You are on page 1of 22

Improved Design Assessment of LTB 4/20/2016

of I-Section Members
via Modern Computational Methods

IMPROVED DESIGN ASSESSMENT OF LTB


OF I-SECTION MEMBERS
VIA MODERN COMPUTATIONAL METHODS

Improved Design Assessment of LTB                     
of I‐Section Members                                 
via Modern Computational Methods
Donald W. White (with credits to Dr. Woo Yong Jeong & Mr. Oguzhan Toğay)

School of Civil and Environmental Engineering
Georgia Institute of Technology
Atlanta, GA    USA

2016 NASCC  The STEEL Conference
2

2016 NASCC
D. White, Georgia Tech 1
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

IMPETUS FOR THIS WORK

• Efficient, more rigorous assessment of 
Web‐tapered members with
 Multiple taper
 Steps in the cross‐section (CS) geometry
 Doubly‐ & singly‐symmetric CS geometry
Ordinary frame members
including
• Impact of general lateral & torsional bracing
• Benefits of end restraint & member continuity across braced points
• Influence of general moment gradient                                                         
and other load & displacement boundary condition effects 3

IMPETUS FOR THIS WORK

• Lack of sufficient rigor of Direct Analysis (DM) type approaches   
for assessment of 3D member limit states & stability bracing 
requirements
• Lack of sufficient computational efficiency of advanced         
(plastic zone) analysis methods 
• Difficulty of correlation between advanced analysis results and 
Specification resistances
• Desire for improvement upon traditional K & Cb factor 
approximations
• Desire for improvement upon traditional strength interaction eqs. 4

2016 NASCC
D. White, Georgia Tech 2
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

INTRODUCTION

• Column inelastic effective length factors have been used 
extensively in the past to achieve improved accuracy and 
economy in  the design of steel frames
• Using a buckling analysis with inelastic stiffness reduction 
factors, , the following effects can be captured quite rigorously 
& efficiently for columns, beams & beam‐columns:
Loss of member rigidity due to the spread of plasticity
Various end restraints 
Various bracing constraints and other load & displacement boundary 
conditions
Continuity across braced points 5

AISC a FOR COLUMNS W/ NONSLENDER ELEMENTS

c Pn  0.9 (0.877) Pe  0.9 (0.877) a Pe Pe  a Pe


a Py
c Pn Pe 4  c Pn  Pn
 0.658 Pe for    0.390  Pe 

c Py Py 9   c Py  0.877

0.877 a Py
c Pn  P   0.877 a Py

 0.658 Pn
ln c n   ln 0.658 Pn 
c Py    
 c Py   
 P  P
ln  c n   0.877a c y ln  0.658  Pu  Pu 
 P  c Pn  P 
 c y for  u  0.390  a  2.724 ln  
 P
 c y

 c Py  c Py 
 P 
for  u  0.390  a  1
 P 
 c y 

2016 NASCC
D. White, Georgia Tech 3
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

ANALYSIS STEPS

• Build a model of the structure
• Apply the desired LFRD factored loads. These loads produce the 
internal axial forces Pu
• Reduce EIx, EIy, ECw and GJ by  SRF  =  0.9 x 0.877x a
• Solve for the inelastic buckling load
 Vary the applied loads by the scale factor 
 Calculate τa at the current load level
 Iterate until the  assumed in the calculation of a is the same as that determined    
from the buckling analysis

The resulting Pu is a rigorous calculation of cPn accounting for       


all member continuity, bracing and/or end restraint effects 7

EXAMPLE COLUMN INELASTIC BUCKLING ANALYSIS

• SABRE2 (using the inelastic reduction 
factor a) gives

cPn = 1153 kip

• This result matches with a traditional 
iterative calculation (Yura 1971) using an 
inelastic  K = 0.861, based on a = 0.633

2016 NASCC
D. White, Georgia Tech 4
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

COLUMN STIFFNESS REDUCTION FACTORS (SRFs)

