See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/274640991

Prefabricated steel structures for low-rise buildings in seismic areas

(Precasteel)

Book · January 2013

DOI: 10.2777/5499

CITATIONS READS

6 9,341

22 authors, including:

Aurelio Braconi Andrea Dall'Asta

EUROFER The European Steel Association University of Camerino
22 PUBLICATIONS   339 CITATIONS    184 PUBLICATIONS   2,624 CITATIONS   

SEE PROFILE SEE PROFILE

Graziano Leoni Benno Hoffmeister

University of Camerino RWTH Aachen University
164 PUBLICATIONS   1,686 CITATIONS    104 PUBLICATIONS   507 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Anti-seismic devices: influence of failure and ultimate capacity on the system reliability View project

Material Choice for Seismic Resistant Structures (MATCH) View project

All content following this page was uploaded by George E. Varelis on 08 April 2015.

The user has requested enhancement of the downloaded file.

 KI-NA-25871-EN-N

EU
Low-rise buildings are used for industrial and commercial activities. Most of
these buildings are built by prefabricated-concrete elements characterised by

Prefabricated steel
low efficiency of connections and a lower ductility performance when subjected
to earthquake loading.

The proposal aim is defining prefabricated steel solutions for single-storey and
low-rise buildings in earthquake-prone areas for industrial and commercial

structures for low-rise

activities. The selection of structural solutions is performed via statistical
analyses in different European countries about solutions requested by the
market. Industrial and commercial solutions are optimised in terms of structural
performance and of construction costs and two of those (technologically

Prefabricated steel structures for low-rise buildings in seismic areas (Precasteel)

buildings in seismic areas
advance and high-performing) are experimentally tested. The solution
selected for commercial activities is a dissipative device working in series with
prefabricated RC-Wall and pendulum steel-structure. The solution selected for
industrial activities is a warehousing system using cold-formed profiles acting
compositely with ribbed steel sheeting as girder web.

Automated integrated design software with cost-effectiveness analysis module

is realised to favour the use of proposed solutions into practice. Solutions
(Precasteel)
proposed by the software are optimised to meet three ‘criteria’:

(1) Prefabrication — structural members and connection types selected in order

to reduce on-site operations;

(2) 
Pre-designed — structural solutions pre-designed and inserted in a
database correlating member/structure with its maximum structural
performance; structural configuration for industrial/commercial buildings are
assembled according to the design inputs;

(3) Standardised — members and structural systems have static schemes

suitable for defining simplified designing procedures (transformed in
technical tables — Precasteel Software): database (catalogue) of solutions/
members may be enlarged enclosing more recent developments.

Studies and reports

EUR 25871

doi:10.2777/5499 Research and

Innovation EUR 25871 EN
 HOW TO OBTAIN EU PUBLICATIONS
Free publications:
• via EU Bookshop (http://bookshop.europa.eu);
• at the European Union’s representations or delegations. You can obtain their
contact details on the Internet (http://ec.europa.eu) or by sending a fax
to +352 2929-42758.
Priced publications:
• via EU Bookshop (http://bookshop.europa.eu).
Priced subscriptions (e.g. annual series of the Official Journal of the
European Union and reports of cases before the Court of Justice
of the European Union):
• via one of the sales agents of the Publications Office of the European Union
(http://publications.europa.eu/others/agents/index_en.htm).

EUROPEAN COMMISSION
Directorate-General for Research and Innovation
Directorate G — Industrial Technologies
Unit G.5 — Research Fund for Coal and Steel

E-mail: rtd-steel-coal@ec.europa.eu
RTD-PUBLICATIONS@ec.europa.eu

Contact: RFCS Publications

European Commission
B-1049 Brussels
 European Commission

Research Fund for Coal and Steel

Prefabricated steel structures for
low-rise buildings in seismic areas
(Precasteel)
A. Braconi and A. Osta
Ilva SpA
Viale Certosa 249, 20151 Milano, ITALY

A. Dall’Asta and G. Leoni

Universitò di Camerino
Via del Bastione 1, 62032 Camerino, ITALY

S. Möller and B. Hoffmeister

Rheinisch-Westfälische Technische Hochschule Aachen
Templergraben 55, 52056 Aachen, GERMANY

S. A. Karamanos and G. Varelis

University of Thessaly
Argonauton & Filellinon, 38221 Volos, GREECE

E. Alderighi, C. Coscetti and W. Salvatore

Università di Pisa
Lungarno Pacinotti 43, 56100 Pisa, ITALY

J. Gracia and E. Bayo

University of Navarra
Campus Universitario, 31080 Pamplona, SPAIN

R. Mallardo, L. Bianco and P. Filipuzzi

Ferriere Nord SpA (Gruppo Pittini)
Via Zona Industriale Fraz., 33010 Rivoli, ITALY

D. Vasilikis and P. Tsintzos

Shelter SA
CHLM Larisas Sykourious 6, 41500 Larisa, GREECE

S. Estanislau and J. Lobo

Instituto de Soldadura e Qualidade, Doutor Cavaco Silva
33, Parque das tecnologias, 2740 120 Porto Salvo, PORTUGAL

L. Fulop and P. Hradil

Technical Research Centre of Finland
Vourimiehentie 3, 02044 Espoo, FINLAND

Grant Agreement RFSR-CT-2007-00038

1 July 2007 to 30 June 2010

Final report
Directorate-General for Research and Innovation

2013 EUR 25871 EN

LEGAL NOTICE
Neither the European Commission nor any person acting on behalf of the Commission is
responsible for the use which might be made of the following information.
The views expressed in this publication are the sole responsibility of the authors and do not
necessarily reflect the views of the European Commission.

Europe Direct is a service to help you find answers

to your questions about the European Union

Freephone number (*):

00 800 6 7 8 9 10 11
(*)  Certain mobile telephone operators do not allow access to 00 800 numbers or these calls may be billed.

More information on the European Union is available on the Internet (http://europa.eu).

Cataloguing data can be found at the end of this publication.

Luxembourg: Publications Office of the European Union, 2013

ISBN 978-92-79-29011-4
doi:10.2777/5499

© European Union, 2013

Reproduction is authorised provided the source is acknowledged.

Printed in Luxembourg

Printed on white chlorine-free paper

Table of contents
Table of contents 3
Final summary 7
1. Statistical analysis and costs definition 17
1.1. Industrial buildings 17
1.1.1. Contribution of partners 17
1.1.2. Morphology and correlated activities 17
1.1.2.1. Main activities performed in the industrial building 17
1.1.2.2. Type of structural system 17
1.1.3. Basic geometric parameters of steel framing 18
1.1.3.1. Number of frame bays (in-plane direction) 18
1.1.3.2. Number of frames (out-of-plane direction) 19
1.1.3.3. Span of each bay 19
1.1.3.4. Distance between consecutive frames 20
1.1.3.5. Height of frame columns 20
1.1.4. Type of girder 20
1.1.4.1. Slope of frame girder 21
1.1.5. Parameters of vertical and crane loading 21
1.1.5.1. Snow loading 21
1.1.5.2. Roof utility loading 22
1.1.5.3. Overhead travelling crane lifting capacity 22
1.1.6. Parameters for horizontal loading 23
1.1.6.1. Type of in-plane frame resistance 23
1.1.6.2. Type of resistance out-of-plane 23
1.1.6.3. Maximum ground acceleration 24
1.1.6.4. Wind load intensity 24
1.1.7. Secondary elements of the steel structure 25
1.1.7.1. Type of roofing 25
1.1.7.2. Type of side cladding 25
1.1.7.3. Type of purlins 26
1.2. Commercial buildings 26
1.2.1. Data collection 28
1.2.2. Morphological and structural data 28
1.2.3. Loading actions 32
1.3. Analysis on potential emerging markets 32
1.3.1. Country specific statistics 33
1.3.2. Country specific study 33
1.3.3. Morphological and structural data 35
1.4. Cost Analysis for designed case studies 37
2. Industrial buildings 41
2.1. General description of selected structural types 41
2.1.1. Structural conceiving and design assumptions 41
2.1.1.1. Design Assumptions 42
2.2. Set of design structures 45
2.2.1. Light gauge steel solutions 47
2.2.1.1. Analysis Results 48
2.2.2. Tapered solutions 48
2.3. Selection of case studies 49
2.4. Purlin design 50
2.5. Optimization of the double-span cases design 50
2.6. Optimization of single-span cases design 54
2.6.1. Structural conceiving and design assumptions for optimization 54
2.6.1.1. General method analysis (GMA) 55
2.6.1.2. Method using EC3 interaction formulae (IFM) 55
2.6.2. Set of designed structures (Members and connections) WT frames 55

3
2.6.2.1. Optimization of designed solutions 56
2.6.2.2. Performance assessment of optimized solutions 56
2.7. Effects of the purlin support on the stability of frames 57
2.8. Analysis of connection systems 57
2.8.1. Improved connections for WT IBs 57
2.8.2. LGS connections design 59
2.9. Improved IB solution: Corrugated web girder 59
2.9.1. Description of the components used 62
2.9.2. Connection to frame column 63
3. Commercial buildings 65
3.1. Structural conceiving and design assumption 65
3.1.1. Main members 65
3.1.1.1. Flooring systems 65
3.1.1.2. Columns 66
3.1.1.3. Braces 66
3.1.1.4. R.c. shear walls 67
3.1.1.5. Dissipative devices 67
3.1.1.6. HDR-based devices 68
3.1.1.7. Flag Shaped Hysteretic Devices (FSHD) 68
3.2. Set of designed structures (Definition of data-base of cross-sections) 69
3.2.1. Flooring systems 70
3.2.2. Column elements 70
3.2.3. Bracing systems 70
3.3. Optimization of structural modulus 70
3.3.1. Flooring systems with composite beams 70
3.3.2. Flooring systems with truss girders 72
3.3.3. Columns 72
3.3.4. Braces 73
3.3.4.1. Concentric braces 73
3.3.4.2. Eccentric braces with two diagonal members 74
3.3.4.3. Eccentric braces with one diagonal member 74
3.3.4.4. Distribution of braces 75
3.3.5. Application of cross-sections database to a defined set of case studies 76
3.3.5.1 Data-base main issues 77
3.3.5.2 Geometry layout of defined frames 78
3.3.5.2.1 Case Type 1 (CS1) – Geometry layout 78
3.3.5.2.2 Case Type 2 (CS2) – Geometry layout 79
3.3.5.2.3 Case Type 3 (CS3) – Geometry layout 79
3.3.5.3. Response spectrum analysis of the analysed set of frames 80
3.3.5.4. Incremental Dynamic Analysis 83
3.3.5.4.1. Planar frame models 83
3.3.5.4.2. The adopted artificial accelerograms 84
3.3.5.4.3 Obtained capacity curves and shear link behaviour 84
3.3.6. Prefabricated walls 85
3.3.6.1. Calculation hypothesis 85
3.3.6.2. Load hypothesis 85
3.3.6.3. Geometrical hypothesis: 86
3.3.6.4. R.C. Walls as alternative bracing systems 88
4. Experimental testing 91
4.1. Experimental assessment of light gauge steel girders for industrial buildings 91
4.1.1. 4-point-bending-tests 91
4.1.2. Results of 4-point-bending-tests 93
4.1.2.1. Comments on experimental evidences 96
4.1.3. Testing of girder-to-column connections 96
4.1.4. Results of tests 97

4
4.2. Experimental characterization of a novel hysteretic dissipative device 98
4.2.1. Prototype description 98
4.2.2. Test setup 101
4.2.3. Testing procedure 102
4.2.4. Results 102
5. Software development 105
5.1. Precasteel Software scope 105
5.1.1. Workflow of the software 105
5.2. Industrial building 105
5.3. Commercial building 107
5.4. Technical issues concerning the software 108
5.4.1. Server side technologies 108
5.4.2. Client side technologies 108
5.4.3. Web application resources structure 108
5.4.4. Statistical case analysis 109
5.4.5. Screen captures of the application 111
5.5. Software for IB optimization 117
6. Design guidelines and final design of applicative examples 129
6.1. Design Guidelines for the IB modulus 129
6.1.1. Cross-sections used 129
6.1.2. Welded-tapered 130
6.1.3. Trussed solutions 130
6.1.4. Welded tapered solutions 131
6.1.5. Type and intensity of actions 131
6.2. Corrugated web girder: Design guidance 132
6.2.1. Design of the web 132
6.2.2. Design of the flanges 133
6.2.3. Design of connections web-flange 133
6.2.4. Calculation of deflection 134
6.2.4.1. Basic principles 134
6.2.4.2. Appliance to CWG 135
6.3. Design guidance of precast r.c. walls in CBs 136
6.3.1. Study case 1 – Dissipative r.c. walls 136
6.3.1.1. Connection decoupling horizontal and vertical loads 138
6.3.1.2. Connection for both horizontal and vertical loads 138
6.3.2. Study case n°2 - Dissipative r.c. walls 139
6.3.3. Study case N°1 (Dissipative devices + Elastic r.c. walls) 140
6.4. General guidelines for commercial buildings 141
6.4.1. Gravity structure 142
6.4.2. Lateral resisting elements 145
6.4.3. HDR Based Dissipative devices: guidelines and examples 147
6.4.4. FSHD-based dissipative devices – guidelines and examples 149
6.4.4.1. Description of the structural system and adopted design issues 149
6.4.4.2 Description of case studies 150
6.4.4.3. Design of frames equipped with FSHD: preliminary considerations 151
6.4.4.3.1 Definition of performance objectives 151
6.4.4.4. The proposed four-steps design procedure 152
6.4.4.4.1 Determination of FSHD monotonic curves 153
6.4.4.4.2 Determination of the capacity curve 154
6.4.4.4.3 Determination of the structural performance point (PP) 154
6.4.4.4.4 Definition of FSHD cyclic curves 156
6.4.4.5. CS1H5 case study: application of FSHD design procedure and seismic performance evaluation 157
6.4.4.5.1 Definition of FSHD monotonic curves 157
6.4.4.5.2 Determination of the equivalent SDOF pushover curve 158
6.4.4.5.3 Determination of the structural performance point (PP) 159

5
6.4.4.5.4 Definition of FSHD cyclic curves 160
6.4.4.5.5. Evaluation of seismic performance: Incremental Dynamic Analysis (IDA) 160
7. Results, conclusions and perspectives 165
References 167
Acronyms and Abbreviations 169
List of figures 171
List of tables 177

6
Final Summary
A correct design approach should consider the safety against actions that the structure may be subjected to
during its lifecycle (like wind, snow, live loads, earthquake and fire) and the consequences on the design
process resulting from functional, maintenance, reparability and cost-efficiency requirements, strongly
depending on the typology of building and of activities carried on inside the building itself.
Therefore, standardization and integrated performance based design represent the key to define modern
constructions truly competitive and attractive with respect to the traditional and common solutions actually
adopted in design practice. Actual seismic codes as Eurocode 8, AISC2005, NTC2008 or VISION2000, in
fact, consider design procedure organized according to a multi-level approach, called Performance Based
Design where the level of actions is related to expected structural performance and to a maximum tolerable
damage level.
This new design philosophy has introduced the integration of economic aspects in the structural design, in
terms of construction costs and economic losses due to damage levels associated to fixed action levels (e.g.
peak ground acceleration of seismic input). In such a context, PBD represents a design approach suitable for
structural types as industrial facilities and low-rise commercial buildings, where the economic losses and the
interruption of productive activities are key requirements under low-medium intensity earthquake in the
same way as life protection under high intensity earthquake. Interruption of productive cycle, damaging of
mechanical components, of (hazardous) chemical substance, transportation systems (e.g. pipes) and of
overhead travelling crane could cause situation of high risk for workers inside the industrial buildings, not
directly correlated to collapse of the structural component, but correlated to a not completely satisfactory
behavior of the structure during the earthquake.
Unfortunately, in European seismic prone areas mainly located in Mediterranean countries, more adopted
structural solutions for the construction of industrial/commercial buildings are prefabricated reinforced
concrete structures which often exhibit severe weakness strongly related to their production process and the
constituent materials. In fact, prefabricated concrete elements, beams and columns, are connected using
corbel systems that realize simply supported conditions for the beams, figure I.

(a) (b)

(d)
(c)
Figure I. (a) typical beam-column connection with corbel and gutter beam; (b) collapsed precast r.c. frames
by earthquake – Turkey 1990; (c) typical Precast r.c. solution for low-rise commercial building in Italy; (d)
typical configuration of one-storey r.c. industrial building
General objectives
The research project was so focused on the definition of prefabricated steel or steel-concrete composite
solutions for realizing single-storey industrial and low-rise commercial buildings in earthquake prone areas
characterized by the following aspects:
‚ geometries and configurations suitable for use category and foreseen operations;
7
 ‚ improved structural performance;
‚ construction costs;
‚ availability of practical support for structural member sizing;
‚ availability of research results on web application.
The structural solutions defined during the research were conceived in order to represent an effective
alternative to reinforced concrete solutions, coupling structural efficiency and costs control of the
construction. On the other hand, the success of structural solutions was connected to its easy and rapid
employment in the day-to-day design practice, and for this reason a user-friendly software for supporting the
designers in the use of PRECASTEEL structural solutions into the practice was created. Finally, the software
application was defined as web application allowing a on-demand structural sizing of selected solutions.

Research plan and work carried out

The research was organized according to the following working phases, presented more extensively in the
figure II:
‚ Phase 1 - analysis of structural and functional requirements for commercial and industrial buildings
and to the definition of steel and steel-concrete most efficient structural schemes on the basis on an
accurate performance analysis;
‚ Phase 2 - numerical and experimental analyses for the assessment and improvement of structural
performance of selected solutions;
‚ Phase 3 - execution of final design of the solutions with improved performance and arrangement of
the simplified design procedures for chosen structural types.
The phase 1 was focused on the selection of most adopted solutions for one-storey industrial buildings and
for low-rise commercial buildings: this selection was organized in two sub-phases. The first was an accurate
statistical investigation about more diffused structural solutions and geometries across European countries
for industrial and commercial activities, see figures III and IV as examples. The collected data came from
Portugal, Spain, Germany, Italy, Greece, Finland and eastern EU countries as Poland, Hungary and Romania.

Figure II. General flow-chart of the research project

The statistical investigation explored also market opportunities for steel solutions, as those developed inside
PRECASTEEL, in new EU countries as Romania, Hungary or Czech Republic.
On the basis of statistical analysis of all collected data, structural configurations were defined fixing
geometries (bays length, storey number, floor configuration or roof slope) in order to be consistent with
housed activities, industrial or commercial, and to be competitive with concrete market shares. The selected
structural solutions for commercial (CB) and for industrial (IB) activities, figure V, were iteratively designed
varying geometrical parameters and resisting static schemes in order to define steel and steel-concrete
composite solutions suitable to be competitive with concrete solutions.

