EBook Horikx

Conceptual structural design
understanding structural performance
Author(s)
Horikx, Michiel Paul
Publication date
2022
Document Version
Final published version
Link to publication
Citation for published version (APA):

Horikx, M. P. (2022). Conceptual structural design: understanding
structural performance.
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)
and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open
content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests,
please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the
material inaccessible and/or remove it from the website. Please contact the library:
https://www.amsterdamuas.com/library/contact/questions, or send a letter to: University Library (Library of the
University of Amsterdam and Amsterdam University of Applied Sciences), Secretariat, Singel 425, 1012 WP
Amsterdam, The Netherlands. You will be contacted as soon as possible.
Download date:29 Apr 2024

Conceptual Structural Design
Understanding
Structural Performance
Systematic thinking Systems thinking Modelling
Scientific method:
1. Problem definition
2. Research framework
3. Hypothesis
4. Validation 2-D
5. Conclusion
Michiel Paul Horikx

ii
Conceptual Structural Design
Understanding
Structural Performance
Michiel Paul Horikx

iv
This textbook was supported by the Faculty of Technology at the Ams-

terdam University of Applied Sciences, the national concrete association
BV, and the national steel association BmS.
Copyright c 2022 by Dr.ir. M.P. Horikx
All rights reserved. No part of this book may be reproduced or transmit-

ted in any form or by any means without permission of the author.
ISBN 9789492644275
A digital version of this book is available at http://hbo-kennisbank.nl
Book and corresponding slide presentations are edited with the free and
open source document preparation system LaTeX, where all figures are
created with the free and open source vector graphics editor Inkscape.
Executive summary
The subject of this textbook is a methodical approach on the complex

problem-solving process of conceptual structural design.
Problem A lack of insight of the professional into structural engineer-

ing is considered to be a main obstacle for an overall performance/cost
optimisation of the built environment.
For a better understanding and an effective solution approach this lack

of insight can be divided into the two partial problems, “lack of insight
into conceptual structural design” and the underlying “lack of insight
into structural performance”.
Solution approach The methodical approach on conceptual struc-

tural design leads to a controlled build-up of insight into the behaviour
of the structure and supports the actual successive design decisions dur-
ing the conceptual design phase on the basis of the following coherent
set of solution components:
Structural design cycle Independent of life cycle phase, complexity

of design, and contractual commitments, the structural engineering
practice can be outlined by a fundamental structural design cycle:
creation, optimisation, and specification. Within each of the major
design phases conceptual design, basic design, and detailed design,
this fundamental structural design cycle is applicable.
vi Executive summary
Basic structural forms One of the main conceptual structural design

activities is the determination of the structural form, based on “un-
derstanding” and “order of magnitude” of basic structural forms.
Characterisation of structural forms, with regard to the capacity
to bear and resist, and with regard to the interfaces with the built
environment, turns out to be feasible on a two-dimensional subsys-
tem level.
Structural design path The structural design process can fundament-
ally be characterised by the two simultaneous processes of spe-
cification and decomposition of the structural form from system
to element. These two processes can be visualised together in a
two-dimensional matrix in which the structural design can be ex-
plored. The design path follows the fundamental dimensioning
routine from structural integrity, via load distribution, to failure
mechanisms.
Structural design loops In order to have a successful solution to a
complex design problem, a cyclic design process is inevitable. Every
cycle goes through the phases of creation, optimisation, and spe-
cification. The cyclic convergent optimum strategy aims for an op-
timisation of both quality of design outcome and number of design
cycles. The individual design loops are appointed and the corres-
ponding basic principles clarified.
Shared knowledge-based conceptual design For an optimisation of
the performance/cost ratio of the life cycle of a structure, an in-
tegral approach on conceptual design is a necessity. Control of
the interfaces between all participating disciplines is largely de-
pendent on experience and intuition. Definition and collection of
the fundamental conceptual design parameters of the most influen-
tial participating disciplines serve as a concurrent-based breeding
ground for integral design solutions.
Conceptual structural design parameters Both for understanding
structural performance and for an effective knowledge exchange
with the other participating disciplines, a balanced set of con-
vii
ceptual structural design parameters is conditional. On a two-

dimensional subsystem level, approximation parameters for con-
ceptual structural design are determined on the basis of timeless
applied mechanics.
Case study This specific case study demonstrates how the methodical
approach leads to a controlled build-up of insight into the behaviour of
the structure and supports the actual successive design decisions dur-
ing the conceptual design of the trusses of the Maeslant storm surge
barrier.
The load paths, overall geometry, and principal detailing on the basis
of performance, structural, and construction demands, are determined.
Subsequently, the structural action in this outlined structure is optim-
ised and the elements are dimensioned. Finally, a thorough risk ana-
lysis is conducted as a demarcation of the conceptual structural design
phase.
Training The required knowledge, skills, and professional attitude have

to be achieved by a mix of the learning methods lecture, training, and
project work. Lectured theories and trained engineering practice can be
applied during project work.
The training programme consists of a series of trainings, including a
zero measurement training. All trainings have approximately the same
entry professional on graduate master level, with a balanced complexity
of geometry and modelling.
Epilogue It is important to study how the design is organised in prac-

tice, and especially the ways in which designers with different disciplin-
ary expertise are able to work together, collaboratively in teams. For an
integral conceptual structural design, the main contributing disciplines
and corresponding interfaces have to be considered
This textbook discusses conceptual structural design on a high level of
abstraction. However, a deepening research on conceptual structural
viii Executive summary
design is valuable and feasible. Recommendations for research are given

with respect to both understanding these complex interdisciplinary in-
terfaces and structural performance.
Contents
Executive summary v
1 Introduction 1
1.1 About this textbook . . . . . . . . . . . . . . . . . . . . . 1
1.1.1 Structural design . . . . . . . . . . . . . . . . . . . 1
1.1.2 Subject and knowledge level . . . . . . . . . . . . . 2
1.1.3 Textbook outline . . . . . . . . . . . . . . . . . . . 4
1.2 Reading guide . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2.1 Partition of the textbook outline . . . . . . . . . . 5
1.2.2 Introductions . . . . . . . . . . . . . . . . . . . . . 6
I Problem, analysis and solution approach 7
2 Introduction to part I 9
3 Present-day problems in structural engineering 13

3.1 Structural (un)safety in the Netherlands . . . . . . . . . . 13
3.1.1 Structural failures . . . . . . . . . . . . . . . . . . 13
3.1.2 Structural safety codes of practice . . . . . . . . . 14
3.1.3 Actual structural safety . . . . . . . . . . . . . . . 16
3.2 Lack of insight of the professional . . . . . . . . . . . . . . 16
3.2.1 Problems, causes, and effects . . . . . . . . . . . . 18
3.2.2 General requirements structural engineering . . . . 19
3.2.3 Recommendations structural design . . . . . . . . . 20
x CONTENTS
3.3 Historical perspective . . . . . . . . . . . . . . . . . . . . . 20

3.3.1 Developments professional structural engineering . 20
3.3.2 Homo universalis . . . . . . . . . . . . . . . . . . . 21
3.3.3 Expanding depth and breadth . . . . . . . . . . . . 22
3.3.4 On-going expanding depth and breadth . . . . . . 23
4 Solution approach 25
4.1 Problem definition . . . . . . . . . . . . . . . . . . . . . . 25
4.1.1 The merit of a proper problem definition . . . . . . 25
4.1.2 An interface control approach . . . . . . . . . . . . 25
4.1.3 Structural engineering activities and missing tools 27
4.1.4 Definition of structural performance . . . . . . . . 27
4.1.5 Definition of conceptual structural design . . . . . 30
4.2 Present-day solutions’ field of practice . . . . . . . . . . . 31
4.2.1 Copying reality . . . . . . . . . . . . . . . . . . . . 31
4.2.2 Planning and control . . . . . . . . . . . . . . . . . 31
4.2.3 Numerical power . . . . . . . . . . . . . . . . . . . 33
4.2.4 Professional higher education . . . . . . . . . . . . 35
4.3 Solution approach structural design . . . . . . . . . . . . . 36
4.3.1 A need for control on system level . . . . . . . . . 36
4.3.2 Shift from calculating to modelling . . . . . . . . . 37
4.3.3 Three pillars of future-proof structural design . . . 37
5 Solution components 41
5.1 In search of a methodical approach . . . . . . . . . . . . . 41
5.1.1 Solution approach . . . . . . . . . . . . . . . . . . 41
5.1.2 Guiding principles . . . . . . . . . . . . . . . . . . 41
5.2 T-shaped professional . . . . . . . . . . . . . . . . . . . . 42
5.2.1 Modern demands . . . . . . . . . . . . . . . . . . . 42
5.2.2 T-shaped professional structural engineer . . . . . 43
5.3 Applied mechanics . . . . . . . . . . . . . . . . . . . . . . 44
5.3.1 Necessity of insight into structural performance . . 44
5.3.2 Timeless applied mechanics . . . . . . . . . . . . . 45
5.4 Designing with progressive insight . . . . . . . . . . . . . . 46
5.4.1 General problem approach . . . . . . . . . . . . . . 46
CONTENTS xi
5.4.2 Progressive insight from estimation to accuracy . . 46

5.5 Decomposition . . . . . . . . . . . . . . . . . . . . . . . . 47
5.5.1 Abstractions of a decomposed system . . . . . . . . 47
5.5.2 Decomposition of complex systems . . . . . . . . . 49
5.5.3 Physical decomposition . . . . . . . . . . . . . . . 49
5.5.4 Process decomposition . . . . . . . . . . . . . . . . 50
5.5.5 Aspect decomposition . . . . . . . . . . . . . . . . 50
II The art of conceptual structural design 53
6 Introduction to part II 55
7 Conceptual design 59
7.1 Determination of the structural form . . . . . . . . . . . . 59
7.1.1 Conceptual structural design . . . . . . . . . . . . 59
7.1.2 Design process . . . . . . . . . . . . . . . . . . . . 62
7.1.3 Flow diagram . . . . . . . . . . . . . . . . . . . . . 63
7.1.4 Experience and intuition . . . . . . . . . . . . . . . 64
7.2 How to capture the intangible . . . . . . . . . . . . . . . . 65
7.2.1 Abstracting conceptual design . . . . . . . . . . . . 65
7.2.2 Directing parameters of conceptual design . . . . . 66
7.2.3 Sharing the knowledge of the built environment . . 67
7.2.4 Back to the fundamentals of conceptual design . . 68
7.3 Shared knowledge-based conceptual design . . . . . . . . . 69
7.3.1 Splitting process and technical breadth . . . . . . . 69
7.3.2 Principal disciplines of the built environment . . . 71
7.3.3 Concurrent-based shared knowledge . . . . . . . . 72
7.3.4 Conceptual structural design parameters . . . . . . 73
8 Process decomposition 77
8.1 Fundamental design cycle . . . . . . . . . . . . . . . . . . 77
8.1.1 Structural life cycle phases . . . . . . . . . . . . . . 77
8.1.2 Structural design phases . . . . . . . . . . . . . . . 77
8.1.3 Cyclic design process . . . . . . . . . . . . . . . . . 79
8.1.4 Fundamental structural design cycle . . . . . . . . 80
xii CONTENTS
8.2 Level of accuracy . . . . . . . . . . . . . . . . . . . . . . . 82

8.2.1 Level of specification . . . . . . . . . . . . . . . . . 82
8.2.2 Fib Model Code 2010 . . . . . . . . . . . . . . . . 83
8.2.3 Accuracy fundamental structural design phases . . 85
8.3 Fundamental design process . . . . . . . . . . . . . . . . . 86
8.3.1 Structural design characteristics . . . . . . . . . . . 86
8.3.2 Fundamental structural design process . . . . . . . 86
9 Physical decomposition 89
9.1 Qualification of the structural form . . . . . . . . . . . . . 89
9.1.1 Classification . . . . . . . . . . . . . . . . . . . . . 89
9.1.2 Form follows function . . . . . . . . . . . . . . . . 90
9.1.3 Interfaces with the built environment . . . . . . . . 90
9.2 Decomposition of the structural form . . . . . . . . . . . . 91
9.2.1 System decomposition . . . . . . . . . . . . . . . . 91
9.2.2 Structural form on subsystem level . . . . . . . . . 92
9.2.3 Basic structural forms . . . . . . . . . . . . . . . . 93
10 Cyclic process control 97

10.1 Exploration of the solubility space . . . . . . . . . . . . . 97
10.1.1 Creation-process requirements . . . . . . . . . . . . 97
10.1.2 Design strategies . . . . . . . . . . . . . . . . . . . 98
10.2 Linear design . . . . . . . . . . . . . . . . . . . . . . . . . 99
10.2.1 Linear process . . . . . . . . . . . . . . . . . . . . 99
10.2.2 Specification of the structural form . . . . . . . . . 99
10.3 Cyclic design . . . . . . . . . . . . . . . . . . . . . . . . . 100
10.3.1 Cyclic process . . . . . . . . . . . . . . . . . . . . . 100
10.3.2 Fundamental structural design path . . . . . . . . 101
10.3.3 Structural design path . . . . . . . . . . . . . . . . 103
10.4 Cyclic convergent design . . . . . . . . . . . . . . . . . . . 105
10.4.1 Volume of complexity . . . . . . . . . . . . . . . . 105
10.4.2 Three steps in conceptual design . . . . . . . . . . 105
10.4.3 Definition step . . . . . . . . . . . . . . . . . . . . 107
10.4.4 Creation step . . . . . . . . . . . . . . . . . . . . . 108
10.4.5 Selection step . . . . . . . . . . . . . . . . . . . . . 108
CONTENTS xiii
10.5 Cyclic optimum design . . . . . . . . . . . . . . . . . . . . 108

10.5.1 Modelling loops . . . . . . . . . . . . . . . . . . . . 110
10.5.2 Typical design loops . . . . . . . . . . . . . . . . . 110
10.5.3 Optimal design cycle . . . . . . . . . . . . . . . . . 111
10.5.4 Risk analysis . . . . . . . . . . . . . . . . . . . . . 112
III Understanding structural performance 115
11 Introduction to part III 117
12 Structural performance 121

12.1 Present-day structural performance . . . . . . . . . . . . . 121
12.1.1 Structural requirements . . . . . . . . . . . . . . . 121
12.1.2 Modern developments . . . . . . . . . . . . . . . . 122
12.2 Force and deformation-driven parameters . . . . . . . . . 123
12.2.1 Equilibrium and strength . . . . . . . . . . . . . . 123
12.2.2 Deformation-driven parameters . . . . . . . . . . . 124
12.2.3 Determination of deformation-driven parameters . 126
12.3 Approximate dimensioning . . . . . . . . . . . . . . . . . . 127
12.3.1 Conceptual structural design approximations . . . 127
12.3.2 Conceptual structural design parameters . . . . . . 127
12.3.3 Dimensioning routine . . . . . . . . . . . . . . . . . 128
13 Structural integrity 131

13.1 Conceptual design on system level . . . . . . . . . . . . . 131
13.1.1 Creation phase of the conceptual design . . . . . . 131
13.1.2 Structural integrity . . . . . . . . . . . . . . . . . . 131
13.2 Load path design . . . . . . . . . . . . . . . . . . . . . . . 132
13.2.1 Load path design on system level . . . . . . . . . . 132
13.2.2 Truss-analogy in load path design . . . . . . . . . . 133
13.2.3 Modelling the system decomposition . . . . . . . . 134
13.2.4 Dimensioning routine of the load path . . . . . . . 134
13.3 Conceptual structural design . . . . . . . . . . . . . . . . 135
13.3.1 Principal details . . . . . . . . . . . . . . . . . . . 135
13.3.2 System configuration . . . . . . . . . . . . . . . . . 136
xiv CONTENTS
13.3.3 Material properties of conceptual design . . . . . . 136
14 Fundamental parameters of load distribution 141

14.1 Conceptual design on subsystem level . . . . . . . . . . . . 141
14.1.1 Load distribution phase of the conceptual design . 141
14.1.2 Load distribution . . . . . . . . . . . . . . . . . . . 141
14.1.3 Parallel load distribution . . . . . . . . . . . . . . . 142
14.2 Load distribution in basic structural forms . . . . . . . . . 144
14.2.1 Load distribution in a frame . . . . . . . . . . . . . 144
14.2.2 Load distribution in a floor slab . . . . . . . . . . . 146
14.2.3 Load distribution in a cable-stayed beam . . . . . . 147
14.2.4 Load distribution in a truss . . . . . . . . . . . . . 150
14.2.5 Load distribution in an arch . . . . . . . . . . . . . 152
14.2.6 Load distribution in a shear wall . . . . . . . . . . 154
14.3 Parallel load distribution on detailed level . . . . . . . . . 155
14.4 Induced deformation . . . . . . . . . . . . . . . . . . . . . 158
14.4.1 Principle of induced deformation . . . . . . . . . . 158
14.4.2 Induced deformation on subsystem level . . . . . . 159
14.4.3 Induced deformation on system level . . . . . . . . 159
15 Fundamental parameters of failure mechanisms 163

15.1 Conceptual design on element level . . . . . . . . . . . . . 163
15.1.1 Dimensioning phase of the conceptual design . . . 163
15.1.2 Failure mechanisms . . . . . . . . . . . . . . . . . . 163
15.2 Shear . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
15.2.1 Shear strength . . . . . . . . . . . . . . . . . . . . 164
15.2.2 Shear deformation . . . . . . . . . . . . . . . . . . 166
15.2.3 Shear deformation on subsystem level . . . . . . . 168
15.3 Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
15.3.1 Stability of the equilibrium . . . . . . . . . . . . . 169
15.3.2 Second-order effects . . . . . . . . . . . . . . . . . 171
15.3.3 Euler-based design approximations . . . . . . . . . 172
15.3.4 Combined compression and bending . . . . . . . . 175
15.4 Bending and compression . . . . . . . . . . . . . . . . . . 175
15.4.1 Conceptual design of concrete in bending . . . . . 175
CONTENTS xv
15.4.2 Conceptual design of concrete in compression . . . 178

15.4.3 Conceptual design of steel in bending . . . . . . . . 179
15.4.4 Conceptual design of steel in compression . . . . . 180
15.4.5 Conceptual design of foundations . . . . . . . . . . 181
IV Case study and training 185
16 Introduction to part IV 187
17 Case study trusses Maeslant storm surge barrier 189

17.1 Demonstration of the methodical approach . . . . . . . . . 189
17.1.1 A case study of conceptual structural design . . . . 189
17.1.2 Actual design method . . . . . . . . . . . . . . . . 190
17.1.3 Methodical approach . . . . . . . . . . . . . . . . . 190
17.1.4 Outcome of the case study . . . . . . . . . . . . . . 191
17.2 Maeslant storm surge barrier . . . . . . . . . . . . . . . . 192
17.2.1 Final piece of the Delta works . . . . . . . . . . . . 192
17.2.2 Requirements . . . . . . . . . . . . . . . . . . . . . 193
17.2.3 Conceptual design Maeslant storm surge barrier . . 193
17.2.4 Performance/cost optimisation . . . . . . . . . . . 195
17.3 Creation of a system outline . . . . . . . . . . . . . . . . . 195
17.3.1 Load-path design on system level . . . . . . . . . . 195
17.3.2 Principal details . . . . . . . . . . . . . . . . . . . 196
17.3.3 Circular hollow section elements . . . . . . . . . . 198
17.3.4 Decomposition in subsystems . . . . . . . . . . . . 199
17.4 Optimisation of the structural action . . . . . . . . . . . . 201
17.4.1 Optimisation on horizontal subsystem level . . . . 201
17.4.2 Optimisation on vertical subsystem level . . . . . . 202
17.4.3 Induced deformation . . . . . . . . . . . . . . . . . 203
17.5 Dimensioning . . . . . . . . . . . . . . . . . . . . . . . . . 203
17.5.1 Cross-sectional area of the bottom chord . . . . . . 203
17.5.2 Global stability . . . . . . . . . . . . . . . . . . . . 203
17.5.3 Cross-sectional area of web and top chord . . . . . 204
17.5.4 Induced deformation of web members . . . . . . . 204
xvi CONTENTS
17.5.5 Section dimensions of conceptual design . . . . . . 207

17.6 Specification and risk analysis . . . . . . . . . . . . . . . . 207
17.6.1 Material demand . . . . . . . . . . . . . . . . . . . 207
17.6.2 Dimensioning and cost weighting . . . . . . . . . . 208
17.6.3 Uncertainties and coverage by dimensioning . . . . 208
17.6.4 Reserves and optimisations . . . . . . . . . . . . . 208
17.7 Further optimisations during basic design . . . . . . . . . 211
17.7.1 Basic design . . . . . . . . . . . . . . . . . . . . . . 211
17.7.2 Extensive modelling . . . . . . . . . . . . . . . . . 212
17.7.3 Coupling truss . . . . . . . . . . . . . . . . . . . . 212
17.7.4 Geometry and Section dimensions . . . . . . . . . 212
18 Training programme 215

18.1 Training programme outline . . . . . . . . . . . . . . . . . 215
18.1.1 Load path design . . . . . . . . . . . . . . . . . . . 215
18.1.2 Programme outline . . . . . . . . . . . . . . . . . . 215
18.2 Training 0 - Spatial struts and ties . . . . . . . . . . . . . 216
18.2.1 Intermediate bracing system with struts and ties . 216
18.2.2 Strength design separate 2-D struts and ties . . . . 218
18.2.3 Deformation-driven design of the ties . . . . . . . . 220
18.2.4 Deformation-driven design of the struts . . . . . . 220
18.2.5 Design of the overall system . . . . . . . . . . . . . 221
18.2.6 Material optimisation . . . . . . . . . . . . . . . . 222
18.2.7 Redundancy analysis . . . . . . . . . . . . . . . . . 222
19 3-D geometry combined with simple modelling 225

19.1 Training 1 - Spatial trusses . . . . . . . . . . . . . . . . . 225
19.1.1 Intermediate bracing system with trusses . . . . . 225
19.1.2 Load distribution in the separate 2-D trusses . . . 227
19.1.3 Strength design of the separate 2-D trusses . . . . 228
19.1.4 Deformation driven design of the chord members . 229
19.1.5 Deformation driven design of the web members . . 230
19.1.6 Design of the overall system . . . . . . . . . . . . . 231
19.2 Training 2 - Box truss with torsion . . . . . . . . . . . . . 232
19.2.1 3-D square box truss with eccentric loading . . . . 232
CONTENTS xvii
19.2.2 Simplified 2-D modelling ULS . . . . . . . . . . . . 234

19.2.3 Simplified 2-D modelling SLS . . . . . . . . . . . . 235
19.2.4 3-D system modelling ULS . . . . . . . . . . . . . 235
19.2.5 3-D system modelling SLS . . . . . . . . . . . . . . 236
19.2.6 2-D versus 3-D modelling . . . . . . . . . . . . . . 237
19.2.7 Stocky or slender system behaviour . . . . . . . . . 237
20 2-D geometry combined with medium modelling 239

20.1 Training 3 - Combined truss and arch . . . . . . . . . . . 239
20.1.1 Truss girder bridge building combined with arch . 239
20.1.2 Strength design of the separate truss girder . . . . 241
20.1.3 Strength design of the separate tied arch . . . . . . 242
20.1.4 Load distribution combined truss girder and arch . 243
20.1.5 Equal load distribution with high strength steel . . 244
20.1.6 Material optimisation on system level . . . . . . . 244
20.1.7 Displacement in the SLS . . . . . . . . . . . . . . . 245
20.1.8 Strength design web members on system level . . . 246
20.2 Training 4 - Second-order effect arch . . . . . . . . . . . . 247
20.2.1 Three-hinged truss arch . . . . . . . . . . . . . . . 247
20.2.2 Analysis load distribution in the 2-D system . . . . 249
20.2.3 N- and M-lines of the arch . . . . . . . . . . . . . . 249
20.2.4 Dimensioning chord members of the truss arch . . 251
20.2.5 Analysis second-order effect . . . . . . . . . . . . . 251
21 1-D geometry combined with abstract modelling 253

21.1 Training 5 - Truss analogy concrete beam . . . . . . . . . 253
21.1.1 Cantilevered reinforced concrete beam . . . . . . . 253
21.1.2 Load distribution strut-and-tie model . . . . . . . 254
21.1.3 Shear strength of the concrete cantilever . . . . . . 255
21.1.4 Shear deformation of the concrete cantilever . . . . 256
21.1.5 Displacement of the cantilever . . . . . . . . . . . . 256
21.1.6 Completion of the strut-and-tie modelling . . . . . 257
21.2 Training 6 - Truss analogy steel beam . . . . . . . . . . . 258
21.2.1 Cantilevered steel beam . . . . . . . . . . . . . . . 258
21.2.2 Truss analogy with an effective width modelling . . 259
xviii CONTENTS
21.2.3 Load distribution truss analogy . . . . . . . . . . . 260

21.2.4 Shear strength of the steel cantilever . . . . . . . . 261
21.2.5 Shear deformation of the steel cantilever . . . . . . 261
21.2.6 Analysis effective width modelling . . . . . . . . . 262
22 Epilogue 263
22.1 Complexity of the built environment . . . . . . . . . . . . 263
22.1.1 Complexity . . . . . . . . . . . . . . . . . . . . . . 263
22.1.2 Complex systems and processes . . . . . . . . . . . 264
22.1.3 Complexity of interdisciplinary interfaces . . . . . 265
22.2 Recommendations for research . . . . . . . . . . . . . . . . 267
22.2.1 In-depth research . . . . . . . . . . . . . . . . . . . 267
22.2.2 Conceptual design parameters built environment . 267
22.2.3 Fundamental behaviour of structural materials . . 268
22.2.4 Transition of stocky to slender beam theory . . . . 269
22.2.5 Adjustment factor of buckling strength . . . . . . . 269
Bibliography 271
Curriculum Vitae 275

List of Figures

1.2 Partition of the textbook outline . . . . . . . . . . . . . . 6
2.1 Reading guide for part I . . . . . . . . . . . . . . . . . . . 10
3.1 Structural safety Eurocode . . . . . . . . . . . . . . . . . . 15

3.2 Expanding depth and breadth . . . . . . . . . . . . . . . . 21
4.1 Structural design as an interface control approach . . . . . 26
5.1 T-shaped professional structural engineer . . . . . . . . . 43

5.2 Problem approach . . . . . . . . . . . . . . . . . . . . . . 46
6.1 Reading guide for part II . . . . . . . . . . . . . . . . . . . 56
7.1 Conceptual structural design process . . . . . . . . . . . . 60

7.2 Directing parameters of conceptual design . . . . . . . . . 67
7.3 T-shaped professional for conceptual design . . . . . . . . 69
7.4 T-shaped professional technical breadth . . . . . . . . . . 70
7.5 Environment and object level . . . . . . . . . . . . . . . . 71
7.6 T-shaped conceptual structural design . . . . . . . . . . . 74
8.1 Structural design process . . . . . . . . . . . . . . . . . . . 80

8.2 Fundamental structural design cycle . . . . . . . . . . . . 81
8.3 Level of specification . . . . . . . . . . . . . . . . . . . . . 83
8.4 Fundamental structural design process . . . . . . . . . . . 88
xx LIST OF FIGURES
9.1 Assembly of an arch bridge . . . . . . . . . . . . . . . . . 92

9.2 Subsystem level . . . . . . . . . . . . . . . . . . . . . . . . 93
9.3 Basic structural forms . . . . . . . . . . . . . . . . . . . . 94
10.1 Design strategies . . . . . . . . . . . . . . . . . . . . . . . 99

10.2 Linear structural design . . . . . . . . . . . . . . . . . . . 99
10.3 Fundamental structural design path . . . . . . . . . . . . 102
10.4 Structural design path . . . . . . . . . . . . . . . . . . . . 104
10.5 Limited volume of complexity per design phase . . . . . . 106
10.6 Fundamental structural design loops . . . . . . . . . . . . 109
10.7 Structural design loops . . . . . . . . . . . . . . . . . . . . 112
11.1 Reading guide for part III . . . . . . . . . . . . . . . . . . 118
12.1 Developments of structural performance . . . . . . . . . . 122

12.2 Force and deformation-driven parameters . . . . . . . . . 124
12.3 Dimensioning routine of the conceptual structural design . 130
13.1 Load path design . . . . . . . . . . . . . . . . . . . . . . . 132

13.2 Truss-analogy in load path design . . . . . . . . . . . . . . 133
13.3 Modelling the system effects . . . . . . . . . . . . . . . . . 134
13.4 Dimensioning routine . . . . . . . . . . . . . . . . . . . . . 135
14.1 Statically determinate structure . . . . . . . . . . . . . . . 142

14.2 Statically indeterminate structure . . . . . . . . . . . . . . 143
14.3 Parallel bending and axial force . . . . . . . . . . . . . . . 144
14.4 Load distribution in a two-way spanning floor slab . . . . 146
14.5 Redistribution in a cable-stayed beam . . . . . . . . . . . 148
14.6 Load distribution in a truss . . . . . . . . . . . . . . . . . 150
14.7 Load distribution in an arch . . . . . . . . . . . . . . . . . 152
14.8 Bending moments in an arch . . . . . . . . . . . . . . . . 154
14.9 Stiffener plate in a beam-to-column joint . . . . . . . . . . 156
14.10Bending deformation of the flange . . . . . . . . . . . . . . 157
14.11Axial deformation of the stiffener . . . . . . . . . . . . . . 157
14.12Shear deformation of the stiffener . . . . . . . . . . . . . . 158
LIST OF FIGURES xxi
14.13Induced deformation torsional clamped beam . . . . . . . 160

14.14Induced deformation in a truss bridge . . . . . . . . . . . 161
15.1 Shear strength . . . . . . . . . . . . . . . . . . . . . . . . 165

15.2 Bending and shear deformation . . . . . . . . . . . . . . . 166
15.3 Transition slenderness for stocky to slender beams . . . . 168
15.4 Shear deformation in a cantilevered truss . . . . . . . . . . 168
15.5 Stability of the equilibrium . . . . . . . . . . . . . . . . . 170
15.6 Second-order effects . . . . . . . . . . . . . . . . . . . . . 171
15.7 Buckling length . . . . . . . . . . . . . . . . . . . . . . . . 173
15.8 Arch and truss analogy in concrete beams . . . . . . . . . 176
15.9 Bending strength of a concrete member . . . . . . . . . . 176
15.10Bending strength of a steel member . . . . . . . . . . . . . 179
15.11Failure mechanism of a spread foundation . . . . . . . . . 181
15.12Failure mechanism of a pile foundation . . . . . . . . . . . 182
16.1 Reading guide for part IV . . . . . . . . . . . . . . . . . . 188
17.1 Maeslant storm surge barrier . . . . . . . . . . . . . . . . 194

17.2 Hydraulic load . . . . . . . . . . . . . . . . . . . . . . . . 196
17.3 Principal joint . . . . . . . . . . . . . . . . . . . . . . . . . 197
17.4 Section properties of a CHS . . . . . . . . . . . . . . . . . 198
17.5 Decomposition in subsystem planes . . . . . . . . . . . . . 199
17.6 Triangular cross-sectional truss configuration . . . . . . . 200
17.7 Optimisation of the supports of the retaining wall . . . . . 202
17.8 Induced deformation of the Maeslant storm surge barrier . 205
17.9 Truss dimensions . . . . . . . . . . . . . . . . . . . . . . . 207
18.1 Bracing system with struts and ties . . . . . . . . . . . . . 217

18.2 Load distribution 2-D struts and ties . . . . . . . . . . . . 218
18.3 Deformation of the ties . . . . . . . . . . . . . . . . . . . . 220
18.4 Deformation of the struts . . . . . . . . . . . . . . . . . . 221
19.1 Bracing system with trusses . . . . . . . . . . . . . . . . . 226

19.2 Section properties of a square RHS . . . . . . . . . . . . . 226
19.3 Load distribution 2-D trusses . . . . . . . . . . . . . . . . 227
xxii LIST OF FIGURES
19.4 Bending deformation of the truss beams . . . . . . . . . . 229

19.5 Shear deformation of the truss beams . . . . . . . . . . . . 230
19.6 Deformation of the web members . . . . . . . . . . . . . . 230
19.7 3-D square box truss with eccentric loading . . . . . . . . 232
19.8 Individual member forces . . . . . . . . . . . . . . . . . . 234
19.9 Displacement due to the elongation of a bar . . . . . . . . 235
19.10Centric force and torsion couple . . . . . . . . . . . . . . . 236
20.1 Truss girder bridge building combined with tied arch . . . 240
20.2 Displacement combined truss girder and tied arch . . . . . 243
20.3 Truss load distribution web members . . . . . . . . . . . . 246
20.4 Arch load distribution web members . . . . . . . . . . . . 247
20.5 Three-hinged truss arch . . . . . . . . . . . . . . . . . . . 248
20.6 Load distribution three-hinged arch . . . . . . . . . . . . . 249
20.7 N-, M-, and, V-lines of enclosing simplified structures . . . 250
20.8 N- and M-lines of a semi-circular arch . . . . . . . . . . . 250
21.1 Cantilevered reinforced concrete beam . . . . . . . . . . . 253

21.2 Load distribution concrete beam . . . . . . . . . . . . . . 255
21.3 Shear deformation of the cantilever . . . . . . . . . . . . . 256
21.4 Contribution shear deformation right span . . . . . . . . . 257
21.5 Displacement of the cantilever . . . . . . . . . . . . . . . . 257
21.6 Cantilevered steel beam . . . . . . . . . . . . . . . . . . . 258
21.7 Bending deformation cantilevered beam . . . . . . . . . . 259
21.8 Shear deformation cantilevered beam . . . . . . . . . . . . 260
21.9 Effective width model . . . . . . . . . . . . . . . . . . . . 260
21.10Load distribution truss analogy . . . . . . . . . . . . . . . 261
List of Tables
4.1 Structural engineering activities and tools . . . . . . . . . 28

4.2 Three pillars of future-proof structural design . . . . . . . 38
6.1 Solution components for conceptual structural design . . . 58
7.1 Principal disciplines of the built environment . . . . . . . 72
8.1 Life cycle phases . . . . . . . . . . . . . . . . . . . . . . . 78

8.2 Design phases . . . . . . . . . . . . . . . . . . . . . . . . . 79
8.3 Fib design phases . . . . . . . . . . . . . . . . . . . . . . . 84
8.4 Level of accuracy . . . . . . . . . . . . . . . . . . . . . . . 86
8.5 Structural engineering characteristics . . . . . . . . . . . . 87
11.1 Solution components for structural performance . . . . . . 120
12.1 Misfit of design parameters in education . . . . . . . . . . 126

12.2 Conceptual structural design parameters . . . . . . . . . . 129
13.1 Material properties for conceptual design . . . . . . . . . . 137
15.1 Conceptual design parameters of failure mechanisms . . . 164

15.2 Critical buckling stress comparison . . . . . . . . . . . . . 174
17.1 Dike reinforcement versus barrier . . . . . . . . . . . . . . 193

17.2 Requirements for the Maeslant storm surge barrier . . . . 193
17.3 Risk analysis: dimensioning and cost weighting . . . . . . 209
xxiv LIST OF TABLES
17.4 Risk analysis: uncertainties and coverage by dimensioning 210

17.5 Risk analysis: reserves and optimisations . . . . . . . . . . 211
18.1 Training programme outline . . . . . . . . . . . . . . . . . 216

18.2 Dimensions 2-D struts and ties . . . . . . . . . . . . . . . 219
18.3 Dimensions 3-D spatial struts and ties . . . . . . . . . . . 221
18.4 Dimensions material optimisation . . . . . . . . . . . . . . 222
19.1 Dimensions 2-D trusses . . . . . . . . . . . . . . . . . . . . 229

19.2 Dimensions 3-D spatial trusses . . . . . . . . . . . . . . . 231
19.3 Standardised square RHS sections . . . . . . . . . . . . . . 231
19.4 2-D versus 3-D modelling . . . . . . . . . . . . . . . . . . 237
Chapter 1
Introduction
1.1 About this textbook

1.1.1 Structural design
Structural design consists of three major sequential design phases: con-
ceptual, basic, and detailed design.
With detailed design - a phase of code checking, detailing, and specifying

- all common material applications have been extensively researched and
recorded in numerous textbooks and design codes of practice.
For basic design - a phase of deepening and optimisation - the main

tools are applied mechanics-based and since Isaac Newton widespread
available as textbook material.
Conceptual design - the creation phase with a complex and partly intu-
itive process and numerous complex interfaces between different fields of
practice - is little touched by technological progress.
Complexity of design Design is to optimise the performance/cost

ratio of the life cycle. For design, it is mostly the outcome that counts,
2 Introduction
rather than the followed path. The design path, however, does control
both duration and flexibility of the complex design process.
The art of design, striving for the ultimate solution, is the process of
getting oversight by abstracting complexity and crossing borders.
Abstracting complexity The standard engineering practice of hand-

ling complex reality is modelling this reality with an abstract and approx-
imate representation. Besides structural modelling, which is the core of
structural engineering, process modelling can attribute to a clarification
of the complex cyclic design process.
Crossing borders between dissimilar things Scientific education

in structural design and corresponding research is compartmented to
such an extent that the interface between applied mechanics and ma-
terial applications is underdeveloped. The interfaces between structural
demand, performance demand, and construction demand are even less
visible. Both professionals and higher education programmes will benefit
from crossing borders and clarification of these interfaces.
Conceptual structural design Conceptual design is the first and

decisive phase of design, providing the overall integrated system. Here,
the proverbial “DNA” of the solution is constructed.
Present-day conceptual design of regularly complex structures with nu-
merous boundary conditions, require a thorough understanding of design
parameters and extensive experience in design.
1.1.2 Subject and knowledge level

