Professional Documents
Culture Documents
.cscworld.com/orion
HANDBOOK
CSC Orion
Disclaimer
Disclaimer
CSC (UK) Ltd does not accept any liability whatsoever for loss or damage arising from any errors which
might be contained in the documentation, text or operation of the programs supplied.
It shall be the responsibility of the customer (and not CSC)
to ensure that program operation is carried out in accordance with the user manuals,
at all times paying due regard to the specification and scope of the programs and to the CSC Software
Licence Agreement.
Proprietary Rights
CSC (UK) Ltd, hereinafter referred to as the OWNER, retains all proprietary rights with respect to this
program package, consisting of all handbooks, drills, programs recorded on CD and all related materials.
This program package has been provided pursuant to an agreement containing restrictions on its use.
This publication is also protected by copyright law. No part of this publication may be copied or
distributed, transmitted, transcribed, stored in a retrieval system, or translated into any human or
computer language, in any form or by any means, electronic, mechanical, magnetic, manual or
otherwise, or disclosed to third parties without the express written permission of the OWNER.
This confidentiality of the proprietary information and trade secrets of the OWNER shall be construed in
accordance with and enforced under the laws of the United Kingdom.
Orion documentation:
CSC (UK) Ltd 2010
All rights reserved.
Orion software:
CSC (UK) Ltd 2010
All rights reserved.
Trademarks
Orion is a trademark of CSC (UK) Ltd.
Fastrak is a trademark of CSC (UK) Ltd.
TEDDS is a registered trademark of CSC (UK) Ltd.
The CSC logo is a trademark of CSC (UK) Ltd.
Autodesk and Revit are registered trademarks or trademarks of Autodesk Inc. in the USA and/or other
countries.
Microsoft and Windows are either trademarks or registered trademarks of Microsoft Corporation in the
United States and/or other countries.
Acrobat Reader Copyright 1987-2010 Adobe Systems Incorporated. All rights reserved.
Adobe and Acrobat are trademarks of Adobe Systems Incorporated which may be
registered in certain jurisdictions.
All other trademarks acknowledged.
Table of Contents
Chapter 1
Overview .
Chapter 2
Modelling Techniques .
Chapter 3
.
.
.
.
.
Introduction
.
.
.
.
.
Modelling Analysis and Design Flowchart
Build the Model .
.
.
.
.
Derive Beam Loads .
.
.
.
General Building Analysis .
.
.
Design and Detailing .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
15
15
15
16
17
17
17
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
18
18
19
19
20
23
24
25
26
28
29
29
31
31
32
35
39
39
41
41
42
42
43
43
44
44
44
45
46
46
47
47
49
49
49
49
50
50
50
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
51
51
51
52
53
54
.
.
.
.
Introduction
.
.
.
.
.
.
.
Modelling Inclined and Lowered Members .
.
Sloping and Lowered Slabs
.
.
.
.
How To Drop Parts of a Slab Panel .
.
.
.
Sloping and Lowered Beams
.
.
.
.
Sloping and Lowered Columns .
.
.
.
Sloping and Lowered Walls
.
.
.
.
Working With Planes
.
.
.
.
.
Modelling Curved Axes and Beams
.
.
.
Curved Axes .
.
.
.
.
.
.
The Curved Axis Generator .
.
.
.
.
Curved Beams .
.
.
.
.
.
.
The Curved Beam Generator
.
.
.
.
How many segments to use?
.
.
.
.
Linking Angled Beams.
.
.
.
.
.
Columns and Walls Spanning More Than One Storey
Example Case Study .
.
.
.
.
.
User Defined Supports
.
.
.
.
.
What is a default support? .
.
.
.
.
When might a default support be inappropriate? .
Specifying a User Defined Support
.
.
.
Applying a User Defined Support
.
.
.
User Defined Supports - Trouble shooting .
.
Stepped Foundation Levels .
.
.
.
.
Default Supports Method .
.
.
.
.
Single Storey Example of the Default Supports Method .
Two Storey Example of the Default Supports Method .
User Defined Supports Method .
.
.
.
Example of the User Defined Supports Method .
.
Beams with Varying Depth .
.
.
.
.
Example Case Study .
.
.
.
.
.
Pinned Member Ends .
.
.
.
.
.
To Pin a Single Column .
.
.
.
.
To Pin Multiple Columns .
.
.
.
.
To Remove the Hinges from the Columns .
.
To Pin a Single Beam
.
.
.
.
.
To Pin Multiple Beams
.
.
.
.
.
To Remove the Hinges from the Beams
.
.
General .
.
.
Modifying Beam Loads
Beam Loads Dialog .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
How to Define a New Point Load
.
How to Define a New Uniformly Distributed Load .
.
.
.
.
.
Table of Contents
.
.
.
.
Using the vertex table
.
.
How to Define a New Load on an Inclined Beam
.
.
Switching between Yield Line and FE load Decomposition.
What is Load Decomposition?
.
.
.
.
.
Why switch to an FE method?
.
.
.
.
.
Example of FE Method for Slab Load Decomposition
.
Why retain the traditional (yield line) method?
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
54
54
56
56
58
58
58
60
65
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
66
66
66
66
66
67
.
.
.
.
.
.
.
.
Limitations - diaphragm modelling and inclined planes .
Case Study 1 - single storey pitched frame .
.
.
.
Case Study 2 - storeys linked by inclined planes
.
.
.
Global Constraints
.
.
.
.
.
.
.
Pattern Loading .
.
.
.
.
.
.
.
Rigid Zones
.
.
.
.
.
.
.
.
Rigid Zones None .
.
.
.
.
.
.
Rigid Zones Reduced by 25%
.
.
.
.
.
Rigid Zones Max
.
.
.
.
.
.
.
Discussion
.
.
.
.
.
.
.
.
Rigid Links
.
.
.
.
.
.
.
.
Shear Walls and Core Wall Systems
.
.
.
.
3D Effects .
.
.
.
.
.
.
.
.
Continuous Beams
.
.
.
.
.
.
.
Effects of one Member on Another .
.
.
.
.
Sway Effects .
.
.
.
.
.
.
.
Transfer Beams .
.
.
.
.
.
.
.
Stiffness adjustments .
.
.
.
.
.
.
Flat Slab Construction .
.
.
.
.
.
.
Transfer Levels .
.
.
.
.
.
.
.
Supports .
.
.
.
.
.
.
.
.
Building Analysis Problems Reviewing/Understanding .
What are Errors? .
.
.
.
.
.
.
.
What are Warnings? .
.
.
.
.
.
.
Building Model Validity Checking Errors .
.
.
.
Beam Load Analysis Errors .
.
.
.
.
.
Building Analysis Errors and Warnings .
.
.
.
Overview of Axial Load Comparison Report .
.
.
Table 1 .
.
.
.
.
.
.
.
.
Table 2 .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. 68
. 68
. 68
. 69
. 70
. 71
. 72
. 72
. 73
. 74
. 77
. 80
. 80
. 81
. 83
. 84
. 85
. 86
. 87
. 89
. 89
. 90
. 92
. 94
. 95
. 95
. 96
. 96
. 96
. 98
. 98
. 98
. 99
. 99
.100
.101
.104
.105
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Slabs to define rigid diaphragms (Default Setting) .
Single rigid diaphragm at each floor level .
.
No rigid diaphragm floor levels
.
.
.
Excluding Specific Slabs from Diaphragms .
.
.
.
.
.
.
.
.
.
Chapter 4
Analysis Methods
Introduction.
.
.
Chapter 5
Introduction.
.
Structural Model .
.
.
Diaphragm Modelling .
.
.
.
Table of Contents
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. 109
. 109
. 110
.
.
Introduction
.
.
.
.
Eigenvalue Analysis Parameters .
Controlling the Storey Mass
.
Model Stiffness
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
114
114
114
114
116
116
117
117
118
119
119
119
.
.
.
.
.
.
.
.
Combining the entire structure into a single stage
.
Setting the Duration of Each Stage
.
.
.
Analysis Properties .
.
.
.
.
.
Modulus of Elasticity.
.
.
.
.
.
FE Merging .
.
.
.
.
.
.
Simulaneously designing for FE merged forces and the results of a completely unstaged analysis. .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
120
120
120
122
122
122
124
125
125
126
126
126
127
128
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
129
129
129
130
131
131
131
131
131
132
132
134
134
134
135
135
135
136
137
138
139
141
144
Chapter 6
.
.
.
.
.
.
Eigenvalue Analysis .
Analysis
.
.
.
Exporting to S-Frame .
.
.
.
.
Exporting to S-Frame Example .
Graphical Results
Numerical Results
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Chapter 7
.
.
.
.
.
.
.
Simple Example
.
.
.
.
.
Staged Construction Modelling and Analysis.
Model Creation
.
.
.
.
.
Staged Loading Creation .
.
.
.
Stage Control .
.
.
.
.
.
Combining multiple floors into a single stage .
Introduction
Chapter 8
.
.
.
Column / Shear Wall Model Type
.
Beam Stiffness Multiplier .
.
.
Slab Stiffness Multiplier .
.
.
Column and Wall Stiffness Multipliers .
Cracking and Creep .
.
.
.
Include Column Sections in FE Model .
Include Slab Plates in FE Model .
.
Consider Beam Torsional Stiffness
.
Include Upper Storey Column Loads .
Upper Storey Column Loads Table
.
Floor Mesh and Analysis .
.
.
Batch FE Chasedown
.
.
.
Meshing and Analysing your Model .
If Slab Plates are NOT included .
.
If Slab Plates are included .
.
.
Validity Checking .
.
.
.
Mesh Density .
.
.
.
.
Mesh Uniformity
.
.
.
.
Effect of Holes and Boundaries
.
.
Mesh Sensitivity
.
.
.
.
Introduction
.
.
.
Model Generation Options .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Table of Contents
.
.
.
.
.
.
.
.
Plots Including Wood and Armer adjustments
.
Steel Reinforcing Requirement Contours .
.
Other Contouring Adjustment Options
.
.
Slab Design
.
.
.
.
.
.
Effects of Adjusting the Beam and Slab Stiffnesses .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.144
.145
.145
.146
.148
.149
.149
.150
.150
.153
.154
.155
.159
.163
.164
.166
.166
.166
.167
.170
.171
.173
.175
.176
.176
.176
.
.
.
.
.
.
.177
.177
.
.
What does this option do? .
.
.
.
When might you use this option? .
Example .
.
.
.
.
Option 1 A Plateless Model
.
Checking the Beam Designs .
.
Solution 2 A Meshed Model
.
Checking the Beam Designs .
.
Merging Column Results
.
.
What does this option do? .
.
When might you use this option? .
Chapter 9
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Analytical Idealisations .
.
.
.
Deflection .
.
.
.
.
.
Cases where one wall option is Preferable
Option to Check Both Ways .
.
.
Wall Panel Design .
.
.
.
.
Forcing Walls to resist all lateral loads .
Adjust Model Stiffnesses
.
.
.
Pin the Columns
.
.
.
.
Sway Effects Under Gravity Load .
.
Fully Framed Structures
.
.
.
Why does this sway happen? .
.
.
Structures Incorporating Flat Slab Areas .
Slab Loads Yield Line Decomposition
.
Slab Loads FE Decomposition
.
.
Discussion
.
.
.
.
.
Load Eccentricity
.
.
.
.
Construction and Creep Effects
.
.
Discrete Cores .
.
.
.
.
Results based on an FE Chase Down
.
Closing Summary .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.178
.178
.178
.178
.179
.179
.180
.180
.181
.181
.181
.182
.184
.184
.186
.187
.187
.188
.189
.190
.191
.192
Introduction.
Chapter 10
Chapter 11
Table of Contents
Introduction
.
.
.
.
.
.
.
.
Comparison of Orion's alternative wall modelling options
Compare analysis results using other software
.
.
Chapter 12
.
Automatic Assessment of Sway Sensitivity .
User Defined Bracing
.
.
.
.
Classification Requirements of each code .
BS8110 (similarly CP65 and HK-2004) .
.
.
EC2
.
.
.
.
.
.
.
Implementation of EC2 Classification in Orion
Setting the Braced/Bracing Members .
.
Assessment of Sway Sensitivity .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Chapter 13
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. 192
. 192
. 193
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
195
195
195
196
197
197
. ACI 318-02198
.
. 199
.
. 201
.
. 201
.
. 201
.
. 202
.
. 203
.
. 203
.
. 205
.
. 207
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
208
209
211
212
214
215
Building Analysis, area factor adjustment methods
216
Why do the two models give such different answers? .
218
How do we eliminate this effect if we want to? .
.
219
Increasing the column area factor
.
.
.
220
What is a reasonable upper limit for the column area factor adjustment?
223
Detailed comparisons of the analysis results .
.
.
.
223
Recommendation .
.
.
.
.
.
.
.
225
Is it acceptable to simply emulate the traditional design result?
226
Ignoring differential axial displacements
.
.
.
.
226
Making allowances for differential axial displacements .
.
226
What is a reasonable lower limit for the area factor adjustment?
227
What is the impact on the design when both upper and lower-bounds are taken into consideration? 227
Consider Case where AF = 2.7 is upper bound and AF = 1.5 is the lower bound:
.
.
.
.
. 228
Consider Case where AF = 4.0 is upper bound and AF = 1.0 is the lower bound:
.
.
.
.
. 228
Consider Case where FE Chasedown is upper bound and AF = 1.5 is the lower bound:
.
.
.
. 228
What is the impact on the design when Pattern Loading is Introduced? .
.
.
.
. 228
Conclusion on Design Impact .
.
.
.
.
.
.
.
.
.
.
. 229
Overall Summary of Suggested Procedure .
.
.
.
.
.
.
.
.
. 230
Closing Discussion .
.
.
.
.
.
.
.
.
.
.
.
.
. 231
Typical Concerns .
.
.
.
.
.
.
.
.
.
.
.
.
. 231
What answers are we trying to get? .
.
.
.
.
.
.
.
.
.
. 231
Fixed Values for Area Adjustment Factor
.
.
.
.
.
.
.
.
.
. 231
Side Effects on Lateral Load Analysis .
.
.
.
.
.
.
.
.
.
. 232
Will FE Chasedown also eliminate differential axial deformation?
.
.
.
.
.
. 232
Is this going to result in uncompetitive over design?
.
.
.
.
.
.
.
. 233
Could I avoid all this complication if Staged Construction Analysis were used? .
.
.
. 233
DADE Analysis & Design Flowchart
.
.
.
.
.
.
.
.
.
.
. 234
Chapter 14
Table of Contents
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.236
.236
.236
.237
.237
.237
.237
.238
.239
.241
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Building Analysis with Area Factor adjustment
.
Staged Construction Analysis Result
.
.
Does Staged Analysis eliminate DADE? .
.
Use of Staged Construction as the upper bound solution .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.242
.242
.
.
.
.
.
.
Overview .
.
.
.
.
.
.
Slab Strip Errors Reviewing/Understanding .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.243
.243
.243
.243
.243
.
.
.
.
1. Check slenderness limits for lateral stability- Cl 3.4.1.6 .
2. Rectangular or flanged - Cl 3.4.1.5
.
.
.
.
3. Analysis of Sections - Cl 3.4.4.1 .
.
.
.
.
4. Design for Bending- Cl 3.4.4.4 and Cl 3.4.4.5 .
.
.
5. Design for Shear- Cl 3.4.5 .
.
.
.
.
.
6. Deflection Checks- Cl 3.4.6
.
.
.
.
.
Worked Example .
.
.
.
.
.
.
.
The Design Model
.
.
.
.
.
.
.
Beam Design Settings .
.
.
.
.
.
.
Analysis Results .
.
.
.
.
.
.
.
Performing the Design
.
.
.
.
.
.
Design for Bending - Cl 3.4.4.4
.
.
.
.
.
Design for Shear - Cl 3.4.5 .
.
.
.
.
.
Deflection Checks- Cl 3.4.6 .
.
.
.
.
.
Output Calculations .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.244
.244
.245
.245
.245
.246
.246
.246
.249
.251
.251
.251
.252
.253
.254
.254
.256
.258
.259
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.260
.260
.260
.260
.261
.262
.262
.263
.264
.264
.265
.265
.267
Chapter 15
Chapter 16
.
.
.
.
.
.
Slab Design .
Chapter 17
.
.
.
.
Beam Design Settings .
.
.
The BS8110 Beam Design Process .
Introduction.
Chapter 18
Beam Detailing .
.
.
Introduction.
.
.
.
The Design and Detailing Process
Overview .
.
.
.
The Design Tab .
.
.
The Parameters Tab .
.
The Bar Selection Tab .
.
The Curtailment Tab .
.
The Detailing Tab
.
.
The Layers Tab .
.
.
Overview of Patterns .
.
.
.
.
.
.
.
.
.
.
.
.
Pattern 1 The Splice Bar Method .
Pattern 2 The Alternative Method
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Table of Contents
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
268
269
270
270
270
272
272
273
274
277
282
284
285
286
288
289
290
292
294
295
298
299
.
.
.
1. Braced or unbraced - Cl 3.8.1.5
.
2. Calculate effective height- Cl 3.8.1.6 .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
3. Check slenderness limits- Cl 3.8.1.7 & 3.8.1.8
.
4. Classify as short or slender- Cl 3.8.1.3
.
.
5. If slender - calculate M_add- Cl 3.8.3.1
.
.
6. Calculate minimum moments - Cl 3.8.2.4 .
.
7a. If braced, calculate design moments about each axis - Cl 3.8.3.2 .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
301
301
302
302
303
303
303
303
304
304
304
304
305
306
306
307
309
310
310
311
312
312
312
314
314
315
316
317
319
321
322
322
322
.
.
.
.
.
.
.
.
Summary .
Chapter 19
.
.
.
.
.
.
.
.
.
Introduction
.
.
.
.
The BS8110 Column Design Process
Worked Examples
.
.
1. Braced or unbraced - Cl 3.8.1.5
.
2. Calculate effective height- Cl 3.8.1.6 .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Table of Contents
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.323
.323
.323
.324
.324
.325
.325
.325
.326
.326
.326
.326
.327
.328
.328
.328
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Mesh and Minimum Reinforcement Requirements .
Points to Consider
.
.
.
.
.
.
Which Mesh Type is Better? .
.
.
.
.
Overview of Possible Reinforcement Arrangements .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Design using Type A Mesh (not considering plain wall design option) .
Design using Type A Mesh (considering plain wall design option)
.
Design using Type B Mesh (considering plain wall design option)
.
Limitation Copy/Paste Bars will not work
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.333
.333
.333
.333
.333
.335
.335
.336
.338
.339
.340
.340
.340
.341
.341
.342
.343
.344
.345
.346
.347
.347
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.348
.348
.348
.349
.350
.351
.351
.351
.351
.352
.352
.353
.353
.354
.354
.
.
2. Calculate effective height- Cl 3.8.1.6 .
.
.
.
Chapter 20
Chapter 21
Foundation Design .
.
Introduction.
.
.
.
.
Conservatism in the design method
Wall Design and Detailing Options .
Design With End Zones
.
.
Design Without End Zones .
.
Should I use End Zones?
.
.
Plain Wall Design
.
.
.
.
.
.
.
.
.
.
.
Option to use Single Layer of Reinforcement .
Design with Mesh Reinforcement .
.
.
The "Revert to Loose Bar" Option
.
.
.
How the "Revert to Loose Bar" Option Works
.
Limitation of "Revert to Loose Bar" Option .
.
Column Steel Details View .
.
.
.
.
.
Foundation Design Settings .
Foundation Depth
.
.
.
.
.
.
The Foundation Forces Table
.
Combining Columns and Walls for Shared Foundation Design .
To combine multiple columns and walls
.
.
.
.
.
To ungroup columns and walls
.
.
.
.
.
.
Calculation of the Combined Footing Design Forces
.
.
.
Creating a Typical Pad/Pile Footing for Multiple Foundations .
To create a typical footing .
.
.
.
.
.
.
Pad Footings
.
.
.
.
.
.
.
.
.
Defining a Pad Footing
.
.
.
.
.
.
.
Pad Footing Details .
.
.
.
.
.
.
.
Combined Pad Footings
.
.
.
.
.
.
.
Introduction.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Table of Contents
Pile Caps .
.
.
.
.
.
.
.
Strip Footings .
.
.
Defining a Strip Footing .
Strip Analysis Options
.
Defining a Pile Cap .
Basic Design Procedure
Limitations .
.
Larger Pile Groups .
Pile Cap Details
.
Combined Pile Caps .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Chapter 22
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Overview of Solution Options and Limitations
.
.
Inclined Beam and Slab Loads .
.
.
.
.
Building Analysis Worked Example
.
.
.
.
Diaphragm Modelling
.
.
.
.
.
.
Simplified Load Decomposition .
.
.
.
.
Analysis .
.
.
.
.
.
.
.
.
Load Comparison Check .
.
.
.
.
.
Switching to FE Beam Load Decomposition
.
.
.
Design and Detailing
.
.
.
.
.
.
Inclined Beam Design
.
.
.
.
.
.
Inclined Beam Detail Drawings and Quantities .
.
.
Inclined Column Design .
.
.
.
.
.
Inclined Column Detail Drawings and Quantities.
.
.
Tapered Wall Design.
.
.
.
.
.
.
Tapered Wall Detail Drawings and Quantities .
.
.
Design and Detailing of the Inclined Slabs
.
.
.
Tapered Wall Modelling .
.
.
.
.
.
General Limitations - Inclined/Lowered Members .
.
Load decomposition for lowered slabs .
.
.
.
FE Analysis Worked Example
.
.
.
.
.
Introduction .
.
.
.
.
.
.
.
Example Model
.
.
.
.
.
.
.
FE Chasedown Analysis .
.
.
.
.
.
FE Model Generation
.
.
.
.
.
.
Load Comparison Check .
.
.
.
.
.
Member Design based on FE Analysis .
.
.
.
Limitations - FE Analysis of Inclined/Lowered Members .
Limitations - Finite Element Analysis and Building Sway .
S-Frame Comparison
.
.
.
.
.
.
Typical Test Model Results .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
355
355
357
357
358
359
359
360
360
364
364
369
370
372
374
375
375
375
375
376
376
377
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
379
379
380
380
381
381
383
383
384
385
387
387
387
388
388
388
389
389
389
392
392
397
397
397
397
397
399
400
401
401
402
403
404
Table of Contents
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.404
.405
.405
Discussion
Conclusions
Chapter 23
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.406
.406
.406
.407
.407
.407
.
Modelling and Analysis .
.
.
.
.
Analysis Model Options
.
.
.
.
Model Tab
.
.
.
.
.
.
Stiffnesses Tab .
.
.
.
.
.
Settings Tab .
.
.
.
.
.
Analysis .
.
.
.
.
.
.
Load Comparison Check
.
.
.
.
Design and Detailing of the Transfer Beams
.
Discussion of Frame Analysis Results
.
.
Gravity Loads (Mid-Pier Wall Modelling) .
.
Building Analysis Results
.
.
.
.
Front Transfer Beam .
.
.
.
.
Rear Transfer Beam .
.
.
.
.
Frame Action .
.
.
.
.
.
Gravity Loads (FE Meshed Wall Modelling)
.
Building Analysis Results
.
.
.
.
Front Transfer Beam .
.
.
.
.
Rear Transfer Beam .
.
.
.
.
Frame Action .
.
.
.
.
.
Limitations Transfer Walls .
.
.
.
Supporting Beam to carry all Wall Load .
.
No supporting Beam Wall to act as a Deep Beam .
Beam and Wall to Work Together .
.
.
.
Limitations Walls Supported by more than 1 Beam
Analysis .
.
.
.
.
.
.
.
Results based on Mid-Pier Wall Idealisation .
.
Results based on Meshed Wall Idealisation
.
.
Alternative Modelling Option Split the wall .
.
Mid-Pier Model .
.
.
.
.
.
.
Meshed Model .
.
.
.
.
.
.
Summary/Recommendations
.
.
.
.
Option 1 Do not split the wall
.
.
.
.
Option 2 Split the wall
.
.
.
.
.
Introduction.
.
.
.
.
.
.
Understanding the Problem and the Limitations
Where Beams support Columns and Walls
.
Where Slabs support Columns and Walls
.
Key Limitation .
.
.
.
.
.
Chapter 24
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.409
.409
.409
.409
.410
.410
.411
.412
.413
.414
.414
.415
.416
.417
.418
.420
.421
.422
.423
.424
.425
.426
.427
.428
.432
.433
.434
.435
.436
.437
.438
.439
.439
.439
Chapter 25
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.440
.440
.441
.443
.444
.445
.447
.447
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Table of Contents
.
.
.
Beam Design .
.
.
.
FE Chase Down with Duplicate Floors
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
448
448
450
452
Chapter 26
Chapter 27
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
454
454
455
459
460
461
461
461
462
462
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
464
464
465
466
466
469
470
470
476
477
477
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
478
478
478
478
478
479
479
480
480
482
482
486
492
498
.
.
.
.
.
Check Shear on Series of Perimeters .
Simple Examples
.
.
.
.
Checking a Typical Internal Column .
Performing the Check
.
.
.
Checking Maximum Shear Capacity
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
499
499
499
499
499
499
500
501
502
502
.
.
.
.
.
.
.
.
.
.
.
.
The FE Analysis and Load Chase Down .
Axial Load Comparison
.
.
.
Merging Column Analysis Results .
.
Merging Beam Analysis Results .
.
Column and Wall Positioning
Slab Insertion .
.
.
Building Analysis .
.
Alternative Modelling Option
Concluding Note on Modelling
Chapter 28
.
.
.
.
.
.
.
.
.
.
Example of a more Irregular Model
Overview
.
.
.
.
Braced Buildings
Un-Braced Buildings .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Column Design
Chapter 29
.
Introduction
.
.
.
.
BS8110 Design Code Requirements
The Design Procedure
.
.
Check Maximum Shear Capacity .
Table of Contents
.
.
.
Checking a Typical Corner Column
.
Column Drop Panels .
.
.
.
Dealing with Openings
.
.
.
Openings which have been modelled
.
Calculation of the Effective Shear Force
.
Allowing for Openings which have not been modelled
.
Final Batch Check and Output
.
.
.
.
Concluding Notes .
.
.
.
.
.
.
Limitations
.
.
.
.
.
.
.
Holes .
.
.
.
.
.
.
.
Dimension x used at the face of the loaded area .
.
Slab Merging .
.
.
.
.
.
.
Specification of Effective Slab Reinforcement
.
.
Providing the Shear Reinforcement .
.
.
.
Punching Perimeters .
.
.
.
.
.
Walls .
.
.
.
.
.
.
.
Overlapping Perimeters
.
.
.
.
.
Discontinuous Columns
.
.
.
.
.
Advantages
.
.
.
.
.
.
.
EC2 Design Code Requirements
.
.
.
.
The Design Procedure .
.
.
.
.
.
Check Maximum Shear Capacity .
.
.
.
Check Shear Capacity at the Basic Control Perimeter.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Calculate Shear Reinforcement Required and Length of Outer Control Perimeter .
Calculation of Magnification Factor .
.
.
.
.
.
.
.
Internal column
.
.
.
.
.
.
.
.
.
.
.
Edge column .
.
.
.
.
.
.
.
.
.
.
.
Corner column .
.
.
.
.
.
.
.
.
.
.
.
Calculation of W1
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Requirements for Models to be Linked for Foundation Design
.
Procedure for Linking Existing Models into a Foundation Project .
Linked Project Manager
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Checking a Typical Edge Column .
Providing Shear Reinforcement
Chapter 30
Chapter 31
.
.
.
.
Report Manager .
Quantity Reports .
Drawings .
.
Examples .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.503
.505
.507
.510
.511
.514
.514
.516
.517
.519
.520
.520
.521
.522
.523
.527
.527
.527
.528
.528
.529
.529
.529
.529
.529
.529
.530
.530
.530
.531
.531
.532
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.534
.534
.536
.536
.536
.538
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.539
.539
.539
.541
.541
.542
.543
Chapter 1 : Overview
Chapter 1
Overview
Introduction
The aim of this handbook is to provide added background information together with hints,
tips, and examples all of which should help you to make the most of Orion. This manual is not
written at a Getting Started level, and it is recommended that you have worked through basic
training examples in order to become familiar with the system and terminologies used, before
addressing the more complex detail provided here.
Copies of the getting started and basic training manuals in pdf format are installed with Orion
and can be accessed by clicking on the links below.
Orion Quick Start guide.pdf
Chapter 1 : Overview
2.
Derive beam loads using the Yield
Line (tributary area) approach
3.
Run the general building analysis to
generate column, wall and beam
design forces
2a - optional
Derive beam loads based on a special
FE model. Choose whether to use
these loads selectively or on all
beams
3a - optional
Use sequential FE floor analyses to
chase gravity loads down through the
structure. Selectively merge/override
column wall and beam design forces
with those of the general building
analysis
3b - optional
Use same FE models to generate and
merge alternative slab design forces
4.1
Beam design and
detailing
4.2
Column/Wall design
and detailing
4.3
Slab design and
detailing
Chapter 1 : Overview
Chapter 2
Modelling Techniques
Introduction
The Standard Training Manual covers many of the modelling techniques required for a
typical structure in sufficient detail that no further mention is required here. It can be accessed
by clicking on the following link: Orion Standard Training Manual.pdf
Note
If the link does not work please browse to find the file name indicated above in
the HELP sub-folder of the Orion Program Folder.
More advanced modelling techniques may also on occasion be required. The following cases
are discussed in this chapter:
1. Modelling Inclined and Lowered Members By default columns and walls are vertical,
beams and slabs are horizontal, however it is possible to specify otherwise to cater for
inclined and lowered members.
2. Modelling Curved Axes and Beams - It is possible to generate axes on a curve to facilitate
the definition of curved slab edges and beams.
3. Linking Angled Beams - If beams connect at an angle the program will attempt to
automatically determine if they should be linked on the detail sheet. However if the
default arrangement is unsatisfactory you can choose to revise it.
4. Columns and Walls Spanning More Than One Storey Each column and wall only spans
one storey in the building unless explicitly specified otherwise. If the column/wall spans
more than one storey, the number of storeys should be defined using Len (Storey) field in
the member properties dialog. This ensures the correct length is used in the slenderness
calculations when the column/wall is designed.
5. User Defined Supports These can be employed to model supports which occur above
the common foundation level. They can also be used to model (linear elastic) ground
springs.
Note
Currently user defined supports are only active in the Building Analysis model
and not in the FE Analysis model.
6. Stepped Foundation Levels Often buildings will be built on sloping sites, or they may
have to accommodate split basement levels. These situations are catered for within Orion
using either of two methods depending on the complexity of the modelling situation.
7. Beams with Varying Depth Generally, beams will have constant section properties
(width and depth) from the beginning of the member to the end, however there may be
occasions when you need to change the beam depth part way along the member.
8. Pinned Member Ends Columns and beams are by default fixed ended members,
however to alter the way forces are distributed the user can introduce pins at specific
locations within the model.
The features described in this section are for the purpose of defining occasional
sloping/lowered elements within a model which still contains distinct horizontal
floor planes. These features are NOT intended to facilitate the modelling of
structures with complex geometries in which the floor planes are not readily
apparent.
Note
Note that the slab design moments obtained (using either the moment coefficient strip
method or Finite Elements strip method) would not be any different for a dropped slab panel
in comparison to an identical panel which had not been dropped. This is because the level
difference would not be recognized in the analysis.
However, because the slab strips would be cut differently in a dropped slab panel the
reinforcement curtailment would be improved, as described later in this section.
How To Drop Parts of a Slab Panel
The example case below illustrates a partial drop.
In order to drop a part of a slab panel, you need to insert dummy axes and dummy beams
surrounding the drop panel.
The original slab layout is shown in the figure below:.
In this case study we will define a partial drop with 2000 x 2000 dimensions at the upper right
corner of slab panel 1S1.
Step 1: Inserting Dummy Axes to Define the Borders of the Drop Panel
Select axis 2 and press the Axis Offset button to offset it by 2000 mm to left direction. Set the
new axis label as 1a. Then, select axis B and offset it by 2000 mm to below direction. Set the
new axis label as A1.
Set the new axes as Not To Plot, since we don't want to include these axes in the output
drawings.
It is always good practice to shorten the dummy axes wherever they are not used to decrease
the number of axis intersections.
Insert the beams 1B1A and 1B1B as shown in the figure below.
If multiple members are to be edited it can be more efficient to first define a Plane to align
them to (see Working With Planes on page 26).
Columns can be inclined by specifying, (from the Column Properties, (General tab), different
axis intersections for each end of the column.
Columns can be associated with Planes (see Working With Planes on page 26) in order to
raise or lower the column end relative to the storey level. Planes are not used to create inclined
columns.
Walls can be inclined by specifying, (from the Properties - General tab) different axis
intersections for the top and bottom of the Wall as shown below.
Walls can be associated with Planes in order to raise or lower the wall ends relative to the
storey level. Planes are not used to create inclined walls.
If the default node points identified are not suitable, one, or all can be reselected using the
appropriate Pick Point icon on the Plane Properties dialog. Once the required node points are
displayed, the Z elevation of each can be updated in order to change the level, or inclination of
the plane.
Having defined the plane and the members which are to be part of the plane, the next step is to
Move Members to the Plane Definition. This command can be accessed from the right click
As shown below, all members contained within, or at the edge of the plane are adjusted to the
plane.
If you look closely at a curved beam you will see that it is actually formed from a series of
straight segments. You can specify how many segments to use when defining the beam.
Curved Axes
Curved axes are required for the definition of curved beams and curved slab edges. They are
formed from a number of linked straight axis segments which approximate to the curve
required. Provided the curve has a constant radius it should be created using the 'Curved Axis
Generator'. If you require a curve which doesn't have a constant radius you are restricted to
placing and then linking each axis segment manually.
The Curved Axis Generator
Straight axes are created by simply clicking and dragging between two points. To create a
curved axis, (or to generate multiple, or offset axes) you follow the same procedure, apart from
you must press and hold down the Shift key while dragging between the points. When you let
go of the mouse a dialog appears as shown below allowing you to define the degree of
curvature and apply offsets, or repeat spacings.
Offset Options
These options can be applied to both straight and curved axes. Instead of the axis passing
through the points clicked, it is drawn offset by the amount specified.
Insertion/Generation Options
These options can be applied to both straight and curved axes. Multiple axes can be created at
equal or varying spacings as required.
Curved Beams
Curved beams are composed of a number of linked straight beam segments approximating to
the curve required. They are defined using the 'Curved Beam Generator'.
The Curved Beam Generator
A straight beam is created by simply clicking and dragging between two points. A curved
beam can be created in the same way, apart from you must press and hold down the Shift key
while dragging between the points. When you let go of the mouse a dialog appears as shown
below allowing you to define the degree of curvature.
Offset Options
These options can be applied to both straight and curved beams. Instead of the beam passing
through the points clicked, it is drawn offset by the amount specified.
Insertion/Generation Options
These options can be applied to both straight and curved beams. Multiple beams can be
created at equal or varying spacings as required.
How many segments to use?
There is no definitive answer to this question for all cases - it will depend on the length of the
beam and the amount of curvature introduced. The default of six segments will often prove
sufficient, (we certainly wouldn't suggest using any less than six), but if you are in doubt you
can check for yourself by examining the effect on the analysis result of introducing more
segments.
Note
Six Segments
Twelve Segments
Model
Deflection
(mm)
Hogging
Moment (kN)
Sagging
Moment (kN)
6 Segments
68
47.3
-14.6
12 Segments
45
47.0
-13.6
24 Segments
40
46.9
-13.4
Although from the above it can be seen that the deflections haven't converged on a stable
answer, the hogging and sagging moments remain fairly constant. If the six segment model
were adopted the beam would be designed for slightly higher moments than if a more refined
model were adopted.
Note
If you change data for one segment then this change is saved when you move to
another segment. If you press cancel you are only cancelling edits made to the
current segment, edits applied to previous segments are not cancelled.
Note
The analysis is completely unaffected by the way the beams are linked together only the beam details are affected.
Once the axes have been linked only the first axis label is displayed. The linked axes can now
be selected/unselected as a single entity.
If at any time you require to return the linked axes to their original unlinked state, this can be
achieved by choosing Separate Linked Axes from the Right Click menu.
After either linking or unlinking, although the analysis results are unaffected, a re-analysis is
still required. This is because the data has to be stored in a different way in preparation for
beam design and detailings. After re-analysis the affected beams should be re-designed.
Having linked axis A1 to A2 the beams are now linked as follows:
Note
It is not always possible to link intersecting axes - for example a Dir 1 axis can
not be linked with a Dir 2 axis and vice-versa. In such cases it may be necessary to
detail the beams individually.
3. Enter the number of storeys that the column spans in the Len (Storey) box. In this example
enter 2 as shown below.
4. Press the Update button to apply the modification to the selected column.
5. Select the floor below, and delete the column that is already covered by the column in the
upper storey. In this example select the 4 th storey and delete the corner column as shown
below. Note that, if the column spans more than two storeys, this step must be repeated for
all the lower floors covered by that column.
P
Whereas, the columns and walls that stop at ST01, (again shaded in grey below) have by
default no external support. The loads within these columns will therefore be transferred
directly into the members in the lower storey.
Footnotes
1. releasing the end of the member is a simpler technique for achieving a pinned connection to the support.
