimeter of the concrete section, and pcp,
the outside perimeter of the concrete cross section, have included the slab portion
(Fig. 18.14.1). Note also that the maximum effective width of slab used is 3 times its
thickness, per Hsu [18.53] and within the limits of ACI-9.2.4.4. Thus,
 2882  1
limit Tu < 0.75(1) 4000  = 3.56 ft-kips
 92  12,000
Since Tu = 7 ft-kips exceeds limit Tu, torsion must be included in the design.
(c) Determine whether the section is large enough for combined shear and torsion.
From ACI-22.7.7.1(a), for solid sections,
2
 Vu   Tu ph   Vc 
 b d  +  1.7 A2  ≤ φ  b d + 8 fc′
w oh w
Assuming 1.5-in. cover to outside of closed #3 stirrups,

Aoh = [18 − 2(1.5) − 0.375][10 − 2(1.5) − 0.375] = 96.9 sq in.
Conservatively, the effect of the integral slab is neglected. The perimeter ph of the centerline
of outermost closed transverse torsional reinforcement, again neglecting the integral slab, is
ph = 2(14.625 + 6.625) = 42.5 in.
Vu 18.0(1000)
= = 115 psii
bw d 10(15.6)
Tu ph 7.0(12, 000)42.5
= = 224 psi
2
1.7 Aoh 1.7(96.9)2
2 2
 Vu   Tu ph 
 b d  +  1.7 A2  = (115) + (224) = 252 psi
2 2
w oh
 V 
φ  c + 8 fc′ = 0.75(2 fc′ + 8 fc′) = 0.75(10 fc′) = 474 psi
 bw d 
Since 474 psi exceeds 252 psi, the section has adequate size.
(d) Compute stirrup reinforcement required for shear.
φVn ≥ Vu
φ (Vc + Vs ) ≥ Vu
1
φVc = φ 2 λ fc′bw d = 0.75(2)(1) 4000 (10)15.6 1000 = 14.8 kips

Av f yt d Av (40, 000)15.6 468 Av
φVs = φ = 0.75 = ( kips)
s s(1000) s
(Continued)
780
Since Vu > φVc, the requirement for transverse reinforcement based on strength is
Av Vu − φ Vc 18 − 14.8
= = = 0.0068
s φ d f yt 468
This amount will be combined with the torsion requirement, and the total Av + 2At must
be at least 0.75 fc′ bw s/fyt but not less than 50bw s/fyt (ACI-9.6.4.2).
(e) Compute transverse torsional reinforcement requirement. Using Eq. (18.7.7),
At T /φ
= u tan θ
s 2 Ao f yt

Ao = 0.85 Aoh = 0.85(96.9) = 82.4 sq in.
Taking θ = 45° as permitted by ACI-22.7.6.1.2(a) gives

7.0(12,000)
At Tu /φ 0.75
= tan θ = (1) = 0.017
s 2 Ao f yt 2(82.4)(40, 000)
(f) Compute transverse reinforcement for combined shear and torsion. The total trans-
verse reinforcement required for strength is
( Av + 2 At )
= 0.0068 + 2(0.017) = 0.041 Controls!
s
 A + 2 At  b 10
min  v  = 0.75 fc′ w = 0.75 4000 = 0.012
 s  f yt 40, 000
bw (10)
but not less than 50 = 50 = 0.013
f yt 40, 000
For #3 closed hoops,
2(0.11)
max s = = 5.4 in.
0.041
Since this exceeds the spacing limitation of ACI-9.7.6.3.3,
ph 42.5
max s = = = 5.3 in.
8 8
or 12 in., the 5.3-in. maximum spacing controls.
Use #3 hoops at 5-in. spacing.
(g) Longitudinal torsional reinforcement. Using Eq. (18.7.16) gives
7.0(12,000)
(Tu /φ) ph (42.5)
A = cot θ = 0.75 (1) = 0.72 sq in.
2 Ao f y 2(82.4)40, 000
Check Aℓ,min
 At   bw 25(10) 
 s = 0.017 ≥ 25 f = 40, 000 = 0.0063
   yt 
(Continued)
781
18.14 EXAMPLES 781
The strength-related value (0.017) exceeds 25bw /fyt; thus
A, min = −  t  ph
fy  s fy
5 4000 (288)  40, 000 

= − 0.017(42.5)  = 2.28 − 0.72 = 1.56 sq in.
40, 000  40, 000 
The amount of 1.56 sq in. is to be distributed around the perimeter of the section at a
spacing not to exceed 12 in. In this case, bars must be placed at middepth,
A, min 1.56
= = 0.52 sq in.
3 3
Use 2–#5 bars at middepth.
The longitudinal steel at the top and bottom in excess of that required for flexure is more
than adequate for the longitudinal torsion requirement (0.52 sq in.). Also, #3 transverse
reinforcement should be placed in the effective flange portion of the slab and should be
anchored within the main rectangle resisting torsion. Though the ACI code requires only
standard 90° hooks if those are restrained from opening by a slab [ACI-25.7.1.6(b)],
Mitchell and Collins [18.83] have recommended the use of 105° hooks (the 105° is the
amount of bend from the initial straight bar). Furthermore, when longitudinal steel is to
carry torsion at the face of support, the steel should be embedded into the support by an
amount Ld [18.83]. The reinforced section is shown in Fig. 18.14.1(b).
EXAMPLE 18.14.2
For the continuous spandrel beam shown in Fig. 18.14.2, design the longitudinal and
transverse reinforcement for the factored moment, factored flexural shear, and fac-
tored torsional moment given in Fig. 18.14.2. Assume that these design loads were
obtained after estimating the torsional stiffness and performing a structural analysis.
Use fc′ = 4000 psi (normal weight) and fy = fyt = 60,000 psi.
SOLUTION
(a) Flexural strength. The effective depth d is approximately 21.5 in. for one layer of
reinforcement. At midspan, neglecting the T-beam effect, the required coefficient of
resistance Rn (assuming φ = 0.90) is
required M n Mu
required Rn = =
bd 2
φ bd 2

83.3(12,000)
= = 200 psi
0.90(12)(21.5)2
From Fig. 3.8.1, the required reinforcement ratio ρ ≈ 0.0048, which gives

required As = 0.0048(12)21.5 = 1.24 sq in.
The minimum reinforcement required (ACI-9.6.1.2) is
3 fc′ 3 4000
As , min = bw d = (12)21.5 = 0.82 sq in.
fy 60, 000
(Continued)
782
but not less than
200bw d 200(12)21.5
As , min = = = 0.86 sq in. Controls!
fy 60, 000
At the support,

required As = 1.24(61.7 / 83.3) = 0.92 sq in.
A 6”
24”
30’–0” 12”
Section A–A
2’–0” A 2’–0”
(a) Beam
100 CL of support
80 Face of support 83.3’k
60
Factored moment (ft-kips)
40
20 CL of span
20
40
60 61.7’k
80
(b) Factored bending moment envelope
21.0k
20 19.4k
Factored shear (kips)
15
10 d
5.6
5
(c) Factored shear envelope
60
Factored torsion
39.2’k
(ft-kips)
40 35.5’k
20
10.2’k
2 4 6 8 10 12 14
(d) Factored torsion envelope
Figure 18.14.2 Spandrel beam including design actions, for Example 18.14.2. (Adapted from
Mattock [18.75].)
(Continued)
783
18.14 EXAMPLES 783
The requirements for the top and bottom reinforcement along the span due to flexure
are shown in Fig. 18.4.3(a) and 18.4.3(b), respectively. The reader may note that less
reinforcement than 0.86 sq in. may be used if the amount used is one-third more than
required for strength (ACI-9.6.1.3). Thus, the sloping straight line of Fig. 18.14.3(a) and
18.14.3(b) represents approximately 4 3 of the required As in the region between the con-
trolling minimum requirements of ACI-9.6.1.3 and 9.7.3.8.2 (or ACI-9.7.3.8.4).
(b) Factored shear Vu and factored torsion Tu at the critical section. Unless a concen-
trated load (or girder framing in) exists between the face of support and the distance
d therefrom, the maximum shear and torsion for which strength must be provided is
at d from the face of support (ACI-9.4.3.2 and ACI-9.4.4.3). From Fig. 18.14.2,
Vu = 19.4 kips

Tu = 35.5 ft-kips
(c) Compute maximum Tu to neglect torsion. The maximum factored torsional moment
Tu that may be neglected (ACI-9.5.4.1) is
 Acp 2 
Tu < φλ fc′   [18.11.1]
 pcp 
Acp = 12(24) = 288 sq in.

pcp = 2(12 + 24) = 72 in.
Note that in calculating Acp, the area enclosed by the outside perimeter of the concrete
section, and pcp, the outside perimeter of the concrete cross section, the slab portion has
been neglected (Fig. 18.14.2). Thus,
 2882  1
limit Tu = 0.75(1) 4000  = 4.6 ft-kips
 72  12, 000
Since Tu of 35.5 ft-kips far exceeds limit Tu, torsion must be included in design.
(d) Determine whether the section is large enough for combined shear and torsion.
2 2
 Vu   Tu ph   Vc 
 b d   1.7 A2  ≤ φ  b d + 8 fc′
w oh w
Assuming 1.5-in. cover to outside of closed #3 stirrups,

Aoh = [24 − 2(1.5) − 0.375][12 − 2(1.5) − 0.375] = (20.625)8.625 = 178 sqq in.
Conservatively, the effect of the integral slab is neglected. The perimeter ph of the centerline
of outermost closed transverse torsional reinforcement, again neglecting the integral slab,
ph = 2(20.652 + 8.625) = 58.5 in.
Vu 19.4(1000)
= = 75 psi
bw d 12(21.5)
Tu ph 35.5(12, 000)58.5
2
= 2
= 463 psi
1.7 Aoh 1.7(178)
2 2
 Vu   Tu ph 
 b d  +  1.7 A2  = (75) + (463) = 469 psi
2 2
w oh
(Continued)
784
(a)
1.24 sq in.
1.0
Area (sq in.)

ACI–9.6.1.3 0.86
Area of bottom
0.5 Min A s reinforcement required
0.31 ACI–9.6.1.2 for flexure
2 4 6 8 10 12 14
(b) 0.92 sq in. Min As (ACI–9.7.3.8.2)
1.0
0.86 (Min As for ACI–9.6.1.2)
Area (sq in.)
Area of top
0.5 ACI–9.6.1.3 reinforcement required
0.31 for flexure
ACI–9.7.3.8.4
(c) 2 4 6 8 10 12 14
8 Face of support ph /8 = 7.3”
Stirrup spacing (in.)
d
6 5”
1.8’ Spacing of #3
4 4”
closed stirrups
3”
2 12@312 ” 12@4” 15@5”
3’–6” 4’–0 6’–3”
2 4 6 8 10 12 14
Feet
(d)
2 – #7,1 – #6, As = 1.64
1.5 1.34 1.57
2 – #7, As = 1.20
Area (sq in.)
0.92 1.29
1.0 1.14
Total area of bottom
0.88 bt + d ACI–9.7.5.3
reinforcement required
d 2.8’
0.5 for beam
1.8’
2 4 6 8 10 12 14
(e)
1.52 4 – #6, As = 1.76
1.5
1.47 bt + d ACI–9.7.5.3
Area (sq in.)
Total area of top

2 – #6, As = 0.88 reinforcement required
1.0 d
0.88 for beam
1.8’ 0.81
0.5
0.43 0.33
0.35 0.28
2 4 6 8 10 12 14
Feet
Figure 18.14.3 Reinforcement requirements for spandrel beam of Example 18.14.2.
Using the simplified expression of AC1-22.5.5.1 for Vc with λ = 1 gives
 V 
(
φ  c + 8 fc′ = 0.75 2 λ fc′ + 8 fc′ = 0.75 10 fc′ = 474 psi
 bw d 
) ( )
Since 474 psi exceeds 469 psi, the section has adequate size.
(e) Compute stirrup reinforcement required for shear.
1
φVc = φ2 λ fc′bw d = 0.75(2)(1) 4000 (12)21.5 1000 = 24.5 kips
(Continued)
785
18.14 EXAMPLES 785
Since Vu > φVc /2 (ACI-9.6.3.1), transverse shear reinforcement is required, but there is
no computed strength requirement because Vu < φVc. The combined shear and torsion
transverse reinforcement must be equal to or greater than the minimum requirement of
ACI-9.6.4.2.
(f) Compute transverse torsional reinforcement requirement. Using Eq. (18.7.7),
At T /φ
= u tan θ
s 2 Ao f yt
Ao = 0.85 Aoh = 0.85(178) = 151 sq in.
Taking θ = 45°, as permitted by ACI-22.7.6.1.2(a), gives at d from the face of support,
At T /φ 35.5(12,000) / 0.75
= u tan θ = = 0.031
s 2 Ao f yt 2(151)60, 000
(g) Compute transverse reinforcement for combined shear and torsion. The total trans-
Av 2 At
+ = 0 + 2(0.031) = 0.062 Controls!
s s
b
12
min( Av + 2 At ) / s = 0.75 fc′ w = 0.75 4000 = 0.0095
f yt 60, 000
but not less than
bw 12
50 = 50 = 0.010
f yt 60, 000
2(0.11)
max s = = 3.5 in.
0.062
The upper limit on spacing is given by ACI-9.7.6.3.3,
ph 58.5
max s = = = 7.3 in.
8 8
but not to exceed 12 in. In this case, the strength requirement controls.
Use #3 hoops at 3.5-in. spacing.
(h) Longitudinal torsional reinforcement. Using Eq. (18.7.16) gives at d from face of
support
 35.5 
(Tu /φ) ph   (12, 000)58.5
0.75 
At = cot θ = (1) = 1.83 sq in.
2 Ao f y 2(151)(60, 000)
Check Aℓ,min:
 At   25bw 25(12) 
 s = 0.031 ≥  f = 60, 000 = 0.005
   yt 
(Continued)
786
The strength-related value (0.031) for At /s exceeds 25 bw /fyt.
A,min = −  t  ph
fy  s fy

5 4000 (288)  60, 000 
= − 0.031(588.5)  = 1.52 − 1.81 = negative
60, 000  60, 000 
The strength requirement (1.83 sq in.) controls. This amount is to be distributed around
the perimeter of the section at a spacing not to exceed 12 in. (ACI-9.7.5.1). In this case,
bars must be placed at middepth,
At 1.83
= = 0.61 sq in.
3 3
Use 2–#5 bars at middepth (one at each side face).
(i) Total longitudinal reinforcement for flexure and torsion. At the face of support,
top steel = 0.92 (flexure) + 0.61(torsion) = 1.53 sq in.
1 
bottom steel = 0.31  of 1.24 flexure + 0.61 (torsion) = 0.92 sq in.
4 
At d from face of support,
top steel = 0.86 (flexure) + 0.61(torsion) = 1.47 sq in.
1 
bottom steel = 0.31  of 1.24 flexure + 0.61 (torsion) = 0.92 sq in.
4 
(j) Transverse steel requirement along the span. Using the same procedure as at the first
critical section, the requirements are computed as shown in Table 18.14.2. The max-
imum spacing curve and selected spacings are shown in Fig. 18.14.3(c). Since the
 Acp 2 
factored torsional moment Tu is never less than φλ fc′   = 4.6 ft-kips [see part
 pcp 
(c) above], closed stirrups are required along the entire span.
TABLE 18.14.2 TRANSVERSE STEEL REQUIREMENT ALONG THE SPAN

FOR EXAMPLE 18.14.2
Requiredb
Location from
Center of At 2At Av max s for
+
Support Vu Tu s s s #3 hoops
(ft) (kips) (ft-kips) (in.) (in.) (in.)
d (1.8) 19.4 35.5 0.031 0.062 3.5
4 17.7 33.0 0.0291 0.0582 3.8
6 15.5 28.9 0.0255 0.0510 4.3
8 13.3 24.7 0.0218 0.0436 5.0
10 11.1 20.6 0.0182 0.0364 6.0
12 8.9 16.4 0.0145 0.0290 7.3a
15 5.6 10.2 0.0090 0.0180 7.3a
a max ph /8 controls.

b (Aυ /s) = 0 for strength requirement because Vu < φ Vc. All values exceed the minimum (0.010) of ACI-9.6.4.2.
(Continued)
78
18.14 EXAMPLES 787
(k) Longitudinal steel requirement along the span. The flexural requirements have
already been shown in Fig. 18.14.3(a) and 18.14.3(b). The torsional requirements
are computed in Table 18.14.3. Because the member depth exceeds 12 in., the lon-
gitudinal torsional steel is placed one-third at each of top, bottom, and midheight.
The sums of the flexural requirement and Aℓ /3 are shown as the total requirement for
longitudinal steel in Fig. 18.14.3(d) and 18.14.3(e).
In determining the length of longitudinal bars, a conservative interpretation has been
made of ACI-9.7.5.3, which requires an extension of (bt + d) beyond the theoretical ter-
mination point, as in Fig. 18.14.3(d) and 18.14.3(e). The theoretical termination point is
taken as that for combined flexure and torsion. It is noted that in some cases ACI-9.7.3.5
regarding cutting bars in a tension zone may control. The crack control provisions of
ACI-24.3.2 must be checked, as well as the deflection if excessive deflection may cause
damage. The design details are shown in Fig. 18.14.4.
TABLE 18.14.3 LONGITUDINAL STEEL REQUIREMENT ALONG

THE SPAN FOR EXAMPLE 18.14.2
Location from Required Required Required

Center of Support Aℓa Aℓ,minb Aℓ /3 As(top) As(bottom)
(ft) (sq in.) (sq in.) (sq in.) (sq in.) (sq in.)
d (1.8) 1.83 negative 0.61 1.47 0.92

4 1.70 negative 0.57 0.88 0.88
6 1.49 0.03 0.50 0.81 1.14
8 1.28 0.24 0.43 0.43 1.29
10 1.06 0.46 0.35 0.35 1.34
12 0.85 0.67 0.28 0.28 1.42
15 0.53 0.99 0.33 0.33 1.57
a Eq. (18.7.16); from ACI Formula (22.7.6.1b).
b Eq. (18.13.8); ACI-9.6.4.3(a).
A B CL of span
6’– 0”
4’– 3”
4 – #6 2 – #6
2 – #7 2 – #7 + 1– #6
2’– 0” 12@3 12 = 3’– 6” 12@4 = 4’– 0” 15@5 = 6’– 3” #3
Closed stirrups
A B
4 – #6 6” 2 – #6
#5 #5
Slab reinforcement
2 – #7 to be well anchored
#7 within closed hoops
12”
#6 (design not shown)
Section A–A Section B–B
Figure 18.14.4 Design details for Example 18.14.2.

78
EXAMPLE 18.14.3
Redesign the spandrel beam shown in Fig. 18.14.2 taking the option permitted in ACI-
22.7.3.2 to allow for redistribution of forces. The torsional member is then designed for
 Acp 2 
a factored torsion Tu of φ 4 λ fc′   (ACI-9.4.4.4; ACI-22.7.3.2; ACI-22.7.5.1). Use
 pcp 
fc′ = 4000 psi (normal weight) psi and fy = fyt = 60,000 psi.
SOLUTION
(a) Flexural strength. Assume that the factored bending moment Mu [Fig. 18.14.2(b)]
and factored shear Vu [Fig. 18.14.2(b)] are not significantly affected by a change in
the assumption for torsional stiffness. Note that most traditional structural analysis
methods do not include the effect of torsional stiffness; thus, this approach is sim-
pler. On completion of the structural analysis, disregarding torsional stiffness, any
member that needs to undergo torsional deformation is then designed to include the
ACI Code–specified Tu.
For this example, the requirements for flexure are as in Example 18.14.2 and are shown
in Fig. 18.14.3(a) and 18.14.3(b).
(b) Factored shear Vu and factored torsion Tu for which strength must be provided. At d
from the face of support,
Vu = 19.4 kips
 Acp 2   2882  1
Tu = φ 4 λ fc′   = 0.75(4)(1) 4000   = 18.2 ft-kips
p
 cp   72  12 , 000
Note that the 18.2 ft-kips is significantly lower than the maximum factored torsional
moment Tu of 35.5 ft-kips, which was designed for at d from the face of support in
Example 18.14.2.
(c) Determine whether the section is large enough for combined shear and torsion.
2 2
 Vu   Tu ph   Vc 
 b d  +  1.7 A2  ≤ φ  b d + 8 fc′
w oh w
This check was made in Example 18.14.2, where the combined stress was close to the
maximum given by the above equation. From Example 18.14.2,
Aoh = 178 sq in.

ph = 58.5 in.
The effect of the integral slab is neglected.

Vu 19.4(1000)
= = 75 psi
bw d 12(21.5)
Tu ph 18.2(12, 000) 58.5
2
= = 237 psi
1.7 Aoh 1.7(178)2
2 2
 Vu   Tu ph 
 b d  +  1.7 A2  = (75) + (237) = 248 psi
2 2
w oh
(Continued)
789
18.14 EXAMPLES 789
and the limit for λ = 1 is
 V 
( ) (
φ  c + 8 fc′ = 0.75 2 λ fc′ + 8 fc′ = 0.75 10 fc′ = 474 psi
 bw d 
)
The 474-psi limit far exceeds the combined stress value of 237 psi. This would permit
a smaller cross section. Under the redistribution procedure of ACI-22.7.3.2, the section
could clearly be smaller than 12 × 24. However, use the larger section for the remainder
of this example.
(d) Compute stirrup reinforcement required for shear. In Example 18.14.2, it was shown
that there is no strength requirement for Vu alone because Vu < φVc. The combined
shear and torsion transverse reinforcement must be equal to or greater than the min-
imum requirement of ACI-9.6.4.2.
(e) Compute transverse torsional reinforcement requirement. Using Eq. (18.7.7),
At T /φ
= u tan θ
s 2 Ao f yt
Ao = 0.85 Aoh = 0.85(178) = 151 sq in.
Taking θ = 45°, as permitted by ACI-22.7.6.1.2(a), gives at d from the face of support,
At T /φ 18.2(12, 000) / 0.75

= u tan θ = = 0.0161
s 2 Ao f yt 2(151)60, 000
(f) Compute transverse reinforcement for combined shear and torsion. The total trans-
Av 2 At
+ = 0 + 2(0.0161) = 0.032 Controls!
s s
b
12
min( Av + 2 At ) /s = 0.75 fc′ w = 0.75 4000 = 0.0095
f yt 60, 000
but not less than
bw 12
50 = 50 = 0.010
f yt 60, 000
2(0.11)
max s = = 6.9 in.
0.032
The upper limit on spacing is given by ACI-9.7.6.3.3,
ph 58.5
max s = = = 7.3 in.
8 8
and always less than 12 in. In this case, the strength requirement controls. Thus for this
beam a practical spacing of 6 in. may be used for the entire beam.
Use #3 hoops at 6-in. spacing along the entire span.
(Continued)
790
(g) Longitudinal torsional reinforcement. Using Eq. (18.7.16) gives at d from face of
support
 18.2 
(Tu /φ) ph   (12, 000)58.5
0.75 
A = cot θ = (1) = 0.94 sq in.
2 Ao f y 2(151)(60, 000)
Check Aℓ, min
 At   25bw 25(12) 
 s = 0.0161 ≥  f = 60, 000 = 0.005
   yt 
The strength-related value (0.0161) for At /s exceeds 25bw /fyt .
A,min = −  t  ph
fy  s fy

5 4000 (288)  60, 000 
= − 0.0161(58.5)  = 1.52 − 0.94 = 0.58 sq in.
60, 000  60, 000 
The strength requirement (0.94 sq in.) controls. This amount is to be distributed around
the perimeter of the section at a spacing not to exceed 12 in. (ACI-9.7.5.1). In this case,
bars must be placed at middepth,
A 0.94
= = 0.31 sq. in
3 3
(a)
Face of support 1.45 1.55
1.5 1.30
Area (sq in.)
1.17
0.95 Total area of bottom
1.0 reinforcement required
0.62 for beam
0.5
(b) d
1.5
1.23
Area (sq in.)
Total area of top

1.17 reinforcement required
1.0
for beam
0.62
0.5 0.31
0.31
2 4 6 8 10 12 14
Feet
Figure 18.14.5 Longitudinal steel requirements for Example 18.14.3.
Note that the amount Aℓ will be divided into three equal parts, top of beam, bottom of
beam, and midheight in accordance with ACI-9.7.5.1.
Use 2–#4 bars at middepth (one at each side face).
(Continued)
791
(h) Total longitudinal reinforcement for flexure and torsion. At the face of support,
top steel = 0.92 (flexure) + 0.31 (torsion) = 1.23 sq in.

bottom steel = 0.31 (flexure) + 0.31 (torsion) = 0.62 sq in.
At d from face of support,

top steel = 0.86 (flexure) + 0.31 (torsion) = 1.17 sq in.

bottom steel = 0.31 (flexure) + 0.31 (torsion) = 0.62 sq in.
For the top reinforcement, note that at d from the face, the ACI-9.6.1.2 minimum
was conservatively used instead of the 4 3 of a lesser value at location d permitted by
ACI-9.6.1.3.
(i) Longitudinal steel requirement along the span. The flexural requirements were
shown in Fig. 18.14.3(a) and 18.14.3(b). The sums of the flexural requirement and
Aℓ /3 are shown as the total requirement for longitudinal steel in Fig. 18.14.5. Note
that this alternate method requires many fewer hoops (i.e., #3 @ 6 throughout);
likewise, less longitudinal reinforcement is needed. When an actual redistribu-
tion is made, the shear and flexure requirements will increase somewhat; however,
for this example, there will still be considerable difference (primarily in closed
stirrups).
SELECTED REFERENCES
General
18.1 Paul Zia. “Torsion Theories for Concrete Members,” Torsion of Structural Concrete (SP-18).
18.2 Paul Zia. “What Do We Know about Torsion in Concrete Members?” Journal of the Structural
Division, ASCE, 96, ST6 (June 1970), 1185–1199.
18.3 K. G. Tamberg and P. T. Mikluchin. “Torsional Phenomena Analysis and Concrete Structure
Design,” Analysis of Structural Systems for Torsion (SP-35). Detroit: American Concrete
Institute, 1973 (pp. 1–102).
18.4 Paul Lampert. “Postcracking Stiffness of Reinforced Concrete Beams in Torsion and Bending,”
Analysis of Structural Systems for Torsion (SP-35). Detroit: American Concrete Institute, 1973
(pp. 385–433). (Also presented at 1971 Annual Convention, American Concrete Institute,
Denver, March 1971.)
18.5 Michael P. Collins and Paul Lampert. “Redistribution of Moments at Cracking—The Key to
Simpler Torsion Design,” Analysis of Structural Systems for Torsion (SP-35). Detroit: American
18.6 David J. Victor, Narayanan Lakshmanan, and Narayanan Rajagopalan. “Ultimate Torque of
Reinforced Concrete Beams,” Journal of the Structural Division, ASCE, 102, ST7 (July 1976),
1337–1352.
18.7 Arthur E. McMullen and B. Vijaya Rangan. “Pure Torsion in Rectangular Sections— A
Re-Examination,” ACI Journal, Proceedings, 75, October 1978, 512–519.
18.8 Madhusudan Chakraborty. “Ultimate Torque of Reinforced Rectangular Beams,” Journal of the
Structural Division, ASCE, 105, ST3 (March 1979), 653–668.
18.9 Michael P. Collins and Denis Mitchell. “Shear and Torsion Design of Prestressed and Non-
Prestressed Concrete Beams,” PCI Journal, 25, September/October 1980, 32–100. Disc., 26,
November/December 1981, 96–118.
18.10 J. Warwaruk. “Torsion in Reinforced Concrete,” Significant Developments in Engineering
Practice and Research (SP-72). Detroit: American Concrete Institute, 1981 (pp. 247–277).
18.11 Himat T. Solanki. “Reinforced Concrete Beams in Pure Torsion,” Journal of the Structural
Division, ASCE, 108, ST12 (December 1981), 2243–2250.
792
18.12 Thomas T. C. Hsu. Torsion of Reinforced Concrete. New York: Van Nostrand Reinhold, 1983.
18.13 Thomas T. C. Hsu and Y. L. Mo. “Softening of Concrete in Torsional Members—Theory and
Tests,” ACI Journal, Proceedings, 82, May–June 1985, 290–303. Disc., 83, July–August 1986,
690–693.
18.14 Peter Marti and Keith Kong. “Response of Reinforced Concrete Slab Elements to Torsion,”
Journal of Structural Engineering, ASCE, 113, 5 (May 1987), 976–993.
18.15 Peter Marti, Peter Leesti, and Waseem U. Khalifa. “Torsion Tests on Reinforced Concrete Slab
Elements,” Journal of Structural Engineering, ASCE, 113, 5 (May 1987), 994–1010.
18.16 Thomas T. C. Hsu. “Softened Truss Model Theory for Shear and Torsion,” ACI Structural
18.17 Atef H. Bakhsh, Faisal F. Wafa, and Ali A. Akhtaruzzaman. “Torsional Behavior of Plain High-
Strength Concrete Beams,” ACI Structural Journal, 87, September–October 1990, 583–588.
18.18 Thomas T. C. Hsu. Unified Theory of Reinforced Concrete. Boca Raton, FL: CRC Press, 1993.
18.19 Joint ACI-ASCE Committee 445. “Report on Torsion in Structural Concrete,” Report 445.1R-
12. Farmington Hills, MI: American Concrete Institute 2013, 92 pp.
Skew Bending Concept

18.20 N. N. Lessig. “Determination of Load Carrying Capacity of Reinforced Concrete Element with
Rectangular Cross Section Subjected to Flexure with Torsion,” Work No. 5. Moscow: Institut
Betonai Zhelezobetona (Concrete and Reinforced Concrete Institute), 1959 (pp. 5–28). (Also
available as Foreign Literature Study No. 371, PCA Research and Development Laboratories,
Skokie, IL.)
18.21 Thomas T. C. Hsu. “Torsion of Structural Concrete—Plain Concrete Rectangular Sections,”
Torsion of Structural Concrete (SP-18). Detroit: American Concrete Institute, 1968 (pp. 203–
238). (Also Portland Cement Association Development Department Bulletin D134.)
18.22 Thomas T. C. Hsu. “Torsion of Structural Concrete—A Summary of Pure Torsion,” Torsion of
Structural Concrete (SP-18). Detroit: American Concrete Institute, 1968 (pp. 165–178). (Also
Portland Cement Association Development Department Bulletin D133.)
18.23 Thomas T. C. Hsu. “Ultimate Torque of Reinforced Rectangular Beams,” Journal of the
Structural Division, ASCE, 94, ST2 (February 1968), 485– 510. (Also Portland Cement
Association Development Department Bulletin D127.)
18.24 Thomas T. C. Hsu. “Torsion of Structural Concrete—
Behavior of Reinforced Concrete
Rectangular Members,” Torsion of Structural Concrete (SP-18). Detroit: American Concrete
Institute, 1968 (pp. 261–306). (Also Portland Cement Association Development Department
Bulletin D135.)
18.25 C. D. Goode and M. A. Helmy. “Ultimate Strength of Reinforced Concrete Beams in Combined
Bending and Torsion,” Torsion of Structural Concrete (SP-18). Detroit: American Concrete
Institute, 1968 (pp. 357–377).
18.26 M. P. Collins, P. F. Walsh, F. E. Archer, and A. S. Hall. “Ultimate Strength of Reinforced
Concrete Beams Subjected to Combined Torsion and Bending,” Torsion of Structural Concrete
18.27 Kevin D. Below, B. Vijaya Rangan, and A. Stanley Hall. “Theory for Concrete Beams in
Torsion and Bending,” Journal of the Structural Division, ASCE, 101, ST8 (August 1975),
1645–1660.
Space Truss Analogy

18.28 E. Rausch. Berechnung des Eisenbetons gegen Verdrehung und Abscheren (Design of Reinforced
Concrete for Torsion and Shear). Berlin: Springer-Verlag, 1929.
18.29 Paul Lampert and Bruno Thürlimann. “Ultimate Strength and Design of Reinforced Concrete
Beams in Torsion and Bending,” Publications, International Association for Bridge and
Structural Engineering, 31-1, 1971, 107–131.
18.30 Paul Lampert and Michael P. Collins. “Torsion, Bending and Confusion—An Attempt to
Establish the Facts,” ACI Journal, Proceedings, 69, August 1972, 500–504.
18.31 CEB-FIP. Model Code for Concrete Structures, CEB-FIP International Recommendations (3rd
ed.). Paris: Comité Euro-International du Beton, 1978, 348 pp.
18.32 Denis Mitchell and Michael P. Collins. “Diagonal Compression Field Theory—A Rational
Model for Structural Concrete in Pure Torsion,” ACI Journal, Proceedings, 71, August 1974,
396–408.
793
18.33 P. Müller. “Failure Mechanisms for Reinforced Concrete Beams in Torsion and Bending,”
Publications, International Association for Bridge and Structural Engineering, 36-II, 1976,
146–163.
18.34 B. G. Rabbat and M. P. Collins. “A Variable Angle Space Truss Model for Structural Concrete
Members Subjected to Complex Loading,” Douglas McHenry International Symposium
on Concrete and Concrete Structures (SP-55). Detroit: American Concrete Institute, 1978
(pp. 547–587).
18.35 Thomas T. C. Hsu. “Shear Flow Zone in Torsion of Reinforced Concrete,” Journal of Structural
Engineering, ASCE, 116, 11 (November 1990), 3206–3226.
Combined Bending and Torsion

18.36 Hans Gesund, Frederick J. Schuette, George R. Buchanan, and George A. Gray. “Ultimate
Strength in Combined Bending and Torsion of Concrete Beams Containing Both Longitudinal
and Transverse Reinforcement,” ACI Journal, Proceedings, 61, December 1964, 1509–1522.
18.37 John P. Klus and C. K. Wang. “Torsion in Grid Frames,” Torsion of Structural Concrete (SP-
18). Detroit: American Concrete Institute, 1968 (pp. 89–101).
18.38 G. S. Pandit and Joseph Warwaruk. “Reinforced Concrete Beams in Combined Bending and
Torsion,” Torsion of Structural Concrete (SP-18). Detroit: American Concrete Institute, 1968
(pp. 133–163).
18.39 A. A. Gvozdez, N. N. Lessig, and L. K. Rulle. “Research on Reinforced Concrete Beams under
Combined Bending and Torsion in the Soviet Union,” Torsion of Structural Concrete (SP-18).
18.40 David J. Victor and Phil M. Ferguson. “Reinforced Concrete T-Beams without Stirrups under
Combined Moment and Torsion,” ACI Journal, Proceedings, 65, January 1968, 29–36. Disc.,
65, 560–566.
18.41 V. Ramakrishnan and Y. Ananthanarayana. “Tests to Failure in Bending and Torsion of

Reinforced Concrete,” ACI Journal, Proceedings, 66, May 1969, 428–431. Disc., 943–944.
18.42 D. W. Kirk and S. D. Lash. “T-Beams Subject to Combined Bending and Torsion,” ACI Journal,
Proceedings, 68, February 1971, 150–159.
18.43 Hota V. S. GangaRao and Paul Zia. “Rectangular Prestressed Beams in Torsion and Bending,”
Journal of the Structural Division, ASCE, 99, ST1 (January 1973), 183–198.
18.44 David J. Victor and P. K. Aravindan. “Prestressed and Reinforced Concrete T-Beams under
Combined Bending and Torsion,” ACI Journal, Proceedings, 75, October 1978, 526–532.
18.45 P. D. Zararis and G. Gr. Penelis. “Reinforced Concrete T-Beams in Torsion and Bending,” ACI
Journal, Proceedings, 83, January–February 1986, 145–155.
18.46 Ming B. Leung and William C. Schnobrich. “Reinforced Concrete Beams Subjected to Bending
and Torsion,” Journal of Structural Engineering, ASCE, 113, 2 (February 1987), 307–321.
Shear and Torsion

18.47 Ugur Ersoy and Phil M. Ferguson. “Behavior and Strength of Concrete L-Beams under
Combined Torsion and Shear,” ACI Journal, Proceedings, 64, December 1967, 797–801. Disc.,
65, 477–479.
18.48 John P. Klus, “Ultimate Strength of Reinforced Concrete Beams in Combined Torsion and
Shear,” ACI Journal, Proceedings, 65, March 1968, 210–215. Disc., 786–791.
18.49 Huey Ming Liao and Phil M. Ferguson. “Combined Torsion in Reinforced Concrete L-Beams
with Stirrups,” ACI Journal, Proceedings, 66, December 1969, 986–993. Disc., 67, 475–478.
18.50 Ahmed A. Ewida and Arthur E. McMullen. “Concrete Members under Combined Torsion and
Shear,” Journal of the Structural Division, ASCE, 108, ST4 (April 1982), 911–928.
Torsion, Shear, and Bending

18.51 Larry E. Farmer and Phil M. Ferguson. “T-Beams under Combined Bending, Shear and
Torsion,” ACI Journal, Proceedings, 64, November 1967, 757–766. Disc., 65, 417–421.
18.52 E. L. Kemp. “Behavior of Concrete Members Subject to Torsion and to Combined Torsion,
Bending, and Shear,” Torsion of Structural Concrete (SP-18). Detroit: American Concrete
Institute, 1968 (pp. 179–201).
794
18.53 Thomas T. C. Hsu. “Torsion of Structural Concrete—Interaction Surface for Combined Torsion,
Shear, and Bending in Beams Without Stirrups,” ACI Journal, Proceedings, 65, January 1968,
51–60. Disc., 566–572.
18.54 M. S. Mirza and J. O. McCutcheon. Discussion of “Torsion of Structural Concrete—Interaction
Surface for Combined Torsion, Shear, and Bending in Beams Without Stirrups,” ACI Journal,
Proceedings, 65, July 1968, 567–570.
18.55 David J. Victor and Phil M. Ferguson. “Beams under Distributed Load Creating Moment,
Shear, and Torsion,” ACI Journal, Proceedings, 65, April 1968, 295–308. Disc., 892–894.
18.56 Michael P. Collins, Paul F. Walsh, and A. S. Hall. Discussion of “Ultimate Strength of
Reinforced Concrete Beams in Combined Torsion and Shear,” ACI Journal, Proceedings, 65,
September 1968, 786–788.
18.57 D. L. Osburn, B. Mayoglou, and Alan H. Mattock. “Strength of Reinforced Concrete Beams
with Web Reinforcement in Combined Torsion, Shear, and Bending,” ACI Journal, Proceedings,
66, January 1969, 31–41. Disc., 593–595.
18.58 M. S. Mirza and J. O. McCutcheon. “Behavior of Reinforced Concrete Beams under Combined
Bending, Shear, and Torsion,” ACI Journal, Proceedings, 66, May 1969, 421–427. Disc.,
940–942.
18.59 Alfred Bishara. “Prestressed Concrete Beams under Combined Torsion, Bending, and Shear,”
ACI Journal, Proceedings, 66, July 1969, 525–538. Disc., 67, 61–63.
18.60 Arthur E. McMullen and Joseph Warwaruk. “Concrete Beams in Bending, Torsion, and Shear,”
18.61 Umakanta Behera and Phil M. Ferguson. “Torsion, Shear, and Bending on Stirruped L-Beams,”
Journal of the Structural Division, ASCE, 96, ST7 (July 1970), 1271–1286.
18.62 Einar Gausel. “Ultimate Strength of Prestressed I-Beams under Combined Torsion, Bending,
and Shear,” ACI Journal, Proceedings, 67, September 1970, 675–678.
18.63 Priya R. Mukherjee and Joseph Warwaruk. “Torsion, Bending, and Shear in Prestressed
Concrete,” Journal of the Structural Division, ASCE, 97, ST4 (April 1971), 1063–1079.
18.64 P. K. Syamal, M. S. Mirza, and D. P. Ray. “Plain and Reinforced Concrete L-Beams under
Combined Flexure, Shear, and Torsion,” ACI Journal, Proceedings, 68, November 1971,
848–860.
18.65 K. S. Rajagopalan and Phil M. Ferguson. “Distributed Loads Creating Combined Torsion,
Bending, and Shear on L- Beams with Stirrups,” ACI Journal, Proceedings, 69, January
1972, 46–54.
18.66 K. S. Rajagopalan, Umakanta Behera, and Phil M. Ferguson. “Total Interaction Method
for Torsion Design,” Journal of the Structural Division, ASCE, 98, ST9 (September 1972),
2097–2117.
18.67 Thomas G. Barton and D. Wayne Kirk. “Concrete T-Beams Subject to Combined Loading,”
Journal of the Structural Division, ASCE, 99, ST4 (April 1973), 687–700.
18.68 Robert L. Henry and Paul Zia. “Prestressed Beams in Torsion, Bending, and Shear,” Journal of
the Structural Division, ASCE, 100, ST5 (May 1974), 933–952.
18.69 Lennart Elfgren, Inge Karlsson, and Anders Losberg. “Torsion-Bending-Shear Interaction
for Concrete Beams,” Journal of the Structural Division, ASCE, 100, ST8 (August 1974),
1657–1676.
18.70 D. Wayne Kirk and David G. McIntosh. “Concrete L-Beams Subject to Combined Torsional
Loads,” Journal of the Structural Division, ASCE, 101, ST1 (January 1975), 269–282.
18.71 B. Vijaya Rangan and A. S. Hall. “Strength of Rectangular Prestressed Concrete Beams in
Combined Torsion, Bending and Shear,” ACI Journal, Proceedings, 70, April 1973, 270–278.
18.72 B. Vijaya Rangan and A. S. Hall. “Strength of Prestressed Concrete I-Beams in Combined
Torsion and Bending,” ACI Journal, Proceedings, 75, November 1978, 612–618.
18.73 K. S. Rajagopalan. “Combined Torsion, Bending, and Shear on L-Beams,” Journal of the
Structural Division, ASCE, 106, ST12 (December 1980), 2475–2492.
18.74 G. Taylor and J. Warwaruk. “Combined Bending, Torsion, and Shear of Prestressed Concrete
Box Girders,” ACI Journal, Proceedings, 78, September–October 1981, 335–340.
Design Methods
18.75 Alan H. Mattock. “How to Design for Torsion,” Torsion of Structural Concrete (SP-18).
18.76 ACI Committee 438. “Tentative Recommendations for the Design of Reinforced Concrete
Members to Resist Torsion,” ACI Journal, Proceedings, 66, January 1969, 1–8. Disc., 66,
576–588.
795
18.77 Thomas T. C. Hsu and E. L. Kemp. “Background and Practical Application of Tentative Design
Criteria for Torsion,” ACI Journal, Proceedings, 66, January 1969, 12–23. Disc., 591–593.
18.78 Michael P. Collins. Discussion of “Tentative Recommendations for the Design of Reinforced
Concrete Members to Resist Torsion,” ACI Journal, Proceedings, 66, July 1969, 577–579.
18.79 Umakanta Behera and K. S. Rajagopalan. “Two-Piece U-Stirrups in Reinforced Concrete
Beams,” ACI Journal, Proceedings, 66, July 1969, 522–524.
18.80 ACI Committee 438. Discussion of “Proposed Revision of ACI 318–
63: Building Code
Requirements for Reinforced Concrete,” ACI Journal, Proceedings, 67, September 1970,
686–689.
18.81 Michael P. Collins and Paul Lampert. “Designing for Torsion,” Structural Concrete Symposium,
Toronto: University of Toronto Civil Engineering Department, May 1971 (pp. 38–79).
18.82 David J. Victor. “Effective Flange Width in Torsion,” ACI Journal, Proceedings, 68, January
1971, 42–46.
18.83 Denis Mitchell and Michael P. Collins. “Detailing for Torsion,” ACI Journal, Proceedings, 73,
September 1976, 506–511.
18.84 Mahmoud A. Reda Youssef and Alfred G. Bishara. “Dowel Action in Concrete Beams Subject
to Torsion,” Journal of the Structural Division, ASCE, 106, ST6 (June 1980), 1263–1277.
18.85 Thomas T. C. Hsu and Y. L. Mo. “Softening of Concrete in Torsional Members—Design
Recommendations,” ACI Journal, Proceedings, 82, July–August 1985, 443–452. Disc., 83,
July–August 1986, 690–693.
18.86 J. G. MacGregor and M. G. Ghoneim. “Design for Torsion,” ACI Structural Journal, 92,
March–April 1995, 211–218.
Spandrel Beams
18.87 M. A. Gouda. “Distribution of Torsion and Bending Moments in Connected Beams and
Slabs,” ACI Journal, Proceedings, 56, February 1960, 757–774. Disc., 1425–1446.
18.88 Robert A. Shoolbred and Eugene P. Holland. “Investigation of Slab Restraint on Torsional
Moments in Fixed- Ended Spandrel Girders,” Torsion of Structural Concrete (SP-18).
18.89 Kolbjorn Saether and N. M. Prachand. “Torsion in Spandrel Beams,” ACI Journal,
18.90 James O. Jirsa, John L. Baumgartner, and Nathan C. Mogbo. “Torsional Strength and Behavior
of Spandrel Beams,” ACI Journal, Proceedings, 66, November 1969, 926–932. Disc., 67,
434–435.
18.91 James O. Jirsa. “Torsion in Floor Slab Structures,” Analysis of Structural Systems for Torsion
18.92 Mario G. Salvadori. “Spandrel-Slab Interaction,” Journal of the Structural Division, ASCE,
96, ST1 (January 1970), 89–106.
18.93 Umakanta Behera, K. S. Rajagopalan, and Phil M. Ferguson. “Reinforcement for Torque in
Spandrel L-Beams,” Journal of the Structural Division, ASCE, 96, ST2 (February 1970),
371–380.
18.94 Ugur Ersoy. “Distribution of Torsional and Bending Moments in Beam–Slab Systems,”
Analysis of Structural Systems for Torsion (SP-35). Detroit: American Concrete Institute,
1973 (pp. 293–324).
18.95 E. L. Kemp and W. J. Wilhelm. “Influence of Spandrel Beam Torsion on Slab Capacity Based
on Yield Line Criteria,” Analysis of Structural Systems for Torsion (SP-35). Detroit: American
18.96 Thomas T. C. Hsu and Kenneth T. Burton. “Design of Reinforced Concrete Spandrel Beams,”
Journal of the Structural Division, ASCE, 100, ST1 (January 1974), 209–229.
18.97 Thomas T. C. Hsu and Ching-Sheng Hwang. “Torsional Limit Design of Spandrel Beams,”
ACI Journal, Proceedings, 74, February 1977, 71–79.
18.98 Alfred G. Bishara, Larry Londot, Peter Au, and Majety V. Sastry. “Flexural Rotational
Capacity of Spandrel Beams,” Journal of the Structural Division, ASCE, 105, ST1 (January
1979), 147–161.
18.99 Charles H. Raths. “Spandrel Beam Behavior and Design,” PCI Journal, 29, March–April
1984, 62–131.
18.100 Gary J. Klein. “Design of Spandrel Beams,” PCI Journal, 31, September–October 1986,
76–124.
18.101 Y. L. Mo and Thomas T. C. Hsu. “Redistribution of Moments in Spandrel Beams,” ACI
796
Inverted T-Girders and Ledger Beams

18.102 Sher Ali Mirza and Richard W. Furlong. “Serviceability Behavior and Failure Mechanisms
of Concrete Inverted T-Beam Bridge Bentcaps,” ACI Journal, Proceedings, 80, July–August
1983, 294–304.
18.103 Sher Ali Mirza and Richard W. Furlong. “Strength Criteria for Concrete Inverted T-Girders,”
Journal of Structural Engineering, ASCE, 109, 8 (August 1983), 1836–1853.
18.104 Ned M. Cleland and Thomas T. Baber. “Behavior of Precast Reinforced Concrete Ledger
Beams,” PCI Journal, 31, March–April 1986, 96–117.
Channel-Shaped Sections
18.105 Petar Krpan and Michael P. Collins. “Predicting Torsional Response of Thin-Walled Open RC
Members,” Journal of the Structural Division, ASCE, 107, ST6(June 1981), 1107–1127.
18.106 Petar Krpan and Michael P. Collins. “Testing Thin-Walled Open RC Structure in Torsion,”
Journal of the Structural Division, ASCE, 107, ST6 (June 1981), 1129–1140.
18.107 Ching-Sheng Hwang and Thomas T. C. Hsu. “Mixed Torsion Analysis of Reinforced Concrete
Channel Beams—A Fourier Series Approach,” ACI Journal, Proceedings, 80, September–
October 1983, 377–386.
Beams with Openings

18.108 M. A. Mansur and A. Hasnat. “Concrete Beams with Small Opening under Torsion,” Journal
of the Structural Division, ASCE, 105, ST11 (November 1979), 2433–2447.
18.109 Mohammad A. Mansur, Seng Kiong Ting, and Seng-Lip Lee. “Ultimate Torque of R/C Beams
with Large Openings,” Journal of Structural Engineering, ASCE, 109, 8 (August 1983),
1887–1901.
18.110 M. A. Mansur. “Combined Bending and Torsion in Reinforced Concrete Beams with

Rectangular Openings,” Concrete International, 5, November 1983, 51–58.
18.111 M. A. Mansur and P. Paramasivam. “Reinforced Concrete Beams with Small Opening in
Bending and Torsion,” ACI Journal, Proceedings, 81, March–April 1984, 180–185.
18.112 Abul Hasnat and Ali A. Akhtaruzzaman. “Beams with Small Rectangular Opening under
Torsion, Bending, and Shear,” Journal of Structural Engineering, ASCE, 113, 10 (October
1987), 2253–2269.
18.113 W. A. M. Alwis and Mohammad A. Mansur. “Torsional Strength of R/C Beams Containing
Rectangular Openings,” Journal of Structural Engineering, ASCE, 113, 11 (November 1987),
2248–2258.
Torsion of Prestressed Concrete Beams

18.114 Alfred Bishara. “Prestressed Concrete Beams under Combined Torsion, Bending, and Shear,”
ACI Journal, Proceedings, 66, July 1969, 525–538.
18.115 Paul Zia and W. Denis McGee. “Torsion Design of Prestressed Concrete,” PCI Journal, 19,
March–April 1974, 46–65.
18.116 S. Timoshenko and J. N. Goodier. Theory of Elasticity (3rd ed.). New York: McGraw-Hill, 1970.
PROBLEMS
All problems are to be worked in accordance with the ACI Code, and all loads given are
service loads, unless otherwise indicated.
18.1 Determine the reinforcement required on a 12 × load. Use fc′ = 4000 psi (normal weight) and
22 in. overall size member to carry a torsional fy = fyt = 50,000 psi.
moment of 3 ft-kips dead load and 12 ft-kips 18.3 For the beam of the figure for Problem 18.3,
live load. Use fc′ = 4000 psi (normal weight) assume that 4–#9 bars are used in the bottom for
and fy = fyt = 60,000 psi. the main flexural reinforcement at midspan, with
18.2 Determine the reinforcement required for 2–#9 bars extended into the support and prop-
the member in the figure for Problem 18.2 to erly anchored. Further assume that 2–#7 are used
carry a torsional moment of 20 ft- kips dead in the top at the supports. What is the nominal
79
PROBLEMS 797
bE = 90”
2 – #7
# 3 @ 5”
5”
24”
36”
4 – #9
4 – #10
2 – #9
15”
14”
Problem 18.2
Problem 18.3
torsional strength Tn of the section according to load, and also that the girder has uniform dead
the ACI Code, assuming no simultaneous trans- load (including its own weight) of 1 kip/ft. Use
verse shear? What is the factored negative bending fc′ = 3000 psi (normal weight) and fy = 40,000
moment Mu that might be permitted to act at the psi. Show design sketch.
supports simultaneously when Tu = φTn, according 18.6 Redesign the transverse reinforcement for the
to the logic of Section 18.8? Use fc′ = 4000 psi beam of Problem 5.1, except consider the total
(normal weight) and fy = fyt = 60,000 psi. loading to be acting along a line at a distance
18.4 Determine the reinforcement required (including of 2 in. from the midwidth of the beam. Use the
torsion) at midspan and at the supports for the beam rough approximation that equilibrium torsion
of Example 18.4.1 (2B4 of Fig. 18.4.1). Assume is acting and that the torsional moment per unit
the shear is the same as that of a simply supported length equals the uniform loading times 2 in.
beam carrying a uniform dead load of 0.65 kip/ 18.7 Redesign the transverse reinforcement for the
ft and a uniform live load of 0.65 kip/ft. Assume beam of Problem 5.8, except consider the line
that flexure alone requires longitudinal steel areas uniform loading to be acting at 3 in. from the
of 1.60 and 2.30 sq in., for positive and negative centerline of the beam cross section. Assume
moment regions, respectively. Use fc′ = 3000 psi equilibrium torsion and use the same approxi-
(normal weight) and fy = fyt = 40,000 psi. Show mation as in Problem 18.6.
sketches of the cross sections. 18.8 Design the reinforcement for the edge beam that
18.5 Design the reinforcement to include torsion on is continuous at both ends shown in the figure
the spandrel girder 2G4 of Fig. 18.4.1. Assume for Problem 18.8. Assume the torsional moment
simple beam shears for this span, which is con- is 76 ft-kips (50% live load) at the face of col-
tinuous at both ends. Assume the reactions to umn and that the torsional moment varies pro-
girder 2G4 from beams 2B1 are concentrated portionally with the flexural shear (this problem
loads of 12.3 kips dead load and 15.9 kips live is similar to that of Hsu and Kemp [18.77]).
Dead load
1.2 kips/ft
(both cantilevers)
120 psf (live load)
18”
12’–0”
4”
16”
30’–0”
4” 14” 4”
Typical joist cross section
28’–8” 18 × 18 2’–0”
columns
Joists and slab
28’–0”
Max. effective = 3hf

A A
hf = 4”
16”
Edge beam 20”
16” 16”
Beam cross section (Section A–A)
Problem 18.8
798
Spandrel beam
(25-ft span)
8”
Symmetrical 60’–0”
6’–3”
about CL
span of
R1 R1 R1 R1 double tees
4’– 0” 4’–0” 4’– 0”
12”
12’– 6” Reactions R1
from tees
R2 (reaction to
the corbel) 16”
1’–0”
(a) Spandrel beam span and loading
Corbel 8”
projecting
from 24” R2
square
column
10”
(b) Spandrel beam cross section (trial)
18.9 Design the parking garage spandrel beam in the member. The final design for a nonprestressed
figure for Problems 18.9 and 18.10 (from Ref. member may require changes from the prelimi-
18.9). The floor system, consisting of double nary given dimensions; those given dimensions
tees spanning 60 ft, carries a live load of 50 psf. may serve as guides for arriving at reasonable
The double tees are supported on the ledge of the proportions.)
spandrel beam. The contribution to R1 from the 18.10 Design a spandrel beam of similar L-shape to
dead load of the floor system is 10.7 kips, and that of Problem 18.9 for loads R1 of 5.5 kips
from the 50 psf live load is 6.0 kips. Consider the dead load and 4.0 kips live load, plus the span-
given cross section as a preliminary trial section, drel girder weight. The span of the spandrel
with the 8-in. eccentricity of R1 with respect to and the locations of the R1 loads are the same
R2 held as constant. The spandrel member will as in Problem 18.9. The eccentricity e is 8 in.
have an attachment to the supporting column, (same as for Problem 18.9). Consider the 8 in.
providing lateral support at each vertical sup- used in Problem 18.9 as the minimum thick-
port of the spandrel girder. Use fc′ = 4000 psi ness; however, the other dimensions should
(normal weight) and fy = fyt = 60,000 psi and the be appropriate for the loads given in this prob-
ACI Code. (Note: The preliminary section is lem. Use fc′ = 4000 psi (normal weight) and
from Ref. 18.9 and is for a prestressed concrete fy = fyt = 60,000 psi.
CHAPTER 19
FOOTINGS
19.1 PURPOSE OF FOOTINGS
Footings are structural elements that transmit to the soil column loads, wall loads, or lateral
loads such as from retained earth. If these loads are to be properly transmitted, footings
must be designed to prevent excessive settlement or rotation, to minimize differential set-
tlement, and to provide adequate safety against sliding and overturning.
19.2 BEARING CAPACITY OF SOIL