The subject of this textbook is the intersection of professional structural
engineering and the conceptual design of the built environment as shown
in figure 1.1. In general, an effective and efficient conceptual structural
design is based upon a profound practice of simplification and decom-
position techniques.
1.1 About this textbook 3
Conceptual
design
engineering
Structural
Figure 1.1: Conceptual structural design
However, the number and complexity of sophisticated high-end com-

puter programs and interfaces with other disciplines, increasingly dom-
inate daily practice of the present-day professional structural engineer.
Due to this on-going expansion of depth and breadth, the simplification
and decomposition techniques of the experienced structural engineer are
steadily disappearing from practice training and higher educational pro-
grammes.
Knowledge level The simplification and decomposition techniques for

conceptual structural design, as covered by the contents of this textbook,
are on an entry professional on graduate master level. Within both pro-
fessional and science graduate education, the focus can vary from a direct
practical relevance to an in-depth academic research relevance.
Zero measurement training Out of educational considerations the

arrangement of contents in this textbook is clustered around major sub-
ject matters, rather than following the overall chronological sequence
of activities of the structural engineering practice. The required know-
ledge, skills, and professional attitude have to be achieved by a mix
of the learning methods lecture, training, and project work. Lectured
theories and trained engineering practice can be applied during project
4 Introduction
work. This textbook supports both lecture and training, in class as well
as self-tuition.
zero measurement training, as listed in table 18.1 on page 216. The zero
measurement training can be applied as a self-check at the beginning of
the learning process.
1.1.3 Textbook outline

This textbook is the follow-up of a previous academic research on the
complex problem-solving process of structural conceptual design [9]. The
so explored methodical approach on conceptual structural design leads
to a controlled build-up of insight into the behaviour of the structure
and supports the actual successive design decisions during the concep-
tual design phase on the basis of the following coherent set of solution
components:
Structural design cycle Capturing structural design by breaking down
this complex process into the essential absolute minimum, result-
ing in a fundamental design cycle as an effective characterisation of
both the overall design process, and the individual design phases.
Basic structural forms Basic set of two-dimensional subsystems, with
each individual subsystem being an assembly of directly connected
structural elements, designed to act together to resist loads.
Structural design path The structural design can be explored in a
two-dimensional matrix, in which the design path follows the fun-
damental dimensioning routine from structural integrity, via load
distribution, to failure mechanisms.
Structural design loops A combination of analysis, check, orienta-
tion, and correction loops to support a cyclic optimum design with
regard to optimisation of both the quality of design outcome and
the number of design cycles.
Shared knowledge-based conceptual design Definition and collec-
1.2 Reading guide 5
tion of the fundamental conceptual design parameters of the most

influential participating disciplines can serve as a joint breeding
ground for integral design solutions.
Load path design A first draft of the structure’s integrity requires

simple and clear three-dimensional modelling on the level of
axial forces, directly or with a truss-analogy, including the three-
dimensional effects.
Load distribution parameters Applied mechanics-based oversight of

force- and deformation-driven load distribution and a matching
balanced set of conceptual design approximations.
Dimensioning parameters Applied mechanics-based oversight of pro-

fessional practice dimensioning and a matching balanced set of con-
ceptual design approximations.
1.2 Reading guide

1.2.1 Partition of the textbook outline
The analysis and definition of the problem, the solution in the form of a
combined physical and process decomposition, the quantification of the
individual design parameters, and a case study and training programme
are divided in four identifiable partitions as shown in figure 1.2.
Part I Problem definition For effective problem solving, the prob-

lem is divided into two partial problems “lack of insight into conceptual
structural design” and the underlying “lack of insight into structural per-
formance”.
Part II Conceptual structural design The solution to the lack of

insight into conceptual structural design consists of a combined physical
and process decomposition.
6 Introduction
Problem Conceptual Structural Case study

definition structural design performance and training
Demands
Mechanics
Materials
Part I Part II Part III Part IV
Figure 1.2: Partition of the textbook outline
Part III Structural performance The solution to the lack of insight

into structural performance consists of qualification and quantification
of the structural engineering fundamentals in the form of conceptual
structural design parameters.
Part IV Case study and training Both qualified and quantified

solutions are clarified by means of a case study and a series of trainings,
including a zero measurement training.
1.2.2 Introductions
Each part starts with a chapter “Introduction” which describes the reason
for its existence and a chapter guide of the specific part within this
textbook.
Accompanying figures and descriptions clarify the coherence between
chapters and the content of individual chapters.
Part I
Problem, analysis and

solution approach
Chapter 2
Introduction to part I
A lack of insight of the professional into structural engineering is con-

sidered to be a main obstacle for an overall performance/cost optimisa-
tion of the built environment. A methodical approach with effective solu-
tions to this lack of insight is the subject of this textbook. Present-day
problems, orienting analyses, and a solution approach with corresponding
solution components are described in part I with a chapter arrangement
as shown in figure 2.1.
Present-day problems High-end computer programs, numerous in-

terfaces with other disciplines, and a tendency towards more implicit
performance-based provisions increasingly dominate daily practice of the
present-day professional structural engineer. Due to this on-going ex-
panding depth and breadth, the simplification and decomposition tech-
niques of the experienced structural engineer are steadily disappearing
from practice training and higher education programmes.
Because of this, the understanding and reliability of the young profes-
sional in structural engineering is on the decline. Computerised designing
without sufficient insight, particularly in the conceptual design phase,
is a dangerous operation both out of economical and safety points of
view.
10 Introduction to part I
Part I
Present-day problems
Chapter 3
Solution approach
Chapter 4
Solution components
Chapter 5
Part II
Figure 2.1: Reading guide for part I
Solution approach A thorough definition of the problem is a cru-

cial step in the problem-solving process, because an effective problem
definition is directing the solution approach.
For a better understanding, and effective problem solving, the lack of
insight of the professional into structural engineering can be divided
into the two partial problems, “lack of insight into conceptual struc-
tural design” and the underlying “lack of insight into structural perform-
ance”.
The field of practice shift from calculating/checking to modelling/design-
ing requires the following three pillars for a future proof structural design:
universal systematic thinking, applied systems thinking, and applied
mechanics-based modelling.
Solution components In general, complex problems can be effect-

ively approached on a relatively high level of abstraction with an elab-
orate knowledge of the fundamentals and a simultaneous working “from
the whole to the part” and “from coarse to fine”.
The methodical approach is aimed to be a conceptual design tool for
11
experienced structural engineers, rather than a black box operated by

data typists. Therefore, the solution is carried out in the field of ab-
stractions of a decomposed system instead of algorithm-based numerical
power.
In order to secure durability of the design method, the solution is focused
on being overall applicable to construction types and the use of timeless
elements such as mechanics.
12 Introduction to part I
Chapter 3
Present-day problems in
structural engineering
3.1 Structural (un)safety in the Netherlands

3.1.1 Structural failures
In the Netherlands, there has been a notable amount of structural failures
over the last decades. For example, a lot of roofs and parking decks
have collapsed while in use, as well as complete buildings, mostly during
construction.
Some news-breaking structural failures include:
Theatre Het Park, Hoorn, 2001 Collapse of the theatre tower dur-
ing construction due to a combination of multiple engineering and
construction errors.
Hotel Van der Valk, Tiel, 2002 Parking deck collapse due to a lat-
eral torsional instability and subsequent horizontal displacement of
the supporting beams.
Bos en Lommerplein, Amsterdam, 2006 Near collapse of support-
14 Present-day problems in structural engineering
ing parking garage beneath a residential complex due to missing

concrete reinforcement.
Stadium De Grolsch Veste, Enschede, 2011 Roof collapse during
construction due to a loading of the incomplete stabilised roof
structure.
Queen Juliana bridge, Alphen aan den Rijn, 2015 Pontoon-based
crane collapse during construction due to severe shortcomings in
construction engineering and management.
Eindhoven Airport, 2017 Floor collapse due to an unusual orienta-
tion of wide-slab flooring in combination with insufficient overlap-
ping of the reinforcement splices.
AFAS Stadium, Alkmaar, 2019 Roof collapse due to engineering er-
rors with respect to wind loading and weld strength of the roof
structure.
3.1.2 Structural safety codes of practice

In Eurocode [19] structural safety is defined as the capacity of a structure
to resist all action(s), as well as specified accidental phenomena, it will
have to withstand during construction work and anticipated use.
Eurocode EN 1990 [14] further defines reliability as the ability of a struc-
ture or a structural member to fulfil the specified requirements, including
the design working life, for which it has been designed. Structural reli-
ability covers in fact four aspects: safety, serviceability, durability, and
robustness of a structure.
The semi probabilistic level I calculations in the material-related Euro-
codes are based on the assumption that an element is sufficiently reliable
if a certain margin is present between the representative values of the
load effect and the resistance as shown in figure 3.1.
The representative value of the load effect S has a 5% probability of
overshooting, whereas the representative value of the resistance R has a
5% probability of undershooting. The use of probability of exceedance
3.1 Structural (un)safety in the Netherlands 15
Probability density
Resistance R
Load effect S
Magnitude
Figure 3.1: Structural safety Eurocode
instead of mean values incorporates the influence of probability distribu-

tion.
The resistance is the capacity of a structure to resist a load effect. The
verification implies that the resistance R has to be greater than or equal
to the load effect S. The risk of failure when R < S should be sufficiently
low.
When there is no margin between the representative values of S and R
the probability of failure approximately amounts to Pf  10 1 , which
is sufficient for the Serviceability Limit State (SLS). However, for the
Ultimate Limit State (ULS) this probability of failure Pf  10 1 is
entirely insufficient and should be upgraded towards Pf  10 5 .
The corresponding margin between the representative values of S and R
can be obtained by global partial factors f and m for load effect and
resistance, respectively. The verification of the safety is then based on
an equation of the following type:
R
S· f  (3.1)
m
With these partial factors stochastic variability is covered, which is re-

lated to uncertainties in loads, materials, geometry, and calculation mod-

els. However, stochastic variability does not include gross human er-
rors.
3.1.3 Actual structural safety

The actual structural safety in the Netherlands with a probability of fail-
ure over the last decades P ⌧ 10 5 satisfies easily the required structural
safety level of the codes of practice with a probability of failure of about
P  10 5 . However, the actual safety (P ⌧ 10 5 ) is not a subset of the
regulated safety (P  10 5 ) as substantiated by Terwel [25]:
An extensive study of structural failures in the Netherlands has shown
that the current number of fatalities among residents due to structural
failures remains within assumed acceptable limits, although a high im-
pact - low probability disaster did not occur in the observed time inter-
val. This study showed also that about 90% of the failures are caused
by human errors, although human behaviour is not included in the prob-
abilistic calculation approach of the Eurocode. It seems a paradox that
the individual risk remains within acceptable limits, although the main
influencing factor, human error, is not included in the calculation ap-
proach. This can be explained because the actual strength of structures
is often higher than the calculated strength due to redundancy.
3.2 Lack of insight of the professional

In recent years, in-depth investigations of a number of specific disasters
have been undertaken by the Inspectorate for Housing, Spatial Planning
and the Environment, research organisations, the Dutch Safety Board,
university professors, expertise firms, and specially convened committees
of enquiry.
These investigations reveal that it is almost never possible to identify
one single cause for a disaster. Mostly, it is a combination of factors
and circumstances - which are an inherent part of the participants of the
building process - that can be identified. All these participants influence
3.2 Lack of insight of the professional 17
the structural safety of a building with their actions and interdependen-

cies. Many failures, however, arise in the design phase.
Structural collapse in the Netherlands appears to be mainly a combin-

ation of the lack of supervision during all project phases and a lack of
insight of the professional into structural engineering, as recorded in the
problem statement “Castle or House of Cards” under the management of
the Ministry of Housing, Spatial Planning and the Environment [23].
For the lack of insight of the professional into structural engineering this
problem statement addresses some major perceptions of the Ministry of
Housing, Spatial Planning and the Environment:
• Many in the construction industry realise that the level of pro-

fessional skill among structural engineers, but also among other
players, is on the decline.
• University professors in the field are noticing a general erosion of

knowledge and command of applied mechanics, the mainstay of the
structural engineering profession.
• The “black box” character of calculation software will further di-

minish people’s understanding of the subject.
Eight organisations, including the Inspectorate for Housing, Spatial Plan-

ning and the Environment, the Concrete Association, and structural en-
gineers and builders’ organisations have published the joint Compendium
for a Structural Safety Strategy [24].
This compendium contains a detailed description of how structural safety

can be guaranteed in the various phases of the design and building pro-
cess and what roles the various participants in the building process can
play with regard to structural safety. Since 2018 this compendium has
evolved into a national internet platform for structural safety “Kennis-
Portaal Constructieve Veiligheid (KPCV)”.
3.2.1 Problems, causes, and effects

The overall problem of structural unsafety in the Netherlands is caused
by many shortcomings and developments. This textbook will discuss the
problem of a lack of insight of the professional into structural engineer-
ing, with corresponding causes and effects, as recorded in the problem
statement [23].
Two partial problems For a better understanding, the lack of insight

of the professional into structural engineering can be divided into two
partial problems:
1. Lack of insight into structural performance on micro level; human

error and inadequacies of people working on building projects.
2. Lack of insight into conceptual structural design on macro level;

problems relating to the structure and culture of the building sector.
Structural performance The lack of insight into structural perform-

ance is generally accepted to be caused by the following:
• General erosion of knowledge and command of applied mechanics,

the mainstay of the structural engineering profession.
• Present-day extensive use of calculation software, essentially a “black

box”, further diminishing people’s understanding of the subject.
Conceptual structural design The lack of insight into conceptual

structural design is generally accepted to be caused by the following:
• Increasing number and complexity of interfaces with other discip-

lines and corresponding collaboration processes.
• Increasing complexity of contractual models such as the Design,

Build, Finance, Operate, Transfer (DBFOT) model and the Value
for Money (VfM) model.
3.2 Lack of insight of the professional 19
Safety and costs Foregoing causes can have considerable effects on

both safety and costs:
• The lack of insight and especially improper use of advanced com-

puter programs can put structural safety at severe risk.
• An uncontrolled design process can bring about insufficient per-

formance/cost optimisation and a lot of failure costs.
• Besides a tendency for excessive numerous functional requirements,

process control and reliability of a tender build-up is obviously
endangered by a lack of insight of participating professionals.
3.2.2 General requirements structural engineering

For competent structural engineering, the compendium [24] outlines the
following general requirements:
• Experience with like projects.
• Adequate knowledge about the required type of structure with re-

gard to the structural behaviour: actions, materials, structural ac-
tion, fire resistance and, if applicable, dynamic effects and fatigue.
• Adequate knowledge about the required type of structure with re-

gard to the integral aspects: construction, architecture, durability,
maintainability, and sustainability.
• With attention to the geotechnical engineering, including interac-

tions between structure, foundation, construction activities, and
structural environment.
• The ability to judge the results of automated design tools.
• Insight into the interaction between detailed design and the beha-
viour of the structure.
3.2.3 Recommendations structural design

With regard to structural design, the compendium [24] outlines the fol-
lowing recommendations:
• Aim for an as-clear-as-possible structural design concept, with re-

gard to the overall load distribution and the decisive failure mech-
anisms.
• Predicated on safety, e.g. structures must not collapse without due

warning.
• Be aware of the potential consequences of a structural failure and

design a second method of support to ensure that the forces are
dispersed elsewhere when a vital structural component can give
way.
• Apply comprehensive approximated design calculations as a check

for black box automated complex design tools.
3.3 Historical perspective

3.3.1 Developments professional structural engineering
With increasing complexity of structures and corresponding design, the
master builder of ancient times inevitably altered into a team of special-
ists, as shown in figure 3.2 and clarified in the following subsections:
Homo universalis Master builder with expertise on the fields of ar-

chitectural, structural, and construction engineering. This master
builder is figured as the “Vitruvian Man” by Leonardo Da Vinci
[28], based on the correlations of ideal human proportions with
geometry described by the ancient Roman architect Marcus Vit-
ruvius Pollio [29].
Expanding depth and breath Structural engineering as a separated

formalised discipline.
3.3 Historical perspective 21
On-going expanding depth and breath Specialisations within the

professional field of structural engineering.
Professional depth
Ongoing expanding
depth & breadth
Specialisations
Expanding
depth & breadth
Disciplines
Interfaces Built Environment
Figure 3.2: Expanding depth and breadth
3.3.2 Homo universalis

Structural engineering has existed since mankind started to construct its
own structures. Throughout ancient and medieval history all architec-
tural, structural and construction design was carried out by one person;
often an artisan in the role of master builder. Structural comprehension
was extremely limited and almost entirely empirically based.
The physical sciences underlying structural engineering began to be un-

derstood during the Renaissance in the late 15th century. It was then
that architectural, structural, and construction design evolved into a
more profound and controllable knowledge level but was still in the hands
of one person, then called “Homo universalis”.
The Latin expression “Homo universalis” can be translated to “Universal

person”, meaning a person with a broad knowledge of several fields and
often with proficiency or accomplishments in at least some of these fields.
Many notable universal persons lived during the Renaissance period such
as Leonardo da Vinci and Michelangelo.
Until the 19th century, only one person was needed to integrally oversee
the design, a generalist with profound expertise on the fields of architec-
tural, structural, and construction engineering. The term “Generalist”
is used to contrast this general approach to knowledge to that of the
“Specialist”.
3.3.3 Expanding depth and breadth

With the development of specialised knowledge of structural theories,
which emerged during the industrial revolution in the late 19th century,
structural engineering came into existence as a more defined and form-
alised discipline.
The volume of knowledge of materials, technologies, and construction

methods was increasing and structures became more complex. Due to the
limited ability of comprehension of each individual professional, the field
of building engineering was inevitably split into the separate disciplines of
architecture, structural engineering, and construction engineering.
The modern structural engineer can rely on a long history of constant

validation of theoretical approaches, building up extensive knowledge
data bases such as applied mechanics-based structural analyses, previ-
ous designs, design rules, design codes of practice, and numerous re-
searches.
To complete any project, it now takes a team of professionals that in-

cludes structural engineers working with other disciplines such as mech-
anical, geotechnical, electrical, and civil engineers, and urban planners
and architects.
3.3 Historical perspective 23
3.3.4 On-going expanding depth and breadth

The volume of knowledge of materials, technologies and building meth-
ods is still increasing enormously. Furthermore, there is a tendency away
from the explicitly deemed to satisfy provisions towards more implicit
performance-based contracting. This inevitably asks for corresponding
expertise and brings with it ever more in-depth specialisation.
Within the field of structural engineering alone there is so much ex-
pertise that a structural engineer can never master it fully, resulting
in specialisations such as geotechnical engineering, pre-stressed concrete
engineering, finite elements engineering, and bridge engineering.
Numerous sophisticated high-end automated design tools increasingly
support the daily practice of the present-day professional structural en-
gineer. In spite of, or perhaps just because of these extensive design
tools, the understanding and reliability of young professionals in struc-
tural engineering decreases dramatically.
They lack a fundamental understanding of structural behaviour, and
they lack an overview and insight into the conceptual design process and
related interfaces. In short, they lack the ability to abstract the basic
design parameters of form, material, and dimension.
Chapter 4
Solution approach
4.1 Problem definition

4.1.1 The merit of a proper problem definition
Albert Einstein is quoted as having said that if he had one hour to save
the world he would spend fifty-five minutes defining the problem and
then five minutes solving it [4].
Furthermore, he is quoted as having said that the significant problems
we face could not be solved at the same level of thinking we were at when
we created them [6].
So finding a methodical approach for the complex conceptual structural
design requires an elaborate problem definition on a high abstraction
level. Subsequently, working out the solution is merely a derivative
activity.
4.1.2 An interface control approach

For a better understanding, and effective problem solving, the lack of
insight of the professional into structural engineering is divided into the
two partial problems “lack of insight into structural performance” and
26 Solution approach
“lack of insight into conceptual structural design”, as appointed in sub-

section 3.2.1 and clarified in figure 4.1.
Customer demand Requirements & conditions
Performance demand: Structural demand: Construction demand:

- Functional reliability - Structural safety ULS - Material demand
- Structural reliability - Serviceability SLS temporary structures
- Redundancy - Durability DLS - Equipment demand
- Architectural demand - Material demand - Manpower demand
- Flexibility - Time demand
- Durability
- Maintainability
- Sustainability

= Interface control
Demand - Knowledge
Structural performance
Mechanics - Material
Material applications
Applied mechanics
= Interface control
Body of knowledge Scientific research

built environment structural engineering
Figure 4.1: Structural design as an interface control approach
The relation of performance, structural, and construction demand with

the performance/cost optimisation of the built environment will be cla-
rified in figure 7.1 on page 60 and corresponding explanatory text.
4.1 Problem definition 27
The problem of structural performance Present-day understand-

ing of structural performance is characterised by a constant expansion of
complex analysis tools, without an adequate control of the fundamental
interface between applied mechanics and the material applications.
The problem of conceptual structural design Present-day con-

ceptual structural design is characterised by a constant expansion of
requirements, related interfaces, and collaboration models, without an
adequate organisation of the brought about possibilities by controlling
the fundamental interface between the body of knowledge of the built
environment and the demands of the customer.
4.1.3 Structural engineering activities and missing tools

Professional structural engineering encompasses both analysis and design;
where analysis is related to structural performance and design is re-
lated to requirements, conditions, and interfaces with the built environ-
ment.
An effective structural design process will be characterised by conver-

gence and optimisation, based on a progressive insight into the beha-
viour of the structure, and an integral control of all influential boundary
conditions.
The corresponding structural engineering activities and tools as listed in

table 4.1 reveal an obvious deficiency with regard to conceptual struc-
tural design tools.
The interface between applied mechanics and material applications is

underdeveloped and the interface between structural demand and the
built environment is not yet developed.
4.1.4 Definition of structural performance

Structural performance and its difference from structural action, is cla-
rified by the following definitions.
Structural engineering activities and tools

Activity Subject Tool
Code Static scheme: Finite Element Method
checking - Load distribution
- Displacement SLS
Code check: Design codes of practice

- Sectional strength
- Stability
Optimisation: Computerised frame an-

- Trial and error alysis and code checking
Structural Fundamental insight into Applied mechanics
analysis structural performance:
- Equilibrium Background documents
- Force-driven parameters design codes of practice
- Deformation-driven
parameters Interface between ap-
- Durability driven plied mechanics and
parameters material applications
is underdeveloped
Conceptual Fundamental overview Interfaces between
structural of basic interfaces with structural demand
design the built environment: and performance and
- Performance demand construction demand
- Structural demand are not developed
- Construction demand
Table 4.1: Structural engineering activities and tools
Structural performance Structural performance is a collective term

for the following structural requirements:
Structural safety - ULS The safety of a structure or structural mem-
ber is prescribed by its Ultimate Limit State (ULS). If structural
4.1 Problem definition 29
behaviour is beyond this undesired state, one of the following fail-

ure mechanisms may occur: collapse due to loss of equilibrium of
the structure; fracture due to excessive internal stresses or fatigue;
or instability due to insufficiency or unbalance of the ground.
Serviceability - SLS The functionality of a structure or structural

member is prescribed by its Serviceability Limit State (SLS). Ser-
viceability in structural design includes the following cases: deflec-
tion of a beam under loads; deformation like swing during vibra-
tion; or tilting of a column under external actions.
Durability - DLS The durability of a structure or structural member

is prescribed by its Durability Limit State (DLS). The ability of
structural members to perform adequately for normal use during
the characteristic working life can among others be endangered by
material deterioration that can within time decrease the strength
capacity and serviceability of these members.
These specifically in the structural design codes of practice prescribed

structural requirements ULS, SLS, and DLS, are part of a more over-
all sustainability objective to minimise the negative environmental im-
pact by enhancing efficiency and moderation in the use of materials and
energy.
Structural action Structural action can be characterised by the way

in which a structure resists the loads acting on it - incorporating the
load distribution within the structure; the corresponding deformation
of the structure; and the strength of structural members to resist these
loads.
The typifications “structural action” and “structural performance” show a

high degree of similarity, except for the modern more specific contractual
specifications with regard to functional behaviour and durability aspects.
For this reason, both typifications are distinguished from each other,
and the broader-ranging structural performance is therefore deliberately
included in the problem definition.
When principally applied mechanics calculations are addressed, the typi-

fication “structural action” instead of “structural performance” will be
used. Particularly, optimisation calculations with respect to load distri-
bution, deformation, and strength will be addressed as an optimisation
of the structural action.
4.1.5 Definition of conceptual structural design

Conceptual structural design and the identical integral design are clari-
fied by the following definitions.
Conceptual structural design Conceptual design refers to the cre-

ation phase of the design of the built environment or objects within this
built environment such as buildings and civil works. Particularly, this
design phase is characterised by a multitude of co-operating disciplines
and mutual interfaces.
Within this multi-disciplinary design phase, all disciplines in general are

responsible for an overall optimisation of the performance/cost-ratio of
the life cycle; the structural engineer in particular is responsible for a safe
and efficient structural design within this overall optimisation.
With regard to a safe and efficient structural design, it should be em-

phasised that understanding structural performance is the starting point
and an absolute prerequisite for conceptual structural design.
Integral design Integrality of design for the benefit of overall optim-

isation should ideally completely take place during conceptual design in
general, and conceptual structural design in particular. In this text-
book therefore, conceptual design is completely synonymous with integ-
ral design and “conceptual structural design” automatically implies full
integrality of design.
4.2 Present-day solutions’ field of practice 31
4.2 Present-day solutions’ field of practice

4.2.1 Copying reality
Most of the present-day solutions reducing failure and corresponding
costs are aimed at a complete and an as-accurate-as-possible procedure
capturing reality.
The most thorough procedure capturing reality is copying reality:
1. Qualify the complete set of structural engineering aspects, includ-
ing all the related interfaces with the built environment.
2. Quantify all these aspects.
3. Describe all the interfaces between these quantified aspects with
unambiguous processes.
4. Combine these processes in one converging flow diagram.
Most of the present-day solutions aim at such a copy of reality, as accur-
ate as possible, often in one of the following forms:
Planning and control A data-driven approach concerned with plan-
ning and controlling all aspects of a process.
Numerical power Mostly in the form of advanced structural analysis
applications or even a modest expert system, capturing an expert’s
knowledge by encoding it in a computer program.
4.2.2 Planning and control

The most usual present-day approach to reduce failure and corresponding
costs is by establishing elaborate planning and corresponding control
activities.
PDCA cycle The concept of planning and control is based on the

scientific method of “hypothesis, experiment, and evaluation” or “plan,
do, and check”. Later developed into the present-day customary Plan,
Do, Check, and Act (PDCA) cycle as an iterative four-step management

method for the control and improvement of processes and products.
For an effective planning-control relationship, an accurate adjustment of
the control activities to the planning is of importance:
• Establishing measurable standards together and corresponding with
the objectives.
• An interaction between planning and control leading to changes
occurs when taking corrective action with the final step of the
control. This can take several forms, but two of the most effective
are to change the objectives or alter the plan.
• A design plan must provide the framework for the design team
control system. When objectives and plans change for whatever
reason, control standards should change accordingly.
Systems engineering Systems engineering is an interdisciplinary ap-

proach and is used for enabling the realisation and deployment of suc-
cessful systems. Systems engineering integrates other disciplines and
specialty groups into a team effort, forming a structured development
process that spans the whole system lifecycle.
Systems engineering became synonymous with the overarching respons-
ibility for the development of the complete end product and enabling
products. This role has increasingly expanded until the present; it now
also being responsible for the interface between the complete device and
the user, and even with the system’s eventual disposal. Interface design
and specification are concerned with assuring that the pieces of a sys-
tem connect and interoperate with other parts of the system, and with
external systems when necessary.
In general, systems engineering proceeds in the following steps:
1. Formalise the approach.
2. Control the overall system, regarding the entire life cycle, all inter-
faces included.
3. Apply explicit, clear, and provable selection processes.
4. Define interactions between requirements, objects, and organisa-

tions.
Building information modelling Building Information Modelling

(BIM) is a digital representation of physical and functional character-
istics of a facility. A building information model is a shared knowledge
resource for information about a facility forming a reliable basis for de-
cisions during its life cycle.
A basic premise of BIM is collaboration by different stakeholders at dif-

ferent phases of the life cycle of a facility to insert, extract, update, or
modify information in the BIM to support and reflect the roles of that
stakeholder.
For the professionals involved in a project, BIM enables a virtual inform-

ation model to be handed from the design team to the main contractor
and subcontractors and then on to the owner/operator; each professional
adds discipline specific data to the single shared model. This reduces in-
formation losses, which traditionally occurred when a new team took
“ownership” of the project, and provides more extensive information to
owners of complex structures.
Traditional design was largely reliant upon two-dimensional (2-D) draw-

ings. BIM extends this beyond the three-dimensional (3-D) physical
geometry with time as the fourth dimension (4-D) and costs as the fifth
(5-D), etc. BIM therefore covers more than just geometry. It also covers
spatial relationships, light analysis, geographic information, and quant-
ities and properties of building components.
4.2.3 Numerical power

Numerous sophisticated high-end automated design tools increasingly
support daily practice of the present-day professional structural engineer.
Finite element method Present-day form-free complex architectural

designs require an elaborate and profound structural analysis on a three-
dimensional level. For a corresponding analysis on stress level, three-
dimensional finite elements are applicable.
Advanced structural analysis may examine:
• Dynamic analysis; natural frequencies and frequency response.
• Geometric nonlinear analysis of second-order behaviour; linear

stress-strain relationship and large displacements.
• Material nonlinear analysis of plastic load-carrying capacity; non-

linear stress-strain relationship and small displacements.
• Induced deformation analysis of the structure, based upon geotech-

nical failure mechanisms and corresponding deformations.
Because of the present-day availability and self-evidence of these sophist-

icated high-end automated structural analysis tools, utilisation of these
tools for even the simplest structural problem seems appealing.
Parametric design Parametric design is a type of rule-based model-

ling where geometric constraints and also scripting are used to ensure
that the main objectives of the design intent within a project are pre-
served. It is about the use of variables and algorithms to generate a
hierarchy of mathematical and geometric relations to explore the whole
range of possible solutions that the variability of the initial parameters
may allow. As a result, design teams are able to generate innovative
forms.
Parametric design is not limited to only constraining geometry; it can

also be used to define and constrict relationships such as thermal prop-
erties and material strength. With the coupling of finite elements, mod-
elling material dimensions and even generative design with optimised
organic structural load paths can be obtained.
Expert systems An expert system represents information and searches

for patterns in that information. They are known as expert systems be-
cause they model how a human expert analyses a particular situation
by applying rules to the facts, or compares the current case with similar
cases, in order to reach a conclusion.
The underlying concept of an expert system is that it is possible via a
series of carefully structured interviews, to capture an expert’s knowledge
and to encode it in a computer system in such a way that the system is
able to mimic the decision-making behaviour of the expert.
Expert systems can include different types of reasoning such as case
based, sequential based, rule based, and fuzzy logic.
Expert systems for conceptual structural design Computers do

not possess common sense, so when an expert system is pushed outside
the bounds of its knowledge, it has no way of judging whether or not
something is sensible.
Much of the knowledge that an expert possesses is in a form that cannot
be expressed clearly. Also the shear volume and depth of the information
is such that it is just not possible to capture it all. Maintaining and
updating knowledge bases is a demanding task that requires specialist
staff.
In reality, matching a previous design to new design requirements is more
complex than can be achieved by simple techniques. Also, automatically
modifying the design proved to be a very difficult task.
4.2.4 Professional higher education

In the hands of experienced conceptual designers, present-day availab-
ility of sophisticated high-end automated structural analysis tools can
contribute to conceptual structural design on a detailed scale such as an
exploratory analysis of complex structural action.
In the hands of inexperienced young professionals however, the concep-
tual design capabilities of these sophisticated tools diminish rather than
improve. Computerised designing with insufficient insight - particularly

in the conceptual design phase - is a dangerous operation both from an
economical and a safety point of view.
Due to the on-going expansion of high-end automated design tools, the
simplification and decomposition techniques of the experienced struc-
tural engineer are steadily disappearing from practice training and higher
education programmes.
University professors in the field are noticing a general erosion of know-
ledge and skills of applied mechanics, the mainstay of the structural
engineering profession [23].
Furthermore, education about structural design in general, and concep-
tual structural design in particular, lacks an integral approach of the
educational programme and corresponding emphasis on the interfaces
between the disciplines.
4.3 Solution approach structural design

4.3.1 A need for control on system level
As with a lot of complex design problems, the standard planning and
control mode is not enough to guaranty a satisfying solution as this
formalistic, more bookkeeping-like approach only supports basic process
control.
With regard to the quality of the design solution even planning and
control with elaborate procedures and explicit supervision protocols - as
proposed in the joint Compendium for a Structural Safety Strategy [24]
- merely gives an illusion of control and reliability.
Furthermore, elaborate process control with an overkill of numerous reg-
ulations and control systems stifles creativity, progress and cooperation.
Especially process control and reliability of a conceptual design - and in
particular the preceding tender build-up - is endangered by a present-
day tendency towards excessive numerous functional requirements and
control procedures.
4.3 Solution approach structural design 37
On the other hand, effective proven design tools such as systems engin-
eering and BIM do offer control and clarity to open the way to creativity,
progress and cooperation. Individual components of these applications,
such as decomposition techniques and applied mechanics-based calcula-
tion routines can be very useful for a solution approach on conceptual
structural design.
4.3.2 Shift from calculating to modelling

As a result of modern sophisticated automated design the field of prac-
tice shifts from calculating/checking to modelling/designing. This shift
requires a fundamental understanding of modelling in combination with
research skills:
1. Modelling of load distribution in complex structures and model-
ling of material behaviour of new structural materials, new applic-
ations of existing materials and new production techniques.
2. Research skills as an effective and efficient problem-solving tool
for complex structural problems.
Partition of research skills For the professional field of practice, re-

search skills can subsequently be divided into universal systematic think-
ing and applied systems thinking:
2a. Systematic thinking based on the scientific research method
about problem definition, research framework, hypothesis, valid-
ation, and conclusion.
2b. Systems thinking as a holistic approach from the whole to the
part and from coarse to fine, regarding complex interfaces, struc-
tural integrity, load distribution, and failure mechanisms.
4.3.3 Three pillars of future-proof structural design

In conclusion, the field of practice shift from calculating/checking to
modelling/designing requires the following three pillars for a future proof
structural design as listed in table 4.2: universal systematic thinking,

applied systems thinking, and applied mechanics-based modelling.
Systematic thinking Systems thinking Modelling
Scientific method:
1. Problem definition
2. Research framework
3. Hypothesis
4. Validation 2-D
5. Conclusion
Table 4.2: Three pillars of future-proof structural design
Systematic thinking Scientific research has to comply with the basic

principles of the scientific method. The essence of the scientific method
is to test a hypothesis, and replication of this testing should get the same
response; this response can be measured and recorded.
The following five steps can outline the scientific method:
1. State a problem and define a corresponding research question.
2. Investigate what is already known and structure the solution find-
ing by means of a research framework.
3. Formulate a hypothesis as a solution to the problem.
4. Test the hypothesis and analyse the results on whether to accept,
adjust, or reject the hypothesis.
5. Conclude, with recommendations for further research, and publish
the results.
The underlying goal or purpose of science to society and individuals
is to produce useful models of reality. To achieve this, one can form
hypotheses based on observations of reality. By analysing a number of
related hypotheses, scientists can form general theories. These theories
benefit society and individuals who make use of them.
4.3 Solution approach structural design 39
Systems thinking Systems thinking is about patterns and relation-

ships to describe how things interact and gain insight into why systems
behave the way they do. Systems thinking refers to various perspectives
or interpretations of reality.
Present-day system theories development is aiming at tools and meth-
ods to better comprehend and manage the complexity of the total sys-
tem life cycle. Modern developments include performance-based design,
multidimensional modelling management, and quantitative risk manage-
ment.
Because of the multitude of parameters and the complexity of interre-
lations, a workable and to optimisation leading numerical power-based
overall design method seems far ahead. In the long run, however, the
ultimate artificial intelligent expert system is surely a possibility.
The time gap between present-day needs and future successful applic-
ations of artificial intelligence will probably be large enough to excuse
for a more simplified applied system thinking for conceptual structural
design.
Applied system thinking for structural design can be based upon de-
composition techniques for a holistic approach from the whole to the
part and from coarse to fine, regarding complex interfaces, structural
integrity, load distribution, and failure mechanisms.
Modelling Due to the complexity of material behaviour, structural

analysis is completely dependent on abstract representations of the ac-
tual structure.
As an abstract representation, modelling has its limitations. For a
reliable application of structural modelling, awareness of these limita-
tions is of great importance; for example, when shear deformation is
dominant.
To perform an accurate analysis, the structural engineer must determine
information about structural loads, geometry, material properties and
support conditions. The results of such an analysis typically include
support reactions, member forces, and displacements. This information

is then compared to criteria that indicate the conditions of failure.
Hand calculations of the structural action are based on analytical for-
mulations that mostly apply to simple linear-elastic and ideal-plastic
analysis models.
Computer calculations of the structural action are generally based on the
finite element method, including the most commonly used displacement
method. It is a numerical method generated by theories of mechanics and
is applicable to structures of arbitrary size and complexity. The finite
element method also helps in producing stiffness and strength visualisa-
tions.
Regardless of approach, the formulation is based on the same three fun-
damental relations of equilibrium, constitutive - stress-strain relation-
ship, and compatibility - strength and stiffness transfer between ele-
ments. The solutions are approximate when any of these relations are
only approximately satisfied, or an approximation of reality.
Chapter 5
Solution components
5.1 In search of a methodical approach