Mechanisms
A mechanism may be introduced if, for example, you have applied a pinned support to a pin
ended member.
Diaphragm restraint
Typically, a rigid diaphragm exists within the floor slab. Hence, if a slab connects to the base of
a column which has a user defined support applied, the support will be directly restraining the
rigid diaphragm itself. This could inadvertently prevent lateral displacements from developing
at that level even if this was not the original intention.
Load Paths
User defined supports are assumed to transfer any reaction directly to the foundation. You
should not apply a user defined support at an upper storey level unless there is means for this
transfer to occur.
A stepped foundation is an example of where a user defined support would be appropriate,
whereas, a transfer column situation (i.e. where the column is supported by another member)
is an example of where it is not.
Depending on the model complexity, and the type of analysis carried out, one of two
modelling methods may be appropriate for achieving this, we shall referto these as:
Default Supports Method
The columns to the right of the model must initally be defined with Len (Storey) = 2 so as to
enable them to connect to the foundation. (If this was not done they would be treated as
unsupported.) Their base levels are then raised to the appropriate level by applying a dZ-Bot
figure as described below.
3. Repeat the above process for the walls changing both dZ-Bot, I and dZ-Bot,
This approach can only be used for models solved by Building Analysis. FE
Chasedown Analysis does not currently recognise user defined supports.
The exact level of the raised foundation is specified by applying a dZ-Bot figure to the columns
which is measured from the current storey level (in the same way as described in the Default
Supports Method).
Because these columns do not connect to St00, they are initially unsupported. Therefore user
defined supports will need to be manually defined and applied. For details of how to specify
these supports refer to the section- User Defined Supports
Note
If the raised foundation level coincides exactly with an existing storey level, it
may be necessary to offset one from the other to avoid a user defined support
being applied within a floor diaphragm. This can result in an illegal constraint at
the joint in question.
Note
User defined supports can be either pinned, fixed, or spring bases as required.
Non-frame axis to
create an intermediate
intersection
st
1 beam with
section 250x500
2nd beam with
section 250x750
1B25
1B25A
250x750
250x750
Note that this procedure is known as Splitting Beams and while it is possible, it will have
implications affecting the resulting beam detail. You will need to review and possibly edit the
automatically produced detail to ensure it is acceptable.
2. Stretch the end-points of the new axis 1x, so that it crosses only the grid line that the beam
will be defined (say A).
Non-frame axis to
create an intermediate
intersection
1x
A
1500 mm
1
1x
3. Remove the check in the View Codes option of this dummy grid line, so that it is defined to
be a ghost axis and check the Not to Plot option so that this axis will not be visible in any
printed output.
4. Instead of creating a single beam, insert two beams, 1B25 from axis 1 to axis 1x and 1B25A
from axis 1x to axis 2.
Non-frame axis to
create an intermediate
intersection
1x
st
1 beam with
section 250x500
2nd beam with
section 250x750
1B25
1B25A
250x750
250x750
1x
Using this button menu, you can either apply the selected end condition to selected columns,
or all columns in the current storey, or all columns in the building.
Using this button menu, you can either apply the selected end condition to selected beams, or
all beams in the current storey, or all beams in the building.
Chapter 3
General
Orion calculates the beam loads automatically using the slab area loads, beam wall load
information and unit weight of the members. Calculated loads can be viewed and modified (if
necessary) using the Edit Member Loads function on the right mouse click menu. The
following distributed loads are calculated automatically:
Beam self weight as a uniformly distributed load (using the unit weight of concrete defined
in the Project Parameters menu and the geometry of the beam),
Dead and imposed (live) loads transferred from the adjacent slabs (as distributed function
loads calculated based on the self weight and area loads of the slabs),
User defined additional dead and imposed (live) loads (any additional loads that cannot be
determined automatically can be defined by the user),
Load Table
Each load in the Load Editor is represented as a row in the Load Table. A hand icon will appear
in the right-most column to mark any manually defined user loads (none shown above).
Typically, Uniformly Distributed Loads and Partial Distributed Loads may exist in this table.
The selected load will be highlighted in the accompanying drawing to the right of the Load
Table. You can only edit or delete manually defined user loads.
Edit Button
You can edit an existing manually defined user load in the Load Editor by clicking on this
button. Note that automatic loads will be loaded as Read-Only to this editor. You cannot make
any modifications to automatic loads.
Delete Button
You can delete an existing manually defined user load in the Load Editor by clicking on this
button. Note that automatic loads cannot be deleted.
OK Button
If you close the Beam Loads dialog by clicking the OK button after making any modifications,
all the changes will be saved.
Cancel Button
If you close the Beam Loads dialog by clicking the Cancel button after making any
modifications, the changes made on the beam will be discarded.
9. Press the OK button to insert the partial distributed load. Alternatively you can abort the
process using the Cancel button.
Note
The vertex table is filled with the details of the partial load you have defined.
2. Press the Add Below button to append a new vertex to the end of the table.
3. Modify the x value to 5.0.
4. Press the Add Below button again to append another vertex to the end of the table.
5. Modify the x value to 7.5, the G value to 6.0 and the Q value to 3.5.
6. Press the Add Below button again to append another vertex to the end of the table.
7. Modify the x value to 5.0, the G value to 6.0 and the Q value to 3.5.
8. Press the Add Below button again to append a closing vertex to the end of the table.
9. Modify the x value to 7.5, the G value to 0 and the Q value to 0.
Load Profile Editor with a 2.5 m partial UDL entered to the end
10. Close the dialog by pressing the OK button to insert the partial distributed load.
Loads applied via the Load Editor are always applied vertically, even if the beam is inclined.
For UDL and VDL loads the magnitude of the load, P (in kN/m) should be entered as a
sloped load value.
The Load Editor requires all loads (including point loads) to be input using a horizontal
projected distance rather than a distance measured on the slope.
For example, a 5m long beam at a slope of 30 degrees would have a horizontal projected length
of 4m. A point load at mid span would therefore be defined with a Reference X, of 2m as shown
below.
Sloped load input using horizontal projected distance along the member
The load is analysed correctly at the mid point of the sloped length - this can be confirmed by
displaying the loading diagram in the beam analysis results.
The boundaries of the tributary areas loading each beam can be viewed graphically as shown
above. These look similar to potential Yield Lines, hence the name given to this default
method.
In this, as in many other areas, Orion effectively applies the sort of engineering methodology
that has been used in hand calculations for many years.
In such circumstances beam loads are more accurately calculated by using a Finite Elements
Model.
Consider the simple 6 m by 4 m slab shown above. Eccentric concentrated point loads are
applied and a large void also has been defined.
The Traditional (Yield Line) method of load transfer would ignore the hole, and it would
average out the point load by calculating an equivalent UDL to apply over the whole area of
the slab. Although the load is not lost, its effect is independent of its proximity to any of the
surrounding beams. This is all reasonably apparent when you examine the beam loads.
The loading on beam 1B1 is a simple trapezoidal profile where the peak UDL is adjusted to
account for the point loads. Based on this method, beam 1B2 sees an identical loading profile.
When you run a building analysis, a completely symmetrical set of column loads is
determined as shown below.
2. Click on the button Determine Loads Transferred from Slabs, this will open up an FE
modelling window.
It is important to note the advice on mesh density that is given in the above dialog.
3. A plate size is suggested by default. Click the Generate Mesh icon and the slab is meshed
automatically.
In this case the suggested plate size results in 234 plates being generated, and although not
well meshed around the opening this is probably sufficient in this simple example for the
intended purpose.
4. Close this FE window and then also the Finite Element Analysis Form. This will result in the
following dialog being displayed.
5. The FE results can be applied to selected beams, all beams in the current storey, or to the
whole model. Choose Apply to All Beams in the Model.
You are then returned to the Graphic Editor.
Having completed the above process, FE information exists and has been applied to all beams
in the model. You can review this on a beam-by-beam basis. Select any beam and open the
beam loads dialog, it will have swapped to the Finite Elements Slab Load Decomposition
Method. The new loading for beam 1B1 is shown below.
As expected the load intensity is peaking towards the ends where the point loads are applied.
To utilise these new loadings it is necessary to reperform the building analysis. On completion
we can see a very different distribution of column loads.
The total IL using the Yield Line method was approximately 267.6 kN. Using the FE method it
has dropped to 265.5 kN, this is slightly lower because no load is applied to the hole.
Note
If you change any slab loads or beam layouts, you must remember to reperform
the Load Decomposition by FE command before reperforming the building
analysis.
Note
Axial load comparisons are performed at the end of each analysis. This is a
comparison of applied loads vs. analysis reactions. Since you can choose
decomposition methods on a beam-by-beam basis you can cause this comparison
to show discrepancies. In the example above, if you choose finite element
decomposition for beam 1B1 only, the comparison will be displayed as shown
below.
The first applied loads table shows the totals for the loads applied to each slab, beam, etc. The
second applied loads table shows the loading totals after decomposition. Since the
decomposition methods have been mixed so that a worst case is used for every beam, the total
loading after decomposition is higher. You can then see that this higher loading is maintained
in the building analysis.
Chapter 4
Analysis Methods
Introduction
Having constructed and loaded your model, an analysis will be required before member
design1 can commence. The analysis method(s) used will depend on a number of factors; for
certain structures (eg flat slabs) two or more analyses may be required.
A brief summary of the analysis methods available within Orion is given below, these are then
discussed in more detail in the next four chapters:
A full description of this method is provided in the chapter General Building Analysis.
Eigenvalue Analysis
An Eigenvalue Analysis can (optionally) be performed as part of General Building Analysis to
determine natural frequencies and mode shapes.
A full description of this method is provided in the chapter Eigenvalue Analysis.
A full description of this method is provided in the chapter Staged Construction Analysis.
Footnotes
1. the exception being slabs, which can be designed directly using tabulated code coefficients, where applicable.
A full description of this method is provided in the chapter Analysis and Design using FE.
Chapter 5
Introduction
This chapter is split into two sections:
The Structural Model section below describes various modelling considerations and
analysis options that have a significant impact on the building analysis result.
Structural Model
A full 3D analysis model is derived from the physical information that you describe when
constructing the model. This may sound like a relatively simple and comprehensive solution
to the analytical modelling of any structure but in fact there are numerous difficulties, options,
and even personal preferences to consider here. The following list introduces most of these
items which are then discussed in more detail in the remainder of this section:
1. Diaphragm Modelling The analysis model is a 3D frame (stick) model, slabs are only
modelled in the sense that their diaphragm action is accounted for. What are the options
relating to how/when diaphragms are included? How are inclined planes handled?
2. Global Constraints What are they and when might you use them?
3. Pattern Loading How is pattern loading catered for?
4. Rigid Zones What are they and why would you use them?
5. Rigid Links What are they?
6. Shear Walls and Core Wall Systems How are walls modelled?
7. 3D Effects A full 3D analysis will sometime give answers that you do not expect, is it
wrong?
8. Stiffness adjustments What options are available and when might you use them?
9. Flat Slab Construction Dealing with structures that incorporate some degree of flat slab
(flat plate) construction.
10. Transfer Levels Dealing with beams and slabs that support discontinuous columns or
walls.
11. Supports Default supports and how to create new support types.
Diaphragm Modelling
In a typical building lateral resistance is provide at a few discrete points and it is assumed that
applied lateral loads will be distributed to the lateral load resisting systems via floor diaphragm
action. Within Orion diaphragm modelling is achieved using diaphragm constraints.
A diaphragm constraint will maintain exact relative positioning of all nodes that it constrains,
i.e. the distance between any two nodes constrained by a diaphragm will never change,
therefore no axial load will develop in any member that lies in the plane of a diaphragm
between any two constrained nodes.
When running the building analysis the model options tab gives options as shown below.
The differences between these options can be demonstrated with the simple model below.
A single floor level has 2 separated slab areas that are linked by 2 beams.
Within the main graphics window Orion will always indicate a single overall centre of mass for
the entire floor level as shown below.
It is important to note that in the case of discrete diaphragms, or when there is no diaphragm
at all, this overall central location is provided for information only.
Slabs to define rigid diaphragms (Default Setting)
If the building is analysed with this setting then Orion will find any discrete areas of
interconnecting slabs and set up discrete diaphragms as appropriate. Separate notional loads
are calculated and applied to each diaphragm area. These applied notional loads and the
resulting sways can all be examined in the model and analysis results display. Perhaps the
easiest way to see that 2 discrete diaphragms have been created is to look at an exaggerated
view of deflections.
For each of the two diaphragm areas a separate centre of mass location is determined and the
notional load is applied at that position. The view above shows the values for the Fy case in
this example. Note that the mass of the walls increases the applied notional load in the left
hand diaphragm area. Clearly the right hand area is moving independently to the left hand
area which is better restrained by the walls as opposed to frame action.
Single rigid diaphragm at each floor level
If the building is analysed with this setting then Orion will find apply a single diaphragm
constraint to every node at any given level. The existence of slabs is completely ignored/
irrelevant. A single notional load is calculated and applied at the overall centre of mass as
shown below. Once again the easiest way to see the effect of this setting is to look at an
exaggerated view of deflections.
This time we can see the entire level translating and rotating as a unit, since the centre of mass
and hence the applied notional load is very eccentric to the core walls, the dominant effect in
this example is one of rotation.
The frame that is restrained by the wall hardly moves at all, other frames move to differing
degrees. Note that the frames which do not include walls have the same stiffness and so the
differing deflections relate to differing notional loads.
Excluding Specific Slabs from Diaphragms
In the example above there was no slab defined in the area linking region between grids 4 and
5. What if there was a slab in this area as shown below?
If the linking slab was substantial you might consider that it maintains the diaphragm action
between the two areas. However, as the link becomes more slender then at some point you
will decide it cannot maintain diaphragm action between the two areas. In this case you can
edit the properties of the slab and exclude it from diaphragm as shown above. In this case you
can still use the default option Slabs to Define Rigid Diaphragms and the resulting deflections
will be the same (ignoring small change due to additional notional load from the added slab)
as were shown above for that option.
If either the Slabs to Define Rigid Diaphragm or Single Rigid Diaphragm at Each Floor Level
option is used a warning message is displayed during the analysis indicating that the nodes
constrained by the diaphragm do not all lie in the same plane.
A single diaphragm is created within the pitched roof as shown below.
On first viewing the above may not appear unreasonable, however, because the rigid arms that
constrain all the nodes to move together are not co-planar there are incorrect side effects as
demonstrated by the bending moment diagram below.
The diaphragm is holding up the mid-span node on the raking beams causing an unexpected
hogging moment at that point. In addition, axial loads develop in the inclined beams (below)
which would not happen if they existed in a properly inclined diaphragm.
Considering a sway case (below), it seems that the diaphragm is constraining displacements in
the horizontal plane only.
The degree to which the results are affected is related to the pitch of the inclined beams. As the
pitch reduces the effect drops, however it should be noted that it is still significant even at
relatively low pitches.
A workaround in this example would be to re-run the analysis using the option No Rigid
Diaphragm Floor Levels. The resulting bending moment diagram is then as shown below.
Alternatively, exclude one of the slabs meeting at the apex (as described in the second case
study below) so that there are two differently inclined but separate diaphragms. (You should
find results are very similar to those obtained using the first workaround.)
Case Study 2 - storeys linked by inclined planes
In this example inclined planes have been used to define ramps between the storey levels.
If either the Slabs to Define Rigid Diaphragm or Single Rigid Diaphragm at Each Floor Level
option is used a warning message is displayed during the analysis indicating that the nodes
constrained by a diaphragm do not all lie in the same plane.
Because the storeys are linked, a single diaphragm has been created constraining all the floor
nodes as shown below.
When a lateral load is applied, each floor moves by the same amount - obviously the model is
not behaving as intended.
Diaphragm action is required to help resist the lateral loads, so specifying No Rigid Diaphragm
Floor Levels is not an acceptable solution. The answer is to restrict diaphragm action to the
horizontal floor planes only. This is achieved by excluding the inclined slabs from the
diaphragm as shown below.
Because the storeys are no longer linked, separate diaphragms are formed at each floor level as
shown below.
By viewing the deflections under lateral load it can be seen that each floor is now moving
independently of the other floors as intended.
Global Constraints
You can use the options in the building analysis dialog shown above to apply general
constraints to the floor translation/rotation:
X/Y and Torsion Permitted,
Pattern Loading
Designs codes generally require that pattern loading is considered in the analysis/design of
continuous beams. Orion has a method of automatically patterning both dead and imposed
loads along continuous beam lines. This subject is discussed in some detail in the Standard
Training Manual.
Rigid Zones
Rigid zones deal with two aspects of the analysis and design of concrete frames:
1. Design codes generally allow you to design for forces at the face of a support.
2. From an analytical modelling point of view it is commonly accepted that a simple
centre-line model does not properly idealise the physical size of members. By modelling
rigid zones we account for the extra stiffness that exist within the 3D block where the
members physically interact with one another.
The chapter dealing with Wall Modelling Considerations shows examples that take this
second point to a more extreme level. Beams attach to the ends of walls (not the centre lines)
and we are happy to models walls and beams in this way. Why would the same not be true
where beams attach to columns?
To demonstrate the effect of rigid zone modelling the following simple example will be used.
A simple symmetrical structure is shown in the rendered view above. When running the
building analysis you have a choice of 3 model options relating to rigid zones, none, reduced
by 25%, and maximum as shown below.
The sections below describe what each of these options will do.
The analysis results are as shown above. The beam bending moment diagrams extend right to
the centre of the column, the beams will be designed for a support moment of 77.91 kNm and
a span moment of 86.56 kNm.
This is the most simplistic analysis model and is likely to be the most similar to any models
you may have created in general analysis packages.
The analysis results are as shown above. Note that the beam bending moment diagrams do
not extend right to the centre of the column, they stop at the face of the column. The column
moment diagram also stop at the underside of the beams. No diagrams are shown on the rigid
zone lengths.
Importantly, for this model type, although the bending moment diagram starts at the face of
the column the actual rigid length within the rigid zone only extends to 75% of this length (i.e.
it is reduced by 25%).
Comparing the moment diagrams to the case where no rigid zones are used we see that:
1. The span moment is reduced from 86.56 kNm to 76.44 kNm. Adding rigid zones at the
ends of beams generally lifts the bending moment diagrams and hence will slightly reduce
span moments.
2. Although the diagram has been lifted the support moment of 77.91 kNm has reduced
slightly to 75.12 kNm because the moment is now being taken at the face rather than the
centre-line of the column.
3. The column design moments are also reduced from 77.91 to 69.59 kNm.
The analysis results are as shown above. Once again the beam bending moment diagrams do
not extend right to the centre of the column, they stop at the face of the column. The column
moment diagram also stop at the underside of the beams. No diagrams are shown on the rigid
zone lengths.
Importantly, for this model type, the bending moment diagram starts at the face of the column
AND the actual rigid length within the rigid zone extends to 100% of this length (i.e. it is
max).
Comparing the moment diagrams to the cases where no rigid zones and reduced rigid zones
are used we see that:
1. The span moment is further reduced from 86.56 kNm to 76.44 to 73.2 kNm. Adding full
rigid zones at the ends of beams increases the lifting of the bending moment diagrams.
2. In this case the diagram has been lifted so much that the support moment of 77.91 kNm
which reduced slightly to 75.12 kNm using reduced rigid zones has now increased to 78.36
kNm.
3. The column design moments which reduced from 77.91 to 69.59 kNm have also increased
again slightly to 72.28 kNm.
Discussion
It should be noted that the simplistic model used above to demonstrate rigid zones does not
deal with the complexities of continuous beams and pattern loading and may to some degree
exaggerate the effects of rigid zone modelling. A more typical four span continuous beam
example is considered below.
Below results when maximum rigid zones are used, span moments reduced another 23%
and support moments increase again by 23% but are still less than the moments when no
rigid zones are used.
In general the use of rigid zones reduced by 25% will result in the maximum reduction in
support moments combined with a less extreme reduction in the span moment. It is
considered that this option will give maximum efficiency and will be the preferred option for
most designers using Orion.
NOTE: Although this option can often result in reduced support and span design moments, it
should not be confused with moment redistribution. In theory there is nothing to these
design forces being further reduced by the use of moment redistribution.
Rigid Links
3D analysis models are still centre-line models, members need to line up to connect to each
other. When you construct a 3D model in a general analysis package you will naturally see
where things do not line up and make adjustments so that connections are made. When
working in Orion you tend to feel that you are working with the physical model rather than the
analytical model that is derived from it. Sometimes members will not line up and in such
cases Orion will attempt to make the required connections by adding rigid links.
The beams on grids 1 and 1a do not line up but are both supported by the same column. In
this example this column is actually inserted at the intersection of 1a and A1. If we look at a
3D stick view we can see the frame that has been defined as shown below.
The column appears to support one of the beams but not the other. However, when the
analysis is carried out Orion will automatically connect the end of any beam that lies within the
physical plan perimeter of any column or wall to that column or wall. So when this model is
analysed a rigid link is added as shown below.
In this example the column was defined in an offset manner directly in line with one of the
beams. To be more precise it could have been defined on a new grid on its own centre line and
then rigid links would have been created from both supporting beams to the column.
Note that rigid links should not be confused with rigid zones, rigid zones are not used in the
above model but could have been.
Note that rigid links will also be created where smaller offset columns sit on larger columns or
walls in the floor below.
3D Effects
In general our traditional engineering expectations are developed from considering simplified
sub-models for analysis and design. When a full 3D model is created unanticipated effects
sometimes creep in, is this wrong?
Some simple examples will help demonstrate these effects.
Continuous Beams
Traditionally continuous beam lines are analysed and designed in isolation. The modelling of
the support conditions in such cases is often unsophisticated. In the example below a series of
3 span secondary beams are supported at different points on primary beams and also directly
by columns.
When analysed in 3D using unadjusted gross member properties the bending moments for a
general UDL case are as shown below.
Note how the end span moments for the internal beams (which all carry the same load) vary
from 38.94 to 49.54 kNm. The smallest value occurs on the central column line and largest
occurs where the beam is supported at the centre of the primary beam. The variation is not
too significant but does start to indicate the importance of all the relative stiffnesses in a 3D
model.
However, note that hogging moments develop at the extreme ends of the secondary beams
where they are supported by a primary beam. Investigation would show that these develop
because torsional forces are developing in the primary beams. A traditional 2D continuous
beam line analysis will generally model this support as a pin and no hogging moment will
develop. In fact torsion is only usually considered in traditional hand calculations where its
development is essential to the local stability of the structure.
In generating the above results the default unadjusted properties of the rectangular beam
sections were used in the analysis. These results are exactly what you would get from any
general 3D analysis package. In Orion you have options to apply adjustment factors to the
default properties of groups of elements as shown below.
These adjustments are discussed further in the next section, but it is worth noting here that by
default Orion suggests adjusting the torsional stiffness factor of beams by a factor of 0.01 (i.e.
reducing to 1%). This is suggested because in most cases this means that the results of a 3D
analysis will be more compatible with the tried and tested analysis/design achieved using older
2D idealisations.
When the model above is reanalysed using this setting the results are as shown below.
The hogging moments have disappeared and the end span moments have increased and
become more consistent at around 67 kNm.
Effects of one Member on Another
The most likely 3D effect is one where the design forces generated are not what you expect.
This sort of effect could happen in many different ways and becomes more likely as the
structural arrangement becomes more complicated and irregular. However, it can be
demonstrated with the very simple model shown below.
A beam is selected along grid A at the front of the model. The loading, shear and moment
diagrams for this beam are shown below.
The loads are slightly offset due to the triangular area of slab being supported, but the bending
moment diagram shows an (unexpected?) sagging moment at the left end support and an
(expected?) hogging moment at the right end support. Can this be correct?
The explanation relates to the diagonal beam across the floor. It has been loaded with an extra
UDL. When the structure is examined in the building analysis post processor we can see how
it deflects and bends.
The heavily loaded diagonal beam is putting a big moment into the supporting columns and
the joint is clearly rotating. This rotation cannot happen without some effect on the other
connected beams. When viewed in this way the design moments in the beam do not seem
unreasonable.
Innumerable examples of this nature could be developed. The point is that if the design forces
in a member seem wrong you cannot assume that the analysis is wrong. You need to review
the results carefully and in particular look at the nature of deflections.
Sway Effects
Many structures will undergo a natural sway under purely vertical (gravity) loads. These
sways can sometimes introduce significant changes to the expected moment diagrams in
beams in much the same way as is shown in the example above.
In such cases it is important to ensure that checks are made for combinations where notional
load cases are applied in sympathy with the natural sway of the structure.
In such situations where you might like to check for differences exposed by a 3D analysis of a
floor in isolation, the chapter Analysis and Design using FE introduces this option.
In more extreme cases buildings stabilised by shear and core walls will sway significantly
under purely vertical load, this topic is discussed in more detail in the chapter Wall Modelling
Considerations.
Transfer Beams
The way in which load accumulates in columns that are in turn supported on transfer beams
in a full 3D analysis is often at odds with initial/traditional expectations. This subject is
specifically dealt with in the series of chapters dealing with transfer levels and in particular the
section Discussion of Frame Analysis Results in the chapter Transfer Beams General
Method.
Stiffness adjustments
When you construct a model you set/control the analysis model properties by:
Using the Building Parameters options to globally set/change material properties.
Some reasons why you might use these options could include:
However, if you have a model that (unusually) actually relies on torsion for stability then using
the above setting will often mean that extremely large deflections accrue.
Investigating Deflections
You could use different values of Youngs Modulus to investigate different loading conditions
for example use the default short-term value when considering wind load conditions, then
reanalyse with multipliers to reduce Youngs Modulus to allow for long-term effects in other
cases and general design.
Transfer Levels
A transfer level is an upper level within the structure at which one or more columns or walls
stop. The discontinuous column or wall is supported by beams or slabs such that its load is
transferred to other columns/walls.
This topic is discussed at length in a series of chapters and worked examples beginning with
the chapter Overview of Solution Options for Transfer Levels.
Supports
In a simple model such as the one shown below, all columns and walls finish at a common
foundation level.
In such models, by default a fixed support is automatically placed underneath each column
and each wall at the top of the common foundation level, (indicated by the grey plane).
If a fixed support is inappropriate, a pinned or spring support can be defined using Support
Type Definitions from the Members menu.
Having defined a new support type, it can then be assigned to a specific column or wall using
the member properties as shown:
There is an option to run validation checks at any stage, if your building analysis
is giving errors or warnings and you have not run the validation checks, it is
recommended that you do so. This may expose the cause of problems.
Tip
Many building validation errors can be avoided by ensuring that the option to
check the validity automatically each time a member is inserted is active. This is
done by clicking on the Graphic Editor View Settings and on the Plan settings
tab, and ensuring that you leave the option Dont check Model During Member
Insertion unchecked.
Note
Note
Beyond the above, it is not anticipated that you need to read this section until
such time as you are trying to understand or resolve errors or warnings.
If any of these conditions occur an error message will be displayed which will clearly indicate
the members that are invalid. You should return to the graphic editor to correct the error and
then re-perform the building model validity check to confirm the problem has been resolved.
On rare occasions a particular error message is displayed:
Error: beam 1B157 is overlapping with beam 1B157
At first this does not appear to make sense since it suggests that a beam is overlapping with
itself. On investigation you will find that there are two beams with identical labels at exactly
the same location. If you encounter this problem you should simply delete one of the beams
while checking that you have not lost any manually applied loads. Note that you will not
immediately see the other beam after deleting the first one, but executing the redraw or regen
function will make it reappear.
Note: If there are no errors as a result of Building Model Validity Checking, it does not follow
that your model is 100% error free. There are many other conditions that are not checked at
this stage, which could result in your model being incorrect. These conditions are checked
during later stages of the building analysis process.
If you see messages about singularities as shown above than it usually means that too many pin
ended members have been defined. It is recommended that you review the structural model
in the Building Analysis postprocessor and eliminate the unnecessary releases.
If there are so many releases that the structure is unstable then this will usually result in the
message shown below.
This message should never be ignored unless you are dealing with a transfer slab structure
see below.
The following warning messages are commonly displayed during Graphical Data Preparation:
The above message informs you that columns (and walls) exist at an upper storey but do not
continue to foundation level and that the means of support at the transfer level are uncertain.
(i.e. a column sitting on a beam does not generate such a warning.)
Action Required
If it is intended that the columns are transfer columns and you are clear as to the means of
support (e.g. a transfer slab) then you can ignore the warning provided that you have taken
account of the related requirements outlined in the chapter Overview of Solution Options
for Transfer Levels later in this handbook.
If it is not intended to be a transfer column return to the graphic editor and remodel so that
you have a support member in the desired location.
A simple model can be created for a simple illustration of the Axial Load Comparison Report.
The 3D view of the model is as above. The plan view of the model is as below.
Beam size:
Column size:
500 mm x 250 mm
Column Height:
2800 mm
Slab Thickness:
Loadings on slab:
The following report is produced by Orion for this model. Note that there are 4 main sections
(tables) of information the headings of which have been emboldened.
Column
ShearWall
Beam
Slab
Rib
Total
(kN)
(kN)
(kN)
(kN)
(kN)
(kN)
St01
-------
33.6
0.0
65.0
117.3
0.0
215.9
Total
215.9
Q - Live Loads:
Storey
------St01
-------
Column
ShearWall
Beam
Slab
Rib
Total
(kN)
(kN)
(kN)
(kN)
(kN)
(kN)
0.000
0.000
0.000
0.000
0.000
Total
0.000
Column
ShearWall
Beam
Slab
Rib
Total
(kN)
(kN)
(kN)
(kN)
(kN)
(kN)
St01
-------
33.6
0.0
182.3
0.0
0.0
215.9
Total
215.9
Q - Live Loads:
Storey
------St01
------Total
Column
ShearWall
Beam
Slab
Rib
Total
(kN)
(kN)
(kN)
(kN)
(kN)
(kN)
0.000
0.000
0.000
0.000
0.000
Delta G
(kN)
-------
(kN)
Delta Q
(kN)
(kN)
St01
-------
215.9
215.9
0.0
0.0
Total
215.9
0.0
Delta G
(kN)
-------
Delta Q
(kN)
(kN)
St01
-------
(kN)
223.6
223.6
0.0
0.0
Total
223.6
0.0
The purposes and validity of the comparisons that can be made between the sections of this
table are discussed in the following sections.
Table 1
SUM OF APPLIED LOADS (Using Un-Decomposed Slab Loads):
-----------------------------------------------------Storey
------St01
-------
Column
ShearWall
Beam
Slab
Rib
Total
(kN)
(kN)
(kN)
(kN)
(kN)
(kN)
0.0
65.0
117.3
0.0
215.9
Total
215.9
Q - Live Loads:
Storey
------St01
-------
Column
ShearWall
Beam
Slab
Rib
Total
(kN)
(kN)
(kN)
(kN)
(kN)
(kN)
0.000
0.000
0.000
0.000
0.000
Total
0.000
Beams Loads
33.6 kN
Slab Loads
65 kN
117.3 kN
Table 2
Differences between this table and table 1 are specifically intended to expose problems in slab
load decomposition. Since there are two methods of slab load decomposition supported we
will look at each of these separately.
Column
ShearWall
Beam
Slab
Rib
Total
(kN)
(kN)
(kN)
(kN)
(kN)
(kN)
St01
-------
33.6
0.0
182.3
0.0
0.0
215.9
Total
215.9
Q - Live Loads:
Storey
------St01
-------
Column
ShearWall
Beam
Slab
Rib
Total
(kN)
(kN)
(kN)
(kN)
(kN)
(kN)
0.000
0.000
0.000
0.000
0.000
Total
0.000
The key difference in this table is that the slab loads are now decomposed and thus counted as
UDLs, VDLs etc. on the supporting beams. Therefore, you will find that the slab loads become
zero but the beam loads increase accordingly.
Decomposed Slab Loads Consider again the model and the decomposed slab loads shown
below.
The plan view shows the yield lines, strictly speaking these are really just load decomposition
lines which are used to show the area of slab loading that will be attributed to each beam. This
method of area load decomposition is commonly known as the Yield Line Method. Looking at
the triangular load distribution generated on the above beam, the beam load calculation
effectively becomes:
Decomposed Slab Load
4 x 29.33 kN
117.3 kN
The beam loading profile and the above calculation clearly get much more complex when
more irregular slab arrangements are used.
The calculations reported in table 2 are therefore as follows:
Columns Self Weight
Beams Loads
UNCHANGED
33.6 kN
65 kN + 117.3 kN
182.3 kN
Column
ShearWall
Beam
Slab
Rib
Total
(kN)
(kN)
(kN)
(kN)
(kN)
(kN)
0.0
182.0
0.0
0.0
215.6
Total
215.6
Q - Live Loads:
Storey
------St01
------Total
Column
ShearWall
Beam
Slab
Rib
Total
(kN)
(kN)
(kN)
(kN)
(kN)
(kN)
0.0
0.0
0.0
0.0
0.0
Both FE and Yield Line decomposition decompose slab loads to beams. If there
are areas without beams (where slabs are supported directly by columns i.e.
Flat Slabs) the decomposition process is guaranteed to lose load. In such cases
full FE analysis must be used and the discrepancy between the totals in tables 1
and 2 should be seen as an indicator of this requirement.
Note
In all but the simplest of models, there will always be a small discrepancy between
table 1 and 2. However, this can be ignored especially if the decomposed load in
table 2 is slightly higher than the un-decomposed load in table 1.
Table 3
BUILDING ANALYSIS COLUMN/SHEARWALL AXIAL LOADS:
----------------------------------------------Storey
G
(kN)
------St01
------Total
Delta G
(kN)
Q
(kN)
Delta Q
(kN)
215.9
0.0
0.0
0.0
The figures in this table reflect the results of the building frame analysis. The building analysis
is a frame analysis where the beams are loaded with all the decomposed slab loads. Therefore
the input is based on either the yield line or FE Load Decomposition method; whichever has
been selected.
When the analysis is complete the accumulated column loads on each storey are shown in the
table.
It is therefore appropriate to compare table 2 with table 3 which is in effect a comparison of
analysis input with analysis results. If the totals are different the building analysis is incorrect
in some way in which case you should check:
If there were warnings during the building analysis, have you ignored them and is it OK to
ignore them?
If transfer levels exist, check whether the discontinuous columns are properly supported.
Note
Flat slab models are a good example of the case where there will be discrepancies
and loads are lost, but this can all be ignored since Finite Element Analysis load
Chase Down is required see next section.
Note
This table is only available if an FE Chase Down (finite element analysis load
chase-down) has been performed:
Table 4
When this option is active as shown above, a meshed up FE model is created as shown below.
Slab loads are applied to all the individual shells and so the slab load decomposition is
effectively inherent in the FE analysis of the floor.
After analysis the total axial loads in columns and walls are derived at each level and reported
in table 4 as shown below.
FINITE ELEMENTS ANALYSIS COLUMN/SHEARWALL AXIAL LOADS:
-----------------------------------------------------Storey
Delta G
(kN)
-------
(kN)
Delta Q
(kN)
St01
-------
(kN)
223.6
223.6
0.0
0.0
Total
223.6
0.0
The total loads and self-weights reported will always be slightly different in FE analysis
because this is a true centre-line model. In this simple example the loads effectively total up as
follows:
Column Self Weight
33.6 kN
(this part is unchanged)
Beams Loads
60 kN
(i.e. self weight only finishes loads are not applied to the beam)
130.0 kN
(see note below)
Total
60 kN + 33.6 kN + 130.0 kN
223.6 kN
Note
The area of the slab used in the FE model is different (greater) than the area used
in tables 1 and 2 because the slab extends to the centre-line (not the edge) of the
beam. This has two effects:
The self weight if the slab in the zone that overlaps with the beam is
double counted small extra load.
The finishes and imposed loads that apply on the outer edge of edge
beams will not be counted small loss of load.
When this option is used it is appropriate to compare the totals in table 1 with table 4 (Load
before decomposition and FE analysis results which embody both decomposition and
analysis). Some degree of difference must be accepted for the following reasons:
In all models, there will be a slight difference for the reasons noted above.
If the option to include column and wall sections in the model has been used, are slab line
loads applied through columns or walls (the load inside the column or wall boundary is
included in table 1 but is not taken into account in the FE analysis because the load could
not physically exist.)
As above, but checking for point loads applied inside column boundary.
Incorrectly modelled cantilever slabs.
Note
The totalling of loads in table 4 will be incorrect if some columns are len(storey)
2 or more, but the results themselves are actually correct.
In this case the FE analysis is simply a 3D frame analysis where the decomposed slab loads are
applied to the beams.
In this case it is valid to compare table 2 with table 4 since this is checking that the
decomposed loads which were intended to be applied to the FE models have been successfully
applied.
Chapter 6
Eigenvalue Analysis
Introduction
An Eigenvalue Analysis can be performed as part of the Building Analysis in order to calculate
natural frequencies and mode shapes; these will be dependent on storey mass and model
stiffness. The Eigenvalue Analysis results can then be used for seismic design purposes and
can also be of value if wind tunnel tests are required.
Note
The G and Q components of the Storey Mass/Weight can be derived from either the
decomposed beam loads, or the undecomposed slab loads. The method used is controlled via
the Settings tab of the Building Analysis Model Options dialog as shown below.