Net SRF

Pu / Py

Pu  Pu   P    Pu 


a  2.724 ln   b  4  u  1    
c Py  c Py 
 P
 y    Py   9

BEAM ltb MODEL, COMPACT & NONCOMPACT WEB MEMBERS

FL Mu
bMn  bMe  0.9ltbMe For m  where m  ltb  1
Fyc bMyc

FL M Y 4X2
For  m  b max .LTB ltb 
Fyc bMyc  2  Fyc 
2

6.76 X   m2  2Y 2 
  E  
where:

  m  Sxc ho
  1    X2 
  Rpc   Lr Lp  Lp   Fyc  1  J
Y  m      
 F   rt rt  rt   E   1.95 
1  L
  Mmax .LTB  Rpc Myc
  Rpc Fyc  
 
10

2016 NASCC
D. White, Georgia Tech 5
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

COLUMN VS. BEAM  FACTORS FOR A W21X44

1.2

1.0

0.8

 0.6
Beam LTB Tau
0.4 Factor
Column Tau 0.223
0.2 Factor

0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Pu Mu
,
c Pye bMmax 11

REQUIREMENTS FOR INELASTIC LTB ANALYSIS

• The software must rigorously include ECw in addition to EIx, EIy & 


GJ in the context of doubly‐symmetric I‐section members
• For singly‐symmetric members, the behavior associated with the 
monosymmetry factor, βx, also must be included 
• The 0.9 x ltb factor should be applied equally to the member 
elastic stiffness contributions EIy, ECw and GJ for the execution of 
the buckling analysis
• Required number of elements:
At least 4 elements per unbraced length are required to capture the 
behavior for frame elements based on cubic Hermitian interpolation of 
the transverse displacements and twists along the element length 12

2016 NASCC
D. White, Georgia Tech 6
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

GENERAL PURPOSE THIN‐WALLED OPEN‐SECTION FRAME ELEMENT

Implemented in SABRE2 (available at white.ce.gatech.edu/sabre)

13

ELASTIC LTB BENCHMARK

14

2016 NASCC
D. White, Georgia Tech 7
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

ELASTIC LTB BENCHMARK

Typ.14 dof prismatic element (10 elem)


stepped using avg. depth in ea. elem.
Typ.14 dof prismatic element (10 elem)
stepped using smallest depth in ea. elem.

15

BASIC BEAM‐COLUMN EXAMPLE USING SABRE2
From AISC/MBMA Design Guide 25
Point Brace, i = 0.825 k/in

11.3 kips

1800 kip-in
90 in 
144 in
Dimensions:
Fy = 55 ksi
At left end: At right end:
Simply-supported
bft = 6 in bft = 6 in end conditions
tft = 0.2188 in tft = 0.2188 in
bfc = 6 in bfc = 6 in
tfc = 0.3125 in tfc = 0.3125 in
SABRE2 available at:
h = 12 in h = 24 in white.ce.gatech.edu/sabre
tw = 0.125 in tw = 0.125 in 16

2016 NASCC
D. White, Georgia Tech 8
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

SRF DIAGRAM

17

SABRE2 VS DG25

vs

1.173 DG25 Solution A

1.138 DG25 Solution C

18

2016 NASCC
D. White, Georgia Tech 9
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

BEAM ltb MODEL – RESULTS (W21X44 BEAMS) 

1.2
Uniform Moment
1.0

0.8
Moment Gradient
Mn /Mp

(Cb = 1.75)
0.6 AISC Specification 
Chapter F

0.4

0.2 Moment Gradient,
Inelastic Buckling 
Analysis (SABRE2)
0.0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
Lb /Lp
19

BEAM ltb MODEL – RESULTS (W21X44 BEAMS) 

1.2
Uniform Moment
1.0

0.8
Moment Gradient
Mn /Mp

(Cb = 1.75)
0.6 AISC Specification 
Chapter F

0.4

0.2 Moment Gradient,
Inelastic Buckling 
Analysis (SABRE2)
0.0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
Lb /Lp
20

2016 NASCC
D. White, Georgia Tech 10
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

W21X44 BEAM LTB RESISTANCE VS. POINT BRACING STIFFNESS

(Flexurally & torsionally simply-supported end conditions, Uniform Moment,


One intermediate point brace at compression flange, Lbr1 = 8.1 ft, Lbr2 = 4.9 ft )