8
 Spain
Prevalent R.C. and P.R.C. Balanced Prev.S and C
1.00 Romania

Portugal
0.80

Poland
0.60 1
R.C.+P.R.C. Italy 2
Steel + Comp.
0.40 Hungary 3
>3
Greece
0.20

Germany
0.00
Finland
y

ia
y

n
nd
e

d
ly

ga
an

ar

ai
ec

an
an
Ita

la

Sp
ng

rtu
m

re

nl
m

Po
er

Fi
Hu

Po
G

Czech Rep.
Ro
G

(a) 0% 20% 40% 60% 80% 100%

(b)
Figure III. (a) most adopted materials for CB solutions; (b) number of storey for structures employed for
commercial activities

Romania
Romania
Poland
Poland
Hungary
Hungary
Steel
Finland Concrete
Finland 1
Timber
2 Spain
Spain Other
>=3
Portugal
Portugal
Italy
Greece
Greece
Germany
Germany

0% 20% 40% 60% 80% 100%

0% 20% 40% 60% 80% 100%
(a) (b)
Figure IV. (a) Number of bays in industrial buildings; (b) type of structural solutions

gravity structure seismic gravity structure

resistant
structure
(a)
(b)
Figure V. (a) Morphological general scheme for IB solutions; (b) Morphological general scheme for CB
solutions.

IB and CB solutions were designed adopting the complete design framework proposed by Eurocodes and
structural profiles and main members were sized in order to optimize also construction costs; the cost
associated to the construction of designed solutions was defined through the definition of a cost model where
all elements contributing to the total cost were referred to a unique parameter: the steel consumption. In such
a way, the iterative design of many structures integrated with the cost analysis was transformed in a complete
performance analysis where structural performance (assessed applying Eurocode design framework) were
harmonized with construction costs, table I and table II, and housing of selected activities based on the
statistical analysis results. The cost model considered information coming from three different countries
(Italy – Southern Europe; Germany – Central Europe; Romania – Eastern Europe) in such a way to adopt

9
standardized reference values and to individuate in which national markets some solutions can be
competitive or not.

Transportation (from
Fabrication Costs Surface treatment Erection
Type of the shop to the field)
€/kg €/kg
section €/kg [each 100 km]

A
Steel production
Concrete Sandblast SA2.5 B
(including drawing
Steel [S355] €/kg [C25/30] primer 40 Pm Hot-deep 100 km
production) [S355]
€/m3 thickness + wet galvanizing
€/kg
coating
HE 0.91 0.40 -- 0.14 0.27 0.011 0.40
CHS 1.45 0.63 -- 0.16 0.40 0.011 0.41
PEHE 0.91 0.40 380.00 0.14 0.27 0.011 0.40
CFCHS 1.45 0.63 150.00 0.16 0.40 0.011 0.41
Cold formed 0.91 0.40 -- 0.14 0.27 0.011 0.40
Table I. Costs of columns for CB in Southern Europe (these costs include a lump-rate estimation of steel
working, bolts and general expenditure)

Transportation (from
Type of Fabrication Costs Surface treatment Erection
the shop to the field)
section €/kg €/kg
€/kg [each 100 km]
Steel production A
B
(including drawing Sandblast SA2.5
Steel [S355] €/kg Hot-deep 100 km
production) [S355] primer 40 Pm
galvanizing
€/kg thickness
IPE/HE 1.32 0.40 0.16 0.29 0.011 0.40
HR trusses 0.93 0.63 0.23 0.34 0.011 0.44
CF trusses 0.93 0.63 0.23 0.34 0.011 0.44
Table II. Costs of floor systems for CB in Southern Europe (these costs include a lump-rate estimation of
steel working, bolts and general expenditure)

After the conclusion of the first phase of the research, partners had at their disposal a huge number of
structural solutions partially optimized in terms of building performance (structural, economic and
utilization). According to the original research plan, structure defined through performance analyses were
deeply analyzed in order to assess their real performance and to improve economic aspects (i.e. reducing
steel consumption) or/and to improve structural performance adopting complete Performance Based
Approach. This complex task was developed through numerical and experimental investigations.
IB solutions were numerically analyzed in order to minimize steel consumption guaranteeing satisfactory
structural requirement: non-linear numerical simulations were applied to individuate and eliminate
potentially weak failure mechanism, to optimize the member size for reducing the steel consumption and to
design the secondary members of the IB solutions (purlins, walls or local braces). More in the detail, the
optimization of IB solutions was executed developing an appropriate numerical procedure that using genetic
algorithms is able to optimize profile size in order reach the minimum steel consumption adopting hot-rolled
elements from catalogue or welded elements from plates.
At the end of this accurate re-design of all IB solutions, the following solution types were made available for
their employment in seismic areas: hot-rolled profiles, welded profiles, cold formed profiles, see figure VI.a,
and cold-formed elements cooperating with corrugated steel sheeting, see figure VI.b. This last solution, in
particular, was deeply investigated also through experimental testing because of its potential application in
industrial warehousing, of its lightness and its internal resisting schemes, see figure VI.b.

10
 (a)

(b)
Figure VI. (a) IB solutions realized with light gauge steel members; (b) part of IB solution with corrugated
steel sheeting tested during the experimental programme.

Concerning the CBs, the solutions derived from the statistical analysis and from the cost analysis were in
general a modular solution in which two main structures were coupled and devoted to different roles: (a) a
gravity structure to sustain vertical loads; (b) a bracing structure to sustain seismic loads, see figure V.b. This
choice allowed the possibility of choosing different member types for each structural component according
to the following list:
‚ beam elements – hot rolled profiles or cold formed profiles compositely acting with floor/roof slabs;
trussed girder using cold– or hot–formed elements;
‚ column elements – bare steel HE profiles; partially encase composite columns; concrete filled
tubular columns;
‚ bracing elements – steel eccentric braces, figure VII.a; prefabricated reinforced concrete walls
(PRCW), figure VII.b; PRCW with additional dissipative devices: Flag Shaped Hysteretic Devices –
FSHD – in which seismic energy is dissipated through steel plastic cycles; High Dissipative Rubber
Bearings – HDRB – in which seismic energy is dissipated through viscous damping of the rubber.
Such structural solutions were characterized by the absence of interaction between gravitational loads and
horizontal (seismic) loads, allowing during the performance analysis also a full structural optimization; in
fact, differently from IB solutions, the columns and beams elements defined at the end of performance
analysis were already optimized, already being minimum steel weight members for a fixed gravitational
loading level.
So, the refined numerical simulations and the studies for optimizing the structural performance of CB
solutions where devoted to the assessment of real seismic performance of different horizontal bracing
systems. The following structural configurations were analyzed:
‚ CB equipped with Eccentrically Braced steel Frames;
‚ CB equipped with prefabricated reinforced concrete walls (PRCW);
‚ CB equipped with innovative hysteretic devices (i.e. Flag Shaped Hysteretic Devices – FSHD)
between gravity structure and PRCWs;
‚ CB equipped with High Dissipative Rubber Bearings (HDRB) between gravity structure and
PRCWs
In particular, the applicability of FSHD system to selected CB configurations was evaluated through the
execution of numerical simulations carried out using incremental dynamic analysis (IDA) method and
through an experimental programme carried out on a FSHD prototype, figure VIII, suitably designed and
assembled for this research project.
It is worth underlining that the development of this innovative dissipative system was carried out in
cooperation with another RFCS research project because conceptual development of the system, feasibility
analyses, fabrication of the prototype and its mechanical characterization were not considered in the original
experimental programme of the project. Anyway, in order to increase the quality and the innovation of the
project, some partners decided to dedicate additional extra work and extra time for studying and testing
FSHD applied to PRECASTEEL CB solutions.
11
Complementary to FSHD, the feasibility and the efficiency of HDRB as protection system against
earthquakes were assessed through numerical studies: the HDRB, see figure IX, was considered as fuse
element between gravity structure and PRCWs.

theoretical link length e

effective link length e’

(a) (b)
Figure VII. (a) eccentric bracing system: shear link; (b) prefabricated reinforced concrete wall

rK 0 E Fy
Fy

K0

'
A

(b)

(a)
Figure VIII. (a) prototype of FSHD system; (b) schematic mechanical behavior of FSHD

100

80

60

40
force (kN)

20

0
-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5
-20

-40

-60

-80

-100
strain

(a) (b)
Figure IX. (a) HDRB system; (b) stable loops of HDRB at different maximum strains.

The final part of the research, according to the general research plan, was focused on the definition of a
design method (procedure) suitable for supporting practitioners and engineers in the use of solutions studied
in PRECASTEEL. Also in this case, respect to the original programme, one partner dedicated extra work and
extra time in order to create an innovative and high quality software for the valorization and utilization of
structural solutions herein considered.

12
The automated software presented in the original proposal became an internet application that every
practitioner and engineer can use from its office in order to preliminary design a complete structure for
industrial and commercial activities using PRECASTEEL solutions. The web application works as a user-
friendly graphical interface, see figure X, that starting from a finite set of parameterized and optimized set of
data, supplies the drawings and estimation of the cost of the structure, floors, roof and connections for
industrial and commercial buildings.
It is important to underline that web application is also able to furnish back to the user (that interacts with the
software at the web-site http://riv-precasteel.rivagroup.com/precasteel/) general drawings of the final
solutions in a *.dxf format (to be directly copied in Autocad® drawing), a short summary about main
characteristic of the structure and its complete economic plan, presented in figure X.c; the drawings
furnished back by the software are front view, side view, plan view and connection types, see figure XI,
general drawings that summarize all relevant technical features.
The statement “integrated tool” means that the program provides drawings encompassing: main structure,
secondary elements, roofing, floors, connections and cost estimation. This is the novelty of the programming
part of this research project, and is in fact an achievement in itself. The software produces the drawings of
the main and secondary structural elements as well as the general drawings, see figure XI, and templates of
the roofing system, floor system, secondary elements, partition and external walls, beam to column
connections and column bases.
The software application relies on a large database from which all the possible building solutions are selected
and retrieved. This database contains all the commercial and industrial cases that were designed, analyzed
and optimized in this research project. It is always possible to implement new structural solutions of or other
structural configuration thank you to the internal architecture of the software that works using simple but
complete information database.

(a)
(c)
(b)
Figure X. Images of the web application containing PRECASTEEL deliverables ready-to-use: (a) selection
of roofing and cladding system for IB; (b) selection of columns type in CB adequate for the vertical loading
level; (c) complete cost analysis of the final structural modulus.

Moreover, the numerical simulations, carried out on IB solutions using genetic algorithms, produced an
optimization tool (e.g. excel sheet – name “EV-Frame” or “EV-tool”) that can be freely distributed; this was
an additional PRECASTEEL deliverable for supporting practitioners and engineers in the application of
chosen steel solutions for IBs. In particular, the optimization procedure is really user-friendly through several
input windows, figures XII.a, , XII.b and XIIca, in which geometry, loads and optimization parameters must
be inserted. At the end of the optimization process, the EV-tool programme furnishes an output window,

13
figure XII.d, where the optimized section of profiles, the total cost of the structure and
sustainability/environmental information are given. The sustainability and environmental information are the
energy consumption during the exercise (working) period of the structural solution and the CO 2 emission
associated to the steel quantity employed for realizing steel profiles.
At the conclusion of the phase 3 of the research project, all partners cooperated in the final design of some
case studies preliminary designed adopting the software packages. In particular, at the end of the research,
after having fixed the structural configurations, having designed and optimized the correspondent solutions
and after having defined the software for the practical use of these solutions, some final case studies were
sized and completely designed as applicative examples. Many IB were designed fixing different structural
solutions: hot-rolled profiles, cold-formed profiles, solutions with haunches and solutions employing web
corrugated sheets, which structural performance were assessed through experimental testing. Moreover, CB
were designed also considering different type of bracing systems: bare steel eccentric braces, prefabricated
walls, walls coupled with additional dissipative devices.
The development of these case studies allowed also the definition of simplified guidelines for supporting the
designers in the adoption of PRECASTEEL solutions for geometries or loading levels outside the variation
ranges explored during the research project.

(b)

(c)
(a)

(d)
Figure XI. Example of IB solution produced by web application – DXF drawing elements: (a) plan view; (b)
front view; (c) side view; (d) connections: apex and eave.

14
 (a) (b)

(c) (d)
Figure XII. EV-tool: (a) Input window - fixing general frame geometry; (b) Input window – introduction of
loading information; (c) Input window – parameters for the optimization procedure of the frame; (d)
summary of optimization process results: steel sections; cost of the structure; sustainability and
environmental impact.

Results, conclusions and perspectives

The results of the research were more relevant and technologically advanced respect those declared in the
original research proposal. In fact, the automated software is now a web application completely free for all
practitioners and engineers, able to furnish a complete cost plan of the structure and general technical
drawings. Additionally, the testing programme firstly addressed to simple connections or details became a
more complex experimental study devoted to the structural assessment of a new and optimized structural
solution to be employed for IB and CB.
Moreover, all technological advanced or high performing solutions as CB with additional dissipative devices
for seismic protection or IB realized with web corrugated sheeting with improved connection details were
equipped also with design guidelines. The suitable combination of all these aspects with the original working
plan, see figure XIII, could have an high impact due to: utilization simplicity of the software, high quality of
information at disposal of the user (drawings, technical info, cost evaluation), suitable choice of geometries
and solutions requested by the market, open architecture of the PRECASTEEL software that can be
continuously updated with new technical solutions. Employment of light gauge steel members in seismic
areas and the application of dissipative devices based on different materials coupled with steel structures and
prefabricated are relevant technical aspects that, beyond PRECASTEEL conclusions, could be further
developed and extended in future researches. For this purpose, ILVA S.p.A. is actually defining a dedicate
web-site on which insert all complete products (deliverables) for a future dissemination action.
The software will be made available to everyone is interested in the exploitation of the PRECASTEEL
results: architects, designers, technicians; in particular, the web interface and the software capabilities have
been designed in order to facilitate as much as possible the selection of structural solutions related to design
inputs.
Moreover, the software has been installed on the RIVA web-site with the complete suite of tables, database
and sub-programme in order to enlarge and enrich, in the future, the internal catalogue of structural solutions.
Additionally, the software can integrate other structural solutions respect to typologies studied inside
PRECASTEEL: it is sufficient to introduce new design principles associated to new structural solutions in a
matrix/tabular form as did for already integrated solutions, relating in a discrete way the size and the
characteristics of elements appropriate for a structural solution with the design inputs as, for example, the
loads.

15
Figure XIII. General organization of the project

16
1. Statistical analysis and costs definition
1.1 Industrial buildings
Industrial Buildings (IBs) include a wide variety of typologies. Depending on their dimension, location,
organization and facilities, they can house one activity or more activities ranging from the simple storage of
light material to the production of heavy materials as steel. Given that each activity need an appropriate
working environment and need a fixed share of free space for machinery movements of for appropriately
storing goods and materials.
In order to individuate data characterizing necessary working condition or space requirements for housing
industrial activities. Towards this purpose, a Census Form has been developed and submitted including
significant issues concerning industrial buildings in order to collect data from different countries. These
Census Forms (CFs) were filled by partners with general information regarding the most common
characteristic of Industrial Buildings in every country as well as with information concerning an individual
building case. These data were provided to the involved partners by Local Councils, design offices,
construction companies and consulting firms as well as in some cases by previous relevant researches. The
collected data were processed in order to identify the most common morphological solutions for industrial
buildings. The data received from different countries were in most cases inhomogeneous, so a
homogenization process was necessary for all the Census Forms.

1.1.1 Contribution of partners

In the table 1.1 the contributions coming from the partners related to the industrial building realized in the
various countries are reported.

Partner Contribution
RWTH German Census Form and 7 individual forms
UniTH, SHE Greek Census Form and 10 individual forms
ILVA, UniCAM,
Italian Census Form
UniPI, FENO
ISQ Portuguese Census Form
4 Spanish Census Forms were provided. For each Seismic
Zone (Zone 1 or 2), two forms were submitted; one
UNAV concerning steel buildings and one concerning concrete
buildings. All the above data were concentrated in one
summary Census Form.
VTT Census Forms from Finland, Hungary, Poland and Romania
Table 1.1 Contributions received by PRECASTEEL partners.

1.1.2 Morphology and correlated activities

1.1.2.1 Main activities performed in the industrial building

Partners from Germany, Greece, Portugal and Spain made a distinction of main industrial activities between
a) Light industrial activities, b) Warehousing and c) Heavy industrial activities. Partners from Italy made a
different distinction of industrial activities as it is seen in the following graphs. Hungary and Romania
provided values without percentages while Finland and Poland did not provide any value. In particular, the
industrial buildings in Hungary and in Romania mainly house the following activities: Warehouse, Factory,
Agriculture (Hungary) and Warehouse, Car service, Light factory halls (Romania).
All information collected using the CFs concerning main activities performed in the IBs is presented.

1.1.2.2 Type of structural system

The main structural system of industrial buildings adopted in most of the countries is steel or concrete
system. As far as the concrete system is concerned, in countries such as Italy, Spain, Hungary and Finland
prefabricated concrete has a great market share. Another adopted system in countries such as Germany,
Finland and Poland is the timber system appearing to have a small market share though. The information
collected by the CFs concerning the type of structural system adopted in the IBs is presented.

17
 Spain

Portugal
Light industrial activities
Warehousing
Heavy industrial activities
Greece

Germany

0% 20% 40% 60% 80% 100%

Activities of private firms

Property developer
Italy
Public activities
Other activities and privates

0% 20% 40% 60% 80% 100%

Figure 1.1. Main activities performed in the industrial buildings.

Romania 60%

Poland
50%
Hungary
Steel 40%
Finland Concrete
Timber 30%
Spain
Other
20%
Portugal

Italy 10%

Greece 0%

Germany
Steel Concrete Timber Other

0% 20% 40% 60% 80% 100% (b)

(a)
Figure 1.2. Type of structural system – a) survey country by country – b) European survey

1.1.3 Basic geometric parameters of steel framing

1.1.3.1 Number of frame bays (in-plane direction)

One-frame or two-frame contiguous buildings are commonly used in industrial activities. The one-frame
solution is however more adopted. The information collected by the CFs concerning the number of frame
bays in the in-plane direction adopted in the IBs is presented. Moreover, a general overview of the number of
frame bays in the in-plane direction adopted in the industrial buildings is also depicted below, condensing all
the collected data (gathered country by country) in homogeneous set.
n
io
ct
re
di
ne
la
f-p
-o
ut

in-plan
O

e directio
n
Figure 1.3. In-plane and out-of-plane direction in industrial building

18
 70%
Romania
60%
Poland
50%
Hungary
40%
Finland 1
2 30%
Spain >=3
20%
Portugal
10%
Greece
0%
Germany 1 2 >=3
0% 20% 40% 60% 80% 100%
(a) (b)
Figure 1.4. Number of frame bays (in-plane direction – along the main resisting system)
( )
60%

Spain
50%

<=11 40%
Portugal
12-15 30%
16-19
Greece 20%
>=20
10%

Germany 0%
<=11 12-15 16-19 >=20

0% 20% 40% 60% 80% 100%

(a) (b)
Figure 1.5. Numbers of frames in the out-of-plane directions – (a) subdivision country by country – (b)
Aggregation of the data (European level)

1.1.3.3 Span of each bay

A large range of values for span of each bay appears. However, in most cases a span of 20-25 m is adopted.
Larger spans are used in the cases where adequate free space is required in order to serve special industrial
activities. The information collected by the CFs concerning the span of each bay adopted in the industrial
buildings is presented. Moreover, a general overview of the span of each bay adopted in the industrial
buildings is also depicted below.