The mass/weight determined for the chosen option can be reviewed by hovering the cursor
over the Center of Gravity of the form plan after running a Building Analysis.
If all slabs transfer their loads to beams or walls, either option should produce a similar mass/
weight. This can clearly be seen not to be the case in the above (flat slab) example - the mass
determined using the Decomposed Beam Loads option is significantly smaller than that from
the Use Undecomposed Slab Loads option. In such models it is important that the latter option
is always selected.
Model Stiffness
The engineer should use section properties that are appropriate for the Eigenvalue Analysis.
For columns and walls this could involve making global stiffness adjustments to model
cracked section properties. The ACI code may be referred to for some guidance in this regard.
These adjustments can be made via the Model Options/Stiffnesses tab of the Building Analysis
dialog.
Analysis
The analysis produces results for both the Static and Eigenvalue Analysis that can be accessed
simultaneously in the graphical post processor.
Graphical Results
To view a mode shape, activate the Displacements button and select the mode shape required
from the Loading menu as shown below. Animation can be activated if required from the
Displacements menu.
Numerical Results
Numerical output from the analysis is accessed from the Report tab Eigenvalue Results
Report Button.
This button shows information relating to the frequencies and mass participations as shown
below.
Exporting to S-Frame
If you export the model to S-Frame to cross check the Eigenvalue analysis, you will need to
make a minor adjustment before running the analysis. This is necessary because the way
S-Frame stores the mass information is slightly different to the way it is held in Orion.
When the model is opened and analysed in S-Frame the frequency is 6.97Hz The difference is
accounted for by the way the mass due to self weight is being modelled in S-Frame.
Correction of Self Weight in S-Frame (for Eigenvalue Analysis only)
Two materials are exported to S-Frame - The first material (MAT1) is assigned to the
columns/walls and has a mass Orion ignores this mass in the vibration calc because it has
included it in the floor mass, S-Frame does not ignore this mass, therefore the mass is double
counted in S-Frame.
The second material (MAT2) used for beams has zero mass.
In order to get compatible Eigenvalue analysis results between Orion & S-Frame you are
required to switch off the mass of MAT1 in S-Frame. This is done by setting the force density
to zero.
If the density of the materials is set to zero in S-Frame the frequency for the first mode
increases to 7.125Hz.
Note
Chapter 7
Introduction
Staged construction analysis is a sophisticated method of solution which, if used with a
reasonable degree of care can more closely reflect the real behaviour of the structure.
A standard 3D analysis takes no account of time dependant effects - it assumes the structure
has been constructed instantaneously. In the real world a concrete building has to be
constructed in stages - typically one floor at a time. Depending on the number of storeys
within the building a significant period of time can elapse from the commencement of first
floor to completion of the top floor. Because concrete matures with time, this means that at the
top floor the concrete has properties which lag behind those of the first floor (and to a
reducing degree each intermediate floor).
Another aspect of the staged nature of concrete construction is that any settlement of existing
lower floors which takes place prior to the pouring of the current floor is allowed for on site
when establishing the current floor level. This effect can not be allowed for in a standard 3D
analysis, however it is catered for automatically in a staged construction analysis.
Simple Example
A model consisting of two blocks is constructed as shown below. The first floor plan and
loading is identical in both blocks. A second floor framework is added to the right hand block
but with no imposed load applied to that level.
A standard 3D analysis can be performed for the above, in which all members are analysed
simultaneously with all the loads applied.
A staged construction analysis can also be performed. If the option were selected to stage the
imposed load only, the analysis would proceed as follows:
In the first stage, only the first floors are analysed for their imposed load.
In the second stage, the second floor is added and the analysis repeated (for the
undeformed model) but this time only for the imposed load at second floor level.
Note
In this way, each stage can only affect the result in its own and previous stages.
In the unstaged result you can see that Q load applied at first floor is causing moments to
develop in the structure above. If that frame existed when the load was applied then this is
completely logical.
In the staged result no moments develop in the frame above the first floor members and in fact
the forces in the first floor structure are identical in the single storey frame and the two storey
frame. This is logical if all the load is applied before the frame above exists AND assuming
that all deflection is instantaneous.
Note
The effect of creep is not considered in the above - in reality the first floor beams
would continue to deflect and the joints continue to rotate after the frame has
been built and loaded. Therefore, although there are no moments in the second
floor frame at the point at which it is constructed, the creep of the first floor
structure should cause some moments to develop over time in the second floor.
Using the controls you are able to stage any gravity load case and then create combinations of
staged and unstaged cases.
Stage Construction Cases Check this box to activate staged construction.
Stage Duration Specify the default duration of each stage, this can subsequently be
amended for individual stages - see Stage Control.
Stage Construction Cases: G Check this box to create an sG staged loadcase.
Stage Construction Cases: Q Check this box to create an sQ staged loadcase.
Create New Combinations for Staged G and Q When not checked one set of
combinations is created which uses the staged cases, when checked two sets of
combinations are created, one with the unstaged cases and the second with the staged
cases.
The result of checking the Create New Combinations for Staged G and Q box is shown below:
Stage Control
Stages are defined within the Load Combination Editor by firstly clicking the Load Cases
button. Next, highlight a staged loadcase, (for example SG as shown below) and click Edit.
You can if required specify different time durations for the individual stages. See Setting the
Duration of Each Stage
3. Stages 3 and 4 are not required and need to be deleted - this is achieved by clicking in the
topmost stage (4) and putting "0" in the storey column.
4. Repeat once more so that you are left with stages 1 and 2.
0.59 Ec(28)
Duration 15 days
0.96 Ec(28)
Duration 28 days
1.00 Ec(28)
Duration 99 days
1.06 Ec(28)
From the above it can be seen that changes due to duration do not have a large impact on the
results after the first few days.
Analysis Properties
As per unstaged construction, factors are applied via the Model Options/Stiffnesses tab
(shown above). These are applied across all construction stages.
Modulus of Elasticity
The Modulus of Elasticity Factor in particular requires careful consideration as it should be
employed in order to make any allowances for creep and cracking. The value entered here is
applied to the short term modulus of elasticity Ec(28) defined in Material dialog.
As described in the previous section Setting the Duration of Each Stage further factors are
then applied to the E value to account for the duration of each construction stage.
Note
Any allowance for creep is only dealt with effectively in the context of an
unstaged analysis..
Note
Footnotes
1. Taken fron Concrete Structures Amin Ghali, Renaud Favre, M. Elbadry
FE Merging
When G and Q loads are merged from the FE analysis they replace all the G and Q loadcases so G-fe replaces G, GP11, GP12, and Q-fe replaces Q, QP11, QP12, etc.
Staged loadcases are never replaced by FE merged results - the consequence of this on the
combinations is as follows:
For combinations of entirely staged cases no merging takes place.
For combinations of entirely unstaged cases - G, GP11, GP12, Q, QP11, QP12, etc. are
replaced.
For combinations that contain both staged and unstaged cases, only the unstaged cases are
replaced.
The latter situation is investigated in more depth below:
Terminology:
G and Q = results from 3D analysis
G-s and Q-s = results from 3D analysis where staged construction has been considered
G-fe and Q-fe = results merged from FE floor analysis where limitations noted above
continue to apply.
1. Initially both G and Q are staged and the option to Create New Combinations for Staged
G and Staged Q is checked:
the result is that before merging there are two sets of combinations:
G +Q
G-s + Q-s
In this case the merged FE forces would affect the unstaged combinations. Therefore:
G-fe + Q-fe
G-s + Q-s
This may be considered to be an attractive mix of combinations by some engineers.
2. You may have chosen to edit the original combinations so that the first combination is
based on staged rather than unstaged G, hence before merging the two sets of
combinations are:
G-s + Q
G-s + Q-s
FE merging only affects the unstaged results so in this case you would end up designing
for an envelope that simultaneously considers
G-s + Q-fe
G-s + Q-s
3. You may have chosen not to stage Q at all in which case before merging there is a single
combination:
G-s + Q
FE merging only affects the unstaged results so in this case you would end up designing
for:
G-s + Q-fe
The above might or might not be considered adequate, that is for you to decide.
Simulaneously designing for FE merged forces and the results of a completely unstaged analysis.
Clearly the above introduces the possibility that FE merged forces can be designed for
simultaneously with a combination where staged conditions are considered. What it does not
allow is the possibility of simultaneously designing for FE merged forces and for the results of
a completely unstaged analysis.
However, a workaround can be employed to achieve this if you run a single staged
construction analysis for both G and Q as follows:
In the Load Generator, check the options to stage both G and Q and also to Create New
Combinations for Staged G and Staged Q:
Edit the two staged loadcases so that the entire model is put in the first stage and set the
duration to 28 days.
Before merging there are now two sets of combinations, (although the staged combination
produces exactly the same result as the unstaged):
G +Q
G-s + Q-s
FE merging only affects the unstaged results so you end up designing for an envelope that
simultaneously considers
G-fe + Q-fe
G-s + Q-s (which in this case is identical to G + Q)
Chapter 8
Introduction
When you run FE Floor analysis you are presented with many options. In this chapter we aim
to help make sense of these options by providing an overview of their meaning/effect and also
indicate when each option would tend to be of most importance. Throughout this chapter we
will also cross reference to more detailed examples that illustrate the use of each option in
practice.
If you have previously run an analysis for the floor the settings will have been saved and will
now be redisplayed. A brief introduction to each item/option is given below.
We strongly recommend that the default Short Frame Model is used unless/until you have
some specific reason to try another option.
For Flat Slab Systems refer to the chapter Flat Slab Models and in particular the discussion
of slab deflection estimation.
moments that are generated over column heads in traditional analysis model idealisations.
Refer to the chapter on Flat Slab Models for an example of the effect of this option. Note that
activating this option may introduce more meshing difficulties during model generation.
For beam and slab models this option has little positive effect and may only introduce
difficulties, hence we recommend it is better not to activate this option for such models.
Note however that in some models excluding the beam torsional stiffness will allow a
mechanism to develop.
In the example shown here a curved elevation is modelled with a series of straight beams. If
the FE model is created without including beam torsional stiffness the deflections are extreme,
indicating the presence of a mechanism.
Around the curve, moments in one element translate to a combination of moments and
torsions in the adjacent elements due to the angular change.
When remodelled including beam torsional stiffness the deflections diagrams start to look
much more reasonable as shown below.
Note
Orion does not consider torsion within beam design, so if significant torsions are
developing these would need separate design checks. You can check torsion levels
by reviewing the FE Analysis Output Report, or if you have S-Frame, you can
export to that and review the results graphically.
Batch FE Chasedown
This option displays a table allowing you to specify all of the model generation options and
mesh parameters required for every floor in the building. By clicking on OK you are then able
to perform the chasedown procedure for all floors in a single batch. An option is provided for
you to check and if necessary adjust the meshing at each floor. It is recommended that the
meshing is reviewed and approved during the first run and also after any significant edits.
If you need to abort the chasedown at any stage simply click on the cross at the top right of the
analysis progress window shown below.
If you wish, you can review the model by activating more of the view options from the menu
on the left of the screen. The options you are most likely to switch on/off are also available as
drop down iconised menus along the top of the screen - you should find these more
convenient to use. However all of this (and more) can also be reviewed in the results window.
For detailed examples using this modelling option, refer to the chapter Transfer Beams FE
Method, Option 1 (Simplest) or to the following section on Merging Beam Results.
Validity Checking
Note that the validity checks accessed from the main Building Analysis dialog will issue
warnings about slabs that may cause problems during FE meshing and analysis.
These types of messages can be very helpful when trying to isolate meshing problems or mesh
regularity issues as discussed in the following sections. Note that messages such as shown
above are only warnings, they can be ignored provided you are happy with the meshing you
generate.
If you get errors referring to overlapping slabs and columns (such as shown below) these will
significantly affect the FE models and cannot be ignored.
Mesh Density
Mesh density is strongly influenced by the target Plate Element Size that is provided as input
before you press the Generate Model button. Orion suggests an initial default of 800mm, and
for a simple regular layout such as shown above where the slabs are placed in squares with the
supporting columns at the corners this default will be very reasonable.
The mesh that gets generated will also be strongly influenced by the the Mesh Uniformity
Factor which is accessed from the Model menu. In the above mesh (which was generated with
a mesh unifomity of 100%) 6 plate segments are produced between the column heads. We
recommend that you generally aim for around 6 to 8 plate segments between column heads.
Mesh Uniformity
If we alter the example above so that there are still 9 panels but they are not all the same size,
then meshing with the default plate element size and keeping the mesh unifomity at 100%
results in the mesh shown below.
This is not an ideal mesh. In several places there are only 4 segments between columns.
At this point we can start to adjust the Plate Element Size and/or the Mesh Uniformity Factor to
achieve a more satisfactory solution.
By retaining a uniform mesh and reducing the plate size to achieve the objective of having a
minimum of 6 plates between the closely spaced columns at top right, we must reduce the
plate size to 500mm as shown below. This produces a total of about 1800 plates; the top right is
now OK but we are generating more plates than we really need elsewhere.
If we swap to the other extreme, i.e. the least uniform mesh setting, and then adjust the plate
element size; a reasonable mesh is achieved when the plate size is set at 2000mm as shown
below. (Although in this example the uniformity factor is over-riding the plate element size
setting, so that the maximum size is not being achieved anywhere within the slab.) The total
number of plates required in this case is 1678.
The objective is to find a uniformity setting that allows you to reduce the number of plates
because you do not have areas with significantly more plates than are really necessary.
The mesh above is achieved with a plate size of 800mm and a uniformity of 50%. It has 866
plates rather than 1678 and is equally as good for the important bottom left corner panel.
Holes can not be placed on a slab boundary, but in the view above some holes have been added
close to a boundary. In the panel containing one hole, the hole is positioned 25 mm from the
edge of the panel. In the panel containing two holes, one is 25 mm from the edge of the panel,
and the two holes are only 25 mm apart.
If we mesh this model with a plate element size of 800mm and the default mesh uniformity
factor of 25% the result is as shown above. The meshing always generates nodes (edges) along
panel boundaries and is therefore trying to create very small plates to work through the small
gaps (as shown below).
The meshing shown here is actually useless for all practical purposes. In such circumstances
Orion will sometimes generate coarse but unusable meshes, other times it may generate a
message such as shown below.
Either way you need to recognise that either the meshing density/refinement needs to be
changed or the model needs to change. In the view below some small but important changes
have been made.
The slabs have been redefined so that there are no edges near to grid 10, this means there
are no small gaps to deal with between the holes and fictitious edges along that grid line.
The two closely spaced holes have been idealised as one larger hole thus avoiding the
problem of the small gap between them. In practice, where there are a cluster of services
openings this idealisation is not only helpful from an analytical point of view, it often
proves to be more economic and flexible from a construction point of view.
This revised model will mesh easily and well as shown above.
Mesh Sensitivity
In general the more elements you use the more precise the answers will be. Considering this
in conjunction with the notes in the previous sections you may conclude that if/when in doubt
the best thing to do is just ask for more plates and create a finer mesh. If you really are in
doubt this is not a bad rule to follow, but this is sometimes also referred to as over meshing
and it has a few drawbacks to consider:
As you add more and more elements you will find that the increased meshing is having
less and less effect on the results checking sensitivity to meshing changes is standard
practice in FE modelling. In essence this is only a speed problem meshing, analysis and
presentation/interpretation of results all get slower as the number of plates in a model
increases.
Stress Point Intensification with highly refined meshes you will begin to expose local
effects (very high forces theoretically applying over very short distances) that are of no
practical interest in RC design work. The most common example of this is the extreme
hogging moments that traditional FE analysis tends to expose over column heads in flat
slab models. The principle is however equally applicable to corners adjacent to holes etc.
In this example we will work with the reasonably simple floor layout shown above. If you want
to work through the example for yourself, you can load model DOC_Example_04. In order
not to destroy the example for someone else you should then save the model with a new name
before proceeding.
One point to notice is that slab angles that have been defined in the angled wing of this model.
The wing is angled at 40 degrees and by defining this angle in the slab information we are
essentially designing the angle of the orthogonal reinforcing system that will be used in this
wing. As we will see later this allows us to see much more relevant contouring details.
FE Mesh Generation
Start with the model settings as shown above. Later in this example we will look again at the
effects of adjusting the slab stiffness multiplier, in this section we will focus on meshing
ensure the Plate Element Size is set to 1000mm and the Mesh Uniformity Factor is set to 100
then click on the Floor Mesh and Analysis button.
Click on the mesh button and you should see a meshed model formed of 614 plates as shown
above this is NOT a good mesh. (See previous section).
The mesh generator tries to refine the mesh adjacent to holes with the result that remote
panels end up quite coarsely meshed as shown above. In beam and slab systems a reasonable
rule of thumb is that there should be at least 6 and preferably 8 segments generated along the
lengths of the beams. (In flat slab systems a similar principle can be applied to the stretches
between columns.)
There are two ways in which you can change the meshing. If you want to increase the
refinement locally around the holes and corners you can reduce the mesh unifomity factor, or
if you want to increase the refinement generally across the whole slab you can reduce the plate
element size.
The the mesh unifomity factor was set to 100, if we reduce this to 25 and re-mesh the result is
as shown below.
This is a big improvement, however there are still a few beams with only 5 and 6 nodes along
their length. Although the total number of plates used has increased to 2488, we might decide
yet more plates are required, and ideally we need a bit more mesh refinement around the
openings.
If we keep the mesh uniformity factor at 25 but reduce the plate element size to 600mm we get
the mesh shown below.
This has had the desired effect. It has to be noted again that meshing is not an exact science, if
in doubt you should reduce the plate element size and accept the slightly slower performance
associated with the larger model.
Having created an acceptable looking model, you can now exit from the FE preprocessor and
the model will be automatically analysed.
Deflection Plots
In any analysis package it is always good practice to look at the deflections if the deflections
do not make sense the results will not either. In this case the contours look very reasonable.
The model has an obvious angled line of symmetry, and the deflection contours are also
symmetrical. The only point that might draw some attention is the deflections of the primary
edge beams. They are not big deflections, but they are more significant than the slab
deflections.
Mx and My Plots
An Mx plot is shown above. Mx and My contours are displayed relative to a single global
coordinate system. Since this model has an angled line of symmetry the contour pattern is not
symmetrical.
If you imagine X direction bars running from left to right (horizontally) in the above view
then Mx moments are the moments that these bars will need to be designed to resist (but see
also the later notes on design moments including Wood and Armer adjustments).
Since it is unlikely that reinforcing bars will be positioned horizontally in the angled wing,
these contours are of little value for that region.
M1 and M2 Plots
In the introduction to this example it was noted that a slab angle (the reinforcing angle) has
been defined for each slab panel.
An M1 plot is essentially an Mx plot where the contours are displayed relative to a local
coordinate system for each slab panel. The local system is defined by the angle of rotation
applied to each slab panel. For this model this means that we can now see a symmetrical
contour plot where the contours in each wing are relevant to the orthogonal reinforcing
system that will be used in each wing.
Notice that in any contour plot you can use the mouse pointer to highlight any node (as shown
above) and the precise information about that node is displayed at the foot of the window.
Plots Including Wood and Armer adjustments
Wood and Armer adjustments take plate torsional moments into account to generate adjusted
design moments. It is beyond the scope of this manual to present the derivation of the
equations for these adjustments. If detailed background on this is required you should refer to
slab/bridge design texts, or to the original papers:
Wood, R.H. The reinforcement of slabs in accordance with a pre-determined field of
moments as published in Concrete, 2. February 1968, pp69-76,
An Md1-bot plot is shown above. This is essentially the M1 sagging (bottom) moments
adjusted to account for plate torsion. The blue zones on the plot are hogging areas, the
maximum sagging moment is 5.4, which is almost identical to the M1 sagging moment (5.3)
shown previously.
An Md1-top plot is shown below. This is the adjusted M1 hogging (top) moments. The red
zones on the plot are sagging areas, the maximum hogging moment is 8.2, which is a little
higher than the M1 hogging moment shown above.
Since they are based on the M1 and M2 moments these design moment plots are also
symmetrical for this floor system.
The degree to which Wood and Armer adjustments affect the design moments is entirely
dependent on the levels of Mxy (or M12) moments these can be regarded as twisting
moments.
Regular beam and slab systems do not tend to have high twisting moments coincident with
peak hogging/sagging moments and Wood and Armer effects are often of little significance.
In this model you can find places where the effect I more significant.
You can review the Mxy or M12 moment contours as shown below.
Corners are always areas where twisting is high. In this model there is also quite a high twist
developing over the beams along the angled line of symmetry. We will see how this affects the
strip design moments later in this example.
Steel Reinforcing Requirement Contours
Before accessing the postprocessor you will have noted the option to apply global adjustment
factors to all Positive (Sagging) or Negative (Hogging) moments. This is not the same as
moment redistribution, but it would potentially allow for redistribution style effects to be
introduced.
It should be noted that FE floor models do not include any pattern loading, it is not feasible/
logical to automate pattern loading to generate every possible worst case scenario for every
conceivable irregular arrangement and any size of model. A more realistic use of these
adjustments may be to amplify the sagging moments (perhaps 10 to 20%) if you are concerned
about allowing for load patterning. This will be discussed further later in this example.
Once you are in the post processor viewing moment contours you will also have noticed a
drop down option relating to Col/Wall Node Interpretation. This has little effect on beam and
slab systems and is a little more relevant to flat slab design. This option is discussed in more
detail in the chapter on Flat Slab Models.
Slab Design
The design approach, which involves creating slab strips and then designing based on
tabulated code coefficients, is covered in the training manual. Discussion there also extends to
the option of using an FE strip. In this section we will look at some comparisons between the
two approaches and also look a little closer at the FE modelling options and how these might
affect design based on an FE analysis.
For this sort of regular slab it would be considered quite reasonable to use the traditional strip
method outlined in the codes. The importance of setting the slab type correctly is noted again,
refer back to the training manual if you have any doubts on this.
When you design the strip X1 on this basis the design moments are given in the output shown
below.
Now we can change the strip to an FE Span Strip by updating the slab strip properties first
check the FE Strip option under the General tab then choose the appropriate type of FE strip on
the FE tab as shown below.
Now remesh the model and then go into the FE Postprocessor to review the analysis results.
In the view above the moment diagram is determined along the actual cut line, in the table
below you can see the peak moments captured from the wider area of slab as indicated on the
plan view in the background.
We can compare the hogging and sagging moments in the table above with those reported
using the coefficient method earlier. Where we now have a hogging moment of 7.5 kNm, it
was 4.6 kNm. Where we now have a sagging moment of 4.8 kNm, it was 4.5. Before
continuing it is worth emphasising that these peak moments indicated above are very close to
or exceed the moments determined by the empirical code approach.
More interestingly, the peak hogging does not occur at the first internal support, at that
support the FE analysis gives a moment of 4.4 kNm,whereas a traditional empirical approach
gives 5.6 kNm. Traditional idealisations will assume that beams are rigid lines of support, but
this support is deflecting. The deflection diagram for the strip cut line can also be viewed as
shown below.
The relative beam deflections are obvious and in a stiffness analysis this will clearly have an
impact on the design shears and moments. In effect we have built-in modelling of support
flexibility in an FE analysis.
Increase the beam stiffness multiplier to 1.5 (arguably to account for the flanges), and decrease
the slab stiffness to 0.01. We would not suggest that such a large adjustment be used in
practice.
In the view below we can see deflection contours and the moments determined on the strip.
These moments are actually more in line with the moments determined by the empirical
approach. The hogging moments are more regular but the peak is still a good bit higher than
determined by the empirical approach, 7.0 kNm as opposed to 5.6 kNm.
The sagging moments are similar but a little lower than determined by the empirical approach,
3.2 kNm as opposed to 3.4 kNm being the most extreme variation. (Ignoring the angled slab).
The empirical approach is deemed to include for load patterning and moment redistribution.
BS8110 also advises that where the loads are not patterned, there should be a 20%
redistribution of support moments with a resulting increase in sagging moments. In the
current model that would mean reducing the 7.0 kNm hogging to 5.7 kNm and increasing the
4.4 kNm sagging (end span) to about 5.0 kNm, and the 3.2 kNm sagging (internal span) to
about 4.4 kNm. This gives very good agreement on the hogging moments, but the sagging
moments are higher than determined by the empirical approach.
Note that for reasons made much clearer in the chapter on Flat Slab Models, the simple
principles of redistribution as traditionally applied to 2D frames do not make sense within the
more complex analytical geometry of a 3D FE model.
Returning to the previous model where the beam and slab stiffnesses might be regarded as
more reasonable, we did not have the extremely high hogging moments to start with. To some
degree by actually accounting for the beam stiffness we have shifted the BMD reducing the
hogging moments and increasing the sagging moments. Would it be safe to reduce the
hogging moments further? Should the sagging moments be amplified?
It is important that we do not regard the empirical approach as the correct answer, it is one
possible method that comes with documented limitations. When we analyse a complete 3D
subframe model we are dealing with a good analytical model, and it probably does not make
sense to start making lots of adjustments to try and make its results fit with those derived from
a table of coefficients.
Before accessing the postprocessor you do have the option to apply moment adjustment
factors as shown below.
We are unaware of any authoritative texts providing analysis and design guidance relating to
FE modelling and analysis. In considering the requirements to adjust the moments we are
probably considering serviceability rather than safety issues. This is all discussed in greater
depth in the chapter on Flat Slab Models. In that chapter we suggest that you should be very
cautious about applying negative adjustment factors of less that 1, it is suggested that you
might apply an adjustment to the positive (sagging) moments of up to around 1.2. Once again,
we emphasise that this is not the same thing as 20% moment redistribution.
If we return to the original beam and slab stiffness settings you can then review the results
after introducing a positive (sagging) moment multiplier as shown above.
In the view below the hogging moments are unchanged, but the sagging moments are factored
up by 1.2.
Orion provides the tools to make such adjustments, but in the absence of any authoritative
guidance on the subject we cannot make any definitive suggestions. In this particular example
the unaltered moments seem to provide a reasonable set of design forces, when the option of
amplifying the sagging moment is used an even safer set is generated. If in doubt we can only
suggest that you consider making this sort of adjustment.
Refer back to the section Plots Including Wood and Armer adjustments in this chapter for an
introduction to this. Consider the same strip (X1) swap to view the Design Moments, this
diagram shows the hogging and sagging moments after applying Wood and Armer
Adjustments. Note that in some places both a hogging and a sagging moment is generated.
In this case the 1.2 factor is still being applied to positive (sagging) moments. Comparing with
the previous diagram all the forces increase slightly, but the moment at the right hand end
where the support beam is angled to the line of the strip (and to the line of the reinforcement)
is more significantly increased.
The adjusted design moments are never (by definition) less than the unadjusted moments.
Designing based on adjusted moments is optional. Comparing the above with the moments
calculated based on tabulate code coefficients, you might feel even more justified in not
applying an amplification factor of as much as 1.2 to the sagging moments.
Reinforcement Design
As you exit from the FE postprocessor a dialog is displayed.
The transfer option needs to be checked in order to pass FE results back to Orion for strip
design. The interpolation option is of little significance in beam and slab work. It is more
important in Flat Slab Models and it is therefore discussed in the chapter Flat Slab Models.
You can choose whether or not to use the design moments including Wood and Armer
adjustments.
Back in the main Orion graphical editor, you can select one or several strips. You can then
right click and select options to run a check design on any existing bars, or delete the bars and
then update the strip to design new bars.
You could for example design strips based on the empirical code method, and then update the
slabs to FE Strips and check the steel provided on the basis of FE results.
This may in fact be a very attractive option in many cases. In this example (since minimum
steel requirements dominate) the original steel works for the revised moments from FE.
Example
To illustrate grillage effects we can continue with DOC_Example_04. In order not to destroy
the example for someone else you should then save the model with a new name before
proceeding.
In this example we will focus on the selected beam, 1B17. At the lower end of this beam, its
reaction is transferred to the long horizontal beam (1B2). At the top end the beam frames into
a wall.
After running the building analysis you can review the results. Beam 1B17 is supported at one
end on beam 1B2. This beam can be seen to deflect vertically, and rotationally it will act like a
pin, because the torsional stiffness factor used in the analysis was left at the default of 0.01.
At the other end Beam 1B17 does not deflect as it is supported by a wall.
The shears and moments in this beam resulting from this analysis are shown below.
In addition to the above it is also worth noting the moments in the supporting beam 1B2 as
shown below.
Alternative Beam Results can be generated in FE and merged with Building Analysis Results in
either of two ways:
1. Using a Plateless model.
2. Using a fully meshed model.
We will look at each of these in turn in the following two examples.
1. After analysis click on the option to merge beam results and merge results for all beams.
2. Now review the merged results for beam 1B17.
There is very good agreement between this result and that obtained by the building
analysis. This is to be expected because in the plateless model the slab does not participate
in the analysis. The slab loads are transferred directly to the supporting members based on
the yield line distribution method.
3. You can make similar comparisons on other beams in this model to see how the design
forces hardly change. On Beam 1B2 for instance, the moments are almost identical as
shown below.
Having merged all the FE results, you can now carry out a batch check design using the option
not to revise the reinforcing and the utilisation ratios hardly change. The moment in beam
1B18 increases marginally so that its utilisation ratio is just greater than 1.0, sufficient to
register as a fail.
For this reasons illustrated in the Merging Beam Analysis Results section in the chapter
Transfer Beams General Method, we recommend that having analysed and designed all the
beams using one method (e.g. Building Analysis) you should then run a check design for
merged results to find any members that fail. At this point you should adjust (increase only)
the steel in those members interactively until they pass. Then you should find that you can
run a check design for either set of analysis results and everything should pass.
Again, it is worth noting that one of the most compelling reasons for not simply using FE
analysis, merging results and then designing for the merged results is that FE analysis does not
deal with patterned load, everything is fully loaded. It is very worthwhile trying to use the
general building analysis and then merge and check for FE results selectively.
Run the FE Floor analysis, use the settings shown previously, then click on the mesh
generation option to access the FE Analysis preprocessor. Mesh up the model and you should
see a mesh as shown below.
Exit from the preprocessor to analyse the model. After analysis click on the option to merge
beam results and merge results for all beams.
In this model all the loads are applied to the shells, (no applied load will be displayed on the
beam loading diagram). However, the effects of the loading from both the shells and the
supported beam are clear to see on the shear force diagram.
In this case the design moments are in reasonably close agreement with those determined by
both the Plateless model and the building analysis model. It is important to note that this
result is related to the use of the slab stiffness multiplier that was set to 0.15. If we reset it to 1.0
and run the analysis again, the merged results for the same beam change as shown below.
The diagrams are similar, but the maximum shear and moments are all lower. This occurs
because the slab is sufficiently stiff that it starts to carry a significant proportion of the load
direct to the columns. Other aspects of this effect relating to slab design were discussed earlier
this chapter. In general, if you are designing beams using FE analysis results with meshed up
models, then you need to be very careful that the beam design forces are what you expect them
to be. This can often be achieved by setting the slab stiffnesses to be low or extremely low and
then perhaps ignore slab deflection results.
Checking the Beam Designs
Having merged the FE results for the meshed model the beam design can be checked in the
same way as described for the plateless model.
Chapter 9
Introduction
The following section of the Engineers Handbook contains five chapters primarily relating to
the topics of Building Sway and Differential Axial Deformation Effects:
The Overview of Bracing and Sway Sensitivity chapter clarifies the bracing and
slenderness classification procedures for columns and walls for each of the design codes.
The subsequent design implications are also examined.
Chapter 10
Analytical Idealisations
Orion supports alternative analytical idealisations of walls.
This topic is discussed in detail within a training presentation that is available in PDF format
alongside this document and accessed via the link below.
Shear Wall Modelling Presentation.pdf
(If the link does not work please browse to find the file name indicated above in the HELP
sub-folder of the Orion Program Folder)
This presentation is given and discussed in detail during Orion Advanced Training Days.
Deflection
For deflection estimation the above presentation demonstrates that the alternative analytical
idealisations yield very similar results for any given input data.
However, if deflection estimates are to be regarded as meaningful they have to be based on
correct input and in this regard allowances must be considered in two areas:
Material Properties Deflections are directly proportional to the Youngs modulus value
you define as input, what is the correct value to use? A short term value may be
appropriate for wind loads, but perhaps a long term value should apply to notional loads?
Allowance for cracking gross section properties are used by default, should some
adjustment be made to allow for cracking? (It is understood that many engineers will
adjust (reduce) stiffnesses by up to around 50% to allow for cracking, the ACI code may be
referred to for some guidance in this regard.)
Provided that the mid-pier element of the wall above sits somewhere on the rigid arms of
the wall below this is not a problem.
In the special case where the mid-pier element of the wall above sits on a coupling beam
over an opening in the wall below the mid-pier model will generate some very
conservative design forces in the coupling beam. (As is discussed in relation to transfer
beams in the chapter Overview of Solution Options for Transfer Levels and onwards).
Hence in some specific circumstances the meshed wall idealisation may be considered
preferable.
In the analysis model options you can adjust the relative stiffnesses of walls/columns/beams.
You can make the walls stiffer and/or you can weaken the columns and beams.
Several Points should be borne in mind if such adjustments are made:
If changing the wall stiffnesses it is better to apply a factor to the E value since this will
apply to both the meshed and mid-pier modelling options.
If you only change the column stiffnesses then you will change the way in which moments
develop at beam/column interfaces. It makes more sense to make relatively equal
adjustments to both the beams and the columns.
If the above adjustment is made at not too extreme a degree (e.g. reduce beams and
column stiffnesses by factor of 0.5), then the analysis results are probably valid for design
of all members.
If extreme adjustments are made than it is probably safest to use these analysis results only
for a cross check on wall design. Therefore the approach to design might be:
The structure above is a very simple 10 storey model with a single C-Shape core which is
completely offset to one side of the building. For the purposes of this example only, all the
beams have been pinned at both ends thus eliminating any frame action which would provide
sway stiffness. The structure is therefore completely reliant on the core walls for stability.
The first point to note is that this is an extreme example. The view below shows how the
building sways and twists under the action of notional loads.
The twisting effect is considerable and the maximum deflections at the top floor actually
exceed 1 m. A real structure cannot be acceptably stabilised by such an eccentric C-Shape
core.
When we consider the dead load case (where only vertical loads are applied) lateral sway is
also evident as shown below.
The design forces in the wall panels at first floor level are shown below. Note that the loads in
the flanges 1W1 and 1W3 are higher than in the web 1W2. The average load per meter in
these shorter flanges is therefore a lot higher.
Looking at this view it is possible to imagine how the effective centre of loading applied to the
C-shape core is offset to the right and above its centroid (which will be quite close to the
centre of rigidity shown in the view above). The eccentric load causes an unresisted curvature
in the core and hence sway develops at the upper floors.
The same effects are evident regardless of whether meshed or mid-pier idealisations are used.
Before moving on to look at a very similar model but using flat slabs it is worth noting that the
total Imposed Load in the three panels of the above core wall is 626 + 229 + 403 = 1258 kN.
As is noted above the sway effect arises because of the eccentric application of load to a core
wall. The magnitude of the sway is related to the magnitude and point of application of the
axial load.
Slab Loads Yield Line Decomposition
If this building is analysed using yield line decomposition of slab loads then a message is
displayed to indicate that the building analysis has not captured all the applied loads.
At this stage the analysis results displayed on plan will show no loads in columns and some
loads in walls.
The amount of load in the walls will be incorrect, it could be too high or too low, in this case
the total Imposed Load in the three panels of the above core wall is 275 +139 + 276 = 690 kN
a lot less than was derived in the model where pinned beams were included.
Basically, as is well documented elsewhere, this building analysis is of no value as regards the
dead and imposed load cases. For a flat slab model the design cases for the dead and imposed
loads would usually come from the merged results of an FE chase down.
The total Imposed Load in the three panels of the above core wall is now 768 11+ 695 = 1452
kN. As a total this is not too dissimilar to the 1258 kN total given for the model that included
the pinned beams, slab continuity effects in the FE analysis mean that more load is attracted to
the core in this case. This load is also more eccentric with the result that tension develops in
the web (for this extreme example.
Once again the sway of the structure for the gravity load case is evident in the analysis
postprocessor view and the variation of stress along the length of the flanges of the C-shape
core is also evident.
In this example the maximum sway that is developing is 65 mm, the discussion below should
put this in some context.
Discussion
Load Eccentricity
As noted at the outset, the model used above is an extreme example. In order to generate the
most extreme offset loading within the C-shape core it is positioned along on edge of the
model. In this position this single core cannot stabilise the structure effectively. The sway
under notional load cases exceeds 1000 mm. This structure is not something that would be
built.
Where structures are stabilised by a single core, the core is normally positioned more centrally
and will tend to attract more uniform loading.
If the example is adjusted as shown above so that the core attracts load from all sides then the
sway under gravity load only drops from 65 to 13 mm (everything else being equal). Even so
this structure still undergoes very significant sway due to notional loads and would again
probably not be something that would actually be constructed. (The open C-Shape of the core
is very susceptible to twist.)
Construction and Creep Effects
In the event that sway induced by gravity loads seems significant and some estimation of the
actual displacement is required then some thought should be given to the sequence of
construction and to subsequent creep effects.