It is not within the scope of this text to discuss the details of arriving at the bearing capacity
of the soil, but there must be reliable information on the safe bearing capacity of the soil
prior to the design of a footing. The allowable bearing capacity of soil is usually determined
by the ruling building code, by comparison with existing footings and with related informa-
tion in the area, by close examination of the soil and study of logs of test borings, by the
application of the science of soil mechanics, by load test, or by combinations of the various
sources and methods mentioned here.
The following are some of the most common reasons for the many uncertainties con-
cerning soil behavior under a footing.
1. There may be wide variations in soil types, which depend on their geological source,
mode of transportation, and sedimentation mechanism.
2. The physical properties and probable behavior under load are unknown and may
require extensive testing.
3. Frost action may cause heaving or subsidence.
4. Vibration may cause consolidation of granular material, which results in nonuniform
settlement.
5. Man-made hazards may exist below the earth surface, such as rock heaps, old sewers,
and questionable fill.
Soil pressure is commonly used in design under a working stress philosophy (Section
2.4). The soil mechanics/foundations specialist (geotechnical engineer) will establish the
ultimate bearing capacity, applying the appropriate margin for safety, and will specify a
service load bearing capacity (allowable bearing capacity) to be used in design.
In general, rock is considered the best foundation material; graded sand and gravel are
good materials; fine particles of sand and silt are generally questionable; and clay should
be studied carefully. The allowable bearing capacity used for design may range from
12,000 psf (575 kN/m2) or higher for rock to 2000 psf (96 kN/m2) for soft clay or silty clay.
Soils unable to carry 2000 psf generally require piling.
80
800 C hapter 1 9 F ootings
Wall and square spread footings visible during construction of the Kurt F. Wendt Engineering
Library (now Wendt Commons Engineering Library) at the University of Wisconsin–Madison.
(Photo by C. G. Salmon.)
19.3 TYPES OF FOOTINGS
Most building footings may be classified as one of the following types (Fig. 19.3.1).
1. Isolated spread footings under individual columns. These may be square, rectangu-
lar, or occasionally circular in plan.
2. Wall footings, either flat or stepped, which support bearing walls.
3. Combined footings supporting two or more column loads. These may be continuous
with a rectangular or trapezoidal plan, or they may be isolated footings joined by a
beam. The latter case is referred to as a strap, or cantilever, footing.
4. A mat foundation, which is one large continuous footing supporting all the columns
of the structure. This is used when soil conditions are poor but piles are not used.
5. Pile caps, structural elements that tie a group of piles together. These may support
bearing walls, isolated columns, or groups of several columns.
19.4 TYPES OF FAILURE
The procedures used for the design of footings in the United States are based primarily on
the work of Talbot in 1907 [19.1], Richart in 1946 [19.2], and Moe in 1957–1959 [19.3].
Moe defines the several types of failure that may occur in a slab acted on by concen-
trated loads. These failure modes are related to the shear span to depth (a/d) ratio, that is,
801
19.4 TYPES OF FAILURE 801
(a) Isolated spread footing (b) Wall footing
Property line
A B A B
(c) Combined footing: rectangular, PA = PB (d) Combined footing: rectangular, PB > PA
Property line
Property line
A B A B
(e) Combined footing: trapezoidal, PB > PA (f) Combined footing: trapezoidal, PA > PB
Property line
(g) Combined footing: strap or cantilever
Figure 19.3.1 Types of footings.
Mu /Vud, and are similar to those described for beams in Section 5.4 (Fig. 5.4.4). The failure
mechanisms may be reviewed as follows.
1. Shear-compression failure [Fig. 19.4.1(a)]. Typical with deep sections of short span

(low a/d ratios), inclined cracks form that do not cause failure but do extend into
the compression zone, thus reducing its size until finally the compression zone fails
under the combined compressive and shear stresses.
2. Flexure failure after inclined cracks form. Also typical with low a/d ratios, inclined
cracks that form first do not cause failure or prevent the development of the nominal
moment strength. If embedment of tension steel is adequate, and no failure in the
compression zone occurs, the tension steel may reach its yield strength.
3. Diagonal tension failure [Fig. 19.4.1(b)]. Commonly called punching shear (Section
16.15), this failure mode is typical with medium-span average depth sections (inter-
mediate values of a/d); the footing fails on formation of the inclined cracks around the
perimeter of the concentrated load. Test results indicate that the critical section can
reasonably be considered at d/2 from the periphery of a column or concentrated load.
4. Flexure failure before inclined cracks form. Typical with large values of a/d, no
inclined cracks form before flexural strength is reached.
In the design of a footing as well as of a beam, a shear failure should not occur prior to
reaching the member flexural strength.
802
d/2 d/2
(a) Shear-compression failure (b) Diagonal tension failure
Figure 19.4.1 Shear-related failure mechanisms in footings.
19.5 SHEAR STRENGTH
Strength requirements for footings are given in Chapter 13 of the ACI Code. Most of these
requirements, however, are given in reference to Chapter 7 (one-way slabs) or Chapter 8
(two-way slabs) of the ACI Code as appropriate. The shear strength of footings may be
computed as that of two-way slab systems, as discussed in Sections 16.15 and 16.18. The
six principal variables involved in the strength of slabs without shear reinforcement are
(a) the concrete strength fc′; (b) the ratio of the side length c of the loaded area to the effec-
tive depth d of the slab; (c) the relationship (V/M) between shear and moment near the
critical section; (d) the column shape in terms of the ratio β of the long side to the short
side of the rectangular column; (e) lateral restraints such as may be provided by stiff beams
along the boundaries of a slab (no such restraints would normally be acting in the case of
footings); and (f) the rate of loading.
As discussed in Section 16.15, the nominal punching shear strength Vc when no shear
reinforcement is used (i.e., when Vn = Vc) is the smallest of
Vc = 4 λ fc′b0 d (19.5.1a)*
 4
Vc =  2 +  λ fc′b0 d (19.5.1b)*
 β
 α d
Vc =  2 + s  λ fc′b0 d (19.5.1c)*
 b0 
where
b0 = perimeter of critical section taken at d/2 from the loaded area (ACI-13.2.7.2)
d = effective depth of slab
β = ratio of long side to short side of the loaded area
αs = 40 for interior columns, 30 for edge columns, and 20 for corner columns
In the application of Eq. (19.5.1c), αs for an “interior column” applies when the perime
ter is four-sided, for an “edge column” when the perimeter is three-sided, and for a “corner
column” when the perimeter is two-sided.
* For SI, with fc′ in MPa and b0 and d in mm, ACI-318-14M gives

Vc = 0.33λ fc′ b0 d (19.5.1a)
 2
Vc = 0.17  1 −  λ fc′ b0 d (19.5.1b)
 β
 α d
Vc = 0.083  2 + s  λ fc′ b0 d (19.5.1c)
 b0 
803
19.6 FLEXURAL STRENGTH 803
Critical section
c
W
d
d
(a) One-way action
Critical
section
c+d c
W
d/2 d/2
L
(b) Two-way action
Figure 19.5.1 Critical section for shear (inclined cracking) in footings.
For footings in which the bending action is primarily in one direction, the procedure
used for one-way slabs should be applied as described in Chapter 8. The critical section for
such an investigation is to be taken at a distance d from the column face (ACI-13.2.7.2).
Figure 19.5.1 summarizes the critical sections to be investigated in regard to shear.
19.6 FLEXURAL STRENGTH AND DEVELOPMENT

OF REINFORCEMENT
Research has shown that critical sections for flexural strength and development of rein-
forcement occur at the face of a reinforced concrete column or wall. The bending in each
direction should be considered separately. Richart’s tests [19.2] showed that, under service
loads, the moment is greater on a strip under the column load than it is out near the corners.
However, failure in flexure does not occur until all of the steel has reached yielding—that
is, after some redistribution of load has taken place. Gesund [19.5, 19.6], Jiang [19.7], and
Subba Rao and Singh [19.8] have presented yield line analyses (see Chapter 17) of footings.
The critical sections for moment and development of reinforcement, however, are to be
taken at (ACI-13.2.7.1) (a) the face of column, pedestal, or wall for footings supporting a
column, pedestal or wall, respectively, (b) halfway between the middle and the edge of the
wall for footings under masonry walls, and (c) halfway between the face of the column and
the edge of the base plate for footings under steel bases.
804
19.7 PROPORTIONING FOOTING AREAS

FOR EQUAL SETTLEMENT
Differential settlement between footings should be minimized because it can adversely affect
the strength of the structure as well as interfere with the fitting of such items as partitions,
doors, and ceilings. It is generally assumed that there will be equal settlement if the unit soil
pressures due to service loads are equal under all footings. However, the service loads in the
columns consist of dead and live loads. The dead load is always there, but in a tall building
there is little chance for a column to receive maximum live load from all floors. Therefore,
most building codes allow a reduction of live loads in columns. Typically, the total live load
would have to be carried by columns supporting the roof and one floor. As the number of
floors carried by a column increases, the percentage of the total live load that the column must
be designed to carry may be reduced from 100% progressively. This percentage might be a
minimum of 50% when at least 8 or 9 floors are carried by the column. The maximum soil
pressure under a footing is, then, due to the sum of the dead load in the column, the maximum
reduced live load in the column, and the weight of the footing itself.
Soil is a substance that may be relatively elastic, such as granular material like sand, or it may
be a relatively plastic substance, such as clay exhibiting time-dependent deformation under sus-
tained load. Often, for design purposes, equal settlement of footings is presumed when the soil
pressure due to sustained loads is the same under all footings. The sustained load may be taken
as the sum of the dead load in the column, the weight of the footing itself, and a certain per-
centage of the maximum reduced live load in the column. In such cases, the relative areas of the
footings should be so proportioned that the unit soil pressures under sustained loads would be
the same under all footings. This requirement is additional to the requirement that the soil pres-
sure under maximum possible load not exceed the allowable bearing capacity at each footing.
19.8 INVESTIGATION OF SQUARE SPREAD FOOTINGS

The investigation of square spread footings can best be treated by an illustrative example in
which the items considered are soil pressure under the footing, shear (inclined cracking),
bending moment, development of reinforcement, and load transfer from column to footing.
EXAMPLE 19.8.1
Check the adequacy of the square footing of Fig. 19.8.1, according to the ACI Code.
The column service level (unfactored) axial load is 300 kips dead load and 160 kips live
load. Both the column and the footing have fc′ = 3000 psi (normal-weight concrete) and
fy = 60,000 psi. The allowable soil pressure is 5000 psf. There is a 2-ft earth overburden
having a unit weight of 100 pcf.
SOLUTION
(a) Soil pressure. The action of soil on the footing is taken to be uniform for isolated
footings under concentric loads. The base of the footing must have an area large
enough so that the allowable soil pressure will not be exceeded under the action of
the column service load, footing weight, and weight of overburden.
Column (300 + 160) 460 kips
Footing weight 10(10)(2.08)0.150) 31
Earth (100 –4)2(0.100) 19
Total weight on soil 510 kips
510
soli pressure p = = 5.1 ksf ≈ 5 ksf OK
(10)(10)
(Continued)
805
1 9 . 8 I N V E S T I G AT I O N O F S QUA R E S P R E A D F O OT I N G S 805
2’– 0”
10’– 0”
2’– 0”
#9 column bars and #8 dowels
3” clear
2’– 1”
3” 3” clear
clear 8 – #8 bars each way
(spacing ~ 16”)
10’– 0”
Figure 19.8.1 Square spread footing for Example 19.8.1.
(b) Shear (inclined cracking) strength. Two possible critical sections must be inves-
tigated: one-way action as a beam and two-way action as a slab, as shown in
Fig. 19.4.1. The shear to be used is the upward soil pressure less the downward
overburden and the footing weight acting outside the critical section. Since the foot-
ing weight and the overburden are usually uniform, the forces acting on the footing
may be obtained by using the net upward pressure, which is caused by the column
load only.
Since the action of a square concentrically loaded footing is symmetrical about both
axes, the reinforcement in each direction is presumed to do the same work. However,
the effective depth cannot be the same for both directions since the bars must cross
each other. The average depth d is commonly used except for very shallow footings
(say, less than 15 in. deep) where the more conservative value should probably be used.
Using a 3-in. clear cover for concrete cast and permanently in contact with ground (ACI-
20.6.1.3.2), the average d is

d = 25 − 3 (i.e.,clear cover ) − 1 (i.e., bar diameter ) = 21 in.
For one-way action, the net earth pressure acting upward due to factored loads is
300(1.2) + 160(1.6)
pnet = = 6.16 ksf
100
(Continued)
806
Using the loaded area shown in Fig. 19.8.2(a) at a distance d from the face of the col-
umn, the factored shear is

Vu = ( pnet )(effective area ) = 6.16(2.25)10 = 139 kips
When no shear reinforcement is used, vc = 2 λ fc′ unless the more detailed procedure
is used; thus
1
Vn = Vc = 2 λ fc′bw d = 2(1.0) 3000 (120)(21) = 276 kips
1000

[φVn = 0.75(276) = 207 kips] > [Vu = 156 kips] OK
For two-way action, the factored shear [Fig. 19.8.2(b)] is

Vu = ( pnet )(effective area )

= 6.16 [100 − 3.75(3.75)] = 529 kips
When no shear reinforcement is used, the strength using Eqs. (19.5.1) is based on
vc = 4 λ fc′ when β ≤ 2 and b0 /d ≤ 20 for a four-sided critical section; thus
b0 /d = 4(3.75) /1.75 = 8.6 < 20
12
Vn = Vc = 4 λ fc′b0 d = 4(1.0) 3000 (4)(3.75)(21) = 828 kips
1000

[φVn = 0.75(828) = 621 kips] > [Vu = 529 kips] OK
Since isolated footings are rarely designed with shear reinforcement, Vc will usually
control the thickness.
(c) Flexural strength. The critical section and loaded area are as shown in Fig. 19.8.2(a).
The factored bending moment is
pnet b 2 6.16(10)(4.0)2
Mu = = = 493 ft-kips
2 2
Mu 493(12, 000 )
φbd 2 0.90(120)(21)2
With reference to Rn, note that footing thickness is rarely controlled by flexure; thus, the
reinforcement ratio ρ will be low enough that φ will be 0.90, except in rare cases where
ρ exceeds ρtc.
Referring to Section 3.8, Eq. (3.8.5),
1 2 mRn 
required ρ = 1 − 1 −  [3.8.5]
m fy 

1  2(23.5)124 
= 1− 1− = 0.0021
23.5  60, 000 
For spread footings, the minimum reinforcement (As min = 0.0018Ag for fy = 60 ksi) for
structural slabs of uniform thickness applies (ACI-8.6.1.1); thus
required As = 0.0021(120)21 = 5.3 sq in.

min required As = 0.0018(120)25 = 5.4 sq in. Controls!

provided As = 8(0.79) = 6.32 sq in.[> min required As = 5.4 sq in.] OK
(Continued)
807
1 9 . 8 I N V E S T I G AT I O N O F S QUA R E S P R E A D F O OT I N G S 807
Load area for one-way

shear action
Load area for two-way
1’– 0” shear action
2’– 0” + d
10’– 0”
3.75’
4’– 0” Load area for
moment
2.25’
d = 21”
1.75’
(a) (b)
Figure 19.8.2 Critical sections and loaded area for square spread footing.
Note that using 7–#8 bars would satisfy the requirement of ≈ 5.4 sq in.; however, they
would be spaced at (120 –6 –1)/6 = 18.8 in. which exceeds the maximum spacing of
18 in., per ACI-8.7.2.2.
(d) Development of reinforcement—
general equation. The general equation,
[Eq. (6.6.1)], is
 

Ld =  d [6.6.1]
 40 λ fc′  cb + K tr   b
  d  
 b 
The cover or spacing dimension cb is the smaller of (1) distance from center of bar being
developed to nearest concrete surface, and (2) one-half center-to-center spacing of bars
being developed. The distance cb is the smaller of the following two values:
bottom and side cover = 3.0 (i.e., clear) + 0.5 (i.e., bar radius) = 3.5 in.
(120 − 6 − 1)
one-half center-to-center spacing ≈ = 8.1 in.
2(7)
Thus, cb = 3.5 in. There are no stirrups; thus, Ktr = 0. Thus,
 cb + K tr 3.5 + 0 
 = = 3.5 > 2.5 max
 db 1.00 
Thus, (cb + Ktr)/db = 2.5, and noting that ψt = ψe= ψs = 1.0
 3 60, 000 1.0(1.0)1.0 

Ld (for #8) =   1.00 = 33 in. (2.75 ft)
 40 (1.0) 3000 2.5
Allowing a 3-in. cover on the end of the #8 bars, the embedment provided is

actual embedment = 48 − 6 = 42 in. > [ Ld = 33 in.] OK
(Continued)
80
(e) Load transfer from column to footing. ACI-16.3.1.1 requires factored forces and
moments acting at the column base to be transferred into the footing. Tensile forces,
if any, must be transferred by developed reinforcement, such as bar reinforcement,
dowels, or mechanical connectors; however, compressive forces may be transmitted
directly by bearing.
The nominal bearing stress fb that the base of the column can withstand is 0.85 fc′
(ACI-22.8.3.2). The nominal strength Pn in compression based on the column concrete
strength fc′ is
Pn = 0.85 fc′ Ag = 0.85(3)(576) = 1470 kips

Pu = 1.2(300) + 1.6(160) = 616 kips
Using φ = 0.65 per ACI-21.2.1, for bearing stresses

[φPn = 0.65(1470) = 956 kips] > [ Pu = 616 kips] OK
Regarding the bearing on the footing concrete, the capacity is increased because the footing
area is much larger than the column area, thus confining the loaded area and permitting a
distribution of the concentrated load. The bearing strength for the footing concrete is thus
based on 0.85 fc′ increased by a multiplier αb that varies between 1 and 2, as follows:
A2
αb = ≤ 2
A1
where A1 is the loaded area, which in this example is the 576 sq in. column area; A2 is
the area of the lower base of the largest frustrum of a pyramid, cone, or tapered wedge
contained wholly within the support and having for its upper base equal to the loaded
area, and having side slopes of 1 vertical to 2 horizontal (ACI-2.2). In this case, A2 is the
entire footing area, 14,400 sq in.
Since both column and footing contain the same strength concrete in this example,
only the check on bearing in the column was necessary. In general, however, the bear-
ing strength is to be computed as the lesser of the nominal concrete bearing strengths
of the supported member or the foundation surface (ACI-16.3.3.4). The bearing stress
on the footing may control when columns of high-strength concrete rest on footings
of low-strength concrete. Bearing in the bottom of the column is important unless the
longitudinal reinforcement is developed by extension into the footing or by embedding
dowels lapped to the column bars. In this case, the load can be carried without using
developed reinforcement.
When the transfer is made by bearing, as in this case, ACI-16.3.4.1 still requires for
cast-in-place columns and pedestals a minimum amount of reinforcement across the
joint between the column and footing. Extended longitudinal reinforcement or dowels at
least equal to 0.005 times the gross cross-sectional area of the supported member must
be provided. When dowels are used, they can be any size #11 and smaller. Though #14
and #18 column bars may be lap spliced at footings, dowels larger than #11 bars cannot
be used (ACI-16.3.5.4). Dowels comparable in size to the bars being developed should
generally be used. The minimum area of developed reinforcement required to cross the
interface of cast-in-place construction is thus (ACI-16.3.4.1)

required As = 0.005(576) = 2.88 sq in.
Using a practical minimum of four bars, one in each corner,
2.88
required As per bar = = 0.72 sq in.
4
(Continued)
809
1 9 . 9 D E S I G N O F S QUA R E S P R E A D F O OT I N G S 809
Thus the #8 dowels shown are adequate. These dowels must be embedded into the foot-
ing a distance equal to the development length Ld in compression for #8 bars. Using the
basic development length Ldc for compression bars not enclosed by transverse reinforce-
ment (ψr = 1.0),
 fy ψ r 
Ldc =   db [6.8.1]
 50 λ fc′ 

 60, 000(1.0) 
Ldc =  1.0 = 22 in.
 50(1.0) 3000 
but not less than
Ldc = 0.0003 f y ψ r db [6.8.2]

= 0.0003(60, 000)(1.0)(1.0) = 18.0 in.
nor less than 8 in. Thus,

Ld (# 8) = Ldc = 22 in. [ = (25 − 3) = 22 in. available] OK
If the available footing thickness is inadequate, a greater number of smaller diameter
bars should be used.
Thus the footing of Fig. 19.8.1 satisfies all ACI requirements.
19.9 DESIGN OF SQUARE SPREAD FOOTINGS

The design of square spread footings involves the determination of the size and depth of
the footing and the amount of main reinforcement and dowels so that all the requirements
described in the preceding sections are fulfilled. Example 19.9.1 illustrates the design
procedures.
EXAMPLE 19.9.1
Design a square spread footing to carry a column dead load of 197 kips and a live load
of 160 kips from a 16-in. square tied column containing #11 bars as the principal col-
umn steel. The allowable soil pressure is 4.5 ksf. Consider that there is a 2-ft overburden
weighing 100 pcf. Use fc′ = 3000 psi (normal-weight concrete) and fy = 60,000 psi.
SOLUTION
(a) Estimate the footing weight and determine the plan of the footing. The total weight
of the footing, plus any overburden, may be estimated and added to the column load
or, as an alternative, the effect of these items in terms of the unit soil pressure may
be estimated. In this case, the footing is estimated to be about 2 ft thick—that is,
300 psf, frequently the minimum used by designers. This leaves the net allowable
soil pressure that must carry the column load as
pnet = 4500 − 200 − 300 = 4000 psf

357
required A = = 89.3 sq ft
4.0
(Continued)
810
Try 9 ft 6 in. square, A = 90.3 sq ft. Note that ACI-13.3.1.1 requires the base area of a
footing to be determined by using service loads (unfactored loads) with the allowable
soil pressure. This is reasonable, since the allowable soil pressure should be determined
by using principles of soil mechanics and may incorporate varying factors of safety
depending on type of soil and condition of loading.
For the design of the reinforced concrete member, factored loads must be used.
Applying overload factors to the column load,
Pu = 1.2(197) + 1.6(160) = 492 kips
492
pnet = = 5.45 ksf
90.3
(b) Determine depth based on shear (inclined cracking) strength. In most cases the
depth necessary for shear without using stirrups controls the footing thickness.
For two-way action [Fig. 19.9.1(a)], assuming a thickness of 24 in.,

average d = 24 − 3 (cover) − 1 (bar diameter) ≈ 20 in.

Vu = ( pnet )(effective area) = 5.45[9.5(9.5) − 3.0(3.0)] = 443 kips
From Fig. 19.9.1(a),

b 0 /d = 4(16 + 20) / 20 = 7.2
For a four-sided critical section with β ≤ 2 and b0 /d ≤ 20,
1
Vn = Vc = 4 λ fc′b0 d = 4(1.0) 3000 [ 4(16 + 20)](20) 1000 = 631 kips

φ Vc = 0.75(631) = 473 kips > [Vu = 433 kips] OK
No shear reinforcement is required.
For one-way action [Fig. 19.9.1(b)],

Vu = 5.45(2.42)(9.5) = 125 kips
and
1
Vn = Vc = 2 λ fc′ bw d = 2(1.0) 3000 (9.5)(12)20 1000 = 250 kips

No shear reinforcement is required. Note that Vc is computed from the simplified proce-
dure. The 24-in. thickness is satisfactory for shear.
(c) Check transfer of load at base of column (ACI-16.3.1.1). The compressive design
strength φ Pn based on the nominal ultimate bearing stress 0.85 fc′ in the column is

φ Pn = φ (0.85 fc′ ) Ag = 0.65(0.85)(3)(256) = 424 kips
where φ = 0.65 for bearing strength.

φ Pn < [ Pu = 492 kips] NG
Thus the column load cannot be transferred by bearing alone. It may well be that the
minimum reinforcement across the interface required by ACI-16.3.4.1 will be adequate
to transfer the excess load.

min required As = 0.005(256) = 1.28 sq in.
(Continued)
81
9’– 6”
16” + d = 36”
29”
16” + d = 36”
2.42’
9’– 6” 20”
8”
(a) Two-way action (b) One-way action
Figure 19.9.1 Critical sections for shear in square footing design.
Neglecting the area of concrete displaced by the reinforcing bars, the excess load to be
carried by the dowels is
excess Pu = 492 − 424 = 68 kips
excess Pu
required As =
φ fy
68
= = 1.74 sq in.
0.65(60)
Logically, to compensate for the displaced concrete, a stress of 0.85 fc′ should be sub-
tracted from fy in the dowels. Thus, more correctly,
excess Pu
φ ( f y − 0.85 fc′ )
Use 4–#7 bars as dowels, As = 2.40 sq in.
The #7 dowels must be developed above and below the junction of column and footing.
The development length Ldc required for bars in compression not enclosed by transverse
reinforcement (ψr = 1.0) is
 fy ψ r  [6.8.1]
Ldc =   db
 50 λ fc′ 
 60, 000(1.0) 
Ldc =  0.875 = 19.2 in. Controls!
 50(1.0) 3000 
but not less than

Ldc = 0.0003 f y ψ r db = 0.0003(60, 000)(1.0)(0.875) = 15.8 in.
and final Ld cannot be less than 8 in. Thus, Ld = Ldc = 19.2 in. The 24-in.-thick footing is
thus adequate for straight dowels.
Hooks or bending of bars are not considered effective in adding to the compressive
resistance of bars (ACI-25.4.1.2). Very often engineers will specify bending of the dowels
as shown in Fig. 19.9.2 to prevent their being pushed through the footing during construc-
tion and thus reducing the effective embedment distance L2. In such cases of bending of
the dowels, full development of the compressive force is required over the distance L1.
(Continued)
812
L2
L1
Figure 19.9.2 Dowel anchorage.
For this design, if such a bend is used, the available length L1 is

L1 = 24 − 3 (cover) − 2(1)(footing bars) − 0.875 (dowels) ≈ 18 in.
This is less than the 19.2 in. required and is therefore unacceptable. Alternatives would
include a thicker footing, a larger number of smaller-sized dowels, or the use of a
pedestal.
Try 6–#5 bars as dowels instead, As = 1.86 sq in.
 60, 000(1.0) 
Ldc =  0.625 = 13.7 in. Controls!
 50(1.0) 3000 
but not less than

Ldc = 0.0003 f y ψ r db = 0.0003(60, 000)(1.0)(0.625) = 11.3 in.
and final Ld cannot be less than 8 in. Thus, Ld = Ldc = 13.7 in., which is less than the
available length L1 = 18 in.
Use 6–#5 bars as dowels.
(d) Design for flexural strength. The critical section for moment is at the face of the
column (Fig. 19.9.3).
9’– 6”
4’– 1” 1’– 4”
24”
Net upward
pressure = 5.45 ksf
1
Figure 19.9.3 Critical section for bending moment and development of reinforcement.

(Continued)
813
1
Mu = (5.45)(9.5)(4.08)2 = 431 ft-kips
2
Mu 431(12,000)
φ bd 2
0.90(114)(20)2
Referring to Section 3.8, Eq. (3.8.5),
1 2 mRn 
ρ= 1 − 1 −  [3.8.5]
m fy 
1  2(23.5)126 
=  1− 1− = 0.0022
23.5  60, 000 
required As = ρbd = 0.0022(114)(20) = 5.02 sq in.
min required As = 0.0018(114)(24) = 4.92 sq in.
Try 12–#6, As = 5.28 sq in.; average d = 24 – 3 – 0.75 = 20.3 in.
C = 0.85 fc′ ba = 0.85(3)(9.5)(12)a = 291a

T = As f y = 5.28(60) = 317 kips
C = T; a = 1.09 in.
1
φ M n = 0.90(317)[20.3 − 0.5(1.09)] = 470 ft-kips
12
φ M n > [ Mu = 431 ft-kips] OK
Use 12–#6 bars (As = 5.28 sq in.) each way.

(e) Development of reinforcement. The general equation, Eq. (6.6.1), often indicates a
smaller Ld than given by the simplified equations. Using Eq. (6.6.1),
 

Ld =  d [6.6.1]
 40 λ fc′  cb + K tr   b
  d  
 b 
The center-to-center spacing of the 12–#6 bars is approximately 9 3 4 -in. For the general
equation, the distance cb is the smaller of the following two values:
bottom and side cover = 3.0 (i.e., clear) + 0.375 (i.e., bar radius) = 3.38 in.
one-half center-to-center spacing ≈ 9.75/2 = 4.9 in.
Thus, cb = 3.38 in. There are no stirrups; thus, Ktr = 0. Thus,
 cb + K tr 3.38 + 0 
 = = 4.5 > 2.5 max
 d b 0 . 75 
Thus, (cb + Ktr)/db = 2.5.
(Continued)
814
Evaluate, using Eq. (6.6.1):
 3 60, 000 1.0(1.0)1.0 

Ld (for #6) =   0.75 = 24.6 in (2.05 ft)
 40 (1.0) 3000 2.5
Note that ψs was conservatively taken as 1.0, although ACI-25.4.2.4 allows a value
of 0.8; ψt and ψe are both equal to 1.0.
Allowing a 3-in. cover on the end of the #6 bars, the available embedment is

available embedment = 49 − 3 = 46 in. > [ Ld = 24.6 in.] OK
(f) Design sketch. A design sketch as shown in Fig. 19.9.4 is necessary to convey the
designer’s decision properly.
12 – #6
(spacing ≈ 9¾”)
each way
1’– 4”
1’– 4”
9’– 6”
3” clear cover
2’– 0”
12 – #6 bars each way
9’– 6”
Figure 19.9.4 Design sketch for spread footing of Example 19.9.1.
For practical design of square footings, design aids are available [2.23, 19.9].
19.10 DESIGN OF RECTANGULAR FOOTINGS

Rectangular footings may be used in locations where restricted space prevents the use of
a square footing. The procedure for their design is essentially identical to that of square
footings, except that one-way shear action and bending moment must be considered in both
principal directions.
815
1 9 . 1 0 D E S I G N O F R E C TA N G U L A R F O OT I N G S 815
EXAMPLE 19.10.1
Design a rectangular spread footing to carry 235 kips service dead load and 115 kips
service live load from an 18-in. square tied column that contains #9 bars. One dimen-
sion of the footing is limited to a maximum of 7 ft. The allowable soil pressure is 5500
psf. Neglect the effect of overburden. Use fc′ = 3000 psi (normal-weight concrete) and
fy = 60,000 psi.
SOLUTION
(a) Determine plan of footing. Assume a footing depth of 2 ft, or 300 psf.
net soil pressure = 5500 − 300 = 5200 psf
350
required A = = 67.3 sq ft
5.2
Space limitation prevents one dimension from exceeding 7 ft; thus
67.3
length = = 9.6 ft
7.0
Try 7 ft × 9 ft 8 in. (area = 67.7 sq ft).
(b) Determine depth required for shear strength. This footing may be long enough for
one-way beam action to govern. In such a case, a direct solution for the effective
depth d is reasonably practical. The factored column load is

Pu = 1.2(235) + 1.6(115) = 466 kips
The net upward pressure under factored load condition is
466, 000
pnet = = 6880 psf
67.7
Using section A-A in Fig. 19.10.1, and making the nominal shear strength Vn = Vc, so that
shear reinforcement is not required, means
Vn = Vc = 2 λ fc′bw d
Vu = φ Vc

6880(7.0)(4.08 − d ) = 0.75[2(1.0) 3000 ](7.0)(d )(144)
d = 1.5 ft (18.0 in.)
Total depth = 18.0 + 3 (cover) + 1 (estimated bar diameter) = 22.0 in. Try 22 in. for
total depth. Check this depth for two-way shear action as a slab, using critical section
B-B-B-B shown in Fig. 19.10.1 with d = 18 in.
1
Vu = 6880 [7.0(9.67) − 3.0(3.0)] = 404 kips
1000
For β = 9.67/7.0 = 1.38 < 2.0, and b0 /d = 4(18 + 18)/18 = 8 < 20 for a four-sided critical

section,
1
Vn = Vc = 4 λ fc′b0 d = 4(1.0) 3000 [ 4(18 + 18)](18) = 568 kips
1000

(Continued)
816
3” clear
A
Critical section
Critical
for two-way
9–#7 bars (spacing ≈ 9½”)

section
action B B for
one-way
18”+18” = 36” action
3.0’
7’– 0”
B B 4.08’ – d 2’– 9”
d/2 d
A
3” clear
1’–6” 4’–1”
L1
1’– 10”
3” clear 9 – #7 bars (spacing ≈ 12”) 3” clear
9’– 8”
Figure 19.10.1 Rectangular footing for Example 19.10.1.
(c) Check transfer of load at base of column. Assuming transfer without aid of dowels,
the design strength is
φ Pn = φ (0.85 fc′ )Ag = 0.65(0.85)(3)(18)(18) = 537 kips

φ Pn > [ Pu = 466 kips] OK
Only the minimum dowels required by ACI-16.3.4.1 are needed.

Use 4–#6 bars as dowels (As = 1.76 sq in.).

Minimum embedment length L1 (Fig. 19.10.1) required for bars in compression [Eq. (6.8.1)],
min L1 = Ldc (# 6) = 16.5 in.

available L1 = 22 − 3 − 1.5 − 0.75 = 16.75 in. > 16.5 in. OK
(d) Design for flexural strength. The reinforcement in the long direction is distributed
uniformly across the 7-ft width, while that in the short direction is concentrated
more heavily under the column in a band equal to the footing width and less heavily
near the ends. ACI-13.3.3.3 prescribes that the portion 2/(β + 1) of the total rein-
forcement in the short direction be placed in the central band (of a width equal to
the short side of the footing) wherein β is the ratio of the long side to the short side
of the footing. The ratio 2/(β + 1) may be derived on the basis that the intensity of
reinforcement in the central band is twice that of the outer portions.
In the long direction,
(4.08)2
Mu = 6.88(7.0) = 401 ft-kips
2
Mu 401(12, 000)
φ bd 2
0.90(84)(18.5)2
(Continued)
817
1 9 . 1 0 D E S I G N O F R E C TA N G U L A R F O OT I N G S 817
Since the longitudinal bars are placed below the transverse bars, d was taken as 18.5 in.
in the above calculation.
Using the trial moment arm method, rather than the formula, Eq. (3.8.5), for ρ,
assume the moment arm as 0.95d = 17.6 in. since the value of required Rn is very low.
Mu 401(12)
φ f y (arm) 0.9(60)(17.6)
Check:
C = 0.85 fc′ ba = 0.85(3)(84)a = 214 a
T = As f y = 5.06(60) = 304 kips
304
a= = 1.42 in.
214
arm = 18.5 − 0.71 = 17.8 in. ≈ 17.6 in. assumed
 17.6 
revisedd required As = 5.06  = 5.0 sq in.
 17.8 
min required As = 0.0018(84)(22) = 3.3 sq in. (ACI-8.6.1.1)
Use 9–#7, As = 5.4 sq in.
In the short direction,
(2.75)2
Mu = 6.88(9.67) = 252 ft-kips
2
assumed arm = 0.95d = 0.95(17.5) = 16.6 in.

Mu 252(12)
φ f y (arm) 0.9(60)(16.6)
Check:
C = 0.85(3)(116)a = 296 a
T = 3.4(60) = 204
a = 0.69 in.

arm = 17.5 − 0.35 = 17.2 in.
 16.6 
revised required As = 3.4  = 3.3 sq in.
 17.2 
For minimum reinforcement, ACI-8.6.1.1 requires ρg = 0.0018. For the short direction,

In this case, the minimum requirement controls.
Try 8–#7, As = 4.8 sq in.
reinforcement in band width 2 2
= = = 0.84
total reinforcement β + 1 9.67 / 7.0 + 1
Number of bars in the 7-ft band = 8(0.84) = 6.7, say 7. If one bar is placed on each side
outside the 7-ft band, a total of 9 bars would be required.
Use 9–#7 bars. (spaced at 12 in. for equal bar spacing within the 7-ft band)
(Continued)
81
(e) Development of reinforcement. The general equation, Eq. (6.6.1), often indicates a
smaller Ld than given by the simplified equations. Use Eq. (6.6.1):
 

Ld =  d [6.6.1]
 40 λ fc′  cb + K tr   b
  d  
 b 
The 9–#7 give an approximate 9 1 2 -in. center-to-center spacing across the 7-ft width, and
12 in. center-to-center spacing across the 9 ft 8 in. width. For the general equation, the
distance cb is the smaller of the following values:
bottom and side cover = 3.0 (i.e., clear) + 0.438 (i.e., bar radius) = 3.44.
9.5
one-half center-to-center spacing ≈ = 4.75 in. (7 ft width)
2
12
one-half center-to-center spacing ≈ = 6.0 in. (9 ft 8 in. width)
2
Thus, cb = 3.44 in. for both directions. There are no stirrups; thus, Ktr = 0. Thus,
 cb + K tr 3.44 + 0 
 = = 3.93 > 2.5 max
 db 0.875 
Thus, (cb + Ktr) /db = 2.5.

Evaluate Eq. (6.6.1), where ψt, ψc, and ψs are all equal to 1.0,
 3 60, 000 1.0(1.0)1.0 

Ld (for #7) =   0.875 = 28.8 in. (2.4 ft)
 40 (1.0) 3000 2.5
Allowing a 3-in. cover on the end of the #7 bars, the available embedment is

available embedment = 49 − 3 = 46 in. > [ Ld = 28.8 in.] (long direction) OK

available embedment = 33 − 3 = 30 in. > [ Ld = 28.8 in.] (short direction) OK
The complete design is shown in Fig. 19.10.1.
19.11 DESIGN OF PLAIN AND REINFORCED

CONCRETE WALL FOOTINGS
Wall footings carrying direct concentric loads may be of either plain or reinforced con-
crete. Those that are required to carry moment, such as for cantilever retaining walls, are
treated in Chapter 15. When the wall footing has bending in only one direction, it may
be designed or investigated by considering a typical 12-in. strip along the wall. Many
typical walls carry relatively light loads, and the supporting footings are proportioned by
using arbitrary minimums. Footings carrying light loads on good soil are often made of
plain concrete and may be designed in accordance with Chapter 14, Plain Concrete, of the
ACI Code.
819
1 9 . 1 1 D E S I G N O F WA L L F O OT I N G S 819
EXAMPLE 19.11.1
Determine the adequacy of the plain concrete wall footing of Fig. 19.11.1 to carry serv-
ice loads of 20 kips/linear ft dead load including the wall weight and 8 kips/linear ft live
load. Use fc′ = 3000 psi (normal-weight concrete), an allowable soil pressure of 6 ksf,
and the ACI Code.
SOLUTION
total service load = 28 + 6(2.5)(0.145) = 30.2 kips/ft
30.2
maximum soil pressure = = 5.0 ksf < 6 ksf OK
6.0
For a concrete wall, the factored bending moment is computed on the critical section at
the face of the wall (ACI-13.2.7.1).

wu = 20(1.2) + 8(1.6) = 36.8 kips/ft
net soil pressure under factored load = 36.8/6 = 6.13 ksf
(2.5)2
Mu = 6.13 = 19.2 ft-kips/ft
2
For computing the moment of inertia of the section, the bottom 2 in. of concrete placed
against the ground is assumed to be of poor quality and must be neglected (ACI-
14.5.1.7). Neglecting the bottom 2 in.,
12(28)3
Ig = = 21, 950 in.4
12
For bending of plain concrete, the design strength is based on a linear stress-strain
relationship for both tension and compression (ACI- 14.5.1.4). In accordance with
ACI-14.5.2.1, the nominal flexural strength is the smaller of the value calculated at the
tension face as

M n = 5λ fc′Sm
and that computed at the compression face as

M n = 0.85 fc′Sm
where Sm is the elastic section modulus.
Therefore,
21, 950 1
M n = 5λ fc′Sm = 5(1.0) 3000 = 35.8 ft-kips Controls!
(28 / 2) 12,000
or
21, 950 1
M n = 0.85 fc′ Sm = 0.85(3000) = 333 ft-kips
(28 / 2) 12,000
For plain concrete, a strength reduction factor φ of 0.6 is prescribed (ACI-21.2). Thus,

φ M n = 0.60(35.8) = 21.5 ft-kips > [ Mu = 19.2 ft-kips/ft] OK
It may be readily verified that shear strength is adequate since the critical section at a
distance d from the face of a wall falls near the edge of the footing.
(Continued)
820
w = 28 kips/ft
12”
2’– 6”
6’– 0”
Figure 19.11.1 Wall footing for Example 19.11.1.
EXAMPLE 19.11.2
Design a reinforced concrete footing for a 12-in. masonry wall carrying 10 kips/lin-
ear ft service dead load including the wall weight and 5 kips/ft service live load. Use
fc′ = 3000 psi (normal-weight concrete), fy = 60,000 psi, and an allowable soil pressure
of 4000 psf.
SOLUTION
(a) Determine the footing thickness. Assume the footing depth to be 10 in. at 125 psf.
Allowable net soil pressure = 4000 –125 = 3875 psf. Footing width = 15/3.875 =
3.87 ft. Use 4 ft. It is probable that the thickness will be governed by one-way shear
action (inclined cracking), which is taken to be critical at a distance d from the face
of the wall (Fig. 19.11.2).
Applying the load factors,

wu = 1.2(10) + 1.6(5) = 20.0 kips/ft
Net soil pressure under factored load = 20.0/4 = 5.0 ksf. When no shear reinforcement is
used, the nominal strength for one-way action, using the simplified procedure, is

Vn = Vc = 2 λ fc′bw d
which requires that
Vu = φVc
( )
5000 (1.5 − d ) = 0.75 2(1.0) 3000 (12)(d )(12)

1.5 − d = 2.37d
d = 0.45 ft (5.4 in.)
(Continued)
821
1 9 . 1 1 D E S I G N O F WA L L F O OT I N G S 821
Critical section for

moment and
development of 12”
reinforcement
1’– 9”
Masonry wall
3” #4 @ 7”
#4
9”
4’– 0”
Figure 19.11.2 Wall footing for Example 19.11.2.
Assuming #4 bars,
total thickness = 5.4 + 3 (cover) + 0.5(bar diameter) = 8.9 in.
Use 9-in. thickness.

check weight = 113 psf [ < 125 psf assumed] OK
(b) Reinforcement for flexural strength. The critical section for bending moment
strength on footings under masonry walls is to be taken halfway between the middle
and the edge of the wall (ACI-13.2.7.1).
1
Mu = (5.0)(1.75)2 = 7.66 ft-kips/ft
2
7.66(12,000)
0.90(12)(5.5)2
The steel area may be obtained by formula, Eq. (3.8.5),

ρ = 0.005(> ρmin )

required As = ρ bd = 0.005(12)(5.5) = 0.33 sq in.
Try #4 @ 7 in. (As = 0.34 sq in./ft). Check strength.
C = 0.85 fc′ ba = 0.85(3)(12)a = 30.6

T = As f y = 0.34(60) = 20.4 kips
C = T; a = 0.66 in.
φ M n = 0.90(20.4)[5.5 − 0.5(0.66)] 121 = 7.9 ft-kips
φ M n > [ Mu = 7.66 ft-kips] OK
(c) Development of reinforcement. For the wide bar spacing 7 in., the general
equation gives

Ld (# 4) = 16.4 in. (1.4 ft)
(Continued)
82
(Note: ψs was conservatively taken as 1.0 in the computation of Ld, although ACI-
25.4.2.4 allows a value of 0.8).
The available embedment distance is 21 in. less 3-in. cover, i.e., 18 in., which is larger
than the 16.4 in. required, and is acceptable.
A minimum temperature and shrinkage reinforcement must be provided per ACI-
7.6.4.1 and ACI-24.4,

As, min = 0.0018(12)(9) = 0.19 sq in. / ft
This temperature and shrinkage reinforcement must have a spacing not greater than 5
times the slab thickness (45 in.) and 18 in. Thus, provide 4–#4 bars at 12 in. spacing (As
= 0.20 sq in./ft) perpendicular to the slab flexural reinforcement. This would leave a 5.5
in. cover on the outer bars, which is greater than the minimum required of 3 in.
See Fig. 19.11.2 for the details of the complete design.
19.12 COMBINED FOOTINGS
A combined footing is one that usually supports two columns. These may be two inte-
rior columns [Fig. 19.12.1(a)], which are so close to each other that isolated footing areas
would overlap. If a property line exists at or near the edge of an exterior column, a rec-
tangular [Fig. 19.12.1(b)] or trapezoidal [Fig. 19.12.1(c)] combined footing may be used
to support the exterior column and its adjacent interior column. The area of the combined
footing may be proportioned for uniform settlement by making its centroid coincide with
the resultant of the respective portions of the two column loads that are sustained for long
duration. It may be noted that for footings of constant thickness, the centroid of the bearing
area always coincides with the resultant of the weight of the footing itself.
Referring to the frequently occurring situation in Fig. 19.12.1(b), the load P1 is close to
the property line; however, there is adequate space to the right of P2. Whenever P2 exceeds
P1, a rectangular combined footing can be used, since it may be made long enough to make
the load resultant and the footing centroid coincide. It may be shown that if 1 2 < P2 /P1 < 1
approximately, a trapezoidal footing could be used. If P2 /P1 < 1 2 approximately, then either
a strap (Fig. 19.12.2) or a T-shaped spread footing would have to be used.
In the strength computation for a combined footing, maximum loads in columns (full
dead load plus reduced live load as discussed in Section 19.7) should be used. Since the
resultant of the maximum column loads does not necessarily coincide with that of the sus-
tained column loads producing uniform settlement, under the former loading condition the
distribution of the net soil pressure along the footing is not uniform. The deviation from
Property Property
line line
P1 P2 P1 P2
(a) (b) (c)
Figure 19.12.1 Combined footings.

823
19.13 DESIGN OF COMBINED FOOTINGS 823
Property line
Pext Pint
Rext Rint
Figure 19.12.2 Cantilever, or strap, combined footing.
uniformity is usually so small that certain approximate shortcut procedures may be used in
determining the shear and moment diagrams in the longitudinal direction.
The ACI Code (see Chapter 13) does not provide full recommendations for combined foot-
ings. Kramrisch [19.10] has provided a detailed treatment of footing design, with particular
emphasis on combined footings. However, ACI Committee 336 [19.11, 19.12, 19.20] has given
design procedures for combined footings and mats, and additional suggestions have been given
by Kramrisch and Rogers [19.13], Szava-Kovats [19.14], and Davies and Mayfield [19.15].
When a series of columns is supported by a single footing extending over a large area, the foot-
ing is referred to as a mat. Mats are treated by Bowles [19.16] as well as by ACI Committee
336. Basically, transverse steel under each column tends to distribute the column load in the
transverse direction. This being considered accomplished, the combined footing itself becomes
a beam in the longitudinal direction. Thus it is suggested that the provisions for isolated footings
be applied to the transverse direction, and those for beams to the longitudinal direction.
The cantilever or strap footing shown in Fig. 19.12.2 is an alternative design to prevent
overturning of an exterior footing placed eccentrically under an exterior column, the edge
of which is at or close to a property line. Overturning of the exterior footing is prevented
by connecting it with the adjacent interior footing by a strap beam. This strap beam is sub-
jected to a constant shearing force and a linearly varying negative bending moment. Thus it
behaves like a cantilever beam; hence the name “cantilever footing.”
In the strength computation for a cantilever footing (Fig. 19.12.2), the weight of the
strap, the exterior footing, and the interior footing is each assumed to be balanced by the
soil pressure caused by it; thus such weight causes no shears and moments in any part of
the structure. In the longitudinal direction, the column loads Pext and Pint are in equilibrium
with the total “net” upward soil pressures Rext and Rint under the exterior and interior foot-
ings; the upward soil pressure is assumed to be uniform over the entire area of each footing.
The exterior footing may be considered as under one-way transverse bending, although
some reinforcement in the longitudinal direction is desirable, and the interior footing is
under two-way bending as in isolated footings.
Use of a cantilever footing may be justifiable under conditions where the distance
between columns is large and a large area of excavation must be avoided. It is usual prac-
tice that the bottom surfaces of the exterior footing, the strap, and the interior footing be
at the same elevation, but the total thickness of each element may be different, depending
on the strength requirements. Certainly it is desirable, unless there are good reasons to the
contrary, to make all three elements of constant thickness.
19.13 DESIGN OF COMBINED FOOTINGS

The design of two common types of combined footings will be shown. The first is a rectan-
gular footing, and the second is a strap or cantilever footing.
824
EXAMPLE 19.13.1
Design a combined footing to support two columns as shown in Fig. 19.13.1(a): PA =

350 kips (40% live load); PB = 400 kips (40% live load); column A is centered 1 ft 3 in.
from the property line, and column B is centered 19 ft 3 in. from the property line. Use
fc′ = 3000 psi (normal-weight concrete), fy = fyt = 60,000 psi, and a maximum soil pres-
sure = 5000 psf. Assume that the ratio of maximum column loads as given is equal to
that of long duration loads in the exterior and interior columns.
SOLUTION
(a) Length and width of footing. ACI-13.3.1.1 indicates base area of footings is to be
determined using service loads and allowable soil pressure.
350(1.25) + 400(19.25)
x from property line = = 10.85 ft
750
length of footing, L = 10.85(2) = 21.70 ft
Use a 21 ft 9 in.-long footing
Since the design for strength of the footing involves factored loads, there will be an
eccentricity no matter how “exact” the length determination. In the design for shear and
bending moment, the soil pressure under factored loads might be taken as linearly vary-
ing to account for the eccentric loading; in the case of a small eccentricity, however, it is
probably sufficient to assume a uniform soil pressure as is done in this example.
Assume the footing thickness to be about 2 ft 6 in., or 375 psf.
net soil pressure = 5500 − 375 = 4625 psf
750, 000
footing width = = 7.46 ft
4625(21.75)
Try a 7 ft 6 in.-wide footing
(b) Longitudinal factored shears and factored moments. For gravity loading,
column A, Pu = 1.2(210) + 1.6(140) = 476 kips
column B, Pu = 1.2(240) + 1.6(160) = 544 kips

1, 020, 000
net soil pressure under factored load = = 6250 psf
21.75(7.5)
The factored shear Vu diagram is computed as for a beam and given in Fig. 19.13.1(c).
For simplicity, the column loads are taken to be acting along the column centerlines,
thus producing the dashed portions within the column widths on the shear diagram. In
the computations, the 20-in.-diameter column is treated as an equivalent square of side
17.7 in. in accordance with ACI-13.2.7.3.
net upward uniform pressure = 7.5(6.25) = 46.9 kips/ft
Vu at centerline of column A = +46.9(1.25) − 476
= +58.6 − 476 = −417.4 kips
Vu at centerline of column B = −46.9(2.5) + 544
= −117.3 + 544 = +426.8 kips
 417.4 
point of zero shear = 18 
 417.4 + 426.8 
= 8.90 ft from centerline of column A
(Continued)
825
The factored moment Mu diagram as computed for a beam is given in Fig. 19.13.1(d). Note
that the numerical values on the two small end portions of the factored moment diagram are
based on assuming all the column loads to be concentrated at the column centerlines.
46.9(10.15)2
max Mu (computed from left side) = − 476(8.90)
2
= −1821 ft-kips
46.9(11.60)2
max Mu (computed from right side) = − 544(9.10)
2
= −1795 ft-kips
1’– 3”
18’– 0” 2’– 6”
16”
Column A, PA = 350k Column B, PB = 400k
7’– 6”
22” 20”
Property line
(a) Footing plan

3” clear Equivalent
20 – #9 spaced 4 21 ” ± square = 17.7”
cover
2’– 5”
1’– 2”6 @ 12” = 6’– 0” 6 @ 12” = 6’– 0” 1’– 3” #4 stirrups
6 – #6 3” clear cover 8 – #6
Spaced @ 9” ± (side and bottom) Spaced @ 8 21 ”±
(b) Footing elevation 426.8k

295k CL of column
184k
8.90’ d=
58.6k
CL of column φVc φVc 24.9” 1.76’
2
d = 2.08’ φVc
φVc 2 (c) Factored shear Vu diagram
117.3k
Vc
1820 ft-kips
Face of
417.4k column
Face of
column Mu
870 ft-kips
+
37 ft-kips
147 ft-kips
(d) Factored moment Mu diagram
Figure 19.13.1 Rectangular combined footing for Example 19.13.1.