Inelastic Buckling Analysis

AISC 2016 Commentary 
Provisions including “Lq”

21

W21X44 BEAM LTB RESISTANCE VS. POINT BRACING STIFFNESS

(Flexurally & torsionally simply-supported end conditions, Uniform Moment,


One intermediate point brace at compression flange, Lbr1 = 8.1 ft, Lbr2 = 4.9 ft )

Inelastic Buckling Analysis

AISC 2016 Commentary 
Provisions including “Lq”

22

2016 NASCC
D. White, Georgia Tech 11
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

W21X44 BEAM LTB RESISTANCE VS. POINT BRACING STIFFNESS

(Flexurally & torsionally simply-supported end conditions, Uniform Moment,


One intermediate point brace at compression flange, Lbr1 = 8.1 ft, Lbr2 = 4.9 ft )

Inelastic Buckling Analysis

AISC 2016 Commentary 
Provisions including “Lq”

23

TRANSFER GIRDER ASSESSMENT

Mp D 2Dc 2Dcp Dc D
 1.40,  160,  109,  32,  3.413,  1.465
My tw tw tw Dcp 2Dc

Lateral brace (TYP) P 0.5P
xx

xx

xx

xx

3 at Lb = 45 ft = 135 ft

Critical Middle Unbraced Length:


Cb = 1.10 K = 0.848

Girder Factored Load Capacity: Pmax = 361 kip … from manual calcs.
24

2016 NASCC
D. White, Georgia Tech 12
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

INELASTIC BUCKLING MODE

Pmax = 376 kip

25

MOMENT & SRF DIAGRAMS

MOMENT

SRF

26

2016 NASCC
D. White, Georgia Tech 13
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

CROSS‐SECTION UNITY CHECK

27

BEAM‐COLUMN SRF

• Calculate the UC value with respect to the cross‐section strength from 
Eqs. H1‐1
UC = Pu /c Pye + 8/9 Mu /b Mmax for  Pu /cPye > 0.2
UC = Pu /2c Pye + Mu/b Mmax for  Pu /cPye <  0.2
• Use the UC value in the a & ltb eqs. instead of  Pu /c Pye & Mu /bMmax
 P / P 
• Determine the angle   a tan  u c ye 
 Mu / bMmax 
• Calculate the net SRF applied to ECw , EIy & GJ as
   A   
SRF   o  0.9 x 0.877 x a e   1  o  0.9Rb ltb
 90  Ag  90  28

2016 NASCC
D. White, Georgia Tech 14
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

BEAM‐COLUMN ltb MODEL – W21X44 RESULTS 

Simply‐supported members, moment gradient loading
1.0
Fully‐Effective Cross‐
0.9
Section Plastic Strength
0.8
0.7 L = 7.5 ft
0.6
cPn /cPy

0.5
0.4 L = 10 ft
0.3
0.2
L = 15 ft
0.1
0.0
0.0 0.2 0.4 0.6 0.8 1.0
bcM cbM
Mnn // Mpp 29

ROOF GIRDER EXAMPLE (ADAPTED FROM AISC 2002)

2 in

(G' = 1 kip/in)

30

2016 NASCC
D. White, Georgia Tech 15
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

ROOF GIRDER EXAMPLE
 = 0.921
b Mn = 230 kip‐ft
c Pn = 18.4 kips

31

SRF, GRAVITY LOAD CASE AT BUCKLING LOAD 
1

0.9

0.8

0.7

0.6
Net SRF

0.5

0.4

0.3

0.2

0.1

0
0 10 20 30 40 50 60 70

Position along girder length (ft)
32

2016 NASCC
D. White, Georgia Tech 16
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

CLEAR‐SPAN FRAME EXAMPLE

SYM

33

BUCKLING MODE & CONTROLLING LIMIT STATE INFORMATION

1.2 (Dead + Collateral + Self-Weight)