The sway is occurring as a result of differential axial deformation within the walls. This axial
deformation is a combination of immediate deformations each time a new floor is supported
and ongoing creep deformations. Both of these effects are time dependent. When a floor load
is added the concrete in the walls that support the floor will be younger in the floor
immediately below and progressively older at lower floors. The younger concrete will see a
greater immediate deformation and a greater long term creep deformation as a result of any
applied load. A detailed discussion of this is beyond the current scope of this handbook, but it
is worth noting that there would be a good deal of common background with the discussion of
flat slab deflections accessed via the chapter on Flat Slab Models.
A very sophisticated assessment of this effect would take account of construction sequencing
and time dependent effects. In such circumstances the result needs to be assessed for
sensitivity to variations in the assumptions on which they are based. To be fully accurate this
analysis would need to take account of the fact that some of the deflection would be
constructed out as the walls are constructed as vertical as is practical during construction.
In fact the result achieved by any such complex analysis will lie somewhere between two
extremes that might be predicted more readily be a simple linear static analysis using different
E (Youngs Modulus) values for the concrete.
A short term E value might be used to estimate a minimum value for the total deflection.
In the case of walls cracking effects can probably be discounted so only additional creep effects
need to be considered. It is likely that the net effect of creep will be something less than a
doubling of instantaneous deflection. Hence the value determined above could be doubled, or
another analysis could be made with E set to half of the short term value.
Note
Where flat slab models are being considered it is common practice to use long
term values of E set at around of the short term value. When this has been
done deflection estimates of sway arising from the building analysis are likely to
be too high, perhaps by a factor of 2 based on the suggestions above.
Discrete Cores
Where structures have discrete cores or numerous discrete walls providing stability the effects
will often counteract each other.
In the above view the model is mirrored so that there are two C-shape cores. This model does
not sway at all in the X direction under gravity load and the vertical load contours indicate
relatively uniform stresses within the cores.
In this case the two cores lean against each other, the diaphragms at each floor stop any
curvature from developing. The axial loads in the wall panels at first floor level are much more
uniform as shown below.
The total Imposed Load in the three panels of the above core wall is now 435 + 663+ 324 =
1422 kN. On a panel by panel basis these are very different forces than were given by the
un-mirrored model, but the total for the 3 panels is very similar to the 1452 kN given
previously. Note that the major axis moments in the wall panels are also much reduced.
Results based on an FE Chase Down
For most models engineers will be happy to use the panel design forces merged from an FE
analysis chase down. In this case the results for this core wall would then be as shown below:
Once again the distribution of forces in the 3 wall panels is different, but the total Imposed
Load is very similar at 561 + 391+ 473 = 1425 kN. The FE Chase down method does not deal
with sway introduced by differential axial compression in the core walls over the full height of
the building. It does however introduce more minor axis bending effects where the walls
interact with the slabs.
For many walls supporting flat slab floors the minor axis bending can be more significant in
the panel design than variations in axial load and major axis moment that may be introduced
by sway effects.
For this simple 10 storey example building, despite the fact that the sway effects have been
forced to be unrealistically extreme, the walls only require nominal reinforcement regardless
of which set of design forces their design is based on.
It is suggested that for flat slab models it is important to have always designed the walls for the
loads generated based on an FE chase down. You then have the option of cross checking these
designs for the results generated by a building analysis where wall loading has been
determined using FE decomposition at each floor.
Closing Summary
In this section we have been discussing Sway Effects Under Gravity Loads, it is important to
emphasise that this relates entirely to sway that is generated in the presence of purely vertical
applied loads.
The possibility of sway developing due to differential axial shortening is something that would
have been routinely ignored in hand calculations by most engineers looking at most
structures.
This has nothing to do with notional loads which are applied separately and will always be
designed for provided you have included them in your design combinations.
For framed models in Orion the effect is always included / exposed by the 3D analysis.
For flat slab models the effect will not be exposed by an FE chase down but can be exposed if
desired. It is emphasised however that we recommend walls are designed for the loads from
an FE chase down and then a subsequent check design can be made for building sway effects if
desired. For many reasons noted within the text of this section it is anticipated that this check
will only produce more critical design in tall buildings and/or quite extreme examples.
Chapter 11
Introduction
In order to produce a sway deflection comparison against other analysis software, a 20 storey
building is analysed in Orion using both mid-pier and meshed idealisations of walls. (The
Example 20 Storey Building described in the later chapter on Differential Axial Deformation
Effects is used for this purpose.)
The results for lateral load cases are then compared against other analysis packages.
Example Model:
Mid-Pier Wall
Model
Meshed Wall
Model
39187
39187
39187
39187
39187
39187
11985
11985
11985
11985
11985
11985
101.6
100.1
-1.5%
101.3
99.6
-1.5%
101.8
100.3
-1.5%
79.9
76.0
-4.9%
35.9
35.8
81.7
77.7
-4.9%
Difference
It can be seen from the above that both midpier and meshed FE shell wall modelling options
work equally well.
S-Frame
SAP
39187
39187
11985
11985
101.6
101.6
101.6
101.3
101.4
101.3
101.8
101.7
79.9
79.9
79.9
35.9
35.9
35.9
81.7
82.5
S-Frame
SAP
ETABS*
39187
39187
11985
11985
100.1
100.2
100.1
104.3
99.6
100.0
99.9
100.3
100.3
76.0
76.2
75.9
79.9
35.8
35.8
35.8
35.7
77.7
78.4
* For the Etabs comparison the model was constructed by a 3rd party. In this model the
default wall idealisation was not used, the walls were meshed more finely. CSC do not
assume responsibility for the accuracy of this model.
Chapter 12
Introduction
The purpose of this chapter is to clarify the bracing and slenderness classification procedures
for columns and walls for each of the design codes. The subsequent design implications are
also examined.
If you do not check this option then Orion will make an assessment of sway sensitivity based
on analytical results and the recommendations of the ACI code.
Note however that this assessment is deflection dependent and that the ACI code gives
guidance on appropriate adjustments to section/material properties to be used in the building
analysis for the purposes of this assessment. Such adjustments will increase the deflection
value used in the checks.
Note that the check can result in different classifications for different storeys which is not a
condition that is recognised by BS8110. We continue to advise that this assessment is used
cautiously and that when the appropriate overall condition is determined that this should be
applied as a user-defined classification (in accordance with BS8110).
ACI 318-02
When the design code is set to BS8110, CP65 or HK-2004; if you uncheck User Defined
Bracing for Columns and Walls, a facility is made available for assessing the susceptibility of
individual storeys to P-Delta effects. This uses the ACI method of classification during the
building analysis.
Bracing Classification using the ACI approach each storey level within a building is
classified as sway or non-sway. The code also provides a method allowing analytical
assessment of this classification based on deflections arising from a linear analysis of the
structure.
Global P-Delta Effects when a storey is classified as "non-sway" then it can be assumed that
global P-Delta effects are small enough to be ignored at that level. When a storey is classified
as "sway" then the frame analysis results need to be amplified in some way, options given are:
A second order analysis (which would inevitably affect all members in the structure)
Approximate moment magnification methods (cl.10.13.2 appears to indicate that this
moment amplification only needs to be applied to the slender members at each floor level
(similar to BS8110) is this logical? - or should this amplify the sway moments in all
columns and walls on a level by level basis?)
Slenderness Classification this is based on the effective length. At "Non-sway" levels
effective lengths are < 1 and at "sway" levels effective lengths are > 1. It is considerably more
likely that a member gets classified as slender when it exists at a "sway" level.
Short (Non-Slender) Members will see no amplification of moment at all even if they are at
"Sway" levels.
Slender Members (Members susceptible to P-Delta effects)
Slender Elements at Non-Sway Levels - additional moments are calculated based on
effective length and are considered to be a maximum at around mid height. These
moments are not added to the highest end moment so this may or may not end up being a
critical design condition.
In essence the approach here is identical to that used for braced slender members in
BS8110.
Elements at Sway Levels - as noted above the end moments of all members may be
amplified to account for Global P-Delta effects. If a member at such a level is classified as
slender, the calculation of the magnified moment is not based on the effective length of
each individual member, moment magnifiers are based either on the stability index for the
floor (cl.10.13.4.2) or an assessment of the average buckling capacity of all members at the
floor (cl.10.13.4.3 - similar to the optional method in BS8110).
The additional moment is added to the highest end moment so this will always end up
being a critical design condition.
Additional check (cl.10.13.5) - having amplified the end moments there is a requirement
to check that intermediate slenderness effects (using effective length = 1.0L) are not more
critical.
While the method of moment amplification is different for slender members at sway levels, the
general principles of moment amplification are the same in BS8110 and ACI and the terms
used for classification are interchangeable:
BS8110 Braced = ACI Non-Sway
EC2
In EC2 similar terminologies are used but the meanings are different:
Cl 5.8.1 - Introduces concept of braced and bracing members.
Cl 5.8.2 - Second Order Effects - this clause distinguishes between global effects (applying
to the whole structure) and isolated member effects (slenderness).
Bracing Classification Bracing members are the members which are assumed to provide
the lateral stability of the structure. Columns and walls that are not bracing members are
classified as braced. Unfortunately there is an element of engineering discretion involved in
this classification which will be discussed later.
Global P-Delta Effects there is some guidance on determining if these effects can be
ignored (For the purposes of this discussion we will classify structures in which global P-Delta
effects cannot be ignored as "sway sensitive"). Cl 5.8.3.3 (1) gives a simple equation that is only
applicable in limited circumstances and is actually also difficult to apply. Initial calculations
using this equation have suggested that it would be too conservative resulting in too many
structures being classified as sway sensitive.
Annex H provides slightly more general guidance. In order to automate the Annex H
classification in Orion, the approach has been modified to become similar in principal to the
ACI classification method. It is noted that a single classification gets applied to the entire sway
resisting structure (the bracing members). If it is determined that global P-Delta effects
cannot be ignored (the structure is sway sensitive) then the approach becomes a user driven
procedure, in which the sway loads are amplified in accordance with Annex H. This is a
relatively simple procedure applied as follows:
1. View the sway sensitivity report to obtain the suggested load amplification factors.
2. Apply this amplification to the existing load combination factors.
3. Re-analyse using the option to over-ride further sway sensitivity assessment and design
the structure as if it is not sway sensitive (because the global P-Delta effects are now
catered for).
Tests have indicated that the sway sensitivity assessment procedure described above results in
a non-sway classification for the vast majority of structures .
Note
Although the classification applies to the bracing members, it is impossible to
isolate these when analysing the structure, so P-delta forces (introduced by load
amplification or P-delta analysis) will accrue in all members (braced or bracing,
short or slender).
Slenderness Classification this is based on the effective length. For braced members
effective lengths are < 1 and for bracing members effective lengths are > 1. It is considerably
more likely that a member gets classified as slender when it has been classified as a bracing
member.
Short (Non-Slender) Members:
As noted above, if these members exist in a sway sensitive frame then there may have been
some amplification of the design forces introduced during the general analysis procedure.
EC2 Bracing = BS8110 Un-Braced (but we would expect that the EC2 amplification might
be lower since the BS8110 amplification at this point mixes both global and local effects
whilst in EC2 any global effects would already have been introduced).
The purpose is to identify Bracing Members in each Global Direction (The member types that
contribute to lateral stability of the building). The default setting is as shown above, (columns
considered to be braced; walls considered to be braced about their minor axis, but to provide
bracing to the structure about their major axis).
The Annex H guidance has been adopted for the first Eurocode release of Orion.
If the structure is classified as sway sensitive then there are two options for dealing with this:
1. Annex H - Application of increased horizontal forces.
2. Do a P-Delta Analysis
In fact there is a third option which might be applied when an engineer discovers a building is
sway-sensitive - they may find a way to add more shear walls and change the classification!
Initially the P-delta option may seem attractive but it must be recognised that EC2 is very clear
on the fact that realistic member properties accounting for creep and cracking must be used
and the calculation of these properties becomes a unique procedure for every member.
Note
For sway sensitive structures, the Annex H guidance has been adopted for the
first Eurocode version of Orion.
A model is constructed as shown above with two 3m wall panels providing stability in each
direction.
Floor to floor ht= 3.0 m
Wall Length / Width= 3m / 0.2m
Concrete Grade= C30/37
G= 7 kN/m2 (total including walls)
Q= 2.5 kN/m2
Beams are provided for load collection only - they are pinned at both ends in order that lateral
loads are focussed in the shear walls.
For the first Orion Eurocode version these adjustments are left entirely to the
user.
Referring to eq 5.26, if we assume theta-ef is around 1.5 then the suggested approximate
stiffness adjustment can be calculated:
Kc = 0.3 / (1 + 0.5*theta-ef) = approx 0.175
For the beams adjustments must be made to allow for creep and cracking - assume:
I-cracked = 0.5 I-conc
Member Properties are adjusted before analysis as shown above. Note that the recommended
column and beam adjustment is 0.7 and for cracked walls it is 0.35.
The report shows the structure is classified as sway-sensitive at all but the lowest floor level.
In the ACI only 5% second order effects are assumed to be ignorable. Q is the measure of this
and at this point it is interesting to note that although Q is only marginally smaller than 0.05 at
the lowest level, it becomes quite significantly greater at the top level.
In fact, if we reduce this to a 4 storey building then the report below shows that the structure is
still classified as sway-sensitive at the upper levels.
Note
Based on the discussion in Model Analysis Properties, member properties are adjusted before
analysis as shown above. Note that although we are using 0.17, you may decide on a higher or
lower value based on your engineering judgement.
The report shows that the 5 storey structure is classified as sway-sensitive at all floor levels.
In EC2 10% second order effects are assumed to be ignorable. Q is the measure of this and so
the actual check is that if Q > 0.1 then the classification is sway-sensitive. For the figures above
we can see this is true at all levels.
It is noted that although Q is only marginally greater than 0.1 at the lowest level, it becomes
quite significantly greater at the top level.
In fact, if we reduce this to a 4 storey building then the report below shows that although Q
becomes less that 0.1 at the lowest level, the structure is still classified as sway-sensitive at the
upper levels.
Although the reduced section properties together with the increased ignorable P-Delta
amplification limit means that the threshold for sway-sensitive/non-sway classification is very
similar for the two codes, the amplification factors that apply to buildings that are classed sway
sensitive are bigger (double) for EC2.
EC2 does not seem to recognise the concept that a building can have different sway sensitivity
at different levels, a single classification and amplification factor is applied to the whole
building. This requirement is catered for in the report by including an extra line for All
storeys. In the above 4 storey example the Q value calculated for All storeys is 0.1497
(therefore sway sensitive).
Total deflection = 5.99mm
Total Axial Load (F-V.Ed)= 30349
Total Shear Load (F-H.Ed)= 101.2
Total height = 12m
Q = 1.5 (30349 * 0.00599) / (101.2 *12) = 0.1497 > 0.1 (therefore sway sensitive).
Chapter 13
Introduction
Differential axial deformation will occur in the 3D analysis of any typical building frame. This
deformation can significantly affect beam bending moments. For the purposes of discussion
the effect is referred to as a "Differential Axial Deformation Effect" (DADE).
During any analysis members subject to compression will shorten. In a concrete frame
columns and walls under compression will both shorten. Highly stressed members will
shorten by greater amounts than lower stressed members. Where two adjacent members
shorten by a different amount there will be an effect (DADE) introduced in any members that
attach to both the shortening members.
Example:
Focusing on DADE, the key points of the discussion in these notes are:
Traditional methods of idealising sub-sections of structures for design purposes inherently
completely ignore this effect. However, the existing building stock which is largely
designed in this way appears serviceable, it seems that ignoring DADE has not caused any
significant problems.
There is a demand to use 3D analysis but at the same time find a way to achieve member
design forces that are reasonably close to the forces that would always have been used.
These notes demonstrate methods of achieving this.
Checks are required to establish whether the method used has had an impact on sway
deflections. If it does not then potentially designers can stop at this point on the basis that
the traditional design approach has been reasonably emulated.
Potentially engineers will also want to consider that DADE should not be completely
ignored. These notes show how a second analysis run could be made so that the initial
design can be checked and strengthened as necessary. In essence this means that the
structure ends up being designed for a wider envelope of design conditions (forces), clearly
this must be a conservative approach.
Concerns have been expressed that designing for this envelope could be uneconomic. For
an example model it is shown that the additional reinforcement associated with this
conservative approach is quite nominal.
It is important to re-emphasise that although this discussion is in relation to Orion, the effect
under discussion will occur in any analysis software. Equally, most of the suggested strategies
for dealing with this effect could be applied in other software.
Finally, the possibility that staged construction analysis will somehow easily deal with this
issue is often raised. This point is covered in the closing discussion the key points being that:
Staged Construction is not a simple analysis technique.
It does not generate results that emulate traditional design (i.e. DADE is not completely
eliminated)
If it is not used with a reasonable degree of care and understanding then a staged
construction analysis can be a sophisticated and time consuming way of getting dubious
answers.
Opening Discussion
In these notes we will establish some traditional sub-frame results for beams at the 20th floor
of a building and then look at how closely these results can be emulated using full 3D analysis.
Several general discussion topics should be borne in mind as the results are compared.
Traditional practice for the analysis and design of multi-storey concrete frames has involved a
significant degree of idealisation of the structure:
Lateral load is often assumed to be exclusively resisted by a selected subset of members:
Special models are constructed to determine the member forces which develop when
lateral loads are applied.
It is quite normal to ignore the sway that can be introduced when pure gravity load is
applied to a non-symmetrical structure.
The result is that designs would be the same for the beams on the 20th floor as on the 1st
floor.
We present the results from the traditional analysis as a target "baseline" result. This should
not be confused with being a "correct" result. A few thoughts on this:
A Quote:
"Engineering (and some may think FE practice also) is the art of modelling materials
we do not wholly understand, into shapes we cannot precisely analyse, so as to
withstand forces we cannot properly assess, this in such a way the public and
(hopefully) the customer has no reason to suspect the extent of our ignorance."
Structural Engineering Modelling and Analysis,
by Arthur T. Murphy
The Structural Engineer - 3rd Feb. 2004
If a structure were loaded to full design load at all levels and the actual forces that
developed in the members were measured what degree of variation with the forces
assumed at design stage would be found to exist?
If two engineers were independently given the same building frame and design brief would
they both end up designing every element for exactly the same forces?
Basically Engineers have traditionally idealised a very complex problem, applied safety factors,
and established design forces which have been used successfully for many years. This is a good
result, but should not be regarded as some sort of scientifically accurate result.
Now there is a trend towards full 3 dimensional modelling and analysis of buildings. Creating
a single analysis model for an entire building will inevitably have the following effects:
Lateral load is shared throughout the structure according to relative stiffness:
Where robust core walls and shear walls exist these provide the vast majority of the
resistance to these loads.
The sway that occurs when pure gravity load is applied to any non-symmetrical
structure will naturally be introduced.
Storey Height
Fdn to G/F
0.5m
G/F to 1/F
4.0m
1/F to Roof
10/F to Roof
Shear Wall
300
225
Core Wall
350
250
10/F to Roof
C1
C2
C3
500 x 1500
400 x 1500
C4
500 x 1500
400 x 1500
C5
500 x 1500
400 x 1500
C6
500 x 1500
400 x 1500
Fdn to 10/F
10/F to Roof
B1 & B2
400 x750
300 x500
B3 & B4
400 x750
400 x500
B5 & B6
450 x700
300 x 450
B7, B8 & B9
350 x650
250 x 500
Slab Thickness
125mm
Height of Openings
All openings to be 2300mm from floor level including the opening at G/F between columns
C1 and C2.
In a separate exercise it has been demonstrated that sway results for this model are in excellent
agreement with results achieved using other software. Refer to the next chapter, Sway
Deflection Verification for details.
Column and wall forces are established by summing up the axial loads determined in each
of the separate analyses. Or, in many cases the axial loads are determined by simply
calculating an assumed supported floor area at each floor.
Both 2D and 3D subframe models can be created and analysed in Orion if required. To
emulate the result that would be achieved by subframe analyses in both directions the top
storey is extracted and analysed as a single floor model.
The analysis results for this 3D subframe are shown below (mid-pier wall idealisation has been
used).
To enable comparisons to be made with other approaches, the bending moment diagram can
be plotted for the highlighted beams.
We shall consider this our "baseline" answer against which other approaches can be
benchmarked.
Once again it is important to note the comments made in the opening discussion about the
assumptions inherent in the above result. It is a target baseline, it should not be regarded as
some sort of perfectly correct result.
Building Analysis
Analysis of a full 3D model
Since Orion's "FE Chasedown" analysis uses a sub-frame approach, we should expect to find it
agrees closely with the "baseline" answer.
Orion's 'building analysis' does not utilise sub-frames, instead it treats the entire structure as a
single 3D frame in which the columns and walls are only fixed at the foundation level.
We will now consider each of the above analysis methods in turn by comparing them against
our "baseline" answer.
This appears to be in very reasonable agreement with the baseline analysis result shown below.
Although the results are not exactly identical, the differences are generally small and are likely
to be down to internal differences in the way FE beam and wall elements are assumed to
connect.
There is however a limitation to this approach - whilst in building analysis pattern load cases
can be automatically derived and included, they are not considered in the FE chasedown
analysis. Consideration of pattern loads would increase the peak design forces but two other
factors could be taken into account:
Support Moments are taken at the centreline of supporting columns. Introducing rigid
zones to the analysis and taking design moments at the face of supports would result in
changes to design moments which are predominantly reductions where beams frame into
columns.
Some engineers may wish to consider using the FE Chasedown results on the basis that the
factors above have some counterbalancing effect. However, other engineers will prefer to
know that pattern load has been considered and for this reason we tend to suggest that the area
factor adjustment method which is illustrated in the next section may be considered
preferable.
If the analysis is run using the above default settings, the bending moment diagram for the
comparison frame at the top storey is as follows:
To gain a better understanding of why this is happening we need to think about the 3D nature
of the analysis model. To begin with let's display the above diagram in 3D along with the other
beams at the top floor:
This result appears to disagree with our "baseline" answer obtained from the analysis of the top
floor as a discrete model, which is shown below:
If we were to accept the results for the baseline model as being our target result, is there
any way to adjust the building analysis model to bring it back in to line with this target?
Why do the two models give such different answers?
To resolve this we need to compare the two deflection diagrams.
Deflections under dead load for the top floor of the building analysis model:
In the baseline model the vertical deflections at the beam support points have been eliminated.
This is not the case in the building analysis model - vertical deflections occur at the beam
support points. Since stresses are greater in the columns, these deflections are greater at
column support locations than at the wall support locations. The differential in these
deflections is most obvious along the line of the frame on which the 'anomaly' in the bending
moments was detected. (shown in dashed lines above.)
The differential tends to increase as you get higher in the building. Evidence for this can be
seen in the change of shape of the bending moment diagram along the line of the highlighted
frame when plotted for the entire building.
There is an obvious change in the bending moment between the ground and the 9th floor.
There is another change at the 10th floor but this is due to a change in beam sizes. It is perhaps
interesting to see that the effect is so noticeable over the first 9 floors. This is partly because
the case of a beam spanning from a wall to a column and back to a wall is typically the most
critical sort of location for exposing this effect.
Differential axial deformations account for the different answers obtained from the two
models. Traditional sub-frame analysis ignores this effect. Sub-frame analyses assume there
is no vertical deflection at supports, whether it is appropriate to ignore this will be discussed
later in these notes. For the time being we will assume that you want to ignore it, in which case
we now need to determine if it is possible to eliminate it.
How do we eliminate this effect if we want to?
We have already seen that this effect is eliminated in Orion's FE chasedown analysis (because
it is a sub-frame analysis).
In order to eliminate it from Orion's building analysis, we need to reconsider the default
stiffness settings applied to member groups in the analysis model.
There are probably several strategies of adjusting member properties that could be employed
to try and reduce / eliminate DADE.
1. Make column vertical deflections consistent with the wall deflections by increasing the
area factor applied to columns.
2. Eliminate column and wall vertical deflections completely by substantially increasing the
area factor applied to columns and walls.
3. Adjust beam stiffnesses so that the differential vertical deflections induce much lower
bending moments.
4. etc?
After various trials it has been found that the first method (column area factor adjustment)
seems effective without readily introducing unwanted side effects. Therefore, these notes now
focus on this method.
Vertical Deflections under dead load at the top floor for AF = 1.0
If the model is re-analysed with the column AF 2.2 as calculated above, the revised deflected
shape indicates a marked reduction in the differential axial shortening.
Vertical Deflections under dead load at the top floor for AF = 2.2
The above deflection diagram for AF 2.2 shows that some differential deformation still
remains. This is explained by the fact the stiffer columns also attract more load (increasing it
towards the value that the sub-frame analyses would determine). A second iteration of the AF
calculation should reduce it.
LH AF = 4.4/4.9 = 0.9,
RH AF = 4.4/2.9 = 1.5
Average AF = (0.9+ 1.5)/2 = 1.2
Applying a further factor of 1.2 to the existing 2.2 the new AF becomes 1.2x2.2 = 2.65 - use 2.7.
If the model is analysed with the column AF 2.7 the differential axial shortening is almost
eliminated.
Vertical Deflections under dead load at the top floor for AF = 2.7
Results for these analyses can again be compared with the result from the baseline model.
With the AF of 2.2 the BM's developed in beams B1 and B2 at the top floor are as follows:
Applying a column AF of 2.2 reduces the effects of differential axial shortening and eliminates
the sagging at the central column support. When the AF is increased to 2.7 the result is even
closer to that of our baseline subframe.
Using this slightly higher AF value of 2.7 ensures a conservative design at the column support.
The same value would not be applicable. Arriving at this value is a matter of some judgement
but the above demonstrates that a logical process can be applied.
Whether the AF should be increased still higher would be a matter of engineering judgement.
For example if the AF is increased to 4.0, the resulting bending moments are as follows:
With this higher area factor the hogging moment at the column increases further, however it is
reduced at the wall supports.
NOTE: An area factor of 4.0 is likely to be extreme for most buildings. In this example it
has reversed the differential deformations - the walls are now shortening by more than the
columns, as shown below:
Vertical Deflections under dead load at the top floor for AF = 4.0
What is a reasonable upper limit for the column area factor adjustment?
The objective in setting an upper limit is to reduce or largely eliminate the differential axial
deformation so that the design forces at each floor more closely resemble those obtained from
a traditional approach.
Although differential axial shortening is not taken into account in traditional subframe
analysis methods, it will still exist to a greater or lesser extent in real buildings. Even taking
into account the staged nature of construction there will always be some amount of differential
shortening that actually does occur.
The objective of setting the upper limit is therefore to obtain a realistic set of design forces,
without necessarily being concerned about completely eliminating the effects of differential
axial shortening.
Gravity Loads
When the area factor is used to counteract axial shortening, it affects the axial force
distribution as well as the bending moments. This is illustrated in the table below, showing
how the dead load results change at the base of the frame containing beams B1 and B2 as the
area factor is increased:
The axial and bending moment results which accrue at the bottom level of the walls and
column along the frame under consideration can be tabulated as follows:
AF 1.0
AF 2.2
AF 2.7
AF 4.0
FE chasedown
(i.e. subframe)
-3219
-3091
-3065
-3025
-3042
-3233
-3954
-4103
-4342
-4112
-3044
-2862
-2828
-2776
-2889
-453
-439
-436
-432
-198
-0.6
-1.6
-1.8
-2.2
-711
-555
-525
-479
btm (kNm)
In this case the FE Chasedown results represent the result for the baseline model.
As the area factor is increased there is a shift of axial load from the walls into the columns.
This is to be expected. It is moving the results back towards those obtained for the baseline
model. For the bending moments however it seems that this method results in much higher
bending moments in the wall panels than the baseline model. This moment occurs for all AF
levels, it is not introduced by the use of an increased AF. It is likely that this moment is
developing as a result of sway effects that will occur in the full 3D analysis but are excluded
from traditional sub-frame analysis.
Lateral Loads
The deflections at the top of the building under lateral load can also be compared:
AF 1.0
AF 2.2
AF 2.7
AF 4.0
FE chasedown
(i.e. subframe)
101.6
99.0
98.4
97.5
N/A
101.3
98.3
97.7
96.7
N/A
101.8
99.2
98.4
97.7
N/A
79.9
78.4
78.3
78.0
N/A
35.9
35.0
34.8
34.5
N/A
81.7
80.4
80.3
80.1
N/A
It can be seen from the above that an increase in the area factor from 1.0 to the value of 2.7 or
even the extreme value of 4.0 has very little impact on the deflections at the top of the building
due to lateral load. This is because the columns are not significant in resisting sway deflection.
If the deflections were changing then some increased caution in the use of the AF would be
appropriate, this is noted again later.
The axial and bending forces in the columns and walls at the bottom of the model can also be
compared for the wind load cases. The effect that increasing the area factor has on these forces
due to wind in X is illustrated below:
AF 1.0
AF 2.2
AF 2.7
AF 4.0
FE chasedown
(i.e. subframe)
1113
965
937
895
N/A
1824
2160
2223
2321
N/A
-153
-225
-239
-261
N/A
681
652
651
648
N/A
124
121
121
120
N/A
9572
9390
9356
9305
N/A
Since the introduction of the AF had only a small effect on the deflection (comparing with
AF=1.0), it has a similarly small impact on the distribution of forces.
Recommendation
Based on the above it is recommended that where DADE is of concern and there is a desire to
largely eliminate it, the strategy adopted for applying an AF adjustment should be:
1. After an initial analysis using AF = 1, examine the beam bending moments and the
vertical deflections of columns and walls at the top floor of the structure. Also record
sway deflections in both directions.
2. Using the approach demonstrated in this chapter, determine an appropriate AF
adjustment to apply.
3. Re run the analysis and review the revised results at the top floor. If the DADE seems to
have been largely eliminated continue with design of all column and wall members.
4. If the deflections are largely unchanged then you might choose to stop at this point,
however, the discussion below should be considered.
5. If the deflections have changed significantly then it is strongly recommended that a
second analysis run is made with a lower AF and that the columns and walls (at least) are
checked for the results of this revised analysis. Within Orion this is a simple procedure
and the implications of this are illustrated in the next section.
FE Chasedown Analysis
Within Orion there is an automatic batch procedure for achieving this. Despite the name "FE
chasedown", the use of Finite Elements to model the slab is optional and unless there is a
specific reason for doing so it is recommended that these are not used. This method will then
very closely emulate the traditional analysis assumptions of sub-frame analysis for gravity
loads.
A building analysis is still required to determine results for the lateral load cases.
Orion allows you to automatically merge the results of these two analyses to generate the
design combinations for all members.
Cons - No pattern loading and no rigid zones.
Consideration of ACI advice on analysis model properties which allow for the more likely
case that beams are cracked while columns/walls are not would have the same effect as
introducing an AF = 1.5. (ACI 318-08 cl 10.10.4.1 the precise clause number varies in older
versions of the ACI, but the guidance appears consistent.)
It was suggested in that presentation that a lower bound of somewhere between 1.5 and 2.0 is
reasonable. 1.5 would be the more conservative value to use (since it widens the envelope
between the lower and upper values).
What is the impact on the design when both upper and lower-bounds are taken into consideration?
To establish this a comparative study can be undertaken using the test model.
The total weight of beam, column and wall steel can be determined using one of the methods
for emulating the traditional approach. (i.e. the upper-bound solution). The model can then be
reanalysed, having made an allowance for differential axial displacements, (i.e. the
lower-bound solution). The reinforcement provision already established is then checked for
the new design forces and if necessary the provision is increased at any failing locations. The
overall impact can be assessed in terms of the changes to the total steel reinforcement weight.
The results of this exercise are as follows:
Upper Bound
Emulation
Method
AF 2.2
AF 2.7
Design for
Upper-bound
Only
44980
45334 (+0.8%)
80185
80211
(+0.03%)
80712
(+0.7%)
44907
45273
(+0.8%)
46438
(+3.3%)
80192
80261
(+0.09%)
80762
(+0.7%)
Upper Bound
Emulation
Method
AF 4.0
FE Chasedown
Design for
Upper-bound
Only
45774
46528
(+1.6%)
47781
(+4.4%)
80187
80344
(+0.2%)
80845
(+0.8%)
N/A
43847
45658
(+4.1%)
N/A
80545
82180
(+2.0%)
Note: the total beam reinforcement weight quoted doesn't include the weight of 2B10, 3B14
and 3B15 as these proved to be inadequate for the forces involved.
Consider Case where AF = 2.7 is upper bound and AF = 1.5 is the lower bound:
In this case the implications of considering the envelope are quite minimal. A 0.8% increase in
beam reinforcement and a 0.01% increase in column and wall reinforcement is introduced.
Consider Case where AF = 4.0 is upper bound and AF = 1.0 is the lower bound:
In this case the envelope being considered includes extremes that have been suggested to be
the most conservative at both ends. The result is that a 4.4% increase in beam reinforcement
and a 0.8% increase in column and wall reinforcement is introduced.
Consider Case where FE Chasedown is upper bound and AF = 1.5 is the lower bound:
FE chasedown only provides results for the gravity loadcase, lateral load results must come
from the full 3D building analysis. In this case the least overall reinforcement weight is
determined. This is likely to be because sway effects introduced by gravity load get eliminated
(as they are in a traditional sub-frame analysis).
The reinforcement weights when both pattern load and rigid zones are considered are as
follows:
..
Upper Bound
Emulation
Method
AF 2.7
AF 4.0
Design for
Upper-bound
Only
40437
40964
(+0.8%)
42295
(+3.3%)
78778
78891
(+0.1%)
79517
(+0.9%)
40951
41568
(+1.6%)
42839
(+4.6%)
78851
78980
(+0.2%)
79606
(+1.0%)
Compare the total weight requirement after designing for upper bound of 2.7 and lower bound
of 1.5. For the beams the weight is now 40964kg, it was 45273kg before rigid zones were
introduced. For the columns the weight is now 78891kg, it was 80261kg before rigid zones
were introduced. The new reinforcement weights are also slightly lower than the weight
previously shown when the FE chasedown method was used (43847 and 80545 respectively).
For this example model, the AF method can be used with double benefit when compared to FE
chasedown:
1. It eliminates any concerns about not having considered pattern loads
2. By introducing the efficiency of rigid zones it also results in the least overall reinforcement
requirement.
Re run the analysis and review the revised results at the top floor. If the DADE seems
to have been largely eliminated continue with design of all column and wall members.
You will however have designed for larger moments in some places, so the steel
may appear to be overprovided at those points.
The above is illustrated in a DADE Analysis & Design Flowchart at the end of this chapter
together with the possible option of using the FE chasedown procedure.
Closing Discussion
Typical Concerns
The sorts of concerns expressed by engineers as they begin to consider this topic for the first
time and then in greater detail are:
1. The suggested area factor is not a defined value, it is just a suggested range of values. I
really want a fixed value or values that can always be used.
2. Making these adjustments to cater for gravity load may be logical but are there side
effects? For example, will it cause completely different force distribution and building
deflection under wind loading cases?
3. Does the FE chasedown option provide another (better?) way to eliminate differential
axial deformation?
4. If I am designing for a wider envelope of forces am I not over-designing the structure?
5. Can this all be solved more accurately by using Staged Construction analysis?
For the upper limit we are considering the case where differential axial deformation is being
largely eliminated. Values of 3-5 are discussed in the presentation but this is actually
something that can be quite structure dependent. However, it is also something where logic
can be applied. If the deflections from an un-modified analysis are examined and a wall is
deflecting down by 2mm while the columns adjacent to it are deflecting 5.5mm then the area
factor required to eliminate this will be a little in excess of 5.5/2 = 2.75. (Bear in mind that the
stiffer columns will attract increased axial load).
In the example the relatively large size of the columns compared to the walls results in quite a
low figure of around 2.7 being a reasonable upper limit, but use of values slightly higher than
this would really only add a little to overall design conservatism.
For the lower limit we are considering the case where differential axial deformation is allowed
for. As is noted above a logical but very conservative lower limit is 1.0. For reasons detailed in
the presentation we suggest that a 1.5 value is actually a very reasonable lower limit.
Generally walls are providing the majority of lateral load resistance and so adjusting
column properties has quite limited impact on deflections.
Once again the impact of this is illustrated in the example. The percentage change in
deflections is relatively small.
A bigger question (not considered here) is the extent to which any analysis accurately predicts
lateral deflection. The result is again entirely dependent on allowances made for creep and
cracking. ACI guidance on this is more thorough and suggests varying member stiffness
adjustments.
Could I avoid all this complication if Staged Construction Analysis were used?
This would not avoid complication - it would just mean that different complications need to
be addressed. A single staged construction analysis can not deliver "correct" answers. The
unknowns associated with the input for such an analysis dictate this. This is illustrated within
our presentation, but the key point to consider is that this type of analysis means you have to
define properties that would not normally be defined for a basic elastic-static analysis, and you
would have to allow for uncertainties associated with these:
Time varying properties of concrete - you have to be able to define what properties should
be used at every time step (stage). This includes consideration of early age creep effects.
You have to define the way in which loading becomes active through the stages.
You must then layer on top of this the need to consider patterning of load. In this context
there is load patterning along beam lines but also the need to consider loaded and unloaded
floors through the height of the building. Can this be realistically achieved as part of a staged
construction analysis of a large structure?
As with any analysis, if it is not used with a reasonable degree of care and understanding then
a staged construction analysis can be a sophisticated and time consuming way of getting bad
answers.