5.1.1 Solution approach
Insight into both structural performance and conceptual structural design
is the primary requirement for a solution, based on universal fundamental
understanding instead of specific applicability.
Present-day innovative accurate and integral design programmes such as
expert systems are still small-scaled, insufficiently developed, and un-
reliable. The black-box character of such systems makes this lack of
reliability worse.
So the methodical approach on conceptual structural design focusses
on fundamental insight, by appointing and organising all factors that
have been identified as important to the problem and to corresponding
relationships.
5.1.2 Guiding principles

In response to the problem definition and subsequent analysis, guiding
principles are searched for, in order to get hold of a coherent methodical
42 Solution components
approach. The following highly-correlated components substantiate the

methodical approach:
T-shaped professional A balanced combination of in-depth under-
standing of structural performance and an in-breadth understand-
ing of conceptual structural design, directly in line with the prob-
lem definition as expressed in figure 4.1 on page 26.
Applied mechanics Insight through simplification by way of an ap-
proximate mechanics-based modelling - the mainstay of the struc-
tural engineering profession - and corresponding conceptual struc-
tural design parameters.
Designing with progressive insight A need for understanding - from
a safety and an economical point of view - based upon a controlled
built-up of insight by working “from the whole to the part” and
“from coarse to fine”.
Decomposition Working “from the whole to the part” through a phys-
ical decomposition of the system and “from coarse to fine” through a
process decomposition of the conceptual structural design process.
5.2 T-shaped professional

5.2.1 Modern demands
The problem of present-day collapses can be attributed to a lack of in-
sight of the professional into structural performance and integral design
as discussed in section 3.2.
The upcoming call for the return of the old-fashioned structural engineer
can be part of the solution. After all, such an engineer has sufficient fun-
damental insight into structural performance to secure structural safety
during conceptual design. Furthermore, this fundamental insight could
prove to be an effective guide to acquire and utilise present-day complex
software tools, codes, and research results.
However, the old-fashioned structural engineer would not be equipped for
5.2 T-shaped professional 43
present-day and future performance-based integral design, with complex

interfaces, contractual conditions, and collaboration processes. For an
effective performance/cost optimisation with regard to service life and
environment, one has to comply with modern demands.
5.2.2 T-shaped professional structural engineer
The problem definition of structural design as an interface control ap-

proach, as shown in figure 4.1 on page 26, represents these modern de-
mands.
Therefore, within the integral Body Of Knowledge of the Built Environ-

ment (BOK BE), the present-day professional in structural engineering
should possess a so-called “T-profile” as given in figure 5.1.
Conceptual structural design In-breadth understanding of

= Interface control conceptual structural design
Demand - Knowledge
In-depth understanding
structural performance
Mechanics - Material
= Interface control
BOK BE BOK BE
Figure 5.1: T-shaped professional structural engineer
This T-profile consists of sufficient in-depth understanding of structural

performance to secure the structural safety, and sufficient in-breadth un-
derstanding of conceptual structural design to function effectively within
an integral design team.
5.3 Applied mechanics

5.3.1 Necessity of insight into structural performance
Structural performance has to comply with structural safety, serviceab-
ility, durability, and sustainability over the life cycle of a structure as
defined in subsection 4.1.4.
Non-conformities with respect to these structural requirements have to
at the least be detected and corrected but preferably prevented by a
preliminary insight into the structural performance.
Structural failure Every failure could lead to a catastrophic struc-

tural collapse. And therein lies the problem for structural engineers:
engineering is the most unforgiving of professions.
Structural collapse is often a combination of causes and even if there is a
single cause, it can lead to a progressive collapse. Usually, the causes can
be attributed to human error: mistakes, misunderstandings, incompet-
ence, ignorance, dishonesty; every facet of human failing is represented
in construction failure.
Disaster is most likely when new designs are based purely on successful
precedents and basic lessons of past failures are ignored, or not fore-
seen; for example the Tacoma Narrows Bridge. Other forces can be at
work such as new materials, competitive pressure, and economical con-
straints.
Other non-conformities With structural safety being a self-evident

prerequisite, serviceability is an upcoming contractual requirement to
be met with regard to a controlled functional behaviour of the struc-
ture.
The further addition of durability as a contractual requirement is ini-
tiated through performance-based design as a logical step towards an
integral performance/cost optimisation of the built environment of the
entire life cycle. As a result, durability is extending to sustainability by
5.3 Applied mechanics 45
minimising a negative environmental impact.
Necessity of insight Present-day sophisticated automated black-box

design tools are an upcoming threat with respect to non-conformities. A
black box permits no insight into the process; the quality of the process
and the design outcome is completely dependent on the understanding
and overall insight of the programmer and at that time current know-
ledge, practices, and perceptions.
Computerised designing with insufficient insight, particularly in the con-

ceptual design phase, is a dangerous operation both from an economical
and a safety point of view. Insight into the conceptual structural design
process is a necessity for structural designers, who are directly respons-
ible for the structural performance.
5.3.2 Timeless applied mechanics

Especially the lack of insight of the young professional during the design
phase can be seen as a shortcoming of present-day professional education.
During the entire design process there must be a constant built-up of

insight into the load distribution in combination with decisive failure
mechanisms of the structure as a whole.
Insight into the load distribution and decisive failure mechanisms can
primarily be based on a thorough knowledge and application of mechan-
ics. The visibility of this application highly contributes to an insightful
controlled structural design process during all design phases, but em-
phatically during the conceptual structural design.
Applied mechanics-based modelling is and will be the mainstay of the

structural engineering profession.
In order to secure durability of the design method, this methodical ap-

proach focuses on an overall applicability for construction types, and the
use of timeless elements and particularly applied mechanics.
5.4 Designing with progressive insight

5.4.1 General problem approach
Conceptual structural design is such a complex problem-solving pro-
cess that it is in urgent need of increased accessibility to the field of
practice.
In general, complex problems can be effectively approached on a relative

high level of abstraction, with an elaborate knowledge of the fundament-
als and a simultaneous working “from the whole to the part” and “from
coarse to fine” as shown in figure 5.2.
From the whole to the part
[Level of decomposition]
s
es
oc
pr
g
in
lv
so
em
bl
o
Pr
From coarse to fine

[Level of specification]
Figure 5.2: Problem approach
The vertical and horizontal axes can be formulated somewhat more ab-
stract as “level of decomposition” respectively “level of specification”.
5.4.2 Progressive insight from estimation to accuracy

An effective structural design process is characterised by convergence
and optimisation, based on a progressive insight into the behaviour of
the structure, and an integral control of all influential boundary condi-
tions.
5.5 Decomposition 47
Working with an increasing number of related interfaces and using com-

plex computational analysis and code checking asks for insight and over-
sight during all design phases.
The three phases of a global structural design process are designing the
primary purpose of the structure, designing the basic structural concept,
and the calculation process.
Progressive insight from estimation to accuracy on (sub)system level:
1. Determination of the structural form and the choice of material.
2. Approximate dimensioning by manual calculation of the load dis-
tribution and failure mechanisms with basic applied mechanics.
3. Computerised two-dimensional framework calculation of the load
distribution and post-processed code checking.
For most buildings this will normally be sufficient, but high-rise buildings
and civil structures may need a more thorough investigation:
4. Computerised three-dimensional framework calculation of the load
distribution (three-dimensional torsion effects) and post-processed
code checking.
5. Computerised Finite Elements calculation on stress level of the
load distribution and post-processed code checking.
6. Computerised Finite Elements calculation on stress level for both
load distribution and material strength, code checking is no longer
applicable.
5.5 Decomposition
5.5.1 Abstractions of a decomposed system
Decomposition is a standard technique when dealing with complex sys-
tems and allows for certain abstractions. One way of decomposing may
allow for natural and elegant abstraction in further system description,
whereas other decompositions allow for less natural abstraction. So the

decomposition has to be chosen with respect to the possible abstractions
later.
Completeness When decomposing a system we need an argument

that the composition of all decomposed components makes the whole
system and its functionality complete again. For certain kinds of decom-
positions the completeness argument is easier than for others.
Loss of information When designing with abstractions of a decom-

posed system, the unavoidable loss of information to the system reality
needs to be taken into account. Decomposing a system inevitably res-
ults in loss of information at the borders of the partitioned subsystems.
Conscientiously chosen locations for the interfaces between subsystems
- often a natural way of decomposing - will decrease insight and reduce
loss, however.
Natural descriptions We use formal language to describe systems.

This language can be based on different disciplines in the formal, math-
ematical world:
• There are logics, where we have basic properties and propositional
and temporal operators to relate basic properties. Theorem provers
are suitable tools that require a logic-based system description.
• There are process algebras, wherein processes and different ways of
synchronisation between processes are the elementary bricks, and
wherein model checkers are the tools corresponding to the process-
algebraic way of system description.
• A physical system can also be described as a set of differential
equations, and classical mathematics forms the basis for solving
sets of differential equations and arguments about the functionality
described.
Additionally, the composition mechanism that corresponds to the chosen
decomposition is relevant. On the one hand, when this composition
mechanism is reflected by a language primitive of the description lan-

guage, the corresponding decomposition fits this language best. On the
other hand, each decomposition corresponds to a most ideal way of sys-
tem description.
5.5.2 Decomposition of complex systems

Particularly the complexity of the interdisciplinary interfaces makes both
the design process and the overall behaviour of a structural system com-
plex.
For an effective analysis of such a complex system, the following coherent

decomposition approach can be applied as brought up by Kickert [13] and
further substantiated by De Ridder [21]:
Physical decomposition Decomposing the physical parts of a system

as the most natural way of decomposition.
Process decomposition Decomposing the process phases of a system

such as design, construction, and operation.
Aspect decomposition Decomposing the aspect parts of a system such

as strength, aesthetics, and durability.
5.5.3 Physical decomposition

This way of decomposition follows the physical parts of a system. Often,
it is a very natural way of decomposition because we easily “see” all the
physical parts.
The completeness criterion of physical parts is easy to check: when we

have processed all physical parts, we have the whole system. Also, failure
of physical parts can be located naturally and therefore described more
straightforwardly.
The actual physical decomposition of a structural system into basic struc-

tural forms is further elaborated in chapter 9.
5.5.4 Process decomposition

When looking at chemical plants or production plants a recipe or a pro-
duction plan forms the “essence” of their functionality. In this case,
process decomposition is the most natural way and allows for the most
effective abstractions.
The focus in the process decomposition lies in the form of causal chains
that are relevant in the system to model.
The actual process decomposition of conceptual structural design into
individual fundamental structural design phases is further elaborated in
chapter 8.
5.5.5 Aspect decomposition

Particularly the interacting of the numerous aspect parts of the built en-
vironment such as functionality, costs, aesthetics, strength, redundancy,
constructability, flexibility, durability, maintainability, and sustainability
becomes complex when crossing disciplinary boundaries.
Although an integral conceptual design is a highly cyclic process, the
complexity lies in the interdisciplinary aspects rather than the process
itself. As a result of this complexity of the interdisciplinary interfaces, the
loss of information at the borders of the partitioned disciplines will be un-
acceptably large and decomposition cannot be applied effectively.
Then concurrent engineering of all system theories in general, and
decomposition-based methods in particular, will be an appropriate solu-
tion. Utilisation and corresponding preconditions is further elaborated
in section 7.3.
Concurrent engineering Concurrent engineering is a systematic ap-

proach to integrated product development that emphasises the response
to customer expectations. It embodies team values of co-operation, trust
and sharing in such a manner that decision-making is by consensus, in-
volving all perspectives in parallel, from the beginning of the product
life cycle.
Essentially, concurrent engineering provides a collaborative, co-operative,

collective and simultaneous engineering working environment. The con-
current engineering approach is based on the five key elements of process,
multidisciplinary team, integrated design model, facility, and software
infrastructure.
Concurrent engineering is recognised as a strategic weapon that busi-
nesses must use for effective and efficient product development. It is not
a trivial task, but a complex strategic plan that demands full corporate
commitment, and therefore strong leadership and teamwork go hand in
hand with successful concurrent engineering programmes.
Part II
The art of conceptual

structural design
Chapter 6
Introduction to part II
Both principal and design team are in search of an overall perform-

ance/cost optimisation. A lack of insight of the professional into struc-
tural engineering is considered to be a main obstacle for such an optim-
isation.
For this relatively unexplored problem, an exploratory research is con-

ducted by a systematically zooming in from the whole to the part and
from coarse to fine. The so explored methodical approach on conceptual
structural design consists of a process control component and an underly-
ing structural performance component. The process control component
is discussed in this Part II, “The art of conceptual structural design”.
The underlying structural performance component is discussed in the
following Part III, “Understanding structural performance”.
Part II starts with present-day complex and ambiguous conceptual struc-

tural design.
Then the conceptual structural design process is subdivided into the indi-
vidual components process decomposition, physical decomposition, and
a cyclic control of the process, as shown on the axes in figure 6.1.
56 Introduction to part II
Part III
Part II
Physical decomposition
ol
tr
10 on
Chapter 9
er c
pt ss
ha ce
C pro
lic
yc
C
Process decomposition
Chapter 8
Conceptual design
Chapter 7
Part I
Figure 6.1: Reading guide for part II
Conceptual design Conceptual structural design is a creative and di-

mensioning process in which the structural form, the materials and the
basic dimensions are determined, taking into account all influential as-
pects such as aesthetics, constructability, sustainability, and costs.
Until the present day, it appears challenging to capture performance-

based solutions out of the multitude of aspects and their complex in-
terrelations. Modelling these interrelations is capturing intuition; the
most ambiguous, and therefore the most intangible aspect of conceptual
design.
A definition and collection of the fundamental conceptual design para-

meters of the most influential participating disciplines can serve as a
57
joint breeding ground for a concurrent “shared knowledge-based concep-

tual design”.
For a controlled conceptual design process, both process and physical
decomposition techniques are inevitable.
Process decomposition Capturing structural design by breaking

down this complex process into the essential absolute minimum, res-
ulting in a “structural design cycle” as an effective characterisation of
both the overall design process and the individual design phases.
Physical decomposition A structure as a three-dimensional physical

system can be decomposed in clear two-dimensional “basic structural
forms”, with each individual structural form being an assembly of directly
connected one-dimensional structural elements, designed to act together
to resist loads.
Cyclic process control A complete control of the present-day com-

plex and highly cyclic conceptual design process is still out of reach.
Success in this design phase is dependent on the level of experience and
intuition of the conceptual structural designer.
Structural design can be explored on a “structural design path”, following
the fundamental dimensioning routine from structural integrity, via load
distribution, to failure mechanisms.
A cyclic optimum design with regard to optimisation of both the quality
of design outcome, and the number of design cycles can be supported
by “structural design loops” such as analysis, check, orientation, and
correction loops.
Solution components The so-obtained coherent set of solution com-

ponents for conceptual structural design, and corresponding chapter ar-
rangement, is listed in table 6.1.
58 Introduction to part II
Chapter arrangement and solution components

No. Chapter Solution component
7 Conceptual design - Shared knowledge-based
conceptual design
8 Process decomposition - Structural design cycle
9 Physical decomposition - Basic structural forms
10 Cyclic process control - Structural design path
- Structural design loops
Table 6.1: Solution components for conceptual structural design

Chapter 7
Conceptual design
7.1 Determination of the structural form

7.1.1 Conceptual structural design
A conceptual design team consists of a representative of the principal
and representatives of each discipline with a major design involvement
with regard to feasibility and the performance/cost ratio of the specific
required system.
After a reduction of the mostly numerous requirements and contractual

conditions into a controllable set of presumably dominant requirements
and conditions, the first conceptual draft on system level can be executed.
Only on this three-dimensional level an integral design is feasible. The
process of creating system outlines by the conceptual design team is
shown in figure 7.1.
Customary, this creation is a composition of basic structural forms on

object level. More innovative structural design, however, requires a sys-
tem creation on aspect level. On such an aspect level, a set of functional
requirements is directing the design of possible load paths instead of just
combining basic forms.
60 Conceptual design
Performance / Cost
optimisation may lead
to modification of the
requirements Requirements
& conditions
Performance demand: Structural demand: Construction demand:

- Functional reliability - Structural safety ULS - Material demand
- Structural reliability - Serviceability SLS temporary structures
- Redundancy - Durability DLS - Equipment demand
- Architectural demand - Material demand - Manpower demand
- Flexibility - Time demand
- Durability
- Maintainability
- Sustainability
Added value design:

Performance surplus
Performance/Cost Performance-based
optimisation cost minimisation
Iterative design loop

Code check
Code check may lead to

correction of the conceptual Form, material
structural design & dimension
Conformity between As built Construction

and requirements should be
secured through conceptual
structural design
As built
Figure 7.1: Conceptual structural design process

7.1 Determination of the structural form 61
For a coherent and complete integral design all fundamental demands,

namely performance demand, structural demand, and construction de-
mand, have to be complied with. Moreover, these three demands are,
one by one, indispensable:
Performance demand A system derives its reason for existence from
its function. Principally, performance demand has to secure this
function by means of explicit reliability demands. Furthermore,
architectural expression, future customisation, service life, and en-
vironmental sustainability can be demanded.
Structural demand The safety and serviceability of the structure are
inevitably conditional for a safe and functional use of the system.
Failing structural safety has proved to cause serious civil disturb-
ances.
Construction demand Design engineers optimise cost drivers of the
scheme - such as simplicity, uniformity, repetition, and phasing - to
be executed with respect to manpower, equipment, and material
demand. The cost optimisation of the execution is implicitly part
of the life cycle costs and is, among other things, dependent on
location and the individual contractor. Constructability in itself,
and corresponding construction time are inevitably conditional for
coming into existence of the system.
The quality of the system can best be valued by its performance/cost
ratio; with a performance in the broadest sense, including structural
and construction demands. Due to the large amount of design freedom,
the performance/cost ratio can best be established and optimised in the
creation phase. In the process of further specification, an optimisation of
the performance/cost ratio will turn out to be increasingly limited.
The system outlines that only comply with the requirements are so-
called “performance-based” solutions. On the basis of these requirements,
striving for full optimisation of the performance/cost ratio will lead to a
performance-based cost minimisation.
However, such a minimisation is not automatically the ultimate goal
of a principal and the design team; sometimes, supplementary added-

value solutions offer a considerable added-value for a relatively small
cost increase. Then, a real performance/cost optimisation, despite the
earlier established requirements, is preferable.
7.1.2 Design process

Conceptual structural design is a creative and dimensioning process, in
which the structural form, the materials, and the basic dimensions are
determined. The structural engineer should be involved in the project
from the start of the conceptual design phase because of the influence of
complex aspects such as functionality, costs, aesthetics, constructability,
and sustainability.
The structural form One of the main conceptual structural design

activities is the determination of the structural form, based on “under-
standing” and “order of magnitude” of standard structural forms, funda-
mental insight into structural behaviour, and an overview of the basic
interfaces with the built environment.
Materialisation The choice of material is dependent on a combina-

tion of structural, construction, and architectural demand parameters;
structural demand with respect to reliability, redundancy, and durabil-
ity; construction demand parameters for both permanent and tempor-
ary structures such as construction time, mass, simplicity, and uniform-
ity; and finally, architectural demand, which can span a wide spec-
trum of expressions from power and massiveness to minimalism and
transparency.
Dimensioning Dimensioning of a materialised structural form is a

quantification process. The load distribution on the system, through
the subsystems into the elements, is one of the flow parameters. The
second parameter in a reversed flow is the determination of the approx-
imated capacity of the materialised elements, the subsystems, and the
system.
7.1 Determination of the structural form 63
The starting point in this process is usually the adoption of dimensions

derived from similar projects. Final dimensioning is obtained through a
“trial and error” procedure involving repeated analysis of the structure.
Only in some uncomplicated cases this process can be rationalised using
an algorithm flow diagram.
7.1.3 Flow diagram

The framework of procedures in conceptual structural design can be mod-
elled by flow diagrams, giving an unambiguous solution.
Flow diagrams can support structured analysis and design, by showing

the flow of data from external entities into the system, and how the data
move from one process to another, as well as its logical storage. Within
processes, optimisation loops can be implemented. There are common
modelling rules creating flow diagrams:
• All processes must have at least one data flow in and one data
flow out.
• All processes should modify the incoming data, producing new

forms of outgoing data.
• Each data store must be involved with at least one data flow.
• Each external entity must be involved with at least one data flow.
• A data flow must be attached to at least one process.
All influential aspects are known in conceptual structural design, the

basic interfaces with the built environment included. The interrelation
and interaction between all these aspects, however, is so complicated
that besides experience, intuition has to be used frequently.
Using a flow diagram is a way of making effective choices out of all

we already know. In doing so, throwing away aspects that cannot be
captured in this way - such as the ambiguous but vital intuition - is
unavoidable.
7.1.4 Experience and intuition

In many cases the interacting of the numerous aspects of the built en-
vironment such as functionality, costs, aesthetics, strength, redundancy,
constructability, flexibility, durability, maintainability, and sustainability
becomes complex when crossing disciplinary boundaries.
Then a lot of experience is advisable, and furthermore and more intan-
gibly, intuition becomes a necessity as stated by famous engineers:
Eduardo Torroja In his book “Philosophy of Structures” [27]: “The
achievement of the final solution is largely a matter of habit, in-
tuition, imagination, common sense, and personal attitude. Only
the accumulation of experience can shorten the necessary labour
or trial and error involved in the selection of one among the dif-
ferent possible alternatives.” “The calculation of stresses can only
serve to check and to correct the sizes of structural members as
conceived and proposed by the intuition of the designer. The work
itself is never born from calculation.”
Pier Luigi Nervi In his book “Structures” [20]: “It is highly regrettable
that some of the highest qualities of the human mind, such as intu-
ition and direct apprehension, have been banned from our schools
and have been overwhelmed by abstract and impersonal mathem-
atical formulas... The essential part of the design of a building
consists in conceiving and proportioning its structural system...
then and only then we can and we should apply the formulas of
mathematical theory of elasticity to specify with greater accuracy
its resisting elements.”
Modern structural engineers can rely on a long history of constant val-
idation of theoretical approaches, building up an immense database of
knowledge.
Until the present day, it appears challenging to capture performance-
based solutions out of the multitude of aspects and their complex interre-
lations. The unsolved topic in modelling these interrelations is capturing
intuition, the most ambiguous, and therefore the most intangible aspect
7.2 How to capture the intangible 65
of conceptual design.
Furthermore, some unambiguous structural aspects such as serial effects,

induced deformation, and the stability of the equilibrium are difficult to
insert in a linear optimisation process. At a certain level, decomposi-
tion and a linear approach to modelling the design process is no longer
feasible.
Consequently, the determination of the structural form is partly a qual-

ification process. The field of application, and the characteristics of the
structural forms with respect to the individual relations with the built
environment, are the variables in this partly unambiguous and partly
ambiguous process.
The ambiguous part is often specified as experience and intuition. The

most tangible of the two - experience - consists of conscious and subcon-
scious knowledge.
It is a reasonable assumption that intuition is nothing more than the

subconscious part of experience. After all, there is a noticeable dif-
ference between the intuitive capacities of a young professional and an
experienced engineer. This difference is contradictory to the sometimes-
suggested idea of a cosmic consciousness.
It is worthwhile revealing the experienced-based subconscious knowledge,

keeping in mind never to take something for granted.
7.2 How to capture the intangible

7.2.1 Abstracting conceptual design
In general, communication can be divided in three basic levels of abstrac-
tion: object, experience, and concept. Going up the levels of abstraction,
ideas increase while reality decreases:
At the object level Communication is about tangible material or an

unambiguous representation of it.
At the experience level Communication is about (common) experi-

ence. Although reality is the point of departure, it has the ab-
straction of interpretation.
At the concept level Communication is about ideas and thoughts.
Concepts can be accepted or rejected.
Considering the levels of abstraction in conceptual structural design,
the basic levels of communication are applicable. The experience level in
engineering practice, however, is mainly restricted to physical behaviour,
described with physical laws.
These laws are scientific generalisations that have become accepted uni-
versally within the scientific community. Describing the observable laws
of nature is based on observations and repeated scientific experiments.
Some extremely important physical laws are simply definitions of the
observable laws of nature, such as the mathematical definition of force by
Newton’s second law of mechanics. Some laws are only approximations of
other more general laws, with a restricted domain of applicability.
Modelling Due to the complexity of load distribution and above all

material behaviour, structural analysis is completely dependent on ab-
stract representations of the actual structure.
As an abstract representation, modelling has its limitations. For a
reliable application of structural modelling, awareness of these limit-
ations is of great importance, for example when shear deformation is
dominant.
7.2.2 Directing parameters of conceptual design

During the conceptual design phase, the individual participants of the
integral design process such as the structural engineer, the architect, and
the contractor, measure, so to speak, with different scales:
Form The structural engineer designs in terms of materialised structural
forms, static schemes, and dimensioning processes; in brief “form”.
7.2 How to capture the intangible 67
Concept The architect designs in terms of architectural concepts, design

philosophy, space, and expression; in brief “concept”.
Process The contractor designs in terms of building techniques, plan-
ning, phasing, and repetition; in brief “process”.
Dependent on the discipline, the parameters “form”, “concept”, and “pro-
cess” give direction to the design solution as shown in figure 7.2.
Concept
Requirements Form Material + Dimension Cost
Process
Figure 7.2: Directing parameters of conceptual design
Present-day misunderstanding between professional disciplines within an

integral design process can partly be ascribed to these almost perpen-
dicular oriented areas of attention.
Participation in this process requires a certain level of abstraction. Par-
ticularly the architect is used to abstraction, considering the nature of
architectural designing. On the contrary, both the structural engineer
and the contractor have to deal with a lot of down-to-earth activities in
the field of code checking and execution. For an effective cooperation,
it is of the utmost importance that every discipline contributes on the
required conceptual design level of abstraction.
7.2.3 Sharing the knowledge of the built environment

For an optimisation of the performance/cost ratio over the life cycle of a
structure, an integral approach and control of all participating influential
disciplines is an absolute necessity. In many cases, the interacting of
the numerous aspects of the built environment is so complex that an
unambiguous flow diagram cannot be applied. Then, a lot of experience

and intuition becomes advisable.
In consequence of gathering the required information for both the con-
sciousness and sub-consciousness of participants of conceptual design,
an accessible knowledge base is to be researched. For this purpose, the
immense complete body of knowledge of the built environment is totally
unsuitable and consequently, resulted in the present-day numerous dis-
ciplines and specialisations.
Nevertheless, the ability to have insight in multiple fields of discipline is
the most effective way of crossing borders between disciplines and thus
more effectively handles the design interfaces between these disciplines.
It is of importance to determine and bring together the required funda-
mental knowledge of these disciplines as an effective basis for conceptual
design.
A mutual knowledge base is possibly feasible, determining the absolute
required minimum:
1. Sharing the knowledge of the built environment on an approxima-
tion level suitable for conceptual design;
2. and restricted to only the knowledge of the disciplines involved
with a major interface during conceptual design.
7.2.4 Back to the fundamentals of conceptual design

The required T-profile of the professional structural engineer as shown
in figure 5.1 on page 43, can be refined by a visualisation of the overlap
between the in-depth and the in-breadth understanding. This overlap,
being the conceptual design parameters, is visualised in figure 7.3.
Conceptual design parameters The conceptual design parameters

are the specific part of the in-depth understanding that contributes to
the conceptual design. These conceptual design parameters have to en-
compass all fundamental aspects of structural engineering. After all, the
7.3 Shared knowledge-based conceptual design 69
In-breadth understanding
In-breadth understanding of
conceptual structural design
structural performance
Conceptual design parameters
(discipline fundamentals)
BOK BE BOK BE
Figure 7.3: T-shaped professional for conceptual design
inherent quality of a design is established during conceptual design, tak-

ing into account all influential interfaces with the other disciplines.
Although subsequently a lot of optimisation, detailing, and verification
has to be carried out, conceptual design and corresponding approximate
parameters establish, so to say, the “DNA” of the final design.
Discipline fundamentals Therefore, the conceptual design paramet-

ers represent the fundamentals of the discipline, in this specific case the
professional field of structural engineering. In view of the nature of con-
ceptual design, this generalisation is applicable to all disciplines within
the built environment.
7.3 Shared knowledge-based conceptual design

7.3.1 Splitting process and technical breadth
For a better understanding of the typical conceptual design activities, the
in-breath understanding of conceptual design can furthermore be divided
into a technical breadth of the built environment and the integral process
control of conceptual design as shown in figure 7.4.
In-breadth understanding Integral process control
Technical breadth
In-depth understanding Conceptual design parameters

(discipline fundamentals)
Technical depth
BOK BE BOK BE
Figure 7.4: T-shaped professional technical breadth
Integral process control The ability to work in a multi-disciplinary

manner with the profile to be “integrator” - besides the standard re-
quired process control abilities with respect to problem solving, project
management, and self-management - includes the following:
• Functional requirements as starting point for conceptual design.
• Decomposition techniques as used in systems engineering.
• Collaboration models such as concurrent engineering and building
information modelling.
Technical breadth The combined technical breadth and depth of the

Built Environment (BE) encompasses the complete Body Of Knowledge
(BOK) as indicated in figure 7.4 with the abbreviation “BOK BE”, applic-
able to all design phases of conceptual, basic, and detailed design.
A part of this body of knowledge, indicated in figure 7.4 as “Technical
breadth”, refers only to the conceptual design phase. On the level of
conceptual design, the technical breadth of the built environment spe-
cifically encompasses a fundamental understanding and application of
the conceptual design parameters of all disciplines.
This understanding and application includes an understanding and con-

trol of the relevant interfaces between the individual disciplines with
attention to constructability, safety, and durability.
However, due to the complexity of present-day design and the limited
extrapolation capacities of previous designs, the consequences of design
decisions in the early phase of conceptual design are difficult to over-
see.
This underlines both the necessity to have insight in multiple fields of
discipline, and to develop numerous concepts in order to bring them to
a lower degree of complexity as further elaborated in section 10.4.
Technical depth structural engineering The technical depth equals

the in-depth understanding of structural performance. This depth en-
compasses the complete body of knowledge of professional structural
engineering as practiced during all design phases of conceptual, basic,
and detailed design.
7.3.2 Principal disciplines of the built environment

For an integral conceptual design, the main contributing, and thus prin-
cipal disciplines and corresponding interfaces, have to be considered.
With respect to the quantity and character of the interfaces between the
participating disciplines within a conceptual design team, a distinction
between environment and object level as shown in figure 7.5 is appropri-
ate.
Object
Environment
Figure 7.5: Environment and object level
Besides this distinction between environment and object level, an ef-

fective classification of principal disciplines can be based upon obvious

requirements such as value, functionality, safety, aesthetics, and realis-
ability. The so-classified principal disciplines for the built environment
are listed in table 7.1.
Principal disciplines of the built environment

Requirements Environment level Object level
Value Property Property
Development Management
Functionality Transportation Installation
Engineering Engineering
Safety Hydraulic Structural
Aesthetics Urban Architectural
Planning Engineering
Realisability Environmental Construction
Table 7.1: Principal disciplines of the built environment
7.3.3 Concurrent-based shared knowledge

On object level, the interfaces of the object-related structural engineering
with the other object-related disciplines are within the system function-
ality of the object and are further substantiated as follows:
Property management The performance/cost ratio is the main driver

for the overall optimisation of conceptual design. Regarding life
cycle costs, maintenance, and management are gaining importance,
and are therefore substantial input for conceptual design.
Installation engineering Particularly, the main ducts of air condition-

ing systems have the same scale as girders and are preferably de-
signed parallelly. Smaller scaled installations normally have minor
to no influence at all on structural dimensioning.
Architectural engineering Architecture is a conscious creation of util-

itarian space and construction of materials in such a way that the
whole is both technically and aesthetically satisfying. Creation
of utilitarian space with materialised forms is a main influential
design interface with the structural form.
Construction engineering The practical feasibility of the execution
focuses on avoiding unnecessary complexity, on influences on di-
mensions and tolerances, and on possible choices between altern-
atives.
On environmental level, the interfaces with the object-related structural
engineering mainly consists of geometrical and loading constraints such
as free space profiles, road cross sections, traffic loads, and hydraulic
loads.
So for the determination of the fundamental shared knowledge with re-
spect to conceptual structural design, the emphasis is on the object-
related disciplines.
Because of the complexity of the interdisciplinary interfaces between
these object-related disciplines, concurrent engineering will be an ap-
propriate solution as discussed in subsection 5.5.5 of the solution com-
ponents.
After all, the concurrent engineering approach provides a collaborative,
co-operative, collective and simultaneous engineering working environ-
ment, based on the five key elements of process, multidisciplinary team,
integrated design model, facility and software infrastructure.
7.3.4 Conceptual structural design parameters

Both for understanding structural performance and sharing our funda-
mental knowledge with the other directly related interface disciplines, a
set of conceptual structural design parameters has to be established.
For professional Structural Engineering (SE), the applied mechanics-
based conceptual structural design parameters can be split into the
structural integrity on system level via the load distribution on sub-

system level to the failure mechanisms on element level, as shown in
figure 7.6.
Professional field
Built Environment
Integral process control

Conceptual design
Technical breadth
Conceptual structural design parameters:

Technical depth
Basic design 3-D Structural integrity

2-D Load distribution
1-D Failure mechanisms
Detailed design
BOK BE
SE
Figure 7.6: T-shaped conceptual structural design
Structural integrity The system as a whole can only be captured

on the corresponding three-dimensional system level. A typical three-
dimensional system effect such as overall torsion has to be incorporated
here. Insight into and control of the load distribution within the three-
dimensional system is merely feasible with axial forces on the level of
load paths.
For a more accurate insight into the structural action and corresponding
analysis of load distribution and dimensioning, the three-dimensional
system is too complex and has to be decomposed into more accessible
two-dimensional subsystems.
Load distribution On the two-dimensional subsystem level, a more

accurate load distribution can be executed, including insight into op-
timisation, redistribution and possibly induced deformation of statically

indeterminate structures. Consequently, a distribution of the prime ac-
tions can be executed to determine the individual element forces and
corresponding required capacity.
Failure mechanisms The failure mechanisms on a one-dimensional

element level represent the ability of materialised elements or cross sec-
tions to resist the distributed loads per element. These failure mechan-
isms are dependent on resistance of materials, resistance of cross sections,
and stability of elements.
The deformation of the individual, materialised elements culminates in
a deformation and corresponding displacements of the subsystems and
subsequently, the whole system.
Chapter 8
Process decomposition
8.1 Fundamental design cycle

8.1.1 Structural life cycle phases
In the life cycle of a structure, the following phases and correspond-
ing structural engineering activities can be chronologically distinguished
as listed in table 8.1: initiation, design and specification, construction,
operation and maintenance, and demolition.
8.1.2 Structural design phases

Out of the functional requirements and boundary conditions, a system is
created and the performance/cost ratio is optimised, taking into account
the interfaces with the built environment. Specifications for construction
are prepared. This process can be divided into three major design phases
with increasing accuracy, as listed in table 8.2: conceptual design, basic
design, and detailed design.
Conceptual structural design The conceptual design phase is of

special interest with respect to the problem definition as identified in
section 4.1. The starting point in this phase is usually the adoption
78 Process decomposition
Life cycle phases of a structure

Life cycle phases Structural engineering activities
Initiation User needs are secured by a set of functional re-
quirements, which define what the system is ulti-
mately supposed to do. At this level of abstrac-
tion, structural requirements are not directly re-
ferred to.
Design and Out of the functional requirements, a system is
specification created and the performance/cost ratio is optim-
ised, taking into account the interfaces with the
built environment. Specifications for construc-
tion are prepared.
Construction Feasibility and cost optimisation of construction
is secured by design, including temporary struc-
tures. Structural site engineers operate on the
level of postponed design activities, including
corrective actions in case of non-conformities.
Operation and After a maintenance or flexibility demand, struc-
maintenance tural engineers design the reconditioning or re-
building of the structure or parts of it, and again
operate on the level of design activities.
Demolition A controlled structural collapse is based upon a
structural action analysis and structural engin-
eers operate on the level of partial design activ-
ities.
Table 8.1: Life cycle phases
of conceptual structural design solutions derived from similar projects.