40%
Romania
35%
Poland
30%

Hungary 25%
<20 m
Finland 20%
20-25 m
15%
Spain 25-30 m
>30 m 10%
Portugal
5%

Greece 0%
<20 m 20-25 m 25-30 m >30 m
Germany

0% 20% 40% 60% 80% 100%

(b)
(a)
Figure 1.6. Span length of each bay recognized during the investigation – (a) subdivision of the results
country by country – (b) Aggregation of the data (subdivision at European level)

19
1.1.3.4 Distance between consecutive frames
The distance between consecutive frames is in most cases bounded between 5 and 7 meters. The information
collected by the CFs concerning the distance between consecutive frames adopted in the industrial buildings
is presented. Moreover, a general overview of the distance between consecutive frames adopted in the
industrial buildings is also depicted as comparison for the picture in which the results are presented country
by country.

70,0%
Spain
60,0%

50,0%
Portugal
5-7 m 40,0%
>7 m
30,0%
Greece
20,0%

Germany 10,0%

0,0%
0% 20% 40% 60% 80% 100% 5-7 m >7 m
(a) (b)
Figure 1.7. Distance between two consecutive frames – (a) subdivision of the results country by country – (b)
Aggregation of the data (subdivision at European level)

1.1.3.5 Height of frame columns

The height of columns is usually 6-8 meters. However, there are cases that due to a special activity, a taller
building is required. For example, when the use of an overhead travelling crane is demanded, a height of
column over 8 meters is necessary. The information collected by the CFs concerning the height of frame
columns adopted in the IBs is presented. Moreover, a general overview of the height of frame columns
adopted in the industrial buildings is also depicted as comparison for the picture in which the results are
presented country by country.

Romania 50%
45%
Poland
40%
Hungary 35%
30%
Finland <6 m
25%
6-8 m
Spain 20%
>8 m
15%
Portugal
10%

Greece 5%
0%
Germany <6 m 6-8 m >8 m

0% 20% 40% 60% 80% 100%

(b)
(a)
Figure 1.8. Height of the columns – (a) subdivision of the results country by country – (b) Aggregation of the
data (subdivision at European level)

1.1.4 Type of girder

The two main types of girder are beams and trusses. Beams appear more frequently and especially for small
or medium spans. If greater spans are necessary, the truss girder solution is used. The information collected
by the CFs concerning the type of girder adopted in the industrial buildings is presented. Moreover, a general
overview of the type of girder adopted in the industrial buildings is also depicted as comparison for the
picture in which the results are presented country by country.

20
 70%
Romania
60%
Poland
50%
Hungary

Finland 40%
beam
Spain truss
30%
Portugal
20%
Greece
10%
Germany

0% 20% 40% 60% 80% 100% 0%

(a) beam truss
(b)
Figure 1.9. Type of girder employed in the main frame – (a) subdivision of the results country by country –
(b) Aggregation of the data (subdivision at European level)

1.1.4.1 Slope of frame girder

The most common angle used varies between 6°-12°. In countries where heavy snowfalls are expected,
greater angled roof solutions are preferred. The information collected by the CFs concerning the slope of
frame girder adopted in the IBs is presented. Moreover, a general overview of the slope of frame girder
adopted in the industrial buildings is also depicted as comparison for the picture in which the results are
presented country by country.

50%
Romania
45%
40%
Hungary
35%
<6 30%
Spain
6-9 25%
9-12 20%
Portugal
>12 15%
10%
Greece
5%

Germany 0%
<6 6-9 9-12 >12

0% 20% 40% 60% 80% 100%

(a) (b)
Figure 1.10. Slope of frame girder – (a) subdivision of the results country by country – (b) Aggregation of
the data (subdivision at European level)

1.1.5 Parameters of vertical and crane loading

1.1.5.1 Snow loading
The snow loading is defined for each country by the corresponding map of EC1. However, the most common
range of values is between 0.6-2 kN/m2. Moreover, as expected, the North countries have greater snow
loading. The information collected by the CFs concerning the snow loading for the IBs is presented.
Moreover, a general overview of the snow loading for the industrial buildings is also depicted as comparison
for the picture in which the results are presented country by country.

21
 50%
Romania
45%
Hungary 40%

Finland
35%

<0.6 30%
Spain
0.6-1.2 25%
1.2-2 20%
Portugal
>2
15%
Italy
10%
Greece 5%
0%
Germany
<0.6 0.6-1.2 1.2-2 >2
0% 20% 40% 60% 80% 100%
(a) (b)
Figure 1.11. Snow loading – (a) subdivision of the results country by country – (b) Aggregation of the data
(subdivision at European level)

1.1.5.2 Roof utility loading

The roof loading depends mainly on the utility of the roof. The most common loading case corresponds to
inspection and repair of roof parts. For example, the main roof utility loading in Spain is 1 m2 over an area of
10 m2. Rarely, the roof is used as a floor and therefore the loading is equal to floor loading (e.g. 2 kN/m2). As
it results from the analysis, the most common roof loading values are between 0.25-1 kN/m2. All information
collected by the CFs concerning the roof utility loading in the IBs is presented. Moreover, a general
overview of the roof utility loading in the industrial buildings is also depicted as comparison for the picture
in which the results are presented country by country.

70%

60%
Greece <0.25
0.25 50%
>0.25
Germany 40%

30%
0% 20% 40% 60% 80% 100%
0.3 20%
Portugal 1
10%
2
0% 20% 40% 60% 80% 100% 0%
(a) <0.25 0.25-1 >1
(b)
Figure 1.12. Live loading on roof system – (a) subdivision of the results country by country – (b)
Aggregation of the data (subdivision at European level)

1.1.5.3 Overhead travelling crane lifting capacity

The overhead travelling crane lifting capacity is highly associated with the activities taking place in each
building. In many cases there is no need for using a travelling crane, because the activities in those buildings
do not demand it (e.g. very light industrial activities). In particular, in Finland, Hungary and Romania a
percentage of 60%, 85% and 90% respectively of industrial buildings do not require crane. Nevertheless, as a
solution in most applications a 100 kN travelling crane is adopted. As far as the travelling crane self-weight
is concerned, it results from the data collected that it highly depends on the crane span and type. The
information collected by the CFs concerning the overhead travelling crane lifting capacity is presented.
Moreover, a general overview of the overhead travelling crane lifting capacity is also depicted in terms of
loading capacities only.

22
 Overhead traveling crane lifting capacity (in kN) <100
100-200
Germany >200 60%

50%
Portugal <100
100 40%
Greece >100
30%

20%
Romania
10%
Poland
<50
0%
>50 <=50 kN 50-100 kN >100 kN
Hungary

Finland (b)
0% 20% 40% 60% 80% 100%
(a)
Figure 1.13. Distribution of the crane capacity in the industrial buildings – (a) subdivision of the results
country by country – (b) Aggregation of the data (subdivision at European level)

1.1.6 Parameters for horizontal loading

1.1.6.1 Type of in-plane frame resistance

The type of in-plane frame resistance that commonly appears is Moment Resisting Frames (MR). That is
because the free space provided from that system is usually necessary for the use of the building. The
information collected by the CFs concerning the type of in-plane frame resistance adopted in the IBs is
presented. Moreover, a general overview of the type of in-plane frame resistance adopted in the industrial
buildings is also depicted as a distribution across the Europe grouping the data only in terms of resisting
system.

Romania 100%
90%
80%
Spain 70%
60%
MR
50%
Portugal CB
40%
EB
30%
Greece 20%
10%
0%
Germany MR CB EB

0% 20% 40% 60% 80% 100% (b)

(a)
Figure 1.14. Resisting system of in-plane direction – (a) subdivision of the results country by country – (b)
Aggregation of the data (subdivision at European level)

1.1.6.2 Type of resistance out-of-plane

In the out-of-plane direction, Concentric Bracing system (CB) is frequently used. Moreover, Eccentric
Bracing system (EB) is also adopted in cases where small openings are required. MR solutions in the out-of-
plane direction are seldom. The information collected by the CFs concerning the type of resistance in the
out-of-plane direction adopted in the IBs is presented. Moreover, a general overview of the type of resistance
in the out-of-plane direction adopted in the industrial buildings is also depicted in term of distribution across
the European territory.

23
 80%
Romania 70%

60%
Spain
50%
MR
40%
Portugal CB
EB 30%

Greece 20%

10%
Germany 0%
MR CB EB
0% 20% 40% 60% 80% 100%
(a) (b)
Figure 1.15. Resisting system of out-of-plane direction – (a) subdivision of the results country by country –
(b) Aggregation of the data (subdivision at European level)

1.1.6.3 Maximum ground acceleration

Earthquake Zones, and as a consequence the ground acceleration of each area, are defined by the
corresponding map in EC8. Nevertheless, the most common values of maximum ground acceleration, as a
fraction of gravity acceleration g, vary between 0.05-0.25g. More specifically, in High Seismicity regions
such as Greece, Portugal, Romania parts of Italy and Spain, the value of ground acceleration is higher than
0.25g (with maximum values of 0.36g). In Low Seismicity regions such as Finland, Germany and Hungary
the value of ground acceleration does not exceed 0.1g. All information collected by the CFs concerning the
maximum ground acceleration is presented. Moreover, a general overview of the maximum ground
acceleration is also depicted in terms of distributed p.g.a. across the Europe.

40,0%
Romania
35,0%
Hungary
30,0%
Finland
25,0%
<0.05
Spain 20,0%
0.05-0.15

Portugal 0.15-0.25 15,0%

>0.25
Italy 10,0%

5,0%
Greece
0,0%
Germany <0.05 0.05-0.15 0.15-0.25 >0.25

0% 20% 40% 60% 80% 100%

(a) (b)
Figure 1.16. Maximum peak ground acceleration – (a) subdivision of the results country by country – (b)
Aggregation of the data (subdivision at European level)

1.1.6.4 Wind load intensity

The main factors that influence the wind load intensity are the distance from the coast, the altitude of the area
and the terrain typology. The EC1 proposes wind load values for each case. The main concentration of
values is between 25-29 m/sec. In coastal areas though, values higher than 29 m/sec also appear. The
information collected by the CFs concerning the wind load intensity is presented. Moreover, a general
overview of the wind load intensity is also depicted in terms of wind intensity groups for all the examined
European countries.

24
 45%
Romania
40%
Hungary
35%
Finland 30%
<25
Spain 25%
25-29
Portugal >29 20%
15%
Italy
10%
Greece
5%
Germany 0%
<25 25-29 >29
0% 20% 40% 60% 80% 100%
(a) (b)
Figure 1.17. Wind load intensity – (a) subdivision of the results country by country – (b) Aggregation of the
data (subdivision at European level)

1.1.7 Secondary elements of the steel structure

1.1.7.1 Type of roofing

Sandwich panels as well as corrugated metal sheets are the most commonly adopted solutions for roof
covering. These two solutions have almost equal market share. Other solutions such as concrete, fiber-
cement, aluminium and asphalt are also used. The information collected by the CFs concerning the type of
roofing coverings adopted in the IBs is presented. Moreover, a general overview of the type of roofing
coverings adopted in the industrial buildings is also depicted grouping all countries in typology groups.

y g
Romania 50%
45%

Spain 40%
35%
30%
Portugal Sandwich panels
25%
Corrugated Metal sheet
20%
Italy Concrete 15%
Other 10%
5%
Greece
0%
Sandwich Corrugated Concrete Other
Germany panels Metal sheet

0% 20% 40% 60% 80% 100% (b)

(a)
Figure 1.18. Type of roofing systems – (a) subdivision of the results country by country – (b) Aggregation of
the data (subdivision at European level)

1.1.7.2 Type of side cladding

Sandwich panels, corrugated metal sheet as well as prefabricated concrete panels are the most commonly
adopted solutions for side cladding. Moreover, other systems such as brick walls, Glass and Aluminium
systems are adopted. The information collected by the CFs concerning the type of side cladding adopted in
the IBs is presented. Moreover, a general overview of the type of side cladding adopted in the industrial
buildings is also depicted grouping all collected data only in temrs of cladding type.

25
 35%
Romania
30%

Spain 25%
Sandwich panels
20%
Portugal Corrugated metal sheet
Precast concrete panels 15%
Italy Other
10%

Greece 5%

0%
Germany
Sandwich Corrugated Precast Other
panels Metal sheet Concrete
0% 20% 40% 60% 80% 100% panel
(a) (b)
Figure 1.19. Type of side-cladding systems – (a) subdivision of the results country by country – (b)
Aggregation of the data (subdivision at European level)

1.1.7.3 Type of purlins

Steel hot-rolled profiles and cold formed elements are the two types of purlins used. The cold formed
elements have a bigger market share compared to the hot-rolled profiles. The information collected by the
CFs concerning the type of purlins adopted in the IBs is presented. Moreover, a general overview of the type
of purlins adopted in the industrial buildings is also depicted grouping the collected data only in terms of
cladding type.

70%
Romania
60%
Finland
50%

Spain 40%
Cold-formed
Profile sections
Portugal 30%

20%
Greece

10%
Germany
0%
0% 20% 40% 60% 80% 100% Cold-formed Profile sections
(a) (b)
Figure 1.20. Type of purlins – (a) subdivision of the results country by country – (b) Aggregation of the data
(subdivision at European level)

1.2 Commercial buildings

Commercial Buildings (CBs) include a wide variety of typologies. Depending on their dimension, location,
organization and facilities, they are divided into: shopping centres, supermarkets and hypermarkets,
department stores, factory outlets, and leisure centres. A brief definition of various kinds of commercial
centres is reported hereafter in order to understand the dimension of buildings and their use.
‚ Shopping centres: they are homogeneous architectural units of independent establishments that are planned,
developed and managed as a single property. Their size, merchandising mix, facilities and parking areas are
determined by the market characteristics.
‚ Supermarkets: they are structures for the direct sale, with surface included between 400 and 2000 sm, with
individual additional spaces.
‚ Hypermarkets: they are covered units with unique volume, on 1-2 storeys and surface variable from 5000 to
50000 sm, placed in areas external to urban centres, endowed with central facilities.

26
 Staff
Customer entrance

Sales area

Warehouse and
preparation

Customer services

Staff services

Customers
Car park

Figure 1.21. Typical supermarket for food-stuff sales layout

Staff

Tradesmen

Hypermarket and
minimarket

Shops

Information offices

Toilets

Restaurant

Paths and stay

areas

Warehouse
Customers
Car park

Figure 1.22. Typical department store layout

‚ Department stores: they are structures placed in urban centres or areas of value; they are articulated for
different commodities (except food-stuffs) with surface variable from 3000 to 20000 sm.
‚ Factory outlets: they are structures where manufacturers and retailers sell merchandises at discounted prices.
‚ Theme–oriented leisure–based centres: include some retail units typically concentrated on a narrow but deep
selection of merchandise within a specific retail category. Leisure–based centres are usually anchored to
multiplex cinemas, restaurants, bars and other leisure facilities.
‚ Retail parks: consist of several mid/large size retail warehouses with direct access from parking or pedestrian
areas; they are open schemes, planned, developed, built and managed as a single entity.
All these commercial centres, are facilities that provide basically a proper environment for the purpose of
salling and storing goods, and often accommodate offices and leisure activities. Buildings for commercial
centres must be designed to contain the loads of the materials to be stored, the associated handling
equipment, the receiving and shipping operations and associated trucking, and the needs of the operating
personnel. The spaces should be planned to best house business service requirements and the products to be
stored/handled. Designers should focus on making the spaces functional and efficient, while providing a safe
and comfortable environment for customers and workers. Building image and aesthetics, landscaping,
customer and worker safety as well as comfort, become important issues in competitive real estate markets
(Figs. 1.21 and 1.22).
Data about CBs have been collected from ten European countries namely: Czech Republic, Finland,
Germany, Greece, Hungary, Italy, Poland, Portugal, Romania and Spain, some of which are characterised by
high seismic risk.
Data were processed in order to identify the morphology of buildings that have the most diffused commercial
activities. This is necessary to (i) understand requirements for each particular commercial activity, (ii) point
out advantages/disadvantages relevant to the various structural solutions adopted, and (iii) define more

27
 suitable steel and steel-concrete solutions that can be used as an alternative to the reinforced concrete and
prestressed reinforced concrete solutions.
The objective of this investigation is to provide a picture of morphological and structural solutions adopted
in the construction of CBs with reference to standard situations.

1.2.1 Data collection

The CBs census regarded relatively young structures constructed after 1990 (and most of them after 2000).
In addition to the general information (location, year of construction and total rental surface) data were
collected concerning the activities carried out in the CBs, the structural typology and the relevant loading
actions. The aim was to get gross information about a large number of cases and detailed information about a
limited number of cases in order to obtain macro- and micro-indicators. These were used to define the
sample structures that were studied in the framework of PRECASTEEL.
The work was carried out differently in the various countries due to the difficulty of collecting data already
processed by national agencies and to the bad feedback with design offices and construction companies.
Data were processed by maintaining separate those from the various countries to avoid difficulties of
interpretation due to the different samples. The results were normalised with respect to the number of cases
for each country.
‚ Data from Czech Republic, Hungary, Romania, Finland and Poland
Data and market information relevant to new low-rise CBs in Finland, Hungary, Poland, Romania and Czech
Republic come from (i) construction statistics for interesting building types, (ii) construction industry
organizations, construction magazines and companies. Information about frame materials, average size and
storeys from interesting buildings were collected.
No specific examples were considered and structural data were collected by considering (i) typology of the
primary structure, (ii) total covered area, (iii) number of storey, (iv) span length. Indication on the most
common roof coverings, floor types and wall claddings were also provided.
‚ Data from Germany
Data were collected by focusing on nine particular cases. Four cases are significant for a large number of
commercial buildings belonging to different chain stores. The other five cases are considered only as
particular examples useful to define typologies.
‚ Data from Greece
The analysis focused on nine particular cases. Two cases are significant for a large number of commercial
buildings belonging to different chain stores. The other seven cases are considered only as particular
examples useful to define typologies.
‚ Data from Italy
No national statistics are available with reference to structural data and only market global information, such
as global rentable surfaces, distribution of services, etc., were collected from national agencies and societies.
Twenty-four particular cases were investigated by contacting design offices of commercial companies
owners of CBs chains and precast concrete companies.
‚ Data from Portugal
The analysis was based on studies elaborated by national agencies and associations, and on interviews to
design offices, precast concrete and metallic construction companies. Data are relevant to CBs with rental
areas between 1000-5000 m2 and 5000-20000 m2.
In addition, in order to have typical examples, details of some relevant cases were collected.
‚ Data from Spain
Data from Spain were collected with reference to ten significant cases.