+ 1.6 Uniform Snow

34

2016 NASCC
D. White, Georgia Tech 17
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

AXIAL FORCES AT STRENGTH LIMIT

35

MOMENTS AT STRENGTH LIMIT

36

2016 NASCC
D. White, Georgia Tech 18
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

SRF VALUES AT STRENGTH LIMIT

37

bMmax VALUES AT STRENGTH LIMIT

38

2016 NASCC
D. White, Georgia Tech 19
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

cPye VALUES AT STRENGTH LIMIT

39

CROSS‐SECTION UNITY CHECKS AT STRENGTH LIMIT

40

2016 NASCC
D. White, Georgia Tech 20
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

ADVANTAGES OF BUCKLING ANALYSIS APPROACH

• More general and more rigorous handling of all types of bracing, 
end restraint & continuity effects
• Substantially cleaner, more streamlined & less error prone 
member strength calculations
• Consistent bracing stiffness & member strength assessments
• More accurate capture of 
Moment gradient and other load & displacement b.c. effects
Tapered & stepped member geometry effects 
via a continuous representation of the corresponding SRF 
values along the member lengths 41

COMPLEMENTARY RESEARCH 

• Trahair, N.S. and Hancock, G.J. (2004). “Steel member strength by inelastic lateral buckling,” Journal of
Structural Engineering, ASCE, 130(1), 64–69.
• Trahair, N.S. (2009). “Buckling analysis design of steel frames,” Journal of Constructional Steel Research,
65(7), 1459-63.
• Trahair, N.S. (2010). “Steel cantilever strength by inelastic lateral buckling,” Journal of Constructional Steel
Research, 66(8-9), pp 993-9.
• Kucukler, M., Gardner, L., Macorini, L. (2014). “A stiffness reduction method for the in-plane design of
structural steel elements”, Engineering Structures, 73, 72–84.
• Kucukler, M., Gardner, L., Macorini, L. (2015a). “Lateral–torsional buckling assessment of steel beams
through a stiffness reduction method”, Journal of Constructional Steel Research, 109, 87–100.
• Kucukler, M., Gardner, L., and Macorini, L. (2015b). “Flexural-torsional buckling assessment of steel
beam-columns through a stiffness reduction method”. Engineering Structures, 101, 662-676.
• Kucukler, M., Gardner, L., and Macorini, L. (2015c). “In-plane design of steel frames through a stiffness
reduction method”. Journal of Constructional Steel Research. in press.
• Gardner, L. (2015). “Design of Steel Structures to Eurocode 3 and Alternative Approaches”. Proceedings,
International Symposium on Advances in Steel and Composite Structures, Hong Kong, 39-51.

42

2016 NASCC
D. White, Georgia Tech 21
Improved Design Assessment of LTB 4/20/2016
of I-Section Members
via Modern Computational Methods

DESIGN METHOD REQUIREMENTS
• Traditional Methods
 Effective Length Method (ELM)  Direct Analysis Method (DM)

} }
 ∅ , , INELASTIC  ∅ , , INELASTIC
BUCKLING BUCKLING
 ∅ , ,  ∅ , ,
ANALYSIS ANALYSIS
 2nd‐Order Elastic Analysis  2nd‐Order Elastic Analysis 
– Joint out‐of‐alignment ∆ – Stiffness reductions 0.8 & 0.8
(gravity only load cases) – Joint out‐of‐alignment ∆

• Advanced Methods
 AISC (2016) App. 1.2 (Elastic Analysis)  AISC (2016) App. 1.3 (Inelastic Analysis)

}
 ∅ ∅ INELASTIC  2nd‐Order Inelastic Analysis
 ∅ , , BUCKLING – 0.9 & 0.9
ANALYSIS
 2nd‐Order Elastic Analysis – Spread of yielding including residual 
– Stiffness reductions 0.8 & 0.8 stress effects
– Joint out‐of‐alignment ∆ – Joint out‐of‐alignment ∆
– Member out‐of‐straightness  – Member out‐of‐straightness 
43
– For TWOS members,  – For TWOS members, 

THANKS FOR YOUR ATTENTION! 

I will be happy to address                                                           
any questions 

SABRE2 available at:


white.ce.gatech.edu/sabre

44

2016 NASCC
D. White, Georgia Tech 22

You might also like