Any new type of structure, however, requires an extended lead-time to
obtain a thorough understanding of the structural action.
It is not expected that corresponding conceptual design calculations and
estimated costs be precise, but rather within accepted tolerance. The
level of acceptance has at least to be qualified, and preferably be quan-
8.1 Fundamental design cycle 79
Design phases of a structure

Design phases Structural engineering activities
Conceptual The purposes of conceptual design of any struc-
design ture are to obtain a clear picture of the structural
action, approximated dimensions of the struc-
ture, principal details, quantities of materials for
making estimates of costs, and a reliable predic-
tion of the basic and detailed design.
Basic design The purposes of basic design are to obtain exact
dimensions of the structure and approximated
dimensions of the details.
Detailed design The purposes of detailed design are to obtain
exact dimensions of the details and specifications
for construction.
Table 8.2: Design phases
tified, with a risk analysis of the costs of the conceptual design.
8.1.3 Cyclic design process

The immense number of individual parameters and mutual interactions
cannot be unambiguously controlled and will inevitably lead to a highly
cyclic process as shown in figure 8.1.
In the process of materialisation from requirements to construction, on-
going design and check activities can influence foregoing activities with
regard to choice of geometry, material, and matching dimensions. In an
effective converging process, however, the number of optimisation loops
will diminish during the process of further specification.
Both principal and design team are in search of an overall perform-
ance/cost optimisation. In some cases, this performance/cost optimisa-
tion may even lead to a modification of the requirements.
The conformity between the as-built situation after completion of the
Requirements
& conditions
Conceptual design
Basic design
Detailed design
Construction
Figure 8.1: Structural design process
construction and the requirements should be secured through conceptual

design.
8.1.4 Fundamental structural design cycle

The objective is capturing structural design by breaking down this com-
plex and unambiguous process to the essential absolute minimum, result-
ing in a fundamental design cycle as an effective characterisation of both
the design process as a whole, and the individual design phases.
The major contribution of professional structural engineering to the over-
all life cycle of a structure is on the level of design activities as clarified
in table 8.1 on page 78.
Each individual design activity starts with requirements as to what the
(part of the) structure is supposed to do and ends with a specifica-
8.1 Fundamental design cycle 81
tion for construction. Independent of the life cycle phase, complexity

of design, and contractual commitments, structural engineering practice
can be outlined by the fundamental structural design cycle as shown in
figure 8.2.
Requirements
& conditions
Creation of a
system outline
Optimisation of the
structural action
Final dimensioning
and specification
Construction
Figure 8.2: Fundamental structural design cycle
Creation of a system outline The most directing functional require-

ments and contractual conditions are interpreted into an overall geo-
metry, choice of materials, and approximate material dimensions. All
influential interfaces with other disciplines are taken into account and
overall performance/cost is optimised.
Optimisation of the structural action Within the boundaries of

the overall integral design, a more thorough analysis of the structural
action leads to optimisation of this structural action and a further ac-
curacy of the material dimensions. An overall structural system check
and compliance with all functional requirements and contractual condi-

tions is carried out.
Final dimensioning and specification Final material and detail di-

mensioning is determined on the basis of code checks with respect to
structural safety, serviceability, and durability. Specification for con-
struction is prepared.
8.2 Level of accuracy

8.2.1 Level of specification
To meet a client’s demand, functional requirements will specify the start-
ing point for the design process. Functional requirements, with function
as a higher order, can lead to form and material. Functional require-
ments are decisive with respect to the quality of a durable structure,
boundary conditions included, and therefore fixed.
The process from functional requirements to as-built can be divided into
identifiable phases as shown in figure 8.3: conceptual design, basic design,
detailed design, and construction.
Time schedule The time schedule on the horizontal axis consists of

the three major design phases and the subsequent construction phase.
These major design phases - conceptual, basic, and detailed design - can
be substituted by the fundamental structural design cycle-based phases
“exploration & creation”, “selection & optimisation”, and “verification &
specification”.
Level of reliable specification The corresponding achieved level of

specification per individual design phase is strongly dependent on the
time schedule. Particularly, in the conceptual design phase during ex-
ploration and creation most of the design is specified. This phase is
decisive for both quality of design and a prosperous completion of sub-
sequent phases.
8.2 Level of accuracy 83
Level of
specification
100 %
Requirements
As built
Exploration Selection Verification
& & &
creation optimisation specification Time
0%
Conceptual Basic Detailed Construction schedule
design design design
Figure 8.3: Level of specification
For an effective decision supporting, the reliability of each design phase

outcome requires a corresponding accuracy. A feasibility study of this
accuracy per design phase is provided by the International Federation
for Structural Concrete.
8.2.2 Fib Model Code 2010

The International Federation for Structural Concrete fib, “fédération in-
ternationale du béton”, is a pre-normative organisation. “Pre-normative”
implies pioneering work in codification. This work has been realised in
2012 with the fib Model Code 2010 [8].
The fib Model Code 2010 includes the whole life cycle of a concrete
structure, from design and construction to conservation and disman-
tlement, in one code for buildings, bridges, and other civil engineering
structures.
In this code, subsection 3.5.3, “Quality Management in Design”, a serious
study is reported on the feasible accuracy of individual design phases.
These individual phases, compatible with the decision process employed
by the owner, are listed in table 8.3.
Fib Model Code 2010 design phases of a structure

Design phases Structural engineering activities
Scouting Initial feasibility scan of the scheme, global func-
tional requirements are specified.
Target accuracy project costs estimation ±30%.
Basis of design Functional requirements, data, and design cri-
teria will be developed, service criteria agreed.
Project Further development into a preliminary design,
specification evaluation alternative structural concepts.
Final design All primary structural members will be specified
and typical details will be designed.
Detailed design The output of this phase will allow construction
of the project.
Table 8.3: Fib design phases
Scouting phase The “Scouting Phase” is an initial feasibility scan of

the scheme. From an abstract level of perception, global functional re-
quirements are specified. In common practice, the design effort expenses
will then be limited, because the feasibility of the project will usually
still be uncertain. At this stage, the target accuracy for the estimate of
overall project costs might typically be ±30%.
Basis of Design phase During this phase, the functional require-

ments, basic data, and design criteria will be developed and the service
criteria agreed upon. A conceptual design will also be developed to sup-
port a more accurate budget estimate. Quite some effort is required as
the “Basis of Design” should be agreed upon, fixed and frozen upon com-
8.2 Level of accuracy 85
pletion. At this stage, the target accuracy for the estimate of overall
project costs might typically be ±20%.
Project Specification phase With the “Basis of Design” as start-

ing point, the design will be developed into a preliminary design. Al-
ternative structural concepts will generally be developed and evaluated
against each other. Specifications for workmanship, materials, and de-
tailed design will then be developed. Significant effort is generally re-
quired. At this stage, the target accuracy for the estimate of overall
project costs might typically be ±10%.
Final Design phase During this phase, all primary structural mem-
bers will be specified and typical details will be designed. The structural
analysis should consider the behaviour of the structure in relation to
the envisaged dimensioning situations, taking into account the relevant
factors that significantly influence the potential performance of the struc-
ture concerned. At this stage, the target accuracy for the estimate of
overall project costs might typically be ±5%.
Detailed Design phase The output of this phase shall allow con-
struction of the project. The level of detail of drawings and specifications
and site instructions shall allow unambiguous understanding by the con-
tractor of what is required and how the scheme must be executed, as well
as how compliance with the documents must be demonstrated.
Target accuracies costs The target accuracies for the estimate of

overall project costs, however, could also be applied to other factors
such as environmental impact and the evaluation sustainability para-
meters.
8.2.3 Accuracy fundamental structural design phases

The fib Model Code 2010 study on the feasible accuracy of correspond-
ing individual design phases can easily be transformed into the required
accuracy of the individual major design phases and structural phases of

the fundamental structural design cycle, as listed in table 8.4.
Structural design phases and corresponding accuracy

Structural phases Major phases Fib phases Accuracy
Creation of a Conceptual Scouting ±30%
system outline design
Basis of design ±20%
Optimisation of the Basic design Project ±10%

structural action specification
Final design ±5%
Final dimensioning Detailed design Detailed design For con-

and specification struction
Table 8.4: Level of accuracy
8.3 Fundamental design process

8.3.1 Structural design characteristics
Independent of life cycle phase, complexity of design, and contractual
commitments, the structural engineering practice can be most effectively
outlined by the structural design characteristics as listed in table 8.5,
directly based upon the fundamental design cycle as shown in figure 8.2
on page 81.
These structural design characteristics can be an effective guide for struc-
turing structural engineering activities and a corresponding professional
profile of structural engineering.
8.3.2 Fundamental structural design process

Within each of the major design phases of conceptual design, basic
design, and detailed design, the fundamental structural design cycle as
8.3 Fundamental design process 87
Professional structural engineering

Structural design characteristics
Conceptual design: Creation of a system outline
- Overall geometry
- Choice of materials
- Approximate section characteristics: ±20%
- Conceptual variant studies of performance-based solutions
- Main optimisation of the performance/cost ratio
- Open mind for ideas and innovative design
Basic design: Optimisation of the structural action
- Thorough analysis of the structural action
- Optimisation of the structural action
- Further dimensioning with an accuracy of ±5%
- Overall structural system check with the complete set of functional
requirements and contractual conditions
Detailed design: Final dimensioning and specification
- Final section and detail dimensioning
- Check calculations structural safety ULS
- Check calculations serviceability SLS
- Check calculations durability DLS
- Specification for construction
- Monitoring decisive parameters
Table 8.5: Structural engineering characteristics
shown in figure 8.2 on page 81 is applicable. Consequently, the struc-

tural design process as shown in figure 8.1 on page 80, can be filled in as
In the process of increasing insight into the system’s functionality, each
design phase is characterised by a change of both accuracy and character
of the activities. The corresponding increasing level of specification is
reflected by the required accuracy of the material dimensioning from
±20%, via ±5%, to final specification for construction.
Requirements
& conditions
Creation of a
Conceptual design:
system outline
system outline
Creation of a
Optimisation of the
structural action
Dimensioning ±20%
and specification
Creation of a
Optimisation of the
system outline
structural action
Basic design:
Optimisation of the
structural action
Dimensioning ±5%
and specification
Creation of a
Final dimensioning
and specification
system outline
Detailed design:
Optimisation of the
structural action
Final dimensioning
and specification
Construction
Figure 8.4: Fundamental structural design process

Chapter 9
Physical decomposition
9.1 Qualification of the structural form

9.1.1 Classification
A physical decomposition of a structure requires a methodical classifica-
tion as a guidance for the partition of the system into subsystems.
There are numerous possibilities of classifying structural systems. The
two most notable classification techniques are the following:
• Classification by physical appearance; in this case structural form.
• Classification by behaviour; in this case structural action.
The majority of applied classifications are directly based on the struc-
tural form with clear identifiable structural elements as beams, columns,
floors, walls, arches, and trusses.
Oskar Büttner und Erhard Hampe performed an extensive study and
reported it in the corresponding publication “Bauwerk, Tragwerk, Trag-
struktur, Band 2: Klassifizierung, Tragqualität, Bauwerkbeispiele” [3] of
a structural action-based classification.
In this classification, specifically the bending flexibility of linear and
90 Physical decomposition
plane structural elements is distinguished. These elements are then as-

sembled into one-dimensional linear elements, and two-dimensional and
three-dimensional structural systems. De facto it is a distinction between
tensile structures and all other structures dominated by bending.
9.1.2 Form follows function

The starting point of the design process is the function expressed by
a set of functional requirements. In order to successfully conceive and
plan a structure it is necessary to investigate and well know its reason
for existence, and its major and minor loads to resist and to bear. The
function, however, is not a static specification but can vary during the
life cycle of a structure.
Eduardo Torroja Stated in his book “Philosophy of Structures” [27]:

“The primary functions of all structure can be summarised as follows:
To enclose a certain space and to protect it from the natural elements of
wind, rain, and snow, from changes in temperature, and from noise. This
function is achieved by the use of walls and roofs. To provide passageways
for the movement of persons and vehicles; floors, staircases, and ramps
of buildings, and bridges and viaducts are used for these functions. To
resist the lateral thrust of earth, water, or other fluids; included in this
category are dams, dikes, reservoirs, storage tanks, silos, and retaining
walls.”
Eladio Dieste Stated in his book “La Estructura Cerámica” [7]: “The
resistant virtues of the structures that we seek depend on their form;
it is through their form that they are stable, not because of awkward
accumulation of material. There is nothing more noble and elegant from
an intellectual viewpoint than this: to resist through form.”
9.1.3 Interfaces with the built environment

The main interfaces with the built environment, which are decisive with
respect to influencing the feasibility in general and the overall perform-
ance/cost ratio in particular, are: reliability, architectural demand, air
9.2 Decomposition of the structural form 91
conditioning system, constructability (practical feasibility), and relative

life cycle costs.
To secure a clear recognisable interface between conceptual structural

design and the mainly form-oriented other conceptual design disciplines
of the built environment, a structural form-based, rather than structural
action-based classification is preferable.
Furthermore, a structural action-based classification can vary completely

on different levels. For example, the global load distribution of a truss
consists of bending moments and shear forces. The corresponding local
load distribution within the truss consists completely of axial compres-
sion and tension forces in the individual members of the truss.
9.2 Decomposition of the structural form

9.2.1 System decomposition
Physical decomposition follows the physical parts of a system. Often, it
is a very natural way of decomposition, because we easily “see” all the
physical parts. Also, the completeness criterion above is easy to check;
when we have all physical parts, we have the whole system.
The whole structure is a three-dimensional system such as a frame build-

ing, multi-storey building, barrier, bridge, and tunnel.
The three-dimensional system can be disassembled in two-dimensional

(plane) subsystems such as a frame, floor slab, cable-stayed beam, truss,
arch, and shear wall.
The two-dimensional subsystems can subsequently be assembled in one-

dimensional (linear) elements such as a compression and tension bar, a
bending beam, and a (shear) corbel.
The system “Arch bridge” as shown in figure 9.1 for example, can be seen
as an assembly of the subsystems frame, truss, arch, and orthotropic
deck.
System
Arch bridge
Subsystem
Truss
Subsystem Subsystem
Frame Arch
Subsystem
Orthotropic deck
Figure 9.1: Assembly of an arch bridge
A two-dimensional subsystem is an assembly of directly connected struc-

tural elements, designed to act together to resist loads.
9.2.2 Structural form on subsystem level

Characterisation of a problem is part of the solution. But character-
isation of a bridge, barrier or building as a whole is nearly impossible
due to the variety and complexity of possible structural forms. Char-
acterisation of individual forms with regard to the capacity to bear and
resist, and with regard to the interfaces with the built environment, ap-
pears feasible. It is plausible that on this subsystem level, as shown
in figure 9.2, enough insight and oversight can be gathered to produce
performance-based solutions.
System
Subsystem
Element
Figure 9.2: Subsystem level
9.2.3 Basic structural forms

On subsystem level the following basic structural forms can be distin-
guished as shown in figure 9.3: frame, floor slab, cable-stayed beam,
truss, arch, and shear wall.
Global and local load distribution On the basis of uniformly dis-

tributed loads the global load distribution of all basic structural forms
consists of bending moments and shear forces. Dependent on the indi-
vidual basic form this global load distribution is internally transferred
differently into a local load distribution.
Each individual basic structural form is characterised by its correspond-

ing local load distribution. This characteristic local distribution of the
loads is described in the following paragraphs, whereby members subjec-
ted to axial compression or tension, are relatively stronger and generally
more economical than those designed for pure bending.
Frame
Floor slab
Truss
Cable-stayed beam
Shear wall
Arch
Figure 9.3: Basic structural forms
Frame Frames consist of bending beams and supporting columns. The

local load distribution of a beam consisting of one element equals the
global load distribution and is therefore subjected to bending and shear.
The supporting columns are subjected to compression. The load distri-
bution in a frame is worked out in subsection 14.2.1.
Floor slab The local load distribution of a floor slab equals the global
load distribution and is therefore subjected to bending and shear. A
two-way spanning floor slab, whether concrete slab or orthotropic steel
deck, is subjected to bending and shear in both spanning directions. The
load distribution in a floor slab is worked out in subsection 14.2.2.
Cable-stayed beam A supporting cable diminishes the supported

length and corresponding local load distribution of beams; by each half,
strength - the effect of section modulus W - is quadrupled and stiffness
- the effect of moment of inertia I - is eight-fold.
The effectiveness of the local load distribution in a cable-stayed, stat-

ically indeterminate continuous beam is dependent on the stiffness of
the cable. The cross-sectional area of the cable is therefore governed by
both tension strength and supporting stiffness. The load distribution in
a cable-stayed beam is worked out in subsection 14.2.3.
Truss Every member of the truss is in pure compression or pure ten-

sion. Shear, bending moments, and other more complex stresses are all
practically zero. This makes trusses physically stronger and stiffer than
other ways of arranging material. Some or all of the joints may be fixed
rather than pinned but the main contribution to the strength and stiff-
ness is provided by the triangulation. The load distribution in a truss is
worked out in subsection 14.2.4.
Arch The form is more or less parabolic and usually, the horizontal
reaction forces are resisted by a tension rod. Uniformly loaded, the entire
arch is in compression and has little bending moments and shear. This
makes arches relatively stronger than other ways of arranging material,
resulting in smaller cross-sectional dimensions. However, an arch requires
a more than average amount of structural space.
So-called “false” arches, when bending is dominant over compression, are
unfavourable: they are circular instead of parabolic arches and/or have
highly-concentrated loads, instead of more or less uniform loads. The
load distribution in an arch is worked out in subsection 14.2.5.
Shear wall The most common application of shear walls is securing the
global stability of a system. For non-sway buildings, these can be in the
form of concrete shear walls or steel wind bracing. The load distribution
of a stocky shear wall consists of dominant shear in combination with
minor bending. The load distribution in a shear wall is worked out in
subsection 14.2.6.
Chapter 10
Cyclic process control
10.1 Exploration of the solubility space

10.1.1 Creation-process requirements
During the creation process, the structural engineer is creating a mater-
ialised form out of the functional requirements and taking into account
the fundamental related interfaces with the built environment.
Key words in the search for a structural form are “insight”, “overview”,
and “breeding ground”.
Insight Insight into each individual aspect of the design process such
as reliability, structural form, load distribution, choice of material, failure
mechanisms, dimensioning, architectural demand, constructability, and
costs.
Overview Overview of the interrelations of the aspects, such as:

• Reliability versus form, load distribution, material, failure mech-
anisms, and dimensions.
• Architectural demand versus form, material, and dimensions.
98 Cyclic process control
• Constructability versus form, material, and dimensions.
• Costs versus reliability, architectural demand, and constructability.
Breeding ground The way of presenting and visualising the indi-

vidual aspects and their interrelations can serve as a fertile breeding
ground for design solutions. All data do not necessarily have to be quan-
tified to fulfil their function. Therefore, hardly quantifiable phenomena
such as intuition, imagination, common sense, and personal attitude can
be implemented in the creation process.
10.1.2 Design strategies

Four typical design strategies with increasing complexity can be distin-
guished as shown in figure 10.1:
Linearity The simplest design strategy describes designing as an un-

ambiguous one-way movement among the design activities. This
strategy is only working for the design of extremely simple struc-
tures.
Cyclic For most of the design problems, a cyclic design process is inevit-
able. Every time the designed structure turns out to be insufficient
a new cycle is appropriate.
Cyclic Convergent In a converging process, the number of loops will

diminish during the process of further specification by making goal-
oriented design choices. Convergence, however, does not automat-
ically mean optimisation.
Cyclic Optimum Optimisation of both quality of design outcome and

number of design cycles is dependent on how strong the initial idea
is and the improvements one chooses to make. For the most suc-
cessful possible outcome, corresponding fundamental design choices
have to be searched for.
10.2 Linear design 99
Linearity Cyclic Cyclic Cyclic

Convergent Optimum
Figure 10.1: Design strategies
10.2 Linear design

10.2.1 Linear process
Each individual design activity starts with requirements as to what the
structure or part of the structure is supposed to do and ends with a spe-
cification for construction. Independent of life cycle phase, complexity
of design, and contractual commitments, structural engineering practice
can be outlined by the fundamental structural design cycle as shown in
figure 8.2 on page 81.
Linear design describes designing as an unambiguous one-way movement
among these design phases as shown in figure 10.2. This strategy is only
working for the design of extremely simple structures.
Creation Optimisation Dimensioning

of a of the and
system outline structural action specification
Figure 10.2: Linear structural design
10.2.2 Specification of the structural form

Essentially, the design process of a structure is a process of composing
the structural form. For an effective control of this composition process
both the level of specification and the level of composition have to be
considered.
The integral composition process of a conceptual draft of the system

outline will inevitably lead to a process of decomposing the structure in
controllable basic structural forms, mostly on a two-dimensional subsys-
tem level.
The immense number of individual parameters and mutual interactions
cannot be unambiguously controlled and will inevitably lead to a cyclic
process.
10.3 Cyclic design

10.3.1 Cyclic process
In order to have a successful solution to a complex structural design
problem a cyclic design process is inevitable. Every cycle goes through
the phases of creation, optimisation, and specification. Then, the concept
is reviewed with respect to the functional requirements in particular, and
a performance/cost optimisation over the life cycle in general.
When the reviewed concept turns out to be insufficient, a new cycle
is appropriate. In the process of materialisation from requirements to
construction, on-going design and check activities can influence forego-
ing activities with regard to choice of geometry, material, and matching
dimensions.
The number of cycles depends on how strong the initial idea is, and on
the improvements one chooses to make; intelligent improvements will
reduce the number of times that one has to go through the cyclic design
process.
The cyclic design process evolves over time in order to produce the most
successful possible outcome but at least a performance-based solution.
In an effective converging process, the number of optimisation loops will
diminish during the process of further specification.
An example of this cyclic design process is when the famous Thomas
Edison, who designed more than ten thousand prototypes of the light
bulb until he was satisfied and knew it worked, said; “I have not failed
10.3 Cyclic design 101
10,000 times, I have successfully found 10,000 ways that will not work”
[30].
10.3.2 Fundamental structural design path

The structural design process can fundamentally be characterised by two
simultaneous processes:
Specification The process of specification of the structural form from
approximate to accurate.
Composition The process of composition or rather decomposition of
the structural form from system to element.
These two processes can be visualised together in a two-dimensional mat-
rix as shown in figure 10.3. Within this matrix, the cyclic structural
design can be explored.
Level of specification On the horizontal axis, the phases of specific-

ation are arranged from creation of the system outline, via optimisation
of the structural action, to dimensioning and specification.
Level of decomposition On the vertical axis, the phases of decom-

position of the structural form are arranged from three-dimensional sys-
tem, via two-dimensional subsystem, to one-dimensional element.
Design path Complex structures can effectively be analysed “from sys-

tem to element” and “from approximate to accurate”. The corresponding
design path is directed by the effectiveness of this combination of breadth
and depth. For an effective converging process, this combination has to
be well balanced.
The design path follows the fundamental dimensioning routine from
structural integrity, via load distribution, to failure mechanisms:
• Structural integrity; three-dimensional system design and decom-
position in subsystems.
decomposition
th
pa
gn
si
Level of
de
al
im
pt
O
lin e
Element
el rat
g
od u
m acc
In
Failure
mechanisms
Subsystem
Load
distribution
ity le
ex b
System
pl lla
m tro
co on
nc
Structural
U
integrity Level of
Creation Optimisation Dimensioning specification
of a of the and
Figure 10.3: Fundamental structural design path
• Load distribution; distribution of the actions within the subsystem

on element level.
• Failure mechanisms; distributed actions can be resisted by mater-

ialising the elements.
Outside the boundaries of this design path, ineffectiveness such as inac-

curate modelling or even uncontrollable complexity can be found.
10.3 Cyclic design 103
Inaccurate modelling The integral creation of a system outline re-

quires a three-dimensional modelling. Due to the huge degree of free-
dom and the complexity of interfaces with other disciplines, this three-
dimensional modelling and corresponding dimensioning is only approx-
imating and without great detail.
For a more refined dimensioning, load distribution and capacity calcula-

tions on element level are appropriate. The required data for an accurate
calculation on element level are not available in such an early conceptual
design stage.
Therefore, a creation of a system outline on element level inevitably leads

to inaccurate and thus unusable modelling.
Uncontrollable complexity The specification of the design is de-

termined on element level in order to get the required detailed depth
and accuracy. Solely on system level, it is not achievable to control the
multitude of information and interrelations from creation, via optimisa-
tion, to specification.
So inevitably, designing the structure on system level only results in

uncontrollable complexity of the dimensioning and specification.
10.3.3 Structural design path

Out of the functional requirements, a system is created and the perform-
ance/cost ratio is optimised, taking the interfaces with the built envir-
onment into account. Then, specifications for construction are prepared.
This process can be divided into three major design phases with increas-
ing accuracy: conceptual design, basic design, and detailed design, as
shown in figure 8.4 on page 88.
The fundamental structural design path as shown in figure 10.3 on page

102 is applicable for all these three design phases. Furthermore, this
fundamental path is also applicable for the structural design process as
a whole - from creation to specification - as shown in figure 10.4.
decomposition
th
pa
gn
si
Level of
de
al
im
pt
O
Detailed
design
lin e
Element
el rat
g
od u
m acc
In
Code
checking
Basic
Subsystem
design
FEM
analysis
Conceptual
ity le
ex b
design
System
pl lla
m tro
co on
nc
Truss
U
analogy Level of
of a of the and
Figure 10.4: Structural design path
In this case, the standard level of decomposition is combined with a level

of specification over the major structural design phases of conceptual,
basic, and detailed design:
Conceptual design Creation of a system outline with the help of truss

analogy.
Basic design Optimisation of the structural action with the help of a

10.4 Cyclic convergent design 105
Finite Element Method (FEM) analysis.

Detailed design Dimensioning with the help of code checking, followed
by a specification for construction.
These include the ineffectiveness that can be found outside the bound-
aries of the design path, as mentioned before, in the form of inaccurate
modelling or uncontrollable complexity.
10.4 Cyclic convergent design

10.4.1 Volume of complexity
The number of interdependent parameters that can be simultaneously
controlled is limited and subject to the complexity of the individual
parameters, the complexity of the interdependency between the para-
meters, and the level of abstraction of both parameters and interde-
pendency.
The limitation of the complexity of either the individual parameters,
or the interdependency between the parameters, has a similarity with
communicating vessels. On the one hand, the control of processes with
complex interdependencies will unavoidably lead to a limitation of the
quantity and/or complexity of the individual parameters. On the other
hand, control of processes with large quantities and/or complex para-
meters, will unavoidably lead to a limitation of the complexity of the
interdependency between the parameters.
Due to the limited volume of complexity per design phase, the integ-
ral design process with complex interdependencies inevitably has to be
separated from the individual architectural, structural, and construction
design processes with complex parameters as shown in figure 10.5.
10.4.2 Three steps in conceptual design

In order to control the conceptual design phase, we can consider three
steps:
Requirements
& conditions
Definition
Definition of the
design problem
Creation
Integral design
Numerous concepts
Optimisation of the Optimisation of the Optimisation of the

Selection
architectural design structural design construction design
Specification for
basic design
Iterative design loop
Conceptual
structural design
Figure 10.5: Limited volume of complexity per design phase
Definition step Defining the design problem.
Creation step Structural engineering-based creation; diverging into a

rough overview of possible solutions.
Selection step Value engineering-based selection; a more refined check-

ing of the boundary conditions, converging into a solution.
10.4 Cyclic convergent design 107
10.4.3 Definition step

Common practice during conceptual design is formulating a quick defin-
ition of the problem in order to save time for a heated search for
performance-based solutions.
Albert Einstein is quoted as having said that if he had one hour to save
the world he would spend fifty-five minutes defining the problem and
then five minutes solving it [4].
Einstein’s wisdom with regard to the problem-solving process is directly

applicable to the conceptual structural design process. Thoroughly for-
mulating the definition of the design problem during the problem defin-
ition phase is a time-consuming but crucial effort before starting the
search for performance-based solutions.
Besides an effective partition of available time, it is a misconception that

extending the duration of the search would lead to a better solution. It
is far more effective to optimise available time, than stretching it.
The conceptual design team has to evaluate the complete set of functional
requirements and contractual conditions, in order to determine a work-
able reduced set of fundamental requirements and conditions: presum-
ably dominant, required performances and aspects - including interfaces
with the adjacent built environment and prerequisite constraints.
The so-determined set of fundamental functional requirements will then

be the starting point of the creation of concepts by the conceptual in-
tegral design team.
Effective functional requirements should be defined on an as-high-as-

possible level of abstraction in order to serve as a starting point for
creative freedom, rather than stifling restriction.
Furthermore, on such a high abstraction level, market tendencies and

other undesirable influences are neutralised.
10.4.4 Creation step

First of all, it is important to create concepts that are likely to meet the
performance requirements. It is hard enough to create a materialised
form out of the functional requirements, and to take the fundamental
related interfaces as reliability, architectural demand, constructability,
and costs, into account.
To be secured of at least one performance-based solution at the end of

the conceptual design process, it is advisable to create numerous con-
cepts and develop them to a lower degree of complexity for a considered
selection. Dependent on the complexity one has to cope with during the
selection phase, the number of initial concepts can be reduced.
10.4.5 Selection step

The goal is convergence by optimising the cost-drivers such as the com-
plexity of execution - shape, repetition and planning, maintenance, and
possibilities regarding future extension capability.
Striving for an optimisation of the performance/cost ratio will lead to a

performance-based cost minimisation. Sometimes, supplementary added-
value solutions offer a considerable added value for a relatively small cost
increase. Then, a real performance/cost optimisation - despite the earlier
established requirements - will be executed.
10.5 Cyclic optimum design

Design cycles encompass complete updates of design. Within a design
cycle, individual design loops can be distinguished representing particu-
lar design actions. Independent of life cycle phase, complexity of design,
and contractual commitments, fundamental design loops can be identi-
fied as shown in figure 10.6.
In principal, design loops can occur at any level of specification, at any

level of decomposition, in any direction, and with any range. Usually, in
10.5 Cyclic optimum design 109
decomposition
Modelling loops
Level of
l) ps
(vertical)
na oo
go l
ia gn
(d esi
D
Element
Analysis
loop
lo ion
op
at
nt
rie
O
Subsystem
n
tio
System
Check
op ec
loop
lo orr
C
Level of
of a of the and
Figure 10.6: Fundamental structural design loops
design practice however, most frequently occurring design loops can be

identified:
Modelling loops Vertical calculation-based loops with regard to struc-

tural safety and serviceability.
Typical design loops Diagonal creation-based loops with regard to a

controlled build-up of the structural system.
10.5.1 Modelling loops

Modelling loops are the backbone of structural design. Within each spe-
cification phase from creation to specification, the complexity of reality
has to be modelled with the help of calculation models with regard to
structural safety and serviceability of the structure, or a specific part of
the structure.
Although the point of attention differs per specification phase, for re-
search on the structural integrity, the whole system or a specific part
of the system is involved, covering all the levels of decomposition from
system to element.
Such modelling loops are vertically orientated in the fundamental struc-
tural design loops as shown in figure 10.6 on page 109.
Refinement requires an analysis loop, whereas keeping the integrity of
the whole system in mind requires check loops.
Analysis loop Going to a higher level of decomposition, complexity

increases and further analysis is necessary to get a grip on the more
detailed structural action.
Check loop With increasing level of decomposition and correspond-

ing further analysis, checking the calculation model on a lower level of
decomposition - preferably the whole system - is necessary in order to
secure the integrity of the structure.
10.5.2 Typical design loops

Typical design loops are directing the design process. Within each spe-
cification phase from creation to specification, requirements-based design
choices are made. Although the point of attention differs per specific-
ation phase, each design choice has to be valued and made ready for
further refinement.
These design loops are diagonally orientated in the fundamental struc-
tural design loops as shown in figure 10.6 on page 109. On-going, more
detailed specification requires an orientation loop, whereas diagnosed

inaccuracies require correction loops.
Orientation loop Going to a higher level of specification in combin-

ation with a higher level of decomposition, complexity increases and
research is necessary in order to orientate on possible further design re-
finements.
Correction loop With on-going detailing, insight into the structural

action increases. With this increasing insight, foregoing structural design
approximations and decisions can turn out to be insufficient or even
incorrect. Such diagnosed inaccuracies require correction.
10.5.3 Optimal design cycle

The cyclic character of the conceptual structural design process is a
continuous process of zooming in and zooming out with (diagonal) design
loops, thus constantly changing both the level of decomposition and the
level of specification.
In between, the dimensioning with the help of calculation models is a
process of zooming in and zooming out with (vertical) modelling loops,
changing only the level of decomposition. This process from creation to
dimensioning is shown in figure 10.7.
The clearly directed fundamental design loops result in small-scaled and
thus shorter, more clarifying individual design cycles. Furthermore, these
small-scaled clarifying cycles potentially lead to a reduction of the total
amount of cycles.
It should be noted that an effective optimal design cycle is a process of
reduction rather than exclusion and limitation. This process of reduc-
tion, however, is completely dependent on known and explicitly unknown
parameters. In the case of implicitly unknown parameters, there is auto-
matically an unwanted exclusion and consequently uncontrollable risk
with corresponding uncertainties in cost estimation.
Creation Optimisation
of a of the
Level of decomposition system outline structural action
Level of decomposition
Level of specification Level of specification
Dimensioning
and
specification
Legend:
Level of decomposition
Modelling loops (vertical):

Analysis loop
Check loop
Design loops (diagonal):

Orientation loop
Correction loop
Level of specification
Figure 10.7: Structural design loops
10.5.4 Risk analysis

The specification of a conceptual design demarcates the concerning pro-
ject phase. Dependent on the contractual arrangement, this specification
can be used as internal transfer documentation or for tenders. In both
cases, a risk analysis is essential.
A risk analysis of the material demand gives an accuracy estimate of the
material quantities and corresponding unit cost indications. The risk

analysis has to be executed on the level of individual components of a
decomposed system, principal details included.
For an effective risk analysis the following, preferably quantified data

have to be determined:
Performed structural analysis Depth and breath of the structural

analysis are dependent on the complexity of the structural action
and the available time schedule, financing, and resources of the
conceptual structural design phase.
Dimensioning The result of the performed structural analysis is an

approximate section-dimensioning of the materialised overall geo-
metry. This dimensioning of the conceptual design is specified by
a list of material quantities.
Cost weighting The cost weighting equals the material quantities times
the unit cost indications, and is a measure for cost optimisation op-
portunities and corresponding risk.
Uncertainties The difference between the performed approximate struc-

tural analysis for conceptual design and the required depth and
breath to meet the in-use requirements for structural safety and
serviceability, can be defined as uncertainties of the conceptual
design. The difference in depth and breath generally concerns load
combinations, load distribution, and failure mechanisms.
Coverage by conceptual dimensioning The uncertainties of the per-

formed approximate structural analysis with respect to the re-
quired depth and breath can be partially, or even not at all, covered.
The corresponding status of the coverage gives an indication of the
risk influence of uncertainties on conceptual design.
Reserves Reserves can be intentionally incorporated or are the result of

rounding-up to the nearest standardised product dimensions. Oc-
casionally, an optimisation of requirements during the conceptual
design phase can result in a reserve.
Optimisations Foreseen but time-consuming cost optimisations can be

postponed to basic design but registered as a potential reserve with
regard to the completed conceptual design.
Part III
Understanding structural
performance
Chapter 11
Introduction to part III
The methodical approach on conceptual structural design consists of a

process control component and an underlying, structural performance
component. The process control component has been discussed in the
previous Part II, “The art of conceptual structural design”. The under-
lying, structural performance component is discussed in this Part III,
“Understanding structural performance”.
Although the individual, conceptual structural design parameters are

fully applied mechanics-based, it is choice, combination, and balanced
depth of these parameters that define their professional applicability.
It is thereto that this part derives its reason for existence within this
textbook.
Part III starts with a qualification of structural performance.
The fundamental structural design path, as shown in figure 10.3 on page

102, gives the basic arrangement of the required conceptual structural
design parameters “3-D structural integrity”, “2-D load distribution”, and
“1-D failure mechanisms”. The quantification of these design parameters
as described in part III is subdivided into the individual parameters as
shown on the diagonal in figure 11.1.
118 Introduction to part III
Part IV
Part III
Chapter 15
Element
Failure
mechanisms
Chapter 14
Subsystem
Load
distribution
Chapter 13
System
Structural
integrity
Creation Optimisation Dimensioning
Chapter 12
Part II
Figure 11.1: Reading guide for part III
Structural performance Structural performance is a collective term

for structural safety, serviceability, and durability. As a major part of
structural performance, structural action with regard to load distribu-
tion, corresponding deformation, and strength is researched.
For an effective and efficient conceptual structural design, approxima-

tions and especially deformation-driven parameters have to be determ-
ined. The choice and combination of these parameters, however, define
their professional applicability.
119
Structural integrity In the course of the conceptual design, and par-

ticularly in the first design loops, the insight into the structural action
has to be built up step by step. Due to the huge degree of freedom and
the complexity of interfaces with other disciplines, conceptual structural
designing requires simple and clear, three-dimensional modelling.
This is possible on the level of axial forces, directly or with a truss
analogy for modelling bending action and more complex forms.
Load distribution The load distribution on the system through the

subsystems into the elements, can best be determined on subsystem level
with design approximations of the load distribution in two-dimensional,
basic structural forms.
For the basic structural forms, the design parameters concerning load
distribution and deformation are determined and guides are given for
the design of parallel load distribution with regard to optimisation, re-
dundancy, and induced deformation.
Failure mechanisms With the determined forces in the individual

one-dimensional elements, the required dimensions of these elements can
be determined by means of design approximations of the load-carrying
capacity with regard to the ultimate limit state.
With the determined deformations, the required dimensions of these ele-
ments can be determined by means of design approximations of the de-
formation with regard to the serviceability limit state.
For the elements of basic structural forms, the design parameters con-
cerning material properties, sectional strength, stability, and deformation
are determined.
Solution components The so obtained coherent set of solution com-

ponents for structural performance, and corresponding chapter arrange-
ment, is listed in table 11.1.
120 Introduction to part III
Chapter arrangement and solution components

No. Chapter Solution component
15 Structural integrity - Load path design
16 Load distribution - Load distribution parameters
17 Failure mechanisms - Dimensioning parameters
Table 11.1: Solution components for structural performance

Chapter 12
12.1 Present-day structural performance

12.1.1 Structural requirements
Structural performance is a collective term for the following structural
requirements as defined in subsection 4.1.4:
Structural safety The safety of a structure or structural member is

prescribed with its Ultimate Limit State (ULS).
Serviceability The functionality of a structure or structural member is

prescribed with its Serviceability Limit State (SLS).
Durability The durability of a structure or structural member is pre-

scribed with its Durability Limit State (DLS).
These specifically in the structural design codes of practice prescribed

structural requirements ULS, SLS, and DLS, are part of a more overall
sustainability objective to minimise the negative environmental impact.
122 Structural performance
12.1.2 Modern developments
The three structural requirements, structural safety, serviceability, and

durability, combined with corresponding directions of development are
Serviceability SLS
Social focus
Modern developments
Structural safety ULS
Durability DLS
Figure 12.1: Developments of structural performance
With structural safety being a self-evident prerequisite, serviceability is

an upcoming contractual requirement with regard to controlled func-
tional behaviour of the structure. Due to the absence of an additional
safety factor for serviceability, the outcome of measurement results typ-
ically is sensitive to downward crossing.
Particularly, the recent addition of a durability limit state to the ultimate

and serviceability limit states, is initiated through performance-based
building design as a logical step towards an integral performance/cost
optimisation of the built environment of the entire life cycle. As a result,
durability is extending to sustainability by minimising negative environ-
mental impact, and further to the Triple Bottom Line (TBL) accounting
framework with a social, environmental, and financial partition.
Due to the amount of structural failures, however, society is shifting its

focus back towards structural safety.
12.2 Force and deformation-driven parameters 123
12.2 Force and deformation-driven parameters

12.2.1 Equilibrium and strength
For an effective and efficient determination of the capacity to bear and
resist, a fundamental insight into structural behaviour is necessary. The
four subjects to be considered profoundly are equilibrium, strength, de-
formation, and parallel load distribution.
Further qualification and coherence of these four subjects as shown in

figure 12.2 will be clarified subsequently.
Equilibrium Equilibrium is the first, most obvious aspect and an

absolute necessity for each building and all civil work; otherwise the
structure or parts of the structure will translate, rotate, or just tumble
down.
Strength Strength is the second most obvious aspect and also an ab-
solute necessity; without enough strength the functional requirements
with respect to the ultimate limit state, and particularly the structural
safety, can never be met.
The two failure mechanisms concerning the strength are sectional strength
and stability. Sectional strength is force-driven, whereas the stability
of the equilibrium - from element up to system level - is deformation-
driven.
This deformation-driven stability is subject to many parameters such as

length, cross-sectional bending stiffness, boundary conditions, deforma-
tion due to fabrication, residual stresses, and material faults. Therefore,
the stability analysis is often merely based upon iterative blind code
checking, without availing the opportunity of essence-based deformation-
driven designing.
The necessity of both equilibrium and strength is described thoroughly in

built environment related legislation by structural safety regulations.
Materialised
structural scheme
Equilibrium
No
Stat. Det.
Yes
Strength Deformation Parallel load distribution
Deformation-driven
Induced deformation
Sectional strength
Deformation ULS
Deformation SLS
Force-driven
Redundancy
Optimisation
Stability
Conceptual
structural design
Figure 12.2: Force and deformation-driven parameters
12.2.2 Deformation-driven parameters

One of the main pitfalls in conceptual designing, however, is focusing
mainly on the combined equilibrium and strength, just because it is an
obvious necessity.
As for stability, deformation-driven parameters in structural designing
are somewhat underexposed in higher education as well as in common,
12.2 Force and deformation-driven parameters 125
modern structural engineering practices.