We would suggest that a staged construction analysis is another way of establishing an
envelope of possibilities. The parameters applied during this process are not specified in any
design code so this approach ought to be subject to a similar degree of judgment and scrutiny
as is required for the area factor adjustment approach.
Which
method to resolve
DADE?
START
01.
FE1.
AF1.
yes
no
AF
FE2.
Determine upperbound AF
02
Are Rigid
Zones set to
None?
FE
Yield Lines or
FE decomposition for Beam Loads
AF2.
03.
Decide on a lowerbound AF
Conservatively use the default AF = 1.0
To reduce DADE, use 1.5 2.0
FE3.
FE4.
AF3.
04.
05.
AF4.
FE6.
sizes
not OK
06.
AF5.
Initial Design
Check
FE7.
sizes OK
AF6.
07.
Is DADE a
concern?
no
yes
FE8.
08.
FINISH
Flowchart Notes.
Box
No.
Comment
03.
Justification for increasing the AF may be derived from ACI advice on analysis
model properties which allows for the more likely case that beams are cracked
while columns/walls are not. Refer to What is a reasonable lower limit for the
area factor adjustment? for an introduction to further discussion on this.
04.
06.
07.
how the beam bending moments change over the height of the structure,
the vertical deflections of vertical elements at the top floor
see Why do the two models give such different answers? for details of how to
do this.
08.
09.
AF method - Applying a single area factor globally may not fully remove
differential axial deformation at all locations.
See Increasing the column area factor for details of how to do this.
AF2.
The deflections at the top of the building under lateral load should be
compared. Provided the differences are small the impact on the distribution of
lateral forces will be small. See Overall Summary of Suggested Procedure for
additional notes on this. If the high AF is affecting the deflections it is more
important that the lower bound solution is checked for all members.
Chapter 14
Introduction
As described in the previous chapter a 'traditional' design will historically have been based on
the results of a series of discrete 2D subframe analyses, in which:
Sub-frame analyses are carried out at each floor from top down (assuming fixed supports)
Worked Example:
In the previous chapter we established some traditional sub-frame results
for beams at the 20th floor of a building and then looked at how closely
these results were emulated using full 3D analysis., the purpose of the
exercise being to quantify the effects of differential axial deformations.
From the results it was possible to then formulate recommendations for
emulating a traditional design.
In this worked example the same model is used to examine 3D staged
construction analysis and compare the results with those obtained
previously. Once again the objective is to derive a means of emulating a
traditional design.
The unstaged G+Q*F result (with 25% rigid arms and AF = 2.2) is:
LH Support
Moment
Span Moment
RH Support
Moment
FE Chasedown
-100.4
67.4
-103.8
BA unstaged
-189
70
-16
BA AF 2.2
-126.2
62.9
-84.4
BA staged
-115.8
62.4
-95
The staged result initially appears to offer an as good or better means of eliminating DADE
than is achieved through the Area Factor approach, (although comparison is difficult as the
staged and AF results in the above table allow for rigid arms whereas the FE result does not).
However, when the staged results are examined in more detail the conclusion is less clear cut.
Each bending moment diagram shows a significant step change at the 10th floor - this can be
accounted for by the step change in beam sizes that takes place at this level.
DADE is clearly affecting the unstaged analysis results from the second floor all the way
through to the top floor of the building.
The staged analysis seems to be largely unaffected by DADE at the upper floor levels, but this
does not hold true from around the 10th floor level and below. (In this particular worked
example it should be noted that the effect is being amplified due to the central column size
being larger over the first 10 storeys.)
The shape of the above bending moment diagrams can be understood more clearly by
considering the extent by which the central column shortens in each analysis.
In the unstaged analysis the column axial shortening must accumulate as you get higher in the
building. The vertical deflection of the column at the top floor is directly affected by the axial
shortening of the lower columns, which is occuring due to the load they each carry.
Unstaged Analysis Deflections
1st Floor
10th Floor
Top Floor
In the staged analysis, the top floor forces and deflections are determined from the analysis of
the final stage alone, in this stage only the top floor loads are applied and the loads carried by
other floors have no influence - hence column shortening is not being accumulated from one
floor to the next.
Staged Analysis Deflections
1st Floor
10th Floor
Top Floor
So in the staged analysis, why are we seeing DADE occurring, particularly at the 10th floor
and below? - To answer this you need to consider how the forces and deflections at the 10th
floor are determined from the combined effect of the analyses of 10th stage to the final stage,
(consisting of all the loads from the 10th floor to the top floor). This significant amount of
load is being carried by a column which is 10 storeys high - there is undoubtably going to be
considerable axial shortening taking place.
Lower in the building, at the 1st floor level the column is carrying a much greater load however, because it is so short the amount of axial dispacement would be minimal.
Conclusions
Staged Construction does not eliminate DADE. (this particular example exacerbates the effect
because of the change in sections - the effect may not be so evident in other models??).
It moves the position where they are most significant (to building mid height ) and reduces
them to some degree.
This result is not technically wrong , but is it what engineers want?
If you are happy to completely ignore differential axial displacements then all you need to
do is adopt one of the methods described previously for emulating the traditional design,
taking note of the associated pros and cons.
If you are happy to take some account of DADE then staged construction may offer the
best solution.
Chapter 15
Introduction
The following section of the Engineers Handbook contains six chapters relating to the design
and detailing of the various member types:
Slab Design looks at slab strip errors and the actions required should you encounter
them.
Beam Design to BS8110 looks at Orions beam design process, relating it to the relevant
code clauses. A worked example is included.
Beam Detailing provides an overview of the beam detailing process, describing the
available patterns and beam detailing settings. A worked example is used to illustrate
how careful application of preferences can produce either: a minimum weight solution
(where the details may be more complex), or a more standardised detail (with a slightly
higher weight of reinforcement).
Column Design to BS8110 looks at Orions column/wall design process, relating it to the
relevant code clauses. Worked examples are included.
Wall Design and Detailing looks at additional aspects of Orions design process that are
unique to walls.
Foundation Design describes the procedures required for defining and designing the
various foundation elements available in Orion.
Chapter 16
Slab Design
Overview
Referring back to the Modelling Analysis and Design Flowchart in the initial Overview
chapter, you can see that slab design may be based either on tabulated code coefficients, or on
finite element analysis (FE) results.
Design based on the tabulated code coefficients is covered in the training manual. Discussion
there also extends to the option of using an FE strip, but design based on FE is covered in
greater depth in this manual in the chapters Analysis and Design using FE and Flat Slab
Models.
In both cases, the design approach involves creating slab strips. Checks are applied as you
create these strips to ensure they are valid. The following brief notes may be of some additional
assistance (assuming that you have already completed the training examples successfully).
Action Required:
For case 1 above, change the end condition to Cantilever and update the strip.
For case 2, split the strip into two half strips. Draw one strip ending in the open area and then
a second strip coming out of the open area.
For case 3, redefine the slab panels so that the strip passes from beam to slab panel to beam etc.
Creating Member (but nothing seems to happen)
This occurs if your strip doesnt entirely traverse one entire slab.
Action Required:
Interrupt the process by clicking on to another command on the Members toolbar, then
redraw the strip but increase its length so that it entirely traverses at least one entire slab.
Chapter 17
Introduction
Orion designs beams bent about the major axis using the code clauses given in BS 8110-1:1997:
Part 1 Section 3.4.
The following table summarises the various stages of the beam design process:
Step
Calculation
Clause
1.
3.4.1.6
2.
3.4.1.5
3.
Analysis of Sections
3.4.4.1
4.
3.4.4.4,
3.4.4.5
Calculate K/K
3.4.4.4
Calculate z
3.4.4.4
Calculate x
3.4.4.4
Calculate As required
3.4.4.4
3.4.4.4
5.
3.4.5
3.4.5.2
3.4.5.4
3.4.5.3
Spacing of links
3.4.5.5
Deflection of Beams
3.4.6
3.4.6.3
3.4.6.5
3.4.6.6
For further details of the implications of adjusting the various settings see the chapter Beam
Detailing
The option to use the rectangular section (rather than the flanged section) when the flange is
in compression will generally result in slightly more conservative steel area requirements,
however if minimum steel requirements apply then design using the flanged section will
introduce greater steel requirements.
If a flanged section is used the effective width is determined from Cl 3.4.1.5.
The design moments for each of the six regions can be displayed by clicking the Design button
shown below.:
The effective depth h to the centre of the bars is worked out separately at each location. This is
automatically reduced if the steel doesnt fit in a single layer.
You can control this calculation via the Design tab of Beam Design Settings as shown below:
Centre of Gravity of Steel Bars the section effective depth will be determined based on
the centre of gravity of the tension bars. In this method, the contribution of each tension
bar in the section will be considered separately.
Centre of Gravity of Layers in this more conservative method the calculation of section
effective depth is based on the average of the distances of the steel bar layers.
Values for K, K, x, As, and As are then all calculated for each region and are also shown in the
same table.
Where the ratio of K/K is less than 1.0 no compression steel is required.
From the Design tab of Beam Design Settings you can specify if you want to design for this
value, or (if the conditions of Cl. 3.4.5.10 are met) you may choose to design for the shear at d
from the column face.
If the latter option is adopted the design shear force at d is then displayed and used in the
design calculations as shown below:
In the above table, the maximum shear stress v is determined from Cl 3.4.5.2 and is always
based on Vd (not Vd@d). This is cross checked against the maximum permissible shear stress
v-max, which is defined in the beam design settings.
vc is calculated using the equation in table 3.8. In BS8110 the concrete grade considered for
shear is capped at 40N/mm2.
Link requirements are established from Table 3.7 as follows - v is determined from either Vd or
Vd@d depending on your choice in the beam design settings and Vnom Nominal shear
capacity (based on minimum link provision) is established.
The maximum shear requirement is then checked at each end of the beam and if the nominal
capacity is not sufficient additional links are provided over a distance (x) extending to the
point where the nominal capacity is adequate (or beyond if user preferences dictate minimum
lengths).
The beam size is not automatically increased to satisfy the deflection criteria - it
is highly unusual for beams to be sized so that deflection is a controlling factor.
Also, it should be noted that increasing the steel is usually a highly uneconomic
way of controlling deflection. It is more normal to conclude that the beam is not
deep enough if beam deflection checks fail.
Worked Example
The Design Model
The example model Doc_Example_4 is opened and saved to a new name (so as not to destroy
the original example). In the calculations that follow, the steel grade has been set to 500 with a
steel material factor of 1.15. Bar diameters permitted for the longitudinal steel in the beams
have been limited to 10, 12, 16, 20 and 25mm. The minimum link diameter is set to 10mm.
The concrete grade is C40. Note that to obtain similar results you will need to apply the same
settings. However, the actual design process would be identical irrespective of which material
grades, material factors and bar diameters are used.
The continuous beam on grid line 2 will be designed and the following calculations will focus
on the design of beam 1B14.
Beam 1B14
span = 4m
beam depth, h = 400mm
beam width, b = 250mm
cover = 20mm
Note
Where the cover shown above has been set to 0 - as noted on the dialog, the cover
is then based on code requirements - (i.e. a default cover of 20mm is applied). You
may prefer to enter a value directly into each field to set the cover required.
Analysis Results
Having run a building analysis using default analysis settings, the design envelope for all
combinations is as follows:
The area of steel calculations are listed below the design moments. These can be confirmed
with simple hand calculations using Cl 3.4.4.4. However, before this can be done we need to
know the bar sizes used to accurately calculate the effective depth.
Click the Steel Bars button to see the steel bars provided:
Similar calculations to those shown above are performed for the other four
regions.
Span
vc is calculated using the equation in table 3.8
d = 365mm, As = 235.6mm2
100As/b.d = 0.258
vc = (40/25)0.33 0.79.(0.258)0.33 (400/365)0.25 /1.25 = 0.48
Link requirements
Determine if minimum bars at maximum spacing provide minimum links
Asv_req / sv = (0.4.bv.)/0.87fyv = 0.23
For minimum bars at maximum spacing, provide H10 links at 250mm centres
Asv_prov / sv = 2 x 78.5 / 250 = 0.63
In this case OK - (if not OK decrease the spacing until requirement is met).
Calculate vnom using the minimum links
capacity from vc
vc b.d = 43.8kN
capacity from links
(0.87 fyv Asv_prov / (bv.sv)). (bv.d) = 99.8N
vnom = 43.8 + 99.7 = 143.6 kN
Determine if shear at the ends exceeds vnom
40.7kN < 143.6kN
In this case OK - hence minimum links apply for whole length of beam.
If the shear at either end had exceeded vnom - the distance, x to the point where vnom is
adequate would be determined as shown below, and a calculation performed to
determine the increased links to apply over that length.
Output Calculations
The following report shows the output for the beams along axis 2:
Chapter 18
Beam Detailing
Introduction
Orion has always included numerous beam design and detailing options that are all controlled
from the main beam Settings and Parameters dialog that is shown below.
Given that there are tabs with sub-tabs and numerous options under each, the control of these
settings can be very daunting. With each Orion release even more options tend to be added in
order to respond to user demand to be able to apply more and more specific preferences.
The objective of this chapter is to provide a brief overview of the options in general and then
more specifically the reinforcement pattern options, the first of which is highlighted above.
5. The bars are arranged and curtailed so that the reinforcement provided is sufficient at all
cross sections.
6. The information is conveyed on a detail drawing.
This may initially seem simple enough but it soon becomes apparent that any two engineers or
draughtsmen are quite likely to apply different preferences (personal preferences or company
standards) and get different results at every stage of the process.
The analysis stage is not the focus for this document but it is worth noting that Orion provides
many analysis options, which will mean the analysis results differ, for example:
Option to use rigid zones (not only generates a more complex analysis model but also
means that design moments at the face rather than the centre-line of the column may be
used).
These settings are generally self evident, they will tend to have a slight influence on the values
of As required that emerge from the design. For example the options to design for the shear at
the column face and to use the rectangular section (rather than the flanged section) when the
flange is in compression will result in slightly more conservative steel area requirements.
Again, these settings are generally self evident, they set limits on the ranges and spacing of bars
which are considered when bars are being selected to provide reinforcement which at least
meets the minimum requirements determined during design.
In this tab we start to apply more specific preferences which will affect the way in which bars
are selected to meet the As requirements determined in design.
Standard Pattern 2 shown above is currently the most commonly used option. Many of the
other options under this tab and also under the curtailments tab are more tuned to standard
pattern 2. Later in this chapter we will look at each of the patterns in turn and at some of the
other options (particularly relating to this tab and the following curtailment tab) in more
detail.
In this tab we apply preferences as to how the reinforcement is curtailed. Although this is not
under the Detailing tab, these sorts of preferences are more traditionally applied by the detailer
than the designer.
In this tab all the preferences relate to detailing presentation options, i.e. changes here only
relate to presentation and not to the reinforcement selection.
Settings in this tab control the layering, line types etc. to be used in the DXF file, which can be
loaded into most general drafting packages.
Overview of Patterns
The concept of different basic patterns, which can then be optimised according to personal
preference, is actually derived from the publication Standard Method of Detailing Concrete
published by the Institution of Structural Engineers (UK) August 1999. In the section on
beam detailing (cl 5.4.20), two basic methods of detailing are introduced:
The Splice Bar Method A pattern aimed at prefabrication of cages and minimisation of
in situ steel fixing.
The pattern is shown diagrammatically in the dialog above and in the typical detail below. The
main attribute of this pattern is that top and bottom bars all stop at each side of columns/
supports and then splice bars are used to maintain the required continuity through the
support.
All sorts of preferences can be applied to the pattern, generally this becomes a balance between
complexity and material efficiency, this concept is explored in detail later in this chapter.
Some general points worth noting at this stage are:
There are various options to merge bars through supports and between spans. If the
objective of using this pattern is to create beam details such that discrete reinforcing cages
for each beam can be constructed then lifted into position the merging options should all
be switched off.
Currently the pattern does not support top span bars that extend to the face of the support
(link hanger bars), the top span bars will always stop short of the support. This may mean
that the links towards each end of the cage are not fixed as securely as might be desired
during lifting. Some extra (un-detailed) top bars may be required.
In general it appears that this pattern has not yet been widely employed by Orion users, or if
used it has generated very little feedback. However since the pattern does encourage off-site or
low-level prefabrication it offers the possibility of reducing site risks by reducing work at
height. There may even be financial benefits to be secured from factory production of cages.
It is considered possible that the use of this pattern may become more dominant in the future.
The pattern is shown diagrammatically in the dialog above and in the typical detail below.
Currently the top bars are the same for this pattern and pattern 1. The bottom bars are
different, the main difference being that they are allowed to run through the support and lap
directly with the bars in the adjacent span.
All sorts of preferences can be applied to the pattern, generally this becomes a balance between
complexity and material efficiency, this concept is explored in detail later in this chapter.
Some general points worth noting at this stage are:
This pattern is the mostly commonly used and so many of the options/preferences
introduced by user request tend to be more applicable to this case. The Detailed Example
and Comparisons are therefore most applicable to this pattern.
There are various options to merge bars through supports and between spans. Since this
pattern is not intended to lend itself to prefabrication these options are more applicable.
The pattern is shown diagrammatically in the dialog above and in the typical detail below.
Currently the bottom bars are the same for this pattern and pattern 2. The top bars are
different, the main difference being that (link) hanger bars are used which can run through
several spans. These bars provide any required span moment resistance. Additional top
reinforcement is then added in as necessary through the supports.
Note that the detailing option to Re-Plot the Bars below the Beam has been used here to show
how top hanger bars extend through span 1 and into the middle of span 2.
All sorts of preferences can be applied to the pattern, generally this becomes a balance between
complexity and material efficiency, this concept is explored in detail later in this chapter.
Some general points worth noting at this stage are:
Since small bars extend right through the support where additional bars are also added,
this pattern may tend to produce cluttered details at the support or even prove to be
impractical from a spacing point of view.
As with pattern 1, it appears that this pattern has not yet been widely employed by Orion users,
or if used it has generated very little feedback.
This pattern is used in high seismic regions and is only shown here for completeness. The
current version of Orion only supports the Turkish earthquake code (in the Turkish language
version of Orion). In non-seismic regions use of this pattern is probably acceptable and is
likely to produce a very robust structure, probably at the expense of both reinforcement weight
and on-site construction complexity.
For now it is recommended that Orion users in non-seismic regions do not use this pattern. In
future releases when other international seismic design codes begin to be supported this
pattern will be more important.
The procedure shown in this section can be applied to any Orion model (or more
usually one floor of a larger model) in a matter of minutes or perhaps hours
(depending on model size).
Basic Setup
First we will get the design and detailing set up so that we have the base data on which further
comparisons can be made. The floor we will look at is shown on the next page.
It is a reasonably simple layout of nearly 50 beams with typical spans in the region of 5 to 7 m.
The applied factored load is approximately 22 kN/m2.
The following subsections show the basic settings used for the initial design.
Design Tab
The above settings are reasonably standard and will not be changed during any of the
subsequent comparisons.
Parameters Tab
In general the above settings are reasonably standard, but as will be shown in the subsequent
comparisons, adjusting or restricting the range of bars sizes to be used can influence the
efficiency of the design.
On the Method sub-tab shown above the bar size minimisation option is selected. This is
generally not the preferred option, but as will be shown it will actually lead to the minimum
weight of steel being determined.
Note that the steel bar area tolerance can be used to introduce an additional safety factor
during preliminary design.
Pattern 2 is selected and the options to use extra bars locally to enhance the support and span
moment capacities are both activated.
The options that are not checked add constraints that will add reinforcement weight.
The options in the Bent-up Bars tab are only relevant to the Bent-up Pattern.
On the Steel Bars sub-tab shown above the option to lap the bottom bars outside the support is
used. This seems to be the preferred option of most Orion users as it tends to reduce clutter
within the column. Although merging will reduce reinforcement weight slightly these options
are not activated yet, we will look at their effect later. The other options that are not checked
add constraints that will add reinforcement weight.
On the Steel Bars 2 sub-tab shown above the options are not checked once again because
adding these constraints will add reinforcement weight.
Note that the minimum Tension Lap Factor. This factor works as follows:
If you set the factor to 1.0, but for the lap in question the code minimum is 1.4, then we
still use 1.4.
If you set the factor to 1.4, but for the lap in question the code minimum is 1.0, then your
user over-ride applies - 1.4 gets used.
Options under the Bob Control sub tab shown above are self-explanatory.
Note the setting for Min. Steel Bob Length shown above is 25 Dia. The actual minimum
specified in BS 8666:2005 Table 2 varies depending on the bar diameter. For 10mm bars it is 12
Dia, for larger bar sizes it reduces further to around 10 Dia. (Potentially then you might
consider reducing this for greater ecomony.)
The Min Steel Bob Length is compared with the bob length that would be required to achieve
the anchorage length, (which for design to BS8110 is taken from Table 3.27). Note that the
anchorage length is measured from the inside face of the supporting member.
The settings shown above are not changed within the comparisons in the next section.
Detailing Tab
As noted earlier, the options under this tab only affect the style of the display and not the
content. These options will not be changed at all during the comparisons in the next section.
The settings used for the comparisons are shown here with some brief discussion. In general
these might be considered a good set of suggested settings, but it is best to experiment a little
with all of these until you have optimised the detailing style to suit your own preferences.
On the General sub-tab shown above note the option to include bar marks. If you want to see
the bar mark appear in the bar label in the usual style you must also ensure that the settings in
the Graphical Editor Settings dialog are set as shown below.
In the Beams sub-tab shown below we are using the option to refer to beams by the axes
(grids) that they lie on and between.
Note
It is probably more usual to activate the option to put the beam label below the
detail, but this option is automatically deactivated when one of the other options
we will use is activated see later.
In the Sections sub-tab shown above we are choosing to see 3 sections on every beam, this can
be cut down. We have also chosen to use the axis (grid) label as a prefix to every section and
so we can restart the section numbers on every axis and still have a unique label on every
section. Since the axis labels are also used to reference the beams everything can be easily
cross-referenced.
Note
If you choose to reference the beams using the beam labels (the numbers generated
as each beam is added to the model), it would be better to add the beam label as a
prefix to the section.
In the Steel Bars and Links sub-tab shown above we have chosen to Re-plot the bars below the
beam. This is an extra option that goes beyond normal detailing practice, it is not suggested in
detailing guides perhaps because the extra drawing work it would generate would not be
justified when the drawings are prepared by hand. In Orion you get this extra/alternative
presentation of the reinforcement detail for no extra effort see below.
At first glance the above does not look standard and perhaps this puts some people off.
However the exploded reinforcement details provide extra (and clearer) information. We
would suggest that many Orion users might seriously consider using this option as a new way
of providing better information to the site. Alternatively, as we will see in the following
section, Orion provides a schedule that gives the overall weight of reinforcement, but it is not a
traditional bar bending schedule. If a traditional schedule is to be prepared the details need to
be taken off the drawings that Orion prepares, the detailer who prepares the schedule might be
able to do this more easily using the exploded view of the reinforcement.
The options under the Side Bars and Dimensioning (shown above) sub-tabs are all self-evident
and will not be changed during the following comparisons.
Using the filter options to filter to the floor of interest, a design is then run of all the beams in
the floor. During this exercise care is taken to always reselect all bars during redesign.
Next a beam detail sheet is created for all the beams (i.e. all the beams on this floor) and the
option to include a quantity table is checked.
Clearly this sheet is not intended for printing, it is a quick way to get a lot of details into one
DXF file which a detailer can then organise onto several drawings for final editing and then
issue.
Notice that at the top corner of the sheet there is a schedule that can be seen in partial close up
below.
At the top there is a summary of reinforcement used, below that the data is broken down for
each bar position (each bar mark) in two columns. For this initial design of all the beams on
this floor the key points to be noted are:
Total Reinforcement Weight 3246 kg
As noted earlier, the option to minimise bar sizes is not often used since this results in lots of
small bars being used at close centres rather than a few larger bars at wider spacing. If we
change to the more usual bar spacing maximisation method as shown above the weight of steel
changes as shown below.
The weight of reinforcement increases to 3467 kg but the number of bar marks reduces to 121.
Also note how the weight of H25 has gone up from 505 to 1130 kg while the H12 have
dropped from 930 to 285 kg.
In this case the weight goes up because the As provided (and hence weight used) changes in
bigger steps. For example if As required is 510 mm 2 then the size minimisation method will
pick 5 H12 (As = 566 mm 2 ), while the spacing maximisation method might pick 2 H20 (As =
628 mm 2 ).
P
We would emphasise again that the size minimisation method is not generally used and so in
the closing summary we have not used it as the base line figure for percentage comparisons.
This option is probably under utilised. In this example we will squeeze the maximum main
bar diameter which can be used down to H20 from H25. Basically this means that in
situations where 2 H25s were chosen we will now get 3 H20, etc. In addition there will be
some saving due to shorter laps.
The weight of reinforcement decreases to 3324 kg and the number of bar marks decreases by 1
to 120.
Note
If you try to restrict the maximum bar size too much lots of beams will start to
appear as failed on the beam design summary see below.
The beams have a utilisation ratio of less that 1 but are being shown as failing because the
smaller bars (maximum size H16) at closer spacing are failing checks on minimum spacing. If
the failing beams are filtered out and redesigned with the maximum bar size increased to H20
then in this model the weight of reinforcement can actually be optimised to 3234 kg but at the
expense of going back up to 136 bar marks.
Ideally you probably want to use a range where the vast majority of beams pass with one
setting and then the remainder can be filtered out and will pass in a second run with a
different setting. The degree to which this would be necessary would be quite structure
dependent.
Merging Bars
This option is widely used. Where bar groups of the same size meet they will be merged into
one bar group in preference to lapping (provided the total bar length is less that the specified
maximum bar length in the parameters tab). Therefore this option will reduce weight due to
savings in laps but will result in the handling and fixing of longer bars on site.
In this example the weight of reinforcement decreases to 3272 kg and the number of bar marks
decreases to 109. As we continue to simplify the detailing, later in the exercise this option will
start to have a greater effect.
The weight of reinforcement increases to 3348 kg - the number of bar marks remains the same
at 109.
Note the shorter top support bars and the extra bottom span bars laid in to the central section
of the span. If the options are now deactivated as shown below, the detailing becomes simpler.
Notice how only one group of top support bars is used and that in the span all the bars extend
full length of the span.
The weight of reinforcement increases to 3614 kg and the number of bar marks drops to 86.
In the bar curtailment tab above all the options to extend the top and bottom span bars right in
to the support have been activated. The options not to use any extra (shorter) bars locally at
the end support are also activated. The view below shows the simplified detail that results in
an end span.
Comparing the view above with the view below (that was obtained when the end support
curtailment options were all unchecked):
In the top view the H20 L bars in the top of the beam are extended and the short H16 span
bars disappear.
the bars in the bottom of the beam are unchanged - this option will not really have any
effect on the bottom bars unless there are significant wind/sway moments that cause more
steel to be required at the support than in the span.
The weight of reinforcement drops slightly to 3608 kg and the number of bar marks drops to
82.
In the bar curtailment tab above the option to extend the support bars symmetrically is
activated. In the view below you can see how support bars are extended to a proportion of the
span length (default 0.3 L may be adjusted to 0.25 L in line with BS8110 simplified detailing
rules the top span bars will by default be designed for the maximum hogging moment in the
central section of the beam see the design tab settings.)
When the bars are extended symmetrically the detail is arguably simpler (as shown below)
with less potential for misplacement when fixing on site. However it is worth noting that this
option will become impractical when the adjacent span lengths vary dramatically it can
mean that the top support bars start running right through the short span.
The weight of reinforcement increases to 3640 kg and the number of bar marks remains
unchanged at 82.
Standardise Link Size
General Note When Orion is asked to Design Link Spacing at Supports it will check the span
region and each support region. If it finds that the reinforcement required in either support
region is the same as in the span region those two regions will be merged. The example below
shows this case.
In the links tab above two additional options have been activated:
Select Symmetrical Links for Support Regions This option means that when support
regions are required at both ends the bar size (but not the same spacing) used at each end
will be the same.
Same Bar Size at Support and Span This option means that the same bar size (but not the
same spacing) will be used throughout the span of each beam.
In the view above you can see how the second of these settings has forced the use of H10 links
throughout this beam, but since the maximum link spacing is 300 mm this just means extra
reinforcement weight.
In the particular model being used here H8 bars were used almost everywhere, so the above
two settings make almost no difference to the total steel quantity.
The weight of reinforcement increases to 3644 kg and the number of bar marks is unchanged
at 82. On a different project it is quite likely that applying these settings will have a bigger
effect on reinforcement weight.
Uniform Links
In the view above you can see how one link size and spacing is now used throughout this span.
Applying this preference to this model causes a reasonably significant increase in
reinforcement weight.
The weight of reinforcement increases to 3737 kg and the number of bar marks is still
unchanged at 82.
Summary
In the detailed example in the previous section we have successively applied more and more
bar selection and curtailment settings preferences and recorded how these changes affect the
total weight of reinforcement in an entire floor. The following table summarises the findings.
Cumulative Adjustments
No of Bar
Marks
Used
Total
Weight
kg
Percentage Weight
Added
134
3246
-6%
121
3467
Base Line
120
3324
-4%
Merging Bars
109
3272
-6%
Standardise Laps
109
3348
-3%
86
3614
+4%
82
3608
+4%
82
3640
+5%
82
3644
+5%
82
3737
+8%
We would suggest the main points to be drawn from this example are:
1. Most Orion users will use the Bar Spacing Maximisation Method hence the use of the 3467
kg weight as the base line in the above table.
2. Most users will probably not have examined the possibilities of limiting the preferred
range of bars used in which case potential efficiencies may have been missed (in this case a
saving of 4% in reinforcement weight).
3. If the weight after Limiting the Bar Range (3324 kg) were used as the base line figure then
the net effect of introducing all the standardisation/simplification up to and including the
option to Standardise Link Size is:
The number of bar marks reduces from 120 to 82 (more than 30% reduction)
Based on this example, the most important message is therefore the need to examine the
effects of reducing the maximum size of bar to be used in the design.
Finally, it is worth re-emphasising that all of these sorts of judgements have traditionally been
made on something of a gut feel basis during the course of design and detailing. The
automated procedures that Orion provides allow you to investigate options, quantify effects,
and optimise the design for any project in a matter of minutes or perhaps an hour.
Chapter 19
Introduction
By default, Orion designs columns bent about a single axis, or bent about both axes using the
code clauses given in BS 8110-1:1997: Part 1 Section 3.8.
The following table summarises the various stages of the BS8110 column design process:
Step
Calculation
Clause
1.
Braced or unbraced?
3.8.1.5
2.
3.8.1.6
3.
Check slenderness
3.8.1.3
4.
3.8.1.3
5.
3.8.3.1
6.
3.8.2.4
7a.
3.8.3.2
7b.
3.8.3.7
3.8.4.5
Member design
3.8.4
For poly-line columns, as shown below, the bi-axial design method will always be adopted. .
The choice of design method is set via the column design settings dialog shown below..
For walls, the braced/unbraced status is applied in the same way as it is for columns. However,
it should be noted that walls can always be considered as braced along their major axis (i.e dir
1).
Mx + My
If the minimum area of steel is satisfactory for every combination, the program
will record combination 1 as being critical, ( irrespective of the relative
magnitude of loads in each combiation) .
Fixed bar layout The bar locations are defined by the user and the program
determines the bar size required
Bar Spacing Maximisation The program determines the bar size and spacing with the
aim to maximise the spacing. This is normally the preferred option.
Bar Size Minimisation The program determines the bar size and spacing with the aim
to minimise the bar size.
The maximum axial load is checked against Cl 3.8.4.3 or Cl 3.8.4.4. The program defaults to
the more conservative capacity determined by Cl 3.8.4.4. The clause used can be changed via
the BS8110 tab of the column design settings as shown:
Worked Examples
The Design Model
The example model Doc_Example_4 is opened and saved to a new name (so as not to destroy
the original example). The copied model is then adjusted so that its storey height is increased
to 5.5m and it is then re-analysed. In this model the steel grade is 460 and the steel material
factor is 1.15. Bar diameters of 13mm are used and 12mm bar diameters are excluded. The
minimum link diameter is set to 10mm. Note however, the actual design process would be
identical irrespective of which steel grade and material factor and bar diameters are used.
Design Parameters
As shown on the design screen above, if only 3 bars are placed in the x direction
the default clear bar spacing limit of 200mm (as specified in the Column Design
Settings) would be slightly exceeded. In the worked examples the Max. Column
Steel Bar Spacing has been relaxed to 205mm in order that the above bar layout
can be used.
In direction 1
Beam stiffness at top of column
L = 5500mm b= 250mm d = 500mm
kb1 =b*d3/(12* L) = 473484.8 mm3
Column stiffness
kc1 = b2*b13/(12 * Lo1) = 520833.3 mm3
calculation using the formulae given in 2.5 of BS 8110-2:1985
c2 = kc1 / kb1 = 1.100
c1 = 1 (fixed base is defined in this example)
cmin = min ( c1, c2 ) =1.000
eq3 = 0.7 +0.05 *( c1 + c2) = 0.805
eq4 = 0.85 +0.05 *( cmin) = 0.900
1 = min (eq3,eq4) = 0.805
In direction 2
Beam stiffness at top of column
L =4250mm b= 250mm d = 400mm
kb1 =b*d3/(12* L) = 313725.5 mm3
Column stiffness
kc1 = b1*b23/(12* Lo1) = 130208.3 mm3
In direction 1
a1 = 1/2000 * ( 1 * Lo1 /b1)2 = 0.032
In direction 2
a2 = 1/2000 * ( 2 * Lo2 /b2)2 = 0.123
Column is slender, hence:
Calculate Madd
K = (Nuz - N) / (Nuz - Nbal) <= 1
Applied Axial Load, N = 151.8 kN
col dimension (in direction under consideration), h = 500 mm
column width, b2 = 250 mm
concrete grade, fcu = 40 N/mm2
steel grade, fy = 460 N/mm2
material factor for steel , s = 1.150
area of steel requiredAs_req = 654.1 mm2
Nuz = 0.45 * fcu * (b1*b2 - As_req )+ 0.87 * fy * As_req = 2500.0 kN
Nbal = 0.25 * fcu * (b1*b2)= 1250.0 kN
(Nuz - N) / (Nuz - Nbal) = 1.879
Hence
K=1
Au = ( a2 * K * b2) = 30.8 mm
Additional MomentMadd2 = N * Au = 4.68 kNm
In direction 1
Smaller end momentM11 = -28.20 kNm
Larger end momentM21 = 57.41 kNm
Mi1 = 0.4 M11 + 0.6 M21 >= 0.4 M21
Mi1 = 0.4*M11 + 0.6*M21 = 23.17 kNm
0.4 * M21 = 22.96 kNm
Hence
Mi1 = 23.17 kNm
Md1eff = Max (abs(M21),abs(Mi1+Madd1),abs(M11+Madd1/2), M1min) = 57.41 kNm
In direction 2
Smaller end momentM12 = 7.85 kNm
Larger end momentM22 = -15.77 kNm
Mi2 = 0.4 * M12 + 0.6* M22 >= 0.4 M21
Reinforcement in Section
In this example the design ultimate axial load is determined using Cl. 3.8.4.3.
Nmax = 0.4*fcu * b1 * b2 + 0.75 * Asprov* fy = 2274.8 kN > N OK
It is important to appreciate that 82% utilisation does not mean that 18% more load can be
added. As is shown on the interaction diagram for this column, a great deal more axial load
could be added.
It is important to note that design stages 1 to 7 are identical to the previous example, hence the
effective design moments about each axis are unchanged:
Md1eff = Max (abs(M21),abs(Mi1+Madd1),abs(M11+Madd1/2), M1min) = 57.41 kNm
Md2eff = Max (abs(M22),abs(Mi2+Madd2),abs(M12+Madd2/2), M2min) = 15.77 kNm
Instead of converting these to a uni-axial design moment (as per stage 8), an exact solution is
determined using the bi-axial moments.
The result is that the area of steel required drops from 650.8mm2 to 328.8mm2.
Thus, this design method can obviously be seen to provide a more economical solution. The
drawback is that because the neutral axis is no longer parallel to either face of the column
verification is more difficult. The cross checks required do not lend themselves to hand
calculation.
This will force the design to adopt the number of bars shown in the 'Qty' cell of the above
table. In this example we will use 8 bars in the design, (Qty = 8). When the design is
performed, the bar sizes will be adjusted to obtain an economic solution based on this layout.
column diameterD = 500 mm
The clear height of column in the two directions takes account of the beams framing in to the
top of the column.
Lo1 = 5500mm - 500mm = 5000mm
Lo2 = 5500mm - 400mm = 5100mm
All the columns in this building should be considered as braced. If the column is designed as
such, the effective length factors in the two directions are calculated as:
1 = 0.846
2 = 0.900
Short hence:
Additional MomentMadd1 = 0 kNm
Additional MomentMadd2 = 0 kNm
In direction 1
Smaller end momentM11 = -19.97 kNm
Larger end momentM21 = 40.56 kNm
As the column is short the effective moment is simply:
Md1eff = Max (abs(M21), Mmin) = 40.56 kNm
In direction 2
Smaller end momentM12 = -39.85 kNm
Larger end momentM22 = 80.60 kNm
Md2eff = Max (abs(M22), Mmin) = 80.60 kNm
When unbraced, the effective length factors in the two directions for this column change to:
1 = 1.439
2 = 1.684
In direction 1
a1 = 1/2000 * ( 1 * Lo1 /D)2 = 0.104
In direction 2
a2 = 1/2000 * ( 2 * Lo2 /D)2 = 0.148
Calculate Madd
K = (Nuz - N) / (Nuz - Nbal) <= 1
Note that now the column is unbraced, combination 1 has become the most critical
combination and hence the value of N has changed in the calculation of K below.