1.2.2 Morphological and structural data

As resulted from the census, reinforced concrete and prestressed reinforced concrete structures are the most
adopted in the major part of the countries for both vertical (Figure1.23a) and horizontal (Figure1.23b)
elements. Other materials are also used in some countries; in particular, even if with minor incidences,
timber structures are used in Finland, Hungary, and Poland. In Germany, masonry is widely used in vertical
structures mixed with reinforced concrete elements.
Figure1.24 shows results obtained by summating data for R.C. and P.R.C. structures and those for Steel and
Composite structures. Two situations may be distinguished: (i) countries in which the major part of the CBs
are constructed with R.C. and P.R.C. structures, and (ii) countries in which Steel and Composite structures

28
are about equal to R.C. and P.R.C. structures. Spain, where steel and composite structures are prevalent,
seems to be an exception.
From these data, it seems that the market of the CB construction may be viewed with a certain interest by
steel manufacturer and producers.

Spain Spain

Romania Romania

Portugal Portugal

Poland Masonry Poland

R.C.+P.R.C.
R.C.+P.R.C. Italy
Italy Steel
Steel
Hungary Composite
Hungary
Composite
Other
Greece Other Greece

Germany Germany

Finland Mixed Masonry & R.C. Finland

Czech Rep. Czech Rep.

0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%
(a) (b)
Figure 1.23 – Material: (a) vertical elements; (b) horizontal elements (% from the number of buildings)

Prevalent R.C. and P.R.C. Balanced Prev.S and C

1.00

0.80

0.60
R.C.+P.R.C.
Steel + Comp.
0.40

0.20

0.00
y

ia
y

n
nd
e

d
ly

ga
an

ar

ai
ec

an
an
Ita

la

Sp
ng

rtu
m

re

nl
m

Po
er

Fi
Hu

Po
G

Ro
G

Figure 1.24. Material: most adopted typologies

Only a small percentage of CBs among those investigated have more than two storeys (Figure 1.25).
Although the major part of the CBs have only one storey, it is important to notice that the typology with two
storeys is quite common and features a significant number of cases in some countries (Romania, Hungary).
In Italy and Greece, two-storeys CBs are even more common than one-storeys CBs.
The minimum and maximum storey heights are scattered between 4 and over 10 m. It is worth noting that for
countries in which only one-storey CBs were inserted in the census, the minimum and maximum storey
heights are very different. For Spanish and Portuguese buildings the storey height is quite high whereas in
the case of German CBs a large part of the buildings is characterised by a somewhat low height (most of
them less than 4 m).
With reference to countries where multi-storey CBs were considered (e.g. Italy), the comparison of results
reported in Figure1.25 with those in Figure1.26 permits to deduce that the major storey heights are
associable to one-storey buildings. Furthermore, the variability of data is explainable by considering that
multi-storey CBs are usually characterised by more complex architectures and distribution of spaces.

29
 Spain 1.20

Romania
1.00
Portugal
0.80
Poland
1
Italy 1 storey
2 0.60
2 storeys
Hungary 3
>3 0.40
Greece

Germany 0.20

Finland
0.00
Czech Rep.

y
nia
y

ain
d

ce
d

ly
ga
an

ar
lan

an

Ita

ee
Sp

ma
rtu

ng
rm

l
Po

Fin

Gr
Po

Hu
0% 20% 40% 60% 80% 100%

Ro
Ge
Figure 1.25 – Number of storeys (% from the number of buildings)

Spain max

Spain min

Portugal max

Portugal min
<4
Italy max 4-7
Italy min 7-10
>10
Greece max

Greece min

Germany max

Germany min

0% 20% 40% 60% 80% 100%

Figure 1.26. Storey height (% from the number of buildings)

As for lateral resisting structures (Figure1.27), braced systems are used mainly in Spain and Portugal where
the major number of structures are steel or composite steel-concrete structures. In all the other cases, where
the most common typology is constituted by r.c. and p.r.c. structures, the lateral resistance is entrusted to
shear walls, moment resisting frames and other systems. In particular the last case (other systems) comprises
p.r.c. structures where beams are simply hinged to columns and the lateral resisting system may be assumed
to be that of the inverted pendulum.
The horizontal elements are mainly constituted by solid girders except for German cases that are relevant to
one-storey buildings with large spans covered with truss girders (Figure1.28). Cases denoted by Other are
related to timber girders used in CBs in Finland, Hungary and Poland. With reference to solid section girders
(Figure1.29a), most of the census cases fall in the range 5-15 m but spans lengths greater than 15 m can be
found when prestressed reinforced concrete girders are used. Truss girders are adopted for span length
greater than 15 m with maximum values that can exceed 25 m (Figure1.29b).

30
 Spain Transv.

Spain Long.

Portugal Transv.

Portugal Long.
Wall
Italy Transv. M.R.F.
Italy Long. Braces
Other
Greece Transv.

Greece Long.

Germany Transv.

Germany Long.

0% 20% 40% 60% 80% 100%

Figure 1.27. Lateral resisting structural systems (% from the number of buildings)

Spain

Romania

Portugal

Poland

Italy Solid sect. girders

Truss girders
Hungary
Other
Greece

Germany

Finland

Czech Rep.

0% 20% 40% 60% 80% 100%

Figure 1.28. Lateral resisting structural systems (% from the number of buildings)

Spain max Spain max

Spain min Spain min

Portugal max Portugal max

Portugal min <5 Portugal min <5

Italy max 5-10 Italy max 5-15

10-15 15-25
Italy min Italy min
>15 >25
Greece max NR Greece max NR

Greece min Greece min

Germany max Germany max

Germany min Germany min

0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%
(a) (b)
Figure 1.29. Span length: (a) solid elements; (b) truss elements (% from the number of buildings)

Flooring systems are the classical used for r.c., p.r.c., steel and composite structures (see Tab.1.2). For what
concern coverings, various light typologies based on sandwich panels are widely used. In some cases,
especially in Germany, these are supported by different substructures like timber trusses and light-gauge
structures.

31
 Main structure Floor system
R.C. and P.R.C. R.C.
Predalles
Hollow core slabs
Steel and Composite Composite slabs
Table.1.2. Main structure and floor systems combinations

1.2.3 Loading actions

Not all the countries considered in the investigation fall in seismic prone areas. Spain and Hungary are
characterised by a medium-low seismicity, whereas Greece, Italy, Portugal and Romania are characterised by
higher seismicity with peak ground acceleration greater than 0.25g (with maximum values of 0.36g)
(Figure1.30).

Spain

Romania

Portugal

Poland
<0.05
Italy 0.05-0.15
Hungary 0.15-0.25
>0.25
Greece

Germany

Finland

Czech Rep.

0% 20% 40% 60% 80% 100%

Figure 1.30. Seismic peak ground acceleration (% from the number of cases)

The reference wind speeds are included in the range 20-40 m/sec but such values are the reference ones and
have to be modified by suitable coefficients in order to take into account specific situations connected to site
configurations and geometry of the building.
The snow loads are widely variable depending on the geographic area, on local topographic conditions and
on geometric configuration of roofs.
Live loads are suggested by technical codes and may be imposed according to the particular use of the
building. Most of cases are characterised by a live load value of 5 kN/m2 but very higher values (till 20
kN/m2) were registered in some cases for the CB sections devoted to storage.

1.3 Analysis on potential emerging markets

The aim was to collect relevant statistics data and other market information in five countries for (1) Single
Storey Industrial Buildings (SSIB - IB) and (2) Low-Rise Commercial Buildings (LRCB - CB)
The goal was approached by looking at the storage, industrial and warehouse markets by:
‚ collecting public information from construction statistics in Finland, Poland, Hungary, Czech Republic and
Romania, and
‚ by conducting a survey of construction industry organizations, construction companies in the above countries
for collecting of relevant information.

The focus was the most adopted morphological and structural solutions for single-storey industrial (SSIB)
and low-rise commercial (LRCB) buildings, paying particular attention to the adopted geometries (height,
spans,…), the typology of roofing systems and infill wall systems, the presence of overhead travelling crane,
the possibility of external operation of the crane and all the requirements related to the free space entrance to
carry out all the industrial activities, lighting advantages and disadvantages.

32
1.3.1 Country specific statistics
The number, m3 and m2 of storage, industrial and commercial new buildings in Finland have handled
including statistics between 2000…2006. Commercial, transport, industry and public buildings can include
also SSIB & LRCB types of buildings. Building statistics in Poland, Czech Republic, Hungary and Romania
include different data of storage, industrial and warehouse buildings by country. They can be a part of
commercial, industry or transport buildings, or exact storage and warehouse buildings.
There is comparison between new storage, commercial and industrial building volumes in Figure.31 and
Figure . In Finland new storage buildings were less than 5% of new building volume in 2006, but in the other
countries they represented about 30 %…40 %.

New Industrial Building Construction in 4 Countries

2500

2000

1500
mill. eur

- Czech Republic
- Hungary
- Poland
1000
- Finland

500

0
2003 2004 2005 2006 2007* 2008*
Source: Euroconstruct 6/2007, VTT
VTT 2007

Figure 1.31. New industrial building volumes (€) is increasing in four countries. Source: Euroconstruct
6/2007.

New Commercial Building Construction in 4 countries

1800
1600
1400
1200
mill. eur

1000 - Czech Republic

- Hungary
800 - Poland
- Finland
600
400
200
0
2003 2004 2005 2006 2007* 2008*
VTT 2007
Source: Euroconstruct 6/2007, VTT

Figure 1.32. New commercial building volumes (€) can include also same kind of building space like storage
buildings. Euroconstruct 6/2007.

33
1.3.2 Country specific study
In Finland the storage construction starts vary between 3…6 mill.m3 building space yearly; or 400 000
…600 000 m2/a . The markets share of bearing structure materials of steel is about 44…47 % in building
space and 35 % of building units, but sheet metal on facades is much more, about 50…75 % of facade area.
An average floor area is about 500…700 m2; a little bit more in case of steel framed buildings.
During last years between 4…8 million m3 new industrial buildings were constructed in Finland. General
bearing structures are concrete and steel. An average size new industry building is about 600…800 m2
(2000…2006), with steel frame buildings a bit smaller an average.
Commercial building starts have increased very much during 2007 in Finland. Commercial structure is
changing and different national and international chains are investing much to secure Finnish market.
Commercial buildings have mostly concrete frames (about 2/3), but there have been also some very big steel
frame buildings. Commercial storages are included to storage buildings in Finland. Typical size of
commercial buildings has increased during last years an average size of commercial buildings have been
between 500…1000 m2/unit.
New warehouses (2004-2006) are mostly larger than size class 5000 m2 in floor area in Finland (Figure).
Commercial warehouses are smaller market than other warehouses, but commercial buildings itself are much
larger markets. Small warehouses are not so common. Most warehouses are 1-storey buildings, because
about 90 % of floor areas are coming from 1-storey buildings and only 10 % have more storeys.

New Warehouses by Size-class in Finland between 2004…2006 (m2)

500000
Other warehouses
450000 Commercial warehouses
Industrial warehouses
400000
350000
300000
m2

250000
200000
150000
100000
50000
0
< 100 100-1000 1000-5000 >5000
VTT 2007
Source: Statistics Finland 2007 Size-class m2

Figure 1.33. New warehouses in sub-sectors in Finland.

An average size of new commercial building in Poland is about 3400 m3, industrial building 11000…14000
m3 and warehouse 5000…8500 m3. The frame material of new buildings is not available from construction
statistics. Typical trade warehouses in building stock are 500 m2/unit, roofed warehouses in private sector are
about 1000 m2/unit, while storage sites can be up to 3000m2/unit.
The most stable segment of the construction market in Hungary is non-residential building construction.
After 2005, a year considered as successful with the development of non-residential building construction, a
large number of projects (office, industry, warehouse, logistics, and commerce) were started or completed in
2006. (Euroconstruct, Finpro b).
Industrial warehouses include to industrial buildings. On the construction of retail stores and shops, 30 % of
the floor space of new retail facilities are made up by shopping centres and ”strip malls”, 30 % by
hypermarkets, 19 % by supermarkets, 13 % by DIY stores and 7 % by stores of interior design, on a total
area of 6-700 thousand m2. (Euroconstruct).
Currently nearly 200 industrial parks are operated in Hungary together with another 11 ones are to be added
in 2007. The industrial parks now account for some 7-10 % of total employment and 30% of total industrial
production and export. Net construction cost of buildings were in industrial buildings 830 €/m2, warehouses
34
604 €/m2 and commerce 845 €/m2 at prices of 2007. Industrial or warehouse construction projects are linked
to infrastructure. (Euroconstruct, 2007)
In the Czech Republic the non-residential output reached 6.99 bills. Euro in the year 2006 in, this is 40 % of
total construction output (only 30.2 % in west Europe). The top output was reached in the year 1996, until
year 2005 declined by 6.3 % and in the year 2006 is again increasing. (Euroconstruct, 2007).
Two sub-sectors prevails in the CzR: industrial and commercial buildings. They together form 62.3 % of the
total non-residential output (57.3 % in the year 2002). The construction output for new industrial buildings is
in comparison with west Europe average more than 3 times higher, followed by commercial and
administrative buildings that is about 2.4 times higher. Construction of buildings for education, health and
agriculture is falling behind. (Euroconstruct, 2007). The number of released building permits in non-
residential buildings decreased a little in 2006 . Warehouses and storages is just one part of non-residential
new buildings.
Construction in Romania has increased during last years more than 9%, but especially in 2006. Retail and
wholesale buildings included to same building category and also industrial and warehouses. That is why it is
impossible to extract more data for SSIB and LRCB from these statistics.
The price level in industrial sheds varies 300…450 €/m2, purpose built industrial units 400…700 €/m2,
supermarket shells 350…580 €/ m2, shopping centres 500…1000 €/m2 in 1st quarter 2007 (Building
Magazines, 2007).
There is enormous need of new space of logistics and storages. Romania has commercial space 25 m2 per
person while in EU the level is 180 m2 per person.

1.3.3 Morphological and structural data

As it can be seen in Figu, structural materials used in SSIB buildings (per building unit) are almost equally
divided between steel and concrete. Only in Finland there is a significant contribution of timber buildings,
mainly in the range of smaller structures (400-500m2), due to the strong tradition of building with timber in
Finland.
It should also be noted, that due to the activity of the data provider, the data set on Romania is probably
biased towards steel buildings, and specifically portal frame buildings. One should also note the significant
presence of light-gauge steel (LGS) buildings in Romania (in the range of small buildings). This is especially
intriguing as Romania has by far largest earthquake loads from the investigated countries, and LGS
structures are generally considered less advantageous.

Concrete- in situ casting Finland

(%):
Concrete - prefabricated
Czech R.
(%):
Steel- roof truss w ith any
type of columns (%): Poland
Steel - portal frames (%):
Romania
Steel - cold-formed steel
(%):
Hungary
Timber (%):

0% 20% 40% 60% 80% 100%

Figure 1.34. Share of different materials/structural typologies in case of SSIB

Buildings included in the survey seem to fall in the group of smaller buildings (up to 5000m2, Figure), and
they are typically single-storey (70-90%). Some fluctuations of the total area can be observed, most probably
due to the different geographical/population characteristics of the countries, trends reflecting the
requirements of the local market environment.

35
 Number of spans
Total area (m 2)

Finland Finland
< 100m2 1

Czech R. Czech R.
100-1000m2
Poland 2 Poland

1000-5000m2
Romania Romania

>3
>5000 m2 Hungary Hungary

0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%
(a) (b)
Figure 1.35. (a) Total area and (b) and number of spans of SSIB’s

Technological processes in the buildings do not require the presence of an overhead crane in 60-80% of the
cases. This would suggest that the majority of the buildings are not intended to host heavy industrial
processes. Single spans structures are also overwhelming in the survey (Figure 1.35.b).
Concerning the span and height of the SSIB buildings, the majority of the buildings are in the small building
range, with spans of up to 25m, and height of 4-6m (Figure 1.36).
Total height (m)
Span (m)

Finland
Finland < 4m
< 15m

Czech R.
Czech R.
15-25m 4-6m
Poland
Poland

25-35m 6-8m
Romania
Romania

>34m >8m Hungary

Hungary

0% 20% 40% 60% 80% 100%

0% 20% 40% 60% 80% 100%
(a) (b)
Figure 1.36. (a) Span and (b) height of SSIB’s

Prefabricated RC is the preferred choice for LRCB buildings in almost all countries (Figure 1.37).
Prefabricated solutions are proffered both RC and steel. This tendency is stronger in Finland. Timber and
LGS retains a considerable market share of LRCB buildings only in Finland.

Concrete- in situ casting (%):

Finland

Concrete - prefabricated
(%): Czech R.
Steel- roof truss w ith any
type of columns (%):
Poland
Steel - portal frames (%):

Romania
Steel - cold-formed steel (%):

Timber (%): Hungary

0% 20% 40% 60% 80% 100%

Figure 1.37. Share of different materials/structural typologies in case of LRCB

The total area of LRCB buildings is typically larger then 1000m2 (Figure 1.38.a). It is interesting to note that
in more populous countries (Romanian & Poland) the built area can be even larger then 5000m2. LRC
36
 buildings are also typically single-storey or at most two-storey buildings with 60-80% being single storey,
and 80-90% two storeys.

Span (m)
Total area (m 2)

Finland
Finland
< 15m
< 100m2
Czech R.
Czech R.
15-25m
100-1000m2 Poland
Poland
25-35m
1000-5000m2 Romania Romania

>34m
>5000 m2 Hungary Hungary

0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%

(a) (b)
Figure 1.38. (a) ) Total area and (b) typical spans for LRC buildings

Typical spans reported in LRC buildings are presented in Figure 1.38b. It is interesting to note that in
Romania larger spans are reported, despite the large snow loads and significant earthquake in that country. It
appears that the morphology is generally more dictated by the market need compared to adaptation to the
environmental conditions.