(Continued)
826
The moments as computed from both sides are not exactly the same because a foot-
ing length of 21 ft 9 in. is used instead of the computed 21.70 ft and because the dis-
tance 8.90 ft to the point of zero shear contains only three significant figures. Use Mu =
1820 ft-kips.
(c) Thickness of footing. For moment, the thickness may be based on a desired rein-
forcement ratio ρ. Select ρ = 0.012, that is, approximately one-half of ρb. For this
value of ρ, using Eq. (3.8.4b),
1
Rn = ρ f y (1 − ρ m)
2
Rn = 0.012(60, 000)[1 − 0.5(0.012)23.5] = 618 psi
Mu 1820 (1000)
reqquired d = = = 20.9 in. ( based on flexural strength
φ Rn b 0.90(618)(7.5) requirements)
The footing is considered a beam in shear computations. One-way action is assumed to

control at the distance d from the face of the columns. The shear at a distance d from the
face of the 17.7-in. equivalent square column (column B) is
1
Vn = 426.8 − (8.85 + d ) 46.9 = 392.2 − 3.91d
12
The nominal shear strength when no shear reinforcement is to be used is

Vn = Vc = 2 λ fc′ bw d
unless the more detailed procedure is used. Then, using λ = 1.0 for normal-weight
concrete,
Vu = φVc
(
392.2 − 3.91d = 0.75 2(1.0) 3000 (7.5) d ) 12
1000
392.2 − 3.91d = 7.39d
d = 34.7 in. (based on shear strength requirements)

It seems desirable to make the footing deep enough at the desirable reinforcement ratio
for the bending moment, but not deep enough to give the extra 14 in. that would be
required to eliminate stirrups.
Try a 29-in.-thick footing. Check the weight, 29(150)/12 = 363 psf.
750, 000
max soil pressure = + 363 = 4961 psf < 5000 psf OK
21.75(7.5)
(d) Main longitudinal reinforcement. At the middle of the span, assuming #9 bars

d ≈ 29 − 3 (cover) − 0.5 (stirrup) − 0.6 (bar radius) = 24.9 in.
Thus,
Mu 1820(12, 000)
φ bd 2 0.90(7.5)(12)(24.9)2
 435 
required As = 0.012  (7.5)(12)(24.9) = 18.9 sq in.
 618 
(Continued)
827
Try 20–#9 (approximate 4 1 2 -in. spacing); As = 20.0 sq in. Check strength.
C = 0.85 fc′ ba = 0.85(3)(90) a = 229.5a

T = 20.0(60) = 1200 kips
a = 5.23 in.
1
φMn = 0.90(1200)[24.9 − 0.5(5.23)]
12
= 2006 ft-kips > 1820 ft-kips OK
The anchorage required from the point of maximum moment to the end of the bars is
the development length Ld for #9 bars. The distance from the bar center to the nearest
concrete surface is [3 (clear cover) + 0.5 (stirrup) + 0.6 (bar radius)] = 4.1 in., which is
greater than half the center-to-center bar spacing of 2.2 in. Therefore, use cb = 2.2 in.
Since the bars have more than 12 in. of concrete cast beneath them, a casting position
multiplier ψt of 1.3 must be applied. Thus, using Eq. (6.6.1) with cb = 2.2 in. and assum-
ing no contribution from stirrups (Ktr = 0),

Ld (for #9) = 47.5(1.3) = 61.8 in. (5.1 ft)
If all 20 of the #9 bars are extended into the centerlines of the columns, the development
length requirement is more than satisfied because an anchorage distance of 108 in. is
provided on both sides of the point of maximum moment.
Use 20–#9 bars at 18 ft long (spaced approximately at 4 1 2 in.)
Check development length requirement at the points of inflection according to ACI-
9.7.3.8.3. The ACI Code provision is checked here even though the reinforcement is
negative moment reinforcement rather than the positive moment reinforcement as pre-
scribed in ACI-9.7.3.8.3. The footing may be visualized for this purpose as an inverted
beam with the soil pressure as loading and the columns as supports. The situation is
similar to the positive moment requirement in the sense that the required embedment
is into the support (column) rather than out into the span as for the negative moment
requirement in an ordinary continuous beam. Furthermore, since the inflection points
are inside the faces of the columns, the 1.3 factor might be used (see Section 6.14); how-
ever, since the column width (22 in.) is only a small fraction of the footing width (7.5 ft),
the authors do not recommend using the 1.3 factor.
Since all 20–#9 bars extend through the inflection points,
M n = (2006 / 0.9) = 2229 ft-kips

Vu = 417.4 kips(at column A)
Vu = 426.8 kips(at column B)
La = 15 in. approx (near both column A and B)
In this case, the actual distance La from the inflection point to the end of the bars
is larger than the maximum La limits of effective depth d (24.9 in.) or 12 bar diam-
eters (13.5 in.); thus 13.5 in. controls. Since Mu and La are the same at both inflection
points, the one near the column B having the larger shear Vu controls. At the column B
inflection point,
Mn 2229(12)
+ La = + 13.5 = 76 in. > [ Ld = 61.8 in.] OK
Vu 426.8
(Continued)
82
(e) Alternative design of the main longitudinal reinforcement using cutoff. If it seems
desirable to cut off some of the tension bars, say 10–#9, and extend the remain-
ing ones into the support, the general anchorage requirements of ACI-9.7.3 must
be applied. The theoretical point at which the 10–#9 bars are no longer required,
indicated in Fig. 19.13.2, corresponds to the flexural design strength, φ Mn, of
1060 ft-kips. The provision in ACI-9.7.3.3 requires an extension of at least the effec-
tive depth d or 12 bar diameters, whichever is greater, beyond the theoretical cutoff
point. In this case, the effective depth of 24.9 in. (or 2.1 ft) controls over 12 bar
diameters = 13.5 in. (or 1.1 ft).
CL of column CL of column
20 – #9
φMn = 2006 ft-kips
φMn diagram
Ld = 4.0’ Ld = 4.0’
Ldh = 24.7” Ldh = 24.7”

2.1’ 2.1’
d d
10–#9
Development
φMn = 1060 ft-kips
over length
X1 Ldh of hooked bars
X2
Theoretical cutoff points
1’ Mu 2’–3”
18’– 0”
Figure 19.13.2 Bar cutoff alternative for Example 19.13.1.
In addition, for such a cutoff in the tension bars to be permitted, ACI-9.7.3.5 must
be satisfied. It is assumed here that stirrups will be provided sufficient to satisfy ACI-
9.7.3.5(b). The continuing bars must be embedded Ld beyond the cutoff of the 10–#9
bars according to ACI-9.7.3.4.
The development length Ld for the bars being cut as well as for the continuing bars,
was computed in part (d) as 5.1 ft without stirrups. Assuming closely spaced stirrups
such that [cb + Ktr]/db is 2.5 max, the required development length of the #9 bars would
be reduced to at most 4.0 ft. At neither end the available embedment distance (≈ 2.1 ft
from end of bars at the column A end of the footing to point X1 and ≈ 3.5 ft from end of
bars at the column B end of footing to point X2) is adequate to develop (Ld = 4.0 ft) the
continuing bars through straight embedment. Thus, hooks would be needed at both ends.
When hooks are provided as anchorage beyond an inflection point (as would be the
case here at both ends of footing), ACI-9.7.3.8.3 need not be satisfied. Referring to
Table 6.10.2 and Fig. 6.10.2 for #9 hooked bars,

Ldh = 24.7 in.
In Fig. 19.13.2, the development of the #9 hooked bars is shown by the sloping dashed
line at each end of the footing, starting from the end of the bar at 3.0 in. from the side
face of footing.
(Continued)
829
Since the cut bars at X1 and X2 are still in the tension zone, ACI-9.7.3.5 must be satisfied.
By inspection of Fig. 19.13.2, φ Mn exceeds twice Mu at the cutoff points, thus, check Vu /
(φ Vn) ≤ 0.75 for the cuts to be acceptable:
 Av f yt d 
φVn = φ  2 λ fc′bw d +
 s 
Using #4 multiple-loop stirrups with N = 8 spaced at 12 in. [(Fig. 19.13.4) in part (h)],
 8(0.2)(60)24.9 
φVn = 0.75  2(1.0) 3000 (90)24.9 1 + 
 1000 12
φVn = 184 + 149 = 333 kips
Since Vu ≈ 368 kips at the proposed cutoff locations (see Figs. 19.13.1 and 19.13.2),
the cuts cannot be made at locations X1 and X2. The 10–#9 bars must be lengthened
to reach the inflection points, in which case they are no longer in the tension zone and
ACI-9.7.3.5 does not apply. [Note that the proposed cutoff locations are closer to the
face of the supports than the first critical section for shear. Thus, for evaluation of shear
strength, Vu = 295 kips < φ Vn as shown later in part (h)].
It is clear in this example that cutting bars is impractical when about 4 ft of straight
length is saved on 10 bars but the remaining 10 must be hooked.
(f) Longitudinal reinforcement at bottom of footing beyond column centers. The bend-
ing moment at the face of column B is
1
Mu = (46.9)(1.76)2 = 72.6 ft-kips
2
Though certainly not always the case, the moment here appears small enough to require
no reinforcement. Assuming a plain concrete element for this check, the design strength
is based on a linear stress-strain relationship for both tension and compression (ACI-
14.5.1.4). In accordance with ACI-14.5.2.1, the nominal flexural strength is the smaller
of the value calculated at the tension face as

M n = 5λ fc′ Sm
and that computed at the compression face as

M n = 0.85 fc′ Sm
where Sm is the elastic section modulus.

Neglecting the bottom 2 in. of thickness,
(7.5)(12)(27.0)2
Sm = = 10, 935 cu in.
6
Therefore,
1
M n = 5λ fc′Sm = 5(1.0) 3000 (10, 935) = 250 ft-kips Controls!
12,000
or
1
M n = 0.85 fc′ Sm = 0.85(3000)(10, 935) = 2324 ft-kips
12,000
(Continued)
830
For plain concrete, a strength reduction factor φ of 0.6 is prescribed (ACI-21.2). Thus,

φ M n = 0.60(250) = 150 ft-kips > [Mu = 72.6 ft-kips] OK
No flexural reinforcement in the longitudinal direction is required for strength at the
bottom of either cantilever.
(g) Transverse reinforcement. Bending in the transverse direction may be treated in a
manner similar to isolated spread footings. The 1940 ACI, Joint Committee [19.17]
recommended that the transverse reinforcement at each column be placed uniformly
within a band having a width not greater than the width of the column plus twice the
effective depth of the footing.
The procedure seems to be reasonable. Certainly the behavior of the footing depends
on the overall length-to-width ratio as well as the spacing of the columns. In this design
example, the large spacing of the columns and the relatively narrow footing mean that
most of the footing between the columns will be subjected to longitudinal curvature only,
while locally, in the vicinity of the concentrated loads, curvature in both directions will
result. Thus, the transverse reinforcement is placed into bands as shown in Fig. 19.13.3.
d = 29 − 3 (cover ) − 0.5 (stirrup) − 0.375 (#6 bar radius assumed) = 25.1 in.
1
column A band width, WA = 1.25 + (8 + 25.1) = 4.0 ft
12
476
net factored load pressure in transverse direction = = 63.5 kips/ft
7.5
1
Mu = (63.5)(2.83)2 = 254 ft-kips
2
Mu 254(12,000)
φ bd 2 0.90(4.0)(12)(25.1)2
From Eq. (3.8.5), ρ = 0.002, thus

required As = 0.002(48.0)(25.1) = 2.4 sq in.
Check minimum area of flexural reinforcement (ACI-8.6.1.1)
As,min = 0.0018(48)(29) = 2.5 sq in. Controls!
Try 6–#6 bars (As = 2.64 sq in.). Check strength.

C = 0.85 fc′ ba = 0.85(3)(48)a = 122.4 a
T = 2.64(60) = 158.4 kips
a = 1.29 in.

φ Mn = 0.90(158.4)[25.1 − 0.5(1.29)] 1 = 291 ft-kips > 254 ft-kips OK
12
Since 2 ft 10 in. is available from face of column, embedment can be provided that
exceeds the required development length Ld for #6 bars (Ld = 32.9 in., Category A,
Table 6.6.1, by the conservative simplified equation).
Use 6–#6 bars (approximate 9-in. spacing).
1
column B band width, WB = 2.50 + (8.85 + 25.1) = 5.3 ft
12
544
net factored load pressure in transverse direction= = 72.5 kips/ft
7.5
(Continued)
831
15”
d = 12.45” d = 12.45”
2 Critical sections 2
Column A Column B
for shear in
two-way action
2’–10” 3.01’
d = 24.9”
Property 2.8’ 2.5’

2.7’ Critical sections
line
WA = 4.0’ for shear in WB = 5.3’
one-way action
d +17.7”
2
Figure 19.13.3 Band width for transverse reinforcement.
1
Mu = (72.5)(3.01)2 = 328 ft-kips
2
Mu 328(12,000)
φ bd 2
0.90(5.3)(12)(25.1) 2
From Eq. (3.8.5), ρ ≈ 0.002, the same as for column A.

required As = 0.002(5.3)(12)(25.1) = 3.19 sq in.
Check minimum area of flexural reinforcement (ACI-8.6.1.1)
As,min = 0.0018(63.6)(29) = 3.32 sq in. Controls!
Try 8–#6 (As = 3.52 sq in.). Check strength.

C = 0.85 fc′ ba = 0.85(3)(63.6)a = 162.2 a
T = 3.52(60) = 211.2 kips
a = 1.30 in.

φ Mn = 0.90(211.2)[25.1 − 0.5(1.30)] 1 = 387 ft-kips > 367 ft-kips OK
12
Anchorage of #6 bars is adequate since available embedment of 3.01 ft exceeds Ld of

32.9 in.
Use 8–#6 bars (approximately 8 1 2 -in. spacing).
(h) Shear reinforcement. The usual approach is to consider the footing as a beam and
to provide shear reinforcement on the assumption that the shear (inclined cracking)
effect is uniform across the width. This approach seems appropriate in this case,
given the large distance between columns and the relatively narrow footing width.
The maximum shear to be provided for is at the critical section a distance d from the
face of the column. From Fig. 19.13.1, Vu = 295 kips. The design shear strength φVc of a
beam without shear reinforcement, using the simplified procedure, is
(
φVc = φ 2λ fc′bw d )

(
φVc = 0.75 2(1.0) 3000 (90)24.9 ) 1
1000
= 184 kips

reequired φVs = Vu − φVc = 295 − 184 = 111 kips
(Continued)
832
When shear reinforcement is required, there must be at least the minimum specified by
ACI-9.6.3.3, as the larger of
1
min φVs = φ (0.75 fc′ )bw d = 0.75(0.75 3000 )(90)(24.9) = 69 kips
1000
and
1
min φVs = φ (50)bw d = 0.75(50)(90)(24.9) = 84 kips Controls!
1000
Thus, the required φVs = 111 kips controls the closest spacing for shear reinforcement.
Even though this is a footing, the rules for beams are believed appropriate here. Thus,
the required amount of shear reinforcement is
Av φVs 111
required = = = 0.1 sq in./in.
s φ f yt d 0.75(60)24.9
For design of shear reinforcement, use a multiple-

loop stirrup as shown in
Fig. 19.13.4. Thus,
Av As N
=
s s
where As is the cross section of the stirrup bar, and N is the number of times the multiple-
loop stirrup crosses the neutral axis of the beam. Try a #4 stirrup, with N = 8 (see
Fig. 19.13.4), spaced at 12 in. This gives
Av 0.20(8)
provided = = 0.133 sq in./in.
s 12
which is greater than the required amount. OK
Figure 19.13.4 Multiple-loop stirrup (N = 8).
Use #4 stirrups @ 12 in. with N = 8, as in Fig. 19.13.4. The 12-in. spacing is adequate
since it is the maximum permitted (ACI-9.7.6.2.2) when Vs does not exceed 4 fc′ bw d ,
which corresponds to φVs = 368 kips.
The first stirrup is placed at s/2 = 6 in. from the face of the column, and the last one
should be within s/2 = 6 in. of the location where Vu = φVc /2, at which point shear
reinforcement is no longer theoretically required for beams (ACI-9.6.3.1). The stirrup
arrangement will be made identical at each column. To space and hold in position the
stirrups and the transverse reinforcement, a few longitudinal bars are placed in the bot-
tom of the footing.
(i) Check shear strength based on two-way action. At the exterior column A, the calcu-
lation based on a perimeter at d/2 (12.45 in.) from the column face on all four sides
would be unrealistic on the side nearest the property line. Instead, a three-sided
perimeter is used at column A in Fig. 19.13.3,

b0 = 2(15 + 8 + 12.45) + [22 + 2(12.45)] = 117.8 in.
(Continued)
83
According to Eq. (19.5.1c), αs = 30 for the three-sided perimeter (“edge column”), and
when b0 /d does not exceed 15 that equation does not control. In this case,

b0 /d = 117.8 / 24.9 = 4.7 < 15
Thus, with b0 /d less than 15 and a loaded area with β < 2, shear strength is based on a
nominal stress of 4 λ fc′ according to Eq. (19.5.1a). Thus,
1
φVn = φVc = φ 4 λ fc′ b0 d = 0.75[ 4(1.0) 3000 (117.8)(24.9)] = 482 kiips
1000
Actually, the resisting section for shear is not symmetrical about the column; therefore,
the shear distribution is nonuniform and eccentricity of shear should be considered in
accordance with ACI-8.4.2.3. In this example, where no moment is assumed at the base
of column and the moment in the slab at the face of column is small, it may be reason
able to neglect the eccentricity of shear. Thus, neglecting any eccentricity of shear,
1
Vu = 476 − 6.25[35.45(46.90)] = 404 kips
144
Neglecting any possible contribution of the stirrups used for φVs in one-way shear action,

φ Vn = φ Vc = 482 kips > [Vu = 404 kips] OK
A similar calculation for two-way shear action at column B will show the thickness to be
adequate without shear reinforcement. Again, it is prudent to consider the failure section
to extend to the end of the footing (see Fig. 19.13.3), rather than merely the circular path
at d /2 from the column.
(j) Design sketch. The complete design details are given in Fig. 19.13.1.
EXAMPLE 19.13.2
Design a cantilever or strap footing for the situation shown in Fig. 19.13.5, where the
property line is at the exterior edge of the exterior column. Column data are given in
Table 19.13.1. The distance between column centers is 18 ft. Equal settlement is assumed
for DL plus 1 2 LL condition at a uniform pressure of 3.35 ksf. Use fc′ = 3000 psf for
footing fc′ = 3750 psi for columns, fy = 40,000 psi.
TABLE 19.13.1 COLUMN DATA FOR EXAMPLE 19.13.2
Column Size Reinforcement LL DL
Exterior 12 × 12 in. 4–#7 70 kips 55 kips

Interior 14 × 14 in. 8–#8 130 kips 80 kips
SOLUTION
(a) Size of exterior and interior footings. According to ACI-13.3.1.1, footing size is
always determined using service loads, rather than factored loads. The size of
(Continued)
834
exterior and interior footings is a function of the assumed width of the exterior foot-
ing (which affects the cantilever action and therefore the reactions required of each
footing) and the assumed total thickness of each footing (which affects the available
net soil bearing capacity). Should these assumed values be revised in the subsequent
computation, the sizes of the footings must be revised accordingly. Assume that
the total thickness of the entire footing is 25 in. The net uniform soil pressure for
both exterior and interior footings is 3.35 –2.08(0.150) = 3.04 ksf. Referring to
Fig. 19.13.5 for the dimensions, taking moments about the interior column using the
service loads in the equal settlement condition gives
(55 + 35)18
Rext = = 99.7 kips
16.25
99.7
area of exterior footing required = = 32.8 sq ft
3.04
Using the assumed exterior footing width of 4 ft 6 in.,

32.8
length of exterior footing = = 7.29 ft (try 7 ft 4 in.)
4.50
Taking moments about the line of action of Rext gives
(80 + 65)16.25 − (55 + 35)(2.25 − 0.50)
Rint = = 135.3 kips
16.25
135.3
area of interior footing required = = 44.5 sq ft
3.04
side of square interior footing = 44.5 = 6.67 ft (try 6 ft 8 in.)
(b) Factored shear and factored moment diagram for strap. Referring to Fig. 19.13.5,
applying load factors U to the maximum load condition gives
[1.2(55) + 1.6(70)]18 (178)18

Rext = = = 197 kips
16.25 16.25
178(1.75)
Rint = [1.2(80) + 1.6(130)] − = 304 − 19 = 285 kips
16.25
Vu in strap = − 178 + 197 = +19 kips
Mu at right end of strap = 19((3.33) = 63 ft-kips
Mu at left end of strap = 19(14) = 266 ft-kips
(c) Design of strap. For the shear requirement, assuming no shear reinforcement is to be
used and the simplified expression for strength is used,

(
φ Vc = φ 2 λ fc′bw d )
Estimate d ≈ 25 –3 (cover) –0.5 (est. bar radius) = 21.5 in. Thus,
19, 000
width bw of strap required = = 10.88 in.
0.75 2(1.0) 3000  21.5
Assume desirable reinforcement, ρ ≈ 0.012, for which Rn = 435 psi.
(Continued)
835
18’ – 6” Section for

12 × 12 col. 2’ – 9”
two-way
14 × 14 col.
shear action
27” 14”
7’ – 4”
6’ – 8”
1.38’
35.5”
1
2’ – 62”
Section for 6’ – 8”
4’ – 6” 2.54’
one-way
k
shear action Pext = 178 (max) Pint = 304k (max)
7 – #6 7 – #8 7 – #7
Loads each way each way
3” cover
given
include
load 2’ – 1”
factors U 43.8 kips/ft (max)
16’ – 3” 42.8 kips/ft (max)
k
Rext = 197 (max) Rint = 285k (max)
(a)
10’ – 8” 137k
≈67k
19k
Vu
Shear diagram, Vu
134k
117k
179 ft-kips
Factored
+
67’k moment diagram, Mu
Mu
63
–
Development 143 Ldh = 14.6”
over length 329
Ldh of hook 226’k φMn diagram
Ldh = 14.6” for negative moment
(b)
Figure 19.13.5 Cantilever, or strap, footing of Example 19.13.2.
At the junction with the interior footing,

Mu 63 (12, 000)
width of strap required = = = 4.2 in.
φ Rn d 2
0.90(435)(21.5)2
At the junction with the exterior footing,
266 (12,000)
width of strap by moment requirement = = 17.6 in.
0.90(435)(21.5)2
Try a strap width varying from 14 to 20 in.
Mu 266 (12, 000)
required Rn at wide end = = = 384 psi
φ bd 2
0.90(20)(21.5)2

63 (12, 000)
required Rn at narrow end = = 130 psi
0.90(14)(21.5)2
(Continued)
836
Approximately,
 384 
required As at wide end ≈ 0.012  (20)(21.5) = 4.6 sq in.
 435 

 130 
required As at narrow end ≈ 0.012  (14 )( 21 . 5) = 1. 1 sq in.
 435 
The minimum requirement of ACI-9.6.1.2 or 9.6.1.3 applies here because the strap is a
beam. Thus,
3 fc′
As , min = bw d [3.7.10]
fy
but not less than 200bw d/fy. Thus, at the narrow end ACI-9.6.1.2 requires
3 fc′ 3 3000
As , min = bw d = (14)21.5 = 1.24 sq in.
fy 40, 000

200 bw d 200(14)21.5
As , min = = = 1.51 sq in. Controls!
fy 40, 000
Alternatively, according to ACI-9.6.1.3,
4 4
(required As ) = (1.1) = 1.50 sq in.
As ,min =
3 3
It may be shown that at the wide end As, min does not govern. Thus, use 7–#8 (As = 5.53 sq in.)
at the wide end. Extend 3–#8 through the narrow end. Two bars would satisfy As,min =
1.50 sq in. but to maintain symmetry, it is necessary to extend three bars.
The development length Ld for these #8 bars depends on their clear spacing and cover.
For the general equation, the distance cb is the smaller of the following two values:
top and side cover = 3.0 ( i.e., clear ) + 0.50 ( i.e., bar radius) = 3.5 in.
20 − 3(2) − 1
one-half center-to-center spacing = = 1.08 in.
2(6)
Thus, cb = 1.08 in. and Ktr = 0. Clearly, development will be a problem with such a small
cb. Increase the strap width from 20 to 27 in. This may permit using smaller bars in the
top of the strap. Recomputing Rn gives
Mu 266 (12, 000)
required Rn at wide end = = = 284 psi
φ bd 2 0.90(27)(21.5)2
From Eq. (3.8.5), ρ = 0.008. Thus,

required As = ρ bd = 0.008(27)21.5 = 4.64 sq in.
The 7–#8 will be retained, because the factored moment diagram increases more steeply
than the assumed linear increase in development of reinforcement. Even with the
increased value of cb for the general equation, the Ld value will still be a significant por-
tion of the roughly 54 in. available. Using the general equation, with
27 − 3(2) − 1
cb = one-half center-to-center spacing = = 1.67 in.
2(6)
Ld (#8) = 32.8(1.3) = 42.6 in. (3.6 ft)
(Continued)
837
The ψt = 1.3 is the modification factor for the bar casting position when more than 12 in.
of concrete lies beneath the bars. Even though the available development length of about
54 in. exceeds Ld, the flexural design strength, (φ Mn diagram) would encroach on the
Mu diagram. The 7–#8 bars should be hooked at the outside edge of the exterior column
footing.
When hooks are provided as anchorage, the development length Ldh for a #8 bar is,
from Table 6.10.2,
 40 
Ldh (# 8) = 21.9   = 14.6 in.
 60 
where the 40/60 ratio accounts for the Grade 40 reinforcement.

In Fig. 19.13.5, the development of the #8 hooked bars is shown by the sloping
dashed line at the exterior end of the footing, starting from the end of the bar at 3.0 in.
from the side face of the footing.
(d) Investigate one-way shear at the distance d (i.e., 21.5 in.) from the face of the exte-
rior column.
1
Vu = 134 − (43.8)(21.5) = 56 kips
12
For no shear reinforcement and the simplified expression for strength,
φ Vc = φ (2 λ fc′ bw d )
φ Vc = 0.75 2(1.0) 3000  (88)(21.5)

1
= 155 kiips > 56 kips OK
1000
(e) Investigate two-way shear action at the critical section d/2 from the face of the exte-
rior column. At this location there is an unsymmetrical three-sided perimeter, and
there is a bending moment on the section across the 7 ft 4 in. width at the column
face. ACI-8.4.4.2 applies. This shear transfer check is an empirical check of stresses
under factored load similar to that for a flat plate at an exterior column. The floor
slab treatment was discussed in Section 16.18, where the details of the computation
are shown.
At the face of the exterior column, the factored moment* is

Mu = 67 ft-kips ( at face of column)
ACI-8.4.4.2.2 and ACI-8.4.2.3.2 require that the fraction of slab (footing) moment to be
transferred to the column by eccentricity of shear be (see Section 16.18):
1 1
γv = 1− γ f = 1− = 1− = 0.355
2 b1 2 12 + 21.5 / 2
1+ 1+
3 b2 3 12 + 21.5
Thus,

Muv = 0.355 Mu = 0.355(67) = 23.8 ft-kips
(Continued)
* This longitudinal moment is taken from the factored moment diagram of Fig. 19.13.5(b). Where the strap joins the exte-
rior footing there is a sudden discontinuity, which may cause an indeterminate redistribution of longitudinal and transverse
moments.
83
Using the symbols of Fig. 16.18.1, for this example b1 = 22.75 in. and b2 = 33.5 in. The
factored shear that must be carried is
197  1 
Vu = 4.5(7.33) − 22.75(33.5)  = 165 kips
4.5(7.33)  144 
The section properties for the three-sided critical section for shear transfer, as shown in
Fig. 19.13.6(b), are
A = 21.5 [2(22.75) + 33.5] = 1699 sq in.
2(22.75)(11.375)
x2 = = 6.55 in.
2(22.75) + 33.5
 2(22.75)3  22.75(21.5)3
J c = 21.5  − (45.5 + 33.5)(6.55)2  +
 3  6
= 133, 600 in.4
Referring to Fig. 19.13.6, the forces to be considered in analysis of the floor slab and the
footing are compared. Note that the direction of the shear in the floor slab is the opposite
of that in the strap footing; however, the moment direction is the same. Thus, in the floor
slab the maximum factored shear stress occurs at the interior side of the critical section
[as shown in Fig. 16.18.1(b)]. In the footing slab, however, the maximum factored shear
stress will occur at the exterior edge.
Using Eq. (16.18.3), the factored shear stress at the exterior side of column is (note
that at this section the stresses due to Vu and Muv are additive and will be maximum in
the footing slab)
165, 000 23.8(12,000) (22.75 − 6.55)
vu1 = + = 97 + 35 = 132 psi
1699 133, 600
Similarly, using Eq. (16.18.4), the factored shear stress at the interior side of column
(here the shear stresses due to Vu and Muv subtract from each other),
165, 000 23.8 (12,000) (6.55)
vu2 = − = 97 − 14 = 83 psi
1699 133, 600
For β not exceeding 2.0, and b0 /d = [2(22.75) + 33.5]/21.5 = 79/21.5 = 3.7, which is
less than 15 for a three-sided critical section, the stress due to factored loads is limited
to φ (4 λ fc′) as a maximum; for this example, 164 psi (ACI-22.6.5.2). The computed
maximum stress of 132 psi is below 164 psi, and therefore acceptable.
(f) Investigate development of reinforcement at the inflection point (ACI-9.7.3.8.3).
The Code provision is checked here, even though the reinforcement is negative
moment reinforcement rather than the positive moment reinforcement as prescribed
in ACI-9.7.3.8.3 because the footing may be visualized as an inverted beam, as dis-
cussed in Example 19.13.1, part (d).
Tension reinforcement is 3–#8 at the narrow end of the strap near the inflection point.
C = 0.85 fc′ ba = 0.85(3)(14)a = 35.7a
T = 3(0.79)40 = 95 kips
95
a= = 2.66
35.7
1
M n = 95(21.5 − 1.33) = 160 ft-kips
12
Vu = 67 kips (Continued)
839
12 × 12 d/2 = 10.75”
Column (Example 19.13.2)
Example 19.13.2 Critical section
for two-way
shear action
d/2
d/2
Footing
Floor slab V
M M
V
Column
(a) Floor slab (b) Cantilever (strap) footing
Figure 19.13.6 Comparison of forces to be transferred between floor slab and column to those
between exterior footing and column in cantilever (strap) footing.
The #8 bars are proposed to be terminated at the interior side of the interior column,
giving about 2.5 ft of embedment from the inflection point. Thus, La = 2.5 ft, but not to
exceed the larger of
d = 1.79 ft and 12db = 1.0 ft
Thus,
Mn  160 
+ La =  + 1.79 12 = 50 in. > [Ld =38.9 in.] OK
Vu  67 
If this does not work, the bars can be extended beyond the interior face of the interior
column, where they can be hooked down.
(g) Investigate development of reinforcement at a simple support (ACI-9.7.3.8.3): that is,
at the top of the exterior column footing. Since it was determined in part (c) that the #8
bars must be hooked at face of the exterior footing, ACI-9.7.3.8.3 need not be satisfied.
Recall the preceding assumption of zero earth pressure under the strap. Hence the region
below the strap should be disturbed during construction. The strap should be formed
on the bottom. Furthermore, liberal anchorage lengths (perhaps even hooks) should be
provided into the exterior footing [as shown in Fig. 19.13.5(a)] to accommodate fully
the tensile force in the top of the footing across to the exterior column. Often some of the
steel extending into the exterior footing is flared to distribute more effectively the load
from the 27-in. width to the 88-in. width.
(h) Design of exterior footing. In the transverse direction,
d = 25 − 3 (i.e.,cover) − 0.5 (i.e., est. bar radius) = 21.5 in.
2
 197  (2.54)
Mu att edge of strap =   = 87 ft-kips
 7.33  2
Mu 87(12,000)
φ bd 2
0.90(54)(21.5)2
(Continued)
840
From Eq. (3.8.5), required ρ = 0.0012 < [min ρg = 0.002 (ACI-8.6.1.1)],
required As = 0.002(54)(25) = 2.70 sq in.
Use 7–#6 bars (As = 3.08 sq in.). Since bars are to extend the full 7 ft 4 in. length (less
minimum cover) of the exterior footing, development of reinforcement for #6 bars is
satisfied; that is,
available embedment = 30.5 in. > [Ld (#6) = 21.9 in. ( 4 6 of value shown in Table 6.6.1)] OK
Also, provide minimum shrinkage and temperature reinforcement; use 7–#6 bars in the
perpendicular direction.
The beam shear (one-way action) at d from the face of the column in the 7.33-ft direc-
tion of the footing is
197
Vu = (1.38) = 37 kips
7.33
φVc = φ (2 λ fc′ bw d )
1
φVc = 0.75 2(1.0) 3000  (54)21.5 = 95 kips > 37 kips OK
1000
(i) Design of interior footing.

285
net soil pressure under overload = = 6.4 ksf
(6.67)2
(2.75)2
Mu = 6.4(6.67) = 161 ft-kips
2
Mu 161(12, 000)
φ bd 2 0.90(80)(21.5)2
From Eq. (3.8.5), required ρ = 0.0015 < [min ρg = 0.002 (ACI-8.6.1.1)]

required As = 0.002(80)(25) = 4.0 sq in.
Use 7–#7 bars each way (As = 4.20 sq in.). The 2 ft 9 in. from face of column to edge of
footing provides adequate embedment to develop the #7 bars (Ld = 31.9 in.).
Check two-way action for shear. When β < 2 and b0 /d < 20 for a four-sided critical sec-
tion, the nominal strength Vc is based on 4 λ fc′. Thus, using λ = 1.0 for normal weight
concrete,
1
φVc = φ 4 fc′ b0 d = 0.75(4 3000 )(4)(35.5)21.5 = 502 kips
1000
Vu = 6.4 [(6.67)2 − (2.96)2 ] = 229 kips < 502 kips OK
(j) Design sketch. The final details of the design are shown in Fig. 19.13.5(a). Note that
the flexural design strength, φMn diagram, in Fig. 19.13.5(b) for the strap portion is
approximate; the #8 bars are extended as far as possible toward the narrow end as
the width narrows from 27 to 14 in.
841
19.14 PILE FOOTINGS
Main provisions for the design of pile footings are given in ACI-13.4.2. The principles and
methods to be used in the design of pile caps are little different from those of spread foot-
ings. The following, however, may be noted.
1. Computations of factored moments and shears may be based on the assumption that
the reaction from any pile is concentrated at the center of the pile (ACI-13.4.2.2).
2. In computing the external shear on any section through a footing supported on
piles, the portion of the pile reaction to be assumed as producing shear on the
section shall be based on straight-line interpolation between full value when the
pile center is at half the pile diameter (i.e., dp /2) outside the section and zero
value when the pile center is at half the pile diameter (i.e., dp /2) inside the section
(ACI-13.4.2.5).
3. In reinforced concrete pile footings, the overall depth of the pile cap above the bot-
tom reinforcement shall not be less than 12 in. (ACI-13.4.2.1).
Pile caps frequently must be designed for shear considering the member as a deep beam.
In other words, when piles are located inside the critical sections d (for one-way action)
or, d/2 (for two-way action) from the face of column, the shear cannot be neglected (ACI-
13.4.2.3). The problem requires the most attention when large loads are carried by a few
piles, such as a two-pile cap having 100-ton piles. For guidance, the reader is referred to
the CRSI handbook [2.21], Rice and Hoffman [9.2, 19.18], and Gogate and Sabnis [19.19].
Alternatively, ACI-13.2.6.3 allows the design of foundations using strut-and-tie models
(see Chapter 14). This approach may be used, in particular, for the design of pile caps.
Guidance for the design of pile caps using strut-and-tie models is provided by Adebar,
Kuchma, and Collins [14.15].
SELECTED REFERENCES
19.1. A. N. Talbot. Reinforced Concrete Wall Footings and Column Footings. Urbana, IL: Engineering.
Experiment Station Bulletin No. 67, University of Illinois, March 1913.
19.2. Frank E. Richart. “Reinforced Concrete Wall and Column Footings,” ACI Journal, Proceedings,
45, October 1948, 97–127; November 1948, 237–260.
19.3. Johannes Moe. Shearing Strength of Reinforced Concrete Slabs and Footings Under
Concentrated Loads (Bulletin No. D47). Chicago: Portland Cement Association Research and
Development Laboratories, April 1961.
19.4. ACI-ASCE Committee 426. Suggested Revisions to Shear Provisions for Building Codes.
Detroit: American Concrete Institute, 1979, 82 pp.
19.5. Hans Gesund. “Flexural Limit Analysis of Concentrically Loaded Column Footings,” ACI
Journal, Proceedings, 80, May–June 1983, 223–228.
19.6. Hans Gesund. “Flexural Limit Design of Column Footings,” Journal of Structural Engineering,
ASCE, 111, 11 (November 1985), 2273–2287.
19.7. Da Hua Jiang. “Flexural Strength of Square Spread Footings,” Journal of Structural Engineering,
ASCE, 109, 8 (August 1983), 1812–1819. Disc., 112, 8 (August 1986), 1925–1926.
19.8. Kanakapura S. Subba Rao and Shashikant Singh. “Lower-Bound Collapse Load of Square
Footings,” Journal of Structural Engineering, ASCE, 113, 8 (August 1987), 1875–1979.
19.9. Richard W. Furlong. “Design Aids for Square Footings,” ACI Journal, Proceedings, 62, March
1965, 363–371.
19.10. Fritz Kramrisch. “Footings,” Handbook of Concrete Engineering (Mark Fintel, Editor).

New York: Van Nostrand Reinhold Company, 1974 (Chapter 5, pp. 111–140).
19.11. ACI Committee 336. “Suggested Design Procedures for Combined Footings and Mats,” ACI
Structural Journal, 85, May–June 1988, 304–324. Disc., 86, January–February 1989, 111–116.
19.12. ACI Committee 336. Design and Performance of Mat Foundations—State-of-the-Art (SP-
152). Farmington Hills, MI: American Concrete Institute, 1995, 274 pp.
19.13. Fritz Kramrisch and Paul Rogers. “Simplified Design of Combined Footings,” Journal of Soil
Mechanics and Foundations Division, ASCE, 87, SM5 (October 1961), 19–44.
842
19.14. Leslie J. Szava-Kovats. “Design of Combined Footings Using Support Reaction and Moment
Influence Lines of Continuous Beam on Elastic Supports,” ACI Journal, Proceedings, 64, June
1967, 312–319.
19.15. Gwynne Davies and Brian Mayfield. “Choosing Plan Dimensions for an Eccentrically Loaded
Footing Slab,” ACI Journal, Proceedings, 69, May 1972, 285–290.
19.16. Joseph E. Bowles. “Mat Design,” ACI Journal, Proceedings, 83, November–December 1986,
1010–1017. Disc., 84, September–October 1987, 449.
19.17. Report of the Joint Committee of Standard Specifications for Concrete and Reinforced Concrete.
Detroit: American Concrete Institute, 1940.
19.18. Paul F. Rice and Edward S. Hoffman. “Pile Caps—Theory, Code, and Practice Gaps,” CRSI
Professional Members’ Structural Bulletin No. 2, February 1978.
19.19. Anand B.Gogate and Gajanan M. Sabnis. “Design of Thick Pile Caps,” ACI Journal,

19.20. ACI Committee 336. Suggested Analysis and Design Procedures for Combined Footings and
Mats. ACI 336.2R-88 (Reapproved 2002). Farmington Hills, MI: American Concrete Institute,
2002, 20 pp.
PROBLEMS
All problems are to be done in accordance with the ACI Code, and all loads given are
service loads unless otherwise indicated.
19.1 Design a square spread footing to support a 14-in. 19.3 Investigate the transfer of load from column
square tied column carrying a dead load of 120 kips to footing for the two conditions of the fig-
and a live load of 90 kips. The column reinforce- ure for Problem 19.3. The column is 20 in.
ment consists of #8 bars. Use fc′ = 4000 psi for square ( fc′ = 4000 psi) containing 12–#10 bars
the column, fc′ = 3000 psi for the footing, and ( fy = 60,000 psi) spirally reinforced. The footing
fy = 60,000 psi. Use a 6 ft 9 in. square footing. is 10 ft square and 28 in. thick ( fc′ = 3000 psi)
19.2 Design a square spread footing to support a and is adequately reinforced. Use the provisions
20-in. square tied column carrying a dead load of ACI-16.3 and ACI-14.5.
of 400 kips and a live load of 264 kips. The col- 19.4 Design a spread footing to carry a load from
umn reinforcement consists of #11 bars. Use an 18 × 32 in. tied column. Dead load and live
fc′ = 3000 psi for both the column and footing, load are each 230 kips. Because of the close-
fy = 60,000 psi, and allowable soil pressure = ness of the property line (see the figure for
5 ksf. Include a design sketch. Problem 19.4), the footing cannot exceed 8 ft
30” × 30” × 12”

20” 20”
pedestal
580k (60% live load) 580k (fc’ = 3,750 psi)
2 2
1 1 1
10’– 0” 10’– 0”
(a) Without pedestal (b) With pedestal
Problem 19.3
Property line
4’– 0” 18”
8’– 0” 32”
Maximum
footing
width
Problem 19.4
843
PROBLEMS 843
perpendicular to that line. Use fc′ = 3000 psi , fy = Equal settlement is taken for DL plus LL con-
60,000 psi, and allowable soil pressure = 5 ksf. ditions at a uniform soil pressure of 3 ksf. Use
19.5 Design a plain concrete footing to carry a long fc′ = 4000 psi and fy = 60,000 psi.
12-in. concrete block wall that must transmit 19.8 Design a rectangular combined footing for
6 kips/ft (60% live load) to the footing. Use the conditions of the figure for Problem 19.8.
fc′ = 3000 psi and allowable soil pressure = Assume that the system is within a building
4 ksf. basement and is to have a 4-in. concrete slab over
19.6 Design a reinforced concrete footing to carry a the footing. Use fc′ = 3000 psi , fy = 60,000 psi,
12-in. concrete wall to carry 20 kips/ft (60% live and allowable soil pressure for the given loads =
load). Use fc′ = 3000 psi , allowable soil pres- 5 ksf.
sure = 3 ksf, and fy = 60,000 psi. 19.9 Rework Example 19.13.2 using fc′ = 5000 psi
19.7 Design a rectangular combined footing for the and fy = 80,000 psi.
situation shown in the figure for Problem 19.7.
Column data are in the following table.
Column Size Reinforcement LL DL
Exterior 12 in. square #9 bars 45 kips 120 kips

Interior 12 in. square #9 bars 90 kips 155 kips
Property line
1’– 4” 18’– 0”
12” × 12” column 12” × 12” column
Problem 19.7
Property line
18” × 18” column 20” diameter column
P = 200k (50% live load) P = 450k (40% live load)
9” 16’– 0”
Problem 19.8
CHAPTER 20
INTRODUCTION TO PRESTRESSED
CONCRETE
20.1 INTRODUCTION
In very simple terms, “prestress” means a stress that acts even though no externally applied
loads are acting. The principle of prestressing has been used for centuries. For example,
wooden barrels may be made by tightening metal bands or ropes around barrel staves. The
tensile stress in the bands causes a compression between the staves, thus making the barrel
tight. In the making of early automobile wheels, the wooden spokes and rim were first held
together by a hot metal tire. Upon cooling, the tensile stress due to shrinkage in the metal
would then compress the wooden rim and spokes together. In bolted joints, the bolt is pre-
tensioned by tightening, which in turn precompresses the elements being joined.
The primary application of prestressing on a large scale today is in concrete construc-
tion. In general, prestress involves the imposition of stresses opposite in sign to those
caused by the subsequent application of service loads. For example, prestressing wires
placed eccentrically in a simple beam, as shown in Fig. 20.1.1, produce in the concrete an
axial compression as well as a negative bending moment. Thus it is possible to keep the
entire section in compression when service loads are added. This is a great advantage, since
concrete is weak in tension but strong in compression. Of course, steel is used to impose
the prestress, though less is required for prestressed concrete than in ordinary reinforced
concrete. In general, prestress provides a means for the most efficient use of material—that
is, steel in tension and concrete in compression.
20.2 HISTORICAL BACKGROUND
The general concepts of prestressed concrete were first formulated in the period 1885–1890
by C. F. W. Doehring in Germany and P. H. Jackson in the United States. These early appli-
cations were handicapped by the low steel strengths obtainable at the time. Steel stressed
to low tension levels will not precompress concrete adequately to maintain its compression
after shrinkage and creep have occurred.
The theory of prestressed concrete was first propounded by J. Mandl of Germany in
1896. It was further advanced by M. Koenen of Germany in 1907 (the first recognition of
losses in prestress force from elastic shortening of concrete) and by G. R. Steiner in the
United States in 1908. Steiner recognized prestress losses due to shrinkage and suggested
retensioning after shrinkage had occurred.
845
2 0 . 3 A DVA N TAG E S A N D D I S A DVA N TAG E S 845
Prestressed concrete T-girders erected on precast columns for a parking garage. (Photo by
C. G. Salmon.)
In practical uses, R. E. Dill of the United States in 1928 produced prestressed planks
and fence posts. Circular prestressing of storage tanks began about 1935, but no significant
linear prestressing (beams, slabs, planks, etc.) was done until about 1950. The Walnut Lane
Bridge in Philadelphia, built in 1949–1950, was the first major use of linear prestressing in
the United States.
In Europe, however, linear prestressing began about 1928 and advanced rapidly with
the work of F. Dischinger, E. Freyssinet, E. Hoyer, G. Magnel, Y. Guyon, P. Abeles, and
F. Leonhardt. With the publication of Magnel’s work on the loss of stress in work-hardened
steels in 1944, the basic theory of prestressed concrete was sufficiently complete for suc-
cessful economical applications. In the United States, T. Y. Lin was a leading proponent
and practitioner.
The use of prestress is now widespread in nearly every type of simple structural element,
as well as in many statically indeterminate structures. Zollman [20.1] has presented his
interesting “Reflections” on the beginnings of prestressed concrete in America. Methods of
inducing prestress are ingenious and unique. Many textbooks are available that describe the
methods and applicable theories [20.2–20.9].
20.3 ADVANTAGES AND DISADVANTAGES OF

PRESTRESSED CONCRETE CONSTRUCTION
The original concept of prestressed concrete was that the material was crack free under
service loads. Especially when a structure is exposed to the weather, elimination of cracks
prevents corrosion. Also a crack-free prestressed member has greater stiffness under serv-
ice loads because its entire section is effective.
Prestressed concrete in several respects is more predictable than ordinary reinforced con-
crete. It permits accommodation of both shrinkage and creep reasonably well. Also, high-
strength concrete may be more efficiently utilized by merely adjusting the prestress force.
Precompression of the concrete reduces the tendency for inclined cracking, and the use
of curved tendons provides a vertical component to aid in carrying shear. Shear strength
exhibits less variability than in ordinary reinforced concrete.
846
846 C hapter 2 0 I ntroduction to P restressed C oncrete
Mb = bending moment from Stress

applied loads distribution
t
(a) Effect of service loads on beam
P e c
P
Prestressed wire Prestressing force
(b) Effect of prestress with eccentricity
Mb
c
P e
P
(c) Combined effects of service load and prestress
Figure 20.1.1 Opposite effects of service load and prestress on a simply supported beam.
Other features of prestressed concrete are its high ability to absorb energy (impact resist-
ance); its high fatigue resistance, due especially to the low steel stress variation resulting
from the high initial pretension; and its high live load capacity, arising from the ability to
design the prestressing tendon to support the dead load. As a result, higher span-to-depth
ratios can be achieved with prestressed concrete elements.
Some of the disadvantages of prestressed concrete construction are as follows: (1) the
stronger materials used have a higher unit cost; (2) more complicated formwork may be
necessary; (3) end anchorages and bearing plates are often required; (4) labor costs are
greater; and (5) more conditions must be checked in design, and closer control of every
phase of construction is required.
Short-span members and single-unit applications of any kind are likely to be uneco-
nomical in prestressed concrete. However, economy is usually achieved when units can be
standardized and the same unit repeated many times. For many situations, the desirability
of achieving a certain advantage is sufficient to justify a higher initial cost.
Currently, prestressing need not create a crack-free structure at service load. In fact,
prestressing to various levels of stress can provide the whole range of results, from “fully”
prestressed, when at service load no cracking occurs, to nonprestressed, as discussed in
preceding chapters. The level of prestressing can be used to accomplish the desired crack
control or stiffness objective. “Partial” prestressing has become common in construction. In
general, this chapter focuses on prestressed concrete where the section is uncracked at serv-
ice load. Mohan Rao and Dilger [20.17] have discussed flexural crack control in cracked
prestressed members.
20.4 PRETENSIONED AND POST-T ENSIONED

BEAM BEHAVIOR
Since discussion in this chapter is limited to beams, it is well to consider at this stage the
behavior of such a member as it relates to the method of inducing the prestress. The most
commonly used procedure is to put a specified tensile force into the wires by stretching
them between two anchorages prior to placing the concrete into the forms. The concrete
is then placed and allowed to cure. As the concrete hardens, the wires become bonded to
847
2 0 . 4 P R E T E N S I O N E D A N D P O S T- T E N S I O N E D B E A M B E H A V I O R 847
the concrete throughout their length. After the concrete has reached a minimum prescribed
strength, the wires are cut at the anchorages. The immediate shortening of the wires trans-
fers through bond (i.e., interaction between the steel and the surrounding concrete) a com-
pressive stress to the concrete. Such a process is called pretensioning.
The behavior associated with pretensioning will be described step by step. In addition,
the terminology and allowable values will be given in accordance with Chapters 20 and 24
of the ACI Code.
In Step 1, as shown schematically in Fig. 20.4.1(a), the wires are stretched between two
anchorages in the casting yard sufficiently to introduce a tensile stress fsi into the wires,
which according to ACI-20.3.2.5.1 may not exceed 94% of the specified yield strength
fpy, but not greater than the lesser of (1) 80% of the specified tensile strength fpu, or (2) the
maximum value recommended by the manufacturer of prestressing tendons or anchorages.
The quality-controlled concrete is then placed in the forms and frequently is steam cured.
Concrete strength must be adequately developed by the time the compression is to be intro-
duced; thus high early-strength cement is usually used. Generally, the concrete strength fci′
at transfer is specified by designers to be 4000 to 4500 psi.
Step 2 is to cut the wires. Acting through bond, the force T0 in the wires acts as a com-
pressive force on the entire effective (transformed) section. The stress in the concrete goes
from zero [Fig. 20.4.1(b)] before the wires are cut to that shown in Fig. 20.4.1(c) after they
have been cut. Once the prestress has been introduced, certain losses of prestress begin to
occur. Loss of prestress may arise from slip at the anchorage, elastic shortening of the con-
crete member, creep and shrinkage of the concrete, relaxation of steel stress, and frictional
losses due to intended or unintended curvature in the tendons. It is true that some small
portion of such losses may occur prior to the transfer of stress to the concrete; however, it is
practical and conservative to assume that the losses occur after the transfer.
Dead load of the flexural member will, of course, be acting simultaneously with the
prestressing force once the wires have been cut and transfer of load to the concrete has
been accomplished. Dead load combined with prestress is shown in Fig. 20.4.1(d), where
the most severe stress situation occurs immediately after transfer and before most losses
have taken place. Limiting values for this temporary situation are (a) a tensile stress of
3 fci′ (approximately 40% of the cracking strength), except at the ends of simply supported
members, where the maximum allowable tensile stress is 6 fci′ (a conservative value for the
modulus of rupture; see Section 1.8) (ACI-24.5.3.2) and (b) a compressive stress equal to
60% of the concrete compressive strength fci′ that has been developed at the time of transfer,
except for ends of simply supported beams, where a compressive stress equal to 0.70 fci′ is
allowed (ACI-24.5.3.1).
One reason for limiting the temporary tensile stress to such a low value is to prevent any
possibility that an upward buckling of the beam will result from sudden cracking at the top.
Step 3 is the service load condition. Since 2002, the ACI Code has classified pre-
stressed members based on the extreme fiber tensile stress ft at service loads as follows
(ACI-24.5.2.1):
If ft ≤ 7.5 fc′ the section is assumed to behave as uncracked—Class U

If ft > 12 fc′ the section is assumed to behave as cracked—Class C
If 7.5 fc′ < ft ≤ 12 fc′ the section is assumed to be in transition between uncracked
and cracked—Class T
except for prestressed two-way slab systems, which are classified as Class U members
with ft ≤ 6 fci′ . Because cracking of the cross section changes its behavior and properties
under service loads, the code establishes different serviceability requirements depending
on whether the section is uncracked or cracked. These requirements are summarized in ACI
Table R24.5.2.1.
For Class U and T members, the stresses at service loads are permitted to be calcu-
lated using uncracked section properties (ACI-24.5.2.2). For Class C members, however,
the stresses should be computed using a cracked transformed section (ACI-24.5.2.3) (see
Section 12.5).
84
Anchorage
Stress fsi
(a) Wires stretched to a stress fsi; then concrete placed
To To = Asfsi
(b) Before cutting wires; no stress in concrete
ft < ft
e Before After
losses losses
To To
fc < fc
(c) After cutting wires; concrete compressed (transformed area effective)
wD
[
ft ≤ fit = 3 fci‘ ]
Before
To To losses
[
fc ≤ fic = 0.60fci‘ ]
(d) Dead load plus prestress acting
wD + sust loads [ ]
fc ≤ ffc = 0.45fci‘ (Class U and T)
After
T < To T < To losses
[ f ‘ ](Class U)
ft ≤ fft = 7.5 c
[f = 7.5
ft f ‘ ]< f ≤[ f = 12 f ‘ ] (Class T)
c t ft c
(e) Dead load, prestress, and sustained live load after all prestress losses
wD + L [ ]
fc ≤ ffc = 0.60fc‘ (Class U and T)
After
T < To T < To losses
[ f ‘ ](Class U)
ft ≤ fft = 7.5 c
[f = 7.5
ft f ‘ ]< f ≤[ f = 12 f ‘ ] (Class T)
c t ft c
(f) Dead load, prestress, and live load after all prestress losses
Figure 20.4.1 Stages of behavior up to service load—pretensioned beam.
Two service load conditions are recognized (ACI-24.5.4.1): (1) dead load + prestress

(after losses) + sustained loads and (2) dead load + prestress (after losses) + all loads.
Although live loads are generally transient, it is possible for a portion of the live load to act
long enough to cause significant long-term deflections. In such a case, that portion of the
live load should be considered to be sustained in design.
849
2 0 . 5 S E RV I C E L OA D S T R E S S E S O N F L E X U R A L M E M B E R S 849
Jacking device
Hollow ducts for tendons
Figure 20.4.2 Post-tensioned member.
Class U and T Members

For the condition of prestress (after losses) + sustained loads, ACI-24.5.4 permits a com-
pressive stress (based on uncracked section properties) not to exceed 0.45 fc′ . For the con-
dition of prestress (after losses) + all loads, the permitted compressive stress is 0.60 fc′ .
According to the ACI Commentary R24.5.4.1, the compressive stress limit (i.e., 0.45 fc′ )
was “originally established to decrease probability of failure … due to repeated loads.”
However, fatigue tests showed that crushing failure in concrete does not control the strength
of prestressed concrete members; thus, the limit was raised to 0.60 fc′ for situations having
large transient live loads.
Class C Members
There are no stress limit requirements for this class of members. However, since Class C
members are expected to be cracked, they must satisfy the crack control requirements of
ACI-24.3.2 (see Section 12.18).
The alternative to pretensioning is post-tensioning. In a post-tensioned beam, the con-
crete is first cast either with a hollow tube enclosed or with unstressed tendons coated
with grease or mastic to prevent bond with the concrete, as shown in Fig. 20.4.2. An end
plate or anchorage is placed against each end of the member; then, once the concrete has
hardened and reached the desired strength, the wires are pulled by jacking against the end
plates. Tensile stress in the prestressing steel immediately after anchorage is not to exceed
70% of specified tensile strength fpu [ACI-20.3.2.5.1]. During the tensioning process, elastic
shortening occurs, frictional losses take place, and the dead load moment becomes partially
active due to the induced curvature. When the prestressing wires or strands are tensioned
in sequence, some loss of prestress will also occur in the reinforcement tensioned first as
further shortening (due to axial shortening and curvature of the member) takes place during
tensioning of additional tendons. These loses can often be accounted for when determining
the jacking force.
For post-tensioning, the stresses induced and the allowable values at the different stages
are essentially the same as those described in detail for pretensioning. However, since the
tendons are not bonded to the concrete, bending of the member will affect the strains and
stresses in the tendons along their entire length rather than locally, as is the case in pre-
tensioned members. Prior to the imposition of live load, the tendons are usually grouted
(i.e., the space in the ducts is filled). If such grouting is properly done, the strains and
stresses in the tendons will vary locally with changes in curvature caused by the live load,
which may be computed on the transformed section, the same as for pretensioning.
20.5 SERVICE LOAD STRESSES ON FLEXURAL

MEMBERS—T ENDONS HAVING VARYING
AMOUNTS OF ECCENTRICITY
To demonstrate some of the attributes of prestressed concrete, in Example 20.5.1 the pre-
stressing elements are placed at the centroid of the section, giving a uniform precompres-
sion of the concrete.
850
EXAMPLE 20.5.1
For the section shown in Fig. 20.5.1 assume that the member is pretensioned by
2.30 sq in. of steel wire having an initial tensile stress of 175,000 psi. The prestress wires
are centered at the centroid of the section. The concrete has fc′ = 5000 psi (modular ratio
n = 7), and it is to be assumed that the concrete has attained a strength of fci′ = 4000 psi
at the time of transfer. Assume that the section will behave as an uncracked section.
500 + fLL 1171 psi 1604 psi
(–)
(–) (–)
30” (–)
Aps = 2.30 sq in.
15”
(+)
(+)
20” 171 psi 530 psi
671 psi 500 + fLL
(a) (b) Initial (c) DL + LL (d) Temporary (e) Final stress

prestress stresses stress after at full service
transfer load
Figure 20.5.1 Section and stresses for Example 20.5.1.
Determine (a) the stresses due to prestress immediately after transfer, (b) the tempo-
rary stresses when the member is used on a 40-ft simple span, and (c) the service live
load moment capacity according to the ACI Code, allowing for a 20% loss of prestress
due to creep, shrinkage, and other sources.
SOLUTION
(a) Stress due to prestress immediately after transfer.