Especially the conceptual designer needs to comprehend the deformation-
driven parameters as shown in figure 12.2 on page 124, in addition to
the standard engineering basics such as applied mechanics, design codes,
and related background documents.
Deformation In case of functional requirements with regard to the

serviceability limit state, displacements of parts of the structure and the
structure as a whole have to be met.
Parallel load distribution Foregoing strength and deformation para-

meters have to meet the requirements concerning structural safety in the
ultimate limit state, and displacements in the serviceability limit state.
The extent of compliance with these requirements determines the effect-
iveness of the design.
Nowadays, a lot of structures are statically indeterminate out of eco-
nomical efficiency and redundancy considerations. Corresponding paral-
lel load distribution in the ultimate limit state is largely determined by
deformation parameters.
Statical indeterminacy The structural system is statically indeterm-

inate when the static equilibrium equations are not sufficient for determ-
ining the internal forces and reactions on the structure. A difference
in deformation will influence the load distribution. Consequently, the
behaviour can be deformation-driven, besides force-driven.
The distribution of the load in detailed design will be on the level of
stress and strain. Due to a high degree of parallel behaviour - statical
indeterminacy - the stress distribution is proportional to the stiffness
distribution, which is inversely proportional to the strains.
Optimisation of the performance/cost ratio The optimisation of

the stiffness distribution is a main factor in the optimisation of the per-
formance/cost ratio. Focusing mainly on the strength will produce a safe
design but not automatically an economical efficient one. Therefore, it

is of great importance that the structural designer, besides perform-
ance and construction demand, focuses on deformation in addition to
strength.
12.2.3 Determination of deformation-driven parameters
The relative small group of conceptual structural design practitioners

does not have the time to analyse its mainly intuition-based designing in
order to make its knowledge accessible for both professionals and higher
Scientific education is compartmented to such an extent that the in-

terface between applied mechanics and material applications is under-
developed. The interfaces between the structural engineering and archi-
tectural demand, constructability and life cycle engineering are even less
visible.
The practice of conceptual structural design with regard to deformation-

driven parameters, is not covered by most of the higher education pro-
grammes, as listed in table 12.1.
Misfit of structural design parameters in education

Structural design parameters: Force-driven Deformation-driven
Field of practice Minor part Major part
Higher education programmes Major part Minor part
Table 12.1: Misfit of design parameters in education
Therefore, a missing set of deformation-driven design parameters has

to be determined as a basis for conceptual structural design in higher
12.3 Approximate dimensioning 127
12.3 Approximate dimensioning

12.3.1 Conceptual structural design approximations
It is of the utmost importance that design assumptions during conceptual
structural design are based on insight into the behaviour of the structure
to assure structural safety and to facilitate design optimisation.
Effective conceptual structural design approximations should address,
and give insight into all influential structural parameters namely material
properties, geometry, support conditions, and loading.
Ineffective approximations So-called “rules of thumb” when solely

based on geometric properties and therefore load indifferent, do not ad-
dress all influential structural parameters, have a low accuracy and are
almost never provided with a clear and comprehensive insight into the
application boundaries.
High-end black box application programs, on the other hand, address all
structural parameters and have a high accuracy but do not give a direct
insight into the behaviour of the structure.
Modified applied mechanics-based approximations Applied

mechanics-based approximations give a calculated insight into load dis-
tribution and likely decisive failure mechanisms. Furthermore, it fa-
cilitates an overview of the optimisation possibilities of the structural
design.
For realistic approximations of complex material behaviour, applied mech-
anics approximations can be modified with identifiable empirical data.
On the level of integral conceptual design, geometrical parameters should
be related to outer dimensions.
12.3.2 Conceptual structural design parameters

Although applied mechanics are textbook material-based, defining an
adequate set of conceptual structural design parameters is appropriate
within this methodical approach.

This set of conceptual structural design parameters has to fulfil the fol-
lowing requirements:
• Be applied mechanics-based to assure structural insight.
• Have clear and unambiguous applicability.
• Have performance-based approximation depth for conceptual design.
• Be balanced with respect to deformation-driven parameters.
• Have manageable limited overall size.
The fundamental structural design path, as shown in figure 10.3 on page

102, gives the basic arrangement of the required conceptual structural
design parameters:
• Load path design on a three-dimensional system level.
• Load distribution on a two-dimensional subsystem level.
• Failure mechanisms on a one-dimensional element level.
This basic arrangement can be extended into an effective set of applied
mechanics-based conceptual structural design parameters as listed in
table 12.2.
12.3.3 Dimensioning routine

In conformity with the fundamental structural design cycle as shown
in figure 8.2 on page 81, the dimensioning routine of the conceptual
structural design can be constructed as shown in figure 12.3.
12.3 Approximate dimensioning 129
Conceptual structural design parameters

Structural integrity Load distribution Failure mechanisms
3-D system level 2-D subsystem level 1-D element level
System design Distribution of Distributed actions
on load path level the prime actions can be resisted
and decomposition within the subsystem by materialising
in subsystems on element level and dimensioning
Basic applied mechanics based
System design: Distribution: Capacity:
- Rigidity - Basic structural - Material strength
- System effects forms and stiffness
- Principal details - Parallel load - Sectional strength
distribution and stiffness
- Induced - Stability
deformation
Table 12.2: Conceptual structural design parameters

Requirements
& conditions
3-D system
load path design
Materialised
structural scheme
2-D subsystem
load distribution
Element forces
& displacements
1-D element
failure mechanisms
Conceptual
structural design
Figure 12.3: Dimensioning routine of the conceptual structural design

Chapter 13
Structural integrity
13.1 Conceptual design on system level

13.1.1 Creation phase of the conceptual design
The creation phase of a system outline starts with requirements and con-
ditions of what the system is supposed to do and ends with one or more
materialised structural schemes as shown in figure 12.3 on page 130.
13.1.2 Structural integrity

Structural integrity is the term used for the performance characteristics
of a structure with regard to the ability to support designed loads and
the redundancy capacity of the whole system.
In an integral conceptual design process, the performance/cost-ratio of

the life cycle has to be optimised. Every structure will have its own
process of creation, influenced by its bearing capacity and resistance, its
economy, its construction site, and last but not least a more or less pro-
nounced aesthetic concept and appearance. Its structural integrity, how-
ever, will always be an undoubted demand and has to be unambiguously
secured by a clear, three-dimensional basic structural concept.
132 Structural integrity
The vertical imposed loads, the self-weight of the structure and natural
impacts such as snow loads have to be borne. In addition to these ver-
tical loads, other equally essential horizontal loadings must be taken into
account due to wind actions and global initial sway imperfections.
On system level, the static resistance and equilibrium during the life cycle
of the structure have to be secured. Particularly, the design of additional
stabilising elements to resist the horizontal loads requires attention and
insight on a three-dimensional system level.
Furthermore, the redundancy of the structural system with respect to

accidental loads has to be secured, based on a risk analysis and dependent
on specific structural performance demands.
13.2 Load path design

13.2.1 Load path design on system level
Due to the huge degree of freedom and the complexity of interfaces with
other disciplines, conceptual structural designing requires simple and
clear three-dimensional modelling. This is merely possible on the level
of axial forces, directly, with a truss-analogy or an arch depending on
the structural form.
Axial forces-based, a rough three-dimensional outline of the load paths

can be constructed as shown in figure 13.1.
Prime actions Load path design
Free space profiles
Points of support
Figure 13.1: Load path design

13.2 Load path design 133
These load paths are directly caused by the prime actions, by-passing the
free space profiles, and borne by the available points of support.
Load path design during the creation phase of conceptual design is the
process of determination and optimisation of the load paths, in close
collaboration with all the influential participating disciplines.
13.2.2 Truss-analogy in load path design
Modelling bending action in more complex forms can be axial forces-

based with the help of a truss-analogy, as shown in figure 13.2.
Figure 13.2: Truss-analogy in load path design
The complex form can be filled out by a statically determinate truss

configuration, two-dimensional or when appropriate three-dimensional.
The load distribution in the truss can then easily be calculated using the
method of sections. The strength-based dimensioning results in sectional
properties and corresponding construction heights. The rigidity of the
truss can further be optimised with a minimal potential energy-based
analysis.
13.2.3 Modelling the system decomposition

Getting a grip on possible three-dimensional load distribution effects in
an early stage of the conceptual design process requires retention of three-
dimensional effects during decomposition of the three-dimensional sys-
tem in two-dimensional subsystems. This system decomposition can be
principally done either by defining the compatibility functions between
the subsystems; or by separating the three-dimensional effect as shown
in figure 13.3.
+ 3-D
effect
2-D 2-D
Compatibility functions Separate 3-D system effects
Figure 13.3: Modelling the system effects
Defining the compatibility functions to such an extent that they can

be used for a neat quantification of the three-dimensional load distribu-
tion is too complex and time-consuming for a conceptual dimensioning.
Therefore, qualification and approximate quantification of a separated
three-dimensional effect is the remaining feasible option.
For example, the three-dimensional system effect of a double track truss

bridge can be separated by defining the overall torsion due to one track
loading, with the help of basic applied mechanics.
13.2.4 Dimensioning routine of the load path

The load distribution on the system, through the subsystems into the
elements can best be determined on subsystem level with design ap-
proximations of the load distribution in basic structural forms. Possible
overall three-dimensional system effects then have to be separately evalu-
ated, estimated and added to the load distribution of the subsystem. The
13.3 Conceptual structural design 135
result is an approximate determination of the forces in each individual

element.
With these forces, the dimensions of the individual elements can be de-
termined by means of design approximations of the load-carrying capa-
city of elements with regard to sectional strength and element stability.
The dimensioning routine, in conformity with figure 12.3 on page 130, is
3-D System 2-D Subsystem 1-D Element
Sectional strength
+ 3-D Element stability
2-D effect
Load path design Load distribution & displacements Dimensioning
Figure 13.4: Dimensioning routine
This determination of the approximated capacity of the elements, the

subsystems and the system is a reversed flow resulting in a dimensioning
of the whole structural system.
13.3 Conceptual structural design

13.3.1 Principal details
The materialisation of the structural form is not only dependent on the
choice of material and corresponding dimensioning, but also on the design
of influential architectural and cost dictating details of the structure, the
so-called “principal details” as shown in figure 13.5.
The design approach of these often three-dimensional details is identical
to that of the structural system. The choice of material and the matching
dimensions of the principal detail can influence the foregoing qualification
process with regard to the field of application and characteristics of the
chosen structural form; then, an optimisation loop is applicable.
Load path design Principal details Conceptual

structural design
Dimensioning
+ +
Routine
Figure 13.5: Conceptual structural design
13.3.2 System configuration

Independent of requirements and performance/cost optimisation of the
conceptual design, structural design should keep the following more gen-
eric rules in mind with regard to a simple and clear structural system
configuration:
Structural economy Members subjected to axial compression or ten-
sion are relatively stronger than other ways of arranging mater-
ial and generally more economical than those designed for pure
bending.
Structural safety In a simple and clear structure, which is rigorously
functional, the transmission of forces in the whole structure and in
each of the elements will be simple, clear, and without twist.
Instinctive security A structure with few and strong elements always
gives an impression of ease and security.
13.3.3 Material properties of conceptual design

In conceptual design, clear and concise material properties are strived
for. The strength of structural elements loaded by axial forces or bend-
ing moments, is based on the material-dependent normal stress strength
fd . For concrete and its reinforcement, the design material strength fc
respectively fs , incorporates a material safety. For structural steel, the
design material strength equals the yielding stress fy .
Other material properties such as shear strength fv , normal stiffness E
and shear stiffness G, can be expressed in this basic strength property

fd or are constant.
The material properties are based upon the applicable Eurocodes for
concrete [15] and structural steel [17].
Elementary material properties for the conceptual structural design of

reinforced concrete and steel structures are listed in table 13.1.
Material properties for conceptual design

Property: Reinforced concrete Structural steel
Normal stress
⇥ N ⇤ Concrete grade
strength fd mm 2 fc = fy = Steel grade
1.5
fs = 435
Shear stress
⇥ ⇤ fy
strength fv N
mm2
fv = 0.2fc fv = p
3
Normal stiffness
⇥ N ⇤ fc
modulus E mm 2 Ec = 1.75 E = 2.1 · 105
1000
Shear stiffness
⇥ N ⇤ fc
modulus G mm 2 Gc = 1.75 G = 8.1 · 104
1000· 2.4
Table 13.1: Material properties for conceptual design
Normal stress strength for reinforced concrete Common mater-

ial grades from conventional to high strength, and their normal stress
strengths fc , are as follows:
⇥ ⇤
Concrete grade fc N
mm2
C30 20
C45 30
C60 40
C90 60
Higher concrete grades offer increasing strength and corresponding in-
creasing stiffness in order to reduce structural dimensions and increase
available space. Furthermore, a shorter cure time allows for quick re-
moval of formwork, and subsequently, putting it into use.
Normal stress strength for structural steel Common material

grades from conventional to high strength, and their normal stress
strengths fy , are as follows:
⇥ N ⇤
Structural steel grade fy mm 2
S235 235
S355 355
S460 460
S690 690
Higher steel grades offer increasing strength in order to reduce strength-
driven structural dimensions.
Shear stress strength for reinforced concrete The shear stress

strength fv for most materials is linearly proportional to the normal
stress strength fd and considerably lower. For reinforced concrete, the
shear stress strength is approximately 20% of the normal stress strength:
fv = 0.2fc .
Shear stress strength for structural steel The shear stress strength
for structural steel can be derived from the Huber Hencky and Von Mises
yield criterion:
p fy
2 + 3⌧ 2  fy ) ⌧  p = fv (13.1)
3
Normal stiffness modulus for reinforced concrete For concrete,

the normal stiffness modulus or Young’s modulus Ec is dependent on
the material grade and its corresponding normal stress strength fc . This
modulus Ec can be derived from the simplified bi-linear modelled stress
strain relationship of concrete with an elastic strain limit of 1.75%� [15],
combined with the practical usage for the modulus Ec , including the
influence of time related effects [1]:
fc
Ec = 1.75 (13.2)
1000
Normal stiffness modulus for structural steel For steel, the nor-
mal stiffness modulus or Young’s modulus E has a constant value of
N
2.1 · 105 mm 2.
Shear stiffness modulus for reinforced concrete For uncracked

concrete the shear modulus Gc is dependent on the material grade and
its corresponding normal stress strength fc :
fc
Ec 1.75
fc
Gc = = 1000
= 1.75 (13.3)
2 (1 + ⌫) 2 (1 + 0.2) 1000· 2.4
Shear stiffness modulus for structural steel For steel the shear
modulus G has a constant value of approximately 8.1 · 104 mm
N
2:
E 2.1 · 105 N
G= = ⇡ 8.1 · 104 (13.4)
2 (1 + ⌫) 2 (1 + 0.3) mm2
Chapter 14
Fundamental parameters of
load distribution
14.1 Conceptual design on subsystem level

14.1.1 Load distribution phase of the conceptual design
The load distribution phase starts with materialised structural schemes
and ends with element forces and displacements as shown in figure 12.3
on page 130.
14.1.2 Load distribution

The distribution of the loads within a two-dimensional subsystem is de-
pendent on the geometrical characteristics of the subsystem. For the
two-dimensional basic structural forms as defined in figure 9.3 on page
94, the approximated load distribution for conceptual design has to be
determined.
In search of a basic set of knowledge and approximation parameters for
conceptual structural design in general, and load distribution in partic-
ular, the following principles are employed:
142 Fundamental parameters of load distribution
Applied mechanics-based to secure insight and durability. Basic ap-

plied mechanics textbook material, being established knowledge, is
not included in the bibliography.
Professional practice factors for improvement of approximation ac-

curacy, including reference in the bibliography or verification in
this textbook.
Emphasis on deformation-driven aspects in accordance with the

force and deformation-driven parameters as shown in figure 12.2
on page 124.
Performance-based and well-balanced with respect to depth and

breadth of underlying knowledge and applicability for conceptual
design.
14.1.3 Parallel load distribution

Statically determinate structures An assembly of directly connec-
ted elements is defined “statically determinate” when the static equi-
librium equations are sufficient for determining the internal forces and
reactions: ⌃H = 0, ⌃V = 0, and ⌃M = 0.
A common statically determinate structure is a simply supported beam

uniform load q
Serial system
Figure 14.1: Statically determinate structure
Practically, a statically determinate structure has just enough (internal)

members and (external) supports to secure structural form and fixation,
in analogy to a serial system.
14.1 Conceptual design on subsystem level 143
As a consequence, the behaviour is completely force-driven and fail-

ure of only one random element immediately leads to failure of the
structure.
Statically indeterminate structures An assembly of directly con-

nected elements is defined “statically indeterminate” when the static
equilibrium equations are not sufficient for determining the internal forces
and reactions on the structure.
A common statically indeterminate structure is a continuous beam as

uniform load q
Parallel system
Figure 14.2: Statically indeterminate structure
Practically, a statically indeterminate structure has more (internal) mem-

bers and/or (external) supports than necessary for structural integrity,
in analogy to a parallel system.
As a consequence, the behaviour can be deformation-driven, besides

force-driven. A difference in stiffness will influence the load distribu-
tion within the parallel system.
Parallel load distribution can be effectively utilised with regard to the

following:
Redundancy Design a second method of support to ensure that the

forces are dispersed elsewhere when a vital structural component
can give way.
Optimisation Optimisation through redistribution by altering the stiff-

ness of the individual elements.
Induced deformation Deformation of a primarily statically indeterm-

inate structure can cause unwanted loading of secondary structures.
14.2 Load distribution in basic structural forms

14.2.1 Load distribution in a frame
Frames consist of bending beams and supporting columns. The local
load distribution of a beam consisting of one element equals the global
load distribution and is therefore subjected to bending and shear. The
supporting columns are subjected to compression.
The effects of bending moments versus axial forces can be analysed with
a statically indeterminate frame structure, consisting of a combined can-
tilever beam and supporting column as shown in figure 14.3.
Fbm
Fcln
Figure 14.3: Parallel bending and axial force
Axial versus bending strength The strength of the individual can-

tilever beam, the individual column and the ratio between both, can be
determined:
2
M Fbm · lbm bm · bbm · hbm
bm = = 1 2
) F bm = (14.1)
Wbm 6 · bbm · hbm
6 · lbm
14.2 Load distribution in basic structural forms 145
N Fcln
cln = = ) Fcln = cln · bcln · hcln (14.2)
Acln bcln · hcln
Fcln cln · bcln · hcln · 6 · lbm

= 2 (14.3)
Fbm bm · bbm · hbm
When both beam and column are equally homogeneous materialised and
dimensioned ⇤ 300 · 300 mm2 with lengths of 3 m:
Fcln ·b·h·6·l 6·l 6·3

= 2
= = = 60 (14.4)
Fbm ·b·h h 0.3
The axial strength exceeds the bending strength by far. In general, mem-
bers subjected to axial compression, or tension, are relatively stronger
than other ways of arranging material.
Axial versus bending stiffness The distribution of a load in a stat-

ically indeterminate structure, will be on the basis of the ratio between
the axial stiffness and the bending stiffness:
F = Fcln + Fbm (14.5)
3
Fbm · lbm 3
EAcln · lbm
Fcln · lcln Fcln
= = ) = (14.6)
EAcln 3 · EIbm Fbm 3 · EIbm · lcln
When both beam and column are equally homogeneous materialised and
dimensioned ⇤ 300 · 300 mm2 with lengths of 3 m:
Fcln E · b · h · l3 4 · l2 4 · 32
= 1 = = = 400 (14.7)
Fbm 3 · E · 12 · b · h3 · l h2 0.32
Because axial stiffness mostly exceeds the bending stiffness by far, a com-
bined load distribution will result in a domination of the axial forces.
14.2.2 Load distribution in a floor slab

The local load distribution of a floor slab equals the global load dis-
tribution and is therefore subjected to bending and shear. A two-way
spanning floor slab, whether concrete slab or orthotropic steel deck, is
subjected to bending and shear in both spanning directions.
The effect of parallel bending can be analysed with a homogeneous two-
way spanning floor slab with a simplified load distribution as shown in
figure 14.4.
F2
F1
l1
l2
Figure 14.4: Load distribution in a two-way spanning floor slab
The load distribution in statically indeterminate structural systems with

parallel bending is dependent on the length of the load paths; the effect-
iveness of the load paths is linearly proportional to the difference in
stiffness.
An approximation of the distribution of the loads in the short span (F1 )
respectively long span (F2 ) of the floor slab:
F = F1 + F2 (14.8)
✓ ◆3
F1 l13 F2 l23 F1 l2
= = ) = (14.9)
48EI 48EI F2 l1
For example, a rectangular two-way spanning slab with spans of l re-

spectively 2l :
✓ ◆3
F1 2l
= =8 (14.10)
F2 l
The short load path is eight times as effective as the long load path, due
to the proportional difference in stiffness.
The outcome is directly appropriate for a slab of reinforced concrete. The
usually low reinforcement percentage has after all a negligible influence
on the stiffness Ec I of a reinforced concrete slab.
14.2.3 Load distribution in a cable-stayed beam

A supporting cable diminishes the supported length and corresponding
local load distribution of beams; by each half, strength - the effect of
section modulus W - is quadrupled and stiffness - the effect of moment
of inertia I - is eight-fold. The required section modulus W , and moment
of inertia I, are worked out in subsection 15.4.3.
The effectiveness of the local load distribution in a cable-stayed statically
indeterminate continuous beam is dependent on the stiffness of the cable.
The cross-sectional area of the cable is therefore governed by both tension
strength and supporting stiffness.
In such a statically indeterminate structure, optimisation by redistri-
bution within the deformation-driven parallel load distribution can be
utilised. An example of such an optimisation by a controlled redistribu-
tion is the cable-stayed continuous beam as shown in figure 14.5.
The effectiveness of the beam is governed by the stiffness of the cable;
the strength of the cable has no influence at all with regard to the redis-
tribution.
uniform load q
l
Δlcable
!
!tot
Figure 14.5: Redistribution in a cable-stayed beam
Regardless of the required cross-sectional area of the cable with respect to

the strived-after redistribution, the structural strength and serviceability
always remain prerequisite requirements.
Optimisation An optimisation of the moment distribution in the con-

tinuous beam can be accomplished by sagging the middle support:
5
· l3
12 ql ql4 ql4
= = (14.11)
3EI 8EI 72EIbm
The corresponding recommended cross-sectional area of the cable with

respect to the stiffness (ULS) amounts to:
7
ql p 2
ql4 lcable Ncable · lcable 6
· l + h2
= = = = sin ↵
72EIbm sin↵ EAcable · sin↵ EAcable · sin ↵
p
84 · Ibm · l2 + h2 Ebm
) Acable, stif f ness U LS = · (14.12)
l3 · sin2 ↵ Ecable
Structural strength The required cross-sectional area of the cable

with respect to the strength (ULS) amounts to:
7
·l
6 qd
Acable, strength U LS (14.13)
sin ↵ · fy
No redistribution and both spans under extreme loading amounts to:
Rmid = 54 qd l 6= 76 qd l (14.14)
Using high-strength steel cables, the required stiffness usually will be

decisive due to the same Young’s modulus for all steel grades.
Serviceability The required cross-sectional area of the cable with re-

spect to the displacements (SLS) amounts to:
7
q l p
6 k
1 qk l 4 1 l 2 + h2
sin ↵ · l
tot ⇡ · + 2 · 
150 EIbm EAcable · sin ↵ 250
p
146 · qk · l2 + h2
) Acable, stif f ness SLS ⇣ ⌘ (14.15)
5qk l3
Ecable · sin2 ↵ · 1 3EI bm
Concerning the displacements of the beam, the stiffness of the beam Ibm
influences the required stiffness of the cable Acable .
14.2.4 Load distribution in a truss
Every member of a truss is in pure compression or pure tension. The

main contribution to the strength and stiffness is provided by the triangu-
lation. Shear and bending moments, and other more complex stresses are
all practically zero. This makes trusses relatively strong and stiff.
The load distribution in a simply supported simplified parallel chord

truss is shown in figure 14.6.
uniform load q
Figure 14.6: Load distribution in a truss

The global load distribution consists of bending moments and shear

forces. The local load distribution within the truss consists completely
of axial compression and tension forces in the individual truss mem-
bers.
A simply supported truss with a uniform load results in the following
maximum axial forces in the chord and the web members:
1
M qd l 2
Nchord, max = = 8 (14.16)
h h
1
V qd l
Nweb, max = = 2 (14.17)
cos ↵ cos ↵
A non-uniform load still results in local axial forces only, with the pos-
sibility of a change of compression and tension in the web members near
the shear transition point.
The global stiffness of the truss can be approximated as follows:
1 2
Itruss ⇡ 0.8 · 2 · Achord · 2h (14.18)
The moment of inertia Itruss has an approximated loss of 0.2 because of

global shear deformation of the truss due to elastic deformation of the
individual web members.
During conceptual design this stiffness parameter Itruss can be applied
to determine the global bending deflection (SLS):
5 qk l 4 2
= · and Itruss ⇡ 0.8 · 2 · Achord · 1
2h (14.19)
384 EItruss
Furthermore, this stiffness parameter Itruss can be applied to determine

the global buckling strength (ULS) as discussed in subsection 15.3.3.
14.2.5 Load distribution in an arch

The form of an arch is more or less parabolic and usually, the horizontal
reaction forces are resisted by a tension rod. Uniformly loaded, the entire
arch is in compression and has little bending moments and shear; this
makes arches relatively strong.
The load distribution in a simply supported arch is shown in figure 14.7.
uniform load q
H H H
Figure 14.7: Load distribution in an arch

The subsystem as a whole is simply supported and therefore globally

statically determinate. The corresponding global load distribution con-
sists of bending moments and shear forces.
The tied arch, consisting of an arch and a tie-rod, is locally statically
determinate. When loaded with a uniform load per m1 span length,
the corresponding local load distribution within a tied parabolic arch
consists solely of axial compression in the arch and axial tension in the
tie-rod.
A uniform load results in the following maximum axial forces in the
tie-rod and the arch:
1
M qd l 2
H= = 8 (14.20)
h h
q
1 2
Narch, max = 2 qd l + H2 (14.21)
When the uniform load is dominant, the arch more or less follows the line
of compression. Loaded with a uniform load per m1 span length it equals
a parabola. Loaded with a uniform load per m1 arch length it equals a
hyperbola (hyperbolic cosine function), the so-called “catenary”.
A non-uniform load results in additional bending of the arch, besides
the axial compression. So-called “false” arches, where local bending is
dominant over compression, are unfavourable:
• They have circular instead of parabolic arches.
• And/or they have highly concentrated loads, instead of more or
less uniform loads.
A practical design value for the maximum bending moment within the
arch can be obtained by half the variable uniform loading as shown in
figure 14.8.
The so-obtained practical design value for the maximum bending mo-
ment within the arch amounts to:
1 2
March, max ⇡ 64 qd l (14.22)
uniform load q
line of compression
H H
Figure 14.8: Bending moments in an arch
The corresponding variable axial forces in tie-rod and arch amount to:
1 2
M 16 qd l
H= = (14.23)
h h
q
3 2
Narch = 8 qd l + H2 (14.24)
14.2.6 Load distribution in a shear wall

A most common application of shear walls is securing the global stability
of a system against wind actions and global initial sway imperfections.
For non-sway buildings these can be in the form of concrete shear walls
or steel wind bracing.
14.3 Parallel load distribution on detailed level 155
The load distribution of a stocky shear wall consists of dominant shear

in combination with minor bending.
Concrete shear wall The load distribution within the shear wall takes
place on element level. For conceptual design, the approximate calcula-
tion of shear strength and deformation in subsection 15.2.1 respectively
15.2.2 is applicable.
Steel wind bracing The load distribution within the steel wind bra-
cing takes place on subsystem level as discussed in subsection 14.2.4 on
the load distribution in trusses.
Because of the stocky dimensions, shear is governing. For conceptual
design, the approximate calculation of shear strength and deformation
in subsection 15.2.3 on the shear deformation in cantilevered trusses is
applicable.
14.3 Parallel load distribution on detailed level

The distribution of the load in detailed design will be on the level of
stress and strain. Due to the high degree of complexity of common three-
dimensional details, these details reveal a corresponding high degree of
statically indeterminacy. The stress distribution within the details is
proportional to the stiffness distribution, which is inversely proportional
to the strains.
Conceptual design of principal details can be interpreted as the design
of a small-scaled system. Thus, an identical methodical approach such
as for system design, with a decomposition in subsystems and elements,
is appropriate.
Because both axial and shear stiffness mostly exceed the bending stiffness
by far, a combined load distribution will result in a domination of the
axial and shear forces.
A study on the optimum design of a stiffening plate in a beam-to-column
joint shows the difference between bending and combined axial/shear

deformation. In this specific case, it is by far preferable to connect the
plate only to the loaded flange, thus avoiding high fitting costs of each
individual stiffener. Both options are given in the same figure 14.9.
100 kN
50
tpl =10 100 kN
30
150 10
Figure 14.9: Stiffener plate in a beam-to-column joint
When the contribution of the flange turns out to be a negligible factor,

the stiffener plate only needs to be connected to the loaded flange for
force transition.
By calculating the deformation of each individual (parallel) failure mech-

anism, the decisive mechanism can be determined under the condition
that there is an order of magnitude difference in deformation. When the
deformations lie in the same range, an accurate finite elements calcula-
tion has to be executed.
In this specific case the following failure mechanisms are applicable:
• Bending of the flange.
• Axial force on the stiffener.
• Shear force on the stiffener.