1.4 Cost Analysis for designed case studies

In this paragraph the costs of the structural elements for CBs and for IBs collected for three different market
zones (defined by different cost tables) across European countries:
ƒ Italy (Southern Europe);
ƒ Germany (Central Europe);
ƒ Romania, Hungary and Bulgaria – named Romania (Eastern Europe).
The unitary costs were referred to the mean values correspondent to the period June 2008 – February 2010
and should be considered as indicative; moreover, it is worth noting that the prices were quite low due to the
influence of the crisis during 2008 and 2009. The model for the cost analysis was conceived referring all the
following unit costs to the steel consumption, €/kg,:
‚ production of worked pieces;
‚ concrete filling for composite elements;
‚ surface treatment (sand blast or galvanizing);
‚ transportation shop-field;
‚ erection;
‚ surface treatment for fire resistance.
The steel price was chosen as the unique variable governing the cost model because it is directly related to
the market conditions (i.e. growth or drop of the steel product sales). In order to transform all the cost voices
in € per steel kg., it was carried out a statistical investigation about the cost related to the galvanization,
sandblasting or concreting of bare steel profiles taken from the complete European catalogue of IPE and HE
series.
For example, the cost for the fire treatment with intumescent paints was generally defined for surface unit
(i.e. €/m2) and is transformed in €/kg applying the original price to all profiles considered in the IPE, HE
catalogues, see figure 1.39. The cost considered in the cost analysis model considered in the software was
calculated assuming the 95% percentile that over estimate notably the cost for heavy profiles giving also the
50% more then averaged price, but it was decided that for a preliminary design and for the definition of a
general first cost plan, useful for definitively fixing a structural solution, it was adequate over-estimate and
then be sure that final prices should notably be decreased. The same procedure was applied to calculate the
cost for galvanization, for concreting and for the sandblast treatment.
All the costs obtained from the three different countries were reported in the tables 1.3 – 1.6 where the cost
of the foundation was also inserted; this costs was approximately estimated and presented in terms of €/m2. It
is important to underline also that the costs tables were differently organized depending on the structural
37
class considered: IB or CB. In fact the costs associated to IB solutions analyzed during the research project
were defined making reference to structural modulus (i.e. the one story IB defined by 6 main frames placed
at constant inter-axis), while the CB costs tables were referred to single structural components because the
structural solutions is a pure assemblage of different member, each one devoted to a different static role.

0.8 1.2

Cost for intumescent paints - REI120

Cost for intumescent paints - REI60

0.7
1
0.6
95% percentile 0.8 95% percentile
0.5
[€/kg]

[€/kg]
0.4 0.6
0.3
0.4
0.2
0.2
0.1 5% percentile 5% percentile
0 0
HE 140 B
HE 160 M

HE 240 B
HE 260 M

HE 340 B
HE 360 M

HE 550 B
HE 600 M
HE 650 B

HE 140 B
HE 160 M

HE 240 B
HE 260 M

HE 340 B
HE 360 M

HE 550 B
HE 600 M
HE 650 B
HE 100 AA
HE 120 A

HE 200 AA
HE 220 A

HE 300 AA
HE 320 A

HE 450 AA
HE 500 A

HE 700 A

HE 100 AA
HE 120 A

HE 200 AA
HE 220 A

HE 300 AA
HE 320 A

HE 450 AA
HE 500 A

HE 700 A
(a) (b)
Figure 1.39 (a) cost of the intumescent paint, €/m2, for REI60 transformed in €/kg; (b) cost of intumescent
paint, €/m2, for REI120 transformed in €/kg

Bare steel Concrete Transportation Treatment for fire

cost, Cut- Production Filling for Surface treatment (from the shop to resistance Foundation
Erection
offs, plates, Assembling composite €/kg the field) [intumescent €/m2 (plan)
bolts solutions max 400 km paints] €/kg
A
B
Sandblast SA2.5 average price
€/kg €/kg €/kg Hot-deep €/kg R60 R90 R120 Pinned Fixed
primer 40 Pm €/kg/(100 km)
galvanizing
thickness
Column HE Rolled profiles 0.91 0.350 - 0.12 0.24 0.030 0.350 0.75 1.20 1.66 20 25
Column CH Welded tubes 0.96 0.555 - 0.14 0.35 0.030 0.360 0.92 1.47 2.02 20 25
Column CH Extruded tubes 1.45 0.555 - 0.14 0.35 0.030 0.360 0.92 1.47 2.02 20 25
Column CFS Cold formed profiles 0.93 0.555 - 0.12 0.24 0.030 0.350 3.20 5.10 7.00 20 25
Column HE Rolled profiles 0.91 0.350 0.530 0.12 0.24 0.030 0.350 0.30 0.47 0.65 20 25
Column CH Welded tubes 0.96 0.555 0.175 0.14 0.35 0.030 0.360 0.92 1.47 2.02 20 25
Column CH Extruded tubes 1.45 0.555 0.175 0.14 0.35 0.030 0.360 0.92 1.47 2.02 20 25
Beam IPE/HE Rolled profiles 1.37 0.350 - 0.12 0.25 0.030 0.350 1.13 1.81 2.49 20 25
Beam CH Rolled profile trussed beam 0.93 0.550 - 0.20 0.30 0.030 0.390 0.83 1.33 1.82 20 25
Beam CFS Cold formed trussed beam 0.93 0.550 - 0.20 0.30 0.030 0.390 3.52 5.61 7.70 20 25
Brace HE/UPN Rolled profiles 0.97 0.350 - 0.12 0.25 0.030 0.360 0.75 1.20 1.66 20 25

Table 1.3. Costs of CB elements for the Italian market

Bare steel Concrete Transportation Treatment for fire
cost, Cut- Production Filling for Surface treatment (from the shop to resistance Foundation
Erection
offs, plates, Assembling composite €/kg the field) [intumescent €/m2 (plan)
bolts solutions max 400 km paints] €/kg
A
B
Sandblast SA2.5 average price
€/kg €/kg €/kg Hot-deep €/kg R60 R90 R120 Pinned Fixed
primer 40 Pm €/kg/(100 km)
galvanizing
thickness
Column HE Rolled profiles 0.77 0.34 0.10 0.20 0.011 0.340 0.64 1.02 1.40 5 10
Column CH Welded tubes 0.82 0.54 0.12 0.30 0.011 0.350 0.78 1.24 1.70 5 10
Column CH Extruded tubes 1.23 0.54 0.12 0.30 0.011 0.350 0.78 1.24 1.70 5 10
Column CFS Cold formed profiles 0.77 0.34 0.10 0.20 0.011 0.340 2.70 4.30 5.91 5 10
Column HE Rolled profiles 0.77 0.34 0.200 0.10 0.20 0.011 0.340 0.25 0.40 0.55 5 10
Column CH Welded tubes 0.82 0.54 0.110 0.12 0.30 0.011 0.350 0.78 1.24 1.70 5 10
Column CH Extruded tubes 1.23 0.54 0.110 0.12 0.30 0.011 0.350 0.78 1.24 1.70 5 10
Beam IPE/HE Rolled profiles 0.89 0.34 0.10 0.21 0.011 0.340 0.95 1.53 2.10 5 10
Beam CH Rolled profile trussed beam 0.79 0.53 0.17 0.25 0.011 0.380 0.70 1.12 1.54 5 10
Beam CFS Cold formed trussed beam 0.79 0.53 0.17 0.25 0.011 0.380 2.97 4.73 6.50 5 10
Brace HE/UPN Rolled profiles 0.77 0.34 0.10 0.21 0.011 0.340 0.64 1.02 1.40 5 10

Table 1.4. Costs of CB elements for Romanian market

38
 Bare steel Concrete Transportation Treatment for fire
cost, Cut- Production Filling for Surface treatment (from the shop to resistance Foundation
Erection
offs, plates, Assembling composite €/kg the field) [intumescent €/m2 (plan)
bolts solutions max 400 km paints] €/kg
A
B
Sandblast SA2.5 average price
€/kg €/kg €/kg Hot-deep €/kg R60 R90 R120 Pinned Fixed
primer 40 Pm €/kg/(100 km)
galvanizing
thickness
Column HE Rolled profiles 1.437 0.22 - 0.32 0.27 0.025 0.450 1.60 2.71 4.60 35 40
Column CH Welded tubes 0.867 0.133 - 0.49 0.36 0.025 0.450 1.95 3.31 5.61 35 40
Column CH Extruded tubes 0.867 0.133 - 0.49 0.36 0.025 0.450 1.95 3.31 5.61 35 40
Column CFS Cold formed profiles 1.62 0.25 - 0.55 0.42 0.025 0.430 6.78 11.52 19.45 35 40
Column HE Rolled profiles 1.435 0.22 0.43 0.32 0.27 0.025 0.450 0.63 1.07 1.81 35 40
Column CH Welded tubes 0.867 0.133 0.145 0.49 0.36 0.025 0.450 1.95 3.31 5.61 35 40
Column CH Extruded tubes 0.867 0.133 0.145 0.49 0.36 0.025 0.450 1.95 3.31 5.61 35 40
Beam IPE/HE Rolled profiles 1.32 0.4 - 0.16 0.29 0.025 0.400 2.40 4.07 6.91 35 40
Beam CH Rolled profile trussed beam 1.28 0.63 - 0.23 0.34 0.025 0.440 1.76 2.98 5.06 35 40
Beam CFS Cold formed trussed beam 1.28 0.63 - 0.23 0.34 0.025 0.440 7.46 12.67 21.40 35 40
Brace HE/UPN Rolled profiles 0.91 0.4 - 0.14 0.27 0.025 0.400 1.60 2.71 4.60 35 40
Table 1.5. Costs of CB elements for German market

Treatment for fire

Bare steel Transportation (from
Production Foundation Surface treatment resistance
cost, Cut-offs, the shop to the field) Erection
Assembling €/m2 (plan) €/kg [intumescent
Country

plates, bolts max 400 km

paints] €/kg
Structural type
A
B
average price Sandblast SA2.5
€/kg €/kg €/kg Pinned Fixed Hot-deep R60 R90 R120
€/kg/(100 km) primer 40 Pm
galvanizing
thickness
Frame with profile sections 0.85 0.44 0.03 0.32 25 30 0.12 0.23 0.75 1.20 1.66
Frame with welded sections 0.89 0.49 0.03 0.3 25 30 0.12 0.23 0.75 1.20 1.66
Italy

Frame with truss girders 0.83 0.56 0.03 0.38 25 30 0.14 0.34 0.83 1.33 1.82
Frame with cold formed profile 0.95 0.46 0.03 0.28 25 30 0.14 0.34 3.52 5.61 7.70
Frame with profile sections 0.8 0.6 0.005 0.243 5 10 0.10 0.20 0.64 1.02 1.40
Romania

Frame with welded sections 0.8 0.7 0.005 0.243 5 10 0.10 0.20 0.64 1.02 1.40
Frame with truss girders 0.75 0.75 0.005 0.28 5 10 0.17 0.25 0.70 1.12 1.54
Frame with cold formed profile 0.78 0.6 0.005 0.243 5 10 0.17 0.25 2.97 4.73 6.50
Frame with profile sections 1.50 0.52 0.03 0.46 35 40 0.32 0.27 1.60 2.71 4.60
Germany

Frame with welded sections 1.51 0.53 0.03 0.49 35 40 0.32 0.27 1.60 2.71 4.60
Frame with truss girders 1.40 0.81 0.03 0.52 35 40 0.55 0.42 1.76 2.98 5.06
Frame with cold formed profile 1.62 0.62 0.03 0.49 35 40 0.55 0.42 7.46 12.67 21.40
Table 1.6. Costs IB modular structures for the three different markets

Looking at the results obtained from the three markets, it is clear that the prices referred to Romanian market
(assumed as representative of Eastern Europe) were the most convenient while more expensive is, as
expected, the German market (Figures 1.40 and 1.41).
1.8 1.80 German Market Romanian market Italian market
Cost [€/kg]

Cost [€/kg]

German Market
Romanian market 1.60
1.6
Italian market 1.40
1.4 1.20
1.2 1.00
0.80
1 0.60
0.8 0.40
0.20
0.6 0.00
0.4
Frame with cold
Frame with truss
Frame with profile

Frame with welded

formed profile

0.2
girders
sections

sections

0
HE

HE

HE/UPN
CFS

CFS
IPE/HE
CH

CH

CH

CH

CH

(a) (b)
Figure 1.40. Comparisons of costs between different structural systems in different markets: (a) CB
elements; (b) IB structural types – material costs

39
 0.7 German Market Romanian market Italian market 0.90 German Market Romanian market Italian market
Cost [€/kg]

Cost [€/kg]
0.80
0.6 0.70
0.60
0.5 0.50
0.40
0.4
0.30
0.3 0.20
0.10
0.2 0.00

Frame with cold

Frame with truss
Frame with profile

Frame with welded

formed profile
0.1

girders
sections

sections
0
HE

HE

HE/UPN
CFS

CFS
IPE/HE
CH

CH

CH

CH

CH
(a) (b)
Figure 1.41. Comparisons of costs between different structural systems in different markets: (a) CB
elements; (b) IB structural types – production and assembling of structural elements.

Data concerning the structure considered also the costs of floors, roof and connections, of industrial and
commercial buildings;. In particular, the economic building performance was evaluated by considering
structural weight, costs of materials, cost of transportation and cost of workmanship necessary for the
assembling of the planar elements as roofing, flooring and cladding systems (as reported in the table 1.7 and
table 1.8, as examples). The unitary costs, expressed as €/m2, were referred to the first months of the year
2010 and should be considered as indicative; in fact they were very variable in time owing to material prices
that are very sensitive to global developing dynamics.
It is worth noticing that regular spacing L for roofing and flooring systems was chosen in the range of 2÷4 m
depending on the primary beam length, in order to enable unpropped construction of the slab that can be
either composite with profiled steel sheeting or constructed on precast predalle panels (considering the same
self weight of 2.85 kN/m2).
The cost analysis and the model defined for estimating the cost of complex structural members as trussed
girders, CFCHS, PEHE, etc. etc. have been applied to all the structural configurations and the structural
elements considered in the database of the PRECASTEEL software; in particular, every element inserted in
the database has been characterized by a proper pricing model in which base material, assembling, additional
treatments and transportations have been singularly referred. According to this information, the optimization
in terms of costs can be made by the designer or by the user in general, choosing between different solutions
able to satisfy the minimum structural safety requirements but characterized by different unit and total costs.
Moreover, it is also important to remind that the cost of the external envelope of the structural solutions will
be considered in the total costs, prices referred to elements with large surfaces as floors, roofs and claddings
will be characterized in terms of € per m2.

40
Subtype Thickness Width Length Element Assembling Transportation
s [m] B [m] L [m] [€/m2] [€/m2] [€/(m2·km)]
0.04 + 0.07 + 1.20 /
Precast Predalle Panel 0.05 2.40 2.00 41.00 10.00 0.02
0.04 + 0.07 + 1.20 /
0.05 2.40 2.50 41.00 10.00 0.02
0.04 + 0.07 + 1.20 /
0.05 2.40 2.67 41.00 10.00 0.02
0.04 + 0.07 + 1.20 /
0.05 2.40 3.00 41.00 10.00 0.02
0.04 + 0.07 + 1.20 /
0.05 2.40 3.33 41.00 10.00 0.02
0.04 + 0.07 + 1.20 /
0.05 2.40 3.50 41.00 10.00 0.02
0.04 + 0.07 + 1.20 /
0.05 2.40 4.00 41.00 10.00 0.02
0.04 + 0.07 + 1.20 /
0.05 2.40 2.00 46.00 10.00 0.02
0.04 + 0.07 + 1.20 /
0.05 2.40 2.50 46.00 10.00 0.02
0.04 + 0.07 + 1.20 /
0.05 2.40 2.67 46.00 10.00 0.02
0.04 + 0.07 + 1.20 /
0.05 2.40 3.00 46.00 10.00 0.02
0.04 + 0.07 + 1.20 /
0.05 2.40 3.33 46.00 10.00 0.02
0.04 + 0.07 + 1.20 /
0.05 2.40 3.50 46.00 10.00 0.02
0.04 + 0.07 + 1.20 /
0.05 2.40 4.00 46.00 10.00 0.02
Composite Steel-
Concrete 0.16 1.00 2.00 41.00 10.00 0.02
0.16 1.00 2.50 41.00 10.00 0.02
0.16 1.00 3.00 41.00 10.00 0.02
0.16 1.00 3.50 41.00 10.00 0.02
0.16 1.00 4.00 41.00 10.00 0.02
0.16 1.00 2.00 46.00 10.00 0.02
0.16 1.00 2.50 46.00 10.00 0.02
0.16 1.00 3.00 46.00 10.00 0.02
0.16 1.00 3.50 46.00 10.00 0.02
0.16 1.00 4.00 46.00 10.00 0.02
0.16 1.00 2.00 49.00 10.00 0.02
0.16 1.00 2.50 49.00 10.00 0.02
0.16 1.00 3.00 49.00 10.00 0.02
0.16 1.00 3.50 49.00 10.00 0.02
0.16 1.00 4.00 49.00 10.00 0.02
Table 1.7. Abstract of data collection about some costs of roofing and flooring systems.

41
Subtype Thickness Width Height Element Assembling Transportation
s [m] B [m] H [m] [€/m2] [€/m2] [€/(m2·km)]
Monolithic Flat 2.00 ÷ 5.00 ÷
Panel 0.12 7.50 15.00 52.00 10.00 0.02
2.00 ÷ 5.00 ÷
0.16 7.50 15.00 65.00 10.00 0.02
2.00 ÷ 5.00 ÷
0.18 7.50 15.00 72.00 10.00 0.02
2.00 ÷ 5.00 ÷
0.20 7.50 15.00 75.00 10.00 0.02
2.00 ÷ 5.00 ÷
0.22 7.50 15.00 85.00 10.00 0.02
2.00 ÷ 5.00 ÷
0.25 7.50 15.00 91.00 10.00 0.02
2.00 ÷ 5.00 ÷
0.30 7.50 15.00 104.00 10.00 0.02
Monolithic Ribbed 0.15 + 0.15
Panel (rib) 2.00 ≤ 15.00 78.00 10.00 0.02
0.20 + 0.15
(rib) 2.00 ≤ 15.00 85.00 10.00 0.02
0.25 + 0.15
(rib) 2.00 ≤ 15.00 92.00 10.00 0.02
Table 1.8. Abstract of data collection – some costs of cladding systems.

42
 2. Industrial buildings
In the present chapter the design of IB types is presented; in particular, the selection of geometrical data and
structural configuration are presented using the main results coming from the preliminary statistical analyses.
After the individuation of relevant structural types, the design process and the optimization procedure are
shortly presented focusing the attention to most interesting aspects. Moreover, some design issues
concerning the definition of the structural connections are reported making reference particular or interesting
types. All structural systems, connections and secondary members herein designed or presented were
inserted in the Precasteel database integrated with the web-application Precasteel 2.0.

2.1. General description of selected structural types

The preliminary design at a series of industrial buildings, that cover a wide range of parameters was based on
the findings of the statistical analysis presented extensively in the previous chapter. The geometric
parameters were based on the statistical analysis and concern the number of bays, the longitudinal direction
of the structure, the span length, the column height, the resisting system in both longitudinal and transverse
direction. Practical considerations were made in order to reduce the extremely large number of all possible
combinations between the various parameters.
At this stage the structural analysis was linear elastic, and the response spectrum analysis was adopted for the
seismic design. The procedure follows EC rules and, in particular, EC3 provisions for the general steel
design checks and verifications, and EC8 provisions as far as seismic design is concerned. Both Ultimate
Limit States and Serviceability Limit States are considered.