T0 = fsi Aps = 175(2.30) = 402.5 kips
This force T0 acts as a compressive force on the transformed section immediately after
the cutting of the wires. The stress in the concrete immediately after transfer can be
determined as
T0 T0 402, 500 402, 500

fc = = = = = 656 psi
Ac + nAps Ag + (n − 1) Aps 600 + 6(2.3) 614
The decrease in steel stress due to elastic shortening of the member is therefore (see also
Section 20.7)

∆fs = nfc = 7(656) = 4590 psi
Thus elastic shortening may be considered to have caused a loss of tensile stress, so that
the remaining tensile stress in the wires is 175,000 –4590 = 170,400 psi. The loss in this
case is 2.6%, but it could be as high as 5%.
(Continued)
851
2 0 . 5 S E RV I C E L OA D S T R E S S E S O N F L E X U R A L M E M B E R S 851
Although it may be theoretically correct to use the transformed section, in ordinary

practice it is common and sufficiently accurate in most cases to use the gross section.
Since the prestressing force is applied at the centroid of the gross section in the present
problem, fc is uniform over the entire section, or
402, 500
fc = = 671 psi
600
which is little different from the value of 656 psi determined above by using the trans-
formed section.
(b) Temporary stress-prestress plus dead load.
20(30)
wD = (150) = 625 plf
144
1
M D = (0.625)(40)2 = 125 ft-kips
8
Using the approximate method with gross moment of inertia Ig and neglecting the trans-
formed area of reinforcement,
1
Ig = (20)(30)3 = 45, 000 in.4
12
402, 500 125(12,000) (15)
f (initial prestress + DL) = − 
600 45, 000
= −671  500

= −1171 psi (compressiion, top)
= −171 psi (compression, bottom)
These stresses are acceptable based on temporary stress restrictions immediately after
transfer and before losses,
fc (max) = 0.60 fci′ = 2400 psi

ft (max) = 3 fci′ = 190 psi
(c) Service live load moment capacity.

f (prestress − losses + DL) = − 0.8(671)  500 = −537  500
= −1037 psi (compression, top)

= −37 psi (compression, bottom)
Based on stress at service load (prestress + DL + LL) [ACI-24.5.4.1 and ACI-24.5.2.1]

for uncracked Class U members, after allowance for all prestress losses,
fc (max) = 0.60 fc′ = 3000 psi

ft (max) = 7.5 fc′ = 530 psi
(Continued)
852
The stress available for live load may then be computed.

f (prestress − losses + DL + LL) = −1037 + fLL = −3000 psi (top)
= −37 + fLL = +530 psi (bottom)
fLL (max,top) = −1963 psi
fLL (max,bottom)) = +567 psi
Controls!
Thus
M L (15)
= 567 psi
45, 000

45, 000(567) 1
ML = = 142 ft-kips
15 12,000
The wide divergence between the maximum acceptable live load stresses of 1963 psi
compression and 567 psi tension indicates the need for an unsymmetrical section or
some arrangement to equalize the stresses better. A study of Fig. 20.5.1 will show that
the most economical arrangement would be for the initial prestress variation to offset the
pattern of final stress under full dead load plus live load.
EXAMPLE 20.5.2
Repeat the solution of Example 20.5.1, except locate the tendons 5 in. from the bottom
of the section (Fig. 20.5.2).
671 500 + fLL +171 1603 psi
(+) (–) (+)
(–)
30”
Aps = 2.30
(–) (–)
5” (+) (+)
20” 2013 500 + fLL 1513 530 psi
(a) (b) Initial (c) DL + LL (d) Temporary (e) Final stress

prestress stresses stress after at full service
transfer load
Figure 20.5.2 Section and stresses for Example 20.5.2.
SOLUTION
(a) Temporary stress (prestress + DL immediately after transfer). Using properties
of the gross section as for Example 20.5.1, the prestressing force T0 = 402.5 kips
applied with an eccentricity of 10 in. gives
402, 500(10)(15)
f (initial prestress + DL) = − 671 ±  500
45, 000
= −671 ± 1342  500

= +171 psi (tension, top)
= − 1513 psi (compression, bottom)
(Continued)
853
20.6 THREE BASIC CONCEPTS OF PRESTRESSED CONCRETE 853
Note that the temporary tensile stress of 171 psi at the top is nearly equal to the allowable
value of 3 fci′ = 190 psi for such stress. Any significantly greater eccentricity, therefore,
would require a reduction in the prestressing force.
(b) Final stress (prestress + DL), after allowance for 20% prestress loss, is
f (prestress + DL − losses) = +171 − 0.2( −671 + 1342)
= +37 psi (top)

= −15513 – 0.2(–671 – 1342)
= −1110 psi (bottom)
(c) Service live load moment capacity (see Example 20.5.1 for allowable stresses).
Based on final dead load plus live load conditions,
+37 + fLL = −3000 psi (top), fLL = −3037 psi

−1110 + fLL = +530 psi (bottom), fLL = +1640 psi
Since the neutral axis for live load resistance is assumed to be at middepth, fLL = +1640
psi controls.
45, 000(1640) 1
ML = = 410 ft-kips
15 12,000
Thus, increasing the eccentricity of the prestressing force increases the live load
capacity until the limit is reached—that is, when the temporary stress at transfer reaches
its maximum permissible value, either at the top or at the bottom of the section.
It is to be noted that the magnitude of the prestress over the concrete section is
constant for the entire span when the tendons are straight, whereas the magnitude
of dead and live load stresses is a maximum at only one point. For straight tendons,
the complete stress situation near the supports on simple spans approaches that of
Fig. 20.5.2(b), less losses, because the superimposed dead and live load stresses van-
ish. Because of this, tendons are frequently placed to have an eccentricity that varies
from zero at points of low external bending moment to a maximum in the region of
high external bending moment. This variation in eccentricity may be accomplished
in pretensioning by holding down the stressed tendons at midspan, or at other loca-
tions such as at the one-third points. In post-tensioning, the ducts or greased tendons
are simply draped (held at the ends and permitted to take a natural deflected shape)
such that desired eccentricities at the ends and at midspan are achieved; the points in
between will lie on a curved path.
20.6 THREE BASIC CONCEPTS OF PRESTRESSED

CONCRETE
When considering the stresses in prestressed concrete under service load conditions, three
general patterns of thought may be applied.
Homogeneous Beam Concept

Section 20.5 used the homogeneous beam concept, wherein the prestressing effectively
eliminates cracking and the combined stress formula, P/A ± Mc/I, may be used to investi-
gate the section. Two examples appear in Section 20.5.
854
Internal Force Concept

The internal force approach uses the equilibrium of internal forces; steel takes the ten-
sion and concrete takes the compression, as shown in Fig. 20.6.1. This approach is
analogous to the internal-couple method used for nonprestressed reinforced concrete.
At service load (assuming linear elastic behavior) in reinforced concrete, the points of
action of the forces C and T (where C = T) are independent of the magnitude of applied
bending moment, depending only on the cross-sectional dimensions and the modulus
of elasticity ratio n; thus the magnitude of the forces is directly proportional to the
applied bending moment. In prestressed concrete under service load, the magnitude of
internal forces is basically independent of applied bending moment, depending only
on the prestress and the percentage of losses. In this case the location of the force C
must vary with the applied loading. The approach may be summarized by the following
steps:
1. A known prestress force put into the steel defines T.

2. A moment M is applied on the beam.
3. For equilibrium, the moment arm = M/T and C = T.
4. Knowing the magnitude and point of action of the force C, the stress in the concrete
may be computed as
C Cec y
f = ± (20.6.1)
A I
C.G. of concrete
section
ec
C
Arm
Figure 20.6.1 Internal force concept of prestressing.
EXAMPLE 20.6.1
Apply the dead load moment of 125 ft-kips and the live load moment of 410 ft-kips (as
computed in Example 20.5.2) to the rectangular beam of Fig. 20.5.2, using the internal
force concept. Determine the service load stresses at transfer and under final conditions.
SOLUTION
(a) At transfer, the prestress force T0 is 402.5 kips.

C = T0 = 402.5 kips
When the applied moment is 125 ft-kips, the moment arm of the internal forces
must be
M D 125(12)
arm = = = 3.73 in.
C 402.5
(Continued)
85
20.6 THREE BASIC CONCEPTS OF PRESTRESSED CONCRETE 855
This means the compressive force C is eccentric to the middepth by an amount,

ec = 15 − 5 (to steel) − 3.73 (arm)
= 6.27 in. (below middepth)
402.5 402.5(6.27)15
f =− ±
600 45, 000
f (top) = − 671 + 842 = +171 psi (tension)
f (bottom) = − 671 − 842 = −1513 psi (compression)
exactly the same as in Example 20.5.2, Fig. 20.5.2(d).
(b) At the final condition, the prestress force Te is 0.8(402.5) = 322 kips after losses.
C = Te = 322 kips
M D + M L (125 + 410)12
arm = = = 19.93 in.
C 322
ec = 15 − 5 − 19.93 = −9.93 in. (above middepth)
322 322(9.93)15
f =− 
600 45, 000
f (top) = − 537 − 1065 = −1602 psi (compression)
f (bottom ) = − 537 + 1065 = +528 psi (tension)
exactly as shown in Fig. 20.5.2(e).
Load Balancing Concept

The load balancing approach visualizes prestressing primarily as a process of balancing
loads on the member. The prestressing tendons are placed so that the eccentricity of the
prestressing force varies in the same manner as the moments from applied loads, which if
exactly done would result in zero flexural stress. Only the axial stress P/A (P is the horizontal
component of force in tendon) would act. Refer to Fig. 20.6.2(a), showing a parabolically
draped prestressing tendon. Figure 20.6.2(b) shows the free body of forces acting on the
concrete due to prestress alone. The prestressing effect may be considered as an upward
uniform load if the tendon is parabolically draped. The maximum prestress moment of T emax
at midspan can be equated to an equivalent uniformly loaded beam moment, wp L2/8; thus
8Temax
wp = = equivalent uniform load (acting upward) (20.6.2)
L2
Let
wnet = w(actual downward load ) − w p
then
wnet L2
M net = (20.6.3)
8
and
C M net y
f =−  (20.6.4)
A I
If the tendons are not parabolically draped, the actual net moment (applied load moment
minus prestress moment) may be used for Mnet in Eq. (20.6.4).
856
Parabolic tendon
T T
emax
L/2
(a) Member having parabolically draped tendon
C C
wp
(b) Forces acting on concrete from prestress alone
Figure 20.6.2 Load balancing concept of prestressing.
EXAMPLE 20.6.2
Compute the stresses on the beam of Fig. 20.5.2 using dead load and live load moments
of 125 and 410 ft-kips, respectively. Use the load balancing concept.
SOLUTION
The maximum moment due to prestress at transfer is
 10 
M prestress = 402.5   = 335 ft-kips
 12 
M net = M D − M prestress = +125 − 335 = −210 ft-kips (negative bending)
402.5 ( −210)(12)15
f =− 
600 45, 000
f (top) = −671 + 842 = +171 psi (ttension)

f (bottom) = −671 − 842 = −1513 psi (compression)
At the final condition,
M net = 125 + 410 − 322(10) /12 = 267 ft-kips (positive bending)
322 267(12)(15)
f =− 
600 45, 000

f (top) = −537 − 1068 = −1605 psi (comppression)
f (bottom) = −537 + 1068 = +531 psi (tension)
The results agree with those previously obtained.
20.7 LOSS OF PRESTRESS
The amount of prestress actually existing in a prestressed concrete member is not easily
measured. The total force in the tendons at the time of prestressing is all that may conven-
iently be determined. Various losses reduce the prestress force during and after initial pre-
stressing of the tendons. The difference between the final available prestress and the initial
value is referred to as the loss of prestress.
In practice the initial prestress is usually determined by a pressure gauge on the jack
and may be verified by a direct measurement of the tendon elongation. In pretensioned
857
20.7 LOSS OF PRESTRESS 857
members, the uniformity of initial prestress may be verified at several points. In certain
post-tensioning procedures, the prestressing force will diminish owing to friction at points
remote from the jacking source, and even to friction in the anchorage device.
Elastic Shortening
The loss of prestress due to elastic shortening can be easily determined. For example, let T0
be the prestressing force that is applied at the centroid of the concrete section in a preten-
sioned member. If Tf is the final tensile force in the tendons just after elastic shortening has
occurred, the strain (unit shortening) in the concrete may be expressed as
fc Tf
εc = = (20.7.1)
Ec Ac Ec
where Ac = Ag – Aps. The change in strain in the tendons as a result of losses is
T0 − T f
∆ε s = (20.7.2)
Aps Es
Equating the expressions for εc and Δεs gives
T0 Ac + nAps AT (20.7.3)
= =
Tf Ac Ac
The loss of prestress Δ fs is

T0 − T f nT f nT0
∆fs = = = (20.7.4)
Aps Ac AT
or
nT0
∆fs ≈ (20.7.5)
Ag
For beams with eccentric tendons, the loss in prestress due to elastic shortening and bend-
ing of the section is obtained as n times the computed compressive stress in the concrete
adjacent to the tendons.
In the post-tensioning case, the tendons are not stretched simultaneously. Thus, tendons
tensioned first will undergone losses due to elastic shortening that takes place as other
tendons are stressed. The various methods of accounting for these gradual losses are ade-
quately described elsewhere [20.2–20.8].
EXAMPLE 20.7.1
Determine the percent loss of prestress due to elastic shortening and bending in the
pretensioned member of Fig. 20.5.2. Use fc′ = 5000 psi with n = 7 and fsi = 175,000 psi.
SOLUTION
(a) Prestress loss due to elastic shortening, neglecting bending. In the “exact” method
[see also part (a) of Example 20.5.1],
AT = 20(30) + (7 − 1)2.30 = 614 sq in.

nT0 7(402.5)
∆fs = = = 4.59 kssi
AT 614
4.59
percent loss = = 2.62%
175 (Continued)
85
In the approximate method,
nT0 7(402.5)
∆fs = = = 4.70 ksi
Ag 600

4.70
175
There is no significant difference in the two results. Three percent is a typical value
for loss due to elastic shortening in a pretensioned beam, whereas such loss in a post-
tensioned beam would be on the order of 1.5%, average.
(b) Prestress loss, including bending due to eccentricity of prestressing force and dead
load. By the approximate method (using gross section instead of transformed sec-
tion), the stress in the concrete adjacent to the tendons is (using the internal force
concept),
M D 125(12)
arm = = = 3.73 in.
T0 402.5
C = T0 = 402.5 kips
C C (15 − 5 − 3.73)10
fc = − −
Ag Ig

402.5 402.5(6.27)10
=− −
600 45, 000
= −671 − 561 = −1232 psi
This is the stress in the concrete 5 in. from the bottom of the beam. The change in steel
stress is nfc.
∆fs = nfc = 7(1.232) = 8.6 ksi

8.6
175
Loss Due to Creep in Concrete

Creep is the time-dependent deformation that occurs in concrete under stress, and has
already been discussed (see Sections 1.11 and 12.11). The strain due to creep will vary
with the magnitude of stress and in general may be assumed to vary with the elastic strain
from about 100% in humid atmosphere to about 300% in very dry atmosphere.
In Chapter 12, Eq. (12.11.1), the creep coefficient Ct is defined as
creep strain, ε cp
Ct = (20.7.6)
initial elastic strain, ε i
The elastic strain in the concrete at the centroid of the section is ( fc = stress at centroid)
fc
εi =
Ec
 f 
ε cp = Ct ε i = Ct  c  (20.7.7)
 Ec 
859
The strain in the concrete due to creep equals the decrease in strain in the steel; thus
 f 
∆ε s = ε cp = Ct  c  (20.7.8)
 Ec 
also
∆fs
∆ε s = (20.7.9)
Es
Then, equating Eqs. (20.7.8) and (20.7.9),
∆fs = Ct nfc (20.7.10)
The coefficient Ct may be determined using the general expressions given earlier (see
Section 12.11), or as suggested by Zia, Preston, Scott, and Workman [20.10], using
Ct = 2.0 for pretensioned members and Ct = 1.6 for post-tensioned members. The stress fc
in Eq. (20.7.10) is recommended [20.10] to be taken as (fcir – fcds), where fcir (the subscript
r means residual) is the net compressive stress in the concrete at the center of gravity of
tendons immediately after the prestress has been applied to the concrete, and fcds (the sub-
script s means superimposed) is the stress in concrete at the center of gravity of tendons
due to all superimposed permanent dead loads that are applied to the member after it has
been prestressed. For example, fcir would correspond to fc = 1232 psi computed in part (b) of
Example 20.7.1 because the prestress and the beam dead load act simultaneously when the
prestress is applied. Typical values for percentage loss of prestress due to creep are from 5
to 6%. Pretensioned beams will exhibit more creep than post-tensioned beams because the
prestress is imposed when the concrete is at an earlier age; age at loading is a major factor
in determining the magnitude of creep.
Loss Due to Shrinkage in Concrete

Shrinkage is the volume change in concrete that occurs with time, as discussed earlier (see
Sections 1.11 and 12.12). The loss of prestress due to shrinkage may be expressed as
∆fs = ε sh Es (20.7.11)
where εsh is the shrinkage strain in concrete (see Section 12.12). Chapter 12 gives general
expressions for evaluating shrinkage strain. Zia et al. [20.10] recommend computing εsh
by starting with a value of 550 × 10–6 as the basic ultimate shrinkage strain, multiplying
by (1 –0.06V/S) to correct for the ratio of volume V to surface S, and then multiplying by
(1.5 – 0.015H) to correct for the relative humidity H. For post-tensioned members, an addi-
tional reduction factor is used to account for the time between the end of moist curing and
the application of prestress.
Loss Due to Relaxation of Steel Stress

Relaxation is taken to mean the loss of stress in steel under nearly constant strain at con-
stant temperature. Loss due to relaxation varies widely for different steels, and such loss
should be provided for in accordance with test data furnished by the steel manufacturers.
The percentage loss of prestress relating to relaxation varies with the type of tendon and
the ratio of initial prestress to tensile strength of tendon. This loss is generally assumed to
be in the range of 2 to 3% of the initial steel stress for stress-relieved strands and about 1%
for low-relaxation strands (see Section 1.13). Zia et al. [20.10] have provided a formula for
this computation.
860
Friction Losses in Post-tensioned Members

There will be frictional losses, which are generally small, in the jacking equipment; also
present will be friction between the tendons and the surrounding material (either duct or
actual concrete member), due to intended or unintended curvature in the tendons. The fric-
tion between tendons and surrounding material is not small and may be considered as partly
a length effect and partly a curvature effect.
Referring to Fig. 20.7.1, let dx be a segment of a curved tendon. Assume that the tendon
is being jacked from the left end by a force Ps, which results in a force Px at some distance
to the right; these forces define the limits for the tension t. The full angle enclosed within
the arc is α .
For equilibrium of the entire segment dx, refer to the force polygon of Fig. 20.7.1(b);
the normal force dN is
 dα   dt  dα
dN = t   +  t + dα (20.7.12)
 2   dα  2
and neglecting infinitesimals of higher order,
 dα 
dN = 2t   = td α (20.7.13)
 2 
The friction force developed along the length dx is
µ dN = µt d α (20.7.14)
Summation of forces along the tendon gives
t − µt d α – (t + dt ) = 0
dt
= −µ d α (20.7.15)
t
Lx dx
dα
t dt dα
µdN = µtdα t+
dα
dt dα
t t+
dα
dα
α dα dN
2
(b) Force polygon
R = radius of curvature
(a) Element of curved tendon
Figure 20.7.1 Friction losses in post-tensioned member.

861
Integrating to obtain the total effect over the entire curved portion included within the
angle α,
Px dt α
∫ Ps t
= ∫ −µ d α
0 (20.7.16)
log e Px – log e Ps = – µα (20.7.17)
Px
= e − µα (20.7.18)
Ps
Replacing the friction force term µα with the following expression, which contains a fric-
tion part due to curvature and a length effect (wobble effect), µα + KLx,
Px
= e − ( µα + KLx ) (20.7.19)
Ps
or
Ps = Px e( µα + KLx ) (20.7.20)
where α = L /R; the length L of the curve divided by R, the radius of curvature; and Lx is
the length of prestressing tendon element from jacking end to any point x along the ten-
don. Equation (20.7.20) was included in the ACI Code up to its 2008 Edition. In the 2014
ACI Building Code, however, ACI-20.3.2.6.2 specifies that “Calculated friction loss in post-
tensioning tendons shall be based on experimentally determined wobble and curvature fric-
tion coefficients”; a reference is given in Commentary ACI-R20.3.2.6.2 to the Post-Tensioning
Manual [20.21], where the same procedure used in the 2008 ACI Code is discussed.
As an approximation, when Ps – Px is small (such as not more than 15 to 20% of the
jacking force Ps), the friction force may be assumed to be constant. If the friction force is
assumed to be proportional to the force Px, then (see Fig. 20.7.2)
µN = µPx α (20.7.21)
and assuming the wobble effect KLx is also proportional to Px, equilibrium requires
Ps = Px + Px (µα + KL x )
or
Ps = Px (1 + µα + KL x ) (20.7.22)
which is permitted in the Post-Tensioning Manual when ( µα + KLx ) does not exceed 0.15.
Lx
µPxα
Ps Px
Jacking
force
α
R
Figure 20.7.2 Approximate procedure for friction losses in post-tensioned members.

862
Two-way post-tensioned slab. (Photo courtesy of Cary Kopczynski & Co.)
Losses Due to Anchorage Seating

In most post-tensioned members, the prestress force is transferred through anchorage fixtures
at the ends of the member (see Fig. 20.4.2). These anchorage devices often utilize wedge
action to anchor the tendons, although other methods, such as threaded bars with nuts or cold-
formed rivet heads, are also used. Friction wedges, such as the one shown in Fig. 20.7.3, will
slip a small amount before the tendon can be fully gripped (seated) at the end. As a result, the
initial prestress force applied by the jack will be reduced by an amount depending on the type
of anchorage device. Typical values of slip range between 0.03 and 0.1 in. [20.6]. In practice,
this loss in prestress can be compensated for by applying a higher jacking force at the time of
prestress. Note that since the slip due to anchorage seating is a fixed amount that depends on
the anchorage device, the loss of prestress will be higher for shorter tendons.
Practical Design Consideration—Total Losses

The total loss in prestress may be expressed in unit strains, total strains, or unit stresses, or
in percentage of initial prestress. Although it is difficult to generalize the amount of prestress
loss, Lin and Burns [20.6] have suggested that for average steel and concrete properties and
for average curing conditions the values in Table 20.7.1 may be taken as representative.
TABLE 20.7.1 AVERAGE PERCENTAGES OF LOSS OF PRESTRESS [20.6]
Pretensioning Post-tensioning
(Percent) (Percent)
Elastic shortening and bending 4 1
of concrete member
Creep of concrete 6 5
Shrinkage of concrete 7 6
Relaxation (creep) in steel 8 8
Totals 25 20
Notes: Zia et al. [20.10] have indicated that the upper limit for total loss in steel stress (not including friction loss) for stress-
relieved strand in normal-weight concrete may be assumed as 50,000 psi.
More detailed treatment of losses of prestress is to be found in the textbooks by Collins and Mitchell [20.2], Libby [20.3], Nawy
[20.4], Naaman [20.5], Lin and Burns [20.6], and Nilson [20.7], as well as in the PCI Committee on Prestress Losses Report
[20.11] and Zia et al. [20.10].
863
Figure 20.7.3 Typical wedge-and-chuck assembly for 7-wire strands. (Photo by José A. Pincheira.)
EXAMPLE 20.7.2
The post-tensioned beam of Fig. 20.7.4 containing a cable of 72 parallel wires, Aps = 3.60

sq in., is to be tensioned 2 wires at a time. The jacking stress is to be measured by a
pressure gauge. The wires are to be stressed from one end of the member to a value f1 to
overcome friction loss, then released to a value f2, so that an initial prestress of 144 ksi
is obtained immediately after anchoring. Compute f1 and f2, as well as the final design
stress after all losses, according to the ACI Code. Assumptions are as follows:
(a) Coefficient of friction μ = 0.50
(b) Wobble coefficient K = 0.0008 (per ft)
(c) Deformation at anchorages and slip of wires = 0.06 in.
(d) Steel relaxation = 5% of initial prestress (144 ksi)
(e) Es = 29,000 ksi; Ec = 5000 ksi
SOLUTION
(a) Loss due to friction. Using the “exact” method, Eq. (20.7.20),
Ps = Px e( µα + KLx )
 45 
µα ≈ 0.5  = 0.1108
 203.125 

KL x = 0.0008(45) = 0..0360
Ps = Px e( 0.0360 + 0.1108 ) = 1.158 Px
The above expression can also be used in terms of unit stress f ; in this case if fx is desired
to be 144 ksi, then the initial stress f1 to overcome frictional loss is

f1 = 1.158(144) = 167 ksi
(Continued)
864
Using the approximate expression, Eq. (20.7.22),

Ps = Px (1 + 0.1468) = 1.1468 Px
or
f1 = 1.1468(144) = 165 ksi
(b) Loss due to anchorage seating. For tensioning from one end,
0.06
εs = = 0.00011
45(12)
∆fs = ε s Es = 0.00011(29, 000) = 3.2 ksi
To allow for anchorage seating, tension to f1 = 167 ksi to overcome friction, then release
to f2 = 144 + 3.2 = 147.2 ksi. The minimum stress fsi = 144 ksi will then exist at both ends.
(c) Elastic shortening due to post-tensioning two wires at a time. The wires tensioned
first will have the greatest loss, having some additional loss as each succeeding pair
of wires is tensioned. The pair tensioned last will have zero loss. Thus Δ fs in the first
pair, given by Eq. (20.7.5), is
nT0
∆fs =
Ag
29, 000
n= = 5.8, say 6 to nearest whole number
5000
35
T0 = (3.60)(144) = 504 kips
36

6(504)
∆fs (first pair ) = = 4.66 ksi
18(36)
∆fs (last pair) = 0 ksi
average ∆fs = 2.33 ksi (1.6% of 144 ksi)
(d) Creep loss. Using the procedure recommended by Zia et al. [20.10], fc in Eq.
(20.7.10) is computed as ( fcir – fcds). The net concrete stress at midspan at the cen-
troid of the steel tendons is
T0 T0 e2 M D e
fcir = + −
Ag Ig Ig
where
T0 = 3.60(144) = 518.4 kips

1 18(36)
MD = (0.15)(45)2 = 170.9 ft-kips
8 144
Then,
518.4 518.4(12)12 170.9(12)12

fcir = + −
18(36) 18(36)3 /12 18(36)3 /112
= 0.80 + 1.07 − 0.35 = 1.52 ksi (at midspan)
(Continued)
865
In this case, if there is no additional superimposed permanent dead load, fcds is zero.
At the support, where there is zero eccentricity of the tendons and zero dead load
moment,
T0
fcir = = 0.80 ksi
Ag
The average value of ( fcir – fcds) should be used as fc in Eq. (20.7.10) for post-
tensioned beams:
1.52 + 0.80
fc = = 1.16 ksi
2
Note again that in this example fcds is taken as zero.
fc 1.16
ε c = elastic strain = = = 0.000232
Ec 5000
and for a post-tensioned beam using Ct = 1.6,
ε cp = Ct ε c = 1.6(0.000232) = 0.000371

∆fs = ε cp Es = 0.000371(29, 000) = 10.8 ksi (7.5%)
(e) Shrinkage loss. Using the procedure of Zia et al. [20.10], the basic shrinkage
strain is

basic ε sh = 550 × 10 –6
The correction factor (CF) for volume /surface (V/S) ratio is
V
(CF)V / S = 1 − 0.06 = 1 − 0.06(6) = 0.64
S
where
V 18(36)
= = 6
S 2(18) + 2(36)
The correction factor (CF) for humidity H is

(CF )h = 1.5 − 0.015 H = 1.5 − 0.015(70) = 0.45
for 70% relative humidity. The adjusted shrinkage strain then becomes
ε sh = (basic ε sh )(CF )V / S (CF )h = (550 × 10 −6 )(0.64)(0.45)

= 0.000158
∆fs = ε sh Es = 0.000158(29, 000) = 4.59 ksi (3.2%)
(f) Relaxation in steel.

∆fs = 0.05(144) = 7.2 ksi (5.0%)
(Continued)
86
(g) Total losses.
Loss of Stress
ksi Percent
Elastic shortening 2.3 1.6
Creep in concrete 10.8 7.5
Shrinkage 4.6 3.2
Relaxation in steel 7.2 5.0
24.9 ksi 17.3%
The final design prestress under dead load plus live load, after losses, is

fse = 144 – 24.9 = 119.1 ksi
R = 203’– 112 ” Symbolic

Assume cable drape about 1’– 6”
0.1108 radians is circular curve midspan
CG
1’– 6” 3’– 0”
1’– 6”
45’– 0” c–c of supports e = 12”
midspan
section
Figure 20.7.4 Post-tensioned beam of Example 20.7.2.
20.8 NOMINAL STRENGTH M n OF FLEXURAL

MEMBERS
The nominal strength Mn of a prestressed concrete flexural member is computed in a man-
ner similar to that discussed in Section 3.4, using the six assumptions stated there. Unlike
nonprestressed reinforced concrete members, where nominal strength is the primary crite-
rion determining whether a design is or is not satisfactory, a prestressed concrete member
must have adequate strength; however, this is usually not the governing factor in a design.
Design must include consideration of all significant load stages. Primarily, these stages
are (1) initial stage, including the period before and during prestressing, as well as the
transfer of prestress to the concrete; (2) intermediate stage during transportation and erec-
tion; (3) final stage under service load, after losses; and (4) overload stage, where cracking
and nominal strength (so-called ultimate strength) are important. Though the actual num-
ber of conditions to be investigated varies with the situation, ordinarily the initial service
condition involves beam dead load plus prestress before losses; the final service condition
involves full dead load plus live load after losses; and the nominal strength must be ade-
quate for possible overload.
The initial and final service load stages were considered in Section 20.5. A beam may
be properly prestressed to carry service loads with little deflection and generally without
cracking but, for example, because of a small moment arm to the centroid of the steel, its
nominal moment strength Mn may not give an adequate margin of safety. On the other hand,
if only the strength Mn were considered, service loads might cause excessive cracking, cam-
ber (upward deflection), or deflection.
867
20.8 NOMINAL STRENGTH Mn OF FLEXURAL MEMBERS 867
Most of the following development applies for members containing only prestressed
reinforcement (so-called fully prestressed members); thus, any nonprestressed steel effect
is omitted. Also, the discussion is limited to pretensioned and post-tensioned construction
with grouted tendons. The flexural strength of post-tensioned members in which tendons
are ungrouted is, in general, less than that of a beam with bonded tendons.
The nominal moment strength may be determined for rectangular sections, in accord-
ance with the principles of Chapter 3 (Section 3.4), as follows:
 a
Mn = T  d p −  (20.8.1)
 2
where T = Aps fps, C = 0.85 fc′ ba , and from C = T,
Aps f ps ρ p bdd f ps ρ p f ps
a= = = dp (20.8.2)
0.85 fc′ b 0.85 fc′ b 0.85 fc′
with Aps, ρp, and fps referring to the area, reinforcement ratio, and tensile stress for the pre-
stressed reinforcement, respectively, and dp to the effective depth measured to the centroid
of the prestressing steel. Note that fps is the average stress in the prestressing steel when
nominal moment strength Mn is reached; it is used instead of fy because the steel usually
exhibits no well-defined yield point (see Fig. 20.8.1).
 ρ p f ps 
M n = Aps f ps  d p − dp (20.8.3)
 1.7 fc′ 
Thus for any concrete cross section, the nominal moment strength depends on ρp and fps.
The actual stress in the prestressed reinforcement when the nominal strength is reached
may not be easily determined, particularly when the specific stress-strain curve for the steel
used is not available. Referring to Fig. 20.8.1, for low percentages of steel and therefore
higher stress fps, the strain may be nearly 0.05, whereas for higher percentages of reinforce-
ment, the strain may be closer to 0.01 (approximately corresponding to yield stress). Thus
fps is not the same for all beams.
In the ACI Building Code, determination of tensile stress in bonded prestressing steel
is addressed in ACI-20.3.2.3.1. For sections in which all prestressed reinforcement is in
the member tension zone and the effective prestress fse (stress in the prestressing steel after
250 1724
1500
200
Unit stress, MPa
Unit stress, ksi
fpu = 250 ksi 1000
100
500
0
1 2 3 4 5
Unit strain, %
Figure 20.8.1 Typical stress–strain curve for 250-ksi steel (stress-relieved) wire.

86
losses) is at least 0.5fpu, the stress fps may be determined as follows, omitting the terms for
nonprestressed reinforcement.
 γ p ρ p f pu 
f ps = f pu  1 − (20.8.4)
 β1 fc′ 
Note the values of γp /β1 in Table 20.8.1.

For members with unbonded tendons, fse ≥ 0.5fpu, and with a span-to-depth ratio of 35 or
less, ACI-20.3.2.4.1 gives
fc′
f ps = fse + 10, 000 + (20.8.5)
100ρ p
but not greater than the lesser of fpy and (fse + 60,000).
For members with unbonded tendons, fse ≥ 0.5fpu, and with a span-to-depth ratio greater
than 35, such as one-way slabs, flat plates, and flat slabs,
fc′
f ps = fse + 10, 000 + (20.8.6)
300ρ p
but not greater than the lesser of fpy and (fse + 30,000).
Harajli and Kanj [20.16] provide a study on the ultimate flexural strength of members
having unbonded tendons.
Balanced Strain Condition, Tension and Compression Control Limits,

Strength Reduction Factors
As first described in Section 3.5, the balanced strain condition (or compression control
limit for the case of one layer of tension steel) occurs when the concrete strain εc at the
extreme compression fiber is 0.003 at the instant the steel reaches its yield strain εy = fy /Es.
In prestressed concrete members, however, the steel used does not exhibit the well-defined
yield point that occurs with ordinary deformed bars, and so the concept of balanced failure
is nebulous. Further, the steel and concrete are strained prior to the application of external
load (i.e., prestrained), and so the calculation of the total strain in the prestressing steel
is not as straightforward as the determination of reinforcement strains in nonprestressed
members.
In the ACI Code, the same measure of strain in the extreme layer of tension reinforce-
ment is used in nonprestressed and prestressed reinforcement for the purpose of determin-
ing whether a section is tension controlled, compression controlled, or in the transition
region. ACI-21.2.2.1 and 21.2.2.2 permit εy = εt = 0.002 for Grade 60 reinforcement and
TABLE 20.8.1 VALUES FOR γp /β1 FOR EQ. (20.8.4)
Concrete Strength, fc′ (psi)
γp 5000 6000 7000 8000
0.55 0.69 0.73 0.79 0.85

(for fpy /fpu ≥ 0.80)
0.40 0.50 0.53 0.57 0.62
0.28 0.35 0.37 0.40 0.43
869
20.8 NOMINAL STRENGTH Mn OF FLEXURAL MEMBERS 869
for all prestressed reinforcement, respectively. For prestressed reinforcement, the strain
εt is not the actual strain in the steel; rather, it is the strain excluding those “due to pre-
stress, creep, shrinkage, and temperature” (ACI-R21.2.2). Thus, for all sections, nonpre-
stressed reinforced, prestressed reinforced, or any combination thereof, and for all shapes
of cross section, and no matter where the steel is located, including compression steel,
when the extreme tensile strain εt is greater than 0.005, the section is “tension controlled”
and φ = 0.90; and when the extreme tensile strain εt is less than 0.002 (for Grade 60 and
prestressed reinforcement), the section is “compression controlled,” and φ will be either
0.75 or 0.65 as required for columns (see Chapter 10). The strain limits were shown earlier
(see Fig. 3.6.2).
When the extreme tensile strain εt is between 0.002 and 0.005, the φ factor is required
to be obtained by linear interpolation, as shown in Fig. 3.6.2. Equations (3.6.3) and (3.6.4)
give φ in terms of the extreme tensile strain εt, and Eqs. (3.6.5) and (3.6.6) give φ in terms
of c /dt, where c is the location of the neutral axis measured from the compression face of
the section and dt is the distance from the extreme compression fiber to the extreme layer
of tension steel.
EXAMPLE 20.8.1
Determine the nominal moment strength Mn of the pretensioned bonded section investi
gated in Example 20.5.2, and evaluate whether the strength is adequate for the given ser
vice loads. The concrete has fc′ = 5000 psi, and the stress-relieved prestressing strands
have fpu = 250,000 psi. Compare the stress in the prestressing steel at nominal condition
calculated according to ACI-20.3.2.3.1 with that obtained using an average stress-strain
relationship for the steel, as given in Fig. 20.8.1, and assuming 20% prestress losses. Use
Es = 29,000 ksi and Ec = 4000 ksi.
SOLUTION
(a) The stress in the prestressing steel when nominal strength is reached may be taken,
according to ACI-20.3.2.3.1, or Eq. (20.8.4), as
 γp f pu 
f ps = f pu  1 − ρ p
 β1 fc′ 

Aps 2.30
ρp = = = 0.0046
bd p 20(25)
For fc′ = 5000 psi, β1 = 0.80; and for stress-relieved strands, fpy /fpu ≥ 0.85 and γp = 0.40.
Thus,
 0.40 250 
f ps = 250 1 − (0.0046) = 250(1 − 0.115) = 221 ksi
 0.80 5 
From Fig. 20.8.2(c),
Cu = 0.85 fc′ ba = 0.85(5)(20)a = 85a

Tu = Aps f ps = 2.30(221) = 508 kips
Cu = Tu

508
a= = 5.98 in.
85
a 5.98
c= = = 7.47 in.
β1 0.8
(Continued)
870
The additional strain εs2 at the centroid of prestressing steel that is due to the super-
imposed nominal moment Mn is
fse Aps  1 e2  dp − c
εs2 =  +  + ε cu
Ec  Ag I g  c
0.8(175, 000)(2.3) 1  1 10 2  17.53

ε s2 =  +  + 0.003 = 0.00735
4000 1000  600 45, 000  7.47
The first term represents the compressive strain in the concrete at the level of the
prestressing steel under the effective prestress. The second term, on the other hand, is
equal to the increase in tensile strain relative to the strain in the prestressing steel at the
time the concrete at that level has zero strain.
The total strain in the prestressing steel is then the strain due to prestress after losses,
εs1, plus the additional strain due to superimposed nominal moment, Mn.
fsi (initial) = 175, 000 psi
∆fs (losses) = 0.2(175, 000) = 35, 000 psii
fse = 175, 000 − 35, 000 = 140, 000 psi

140
ε s = ε s1 + ε s 2 = + 0.00735 = 0.0122
29, 000
Using a typical stress-strain relationship for a steel with fpu = 250 ksi, such as in
Fig. 20.8.1, the strain εs of 0.0122 corresponds approximately to that for a stress fps of
225 ksi. The value of fps agrees closely with the starting assumption of 221 ksi based on
Eq. (20.8.4) in this case.
Thus, the nominal moment strength is
 a   5.98   1
M n = Tu  d p −  = 508  25 −  = 932 ft-kips
 2   2   12
(b) The factored moment Mu based on the service loads of Example 20.5.2 is

Mu = 1.2(125) + 1.6(410) = 806 ft-kips
Evaluate the strength reduction factor φ to be used. The neutral axis distance c for
the section is

c = 7.47 in.
This section has h = 30 in. and dp = 25 in. for, say, two layers of strands. The distance
dt from the extreme compression fiber to the extreme tension steel will be larger than dp,
say dt = 27 in. The strain εt at the extreme tension steel is
dt − c 27 − 7.47
ε i = 0.003 = 0.003 = 0.0078
c 7.47
Since the strain εt exceeds 0.005, the section is “tension controlled” and φ = 0.90.
Had the strain εt been less than 0.002, the section would be “compression controlled”
and φ would have been either 0.65 or 0.75, depending on whether the section was tied
or spirally reinforced (typically tied for the case of beams). If εt is between 0.005 and
0.002, Eq. (3.6.3) or (3.6.4) would be used for the linear interpolation between tension-
controlled and compression-controlled sections.
(Continued)
871
20.9 CRACKING MOMENT 871
As mentioned earlier, in checking the strain εt in the extreme layer of steel in ten-
sion against the 0.005 limit, only the strain change relative to the steel strain at the time
the concrete at the level of the prestressing steel has zero strain is considered. Since
the strain in the concrete due to prestress at the level of the prestressing steel is very
small compared to the increase in strain due to externally applied loads, the strain εt is
an approximation of the strains induced after loading. The intent is thus to provide the
beam with sufficient ductility once the external loads have been applied.
(c) Compare φ Mn with Mu.

[φ M n = 0.90(932) = 839 ft-kips] > [Mu = 806 ft-kips] OK
20” 0.003 0.85fc’
a = β1c Cu = 0.85fc’ ba
c
dp = 25”
dp – a
30” 2
Aps = 2.30 sq in.

Tu = Aps fps
(a) Section (b) Strain when Mn is reached (c) Internal forces
Figure 20.8.2 Strain and stress distributions, and internal force resultants at nominal
strength Mn of section for Example 20.8.1.
20.9 CRACKING MOMENT
One of the features of prestressed concrete is that under service load it is usually crack
free. To be sure that an adequate reserve exists against cracking, ACI-9.6.2.1 requires that
for beams, the total amount of prestressed and nonprestressed reinforcement be adequate
to develop a factored load in flexure at least 1.2 times the cracking load calculated using a
modulus of rupture equal to 7.5λ fc′. The same concept applies for one-way slabs (ACI-
7.6.2.1) and two-way slabs (ACI-8.6.2.2). The basic cracking moment requirement is a
means of ensuring that cracking will not lead to failure and that a margin large enough
exists between flexural and cracking strength so that significant deflection will occur to
warn that the strength Mn is being approached. The typical member will have a fairly large
margin between cracking strength and flexural strength, but the designer must be certain
by checking it.
The ACI Code (ACI-9.6.2.2 for beams, ACI-7.6.2.2 for one-way slabs, and ACI-8.6.2.2.1
for two-way slabs) waives the above requirement when φVn and φ Mn is at least twice Vu
and Mu, respectively.
EXAMPLE 20.9.1
Compute the cracking moment Mcr and check its acceptability according to the ACI
Code, for the beam of Example 20.8.1 (Fig. 20.8.2). Use fc′ = 5000 psi (normal-weight
concrete) and assume that the effective prestress, after losses, is 140 ksi.
(Continued)
872
SOLUTION
(a) Use the homogeneous beam concept (Section 20.6). The effective prestress force is

Te = Aps fse = 2.30(140) = 322 kips
Using the following equation to find the cracking moment:
 axial   moment   external  cracking

− − + =
 prestress  prestress  loads   stress 
Te Te eyt M cr yt
− − + = fr = 7.5 5000 = 530 psi
Ag Ig Ig
322 322(10)15 M cr (15)

− − + = 0.530
600 45, 000 45, 000
 0.530(45, 000) 322(45, 000)  1

M cr =  + 322(10) +
 15 600(15)  12

= 133 + 268 + 134 = 535 ft-kips
(b) Use the internal force concept (Section 20.6). Consider that the cracking moment
Mcr has two parts,

M cr = M1 + M 2
where M1 is the superimposed moment necessary to give zero stress in the precom-
pressed tension zone (see Fig. 20.9.1), and M2 is the additional moment to cause cracking,
assuming that zero stress exists at the tension face when M2 is applied (see Fig. 20.9.1).
C = Te = 322 kips
 10 + 5 
M1 = Te (e + ec ) = 322  = 402.5 ft-kips
 12 
fr I g 0.530(45, 000)
M2 = = = 132.5 ft-kips
yt 15(12)
M cr = M1 + M 2 = 403 + 133 = 535 ft-kips
(c) Check ACI-9.6.2.1. From Example 20.8.1, the nominal moment strength Mn pro-
vided is 932 ft-kips. Since ACI-9.6.2.1 requires reinforcement “to develop a factored
load at least” 1.2 Mcr,
φ M n ≥ 1.2 M cr
0.90(932) > 1.2(535)

839 ft-kips > 642 ft-kips OK
Thus cracking will occur soon enough before reaching nominal moment strength so that
large deflection will give a warning of impending failure.
(Continued)
873
20.10 SHEAR STRENGTH 873
fr
Centroid
10” (–) (–) (–)
of beam C
cross section ec = 5” + =
30”
e = 10”
Te (+)
0 fr fr
Midspan M1 + M2 = Mcr
Figure 20.9.1 Computation of cracking moment.
20.10 SHEAR STRENGTH OF MEMBERS WITHOUT

SHEAR REINFORCEMENT
In general, the ideas presented in Chapter 5 regarding shear strength for nonprestressed
beams are also applicable to prestressed beams. An excellent summary of background for
the ACI Code expressions for shear strength of prestressed concrete beams is given by
MacGregor and Hanson [5.63]. More information on shear strength of prestressed concrete
members is available in the ASCE–ACI Committee 426 Reports [5.4, 5.16]. Collins and
Mitchell [18.9] have presented an excellent unified treatment of shear and torsion behavior
for both prestressed and nonprestressed concrete beams, along with proposed design rules.
Use of the strut-and-tie model for prestressed concrete members has been presented by
Alshegeir and Ramirez [20.18] and by Ramirez [20.20].
It is known, however, that a prestressed concrete beam generally performs better under
high shear conditions than does an ordinary reinforced concrete beam. Consider Eq. (5.3.1)
as derived in Chapter 5 for maximum principal stress,
2
ft f 
ft (max) = +  t  + v2 [5.3.1]
2  2
+v
(+ft’, +v)
v
v v 2α
α
ft α ft +ft
α
ft (max)
v
(0, –v)
(a) Principal stress–reinforced concrete beam
+v
v (–ft’, +v)
2α
v α +ft
–ft –ft
v α
α
v
ft (max) (0, –v)
(b) Principal stress–prestressed concrete beam
Figure 20.10.1 Comparison of directions of principal tensile stress: (a) reinforced and

(b) prestressed concrete beams.
874
In nonprestressed concrete, the normal stress ft is a tensile stress in the tension zone of the
beam. When the concrete is prestressed, ft is a compressive stress throughout the member
depth, or at least over most of it. On replacing ft with – ft, it is apparent that the magnitude
of the principal tensile stress decreases,
2
−f f 
ft (max) = t +  t  + v 2 (20.10.1)
2  2
The angle α that the principal tensile stress makes with the beam axis is greater for
a prestressed concrete beam than for an ordinary reinforced concrete beam, as shown in
Fig. 20.10.1. Inclined cracking will be less likely to occur and, if it occurs, will be at a lower
angle with respect to the longitudinal axis than in nonprestressed members.
Inclined cracks of two types are possible in prestressed concrete beams: (1) the flexure-
shear crack that occurs in a beam previously cracked owing to flexure and (2) the web-
shear crack that occurs in the thin web of a previously uncracked beam [see Fig. 5.4.1(a)].
Whereas only the flexure-shear crack is common in nonprestressed beams, both types rep-
resent potential cracks in prestressed concrete beams.
Flexure-Shear Cracking Strength

The flexure-shear crack arises from high principal tensile stress near the interior extrem-
ity of a flexural crack. Experimental studies have shown that the shear corresponding
to the flexure-shear crack is that which causes flexural cracking at approximately a dis-
tance dp /2 from the load point, in the direction of decreasing bending moment [20.13]
(see Fig. 20.10.2). The 1963 ACI Code formula for flexure-shear cracking strength was
the experimentally determined linear relationship of Fig. 20.10.3. Because the relationship
between moment and shear (M/V) was not the same for both dead and live loads in the tests,
the effects of dead load were not included in the linear relationship.
The ordinate of the plot (see Fig. 20.10.3) is thus in terms of Vci – Vd, where Vd is the ser
vice dead load shear force and Vci is the total nominal shear strength. The abscissa involves
the net cracking moment Mcr. The net cracking stress used for computing Mcr equals the
modulus of rupture (assumed conservatively at 6 λ fc′ psi), plus the compressive stress fpe
provided by the prestressing force (after losses) occurring at the extreme fiber at which
tensile stresses are caused by the applied loads, less the tensile stress fd due to service dead
load. Thus
I
M cr = (6 λ fc′ + f pe − fd ) (20.10.2)*
yt
which is ACI Formula (22.5.8.3.1c).

The abscissa in Fig. 20.10.3 involves Mcr /(M/V – dp /2), which represents the shear corre-
sponding to flexural cracking developing at dp /2 from the cross section being investigated.
Thus the linear relationship for the flexure-shear cracking strength resulting from high
principal tensile stress in the vicinity of a flexural crack is, according to Fig. 20.10.3,
Vct − Vd M cr
= 0.6 + (20.10.3)
b w d p fc′ ( M /V − d p / 2) bw d p fc′
Since 1971, this expression has been simplified by eliminating the subtracted dp /2 term.
* For SI, ACI 318-14M, for fc′ , fpe, and fd in MPa, Mcr in N∙mm, I in mm4, and yt in mm, Eg. (20.10.2) may be rewritten as
I 1  (20.10.2)
M cr =  λ fc′ + f pe − fd 
yt 2
875
Thus the nominal flexure-shear cracking strength Vci, (ACI-22.5.8.3.1), is
Vi M cr
Vci = 0.6 λ fc′bw d p + + Vd ≥ 1.7λ fc′bw d p (20.10.4)*
M max
In Eq. (20.10.4), Mmax replaces M of Eq. (20.10.3) and represents the maximum moment
that can occur at the section under consideration, due to externally applied factored loads
(i.e., applied loads other than beam weight and prestress, unless the prestress causes an
external reaction). Vi replaces V of Eq. (20.10.3) and represents the shear force at the sec-
tion due to the factored loading that caused maximum moment. In other words, one uses
the moment envelope values for Mmax along with the corresponding shears, rather than
the shear envelope values which would be larger. Of course, where partial span loadings
are not considered, the full-span loading gives Vi and Mmax for each point along the span.
When full-span uniform loading is used in a simply supported span, Vi = w(L – 2x)/2 and
Mmax = wx(L – x)/2, and
Vi L − 2x
= (20.10.5)
M max x( L − x )
In effect, Eq. (20.10.4) gives the flexure-shear cracking strength as the sum of (1) the
shear due to actual dead load (beam weight) without overload factor, (2) the shear due
to superimposed factored load that is sufficient to cause flexural cracking, and (3) the
additional shear ( 0.6λ fc′ bw d p) that will cause the flexural crack to initiate the inclined
flexure-shear crack.
Web-Shear Cracking Strength

The web-shear crack arises in a beam previously uncracked due to flexure. Such a crack is
typical near the support of a thin-webbed section (see Fig. 5.4.1) as a result of high princi-
pal tensile stress.
For this development the beam may reasonably be considered as homogeneous. Thus
Eq. (20.10.1) may be directly applied,
2
− ft f 
ft (max) = +  t  + v2 [20.10.1]
2  2
Decreasing moment
Flexure-shear
crack
Initiating
flexural
crack
dp A
2
Figure 20.10.2 Flexure-shear type of inclined crack.
* For SI, ACI 318-14M, for fc′ in MPa, bw and dp in mm, and V in N, gives
1 VM 1
Vci = λ fc′ bw d p + i cr + Vd ≥ λ fc′ bw d p [20.10.4]
20 M max 7
876
Vci – Vd 5
bwdp fc’ 4
0
0 1 2 3 4 5 6
Mcr
( M
V
–
dp
2
) bwdp fc’
Figure 20.10.3 Comparison of the shear corresponding to flexure-shear cracking in

prestressed concrete beams with the ratio of the flexural cracking moment to the shear
span (both axes nondimensionalized by dividing by bw dp fc′ ). (From Sozen and Hawkins
[20.13].)
Since tests have demonstrated that cracks of this type usually originate near the centroid of
the section, the stresses refer to that location. Solving Eq. (20.10.1) for v,
2 2
 ft   ft 
 ft (max) + 2  =  2  + v
2
 
2 2
f  f 
ft (max) + ft (max) ft +  t  =  t  + v 2
2
 2  2
ft
v = ft (max) 1 + (20.10.6)
ft (max)
where ft is the compressive stress at the level of the centroid, and ft(max) is the principal
tensile stress, which should be less than the tensile strength of concrete if no cracks are
to form.
Since Eq. (20.10.6) should agree with the criteria for shear strength of nonprestressed
beams, ft (max) is taken as 3.5λ fc′. The equation then becomes
ft
v = 3.5λ fc′ 1 + (20.10.7)
3.5λ fc′
A plot of Eq. (20.10.7) for λ = 1.0 is given in Fig. 20.10.4, illustrating that this equation
may be approximated by a straight line,
v = 3.5λ fc′ + 0.3 ft (20.10.8)

87
10
v = 3.5√ fc’ ft
1+ [Eq. (20.10.7)]
8 3.5√fc’
6
v
fc’ v = 3.5√fc’ + 0.3ft [Eq. (20.10.8)]
4
0 2 4 6 8 10 12 14 16 18 20
ft
fc’
Figure 20.10.4 Comparison of theoretical maximum shear stress with straight-line approximation.
or, in ACI Code terminology, multiplying by bw dp to give nominal shear strength,
Vcw = (3.5λ fc′ + 0.3 f pc )bw d p (20.10.9)
where fpc is defined as the compressive stress (psi) in the concrete, after losses, at the cen-
troid of the section resisting the applied loads, or if the centroid lies within the flange on a
T-section, fpc is the stress at the junction of the flange and web.
If prestressing tendons are draped, a vertical component Vp arises which assists in carry-
ing the shear. Therefore, ACI-22.5.8.3.2 gives
Vcw = (3.5λ fc′ + 0.3 f pc )bw d p + Vp (20.10.10)*
Alternatively, the web-shear cracking strength may be determined (ACI-22.5.8.3.3) via

the principal stress equation, Eq. (20.10.6), where the principal stress ft (max) is limited to
4 λ fc′. This would give Eq. (20.10.7) with coefficients 4 instead of 3.5. Thus using fpc for
ft, Eq. (20.10.7) multiplied by bwdp with Vp added becomes
f pc
Vcw = 4 λ fc′ 1 + bw d p + Vp (20.10.11)*
4 λ fc′
which is an acceptable alternate to using Eq. (20.10.10).