14.3 Parallel load distribution on detailed level 157
Bending of the flange The bending deformation of the flange is

̴100
100 kN
50
10
Figure 14.10: Bending deformation of the flange
The displacement, due to bending deformation of the flange, amounts to:
F l3 100 · 103 · 503

= = 1 = 2.38 mm (14.25)
3EI 3 · 2.1 · 105 · 12 · 100 · 103
Axial force on the stiffener The axial deformation of the stiffener

is shown in figure 14.11.
80 tpl =10 100 kN
150
!
Figure 14.11: Axial deformation of the stiffener

The displacement, due to axial deformation of the stiffener, amounts to:

N
="·E ) = ·E
A l
1
N ·l · 100 · 103 · 150
) = = 2 = 0.04 mm (14.26)
E·A 2.1 · 105 · 80 · 10
Shear force on the stiffener The shear deformation of the stiffener
is shown in figure 14.12.
!
100 kN
tplt=10
pl =10 50
150
Figure 14.12: Shear deformation of the stiffener
The displacement, due to shear deformation of the stiffener, amounts to:

V
⌧= ·G) = ·G
Av h
V ·h 100 · 103 · 50
) = = = 0.04 mm (14.27)
G · Av 8.1 · 104 · 150 · 10
The deformation due to bending of the flange is an order of magnitude

higher; consequently, the contribution of the flange is negligible.
14.4 Induced deformation

14.4.1 Principle of induced deformation
An induced deformation may only occur in a statically indeterminate
and thus deformation-driven structure.
14.4 Induced deformation 159
Problem of induced deformation Within a statically indetermin-

ate subsystem, a stiff element can cause an induced deformation of an
adjacent relatively flexible element. The same can occur on a higher sys-
tem level with an induced deformation of a relative flexible subsystem
within a statically indeterminate system.
As a result this induced deformation can cause a structural failure of the

concerning relatively flexible element, respectively subsystem.
Solution to the problem of induced deformation Failure is com-

monly opposed by an increase of the cross-sectional strength and ac-
companying stiffness. When the stiffness and corresponding structural
action increases more than the strength of the structure, the solution is
to reduce the stiffness.
14.4.2 Induced deformation on subsystem level

An example of induced deformation on subsystem level is a concrete floor
slab on a torsional clamped steel beam as shown in figure 14.13.
When the problem of possible torsional failure is force-driven, the only

solution is the necessary increase of the torsional strength to oppose
structural failure.
However, when the problem of possible torsional failure is deformation-

driven, the solution is to reduce the torsional stiffness to prevent struc-
tural failure.
14.4.3 Induced deformation on system level

Within a statically indeterminate system a stiff subsystem can cause an
induced deformation of a relatively flexible subsystem.
A common application is, for example, a bridge with a top lateral bracing
as shown in figure 14.14. The effect of the induced deformation, however,
also appears in other types of bending systems.
Problem: Force driven Problem: Deformation driven

e
F F
e
e
F F
Solution: Increasing strength Solution: Reducing stiffness
Figure 14.13: Induced deformation torsional clamped beam
The position of the neutral axis is dependent on the difference in stiffness

between the top lateral bracing and the bottom deck.
The shortening of the neutral axis lneutral axis due to the pure math-
ematical deflection is a negligible factor with regard to the mechanical
deformations lbottom edge and ltop edge related to the cross-sectional
and material stiffness parameters.
The lengthening of the bottom edge lbottom edge gives an induced dis-
placement of the bearings. The shortening of the top edge ltop edge gives
an induced deformation of the top lateral bracing.
Triangular bracing When the system as a whole is statically determ-

inate, induced deformations cannot occur. The matching deformation of
a triangular hinged top lateral bracing is shown in figure 14.14.
14.4 Induced deformation 161
K-bracing
Triangular bracing
Deformation of the neutral axis
Figure 14.14: Induced deformation in a truss bridge
Normally however, the top chord is continuous over the truss span length,
making the structure statically indeterminate and inducing a deforma-
tion of the top lateral bracing. When designing a stiff truss bracing, the
compressed members will probably fail due to this induced deformation.
K-bracing Therefore, a more flexible K-bracing is an effective design

to reduce the overall bracing stiffness by introducing flexible bending
elements among the stiff axial elements, thus reducing the load on all
bracing elements. The deformation of the top lateral K-bracing is shown
in figure 14.14.
Chapter 15
Fundamental parameters of
failure mechanisms
15.1 Conceptual design on element level

15.1.1 Dimensioning phase of the conceptual design
The dimensioning phase starts with element forces and displacements,
and ends with the required section properties for conceptual design, as
shown in figure 12.3 on page 130.
15.1.2 Failure mechanisms

Failure mechanisms typically include structural failures in the Ultimate
Limit State (ULS) and serviceability failures in the Serviceability Limit
State (SLS). Corresponding minimum required section properties have
to be defined to comply with functional requirements and contractual
conditions.
In search of a basic set of knowledge and approximation parameters
for conceptual structural design in general, and failure mechanisms in
particular, the following principles are employed:
164 Fundamental parameters of failure mechanisms
Applied mechanics-based to secure insight and durability. Basic ap-

plied mechanics textbook material is considered established know-
ledge and is therefore not included in the bibliography.
Professional practice factors for improvement of approximation ac-

curacy, including reference in the bibliography or verification in
this textbook.
Emphasis on deformation-driven aspects in accordance with the

force and deformation-driven parameters as shown in figure 12.2
on page 124.
Performance-based and well-balanced with respect to depth and

breadth of underlying knowledge and applicability for conceptual
design.
The following appropriate applied mechanics-based knowledge and ap-

proximation parameters are discussed as listed in table 15.1.
Conceptual design parameters of failure mechanisms

Shear Stability Bending & axial
Shear strength Principle Concrete elements
Shear deformation Euler based Steel elements
Stocky beam Buckling Foundations
Table 15.1: Conceptual design parameters of failure mechanisms
The material properties are based upon the applicable Eurocodes for
concrete [15] and structural steel [17].
15.2 Shear
15.2.1 Shear strength
For a reliable determination of the shear strength it is essential to com-
prehend the two major parameters fv and Av as shown in figure 15.1.
15.2 Shear 165
Figure 15.1: Shear strength
Material shear strength fv The shear strength fv for most mater-

ials is linearly proportional to the axial strength f and considerably
lower. For reinforced concrete the shear strength is approximately 20%
of the axial strength. Decisive for this global shear strength is the local
compressive strength of the concrete diagonals in the truss-like load dis-
tribution of the global shear force, provided that the shear reinforcement
is sufficient. For structural steel the shear strength is approximately 60%
of the axial strength.
Effective shear area Av The stiffness contribution of the perpendic-

ular to shear force orientated material such as flanges, is negligible with
regard to the material parallel to the shear force. Therefore, the shear
force will be borne solely by the shear area Av parallel to this force. Fail-
ure will result in rupture of the shear area before the flange material will
contribute, resulting in a propagating shear rupture of the whole cross
section.
The required effective shear area Av can be calculated as follows:
Vd Vd Vd
=  1 ) Av
Vu Av · f v fv
Vd Vd
Reinforced concrete Av , steel Av (15.1)
0.2fc 0.6fy
For standard slender beams, however, bending strength is mostly decisive

over shear strength.
15.2.2 Shear deformation

A beam as shown in figure 15.2 is subjected to a combined bending and
shear deformation.
stocky slender beam

F
Figure 15.2: Bending and shear deformation
For slender beams the bending deformation is decisive and the shear
deformation may be neglected, for stocky beams the shear deformation
is decisive and the bending deformation may be neglected.
15.2 Shear 167
For a slender beam, the bending deformation is in general:

M
¨
bending = (15.2)
EI
l
And in this specific case:
F l3
bending = (15.3)
3EI
For a stocky beam, the shear deformation is in general:

V
ˆ
shear = (15.4)
GA
l
And in this specific case:
Fl
shear = (15.5)
GA
The transition from “stocky” to “slender” depends on the bending versus

shear stiffness of the material. The transition slenderness cannot be de-
termined by an equilibrium of foregoing shear and bending formulae,
because the abstraction of bending is based upon slender beam the-
ory and the transition slenderness is beyond the validity range of this
theory.
In mechanics-based analyses of the transition slenderness from “stocky”
to “slender” of a simply supported beam, the span-over-height ratio hl
varies between 3 and 4. The Dutch National Annex to the Eurocode
2 [16] defines this transition slenderness for simply supported concrete
beams explicitly at a span-over-height ratio of hl = 3.
Considering the relatively small difference in E G = 2 (1 + ⌫) between con-
crete and steel, 2 (1 + 0.2) = 2.4 respectively 2 (1 + 0.3) = 2.6, the cor-
responding transition slenderness of the two will be approximately sim-
ilar, resulting in span-over-height ratios hl as shown in figure 15.3.
l l
Figure 15.3: Transition slenderness for stocky to slender beams
15.2.3 Shear deformation on subsystem level

Shear deformation on subsystem level occurs in cantilevered truss struc-
tures. For a comprehensible conceptual design, the shear contribution on
element level has to be appointed. An example of a usual, present-day
cantilevered parallel chord building design is shown in figure 15.4.
uniform load q
H
h
H
Δl
ql "
Figure 15.4: Shear deformation in a cantilevered truss
Because of the usually heavy dimensioned floor-carrying chord members

in combination with the nearby stocky dimensions of this cantilevered
truss, the displacement at the cantilevered end of the truss will be domin-
ated by shear deformation, rather than bending deformation. To realise
this, the design in the serviceability limit state should focus on shear de-
formation and the corresponding required section properties of the web
15.3 Stability 169
members. The linearity of deformation is based on a strength-optimised

design of the web members.
Approximated loads (ULS):
1 2
2 qd l
Nchord, max = H = (15.6)
h
1
qd · npanel 2 ·l
Nweb, max = (15.7)
sin ↵
Displacement due to shear deformation (SLS):
l Nweb, k · lweb
= npanel · = npanel ·
sin↵ EAweb · sin ↵
r⇣ ⌘2
qk ·(npanel 12 )·l l
sin ↵ · npanel + h2
= npanel ·  required (15.8)
EAweb · sin ↵
Required cross-sectional area with regard to shear deformation (SLS):

r⇣ ⌘2
qk ·(npanel 12 )·l l
sin ↵ · npanel + h2
Aweb, max (15.9)
required
npanel · E · sin ↵
15.3 Stability
15.3.1 Stability of the equilibrium
The basic principles of the stability of the equilibrium can be demon-
strated by the uncomplicated structural element as shown in figure 15.5.
A body may be in one of three states of equilibrium: stable, unstable, and
neutral. Analysis on the state of stable equilibrium can be effectively ex-
ecuted by a displacement over a slight distance . The structural element
is in stable equilibrium if it returns to its equilibrium position.
!
N
instability N
neutral equilibrium
stable equilibrium
!
Figure 15.5: Stability of the equilibrium
Driving moment leading to an instability failure mode:
Nu · (15.10)
Resisting moment providing a possible equilibrium:
H ·l =k· ·l (15.11)
Equilibrium and corresponding ultimate failure load:
Nu · = (k · ) · l ) Nu = k · l (15.12)
What we can learn from the equilibrium of this uncomplicated structural

element is applicable for stability problems in general:
Stability is deformation-driven The ultimate failure load Nu with
respect to stability has no relation whatsoever with strength para-
meters, but is solely determined by flexural stiffness parameters
15.3 Stability 171
such as length, cross-sectional bending stiffness and boundary con-

ditions.
Cross-sectional strength is naturally determined by the cross-sectional

area in combination with the material strength: Nu = A · fd .
In addition, this analysis shows that every rule has p

its “field of applica-
tion” to be considered; in this case, the assumption l2 2 ⇡ l is valid
under the condition of small displacements with respect to the overall

geometry.
15.3.2 Second-order effects

There are second-order effects when the first-order deformation due to
loading affect the distribution of internal forces.
The basic principles of second-order effects can be demonstrated by a

column with an initial deformation due to fabrication e0 and subjected
to an axial load N as shown in figure 15.6.
N N N
N N
Figure 15.6: Second-order effects

The initial deformation e0 , combined with the axial load N , causes a

first-order bending moment distribution with M = N · e0 in the middle
of the column. The corresponding bending deformation is additional to
the initial deformation and consequently leads to an amplification of the
bending moments.
In a stable structure, the second-order effects are self-limiting; the second-
order deformations are small enough that they do not continue to amp-
lify. Unstable structures exhibit a different behaviour; second-order
deformations are large enough to induce even larger forces, leading to
greater and greater deformations and ultimately to collapse.
For such a combined compression and bending, the quantification of the
second-order effects is substantiated with a magnification factor nn 1 in
subsection 15.3.4.
15.3.3 Euler-based design approximations

The ultimate failure load with respect to stability, is defined by Euler as
follows:
⇡ 2 EI
NEuler = (15.13)
L2cr
Where Lcr is the buckling length with design approximations as shown

in figure 15.7.
For braced frames, a buckling length of l is a safe approximation. For
unbraced frames, there is no absolutely safe approximation; for common
frames a buckling length of 2.5 l is appropriate [12].
For in-plane buckling of arches, a buckling length of 0.5 larch is appropri-
ate [26]. For out-of-plane buckling the overall arch length is appropriate,
which is normally reduced by adding lateral supports, such as bracings
between the arches.
The mathematical formula of Euler is valid for a slender, perfectly straight
member of homogeneous material that is free from initial stress. Because
15.3 Stability 173
Braced frame Unbraced frame
Figure 15.7: Buckling length
of deformation, material inhomogeneity, and residual stress due to fab-

rication, the Euler critical load has to be reduced by an adjustment
factor k:
⇡ 2 EI
Nu = (15.14)
kL2cr
Whereby the lowest Nu of both axes is decisive for the failure load.
In the professional practice, the value for the adjustment factor k can be
set at 1.7, based on experimental results of buckling failure as included
in Eurocode 3 [17] and listed in table 15.2.
For both an adjustment factor k = 1.7 and a buckling curve b, the
critical buckling stress cr is determined for structural steel grade S355.
The accuracy of the adjustment factor k in relation to the Eurocode-
based experimental results amounts to ±30%.
The sectional strength-related, basic applied mechanics-based, design
approximations have an accuracy far within the required ±20% as listed
in table 8.5 on page 87.
The stability-related design approximations, however, are more difficult
to capture. Probably the most sensitive is the buckling design approx-
imation with an accuracy of ±30%.
Critical buckling stress steel grade S355

EC3 buckling curve b Adjustment factor k = 1.7
cr cr %
0 355 355 0
20 347 355 +2
40 310 355 +15
60 260 339 +30
80 201 190 -6
100 150 122 -19
150 77 54 -30
Table 15.2: Critical buckling stress comparison
Global buckling strength Whether it is local buckling on element

level or global buckling on subsystem level, the same conceptual design
formulae for buckling strength Nu and buckling length Lcr are applicable.
Both the moment of inertia I and buckling length Lcr have to be related
to the subsystem as a whole.
Lateral torsional buckling strength Lateral torsional buckling of a

beam in bending is dependent on both buckling failure of the compressed
flange and opposing effects such as torsional resistance and stabilisation
of the tensioned flange.
As a safe approximation for conceptual design, opposing effects can be
neglected. Lateral torsional buckling can then be determined on the basis
of the buckling of the compressed flange. The same conceptual design
formulae for buckling strength Nu and buckling length Lcr can then be
applied.
Lateral buckling restraints All lateral restraints of buckling-sensitive

structures or parts of structures have the function of a stiffener since sta-
bility in general, and buckling in particular, is deformation-driven.
However, the required stiffness can be obtained with a strength-driven
15.4 Bending and compression 175
design approximation Fd for the lateral restraint: Fd = 1%·Nd [18].

It should be noted that a lateral restraint is not just a detailing matter
but a noticeable structural element, which has to be connected to a fixed
support.
15.3.4 Combined compression and bending

The combination of compression and bending with respect to stability
is too complex to handle on the level of conceptual design approxima-
tions.
However, when possibly decisive, this combination can be designed on
stress level, including the estimation of an initial deformation e0 . Con-
sidering the relative complex calculation, such an analysis functions more
as a check than as a design.
On stress level, the combined compression and bending can be described
with the following equation:
Nd n Nd · e0 n Md
= + · + · f
A n 1 W n 1 W
⇡ 2 EI
NEuler L2
and n = = cr (15.15)
Nd Nd
The following corresponding unity check is equivalent:

Nd n Nd · e 0 n Md f
+ · + · 
A·f n 1 W ·f n 1 W ·f f
Nd n Nd · e0 n Md
) + · + ·  1 (15.16)
Nu n 1 Mu n 1 Mu
15.4 Bending and compression

15.4.1 Conceptual design of concrete in bending
Bending strength of concrete members Because of the insignific-
ant tensional strength of concrete, global bending moments and shear
forces in a concrete member cannot be resisted through homogeneous

material behaviour.
The local load distribution within a concrete beam consists of axial com-
pression and tension forces in the concrete and the reinforcement, re-
spectively, in complete analogy with the load distribution in an arch for
stocky beams and a truss for slender beams, as shown in figure 15.8.
Figure 15.8: Arch and truss analogy in concrete beams
A design approximation of the ultimate strength of the usually govern-

ing bending rather than shear, can be effectively modelled by a lever
arm with an approximated value of 0.8h between the centres of gravity
of the compressed concrete zone and the reinforcement [2] as shown in
figure 15.9.
Figure 15.9: Bending strength of a concrete member
Within the cost optimal limits of reinforcement percentage, the strength

of the reinforcement will be decisive for the bending strength of concrete
members. As a professional practice-based average for cost optimal con-
crete design for an individual reinforcement direction and layer, a rein-
forcement percentage of 0.5% of the gross concrete cross section b · h will
be applied.
Consequently, the amount of reinforcement As to provide the required
bending strength Md can be based on the reinforcement strength fs and

the lever arm of 0.8h:
Md Md Md
=  1 ) As (15.17)
Mu (As · fs ) · 0.8h fs · 0.8h
The required section height h, relative to the member length l, for a

simply supported beam under a uniform load can be derived as follows:
s
1 2 1
q d l 0.5 h 8 qd
As 8
= 100 bh ) 0.5 (15.18)
fs · 0.8h l b · 435 · 0.8 · 100
The required section height can be adjusted to resist load distributions

due to non-uniform loading or beam end restraints, for example for the
end field of a continuous beam:
s
1 2 1
q
10 d l 0.5 h 10 qd
As = 100 bh ) 0.5 (15.19)
fs · 0.8h l b · 435 · 0.8 · 100
The width b of a poured-in-place concrete beam will mostly be adapted

to the supporting concrete structure for reasons of formwork economics.
In any case, lateral torsional failure must be prevented. A common value
for the hb -ratio is between 12 and 34 [1].
For a floor slab, the load qd is directly proportional to the width b, so usu-
ally, the reinforcement design will be executed per m1 unit width.
Displacement of concrete members For the displacement of a con-

crete member, the material and cross-sectional related stiffness EIc has
to be approximately determined. The material stiffness Ec is derived
from the simplified bi-linear modelled stress-strain relationship of con-
crete with an elastic strain limit of 1.75%�, including the influence of
time-related effects, as discussed in subsection 13.3.3 about material
properties for conceptual design. The cross-sectional stiffness Ic is based
on an uncracked cross-section in the serviceability limit state.
5 qk l 4 fc 3
= · and EIc = 1.75 · 1
12 bh (15.20)
384 EIc 1000
The required section height h, relative to the member length l, for a

simply supported beam can be derived as follows:
5 qk l 4 l
= · fc 1

384 · 3 250
1,75 12 bh
1000
r
h 5 qk
) 3
· 250 · 0.021 (15.21)
l 384 b · fc
The required section height can be adjusted to resist load distributions

due to non-uniform loading or beam end restraints, as for the end field
of a continuous beam:
1 qk l 4 l
= · fc 1

150 · 3 250
1,75 12 bh
1000
r
h 1 qk
) 3
· 250 · 0.021 (15.22)
l 150 b · fc
15.4.2 Conceptual design of concrete in compression

Axial strength The axial strength capacity of a more or less homo-
geneous material is dependent on the strength fd of the material in com-
bination with the area A of the cross section.
The required area of concrete can simply be calculated as follows:
Nd Nd Nd
= 1)A (15.23)
Nu A · fc fc
This simple model can also be applied as a design approximation for

reinforced concrete whereby the contribution of the reinforcement may
be neglected.
Buckling strength of concrete members For reinforced concrete

the Euler critical load has to be reduced by an adjustment factor of
k = 1.7.
The required sectional height can be derived as follows:
s
Nd Nd 1.7 · L2cr
= 2 1 3
 1 ) h weak axis
3
N d · b
(15.24)
Nu 1 ⇡ Ec 12 bh
· ⇡ 2 Ec · 12
k L2cr
15.4.3 Conceptual design of steel in bending

Bending strength of steel members Global bending moments and
shear forces in a steel member can be resisted through nearly homogen-
eous material behaviour with a plastic stress distribution for standard
rolled sections with the lowest steel grade S235 and a mostly elastic stress
distribution for higher grades [17] as shown in figure 15.10.
Figure 15.10: Bending strength of a steel member
Due to material costs and weight of structural steel, the design of cross
sectional areas has to be fully optimised. In order to support this optim-
isation process, design approximations have to be based upon the section
modulus W :
Md Md Md
= 1)W (15.25)
Mu W · fy fy
The required section modulus for a simply supported beam under a uni-
form load is as follows:
1 2
8 qd l
W (15.26)
fy
The required section modulus can be adjusted to resist load distributions

due to non-uniform loading or beam end restraints, for example for the
end field of a continuous beam:
1 2
10 qd l
W (15.27)
fy
Displacement of steel members For the displacement of a steel

member, the material and cross-sectional related stiffness EI is govern-
ing. For steel, the material stiffness E has a constant value of 2.1·105 mm
N
2
and is completely independent of the steel grade.
The required cross-sectional stiffness or moment of inertia I for a simply

supported beam can be derived as follows:
5 qk l 4 l 5 qk l 3
= ·  )I 250 · · (15.28)
384 EI 250 384 E
The required moment of inertia can be adjusted to resist load distribu-

tions due to non-uniform loading or beam end restraints, for example for
the end field of a continuous beam:
1 qk l 4 l 1 qk l 3
= ·  )I 250 · · (15.29)
150 EI 250 150 E
15.4.4 Conceptual design of steel in compression

Axial strength The axial strength capacity of a more or less homo-
geneous material is dependent on the strength fd of the material in com-
bination with the area A of the cross section.
The required area of steel can simply be calculated as follows:
Nd Nd Nd
= 1)A (15.30)
Nu A · fy fy
Buckling strength of steel members For structural steel, the Euler

critical load has to be reduced by an adjustment factor of k = 1.7.
The required moment of inertia can be derived as follows:
Nd Nd L2cr
= ⇡ 2 EI
 1 ) Iweak axis Nd · 1.7 · (15.31)
Nu 1
· ⇡2E
1.7 L2cr
15.4.5 Conceptual design of foundations

Spread foundation strength The structural action of a spread found-
ation consists of a load spread, a stabilisation by a possibly present top
loading, and a complex failure mechanism of slipping soil particles as
top load q
ce
urfa
slips
2:1
load spread
Figure 15.11: Failure mechanism of a spread foundation
Soil failure occurs with the formation of a load and soil dependent slip
surface, characterised by the slope of the slip surface and the shear
strength along this slip surface.
A safe and considerably simplified, ultimate limit state design approxim-
ation is a maximum of uniform assumed soil stresses at the foundation
contact area of 0.2 mm
N
2.
This professional practice-based conceptual design approximation is based

on a common foundation type on sand ground, a base width of about 1
metre and a minimal embedment depth of 0.3 metre.
In combination with horizontal loading, intermediate layers or adjacent
(future) excavations, a more deepening analysis is required.
Pile foundation strength The structural action of a pile foundation

consists of a pile head bearing capacity, a pile shaft friction-bearing capa-
city and a complex failure mechanism of slipping soil particles as shown
in figure 15.12.
ce
fa
sur
p
sli
Figure 15.12: Failure mechanism of a pile foundation
Soil failure occurs with the formation of a load and soil dependent slip
surface, characterised by the slope of the slip surface and the shear
strength along this slip surface.
A safe and considerably simplified ultimate limit state design approx-
imation is a maximum of uniform assumed soil stresses at the pile head
area of 5 N
mm2
.
This professional practice-based conceptual design approximation is based
on a common typical Dutch foundation type on a deeper sand layer with
substantial base resistance, partial shaft resistance, and partial negative
skin friction.
This design approximation can even be applied to the conceptual design
of tension piles, despite the difference in failure mechanism.
Part IV
Case study and training

Chapter 16
Introduction to part IV
Both qualified and quantified solutions are clarified by means of a case

study and a series of trainings, including a zero measurement training.
The resulting chapter arrangement is visualised in figure 16.1.
Case study This specific case study demonstrates how the methodical
approach leads to a controlled build-up of insight into the behaviour of
the structure and supports the actual successive design decisions dur-
ing the conceptual design of the trusses of the Maeslant storm surge
barrier.
of performance, structural, and construction demands, are determined.
phase.
Training The required knowledge, skills, and professional attitude have

to be achieved by a mix of the learning methods lecture, training, and
project work. Lectured theories and trained engineering practice can be
188 Introduction to part IV
Part III
Part IV
Case study
Chapter 17
Training programme
Chapter 18
3-D + simple model 2-D + medium model 1-D + abstract model

Chapter 19 Chapter 20 Chapter 21
Epilogue
Chapter 22
Figure 16.1: Reading guide for part IV
The training programme consists of a series of trainings, including a zero

measurement training. All trainings have approximately the same entry
professional on graduate master level, with a balanced complexity of
geometry and modelling. The zero measurement training can be applied
as a self-check at the beginning of the learning process.
Epilogue It is important to study how the design is organised in prac-

tice, and especially the ways in which designers with different disciplin-
ary expertise are able to work together, collaboratively in teams. For an
integral conceptual structural design, the main contributing disciplines
and corresponding interfaces have to be considered
This textbook discusses conceptual structural design on a high level of
abstraction. However, a deepening research on conceptual structural
design is valuable and feasible. Recommendations for research are given
with respect to both understanding these complex interdisciplinary in-
terfaces and structural performance.
Chapter 17
Case study trusses Maeslant

storm surge barrier
17.1 Demonstration of the methodical approach

17.1.1 A case study of conceptual structural design
The purpose of this specific case study is to demonstrate how the meth-
odical approach can lead to a controlled build-up of insight into the
behaviour of the structure and supports the actual successive design de-
cisions during conceptual design.
For a clear comparison between standard practice and the methodical

approach, diffuse conceptual design variables such as experience, and es-
pecially intuition, have to be eliminated. Therefore, (a part of) an actual
personally executed conceptual structural design such as the trusses of
the Maeslant storm surge barrier is preferred rather than a new project.
of performance, structural, and construction demands are determined.
190 Case study trusses Maeslant storm surge barrier

phase.
17.1.2 Actual design method

The actual design method at the time of the conceptual structural design
of the Maeslant barrier doors can be characterised by the following:
3-D Modelling From the beginning, full three-dimensional modelling
is applied for all conceptual design activities as geometry design,
load distribution analysis, and code checking.
Force-driven The initial geometry design and optimisation is force-
driven based. Subsequently, code-based unity checks on sectional
strength, stability, and the hollow section joints are conducted.
Reactive The reactive unity check-based conceptual design indicated a
severe failure problem of the web members. Analysis revealed an
induced deformation problem and corresponding solution.
Risk analysis As one of the first Design and Construct contracts, a
thorough risk analysis of the material demand was conducted for
tendering, and subsequently, as internal transfer documentation for
the basic design team.
17.1.3 Methodical approach

The methodical approach as practiced on the conceptual structural design
of the Maeslant barrier door trusses, can be characterised by the following:
2-D Modelling Decomposition-based conceptual design ensures a con-
trolled build-up of insight into the behaviour of the structure, with
a progressive insight from two-dimensional estimations to three-
dimensional accuracy.
Deformation-driven Deformation-driven conceptual design with con-
tinuous consideration of aspects such as local stability, global sta-
bility, and induced deformation.
17.1 Demonstration of the methodical approach 191
Pro-active A pro-active methodical approach including principal de-

tails and foreseen induced deformation, with the help of orienta-
tion, analysis, and check loops.
Risk analysis For complex design and construct-based contracts - be-
ing today’s standard in the field of civil engineering - a thorough
risk analysis of the material demand is inevitable as transfer doc-
umentation for basic design.
17.1.4 Outcome of the case study

The methodical approach leads to a controlled build-up of insight into
the behaviour of the structure and supports the actual successive design
decisions during conceptual design, on the basis of the following coherent
set of solution components:
Structural design path The structural design is effectively explored
from structural integrity, via load distribution, to failure mechan-
isms.
Structural design loops Conceptual structural design is based on a
progressive insight with orientation, analysis, check and, if neces-
sary, correction loops.
Load path design Out of the functional requirements, an initial sys-
tem outline is created based on major load paths, an order of mag-
nitude, and principal details.
Load distribution parameters After a decomposition in two-dimen-
sional subsystems, further geometrical optimisations are clarified,
and the load distribution is established.
Dimensioning parameters On the basis of so-obtained member loads
in combination with deformation-driven conditions, conceptual di-
mensioning can eventually be specified.
Structural design cycle The fundamental design cycle with the de-
composed design phases creation, optimisation, and dimensioning
proved to be viable and effective.
Basic structural forms Decomposition of the three-dimensional struc-

tural system into two-dimensional basic structural forms - in this
case trusses - proved an effective basis for structural analysis.
Shared knowledge-based conceptual design The interface of
structural demand with architectural, and in this case particular
construction demand, relies mainly on professional experience. A
successful material demand and risk analysis is conducted.
17.2 Maeslant storm surge barrier

17.2.1 Final piece of the Delta works
After the North Sea flood of 1953, a commission was installed which
had to come up with a plan to research the causes and seek measures
to prevent such a disaster in future; they came up with a plan for the
so-called “Delta works”.
The plan consisted of blocking the estuary mouths of the Oosterschelde,
the Haringvliet, and the Grevelingen. This reduced the length of the
dikes exposed to the sea by 700 kilometres. The mouths of the Nieuwe
Waterweg and the Westerschelde were to remain open because of the
shipping routes to the ports of Rotterdam and Antwerp. The dikes along
these waterways were to be heightened and strengthened.
The construction of the Maeslantkering was the final stage of these Delta
Works. The main objective was improving the safety against flooding
of the Rotterdam harbour and the surrounding towns and agricultural
areas. In the original plan, this had to be carried out by the reinforce-
ment of existing dikes as far as 50 kilometres inland. During the 1980s,
it became clear that dike reinforcement would take at least 30 years
and would only have disadvantages compared to a barrier as listed in
table 17.1.
Therefore, the initial plan was put aside and the Ministry of Waterways
and Public Works organised a competition in which construction com-
panies could make plans for the construction of a reliable, yet relatively
17.2 Maeslant storm surge barrier 193
inexpensive, storm surge barrier.
Dike reinforcement versus barrier

Dike Barrier
Costs (in billion euros) 0.82 0.45
Uncertainty costs ±20% ±10%
Delta safety schedule 2020 2000
Uncertainty schedule 10 years 2 years
Storm surges exposure 300 km 35 km
Environmental impact Large scale Limited
Table 17.1: Dike reinforcement versus barrier
17.2.2 Requirements
The storm surge barrier had to be located in the waterway that connects
Rotterdam with the North Sea. As this waterway is the main route to the
port of Rotterdam, a wide opening, unlimited headway, and a minimum
disturbance of ship movements, among others, were required as listed in
table 17.2.
Functional requirements for the Maeslant storm surge barrier

Unlimited headway
360 metres’ wide opening
Barrier failure 1/1,000,000 in any one year
Minimum disturbance of ship movements
100 years’ lifetime
Table 17.2: Requirements for the Maeslant storm surge barrier
17.2.3 Conceptual design Maeslant storm surge barrier

The winning barrier design with two huge hollow floating barrier doors
was put forward by the Bouwcombinatie Maeslant Kering (BMK) con-
sortium and became one of the first large Design and Construct projects
in the Netherlands.
Each barrier door consists of a 210 metres’ long retaining wall with flood-
able buoyancy chambers, supported by 250 metres’ long trusses and end-
ing in a ball joint with a diameter of 10 metres, and embedded in a
concrete caisson as shown in figure 17.1.
Figure 17.1: Maeslant storm surge barrier
Under normal conditions, the barrier doors are fully opened to allow
ships to sail to and from Rotterdam. However, if the water level rises
by three metres above the designated norm, the barrier doors are closed
and flooded with water. This causes them to sink slowly onto the sill
blocks at the bottom of the waterway. The entire process takes about
90 minutes.
During a storm surge, the water level on the North Sea side rises rel-
ative to the water level on the Rotterdam side. The corresponding hy-
draulic design load against one door equals approximately 350 meganew-
ton (MN).
A major advantage of this design was that construction of the storm
surge barrier could take place under dry conditions, in dry docks. Other
advantages were that no vital parts of the barrier had to be placed under
water, and maintenance of the barrier would be easy because of the
dry docks. Finally, there would be almost no inconvenience for passing
ships.
17.3 Creation of a system outline 195
17.2.4 Performance/cost optimisation

For a coherent and complete integral design, all fundamental demands -
namely performance demand, structural demand, and construction de-
mand - have to be complied with. Within a Design and Construct con-
tract, the performance/cost optimisation within the functional require-
ments is up to the integral constituted conceptual design team.
With respect to performance and construction demand, the following

major performance/cost optimisations are essential for the conceptual
design of the trusses:
Architectural demand The functionalism-based concept of architect

Wim Quist for the design of the trusses, consists of a combination of
tubular offshore-like appearance and a powerful clear transmission
of the enormous hydraulic forces.
Maintainability Maintenance costs are a major issue for such a func-

tional storm surge barrier. Therefore, the conceptual design of the
steel trusses demands special attention to the coating by minim-
ising surface area and avoiding sharp edges.
Construction demand Welding is, for the construction of steel trusses,

a major laborious and thus cost-dominant design parameter. The
design objective is thus an optimisation of weldability by minim-
ising weld volume and maximising welder accessibility.
17.3 Creation of a system outline

17.3.1 Load-path design on system level
The primary design load consists of a combination of approximately 6
metres of hydraulic head and approximately 3 metres of transversing
wave load, and can initially be simplifiedly modelled by a quasi-static
hydraulic head design load of 9 metres, and a corresponding uniform
load of qd = 90 kN
m in accordance with figure 17.2.
21
0m
250 m
22 m
360 m
Figure 17.2: Hydraulic load
The resulting force of the ball joint amounts to 320 MN as a result of

the projected hydraulic head in combination with the influence of the
top and bottom of the wave.
Resisting this primary load and bridging 250 metres to the balancing con-
crete caisson requires a large amount of material. The corresponding self-
weight of the structure is an inevitable secondary load of importance.
17.3.2 Principal details

Circular Hollow Section (CHS) members and joints are by far preferable,
in analogy with offshore structures and based on the following perform-
ance/cost optimisation considerations:
Coated surface Surface area and sharp edges highly affect the influen-
tial costs of coating and corresponding maintenance. Minimisation
of the coated surface, and avoiding sharp edges evidently, leads to
a circular section.
Water pressure To avoid coating the inner surface, and the corres-
ponding need for corrosion inspection and maintenance on the in-
side of such a complex structure, the trusses are completely water-
tight. Consequently, a circular section is the best way to resist the
water pressure.
Drag coefficient The drag coefficient quantifies the drag or resistance

of an object in a fluid environment such as air or water. A circular
section has a very low drag coefficient and is therefore less prone to
both wind loading and dynamic hydraulic loading when submerged.
Element buckling Circular hollow sections have an excellent element

buckling strength due to their all-directional high moment of iner-
tia, low geometrical tolerances and low residual stresses.
Plate buckling A higher moment of inertia can be obtained by making

sections thin-walled. Due to their circular shape, circular hollow
sections effectively combine a high moment of inertia and excellent
plate buckling resistance. Plate buckling, however, imposes a limit
to the extent to which sections can be made thin-walled; dt 5 50.
Uniformity To obtain cost-saving identical tubular gap joints and web

member lengths, the trusses are designed with parallel bottom and
top chords.
Joint geometry Aiming for a full capacity connection, the multi-planar

gap joints are designed with canned sections and an average ddchord
web
ratio of 0.6.
In this way, a principal joint as shown in figure 17.3 leads to a life cycle
design that is both economically and architecturally satisfying.
Figure 17.3: Principal joint

17.3.3 Circular hollow section elements

For an effective design of trusses with CHS members, section properties,
steel grade, and the transition between sectional strength and element
buckling have to be determined.
Section properties CHS The section properties of a CHS, with ap-

proximations based on a relatively small wall-thickness, are given in fig-
ure 17.4.
Figure 17.4: Section properties of a CHS
Steel grade At the time of the design of this storm surge barrier,
standardised grades for structural steel varied from S235 up to S355.
Given the functional requirements with respect to failure, this specific
design is primarily force-driven instead of deformation-driven. Thus, the
highest steel grade S355 is appropriate, thereby reducing the enormous
self-weight of the structure.
Transition slenderness CHS The slenderness el = ldel , where the

axial strength equals the buckling strength, defines the transition between
both failure mechanisms.
Axial strength:
Nu = A · fd (17.1)
Buckling strength of an element in a braced structure:
1 ⇡ 2 EI 1 ⇡ 2 EI
Nu ⇡ · 2 ⇡ · 2 (17.2)
1.7 Lcr 1.7 lel
Transition slenderness el,trans CHS:
1 ⇡ 2 EI 1 ⇡ 2 E · 18 Ad2 1 ⇡ 2 E · 18 A
A · fd ⇡ · 2 ⇡ · 2 = · 2
1.7 lel 1.7 lel 1.7 el
s r
lel E 210000
) el,trans = = 0.85 = 0.85 = 20.7 (17.3)
d fd 355
17.3.4 Decomposition in subsystems

For a clear and effective design of the primary and secondary load paths,
the three-dimensional system is decomposed into two-dimensional hori-
zontal and vertical subsystem planes, as shown in figure 17.5.
Figure 17.5: Decomposition in subsystem planes
Horizontal plane The horizontal plane directs the load paths of the
primary hydraulic load. To provide a short transition of this primary
hydraulic load to the ball joint, the level of this horizontal plane coincides
with the resulting hydraulic design load.
The total required cross-sectional area of the bottom chords per barrier
door amounts to:
l · h · qd 210 · 22 · 90
Abottom = = · 103 = 117 · 104 mm2 (17.4)
fd 355
Vertical plane The vertical plane directs the load paths of the sec-
ondary structural self-weight. Furthermore, this plane directs the load
paths caused by an eccentricity of the resulting hydraulic load due to the

transversing wave.
The secondary top chords are less loaded than the primary bottom
chords. Nevertheless, out of strength and fabrication considerations
the preferred principal CHS joints are appropriate for both bottom and
top chord joints. In combination with stability considerations, a three-
dimensional truss configuration as shown in figure 17.6 is the logical
consequence.
Figure 17.6: Triangular cross-sectional truss configuration
Intersection of horizontal and vertical plane The intersecting line

between the horizontal and vertical plane gives an interaction between
the hydraulic load and the structural self-weight.
The maximum hydraulic load in the horizontal plane is relieved by the
structural self-weight in the vertical plane. However, when submerged,
the buoyancy of the bottom chords will partly neutralise this relief since
the buoyancy of the submerged members approximately equals their own
weight.
The before mentioned eccentricity of the resulting hydraulic load on the
spot of the bottom of the transversing wave, and the corresponding re-
17.4 Optimisation of the structural action 201
duced water level also generates a relieving load with respect to the
structural self-weight. Because of its temporal character, this relief will
not be taken into account.
3-D system effects Possible three-dimensional load distribution ef-

fects require retention of these effects during decomposition of the three-
dimensional system into two-dimensional subsystems.
In case of such a relatively flat structural system, three-dimensional sys-

tem effects, and especially overall torsion, cannot be expected to be
influential.
17.4 Optimisation of the structural action

17.4.1 Optimisation on horizontal subsystem level
In search of an optimal configuration, the absolute minimum performance-
based configuration will be determined first; and then the possibility of
redundancy as an added value, will be explored.
Performance-based cost minimisation The absolute minimum con-

figuration consists of a statically determinate arrangement of two sup-
ports of the retaining wall. To secure global stability over a length of 250
metres, in plane horizontal stabilisation elements have to be designed in
addition to the vertical weight bearing structures. However, the great
distance between both supports requires heavy bracings.
Therefore each support is designed as a three-dimensional truss con-

figuration. An efficient design of the retaining wall, however, requires
more supports. Considering the long load path between the retaining
wall and the ball joint, additional bracings instead of complete trusses
are the most economical solution. Additional supporting bracings are
designed as shown in figure 17.7.
Figure 17.7: Optimisation of the supports of the retaining wall
Added-value design Nevertheless, one or more extra supporting

trusses can be included to incorporate redundancy in this statically
determinate design. However, even with four instead of two three-
dimensional supporting trusses, failure of only one truss will inevitably
lead to failure of the whole system due to an insurmountable loss of re-
taining wall capacity. So extra supporting trusses are still useless and
costly.
17.4.2 Optimisation on vertical subsystem level

The vertical height h and the horizontal width w of the three-dimensional
truss configuration as shown in figure 17.6 on page 200, have to be de-
termined.
Vertical truss height The optimal height/span ratio of trusses varies

between 101
and 15
1
. Because of the secondary load character of the truss
design, a height/span ratio near 15
1
is appropriate with a height h of 18
metres.
The resulting corresponding equilateral triangle has a web member, and

bottom chord and top chord lengths of approximately 20 metres.
17.5 Dimensioning 203
Horizontal truss width For a sufficient connection angle of both web

members to the top chord, with regard to the cross-section of the truss as
shown in figure 17.6 on page 200, the width w requires 15 metres.
17.4.3 Induced deformation