2.1.1. Structural conceiving and design assumptions

In this preliminary design stage, cross-sections for the main structural members are determined, so that total
weight of steel used in each structure is determined. Based on the amount of steel, the construction cost is
estimated. Furthermore, the cost of workmanship, as well as the cost of transportation is estimated, so that
we arrive at a final cost for each structural case.
The proposed module for the industrial buildings consists of five single/double span frames which are
repeated in the out-of-plane direction having a constant distance between them. In the in-plane direction the
Moment Resisting Frames withstand the horizontal actions due to wind of earthquakes. In the out-of-plane
direction a Bracing System for every five consecutive frames is responsible to withstand all the horizontal
forces. The configuration described above is considered as the basic module which by repeating can form an
industrial building as shown in the figure 2.1.
For the Moment Resisting Frames in High and Medium Seismicity areas, hot-rolled sections are selected. For
Low Seismicity areas, cold-formed sections could possibly be also used. The typical section category for
columns is HEA and for beams is IPE. Steel material S275 is used in the design. Currently, steel S235 is also
used for industrial structures, but steel S275 is more likely to be used for the majority of industrial buildings
in the years to come. Columns are considered as fixed at their base in the in-plane direction in most cases.
Hinged column bases are also taken into account only for low seismicity areas. In the out-of-plane direction
all column bases are considered hinged.

Figure2.1: Basic module for single and double bay industrial building configuration

According to the statistical analysis for the Industrial Buildings, the preliminary design of the selected
combinations will be executed by varying the chosen parameters within the limits described in the following.
‚ Number of bays: The number of bays that are most commonly used in the industrial buildings is 1 or 2
frame bays. These values cover the 85% of the total structures that were taken into consideration in WP1
statistical analysis.

43
‚ Span of bay: According to the distribution of values that resulted from the statistical analysis, the most
common span lengths fall between 20-25 m. It is decided that the span values to be used in the parametric
analysis are 16 m, 20 m, 24 m, 27 m, 30 m and 32 m, thus covering a wide range of building solutions.
‚ Frame distance: In the out-of-plane direction of most industrial buildings a typical main frame is repeated
in regular intervals. According to the statistical analysis, the most commonly adopted value for the distance
between consecutive frames in the out-of-plane direction is 6 m.
‚ Height of frame columns: The two typical values for the column height that serve industrial activities of
any kind are 6m and 8m. Heights above 8 m are considered rather special or exceptional.
‚ Slope of girder: The most common value used for the realization of the roofing slope is 15% which
corresponds to 8.53°. This value covers most of the cases of industrial buildings across Europe and is used in
the parametric study.
‚ Type of girder: Both beams and trusses for the horizontal elements (girder) of the moment resisting frame
are considered for the preliminary design of the selected cases. Following current design practice, truss
girders are used for long spans.
‚ In-plane resistance: The in-plane structural resisting system that has been chosen is the Moment Resisting
Frame. This system provides the necessary free space for the industrial activities.
‚ Out-of-plane resistance: The out-of-plane structural resisting system is based on bracing. In this work the
Concentric Bracing System has been chosen. This system is the most preferred among the other possible
solutions (MR, EB) due to its simplicity in fabrication.
‚ Actions on industrial structures
The definition of the loading values is described in the following Table 2.1:

Seismicity (p.g.a.) High Medium Low

0.32 g 0.16 g 0.08 g
Snow loads High Low
1.5 kN/m2 0.75 kN/m2
Wind load
Average EU reference velocity value: 30 m/s
Crane loads
0 kN 50 kN 100 kN 250 kN
Cladding
0.15 kN/m2
Table 2.1. Assumed design value of acting loads

2.1.1.1. Design Assumptions

The analysis and design of all cases has been executed with the CCS INSTANT structural analysis software.
During this analysis the following assumptions were taken into consideration:
‚ General Structure geometry: The modular structure consists of 6 sequential frames. The distance between
these frames was taken equal to 6 m. Diagonal bracing system is used to provide stiffness to the structure in
the out-of-plane direction. This bracing was placed at the middle of the structure in the vertical elements as
well as in the roof.
‚ Steel grade: The S 275 steel grade was used as the preselected steel grade for all the structural elements.
‚ Cross-sections used: HEA profiles were used for the columns of the structure, IPE profiles were used for
the beams of the structure, CHS profiles were used for the bracing system of the structure, IPE 120 was used
as a typical cross-section for all the purlins that support the roof cladding. At this stage no purlin design was
conducted. UPN 80 was used as a typical cross-section for all the purlins that support the side cladding. At
this stage no purlin design was conducted.
‚ Haunch selection: Haunches were used to strengthen the beam cross-section only at the points where the
beams were connected with the columns of the structure. No joint design was executed at this stage. The
height of the haunch cross-section at the end of the beam element was considered as 1.5 – 2 times the height
of the beam cross-section that was used for the rest part of the beam. The other cross-sectional characteristics
(tf, b, bw etc.) were taken same as the characteristics of the steel profile used for the beam. The length of the
haunch was taken between L/5 – L/6, where L is the length of the half frame opening in the in-plane
direction.

44
‚ Response spectrum parameters: The ground acceleration was taken equal to 0.08 g, 0.16 g and 0.32 g
depending on the case under examination. Ground type B was considered in all cases during the analysis. All
the buildings analyzed were considered as ordinary buildings (importance class factor II). Type 1 elastic
response spectrum was used according to EC8. In all cases the q factor was taken equal to 1.5. Moreover the
vertical component of the seismic action was neglected.

2.2. Set of design structures

According to the above selected parameters and their varying range, the parameters sets for conducting the
preliminary analyses are selected as follows:
‚ the IDs references for the trussed elements and for the elements of secondary frame are represented in the
figure 2.2.

(a) (b)
Figure 2.2. (a) Truss girder scheme adopted in the preliminary design, (b) Beam that connects the heads of
the columns in the longitudinal direction

‚ the execution of structural design of IBs was carried out fixing a variation range of geometrical parameters
and load level according to the evidences came out from statistical analysis; in particular, the value of
geometrical variables and load levels are reported in the charts presented in the figures 2.3-2.8. In these
figures the value of the steel consumption is reported to different parameters as span length, column height,
PGA of the seismic action, wind load and crane load. The charts are the condensed representation of all
structures assumed as case studies and preliminary designed.

45.00 60.00

40.00
50.00
Steel consumption (kg/m )

Steel consumption (kg/m )

2

35.00 Single span

Single span
Snow: 0.75 kN/m2
30.00 Height: 6 m 40.00 Snow: 1.5 kN/m2
Height: 6 m
25.00
30.00
20.00

15.00 0.32 g 20.00 0.32 g

0.16 g 0.16 g
10.00
0.08 g
10.00
5.00

0.00 0.00
16 20 24 27 30 32 16 20 24 27 30 32
Span length (m) Span length (m)

Figure 2.3. Comparison between assumed span length and steel consumption with variable snow loads and
variable seismic action (single span)

45
 35.00 45.00

40.00
30.00

Steel consumption (kg/m )

Steel consumption (kg/m )

2
2

Double span 35.00

Double span
25.00 Snow: 0.75 kN/m2
Height: 6 m 30.00 Snow: 1.5 kN/m2
Height: 6 m
20.00 25.00

15.00 20.00

0.32 g 15.00 0.32 g

10.00 0.16 g 0.16 g
10.00
0.08 g
5.00
5.00

0.00 0.00
16 20 24 16 20 24
Span length (m) Span length (m)

Figure 2.4. Comparison between assumed span length and steel consumption with variable snow loads and
variable seismic action (double span)

80.00 80.00

70.00 70.00

Steel consumption (kg/m )

Steel consumption (kg/m )

2
2

Single span
60.00 60.00 Single span
Snow: 0.75 kN/m2
Height: 6 m Snow: 1.5 kN/m2
50.00 50.00 Height: 6 m
Crane: 100 kN
Crane: 100 kN
40.00 40.00

30.00 30.00
0.32 g 0.32 g
0.16 g 0.16 g
20.00 20.00
0.08 g
10.00 10.00

0.00 0.00
16 20 24 16 20 24
Span length (m) Span length (m)

Figure 2.5. Comparison between assumed span length and steel consumption with variable snow loads,
variable seismic action and crane load influence (single span)

60.00 50.00

45.00
50.00
Steel consumption (kg/m2)
Steel consumption (kg/m2)

Single span 40.00 Single span

Snow: 1.5 kN/m2 35.00 Snow: 1.5 kN/m2
40.00 Seismicity: 0.08g Seismicity: 0.32g
Height: 6 m 30.00 Height: 6 m

30.00 25.00

20.00
fixed base fixed base
20.00
pinned base 15.00 pinned base
10.00
10.00
5.00

0.00 0.00
16 20 24 27 30 32 16 20
Span length (m) Span length (m)

Figure 2.6. Influence of fixing condition at the column bases and PGA on steel consumption

45.00 60.00

40.00 Single span Single span

Snow: 0.75 kN/m2 50.00 Snow: 1.5 kN/m2
Steel consumption (Kg/m2)
Steel consumption (kg/m2 )

35.00 Span length: 32 m Span length: 32 m

Height: 6 m Height: 6 m
30.00 40.00

25.00
30.00
20.00 beam beam
truss truss
15.00 20.00

10.00
10.00
5.00

0.00 0.00
0.32 g 0.16 g 0.32 g 0.16 g 0.08 g

Figure 2.7. Direct comparison between steel consumption and PGA

46
 70.00

60.00

Steel consumption (Kg/m2)

Single span
50.00 Snow: 1.5 kN/m2
Span length: 20 m
Height: 6 m
40.00

30.00
0.16 g
20.00 0.08 g

10.00

0.00
50 kN 100 kN 250 kN

Overhead travelling crane (kN)

Figure 2.8. Influence of crane load on steel consumption

2.2.1 Light gauge steel solutions

Five additional cases were examined adopting light gauge steel profile solutions. The goal of this extra case
consideration was to examined the applicability of light gauge steel solutions to spans 16 m and 20 m.
Therefore the heavy snow loading was adopted as the most critical loading condition.
The additional five cases consist of three cases with 16 m span length and high snow loading (1.5 kN/m 2)
and two cases with 20 m span length and low snow loading (0.75 kN/m2), see table 2.2. Cases with 20 m
span length and high snow loading did not satisfy the strength requirements. All the above cases are directly
comparable to the reciprocal originally considered cases.
Base Constraint (in-plane)

ground

Overhead traveling crane

Snow loading (in kN/m2)
Height of frame columns

Resistance out-of-plane
Number of frame bays

(1 brace per 5 frames)

Slope of frame girder

Wind load intensity

Resistance in-plane
Span of each bay

Base Constraint
Frame distance

Type of girder

(out-of-plane)

acceleration
Maximum

4.1 1 16 m 6 m 6 m 15 % Beam MR CB F H 1.5 0.32 g 30 m/s 0

4.2 1 16 m 6 m 6 m 15 % Beam MR CB F H 1.5 0.16 g 30 m/s 0
4.3 1 16 m 6 m 6 m 15 % Beam MR CB F H 1.5 0.08 g 30 m/s 0
4.4 1 20 m 6 m 6 m 15 % Beam MR CB F H 0.75 0.32 g 30 m/s 0
4.5 1 20 m 6 m 6 m 15 % Beam MR CB F H 0.75 0.16 g 30 m/s 0
Table 2.2. Parameters assumed for designing light gauge steel solutions

The FeE 350G steel grade was used for all the structural elements. All the profiles used belonged to the
KONTI 2B and KONTI C library (see figure 2.9). Haunched solutions were not adopted.

KONTI 2B KONTI C

Figure 2.9. Type of cold formed sections considered for the members sizing

47
2.2.1.1 Analysis Results
The design of light gauge steel solutions was executed adopting the same assumptions and hypothesis
employed for the hot-rolled solutions. The results are presented in the table 2.3.

Cases
4.1 4.2 4.3 4.4 4.5
Columns KO-2KB 600X50 KO-2KB 600X50 KO-2KB 600X50 KO-2KB 550X50 KO-2KB 550X50
Beams KO-2KB 550X50 KO-2KB 550X50 KO-2KB 550X50 KO-2KB 550X50 KO-2KB 550X50
K KO-2KB 250X15 KO-2KB 250X15 KO-2KB 250X15 KO-2KB 250X15 KO-2KB 250X15
Vertical Braces KO-2KB 250X15 KO-2KB 250X15 KO-C175-2.5 KO-2KB 250X15 KO-2KB 250X15
Horizontal Br KO-C205-2.5 KO-C175-1.8 KO-C140-2.0 KO-C205-2.5 KO-C175-1.8
Purlins (roof) KO-2KB 250X20 KO-2KB 250X20 KO-2KB 250X20 KO-2KB 250X15 KO-2KB 250X15
Purlins (side) KO-2KB 250X20 KO-2KB 250X20 KO-2KB 250X20 KO-2KB 250X15 KO-2KB 250X15
Table 2.3. Profiles size obtained from structural design

Cases with cold-formed profiles gave a 5% average weight reduction compared with the hot-rolled profile
solutions. A detailed steel consumption analysis is depicted in the figure 2.10.

Light gauge steel solutions

40.00
39.50
Single span
Steel consumption (kg/m 2)

39.00 Span length: 16 m

38.50 Snow: 1.5 kN/m2
38.00
hot-rolled
37.50
cold-formed
37.00
36.50
36.00
35.50
35.00
34.50
0.32 g 0.16 g 0.08 g
Case
Figure 2.10. Comparison between hot-rolled and cold-formed solutions

2.2.2 Tapered solutions

In addition to the light gauge steel solutions, four extra cases were also examined adopting tapered column-
beam sections. The tapered section cases had 32 m span length and hinged in-plane column bases. Two
different haunched sections were adopted for the column sections.
Case 5.1: The column’s cross-section height varied from 225 to 900 mm. The beam’s cross-section height
varied from 900 to 400 mm in the haunch and then a built-up section of 400 mm height was used. This
section did not pass the capacity check.
Case 5.2: The column’s cross-section height varied from 275 to 1100 mm. The beam’s cross-section height
varied from 1100 to 500 mm in the haunch and then a built-up section of 500 mm height was used.
Case 5.3, 5.4: An alternative to the above cases was the use of the tapered column sections (225 to 900 mm
and 275 to 1100 mm) in combination with the original 3.25 case rafter’s cross-sections.
In all cases, the built-up cross-section width was 380 mm, the flange thickness was 14 mm and the web
thickness was 12 mm. All cases examined are presented in the table 2.4.
In the following graph (figure 2.11), the steel consumption analysis for the tapered solutions is depicted. An
8% average weight reduction compared with standard profiles solutions is observed. In comparison with case
3.25, the tapered or tapered-standard profiles combination is an effective alternative for the 32 m span length.

48
 Cases hc (mm) hh (mm) hb (mm) b (mm) tf (mm) tw (mm)
5.1 225 900 400 380 14 12
5.2 275 1100 500 380 14 12
5.3 225 900 Haunch IPE 600 * IPE 600 IPE 600 IPE 600
5.4 275 1100 Haunch IPE 600 * IPE 600 IPE 600 IPE 600
Table 2.4. Geometrical parameters of tapered solutions with 32 m spans. Note: Haunch IPE 600 * refer to the
haunched section used for original cases as described in paragraph 6.1

Tapered solutions
60
Steel consumption (kg/m2)

50

40 Single span
Span length: 32m
Height: 6 m
30
Snow: 1.5 kN/m2
Seismicity: 0.08g
20

10

0
3.25 5.2 5.3 5.4
Case
Figure 2.11. Steel consumption for all designed tapered cases with 32 m span.

2.3 Selection of case studies

Taking into account the preliminary design and the cost analysis of all presented cases, executed on the basis
of cost model, 7 cases were selected as case studies. The definition of the structures that present better
performances was based also on the conducted statistical analysis. The complete list of the selected cases is
reported in the table 2.5.
(out-
Base Constraint (in-plane)

ground
Resistance out-of-plane (1

Overhead travelling crane

Snow loading (in kN/m2)
Height of frame columns
Original Case Number
Number of frame bays

Fixed (F), Hinged (H)

Slope of frame girder

Wind load intensity

Resistance in-plane

brace per 5 frames)

Span of each bay

Base Constraint
Frame distance

Type of girder

acceleration
Profile type

Maximum
of-plane)

1. 1.2 1 20 m 6 m 6 m 15 % HR Beam MR CB F H 0.75 0.32 g 30 m/s 0

2. 2.34 2 20 m 6 m 6 m 15 % HR Beam MR CB F H 1.5 0.16 g 30 m/s 0
3. 3.7 1 32 m 6 m 6 m 15 % HR Truss MR CB F H 1.5 0.08 g 30 m/s 0
4. 2.12 1 24 m 6 m 8 m 15 % HR Beam MR CB F H 0.75 0.16 g 30 m/s 100 kN
5. 3.20 1 16 m 6 m 6 m 15 % HR Beam MR CB H H 1.5 0.08 g 30 m/s 0
6. 4.3 1 16 m 6 m 6 m 15 % LGS Beam MR CB F H 1.5 0.08 g 30 m/s 0
7. 5.3 1 32 m 6 m 6 m 15 % HR Beam MR CB H H 1.5 0.08 g 30 m/s 0
Table 2.5. Geometric parameters and loads considered in the design of 32m tapered solutions.

49
2.4 Purlin design
In order to integrate the preliminary analysis results with the designed purlin member, a simplified purlin
design analysis was conducted. In the purlin design it was assumed that: 1) The purlins are simply supported
on the main frame girders. This will also facilitate the erection phase. 2) The cladding stiffness is not taken
into account due to uncertainties on the way it might be connected on the purlins and due to the wide variety
of cladding profiles and corresponding stiffness. It is believed though that if the cladding stiffness is taken
into account, the resulting purling cross sections will be smaller.
Based on the above assumptions, that the roof hot rolled purlin profile suitable for the cases with low snow
loading is IPE 140, while for the cases with high snow loading is IPE 180. For the support of the side
cladding IPE 120 profiles can be used.
LGS purlin solutions were also examined. The analysis results show that the appropriate LGS profile for
cases with high snow loading is a double back-to-back C profile with general geometry of 250x157x20 mm,
while for cases with low snow loading the proposed profile of the same type has dimensions 250x156x15
mm. In the present analysis the Konti 2KB 250x15 and 2KB 250x20 profiles were used.
Based on the above results, all the reported tables and graphs have been updated before their implementation
to the developed software database.