The nominal shear strength Vn at which inclined cracking is imminent is given by the
lesser of Vci and Vcw. When applying Eqs. (20.10.4) and (20.10.10) or (20.10.11) for Vci and
Vcw, the effective depth dp is to be taken as the distance from the extreme compression fiber
to the centroid of the prestressing tendons, or as 80% of the overall depth of the member,
whichever is greater.
In computing the prestress effect fpc, full prestress after losses may be used only when
the prestressing tendons are embedded a distance that exceeds their required transfer length
from the section being investigated. The critical section for maximum shear is generally at
* For SI, ACI 318-14M, for fc′ and fpc in MPa, bw and dp in mm, and V in N, gives
Vcw = (0.29λ fc′ + 0.3 f pc )bw d p + Vp (20.10.10)
1 3 f pc
Vcw = λ fc′ 1 + bw d p + Vp (20.10.11)
3 λ fc′
87
h/2 from the face of support (ACI-9.4.3.2), so that the region between the face of support
and h/2 therefrom must be designed for the shear at the critical section.
ACI Code Simplified Alternative for Shear Strength

When the member has an effective prestress fse at least equal to 40% of the tensile
strength fpu of the flexural reinforcement, the nominal shear strength may be taken as
(ACI-22.5.8.2)
 Vu d p 
Vc = 0.6 λ fc′ + 700  bw d p ≤ 5λ fc′bw d p (20.10.12)*
 Mu 
but need not be taken less than 2 λ fc′bw d p . Supporting data for this alternate relationship
are shown in Fig. 20.10.5 from MacGregor and Hanson [5.63].
When applying Eq. (20.10.12) from ACI-22.5.8.2, Vu is the maximum shear due to fac-
tored loading at the section and Mu is the simultaneously occurring moment; Vudp/Mu is also
limited to a maximum value of 1.0; and dp is the actual effective depth to the centroid of
prestressed reinforcement. Note that d in Eq. (20.10.12) is the larger of dp or 0.8h. Equation
(20.10.12) represents essentially the flexure-shear cracking strength, expressed in a manner
similar to that for nonprestressed concrete. Since the percentage of reinforcement is low in
prestressed concrete, a constant has been used instead of the variable ρ.
14
fse
= ≥ 0.4
12 fpu
fse
= < 0.4
fpu
10
8
vc
fc’
6
2 Vdp
vc = 0.6 fc’ + 700
M
0 2 4 6 8 10 12
1000 Vdp
M f’
c
Figure 20.10.5 Alternate equation for computing vc for prestressed beams (l = 1.0). (From
MacGregor and Hanson [5.63].)
* For SI, ACI 318-14M, for fc′ in MPa, bw and dp in mm, gives

1 Vu d p 
Vc =  λ fc′ + 4.8  b d ≤ 0.42λ fc′ bw d p (20.10.12)
 20 Mu  w p
879
EXAMPLE 20.10.1
Determine the nominal shear strength Vn = Vc at a section 4 ft from the supports of the
40-ft simple span shown in Fig. 20.10.6. The effective prestress force Te (after losses)
is 322 kips. The beam is to support a service live load of 1.50 kips/ft in addition to the
beam weight of 0.675 kip/ft. The concrete has normal weight with fc′ = 5000 psi.
SOLUTION
(a) Simplified alternate procedure (from ACI-22.5.8.2),
Vu d p
Vc = [0.6 λ fc′ + 700 ]bw d ≤ 5λ fc′bw d (20.10.12)
Mu
Aps = 2.30 sq in. 15”
12 1 30”
5”
4’– 0”
15” 20”
10’– 0” 20’– 0” 10’– 0”
Figure 20.10.6 Beam, for Example 20.10.1.
In the application of the above formula, the symbols Vu and Mu indicate use of factored
shear and moment. Assume partial span loading is not considered,
wu = 0.675(1.2) + 1.50(1.6) = 0.81 + 2.40 = 3.21 kips/ft
Vu = 3.21(20 − 4) = 51.4 kips
1
Mu = (3.21)(4)(36) = 231 ft-kips
2
Vu d p 51.4(30 − 11)
= = 0.35 < 1.0
Mu 231(12)
where dp in this term is the actual dp = 19 in. at 4 ft from the support.

The nominal unit shear stress capacity is

vc = 0.6(1.0) 5000 + 700(0.35) = 42 + 245 = 287 psi

vc = (upper limit ) = 5(1.0) fc′ = 354 psi > 287 psi OK

vc = (lower limit ) = 2(1.0) fc′ = 141 psi < 287 psi OK
Thus at this section the prestressed member has a contribution to the shear strength
attributable to the concrete, vc, of 287 psi. The nominal shear strength Vc at 4 ft from the
support would be

Vc = vc bw d = 0.287(20)(24) = 138 kips
where d is taken as the larger of 0.8h [i.e., 0.8(30) = 24 in.] and the actual dp (i.e., 19 in.).
(b) Flexure-shear cracking strength using the more elaborate procedure (ACI-22.5.8.3.1),
Vi M cr
Vci = 0.6 λ fc′bw d p + + Vd ≥ 1.7λ fc′bw d p [20.10.4]
M max
(Continued)
80
Using Eq. (20.10.2),

I
M cr = (6 λ fc′ + f pe − fd )
yt
1
I= (20)(30)3 = 45, 000 in.4 (neglectiing steel)
12
yt = 15 in.
f pe = compressive stress due to prestressingn force
322, 000 322, 000(4)(15)
= + = 537 + 429 = 966 psi
20(30) 45, 000
e = 4 in. at 4 ft from center of support
My 675(4)(36)(12)(15)
fd = dead load stress = = = 194 psi
I 2(45, 000)
45, 000
M cr = [6(1) 5000 + 966 − 194]
15(12, 000)
= 0.25(424 + 966 − 194) = 0.25(1196) = 299 ft-kips
M max = maximum moment due to externally applied factored loads,
except beam weight at section being investiigated
1
= (1.50)(4)(36)(1.6) = 173 ft-kips
2
Vi = corresponding shear due to factored load at section investigated
= 1.50(20 − 4)(1.6) =38.4 kips
Vi M cr 38.4(299)
= = 66.0 kips
M max 173
Vd = dead load shear = 0.675(20 − 4) = 10.8 kips
d p = 30 − 11 = 19 in. or 0.80 h = 0.80(30) = 24 in. (use larger value)
1
Vci = 0..6(1.0) 5000 (20)(24) + 66.0 + 10.8 = 97.2 kips
1000
The flexure-shhear cracking strength is not to be taken less than

min Vcii = 1.7(1.0) fc′bw d p
1 OK
min Vci = 1.7(1.0) 5000 (20)(24) = 57.7 kipps < 97.2 kip
1000
(c) Web-shear cracking strength using the more elaborate procedure (ACI-22.5.8.3.2),
( )
Vcw = 3.5λ fc′ + 0.3 f pc bw d p + Vp [20.10.10]
fpc = compressive stress in concrete at centroid of section resisting live load, due

to all applied loads (zero due to all except Te /Ac in this case)
322, 000
= = 537 psi
20(30)

d p = 19 in. or 0.80(30) = 24 in.. (use larger value)
322
Vcw = 3.5(1.0) 5000 + 0.3(537) (20)(24)
1
+ = 223 kips
1000 12
(Continued)
81
20.11 SHEAR REINFORCEMENT 881
Since Vci < Vcw, Vc = Vci = 97.2 kips at 4 ft from the support for the member without
shear reinforcement. The simplified alternate equation gives a nominal shear strength
42% higher than obtained from the more elaborate procedure: 138 kips compared to
97.2 kips.
20.11 SHEAR REINFORCEMENT FOR PRESTRESSED

CONCRETE BEAMS
The computations for shear reinforcement are essentially the same as for nonprestressed
concrete as developed in Chapter 5. The total nominal shear strength Vn may be represented
as the sum of the amount Vc attributable to the concrete and the amount Vs attributable to
the shear reinforcement; thus Eq. (5.8.1) still applies,
Vn = Vc + Vs [5.8.1]
The nominal strength Vc attributable to the concrete may be determined by (1) Eq. (20.10.12)
when the effective prestress is at least 40% of the tensile strength of the steel, or (2) the
smaller of Eqs. (20.10.4) and (20.10.10) for Vci and Vcw, respectively. The second and more
accurate method using Vci and Vcw may be used whatever the magnitude of prestress.
For the shear reinforcement contribution, Eq. (5.10.7) is used,
Av f yt d
Vs = [5.10.7]
s
where
A v = effective area of shear reinforcement at any section
s = spacing of the shear reinforcement
Just as for nonprestressed reinforced concrete beams, prestressed concrete beams must
have at least a minimum amount of shear reinforcement whenever Vu > φVc /2 (ACI-9.6.3.1).
However, this requirement may be waived if tests are made showing that the required flex-
ural and shear strengths can be developed when shear reinforcement is omitted. Since
prestressed concrete members are usually of standardized shapes with many identical or
similar members manufactured, manufacturers frequently have made tests demonstrating
that no shear reinforcement is required.
When minimum shear reinforcement is required, ACI-9.6.3.3 requires using the lesser
of the following items:
1. ACI Table 9.6.3.3 rows (a) and (b):
bw s 50bw s
min Av = 0.75 fc′ ≥ [5.10.8]
f yt f yt
which is also used for nonprestressd concrete, and

82
2. where the effective prestress force equals at least 40% of the tensile strength of the
steel based on fpu, ACI Table 9.6.3.3 row (e), as follows:
Aps  f pu   s  d
min Av =    (20.11.1)
80  f yt   d  bw
where
Aps = area of prestressed reinforcement
fpu = tensile strength of prestressed reinforcement
fyt = specified yield strength of shear reinforcement
s = stirrup spacing, which may not exceed 0.75 of the depth of the member, or 24 in.,
whichever is smaller, if Vs ≤ 4 fc′ bw d (ACI- 9.7.6.2.2)
d = distance from the extreme compressive fiber to the centroid of the prestressed rein-
forcement (but need not be taken less than 0.8h)
EXAMPLE 20.11.1
Determine the maximum spacing of #3 U stirrups at a point 4 ft from the support on the
beam of Example 20.10.1 (Fig. 20.10.6). The prestressing steel has an area of 2.30 sq in.
and a tensile strength fpu = 250,000 psi. The yield strength of the stirrup reinforcement
fyt = 60,000 psi, and the concrete (normal weight) has fc′ = 5000 psi. The service live
load is 1.50 kips/ft, and the dead load is 0.675 kip/ft.
SOLUTION
The inclined cracking strength Vc = Vci as determined in Example 20.10.1 is

Vci (controls) = 97.2 kips (more elaborate precedure)
At 4 ft from centerline of support,
Vu = [1.2(0.675) + 1.6(1.50)](20 − 4) = 3.21(20 − 4) = 51.4 kips
 φV 0.75(97.2) 
Vu = 51.4 kips >  c = = 36 kips
 2 2 
unless tests are made to justify its omission, a minimum amount of shear reinforcement
must be provided. Thus using Eq. (5.10.8)
Av b 50bw s
min = 0.75 fc′ w ≥
s f yt f yt
Av   20  
min = 0.75 5000   = 0.018 in.
s   60, 000  

 50(20) 
≥ = 0.017 in.
 60, 000 
Av
min = 0.018 in.
s
or, alternatively, using the more elaborate formula, Eq. (20.11.1),
Av Aps  f pu   1  d
min =   
s 80  f yt   d  bw

A 2.30  250   1  24
min v =    = 0.0055 in.
s 80  60   24  20
(Continued)
83
20.12 DEVELOPMENT OF REINFORCEMENT 883
Using the smaller of 0.018 in. and 0.0055 in., the maximum spacing s for #3 U stirrups is
2(0.11)
s= = 40.0 in.
0.0055
but, according to ACI-9.7.6.2.2, the spacing may not exceed 0.75h = 0.75(30) = 22.5 in.
or 24 in., whichever is smaller. Thus #3 stirrups could be placed no farther apart than
22.5 in.
20.12 DEVELOPMENT OF REINFORCEMENT
Following the basic concepts established for nonprestressed reinforced concrete in
Chapter 6, development of reinforcement must also be considered in prestressed members.
The prestressing force must be transferred into the concrete by bond (interaction between
concrete and steel strand) in the pretensioned beam, and the length required to accomplish
this is called the “transfer length.” This, of course, occurs in end regions and, in effect,
anchors the tendons.
The mechanism of transfer is summarized by Zia and Mostafa [20.14]. The high stress
in the pretensioned tendon must, on cutting of the tendon, be transferred to the concrete so
that equilibrium is achieved. The situation in the transfer zone is quite different from that in
the anchorage zones of nonprestressed concrete.
In prestressed concrete, the wires or strands are pretensioned to a high level of stress,
thus initially reducing the wire diameter prior to the placing of the concrete. Once the con-
crete has cured and the wires have been cut at the ends, the wires tend to shorten and cor-
respondingly increase in diameter. Thus a compression between the wires and concrete is
produced. The stress in the cut wire must increase from zero at the free end to the prestress
value at a certain distance from the free end. This distance is known as the “transfer length.”
During the accomplishment of the transfer, probably in the first few inches from the free
end, it is solely the friction that is developed as the slipping wires compress against the con-
crete. Farther from the end, adhesive bond—that is, slip resistance—certainly contributes
to the transfer.
The transfer length Lt typically is about 50 diameters for a strand, and about 110 to 120
diameters for a single wire [20.14]. These values assume a clean strand (or wire) surface,
a gradual release of the prestressing force to the concrete, and a steel stress of 140 to 150
ksi after transfer. For strands with a slightly rusted surface, the transfer distance is certainly
less, and for a sudden release of stress such as can occur by burning the strand, the transfer
length may easily be 20% greater. The transfer length does not seem to be affected by var-
iation in concrete strength [20.15].
The importance of the transfer length depends on the member under consideration. It
seems to be of little importance except on short members and in situations where flexural
cracking may occur in the transfer zone.
Development Length for Prestressing Strand

Exactly as for nonprestressed reinforcement, the tension in the prestressing steel necessary
to achieve nominal flexural strength Mn must be developed, typically by embedment length.
The purpose is to prevent general slip prior to achieving the necessary moment strength Mn;
thus the steel stress must increase from the effective prestress value fse to the value fps used
in computing the moment strength. The net anchorage length available to accommodate a
change in the tensile force is the distance to the end of the strand from the section in ques-
tion, less the transfer length Lt.
84
The following empirical relationship was established based on data by Hanson and
Kaar [20.15],
Ld − L t
( f ps − fse ) in ksi = (20.12.1)
db
where db is the nominal strand diameter. Thus,

L d − Lt
f ps ( ksi) ≤ fse (ksi) + (20.12.2)
db
Next, it is necessary to obtain an expression for Lt /db. Applying the concept used in
deriving Eq. (6.2.1), that is, the tensile force Aps fse in the strand must be transmitted
to the surrounding concrete by a “stress” us acting around the perimeter πdb over the
embedment length Lt,
Aps fse = us πdb Lt
 π db 2 
 4  fse = us πdb Lt (20.12.3)
 
Since the actual strand (3-or 7-wire) properties differ from those based on the nominal
diameter, a correction must be applied. The actual circumference is taken as 4 / 3 πdb ; and
the actual area Aps as 0.725πdb2/4.* Thus
 πd 2   4
0.725  b  fse = us   πdb Lt (20.12.4)
 4   3
Lt 2.175 fse
= (20.12.5)
db 16us
If the average bond stress us in the transfer zone is taken as 400 psi for clean strands
[20.15],
Lt 2.175 1 1
= fse = fse say, fse (20.12.6)
db 16(0.4) 2.94 3
where fse is measured in ksi.

Substituting Eq. (20.12.6) in Eq. (20.12.2) gives
Ld 1
f ps ≤ fse + − fse
db 3
and solving for Ld, the necessary development length to the end of the strand,
 2  (20.12.7)
Ld =  f ps − fse  db
 3 
which is equivalent to ACI Formula (25.4.8.1) for 7-wire pretensioning strands. Note that
the expression in parentheses uses stresses fps and fse in ksi units, but the resulting expres-
sion ( f ps − 2 / 3 fse ) is treated as a dimensionless constant. Investigation may be restricted to
cross sections nearest each end of the member that are required to develop their nominal
strengths Mn.
* The relationship between the nominal diameter and the actual area used herein corresponds to that at the time the ACI Code
provisions for transfer length were developed.
85
2 0 . 1 3 P R O P O R T I O N I N G O F C R O S S S E C T I O N S F O R F L E X U R E 885
EXAMPLE 20.12.1
Investigate the development of reinforcement for the beam of Example 20.10.1 (see
Fig. 20.10.6). Assume fps =221 ksi as computed in Example 20.8.1. The effective pre-
stress, after losses, may be assumed to be 0.8(175) = 140 ksi.
SOLUTION
Since only at midspan is the maximum moment strength Mn required, the total embed-
ment from midspan must exceed the development strength Ld,
 2 
Ld = 221 − (140) db
 3 

Ld = [(221 − (93.3)](0.5) = 64 in. < 240 in
n. OK
20.13 PROPORTIONING OF CROSS S ECTIONS FOR

FLEXURE WHEN NO TENSION IS PERMITTED
This section is intended to give further insight into the variables involved by discussing
the proportioning of the cross section. All cross sections used here consist of rectangular
flanges and web, though in practical design 90° junctions between flanges and web are
avoided because of forming difficulties.
This brief treatment presents only one case: that of no tension permitted, either at the
initial condition (at transfer of prestress before losses) or at the final condition (full dead
and live load after losses). Thus the desired stress distributions are shown in Fig. 20.13.1.
It can be shown that on any cross section having an axis of symmetry, there are two
points on the axis of symmetry, known as the “kern points,” that are located at distances kb
and kt from the centroid, as shown in Fig. 20.13.1. Each kern point represents the farthest
distance from the centroid at which a resultant force can act without inducing a stress of
opposite sign at the extreme fiber in the opposite direction from the centroid. This means
that the stress distribution varies from zero at the top to maximum at the bottom, when the
resultant force acts on the lower kern point; or it varies from zero at the bottom to maximum
at the top when the resultant force acts on the upper kern point. Referring to Fig. 20.13.1,
and using Eq. (20.6.1) with the internal force concept,
C Ckb yt
f = − = 0 (20.13.1a)
A I
and
C Ckt yb
f = − = 0 (20.13.1b)
A I
Solving Eqs. (20.13.1) for kb and kt,
I r2
kb = = (20.13.2a)
Ayt yt
and
I r2
kt = = (20.13.2b)
Ayb yb
in which r is the radius of gyration of the cross section.

86
Te h
Ac yb
Upper kern point
Te/Ac
yt T0/Ac
C = Te
CG kt
h kb
Lower (arm)D+L
e C = T0
kern yb
point (arm)D
Aps T0
Te = ηT0
T0 h
Section Ac yt
Initial condition Final condition
(stress distribution (stress distribution
at transfer) under full load
after losses)
Figure 20.13.1 Stress distributions when no tension is permitted—small girder moment.
The reader may recall from Section 20.6 that as the load applied to a prestressed con-
crete beam changes, the position of the internal force C must also change; that is, the
moment arm (see Fig. 20.13.1) measured from the position of the prestressing steel must
increase as the load increases. Thus when the moment MD due to dead load of the girder
is acting at the initial condition (immediately after transfer of prestress), the moment arm
of the internal couple is (arm)D; and when moment due to dead plus live load is acting, the
moment arm is (arm)D+L.
To achieve the triangular distribution at the initial condition, the prestressing steel
must have its centroid below the lower kern point by the amount (arm)D = MD /T0 (see
Fig. 20.13.1). If the girder moment MD is small, it may be possible and practical to position
the steel at e = kb + (arm)D below the centroid of the section. If the moment MD is large,
the distance [kb + (arm)D] may extend so far below the centroid of the section that it is too
close to the bottom for proper cover, if not outside the section. Thus for larger girder weight
(large MD), the optimum condition of triangular stress distribution at the initial condition
may be impossible to achieve.
Preliminary Design for Small Girder Moment

When the girder moment MD is small (say, MD representing 0.2 or less of the total, MD +
ML ), the (arm)D will be small and C can probably be located at the lower kern point; that
is, the steel will be located at e = kb + MD / T0 from the centroid of the section. Observing
Fig. 20.13.1 again,
(arm ) D + L − (arm ) D = kt + kb
MD + ML MD
− = kt + kb
Te T0
If MD is small, MD /T0 in the above equation may be assumed to be MD /Te, then
ML
≈ kt + kb
Te
87
or
ML
Te ≈ (20.13.3)
kt − kb
The maximum stress in Fig. 20.13.1 may be obtained from the stress at the centroid
(which is T/Ac) by the linear relationship. Assuming kt + kb ≈ 0.5h, Eq. (20.13.3) becomes
ML
Te ≈ (20.13.4)
0.5h
If the section is symmetrical, the stress at the centroid will be half of the maximum; thus the
required area Ac of the section would be, for the initial condition at transfer,
T0
required Ac = (20.13.5)
0.50 fic
and for the final condition of dead plus live load after losses,
Te
0.50 f fc
For the preliminary selection of cross section in case of small girder moment MD, the
following steps may be followed:
1. Select the overall depth h of the section. This is somewhat arbitrary but in the absence
of other limitations, the guidelines of Lin and Burns [20.6] may be followed:
(a) Use 70% of the depth that would be used for nonprestressed reinforced concrete
construction.
(b) For slabs having light loading, h ≥ 1/55 of the span L.
(c) For slabs having heavy loading, h ≥ L /35.
(d) On bridges, h ranges between L /15 and L /25.
(e) As a rule of thumb, h (inches) of beams can be approximated by the empirical
formula, 1.5 M (ft-kips) to 2.0 M .
2. Compute the approximate Te from Eq. (20.13.4).

3. Determine the approximate Ac from Eq. (20.13.6).
4. Proportion a symmetrical I-shaped section.
5. With the preliminary section, compute the section properties and locate the desired
distance e of the steel centroid from the section centroid (CG),
MD
e = kb + (arm ) D = kb + (20.13.7)
T0
where T0 = Te /η and η is the proportion of initial prestress remaining after losses.
6. If the steel can be located at the desired e, then Te is more correctly determined:
MD + ML MD + ML
Te = = (20.13.8)
(arm) D + L e + kt
Then T0 = Te /η and a new value of e is established; the iterative process is repeated until the
desired accuracy has been obtained.
8
7. Equations (20.13.5) and (20.13.6) are then used to determine the required Ac.
When the equations give significantly different requirements, the section may be
changed to become somewhat unsymmetrical with respect to the centroidal axis.
In such a case, the average stress is not half of the maximum. In general, from
Fig. 20.13.1,
T0 h
fic yt
Te h
f fc yb
The minimum area Ac is obtained when Eqs. (20.13.9) and (20.13.10) give the same result.
Preliminary Design for Large Girder Moment

When the girder moment MD exceeds about 0.2 to 0.3 of the total moment MD + ML, the
(arm)D will be too large to permit the steel distance e to be at kb + MD /T0 from the centroid
of the section. Thus the initial stress distribution at transfer cannot be triangular but instead
will be trapezoidal (see Fig. 20.13.2). The final condition, which can still give a triangular
stress distribution, will probably govern. Thus
MD + ML
Te = (20.13.11)
e + kt
When the final condition controls, more of the area Ac should be located at the top where the
highest stress occurs. Thus an unsymmetrical section is indicated—for instance, a T-shaped
section. As an approximation, Lin and Burns [20.6] suggest (e + kt) ≈ 0.65h for use in Eq.
(20.13.11). Generally e + kt will vary from 0.3h to 0.8h, with the average about 0.65h. For
preliminary design, Eq. (20.13.11) then becomes
MD + ML
Te ≈ (20.13.12)
0.65h
The required area Ac would then be
Te MD + ML
required Ac = = (20.13.13)
avg fc 0.65h(avg fc )
where avg fc is the compressive stress at the centroid of the section. For the unsymmetrical
section, yb /h in Eq. (20.13.10) will be greater than 0.5, say, 0.6. Equation (20.13.13) for
design would then be
MD + ML 2.6( M D + M L )
required Ac = = (20.13.14)
0.65h(0.6 ff c ) h ff c
For the preliminary selection of cross section in case of large girder moment, the follow-
ing steps may be followed:
1. When MD /(MD + ML) > 0.2 to 0.3, establish the overall depth according to Step 1 for
small girder moment and use Eq. (20.13.14) to estimate Ac.
2. Proportion an unsymmetrical section; a T-section may be a practical choice.
89
20.13 PROPORTIONING OF CROSS SECTIONS FOR FLEXURE 889
3. With the preliminary section, compute the section properties and establish the dis-
tance e from the centroid of the section to the centroid of the prestressing steel. For
large girder moment, one will find
 M 
e <  kb + D 
 T0 
If e can be equal to kb + MD /T0, then the procedure for the small girder moment case is to
be followed.
4. With the steel located, Te can be determined using Eq. (20.13.8). From that, T0 = Te /η.
5. The required area Ac based on the final condition is then determined using
Eq. (20.13.10).
6. Check the required area based on the initial condition. Referring to Fig. 20.13.2, a
trapezoidal stress distribution should occur.
The maximum compressive stress f is, using Eq. (20.6.1) with the internal force concept,
C Cec y
f = + (20.13.15)
A I
Since C = T0, A = Ac, I = Ac r2, and y = yb,

T0 T0 ec yb
f = + ≤ fic
Ac Ac r 2
T0  e − M D /T0  (20.13.16)
= 1 +  ≤ fic
Ac  kt 
The required area Ac based on the initial condition is
T0  e − M D /T0 
required Ac = 1 +  (20.13.17)
fic  kt 
Again the minimum area Ac will be obtained when Eqs. (20.13.10) and (20.13.17) give the
same result.
ec = e – MD/T0
C = Te
yt
kt
CG
Te
h kb Ac
C = T0
(arm)D+L
yb e
(arm)D
Aps
T0 Te
Initial condition Final condition
Figure 20.13.2 Stress distributions when no tension is permitted—large girder moment.

890
EXAMPLE 20.13.1
Design a cross section for a 30-in.-deep girder whose girder moment MD = 45 ft-kips.
The live load moment to be carried is 300 ft-kips. The initial prestress fsi = 175,000 psi;
assume 20% losses. The allowable service load stresses are fit = 0, fic = 2400 psi, fft = 0,
and ffc = 2250 psi. Omit checks of moment strength, cracking moment, shear strength,
and development of reinforcement.
SOLUTION
(a) Preliminary design. The girder moment as a percent of the total moment is
MD 45
= = 0.13 < 0.2
M D + M L 45 + 300
Approach as a small girder moment design. Using Eqs. (20.13.4) and (20.13.6),
ML 300(12)
Te ≈ = = 240 kips
0.5h 0.5(30)
Te 240
T0 = = = 300 kips
η 0.8
T0 300
required Ac = = = 250 sq in.
0.5 fic 0.5(2.40)
Since ffc /fic = 2.25/2.4 = 0.94 is greater than Te /T0 = 0.8, the equation based on the ini-
tial condition controls if the section is symmetrical. Though a slightly unsymmetrical
section (with the centroid below the middepth in this case) would give the section of
minimum area, the procedure is illustrated using a symmetrical one.
Try a 30-in.-deep section with flanges 5 × 17 in. and a 4-in.-thick web, Ac = 250 sq in.
(Fig. 20.13.3).
(b) Determine the section properties.
Area (sq in.) I (in.4)
17 × 30 rectangle 510 38,250

13 × 20 sides –260 –8677
250 29,583
I r 2 r 2 118.3
r2 = = 118.3 sq in., kt = kb = = = = 7.89 in..
Ac yb yt 15
(c) Locate centroid of prestressed reinforcement. Using approximate Te, find

approximate T0.
Te 240
T0 ≈ = = 300 kips
η 0.8

M 45(12)
desired e = kb + D = 7.89 + = 9.69 in.
T0 300
(Continued)
891
The available distance is 15 in., less appropriate cover. Thus e = 9.69 in. would be
acceptable. Recalculate Te using Eq. (20.13.8),
M D + M L (45 + 300)12
Te ≈ = = 235 kips
e + kt 9.69 + 7.89
235
revised T0 = = 294 kips
0.8
45(12)
revised e = 7.89 + = 9.73 in.
294
This is in close agreement with the previous value of 9.69 in. If the first estimate had
been farther off, additional iterations might have been needed.
(d) Check whether the area Ac is adequate. Using Eqs. (20.13.9) and (20.13.10).
T0 h 294(30)
required Ac (initial condition) = = = 245 sq in.
fic yt 2.4(15)
Te h 235(30)
required Ac (final condition) = = = 209 sq in.
f f c yb 2.25(15)
The section is adequate.

Use 5 × 17 in. flanges with a 4-in. web (Ac = 250 sq in.). If made slightly unsymmetrical,
the area could be reduced to somewhere between 209 and 245 sq in. A final check of
initial and final stresses should be made, as was done in Section 20.6, but the illustration
is omitted here.
17”
5”
30”
4”
5”
17”
Figure 20.13.3 Section for design example 20.13.1.

892
EXAMPLE 20.13.2
Redesign the cross section of Example 20.13.1 for MD = 200 ft- kips and ML =
145 ft-kips. Note that the total MD + ML = 345 ft-kips is the same as in Example 12.13.1.
SOLUTION
(a) Preliminary design. The girder moment as a percentage of the total moment is
MD 200
= = 0.58 > 0.2 to 0.3
M D + M L 345
Approach as a large-girder moment design. Using Eq. (20.13.14),
2.6( M D + M L ) 2.6(345)12
required Ac = = = 159 sq in.
h f fc 30(2.25)
An unsymmetrical shape is to be selected; try flanges 5 × 14 in. and 5 × 8 in. with a 4-in.
web having Ac = 190 sq in. (Fig. 20.13.4). The larger the ratio of MD to (MD + ML), the
larger may be the numerator coefficient (2.6, used above); thus an area somewhat larger
than indicated by the formula was used.
14”
5”
yt =
13.03”
30”
4”
yb =
16.97”
5”
8”
Figure 20.13.4 Section for design example 20.13.2.
(b) Determine section properties. Referring to Fig. 20.13.4, first locate the centroid of
the area measured from the top.
Area, Ac Arm, y Ac y I

(sq in.) (in.) (in.3) (in.4)
5 × 10 top flange projection 50 2.5 125 313

I0 104
5 × 4 bottom flange projection 20 27.5 550 15,125
I0 42
4 × 30 web (full depth) 120 15 1,800 27,000
I0 9,000
190 2475 51,584
(Continued)
893
yt = y =
∑ A y = 2475 = 13.03 in.
∑ A 190
I 0 = I − Ac y 2 = 51, 584 − 190(13.03)2 = 19, 300 in.4
I0
r2 = = 101.7 sq in.
Ac
2
r 101.7
kt = = = 5.99 in.
yb 16.97
r 2 101.7
kb = = = 7.81 in.
yt 13.03
(c) Locate centroid of prestressed reinforcement. Assume that with adequate cover the
steel may be centered 4 in. from the bottom of the section. Then

e = yb – 4 = 16.97 – 4 = 12.97 in.
and using Eq. (20.13.8),
MD + ML 345(12)
Te = = = 218 kips
e + kt 12.97 + 5.99
then
218
T0 = = 273 kips
0.8
Check:
MD 200(12)
kb + = 7.81 + = 7.81 + 8.79 = 16.60 in. > e
T0 273
This shows that the tendons cannot be located far enough from the centroid of the sec-
tion to give a triangular stress distribution at the initial condition.
(d) Check whether the area Ac is adequate. Using Eqs. (20.13.10) and (20.13.17),
Te h 218(30)
required Ac (final condition) = = = 171 sq in.
f fc yb 2.25(16.97)
T0  e − M D / T0 
required Ac (initial condition) = 1 + 
fic  kt 

273  12.97 − 8.79 
= 1 + = 193 sq in.
2.4  5.99 

This is close enough but shows that the initial condition governs in this case, the reason
being that area has been shifted from the bottom of the section to the top, where it is
needed for the final condition. The minimum area section for this case would be slightly
more symmetrical than the chosen one.
To verify the result, make a final check of stresses for the initial and final conditions
(see Section 20.6).
The general line of reasoning presented here may also be used when tension is per-
mitted at initial or final conditions, or both. Lin and Burns [20.6] and Naaman [20.5]
provide a detailed treatment of design of sections when tension is permitted.
894
20.14 ADDITIONAL TOPICS
Many other topics have been omitted from this introductory treatment of prestressed con-
crete. Topics such as the practical design approaches, use of I-shaped and nonsymmetrical
sections, prestressing of continuous members, stresses in end blocks, partial prestressing,
deflections, composite construction, and other specific applications are adequately and
extensively treated in textbooks devoted entirely to the subject [20.2–20.7].
SELECTED REFERENCES
20.1 Charles C. Zollman. “Reflections on the Beginnings of Prestressed Concrete in America,” PCI
Journal, Part 1, 23, May/June 1978, 22–48, Part 2, 23, July/August 1978, 29–63; Part 9, 25,
January/February 1980, 124–145, March/April 1980, 94–117, May/June 1980, 123–152.
20.2 Michael P. Collins and Denis Mitchell. Prestressed Concrete Structures. Canada: Response
Publications, 1997, 766 pp.
20.3 James R. Libby. Modern Prestressed Concrete, Design Principles and Construction Methods
(4th ed.). Princeton, NJ: Van Nostrand Reinhold, 1990, 872 pp.
20.4 Edward G. Nawy. Prestressed Concrete, A Fundamental Approach (2nd ed.). Upper Saddle
River, NJ: Prentice-Hall, 1996, 789 pp.
20.5 Antoine E. Naaman. Prestressed Concrete Analysis and Design—Fundamentals. Ann Arbor;
MI: Techno Press 3000, 2012, 1173 pp.
20.6 T. Y. Lin and Ned H. Burns. Design of Prestressed Concrete Structures (3rd ed.). New York: John
Wiley & Sons, 1981, 646 pp.
20.7 Arthur H. Nilson. Design of Prestressed Concrete. New York: John Wiley & Sons, 1978,
526 pp.
20.8 Yves Guyon. Prestressed Concrete, Vols. 1 and 2. New York: JohnWiley & Sons, 1960.
20.9 Gustave Magnel. Prestressed Concrete (2nd ed.). London: Concrete Publications, 1950.
20.10 Paul Zia, H. Kent Preston, Norman L. Scott, and Edwin B. Workman. “Estimating Prestress
Losses,” Concrete International, 1, June 1979, 32–38.
20.11 PCI Committee on Prestress Losses. “Recommendations for Estimating Prestress Loss,” PCI
20.12 J. R. Janney, E. Hognestad, and D. McHenry. “Ultimate Flexural Strength of Prestressed and
Conventionally Reinforced Concrete Beams,” ACI Journal, Proceedings, 52, January 1956,
601–620.
20.13 M. A. Sozen and N. M. Hawkins. Discussion of “Shear and Diagonal Tension Report,”
Report of ACI- ASCE Committee 326, ACI Journal, Proceedings, 59, September 1962,
1341–1347.
20.14 Paul Zia and Talat Mostafa. “Development Length of Prestressing Strands,” PCI Journal, 22,
September/October 1977, 54–65.
20.15 Norman W. Hanson and Paul H. Kaar. “Flexural Bond Tests of Pretensioned Prestressed
20.16 M. H. Harajli and M. Y. Kanj. “Ultimate Flexural Strength of Concrete Members

Prestressed with Unbonded Tendons,” ACI Structural Journal, 88, November–December
1991, 663–673.
20.17 S. V. Krishna Mohan Rao and Walter H. Dilger. “Control of Flexural Crack Width in Cracked
Prestressed Concrete Members,” ACI Structural Journal, 89, March–April 1992, 127–138.
20.18 A. Alshegeir and J. A. Ramirez. “Strut-Tie Approach in Pretensioned Deep Beams,” ACI
20.19 Ravi K. Devalapura and Maher K. Tadros. “Critical Assessment of ACI 318 Eq. (18-3) for
Prestressing Steel Stress at Ultimate Flexure,” ACI Structural Journal, 89, September–October
1992, 538–546.
20.20 Julio A. Ramirez. “Strut-Tie Design of Pretensioned Concrete Members,” ACI Structural
Journal, 91, September–October 1994, 572–578.
20.21 Post-Tensioning Institute. Post-Tensioning Manual (6th ed.) PTI TAB.1-06, Phoenix,
AZ, 2006.
895
PROBLEMS 895
PROBLEMS
All problems are to be worked in accordance with the ACI Code, assuming Class U mem-
bers and using the allowable concrete stresses in ACI-24.5 unless otherwise indicated.
Flexural strength-related provisions are to be in accordance with ACI-20.3.2 and ACI-22.3,
as indicated by the instructor.
20.1 The rectangular beam of the figure for Problem distributed live load may be safely carried on a
20.1 contains pretensioned steel with an initial 40-ft simple span? Omit consideration of flex-
tensile stress of 160 ksi ( fpu = 250 ksi; low- ural strength Mn.
relaxation strand having fpy = 0.90 fpu). The
concrete has fci′ = 4000 psi and fc′ = 6000 psi
(n = 6.5). The beam is on a simple span of 35 ft.
(a) Determine the concrete stresses at top and
bottom, and the steel stress at transfer imme- 24”
diately after the wires are cut at the ends. 5”
(b) Recompute the stresses in part (a) after a
Aps = 2.5
20% loss in prestress. What is the maximum sq in.
service live load that can be superimposed on 12”
the beam? Consider only the section of maxi-
mum bending moment; omit consideration of Problem 20.3
flexural strength Mn.
20.4 For the live load determined in Problem 20.1,
determine the maximum permissible eccentricity
of the prestressing tendons at the 1 4 point of the
span to satisfy service load allowable stress lim-
18”
Aps = 0.8 its. Omit consideration of flexural strength Mn.
sq in. 20.5 For the live load determined in Problem 20.3,
determine the maximum permissible eccentricity
12” 2”
of the prestressing tendons at the 1 8 point of the
span to satisfy service load allowable stress lim-
Problems 20.1, 20.2 and 20.9 its. Omit consideration of flexural strength Mn.
20.6 A straight pretensioned member 35 ft long is
20.2 Based on the midspan cross section of the figure 18 in. square in cross section. It is concentri-
for Problem 20.1, investigate whether it is pos cally prestressed with 2.24 sq in. of high ten-
sible to increase the live load moment capacity sile strength steel wire. The wires are stressed
by either or both of the following: originally to 170 ksi and are anchored to end
(a) Increasing the initial prestress above bulkheads. Use both the approximate and the
160 ksi. “exact” methods to calculate the loss and per-
(b) Decreasing the eccentricity. centage of loss of prestress in the wires due to
elastic shortening of the concrete at transfer. Use
Assume a 20% loss of initial prestress. Determine fci′ = 6000 psi (n = 6.5).
the maximum service live load capacity pos- 20.7 An 18- in. square concrete member is post-
sible by adjusting the prestress or the eccen- tensioned by four cables, each with an area of
tricity or both, but still not violating the ACI 0.56 sq in. These cables are stressed one after
Code limitations. Omit consideration of flexural another, each to a stress of 170 ksi. Without tak-
strength Mn. ing any account of the eccentricity of the cables,
20.3 The rectangular section of the figure for Problem compute the loss and percentage of loss of pre-
20.3 has been pretensioned by a force of 300 kips stress in each cable due to the elastic shorten-
after all losses. If the concrete has fci′ = 4000 psi ing of the concrete. Compute the average loss of
and fc′ = 6000 psi (n = 6.5), what uniformly prestress. Assume n = 6.5.
896
20.8 The symmetrical double cantilever beam shown U stirrups for the beam of Problem 20.3 if
in the figure for Problem 20.8 is to be post- the maximum service load computed in that
tensioned by a single cable ABCDE. The cable problem is acting. Use and compare both the
consists of 12 wires, each of 0.20 in. diameter, alternate (ACI- 22.5.8.2) and more accurate
and is to be prestressed simultaneously from (ACI-22.5.8.3.1 and ACI-22.5.8.3.2) proce-
both ends of the member. It is desired that the dures in the ACI Code.
minimum stress in the cable immediately after 20.12 Design of section for small dead load moment.
stressing and before any creep or shrinkage (a) Make a preliminary design (use rectan-
losses take place be 145 ksi. The cable is such gular flanges and a web) for a section of
that the friction constant µ = 0.50 and the wob- a prestressed beam to resist a total bend-
ble effect K = 0.0010. Determine the steel stress ing moment of 960 ft-kips, assuming that
at the jack, the percentage of friction losses, and the moment due to the girder weight is
the extension that will be required at each jack. 70 ft-kips. The overall depth of the section
Solve by each of the following methods. is to be 42 in. and the effective prestress fse
(a) Neglecting variation in tension throughout in the steel is 136 ksi. In selecting a section
the length, using Eq. (20.7.22), Ps = Px (1 + assume the minimum thickness of compo-
µα + KLx). nents (flanges or web) is 5 in.
(b) Neglecting variation in tension at every (b) Make the final design for the preliminary
point along the length of the curve but con- section you selected for part (a), revising
sider variation from segment to segment, as you find necessary, with the objective of
using Eq. (20.7.22). obtaining a minimum cross-sectional area.
(c) Using the “exact” expression, Eq. (20.7.20). The selected cross- sectional area should
20.9 Determine the nominal moment strength Mn not exceed 460 sq in.; however, the “best”
and the cracking moment Mcr for the preten- design will be presumed to be the one hav-
sioned bonded section of the figure for Problem ing the least cross-sectional area.
20.1. The concrete has fc′ = 6000 psi and the (c) Make a check of stresses at initial (trans-
steel has fpu = 250 ksi (stress-relieved strand). fer) and final conditions. Use the following
Assume the average stress– strain curve of control stresses:
Fig. 20.8.1 is to be used for the steel. fic = 2400 psi ffc = 2250 psi fsi = 160 ksi
20.10 Assuming that no special tests are to be made, fit = 0 psi fft = 0 psi fse = 136 ksi
determine the number and spacing for #3 U stir- 20.13 Design of section for large load moment. The
rups for the beam of Problem 20.1 if a service load data are the same as Problem 20.12, except that
of 0.48 kip/ft is acting. Use the alternate procedure the moment due to the girder weight is 650 ft-
for Vc of ACI-22.5.8.2, as well as the more accurate kips, instead of 70 ft-kips (total moment is the
procedure of ACI-22.5.8.3.1 and ACI-22.5.8.3.2. same); the minimum thickness for components
20.11 Assuming that no special tests are to be made, (flanges or web) is 4 in.; and the cross-sectional
determine the number and spacing for #3 area should not exceed 340 sq in.
R = 200 ft
Symbolic about
1’– 6” midspan
3’– 0” A C D 3’– 0”
R = 100 ft
B E
1’– 6”
1’– 6”
30’– 0” 10’– 0” 20’– 0” 20’– 0”
100’– 0”
Problem 20.8
CHAPTER 21
COMPOSITE MEMBERS
AND CONNECTIONS
21.1 INTRODUCTION
In the context of reinforced concrete construction, the term composite members refers to
those constructed with two or more materials, at least one of them being concrete (with
or without steel bars), interconnected in a way that allows transfer of stresses between the
different materials. Based on the materials used, two broad categories of concrete com-
posite members can be defined: concrete-concrete composite members and concrete-steel
composite members.
Concrete-concrete composite members, sometimes simply referred to as concrete com-
posite members, are those made of concrete cast at different times. This practice is common
in precast girder construction, in which a cast-in-place concrete slab or topping is added on
site once the precast girder is in place.
Concrete-steel composite members, on the other hand, combine concrete (with or with-
out reinforcing bars) and structural steel shapes. These are commonly found in floor sys-
tems of steel structures [Fig. 22.1.1(a)], in which a concrete slab, typically supported on a
metal deck, is constructed to act compositely with steel beams. Concrete-steel composite
members are also used as columns, where either a structural shape is embedded in rein-
forced concrete [Fig. 22.1.1(b)] or a steel tube is filled with concrete [Fig. 22.1.1(c)].
This chapter deals with the behavior and design of concrete composite flexural members,
as well as composite columns that consist of either concrete-encased steel shapes or
concrete-filled steel tubes [Fig. 22.1.1(b) and 22.1.1(c)]. Connections between composite
columns and steel beams are also discussed, as the force transfer mechanisms differ from
those in reinforced concrete or steel construction. Design of composite floor systems con-
sisting of steel beams and a concrete deck is well covered in the AISC Steel Construction
Manual [21.1] and in other textbooks [21.30], and is therefore not included here.
21.2 COMPOSITE ACTION
Before discussing the behavior and design of composite members, it is necessary to under-
stand the fundamentals of composite action and the forces developed at the interface
between the various materials in the composite member. Consider a simply supported beam
made of two different materials, as shown in Fig. 21.2.1(a). Three cases are considered with
respect to the interface between the two materials: frictionless interface, “rough” interface,
and “slip-free” interface.
89
898 C hapter 2 1 C omposite M embers and C onnections
Shear studs Hollow steel section

Bars Hollow steel section
Shear studs Concrete
Concrete
Metal deck Concrete
Steel
Steel section section
Ties
(a) Composite beam (b) Concrete-encased steel (c) Concrete-filled tube
composite column
Figure 21.1.1 Common concrete-steel composite members.
b2
Material 2
Material 1
b1
Cross section
(a) Beam made of two materials
Material 2 M2
M1
Material 1 Strains
(b) Beam with frictionless interface (no composite action)
Material 2 C2
M2
M1
T1
Material 1
Strains
(c) Beam with “roughened” interface (partial composite action)
Material 2 C2
M2
M1
T1
Material 1
Strains
(d) Beam with slip-free interface (full composite action)
Figure 21.2.1 Composite action in flexural members.
Under the action of transverse loads, the member with frictionless interface will behave
as two beams in parallel, with relative sliding free to occur along the interface between the
two materials [Fig. 21.2.1(b)]. Thus, only normal stresses develop between the two material
components. Even though two materials are used in this structural element, the member is
considered to be noncomposite because the two material components work separately to
89
21.2 COMPOSITE ACTION 899
resist the applied loading. In this case, the moment acting on the member is equal to the
summation of the moments of the individual components [M1 and M2 in Fig. 22.2.1(b)].
When some roughness is introduced at the interface between the two materials, slid-
ing may still occur, but not freely, leading to the transfer of force parallel to the interface
(shear). In this case, the moment is resisted by a combination of bending in each material
component [M1 and M2 in Fig. 22.2.1(c)], as well as by a couple created by the transfer of
shear between the two materials [T1 and C2 in Fig. 21.2.1(c)]. This member is said to be
partially composite.
The case depicted in Fig. 21.2.1(d) corresponds to an interface at which relative slip is
prevented: the section will behave as a single unit, with continuity of normal strains across
the depth of the section. The moment carried by the composite beam can be calculated as
for the case of partially composite action, in which each component carries a moment and
an axial force. The amount of shear transferred along the interface is that required for equi-
librium when the two material components act as a single unit. These members are referred
to as fully composite.
A slip-free interface in concrete composite members is not feasible in practice unless
a “rough” interface is provided and the interface shear stresses are relatively low, up to
approximately 200 to 250 psi [21.2, 21.3]. The requirement for full composite action, there-
fore, is typically based on the shear transfer along the interface between the material com-
ponents, not on achieving a slip-free interface. As will be discussed later, interface slip is
often ignored in the calculation of flexural strength of composite members. Its effect on
member stiffness, however, may be important enough in some cases to require explicit
consideration.
Interface Shear
Calculation of shear developed at the interface between two material components behaving
as a unit can be done using statics. Let us consider again the beam shown in Fig. 21.2.1(a).
Transverse forces are applied on the composite beam only (i.e., any loads acting sepa-
rately on each material component are not considered). For simplicity, let us transform
the composite beam into a beam made of the material used in the bottom part of the beam
(material 1). This can be done by changing the width b2 to b2 (E2 /E1), where E1 and E2 are
the moduli for materials 1 and 2, respectively. Taking a slice of thickness dx at a distance x
from the left support [Fig. 21.2.2(a)] leads to the free-body diagram shown in Fig. 21.2.2(b).
Only normal stresses are shown acting on the cut faces, as these are the only stresses needed
in the calculation of interface shear. For the case of linear elastic behavior, normal stresses
are assumed to increase linearly with the distance from the centroid of the transformed sec-
tion, y. In the presence of shear, the moments at the two sections, M(x) and M(x + dx), will
differ, and so will the magnitudes of the normal stresses.
By taking a horizontal cut along the interface between the two material components
(Fig. 21.2.2(b), it can be seen that a force Vh(x) is required to balance the resultant normal
forces on the two cut faces, F(x) and F(x + dx). These forces can be calculated as follows
[Fig. 21.2.2(b)],
y2 M ( x) y
F ( x) = ∫ b( y) dy (21.2.1)
y1 I tr
y2 M ( x + dx ) y y2 [ M ( x ) + dM ( x )] y
F ( x + dx ) = ∫ b( y) dy = ∫ b( y) dy (21.2.2)
y1 I tr y1 I tr
dM ( x ) y2
I tr ∫y1
Vh ( x ) = F ( x + dx ) − F ( x ) = yb( y) dy (21.2.3)

where y is the distance from the centroid of the transformed section to the location where
normal stress is calculated, b( y) is the section width at a distance y from the centroid, and Itr
is the moment of inertia of the transformed section. The shear force Vh(x) can be expressed
in terms of a shear stress vh(x) acting on a surface of area bv dx, where bv is the width of the
90
contact surface. Also, the integral in Eq. (21.2.3) represents the first moment of the area
above the interface with respect to the centroid of the transformed section, Q1. Thus,
dM ( x ) y2 dM ( x )
I tr ∫y1
vh ( x )bv dx = yb( y) dy = Q1 (21.2.4)
I tr
Solving for vh(x),
dM ( x ) Q1 V ( x )Q1
vh ( x ) = = (21.2.5)
dx I tr bv I tr bv
Equation (21.2.5) is valid for linear elastic behavior and over regions without concentrated
forces. For composite members with the same cross section along their length, Q1, Itr and
bv are constant; thus the shear stress at the interface varies in the same proportion as the
shear force V. It should be noted that Vh(x) may also be calculated from the normal resultant
forces below the interface between the two materials.
For members behaving beyond the linear elastic range, the expressions above are not
valid. The horizontal force Vh, however, may be calculated between two sections separated
by a distance, say L1–2, based on equilibrium (Fig. 21.2.3). In practice, the horizontal shear
force Vh is calculated between a section of maximum moment and the inflection point
(i.e., M = 0), and also between concentrated load points. Considering the composite beam
shown in Fig. 21.2.3, the interface shear force Vh and average shear stress vh between the
sections of maximum moment and M = 0 are calculated as follows,
Vh = F ( x = L /2) − F ( x = 0) = F2 − F1 = F2 (21.2.6)
Vh F2 (21.2.7)
vh = =
bv L1- 2 bv L / 2
where F2 is the resultant normal force above the interface at the midspan section, which can
be calculated based on the nominal stress condition. For T-sections exhibiting a rectangular
b2 E2/E1
b1
x dx Cross section
(a) Composite beam (transformed into material 1)
bv
b(y) dy M(x) M(x+dx)

y2
cgtr y
y1
dx
F(x) F(x+dx)
y1
Vh(x)
(b) Normal stresses in differential beam segment (above) and interface shear force (below)
Figure 21.2.2 Interface shear force at a section in a composite beam.