A timely determination of deformation-driven design parameters is of
importance for an effective optimisation of the structural action during
design. This statically indeterminate truss structure with primary loaded
chord members and secondary loaded web members can be potentially
sensitive to induced deformation of the web members.
Because the primary loaded chord members can induce an impermissible
deformation and corresponding failure of the web members, a sensitivity
study, as input for final dimensioning of these web members, has to be
conducted.
17.5 Dimensioning
17.5.1 Cross-sectional area of the bottom chord
In the statically determinate arrangement, the hydraulic load will be
divided over 2 · 2 = 4 cross sections. Each cross section requires the
following area:
1
Abottom = · 117 · 104 = 29.3 · 104 mm2 (17.5)
4
17.5.2 Global stability

The load-bearing capacity of the truss with respect to global buckling
can be approximated as follows:
1 ⇡ 2 EI 2
Nu ⇡ · 2 and Itruss ⇡ 0.8 · 2 · Achord · 1
2w (17.6)
1.7 lsys
With known cross-sectional area Abottom of the bottom chord members,

the minimum required truss width w with respect to global buckling
amounts to:
s r
2
4.25 · Nd · lsys 4.25 · 207.9 · 106 · 2502 · 106 3
w = · 10
⇡ 2 · E · Abottom ⇡ 2 · 2.1 · 105 · 29.3 · 104
= 9.5 5 actual 15 m (17.7)
17.5.3 Cross-sectional area of web and top chord

The uniform structural self-weight load of the bottom chords per barrier
door amounts to:
kN
qbottom = Abottom · ⇢s = 1.17 · 78.5 = 91.8 (17.8)
m
An additional top chord and a three-dimensional web member config-

uration will approximately double this weight. On the basis of post
calculation, the design value of the uniform structural self-weight load
amounts to approximately 200 kN m per barrier door.
The uniform structural self-weight load of approximately 100 kN m per

three-dimensional truss results in the following maximum cross-sectional
areas in the top chord and the web members:
1 2 1
8 qd l · 100 · 2502 · 106
Atop = = 8
= 12.2 · 104 mm2 (17.9)
h · fd 18 · 103 · 355
1
· 12 qd l
2
1 1
· · 100 · 250 · 103
Aweb = = 2 2
cos ↵ · cos · fd cos 22.6 · cos 30 · 355
= 2.2 · 104 mm2 (17.10)
17.5.4 Induced deformation of web members

Because the primary loaded chord members can induce an impermissible
deformation and a corresponding failure of the web members, a sensit-
ivity study is conducted.
17.5 Dimensioning 205
Phenomenon An induced deformation on subsystem level is applic-

able for the web members of the trusses of the Maeslant storm surge
barrier as shown in figure 17.8.
M !
Figure 17.8: Induced deformation of the Maeslant storm surge barrier
The angle ' = 6EI

Ml
is induced because of the elastic deformation of the
bottom chords due to the immense horizontal water load, in combination
with the stiffness of the welded tubular hollow section joints.
As a consequence of the induced angle ', fixed web member length l, and
fixed normal stiffness modulus E, the fraction M I , is a constant.
So reducing the bending moment M , as caused by the induced angle, is

only possible by reducing the cross-sectional moment of inertia I of the
web member.
Modelling During conceptual design the barrier door is modelled with

an overall finite elements model including retaining wall, trusses, and ball
joint. The load cases consist of hydraulic head, transversing wave, and
the structural self-weight.
The results are imported into a computerised code check including the
failure mechanisms of the individual truss members and their connection
to the canned joints.
Analysis To be on the safe side for the web members, the computerised
calculation input is initially based on CHS sizes of ? 900 ⇥ 30 mm.
The corresponding load distribution within the trusses results in a cross-
sectional failure of these web members, primarily due to bending.
Subsequently, the wall thickness is incremented until the bending strength

of the web members is sufficient. Even with a massive section of ? 900
mm, cross-sectional failure still occurs. This huge sectional area is com-
pletely out of proportion with the initial required axial strength.
The problem of induced deformation where the bending stiffness and the
corresponding bending moments increase more than the strength of the
structure, can be effectively solved by reducing the bending stiffness of
the concerning member.
Reducing the bending stiffness of the web members by reducing the wall
thickness to CHS sizes of ? 900 ⇥ 20 mm, the corresponding strength
proves more than sufficiently that it resists all forces including the in-
duced deformation-driven bending moments.
Design solution The potential induced deformation-driven problem

of cross-sectional failure of CHS web members with high plate thick-
nesses, can be effectively prevented by applying the relatively low plate
thickness of a ? 900 ⇥ 20 mm web member.
To prevent this potential induced deformation-driven problem during

further design optimisation, the cross-sectional bending stiffness of the
web members may not exceed the moment of inertia of a CHS ? 900⇥20
mm.
17.6 Specification and risk analysis 207
17.5.5 Section dimensions of conceptual design

The induced deformation-based dimensions of the web members, in com-
bination with the average ddchord
web
ratio of 0.6, results in the dimensions
Figure 17.9: Truss dimensions
17.6 Specification and risk analysis

17.6.1 Material demand
The specification of conceptual design demarcates this project phase. As
a Design and Construct project with a separated competitive tendering
and assignment phase, this specification is used as a risk analysis for
tendering, and subsequently, as internal transfer documentation for the
basic design team.
A risk analysis of the material demand gives an accuracy estimation of
the material quantities and corresponding unit cost indications. The risk
analysis has to be conducted on the level of individual components of a
decomposed system, principal details included.
For an effective risk analysis, as discussed in subsection 10.5.4, quantified
data have to be determined.
17.6.2 Dimensioning and cost weighting

The result of the performed structural analysis is an approximate section
dimensioning of the materialised overall geometry. This dimensioning of
the conceptual design is specified by a list of material quantities.
An executed example of the dimensioning and cost weighting during the
conceptual design of the trusses of the Maeslant storm surge barrier is
listed in table 17.3.
The cost weighting equals the material quantities times the unit cost
indications, and is a measure for cost optimisation opportunities and
corresponding risk.
17.6.3 Uncertainties and coverage by dimensioning

The difference between the performed approximate structural analysis
for conceptual design and the required depth and breath to meet the in-
use requirements for structural safety and serviceability can be defined
as uncertainties of the conceptual design. The difference in depth and
breath generally concerns load combinations, load distribution, and fail-
ure mechanisms.
The uncertainties of the performed approximate structural analysis with
respect to the required depth and breath cannot, or only partially, be
covered. The corresponding status of the coverage gives an indication of
the risk influence of the uncertainties of the conceptual design.
An executed example of the uncertainties and corresponding coverage by
dimensioning during the conceptual design of the trusses of the Maeslant
storm surge barrier is listed in table 17.4.
17.6.4 Reserves and optimisations

Reserves can be intentionally incorporated or are the result of rounding
up to the nearest standardised product dimensions. Occasionally, an
17.6 Specification and risk analysis 209
Risk analysis conceptual design trusses Maeslant storm surge barrier

Dimensioning and cost weighting
Description Weight Costs Weight
indication ⇥
[ton] [mhr/ton] Costs
Top chord members 2550 15 12%
? 1500 ⇥ approx. 30 mm
1 splice-weld/member
Bottom chord members 5790 10 18%
? 1500 ⇥ approx. 70 mm
1 splice-weld/member
Web members 2830 35 31%
? 900 ⇥ approx. 20 mm
2 splice-welds/member
Hollow section joints 600 60 11%
in top chord
1 splice-weld/joint
Hollow section joints 1170 45 17%
in bottom chords
1 splice-weld/joint
Ball joint ? 10 m 1010 35 11%
Performed analysis:
3-D framework analysis with post-processing on stress level
Table 17.3: Risk analysis: dimensioning and cost weighting
optimisation of requirements during the conceptual design phase can

result in a reserve.
Foreseen, but time-consuming cost optimisations can be postponed to
basic design, but registered as a potential reserve with regard to the
completed conceptual design.

Uncertainties and coverage by dimensioning
Uncertainties Coverage by dimensioning
Top chord, bottom chords and web members:
Decisive load combination head + Average influence based on known
wave based on 5% of the members member forces
Fatigue No specific coverage, anticipation
not decisive
Geometrical non-linearity Average influence based on frame-
work analysis of one load case
Serial effect Assumption based on TNO re-
port B-89-455
Global buckling Average influence through en-
largement factor
Hollow section joints in top and bottom chords:
Decisive load combination un- No specific coverage
known
Fatigue No specific coverage, anticipation
not decisive
Net area due to man holes 10% extra plate thickness with a
length of 3 m per hole
Ball joint:
Load life cycle with FEM model Order of magnitude
stress distribution
Friction coefficient of bearing ma- 0.3
terial
Table 17.4: Risk analysis: uncertainties and coverage by dimensioning
An executed example of the incorporated reserves and possible optimisa-

tions during the conceptual design of the trusses of the Maeslant storm
surge barrier is listed in table 17.5.
17.7 Further optimisations during basic design 211

Reserves and optimisations
Reserves Optimisations
3% reserve bottom chords Unity Cross-section members
Check axial stress Ultimate Limit
State Geometry web members with em-
phasis on length of chord mem-
Slight over-dimensioning of some bers versus number of joints
top chord and web members
due to minimum plate thickness Eccentricity connections
Weight reduction of 830 ton due

to head reduction (14-07-89) and
disposed inlet slides (11-10-89)
Table 17.5: Risk analysis: reserves and optimisations
17.7 Further optimisations during basic design

17.7.1 Basic design
During conceptual design, the functional requirements and contractual
conditions are evolved into a materialised overall system geometry, prin-
cipal details included, with approximated modelling-based dimensions,
quantities of materials, and corresponding risk analysis.
The subsequent basic design phase is characterised by a thorough analysis-
based optimisation of the structural action and a thorough overall struc-
tural system check.
Furthermore, the basic design is assisted by testing of the ductility
of the thermo-mechanical steel and the cast steel, the pre-stressed in-
jection bolts, and the friction coefficient of the coating of the sliding
surfaces.
Corresponding gained insight can reveal a need for adjustment and ap-
parently even an addition of structural elements.
17.7.2 Extensive modelling

During basic design, the stiffness influence of both retaining wall and ball
joint on the structural action in the trusses is refined within the overall
finite elements model of the barrier door:
Retaining wall An accurate stiffness contribution of the combined bend-
ing plate and longitudinal stiffener, the combined shear plate and
longitudinal stiffener, and the transverse stiffener, is individually
modelled and condensed into the model of the retaining wall. Sub-
sequently, the tubular bracings and partition bulkheads are in-
cluded.
Ball joint The effects of slip-stick, variation of the friction coefficient,
and actual rolling instead of slipping is determined with numerous
life cycle calculations with partial models and the condensed overall
model.
17.7.3 Coupling truss

During basic design, the refined modelling of the retaining wall in com-
bination with more refined load cases and more load combinations, re-
vealed unacceptable high longitudinal stresses in the retaining wall. These
high stresses were mainly caused by the transversing wave load, with a
top and bottom of approximately ±3 metres.
For a more uniformly distributed support of the retaining wall, coupling
of both supporting multi-planar trusses was required. Addition of an
extra multi-planar coupling truss resulted in the necessary reduction of
these longitudinal stresses.
17.7.4 Geometry and Section dimensions

During basic design, a more detailed code check was executed, includ-
ing all individual failure mechanisms of the individual members, and
particularly, their connection to the canned joints.
For the trusses, welding in situ was inevitable. Large lengths of the
17.7 Further optimisations during basic design 213
chord members could be welded on ground level and hoisted in place

afterwards. However, all connections of the web members had to be
welded in their permanent position.
Because of the difficultly accessible and laborious welding, a profound
optimisation of geometry in combination with weld profile, volume, pro-
cess, tolerances, and position was executed during basic design. This
resulted in a reduction of 10% of the total weld volume and correspond-
ing costs.
As a result of this optimisation of strength and welding of the tubular
joints, the diameter of the web members was altered to 800 mm, and the
diameter of the chords to 1800 mm.
Chapter 18
Training programme
18.1 Training programme outline

18.1.1 Load path design
Due to the huge degree of freedom, conceptual structural designing re-
quires simple and clear modelling. This is merely possible on the level
of axial forces, directly, with a truss-analogy or an arch depending on
the structural form. Load path design on system level as discussed in
section 13.2 is included in this training programme.
18.1.2 Programme outline

The required knowledge, skills, and professional attitude have to be
achieved by a mix of the learning methods lecture, training, and pro-
ject work. Lectured theories and trained engineering practice can be

zero measurement training, as listed in table 18.1. All trainings have
approximately the same entry professional on graduate master level, with
a balanced complexity of geometry and modelling:
216 Training programme
• Complex 3-D geometry combined with simple modelling.

• Medium 2-D geometry combined with medium modelling.
• Simple 1-D geometry combined with abstract modelling.
Training programme outline

No. Training Geometry Modelling
0 Spatial struts and ties 3-D Simple
1 Spatial trusses 3-D Simple
2 Box truss with torsion
3 Combined truss and arch 2-D Medium
4 Second-order effect arch
5 Truss analogy concrete beam 1-D Abstract
6 Truss analogy steel beam
Table 18.1: Training programme outline
The training requirements can be identified by utilising a zero measure-

ment; the zero measurement training 0 can be applied as a self-check at
the beginning of the learning process.
18.2 Training 0 - Spatial struts and ties

18.2.1 Intermediate bracing system with struts and ties
Given a 3-D bracing system with struts and ties as shown in figure 18.1.
Such a structure typically can be utilised as an intermediate load distri-
bution layer between a new and an underlying existing building.
For this training, the following requirements are applicable:
• Cross-sectional dimensions have to be determined for the ULS, the
SLS is not applicable.
• The load-bearing capacity of the 4 bearings is limited to 400 kN
each.
18.2 Training 0 - Spatial struts and ties 217
800 kN
1600 kN 4m
2 x 6 = 12 m
400 kN
800 kN
400 kN
400 kN 4m
400 kN 2 x 9 = 18 m
Figure 18.1: Bracing system with struts and ties
• The 4 struts consist of structural steel CHS with wall-thickness

t = 10 mm and normal stress strength fy = 355 N/mm2 . The sec-
tion properties of a CHS, with approximations based on a relatively
small wall-thickness, are given in figure 17.4 on page 198.
• The 2 crossing ties consist of structural steel round bars with nor-
mal stress strength fy = 355 N/mm2 .
• The structural self-weight may be neglected in comparison with
the loading of 1600 kN.
Analyse and determine the cross-sectional dimensions of this statically
indeterminate structure with the following problem-solving approach:
1. Consider the separate statically determinate simply supported 2-D
subsystems and determine the individual member forces and cor-
responding cross-sectional dimensions.
2. Now, model the complete 3-D system and determine the cross-
sectional dimensions of the ties, assuming the struts are infinitely
stiff.
3. With this 3-D model, determine the cross-sectional dimensions of

the struts, this time assuming the ties are infinitely stiff.
4. Determine the cross-sectional dimensions of the overall system and

the corresponding displacement at the location of the loading of
1600 kN.
5. Analyse and optimise the total amount of structural material for

this system.
6. Perform a redundancy analysis for the optimised system and con-

clude whether the optimisation is still valid when a second method
of support is required.
18.2.2 Strength design separate 2-D struts and ties

For the statically determinate simply supported 2-D subsystems, the
individual member forces for both the short and the long span can be
determined as shown in figure 18.2.
1.1 .9
72 400 984 400
600 900
Figure 18.2: Load distribution 2-D struts and ties
For the short span, the cross-sectional dimension of the tie amounts to:
Nd 600 · 103
A = = 1690 mm2
fy 355
) ? 50 mm (1963 mm2 ) (18.1)
For the short span, the cross-sectional dimension of the strut amounts to:
Nd 721.1 · 103
A = = 2031 mm2 and
fy 355
L2 72112
I Nd · 1.7 · 2cr = 721.1 · 103 · 1.7 · 2 = 3076 · 104 mm4
⇡ E ⇡ · 2.1 · 105
) ? 200 ⇥ 10 mm (6283 mm2 ) (18.2)
For the long span, the cross-sectional dimension of the tie amounts to:
Nd 900 · 103
A = = 2535 mm2
fy 355
) ? 60 mm (2827 mm2 ) (18.3)
For the long span, the cross-sectional dimension of the strut amounts to:
Nd 984.9 · 103
A = = 2774 mm2 and
fy 355
L2 98492
I Nd · 1.7 · 2cr = 984.9 · 103 · 1.7 · 2 = 7836 · 104 mm4
⇡ E ⇡ · 2.1 · 105
) ? 280 ⇥ 10 mm (8796 mm2 ) (18.4)
The cross-sectional dimensions based on the strength design of the sep-

arate 2-D subsystems are listed in table 18.2.
Cross-sectional dimensions 2-D subsystems design

Element Short span Long span
Tie ? 50 mm ? 60 mm
Strut ? 200 ⇥ 10 mm ? 280 ⇥ 10 mm
Table 18.2: Dimensions 2-D struts and ties

18.2.3 Deformation-driven design of the ties

For such a statically indeterminate and thus deformation-driven 3-D sys-
tem, the required stiffness and corresponding cross-sectional dimensions
of the ties can be determined assuming the struts are infinitely stiff.
For both the short and the long span, the elongation l of the tie and
corresponding displacement at the location of the loading can be de-
termined as shown in figure 18.3.
4 4
6 9
Figure 18.3: Deformation of the ties
For the long span, the cross-sectional dimension of the tie amounts to:
1 = 2 with tie 1 ? 50 mm (1963 mm2 )

6 6 F1 · l 1 6 600 · 103 · 6000
1 = · l1 = · = · = 13.1 mm
4 4 E · A1 4 2.1 · 105 · 1963
9 4
1 = 13.1 = 2 = · l2 ) l2 = · 13.1 = 5.8 mm
4 9
F2 · l 2 F2 · l 2 900 · 103 · 9000
l2 = ) A2 = = = 6650 mm2
E · A2 E · l2 2.1 · 105 · 5.8
) tie 2 ? 95 mm (7088 mm2 ) (18.5)
18.2.4 Deformation-driven design of the struts

of the struts can be determined assuming the ties are infinitely stiff.
For both the short and the long span, the shortening l of the strut
and corresponding displacement at the location of the loading can be
11 49
7.2 9.8
4 4
6 9
Figure 18.4: Deformation of the struts
For the long span, the cross-sectional dimension of the strut amounts to:
1 = 2 with strut 1 ? 200 ⇥ 10 mm (6283 mm2 )

7.211 7.211 721.1 · 103 · 7211
1 = · l1 = · = 7.1 mm
4 4 2.1 · 105 · 6283
9.849 4
1 = 7.1 = 2 = · l2 ) l2 = · 7.1 = 2.9 mm
4 9.849
F2 · l 2 F2 · l 2 984.9 · 103 · 9849
l2 = ) A2 = = = 15928 mm2
E · A2 E · l2 2.1 · 105 · 2.9
) strut 2 ? 510 ⇥ 10 mm (16022 mm2 ) (18.6)
18.2.5 Design of the overall system

The actual system behaviour combines the deformations and correspond-
ing cross-sectional dimensions of both ties and struts.
These cross-sectional dimensions based on the deformation-driven design
of the 3-D system are listed in table 18.3.
Cross-sectional dimensions 3-D system design

Tie ? 50 mm ? 95 mm
Strut ? 200 ⇥ 10 mm ? 510 ⇥ 10 mm
Table 18.3: Dimensions 3-D spatial struts and ties
For the actual displacement at the location of the loading of 1600 kN,
both deformations of ties and struts have to be combined serially:
tot = 1,tie + 1,strut = 2,tie + 2,strut = 13.1 + 7.1 = 20.2 mm (18.7)
18.2.6 Material optimisation

The cross-sectional dimensions based on the strength design of both sep-
arate 2-D subsystems, as listed in table 18.2 on page 219, must not be
undershooted.
However, the extra material for the required stiffness and corresponding
displacement 2 of the long span, as listed in table 18.3 on page 221, is
interchangeable:
9 900 · 103 · 9000 86786
2,tie = · 5
= (18.8)
4 2.1 · 10 · A2,tie A2,tie
9.849 984.9 · 103 · 9849 113736
2,strut = · 5
= (18.9)
4 2.1 · 10 · A2,strut A2,strut
113736
The tie appears = 1.31 more effective than the strut.
86786
So, when strut 2 remains ? 280 ⇥ 10 mm (8796 mm2 ) instead of the
required 15928 mm2 ,
15928 8796
then tie 2 has to be increased with = 5444 mm2 ,
1.31
resulting in A2,tie = 6650 + 5444 = 12094 mm2 ) ? 125 mm.
The cross-sectional dimensions based on this material optimisation of
the 3-D system are listed in table 18.4.
Cross-sectional dimensions optimised 3-D system design

Tie ? 50 mm ? 125 mm
Strut ? 200 ⇥ 10 mm ? 280 ⇥ 10 mm
Table 18.4: Dimensions material optimisation
18.2.7 Redundancy analysis

The cross-sectional dimensions as listed in table 18.2 on page 219 are
based on a statically determinate strength design and consequently have
none residual strength. Only the extra material for the required stiffness
of the long span within the statically indeterminate system design does
have residual strength.
In case of an overload of the short span subsystem and corresponding
yielding of tie 1, the following redundancy can generally be generated by
the long span subsystem:
Short span The short span subsystem will maintain its full load-bearing
capacity but with zero stiffness, whereby the short span struts still
stabilise the top joint.
Long span Remains a statically determinate force driven 3-D system
with residual strength due to the extra material of the long span
subsystem, but with an overload of the corresponding bearings.
The non-optimised system with extra material for both struts and tie of
the long span subsystem, as listed in table 18.3 on page 221, can generate
such a redundancy.
However, the material-optimised system has no residual strength at all
due to a lack of extra material for the struts of the long span subsys-
tem, as listed in table 18.4 on page 222. So material optimisation and
redundancy perform like communicating vessels.
Chapter 19
3-D geometry combined with

simple modelling
19.1 Training 1 - Spatial trusses

19.1.1 Intermediate bracing system with trusses
Given a 3-D bracing system with trusses as shown in figure 19.1. Such
a structure typically can be utilised as an intermediate load distribution
layer between a new and an underlying existing building.
• Cross-sectional dimensions have to be determined for the ULS, the
SLS is not applicable.
• The load-bearing capacity of the 4 bearings is limited to 400 kN
each.
• The truss members consist of structural steel square RHS with wall-
thickness t = 5 mm and normal stress strength fy = 355 N/mm2 .
The section properties of a square RHS, with approximations based
on a relatively small wall thickness, are given in figure 19.2.
226 3-D geometry combined with simple modelling
800 kN
beam 1
1600 kN 4m
beam 1
4 x 3 = 12 m
400 kN
800 kN
beam 2
400 kN 400 kN 4m
beam 2
400 kN 6 x 3 = 18 m
Figure 19.1: Bracing system with trusses
• The top joints of the trusses are laterally supported by a concrete

floor slab.

the loading of 1600 kN.
Figure 19.2: Section properties of a square RHS

1. Consider the separate statically determinate simply supported 2-D

subsystems and determine the individual member forces.
19.1 Training 1 - Spatial trusses 227
2. Then, with these member forces, determine the corresponding cross-

sectional dimensions.
3. Now, model the complete 3-D system and determine the cross-
sectional dimensions of the chord members, assuming the web
members are infinitely stiff.
4. With this 3-D model, determine the cross-sectional dimensions of

the web members, this time assuming the chord members are in-
finitely stiff.
5. Determine the cross-sectional dimensions of the overall system,

based on standardised square RHS sections.
19.1.2 Load distribution in the separate 2-D trusses
For the statically determinate simply supported 2-D subsystems, the

individual member forces for both the short and the long span can be
beam 1 8(00) kN beam 2 8(00) kN

-3 -6 -6 -3 -3 -6 -9 -9 -6 -3
+5 +5 -8 -4 -4 +5 +5 +5 -8 -4 -4 -4
4m
-4 -4 +5 +5 -4 -4 -4 +5 +5 +5
0 +3 +3 0 0 +3 +6 +6 +3 0
4(00) kN 4(00) kN 4(00) kN 4(00) kN
4 x 3 = 12 m 6 x 3 = 18 m
Figure 19.3: Load distribution 2-D trusses

19.1.3 Strength design of the separate 2-D trusses

For the short span, the cross-sectional dimension of the chord members
amounts to:
Nd 600 · 103
A = = 1690 mm2 and
fy 355
L2 30002
I Nd · 1.7 · 2cr = 600 · 103 · 1.7 · 2 = 443 · 104 mm4
⇡ E ⇡ · 2.1 · 105
) ⇤ 120 ⇥ 5 mm (19.1)
For the short span, the cross-sectional dimension of the web members
amounts to:
Nd 500 · 103
A = = 1408 mm2 and
fy 355
L2 40002
I Nd · 1.7 · 2cr = 400 · 103 · 1.7 · 2 = 525 · 104 mm4
⇡ E ⇡ · 2.1 · 105
) ⇤ 120 ⇥ 5 mm (19.2)
For the long span, the cross-sectional dimension of the chord members
amounts to:
Nd 900 · 103
A = = 2535 mm2 and
fy 355
L2 30002
I Nd · 1.7 · 2cr = 900 · 103 · 1.7 · 2 = 664 · 104 mm4
⇡ E ⇡ · 2.1 · 105
) ⇤ 130 ⇥ 5 mm (19.3)
For the long span, the cross-sectional dimension of the web members
amounts to:
Nd 500 · 103
A = = 1408 mm2 and
fy 355
L2 40002
I Nd · 1.7 · 2cr = 400 · 103 · 1.7 · 2 = 525 · 104 mm4
⇡ E ⇡ · 2.1 · 105
) ⇤ 120 ⇥ 5 mm (19.4)
The cross-sectional dimensions based on the strength design of the sep-

arate 2-D subsystems are listed in table 19.1.
Cross-sectional dimensions 2-D subsystems design

Member Short span Long span
Chord ⇤ 120 ⇥ 5 mm ⇤ 130 ⇥ 5 mm
Web ⇤ 120 ⇥ 5 mm ⇤ 120 ⇥ 5 mm
Table 19.1: Dimensions 2-D trusses
19.1.4 Deformation driven design of the chord members

of the chord members can be determined assuming the web members are
infinitely stiff.
For both the short and the long span, the bending deformation is curved
due to axial deformation of the chord members and with a shared dis-
placement at the location of the loading as shown in figure 19.4.
beam 2 beam 1
Figure 19.4: Bending deformation of the truss beams
The required stiffness and corresponding cross-sectional dimensions of

the chord members amount to:
F · 123 F · 183 I1
1 = 2 ) = ) = 0.30 and
48 · EI1 48 · EI2 I2
1 2
Itruss = 2 · Achord · 2h ) Itruss ⇠ Achord (19.5)
(h ⇥ t)chord,1 = 120 ⇥ 5 mm
✓ ◆
120
) (h ⇥ t)chord,2 = ⇥ 5 ⇡ 400 ⇥ 5 mm (19.6)
0.30
19.1.5 Deformation driven design of the web members

of the web members can be determined assuming the chord members are
infinitely stiff.
For both the short and the long span, the shear deformation is linear due
to axial deformation of the web members and with a shared displacement
at the location of the loading as shown in figure 19.5.
beam 2 beam 1
Figure 19.5: Shear deformation of the truss beams
For both the short and the long span, the linear shear deformation, un-
derlying axial deformation of the web members, and corresponding dis-
placement at the location of the loading are shown in figure 19.6.
beam 1 beam 2
Figure 19.6: Deformation of the web members
The required stiffness and corresponding cross-sectional dimensions of

the web members amount to:
A1 2
1 = 2 )2 v,1 +2 d,1 =3 v,2 +3 d,2 ) = (19.7)
A2 3
(h ⇥ t)web,1 = 120 ⇥ 5 mm
✓ ◆
3
) (h ⇥ t)web,2 = · 120 ⇥ 5 ⇡ 180 ⇥ 5 mm (19.8)
2
19.1.6 Design of the overall system
The truss beams, with a slenderness between slender and stocky, are
subject to both significant bending and shear deformation. The actual
system behaviour combines these deformations and corresponding cross-
sectional dimensions of both chord and web members.
These cross-sectional dimensions based on the deformation-driven design

of the 3-D system are listed in table 19.2.
Cross-sectional dimensions 3-D system design

Chord ⇤ 120 ⇥ 5 mm ⇤ 400 ⇥ 5 mm
Web ⇤ 120 ⇥ 5 mm ⇤ 180 ⇥ 5 mm
Table 19.2: Dimensions 3-D spatial trusses
The final cross-sectional dimensions based on standardised square RHS

sections in accordance with the required strength and stiffness ratios are
listed in table 19.3.
Cross-sectional dimensions optimised 3-D system design

Chord ⇤ 120 ⇥ 5 mm ⇤ 200 ⇥ 10 mm
Web ⇤ 120 ⇥ 6.3 mm ⇤ 140 ⇥ 8 mm
Table 19.3: Standardised square RHS sections

19.2 Training 2 - Box truss with torsion

19.2.1 3-D square box truss with eccentric loading
Given a cantilevered 3-D square box truss with eccentric loading F as
shown in figure 19.7. The square frames are connected with chords and
the top, bottom, and side planes are stabilised with X-bracings (not all
bracings are shown in figure 19.7).
3m
3m
4m
4m
Figure 19.7: 3-D square box truss with eccentric loading
• The load-bearing capacity F has to be determined both for the

ULS with F = Fd /1.5 and the SLS with F = Fk .
19.2 Training 2 - Box truss with torsion 233
• For the SLS the displacement at the location of the loading F is

limited to l/125.
• The square frames and chords consist of structural steel RHS ⇤120⇥
120 ⇥ 8 mm (A = 3515 mm2 and I = 725.8 · 104 mm4 ) with normal
stress strength fy = 275 N/mm2 .
• The X-bracings consist of high strength steel round bars ? 20 mm

(A = 314.2 mm2 ) with normal stress strength fy = 1200 N/mm2 .

the loading F .
Analyse and determine the load-bearing capacity F of this statically de-

terminate structure with the following problem-solving approach:
1. Consider only the right side plane as a simplified 2-D model for
the load-bearing capacity of the complete 3-D system. With this
2-D model, determine the individual member forces expressed in
F and then determine the load-bearing capacity F for the ULS.
2. With this 2-D model, determine the displacement at the location

of the loading F and then determine the load-bearing capacity F
for the SLS.
3. Now, model the complete system with a 3-D model. Split the
eccentric loading F in a centric force and a torsion couple and
then determine the load-bearing capacity F for the ULS.
4. With this 3-D model, determine the displacement at the location

of the loading F and then determine the load-bearing capacity F
for the SLS.
5. Analyse and explain the difference in load-bearing capacity between

the simplified 2-D model and the 3-D model.
6. Explain why the transition slenderness of l/h = 1.5 for cantilevered

stocky to slender beams as shown in figure 15.3 on page 168 is not
applicable for this case.
19.2.2 Simplified 2-D modelling ULS

Only the right side plane is considered as a simplified 2-D model for the
load-bearing capacity of the complete 3-D system.
With this 2-D model, the individual member forces expressed in F can
be determined as shown in figure 19.8.
F F
F F
Figure 19.8: Individual member forces
Axial strength:
3
Nu = A · fs = 314.2 · 1200 · 10 = 377.0 kN
3
 3515 · 275 · 10 = 966.6 kN (19.9)
Buckling strength:
⇡ 2 EI ⇡ 2 · 2.1 · 105 · 725.8 · 104 3

Nu = = · 10
1.7L2cr 1.7 · 40002
= 553.1 kN (19.10)
Then, the load-bearing capacity F for the ULS can be determined:
Fd 1 3
F = = · · 553.1 = 138.3 kN
1.5 1.5 8
1 3
 · · 377.0 = 150.8 kN (19.11)
1.5 5
19.2.3 Simplified 2-D modelling SLS

With equal normal stiffness modulus E and more than an order of mag-
nitude difference between Atube (3515 mm2 ) and Abar (314.2 mm2 ), a
displacement at the location of the loading F will be largely caused by
elongation of the bars.
5
3
4
Figure 19.9: Displacement due to the elongation of a bar
The displacement due to the elongation of the bars, as shown for one bar
in figure 19.9, amounts to:
5 53 F · l 2 · 4000
tot =2· ·  = 64 mm (19.12)
3 E·A 125
Then, the load-bearing capacity F for the SLS can be determined:
✓ ◆2 ✓ ◆2
1 3 E·A 1 3 2.1 · 105 · 314.2
F = 64 · · · = 64 · · · · 10 3
2 5 l 2 5 5000
= 152.0 kN ( 138.3 kN) (19.13)
19.2.4 3-D system modelling ULS

Now, the complete system is 3-D modelled.
The eccentric loading F is split in a centric force and a torsion couple as
Geometry, section properties, and loading pattern of top, bottom, and
side planes are completely identical. Only magnitude and direction of
the loading varies.
+ =
Fe
Figure 19.10: Centric force and torsion couple
Maximum load and strength of frames and chords:

3 8 1 4 5
Nmax = · F (compression) · F (tension) = F
4 3 4 3 3
and Nu = 553.1 kN (19.14)
The maximum loaded rear bottom right chord is in compression by a

downwards force 34 F on the right side plane and in tension by a leftwards
force 14 F on the bottom plane.
Maximum load and strength of X-bracings:
3 5 5
Nmax = · F = F and Nu = 377.0 kN (19.15)
4 3 4
Then, the load-bearing capacity F for the ULS can be determined:

Fd 1 3
F = = · · 553.1 = 221.2 kN
1, 5 1.5 5
1 4
· · 377.0 = 201.1 kN (19.16)
1.5 5
19.2.5 3-D system modelling SLS

The governing displacement of the right side plane amounts to:
5 54 F · l 2 · 4000
tot =2· ·  = 64 mm (19.17)
3 E·A 125
Then, the load-bearing capacity F for the SLS can be determined:
1 3 4 E·A 1 3 4 2.1 · 105 · 314.2

F = 64 · · · · = 64 · · · · · 10 3
2 5 5 l 2 5 5 5000
= 202.7 kN ( 201.1 kN) (19.18)
19.2.6 2-D versus 3-D modelling

The load-bearing capacity F of the simplified 2-D model versus the 3-D
system model is listed in table 19.4.
Load-bearing capacity F
Limit state Simplified 2-D model 3-D system model
ULS F = 138.3 kN F = 201.1 kN
SLS F = 152.0 kN F = 202.7 kN
Table 19.4: 2-D versus 3-D modelling
The doubling of load-bearing material with an extra left side plane has
not resulted in a doubling of the load-bearing capacity F , due to an
eccentricity of the load and corresponding introduction of torsion in the
3-D modelling.
However, the actual additional material will generally generate more
strength and stiffness as quantified in table 19.4.
19.2.7 Stocky or slender system behaviour

The transition slenderness of l/h = 1.5 for cantilevered stocky to slender
beams, as shown in figure 15.3 on page 168, is not applicable for this
case.
As a result of the high strength steel, the X-bracings are an order of mag-
nitude more flexible than the frames and chords. So, shear deformation
of the 3-D square box truss is governing, resulting in a stocky behaviour
of the structure despite the fact that l/h = 2.7 1.5.
Chapter 20

medium modelling
20.1 Training 3 - Combined truss and arch

20.1.1 Truss girder bridge building combined with arch
Given a truss girder bridge building combined with a tied arch as shown
in figure 20.1. The truss girder is combined with an arch out of archi-
tectural considerations.
• The floor and roof loading amounts to qd = 1.2 · 15 + 1.5 · 10 =
33 kN/m1 respectively qd = 1.2 · 10 + 1.5 · 4 = 18 kN/m1 , the
structural self-weight may be neglected in comparison with this
loading.
• For reasons of redundancy both the truss girder and the tied arch
have to bear half the load.
• For the SLS the displacement is limited to l/500.
• The truss members and arch consist of structural steel CHS 20  d
t  30
240 2-D geometry combined with medium modelling
12 m
4m
12 x 3 = 36 m
Figure 20.1: Truss girder bridge building combined with tied arch
with normal stress strength fy = 460 N/mm2 . The section prop-

erties of a CHS, with approximations based on a relatively small
wall-thickness, are given in figure 17.4 on page 198.
• The hangers consist of structural steel round bars with normal
stress strength fy = 460 N/mm2 .
• For in-plane buckling of the arch, a buckling length of 0.5larch ⇡
23 m is appropriate. Out-of-plane buckling is prevented by lateral
bracing.
• The deformation of the arch is negligible relative to the elongation
of the hangers.
1. Consider the separate statically determinate simply supported truss
girder subsystem with half the load and determine the member
forces, corresponding cross-sectional dimension of the chord mem-
bers, and the displacement in the ULS.
2. Consider the separate statically determinate simply supported tied
arch subsystem with half the load and determine the member
forces, corresponding cross-sectional dimensions, the influence of
20.1 Training 3 - Combined truss and arch 241
an eccentric loading, and the displacement in the ULS.