2.5 Optimization of the double-span cases design

The analysis results for the double-span cases have been verified with the use of nonlinear Pushover analysis.
All the analyses have been conducted with the use of SAP2000 structural analysis software with the use of
3D beam elements. For each examined case, a single frame has been derived from the initial geometry. The
roof and side purlins have been also implemented in the model as shown in figure 2.12. In those cases where
crane supporting girders are used, the corresponding elements have been also added to the developed model.
In all cases, the appropriate boundary conditions of the elements have been introduced with respect to the
symmetry of the structure in the out-of-plane direction.
For the verification of the results derived from the initial analysis conducted in INSTANT structural analysis
software, the same geometry and loading conditions have been introduced for each case examined. Before
proceeding with the nonlinear analysis, a design check according to EC3 has been conducted and both the
Ultimate Limit State (ULS) and the Serviceability Limit State (SLS) have been checked. The results of the
initial structural analysis have been verified and all the cases examined passed the all EC3 checks.
Additionally, the prescribed initial bow and sway imperfections implied by the EC3 have been introduced in
terms of lateral loads applied to the structure. Their contribution to the resulting stress state of the elements
has been taken into consideration for the development of the horizontal and vertical pushover curves. The
pushover curves have been derived by taking into account the geometric nonlinearities throughout the
analysis (P-Delta effect or P-Delta plus Large Displacements effects). Finally, for all the analyses conducted,
the purlins are assumed as pinned to the main frame beam and no purlin bracing system is adopted.
The pushover curves have been produced with the use of plastic hinges appropriately positioned on the
structural elements. Two types of hinges have been adopted. The first type is a bending hinge that takes into
consideration both the in-plane and out-of-plane bending of the element it is applied on (interacting M2-M3).
Three bending hinges are assigned at the beginning, the middle and the end of every structural element. They
are only activated if the developed moment exceeds the automatically calculated yield moment. Therefore,
their increased number does not increase the computational time. In addition to the bending hinges, axial
load hinges are assigned at the middle of each structural element.

(b)

(a)
Figure 2.12. (a) The developed model in SAP2000, (b) Imperfection load application - Deformed shape

50
The resulting pushover curves show the behaviour of each case examined when it is loaded with the
corresponding horizontal or the vertical loading pattern. In many cases only one is the prevailing and critical
loading pattern which governs the whole design of the structure. In other cases though, the design is not
clearly governed by the horizontal or the vertical loading pattern as both of them lead the structure to an
almost critical state. The ULS (1.35G+1.5Q) and EQ (G+0.3Q+E) loading combinations resulting force
levels are depicted in every graph in order to facilitate the verification of the secure design of each case.
Illustrative pushover graphs for horizontal and vertical loading are presented in the figure 2.13.

(a) (b)
Figure 2.13. (a) PushOver for seismic performance assessment; (b) loading of IB structure with static loads
at ULS – comparison between demand and capacity

The developed pushover curves verify that all the cases have been designed safely in the preliminary level.
Some of them are optimally designed while other seem somewhat conservative. This fact is attributed to the
choice of the cross-section families used in the design of the structures (HEA sections for the column
elements and IPE sections for the beam elements) and the variety of cross-section sizes each family has. In
all cases examined, the selection of the cross-section size is optimized for the given cross-section families.

2.6 Optimization of single-span cases design

Design and optimization procedures were developed in order to analyze HR frames and WT frames. Due to
the sophistication of the design of LGS structures, they were not addressed with the formal, genetic
algorithm based, optimization procedures associated to EV-tool presented in the chapter 5.
HR frames, figure 2.14.a, with haunches on rafters are typically built with a fixed base in earthquake areas,
in order to decreases the overall steel consumption, but with the penalty of the increased demands on
foundations. On the other hand WT frames are typically pinned at the base, figure 2.14.b; with the variation
of column and rafter heights made to agree with the bending moment distribution and allowing for weight
savings. Fabrication of WT frames is usually more difficult and expensive, even impossible if proper
welding technology is not available to the fabricator.
The performance comparison of the two configurations can only be made after a thorough economic study.
However, even the results of such study are influenced by the economic environment (e.g. labour cost vs.
material cost), and the equipment which the fabricator uses. Therefore, the question “which of the two frame
configurations is more economical?” can only be answered for a particular fabricator operating in a particular
market.

51
 D D

H
H

S/2 S/2

(a) (b)
Figure 2.14. (a) Hot-rolled (HR) frames; (b) Welded tapered (WT) frames

Transversal
Middle Longitudinal Purling beam
Case crane Column profile Beam profile
Column profile beam profile profile
loading
1.12 0 kN HEA 260 HEA 280 IPE 300 IPE 160 IPE 140 each 2 m
1.13 0 kN HEA 320 HEA 320 IPE 360 IPE 160 IPE 140 each 2 m
1.14 0 kN HEA 400 HEA 400 IPE 400 IPE 180 IPE 140 each 2 m
1.15 100 kN HEA 400 HEA 450 IPE 400 IPE 300 IPE 140 each 2 m
1.27 0 kN HEA 320 HEA 340 IPE 360 IPE 200 IPE 180 each 2 m
1.28 0 kN HEA 400 HEA 400 IPE 450 IPE 200 IPE 180 each 2 m
1.29 0 kN HEA 500 HEA 500 IPE 500 IPE 200 IPE 180 each 2 m
1.30 100 kN HEA 550 HEA 550 IPE 500 IPE 360 IPE 180 each 2 m
2.14 0 kN HEA 260 HEA 280 IPE 300 IPE 160 IPE 140 each 2 m
2.15 0 kN HEA 300 HEA 300 IPE 360 IPE 160 IPE 140 each 2 m
2.16 0 kN HEA 360 HEA 360 IPE 400 IPE 160 IPE 140 each 2 m
2.17 50 kN HEA 320 HEA 320 IPE 360 IPE 240 IPE 140 each 2 m
2.18 100 kN HEA 360 HEA 340 IPE 400 IPE 240 IPE 140 each 2 m
2.19 250 kN HEA 400 HEA 550 IPE 400 IPE 400 IPE 140 each 2 m
2.33 0 kN HEA 320 HEA 340 IPE 360 IPE 200 IPE 180 each 2 m
2.34 0 kN HEA 400 HEA 400 IPE 450 IPE 200 IPE 180 each 2 m
2.35 0 kN HEA 500 HEA 500 IPE 500 IPE 200 IPE 180 each 2 m
2.36 50 kN HEA 400 HEA 400 IPE 450 IPE 240 IPE 180 each 2 m
2.37 100 kN HEA 450 HEA 450 IPE 500 IPE 240 IPE 180 each 2 m
2.38 250 kN HEA 450 HEA 550 IPE 450 IPE 450 IPE 180 each 2 m
3.14 0 kN HEA 320 HEA 340 IPE 360 IPE 200 IPE 180 each 2 m
3.15 0 kN HEA 400 HEA 400 IPE 450 IPE 200 IPE 180 each 2 m
3.16 0 kN HEA 500 HEA 500 IPE 500 IPE 200 IPE 180 each 2 m
3.17 50 kN HEA 400 HEA 400 IPE 450 IPE 200 IPE 180 each 2 m
3.18 100 kN HEA 400 HEA 400 IPE 500 IPE 220 IPE 180 each 2 m
3.19 250 kN HEA 400 HEA 550 IPE 450 IPE 300 IPE 180 each 2 m
3.26 0 kN HEA 340 HEA 340 IPE 360 IPE 200 IPE 180 each 2 m
3.27 0 kN HEA 450 HEA 450 IPE 450 IPE 200 IPE 180 each 2 m
3.28 0 kN HEA 550 HEA 550 IPE 550 IPE 200 IPE 180 each 2 m
Table 2.6. The final selection of cross-sections for all the double-span cases examined.

The use of the two back-to back C shaped profiles as beams and columns were investigated for LGS frames.
The members in this case were connected by welded corner pieces, configured so that they do not fail before
members, figure 2.15.

52
 (b)

(a)
(c)
Figure 2.15. LGS frame with back-to-back C elements and corner fixings

From among the frame geometries the lowest spans (S=16, 20m) were used, but both 0.75 kN/m 2 and 1.5
kN/m2 snow load has been considered. The initial shapes of the LGS profiles, figure 2.16, were simplified to
a simple C (Figureb) but maintaining the overall dimensions. The yield stress was considered fy = 350
N/mm2, a typical value for most LGS steel profiles.

C1 C1

t
t
B

C2 C3 C4 C3 C2
H H
(a) (b)
Figure 2.16. a) Initial profile; (b) and simplified shape for the frame analysis

Single story portal frames present design challenges especially due to buckling behavior, which is not easily
evaluated and incorporated into the design calculations. The most significant mode of buckling for portal
frames, especially for WT frames is Lateral Torsional Buckling (LTB – figure 2.17). LTB is difficult to
account for. In the project modelling techniques have been developed for analyzing portal frames with focus
on more precisely evaluation of the buckling behaviour.

Figure 2.17. Typical LTB failure of a portal frame structure

53
 2.6.1 Structural conceiving and design assumptions for optimization
The design methods used are divided in three levels according their sophistication, accuracy and time
required to perform them. All design methods conform to the Eurocodes and more specifically to EN 1993
and EN 1998. Scripts were developed under Microsoft Excel and Abaqus in order to carry out automatically
the finite element analysis (FEM), and design checks. These scripts have been integrated into genetic
algorithm based optimization procedures for HR and WT frames. The design methods were:
‚ Method 1: Global nonlinear analysis (GNLA) Straightforward numerical calculation by modelling
frames with shell type finite element (FE). The calculation is material and geometrically nonlinear, and it
takes into account initial imperfections.
‚ Method 2: General method (GMA) Linear or nonlinear in-plane analysis of the frame, with out-of-
plane stability being taken into account by a global reduction factor. Beam model with stepped cross-sections
(10 divisions) is used for the in-plane analysis, while shell based model is used for the out-of-plane effects.
‚ Method 3: EC3 interaction formulae (IFM) Linear in-plane analysis using cross-sectional checks
as limit states conditions. Frame modelled using beam type finite elements with stepped cross-sections (4
divisions).
‚ Global nonlinear analysis (GNLA) Geometrically and Materially Nonlinear Analysis on Imperfect
structure (GMNIA) is carried on the shell model, figure 2.18.a. The required initial imperfection is obtained
by dislocating the nodes of the mesh with values previously obtained from a buckling calculation of the same
FE model, figure 2.18.b.

(a) (b)
Figure 2.18. Steps of the Global non-linear analysis

GNLA starts with a buckling step in the Abaqus. The first positive eigenvalue is searched, and the
corresponding buckling shape is used as source for the initial imperfection. In the second step of the analysis
the loads from accompanying actions are placed on the model, in a static step. Accompanying loads can be:
‚ The wind load and the crane load, if the analysis is performed to determine the behaviour of the frame in
fundamental load combination. In this case it is expected, that the intensity of the accompanying actions is
much smaller then that of the leading load (i.e. snow).
‚ The reduced value of the snow load, which accompanies earthquake loads, if the aim of the analysis is to
determine the performance under earthquake loads. In this case snow load is a small fraction of the design
snow load, 0% or 20% according to EN 1990
The third step is a non-linear analysis the leading load is gradually increased until the frame collapses, while
the accompanying load is kept constant. Initial imperfections were scaled to correspond to EN 1993-1-1 bow
imperfections. Additional sway imperfections, conforming to EN 1993-1-1, are applied before the second
step.

2.6.1.1 General method analysis (GMA)

According to EN 1993-1-1, overall resistance of a structural component can be verified using following
condition:
F op ˜ D ult, k
t 1,0 [2.1]
J M1

54
Where: D ult,k - is the minimum load amplifier of the design loads to reach the characteristic resistance of the
most critical cross-section without taking out-of-plane buckling into account.
F op - is the reduction factor to take into account out-of-plane buckling.
J M 1 - is the safety factor.
The reduction factor F op can be calculated using EN 1993 buckling curves, and the global non-dimensional
slenderness for out-of-plane buckling, O op :
D ult,k
O op [2.2]
D cr ,op
where D cr,op is the minimum load amplifier of the design loads to reach the elastic critical resistance with
regards to lateral or lateral torsional buckling (LTB). In the GMA, the first step (LBA-z) uses the shell model
but only the critical load multiplier D cr,op is retained. The second and third steps of the GMA is carried out
on a beam model of the frame. In the second step the structure is preloaded with accompanying loads in a
static step, and in the third step the frame is gradually loaded with the leading action until failure. Before the
loading phase starts initial imperfections resulting from an in-plane buckling analysis of the wire model are
applied (LBA-y), taking into account in-plane sway effects.
The failure of the frame is defined depending on the cross-section classes of the members. If the frame is
made of Class 1 or Class 2 members, failure occurs at full plasticisation of a cross-section (i.e. forming of a
plastic hinge). Plastic hinges are identified by a rapid increase of equivalent plastic strain on the wire model
(plastic strain 5 times the elastic strain is used). If the frame has Class 3 members, then the first yielding
corresponds to failure.

2.6.1.2 Method using EC3 interaction formulae (IFM)

This method is the most used in the design practice, and it is based on static calculation of the structure using
beam elements in order to determine the internal forces. ULS design checks are then carried out on members
isolated from the structure (e.g. columns, beams, braces etc.) From the point of view of the member check,
the interaction with the rest of the structure is taken into account in a simplified way.
For the buckling check of members subjected to combined axial compression (N) and bending (M), EN1993-
1-1 is used. The calculation uses the second order effect amplification factor for slender frames, where the
in-plane critical multiplier D cr d 10 for elastic or D cr d 15 for plastic analysis. The calculation of the
critical multiplier ( D cr ) in more sophisticated then the one in EN 1993 taking into account the presence of
the axial force in the rafters using the formulas proposed by Davies.

2.6.2 Set of designed structures (Members and connections) WT frames

The study focuses on optimization of 96 cases of single-span or double-span WT frames. The different eaves
heights (6 m and 8 m) as well as different span lengths (varying from 16 to 32 m in case of single-span
frames and 16 to 24 m in case of double-span frames) were considered. The loading scenarios included
different characteristic snow loads (750 and 1500 N/m2) and seismic loads corresponding to different PGA
from 0,08g to 0,32g.
Each frame was calculated twice: (1) loaded with fundamental combination of dead, snow and reduced wind
load, and (2) seismic combination of accidental, dead and reduced snow load. Single-span frames were
loaded with uniform snow load and the results checked with non-symmetric snow, where only one half of
snow mass was applied on the downwind side. In case of double-span frames, the drifted snow load case was
used for design and the results were checked with uniform snow load.
The unique names of column and rafter are composed of their dimensions in mm, figure 2.19. For example
column 200x800x260x12x6 means that tapered section’s shallow end is 200 mm high while the deep end is
800 mm high, flanges are 260x12 mm and web thickness is 6 mm. Also the length of the haunch (L h) has to
be provided as the rafter is tapered only at the frame corner. Sometimes the length is expressed as the haunch
ratio (S/Lh).
The frame cross-sections were limited to section class 3 and lower in order to effectively use shell FEM’s
without need of local buckling calculation. The steel grade of all calculated cases is S275. All optimization
55
results were calculated using Genetic Algorithm (GA) method with fixed number of 50 generations and
population of 40 individuals in each generation. Although there is no guarantee that the frame configurations
resulting from a GA run represents the global minimum. For this reason, each optimization was carried out
twice and results were compared and refined. In later studies of double-span frames, the local search was
implemented automatically into the optimization algorithm.

2.6.2.1 Optimization of designed solutions

The lowest mass of individual frame was calculated for selected spans from 16 to 32 m with different snow
load (750 and 1500 N/m2) and frame height (6 and 8 m). From the optimized single portal frame the steel
consumptions in kg/m2 were calculated for the hall including purlins, side rails, horizontal and vertical
bracing, figure 2.20 and figure 2.21. Purlins, side rails and bracing were the same as for HR frames.

Constants:
Optimized variables:
Distance between frames: 6 m
Column bottom-end height (hc)
Roof pitch: 15.0 % (8,53°)
Rafter constant part height (hb),
Characteristic dead load: 380 N/m2
Corner height of the haunches (hhb,hhc),
Wind load: 30 m/s, terrain type 1
Flange width (b)
Seismic load: spectrum type 1, ground type B, q =
Flange thickness (tf)
1.5
Web width (tw)
Material: S275
Haunch ratio (S/Lh)
Base support: Pinned
hh b

hb

tw

hhc Column: hc x hhc x b x tf x tw

Rafter: hb x hhb x b x tf x tw
hc
tf

Figure 2.19. Column and rafter names (WT frames)

2.6.2.2 Performance assessment of optimized solutions

As the default output of AP-Frame software, the ultimate limit state (ULS) and serviceability limit state
(SLS) capacities are stored either in N/m2 of vertical load (in fundamental design situations) or in kN of
horizontal load in seismic design situations. The utilization factor is calculated using those capacities and
corresponding (design or characteristic) load. Their physical meaning is, therefore, related only to the
vertical or horizontal component of load while the other actions on structure are considered to be constant.
The selected frames were optimized using fundamental design situation with symmetrical snow load and
wind from the left side altogether with seismic design situation. Results were checked with unsymmetrical
snow loads and in two cases (24/6 and 24/8) where the SLS check was failing, the frames were redesigned to
satisfy SLS condition. The optimized WT frame configurations were included to the results of
PRECSATEEL (http://riv-precasteel.rivagroup.com/precasteel/).

56
 Steel consumption of double-span
Steel consumption of single-span industrial buildings with welded tapered
industrial buildings with welded tapered portal frames, height 6 m and 8 m
portal frames, height 6 m and 8 m 70 Height 6 m, Snow 750 N/m2
70
Height 6 m, Snow 750 N/m2 65 Height 6 m, Snow 1500 N/m2
65 Height 6 m, Snow 1500 N/m2 Height 8 m, Snow 750 N/m2
Height 8 m, Snow 750 N/m2
60
60 Height 8 m, Snow 1500 N/m2
Height 8 m, Snow 1500 N/m2
55
55
Steel consumption (kg/m2)

Steel consumption (kg/m2)

50
50
45
45
40
40
35 35

30 30

25 25

20 20
14 16 18 20 22 24 26 28 30 32 2x14 2x16 2x18 2x20 2x22 2x24 2x2
Span (m) Span (m)

Figure 2.20.Steel consumption of single-span Figure 2.21. Steel consumption of double-span

buildings buildings

2.7. Effects of the purlin support on the stability of frames

Unless other connecting elements are provided in the longitudinal direction of a hall (e.g. longitudinal
bracing ties, eaves beam or tie etc.); purlins alone provide lateral support to the frames. The effectiveness of
this support depends on the type of the purlin and its connection to the frame. As a basic case it was
supposed that LGS purlins are used, figure 2.22.a , but the effects of other stabilising elements were also
assessed between the frames, figure 2.22.b. The frame in the connection point of a purlin will have (1) a
lateral and (2) a torsional support. A configuration like in figure 2.22.b can be used (3) to provide lateral
support to the lower flange.
The stiffness provided to the frame cross-section is presented in 2.23. The tension flange is laterally
supported with a translational stiffness Ktf, and a rotational stiffness Kt. The blockings connecting the
compressed flange, figure 2.22.b, provide a translational support Kcf.

e.g.Z150/2

T=6000 T=6000

(a)

(b)
Figure 2.22. Connection to purlins to the frame

57
 Kt Ktf

h
Kcf

Figure 2.23. Stiffness from purlin to frame

The value of these stiffnesses and the effectiveness of the support, depend on the stiffness of the bolted
connections, and the axial and bending stiffness of the purlins. The possibility of sliding of the bolted
connections has also to be taken into account. It was found that the lateral (K tf) support is effective even
when provided by thin walled Z150/2 purlins, but the torsional support (Kt) is ineffective because of the slip
in the bolted connection allow significant initial rotation. Even support as in figure 2.22.b has reduced
efficiency in terms of Kct, and providing torsional blocking to the frame, because of the bending flexibility of
the purlin. The efficiency can be increased if different schemes are adopted for the supporting elements,
figure 2.24.