901
21.3 CONCRETE COMPOSITE FLEXURAL MEMBERS 901
1 2
Material 2
Material 1
L1–2
F1 F2
Vh
Figure 21.2.3 Calculation of interface shear force between two sections in a composite beam.
section behavior (i.e., a ≤ hf ; see Chapter 4), and ignoring the effect of compression steel in
the flange (if any), F2 is calculated as
F2 = 0.85 fc′ bE a (21.2.8)
For T-section behavior,
F2 = 0.85 fc′ bE h f (21.2.9)
where bE and hf are the flange effective width (see Section 4.3) and thickness, respectively.
21.3 CONCRETE COMPOSITE FLEXURAL MEMBERS

The discussion on concrete composite members will focus on those consisting of a precast
(either nonprestressed or prestressed) beam and a cast-in-place slab, which are commonly
found in precast construction. However, the same concepts apply to composite cast-
in-place concrete members in which concrete is cast in stages.
Assuming sufficient shear can be transferred along the interface between the precast
beam and the cast-in-place slab to achieve full composite action, the design at ultimate
for flexure and shear of concrete composite beams is basically the same as for monolithic
members. Special attention should be paid to the loads expected to be resisted by the pre-
cast beam during construction, however, as these will depend on whether the member is
shored or unshored during construction.
For the case of shored construction, any load in addition to the self-weight of the precast
beam is resisted by the composite member. In unshored construction, on the other hand,
the precast beam will need to be capable of supporting, in addition to its self weight, the
weight of the cast-in-place slab and any other load applied before the slab behaves compos-
itely with the precast beam. The design of the precast beam will thus need to be adequate to
resist the loads expected prior to the achievement of composite action. Stresses and deflec-
tions under service loads will also depend on the construction process.
Interface Shear Strength

Shear strength along the interface between concretes cast at different times has been the
subject of extensive research [21.2–21.8]. Shear along the interface between two concretes
cast at different times (e.g., precast beam and cast-in-place slab concretes) is transferred
primarily through friction and dowel action of reinforcement crossing the interface [21.6].
Thus, interface shear strength depends primarily on interface roughness, aggregate type,
and amount and strength of reinforcement crossing the interface.
For several decades, the use of a shear-friction analogy (see Section 5.15) has been
used for calculating interface shear strength [21.5], where a normal or clamping force is
902
calculated based on the yield strength of the steel crossing the interface. This analogy is
supported by the fact that as relative slip occurs along the interface between the two con-
cretes, the roughness of the concrete surfaces causes a vertical separation, which in turn
produces an elongation of the steel crossing the interface, provided it is well anchored on
both sides [21.5]. The tensile force developed in the reinforcement translates then into a
normal or clamping force. Experimental data [21.3, 21.6], however, have indicated that
even in the absence of transverse reinforcement, some amount of shear can be transferred
when a rough interface is provided.
ACI 318-14 provides two design approaches for interface shear in concrete composite
members (ACI-16.4). One of these is a sectional method, wherein the interface shear stress
is calculated based on the shear acting on the section (ACI-16.4.3 and ACI-16.4.4). The
alternative method, on the other hand, applies to a segment of the member (ACI-16.4.5); it
is based on the difference between resultant normal forces above (or below) the interface
between two sections.
Sectional Method
In the sectional method, the factored shear stress acting on the interface, vuh, is calculated
as follows,
Vu
vuh = (21.3.1)
bv d
where Vu is the factored shear acting on the section, bv is the width of the contact sur-
face, and d is the effective depth of the composite member, but not to be taken less than
0.8h for prestressed concrete members (ACI-16.4.4.3). At any section, Eq. (21.3.2) must
be satisfied,
φvnh ≥ vuh (21.3.2)
where vnh is the nominal shear stress capacity of the concrete interface and φ = 0.75
(ACI-21.2).
The nominal interface shear strength depends primarily on the roughness of the interface
and the amount and yield strength of the transverse reinforcement crossing the interface
(if any). At a minimum, the contact surface of the hardened concrete must be clean and
free of laitance, and either intentionally roughened to an amplitude of approximately 1 4 in.
(referred to simply as intentionally roughened) or crossed by transverse steel with area
Av ≥ Av,min. Three cases are listed in ACI 318-14 for determination of interface shear stress
capacity as follows (ACI-16.4.4.1 and ACI-16.4.4.2).
1) For concrete placed against an intentionally roughened hardened concrete surface

and Av < Av,min, or for concrete placed against not intentionally roughened hardened
concrete and Av ≥ Av,min [ACI Table 16.4.4.2(c) and (d)],
vnh = 80 psi (21.3.3)
2) For vuh ≤ φ (500 psi), concrete placed against an intentionally roughened hardened
concrete surface and Av ≥ Av,min [ACI Table 16.4.4.2(a) and (b)],
 0.6 Av f yt 
vnh = λ (260 psi) +  ≤ 500 psi (21.3.4)
 bv s 
where λ is the lightweight concrete strength reduction factor (see Section 1.8) to reflect
the lower interface shear strengths exhibited by lightweight concrete [21.8], Av is the
area of transverse reinforcement perpendicular to the concrete surface at a spacing s, and
fyt is the yield strength of the transverse reinforcement.
903
3) For vuh > φ (500 psi) (ACI-16.4.4.1 and ACI-22.9),
µ Av f yt
vnh = (21.3.5)
bv d
where µ = 1.0λ for concrete placed against an intentionally roughened concrete surface
and 0.6λ for concrete placed against a not intentionally roughened concrete surface. For
cases in which normal-weight concrete is cast against intentionally roughened hardened
concrete, the interface nominal shear stress capacity in Eq. (21.3.5) is limited to the least
of [ACI-Table 22.9.4.4(a)–(c)]:
a) 0.2 fc′
b) (480 psi + 0.08 fc′ )
c) (1600 psi)
For all other cases, the interface nominal shear stress capacity is limited to the lesser of
0.2 fc′ and 800 psi [ACI Table 22.9.4.4(d) and (e)]. When concretes with different com-
pressive strengths are used, choose the lesser fc′ .
For cases where vuh ≤ φ (500 psi), an intentionally roughed surface is present and at least
minimum transverse reinforcement is provided, the calculation of nominal interface shear
strength in ACI 318-14 [Eq. (21.3.4)] is based on a modified shear friction analogy, where a
minimum shear stress of 260λ psi, equivalent to a cohesion term, is used. The second term
in Eq. (21.3.4) represents the interface shear strength based on a clamping force equal to the
yield strength of the transverse steel and a coefficient of friction µ = 0.6λ. For vuh > φ (500 psi),
on the other hand, the nominal shear strength is based entirely on a shear friction analogy.
Alternative (Segment) Method
Calculation of factored shear stress vuh in the alternative method of ACI 318-14 (ACI-
16.4.5) is based on the difference in resultant normal force between two sections, above
(or below) the interface (Fig. 21.2.3). In this case, the interface shear stress is calculated as
F2 − F1
vuh = (21.3.6)
bv L1−2
where F1 and F2 are the resultant normal forces at sections 1 and 2, respectively, and L1-2 is
the distance between the two sections. In general, sections 1 and 2 correspond to the sections
where M = 0 (inflection point) and M = Mmax. Thus, for simply supported beams subjected
to uniform distributed loading, sections 1 and 2 correspond to the support section and the
midspan section, respectively, while L1-2 is equal to half the span length. When concentrated
forces are present, interface shear stresses should also be checked between sections of maxi
mum moment and adjacent concentrated force, as well as between concentrated forces.
The nominal shear stress capacity vnh in the alternative method is the same as that used in
the sectional method, except for vuh > φ (500 psi), where Eq. (21.3.7) should be used instead
of Eq. (21.3.5). This is because in the segmental method, the stress capacity is equal to the
shear friction strength divided by the interface area between the two sections considered
(i.e., bv L1-2).
µ Av f yt
vnh = (21.3.7)
bv L1-2
where Av in this case is the total area of transverse reinforcement crossing the interface
between sections 1 and 2.
Minimum Area of Interface Transverse Reinforcement

Transverse reinforcement used in the precast member may be extended into the cast-in-
place slab and used also for interface shear transfer. To take advantage of its contribution
904
to interface shear transfer, the area of transverse reinforcement must be at least Av,min for
transverse shear (see Section 5.10) as follows (ACI-16.4.6.1),
bw s 50bw s
min Av = 0.75 fc′ ≥ [5.10.8]
f yt f yt
Transverse reinforcement must be developed in the two elements being connected (e.g.,
precast beam and cast-in-place slab) and spaced no farther than 4 times the minimum
dimension of the supported element, typically the slab thickness, or 24 in., whichever is
smaller (ACI-16.4.7.2).
Flexural and Shear Strength

The sequence of loading and the construction process (shored vs. unshored) does not affect
the calculation of required and nominal flexural and shear strengths in the composite beam.
However, the design of the precast beam must be adequate to resist the loads applied to it
prior to the achievement of composite action.
The flexural and shear strength of concrete composite beams is calculated using the
same principles discussed in Chapters 3 and 5 for monolithic reinforced concrete sections
and in Chapter 20 for prestressed concrete sections. Although a discontinuity in the strain
distribution is expected at the interface between the two materials owing to strains in the
precast beam prior to composite action and interface slip, satisfactory results are obtained
assuming a monolithic concrete section behavior. As discussed earlier, negligible slips have
been reported along intentionally roughened concrete interfaces subjected to interface shear
stresses below approximately 250 psi [21.2, 21.3]. For larger interface shear stresses, even
though some interface slip will occur, flexural and shear strength will not be appreciably
affected as long as the interface shear strength is adequate.
Effective depth d used in the calculation of flexural and shear strength of concrete
composite beams is that corresponding to the same cross section as if it had been cast
monolithically. Typically, the concrete strength in the precast section is greater than the
slab concrete strength. In that case, the lower concrete strength (slab) should be used in the
calculation of positive flexural strength (slab in compression), while for negative flexural
Precast concrete girder, prior to casting of topping slab. Note the girder transverse reinforcement
extending beyond top surface for composite action with cast-in-place slab. (Photo courtesy of Luis
Fargier-Gabaldón.)
905
strength (slab in tension), the precast concrete strength should be used. In most cases, the
depth of the precast beam represents 3 4 or more of the depth of the composite beam. Thus,
the use of the precast concrete strength is often adequate for calculation of shear strength.
As an alternative, a weighted concrete strength, based on the depth of each component,
may be used.
Deflections
Calculation of deflections in concrete composite beams follow the same principles outlined
in Chapter 12. However, some loads are first resisted by the precast beam while others by
the composite member, and this condition requires careful consideration of the loading
sequence, which will depend on whether shored or unshored construction is used. Further,
differential shrinkage between the slab and precast beam will lead to additional deflections,
which are not considered in monolithic construction. The following description of basic
steps for calculating deflections in composite beams consisting of a precast reinforced con-
crete section supporting a cast-in-place slab for both unshored and shored construction is
based on research by Branson [21.9, 21.10]. The same principles can be extended to pre-
cast prestressed concrete beams. Detailed information about calculation of deflections in
composite beams consisting of prestressed concrete sections and cast-in-place slabs can be
found in References 21.9 through 21.11.
Unshored Construction
In unshored construction, the precast beam must be capable of supporting, in addition to
its self-weight, the weight of the cast-in-place slab and other construction-related loads. By
the time the slab concrete has hardened, the deflection in the precast beam will consist of an
instantaneous deflection due to the self-weight of the beam and slab, plus time-dependent
deflections caused by creep and shrinkage. Once the slab concrete has gained sufficient
strength and the beam has become composite, there will be additional instantaneous deflec-
tion due to superimposed dead load, if any, and time-dependent deflections caused by sus-
tained dead load (i.e., precast beam and slab self-weight, plus any additional superimposed
dead load) and differential shrinkage between the precast beam and the slab. Deflections
will also occur as the composite beam is subjected to live load. Since live loads are tran-
sient, however, these deflections are only instantaneous, unless part of the live load may be
considered to be permanent for a significant period of time.
Section 9.3.2.2 of ACI 318-14 specifies that for cases in which both the depth of the
composite beam and the precast section satisfy the minimum depth in ACI-9.3.1 (see
Table 12.15.2), there is no need to check deflections unless the beam supports or is attached
to partitions or other elements likely to be damaged by large deflections. On the other hand,
composite beam deflections must be investigated when the depth of the composite beam
does not satisfy the minimum depth in ACI-9.3.1 or elements likely to be damaged by large
deflections are supported by or attached to the beam. Further, for cases in which only the
composite beam depth satisfies the minimum depth, deflections occurring before the beam
became composite should be investigated.
Basic steps for calculating deflections in unshored construction are outlined below, with
application to the particular case of a simply supported beam of span L consisting of a
precast beam supporting a cast-in-place slab. Attention should be paid to the use of differ-
ent moments of inertia depending on whether the beam is uncracked or cracked and, for
cracked beams, on the maximum moment for the case considered.
1. Instantaneous deflection of precast beam due to self-weight,
5[(wD ) pb ]L4
( ∆ i ) pb -precast = (21.3.8)
384 Ec ( I g ) pb
where (wD)pb is the load corresponding to the self-weight of the precast beam, Ec is
the modulus of elasticity of the precast concrete, and (Ig)pb is the gross moment of
906
inertia of the precast section. If the beam is expected to be cracked under its own
weight, a cracked moment of inertia must be used.
2. Time-dependent deflection of precast beam due to self-weight (up to just before cast-
ing the slab concrete),
( ∆ td ) pb = (λ ∆ ) pb ( ∆ i ) pb -precast (21.3.9)
where (λΔ)pb is a time-dependent factor that accounts for the effect of creep and
shrinkage on deflections (see Section 12.13 and ACI-24.2.4.1.1). The sustained load
duration used in the calculation of (Δtd)pb should be the period between the time the
precast beam is first subjected to its own weight and the time the slab is cast.
3. Instantaneous deflection of precast beam due to slab self-weight, ( ∆ i ) pb-slab . In this
case, it is assumed that the precast beam will be cracked under the action of its own
weight plus the slab self-weight. The effective moment of inertia of the precast beam,
(Ie)pb, can be calculated by using Eq. 21.3.10 below (see Section 12.9) with Ma equal
to the maximum moment due to the weight of the precast beam and slab.
M 
3
 M  3


( I e ) pb =  cr  ( I g ) pb + 1 −  cr   ( I cr ) pb

(21.3.10)
 Ma    M a  
where Mcr and (Icr)pb are the cracking moment and the cracked moment of inertia of
the precast section, respectively. The instantaneous deflection of the precast beam
due to slab self-weight, (wD)slab, is then
5[(wD )slab ]L4
( ∆ i ) pb -slab = (21.3.11)
384 Ec ( I e ) pb
It should be noted that flexural cracking in the precast beam will cause an increase
in the deflection due to its own weight, previously calculated assuming uncracked
behavior [see Eq. 21.3.8]. Thus, the total instantaneous deflection in the cracked
precast beam due to its own weight and the slab self-weight is
5[(wD ) pb + (wD )slab ]L4

( ∆ i ) pb = (21.3.12)
384 Ec ( I e ) pb
4. Instantaneous deflection in composite beam due to additional dead load (i.e., super-
imposed dead load, wSD),
5(wSD )L4
( ∆ i )cb -SD = (21.3.13)
384 Ec ( I e )cb -D
where (Ie)cb-D is the effective moment of inertia of the composite beam when sub-
jected to the total dead load (self-weight plus superimposed dead load). In general,
the effective moment of inertia of the composite beam, (Ie)cb, can be calculated by
assuming a monolithic section with the same dimensions as the transformed compos-
ite beam section. From Eq. (21.3.10), (Ie)cb can be calculated as
M 
3
  M 3
( I e )cb =  cr 
 Ma 
( )
Ig
cb
+ 1 −  cr   ( I cr )cb (21.3.14)
  M a  
where (Ig)cb and (Icr)cb are, respectively, the gross and cracked moment of inertia
of the composite section. For the particular case of (Ie)cb-D, Ma in Eq. (21.3.14) is
equal to the moment due to the self-weight of the composite beam plus superimposed
dead load.
5. Time-dependent deflection in composite beam due to sustained load, ( ∆ td )cb. This
deflection represents the summation of time-dependent deflections due to the precast
907
beam self-weight (in addition to those that took place before the slab was cast), the
slab weight, and any additional permanent load:
 ( I g ) pb 
( ∆ td )cb = [(λ ∆ )cb − (λ ∆ ) pb ] ( ∆ i ) pb -precast 
 ( I e )cb -D 
 ( I e ) pb 
+ (λ ∆ )cb ( ∆ i ) pb -slab + ( ∆ i )cb -SD  (21.3.15)
 ( I e )cb -D 
where (λ ∆ )cb is a time-dependent factor that accounts for the effect of creep and
shrinkage on deflections in the composite beam, which is determined based on the
time period of interest, typically 5 years or longer. In that case, (λ ∆ )cb = 2.0. Note
that in Eq. (21.3.15) a single factor (λ ∆ )cb is used for simplicity, which is acceptable
when sustained loads are applied for several years.
6. Deflection due to differential shrinkage between the slab and the precast beam, ( ∆ td )sh.
Since the concrete in the slab shrinks soon after casting and most of the shrinkage
has already occurred in the precast beam, the precast beam will provide some restrain
against shrinkage of the slab, leading to additional bending (and deflection) of the
composite beam. Different methods have been proposed to account for deflections
induced by differential shrinkage [21.9]. It should be kept in mind, however, that only
rough estimates of shrinkage-induced deflections are possible, given the uncertain-
ties involved. A simple procedure to estimate deflections due to differential shrinkage
consists of applying an equivalent shrinkage-induced force, Q, at the centroid of the
slab (Fig. 21.3.1). For a differential shrinkage strain εsh, the force Q is calculated as
Q = ε sh Ecs Aslab (21.3.16)
where Ecs is the modulus of elasticity of the slab concrete and Aslab is the cross-
sectional area of the slab, calculated as the product of the effective slab width, bE ,
and the slab thickness.
The deflection due to differential shrinkage is then calculated as
Qys L2
( ∆ td )sh = (21.3.17)
8 Ec ( I e )cb -D
where ys is the distance between the slab centroid and the centroid of the composite
(transformed) section.
7. Deflection due to live load. Given the transient nature of live load, the deflection
associated with it is instantaneous unless a portion of the live load can be considered
as permanent for a period of time. The instantaneous deflection due to live load, wL,
is calculated as
5wL L4
( ∆ i )L = (21.3.18)
384 Ec ( I e )cb -L
where ( I e )cb -L is calculated from Eq. (21.3.14) with Ma equal to the moment due to the
total dead load plus live load.
btr
Q Q cgslab
cgtr ys
ys
Figure 21.3.1 Equivalent shrinkage-induced force in composite section.

908
As discussed in Section 12.6, two deflection checks need to be performed, one

involving the instantaneous live load deflection, ( ∆ i )L, and the other all deflections
that occur after attachment of nonstructural elements, ∆ t . Assuming (conservatively)
that all shrinkage- related deformations occur after attachment of nonstructural
elements,
∆ t = ( ∆ i )cb -SD + ( ∆ td )cb + ( ∆ td )sh + ( ∆ i )L (21.3.19)
Limits associated with ( ∆ i )L and ∆ t are listed in ACI Table 24.2.2 and discussed in
Section 12.6.
Shored Construction
In shored construction, all the loads other than the self-weight of the precast beam are
resisted by the composite member. The procedure discussed above for unshored construc-
tion can also be used for shored construction with the exception that Eq. (21.3.11) must be
replaced by Eq. (21.3.20). Also, since the self-weight of the slab is resisted by the compos-
ite beam, Eq. (21.3.12) is not applicable for shored construction.
5[(wD )slab ]L4

( ∆ i )cb -slab = (21.3.20)
384 Ec ( I e )cb -slab
where ( I e )cb-slab is calculated from Eq. (21.3.14) with Ma equal to the moment due to the
self-weight of the precast beam and slab.
There is no need to check deflections in shored construction if the depth of the compos-
ite beam satisfies the minimum depth in ACI-9.3.1 (see Table 12.15.2) and supports or is
attached to elements not likely to be damaged by large deflections.
EXAMPLE 21.3.1
The composite beam shown in Fig. 21.3.2 is part of a floor system that consists of a
series of precast reinforced concrete beams spaced 10 ft apart that support a 5-in.-thick
cast-in-place slab. The beam has a center-to-center span length of 30 ft and is supported
on 12-in.-long supports (i.e., clear span length = 29 ft). The floor system is subjected
to a superimposed dead load of 20 psf and a live load of 60 psf. Consider a 20-psf
live load during construction. Check the adequacy of the design in terms of flexural
and shear strengths, as well as deflections. Assume unshored construction and that the
floor system supports nonstructural elements not likely to be damaged by large deflec-
tions. For the precast beam: fc′ = 6000 psi; fy = fyt = 60 ksi. For the cast-in-place slab:
fcs′ = 3500 psi; fy = 60 ksi. Concrete is normal weight. Cover to center of longitudinal
bars = 2.5 in.
#3 @ 6.5 in. Cast-in-place slab
5”
Precast 16”
2 #9 + 2 #8 (Slab bars not shown)
beam
10”
Figure 21.3.2 Cross section for composite beam of Example 21.3.1.
(Continued)
90
SOLUTION
(a) Determine required flexural and shear strengths.
Precast beam self-weight:
10(16)
(wD ) pb = (0.15) = 0.17 kips/ft
144
Slab weight:
5(120)
(wD )slab = (0.15) = 0.63 kips/ft
144
Superimposed dead load:

wSD = 0.020(10) = 0.20 kips/ft
Construction live load:

(wL )const = 0.020(10) = 0.20 kips/ft
Service live load:

wL = 0.060(10) = 0.60 kips/ft
Factored loads:
- Precast beam:
(wu ) pb = 1.2[(wD ) pb + (wD )slab ] + 1.6(wL )const
= 1.2(0.17 + 0.63) + 1.6(0.20)

= 1.27 kips/ft
- Composite beam:
(wu )cb = 1.2[(wD ) pb + (wD )slab + wSD ] + 1.6 wL
= 1.2(0.17 + 0.63 + 0.20) + 1.6(0.60)

= 2.16 kips/ft
Maximum factored moment (at midspan section):
- Precast beam:
(wu ) pb L2 1.27(30)2
( Mu ) pb = = = 143 ft-kips
8 8
- Composite beam:
(wu )cb L2 2.16(30)2

( Mu )cb = = = 242 ft-kips
8 8
Maximum factored shear (at one effective depth from support face):
Note that the effective depth increases once the slab becomes composite with the
precast beam.
- Precast beam:
L  13.5
(Vu ) pb = (wu ) pb  − d pb  = 1.27(15 − ) = 17.6 kips
2  12
(Continued)
910
- Composite beam:
L   18.5 
(Vu )cb = (wu )cb  − dcb  = 2.16  15 − = 29.0 kips
2   12 
Required flexural and shear strengths. Assume both precast and composite beam sec-
tions are tension controlled (i.e., φ = 0.90). For shear, φ = 0.75.
( Mu ) pb 143
= = 159 ft-kips
φ 0.90
( Mu )cb 242
= = 269 ft-kips
φ 0.90
(Vu ) pb 17.6
= = 23.5 kips
φ 0.75
(Vu )cb 29.0
= = 38.7 kips
φ 0.75
(b) Check flexural strength.

- Precast beam:
As = 2(1.0) + 2(0.79) = 3.58 sq in.

As f y 3.58(60)
a pb = = = 4.21 in.
0.85 fc′ b 0.85(6)(10)
 a pb   4.21 1
(M n ) pb = As f y  d pb − = 3.58(60)  13.5 −  = 204 ft-kips
 2   2  12

 a pb   4.21 
 d pb − β  
13.5 −
ε t = 0.003  1
 = 0.003  0.75  = 0.0042 > 0.004
 a pb  4.21 
 
 β1 
  0.75 
Section is in the transition region. From ACI Table 21.2.2(d), φ = 0.83.

( Mu ) pb 143
( M n ) pb = 204 ft-kips > = = 172 ft-kips OK
φ 0.83
-Composite beam (neglecting any reinforcement in slab):

Effective slab width. According to ACI-6.3.2.1, the effective slab width, bE, for an inte-
rior beam is the least of
bw + 16t = 10 + 16(5) = 90 in.
Ln 29(12)
bw + = 10 + = 97 in.
4 4
center-to-center spacing of beams = 120 in.

where t is the slab thickness and Ln is the clear span length. Thus, bE = 90 in.
(Continued)
91
Assuming rectangular section behavior,
As f y 3.58(60)
acb = = = 0.80 in. < t OK
0.85 fcs′ bE 0.85(3.5)(90)
 a   0.80  1
( M n )cb = As f y  dcb − cb  = 3.58(60)  18.5 −  = 324 ft-kipps
 2   2  12
 a   0.80 
d − cb 18.5 −
 cb β   0.85  = 0.049
ε t = 0.003  1
 = 0.003 
 acb  0.880 
 β   
 0.85 
1
Section is tension controlled. Thus, φ = 0.90.
( Mu )cb 242
( M n )cb = 324 ft-kips > = = 269 ft-kips OK
φ 0.90
(c) Check shear strength.

- Precast beam:
(Vc ) pb = 2 λ fc′bw d pb = 2(1) 6000 (10)(13.5) = 20.9 kips

d pb 0.22(60)(13.5)
(Vs ) pb = Av f yt = = 27.4 kips
s 6.5
(Vu ) pb
(Vn ) pb = (Vc ) pb + (Vs ) pb = 48.3 kips > = 23.5 kips
φ
Because the design is the same throughout the span and satisfies shear requirements near
the support, where shear force is maximum, there is no need to check shear design at
other beam regions.
Check maximum spacing and minimum amount of transverse steel. Since

(Vu ) pb < 4 fc′bw d pb , smax = d pb / 2 = 6.75 in. > s OK
bw s 0.75 6000 (10)(6.5)

min Av = 0.75 fc′ =
f yt 60, 000
50bw s 50(10)(6.5)
= 0.063 sq in. ≥ = = 0.054 sq in.
f yt 60, 000
min Av = 0.063 sq in. < Av = 0.22 sq in. OK
- Composite beam:
It should be noted that in this case (Vn)pb > (Vu)cb /φ and thus, no further check is neces-
sary. For illustration purposes, however, the procedure will be shown here.
As discussed earlier, calculation of shear strength in composite beams with concretes
of different compressive strengths may be performed using a weighted strength based on
(Continued)
912
the precast and composite beam depths. The use of such an approach in this case would
lead to a weighted compressive strength of approximately 5300 psi. Use 5000 psi.
(Vc )cb = 2 λ fc′bw dcb = 2(1) 5000 (10)(18.5) = 26.2 kips
dcb 0.22(60)(18.5)
(Vs )cb = Av f yt = = 37.6 kips
s 6.5
(Vu )cb
(Vn )cb = (Vc )cb + (Vs )cb = 63.7 kips > = 38.7 kips
φ
Because the design is the same throughout the span and satisfies shear requirements near
the support, where shear force is maximum, there is no need to check shear design at
other beam regions. Because, further,

(Vu )cb < 4 fc′bw dcb , smax = dcb / 2 = 9.25 sq in. > s OK
Amount of transverse steel satisfies minimum required, as checked above for precast
section.
d) Check interface shear.
-Sectional method:
Vu 29.0(1000)
vuh = = = 156 psi < φ (500 psi)
bv dcb 10(18.5)
 0.6 Av f yt   0.6(0.22)(60, 000) 

vnh = λ  260 +  = 1 260 +  = 382 psi
 bv s   10(6.5) 

φvnh = 286 psi > vuh OK
-Segment method:
In this case, the interface shear force is calculated between the support (M = 0) and
midspan (M = Mmax). Because the entire concrete compressive stress block falls within
the beam flange, the horizontal force at midspan, F2, in Eq. (21.3.6) is equal to the yield
strength of the longitudinal steel (As fy) while F1 = 0. Thus,
F2 − F1 As f y 3.58(60, 000)
vuh = = = = 119 psi < φvnh OK
bv L1- 2 L (30)(12)
bv 10
2 2
e) Check deflections.
-Instantaneous deflection of precast beam due to self-weight:
Check first whether the precast beam is uncracked or cracked under its own weight.
fr bw (h)2pb
( M cr ) pb =
6
(10)(16)2 1
= 7.5(1) 6000 = 20.7 ft-kipss
6 12,000
(Continued)
913
(wD ) pb L2
( M a ) pb = = 18.8 ft-kips < ( M cr ) pb Beam is uncracked.
8
bw (h)3pb 10(16)3
( I g ) pb = = = 3413 in.4
12 12
57, 000 6000
Ec = 57, 000 fc′ = = 4415 ksi
1000
 
5(0.17)  1  (30 × 12)4
5(wD ) pb L4  12 
(∆ i ) pb--precast = = = 0.20 in.
384 Ec I g 384(4415)(3413)
-Time-dependent deflection of precast beam due to self-weight (up to time of beam

becoming composite):
Assuming a period of 3 months between the time the precast beam is subjected to its
own weight and the time the beam becomes composite, (λ ∆ ) pb = 1.0 (ACI-24.2.4.1). Thus,
( ∆ td ) pb = (λ ∆ ) pb ( ∆ i ) pb -precast = 1.0(0.20) = 0.20 in.
-Instantaneous deflection of precast beam due to slab self-weight:

The precast beam will experience flexural cracking once it is subjected to the slab
weight. From Eq. (21.3.10), the effective moment of inertia is calculated as
M 
3
 M  3

( I e ) pb =  cr  ( I g ) pb + 1 −  cr   ( I cr ) pb
 Ma    M a  
where
(wD ) pb + (wD )slab  L2 (0.17 + 0.63)(30)2

Ma =  = = 89.1 ft-kips
8 8
The cracked moment of inertia of the precast beam section, (Icr)pb, can be obtained by
following the procedure in Section 12.5. The neutral axis depth c and cracked moment
of inertia are obtained as follows:
c2
bw = ηs As (d pb − c) = 0
2
where ηs is the reinforcing steel-precast concrete modular ratio (ηs = 6.57). Solving for
c leads to c = 5.96 in.
2
bw c 3  c
( I cr ) pb = + bw c   + ηs As (d pb − c)2
12  2
The cracked and effective moments of inertia for the precast section are then
(10)(5.96)3 10(5.96)3
( I cr ) pb = + + 6.57(3.58)(13.5 − 5.96)2 = 2043 in.4
12 4
(Continued)
914
and
 20.7 
3
  20.7  3

( I e ) pb =   (3413) + 1 −    (2043) = 2060 in.
4
 89.1    89.1  
The instantaneous deflection due to the slab weight is then
5[(wD )slab ]L4

(∆ i ) pb -slab =
384 Ec ( I e ) pb
 
5(0.63)  1  (30 × 122)4
 12 
=
384(4415)(2060)
= 1.25 in.
Because the beam exhibits flexural cracking when subjected to the weight of the slab, the
deflection of the precast beam due to its own weight will increase. From Eq. (21.3.12), the
total instantaneous deflection of the precast beam due to its own weight and that of the slab is
5[(wD ) pb + (wD )slab ]L4

(∆ i ) pb =
384 Ec ( I e ) pb
 
5(0.17 + 0.63)  1  (30 × 12)4
 12 
=
384(4415)(2060)

= 1.59 in.
-Instantaneous deflection in composite beam due to superimposed dead load. Because
the precast beam is already cracked, the effective moment of inertia of the compos-
ite beam under the action of dead load, ( I e )cb -D , needs to be calculated. This requires
first the calculation of the properties of the transformed section (both uncracked and
cracked) as follows.
Effective slab width of transformed section:
(57, 000 3500 )

(bE )tr = bE ηc = 90 = 68.7 in.
4415 × 1000
For uncracked transformed section, it can be shown that the centroid is located at
ytr = 14.68 in., measured from the bottom of the section. The transformed moment of
inertia and cracking moment are
( I tr )cb = 19, 240 in.4 ≈ ( I g )cb
fr ( I tr )cb (19, 240)

( M cr )cb = = 7.5(1) 6000 = 63.4 ftt-kips
ytr 14.7(12,000)
The moment of inertia for the transformed cracked section for use in Eq. (21.3.14) can
be determined following the same procedure as for the precast section. The neutral axis
depth and crack moment of inertia for the cracked transformed composite section are
c = 3.23 in. (ytr = 17.77 in.)
( I cr )cb = 6255 in.4
(Continued)
915
The acting moment Ma to be used in Eq. (21.3.14) is

[(wD ) pb + (wD )slab + wSD ]L2
Ma =
8
(0.17 + 0.63 + 0.20)(30)2
=
8

= 112 ft-kips
From Eq. (21.3.14),
 63.4 
3
  63.4  3 
( I e )cb -D = (19, 240) + 1 −    (6255)
 112    112  

= 8644 in.4
The instantaneous deflection of the composite beam due to superimposed dead load is then
5(wSD )L4
(∆ i )cb -SD =
384 Ec ( I e )cb -D
 
5(0.20)  1  (30 × 12)4
 12 
=
384(4415)(8644)

= 0.096 in.
-
Time-dependent deflection of composite beam due to sustained load. From Eq.
(21.3.15) and assuming a period of 5 years or more [i.e., (λ ∆ )cb = 2.0.],

( ∆ td )cb = 0.87 in.
-Deflection due to differential shrinkage between the slab and the precast beam.
Assuming a shrinkage strain εsh = 400 × 10–6,
1
Q = ε sh Ecs Aslab = 400(10)−6 (57, 000 3500 )(5)(90) = 607 kips
1000
Qys L2
( ∆ td )sh =
8 Ec ( I e )cb -D
 5 
607  21 − − 17.8 (30 × 12)2
 2 
=
8(4415)(8644)
= 0.19 in.
- Instantaneous deflection due to live load. From Eq. (21.3.14) with Ma corresponding to
the moment caused by the total dead load plus live load,
[(wD ) pb + (wD )slab + wSD + wL ] L2
Ma =
8
(0.17 + 0.63 + 0.20 + 0.60)(30)2
=
8
= 179 ft-kips
 63.4 
3
  63.4  3 
( I e )cb -D + L =  (19, 240 ) + 1 −    (6255) = 6833 in.
4
 179    179 
(Continued)
916
5(wL )L4
(∆ i )L =
384 Ec ( I e )cb -D + L
 
5(0.6)  1  (30 × 12)4
 12 
=
384(4415)(6833)
= 0.36 in.
Deflection checks:
Live load deflection
L
(∆ i )L = 0.36 in.< = 1.0 in.
360 OK
Total deflection occurring after attachment of nonstructural elements,
L
∆ t = ( ∆ i )cb -SD + ( ∆ td )cb + ( ∆ td )sh + ( ∆ i )L = 1.51 in. ≈
240

= 1.5 in. OK, but right at the limit
21.4 CONCRETE-S TEEL COMPOSITE COLUMNS

Columns that combine structural steel shapes and concrete have been used for over
100 years. Early in the twentieth century, concrete around structural steel was often used
for fireproofing, and thus its contribution to column strength and stiffness was generally
neglected. According to Eggemann [21.12], however, the structural advantages of encasing
a metal structural section in reinforced concrete had been reported by Fritz von Emperger as
early as 1915, based on research conducted for over 10 years and followed by applications
of his proposed composite column design in the United States and Europe around 1930.
The concept of combining structural steel and concrete in columns was already rec-
ognized in the 1910 Standard Building Regulations for the Use of Reinforced Concrete
[21.13], although it was not until the 1920 ACI Building Code [21.14] that more general
design provisions were available for columns consisting of concrete-encased steel sections
as well as of concrete encased by steel. A summary of the evolution of design provisions for
composite columns can be found in References 21.15 and 21.16. Although the 2014 ACI
Building Code includes design provisions for composite columns, these are fairly limited in
scope and have remained basically unchanged in the past few editions. On the other hand,
design provisions for composite columns in the AISC Specification for Structural Steel
Buildings (21.17) have undergone significant changes based on research in the past two
decades, becoming the main design document for composite columns in the United States.
Types of Concrete-Steel Composite Columns

Composite columns can be lumped in two major categories: concrete-encased steel col-
umns and concrete-filled steel tube columns. As shown in Fig. 21.4.1(a), concrete-encased
steel columns consist of a structural steel shape surrounded by reinforced concrete that
includes both longitudinal and transverse reinforcement. These columns are commonly
referred to as SRC columns. Concrete-filled tube (CFT) columns, as the name implies, are
structural hollow steel sections (HSS), either circular or rectangular, filled with concrete
[see Fig. 21.4.1(b)]. In some cases, longitudinal steel is added to the concrete core for
increased strength (or to ensure adequate strength under fire conditions) and reduction of
creep-induced deformations.
917
21.4 CONCRETE-STEEL COMPOSITE COLUMNS 917
Headed anchor
Ties Longitudinal bars Hollow steel section
Concrete
Steel
shape
(a) Concrete-encased steel (b) Concrete-filled tube

(SRC) section (CFT) section
Figure 21.4.1 Types of composite columns.
Concrete-encased steel composite columns have several advantages with respect to either
steel columns or reinforced concrete columns. The surrounding reinforced concrete con-
tributes significantly to column strength and stiffness, and provides lateral support to the
encased steel shape, enhancing column stability and ductility, as well as providing fire pro-
tection. The steel section, on the other hand, greatly increases column strength and ductility.
Although concrete- encased steel composite columns have been used in Japan for
decades, their use in the United States has been restricted primarily to high-rise structures,
where the use of composite columns offers significant advantages from the construction
and cost viewpoints. As reported by Griffis [21.18], the use of a steel section allows rapid
erection of a steel frame, several stories ahead of the reinforced concrete encasing and thus,
the spread of different construction tasks over several stories.
In the case of concrete-filled tube columns, the steel section serves not only as reinforce-
ment but also as formwork, which greatly facilitates the erection of these columns. Also,
when circular sections are used, the steel tube offers substantial confinement to the concrete
core once it cracks, increasing its ductility and strength. The concrete core, on the other
hand, increases both the stability of the steel shape (by preventing inward buckling) and the
overall stability of the column (by significantly increasing column stiffness).
Design Strength of Composite Columns

As mentioned above, the AISC Specification for Structural Steel Buildings [21.17] cov-
ers more comprehensively the design of composite columns than ACI 318. Thus, most of
the provisions referred to in this and following sections are from the AISC publication.
Pertinent sections in this document will be referred to as AISC, followed by the section
number. Background information on the calibration of the provisions for composite col-
umns in AISC can be found in Reference 21.26.
The design strength of composite columns in the AISC Specification [21.17] is the prod-
uct of the nominal strength and the corresponding strength reduction factor, as for rein-
forced concrete columns in ACI 318. The AISC strength reduction factors that apply to
composite columns, however, are independent of section strains, which makes their deter-
mination substantially simpler than in ACI 318. Further, unlike the design of reinforced
concrete sections in ACI 318, different strength reduction factors are specified in AISC for
axial force and bending. The strength reduction factors in AISC are as follows:
For axial compression: φc = 0.75 (AISC-I2.1b; AISC-I2.2b)

For axial tension: φt = 0.90 (AISC-I2.1c; AISC-I2.2c)
For bending: φb = 0.90 (AISC-I3.3; AISC-I3.4b)
For sectional shear (compact rolled I-shaped sections): φv = 1.0 [AISC-G2.1(a)]; for
CFT columns: φv = 0.90 (AISC-G1), and where a contribution of the concrete and/or bar-
type transverse reinforcement is considered: φv = 0.75 (AISC-I4.1).
For shear transfer between steel and concrete through headed stud anchors: φv = 0.65
(AISC-I8.3a).
For bond between steel and concrete in concrete-filled tubes, φ = 0.50 (AISC-I6.3c).
918
Concrete-encased steel composite column. (Photo courtesy of Tiziano Perea.)
21.5 CONCRETE-E NCASED STEEL COMPOSITE

COLUMNS
The behavior of concrete-encased steel composite columns has been experimentally inves-
tigated for decades (see, e.g., Refs. [21.19–21.25]. Although bond conditions between the
encased steel shape and the surrounding concrete differ from those between deformed steel
bars and concrete, results from evaluations [21.22, 21.24] of different bonding conditions
between the steel shape and the surrounding concrete indicated little effect on ultimate
strength under either monotonic or reversed cyclic loading. Further, the use of a strain com-
patibility approach has been shown to be adequate for calculating the strength of concrete-
encased steel composite columns subjected to combined axial force and bending [21.22,
21.24, 21.25]. The strength of concrete-encased steel composite columns under either axial
force or combined axial force and bending can thus be calculated by following the same
principles discussed in Chapter 10.
Reinforcement Limits
According to AISC-I2.1a, the steel core of a concrete-encased steel column must represent
a minimum of 1% of the gross area of the column, while the ratio of longitudinal reinforc-
ing bar area to the gross column area must be at least 0.4%. ACI-10.6.1.2, on the other
hand, requires a minimum area of longitudinal steel bars equal to 1% of the concrete area,
taken as the gross area minus the area of the encased steel shape. Given that the minimum
area of longitudinal steel bars of 1% is meant to prevent yielding due to creep of longitu-
dinal reinforcement under service conditions (see Section 10.10), the use of a combined
longitudinal reinforcement ratio of 1.4%, with a 0.4% minimum percentage of bar area
should be adequate to account for the reduced bond between the steel shape and surround-
ing concrete in comparison to that between deformed bars and concrete. To ensure adequate
bond between the reinforcing bars and the surrounding concrete, a minimum clear distance
between the encased steel shape and the longitudinal bars of 1.5 bar diameters or 1.5 in.,
whichever is greater, is required (AISC-I2.1e).
91
21.5 CONCRETE-ENCASED STEEL COMPOSITE COLUMNS 919
The same provisions for lateral ties and spiral reinforcement discussed in Sections 10.8
and 10.9, respectively, apply to concrete-encased steel sections. In addition, the following
minimum requirements apply [AISC-I2.1a(b)]:
• #3 ties at 12-in. spacing or #4 ties at 16-in. spacing

• Tie spacing not greater than half the least cross-sectional dimension
Although specific provisions in ACI 318 for composite columns are exempted when
applying the AISC Specification, the writers recommend the application of ACI-10.7.6.1.4,
which requires that the tie diameter be not less than 0.02 the maximum column-side dimen-
sion, but not smaller than #3, and need not be taken larger than #5.
Material Strength Limits
In terms of material properties, AISC-I1.3 requires a minimum specified concrete strength
fc′ of 3000 psi. For strength calculation purposes, fc′ cannot exceed 10,000 psi for normal-
weight concrete and 6000 psi for lightweight concrete. However, higher concrete strengths
may be used in the calculation of stiffness. Finally, the specified yield strength of the struc-
tural steel shape and reinforcing bars used in strength calculations cannot exceed 75 and
80 ksi, respectively. In most cases, these limits are due to lack of experimental data that
support the use of higher strength values.
Axial and Flexural Strength (Short Columns)

The calculation of compressive strength of concrete-encased steel composite columns in
AISC requires the computation of Euler’s elastic buckling load, Pe, in order to account for
slenderness effects. For simplicity, the axial and bending strength of short columns, Pns and
Mn, respectively, will be discussed first. Column nominal compressive strength, in the pres-
ence or absence of bending, will then be adjusted to account for slenderness effects. Since
nominal bending strength is not adjusted for slenderness effects, Mn, rather than Mns, will
be used for short columns.
The compressive strength of a concrete-encased steel composite column under pure
axial load is calculated as for a reinforced concrete column [Eq. (10.5.1)], adding a term to
account for the axial strength of the steel shape as follows,
Pnso = 0.85 fc′( Ag − As − Asr ) + As f ys + Asr f yr (21.5.1)
where As is the cross sectional area of the encased steel shape, Asr is the area of longitu-
dinal bars, and fsy and fysr are the yield strength of the steel shape and reinforcing bars,
respectively. As discussed in Chapter 10, the stress of 0.85 fc′ corresponds to the assumed
compressive strength of concrete in the column. At the point at which the concrete reaches
its peak stress, both the reinforcing bars and steel shape are assumed to yield. Note that to
avoid confusion between the two types of reinforcement used, this chapter’s notation is
slightly different from that in Chapter 10.
In the presence of combined axial compression and bending, AISC-I1.2 allows the use of
four methods: plastic stress distribution, strain compatibility, elastic stress distribution, and
effective stress-strain. The strain compatibility method is the same as that used for reinforced
concrete columns (see Sections 10.13 and 10.14) and will thus be followed here. The only
difference with respect to reinforced concrete columns is the presence of the steel shape,
which typically has different yield strength than the reinforcing bars. The steel shape, how-
ever, can be divided into several equivalent bars representing the section flanges and web.
Each point in the Pns–Mn interaction diagram other than that corresponding to the pure
axial load condition is determined based on a linear strain distribution with a maximum com-
pressive strain of 0.003. The concrete compression zone can be modeled using the Whitney
rectangular stress block or any other suitable stress distribution, as in reinforced concrete
sections. The behavior of the steel components (i.e., bars and encased section), on the other
hand, is assumed to be linear elastic–perfectly plastic (i.e., strain hardening is neglected).
Figure 21.5.1 shows the strain and stress condition for an arbitrary point in the Pns–Mn
interaction diagram. For this particular strain condition, the extreme layer of tension
920
εcu = 0.003 0.85 fc’

εsr1 Fsr1
β1c fstf
c εstf Cc Fstf
d Fsr2 Fswc
h εsr2
Fswt
εsy
εsbf
Fsr3
εsry fsbf = fsy Fsbf
εsr3
b Strains Stresses/forces Stresses/forces

(reinforced concrete) (steel section)
Figure 21.5.1 Strains, stresses, and forces for calculation of a point in Pns–Mn diagram for
concrete-encased steel composite columns.
reinforcing bars has experienced yielding, as has the bottom portion of the steel shape. The
axial force Pns corresponds to the net axial force in the section, while the moment Mn is typ-
ically calculated with respect to the plastic centroid of the section. The detailed procedure
for calculating a point in the interaction diagram is illustrated at the end of this section (see
Example 21.5.1). Other points in the Pns–Mn diagram can be calculated by either reducing
or increasing the neutral axis depth c.
As discussed in Section 10.6, the point in the interaction diagram of a reinforced con-
crete section corresponding to the balanced condition (i.e., compression-controlled limit) is
that in which the strain in the extreme compressive fiber is equal to 0.003, while the strain
in the extreme tensile layer of steel is equal to the yield strain of the reinforcing steel. In
concrete-encased steel columns, however, two balanced conditions exist, one correspond-
ing to first yielding of the extreme layer of tension steel bars (as in reinforced concrete),
and the other corresponding to first yielding of the extreme tension fiber of the encased
steel shape. In general, these two points are close to each other because the strain diagram
is nearly the same for both conditions.
Slenderness Effects
The design of composite columns in AISC includes the use of an axial compressive strength
reduction factor originally developed for steel columns to account for initial imperfections
and residual stresses. Whether such reduction should be applied to concrete-encased steel
composite columns is debatable, however, because the effect of initial imperfections in the
encased steel section should be significantly reduced by the presence of the reinforced con-
crete encasing. This approach, on the other hand, allows uniform treatment of composite
columns and bare steel columns.
The axial compressive strength reduction factor to account for initial imperfections, κ, is
based on the ratio between the pure axial compressive strength of a short column, Pnso, and
the elastic buckling load, Pe. For cases in which Pnso /Pe ≤ 2.25, which covers most typical
concrete-encased steel composite columns,
Pnso
κ = 0.658 Pe
(21.5.2)
where
π 2 EI eff
Pe = (21.5.3)
( Lc ) 2
EIeff in Eq. (21.5.3) is the effective flexural rigidity of the composite section, while Lc is the
effective column unsupported length, equal to kL, where L is the unsupported length of the
column and k depends on the column boundary conditions. For cases in which the required
strength is obtained through a second-order analysis that satisfies the requirements of the
Direct Analysis Method of AISC-C2, k should be taken as 1.0, unless a lower value is jus-
tified by analysis (AISC-C3).
921
There has been extensive work on the development of expressions for estimating the
effective flexural rigidity EIeff (see, e.g., Refs. 21.26–21.28). It has been reported [21.27,
21.28] that the effective flexural rigidity depends not only on the flexural rigidity of the
individual components (i.e., concrete, steel bars and steel shape), but also on the axial load
eccentricity and column slenderness. Accounting for eccentricity in the calculation of EIeff,
however, could be impractical given the number of load combinations that must be accounted
for. Thus, a simplified expression has been adopted in AISC [AISC-Eq. (I2-6)] as follows
EI eff = Es I s + Es I sr + C1 Ec I c (21.5.4)
where
 A + Asr 
C1 = 0.25 + 3  s  ≤ 0.70 (21.5.5)
 Ag 
In Eqs. (21.5.4) and (21.5.5),

Es = modulus of elasticity of the steel
Ec = modulus of elasticity of the concrete
Is = moment of inertia of the steel shape with respect to the elastic centroid of the com-
posite section
Isr = moment of inertia of the steel bars with respect to the elastic centroid of the com-
posite section
Ic = moment of inertia of the concrete section with respect to the elastic centroid of the
composite section
As = cross-sectional area of steel shape
Ac = cross-sectional area of concrete
Asr = cross-sectional area of reinforcing bars
Contrary to design provisions in ACI 318, AISC does not account for the effect of sus-
tained load on stiffness. It could be argued, however, that creep in concrete-encased steel
columns should be less critical than in reinforced concrete columns because composite
columns typically have a much higher reinforcement ratio due to the encased steel shape.
Once EIeff has been determined, Euler’s buckling load Pe [Eq. (21.5.3)] and the slender-
ness strength reduction factor κ [Eq. (21.5.2)] can be calculated. The nominal axial strength
for columns with Pnso /Pe ≤ 2.25, in the presence or absence of moment, is then
Pn = κ Pns (21.5.6)
For the particular case of pure axial load, Pn = Pno and Pns = Pnso.
Axial Tensile Strength

The strength of a concrete-encased steel composite column under pure tension is calculated
as the summation of the yield strength of the steel shape and longitudinal bars as follows.
Pnt = f y As + f ysr Asr (21.5.7)
Shear Strength
Three major contributions to the shear strength of concrete-encased steel composite col-
umns can be identified: steel shape; concrete; and steel ties.
The encased steel shape often represents the main contributor to the shear strength of
composite columns. Its contribution comes primarily from shear yielding of the web (if
web is parallel to the direction of shear) or flanges (if flanges are parallel to the direction
of shear). Concrete and steel ties, on the other hand, also contribute to the shear strength
of a concrete-encased composite column much as in a reinforced concrete column (see
Chapter 5). The concrete contribution consists of shear resisted in the member compression
zone, aggregate interlock, and dowel action, while the steel ties contribute to shear strength
through the activation of a truss mechanism.
92
The shear strength of a concrete-encased steel composite column, however, cannot

be taken as the summation of these three mechanisms, which would imply that all three
can provide their assumed contributions simultaneously, which is not likely to be true.
Unfortunately, experimental data on the ultimate shear strength of concrete-encased steel
composite columns is scarce. From tests of composite columns under reversed cyclic load-
ing, Ricles and Paboojian [21.24] indicated that a truss mechanism contributing to shear
resistance does not develop in concrete-encased steel composite columns because of the
large shear stiffness of the encased steel shape. They thus suggested that the shear strength
be taken as that of the encased steel shape.
Given the uncertainty about the ultimate shear strength of concrete-encased steel compos-
ite columns, AISC has taken the approach of assuming that only two of three mechanisms are
additive. According to AISC-I4, the nominal shear strength of a concrete-encased steel com-
posite section can be taken as either: (1) the shear strength of the encased steel section; (2) the
shear strength of the encased steel section plus the shear strength contributed by the lateral
ties through truss action; or (3) the shear strength of the reinforced concrete portion. As dis-
cussed earlier, different shear strength reduction factors φv apply, depending on the strength
option selected (see Section 21.4, subsection on Design Strength of Composite Columns).
Shear Strength of Encased Steel Section

In AISC-G2.1, the nominal shear strength of the concrete-encased steel section is calcu-
lated assuming shear yielding of an effective shear area Aw. As shown in Fig. 21.5.2, for
cases in which direction of shear is parallel to the web, Aw corresponds to the area of the
web. For cases in which the direction of shear is parallel to the flanges, Aw is taken as the
combined area of the flanges. Thus, for direction of shear parallel to web,
Vn = 0.60 f ys Aw Cv = 0.60 f ys (t w ds )Cv (21.5.8)
For direction of shear parallel to flanges;
Vn = 0.60 f ys Aw Cv = 0.60 f ys (2t f b f )Cv (21.5.9)
where 0.60fy is the yield shear strength of steel, tw is the web thickness, ds is the overall depth
of the steel shape, and bf and tf are the width and thickness of the steel flanges, respectively.
The web shear coefficient Cv is meant to account for potential shear buckling. In general,
Cv depends on the slenderness of the web and is equal to 1.0 for most commercially available
wide-flange sections. Since, however, this coefficient was originally intended to apply to the
design of bare steel sections, it is believed that the lateral support provided by the surrounding
concrete is sufficient to ensure shear yielding of the web (or flanges): that is, Cv = 1.0.
The applicable strength reduction factor when the shear strength of a composite column
is taken as that of the bare steel section is φv = 1.0.
bwE/2 bf bwE/2 bwE/2 ds bwE/2
Direction of
shear
Aw
Aw
h h
b b
Figure 21.5.2 Effective shear area Aw and effective web width bwE.

923
Shear Strength of Steel Section and Lateral Ties

When shear strength is assumed to be provided by the steel section and lateral ties, the
shear strength of the composite section is calculated as the summation of Vn in Eq. (21.5.8)
or (21.5.9), whichever is applicable, and the contribution of the ties, Vs. The applicable
strength reduction factor is that used for shear in ACI 318 (φv = 0.75).
The contribution of the lateral ties to shear strength, which is calculated using Eq.
(5.10.7) [(ACI-22.5.10.5.3)], is not to exceed the limit in ACI-22.5.1.2 as follows,
Av f yt d
Vs = ≤ 8 fc′bwE d (21.5.10)
s
It should be noted that an effective web width bwE (Fig. 21.5.2) is used in Eq. (21.5.10).
This practice has been recommended [21.29] to account for the formation of the truss mech-
anism outside the width of the steel shape. As shown in Fig. 21.5.2, for cases in which the
direction of shear is parallel to the steel section web, bwE is equal to the column width minus
the flange width (b – bf). On the other hand, when the direction of shear is parallel to the
flanges, bwE is equal to the column width minus the overall depth of the steel section (b – ds).
The area of tie reinforcement, Av, should also satisfy the minimum requirement dis-
cussed in Section 10.21 for reinforced concrete columns. Transverse reinforcement spacing
s must also be checked against the maximum allowed for shear as follows:
• For Vs ≤ 4 fc′bwE d , s ≤ d / 2 ≤ 24 in.