3. Now, model the complete statically indeterminate system with
combined truss girder and tied arch and determine the actual load
distribution between the truss girder and the tied arch.
4. With this modelled system, analyse and determine how the util-
isation of high strength steel can realise the required equal load
distribution between the truss girder and the tied arch.
5. For the chord members of the truss girder, analyse and optimise
the amount of structural material on system level.
6. With this modelled system, determine and check the displacement
in the SLS.
7. For the web members of the truss girder, determine the actual
member forces on system level and corresponding cross-sectional
dimension.
20.1.2 Strength design of the separate truss girder

For the separate statically determinate simply supported truss girder
subsystem with half the load, the maximum chord and web member
forces amount to:
1 2 1
8 qd l · 25.5 · 362
8
Nchord, max = = = 1032.8 kN and
h 4
1
qd l 5
Nweb, max = 2 = · 12 · 25.5 · 36 = 573.8 kN (20.1)
cos ↵ 4
The cross-sectional dimension of the chord members amounts to:
Nd 1032.8 · 103
Achord = = 2245 mm2 and
fy 460
L2 1.7 · 60002
Ichord Nd · 1.7 · 2cr = 1032.8 · 103 · 2 = 3050 · 104 mm4
⇡ E ⇡ · 2.1 · 105
) ? 200 ⇥ 10 mm Achord = 6283 mm2 (20.2)
The deformation of the separate truss girder in the ULS amounts to:
5 qd l 4 2
= · and Itruss ⇡ 0.8 · 2 · Achord · 1
2h
384 EItruss
5 qd l 4
) truss = ·
384 E · 0.8 · 2 · Achord · 1 h 2
2
5 25.5 · 360004
= · = 66.0 mm (20.3)
384 2.1 · 105 · 0.8 · 2 · 6283 · 20002
20.1.3 Strength design of the separate tied arch

For the separate statically determinate simply supported tied arch sub-
system with half the load, the tie and maximum arch forces amount to:
1 2 1
8 qd l · 25.5 · 362
8
H= = = 344.3 kN = Ntie = Narch, mid
h 12 q
1 2
and Narch, max = 2 qd l + H 2 = 573.8 kN (20.4)
The cross-sectional dimension of the arch amounts to:
1
Nd (344.3 + 573.8) · 103
Aarch = 2
= 998 mm2 and
fy 460
L2 1.7 · 230002
Iarch Nd · 1.7 · 2cr = 12 (344.3 + 573.8) · 103 · 2
⇡ E ⇡ · 2.1 · 105
4 4
= 19918 · 10 mm ) ? 320 ⇥ 16 mm (20.5)
And the cross-sectional dimension of the hanger amounts to:
Nd 25.5 · 6000
Ahang = = 333 mm2
fy 460
) ? 21 mm Ahang = 346 mm2 (20.6)
A bending moment within the arch, due to an eccentric loading of half

the variable uniform loading, appears not to be decisive:
✓ ◆
2 1.5 · 10
1 1
March, max ⇡ 64 qd l = 64 · 362 = 151.9 kNm
2
March, max 151.9 · 106
Warch = = 330.2 · 103 mm3
fy 460
) Buckling is decisive ? 320 ⇥ 16 mm (20.7)
The deformation of the arch is stated negligible relative to the elongation

of the hangers, so the deformation of the separate tied arch in the ULS
is equal to the elongation of the middle hanger and amounts:
Nd l (25.5 · 6000) · 12000
arch = hang = = = 25.3 mm (20.8)
EA 2.1 · 105 · 346
20.1.4 Load distribution combined truss girder and arch
The complete system with combined truss girder and tied arch is a stat-
ically indeterminate structure and therefore the load distribution is de-
formation driven, whereby the deformation of the truss equals the elong-
ation of the middle hanger as shown in figure 20.2.
Figure 20.2: Displacement combined truss girder and tied arch
The distribution of the total load of qd = 51 kN/m1 between the truss

girder and the tied arch amounts:
arch 25.3
qtruss = · qd,tot = · 51
truss + arch 66.0 + 25.3
= 14.1 kN/m1 (20.9)
truss 66.0
qarch = · qd,tot = · 51
truss + arch 66.0 + 25.3
= 36.9 kN/m1 (20.10)
20.1.5 Equal load distribution with high strength steel

For reasons of redundancy both the truss girder and the tied arch have to
bear half the load, so the actual difference in stiffness between these two
subsystems can possibly be corrected with another steel grade.
A lower steel grade for the truss girder is useless due to the decisive
buckling strength.
A higher steel grade for the hangers of the tied arch is highly effective,
whereby the round bars can also be replaced by cable strands:
Nl N ·l
arch = hang = = truss ) Ahang = (20.11)
EA E · truss
N N N E· truss
Ahang = ) fy = = N ·l
=
fy Ahang E·
l
truss
2.1 · 105 · 66.0

= = 1155 N/mm2 (20.12)
12000
20.1.6 Material optimisation on system level

On system level the interaction between the top chord member in com-
pression and the arch tie in tension can lead to a material optimisation
of the chord members:
L2cr
Ichord (Nchord Ntie ) · 1.7 ·
⇡2E
60002
= (1032.8 344.3) · 103 · 1.7 · 2 = 2033 · 104 mm4
⇡ · 2.1 · 105
) ? 200 ⇥ 7 mm Achord = 4398 mm2 (20.13)
For the bottom chord member in full tension this cross-sectional area is
still sufficient:
Achord = 4398 mm2 Amin = 2245 mm2 (20.14)
Then, the higher steel grade and corresponding cross-sectional area of

the hangers amount to:
5 qd l 4
truss = · 2
384 E · 0.8 · 2 · Achord · 1
2h
5 25.5 · 360004
= · = 94.3 mm (20.15)
384 2.1 · 10 · 0.8 · 2 · 4398 · 20002
5
E· truss 2.1 · 105 · 94.3

fy = = = 1650 N/mm2
l 12000
Nd 25.5 · 6000
) Ahang = = = 93 mm2 (20.16)
fy 1650
20.1.7 Displacement in the SLS

Both Parallel subsystems truss girder and tied arch bear half the load:
(15 + 10) + (10 + 4)

qk,arch = qk,truss = = 19.5 kN/m1 (20.17)
2
Both Parallel subsystems have the same corresponding displacement and

satisfy the requirement for the SLS:
5 qk,truss l4
= arch = truss = · 2
384 E · 0.8 · 2 · Achord · 1
2h
5 19.5 · 360004
= · = 72 mm
384 2.1 · 105 · 0.8 · 2 · 4398 · 20002
l 36000
 = = 72 mm (20.18)
500 500
The tied arch, with equal displacement as the truss girder, can serve as
a verification:
Nk l (19.5 · 6000) · 12000

arch = hang = = = 72 mm (20.19)
EA 2.1 · 105 · 93
20.1.8 Strength design web members on system level
The truss load distribution in the web members can be determined by

a transformation of the uniform floor and roof loading into point loads
onto the truss joints as shown in figure 20.3.
27 54 54 27
34
.8
.3
15
50
4m
5
33
6
8.8
7.5
9
22
41
.1
24.75 74.25 99 99
459 3m 3m
Figure 20.3: Truss load distribution web members
The arch load distribution in the web members can be determined by

a transfer of the floor loading via the web members into the hangers as
The actual load distribution in the web members is a combination of the

truss load distribution and the arch load distribution, with a maximum
tension force of 509.1 + 61.9 = 571.0 kN and a maximum compression
force of 416.3 61.9 = 354.4 kN.
20.2 Training 4 - Second-order effect arch 247
99 kN
61
.9 4m
61
3m .9
3m
Figure 20.4: Arch load distribution web members
The corresponding critical cross-sectional dimension amounts to:
Nd 571.0 · 103
Aweb = = 1241 mm2 and
fy 460
L2 50002
Iweb Nd ·1.7· 2cr = 354.4·103 ·1.7· 2 = 727·104 mm4
⇡ E ⇡ · 2.1 · 105
) ? 150 ⇥ 6 mm (20.20)
20.2 Training 4 - Second-order effect arch

20.2.1 Three-hinged truss arch
Given a three-hinged truss arch as shown in figure 20.5. The point
load is transferred via a semi-circular arch out of architectural consider-
ations.
• Cross-sectional dimensions of the chord members have to be de-
termined for the ULS, the SLS is not applicable.
• The semi-circular truss arch has a radius r = 8 m and truss height
h = 1 m with web members at a 45 degree angle.
• The point load amounts to Fd = 1.2 · 360 + 1.5 · 270 = 837 kN and
the structural self-weight may be neglected in comparison with this
2r
Figure 20.5: Three-hinged truss arch
load.
• The truss members consist of structural steel RHS 15  bt  25
with normal stress strength fy = 355 N/mm2 . The section prop-
erties of a RHS, with approximations based on a relatively small
wall-thickness, are given in figure 19.2 on page 226.
Analyse and determine the cross-sectional dimension of the chord mem-
bers, including the deformation driven second-order effect of this stat-
ically determinate 2-D system, with the following problem-solving ap-
proach:
1. Consider the statically determinate semi-circular three-hinged arch
and analyse the load distribution.
2. Then, determine the corresponding N- and M-lines of the arch.
3. With these N- and M-lines, determine the cross-sectional dimension
of the chord members of the truss arch.
4. Analyse the second-order effect and determine whether this effect
has to be taken into account with respect to the final cross-sectional

dimension of the chord members.
20.2.2 Analysis load distribution in the 2-D system

The load distribution of the point load F into the semi-circular three-
hinged arch brings about a substantial first order moment M = pF2 · e
due to the large eccentricity e as shown in figure 20.6.
Figure 20.6: Load distribution three-hinged arch
20.2.3 N- and M-lines of the arch

For getting grip on the more complex N- and M-lines of a semi-circular
arch, the N-, M-, and, V-lines of enclosing simplified structures can be
exercised as shown in figure 20.7.
Now, the actual N- and M-lines of the semi-circular arch can be determ-
ined as shown in figure 20.8.
The corresponding quantification of the maximum values for the axial

F F e F F
F
N-line
M-line
V-line
F
Figure 20.7: N-, M-, and, V-lines of enclosing simplified structures
force N and the bending moment M amount to:
F 837 p
N = p = p = 591.8 kN and e = 8 4 2 = 2.343 m
2 2
) M = N · e = 591.8 · 2.343 = 1386.7 kNm (20.21)
8 4 N-line M-line
Figure 20.8: N- and M-lines of a semi-circular arch

20.2.4 Dimensioning chord members of the truss arch

The critical axial force in the chord members, as a result of the combined
axial force in the arch and the bending moment due to the eccentricity
of this axial force, amounts to:
N M 591.8 1386.7
Nd = + = + = 1682.6 kN (20.22)
2 h 2 1
The corresponding cross-sectional dimension amounts to:
Nd 1682.6 · 103
A = = 4740 mm2 and
fy 355
L2 20002
I Nd · 1.7 · 2cr = 1682.6 · 103 · 1.7 · 2 = 552 · 104 mm4
⇡ E ⇡ · 2.1 · 105
) ⇤ 160 ⇥ 8 mm Achord = 5120 mm2 (20.23)
20.2.5 Analysis second-order effect

The first-order deformation of the hinged arch section can be approxim-
ated with the deformation of a fictive uniformly loaded simply supported
beam and amounts to:
8M 8 · 1386.7
qf ictive = 2 = p 2 = 86.7 kN/m (20.24)
l 8 2
5 qf ictive · l4
= · 2
384 E · 0.8 · 2 · Achord · 1
2h
p 4
5 86.7 · 8000 · 2
= · = 43 mm (20.25)
384 2.1 · 105 · 0.8 · 2 · 5120 · 5002
There are second-order effects when the first-order deformation due to
loading affect the distribution of internal forces.
The approximate first-order deformation of 43 mm is more than an order
of magnitude smaller than de eccentricity of 2343 mm. So, the second-
order deformation has a negligible influence on the load distribution and
corresponding dimensioning.
Chapter 21

abstract modelling
21.1 Training 5 - Truss analogy concrete beam

21.1.1 Cantilevered reinforced concrete beam
Given a cantilevered reinforced concrete beam on two supports as shown
in figure 21.1.
500
500 1500
Figure 21.1: Cantilevered reinforced concrete beam

254 1-D geometry combined with abstract modelling

• Shear strength and corresponding shear deformation of the canti-
lever have to be determined for the ULS, the SLS is not applicable.
the point load.
• Within a truss analogy, the concrete diagonal struts have an effect-
ive cross-sectional area Ac = 5000 mm2 and normal stress strength
N
fc = 20 mm 2.
• The steel stirrup reinforcements have a cross-sectional area As =

200 mm2 and normal stress strength fs = 435 mm N
2.
Analyse and determine the shear strength and corresponding shear de-
formation of the (externally) statically determinate 1-D cantilevered con-
crete beam on two supports, modelled by a (internally) statically de-
terminate 2-D strut-and-tie model, with the following problem-solving
approach:
1. Consider a truss analogy with a 2-D strut-and-tie model and de-
termine the load distribution.
2. Determine the shear strength of the separate cantilever.
3. Then, determine the corresponding shear deformation of this sep-
arate cantilever.
4. Analyse the shear deformation caused by the concrete struts of
the right span and determine the displacement of the cantilever,
including the contribution of this right span.
5. Appoint the contribution of the other elements within the strut-
and-tie model which have then to be analysed.
21.1.2 Load distribution strut-and-tie model

The member forces of the statically determinate 2-D strut-and-tie model,
and corresponding V- and M-lines of the statically determinate 1-D con-
21.1 Training 5 - Truss analogy concrete beam 255
crete beam, are shown in figure 21.2.
V-line
M-line
Figure 21.2: Load distribution concrete beam
21.1.3 Shear strength of the concrete cantilever

The axial strength of the concrete strut amounts to:
3
Nu,c = Ac · fc = 5000 · 20 · 10 = 100 kN (21.1)
The axial strength of the steel tie amounts to:

3
Nu,s = As · fs = 200 · 435 · 10 = 87 kN (21.2)
The shear strength of the cantilever is completely dependent on the axial

strength of the concrete strut and amounts to:
Nu,c 100
Fu = p = p = 70.7 kN (21.3)
2 2
21.1.4 Shear deformation of the concrete cantilever

The shear deformation of the cantilever is completely dependent on the
axial deformation of the concrete strut as shown in figure 21.3.
Figure 21.3: Shear deformation of the cantilever
Then the shear deformation of the cantilever amounts to:

p
N ·l 100 · 103 · 500 · 2
l= = 20 = 1.2 mm
Ec · Ac 1.75·10 3
· 5000
p p
) = l 2 = 1.2 · 2 = 1.7 mm (21.4)
21.1.5 Displacement of the cantilever

Besides the cantilever, also the right span with corresponding axial de-
formation of the concrete struts is subject to shear deformation as shown
in figure 21.4.
The right span is 3⇥ longer, but the shear force is 3⇥ smaller, so the
shear deformation of the right span is equal to the shear deformation
of the cantilever.
As a result of this shear deformation of the right span in combination

with the right roller support, the displacement of the cantilever due to
both the axial deformation of the concrete strut of the cantilever and
the concrete struts of the right span amounts to tot = 43 as shown in
figure 21.5.
21.1 Training 5 - Truss analogy concrete beam 257
Figure 21.4: Contribution shear deformation right span
Figure 21.5: Displacement of the cantilever
21.1.6 Completion of the strut-and-tie modelling

This problem-solving approach first focussed on the contribution of the
concrete diagonal struts to the deformation of the cantilever.
The contribution of the other members within the strut-and-tie model

to this deformation of the cantilever has then to be analysed:
1. Deformation of the stirrup reinforcement of the right span.
2. Deformation of the longitudinal reinforcement.
3. Deformation of the longitudinal compressed concrete area.
21.2 Training 6 - Truss analogy steel beam

21.2.1 Cantilevered steel beam
Given a cantilevered steel beam on two supports as shown in figure 21.6.
Figure 21.6: Cantilevered steel beam

• Shear strength and corresponding shear deformation of the can-
tilever have to be analysed with a truss analogy and an effective
width be modelling of the web members.
the point load.
Analyse and determine the shear strength and corresponding shear de-
formation of the (externally) statically determinate 1-D cantilevered steel
beam on two supports, modelled by a (internally) statically determ-
21.2 Training 6 - Truss analogy steel beam 259
inate 2-D strut-and-tie model, with the following problem-solving ap-

proach:
1. Consider a truss analogy with a 2-D strut-and-tie model, whereby

for the determination of both shear strength and shear deformation
an effective width be modelling of the web members is applicable.
2. Determine the load distribution within this 2-D strut-and-tie model.
3. Determine the effective width be with regard to the shear strength

of the separate cantilever.
4. Then, determine the effective width be with regard to the corres-

ponding shear deformation of the separate cantilever.
5. Analyse the effective width be modelling with regard to both shear

strength and shear deformation and conclude whether this model-
ling is viable.
21.2.2 Truss analogy with an effective width modelling

The cantilevered steel beam as shown in figure 21.6 on page 258 is sub-
jected to a combined bending and shear deformation. For a slender beam
the bending deformation is decisive and the shear deformation may be
neglected as shown in figure 21.7.
Figure 21.7: Bending deformation cantilevered beam
For a stocky beam the shear deformation is decisive and the bending
deformation may be neglected as shown in figure 21.8. With a displace-
ment of the cantilever due to both the shear deformation of the individual
cantilever and the contributing shear deformation of the right span as
elaborated in figure 21.5 on page 257.
Figure 21.8: Shear deformation cantilevered beam
A beam between slender and stocky dimensions can best be analysed

with a truss analogy, where the load distribution completely consists of
axial forces in the truss members and therefore not limited by bending
formula based on slender beam theory. For the determination of both
shear strength and shear deformation an effective width be modelling of
the web members can be applied as shown in figure 21.9.
h V
V
δ
h
Figure 21.9: Effective width model
21.2.3 Load distribution truss analogy
Within the truss analogy the member forces of the statically determin-
ate 2-D strut-and-tie model can be determined and are shown in figure
21.10.
21.2 Training 6 - Truss analogy steel beam 261
Figure 21.10: Load distribution truss analogy
21.2.4 Shear strength of the steel cantilever

The shear strength of the actual web plate of the beam can be modelled
as follows:
fy
V u = Av · f v = h · t w · p (21.5)
3
The shear strength within a truss analogy can be modelled with an ef-
fective width be of the web members as follows:
b e · tw · f y
Vu = p (21.6)
2
Then, with regard to shear strength the effective width be can be de-
termined and amounts to:
p
fy be · tw · f y 2
V u = h · tw · p = p ) be = p · h ⇡ 0.8h (21.7)
3 2 3
21.2.5 Shear deformation of the steel cantilever

The shear deformation of the actual web plate of the beam can be mod-
elled as follows:
V ·h V ·h
v = = (21.8)
G · Av G · h · tw
The shear deformation within a truss analogy can be modelled with an

effective width be of the web members as follows:
p p
F ·l p V · 2·h· 2 p p V ·h
v = · 2= · 2=2 2 (21.9)
E·A E · be · tw E · be · tw
Then, with regard to shear deformation the effective width be can be

determined and amounts to:
V ·h p V ·h
v = =2 2
G · h · tw E · b e · tw
p G p 8.1 · 104
) be = 2 2 · · h = 2 2 · · h ⇡ 1.1h (21.10)
E 2.1 · 105
21.2.6 Analysis effective width modelling

This problem-solving approach focussed on a representative model with
a constant value for the effective width be . The difference between both
effective width be values, 0.8h respectively 1.1h, is too large for a viable
effective width modelling.
Chapter 22
Epilogue
22.1 Complexity of the built environment

22.1.1 Complexity
Complexity describes the behaviour of a system or model whose com-
ponents interact in multiple ways and follow local rules, meaning there
is no reasonable higher instruction to define the various possible inter-
actions [11].
These are typically large collections of connected elements that influence
each other. Examples are the brain; society; traffic; the financial system;
interacting institutions; climate; ecosystems; interacting atoms or mo-
lecules; the World Wide Web. These diverse examples have surprisingly
many features in common. As a rule, they show various properties that
make complex systems more than the sum of their parts.
Complexity is generally used to characterise something with:
1. Many components.
2. Where those components interact with each other in multiple ways.
3. Culminating in a higher order of emergence greater than the sum
264 Epilogue
of its parts.
However, many simple components interacting with each other in mul-
tiple simple ways cannot be characterised as a complex system. Therefore
the definition of complex systems by Herbert Simon as a “large number
of parts that interact in a non simple way. In such systems the whole is
more than the sum of the parts” [22], clarifies the higher order charac-
terisation.
22.1.2 Complex systems and processes

Design is an activity that plays a fundamental part in the creation of
the built environment. The interdisciplinary design of the built envir-
onment consists of cyclic design processes, culminating in a physical
system.
In the ISO 9000:2015 a system is defined as a “set of interrelated or inter-
acting elements” and a process is defined as a “set of interrelated or inter-
acting activities that use inputs to deliver an intended result” [10].
Complex system The traditional approach to dealing with complex-

ity is to reduce or constrain it. Typically, this involves decomposition:
dividing a large system into separate elements. The actual physical sys-
tem itself - although consisting of numerous physical elements - is not
complex.
The individual system elements of the built environment, however, can be
classified as physical or non-physical. For example architectural demands
can include non-physical elements, as aesthetics can neither be classified
as physical, nor as a process.
Particularly the interacting of these non-physical elements such as aes-
thetics, load paths, and constructability becomes complex when the cor-
responding traditional disciplinary boundaries have to be crossed.
Complex process It is important to study how the design is organised

in practice, and especially the ways in which designers with different
22.1 Complexity of the built environment 265
disciplinary expertise are able to work together, collaboratively in teams.

A motivation in these studies is not only to improve design processes but
also the designed system.
22.1.3 Complexity of interdisciplinary interfaces

For an integral conceptual structural design, the main contributing dis-
ciplines and corresponding interfaces have to be considered. Architec-
tural design is par excellence - besides construction engineering - an
important and much discussed interface.
In order to get hold on the complexity of interdisciplinary interfaces in
general, some fundamental characteristics of the interface between struc-
tural and architectural demand are researched; the creation of utilitarian
space with materialised forms as a main influential architectural design
interface with the structural form.
Fundamental architectural demand The following terms can serve

as directional guidance for the research on architectural demand for con-
ceptual structural design:
Architectural concept A guiding concept implies an idea or range of
ideas, a design intent or a development approach. It resolves the
issues of “what” and “how much” and begins to set the stage for
understanding “how”.
Architectural imaging In architecture, imaging often stands for a phys-
ical or digital visualisation. On a more conceptual level, it is related
to aesthetics, interpretations, and perspectives.
Structural form “The resistant virtues of the structures that we seek
depend on their form; it is through their form that they are stable,
not because of awkward accumulation of material” as stated by
Eladio Dieste [7].
Space in architecture The search for a definition which covers the

basic idea of architecture, and corresponds with the enclosure of space
266 Epilogue
with three-dimensional structural forms, results in the following more or

less common definitions:
Function, structure, and aesthetics Architecture is a combination
of function, structure, and aesthetics; these factors moving to-
gether through time creates architecture. A building exists as a
crystallisation of a given moment of society, technology, and art.
Utilitarian space Architecture is a conscious creation of utilitarian
spaces and construction of materials in such a way that the whole
is both technically and aesthetically satisfying.
Utilitarian art Frank Gehry and Santiago Calatrava design in three
dimensions. They create art that is tailored to provide shelter.
Their designs also serve certain programmatic needs.
An introduction to architecture and space in architecture is given by
Frank Ching in his book “Architecture: Form, Space, and Order” [5].
Complexity of the interdisciplinary interface Architectural de-

cisions are those that need to be made from an overall system perspect-
ive. Essentially, these decisions identify the key structural elements of the
system and their externally visible properties and relationships.
Further, they define how the architecturally significant requirements will
be achieved. During the conceptual design phase, the architectural de-
signer should focus on the capacity to bear and resist, in addition to the
architectural requirements.
Nowadays, both disciplines are seldom combined in one person, so mutual
cooperation is necessary. Although organising the physical meeting in an
early stage of the design process is a necessity, it is not enough; a mental
meeting for mutual understanding has to be arranged as well.
An architectural design is often concept-based, whereas a structural
design is form-based: a materialised structural form with the emphasis on
the internal distribution of the loads. In order to connect these manifest
differences in design attitude between architects and structural engineers,
22.2 Recommendations for research 267
a recognisable intersection of both disciplines has to be defined.

Furthermore, consumer ideology has turned architecture into fast cycles
of fashion and signature styles. Successive architectural movements claim-
ing an avant-garde position have emerged, one after the other, including
Minimalism, High Tech, Deconstructionism, and most recently, computer
generated Blobitecture.
Consequently, a workable intersection of architectural concepts and struc-
tural forms can be characterised as a complex interface.
22.2 Recommendations for research

22.2.1 In-depth research
This textbook discusses a methodical approach on conceptual structural
design on a high level of abstraction. However, a deepening research on
conceptual structural design is valuable and feasible. Recommendations
for research are given with respect to both understanding structural per-
formance and conceptual structural design.
Out of the multitude of missing in-depth knowledge, the following most
direct completing topics with respect to the discussed methodical ap-
proach are recommended for research:
• Conceptual design parameters built environment.
• Fundamental behaviour of structural materials.
• Transition of stocky to slender beam theory.
• Adjustment factor of buckling strength.
22.2.2 Conceptual design parameters built environment

A definition and collection of the fundamental conceptual design para-
meters of the influential object-related participating disciplines can serve
as a joint breeding ground for a concurrent “shared knowledge-based con-
ceptual design” as discussed in section 7.3.
268 Epilogue
Because of the complexity of the interdisciplinary interfaces between

these object-related disciplines, concurrent engineering will be an ap-
propriate solution as discussed in subsection 5.5.5 of the solution com-
ponents.
As given in figure 7.6 on page 74 the T-shaped in-breadth understanding
in general, and the conceptual structural design process control in partic-
ular, is qualitatively modelled in part II of this textbook. Subsequently
the conceptual structural design parameters are approximately quanti-
fied in part III.
Both the individual sets of conceptual design parameters of the other
influential disciplines and the integral process of interacting and control
of these sets are in need of approximate quantitative modelling.
22.2.3 Fundamental behaviour of structural materials

Design approximations of load distribution are common property through
standard applied mechanics textbooks. Design approximations of strength
and stiffness behaviour of common structural materials are widely access-
ible through numerous textbooks and design codes of practice.
Both design approximations of load distribution, and material strength
and stiffness of concrete as well as of structural steel are organised and
supplemented into a set of conceptual structural design parameters in
this textbook.
Approximated behaviour and strength of new structural materials, how-
ever, have to be modelled with care. Especially brittle material be-
haviour, and corresponding approximate conceptual modelling, is not
widely accessible. Brittle material behaviour requires far more in-depth
modelling to detect and prevent high-peak stresses with consequent pro-
gressive tearing failures.
In general, the relationship between the degree of ductility - quantified
by the length of the ductile or plastic zone - and the required corres-
ponding degree of in-depth modelling, has to be researched. In par-
ticular, the fundamental behaviour of such structural materials and an
22.2 Recommendations for research 269
effective approximate modelling for conceptual design should become

available.
22.2.4 Transition of stocky to slender beam theory

For slender beams, the bending deformation is decisive and the shear
deformation may be neglected; for stocky beams, on the other hand,
the shear deformation is decisive and the bending deformation may be
neglected. The transition from “stocky” to “slender” depends on the
bending versus shear stiffness of the material. This transition cannot be
determined with the common abstraction of bending based upon slender
beam theory.
For a better understanding and application of deformation in general,
and shear deformation in particular, research on the transition slender-
ness with the help of proper constructed in-depth truss-analogy and/or
finite element modelling, is advisable.
22.2.5 Adjustment factor of buckling strength

The mathematical formula of Euler gives a proper insight into structural
behaviour but has to be adjusted by a factor k for accuracy reasons.
Even then, buckling strength-related design approximations are difficult
to capture and often have insufficient accuracy with regard to conceptual
design.
Therefore, more thoroughly substantiating the material-dependent ad-
justment factor k, is desirable; first of all, an appropriate adjustment
factor k for the buckling strength of concrete members, in combination
with a corresponding stiffness modulus Ec ; and subsequently, material-
dependent adjustment factors for other common structural materials.
270 Epilogue
Bibliography
[1] Braam, C.R., P. Lagendijk. Constructieleer Gewapend Beton (CB2).

Boxtel: Aeneas Media, 2011.
[2] Braam, C.R. Ontwerpen in Gewapend Beton (CB4). Boxtel: Aeneas

Media, 2012.
[3] Büttner, O., E. Hampe. Bauwerk, Tragwerk, Tragstruktur, Band

2: Klassifizierung, Tragqualität, Bauwerkbeispiele. Berlin: Ernst &
Sohn, 1985.
[4] Calaprice, A., F. Dyson, A. Einstein. The New Quotable Einstein.

Princeton, NJ: Princeton University Press, 2005.
[5] Ching, F.D.K. Architecture: Form, Space, and Order. Hoboken, NJ:
John Wiley & Sons, 2007.
[6] Couger, D. Creative Problem Solving and Opportunity Finding. Dan-

vers, MA: Boyd & Fraser Publishing Company, 1995.
[7] Dieste, E. La estructura cerámica. Bogotá: Escala, 1987.
[8] fib Bulletin 65: Model Code 2010, Volume 1. International Federa-
tion for Structural Concrete fib “fédération internationale du béton”,
2012.
[9] Horikx, M.P. A Methodical Approach on Conceptual Structural

Design. Delft: Delft University of Technology, 2017.
272 BIBLIOGRAPHY
[10] ISO 9000:2015 Quality management systems - Fundamentals and

vocabulary. International Organization for Standardization, 2015.
[11] Johnson, S.B. Emergence: The Connected Lives of Ants, Brains,

Cities, and Software. New York: Scribner, 2001.
[12] Kamerling, M.W., J.W. Kamerling. Jellema deel 9 - Bouwmethoden

Utiliteitsbouw. Amersfoort: ThiemeMeulenhoff, 2011.
[13] Kickert, W.J.M. Organisation of decision-making: a systems-

theoretical approach. Amsterdam: North Holland Publishing Com-
pany, 1979.
[14] NEN-EN 1990+A1:2006+A1:2006/C2:2019 Eurocode: Basis of

structural design. Nederlands Normalisatie-instituut, 2019.
[15] NEN-EN 1992-1-1+C2:2011 nl: Eurocode 2: Design of concrete

structures - Part 1-1: General rules and rules for buildings. Neder-
lands Normalisatie-instituut, 2011.
[16] NEN-EN 1992-1-1+C2:2011/NB:2016+A1:2020 nl: National An-

nex to NEN-EN 1992-1-1+C2 Eurocode 2: Design of concrete struc-
tures - Part 1-1: General rules and rules for buildings. Nederlands
Normalisatie-instituut, 2020.
[17] NEN-EN 1993-1-1:2006 en: Eurocode 3: Design of steel struc-

tures - Part 1-1: General rules and rules for buildings. Nederlands
Normalisatie-instituut, 2006.
[18] NEN-EN 1993-1-1+C2+A1/NB:2016 nl: Eurocode 3: Design of

steel structures - Part 1-1: General rules and rules for buildings.
Nederlands Normalisatie-instituut, 2016.
[19] NEN-ISO 6707-1: 2017 en: Buildings and civil engineering works
- Vocabulary - Part 1: General terms. Nederlands Normalisatie-
instituut, 2017.
[20] Nervi, P.L. Structures. Transl. G. Salvadori, M. Salvadori. New

York: F.W. Dodge Corp., 1956.
BIBLIOGRAPHY 273
[21] Ridder, H.A.J. de. Design & Construct of Complex Civil Engineer-
ing Systems: A new approach to organization and contracts. Delft:
Delft University Press, 1994.
[22] Simon, H.A. The Sciences of the Artificial - 3rd Edition. Cambridge,
MA: The MIT Press, 1996.
[23] Spekkink, D. Castle or House of Cards?: Strengthening the struc-
tural safety chain. The Hague: Ministry of Housing, Spatial Plan-
ning and the Environment, 2009.
[24] Spekkink, D. Compendium Aanpak Constructieve Veiligheid. Gouda:
Betonvereniging, 2011.
[25] Terwel, K.C. Structural Safety: Study into critical factors in the
design and construction process. Rotterdam, 2014.
[26] Timoshenko, S.P., J.M. Gere. Theory of Elastic Stability. New York:
Dover Publications Inc., 2009.
[27] Torroja, E. Philosophy of structures. Transl. J.J. Polivka, M.
Polivka. Los Angeles: University of California Press, 1958.
[28] Vinci, L. Da. The Notebooks of Leonardo Da Vinci, Volumes I and
II. Transl. J.P. Richter. New York: Dover Publications Inc., 1970
(1883).
[29] Vitruvius. Vitruvius Handboek bouwkunde. Transl. T. Peters. Ams-
terdam: Athenaeum, 2003.
[30] World Bank. World Development Report 1994: Infrastructure for
Development. New York: Oxford University Press, 1994.
274 BIBLIOGRAPHY
Curriculum Vitae
Michiel Paul Horikx was born in 1956, in The Hague, the Netherlands.
He attended the Lyceum Augustinianum in Eindhoven and completed
his secondary education in 1976. Subsequently, he first studied Archi-
tectural, and later Structural Engineering at the Eindhoven University
of Technology. He completed his master’s thesis in 1983.
After completing his military service he worked with the Hollandsche
Beton Groep, at that time the largest civil engineering contractor in the
Netherlands. As a structural designer, he was involved in large scale
projects, including offshore and bridge design. From 1988 up to 1992 he
held the position of conceptual designer and engineering manager of the
steel structures - retaining wall, trusses, and ball joint - of the Maeslant
Storm Surge Barrier.
Since 1992 he has worked as a senior lecturer and manager at the Am-
sterdam University of Applied Sciences and has been responsible for
the design, implementation and management of the following successful
higher education programmes: Bachelor in Civil Engineering; Bachelor
in Structural Engineering; and Master in Structural Engineering.
In 2019 he has been appointed as professor of the national Lectorate
Structural Safety, commissioned by the national concrete association
BV, the national steel association BmS, and the Amsterdam University
of Applied Sciences, including their corresponding professional master’s
programmes in structural design.

EBook Horikx

Uploaded by

Copyright:

Available Formats

EBook Horikx

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

EBook Horikx

Uploaded by

Copyright:

Available Formats

Conceptual structural design

understanding structural performance

Citation for published version (APA):

Download date:29 Apr 2024

Systematic thinking Systems thinking Modelling

Michiel Paul Horikx

Michiel Paul Horikx

This textbook was supported by the Faculty of Technology at the Ams-

Copyright c 2022 by Dr.ir. M.P. Horikx

All rights reserved. No part of this book may be reproduced or transmit-

A digital version of this book is available at http://hbo-kennisbank.nl

The subject of this textbook is a methodical approach on the complex

Problem A lack of insight of the professional into structural engineer-

For a better understanding and an eﬀective solution approach this lack

Solution approach The methodical approach on conceptual struc-

Structural design cycle Independent of life cycle phase, complexity

Basic structural forms One of the main conceptual structural design

ceptual structural design parameters is conditional. On a two-

Training The required knowledge, skills, and professional attitude have

Epilogue It is important to study how the design is organised in prac-

design is valuable and feasible. Recommendations for research are given

I Problem, analysis and solution approach 7

3 Present-day problems in structural engineering 13

3.3 Historical perspective . . . . . . . . . . . . . . . . . . . . . 20

5.4.2 Progressive insight from estimation to accuracy . . 46

II The art of conceptual structural design 53

8.2 Level of accuracy . . . . . . . . . . . . . . . . . . . . . . . 82

10 Cyclic process control 97

10.5 Cyclic optimum design . . . . . . . . . . . . . . . . . . . . 108

III Understanding structural performance 115

11 Introduction to part III 117

12 Structural performance 121

13 Structural integrity 131

13.3.3 Material properties of conceptual design . . . . . . 136

14 Fundamental parameters of load distribution 141

15 Fundamental parameters of failure mechanisms 163

15.4.2 Conceptual design of concrete in compression . . . 178

IV Case study and training 185

16 Introduction to part IV 187

17 Case study trusses Maeslant storm surge barrier 189

17.5.5 Section dimensions of conceptual design . . . . . . 207

18 Training programme 215

19 3-D geometry combined with simple modelling 225

19.2.2 Simplified 2-D modelling ULS . . . . . . . . . . . . 234

20 2-D geometry combined with medium modelling 239

21 1-D geometry combined with abstract modelling 253

21.2.3 Load distribution truss analogy . . . . . . . . . . . 260

Curriculum Vitae 275

1.1 Conceptual structural design . . . . . . . . . . . . . . . . 3

2.1 Reading guide for part I . . . . . . . . . . . . . . . . . . . 10

3.1 Structural safety Eurocode . . . . . . . . . . . . . . . . . . 15

4.1 Structural design as an interface control approach . . . . . 26

5.1 T-shaped professional structural engineer . . . . . . . . . 43

6.1 Reading guide for part II . . . . . . . . . . . . . . . . . . . 56

7.1 Conceptual structural design process . . . . . . . . . . . . 60

8.1 Structural design process . . . . . . . . . . . . . . . . . . . 80

9.1 Assembly of an arch bridge . . . . . . . . . . . . . . . . . 92

10.1 Design strategies . . . . . . . . . . . . . . . . . . . . . . . 99

11.1 Reading guide for part III . . . . . . . . . . . . . . . . . . 118