Applied Axial LoadN = 285.1 kN
Area of sectionA = pi * (D/2)2 =196349.5 mm2
concrete gradefcu = 40 N/mm2
steel gradefy = 460 N/mm2
material factor for steel s = 1.150
area of steel requiredAsreq = 845.1 mm2
Nuz = 0.45 * fcu * (A - Asreq )+ 0.87 * fy * Asreq = 3857.3 kN
Nbal = 0.25 * fcu * A = 1963.5 kN
(Nuz - N) / (Nuz - Nbal) = 1.886
Hence K = 1
Au1 = ( a1 * K * D) = 51.8 mm
Additional MomentMadd1 = N * Au1 = 14.76 kNm
Au2 = ( a2 * K * D) = 73.8 mm
Additional MomentMadd2 = N * Au2 = 21.03 kNm
In direction 1
Smaller end momentM11 = 0.00 kNm
Larger end momentM21 = 0.00 kNm
Md1eff = Max (abs(M21+Madd1),abs(M11+Madd1), Mmin) = 14.76 kNm
In direction 2
Smaller end momentM12 = -47.50 kNm
Larger end momentM22 = 96.21 kNm
Md2eff = Max (abs(M22+Madd2),abs(M12+Madd2), M2min) = 117.24 kNm
The area of concrete in compression (the shaded area above) is determined from the equation:
A = R2 tan-1 (squareroot (R/r)2 -1) - r (squareroot (R2 - r2)
R = D/2 = 250 mm
r = R - Ybar = 146 mm
A = 29568 mm2
Chapter 20
Introduction
Orion designs walls in the same way as columns, therefore for general details of the design
process refer to the previous chapter - Column Design to BS8110.
In this chapter additional aspects of the process that are unique to walls are examined.
In the column/wall design settings you can choose between shear wall reinforcing styles
shear walls with or without End Zones.
The column application plan view below shows the reinforcing for wall panels when end zones
are requested. Note that there is significant overlapping of the bars where the panels overlap.
As noted earlier, this is one of the areas where manual adjustment of the detailing is required.
The detailer needs to look at the total reinforcement required in the overlapping zone and
provide the same total in a revised corner detail. For example, at the corner of walls 1 and 2,
wall 1 has 7 T20 bars and wall 2 has 13 T16, the detailer might change this to 10 T25. As noted
previously this summing of the individual panel requirements is a safe engineering approach.
The view above shows the column application plan view after redesigning all the walls having
changed to the design setting where End Zones are not considered.
When end zones are not used the same bar size will be used throughout the length of the wall
with the exception of the end corner bars which may be larger.
Once again there is a certain amount of overlapping reinforcing between wall panels that
needs to be rationalised, but this tends to be a simpler process where end zones are not used.
Note
Even when end zones are not used Orion will still tend to suggest a larger bar size
at the extreme ends of the wall. Sometimes this might be needed because it
increases the supplied steel just enough to exceed the minimum steel requirement.
Sometimes it may also help to satisfy detailing (maximum spacing) requirements
across the end of the wall. However Orion will often provide larger bars here for
no other reason than it is considered Good Practice. In the case of intersecting
walls it is particularly felt that this is good practice. However, it is acknowledged
that some engineers would prefer to use a single main bar size throughout and the
option to set this preference will be added in a future release of Orion. For now it
is possible to interactively adjust the sizes of the end bars to achieve this objective.
The British Standard does not really acknowledge or provide guidance on End Zones. It is
written from the point of view that walls will be quite uniformly reinforced, hence many Orion
users will prefer to use the option NOT to consider end zones. As is demonstrated in the
above example, for many structures this will actually give a lesser reinforcing requirement.
The only argument that can be set against this approach is in the case where the walls are
actually resisting significant in-plane moments. In such cases not only does the end zone
approach become more efficient, it also puts the steel where the strain is and so potentially
limits cracking more efficiently.
BS8110 cl 1.3.4.7 defines a plain wall as one that contains either no reinforcement, or
insufficient to satisfy the criteria in cl 3.12.5. Any reinforcement supplied is ignored when
considering the strength of the wall. Where no reinforcement is required in the design of the
wall cl 3.9.4.19 indicates that if reinforcement is provided at all, it should be at least 0.25% (for
Gr460/500 steel) and 0.3% (for Gr250 steel)
Therefore, if design can be based on no reinforcement and this indicates a capacity in excess of
that required for the design combinations, a lower minimum reinforcement level can be used.
Note
Orion will determine if a wall section is in compression throughout for each applied load
combination. When compression throughout is proven then design will utilise this lower
minimum steel provision.
The "detailed design report" option includes a statement that wall is in compression
throughout and that plain wall design applies as shown below:
Although plain wall design applies, the capacity of the section (and hence the utilisation
ratios) are still based on the provided area of steel.
The final calculations, interaction curves, etc relate to design of the section using the actual
supplied reinforcement. These calculations also show the NA position calculated for the
critical combination.
This NA position effectively reflects the theoretical capacity moment for the given axial load
and the required reinforcement. Therefore this may often indicate that the NA lies in the
section hence implying that the section is not in compression throughout - when designed as a
plain wall this is a false implication and can be ignored.
The single layer may be either mesh or loose bars and it may be used in conjunction with end
zone reinforcement if required.
In the column/wall design settings you can then set preferences relating to mesh
reinforcement. The settings shown below would be quite reasonable in this regard.
Note the Min. Wall Length for Mesh Steel - a wall of this length or greater will attempt to adopt
mesh reinforcement - provided you have chosen to use a fabric for wall web steel in the
material properties dialog.
You should set the mesh width to that of the standard panel width you intend to use.
The "Revert to Loose Bar" Option
The option to use loose bars (shown checked above) applies during the main steel design
routines when mesh has been specified. If no mesh can be found with sufficient main steel
reinforcement to meet the design requirements then the wall is set to use loose bar and
re-designed on that basis.
How the "Revert to Loose Bar" Option Works
In the properties for any wall, the Web Steel (Longitudinal) property will initially be set as
"default". The default setting results in loose bars being used if the wall in question is shorter
than the minimum length for mesh steel with mesh steel being used for longer walls. However,
in the latter case, if it is determined that no available mesh is adequate and the "Revert to Loose
Bar" option has been checked then the Web Steel property for the wall be re-set to a relevant
non-default value as shown below.
Once this is done then loose bar will always be selected for that wall until the property is re-set
back to default. Re-setting these properties would be tedious if done interactively so when
running batch design we have added the option to "Try Using Mesh in all Walls". When this is
checked the properties are re-set to default before the batch design and so if the design forces
have changed mesh could be selected in a wall where loose bar has previously been selected.
Limitation of "Revert to Loose Bar" Option
The design will not revert to loose bars in the following cases:
1. If the main bars of a mesh provide sufficient main reinforcement but subsequently it is
found that lateral bars do not provide sufficient transverse reinforcement, then the design
will show the wall as failed and will not automatically revert to using loose bar.
Generally this possibility only applies to B series meshes in thicker walls (as indicated by
the * in the Minimum Reinforcement Requirement tables below).
2. If the main bars of a mesh provide sufficient main reinforcement to meet combined Axial
load and bending, but subsequently it is found that the limiting axial load is not sufficient,
then the design will show the wall as failed and will not automatically revert to using loose
bar. This requires walls to be subject to unusually high axial loads and should occur very
rarely.
Column Steel Details View
The view below shows the column steel details view for a wall with mesh reinforcement. This
view is obtained from right mouse click menu within the column application plan view.
Plain walls - min horizontal and vertical reinforcement is 0.25% (basically an anti-crack
provision)
Plain Walls
Reinforced Walls
A series mesh
B series mesh
A series mesh
B series mesh
125mm
A393
150mm
A393
175mm
200mm
Plain Walls
Reinforced Walls
A series mesh
B series mesh
A series mesh
B series mesh
125mm
A193
B196
150mm
A193
B196
A393
B385
175mm
A252
B503
A393
B503
200mm
A252
B503
B503
250mm
A393
300mm
A393
B283
2. For walls thicker than 150mm where double layer reinforcement is being used:
If the plain wall design option is activated then potentially A-series meshes can be
selected in quite thick walls.
However, where plain wall design is not selected or not applicable, then A series
meshes will not be picked in any of the thicker walls.
If B-series meshes are selected then mesh will never be picked in walls thicker than
200mm. (This is because no B mesh satisfies the min horizontal reinforcement
requirement for thicker walls).
3. Where no mesh is suitable the design will either fail (which could be many/most walls in a
building) or design may revert to using loose bar if that option has been activated.
The options shown above allow for a large range of reinforcement configurations to be
considered automatically. If a single layer option is selected (as highlighted above) it will only
be used in walls thinner than a specified max thickness (shown as 150mm above), for thicker
walls a double layer of mesh will be used.
The example below demonstrates the range of design outcomes that are possible..
A simple model is created with 8 walls (as above) set with lengths and widths such that
different design results are expected:
Walls 1 - 4 are 150 mm thick - with max thickness of single layer walls set to 150 mm we
expect to see single layer reinforcement in all these walls. Walls 4 - 8 are 175 thick so
double layer reinforcement is used.
Min length of wall in which mesh can be used is set to 1300mm - so we only ever expect to
see mesh in walls 3,4,7 and 8.
W1 (900x150) and W5 (1050x175) - are both short walls - setting allows use of end zones
in short walls.
W4 is identical size to W3 - the loads in this wall have been increased in order to
demonstrate that when mesh is specified as the default preference, the design will revert to
using loose bars if the largest available mesh is not adequate. (Same applies to W8 and
W7)
View above shows design result when plain wall design is not active and when loose bar is set
as the general preference - a successful design is achieved for all walls. (Note that the
reinforcement requirement in wall 8 exceeds 2% resulting in the automatic introduction of tie
bars in the wall.) View below shows how design changes if the plain walls design option is
activated.
NOTE: Review of the designs at this point confirms that plain wall design is applicable to
walls 1,2,3,5,6,7 and hence the steel supplied in these walls drops as shown above.
Design using Type A Mesh (not considering plain wall design option)
Refer to the Mesh and Minimum Reinforcement Requirements table earlier in this chapterthis indicates that the heaviest mesh (A393) will not be sufficient in single layer walls, but that
A393 may be OK as double layer mesh in the 175 thick walls.
Unexpectedly in 1W3 A393 is selected, this is achieved because heavy end bars are selected to
increase the area of steel provided to beyond 0.4%, this would not be possible on longer walls.
In walls W4 and W8 the heaviest available meshes are not adequate and so the design has
successfully reverted to the use of loose bar.
In W3, A393 mesh is still used, but the end bar sizes reduce.
In W7, A252 is now used instead of A393.
As expected design fails in walls W3 and W4. For reasons stated earlier (Limitations of Revert
to Loose Bar Option), since these failures relate to the failure of any mesh to provide minimum
horizontal steel, we do not expect the design to revert to use of loose bar in this case.
Chapter 21
Foundation Design
Introduction
In addition to building superstructure analysis and design, Orion includes foundation design
calculations for the following:
Pad Footings /Combined Pad Footings,
Some further detailed preferences are set in the Graphical Editor Settings Foundations tab.
Note in particular the Allowable Soil Stress Ultimate Strength Factor - in Orion the foundation
design procedures assume that the permissible bearing pressure is used in conjunction with
ultimate loads.
Hence if the allowable soil stress is given as 200 kN/m2 the program then checks factored
combinations against a factored permissible bearing pressure of 1.4 x*200 = 280 kN/m2. This
is usually conservative.
Foundation Depth
The foundation depth is set up globally for the whole structure a via the Edit Storey dialog.
In order to create detail drawings including starter bars, a reasonable (non zero) foundation
depth is required.
Although the foundation depth is added onto the column length displayed in the 3D window,
it does not effect the column design length.
A column kicker depth cn be set in in Column Design Settings Settings Detail Drawings General tab.
If different footing heights have been specified in the same model, the undersides are all set
out at the foundation depth - the topsides will thus vary in elevation.
The forces and moments displayed are initially those determined from the analysis. If required
they can be revised in one of two ways:
Applying a Force Override To apply a Force Override simply overtype the displayed
value with the override value. Columns which have had a force override applied are
indicated by a tick in the Edited checkbox. If you remove the tick the value determined
from the analysis is restored.
Added Axial Load If you require, an extra axial load can be added to the value
displayed. The extra load is entered in the Added Axial Load column. If you want to add
the same amount of axial load to all columns this can be achieved by clicking the 'Apply
added axial load to all columns' button. This feature can be particularly useful for piled
foundations.
Note
If the loads are edited via either of the above methods, the revised values are the
ones displayed on the plan view at foundation level.
4. The members are now grouped together for the purposes of foundation design.
To ungroup columns and walls
1. Select any member in the group (this selects all members of the group).
2. From the right click menu select Remove Foundation Column Grouping.
3. The members are now ungrouped.
Calculation of the Combined Footing Design Forces
The axial forces in the individual group members are summed together to provide the design
axial force for the footing.
The design moment is calculated as the sum of the moments in the individual columns, plus
an additional moment due to eccentricity of the axial force in each member from the
geometric centroid of the group.
In the example above two pinned columns are supported on a single pad base. The combined
design forces for load combination 1 are thus:
N = 19.4kN + 345.1kN = 364.5kN
Mx = 345.1kN x 0.9m - 19.4kN x 0.9m = 293.1 kNm
My = 0 kNm
Pad Footings
The footing is sized to ensure the soil is not overstressed. If the applied moment is greater in
one direction, it will usually be rectangular, although you can restrict the design to a square
footing if you prefer.
Reinforcement to resist bending is determined in each direction and punching and shear
checks are performed.
If required, more than one column can be supported by a single footing. For details refer to the
section on Combined Pad Footings
The requirements/procedure for defining a pad base are in the help system. The steps
involved are summarised below:
1. A valid building analysis (or FE chase down procedure) must have been completed so that
the relevant foundation design loads exist. Column reactions are simply fed into the pad
base design module, there is no more complex interaction of superstructure and the
foundation analysis models.
2. Display the foundation level, (ST00) in the Plan View.
3. Select the column under which the pad base will be inserted. More than one column may
be selected. In that case, the calculated pad base (based on the effects transferred from all
the selected columns) can optionally be inserted to all selected columns.
4. Display the right click menu and select the Insert Pad Footing option.
5. The Pad Base Properties menu will be displayed. Check the parameters in this menu and
then press the Analysis button to design the footing.
6. Use the up arrow buttons located to the right of the "Lx" and "Ly fields to decide on the
desired footing size.
7. Press OK to close the Pad Base Design dialog, then press OK once more to complete the
insertion of the footing. The base is then displayed on the Plan View as shown above.
Pile Caps
Defining a Pile Cap
Define the pile and pile cap data as required. The pile cap dimensions can be calculated
automatically - if entered manually the initial values may be rounded up or down during
the analysis process, (if this happens try modifying the inital value and analysing again).
Note
The reinforcement defined here is neither designed or checked in any way
6. On the Loads tab you can use the existing combination data or input manually defined
loads. Surcharge is also applied on this tab.
7. On the Parameters tab you can specify the minimum number of piles, spacing
requirements and other design parameters.
Note that the max pile spacing multiplier only applies to walls , so as to achieve efficient
transfer of load through the pile cap.
If you check the box to allow a tapered footing, then the corners will be chamfered (as
illustrated by the footing shown on the right at the start of this section).
8. Press the Analysis button to design the pile cap.
9. A report is displayed as shown below. If required, a column forces table and the pile axial
forces can be added to this report - options for this exist on the Parameters tab.
10. The number of piles used can be adjusted on the fly by clicking the (+)Increase or
(-)Decrease buttons. It is not possible to decrease to less than the original number of piles
calculated. In other words, (+)Decrease can only be used to reduce an (+)Increase.
11. Click OK to insert of the Pile Cap. The footing is then displayed on the Plan View.
Limitations
Within the above apparently simple process there are a number of limitations to be
considered:
1. Piles are positioned on a fixed spacing basis, spacing is not adjusted (increased) to resist
bending, the number of piles is always increased.
2. Pile groups are centred on the supported member - No offset grouping to resist applied
moments.
3. No ability to model ground beam systems in such a way that analysis is performed and
additional loads are introduced to the pile caps.
4. Design relates to pile selection only, no design is carried out for the pile cap:
No shear checks or cap depth checks
Design is carried out using factored combinations with a factor applied to SWL
(default 1.4).
The Factor applied to SWL is specified on the Foundations tab in Graphic Editor Settings .
Note
There is potential for a significant error if you incorrectly provide the ultimate
design capacity of the pile (which may be 3 times SWL)
6. No general summaries (pile schedules).
Currently there is no way to apply preferences to prevent non-preferred layouts from being
proposed and no way to interactively define an alternative preferred layout.
Strip Footings
For the most part strip foundation design is relatively straightforward. However, Orion offers
some options in relation to strip analysis and the existence of such options can introduce
uncertainty in some cases. This section therefore focuses primarily on the strip analysis
options that are offered during the course of strip analysis and design.
The requirements/procedure for defining a strip are in the help system. In this example we
will consider a strip running between columns C2, C5, C7 and C9 as shown above. The steps
involved are as follows:
1. A valid building analysis (or FE chase down procedure) must have been completed so that
the relevant foundation design loads exist. Column reactions are simply fed into the
foundation design modules, there is no more complex interaction of superstructure and
foundation analysis models.
2. Beams are defined between the columns along this line (as shown in the previous screen
shot) Note that these beams will be required to resist the longitudinal moments developed
along the strip. Hence the sizing of these beams needs to be realistic and getting the size
right may prove to be an iterative and interactive procedure we are starting here with a
beam strip 800 wide and 800 deep.
3. The 3 beams need to be selected and then a right click exposes the option to Insert Strip
Footing and the strip footing design dialog appears as shown below.
4. The strip design is performed, as discussed in more detail in the following sections. When
the design is saved the strip appears on the Plan view at foundation level as shown below
5. The foundation beam design is then carried out in a similar manner to the design of
superstructure beams, but it is a 2 stage process:
a. First use the Create/Update Footing Beam Records option as shown above. (This
transfers the strip analysis results data into the beam design module.)
b. Then run the Foundation Beams design option. In this case only one strip has been
defined and so the beam design dialog appears as shown below.
6. You can design the foundation beams interactively or in a batch mode in exactly the same
way as you would for beams in the supported building and drawings can be created/
exported in the usual way.
7. Elevations and sections are created as shown above, it is probably best to activate the
option to label bars within the section since this means that the cantilever footing steel is
also noted up as shown below. It is also suggested that it is better to use the detailing
option to Re-plot the bars below the beam resulting in a display as shown above. The
foundation beam detailing is not yet as automated as the mains structure beam detailing
and this ensures that the maximum amount of information is readily available when final
edits are made.
We will now examine some of the options and the design philosophy in a little more detail.
There are various options at this point the left and right extension length and the footing
depth are not determined or optimised automatically during design these could be set now or
reset after an initial design is examined.
The assumed design philosophy is that a footing cantilevers out to each side of the 800 by 800
longitudinal beam, in the example above the assumed initial depth for this footing is 400 mm.
If the footing width is set to zero a minimum width will then be calculated automatically.
Note that the axial loads listed in the table of design information relate to one currently
selected load combination. A design carried out at this point is for the one selected
combination only. To design for all combinations the design envelope option must be
checked. We will look at this later, for now we will make analysis comparisons based on the
G+Q*F combination only.
Adjusting the Subgrade Coefficient
The analysis method accounts for the assumed elastic deformation of the soil and the model
used is that of a beam supported by springs. The springs model the stiffness of the soil which
you define by setting the subgrade coefficient. The subgrade coefficient defined in the
building parameters is displayed here and can be edited.
When you click on design a tabbed dialog is displayed, on the first tab calculations are shown
for review/printing, on the second tab diagrams are displayed for review/printing. The
diagrams for this strip are initially as shown below.
Note that the bearing pressures are varying, with peak pressures occurring under the load
positions.
The permissible bearing pressure was given as 200 kN/m2. Since the applied column loads are
factored the design is aiming to keep the peak bearing pressure below a factored allowable
pressure which is conservatively assumed to be 1.4*200 = 280 kN/m2. (The 1.4 factor can be
adjusted in the foundation settings.)
In this initial run this has been slightly exceeded, but as is shown below, some degree of
manual optimisation is always required.
When the initial design is reviewed you can see that the footing width is not a practical sort of
number you might change this to something like 3000 mm. You might also think that the 1
m extensions are a bit long, lets reduce those to 800 mm. The revised diagrams are then as
shown below.
You may not be confident of what value to define for the sub-grade coefficient, the table below
indicates the potential range of values that might be considered.
We can easily examine the effect of changing this coefficient, the diagram below shows the
result when this strip is rechecked with the coefficient reduced from 80000 to 8000 .
This results in a more uniform bearing pressure with lower peak values. By assuming weaker
soil springs we are in effect considering the footing as being relatively more rigid and so the
pressures get spread more evenly. If the design were carried out on this basis a narrower strip
footing could be used. But notice that the span bending moments in the strip have increased
significantly while the moments under the columns have reduced. In fact this result is
probably more akin to the results achieved in traditional hand calculations.
In practice you may want to design for a range of possibilities, this can be achieved
automatically in Orion by using the step option as shown below.
We have asked the design to consider a range of subgrade coefficients from 8000 to 80000 and
the resulting diagrams for this one load combination are then as shown below.
The bearing pressure diagram shows a single case and indicates the maximum bearing
pressure from all the cases considered. The diagrams indicate show the envelope of design
conditions that have been developed.
As noted previously the pressure diagram displayed at this point is not necessarily the most
critical one, but the maximum pressure is shown. The envelope diagrams for design forces
have now widened to include the worst cases for all load combinations and the required range
of sub-grade coefficients.
In practice you may not want to consider such an extreme range of sub-grade coefficients as
have been used here, but the points to note are:
Higher sub-grade coefficients result in higher local bearing pressures and hence
potentially wider footings.
Lower sub-grade coefficients result in some higher strip design forces which may be closer
to those that you would expect on the basis of traditional hand calculations.
Note
General Limitation The analysis never attempts to take any account of column
base moments transverse to the strip direction. It is anticipated that such forces
will not be allowed to develop where the intention is to use a strip footing i.e. it
is assumed that the column bases will have been defined as pinned or that the
moments are nominal. If you wish to consider wide strips where biaxial moments
get applied from the columns the FE foundation design option should be used.
The above cross section illustrates the two aspects of the strip design:
1. The longitudinal beam that spreads the column point loads along the length of the strip
in this case an 800 x 800 beam.
2. The cantilever edges that in effect spread the line load developed by the beam out to the
edges of the strip. Within Orion this part is referred to as the footing design.
When the design is run from Strip Footing dialog shown above, the results dialog appears as
shown below.
In the previous section we looked at the results shown on the diagrams tab, here we are
looking at the numerical results. In the previous section we saw that the maximum bearing
pressure was established as 278 kN/m2. This peak figure is repeated in the numerical results
above.
The design of the steel required in the cantilever footing is carried out at this point. Note that
the cantilever is considered as the distance from the face of the beam and not the centre-line of
the strip and that the peak bearing pressure is used. For the bar size defined in the input
dialog, the minimum spacing requirement is determined. You can change the bar size and the
footing depth until you have an arrangement that you are comfortable with. These sorts of
changes do not affect the bearing pressures.
Beam Design
Scrolling further down the strip design results you will see the design forces for the beam
being reported at the end, but the beam design is not carried out at this point. These are the
forces that are transferred to the beam design module when you Create/Update Footing Beam
Records. You may notice that the design moments noted above for transfer are not identical to
the moments indicated on the envelope diagram in the previous section. This is because
alternative Minimum Moments are also calculated and shown in the results at the top of the
data shown above.
If the minimum moments exceed the analytical results then they are used in design. The
minimum moments are based on average bearing pressures (W) and are calculated as follows:
Minimum Span Moment = WL2/16 (where L is the span length)
Minimum Support Moment = WL2/12 (where L is the longer of the 2 spans to each side of
the support in question)
In cases with multiple equal spans these minimums are unlikely to dominate significantly if at
all. In cases with unequal spans they may be more significant. In this example the minimum
moments are greater in several cases. Application of such minimums may be regarded as good
practice since they ensure that the foundation strip is more capable of dealing with loading
variations (pattern loads) and variable ground conditions (soft spots). If you feel that these
minimums are introducing too much conservatism you can adjust them in the Foundation
Design Settings dialog shown at the start of this chapter.
Having transferred the design forces the beam is designed in the same way as superstructure
beams. Design reports are available (as shown below) as are options to print/export
reinforcement details.
Strip footings with deep stiffening beams are more structurally efficient but may be considered
more costly to construct in some circumstances. Many engineers prefer wide rectangular strip
footings as indicated above, particularly for lower rise construction.
Wide strip footings can be generated in Orion, but the following should be noted:
1. You need to start by defining wide beams between the supported columns, and then select
and insert the strip footing in the usual way.
2. These strips tend to be shallow and therefore flexible so you may find a wide variation of
bearing pressures on stiffer ground (when the sub-grade coefficient is higher).
3. When you do the design you do not want a cantilever footing to be introduced, you will
probably need to keep changing the beam width until an acceptable solution is achieved.
4. Links are introduced to satisfy longitudinal shear only. As with superstructure beams
there is a maximum of 3 links (6 legs) across the section. For wide footings this may not
be sufficient.
5. No design across the width of the strip is carried out i.e. the links are not designed to
resist any transverse bending.
6. Punching Shear under the columns may be a more critical design condition than shear
across the full section of the beam. This it is NOT checked at all.
Piled Rafts
Defining a Piled Raft
Define the pile and pile spacing data as required. A preview of the resulting pile layout is
displayed in the Plan View.
7. Click OK to create the raft.
Where load exceeds capacity the pile node is shown red, where it is greater than 90% the pile
node is orange, for < 90 % the pile node is green.
Chapter 22
Introduction
The objective of this chapter is to demonstrate the solution options available for structures
containing inclined/lowered members, of which any of the following are possible:
sloping and/or lowered slabs
Consider the model shown above, which features a number of inclined members:
inclined beams, slabs and columns have been defined at storey 2,
these columns and also the end wall have been lowered beneath the general floor level,
a tapered wall has been defined at storey 1.
Note
The features described in this chapter are for the purpose of defining occasional
sloping/lowered members within a model containing distinct horizontal floor
planes. These features are NOT intended to facilitate the modelling of structures
with complex geometries in which the floor planes are not readily apparent.
The method used for creating inclined/lowered members is described elsewhere - see
Modelling Inclined and Lowered Members
Area and line loads applied via the Slab Load tool are always applied to the projected slab
area and therefore take no account of the incline.
For slab loads defined via the slab properties dialog (including slab self weight) - whether
or not the incline is taken into account when determining the loaded area will depend on
the method of solution:
building analysis takes the incline into account (irrespective of whether yield line, or FE
beam load decomposition is used).
The model described at the start of this chapter is used in this example.
Diaphragm Modelling
One of the key issues to consider for any model which incorporates inclined planes is the
diaphragm modelling. There is a potential for errors to result if those nodes being constrained
by a diaphragm do not all lie in the same plane. A warning is displayed during the building
analysis if such a situation exists:
In this example there is a potential problem at storey 2. The deviation from a common plane is
small and the error introduced will not be too great, however, for the purpose of the example it
is convenient to remove the problem. Corrective action requires the physical exclusion of the
two inclined slabs from the floor diaphragm as shown below.
Note that the removal of slabs from a diaphragm in this manner is only effective if the Storey
Diaphragm Model is set to Slabs to Define Rigid Diaphragm in the Building Analysis Model
Options.
The diaphragm model that results as a consequence of the above settings is shown below, note
that all nodes constrained by each diaphragm are constrained within a single plane:
For further guidance on the topic of working with diaphragms and inclined planes see
Diaphragm Modelling
Analysis
A building analysis is to be performed using yield line load decomposition.The diaphragm
model is set as Slabs to Define Rigid Diaphragm and the other Analysis Model settings are as
shown below.
This report shows up to 4 tables, the first indicates the sum of loads as they are applied to the
structure (the "un-decomposed" loads). The second indicates the sum of loads as they are
applied to beams at each level (after "decomposing" slab loads). The third indicates the total
column/wall loads derived at each level after the building analysis. The fourth (not shown
above) indicates the total column/wall loads derived at each level if an "FE Chase Down" has
been performed.
In the example above you can see that no significant loading discrepancies have occurred in
either the load decomposition phase or the building analysis phase, the analysis is complete.
The following points should be noted:
The reduction in self weight due to the wall and two columns being lowered below the
general floor level at storey 2 is accounted for.
The slight increase in self weight where beams, slabs and columns have been inclined is
accounted for.
For the two inclined slabs, the sloped area rather than the projected area has been used
when calculating the full area dead and live loads.
The total column/wall loads reported in the third table are adjusted to take into
consideration any non-vertical columns or walls. Only the axial force component is
adjusted, no allowance is made for the inclusion of a vertical shear component. This may
result in a slight discrepancy between the second and third tables.
Switching to FE Beam Load Decomposition
For the purposes of load decomposition irrespective of the method used, inclined/lowered
beams and slabs are considered to be in the same plane, this is clearly apparent when viewing
the FE load decomposition model for storey 2 as shown below:
If the analysis is re-run having been loaded on this basis, the same adjustments for inclined
members are made as were described previously and hence the building analysis result should
be similar to that obtained on the basis of yield line decomposition.
The axial load comparison report based on FE beam load decomposition is as follows:
Although there is a slight increase in the load applied when the FE decomposition method is
used, this is due to a slight conservatism in the FE decomposition method generally (and is
unrelated to the presence, or otherwise, of inclined members).
Note
The concrete and formwork quantities do not take account of the sloped length
of the beams.
Each section is drawn perpendicular to the face of the member - for the inclined columns the
sections are therefore inclined as opposed to horizontal.
By examining the longitudinal steel details, it is clear that the drop has been accounted for, (the
bar lengths differ by exactly 500mm). However, it is also clear that these lengths do not take
into account column inclines (if they did the H16-2 bars would be slightly longer). This
limitation also applies in the column steel quantity reports.
Note
Similarly, the concrete and formwork quantities take into account the column
drops but do not take account of the column inclines.
Note
The tapered parts of the wall are not included in the concrete quantities. The
formwork also follows the same sizing ignoring the taper.
Although the overall results are only affected very slightly, there is a local increase in the axial
force due to dead load in the tapered wall.
This increase is attributable to the chosen method of modelling. The mid-pier modelling of
the tapered wall uses an average width defined by top width which means a reduction in the
wall self weight used in the analysis.
The section of the wall used in the analysis based on mid-pier modelling is shown hatched,
above.
The FE shell model gives a closer approximation to the actual self weight of the tapered wall.
We would not suggest that this is sufficient reason to use FE Meshed Wall modelling generally.
Tapered walls are unlikely to feature in a typical model and even if they do, the amount of
error introduced by the mid-pier idealisation is quite small - the benefits of simplicity and
speed of solution that are offered by the mid-pier wall solution should not be ignored.
Note
Note that there are two transverse beams in the middle bay, one directly underneath the other.
The structure may appear valid but a problem can be seen in the load decomposition views.
Both transverse beams in the middle bay are considered to be in the same plane for the
purposes of load decomposition - hence they are both conservatively loaded by the slabs from
both sides.
FE load decomposition
When FE decomposition is used, once again both transverse beams in the middle bay are
considered to be in the same plane - however using this method one of the two beams attracts
all of the slab load while the other is unloaded (apart from it's own self weight) - using this
method there is potential for some beams to be under designed.
Suggested Workaround
The recommended method of modelling in this scenario is to use floors defined by separate
storey levels and limit the use of planes to sloping and minor offsetting of slabs.
Alternative Workaround - Returning to the single storey method with a lowered plane, but
with 2 grid lines defined at close proximity so each beam is on a different grid line (and hence
receives load from one slab only). The program automatically detects that both beams are
connected to the same columns.
Plan layout showing the additional grid lines.
From both the above load decomposition diagrams, it can be seen that the correct loads are
applied to each beam in the centre of the model.
Example Model
The model from the building analysis example is re-used:
FE Chasedown Analysis
The analysis uses an idealised model in which sloped floor planes are adjusted to the
horizontal.
The following simplifications are applied:
Inclined beams and slabs are modelled horizontally at the general floor level.
Lowered (or raised) beams and slabs are modelled horizontally at the general floor level.
Any columns or walls which have had their top ends lowered (or raised) are modelled as if
they still connect at the general floor level.
tapered walls are modelled as rectangular walls, based on the top cross section of the wall.
FE Model Generation
The analysis models are created with a reduced slab stiffness multiplier (0.15) the aim of the
reduction being to transfer most of the slab load directly into the beams - (see Worked
Example Beam and Slab Systems in the Analysis and Design using FE chapter). When
creating the analysis models the following warning message is displayed:
The FE Analysis model for storey 1 is shown below. Note that the inclined columns are
positioned vertically at the point where they connect to the floor. The loads being applied
from the storey above via these columns are displayed.
Partly because of the previously mentioned simplifications, the sum of axial loads for the FE
Analysis do not exactly match the previous Building Analysis results.
When any beam is designed, the diagram moments are automatically cross checked against the
analysis moments. If merged FE results are used for steeply inclined beams, there is a high
likelihood that a mis-match will occur. In such circumstances, because the design forces are
invalid the beam fails. If you attempt to examine the results an error will be displayed similar
to the one shown below.
When the beams in the example model are re-checked after merging the FE results no
analysis/diagram moment mis-matches occur (because the beam slopes are small). Although
the beam utilisation ratios change, the reinforcement previously determined adequate for the
building analysis results remains adequate for all but one of the beams.
Column Design
For columns (and walls) a similar disparity exists - the building analysis model uses the sloped
member length, whereas the FE model uses the shorter vertical length. However, in this
situation an analysis/diagram moment mis-match will not arise - this is because the design
only requires the forces at the ends of the column - the forces at other points along the
member length are not considered.
In the example model, when the columns and walls are re-checked after merging, again the
utilisation ratios change. This result is not unexpected as different axial loads and moments
are bound to be generated by the different analytical approaches.
See also the general limitations for inclined members General Limitations Inclined/Lowered Members
The FE analysis model for this is as shown below, note the inclined columns in the last bay.
S-Frame Comparison
To investigate how significant this concern may be it is necessary to perform a comparative
study. The results from the Orion model for varying degrees of column slope are compared to
those obtained after exporting the model to S-Frame (a general analysis program). In the
S-Frame model multiple analysis runs are performed having removed the translational
restraint at the floor level.
The affect of varying the lateral stiffness is also investigated by changing the length of the walls
and also by remodelling them using shells as shown below:
The moment at the top of one of the inclined corner columns is recorded for 3 different
column inclines (45, 30 and 15 deg) and for a wall length of 1.5m and then 3m. These results
can be compared to the Orion model results (in which the floor is translationally restrained).
Typical Test Model Results
Column slope
Wall model
sway direction
(kNm)
perpendicular
to sway (kNm
0 degrees
126.6
-126.5
126.3
-126.7
126.0
-128.2
125.7
-128.1
117.4
-121.4
117.0
-121.3
101.6
-102.9
101.3
-102.8
15 degrees
35 degrees
45 degrees
translationally restrained
translationally unrestrained
sway direction
(kNm)
perpendicular
to sway (kNm
sway direction
(kNm)
perpendicular
to sway (kNm
Short wall
126.7
-128.9
113.3
-127.3
Short wall
(meshed)
115.2
-127.4
Long wall
126.4
-128.8
117.1
-127.5
Long wall
(meshed)
118.8
-127.6
translationally restrained
translationally unrestrained
sway direction
(kNm)
perpendicular
to sway (kNm
sway direction
(kNm)
perpendicular
to sway (kNm
Short wall
118.1
-122.3
89.7
-119.3
Short wall
(meshed)
95.6
-119.4
Long wall
117.8
-122.2
98.4
-120.0
Long wall
(meshed)
103.2
-120.2
translationally restrained
translationally unrestrained
sway direction
(kNm)
perpendicular
to sway (kNm
sway direction
(kNm)
perpendicular
to sway (kNm
Short wall
102.3
-103.8
55.0
-100.3
Short wall
(meshed)
65.6
-9100.9
Long wall
102.1
-103.8
69.7
-101.4
Long wall
(meshed)
78.0
-101.8
Discussion
By modelling the floor as translationally restrained, the Orion FE model does not cater for
building sway. The significance of this can initially be examined by comparing for the "short
wall" model the Orion results with the translationally unrestrained S-Frame results. These
show that the elimination of building sway in the Orion model in this example produces a
conservative answer for the column under investigation. In this example the degree of
conservatism increases as the amount of sway increases (i.e. as the column slope increases 12% for a 15 degree slope rising to nearly double the moment for a 45 degree slope.)