• For Vs > 4 fc′bwE d , s ≤ d / 4 ≤ 12 in.
Shear Strength of Reinforced Concrete Portion

The nominal shear strength of the reinforced concrete portion of the concrete-encased com-
posite column can be calculated following the procedure in Section 10.21 and Chapter 5
as the summation of the contribution of the concrete, Vc, and the lateral ties, Vs , as follows,
Vn = Vc + Vs [5.8.1]
Accounting for the influence of axial compression on shear strength, Vc can be calculated
by using Eq. (5.13.1), given in ACI-22.5.6.1.
 Nu 

Vc = 2  1 +  λ fc′bw d [5.13.1]
 2000 Ag 
The contribution of the lateral ties is calculated using Eq. (21.5.10). As discussed in the
preceding section, the area of tie reinforcement must satisfy the minimum requirement
for reinforced concrete columns (Section 10.21), and the selected tie spacing shall not be
greater than either d/2 or d/4, depending on the shear force required to be resisted by the
transverse reinforcement through a truss mechanism.
The applicable shear strength reduction factor when the shear strength of the composite
column is taken as that of the reinforced concrete component is the same φ factor used for
shear in ACI 318 (φv = 0.75).
Load Transfer
In the discussion above on determination of axial and bending strength of concrete-
encased steel composite columns, it was assumed that load was introduced to the rein-
forced concrete portion and encased steel shape based on their relative contributions to
column strength. In reality, however, load is often introduced into the composite column
through the steel section. This is the case, for example, of a gravity load frame consist-
ing of composite columns and steel girder-composite deck floor systems, where gravity
load is introduced into the column through a shear connection between the steel floor
girder and the encased steel shape (Fig. 21.5.3). Another example corresponds to the
924
case of a composite column that is part of a composite braced frame system in which
the braces are made of steel. Part of the vertical component of the force in the braces,
transferred into the steel column as an axial force, must then be transferred to the sur-
rounding reinforced concrete portion.
Attention must therefore be paid to ensuring that the load introduced to either the
embedded steel shape or the concrete is transferred to the other component according to
its relative contribution to column strength. Otherwise, it may be unconservative to assume
that both components of the composite column contribute to resist the applied loads as cal-
culated using either strain compatibility or a plastic stress distribution.
The discussion that follows will focus on cases in which the load is introduced through
the encased steel shape, as this is most common in practice. Design of the connections
(bolted, welded, or a combination of both) between the steel components is outside the
scope of this book. The reader is referred to References 21.1 and 21.30 for information
about design of connections between steel components.
Sharing of Axial Load between Reinforced Concrete and Steel Shape Components

The relative contribution of the encased steel section and the reinforced concrete encasing
to nominal axial column strength varies depending on column axial load. A reasonable esti-
mation, however, can be made by assuming that their relative strength contribution is inde-
pendent of column axial strength and equal to that corresponding to the case of pure axial
load. This is the principle followed in the AISC Specification for determining the magnitude
of the axial force required to be transferred, through longitudinal shear, from the steel shape
to the surrounding reinforced concrete (AISC-I6.2a). Taking the axial load transferred to the
steel column via a floor girder (Fig. 21.5.3) or diagonal braces as Pr , the required longitudi-
nal shear to be transferred to the surrounding reinforced concrete, Vr′ , is [AISC Eq. (I6-1)]
 As f y 
Vr′ = Pr  1 − (21.5.11)
 Pnso 
where Pnso is obtained from Eq. (21.5.1).
Longitudinal Shear Transfer

Transfer of forces between the embedded steel shape and the surrounding concrete is
achieved through shear connectors attached to the steel shape (e.g., steel headed stud anchors
or channels) or steel plates bearing on concrete. The following discussion focuses on the
use of headed studs for longitudinal shear transfer, as these are typically used for shear
transfer. Provisions for steel channel anchors can be found in AISC-I8.3d and AISC-I8.3e.
Provisions for determining the shear strength of headed stud anchors welded to the
encased steel shape are given in AISC-I8.3a. When concrete breakout failures do not con-
trol, as is typically the case in concrete-encased steel composite columns, the shear strength
of a single stud anchor, Qnv, is calculated as
Qnv = fua Asa (21.5.12)
where fua and Asa are the specified minimum tensile strength and cross-sectional area,
respectively, of the steel headed stud anchor. The applicable strength reduction factor for
headed stud anchors, φv, is equal to 0.65. The required number of headed stud anchors
required to transfer the axial force Vr′ is calculated from Eq. (21.5.13).
Vr′
na ≥ (21.5.13)
φ v Qnv
where na is the number of headed stud anchors. The minimum required stud length Lsa,
measured from the base of the stud to the head, is 5 times the stud shank diameter (AISC-
I8.3). Longer studs are specified when these are required to resist tension and shear, as well
as in members constructed with lightweight concrete, which would be a rare case for com-
posite columns. The procedure above is illustrated later (see Example 21.5.2).
925
In general, moments due to eccentricity of the girder shear force transferred into the
column do not affect the design of the headed studs, either because girders from opposite
faces frame into the column and moments cancel out, as in Example 21.5.2, or because the
magnitude of the moment is small enough to be neglected. At times, however, the moment
transferred into the composite column due to the eccentricity of shear is not negligible.
In this case, the number of headed studs must be calculated such that these are capable
of transferring not only the force Vr′ , but the additional force caused by the moment to be
transferred from the encased steel shape to the reinforced concrete encasing. This will be
illustrated in Example 21.5.3.
Placement of Headed Stud Anchors

Headed stud anchors must be placed on at least two faces of the encased steel shape in
a symmetrical arrangement about the axis of the shape (AISC-I6.4a). The studs must be
placed over a distance referred to as the load introduction length, which is to be taken no
greater than twice the minimum section dimension of the composite column above and
below the region where load is transferred (Fig. 21.5.3).
In some cases, however, the magnitude of the force Vr′ is such that it cannot be trans-
ferred over the load introduction length. In this situation, the column should be treated as
noncomposite over the length required to transfer the force Vr′ (AISC Commentary I6.4).
In terms of headed stud anchor placement, AISC-I8.3e requires a minimum spacing
between stud anchors, in any direction, of 4 times the diameter of the shank, dsa, and a
maximum spacing of 32dsa. AISC-I8.3e also refers to ACI 318 for minimum clear cover
for headed stud anchors which, according to ACI-20.6.1.3.5, is the same as that for any
other reinforcement in the member. For cast-in-place columns not exposed to weather or in
contact with ground, the minimum cover is then 1.5 in. (ACI-Table 20.6.1.3.1). The writers
recommend, however, that in addition to satisfying this minimum cover, the head of the
stud be located within the core of the column.
Even though the flexural strength of concrete-encased steel composite columns under
combined axial force and bending has been shown not to be affected by the presence of
headed studs [21.24], AISC-I6.4a requires the placement of headed studs outside the load
introduction length to help maintain composite action in the column. As discussed above,
stud spacing must not exceed 32 times the dimeter of the stud shank, dsa.
≤ 32dsa
Headed stud
≤ 2b ≥ 4dsa
≤ 32dsa
V = Pr’
Load introduction
length
Steel beam
≤ 2b Shear tab
Encased steel shape
b: minimum column section dimension; reinforcing bars

in column and concrete slab not shown for clarity.
Figure 21.5.3 Load introduction length and stud spacing requirements in concrete-encased steel
composite columns.
926
EXAMPLE 21.5.1
Verify that the design of the concrete- encased steel composite column shown in
Fig. 21.5.4 is adequate according to the AISC Specification for Structural Steel Buildings
to resist the following factored loads obtained from a second-order analysis satisfying the
requirements of the Direct Analysis Method of AISC (not including the effect of initial
imperfections): Pu = 935 kips; Mu = 1100 ft-kips (about strong axis of steel shape); Vu =
130 kips. Unsupported column length for bending about strong axis L = 12 ft. (column
is laterally supported in perpendicular direction). For the W14 × 132 shape: ds = 14.7 in.;
As = 38.8 sq in.; tw = 0.645 in.; bf = 14.7 in.; tf = 1.03 in.; Is= 1530 in.4. Material proper-
ties: fc′ = 6000 psi (normal weight); fys = 50 ksi (A572 Grade 50 steel); fyr = fyt = 60 ksi.
#4 @ 10” 12 #7
2.5”
4”
24” 11”
4”
2.5”
W14 × 132
24”
Figure 21.5.4 Cross section of concrete-encased steel composite column of Example 21.5.1.
SOLUTION
(a) Check that reinforcement ratios satisfy minimum requirements.
Longitudinal bars:

Asr = 12(0.6) = 7.2 sq in. > 0.004 Ag = 0.004(24)(24) = 2.3 sq in. OK
Steel shape:
As = 38.8 sq in. > 0.01Ag = 0.01(24)(24) = 5.76 sq in. OK
(b) Check axial force and moment strength. The determination of the nominal axial
strength of the column requires first the calculation of the equivalent flexural rigidity
EIeff, Euler’s buckling load Pe, and axial strength reduction factor κ.
The effective flexural rigidity, EIeff, is calculated using Eq. (21.5.4), with the variables
calculated as follows,
Es = 29, 000 ksi

I s = 1530 in.4
I sr = 2[ 4(0.60)(12 − 2.5)2 + 2(0.60)(12 − 6.5)2 ] = 506 in.4
24(24)3
Ic = − 1530 − 506 = 25, 612 in.4
12
 38.8 + 12(0.6) 
C1 = 0.25 + 3   = 0.49 < 0.70
 (24)(24) 
57, 000 6000
Ec = = 4415 ksi
1000
(Continued)
927
Thus,
EI eff = (29, 000)(1530 + 506) + 0.49(4415)(25, 612)

= 1.14 × 108 kip-sq in.
Since Pu, Mu have been obtained from the Direct Analysis Method, k may be taken as
1.0 and
π 2 EI eff π 2 (1.14 × 108 )

Pe = = = 54, 450 kips
( Lc ) 2 [1(12)(12)]2
The pure axial load capacity of the column neglecting slenderness effects, Pnso, is
obtained using Eq. (21.5.1).
Pnso = 0.85 fc′( Ag − As − Asr ) + As f ys + Asr fsr

= 0.85(6)[(24)(24) − 38.8 − 12(0.60)] + 38.8(50) + 12(0.60)(60)

= 5075 kips
Pnso /Pe = 0.093 < 2.25. The axial strength reduction factor κ is then

Pnso
 5075 
Pe  
κ = 0.658 = 0.658 54, 450  = 0.96
The adequacy of the column section design will be checked by calculating the design
moment strength, φb Mn, corresponding to a design axial strength φc Pn = φc κPns = Pu. In
other words, the design will be satisfactory if Mn ≥ Mu /φb at an axial force Pns = Pu /(φc κ).
For this particular case, the neutral axis depth, c, is equal to 12.9 in. The resultant forces
in the concrete and steel reinforcement are as follows (Fig. 21.5.5):

Cc = 0.85 fc′ bβ1c = 0.85(6)(24)(0.75)(12.9) = 1184 kips
The resultant axial force is then be determined as,

i=4
Pns = Cc + ∑ Fsri + Fstf + Fsbf + Fsw
i =1
= 1184 + 1.0 + 680 + 36.3 − 606

= 1295 kips
TABLE 21.5.1 STRAINS, STRESSES AND FORCES IN REINFORCING BARS

OF EXAMPLE 21.5.1
Layer i Asr,i (di) (yi) εsr,i fsri (ksi) Fsri (h/2 –di)

(sq in.) (in.) (in.) (kips) (in.)
a a
1 2.4 2.5 10.4 0.00242 60–5.1 132 9.5
a a
2 1.2 6.5 6.40 0.00149 43.2–5.1 45.7 5.5
3 1.2 17.5 –4.60 –0.00107 –31.0 –37.2 –5.5
4 2.4 21.5 –8.60 –0.00200 –58.0 –139 –9.5
a Stress/force adjusted to account for displaced concrete. Tensile strains, stresses and forces are negative.
(Continued)
928
TABLE 21.5.2 STRAINS, STRESSES AND FORCES IN STEEL SHAPE

OF EXAMPLE 21.5.1
As (y) εs fs (ksi) Fs (kips) (h/2 –di )

(sq in.) (in.) (in.)
a a
Top flange 15.1 7.74 0.00180 50–5.1 680 6.83
a
Web (top) 7.23 0.00168 48.7–5.1 a
8.15 36.3 —b
Web (bottom) –5.42 –0.00126 –36.6
Bottom flange 15.1 –5.93 –0.00138 –40.0 –606 –6.83
a Stress/force adjusted to account for displaced concrete. Tensile strains, stresses, and forces are negative.
b In this example, the web has been modeled as a continuous steel element, not as discrete equivalent bars. Thus, a single
lever arm cannot be used to calculate its contribution to the section moment capacity. This is discussed below. See also
Fig. 21.5.6.
εcu = 0.003 0.85 f c’ = 5.1 ksi
di εsr1 Fsr1 = 132 k

β1c = fstf = 50 ksi
εstf Cc = 1184 k
c = 12.9” 9.68” Fstf = 680 k
εsr2 Fsr2 = 45.7 k
Msw = 56 ft-k
24” y
Fsw = 36.3 k
Fsr3 = 37.2 k
εsr3 Fsbf = 606 k
εsbf Fsr4 = 139 k
fsbf = 40 ksi
εsr4
24” Stresses/forces Stresses/forces

Strains
(reinforced concrete) (steel section)
Figure 21.5.5 Strains, stresses, and forces for calculation of Pns–Mn in Example 21.5.1.
Taking into account slenderness effects due to initial imperfections, the nominal axial
strength is calculated as

Pn = κ Pns = 0.96(1295) = 1246 kips
The design axial strength is then

φc Pn = 0.75(1246) = 935 kips = Pu
The corresponding nominal flexural strength can be calculated by taking moments about
the column plastic centroid (i.e., column center) of the various forces. The moment
con-
tribution of the concrete, reinforcing bars and steel shape, Mc, Msr, and Ms, respectively,
are calculated as follows:
 h β c  24 0.75(12.9)  1
M c = Cc  − 1  = 1184  −  12 = 707 ft-kipss
2 2  2 2 
i=4
h 
M sr = ∑ Fsri  − di  = 253 ft-kips
i =1  2 
(Continued)
92
εcu = 0.003
0.85 fc’
dtf = 5.17”
εstf β1c =
c =12.9” 9.68”
Fstf = 680 k
dbf =
17.84” Msw = 56 ft-k
24”
Fsw = 36.3 k
fs(c,z)
dz Fsbf = 606 k
εsbf z
Strains Stresses Forces/moments

24”
Figure 21.5.6 Stresses, resultant forces, and moments for steel section in Example 21.5.1.
The moment contribution from the encased steel section is divided into the contributions
from the section flanges, Msf , and that from the web, Msw (Fig. 21.5.6)
h  h 
M sf = Ftf  − dtf  + Fbf  − dbf  = 732 ft-kips
2  2 
h + ds
−t f  h
M sw = ∫ h −2ds fs (c, z )tt w  z −  dz = 56.3 ft-kips
2
+t f  2
M s = M sf + M sw = 732 + 56.3 = 788 ft-kips
where z is the vertical distance from the base of the section to the point on the web being
evaluated (Fig. 21.5.6), and fs(c, z) is the stress at that point, which depends on the cur-
vature (and thus, neutral axis depth c) and the distance between the point on the web
being considered and the neutral axis [z – (h – c)]. For web locations within the concrete
compressive stress block, the stress in the web is reduced by 0.85 fc′ .
The total flexural strength for the design axial strength φPn = 935 kips is then
M c = M c + M sr + M s
= 707 + 253 + 788
= 1748 ft-kips

φb M n = 1573 ft-kips > Mu = 1100 ft-kips
The design is thus adequate to resist the factored axial force and moment.
(c) Check shear capacity. For illustration purposes, the nominal shear strength of the
composite column will be calculated following the three options given in AISC.
-Steel section alone:
Vn = 0.60 f ys Aw Cv
= 0.60(50)(0.645)(14.7)(1.0)
= 284 kips
φ vVn = (1.0)(284) = 284 kips >Vu = 130 kips (Continued)
930
-Steel section plus lateral ties:

Av f yt d
Vn = 0.60 f ys Aw Cv +
s
0.20(2 + 2 )(60)(0.8)(24)
Vn = 284 +
10
= 284 + 78.7 kips
= 363 kips
Note that a factor equal to 2 / 2 is applied to the area of shear reinforcement provided by
the octagonal shaped hoop to account for the portions near the column corners oriented
at 45° from the direction of shear. Also, d was taken as 0.8h (see Section 10.21).
Verify that Vs is less than upper limit.
8 6000 (24 − 14.7)(19.2)

Vs < 8 fc′bwE d = = 111 kips OK
1000
Thus,
φ vVn = (0.75)(363) = 272 kips > Vu = 130 kips
Verify that the selected reinforcement satisfies minimum shear reinforcement require-
ment. From Eq. (5.10.8) and noting that 0.75 fc′ > 50 psi for fc′ = 6000 psi,
0.75 fc′ 0.75 6000

min Av =
= bw s = 0.23 sq in. < 0.20(2 + 2 ) = 0.68 sq in. OK
f yt 60, 000
Verify maximum tie spacing. Using an effective web width bwE = (24 – 14.7) = 9.3 in.,
the tie shear contribution Vs corresponds to 5.7 fc′bwE d. This leads to a maximum tie
spacing of d/4 = 4.8 in., which is less than the spacing provided (10 in.). According to
this procedure, the shear design would thus be inadequate.
-Reinforced concrete section:
 Nu 
Vc = 2  1 +  λ fc′bw d
 2000 Ag 
 935, 000  1
= 2 1 + (1.0) 6000 (24)(19.2)
 2000(24)(24)  1000

= 129 kips
Av f yt d 0.20(2 + 2 )(60)(19.2)
Vs = = = 79.8 kips
s 10
Vs < 8 fc′bwE d OK

φ vVn = φ v (Vc + Vs ) = 0.75(144 + 79.8) = 168 kips > Vu = 130 kips
As for the case of shear strength assumed to be provided by the steel section and tie
reinforcement, the area of tie reinforcement satisfies the minimum required. However,
tie spacing is limited to d/4, which is smaller than the spacing provided.
In this example, the shear design is adequate according to the first procedure. The other
two procedures, on the other hand, would require a reduction in tie spacing. Given that the
shear strength of the bare steel section is adequate, the shear design is considered acceptable.
(Continued)
931
(d) Check minimum tie requirements for lateral support of longitudinal bars. ACI-
10.7.6.1.2 and ACI-25.7.2 (see Section 10.8) require the following.
– Minimum #3 size tie for #10 or smaller longitudinal bars. Because #4 ties are used
with #7 longitudinal bars in this case, this requirement is satisfied.
– Lateral tie spacing shall not exceed 16 longitudinal bar diameters, 48 tie bar diam-
eters, or the least dimension of the column, which correspond to 14 in., 24 in., and 24
in., respectively. Thus, the tie spacing of 10 in. satisfies this requirement.
– Every corner and alternate longitudinal bar shall have lateral support provided by the
corner of a tie having an included angle of not more than 135° and no bar shall be
farther than 6 in. clear on either side from such a laterally supported bar. All longi-
tudinal bars are laterally supported by the corner of a tie having an included angle of
not more than 135°. This requirement is therefore satisfied.
In addition to the requirements in ACI-10.7.6.1.2 and ACI-25.7.2, the following mini-
mum requirements apply (AISC-I2.1b):
– #3 ties at 12-in. spacing or #4 ties at 16-in. spacing. As #4 ties at 10 in. spacing are
provided, this requirement is satisfied.
– Tie spacing not greater than half the least cross-sectional dimension. The provided
spacing of 10 in. is less than half the least cross-sectional dimension (12 in.)
It is also recommended, based on ACI-10.7.6.1.4, that the tie diameter be not less than
0.02 times the maximum column side dimension, but not smaller than #3. This provision
would require a minimum diameter of 0.48 in., which is slightly less than the tie diameter
(0.5 in.).
EXAMPLE 21.5.2
Two W14 × 53 floor girders, framing into opposite faces of the composite column of
Example 21.5.1, are shear-connected to the flanges of the encased steel shape (see
Fig. 21.5.7). If the factored shear at the end of the floor girders is 55 kips, determine the
number and spacing of headed stud anchors required to transfer the corresponding force
Vr′ to the reinforced concrete encasing. Determine also the required stud spacing outside
of the load introduction length. Tensile strength of studs, fua = 65 ksi.
8.5”
5”
Vu = 55 k Vu = 55 k
W14 × 132 W14 × 53
Shear tab
24”
Reinforcing bars in column and concrete slab

not shown for clarity.
Figure 21.5.7 Connection between steel beam and composite column of Example 21.5.2.
(Continued)
932
SOLUTION
(a) Determine force to be transferred from encased steel shape to reinforced concrete
encasing, Vr′ . From Eq. (21.5.11) and Example 21.5.1,
 As f y   (38.8)(50) 
Vr′ = Pr  1 − = 55(2) 1 − = 68.0 kipps
 Pnso   5075 
(b) Determine diameter and required number of headed stud anchors to transfer force
Vr′ . As required by AISC-I8.3e, which refers to ACI 318-14, at least a 1.5-in. clear
cover to the outside edge of the stud must be provided (assuming cast-in-place col-
umn not in contact with ground; ACI-20.6.1.3.5). Further, as discussed above, it is
desirable to keep the head of the studs within the core of the column. Assuming a
clear cover of 1.5 in. to the column ties, the maximum possible headed stud length is
[(24–14.7 in.) /2–1.5 in.] = 3.15 in. The maximum stud diameter for a minimum stud
length of 5dsa (AISC-I8.3) is then 3.15 in./5 = 0.63 in. Use 5 8 -in.-diameter headed
stud anchors with a length Lsa = 5dsa = 3 1 8 in.
The shear strength of a single stud is calculated form Eq. (21.5.12), where Asa = 0.31 sq in.
and fua = 65 ksi.
Qnv = fua Asa = 65(0.31) = 20.2 kips
The required number of shear studs is calculated from Eq. (21.5.13).
Vr′ 68.0
ηa ≥ = = 5.2 (Use 6 studs minimum)
φ v Qnv 0.65(20.2)
A minimum of three studs on each flange of the encased steel shape is thus required to
transfer the force Vr′ . To maintain symmetry about the beam axis also, use four studs on
each flange, two above and two below the shear tab.
(c) Headed stud layout. As only eight studs are required over the load introduction length
(not to exceed 96 in. or twice the minimum column cross-sectional dimension above
and below the shear tab), four above and four below the shear tab, maximum spacing
considerations will govern in this case. According to AISC-I8.3e, headed stud spacing
should not exceed 32dsa = 24 in. This same requirement applies to the column regions
outside the load introduction length. The final headed stud layout is shown in Fig. 21.5.8.
48”
20”
Load introduction
length ≤ 96”
20”
48”
Headed stud anchors at
20 in. spacing (dsa = 5/8”, Lsa =
3 1/8”)
24”
Reinforcing bars in column and concrete slab not shown for clarity.
Figure 21.5.8 Headed stud layout in composite column of Example 21.5.2.

93
EXAMPLE 21.5.3
Determine the number and layout of headed studs in the connection of Example 21.5.2
assuming factored shear forces of 65 and 45 kips for the left and right girder, respec-
tively (Fig. 21.5.9).
Vr’/2 + Mva/ds Vr’/2 – Mva/ds
Vu = 65 k Vu = 45 k
8.5” ds 8.5”
Reinforcing bars in column and concrete slab not

shown for clarity.
Figure 21.5.9 Beam-column connection and transfer of forces between steel shape and concrete
encasing for Example 21.5.3.
SOLUTION
(a) Headed studs required to transfer force Vr′ . The total axial force introduced into the
column, Pr , and Vr′ , are the same as in Example 21.5.2 (110 kips and 68 kips, respec-
tively). Thus, six studs are required for transferring the force Vr′ .
(b) Determine number of studs required to transfer moment from encased steel shape to
reinforced concrete encasing. The total moment transferred from the steel girders to
the encased steel section, Mv, is
 14.7 
M v = (65 − 45)  + 8.5 = 317 in.-kips
 2 
The fraction of Mv to be transferred to the reinforced concrete encasing, Mva, can be cal-
culated assuming the same proportion as for axial load strength between the steel shape
and reinforced concrete encasing. Thus,
 As f y   (38.8)(50) 
M va = M v  1 − = 317 1 − = 196 in.-kiips
 Pnso   5075 
The transfer of a moment Mva, counterclockwise, will increase the force in the studs
on the left flange of the encased shape while decreasing the force in the studs on the
right flange (Fig. 21.5.9). It is recommended, however, that the same number of studs
be added to both flanges to maintain symmetry as well as to eliminate the possibility
of placement errors. The required number of additional studs on the left flange of the
encased steel shape, (na)Mva, are
 M va   196 
 d   
14.7 
(na ) Mva ≥
s
= = 1.0
φ v Qnv 0.65(20.2)
(Continued)
934
A minimum of four studs on each flange (two above and two below the shear tab) is
required within the load introduction length. The total number of studs within the load
introduction length is therefore 8, two more than the minimum required to transfer the
force Vr′ , while the number of studs in each flange is 4, which satisfies the required num-
ber of studs on the left flange to transfer both Vr′ and the moment Mva (2.6 studs for Vr′
and 1 stud for Mva). The headed stud layout shown in Fig. 21.5.8 satisfies both strength
and spacing requirements.
21.6 CONCRETE-F ILLED TUBE COLUMNS

The ease of construction of concrete-filled steel tube (CFT) columns has made these
composite members a very attractive column design alternative. CFT columns have been
increasingly used in the past four decades, finding applications in both low-rise and high-
rise construction [21.31]. As interest by structural designers and contractors in field appli-
cations of CFT columns grew, so did the need for research to understand their behavior
under various loading conditions. Research on CFT columns covers a wide variety of areas,
including strength, stiffness, concrete-steel tube bond, stability, and long-term behavior
(see, e.g., Refs. 21.32–21.49). For an excellent summary on research work conducted on
CFT columns in the past several decades, the reader is referred to Gourley et al. [21.50].
From tests of square and circular CFT columns, Furlong [21.32, 21.33] indicated that
the infill concrete and the steel tube tend to behave independently under loading and that
adhesion between the concrete and the steel does not prevent relative sliding between the
two materials. As reported in Roeder et al. [21.45], however, mechanical bond caused by
interior roughness of the steel tube will develop in typical applications, but at a decreas-
ing rate with an increase in the ratio between the tube diameter D (for circular tubes) and
thickness t. Further, as the compressive stress in the concrete approaches its peak strength,
lateral expansion of the concrete will induce lateral pressure on the tube walls, creating
additional friction at the concrete-steel interface and confinement of the concrete by the
steel tube. As noted by Knowles and Park [21.34], however, such an expansion can develop
only if the steel tube has not buckled by the time the stress in the concrete approaches its
compressive strength.
The friction that develops between the infilled concrete and surrounding steel tube seems
to be sufficient to allow the use of a strain compatibility approach to calculate the strength
of CFT columns under combined axial force and bending, as allowed in AISC-I1.2b. Even
though this approach is based on the assumption of perfect bond between the concrete and
steel tube, its suitability has been demonstrated through experimental research, where the
strength of CFT short columns subjected to combined axial force and bending has been
calculated with reasonable accuracy by means of strain compatibility [21.32, 21.35]. For
consistency with the treatment of reinforced concrete columns in Chapter 10, as well as
concrete-encased steel composite columns earlier in this chapter, the strain compatibility
approach will be used for calculation of strength of CFT columns under combined axial
force and bending.
Reinforcement Limits
In CFT columns, the minimum cross-sectional area of the steel tube must be at least 1%
of the gross area of the column (AISC-I2.2a). In typical CFT columns, no longitudinal
reinforcing bars are used, and no minimum is required in the AISC Specification. When
longitudinal reinforcement is used, as in CFT columns of high-rise structures, a minimum
area equal to 0.4% of the gross area of the column is recommended, as for concrete-encased
steel composite sections.
935
21.6 CONCRETE-FILLED TUBE COLUMNS 935
Material Strength Limits
In the AISC Specification for Structural Steel Buildings, the same concrete strength limi-
tations for concrete-encased steel composite columns apply to CFT columns (AISC-I1.3).
Specified concrete compressive strength fc′ must be at least 3000 psi. For strength calcula-
tion purposes, fc′ is limited to 10,000 psi and 6000 psi for normal-weight and lightweight
concrete, respectively. No upper limit in concrete compressive strength is specified for
stiffness calculations. The specified yield strength of the steel section and reinforcing bars
used for strength calculations cannot exceed 75 ksi and 80 ksi, respectively.
Design Wall Thickness and Limits on Tube Width-to-Thickness Ratio

The design wall thickness t for CFT sections is generally smaller than the nominal tube thick-
ness. Section AISC-B4.2 requires that the design wall thickness be taken as 0.93 times the
nominal wall thickness for tube sections, except those produced according to ASTM A1065
or A1085, for which the design thickness may be taken equal to the nominal thickness.
The behavior of a tube section under compression and/or bending is dependent on the
stability of the tube walls. Tube sections with thick walls relative to their width are likely
to be able to yield in compression without buckling, while thin-walled tubes are likely to
exhibit buckling prior to yielding. The AISC Specification for Structural Steel Buildings
thus establishes width-to-thickness limits for purposes of classifying the walls of CFT
members in terms of local buckling as compact, noncompact, or slender (AISC-I1.4).
These limits are given in Table 21.6.1. For rectangular sections, the limits apply to the
wall width-to-thickness (b/t) ratio, where the wall width is taken as the clear distance
between the perpendicular tube walls (e.g. clear distance between flanges when calculat-
ing web width) minus the inside-corner radius on each side. For circular sections, these
apply to the diameter-to-thickness (D/t) ratio. A compact section is expected to be able to
undergo significant yielding prior to local buckling. Noncompact sections, although able
to undergo compression yielding prior to buckling, are not expected to provide adequate
confinement to the concrete at large compressive stresses. Slender sections will undergo
local buckling without any yielding.
The limits in Table 21.6.1 are higher than those applicable to hollow steel sections
(HSS). This is because in hollow steel members, the tube walls may buckle inward and
outward. In CFT members, on the other hand, the buckled shape is different because no
TABLE 21.6.1 WIDTH-TO-THICKNESS RATIO LIMITS (b/t OR D/t) FOR CFT MEMBERS
Element Compact Noncompact Slender Maximum limit
Members Walls of
Es Es E Es Es
subjected rectangular ≤ 2.26 2.26 to 3.00 s > 3.00 5.00
to axial sections f ys f ys f ys f ys f ys
compression Es Es E Es Es
Circular ≤ 0.15 0.15 to 0.19 s > 0.19 0.31
sections f ys f ys f ys f ys f ys
Members Flanges of Es Es E Es Es
subjected to rectangular ≤ 2.26 2.26 to 3.00 s > 3.00 5.00
f ys f ys f ys f ys f ys
bending sections
Webs of Es Es E Es
rectangular ≤ 3.00 3.00 to 5.70 s — 5.70
f ys f ys f ys f ys
sections
Circular Es Es E Es
≤ 0.09 0.09 to 0.31 s — 0.31
sections f ys f ys f ys f ys
936
inward displacement is possible, which increases the stability of the tube walls. The vast
majority of the commercially available steel hollow sections made out of ASTM A500
steel in the United States qualify as compact sections. It should also be mentioned that by
satisfying the limits for compact sections in Table 21.6.1, the required minimum area of
steel of 1% of the gross column area is also satisfied. The discussion that follows will thus
focus on compact sections.
Axial Compressive Strength (Short Columns–Compact Sections)

As for concrete-encased steel composite columns, the strength of short CFT columns is first
discussed, followed by slenderness effects.
The strength of a compact CFT short column subjected to pure axial load is calculated
using Eq. (21.6.1), which is the same as Eq. (21.5.1), but using a generic concrete stress
equal to C2 fc′ in order to account for an increase in concrete compressive strength due to
lateral confinement in circular tubes.
Pnso = C2 fc′( Ag − As − Asr ) + As f ys + Asr f yr (21.6.1)
where C2 = 0.85 for rectangular sections and 0.95 for circular sections. For CFT sections
without steel reinforcing bars, Asr = 0. It should be noted that Eq. (21.6.1) differs from
Eq. (I2-9b) in AISC-I2.2b in the treatment of the contribution from longitudinal reinforcing
bars, where the reinforcing bar stress is taken equal to (C2 fc′ )Es /Ec. This wrongly implies
that the concrete exhibits linear elastic behavior up to its peak stress. Assuming such elas-
tic behavior leads to a significant underestimation of the strain in both the concrete and
longitudinal steel at peak concrete stress. The writers thus disagree with this approach and
recommend the use of Eq. (21.6.1) instead.
Combined Axial Compression and Bending

(Short Columns–Compact Sections)
The strength of CFT sections under combined axial force and bending can be calculated
by using strain compatibility, a plastic stress distribution, an elastic stress distribution,
or an effective stress-strain method (AISC-I1.2). As mentioned above, good agreement
has been obtained between experimental and calculated strengths when strain compat-
ibility was used [21.32, 21.35]. Based on this experience, and for consistency with the
treatment of reinforced concrete columns, the strain compatibility approach is used here
to illustrate the calculation of the strength of CFT sections under combined axial force
and bending.
The strain and stress distribution at nominal condition for a circular CFT column section
subjected to combined axial force and bending is shown in Fig. 21.6.1. The steel tube can
be modeled as a series of discrete equivalent reinforcing bars or as a continuous element.
The latter has been chosen in the calculation of forces below. The stress intensity in the
concrete compression block is generally taken as 0.85 fc′ . In circular sections analyzed by
using the plastic stress distribution method of AISC-I1.2a, a stress intensity of 0.95 fc′ is
allowed because of the concrete confinement provided by the steel tube. This implies that
such an increase in stress is also allowed when strain compatibility is used. For uniformity
of treatment in the analysis of circular and rectangular sections, however, a single concrete
stress intensity of 0.85 fc′ will be used. The procedure can be easily modified to account for
the increased concrete stress intensity in circular sections.
The calculation of the resultant concrete compressive force requires the determination of
the concrete core area within a depth a from the extreme concrete compressive fiber, Acomp,
as discussed in Section 10.18 (see Figures 10.18.1 and 10.18.2). Note that the diameter of
the concrete core Dc = D - 2t, rather than the outside tube diameter D, should be used in
937
0.85fc’ *
εcu = 0.003
Location of Cc fys
εsy
Acomp c a = β1(c–t) Cc
θs yc
α
D
t
Dc/2
dθs εsy
R
εs,max f ys
Cross section Strains Stresses Stresses

(concrete) (steel)
* A stress intensity of 0.95fc’ is allowed in circular sections.
Figure 21.6.1 Strain and stress distribution at nominal for circular CFT column section.
the determination of the concrete area in compression. The resultant concrete compressive
force and its moment about the column centroid are
Cc = 0.85 fc′Acomp (21.6.2)
M c = Cc yc (21.6.3)
The strain εs and stress fs at any point on the steel tube defined by the angle θs in Figure 21.6.1
can be determined as follows:
ε cu
ε s (θ s ) = [c + R(cos θ s − 1)] (21.6.4)
c−t
− f ys ≤ fs (θ s ) = ε s (θ s ) Es ≤ f ys (21.6.5)
The resultant axial force in the steel tube and the contribution of the steel tube to the section
moment capacity are
π
Fs = 2∫ fs (θ s )tR dθ s (21.6.6)
0
π
M s = 2∫ fs (θ s )tR( R cos θ s )dθ s (21.6.7)
0
The nominal axial strength and moment in the composite section are then obtained as the
summation of the contributions from the steel tube and the concrete core.
Pns = Cc + Fs (21.6.8)
Mn = Mc + Ms (21.6.9)
Calculation of the nominal axial strength and moment in a rectangular CFT section is eas-
ier. In that case, the same procedure used for rectangular concrete-encased steel sections
can be followed, with the tube walls parallel to the axis of bending representing the flanges
and the other two tube walls being the web.
When longitudinal bars are used in the core of the CFT column, the contribution of the
reinforcing bars to axial force and moment strength can be determined as for a reinforced
concrete section.
938
Slenderness Effects
The effect of initial imperfections in CFT columns on axial compressive strength is
accounted for in the same manner as for concrete-encased steel composite columns.
However, a different expression must be used for the calculation of effective flexural rigid-
ity (AISC-I2.2b):
EI eff = Es I s + Es I sr + C3 Ec I c (21.6.10)
where
 A + Asr 
C3 = 0.45 + 3  s  ≤ 0.90 (21.6.11)
 Ag 
On comparing Eq. (21.6.11) with Eq. (21.5.5), it can be seen that a higher contribution of
the concrete to the section flexural rigidity is used in the case of CFT columns because of
the confinement provided by the steel tube. Once the effective flexural rigidity has been
determined, the elastic buckling load Pe is calculated by using Eq. (21.5.3), and the strength
reduction factor due to initial imperfections, κ, is determined (for compact CFT sections
with Pnso /Pe ≤ 2.25) by using Eq. (21.5.2).
As was the case for concrete-encased steel composite sections, AISC does not account for
the effect of sustained load on the stiffness of CFT columns. Creep strains in CFT columns
substantially lower than those in reinforced concrete columns, as well as small shrinkage-
induced strains, have been reported [21.48]. Further, a negligible increase in steel strain due
to concrete creep was reported in circular CFT columns under sustained loading [21.49].
Axial Tensile Strength

The strength of a CFT column under pure tension is calculated as the summation of the
yield strength of the steel shape and longitudinal bars (see Eq. 21.5.7).
Shear Strength
Shear rarely controls the design of a CFT section. Thus, the shear strength of CFT column
sections is typically conservatively taken as that of the bare steel section, as specified in
AISC-I4.1(a). Although AISC-I4.1 also allows the calculation of the shear strength of a
CFT column as that of the reinforced concrete portion or as the summation of the shear
strength of the steel section and the reinforcing steel, these two options are often not appro-
priate, since traditional bar-type shear reinforcement is not typically used in CFT columns,
even when longitudinal bars are present.
The nominal shear strength of a rectangular CFT section is calculated based on Section
AISC-G4 as follows,
Vn = 0.60 f ys Aw Cv (21.6.12)
where Cv = 1.0 for all compact sections. The effective shear area Aw = 2ht, where h is the
length of the straight portion of the tube walls parallel to the direction of shear, and t is the
design thickness. The distance h can be calculated as the clear distance between flanges minus
2 times the inside corner radius of the tube. If the corner radius is not known, h is taken as the
overall depth of the section in the direction of shear minus 3 times the wall thickness.
For compact circular sections, the nominal shear strength is calculated as (AISC-G5)
Vn = 0.60 f ys Aw (21.6.13)
where Aw is taken as As /2.

The applicable shear strength reduction factor φv for CFT sections is 0.90.
93
EXAMPLE 21.6.1
A composite column consisting of an HSS20 × 0.50 steel section filled with concrete is
subjected to the following factored loads obtained from a second-order analysis satisfying
the requirement of the Direct Design Method (not including the effect of initial imperfec-
tions): Pu = 735 kips; Mu = 680 ft-kips; Vu = 85 kips. Unsupported column length L = 12 ft.
For the HSS20 × 0.50 steel shape: D = 20.0 in.; As = 28.5 sq in.; t = 0.465 in.; Is= 1360 in.4.
Material properties: fc′ = 8000 psi (normal weight); fys = 42 ksi (A500 Grade B steel).
Verify that the concrete-filled tube column is adequate to resist the factored loads
according to the AISC Specification for Structural Steel Buildings.
SOLUTION
(a) Check compactness criterion. From Table 21.6.1,
D 20.0 E
= = 43.0 < 0.09 s = 62.1
t 0.465 Fys
Section is compact for both compression and bending.

(b) Check minimum area of steel. As discussed earlier, all compact sections satisfy the
minimum steel area requirement of 1% of the column area. For this example,
As 28.5
= = 0.09 > 0.01 OK
Ag π(20)2
4
(c) Check axial force and moment strength. The determination of the nominal axial
strength of the column requires first the calculation of the equivalent flexural rigidity
EIeff , Euler’s buckling load, Pe, and axial strength reduction factor, κ.
The effective flexural rigidity, EIeff, is calculated using Eqs. (21.6.10) and (21.6.11), with
the variables calculated as follows:
Es = 29, 000 ksi

I s = 1360 in.4
π Dc4 π [20 − 2(0.465)]4
Ic = = = 6492 in.4
64 64
 
 28.5 
C3 = 0.45 + 3   = 0.72 < 0.90
 π (20)
2

 4 
1
Ec = 57, 000 8000 = 5098 ksi
1000
Thus,

EI eff = (29, 000)(1530) + 0.72(5098)(6492) = 6.34 × 10 7 kip-sq in.
and
π 2 EI eff π 2 (6.34 × 10 7 )
Pe = = = 30,180 kips
(kL )2 [1(12)(12)]2
where k has been taken as 1.0.

(Continued)
940
The pure axial load capacity neglecting slenderness effects, Pnso, is obtained using
Eq. (21.6.1) with Asr = 0.
Pnso = C2 fc′( A g − As ) + As f ys
 π [20 − 2(0.465)]2 
= 0.95(8)   + 28.5(42)
 4

= 3368 kips
Pnso /Pe = 0.11 < 2.25. The axial strength reduction factor κ is then,
Pnso  3368 
Pe  
κ = 0.658 = 0.658 30,180  = 0.95
As in Example 21.5.1, the adequacy of the column section design will be checked by

calculating the design moment strength, φb Mn, corresponding to a design axial strength
φc Pn = φc κPnso = Pu. The strain and stress distribution corresponding to φc Pn = 735 kips is
shown in Fig. 21.6.2. As can be seen, the maximum compressive strain in the concrete is
taken as 0.003, which means that the strain in the extreme compression fiber of the steel
tube is slightly larger than 0.003. For this condition, the neutral axis depth, measured
from the extreme tube compression fiber is c = 12.85 in.
Dc = 19.07” 0.85 fc’ = 6.8 ksi
εcu = 0.003
42 ksi
Acomp =115 in.2 4.68”

εsy a = 8.05”
c = 12.85” Cc = 779 k
D = 20” yc = 4.86”
α = 1.414 rad
εsy = 0.00145
42 ksi
t = 0.465”
Cross section Strains Stresses Stresses
(concrete) (steel)
Figure 21.6.2 Strains and stresses for calculation of Pns–Mn in Example 21.6.1.
The resultant concrete compressive force Cc can be determined by using Figure 10.18.1

with h = Dc and a depth of compressed concrete ηh = β1(c – t) = 8.05 in., where β1 = 0.65.
 α − sin α cos α 
Acomp = Dc2  
 4 
where
 D / 2 − β1 (c − t ) 
α = cos −1  c 
 Dc / 2 

 9.54 − 0.65(12.85 − 0.465) 
−1
= cos   = 81.0 º = 1.414 rad
 9.54 
Thus,
1.414 − sin(1.414) cos(1.414) 
Acomp = (19.07)2   = 115 sq in.
 4 

Cc = 0.85 fc′Acomp = 0.85(8)(115) = 779 kips
(Continued)
941
The contribution of the resultant concrete compressive force to the section moment
capacity is [Eq. (21.6.3)],

M c = Cc yc
From Fig. 10.18.2,
Dc3 (sin α )3 (19.07)3 (0.988)3

yc = = = 4.86 in.
12 Acomp 12(115)
Thus,
 4.86 
M c = Cc yc = 779  = 316 ft-kips
 12 
The resultant axial force in the steel tube and its contribution to the moment capacity of
the section are calculated using Eqs. (21.6.6) and (21.6.7).
π π
Fs = 2∫ fs (θ s ) t R dθ s = Dt ∫ fs (θ s ) dθ s = 246 kips
0 0
π π
D2 t
M s = 2∫ fs (θ s )t R( R cos θ s ) dθ s = ∫ fs (θ s ) cos θ s dθ s = 575 ft-kips
0
2 0
where fs (θ s ) is calculated from Eqs. (21.6.4) and (21.6.5).

The resultant axial force and moment are then
Pns = Cc + Fs = 779 + 246 = 1025 kips

M n = M c + M s = 316 + 575 = 891 ft-kips
Taking into account slenderness effects due to initial imperfections, the nominal and
design axial strengths are
Pn = κ Pns = 0.95(1025) = 978 kips

φc Pn = 0.75(978) = 733 kips ≈ Pu
and the design moment capacity is
φb M n = 0.9(891) = 801 ft-kips > Mu = 680 ft-kips
The design is thus adequate to resist the factored axial force and moment.
(d) Check shear capacity. From Eq. (21.6.13),
Vn = 0.60 f ys Aw
As
= 0.60 f ys
2
 28.5 
= 0.60(42)  = 359 kips
 2 
φ vVn = (0.9)(359) = 323 kips > Vu = 85 kips
The design is also adequate to resist the factored shear force.

942
Load Transfer
Typical shear connections between steel girders and CFT columns consist of a shear tab
welded to the tube section. Thus, transfer of load from the steel tube to the concrete core
must occur if the column is to behave as a composite member. Bond between steel and
concrete in CFT columns has been studied experimentally, typically through push-out tests
with or without a shear tab [21.43–21.47]. While some friction develops between the steel
and concrete depending on the roughness of the interior surface of the tube, rotation of the
shear tabs leads to high concrete bearing stresses at the bottom of the shear tab, increas-
ing shear transfer. Also, as the top of the shear tab pulls the tube in one direction, the tube
pushes against the concrete in the lateral direction, further increasing bond.
Current bond strength provisions for CFT columns in AISC are based on the work by
Zhang et al. [21.46]. The shear transfer strength along the load introduction length is (AISC-I6.3c)
Rn = pb Lin Fin (21.6.14)
where pb is the perimeter of the steel-concrete interface, Lin is the load introduction length,
and Fin is the nominal bond stress. The load introduction length Lin is taken as twice the min-
imum cross-sectional dimension (rectangular sections) or twice the diameter (circular sec-
tions) above and below the shear transfer region (AISC-I6.4b). The nominal bond stress is
12 t
Fin = ≤ 0.1 ksi (rectangular sections) (21.6.15a)
H2
30 t
Fin = ≤ 0.2 ksi (circular sections) (21.6.15b)
D2
where H is the maximum cross-sectional dimension, D is the tube diameter, and t is the
design tube thickness. As can be seen, the bond stress equations reflect the trend observed
in tests, where an increase in cross-sectional dimensions is accompanied by a decrease in
bond [21.46]. Also, larger bond stresses are developed in circular tubes than in rectangular
sections. The strength reduction factor applicable to bond is 0.50.
In most cases, bond between the steel tube and concrete core is sufficient to transfer
the required force from the steel tube to the concrete. However, when the force transferred
through bond is not sufficient, shear connectors must be used, as discussed for concrete-
encased steel composite columns. In that case, the shear strength along the tube-concrete
interface over the load introduction length is taken solely as that provided by the shear
connectors. This is because slips associated with peak bond stresses in tests are much lower
than slips required to develop the strength of shear connectors [21.46].
EXAMPLE 21.6.2
For the CFT column in Example 21.6.1, calculate the maximum shear force that can be
transferred by steel girders without the need for shear connectors. Assume the column
extends above the steel girders.
SOLUTION
(a) Determine nominal bond strength along load introduction length.
30t 30(0.465)
Fin = = = 0.035 ksi < 0.2 ksi
D2 (20)2

Lin = 2(2 D) = 2(40) = 80 in.

pb = π ( D − 2t ) = π [20 − 2(0.465)] = 59.9 in.

Rn = pb Lin Fin = 59.9(80)(0.035) = 167 kips
(Continued)
943
21.7 MOMENT CONNECTIONS WITH COMPOSITE COLUMNS 943
The maximum shear force that can be transferred by the steel girders, Vu, is calculated
from Eq. (21.5.11) by setting Vu = Pr and Vr′ = φRn.
φ Rn 0.50(167)
Vu = = = 130 kiips
 As f y   28.5(42) 
1 −
 1 − P   3368 
nso
21.7 MOMENT CONNECTIONS WITH COMPOSITE

COLUMNS
Extensive research on connections between steel girders/beams and composite columns,
either concrete-encased steel composite columns or concrete-filled tubes, was conducted
in the late 1980s and 1990s, primarily under earthquake-type loading [21.51–21.61].
Connections between concrete-encased steel composite columns and steel beams can be
lumped in two main groups, through-column and through-beam connections. In the former,
the encased steel column is continuous through the joint, while in the latter the steel beam
passes continuously through the joint. Through-column type connections are similar to
moment connections in steel frames [21.55, 21.62], involving welding of the steel beam
flanges to the encased steel shape. Transverse reinforcement is necessary to provide con-
finement to the joint, although research [21.55] indicates that lower amounts than those
required for reinforced concrete joints (see Chapter 11) have led to adequate seismic per-
formance. In some cases, additional steel plates, welded to the steel beam flanges and paral
lel to the beam web, are used for increased shear resistance [21.55].
Through-beam connections are simpler than through-column connections and do not
rely on welding for force transfer between the steel beam flanges and the encased steel
column. Typical details in these connections consist of either face bearing plates in com-
bination with transverse reinforcement for confinement, or face bearing plates and steel
band plates wrapping around the column just above and below the steel beam (Fig. 21.7.1).
Because the steel beam is continuous through the column, moments are transferred in one
direction only, although perpendicular beams may transfer shear into the column.
Joint shear strength in through-beam-type connections is provided by the steel web
panel, an inner concrete strut activated through concrete bearing of the face bearing plates
and beam flanges, and an outer concrete panel, activated through concrete bearing of the
encased steel section in combination with steel hoops above and below the steel beam, or
through concrete bearing of the steel band plates. For connection with no transverse beams,
joint confinement can be provided through overlapping U-shaped ties passing through
holes drilled on the web of the steel beam [Fig. 21.7.1(a)]. For connections with transverse
beams, on the other hand, joint confinement may be provided by steel band plates welded
to the steel beam flanges (Fig. 21.7.1(b)]. Additional ties, however, are needed to provide
lateral support to the longitudinal column reinforcement over the beam depth. These ties
may be anchored in the column core and do not need to pass through the steel beam web.
Design guidelines for through-beam type connections were published by an ASCE Task
Group in 1994 [21.63]. Additional design recommendations are available in References
21.58 and 21.62.
To ensure adequate behavior in moment connections between steel girders/beams and
CFT columns, both the steel tube and the concrete core must participate in the trans-
fer of forces. Tensile and compressive forces in the beam flanges must be transferred to
the concrete through bearing, typically using a bolted connection. An effective connec-
tion detail that has been shown to behave well under large shear reversals is shown in
Fig. 21.7.2 [21.60]. In this case, the bolts passing through the column are unbonded and
94
Hoops Steel Steel

column Transverse beam column
Steel Steel
beam beam
Face Face
U-shaped bearing Ties bearing
ties plates plates
Column
bars
Steel band plates Column bars
(a) Connection with no transverse beams (b) Connection with transverse beams
Figure 21.7.1 Typical details for through-beam connections between steel beams and concrete-
encased steel composite columns.
CFT column
Through rods Split-tee

(unbonded)
Shear tab
Headed studs Steel beam
Figure 21.7.2 Steel beam-to-CFT column moment connection.
must be designed to behave elastically, thus avoiding an unacceptable decrease in stiffness.

Although the connection shown in Fig. 21.7.2 includes shear connectors for shear trans-
fer, these may not be necessary in a connection transferring gravity loads, as discussed in
Section 21.6. Further information on the design of this type of connection is available in
References 21.62 and 21.64.
SELECTED REFERENCES
21.1. American Institute of Steel Construction. Steel Construction Manual (14th ed.). Chicago:
AISC, 2011.
21.2. J. C. Saemann and George W. Washa. “Horizontal Shear Connections Between Precast Beams
and Cast-in-Place Slabs,” Journal of the American Concrete Institute, 61, November 1964,
1383–1408.
21.3. Robert E. Loov and Anil K. Patnaik. “Horizontal Shear Strength of Composite Concrete Beams
With a Rough Interface,” PCI Journal, 39, January–February 1994, 48–69.
21.4. N. W. Hanson (1960). “Precast-Prestressed Concrete Bridges. 2—Horizontal Shear Connections,”
Journal, PCA Research and Development Laboratories, 2, May 1960, 38–58. PCA Development
Department Bulletin D35, 1960, 21 pp.
21.5. Philip W. Birkeland and Halvard W. Birkeland. “Connections in Precast Concrete Construction,”
Journal of the American Concrete Institute, 63, March 1966, 345–367.
21.6. Alan H. Mattock and Neil M. Hawkins. “Shear Transfer in Reinforced Concrete—Recent
Research,” PCI Journal, 17, March–April 1972, 55–75.
945
21.7. Joost Walraven, Jerome Frénay, and Arjan Pruijssers. “Influence of Concrete Strength and Load
History on the Shear Friction Capacity of Concrete Members,” PCI Journal, 32, January–
February 1987, 66–84.
21.8. Alan H. Mattock, W. K. Li, and T. C. Wang. “Shear Transfer in Lightweight Reinforced
Concrete,” PCI Journal, 21, January–February 1977, 20–39.
21.9. Dan E. Branson. “Time-Dependent Effects in Composite Concrete Beams,” Journal of the
American Concrete Institute, 61, February 1964, 213–229.
21.10. Dan E. Branson. “Design Procedures for Computing Deflections,” ACI Journal, 65, September
1968, 730–742.
21.11 Antoine E. Naaman. Prestressed Concrete Analysis and Design—Fundamentals. Ann Arbor,
MI: Techno Press 3000, 2012, 1173 pp.
21.12 Holger Eggemann. “Development of Composite Columns. Emperger’s Effort,” Proceedings of
the First International Congress on Construction History, Madrid, January 2003, 787–797.
21.13 National Association of Cement Users, Philadelphia. Standard No. 4—Standard Building
Regulations for the Use of Reinforced Concrete. Adopted by the American Concrete Institute in
February, 1910, 14 pp.
21.14 American Concrete Institute. Standard Specifications No. 23—Standard Building Regulations
for the Use of Reinforced Concrete, 1920, 20 pp.
21.15 Richard W. Furlong. “Design Rules for Steel-Concrete Composite Columns: 1910 to 1963,”
21.16 Richard W. Furlong. “Design Rules for Steel-Concrete Composite Columns: 1971 to 2011,”
21.17 American Institute of Steel Construction. Specification for Structural Steel Buildings, ANSI/
AISC 360-16, Chicago: AISC, July 2016, 620 pp.
21.18 Lawrence G. Griffis. “Some Design Considerations for Composite-
Frame Structures,”
Engineering Journal, Second Quarter, 1986, 59–64.
21.19 Oscar Faber (1956). “Savings to Be Effected by the More Rational Design of Cased Stanchions
as a Result of Recent Full Size Tests,” The Structural Engineer, 34, March, 88–109.
21.20 Royston Jones and A. A. Rizk (1963). “An Investigation on the Behaviour of Encased Steel
Columns Under Load,” The Structural Engineer, 41, January, 21–33.
21.21 Shosuke Morino, Chiaki Matsui, and Hidehiko Watanabe. “Strength of Biaxially Loaded SRC
Columns,” in Composite and Mixed Construction, Reston, VA: American Society of Civil
Engineers, 1985), Charles W. Roeder, Editor, pp. 185–194.
21.22 S. Ali Mirza, Ville Hyttinen, and Esko Hyttinen. “Physical Tests and Analyses of Composite
Steel-Concrete Beam Columns,” Journal of Structural Engineering, 122, November 1996,
1317–1326.
21.23 Cengiz Dundar, Serkan Tokgoz, A. Kamil Tanrikulu, and Tarik Baran. “Behavior of Reinforced
and Concrete-Encased Composite Columns Subjected to Biaxial Bending and Axial Load,”
Building and Environment, 43, 2008, 1109–1120.
21.24 James M. Ricles and Shannon D. Paboojian. “Seismic Performance of Steel-Encased Composite
Columns,” Journal of Structural Engineering, 120, August 1994, 2474–2494.
21.25 Cheng-Chih Chen, Jian-Ming Li, and C. C. Weng. “Experimental Behaviour and Strength
of Concrete-Encased Composite Beam-Columns with T-Shaped Steel Section under Cyclic
Loading,” Journal of Constructional Steel Research, 61, 2005, 863–881.
21.26 Roberto T. Leon, Dong Keon Kim, and Jerome F. Hajjar. “Limit State Response of Composite
Columns and Beam-Columns. Part 1: Formulation of Design Provisions for the 2005 AISC
Specification,” Engineering Journal, Fourth Quarter, 2007, 341–358.
21.27 Timo K. Tikka and S. Ali Mirza. “Nonlinear EI Equation for Slender Composite Columns
Bending about the Minor Axis,” Journal of Structural Engineering, 132, October 2006,
1590–1602.
21.28 Timo K. Tikka and S. Ali Mirza. “Effective Flexural Stiffness of Slender Structural Concrete
Columns,” Canadian Journal of Civil Engineering, 35, 2008, 384–399.
21.29 Ivan M. Viest, Joseph P. Colaco, Richard W. Furlong, Lawrence G. Griffis, Roberto

T. Leon, and Loring A. Wyllie Jr. Composite Construction. Design for Buildings.
New York: McGraw-Hill, 1996.
21.30 Charles G. Salmon, John E. Johnson, and Faris A. Malhas. Steel Structures. Design and
Behavior (5th ed.). Upper Saddle River, NJ: Prentice Hall, 2008.
21.31 Bungale S. Taranath. Steel, Concrete, & Composite Design of Tall Buildings, (2nd ed.).
New York: McGraw-Hill, 1998, 998 pp.
21.32 Richard W. Furlong. “Strength of Steel-Encased Concrete Beam Columns,” Journal of the
Structural Division, ASCE, 93, October 1967, 113–124.
946
21.33 Richard W. Furlong. “Design of Steel-Encased Concrete Beam Columns,” Journal of the
Structural Division, ASCE, 94, January 1968, 267–281.
21.34 Robert B. Knowles and Robert Park. “Strength of Concrete Filled Steel Tubular Columns,”
Journal of the Structural Division, ASCE, 95, December 1969, 2565–2587.
21.35 Amit H. Varma, James M. Ricles, Richard Sause, and Le-Wu Lu. “Experimental Behavior
of High Strength Square Concrete-Filled Steel Tube Beam-Columns,” Journal of Structural
Engineering, 128, March 2002, 309–318.
21.36 Kenji Sakino, Hiroyuki Nakahara, Shosuke Morino, and Isao Nishiyama. “Behavior of Centrally
Loaded Concrete-Filled Steel-Tube Short Columns,” Journal of Structural Engineering, 130,
February 2004, 180–188.
21.37 Eiichi Inai, Akiyoshi Mukai, Makoto Kai, Hiroyoshi Tokinoya, Toshiyuki Fukumoto, and
Koji Mori. “Behavior of Concrete-Filled Steel Tube Beam Columns,” Journal of Structural
Engineering, 130, February 2004, 189–202.
21.38 Toshiaki Fujimoto, Akiyoshi Mukai, Isao Nishiyama, and Kenji Sakino. “Behavior of