By making the comparison using the stiffer "long wall (meshed)" model instead, the
conservatism is 7% for a 15 degree slope rising to 25% for a 45 degree slope.
Several Points should be noted:
1. Even 2 discrete 3m long panels is probably providing an unrealistically low resistance to
sway and hence this example is probably exagerating this effect - in a practical structure it
is likely to be much smaller.
2. However, the sway effect would not necessarily always reduce column moments. In cases
where ignored sway might increase the moments this could be offset against the
conservatism.
3. In an un-braced structure (where sway stability relies on the interaction of columns and
slabs) this effect could not be ignored. As is noted elsewhere, unbraced flat slab
construction is beyond the intended scope of Orion.
4. Finally, inclined columns will clearly cause tensions and compressions to develop in the
slab. This effect will not develop in the idealised Orion model. Some consideration
should be given to this to ensure that tying details are adequate.
A final more general point is that the comparisons made also highlighted the increased
modelling sensitivities that arise when inclined members meet FE meshes. This issue is very
similar to the issues that arise when beams intersect with walls. There is extensive discussion
of this in the wall modelling presentation accessed from the wall modelling chapter.
Conclusions
Sway effects are catered for in Orion's Building Analysis, however they are not catered for in its
Finite Element Analysis of a sub-floor structure. (In exactly the same way as a traditional
sub-frame analysis would ignore sway effects arising from vertical loading).
Although in this example the effect was conservative, this may not always be the case for other
structural geometries.
Chapter 23
Introduction
A transfer level occurs wherever a column or wall is supported by a beam or slab. Consider
the simplified model shown here.
In this model there are walls along the rear and side elevations that stop at first floor level.
There are also two columns on the front elevation that sit on a first floor level transfer beam.
During a general building analysis of this structure, a completely coherent 3D analysis model
can be constructed and solved. Orion will not any issue warning or error messages for a
structure as simple as this one. However, in any model containing transfer levels there are
potential problems and limitations to bear in mind. Orion provides more than one way of
dealing with such models. This chapter is intended to introduce the options available to you
and the potential limitations.
Walls supported by more than one beam (An example to show an arrangement that needs
to be handled with caution).
Subsequent chapters go on to illustrate the optional FE Chase Down methods for assessing
transfer beams:
A simpler/faster way that you would use for more regular buildings and in particular
where you do not intend to make use of FE analysis results for slab design. A worked
example for this is provided in the chapter Transfer Beams FE Method, Option 1
(Simplest).
A more complex variation of the method above which uses the meshed up FE Floor
models. If it is your intention to use FE results for slab design, then you will find it easier
to use this method since it reutilises the same floor models. A worked example for this is
provided in the chapter Transfer Beams FE Method, Option 2.
In these examples we show how beams can initially be designed using one method and then
cross checked and amended using another.
Key Limitation
Before delving into the detail of the specific examples in the following chapters, it is worth
noting the key limitation that continues to apply to structures containing transfer levels.
The FE Chase Down method does not deal with sway (lateral) loads in any way. Sway
resistance is always assessed using the building analysis and is therefore subject to the
limitations of that analysis. If the columns/walls that are discontinuous are an important
element of the sway resistance system for your building then you will need to consider your
design with some care. For example:
1. If continuous walls brace your building then columns would be regarded as braced for the
purposes of design. Such columns are almost unaffected by sway loads and designed
exclusively for gravity loads. Where such columns stop at transfer levels, Orions iterative
building analysis and/or FE Chase Down methods will collect all the gravity loads and
apply them to the beams or slabs as appropriate. This sort of condition is not a concern.
2. It is much less common to see un-braced columns i.e. buildings where sway resistance is
provided by frame action between columns and beams/slabs. In such cases concerns
relating to discontinuous columns would be greater the same issues as described for
walls below would apply.
3. Discontinuous walls are more of a concern. Walls generally attract more significant sway
loads thus generating moments (or couples) at the transfer level. In the chapter Transfer
Beams General Method examples show how this condition is catered for analytically
while the design may prove to be conservative. However, in such circumstances it is
recommended that an increased level of attention is paid to the parallel checks that are
required when using any analysis/design software.
Chapter 24
If you want to work through this example for yourself, you can load model DOC_Example_01.
In order not to destroy the example for someone else you should then save the model with a
new name before proceeding.
Analysis
Check the options as shown above and start the building analysis which should then run with
no error or warning messages and be completed as shown below.
This report shows up to 4 tables, the first indicates the sum of loads as they are applied to the
structure (the un-decomposed loads). The second indicates the sum of loads as they are
applied to beams at each level (after decomposing slab loads). The third indicates the total
column/wall loads derived at each level after the building analysis. The fourth (not shown
above) indicates the total column/wall loads derived at each level if an FE Chase Down has
been performed.
In the example above you can clearly see that no loading discrepancies have occurred in either
the load decomposition phase or the building analysis phase, the analysis is complete.
Note
In the analysis settings you have the option to define an Axial Load Comparison
Tolerance, in this example it was set to the default 5%. If the overall totals in the
above tables vary by more than this amount Orion will issue warnings at the end
of the analysis.
Transfer beams are designed and detailed like any other beam; this procedure is controlled by
the beam design settings and preferences as covered in the Orion Training Manual. Details
along the lines of those shown below are produced automatically.
The columns above the beam are shown automatically and slab and beam lines behind the
beam are optionally dotted in.
Does it seem strange that the internal columns along grid A do not develop higher axial loads
than the corner column?
The results can be reviewed graphically to see shear force diagrams as shown above and
bending moment diagrams as shown below. Note how the side and rear walls are idealised
with a single mid-pier element.
The member force diagrams for the front transfer beam are shown above. It all looks very
reasonable a peak sagging moment of 579 kNm, and end shears of up to 336 kN.
Notice that the loading diagram does not show applied loads at the transfer column positions.
The transfer columns are a part of the analysis model and so the loads they transfer can only
be seen in terms of the steps in the shear force diagram, which are clearly visible.
Notice the small steps in the bending moment diagram under the supported column positions,
these are a small indication of the frame action being developed with the columns themselves,
this step could be eliminated by pinning the bottom of the transfer columns.
It is noted that there is currently no (easy) way to account for IL reductions within the
columns that load the transfer beam, in beams supporting columns with many levels above
this means that the transfer beam design will be conservative by default.
The member force diagrams for the rear transfer beam are shown above. Again, it all looks
very reasonable a peak sagging moment of 1155 kNm, and end shears of up to 416 kN.
Note that the wall is idealised as single mid-pier element therefore its load is concentrated at
the centre of the beam as can be seen from the shear force diagram. Clearly this idealisation
will tend to result in conservative design of the transfer beams.
Frame Action
Now consider the moment diagrams for the continuous beam line at second floor level above
the front transfer beam.
Perhaps not what some might initially expect there is no hogging across the transfer column
positions. This is a logical result of any full 2D or 3D analysis, the transverse beam is
deflecting, hence the supported columns are deflecting and this has an effect on the beams
above. In essence the loads are being shared according to the stiffnesses of the beams at all
levels. This effect makes more sense when the deflections for the frame are viewed as shown
below.
This also explains the apparently low axial loads noted in the transfer columns at the start of
this section.
It is an entirely logical result the same as would be given by any general 3D frame analysis
package. It is probably not the same result as would be derived by traditional hand calculation
methods. An engineer carrying out hand calculations would probably consider each floor
independently designing the continuous beam lines on the assumption that the columns
provide a rigid support producing a moment diagram along the lines shown below.
This may be more in line with traditional expectation but it has not considered the vertical
deflection of columns supported by beams or slabs at a lower level. A 3D analysis inherently
considers this effect.
Is there a correct answer what is it? Since the answer is related to deflection it relates to
stiffness assumptions. This can become an extremely complex subject which is discussed in
some detail in the context of Flat Slab Analysis and Design in the chapter Analysis and Design
using FE of this handbook. A very sophisticated assessment would take account of
construction sequencing and time dependent effects. In such circumstances the result needs
to be assessed for sensitivity to variations in the assumptions on which it is based. In fact the
result achieved by any such complex analysis will lie somewhere between the two extremes
shown above. A simpler approach to satisfying these extremes is to ensure that your design
covers for both possibilities. This can be achieved in Orion in one of two ways.
1. Carry out a building analysis on the basis described above and design all members.
2. EITHER:
Edit the properties of the transfer beam and artificially increase its stiffness (by increasing
both the inertia and the shear area) and hence largely eliminate transfer beam deflection.
Reanalyse and examine results to see that this has had the desired effect then run a design
check on all members in the structure if any member fails investigate and increase
reinforcement accordingly.
There are very small changes in the distribution of axial loads in the columns and walls. The
effect where the internal columns along grid A do not develop higher axial loads than the
corner column still applies.
Once again, the results can be reviewed graphically to see shear force diagrams as shown above
and bending moment diagrams as shown below. Note how the side and rear walls are now
idealised with a mesh of shell elements.
The member force diagrams for the front transfer beam are very similar to those given by the
mid-pier idealisation. The peak sagging moment has increased from 579 kNm to 584 kNm,
and the maximum end shears has decreased from 336 kN to 326 kN.
Overall the change from mid-pier to meshed wall modelling has had no significant impact on
the design forces developing in this beam despite the fact that one end of the beam interacts
with a wall.
In this case the shear forces at the ends of the beam are almost unaffected by the change for
mid-pier to meshed wall idealisation. The maximum end shear decreases from 416 to 410 kN.
However, over the length of the wall we can see a series of step changes in the shear force
diagram as the wall interacts with the beam. This interaction also appears to have the effect of
smoothing the BMD and the peak sagging moment drops from 1155 kNm to 947 kNm.
Within the length of the wall the wall itself will provide a much deeper compression zone thus
reducing steel tension reinforcement requirements. Within Orion the beam is designed in
isolation ignoring this interaction of beam and wall elements. Orions design for a moment of
947 kNm could therefore still be argued to be conservative.
This situation needs a little more consideration as the wall becomes proportionately longer in
relation to the transfer beam. This is discussed further in the Limitations Transfer Walls
section later in this chapter.
Frame Action
Consider again the moment diagrams for the continuous beam line at second floor level above
the front transfer beam.
The numbers change slightly but the same effect occurs, there is no hogging across the transfer
column positions. As before this effect makes more sense when the deflections for the frame
are viewed as shown below.
Consider the wall at the left hand side of the above model. It stops at first floor level and a
beam is placed under this wall despite the fact that there is a column under each end of the
wall. On the whole there seems to be 3 engineering approaches adopted in such situations:
1. A beam is defined, so the engineer expects the beam to carry all of the wall load.
2. No beam is defined the wall must be designed to act as a deep beam spanning between
the supporting columns.
3. A beam is defined but only for detailing reasons, or perhaps to provide an inbuilt
construction stage support for the wall above. In essence the engineer expects the beam to
be ignored or to carry a very small proportion of the load and that the wall should then be
designed as a deep beam for the greater part of the load.
Orion provides different levels of support for each of these approaches, we will consider each
in turn in more detail.
As was shown earlier in this chapter, if the model is analysed using the mid-pier idealisation of
walls then the load is concentrated at the centre of the wall and the beam is designed for a
moment that would normally be regarded as conservative.
In this case the moment is 905 kNm and the sum of the end shears is 715 kN.
If you want to ensure that a supporting beam is conservatively designed to support all the
loads from a wall above we recommend that you use the mid-pier idealisation. If this is going
to be too conservative an approach then the meshed wall modelling option can be used
provided the points made in the Beam and Wall to Work Together section below are taken
into account.
Once again it is worth noting that even in meshed walls Orion inserts beam elements at each
floor level for numerous reasons discussed elsewhere. In the absence of such elements (and in
particular in more complex models which include transfer walls) the comparison of results
between the meshed and mid-pier idealisation would not be so good.
LIMITATION The model has been solved analytically, but Orion does not automatically
recognise walls which require deep beam design. In the current version it remains
essential that a manual design is carried out for all Transfer Wall Panels.
The moment has dropped from 905 kNm to 263 kNm and the sum of the end shears has
dropped from 715 kN to 562 kN. Given that the mid-pier model generates a single central
point load where a UDL might be a more realistic we should expect the reduction in moment
to be at least in the ratio of WL/4 reduced to WL/8 i.e. something less than 50% for this
single span example. In fact the moment has dropped by nearly 70%. The reduction may not
be as significant for more complex situations with continuous transfer beams.
The stepped shape of the shear force diagram is indicative of the interaction between the shells
and the beam. The drop in moments and shears is an indication that the meshed wall has a
stiffness in its own right and is carrying loads direct to the supporting columns.
In fact, these are probably still conservative design forces for the supporting beam. When
Orion constructs the analytical model of meshed walls it creates relatively rigid beams at each
floor level and where these beams overlap with real beams it replaces the properties of the real
beam with those of a more rigid beam. Therefore the more rigid beam is going to attract a
greater share of the design forces and the design will tend to be conservative.
You have the option to apply adjustments to the rigid beams used in these walls using the
analysis option to adjust the moment of inertia as shown below.
Note that a Modulus of Elasticity adjustment will not affect the merged beam because it retains
the material properties of the transfer beam.
If a 0.2 factor is applied as shown above then the results in the transfer beam change as shown
below.
The moment has dropped from 263 kNm to 152 kNm and the sum of the end shears has
dropped from 562 kN to 478 kN.
Caution
A great deal of effort has gone into establishing the rigid member properties that
are generated by Orion. As is shown elsewhere Orion will generate very
compatible results using either the mid-pier or meshed models when the default
properties are accepted. Applying significant adjustments to these multipliers
could seriously affect the overall validity of the analysis.
In conclusion, by using the meshed wall option the transfer beam and wall will both carry
load. As is shown in this example the design forces developed in the transfer beam are still
likely to be conservative. However, the wall is required to carry some of the load and in this
respect the same limitation as is noted for the case where there is no supporting beam is
applied.
LIMITATION Where meshed walls are supported by transfer beams, the model is solved
analytically, but Orion does not automatically recognise that such walls require some
degree of deep beam design checking. In the current version it remains essential that a
manual design is carried out for all such Transfer Wall Panels.
However, assuming the transfer beam reinforcement can be considered as tension
reinforcement in the deep beam design, it is considered likely that manual design checks in
such situations will tend to show that little or no additional reinforcement is required.
In the above view you can see a wall supported at first floor level but it sits on two different
transfer beams.
If you want to work through this example for yourself, you can load model DOC_Example_07.
In order not to destroy the example for someone else you should then save the model with a
new name before proceeding. The first floor plan view is shown below.
Analysis
You can immediately launch an analysis of this building with the settings as shown below and
using the mid-pier idealisation for walls.
The message is not completely clear as to the problem. Orion is checking to ensure that where
a wall is not supported by at least 2 columns that it is supported by the same beam at both
ends. However, there are many permutations of this sort of support condition and not all will
generate warnings. In fact the analysis will continue to run quite happily. It is better to
understand the issue which can be explained by examining the results.
In this case the wall is symmetrically positioned over a supporting column below. Hence the
mid-pier element of the wall sits directly on this column. This is likely to mean that lower
moments and shear forces develop in the transfer beams to each side of the column. These
forces are shown below for the Factored G + Q combination.
In this case the meshed wall elements will obviously be interacting with the supporting beams.
The forces that develop in the transfer beams are shown below for the Factored G + Q
combination.
The span moment only increases slightly from 145 to 158 kNm. However the hogging
moment increases greatly from 649 to 1374 kNm and the end shears increase from 237 to
789kN.
As was noted earlier in the chapter, these are likely to be conservative transfer beam design
forces due to the rigid link modelling that is used at the bottom of the meshed wall.
However, the wall itself needs to be considered as an inverted deep beam hogging over the
support. Once again this requires some manual check design calculations.
The split walls are defined face to face no wall extension is defined at the interface.
With the edits made, the building analysis can be rerun for each of the wall modelling options.
Using this alternative modelling the analysis will run with no error/warning messages. We can
now examine the effects that this has on the results.
Mid-Pier Model
Now the mid-pier elements of the two separate walls are sitting part way along each of the
transfer beams. The forces that develop in the transfer beams are shown below for the
Factored G + Q combination.
The span moment only increases slightly from 145 to 168 kNm. However the hogging
moment increases greatly from 649 to 1584 which is now higher than the 1374 kNm given by
the meshed wall. The end shears increase from 237 to 865 which is again greater than the
789kN given by the meshed wall.
Meshed Model
The model looks identical to that generated for the un-split wall. In other examples it may
differ slightly because the meshing would not break down to the same sizes. In fact the
properties of the rigid elements created within the walls are related to the dimensions of the
wall so we may see slight differences in the results shown below.
However, in this example the results for the split wall when meshed are identical to those given
when the wall was not split.
Summary/Recommendations
Where a transfer wall has numerous points/lines of support there are two basic options:
Option 1 Do not split the wall
The analysis may give warnings.
In this case the Mid-Pier idealisation may not be appropriate (regardless of whether or not
warnings are given).
The meshed idealisation may be preferable but might not be an upper bound.
An advantage of this method is that the wall is then designed as a single long element.
Consideration should be given to manual design checks regarding the design of the transfer
wall panel as a deep beam.
Option 2 Split the wall
In this case the analysis will start to run without error messages.
The mid-pier idealisation will tend to provide a most conservative upper bound for design of
the transfer beam.
The meshed idealisation will generate the same or very similar results as it did when the wall
was not split.
A potential disadvantage of this method is that the wall is then designed as a series of discrete
elements and some manual adjustment of the design details will be required to reintegrate.
However, for very long walls (particularly those with varying load intensity) this may be
deemed appropriate and in fact something of an advantage.
Consideration should again be given to manual design checks regarding the design of the
transfer wall panel as a deep beam.
Chapter 25
If you want to work through this example for yourself, you can load model DOC_Example_01.
In order not to destroy the example for someone else you should save the model with a new
name before proceeding. This is the same model as was used for illustration of the General
Method in the chapter Transfer Beams General Method.
In order to chase the loads down through the building using the simpler modelling option you
must start at the top floor level and work downwards. At each level you must set/consider the
options as shown above, noting that:
This model has no duplicate floors, duplicates can be accommodated but there are rules to
note, see the section FE Chase Down with Duplicate Floors section later in this chapter
for more information.
Removing the check against the option to Include Slab Plates in FE model means that you
work with a simple stick model as shown later.
Including or excluding beam torsional stiffness will make little difference to the overall
validity of the load chase down. (However, engineers have traditionally calculated forces
in floor grillage systems without allowing for torsion and Orion does not consider torsion
within beam design, so lets exclude it for the purposes of this example.)
Checking the option to Include Upper Storey Column Loads means that the reactions
established by FE analysis of the level above will be reapplied as loads within the model
you create for the current level. It also adds in the self-weight of the columns for the
current level. You can review these loads by clicking on the Column Loads Table button.
Since there will be no slab elements, adjusting the slab stiffness multiplier will have no
effect, you could however adjust the relative stiffness of the beams. Orion uses the gross
sectional area of the beams (ignoring flanges) and columns by default, so you might make
this adjustment to allow for the flanges.
You can run the FE Chase Down manually starting at the top level by making the appropriate
model generation settings and clicking the mesh generation button. This creates the model for
the top level and opens a window so that you can review it. Simply exit from this window to
analyse. Repeat this process for each non duplicate level in the building.
Alternatively you can run the process automatically by clicking on the Batch FE Chasedown
button. The batch method also allows you to review and adjust the mesh at each floor level if
required.
After analysis, options have been added to allow graphical review of the applied loads. The
picture below shows the accumulated dead loads applied at the first floor level in this example
model.
Note that the dead loads shown in the discontinuous columns above are similar, but slightly
higher than the 63 to 68 kN loads determined in the chapter Transfer Beams General
Method.
For this model there are no discrepancies, as you will see in the chapters Overview of Solution
Options for Transfer Levels and Transfer Beams General Method this is not always the case.
The most important comparison when reviewing the results of the FE Chase Down (shown in
the last table in the report above) is with the first table (the un-decomposed slab loads).
Having completed the sequential floor analysis column design forces can be updated by
merging the results from FE with those from the main building analysis. Column results are
merged using the appropriate button on the postprocessor tab of the FE analysis dialog. Lets
look at how the results change before and after merging.
The view above shows column loads before merging. These are the loads from the general
building analysis carried out using the mid-pier idealisation for walls and with rigid zones set
to none.
The view above shows column loads after merging. Obviously the loads on the transfer beams
have increased, but as the axial load comparison shows, the total load is the same at all levels.
Column Design
Having merged the column results, the columns can be designed in the usual way. Note that
you have the option to design them based on building analysis results and then check them
based on the merged FE results. The view below shows the column design summary based on
the building analysis results. Note that the reinforcement and utilisation ratios may be
different on your machine because of different design settings/preferences that you may have
set.
After merging the analysis results from the FE Chase Down and using the batch mode check
design option, the reinforcement is unchanged, but the utilisation ratios change as shown
below.
Some utilisation ratios are higher, some are lower, it is simply a function of the different axial
loads and moments that are generated by the different analytical approaches. Clearly one wall
is now failing. This is occurring since the FE chase down is generating higher moments in the
wall because of the increased bending in the transfer beam that is connected to it. Any failed
members can be filtered out and selectively redesigned.
Note that you could also have used the options to create and save reports so that you have a
complete set of column design calculations, one set based on the building analysis, a second
set based on the FE Chase Down.
When you merge beam results a message box will appear confirming whether or not the
merge has been successful at each level.
It should be noted that when results are merged there will no longer be any pattern load cases
results for the beams.
Note that the beam label is annotated (FE) indicating that the diagrams are based on the results
of FE analysis. Once again, the shear force diagram clearly indicates the presence of high point
loads at the expected positions.
The diagrams are quite similar to those shown in the chapter Transfer Beams General
Method, in this case the shear forces and moments are a little higher (maximum sagging
moment increasing from around 579 kNm to 712 kNm). The left hand end hogging moment
has increased from around 258 to 360 kNm. The transfer of this force into the wall is the
explanation for the wall failure noted above. The maximum shear force has increased from
around 336 kN to 377 kN.
All this is to be expected since the FE chase down accumulates all the loading in the columns
and applies it to the transfer beam. There is no frame action that generates sharing of load
back up to the floor above. In fact, this approach probably more closely emulates traditional
hand calculations.
It is noted that there is currently no (easy) way to account for IL reductions within the
columns that load the transfer beam, in beams supporting columns with many levels above
this means that the transfer beam design will be conservative by default. If it is essential to try
assess the effect of IL reduction this can be achieved by creating a copy model with a new set of
load combinations where the IL factor is reduced from 1.6 to say 0.96 if a 40% reduction is
required. This model can then be only be used for the design of the transfer beams.
Now we can compare the diagrams for the continuous beam line at second floor level above
the transfer beam.
These results may look more in line with initial expectations than those shown in the chapter
Transfer Beams General Method, but what is the correct answer? Some might say that the
sequential nature of construction means that the above style of diagrams is more appropriate
for the self-weight, this would mean that the correct answer lies somewhere in the middle of
the two extremes. This particular model may be an extreme example of this effect, but we
would recommend that the possible effects of load redistribution be quite carefully considered
in any transfer situation.
Once again, the check design mode for beams can be used, so if you wanted to design for one
extreme and then check for the other you can do it as shown below.
Beam Design
The design summary for the beams in this model based on the building analysis results in the
chapter Transfer Beams General Method is shown below.
Having merged the beam results the design and detailing can again proceed as for any other
beam, the summary below shows the changes after running a batch mode check design
without allowing any of the reinforcing to change.
Now lots of the beams fail, so we can rerun the batch design and use the option to increase
steel only when required, if you do that the summary may change as shown below to show that
everything passes.
However, in the batch mode when a beam fails all the steel in the beam is reselected, so if you
un-check the merge option and then check the beams again you might find that beams which
previously passed now fail, as shown below.
For this reason we recommend that having analysed and designed all the beams using one
method (e.g. Building Analysis) you should then run a check design for merged results to find
any members that fail. At this point you should adjust (increase only) the steel in those
members interactively until they pass. Then you should find that you can run a check design
for either set of analysis results and everything should pass.
Finally, it is worth noting that one of the most compelling reasons for not simply using FE
analysis, merging results and then designing for the merged results is that FE analysis does not
deal with patterned load, everything is fully loaded. It is very worthwhile trying to use the
general building analysis and then merge and check for FE results selectively.
Consider the 10-storey model above. The folders indicate that information is defined at upper
levels 10, 9, 6, and 1.
From a beam layout point of view level 9 might be identical to level 10. However there are no
columns above level 10 so the FE models at these two levels will never be identical. Hence it is
always necessary to create information and generate the FE models at the top 2 levels in any
model if an FE Chase Down is to be used.
In the model above 7 and 8 are valid duplicates of level 9.
Level 6 has information, the model takes up an increased plan area and new columns start
from this level.
For the same reasons as at the top two floors, level 5 is not a valid duplicate of level 6. If you
try to work through an FE Chase Down for this model then you will get a message as shown
below when you try to create the model at level 1.
By copying storey information from level 6 to level 5 this model will have information at
enough floors for an FE Chase Down to be completed.
Note that information is always required at level 1 and therefore as an absolute minimum in
any model information is required at 3 levels, Top, 2nd Top, and First.
Chapter 26
If you want to work through this example for yourself, you can load model DOC_Example_01.
In order not to destroy the example for someone else you should save the model with a new
name before proceeding.
This is the same model as was used for illustration of the other methods in the previous
chapters. There is no need to make special changes in support of this new method. Refer to
the opening notes in the chapter Transfer Beams FE Method, Option 1 (Simplest) for an
overview of the problem.
In order to chase the loads down through the building you must once again start at the top
floor level and work downwards. At each level you must set/consider the options as shown
above, noting that:
This model has no duplicate floors, duplicates can be accommodated but there are rules to
note, see the section FE Chase Down with Duplicate Floors at the end of the chapter
Transfer Beams FE Method, Option 1 (Simplest).
Checking the option to Include Column and Wall Sections in the Model means that a more
sophisticated model which idealises a rigid zone extending to perimeter of the column or
wall is used. This modelling sophistication is really only of interest in flat plate models and
may introduce more meshing difficulties. For beam and slab models it is better not to
activate this option.
Checking the option to Include Slab Plates in FE model means that you will need to mesh
up the floor slab as shown later.
Including or excluding beam torsional stiffness will make little difference to the overall
validity of the load chase down. (However, engineers have traditionally calculated forces
in floor grillage systems without allowing for torsion and Orion does not consider torsion
within beam design, so lets exclude it for the purposes of this example.)
Checking the option to Include Upper Storey Column Loads means that the reactions
established by FE analysis of the level above will be reapplied as loads within the model
you create for the current level. It also adds in the self-weight of the columns for the
current level. You can review these loads by clicking on the Column Loads Table button.
The theoretical stiffness of the slab relative to the beams can make a big difference to the
load paths and influence the design moments determined in the beams. Refer to the
chapter Analysis and Design using FE for more detailed discussion of this effect. For this
example set the slab stiffness multiplier to 0.2. At the end of this chapter we will also look
at the effect of changing this setting for this model. You could also adjust the relative
stiffness of the beams. Orion uses the gross sectional area of the beams (ignoring flanges)
and columns by default, so you might make this adjustment to allow for the flanges. In
this example leave the beam stiffness multiplier set to 1.0.
At each level in turn you click on the mesh generation button. This takes you into the FE
Preprocessor where you can mesh up your model. You can use the mouse controls to zoom/
pan/rotate the model until you see a view as shown below for the third floor level.
When you first access any model Orion will suggest a default number of plates. Meshing is not an exact science, to a large degree the more plates the better, but as a rule of
thumb you should aim to see at least 6 to 8 nodes generated along the length of each
beam. In this model changing from the initial default to 600 plates, the resulting mesh
is as shown below.
Note
Click buttons in the menu bar to toggle on and off the display content.
You can keep changing the suggested number of shells and re-mesh to review the result until
you are happy. Since we have 6 nodes along the length of the middle beam at the front we will
use this mesh density.
Refer to the chapter Analysis and Design using FE for more discussion on mesh density in FE.
When you exit from the preprocessor the model of this floor will be automatically analysed.
Repeat the modelling options and meshing at the second and then the first floor levels.
Alternatively you can run the above process for all floors automatically by clicking on the
Batch FE Chasedown button. The batch method also allows you to review and adjust the mesh
at each floor level if required.
After analysis, the postprocessing option allows graphical review of the applied loads and
other results. The picture below shows deflection contours together with the applied nodal
loads for the Dead Load case.
Note that the dead loads shown in the discontinuous columns above are quite similar to those
determined in the chapter Transfer Beams FE Method, Option 1 (Simplest).
Within the FE Post Processor you can of course also review contour diagrams for all sorts of
shell results. Once again you should refer to the chapters Analysis and Design using FE and
Flat Slab Models later in this handbook for more detailed discussion.
For this model there are small discrepancies between the loads developed in the FE Chase
Down (shown in the last table in the report above) when compared with the sum of applied
loads in the first table (the undecomposed slab loads).
These differences are explicable and unavoidable. The FE model is a traditional centre-line
analysis model. For the meshed up model the loads are applied to the plates, the plates are
then supported by the beams and columns. The model is constructed on outer grids spaced at
5 m by 8 m, so the total plan area for the centre-line model is 40 m 2.
P
Consider the imposed loads: the imposed load at each level is 2.5 kN/m 2, so the load at each
floor is 200 kN. However, when Orion calculates beam loads using the yield line
decomposition method in preparation for the general building analysis it takes account of the
size of the beams and adds in imposed loads on the outer half of the perimeter beams. So in
this example the FE Chase Down finds a little less imposed load than the building analysis.
P
Consider the dead loads: in the FE analysis the slabs exist right up to the centre line of the
beams. However, when Orion calculates beam loads using the yield line decomposition
method in preparation for the general building analysis it takes account of the size of the
beams and only considers the self weight of the slab up to the face of the beam. So in this
example the FE Chase Down effectively double counts a small amount of slab weight where
the slab and beam overlap and hence finds a little more dead load than the building analysis.
Using either of these methods it is quite likely that Orion has been more rigorous in
decomposing and maintaining loads than most Engineers would have deemed necessary in
hand calculations or more traditional approaches.
A more detailed discussion of the Axial Load Comparison can be found in the chapter
General Building Analysis.
Having completed the sequential floor analysis column design forces can be updated by
merging the results from FE with those from the main building analysis. You can then design
columns or check previously designed columns for these new slightly different loads. This
procedure is identical regardless of whether the floors have been meshed and can therefore be
followed for this meshed up version of the model by referring back to the section Merging
Column Analysis Results in the chapter Transfer Beams FE Method, Option 1 (Simplest).
Note that the beam label is annotated (FE) indicating that the diagrams are based on the results
of FE analysis. Once again, the shear force diagram clearly indicates the presence of high point
loads at the expected positions.
The diagrams are quite similar to those shown in the previous chapter, in this case the shear
forces and moments are a little higher (maximum sagging moment increasing from around
712 kNm to 741 kNm). The left hand end hogging moment has decreased from around 360
kNm to 306 kNm. The maximum shear forces have also decreased slightly.
The notes in the chapter Transfer Beams FE Method, Option 1 (Simplest) regarding the way
these FE chase down methods concentrate the loading in the lowest beam are equally
applicable here.
We can again compare the diagrams for the continuous beam line at second floor level above
the transfer beam.
These results may look to be in line with initial expectations but you should refer again to the
discussion in the chapter Transfer Beams FE Method, Option 1 (Simplest) for an alternative
view on this.
Beam Design
Having merged the beam results you can then design the beams or check previously designed
beams for these new slightly different loads. This procedure is identical regardless of whether
the floors have been meshed and can therefore be followed for this meshed up version of the
model by referring back to the Beam Design section in the chapter Transfer Beams FE
Method, Option 1 (Simplest).
However, lets now look at the beam design forces that will get merged. The member force
diagrams for the front transfer beam are shown below.
The hogging and sagging moments on the transfer beam have both reduced slightly. The
shears at each end of the beam have also dropped. In this example the differences are not very
dramatic, but in other examples you may well find that they can be.
The effect that is occurring is that the slabs are carrying a proportion of the load straight to the
supporting columns and by stiffening the slab the effect gets bigger. Once again you should
refer to the chapter Analysis and Design using FE for a more detailed discussion of this and
other aspects of FE analysis.
Chapter 27
If you want to work through this example for yourself, you can load model DOC_Example_02.
In order not to destroy the example for someone else you should save the model with a new
name before proceeding.
All the beams have been removed at first floor level and a deep flat (transfer) slab has been
created.
There are a few points to note about the modelling of this simple layout that can often have
greater implications in larger and more complex models:
Slab Insertion
The default slab insertion method is Beam Region in which case Orion will look for a closed
perimeter of beams around the area in which you click and insert a slab extending to that
perimeter. This option does not work where there are no beams.
For Flat Slab models you need to change the insertion method to either Axis Region or Pick
Axes.
Using Axis Region Orion will look for a closed perimeter of axes (grids) around the area in
which you click and insert a slab extending to that perimeter. If you did that for the model
shown above you would end up with 3 slabs. That would not necessarily be a problem, but one
slab will do. Using the axis region method you can merge these slabs as you insert them if you
press and hold the control key while you click in each of the 3 areas.
For more complex boundaries you might use the Pick Axes option in which case you pick the
axes that define the boundary of a slab by clicking on them one at a time.
Overall it is better to have larger and hence fewer slabs and to avoid small or thin infill strips.
This will generally result in better meshing in the FE module, refer to the next chapter
Analysis and Design using FE for more guidance on this.
Building Analysis
Run the building analysis as described in the Modelling and Analysis section in the chapter
Transfer Beams General Method. During the analysis errors will be displayed as shown
below, but you will have the option to continue.
The first message shown below lists columns and walls that the system regards as
unsupported. This basically means columns and walls that stop at an upper floor level and
which do not sit on a beam.
Note that the error suggests that the analysis results will not be reliable this cannot be
ignored if you expect to use the building analysis results for design. In the case of any flat slab
structure and in particular transfer slabs you will use FE Chase Down analysis so you will not
going to be relying on these results for gravity load design, press Yes to continue.
At the end of the analysis you should see an axial load comparison warning.
The first warning is telling us that only 35% of the applied load is being applied successfully
decomposed onto beams and hence included in the building analysis. Essentially this is also
telling us that the building analysis results are of little value.
At this point it is worth examining the structural model that has been created and analysed.
The problem is very obvious, the two front columns and the wall at the rear do not sit on
anything, so the loads in these frames actually reach the end wall and columns by virtue of
beam and vierendeel action at the upper level.
This is not a mechanism, it is potentially a valid structure, but it is ignoring the transfer slab
and so it is not the structure we want to consider here.
Band beams are defined along grids A and B. Support band beams are in a sense fictitious
beams that allow you to generate a coherent building analysis model but these beams are not
passed to the FE model, so they do not affect the FE results.
For this particular model the edits shown above mean that the Building Analysis will run with
no errors or warnings and the axial load comparison will be perfect no missing loads.
This may initially seem to be the best option for dealing with transfer slabs, and for simple/
regular models it probably is. However, most of the real models we see, and especially those
involving flat slabs and transfer slabs, are not regular. It can often be very difficult to insert a
logical system of band beams.
After analysis use the options for graphical review of the applied loads. The picture below
shows the accumulated dead loads applied at the first floor level in this example model
together with the deflected contours for that case.
Although a little difficult to see in the above view, the column loads are the same as
determined in the chapter Transfer Beams FE Method, Option 1 (Simplest). This is as
expected, since the 3rd and 2nd floor FE models are identical in this chase down, and the
analysis at those levels is unaffected by what goes on at the first floor.
U
More importantly, after analysis, you can see moment and deflection contours for the transfer
slab that take account of the supported column and wall loads. At this point we enter into the
general topics associated with FE analysis and Flat slab design which are described in much
more detail in the following two chapters, Analysis and Design using FE, and Flat Slab
Models.
For this example we will look quickly at some of the results, which can quickly be exposed.
In the view above we can see a contour diagram in the background showing design sagging
(tension in bottom) moments generated in the longer span direction of the slab. On top of
that is a section view showing design moments on a line (red dots along the centre of the strip)
cut in the same direction. Note the high local moment at the column head at the right had end
of the line.
By changing to an Integral strip, we can see the results captured and averaged over a specified
the width of strip, in this case a 1 m wide strip at the edge of the slab. This shows how the high
moment at the RH end is a local effect and in fact higher average hogging moments are
generated at the slab to wall boundary at the other end.
This next view shows the top and bottom reinforcing requirements averaged over this 1 m
wide strip. The view below shows As required contours for bottom steel in direction 1.
Note
While in some respects a deep slab such as illustrated in this example can be
treated as any other flat slab it is noted that such slabs may well also require
some different/extra design and detailing considerations in areas such as
deflection, punching, and perhaps reinforcement bundling. It is recommended
that the details of such important structural elements be thoroughly cross
checked.
For this model (before introduction of any band beams at first floor level) we still have the
discrepancies occurring during load decomposition and building analysis that are discussed
earlier in this chapter. However, the results for the FE Chase Down (shown in the last table in
the report above) compare perfectly with the sum of applied loads in the first table (the
undecomposed slab loads).