Eccentrically Loaded Concrete-Filled Steel Columns,” Journal of Structural Engineering, 130,
February 2004, 203–212.
21.39 Amit H. Varma, James M. Ricles, Richard Sause, and Le-Wu Lu. “Seismic Behavior and Design
of High-Strength Square Concrete-Filled Steel Tube Beam Columns,” Journal of Structural
Engineering, 130, February 2004, 169–179.
21.40 Tiziano Perea. Analytical and Experimental Study on Slender Concrete-Filled Steel Tube
Columns and Beam-Columns. Ph.D. thesis, Georgia Institute of Technology Atlanta, December
2010, 667 pp.
21.41 Tiziano Perea, Roberto T. Leon, Jerome F. Hajjar, and Mark D. Denavit. “Full-Scale Tests of
Slender Concrete-Filled Tubes: Axial Behavior,” Journal of Structural Engineering, 139, July
2013, 1249–1262.
21.42 Tiziano Perea, Roberto T. Leon, Jerome F. Hajjar, and Mark D. Denavit. “Full-Scale Tests of
Slender Concrete-Filled Tubes: Interaction Behavior,” Journal of Structural Engineering, 140,
No. 9, 04014054, 2014.
21.43 K. S. Virdi and P. J. Dowling. “Bond Strength in Concrete Filled Steel Tubes,” IABSE
Proceedings P-33/80, 1980, 125–139.
21.44 Michelle Ann Parsley. Push-Out Behavior of Concrete-Filled Steel Tubes, Master of Science in
Engineering thesis, University of Texas at Austin, August 1998, 145 pp.
21.45 Charles W. Roeder, Brad Cameron, and Colin B. Brown. “Composite Action in Concrete Filled
Tubes,” Journal of Structural Engineering, 125, May 1999, 477–484.
21.46 Jie Zhang, Mark D. Denavit, Jerome F. Hajjar, and Xilin Lu. “Bond Behavior of Concrete-
Filled Steel Tube (CFT) Structures,” Engineering Journal, Quarter, 2012, 169–185. Also,
Errata, Third Quarter, 2013, 201–213.
21.47 M. H. Mollazadeh and Y. C. Wang. “New Insights into the Mechanism of Load Introduction into
Concrete-Filled Steel Tubular Column Through Shear Connection,” Engineering Structures,
Elsevier, 75, 2015, 139–151.
21.48 Gianluca Ranzi, Graziano Leoni, and Riccardo Zandonini. “State of the Art on the Time-
Dependent Behaviour of Composite Steel-Concrete Structures,” Journal of Constructional
Steel Research, 80, 2013, 252–263.
21.49. Dawn E. Lehman, Katherine G. Kuder, Arni K. Gunnarrson, Charles W. Roeder, and Jeffrey
W. Berman. “Circular Concrete- Filled Tubes for Improved Sustainability and Seismic
Resilience,” Journal of Structural Engineering, 141, March 2015, B4014008.
21.50 Brett C. Gourley, Ceng Tort, Mark D. Denavit, Paul H. Schiller, and Jerome F. Hajjar. A Synopsis
of Studies of the Monotonic and Cyclic Behavior of Concrete-Filled Steel Tube Members,
Connections, and Frames, Department of Civil and Environmental Engineering, University of
Illinois at Urbana-Champaign, Report No. NSEL-008, April 2008, 364 pp.
21.51 T. M. Sheikh, J. A. Yura, J. O. Jirsa. Moment Connections Between Steel Beams and Concrete
Columns, PMFSEL Report No. 87-4, University of Texas at Austin, 1987.
21.52 G. G. Deierlein, J. A. Yura, and J. O. Jirsa. Design of Moment Connections for Composite
Framed Structures, PMFSEL Report No. 88-1, University of Texas at Austin, 1988.
21.53 R. Kanno. Strength, Deformation, and Seismic Resistance of Joints between Steel Beams and
Reinforced Concrete Columns, Ph.D. thesis, Cornell University, Ithaca, NY, 1993.
21.54 S. S. F. Mehanny. Modeling and Assessment of Seismic Performance of Composite Frames with
Reinforced Concrete Columns and Steel Beams, Ph.D. thesis, Stanford University, Stanford,
CA, 2000.
21.55 Chung-Che Chou and Chia-Ming Uang. “Cyclic Performance of a Type of Steel Beam to Steel-
Encased Reinforced Concrete Column Moment Connection,” Journal of Constructional Steel
Research, 58, 2002, 637–663.
947
PROBLEMS 947
21.56 G. Parra-Montesinos and J. K. Wight. “Seismic Response of Exterior RC Column-To-Steel
Beam Connections,” Journal of Structural Engineering, 126, October 2000, 1113–1121.
21.57 G. Parra-Montesinos and J. K. Wight. “Modeling Shear Behavior of Hybrid RCS Beam-
Column Connections,” Journal of Structural Engineering, 127, January 2001, 3–11.
21.58 G. J. Parra-Montesinos, X. Liang, and J. K. Wight. “Towards Deformation-Based Capacity
Design of RCS Beam-Column Connections,” Engineering Structures, 25, 681–690.
21.59 Atorod Azizinamini and Stephen P. Schneider. “Moment Connections to Circular Concrete-
Filled Steel Tube Columns,” Journal of Structural Engineering, 130, February 2004, 213–222.
21.60 J. M. Ricles, S. W. Peng, and L. W. Lu. “Seismic Behavior of Composite Concrete Filled Steel
Tube Column-Wide Flange Beam Moment Connections,” Journal of Structural Engineering,
130, February 2004, 223–232.
21.61 Isao Nishiyama, Toshiaki Fujimoto, Toshiyuki Fukumoto, and Kenzo Yoshioka. “Inelastic
Force-Deformation Response of Joint Shear Panels in Beam-Column Moment Connections to
Concrete-Filled Tubes,” Journal of Structural Engineering, 130, February 2004, 244–252.
21.62 American Institute of Steel Construction. Seismic Provisions for Structural Steel Buildings,
ANSI/AISC 341-16. Chicago: AISC, 2016, 430 pp.
21.63 ASCE Task Committee on Design Criteria for Composite Structures in Steel and Concrete.
“Guidelines for Design of Joints between Steel Beams and Reinforced Concrete Columns,”
Journal of Structural Engineering, 120, August 1994, 2330–2357.
21.64 Erica C. Fischer and Amit H. Varma. “Design of Split-Tee Connections for Special Composite
Moment Frames,” Engineering Journal, Third Quarter, 2015, 185–202.
PROBLEMS
FOR PROBLEMS 21.1 and 21.2, use the provisions in ACI 318-14.
21.1 Consider the concrete composite beam shown in For the cast-in-place slab: fcs′ = 4000 psi; fy =
the figure for Problem 21.1. The beam is part of 60 ksi. Concrete is normal weight. Cover to cen-
a floor system with parallel, simply supported ter of longitudinal bars = 2.5 in.
beams spaced at 10 ft. The center-to-center span 21.2 Repeat Problem 21.1 but assuming shored
length of the composite beam is 40 ft with a clear construction.
span length of 38 ft. During construction, the
beam must support a live load of 25 psf. Once FOR PROBLEMS 21.3 through 21.8, use the 2016
the beam becomes composite, it must sustain, in AISC Specification for Structural Steel Buildings.
addition to its self-weight, a superimposed dead
21.3 Consider the concrete-encased steel composite
load of 20 psf and a live load of 100 psf. The
column shown in the figure for Problem 21.3.
slab is reinforced with #4 bars at 14 in. spacing
For bending about the strong axis of the encased
in the direction of the beam span.
steel section, determine:
Check the adequacy of the design in terms of
(a) Nominal axial force–moment strength inter-
flexural and shear strengths, as well as deflec-
action diagram; use the strain compatibility
tions. Assume unshored construction and that
approach and neglect slenderness effects.
the floor system supports nonstructural elements
(b) Nominal axial force–moment strength inter-
not likely to be damaged by large deflections. For
action diagram; use the strain compatibility
the precast beam: fc′ = 8000 psi; fy = fyt = 60 ksi.
approach, including slenderness effects for
an unsupported column length of 12 ft. for
6” bending about the strong axis.
(c) Design shear strength for shear parallel to
#4 @ 12 in. Cast-in-place slab the web of the steel shape. Compare the
results using the three approaches allowed
30”
Precast in the AISC Specification (i.e., bare steel
beam section, steel section plus ties, reinforced
5 #9
concrete section).
16” Material properties: fc′ = 8000 psi (normal
weight); fys = 50 ksi (A572 Gr. 50 steel); fyr =
Problems 21.1 and 21.2 fyt = 60 ksi.
948
#4 @ 10” 21.6 For the column of Problem 21.3, determine the

12 #10 diameter, length, number, and spacing of headed
3” stud anchors required to transfer the shear force
6” carried by two W16 girders framing into the
two flanges of the concrete-encased steel sec-
5”
28” tion. The factored shear force in each girder is
5” 72 kips. Tensile strength of studs, fua = 65 ksi.
6”
Draw a sketch of the connection, indicating the
layout of headed studs.
3”
21.7 For architectural reasons, the steel tube of a
28” rectangular CFT composite column is to be
an HSS18 × 12 section. Determine the lightest
W14 × 145 tube section capable of resisting the following
Problems 21.3 and 21.6 factored loads obtained from a second-order
analysis (not including the effect of initial
imperfections): Pu = 270 kips; Mu = 225 ft-kips;
21.4 Repeat Problem 21.3 but for bending about the Vu = 28 kips (section is subjected to bending
weak axis of the steel section. about its strong axis). Calculate the maximum
21.5 For the concrete-encased steel composite column shear force that can be transferred by steel
section in the figure for Problem 21.3, determine girders without the need for shear connectors.
the lightest W14 steel shape required to resist Assume the column extends above the steel gird-
the following actions obtained from a second- ers. Unsupported column length L = 12.5 ft. for
order analysis (not including the effect of initial bending about strong axis (column is fully sup-
imperfections): Pu = 1550 kips; Mu = 1310 ft-kips ported in the perpendicular direction). Material
(about strong axis of steel shape); Vu = 160 kips. properties: fc′ = 6000 psi (normal weight);
Unsupported column length L = 12 ft. for bend- fys = 42 ksi (A500 Grade B steel).
ing about strong axis (column is fully supported 21.8 Repeat Problem 21.7 but using a circular HSS16
in the perpendicular direction). tube section.
94
INDEX
Note: Page references followed by a “t” indicate table; “f ” indicate figure.
active pressure, on retaining bar details bearing stress, 179

walls, 597 in floor beams, 287 on square spread footings, 808
admixtures, 6–9, 7f in one-way slabs, 264 bearing walls, 569, 571f
accelerating, 8 beams. See also deep beams design of, 573–76
air-entraining, 8 balanced strain condition for, strength reduction factors for, 574
chemical, 8–9 51–52, 51f biaxial bending
for flowing concrete, 9 behavior of beams without shear axial compression and, 360–67,
for self-consolidating concrete, 9 reinforcement, 114–22 360f–65f, 367f, 494
mineral, 7–8 behavior of beams with shear biaxial strength of concrete, 14, 14f
set-retarding, 8–9 reinforcement, 122–29 bond, See development of
water-reducing, 8–9 compression-controlled sections reinforcement
aggregates, 6 of, 52–55, 78–80 braced frames, 470f. See also nonsway
aggregate interlock, 115–16, 118, crack control for, 451–53 frames.
124, 128 deflections of, 411, 446–51, brackets and corbels, 157–68,
all-lightweight concrete, 6 446f, 449f 158f, 161f, 162f, 165f, 167f,
shear friction and, 154 flexural strength of, 43–88, 177, 559–65
shear strength and, 121 182–85 B-regions, 127
tensile strength of, 13 minimum tension reinforcement
arch action for shear strength, 115, for, 58–59 cantilever retaining walls, 580–81, 581f
117–18 minimum shear reinforcement for, friction in, 608–9, 608f
axial compression 131–135 preliminary proportioning
bending moment and, 326–28, 470–79 moment-curvature response, 86–88 of, 603–18, 604f, 606f–8f,
biaxial bending and, 360–67, nominal flexural strength for, 48–50 606t, 610f, 613f, 616f, 619f
360f–65f, 367f, 494 post-tensioning of, 849, 849f, 862 safety factors for, 606–8
in CFT columns, 936–37 of prestressed concrete, shear service loads on, 608
in concrete-encased steel reinforcement for, 881–83 shear reinforcement for, 605
composite columns, 919–20 pretensioning of, 847–49, 858f shear strength of, 605
shear strength and, 147–48, 588–89, reinforcing bars for, 64–67, soil pressure on, 604, 606f, 607f
923 65t–67t transverse reinforcement for, 618
slenderness effect and, 328, 464–66, shear failures in, 111, 117–18, cellular concretes, 6
479–80, 921, 938 117f, 119f cement, 5
axial tension shear strength of, 136–46, 137f–39f, cement slag, as mineral admixture, 7
bending moment and, 357–59 143f, 144t CFT columns. See concrete-filled tube
in brackets and corbels, 158–59 shear stress on, 113–14, 113f, 114f columns
in CFT columns, 938 skin reinforcement in, 455 circular columns, 354–57, 355f, 356f,
in concrete-encased steel composite tension-controlled sections of, 356t, 937–38
columns, 921 52–57, 78–84, 80f circular sections. See also circular
shear strength and, 149 truss model for, 125–27, 126f, 127f, columns
torsional stiffness and, 753 128f torsional stiffness in, 751
in walls, 588–90 two-way floors with, 624, 629–31, torsional stress in, 749–50, 750f
629f, 630f, 633–43, 634f–36f, coarse aggregate (gravel), 6
balanced strain condition 634t, 638f, 639f, 642f, 669–71, coefficient of friction, 153–54, 600,
(compression control limit) 670f 903
for beams, 51–52, 51f bearing capacity of soil, footings collapse mechanism, 312–14, 721f,
for columns, 327, 333–36, 333f–35f, and, 799 720–27
336, 357 bearing strength, 163, 542 column capital, 623, 627, 629,
for prestressed concrete, 868–69 for square spread footings, 808 639–43, 640f
950
950 INDEX
columns, 321f. See also circular concrete-encased steel composite for monolithic beam-column joints,
columns; composite columns; columns, 917–34, 920f, 383–87
concrete-encased steel 922f, 925f–27f, 927t, 929f, spirally reinforced columns and,
composite columns; concrete- 931f–34f 323–25
filled tube columns; CFT concrete-steel composite columns, effect on bond, 181, 186
columns; spirally reinforced 916–17, 917f conjugate beam method, for
columns; tied columns design strength of, 917 deflections, 412–13, 412f
alignment charts, 490–92 moment connections with, 943–44, corbels. See brackets and corbels
axial compression in, 326, 360–67, 944f corner connections, in flat slab floors,
360f–65f, 367f, 368 composite members, 897–944, 898f 697, 698f
axial load on, 34, 322–24, 323f–25f, composite action in, 897–901 corner effects, yield line theory and,
333 CFT columns, 916, 934–43, 937f, 742–43, 742f
axial tension in, 357–59 940f corner reinforcement, for two-way
balanced strain condition for, 327, concrete composite flexural floors, 664–65, 665f
333–36, 333f–35f, 357 members, 901–16 corrosion
bending moment in, 321, 326, 345, concrete-encased steel composite reinforcement protection for, 24
357–59, 463, 464f, 470–76, columns, 917–34, 920f, 922f, crack control and, 451–52
471f–73f, 475f, 477f 925f–27f, 927t, 929f, coupled walls, 579
biaxial bending in, 360–67, 931f–34f coupling beams, 579
360f–65f, 367f concrete-steel composite columns, crack control
compressive strength of, 323, 333–34 916–17, 917f for beams, 451–54
concrete-steel composite, 916–17, interface shear in, 899–901, 900f, for one-way slabs, 451–54
917f 901f side face, for deep beams, 455–56,
concrete-encased composite, compression-controlled sections 455f
918–34, 943–44 of beams, 52–57, 53f, 78–84, 80f for two-way floors, 666
concrete-filled tube, CFT, 934–943 nominal flexural strength of, 72–78, cracking moment, 416
critical buckling load of, 483, 484, 73f, 75f, 77f minimum tension reinforcement
506, 509, 514–15 compression control limit. See and, 58–59
design strength for, 515 balanced strain condition for prestressed concrete beams,
ductility of, 323 compression lap splices, 216, 216t 871–72, 873f
effective length for, 468–69, 469f, compressive strength, 9–12 creep, 16–17
470f biaxial strength and, 14 beam curvature and, 428–29, 429f
flexural stiffness of, 702–3 of columns, 323, 333–34 compression steel and, 430–31
lateral ties in, 329–30, 330f of concrete-encased steel composite deflection and, 323, 428–31, 428f,
longitudinal reinforcement for, 326, columns, 919–20 429f, 430t, 436–39, 442, 448
332 of concrete filled tubes, 936 humidity and, 429–30, 430t
nominal flexural strength of, 357 effect of confinement on, 323 prestressed concrete and, 858–59
plastic centroid in, 327, 334, 336, of nodal zones, in strut-and-tie critical section
347 models, 537, 537t, 563 for brackets and corbels, 161
shear in, 368–70, 369f of struts, 535–37, 536f, 537f, 544, for footings, 803, 803f
shear stress in, 332 548 for nominal shear strength, 135,
slenderness effects for, 463–523 concrete-encased steel composite 136f
slenderness ratio for, 326, 328–29, columns, 917, 918, 918–34, for torsion, 772
516 920f, 922f, 925f, 926f, 927f, cutoff points, 198–203
stiffness reduction factor for, 481, 927t, 929f, 931f–34f
483 axial tensile strength of, 921 DDM. See Direct Design Method
strength interaction diagram for, compressive strength of, 919–20 deep beams, 151, 542–59
326–27, 327f, 328f, 479–80, flexural strength of, 919–20 deflections. See also immediate
480f headed shear studs for, 925 deflection
strength reduction factors for, shear strength of, 921–23 of beams, 410–413, 446–51, 446f,
325–26, 352, 481 slenderness effects of, 920–21 449f
column strips, in two-way floors, concrete-filled tube (CFT) columns, of concrete composite flexural
656–62, 656f, 658t–62t, 664 916, 934–43, 937f, 940f members, 905–08
combined footing, 800, 801f, 822–41, axial compressive strength of, compression steel and, 430–31, 435
823f, 825f, 828f, 831f, 832f, 936–37 conjugate beam method
833t, 835f, 839f axial tensile strength of, 938 for, 412–13, 412f
compatibility torsion (statically shear strength of, 938 creep and, 17, 323, 428–31, 428f,
indeterminate torsion), 313, slenderness effects for, 938 429f, 430t, 436–39
748, 752–53 width-to-thickness ratio for, effective moment of inertia for
composite columns, 322, 323f 935–36, 935t calculating, 415–18
concrete filled tube or CFT columns, confinement, 155 immediate (instantaneous)
916, 934–43, 937f, 940f effect on compressive strength, 323 deflection, 417–19
951
INDEX 951
maximum reinforcement ratio for epoxy-coated reinforcing bars, 24 anchorage with hooks in onolithic
control of, 427–28 development length factors for, beam-column joints, 390, 391f
minimum depth and, 439–41, 440t, 190–91, 196 for ties and stirrups, 211–12
441t, 446 for cantilever beams, 205, 207 humidity
moment area theorems for, 433–35 equilibrium torsion (statically creep and, 16–17, 428, 429–30, 430t
of one-way slabs, 256 determinate torsion), 748–49 shrinkage and, 17–18, 432, 859
shrinkage and, 431–39 Equivalent Frame Method (EFM), hydraulic cements, 5
of two-way floors, 637–39, 638f, 624–25
639f, 704 for gravity loads, 698–704 immediate deflection, 417–19
deformation capacity, 87, 324 models for, 710–711 influence lines, for continuous beams
development of reinforcement, 176, lateral loads, 711 for bending, 241–242
177–80 equivalent frame models, 710 for shear, 250–251
ACI Code on, 185–92 integrity steel
bond stresses, 176–79 fiber-reinforced concrete, 25–26 for beams, 287–88
bond failure mechanisms, 180–81 as shear reinforcement, 129, 130f for two-way slabs, 664
for bundled bars, 195 fiber reinforcement, 24 interaction diagrams. See strength
for compression reinforcement, 194 fine aggregate (sand), 6 interaction diagram
for hooked bars, 195–97, 195f–97f, flat-plate floors, 623, 623f interface shear, of composite
198t shear reinforcement in, 676–85, members, 899–901, 900f, 901f
modification factors for, 189–92 677f, 680f, 681f, 683f, 684f interface shear strength, of concrete com
of shear reinforcement, 211–13 shearheads for, 695–96, 697f posite flexural members, 901–4
for prestressed reinforcement, stirrups for, 678 interface shear transfer in beams
883–84 flat slab floors, 623, 623f (aggregate interlock), 115
splitting failure, 181 column capital for, 623
for tension reinforcement, 195–97, with drop panels, 622, 623, 623f Joints, 380
195f–97f, 198t openings in, 697, 698f beam-to-column, 380
diagonal tension. See shear strength fly ash, as mineral admixtures, 7 confinement of, 383–87
dimensions, 40 footings, 800 core of, 385–86
Direct Design Method (DDM) bearing capacity of soil and, 799 development of reinforcement in,
moments and, 644–51, 646f, 647, combined, 800, 801f, 822–41, 823f, 390–91
649f, 650t, 651f, 656–62, 656f, 825f, 828f, 831f, 832f, 833t, effective joint width, 388–39
657t–62t 835f, 839f forces on, 381–83
moments in columns and, 686–87 critical section for, 803, 803f shear in, 381–83
for two-way floors, 624–25, 644–51, pile, 800, 841 shear strength of, 387–89
686–87 proportioning for equal settlement, column-to-beam minimum strength
discontinuity. See D-regions 804 ratio, 389
D-regions, 125–26 rectangular, 814–18, 816f transverse reinforcement in, 383–87
of deep beams, 528, 542 shear strength of, 802–3 joist floor systems, 306–07
drop panels, 622–23, 639–40 types of, 800
dual system, for shear walls, 576–77 types of failures in, 800–801 kern points, 885
for walls, 800, 818–22, 820f, 821f
effective compressive strength friction, loss of prestress, 860–61 lateral ties, in columns, 329–30, 330f
of nodal zones, 537, 538f lightweight aggregate concrete, 6
of struts, 535–36 galvanized reinforcing bars, 24 lightweight concrete. See also all-
effective flange width (or effective gravel (coarse aggregate), 6 lightweight concrete; sand-
slab width) lightweight concrete
for monolithic beam-column joints, headed shear studs interface shear strength in members
382 for concrete-encased steel with, 902
for T-beams, 95–97, 96f, 97f composite columns, 924–25 tensile strength of, 12–13
for two-way floor systems, 688–89, two-way slabs, shear strength modulus of elasticity of, 15
711 contributed by, 679–80, 680f shear strength of members with,
effective length heavyweight, high-density concrete, 6 121, 130, 134, 588
alignment charts for, 490–92, 491f high-strength concrete, 12 shear friction in members with, 154
factor for columns, 468–69, 469f, shear strength of members with, 122 maximum shear strength in brackets
470f high-performance fiber-reinforced and corbels, 161
effective moment of inertia, 415–17 concrete (HPFRC), 25 minimum reinforcement ratio for
for computation of deflections, 411, hollow steel sections (HSS), 916. deflection control in members
413–18, 441–42, 451 See also CFT columns with, 428
for evaluation of slenderness hooks, standard, 195–98. See also modification factor for bond in
effects, 463, 465, 481–82, 484, Development length members with, 191
492, 496 development length for, 195–97, torsional cracking strength of
EFM. See Equivalent Frame Method 195f–97f, 198t members with, 753
952
952 INDEX
load balancing concept, for prestressed floor systems and, 391 deflection in, 256
concrete, 855–56, 856f hooks for, 390, 391f design of, 255–58, 256t, 257f
load combinations, 36–37 moment strength ratio for, 389, 393 flexural reinforcement for, 258–59
load contour method, 360, 364–65, 365f nominal strength for, 393–94, 398 shear in, 256
L-sections shear for, 392, 393f shrinkage and temperature
torsional stiffness in homogeneous, shear strength for, 387–88, 388f, reinforcement for, 259, 260f
752 389f, 389t, 398 yield line theory for, 721, 725–28,
torsional strength of homogeneous, transfer of axial forces, 391 726f, 728f
751 transverse reinforcement for, openings, in flat slab floors, 697, 698f
383–87, 384f, 385f, 387f, 394, overloads, safety provisions and, 34, 36
mat foundation, 800 396–97, 396f overreinforced, 46
maximum reinforcement ratio
for deflections, 421, 427–28 natural stone aggregates, 6 partial composite, 899
for tension-controlled sections, 53 negative moment region of continuous passive pressure, on retaining walls,
for any maximum tensile strain, 53 beams, cutoff points in, 597
mineral admixtures, 7–8 198–200, 199f, 200f passive resistance, of retaining walls,
modulus of elasticity, 14–15, 15f, 15t neutral axis 600, 600f
of concrete, 14–15 transformed section method and, pile footings, 800, 841
of prestressing reinforcement, 24 409–10, 409f, 410f plastic analysis, 312–17, 313f–16f
of nonprestressed steel nodal forces, yield line theory and, plastic centroid
reinforcement, 24 729–36, 730f–35f in columns, 327, 334, 336, 347
ratio, 404 nodal zones, in strut-and-tie model, plastic flow, creep and, 16
in calculation of deflections, 414 528, 531–33, 529f, 532f, 533f, plastic hinges, 313–14
shear, 751 534f, 538f, 546–57, 555f inelastic deformation and, 313
modulus of rupture, 13, 13f, 58, 416, nominal strength of, 537–38, 537t yield line theory and, 720
757, 847, 871, 874 dimensioning of, 541–42 plasticizing admixtures, 9
moment area theorems, for effective compressive strength of, plasticizing and retarding
deflections, 433–35 553 admixtures, 9
moment coefficients for continuous hydrostatic nodal zones, 531–32, 532f plastic rotation capacity, 313
beams, 246–47, 247t nodes, in strut-and-tie models, 528, portland blast-furnace slag cement, 5
moment connections, with composite 537–38 portland cement, 5, 5t
columns, 943–44, 944f nominal flexural strength positive moment of continuous beams
moment-curvature analysis, 86–87, 87f for beams, 48–50 bar cutoff points for, 201, 201f, 202f
moment magnifier for columns (strength interaction postcracking stiffness, torsional
for nonsway frames, 497–512, 497f, diagram), 326 moment and, 752
500, 505–6, 505f, 508f, 511t, for rectangular sections, 72–78, 73f, post-tensioning, of beams, 849, 849f,
512f, 513 75f, 77f 862
for sway frames, 497–512, 497f, of T-beams, 97–105 friction on, 859–62, 860f, 862f
500, 502–3, 505f, 508f, 511t, of walls, 581 pozzolan, 5, 7
512f nominal shear strength prestressed concrete members, 844–93
moment of inertia, 58, 267, 270–71. critical section for, 135, 136f advantages and disadvantages of,
See also effective moment of for two-way floors, 672 845–46
inertia nominal shear stress, for two-way balanced strain condition for,
for two-way floor systems, 633–34, floors, 669 868–71
637, 647, 700–04 nominal strength, 35 cracking moment for, 871–72, 873f
moments in columns nonbearing walls, 569, 571f creep and, 844, 845, 858–59
DDM and, 686–87 design of, 573 deflection of, 866
two-way floors and, 686–97, 688f, lateral load of, 573 development length for, 883–84
690f, 693f, 697f noncomposites, 898 elastic shortening of, 857–58
moment strength ratio nonrectangular sections, 84–86, 85f flexural behavior of, 885–93, 886f,
for monolithic beam-column joints, nonsway frames, 470 889f, 891f
389, 393 effective length factor, 490–92 flexure-shear cracks in, 874–75
monolithic beam-column joints, minimum design eccentricity, homogenous beam concept for, 854
380–99, 381f, 392f 493–94 inclined cracks in, 875, 875f
actions of, 381–82, 382f, 383f moment magnification, 470–77, 482 internal force concept for, 854–55,
anchorage of reinforcement, 390–91, slenderness ratio limits, 494–495 854f
391f, 392 load balancing concept for, 855–56,
axial load in, 381, 395, 398 one-way slabs, 254–65, 623f 856f
confinement for, 383–85 analysis methods for, 254–55 loss of prestress in, 856–66, 862t
ductility of, 399 bar details for, 264, 264f nominal strength of, 866–71
eccentricity in, 396 collapse mechanism of, 721f, 722 post-tensioned beams, 849, 849f,
effective width for, 395 crack control for, 451–54 859–62, 860f, 862, 862f
953
INDEX 953
pretensioned beams, 847–49, 858f reinforcing bars. See also epoxy- continuous beams and, 250–52, 251f
proportioning of cross sections of, coated reinforcing bars corbels and, 559
885–93, 886f, 889f, 891f for beams, 64–71, 65t–67t, 70f, 72f fiber-reinforced concrete and, 26
reinforcement for, 883–85 concrete cover and, 40, 452 in flat slab floors, 624
relaxation of steel in, 859 deformed, 19–20, 20f, 21t, 22t, 180 live loads and, 250
service loads for, 846f, 849–53, galvanized, 24 in monolithic beam-column joints,
852f, 866 relaxation, of prestressing steel, 859 381–82, 392, 393f
shear reinforcement for, 881–83 required average compressive strength, in one-way slabs, 256
shear strength of, 873–81 18–19 stirrups and, 789
web-shear cracks in, 875–78, 876f reserve capacity, safety provisions strength interaction diagram for
yield points for prestressing and, 34 torsion, bending, and, 768–69,
reinforcement, 866 retaining walls, 569, 571f, 597–601. 768f, 769f
prestressing reinforcement, 23–24, 23f See also cantilever walls torsion and, 767–68
pretensioning, of beams, 847–49, coefficient of friction for, 600, 600t transverse reinforcement and, 1,
858f safety factors for, 599–601 124, 124f, 368, 785, 789
prismatic struts, 529, 530f soil pressure on, 601, 601f two-way floors and, 669–71, 687–97,
pullout failures 688f, 690f, 693f, 697f
bundled bars and, 195 safety provisions, 34–38 shear failure
development length and, 186 sand (fine aggregate), 6 in beams, 111, 117–18, 117f, 119f
with lightweight aggregate sand-lightweight concrete of cantilever walls, 580, 581f
concrete, 191 shear friction and, 154 of deep beams, 117–18, 117f
pullout resistance, development length tensile strength of, 13 of fiber-reinforced concrete beams,
and, 185 sand replacement, 6 129, 130f
pullout tests, for development sanitary structures, crack control of footings, 801, 802f
length, 181 for, 452 shear friction, 152–57, 152f, 153f,
punching shear, of footings, 801 secant modulus, modulus of elasticity 156f
and, 14–15, 414 for brackets and corbels, 159, 166
quality control, for concrete, 18–19 second-order analysis, ACI Code shearhead reinforcement, 623
on, 493 shear reinforcement
radiation, heavyweight, high-density self-consolidating concrete, beams without, 114–22, 115f, 117f,
concrete for, 6 admixtures for, 9 119f
radius of gyration, 463, 496 serviceability, 403–56 for cantilever walls, 605
slenderness ratio and, 497 crack control and, 451–54 for columns, 369
reciprocal load method, 360, 366–67, deflection and, 38–39, 410–11 for combined footings, 831–32
367f deflection and creep and, 428–31, development of, 211–13, 212f, 213f
rectangular footings, 814–18, 816f 428f, 429f, 430t fiber reinforcement as, 129, 130f
rectangular sections effective moment of inertia and, in flat-plate floors, 676–85, 677f,
balanced strain condition for, 414–17, 414f, 415f 680f, 681f, 683f, 684f
51–52, 333, 334f, 336–40, flexural behavior and, 403–4 for prestressed concrete beams,
336f, 339f floor system vibration control, 456 881–83
flexural strength of, 44–50, 45f immediate deflections and, 417–28, shear strength from, 123, 128–29,
minimum tension reinforcement 418f–20f, 423f, 424f 128f
for, 59 linear elastic behavior and, 407 shear strength without, 114–22,
nominal flexural strength of, 72–78, modulus of elasticity and, 404, 115f, 117f, 119f, 149t–51t
73f, 75f, 77f 404t, 406, 414 for slab-beam-girder floor systems,
skew bending theory for, 756, 756f side face crack control and, 454–55, 285–87
space truss analogy for, 761–65, 455f for square spread footings, 806
762f–64f time-dependent deflections, 428 stirrups as, 211–13, 212f
tension reinforcement for, 60–64, transformed section method and, transverse reinforcement as,
61f, 64f 407–10, 407f, 409f, 410f 123–24, 123f, 140–46, 140t,
torsional stiffness of, 751 for two-way floors, 644 143f, 144t, 146f
torsional stress in, 750, 750f, 750t service-level, unfactored loads, 37, 39, truss model and, 122
torsion and shear on, 767–68 403–4. See also service loads for wall footings, 820
for walls, 581–83, 582f, 583f crack control and, 452 for walls, 591–95, 592f, 596f
rectangular walls, shear strength service loads, 33 welded wire reinforcement and,
of, 588 set-retarding admixtures, 8–9 212–13, 213f
redistribution of moments, 312–17, shear shear-related inclined cracks, 180
313f–16f axial tension and, 13 shear resistance, 115–17. See also
reinforcement. See also specific bending moment and, 772–73, 773f shear strength
topics in columns, 368–70, 369f of deep beams, 151
areas, tables of, 21, 65 compressive stress and, 110–11, of monolithic beam-column
reinforcement ratio, 53–54, 54t 110f, 111f joints, 387
954
954 INDEX
shear span, 116, 116f T-sections of, 577–78 sway frames and, 477–79, 478f,
brackets and corbels and, 161 wall-to-floor area ratio for, 576 485–90, 486f, 495, 496
of continuous beams, 121 shored construction, of concrete slenderness ratio, 326, 328, 342, 463,
shear span to depth ratio, 116–119 composite flexural members, 465, 467, 480, 494
shear strength, 110–68. See also 908–16 slippage failure, 178
nominal shear strength, shear shrinkage, 17–18, 18f slump, 6–10, 16–18
resistance correction factors for, 432, 432t soil pressure
axial compression and, 588 crack control and, 451 on cantilever retaining walls, 601,
axial load and, 146–51, 149t creep and, 16 601f, 604, 606f, 607f
bending moment and, 146–51, 149t effect on deflections, 431–37 on square spread footings, 804
of beams, 136–46, 137f–39f, multiplier procedure for, 435–39, on wall footings, 819, 820
143f, 144t 437t space truss analogy (torsion), 761–65,
of CFT columns, 938 prestress loss due to, 859 762f–64f
of columns, 369 and temperature reinforcement, 191, spandrel beams
of concrete composite flexural 215, 259, 306–308, 432, 822 torsional moment and, 754
members, 904–5 warping from, 433–35, 434f torsional stiffness of, 778–91, 778f
of concrete-encased steel composite shrinkage curvature, 432 torsion in, 748
columns, 921–23 shrinkage strain, 431–32 span-to-depth ratio
of continuous beams, 121, 122f beam curvature and, 433 for brackets and corbels, 559
of deep beams, 119 side face crack control, for deep for deep beams, 528, 542
of footings, 802–3 beams, 455–56, 455f for prestressed concrete members,
from headed shear studs, 679–80, silica fume, for mineral admixtures, 846
680f 7–8 shearhead reinforcement
from shear reinforcement, 123, simple supports, for flat-plate floors, 695–96, 697f
128–29, 128f development of positive moment shear strength from, 678–79,
of high-strength concrete, 122 reinforcement, 209–10 695–96, 697f
for joist floor systems, 307 single curvature, 329, 471, 474, special moment frames, monolithic
of monolithic beam-column 476–77, 483, 487, 490–491, beam-column joints in, 386
joints, 387–88, 388f, 389f, 495–496 specified compressive strength, 18–19
389t, 398 singly reinforced sections, spirally reinforced columns, 322, 323,
of prestressed concrete members, rectangular, 48 323f, 324f
873–81 T-beams, 98–99 axial load in, 330
of rectangular spread footings, 815 size effect, for shear strength, 122 clear cover for, 332
without shear reinforcement, skew bending theory, 757 confinement and, 323–24, 330–31
114–22, 115f, 117f, 119f, skin reinforcement, for side face crack ductility of, 324
149t–51t control, 455–56, 455f lap splices for, 332
size effect and, 122 slab-beam-girder floor systems, 240f, longitudinal reinforcement in,
skew bending theory and, 758 266–305 330–32, 331f, 332f
from shearheads, 678–79, 695–96, bar details for, 287–94, 290f mechanical splices for, 332
697f beam web size for, 267–70 stiffness reduction factor for, 483
of square spread footings, 805 continuous frame analysis for, strength reduction factor for, 352
of two-way floors, 640, 671–76, 270–72, 272f–74f, 272t splice length, 213
673f, 674f, 676f, 697 shear reinforcement design for, splices
of walls, 588–90, 589f–91f 285–87 under compression and bending,
shear stress slabs. See also flat slab floors; 217
axial load and, 148f two-way floor systems; one- compression lap, 216, 216t
in beams, 113–14, 113f, 114f way slabs lap, for spirally reinforced columns,
in columns, 332 effective width of, 711 332
linear elastic behavior and, 111–13, yield line theory of, 720–46 length, 213
112f slenderness effects mechanical, 215–16, 217, 332
skew bending theory and, 758 in CFT columns, 938 tension lap, 213–15, 214f, 215t
in thin-wall sections, 762 in columns, 463–523 welded tension, 215–16, 217, 332
torsion and, 756, 772–73, 773f in concrete-encased steel composite split-cylinder test, 12–13
shear walls, 569, 571f, 584–87, 587f columns, 920–21 splitting failure, 181f
buckling of, 579f effective length alignment charts, bearing area and, 178
flexural strength of, 581–87 490–92, 491f bundled bars and, 195
in-plane bending in, 576 interaction diagrams including, concrete cylinder hypothesis and,
interaction diagrams for, 585, 585f 479–80, 480f, 515, 515f 182, 182f
level of coupling for, 579–80 nonsway frames and, 470–76, 472f, hooks and, 195
longitudinal reinforcement for, 584 473f, 475f, 477f, 481–85, 485f, lugs and, 180–81
shear strength of, 588 495, 496 splitting resistance, development
transverse reinforcement for, 584–85 second-order analysis, 493 length and, 185
95
INDEX 955
square spread footings, 804–14, 805f, nonprestressed steel reinforcement, positive moment region of
807f 21f continuous beams and, 427
design of, 809–14, 811f, 812f, 814f prestressing steel reinforcement, 23f singly reinforced, 98–99
shear strength of, 805 welded wire reinforcement, 23f with compression reinforcement,
soil pressure on, 804 structural design process, 31 102
squat walls, 581, 589 strut-and-tie models, 528–65. See also temperature and shrinkage
stability index, 482, 488, 498–99, 505 struts; ties, nodal zones reinforcement
for sway frames, 509 ACI Code provisions, 534–39 for joist floor systems, 308, 308t
statically determinate torsion beam-column joints and, 388f for one-way slabs, 259, 260f
(equilibrium torsion), 312, 748 bottle-shaped struts, 529 529f, for two-way floor systems, 622
statically indeterminate torsion 530f tensile strength,
(compatibility torsion), 313, compression fields in, 529, 530f concrete,
748, 752–53 fan-shaped action, 529, 530f biaxial strength and, 14
torsional stiffness and, 752–55, 754f for brackets and corbels, 559–65, split-cylinder test, 12
steel reinforcement, 19–24, 226 560f lightweight, 121
corrosion of, 24 for deep beams, 528, 529f, 542–59, modulus of rupture, 13
ductility (deformability) of, 20, 21f 543f, 544f, 547f, 551f, 553f, steel reinforcement, 19–24
modulus of elasticity of, 49, 404 553t, 554t, 555f, 556f, 559f tension-controlled sections, 52–57,
stress-strain relationship, 21f, 23f nodal zones, 531–534 53f, 78–84, 80f, 227
stiffness reduction factor pile cap, 541 tension lap splices, 213–15, 214f, 215t
for columns, 481, 483 selection of, 539, 539f–41f tension zone, cutting bars, 199–200
stirrups, 3, 54–55, 65–67,114, 122–135 for shear walls, 541f ternary blended cement, 5
See also shear reinforcement, struts, 529–31 thermal expansion coefficients, 2
hoop reinforcement; transverse ties, 531 thermal stress, crack control and, 451
reinforcement transition stress fields, 532–33, thin-wall sections
for brackets and corbels, 160 533f shear stress in, 762
cutoff points and, 199–200 struts, in strut-and-tie models, space truss analogy for, 762
development of, 211–213 bottle-shaped, 529, 530f torsion in, 756, 762
for flat-plate floors, 676, 678, 677f compressive strength of, 535–37, threshold torsion, 770
for shear reinforcement, 122–135 536f, 537f, 544, 548 tied columns, 322, 323, 323f, 324f
for shear strength, 122–135 dimensioning of, 541–42 stiffness reduction factor for, 483
hooks for, 212 effective compressive strength of, strength reduction factor for, 352
for thin-walled sections (torsion), 535 ties, in strut-and-tie models, 531, 531f
762 in flanges of T-sections, 536 dimensioning of, 541–42
strap footing. See combined footings. nominal strength of, 535–37 nominal strength of, 538–39
strength design method, 32–34 prismatic, 529f tolerances, 40
strength interaction diagram reinforcement for, 536, 536f torsion, 748–91
combined bending and axial force, tapered, 529f combined bending and, 765–67,
326–27, 327f, 328f superplasticizers, 8 766f, 767f
combined bending and torsion, sustained loads, combined shear and, 767–68
766–67 creep under, 16 combined bending, shear, and, 768
combined bending, shear, and effect on deflections, 411, 418, in cantilever beams, 748
torsion, 768–69, 768f, 769f 428–39 compatibility torsion, 748, 752–53
strength reduction factors, 37–38 effect on stiffness, 481, 484–85, 489 critical section for, 772, 783
for brackets and corbels, 161 stability index and, 499 inclined cracks from, 749f
for development length, 188 sway frames, 477–79, 496, 470f longitudinal reinforcement and,
for lightweight concrete, 12–13 effective length factor, 490–92 772, 780, 786, 791
for tension- and compression- moment magnification factor, 488 strength reduction factor for, 37, 38
controlled sections, 54–55, 55f slenderness ratio limits, 494–95 shear stress and, 756, 772–73, 773f
for sections in the transition zone, stability index, 482, 488 skew bending theory for, 758
327, 352 space truss analogy for, 761–65,
for struts, ties, and nodal zones in tall walls, 580 762f–64f
strut and tie models, 535 T-beams, 59 in spandrel beams, 748
stress, principal tensile, 110–11, in bending, 94–108, 95f statically determinate, 312, 748
113–14, 116, 119–20, 148, comparison of rectangular sections statically indeterminate, 313, 748,
873–874, 876–877 and, 95 752–55, 754f
stress-strain relationship, design of, 105 strength interaction diagram for,
concrete in compression, 11–12, effective flange width for, 95–97, 768–69, 768f, 769f
11f, 12f, 428f 96f, 97f tensile stress and, 756
Hognestad’s, 465–66, 467f minimum tension reinforcement in thin-wall sections, 756, 762
fiber-reinforced concrete, 25f for, 59 transverse reinforcement and, 780,
for design, 48f nominal flexural strength of, 97–105 785, 789
956
956 INDEX
torsional moment confinement from, 181, 186, 383 flexural rigidity of, 657
compressive stress and, 753 for continuous beams, 781 flexural stiffness of, 633–37,
girders and, 754 development length and, 186 634f–36f, 634t, 644, 647, 657,
postcracking stiffness and, 752 for monolithic beam-column joints, 699, 702–3
spandrel beams and, 754 383–87, 384f, 385f, 387f, 394, flexural strength of, 640
stirrups and, 762–63, 763f 396–97, 396f gravity load in, 624, 625, 691,
tensile stress and, 753 shear and, 1, 124, 124f, 368, 785, 698–708, 701f–3f, 705f, 706f,
torsional moment strength 789 708f, 709f
at cracking, 755–57, 756f shear friction and, 154 longitudinal moments in, 645–51,
nominal, 771, 777 as shear reinforcement, 123–24, 646f, 649f, 650t, 651f, 656–62,
torsional rigidity. See torsional 123f, 140–46, 140t, 143f, 656f, 657t–62t
stiffness (rigidity), 240, 749, 144t, 146f moment of inertia for, 633
751–53 for shear walls, 584–85 moments in columns and, 686–97,
statically indeterminate torsion and, splitting planes and, 186–87 688f, 690f, 693f, 697f
752–55, 754f for struts, 549, 557–58, 565 moment strength ratio for, 689
of transverse beams in two-way torsional strength from, 770–72, negative moment in, 627, 640, 664,
floors, 651–57, 652f, 653f, 775 704
655f, 700–702 torsion and, 780, 785, 789 nominal shear strength of, 672
torsional strength for walls, 572, 593 nominal shear stress in, 669
skew bending theory, 758 yield strength of, 223, 368 openings in, 697–8
space truss analogy, 761 trial mixes, 18 positive moment for, 625–26, 647
torsional stress triaxial strength of concrete, 14 reinforcement for, 662–68, 663f,
in circular sections, 749–50, 750f stress-strain relationship and, 14f 665f, 667t–68t, 669f
in homogenous sections, 749–51, truss model shear reinforcement for, 676–85,
750f, 750t for beams, 125–27, 126f, 127f, 128f 677f
in I-sections, 751 shear reinforcement and, 122 shear strength of, 640, 671–76,
in L-sections, 751 T-sections. See also T-beams 673f, 674f, 676f, 697
in rectangular sections, 750, 750f, 750t beam web in, 267–70 slab thickness in, 637–43, 638f,
in T-sections, 751 effective flange width for, 95–7 639f, 642f, 662–68, 663f, 665f,
torsion reinforcement effective moment of inertia 667t–68t, 669f
longitudinal reinforcement and, 771 for, 414, 414f temperature reinforcement
minimum requirements for, 774–75 nominal moment strength of, 97 for, 622
total factored static moment torsional stiffness of, 752 torsional stiffness of, 651–55, 652f,
for flat slab floors, 628–29, 628f torsional stresses in, 751 653f, 655f, 700–702
for two-way floors, 625–33, 626f, in two-way floors, 669 total factored static moment for,
627f, 644, 647–48 two-way floors, 622, 622–711, 623f 625–33, 626f, 627f, 644,
transfer length, 887, 883 with beams, 624, 629–31, 629f, 647–48
transformed section method 630f, 633–43, 634f–36f, transverse beams in, 652–55, 652f,
effective moment of inertia and, 442 634t, 638f, 639f, 642f, 669–71, 653f, 655f
neutral axis and, 409–10, 409f, 410f 670f transfer of moment and shear in,
serviceability and, 407–10, 407f, column capital for, 639–43, 640f 687–95
409f, 410f column strips in, 656–62, 656f, T-sections in, 669
transition zone, 338 658t–62t, 664 yield line theory for, 722, 722f,
balanced strain condition and, 333 concentric shear in, 673 736–41, 736f–38f, 740f
bending moment and, 327f corner reinforcement for, 664–65,
design strength for, 351–54, 354f 665f ultimate load analysis, yield line
strength reduction factors and, 327, crack control for, 666 theory and, 720
352 direct design method.See Direct ultimate strength method. See strength
transverse distribution, of longitudinal Design Method (DDM) design method
moments, 656–62, 656f, deflections in, 637–39, 638f, 639f, unbraced frames. See sway frames.
657t–62t 704 underreinforced beams, 46
transverse reinforcement (vertical drop panels for, 639–43, 640f unintentional roughening, 903
stirrups), 3 edge beams in, 639–43 unshored construction, of concrete
for beams, 125, 128, 128f edge columns supporting, 689 composite flexural members,
for brackets and corbels, 159 effective depth for, 674 905–8
for cantilever walls, 580, 618 equivalent frame method. See
center-to-center spacing for, 186, Equivalent Frame Method vertical stirrups. See transverse
206, 386, 387 (EFM). reinforcement
in columns, 368 factored moment in, 625, 627, 663, vibration control, for floor systems,
for combined footings, 830, 831f 667, 667t–68t, 686–88, 704 456
for concrete composite flexural flexural behavior of, 669–71, virtual work method, for yield line
members, 903–4, 689–91 theory, 724–27
957
INDEX 957
wall boundary elements, 577, 578f web reinforcement, 122–24, 128, 123f, free edge effect and, 729, 735, 730f,
walls, 569–618, 571f. See also 124f 731f
bearing walls; cantilever walls; for bottle-shaped struts, 536 fundamental assumptions on,
nonbearing walls; retaining torsion and shear and, 768 723–24
walls; shear walls; thin-wall web-shear cracks, 114, 115f intersection of three yield lines,
sections in prestressed concrete, 875–878 729, 733
axial tension in, 588, 590 in walls, 589–590, 590f-591f nodal forces and, 729–36,
barbell, 578 welded splices 730f–35f
with coupling beams, 579 in compression, 217 one-way slabs and, 721, 725–28,
effective length for, 574t in tension, 215–216 726f, 728f
flexure-shear cracks in, 590, 591f welded wire reinforcement, 22–23 patterns and, 722, 722f, 724f, 732f,
footings for, 800, 818–22, 820f, 821f shear reinforcement and, 212–13, 736f, 742f, 743f
longitudinal reinforcement for, 576, 213f plastic hinges and, 720
593, 596, 596f stress-strain relationship for, 23f rectangular two-way slabs and, 736,
rectangular sections for, 581–83, Whitney rectangular stress 742
582f, 583f distribution, 47–48, 47f special cases, 743–46, 744f, 746t
reinforcement for, 570–72, 572t for T-beams, 97 triangular slab and, 730, 731f
shear reinforcement for, 591–95, wide-beam connections, 380 twisting moment and, 720, 723–24,
592f, 596f wind load, 37, 39 723f
shear strength of, 588–90, 589f–91f wire reinforcement, 22–23 virtual work method and, 724–27
squat, 581, 589 workability, 5, 7, 8, 10, 25 yield point
tall, 580 working load. See service loads in columns, 323
thickness of, 569, 572f working stress method, 32–33, 404 of deformed bars, 22t
transverse reinforcement for, 593 of prestressing reinforcement, 23
web-shear cracks in, 590, 591f yield line theory, 720 of welded wire reinforcement,
wall-to-floor area ratio, for shear analysis methods and, 724–25 22–23
walls, 576–77 collapse mechanism and, 722, 724, yield strain, 44, 46, 52, 55, 87
warping, from shrinkage, 433–35, 434f 743, 747, 721f yield strength, 20, 22, 23
water to cement (or to cementitious) corner effects and, 742–43, 742f yield stress, 20–22
ratio, 10, 16, 10f equilibrium method and, 724–726,
water-reducing admixtures, 8–9 729–732, 736, 742, 744, 726f zinc-coated reinforcing bars, 24
958
95
960
961
962

Wang, Chu-Kia - Reinforced Concrete Design-Oxford University Press, USA (2017 - 2018)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Wang, Chu-Kia - Reinforced Concrete Design-Oxford University Press, USA (2017 - 2018)

Uploaded by

Copyright:

Available Formats

i

Oxford University Press is a department of the University of Oxford. It furthers

Published in the United States of America by Oxford University Press

© 2018 by Oxford University Press

For titles covered by Section 112 of the US Higher Education

All rights reserved. No part of this publication may be reproduced, stored in

You must not circulate this work in any other form

Library of Congress Cataloging-​in-​Publication Data

1 INTRODUCTION, MATERIALS, AND PROPERTIES 1

2 DESIGN METHODS AND REQUIREMENTS 31

3 FLEXURAL BEHAVIOR AND STRENGTH OF BEAMS 43

4 T-​S ECTIONS IN BENDING 94

5 SHEAR STRENGTH AND DESIGN FOR SHEAR 110

6 DEVELOPMENT OF REINFORCEMENT 176

7 ANALYSIS OF CONTINUOUS BEAMS AND ONE-​WAY

8 DESIGN OF ONE-​WAY SLABS 254

9 DESIGN OF SLAB–​B EAM–​G IRDER AND JOIST FLOOR

10 MEMBERS IN COMPRESSION AND BENDING 321

11 MONOLITHIC BEAM-​C OLUMN CONNECTIONS 380

13 SLENDERNESS EFFECTS ON COLUMNS 463

14 STRUT-​A ND-​T IE MODELS—​D EEP BEAMS, BRACKETS,

15 STRUCTURAL WALLS 569

16 DESIGN OF TWO-​WAY FLOOR SYSTEMS 622

17 YIELD LINE THEORY OF SLABS 720

20 INTRODUCTION TO PRESTRESSED CONCRETE 844

21 COMPOSITE MEMBERS AND CONNECTIONS 897

1 INTRODUCTION, MATERIALS, AND PROPERTIES 1

1.1 Reinforced Concrete Structures 1

1.2 Historical Background 2

1.7 Compressive Strength 9

1.8 Tensile Strength 12

1.9 Biaxial and Triaxial Strength 14

1.10 Modulus of Elasticity 14

1.11 Creep and Shrinkage 16

1.12 Concrete Quality Control 18

1.13 Steel Reinforcement 19

1.14 Fiber-​Reinforced Concrete 25

2 DESIGN METHODS AND REQUIREMENTS 31

2.1 Structural Design Process—​General 31

2.2 ACI Building Code 31

2.3 Strength Design and Working Stress Methods 32

2.4 Working Stress Method 33

2.5 Strength Design Method 33

2.6 Safety Provisions—​General 34

2.8 Serviceability Provisions—​General 38

2.9 Serviceability Provisions—​ACI Code 39

2.10 Handbooks and Computer Software 39

2.11 Dimensions and Tolerances 40

2.12 Accuracy of Computations 41

3 FLEXURAL BEHAVIOR AND STRENGTH OF BEAMS 43

3.1 General Introduction 43

3.2 Flexural Behavior and Strength of Rectangular Sections 44

3.3 Whitney Rectangular Stress Distribution 47

3.4 Nominal Flexural Strength Mn—​Rectangular Sections Having Tension

3.5 Balanced Strain Condition 51

3.6 Tension-​and Compression-​Controlled Sections 52

3.7 Minimum Tension Reinforcement 58

3.8 Design of Rectangular Sections in Bending Having Tension Reinforcement

3.10 Nominal Flexural Strength Mn of Rectangular Sections Having Both Tension

3.11 Design of Beams Having Both Tension and Compression Reinforcement 78

3.12 Nonrectangular Sections 84

3.13 Effect of As, As′ , b, d, fc′ , and fy on Flexural Behavior 86

4 T-​S ECTIONS IN BENDING 94

Library of Congress Cataloging-in-Publication Data

4 T-S ECTIONS IN BENDING 94

7 ANALYSIS OF CONTINUOUS BEAMS AND ONE-WAY

8 DESIGN OF ONE-WAY SLABS 254

9 DESIGN OF SLAB–B EAM–G IRDER AND JOIST FLOOR

11 MONOLITHIC BEAM-C OLUMN CONNECTIONS 380

14 STRUT-A ND-T IE MODELS—D EEP BEAMS, BRACKETS,

16 DESIGN OF TWO-WAY FLOOR SYSTEMS 622

1.14 Fiber-Reinforced Concrete 25

2.1 Structural Design Process—General 31

2.6 Safety Provisions—General 34

2.8 Serviceability Provisions—General 38

2.9 Serviceability Provisions—ACI Code 39

3.4 Nominal Flexural Strength Mn—Rectangular Sections Having Tension

3.6 Tension-and Compression-Controlled Sections 52

4 T-S ECTIONS IN BENDING 94

4.2 Comparison of Rectangular and T-Sections 95

4.4 Nominal Moment Strength Mn of T-Sections 97

4.5 Design of T-Sections in Bending 105

5.5 Shear Strength of Beams without Shear Reinforcement—ACI Approach 119

5.8 Shear Strength of Beams with Shear Reinforcement—ACI Approach 128

5.12 Shear Strength of Beams—Design Examples 136

6.5 Flexural Strength Diagram—Bar Bends and Cutoffs 182

6.6 Development Length for Tension Reinforcement—ACI Code 185

7 ANALYSIS OF CONTINUOUS BEAMS AND ONE-WAY

7.4 ACI Code—Arrangement of Live Load and Moment Coefficients 246

8 DESIGN OF ONE-WAY SLABS 254

9 DESIGN OF SLAB-B EAM-G IRDER AND JOIST FLOOR

9.10 One-Way Joist Floor Construction 306

9.12 Redistribution of Moments—Introduction to Limit or Plastic Analysis 312