This means that the FE Chase Down procedure appear to have been successful and we can
now continue to merge column and beam design forces.
Having completed the sequential floor analysis column design forces can be updated by
merging the results from FE with those from the main building analysis. You can then design
columns or check previously designed columns for these new slightly different loads. This
procedure is identical regardless of whether the floors have been meshed and can therefore be
followed for this meshed up version of the model by referring back to the section Merging
Column Analysis Results in the chapter Transfer Beams FE Method, Option 2.
Chapter 28
Introduction
When tackling the subject of flat slabs some fundamental issues routinely arise:
The building codes provide simplified methods of dealing with regular flat slab models. In
practice a large proportion of models do not seem to fit within these limitations and so
engineers turn to more advanced FE methods.
Having turned to more advanced FE methods in order to look at models with more irregular
geometry, engineers are often faced with FE results that seem unreasonable to them.
Then there is the need to consider compliance with deflection limits.
These topics and more are all discussed in detail within a training presentation that is available
in PDF format alongside this document and accessed via the link below.
Orion Flat Slab Training.pdf
(If the link does not work please browse to find the file name indicated above in the HELP
sub-folder of the Orion Program Folder).
The above presentation extract is given and discussed in detail during Orion Advanced
Training Days. The extract assumes that the content of the chapter Analysis and Design using
FE of this handbook has already been covered.
The remainder of this chapter focuses on a practical example in which the techniques
described in the presentation are put to use to design an irregular model.
Orion is also able to undertake the Punching Shear Checks associated with flat slab design.
The implementation of these checks and associated limations are discussed in the succeeding
chapter.
made use of Orions Band Beam option which models fictitious beams within the slab.
However, it must be accepted at the outset that this form of design will require more
thoughtful, interactive, and time consuming design involvement by the engineer.
Consider the model shown above. A core wall system is assumed to brace the structure and
therefore the flat slabs and peripheral columns are to be designed for gravity loads only.
Although it looks like a simple layout, it would actually be very difficult to apply the
idealisation of column and middle strips and hence it becomes almost impossible to apply the
codes simplified design methods.
Ensure that the concrete material properties are adjusted so that a long term E value is used, in
this case C40 concrete is used with E set to 7000 N/mm2. At the 4th floor level use settings as
shown above.
You could adjust the mesh density from the default down to a lower setting that you think
might be acceptable. However, using the default 1500 shells and setting the regularity factor to
1.0 produces a mesh as shown below which is clearly more than adequate.
Analyse the model and then access the postprocessor setting the Positive (sagging) Moment
Factor to 1.2.
Deflection
Estimated total deflection is peaking at around 70 mm along the perimeter lengths between
columns. The distance between these columns is about 11.3 m therefore we should be looking
to see total deflection restricted to around 11300/250 = 45 mm. A 300 thick slab is probably
not adequate for these long edges. However we will continue with the example unaltered.
Bottom Steel Reinforcement Provision
Before starting to review reinforcement requirements right click in the graphics area and set
the effective depth information to some reasonable values.
We will start by assuming that the whole slab is to be reinforced orthogonally in the Global X
and Y directions.
At this point the slabs reinforcement angles have not been adjusted, so direction 1 steel is
aligned with global X, which runs from left to right in the view below.
This view shows bottom steel requirements in direction 1 and as might be anticipated the peak
requirements are occurring along the longest free edges. In direction 2 (below) we see a very
similar situation and a very similar peak requirement of just over 1300 mm2/m.
In practice you may want to consider the reinforcing requirements if the main steel along the
angled edges is aligned to the edge. This can be done and will be shown in the Additional
Notes on Bottom Steel Provision section at the end of this example.
We can now use the custom contours option to show where different steel reinforcement
provisions would be adequate. In general the strategy would be to decide on some general
lower level of reinforcement to be provided continuously throughout the slab and identify the
regions where an increased provision is required.
In the view above we have set the lower general provision to T12 at 200 and we can see that
this is sufficient over a large proportion of the slab.
We have then considered the possibility of laying in extra T16 at 200 (providing T12 and T16
alternate bars) and the view shows that this is adequate everywhere else.
These requirements can be simply communicated to the detailer by exporting the above
contours to the main graphical editor and then to DXF.
The same procedure would be repeated for the bottom steel in both directions.
As expected, a review of the top reinforcing requirements shows that in theory no steel is
required over large areas of the slab and that the hogging moments intensify rapidly over the
column heads and the core wall. However, many (most?) engineers would tend to provide
minimum reinforcement throughout the top of an irregular flat slab and so the user-defined
contours shown below show this as the minimum level.
In the view below a 2 m wide strip is cut across the head of the most critical internal column.
Based on the proportions and span of the slabs involved this might be considered as a
reasonable strip width and the average steel requirement in this width is 1772 mm2/m.
However, if this steel were provided, it must be provided over at least the 2 m width of the
strip. Clearly this strip strays beyond the contour boundaries where a much lower steel
requirement has been shown to be adequate.
This strip is cut just within the boundaries where the lower provision of T10@200 + T16@200
is shown to be adequate. The average requirement in this narrower strip has increased to 2187
mm2/m. (This is also more in line with the rule of thumb expectation that the peak hogging
steel provision will be in the order of double the peak sagging steel provision).
Peak requirements tend to occur around the ends/corners of walls, and a little unusually in
this example the peak requirement at the wall corner is actually higher than at the column.
Once again there would be arguments for cutting a wider strip and determining a lower
average steel requirement in the vicinity of the wall, but given that we are working with a
standard 200 spacing elsewhere, the choice here lies between T20@200 and T25@200 clearly
T25@ 200 are required.
In order to achieve standardisation throughout you may then choose to provide T25 at 200 as
a standard patch of reinforcement across all column heads. Once again this information can
be communicated to the detailed by adjusting the user defined contours as shown above. The
yellow patches indicate the minimum zones in which the peak reinforcement is required.
However, it would probably be good practice to apply a standard minimum patch throughout,
in this case something like a minimum of 8 T25@200 across all column heads. To be effective
these bars must also extend a full anchorage length beyond the point where they are required.
It was noted that for this slab some of the highest sagging moments and hence the greatest
bottom steel requirements occur along the angled slab edges. This requirement was being
established on the assumption that direction 1 and 2 steel will be provided in the global X and
Y directions throughout. The steel requirement at this point is high in both directions.
If you check you will find high design moments in both directions at these points, and if you
look at the unadjusted moments you will find that Mxy is very high and that the Wood and
Armer adjustment is having a big impact on the design moments for reinforcement if it is not
going to be provided parallel to the free edge.
Logically you would expect that steel provided parallel to the free edge would be the most
efficient solution in an area like this.
When you cut a strip at an angle as shown above you will automatically be looking at moments
and hence reinforcing requirements along the cut line.
Alternatively you can define the intended reinforcing angles in any/every slab independently
and view contours on that basis. In the view below, you can see how the reinforcing angle has
been reset in the four panels adjacent to these edges.
After re-meshing and reanalysis, you will see contours as shown below.
The direction 1 steel area requirements are shown above appear very similar to the
requirements shown earlier before this adjustment was made. The direction 2 requirements
shown below are quite different. Direction 2 is perpendicular to the free edges, and as
expected the reinforcement requirement becomes nominal at the edges where a peak was
previously exposed.
In other examples the feature that allows contours to be simultaneously displayed relative to
different axis systems can seem a little strange until you get used to them, refer back to the
Review of Contouring Options in the chapter Analysis and Design using FE for more detail
if necessary.
X
UX
In terms of the weight of reinforcing provided, it would be more efficient to design a perimeter
ring and then place orthogonal infill steel, but this may not be the fastest/simplest or even the
cheapest construction solution. Using Orion as shown above you can quite easily investigate
such options and you could even create extra slabs to define a ring beam zone.
As was noted in this example and in the training presentation, it is relatively easy to generate
the design information (the information that can be passed to a detailer) in Orion. This
example should reinforce the view that for flat slabs it may be better to restrict your use of
Orion to this level of detail and then revert to traditional detailing.
Column Design
By following the procedure for sequential FE floor analysis and then export of column design
forces described in the chapter Transfer Beams FE Method, Option 2 column loads can be
made available for design.
The view above shows that axial loads and moments are transferred from the FE analysis
(short frame model) and that the braced columns can now be designed in Orion including for
IL reduction factors if desired.
Since there are no beams attached to these columns Orion assumes a small strip of slab is
effective as the beam and calculates effective length factors on this basis.
Chapter 29
Introduction
In Orion R13.0 a basic interactive punching shear check was introduced. Since then some
enhancements have been made which we believe have made it a more attractive design aid.
However, the nature of the design code requirements tend to dictate that punching shear
checks can not be safely automated in batch runs. For this reason Orions punching shear
check is an interactive design tool, you need to use it carefully on one column at a time to
review/establish the correct design settings for each column. Having done so rechecking can
be carried out in a more automatic fashion.
In this chapter we will show what this check can do for you and illustrate the limitations of
which you need to be aware.
The check fails but the required area of shear reinforcement can be calculated and the
checking procedure moves on to the next check perimeter.
The check passes in which case no shear reinforcement is required and no further checks
on subsequent perimeters are required.
Note
Simple Examples
In this section we will look at a series of relatively simple examples using the model shown
below.
However, although it looks simple, this model is sufficient to illustrate most aspects of the
checking procedure including:
A typical internal column.
For circular columns, the same principle applies - BS8110 only considers idealised
rectangular perimeters so checks should be made using the rectangular as opposed
to the circular setting.
Check Column Perimeter and Check Column Drop Panel Perimeter (only active if a drop
has been defined) These options allow the check on maximum shear capacity at the face
of the loaded area to be toggled on and off.
Periphery Reduction Amount An allowance for holes that have not been modelled or
may be cut in the future see the section on Dealing with Openings.
Include Load Within Punching Perimeter When this option is checked the shear load
considered on each successive punching perimeters will reduce (due to the increasing area
of loading that falls within the punching perimeter. When un-checked the shear load will
be the same on all perimeters i.e. the check becomes more conservative.
Column Location
As will be illustrated in the following examples, it is the users responsibility to indicate the
location (internal/edge/corner) and hence the more applicable shear checks as detailed in
BS8110 cl 3.7.6.2 and 3.7.6.3.
If no changes are made and the column is checked with the settings as shown the result is as
shown below.
The output indicates the design forces and then makes the first punching check (at the face of
the loaded area). For this and most reasonable examples this check passes.
The red perimeter is indicating that the checks then fail at the next perimeter see below.
Check Shear on a Series of Perimeters
Continuing with the same example, the code requires the first perimeter check to be made at
1.5 d from the face of the column (the loaded area). Note that Vt is slightly reduced for this
perimeter, from 1139.5 kN at the face of the column to 1112.8 kN (because the option to
include the load within the punching perimeter is active i.e. we have included its exclusion.)
The output above indicates that v > 2vc reference should be made to BS8110 cl 3.7.7.5, once
v exceeds this level justification of shear resistance moves beyond the limits of the code.
Hence in this situation Orion simply indicates the perimeter in red on the plan view and the
output suggests that something must be done to increase vc (or reduce v).
The shear capacity vc is influenced by concrete grade and provided steel. For this check we
accepted the default bars T8@250. Clearly a much higher steel provision would be normal
across a column head so we can now rerun the check with more realistic steel say T20@200
(probably still a little low).
The 3rd perimeter at 3.0 d is checked and it passes so no further perimeters need to be
checked.
The plan view reflects these key results the first two perimeters are shown red and the Asv
requirement is written on each perimeter. The 3rd perimeter is shown green. This sort of red,
red, green series of perimeters is ultimately what you will see against all of the columns.
Note
Since the punching shear perimeter length increases and the shear load decreases
on each successive perimeter you may expect to see a reducing Asv requirement on
the successive perimeters. For higher shear loads this will be true, however, where
nominal shear reinforcement is required
(Asv min = 0.4ud/0.95fyv) the requirement will increase as a function of
perimeter length (u).
Note
The input of main reinforcement has 2 effects. The bar size influences the value
of effective depth (d) used in the checks, and the spacing sets the total effective
tension steel (Ast). In practice you might actually have alternate bars of different
sizes providing a total Ast (say 1850 mm2/m for the purposes of an example). In
this case you would simply use the larger bar size and adjust the spacing so as to
achieve the correct Ast value as shown below. It is the Ast value that is
important, the spacing is irrelevant in the context of this check.
Note that on 2C10 the very slightly higher punching load is sufficient to mean that the Asv
requirement on the first perimeter is greater than the minimum requirement which applies on
2C9. On the second perimeter however the minimum requirement applies to both columns.
In practice punching shear resistance may be provided in different ways using various
proprietary systems. If it is to be provided using links, the detailing requirements of BS8110 cl
3.7.7.6 and fig 3.17 must be adhered to. This remains a detailing exercise but the following
notes indicate the points to watch continue to consider column 2C9:
Check Perimeter at 1.5 d
The 1088 mm2 required at the first check perimeter should be provided on at least 2
reinforcement perimeters:
At 0.5 d at least 40% i.e. approx. 435 mm2
At 1.25 d the balance of approx. 653 mm2
Check Perimeter at 2.25 d
The 1445 mm2 at this second check perimeter must again be provided on 2 reinforcement
perimeters the first of which is at 1.25 d and is therefore common to the previous check
perimeter.
At 1.25 d at least 40% i.e. approx. 578 mm2 but 653 mm2 already provided above.
At 2.0 d the balance of approx. 1445 653 = 792 mm2 is required.
Now consider the detailing requirements on each perimeter. The main restriction to note is
that the maximum spacing of link legs along any perimeter is 1.5 d which in this example is a
spacing of 380 mm. The spacing of main bars is likely to further dictate the maximum spacing
that can be used say main bars are at 100 crs then maximum reasonable spacing is 300 mm.
The table below indicates the minimum area of shear reinforcement that will be provided on
each perimeter if a 300 mm link leg spacing is used.
Perimeter at
(distance
from face of
column)
Minimum
Asv reqd
mm2
Perimeter
Length
(mm)
Min No of
Link Legs
(at 300 mm
crs)
0.5 d
435
2620
1.25 d
653
2d
792
T8
T10
254 mm2
452 mm2
707 mm2
4150
14
396 mm2
704 mm2
1100 mm2
5680
19
537 mm2
955 mm2
1491 mm2
There are some further points to consider in relation to the provision of shear
reinforcement around holes. See further comments in the section Dealing with
Openings.
In the above view a column close to an edge has been selected and a check has been made
using the same settings as were used for the internal columns above.
The view shows how Orion has correctly detected the shortest punching perimeter in
accordance with BS8110 figure 3.19, however, clauses 3.7.6.2 and 3.7.6.3 require that a
distinction is made between internal and edge columns. This affects the calculation of Veff in
both directions. Orion cannot make this distinction automatically for all potential geometries,
and so it never attempts to make it at all. THE USER MUST ALWAYS CONTROL THIS
SELECTION.
In this simple case the red and green perimeters clearly indicate that the free edge is close
enough that this check should be made as an edge condition on all perimeters. If we rerun the
check having changed to the appropriate column location setting the result changes as shown
below.
The result looks similar, but the Asv requirements have actually changed. In actual fact in this
example the requirement on the first perimeter has reduced, and on the second perimeter min
requirements continue to apply. In fact, because of the way Orion calculates the dimension x
used in the calculation of Veff (see later discussion) this is likely to be the case, but this result
cannot be guaranteed for all edge conditions.
As an example of where some care needs to be taken with the selection of edge conditions,
consider the column shown below.
The first perimeter is that of an internal column, the second is an edge condition. When
checked as an internal column the Asv requirement shown above is 1088 mm2 (min
reinforcement). When checked as an edge column the calculation of v changes, but the
minimum Asv requirement continues to apply i.e. the result is unchanged in this example.
Beyond the requirement to consider and apply the appropriate check there is no difference
between the checks for internal and edge columns.
The example shown above illustrates a simple corner column case. The points made for the
edge column in the previous section are equally applicable here, the obvious difference being
that the corner column location must be selected. In this example, the corner column with
significant overhangs passes the punching shear check with no reinforcement being required.
In the view above column 2C10 has been selected and a 2m square by 500mm deep drop panel
has been defined at the top of the column.
The purpose of a drop panel is typically to remove the need for additional punching shear
reinforcement. This can be achieved by adjusting the geometry of the column drop panel.
When the checks are performed the objective is therefore to obtain two green perimeters as
shown above. The inner perimeter (at 1.5d from the column) confirms that the drop is
sufficiently deep and the outer (at 1.5d from the drop) confirms that it is sufficiently wide.
Note
Depending on the size of panel, it is possible for the first perimeter at 1.5d to
occur outside the drop panel. The calculations are then performed at this
perimeter based on the increased slab depth. however, in such a case the
calculation becomes irrelevant and can be ignored.
Note
For an output report to be created for the results around the drop panel
perimeter (as shown above), the option Check Column Perimeter should be left
unticked.
If both options Check Column Perimeter and Check Column Drop Panel
Perimeter are ticked, all the required checks are performed and the graphical
display of the perimeters is created accordingly, however, the accompanying
report will only display the results around the column perimeter.
In the view above a small hole has been positioned close to a column. On the punching
perimeters that are drawn you can then see sections in black (rather than red or green)
indicating the lengths on the check perimeters that are excluded. The excluded length is
established as the length between lines radiating from the centre of the column past each of the
extreme edges of the hole.
In actual fact this is quite a small hole (150 mm square) and is actually something we would
not recommend complicating the FE modelling with. We would tend to recommend that you
allow for something like this by using the Periphery Reduction Amount option. Before moving
on to that, it is worth noting just how significant the reductions due to this (any) small hole
can be.
4. At 3.00 d from face The perimeter length is reduced by 651 mm (from 7720 mm to 7069
mm).
As expected the reduction length is increasing. The main point to note is that a small hole of
150 mm side quite close to the face of the column reduces the punching perimeter by much
longer lengths.
Note
Note
Due to the almost infinite variations of intersecting slab shapes and hole shapes
and positioning, it is quite possible that Orion will have difficulties accurately
predicting punching perimeter reductions, this point is illustrated in the
concluding notes.
Internal Columns:
For bending in both directions:
Veff = Vt (1 + 1.5 Mt/(Vt*x))
(eq25)
Edge Columns
Where bending is about an axis parallel to the free edge:
Veff = 1.25 Vt
(eq26)
Corner Columns
For bending in both directions:
Veff = 1.25 Vt
BS8110 defines x as the length of the side of the perimeter considered parallel to the axis of
bending. There is no guidance as to whether or not this dimension should be shortened if
openings affect the perimeter. There is no guidance on what to do when the length is
shortened on one face but not on the other. Where opinions have been sought on this there is
a view that x should not be shortened to account for openings, however no authoritative
guidance has been found to this effect.
Looking to the above example, for direction 1 bending on the first perimeter one side is
shortened by 393 mm. The most conservative view would be that x should be taken as 1165 393 = 772 mm. In Orion x is taken as the average of the two relevant sides, in this case (1165 +
772) / 2 = 968 mm. This is believed to be a reasonably conservative interpretation of the code
requirements.
2. Despite the above, for the first perimeter the calculated Asv requirement is identical. This
is because the design in direction 2 is dominating and in that direction x was not reduced
when the hole was modelled.
3. For the second perimeter the Asv requirement is a little higher because using this option
the length reduction is constant on all perimeters (hence it is 393 mm rather than 522
mm). On this perimeter min Asv requirements are being used and so the requirement
increases as the perimeter length increases. Given that we are only allowing for the
possibility of a hole, we should check the requirements when no hole is allowed for as
shown below.
In essence the above is exemplifying the note made earlier regarding the practicalities of
providing the shear reinforcement. If the Asv requirements obtained by excluding a perimeter
length are factored up by the ratio of the full perimeter to the reduced perimeter so that
reinforcement can then be provided on the full perimeter a very reasonable result will be
achieved, perhaps erring on the side of conservatism.
In conclusion, for this model if an allowance had to be made for the possibility of holes (up to
say 150 mm) adjacent to any column, it would be quite reasonable to specify a periphery
reduction of around 400 mm to all columns rather than take the time to attempt to model
every possible hole in detail. The reported Asv requirement might then be factored up slightly
to allow for the point made above.
When accessed this option will show a table where the key design information can be reviewed
and adjusted if necessary.
So for the final run on this model a perimeter reduction of 400 mm is applied to all columns.
Concluding Notes
Limitations
Three points emphasised throughout this chapter are:
1. The punching shear checks required by design codes are quite empirical by nature and
engineering judgement is often required as to which part of the checks are most
applicable.
2. Potentially there are an almost infinite number of variations of intersecting slab shapes,
hole shapes, discontinuous edges, etc. It is quite possible that Orion will have difficulties
accurately predicting punching perimeters in many circumstances.
3. As a result, the checks must initially be carried out on one column at a time so that the
engineer can take proper control over the input and review the results.
The areas in which we would particularly highlight the limitations of the current checking are:
Holes
It is possible that Orion will have difficulty identifying correct punching perimeters as the
geometry of slabs and openings become more complex. The example below shows the sort of
issue that could occur.
Two holes are positioned near to a column. Orion has identified the circular hole outside the
check perimeters and has reduced the perimeters accordingly, the black sections of the
perimeter lines can be seen on close examination.
For the rectangular opening however, reduced lengths are also shown that extend above and
below the edges of the opening. The above view was created using an older release of Orion.
However, there is some debate as to whether these extended exclusions are actually required,
and in fact introducing these extensions was found to cause other difficulties.
The view below shows the punching perimeters that are now generated in the current release.
There are no extensions to the excluded lengths where a check perimeter is actually broken by
a hole, the splayed reduction of length is only applied to unbroken perimeters. Note that an
additional periphery reduction is still applied this value could be further increased if you
wished to make more allowance for extended exclusions at the sides of the hole.
If you identify or are concerned about problems of this nature we would appreciate being sent
the example so that we can look into the matter. In the meantime we would suggest using the
options to either simplify the holes or to remove them and specify a periphery reduction
length.
Dimension x used at the face of the loaded area
At the face of the loaded area you will find that value of x used is the same as at the first
check perimeter. It is not taken as just the column dimension. The justification for this
assumption is discussed below:
We found that when you do the check using the actual column dimensions then columns in
fairly ordinary situations start to fail. When we talked to people in the industry about this we
found that in practice they tended to ignore this check completely (assuming it to be ok by
inspection) or if they did it they just used the actual shear force and did not bother to amplify
it at all, or sometimes that they just assumed the 1.15 amplification noted in Clause 3.6.7.2
would cover it. Everyone felt that the check seemed incorrect (over conservative) when
applied more rigorously using the column dimensions as "x" at the face of a column.
Basically as you approach the column "x" decreases rapidly and so the amplification calculated
in Veff becomes extreme. Internal columns with ordinary transfer moments are treated as
worse cases than edge columns with very high transfer moments where the amplification of
Veff is a fixed value of 1.25Vt. It also seems to work back to front in terms of the way it
amplifies V - for example using a 600x300 column, if the same transfer moment is applied in
both directions the amplification of Veff will be greater when bending is in the strong
direction (because x=300) than when bending in the weak direction - seems counter intuitive?
Furthermore there is the complication that x is unquantifiable for irregular and even circular
column perimeters.
After investigations at CSC and external consultations, the conclusion arrived at was that
making the same amplification as at the first check perimeter seemed reasonable and probably
more conservative than what had previously been done/assumed in practice for many years.
Slab Merging
For Orion to be able to resolve the punching perimeter around any column all of the slabs
inserted around that area need to be merged into one. For simple geometries this is not a
challenge. For more complex cases it is a significant challenge where improvements have been
made on an ongoing basis.
Where difficulties are encountered the floor slab merge options may help to either resolve the
issue completely, or isolate the problem so that the modelling can be adjusted.
The examples below are exaggerated in some respects but they do show how the features work.
Notice that some of the punching perimeters in the slab area shown above are clearly
incorrect. The option to Hatch Floor Area can be used as shown below and this shades all the
slab areas that Orion has managed to merge.
Clearly one of the slabs has not been included, the reasons why this might happen vary, but
usually become apparent if the boundary of the slab is examined in close up. In this example
the slab and some of the adjacent small strips do not meet properly, the slabs can be redefined
and the merged floor area and punching perimeters are all then resolved as shown below.
Note
In general it is good practice to avoid thin strips of slab and have fewer slabs
wherever possible. Not only does it avoid the sort of problem shown above, it is
also less likely to cause meshing difficulties during the FE analysis.
Tip
The Slab Merging option can be used at any stage during model creation,
therefore you can check your model for potential problems as you create it rather
than finding problems later on when much more editing (and reanalysis) may be
required to eliminate them.
A second typical problem occurs where lots of grids meet at or near a single point. The view
below shows a column where 6 grids are meeting in close proximity and the punching shear
check is showing a long section of black (missed) perimeter.
When the hatching option is used you can begin to see more clearly how the slabs have not
been defined so that a triangle has been missed within the column. (This sort of problem
usually occurs in smaller triangles than shown here.)
If we now check the option to include the column wall and beam edges and then re-hatch, the
area inside the column is considered as another slab area and so the merged boundary works
round the perimeter of the column. This will potentially avoid all problems relating to local
inaccuracies where grids meet.
If the punching shear check is them made with this option active the perimeters seem much
more reasonable. Note that in this example the missed triangle actually extends outside the
column and so is considered as a hole with the resultant shortening of the punching perimeter.
Specification of Effective Slab Reinforcement
When defining the effective slab reinforcement you can only provide one value which will
apply at all the check perimeters. In practice it is possible that lesser values of effective tension
reinforcement may be applicable at perimeters further from the column perimeter. In such
cases results from multiple checks may need to be used, or the more conservative (lower) value
could be used for a single check across all perimeters.
Providing the Shear Reinforcement
Orion provides the design checks and information to allow the detailing to commence. As
noted within the examples, the detailing is not necessarily a trivial matter, especially when
allowances for holes need to be made.
Punching Perimeters
The displayed punching perimeters should look correct and complete. No examples of
incorrect perimeters are shown in this document (other than the debatable perimeters shown
above).
The presence of holes and/or complex irregular slab geometry may trigger problems in the
perimeter identification. Where holes appear to be the source of the problem it is again worth
considering removing the holes and using the option to specify a periphery reduction. Where
slab geometry appears to be the source of the problem it is possible that the slabs can be
redefined locally to avoid the problem.
Walls
Checks on walls are made but should be viewed with particular caution.
The example above shows 3 discrete walls being checked. In this example the perimeter of the
wall on grid D looks suspicious (line being drawn outside the slab). On closer checking you
would find that the effective length of the punching perimeter is actually correct.
It is particularly advisable to check the lengths used for the punching perimeters in the case of
walls, and it is also noted that overlapping perimeters (see Overlapping Perimeters below) will
almost always apply where core walls are used.
In addition there could well be some debate as to the applicability of a concentrated punching
load check in the case of long walls, BS8110 provides no additional guidance on this. Could
there be a stress concentration at the ends that is not being checked?
Overlapping Perimeters
Where supports are closely spaced and perimeters overlap the code requires that a single
combined perimeter be checked. Orion does not identify such conditions and will simply
draw overlapping perimeters on the screen. If this condition occurs additional hand
calculations are required to substantiate the design.
Discontinuous Columns
In the special case where a slab supports a column or wall Orion will not make any checks at
all. The case remains beyond the current scope and additional hand calculations are required
to substantiate the design.
Advantages
Checking punching perimeters by hand can be a very tedious and time-consuming task. It is
not yet something that we can completely automate in Orion, and so it must still be treated as
an interactive activity. This interactive checking procedure will also be subject to limitations
in many projects.
However, we hope that this chapter still illustrates the potential for significant time saving if
punching shear checks are made using Orion. It is especially noted that once the initial checks
have been made (thus setting up the correct check data), subsequent rechecking can be carried
out very quickly indeed.
It is also noted that by following the training guidance notes and making checks as models are
constructed limitations can be determined earlier and hence time can be saved overall.
We look forward to receiving comments on this feature so we can continue to refine it to meet
the demands of designers.
If vEd,1 > 2vRd,c the slab thickness is inadequate, the check has failed and no further
calculations possible for the particular column.
else
If vEd,1 <= vRd,c the punching shear resistance of the slab is aedquate without the provision
of shear reinforcement. No further calculations are necessary.
else
the punching shear resistance of the slab is inadequate without the provision of shear
reinforcement. Shear reinforcement is required, and the checking procedure moves on to
determine a further perimeter where shear reinforcement is no longer required.
= 1 + k1.(Mt1/Vt).(u1/W11) + k2.(Mt2/Vt).(u1/W12)
Where ky & kz are factors from Table 6.1 depending on the column aspect ratio for the
particular direction
Mt1 & Mt2 are the transferred moments in the two orthogonal directions
= 1 + (u1/u1,red).[k1.(Mt1/Vt).(u1/W11) + k2.(Mt2/Vt).(u1/W12)]
If the user chooses not to select that the column be treated as an edge column the value of
will be calculated as for an internal column as detailed above taking account of the presence of
the edge of the slab.
If the user selects that the column be treated as an edge column then the component of
resulting from the transferred moment about the axis parallel to the edge of the slab is taken as
1.4 ie as in Figure 6.21 of EC2. The component of to take account of any transferred bending
moment about the axis perpendicular to the slab edge is calculated using the right hand term
of equation 6.44 ie k.(u1/W1).epar which is effectively the generic equation 6.39. The effects of
the moments about the two orthogonal axes are then combined using simple addition as
follows:
Magnification factor
= 1.4 + k.(Mt/Vt).(u1/W1)
where k, Mt and W1 all relate to bending about the axis perpendicular to the edge of the slab.
Where actual physical holes are present or the Perimeter Reduction is greater than zero, W1 is
calculated as described for an internal column but taking account of the location of the edge of
the slab.
Corner column
If the user chooses not to select that the column be treated as a corner column the value of
will be calculated as for an internal column as detailed above taking account of the presence of
the edge of the slab.
If the user selects that the column be treated as a corner column the value of will be taken as
1.5 from Figure 6.21 of EC2 with no additional calculation. The value of 1.5 is not adjusted if
actual physical holes are present or if the Perimeter Reduction is greater than zero.
Calculation of W1
EC2 describes W1 as a factor which 'corresponds to a distribution of shear as illustrated in
Figure 6.19 and is a function of the basic control perimeter u1'.
arrangement examined, the values of for an edge and corner column with W1 calculated
about the column centreline were 1.782 and 2.049 respectively but with W1 calculated about
the perimeter plastic neutral axis these increased to 2.727 and 3.644 respectively. These values
were calculated assuming no holes present to affect the perimeter. Some comparison ought,
therefore, to be able to be made between the values of for edge and corner columns
presented in Figure 6.21 (for a regular structural layout) shown below and the calculated
values.
Chapter 30
Introduction
This chapter provides an overview of the reasons why, and the way in which, model linking
and merging is used in Orion. Although the two features are related they are used for quite
different purposes:
The Merging (Complete) Models feature has yet to be enabled in the market
version of the program.
More than one engineer to work on the creation of model data for a large project.
Where there is repetition of the floor plan within a project, a standard sub-section can be
defined and then merged several times to create the full model (in effect a "copy/paste"
function).
Model relocation - If for whatever reason a model has been created a great distance from
the origin, it can be relocated to a more convenient position. The original model is simply
merged into a new empty model using the offsets, once again emulating a "copy/paste"
function.
Statically, free chunks of uncoupled 3D systems in space will yield the same analysis results
regardless of separate modelling. However, modelling those together will make the model
unnecessarily crowded.
There will be separate blocks of buildings in a large construction complex. Some of these
blocks are identical. Engineers tend to analyse and design only one of those blocks. The
typical project is then used for other similar blocks.
Linking models allows a common foundation to be modelled for these kinds of buildings.
Separate models are maintained for each block, one of the models contains a common
foundation which links to the other models.
Not all members or storeys are transferred from the linked models - only the supported
columns and walls at foundation level, with their loadings.
The project in which the links are created is not restricted in any way - you are free to
construct, analyse and design the model in the usual way. The foundation columns and
walls from the linked models do not get considered in the Building Analysis or Post
Analysis processes, they are only considered in the foundation design.
It is not essential to define any superstructure model for the foundation project, you are
able to link from a blank (master) project to other projects if you prefer to do so.
Models that are linked in to the foundation project can be re-located by translation but not
rotation. You are responsible for setting up the source models at the appropriate relative
angles.
During linking the foundation design forces created in the linked model populate a
foundation forces data table thus creating a complete composite set of data which can be
used in foundation design.
At foundation level, new slabs, beams and foundation systems are able to be defined as
required.
If the position of the column in the linked model is altered then the link will be broken in
foundation project.
Force overrides are applied via the The Foundation Forces Table.
For a linked column the link to it's original project is retained, however, the analysis results
are overridden.
You can select to cancel the override operation at a later stage. In which case, the analysis
results will be restored from the linked model.
3. Select the model to import. Note that models can reside in different drives and directories
of the computer.
4. Enter a Block Character for the imported model.
5. Choose whether the imported columns and walls are to have the block character added to
their reference, or if they are to retain their existing labels (where this is possible).
6. Click the Load Model button.
7. Initially the axes for the linked model will be loaded as a ghost picture.
8. Before import, the linked model must be located in its correct position on the foundation
project.
9. Rough positioning can be done by clicking the Drag Two Points button. Click Move Model
to see the effect.
10. Fine tuning can be done by entering X and Y entries directly. Again click Move Model to
see the effect.
11. Click Finish to import the model.
Imported Grids
When a new model is linked into the foundation project, the grid system is merged first. That
is, colinear and overlapping grids are combined, the existing parent grid is taken as the
reference.
In this way, since the existing grid system is not changed, the references of columns inserted
before will stay unchanged.
If three or more axes pass through the insertion point, the axes adopted by the
column in the imported model may differ from the original model.
If one of the linked projects is highlighted, the Link Management button becomes active.
When clicked the following dialog is displayed:
Update the Link Information of All Columns and Walls of the Selected Project
This option deletes any user defined loads in the highlighted project.
Chapter 31
Introduction
Various forms of output can be generated from the program:
Orion Reports
These can printed individually, or combined together using the Report Manager.
Quantity Reports
These can either be printed directly, or exported to Microsoft Access or Excel.
Drawings
These can either be plotted directly, or exported to DXF
Orion Reports
Orion Reports can be created from each of the modules. The main reports are formatted and
consist of a header & footer and a main body containing text, graphics and/or tables.
From the Report Print Preview the report can be configured and printed, or it can saved for
later printing via the Report Manager, (allowing it to be merged together with other Orion
Reports).
Options also exist for exporting the report to different formats as follows:
PDF Saves the report in pdf format.
TXT opens the report in a text editor, but strips out any formatting. Reports containing
graphics can not be opened in a text editor.
TDF Exports as a TAB delimited file.
Note
Certain data output (e.g. the FE Analysis Output Report) is unlikely to form
part of your calculation submission and is only created as plain text - it cannot
therefore be incorporated into a formatted report.
Module
Report
Type
Load Decomposition
Output Report
Plain text
Building Analysis
Storey Loads
Plain text
Building Analysis
Validity Check
Plain text
Module
Building Analysis
Report
Analysis Output Report
Type
Formatted
Nodal Results
Frame Element Results
Shell Element Results
Structural Member Results
Building Analysis
and FE Analysis
Formatted
Building Analysis
Formatted
Building Analysis
Formatted
Building Analysis
Formatted
Building Analysis
Formatted
Building Analysis
Formatted
Building Analysis
Formatted
Building Analysis
Formatted
FE Analysis
Output Report
Plain text
Column Design
Formatted
Column Design
Plain text
Column Design
Formatted
Column Design
Plain text
Column Design
Plain text
Beam Design
Formatted
Beam Design
Plain text
Beam Design
Formatted
Slab Strip
Plain text
Formatted
Column Punching
Check
Plain text
Pad/Pile Footing
Formatted
Strip Footing
Formatted
Stair Design
Plain Text
Module
Report
Quantity Extraction
Tables
Type
Formatted
Report Manager
Formatted reports (once generated by the different modules), will be listed in the Report
Manager. Three options exist for creating a combined report.
All Reports
With this option selected, individual reports can be selected and added to a combined report.
Txt Files
With this option selected, only those formatted reports that can be displayed as text files will
be listed. Individual reports can be selected and added to a combined report, however all
formatting and graphics will be stripped from this report.
Merged Report
If the All Reports option has been previously used to create one or more combined reports and
these reports have themselves been saved, then with this option selected, these combined
reports can be merged with each other.
Quantity Reports
Quantity Reports can be created for concrete, formwork and steel bars.
By clicking on the Quantity Extraction Tables icon (or from the menu File/Quantity Extraction
Tables), formatted reports can be created and printed or the data can be exported to either
Microsoft Access or Excel.
Note
Steel bar schedules are only available for those members/slab strips that have
been successfully designed.
Drawings
The following table summarises the kinds of drawing available:
Module
Form Plan
Drawings
1. General arrangements and sections including slab
reinforcement and dimensions. Slab steel quantity tables
can optionally be included.
2. Flat slab general arrangements including area of steel
contours.
3. Flat slab punching shear perimeters.
Column Application
Plan
Foundation Details
Column Design
Column Design
Beam Design
Examples
Form Plan - GAs & Sections