(a)

T=6000 T=6000

(b)
Figure 2.24. Improved connection to purlins to the frame

2.8. Analysis of connection systems

2.8.1 Improved connections for WT IBs

Traditional connections of the WT frames use extended flange for extra bolts that also increase the lever arm
of the connection, figure 2.25.a. Its main disadvantage is that the end plate is loaded in bending resulting in
thick endplates. The bending effect can be decreased by inserting additional stiffeners near the bolts in
tension, figure 2.25.b , or a connection type exploiting the shear capacity of the bolts can be used, figure
2.25.c. The last solution also has the advantage of providing the horizontal block when assembling the frame.

58
 (a) (b) (c)
Figure 2.25. Corner connection typologies for the WT frames

Connections in figure 2.25 have horizontal connection plane, but vertical connections are also common, as
presented by Dowling. The disadvantage of vertical connection over the horizontal ones is technological. In
case of horizontal connections the beam can be placed on the surface of the columns end plate and the bolts
can be fixed, while in case of the vertical connections the bolts have to be fixed while the beam is hanging in
the crane.

2.8.2 LGS connections design

The main challenge when designing connections for the LGS frame, table is that forces have to be
transmitted to the thin plate of the C profile. The typical failure mode of bolted connections in this case is
bearing. For economical solutions the connections should transmit forces comparable to the capacity of the
connected member; or they have to transmit even larges forces in some earthquake situations.
Hand calculations backed by FEM modeling has been carried out for connections of the LGS frames, figure
2.15. The FEM model has been prepared in ABAQUS Explicit, in order to be able to handle the multiple
contact situations. The material stress-strain curves used in the model were different for: (1) the M20 grade
8.8 bolts, (2) the S350 steel for the corner elements and the (3) S350 steel of the C profiles including
overstrength.
A 3 bolt row configuration was adopted for the connections, based on some preliminary calibration (table
2.7). In some models the yield stress of the LGS member has been increased, in order to test the effect of the
overstrength on failure of other component in the joint. According to EN 1998 capacity design rules the
failure should take place in the beam and the joint should be able to handle the overstrength of the connected
member. Seven failure modes of components were considered, table 2.7, and the corresponding overstrength
evaluated.
For most components the response was satisfactory. For mode 4 the shear failure of bolts developed, and bolt
quality had to be increased from grade 8.8 to grade 10.9. For mode 5 the net-section failure develops in the
bolt region at an overstregth value less then the 1.375 recommended by the code.
With the above observations the have been designed aiming for capacity design requirements of EN 1998.
The proposed configuration for members in the LGS frames is based on this configuration of M20 gr.10.9.
All holes are 22mm, bolts are not working as friction bolts, so they do not need to be pretension. As a good
practice measure it is recommended that the tension in all bolts is equal (e.g. half of the pretension force
recommended for friction bolts M20.gr.10.9). The corner piece has web 15mm, and flanges 8mm thick.

2.9. Improved IB solution: Corrugated web girder

The structure designed using cold-formed members - light gauge steel solution – considered a particular
morphology, inspired form performance assessment carried out on other structural type and from statistical
analysis: the corrugated web girders (CWG) with bolted web-to-flange connections. This solution was
chosen because it provides sufficient resistance to global and local buckling phenomena as well as joint
systems for the connection of the girder to the frame column. The general layout of the girder is depicted in
figure 2.26.

59
Nr. Comp. Yield strength of Msection Scale Plastic strain at Mmax Observations
LGS beam (kNm) factor
(N/mm2)
1 LGS beam 350 386.6 NA OK
yielding (due
to bending)

2 Corner 1.1*Ωov*fy= 488.31 1.31 OK

element (CE) =1.375*350
net section
3 CE yielding OK
(due to
bending)
4 Bolt shear γM2*1.1*Ωov*fy= 607.05 1.63 Increase number
failure. 1.25*1.375*350 of bolts or quality.
Increase to 10.9
DETECTED! solves it.

5 CE bearing 604.15 1.62 Very limited

failure plastic strain.
Finally the failure
is LGS yielding.

6 LGS net γM2*1.1*fy= 491.35 1.32 Net section failure

section failure 1.25*1.1*350 develops in the
Ωov not applied LGS. No solution
because to avoid it!
7 LGS bearing components in the OK
failure LGS are checked.

Table 2.7. Component performance in case of LGS frame connection

60
 Figure 2.26: General layout of proposed CWG

CWG with welded web-to-flange connections are already used today, as presented by Pasternak, H. And
Branka, P. However, the equipment necessary to create the welding seam is a barrier for the access to this
technique, figure 2.27.

Figure 2.27. Welding procedure

Investigations concerning CWG with bolted web-to-flange connections have been carried out in a research
project in 1997 [U. Peil, Trapezstegträger mit Steg-Gurt-Verbindung mit mechanischen Verbindungsmitteln,
DAST Forschungsbericht 3/1997] where the flanges were welded structures made of hot-rolled profiles. The
present campaign aims at providing an improved design of CWG using standardised cold formed sections
providing high bending stiffness and resistance against global and local buckling phenomena. In addition, a
bolted connection between CWG and frame column was developed that allows for modular assembly as
foreseen in the proposal to this project.
The bolted connections between web and flanges as well as the standardised products used offer easy access
to this technique. The girder is expected to benefit from its low mass in case of earthquake and is intended to
be used in the industrial buildings configurations listed in table 2.8 which were defined in the previous
performance assessment considering other structural solutions.

61
 Base Constraint (in-plane)

Overhead traveling crane

Snow loading (in kN/m2)
Height of frame columns

Resistance out-of-plane
Number of frame bays

Fixed (F), Hinged (H)

(1 brace per 5 frames)
Slope of frame girder

Wind load intensity

Resistance in-plane
Configuration no.

Maximum ground
Span of each bay

Base Constraint
Frame distance

Type of girder

(out-of-plane)

acceleration
2.2 1 20 m 6m 6 m 15 % Beam MR CB F H 0.75 0.16 g 30 m/s 0
2.8 1 20 m 6m 8 m 15 % Beam MR CB F H 0.75 0.16 g 30 m/s 0
2.20 1 16 m 6m 6 m 15 % Beam MR CB F H 1.5 0.16 g 30 m/s 0
3.20 1 16 m 6m 6 m 15 % Beam MR CB H H 1.5 0.08 g 30 m/s 0
3.21 1 20 m 6m 6 m 15 % Beam MR CB H H 1.5 0.08 g 30 m/s 0
Table 2.8. Selected case studies for the application of CWG

Config. Flanges Web Bolts per flange

no.
2.2 CL 150-30 S 355 TKH HP 107 S320, t=1.5 mm 2 M 10 12.9
2.20 CL 150-30 S 355 TKH HP 107 S320, t=1.5 mm 2 M 10 12.9
2.8 CL 150-30 S 355 TKH HP 107 S320, t=1.5 mm 2 M 10 12.9
3.20 CL 150-30 S 355 TKH HP 107 S320, t=1.5 mm 2 M 10 12.9
3.21 CL 150-40 S 355 TKH HP 107 S320, t=1.5 mm 2 M 10 12.9
Table 2.9. CWG solutions

The general load bearing behaviour corresponds to truss structures. The transverse force is covered by the
shear resistance of the trapezoidal web while the bending moment is split into a pair of tensile and
compression forces in the flanges. Accordingly, the bolted connection between flange and web must be
capable to withstand the corresponding shear flow.
Detailed information concerning the profiles used can be found in the following sections.
Design of CWG with welded web-to-flange connection under static loading is included in EN 1993-1-5,
Annex D. The design checks given include global and local buckling of the web as well as yielding and
buckling of the flanges.

2.9.1 Description of the components used

In order to avoid problems connected to availability of taylor-made profiles in practice, it was decided to
concentrate on the use of standard shaped profiles in the experimental campaign. For the web, corrugated
sheet metal of the type HP 107 fabricated by ThyssenKruppHoesch were chosen, while for the flanges C-
shaped profiles with varying web thickness were used, see figure 2.28 and figure 2.29, respectively. The
limited width of the trapezoidal sheeting of 750mm demands the arrangement of longitudinal joints in the
web. For these joints, blind rivets ‘Goebel Go-lock’ of diameter 6.4mm were chosen.

Figure 2.28. ThyssenKruppHoesch HP 107

62
 Figure 2.29. Schrag CL-profiles

However, in order to improve performance with regard to local buckling phenomena or lateral stiffness, the
use of taylor made profiles as depicted in figure 2.30 may be advantageous in practice.
The experimental campaign showed that due to the torsional deformation of the corrugations during loading
it is advantageous to dispose at least two fasteners each rib, figure 2.31. In this way, the flange of the
corrugated sheet metal is held in place. Further, it is advantageous to oversize the connection in order to
allow for redistribution of forces after failure of connectors. In this way, a more ductile load-deflection
behaviour is achieved.

Figure 2.30. Example of tailor made profiles for flanges

Figure 2.31. Two fasteners M 10 12.9 each rib

2.9.2 Connection to frame column

The connection between CWG and frame column was designed to allow for modular transport and assembly
of the components. As can be seen in figure 2.32, adaptor pieces are arranged between CWG and frame
63
 column which can be adjusted and preassembled in shop, deconstructed for transport and reassembled on
site.
The adaptor pieces consist of a plate with welded fingers extending into the flange profiles. For each flange,
one adaptor plate is used, thereby offering the possibility to even out production tolerances. The adaptor
pieces are designed to transfer compression and tensile forces from the flanges to the column. In order to
provide a rigid connection, additional plates reinforcing the column’s flange as well as stiffeners are welded
to the column in the area of load introduction from the girder.
The transverse forces from the web are being transferred to the column by means of a plate that is welded to
the column in vertical orientation. The trapezoidal sheeting can then be bolted to this plate using metric bolts.
Again, the layout of this connection enables the user to even out tolerances in production and assembly.
The order of assembly is arbitrary in general and depends on the conditions on site. However, the following
order will be applicable in most cases:
‚ Erection and alignment of columns
‚ Mounting of adaptor plates to CWG
‚ Placing of CWG between columns and final mounting of connections flange-column and web-column

Figure 2.32. Connection girder-column (see also figure 2.23)

64
3. Commercial buildings
In the present chapter the design of CB types is presented; in particular, the selection of geometrical data and
structural configuration are presented using the main results coming from the preliminary statistical analyses.
After the individuation of relevant structural types, the design process and the optimization procedure are
shortly presented focusing the attention to most interesting aspects. Moreover, some design issues
concerning the definition of the structural connections are reported making reference particular or interesting
types. All structural systems, connections and secondary members herein designed or presented were
inserted in the Precasteel database integrated with the web-application Precasteel 2.0.

3.1. Structural conceiving and design assumption

Structures for low-rise commercial buildings considered in PRECASTEEL are conceived to facilitate
prefabrication processes and to comply with designers’ requirements. For this purposes, very simple
arrangements of elements, connections and static schemes are chosen. Particular elements, as well as
structural arrangements, should consequently be designed case-by-case.
The gravity structure is constituted by beams hinged to continuous columns. The seismic resistant structure
may be constituted by steel concentric or eccentric braces or by shear reinforced concrete walls, figure 3.1.
In the latter case, the structure may be rigidly connected to the shear walls or interfaced by dissipative
devices in order to enhance the structural behaviour while limiting the design horizontal forces. Moment
resisting frames are not considered since they are characterised by high lateral deformability and high cost of
beam-to-column connections. On the contrary, braces may be easily accommodated thus reducing
considerably the structure weight and the connection complexity.
Flooring systems and columns are designed to withstand gravity actions whereas the braces are designed to
resist the assigned base shear forces. Dimensions and actions are varied within suitable ranges in accordance
with the statistical analysis previously presented.

gravity structure seismic gravity structure gravity structure seismic gravity structure
resistant resistant
structure structure

(a) (b)

gravity structure seismic gravity structure gravity structure seismic gravity structure
resistant resistant
structure structure

(c) (d)
Figure 3.1. Generic structural scheme for commercial buildings: (a) eccentric braces; (b) concentric braces;
(c) r.c. shear walls; (d) r.c. shear walls and dissipating devices

3.1.1. Main members

3.1.1.1 Flooring systems

Flooring systems considered are constituted by primary and secondary elements that are supposed to be
simply hinged at the ends. The plan scheme is regular and complies with a rectangular layout (figure 3.2a).
Different solutions have been evaluated such as composite elements obtained with hot rolled profiles, cold

65
formed beams (ZKU, ZKUG, rectangular), and trusses constructed with hot rolled profiles and cold formed
profiles (figure 3.2b).
Double web-angle connections have been adopted to facilitate both shop and erection operation in the
construction site (figure 3.2c). For secondary beams continuity of reinforcements (wire fabrics and transverse
reinforcements at the shear connection of the primary beam) is neglected while for primary beams a suitable
gap between slab and columns is considered in order to avoid undesired composite actions.

(b)

(a) (c) (d)

Figure 3.2. (a) Flooring system pattern; (b) geometry of truss girders; (c) double web-angle connections for
flooring systems; (d) bolted connections for trusses obtained with cold formed profiles

3.1.1.2 Columns
Both hot rolled (HE) and circular hollow sections (CHS) are considered for one-storey and two-storey
buildings. Structural sections considered include bare steel profiles, partially encased sections (PEHE) and
concrete filled circular hollow sections (CFCHS).
The advantage of considering CHS columns relies on their ability to be used not only for the rectangular grid
previously considered but also in more complex patterns which may be not rare for commercial buildings
due to architectonic requirements.
Columns have been designed by considering loads applied on the tributary areas descending from the beam
patterns considered for the flooring systems (figure 3.3.a).
Due to the beam to column connection and to the presence of horizontal diaphragms and seismic braces (or
shear walls), the structure is non-sway and the columns are considered pinned at the ends (figure 3.3.b).

H
L1
L2

(a) (b)
Figure 3.3. (a) Columns tributary areas; (b) schemes adopted in the design

3.1.1.3 Braces
Concentric and eccentric braces are considered for one-storey and two-storey buildings (figure 3.4). The
geometry of the braces is defined by varying the bay length (B) and the storey height (H); in particular the
bay widths may be different from the span lengths defined in the design of flooring systems, however, they
have been chosen equal to a fraction of the dimensions of the floor rectangular pattern in order to be easily
accommodated in the plan.
66
 B B B
e e

H H H

H H H

Figure 3.4. Braces considered in the project

3.1.1.4 R.c. shear walls

The design of prefabricated concrete walls, as an alternative of steel bracing system in CB, figure 3.5, was
performed in accordance with EC2 and EC8. Two different case-studies were examined:
‚ dissipative wall in the three ductility classes DCL, DCM, and DCH;
‚ elastic (completely undamaged) wall (coupled with dissipative devices) in DCL.
In accordance with different specific earthquake resistant provisions and with the storey-shear forces derived
from preliminary design, the wall solutions (in order to defined solutions adaptable to employing them in
automated software) are characterized by different shear forces (rating loads) at the base of the wall. The
values chosen for this analysis are: VBASE= 500kN; 1.000kN; 1.500kN; 2.000kN. Prefabricated walls were
realized adopting a Precast double slab walls able to substitute the steel bracing members, see figure 3.6.

Figure 3.5. Individuation of most pre-stressed wall element in the commercial building plan

Figure 3.6. the precast double slab wall solution

3.1.1.5 Dissipative devices

Dissipative devices may be profitably used when the lateral resisting elements are r.c. shear walls. For this
purpose different kinds of dissipators may be used. In the PRECASTEEL project only two particular devices
are considered: High Damping Rubber-based devices and Hysteretic devices. The former are obtained by
coupling thin HDR layers, reinforced with steel sheets, with steel plates in order to obtain devices capable of
undergoing important shear strains (figure 3.7). The latter was defined combining in a mechanical systems
dissipative steel fuses working both in compression and in tension and a steel wire element able to recover
final deformation and recentering the whole system.

67
 (b)
(a)
Figure 3.7. (a) HDR-based device; (b) hysteretic device

3.1.1.6 HDR-based devices

The use of HDR-based devices for structural systems is very promising in controlling the structural response
under actions like wind and earthquakes. Contrary to elastoplastic materials, HDR can dissipate energy for
low strains (figure 3.8.a) and is a fading memory material which provides the devices with interesting
recentering capabilities. Compared to visco-elastic and viscous devices, HDR-based devices are less
sensitive to the strain rate (figure 3.8.b). Consequently, the HDR-based devices can conveniently be used to
control the structural response in case of earthquakes falling within a wide range of intensities.

100 100
80 80
60 60
40 40
force (kN)

20 20
force (kN)

0 0
-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 -2.5 -1.5 -0.5 0.5 1.5 2.5
-20 -20
-40 -40
0.01 0.1 0.5
-60 -60
1 5 10
-80 -80
-100 -100
strain strain

(a) (b)
Figure 3.8. Stable loops of an HDR-based device: (a) different strain amplitudes; (b) different strain rates

On the other hand, simply hysteretic based devices are no more competitive is compared to HDR systems,
while high quality and high performing hysteretic dissipative members could be competitive if suitable
integrated in those lacking aspects as restoration of initial mechanical properties before the seismic event and
the residual deformation after the seismic event. For such a reason a original new steel hysteresis base
dissipative devices was considered in such a context, characterized by re-centering capabilities (no-residual
deformation at the end of the earthquake) and by replace ability of steel fuses (restoring of initial
performance), figure 3.9. The hysteretic device is characterized by a force-displacement law presenting flag
shaped cycle always passing through the origin and so reducing residual deformation.

3.1.1.7 Flag Shaped Hysteretic Devices (FSHD)

The development of dissipation devices, connections and entire seismic resisting systems exhibiting “flag-
shaped” behavior, characterized by the combination of self-centering and energy dissipation capacity, can
significantly reduce the expected level of damage, if compared with traditional monolithic ductile systems.
This is achieved by controlling the maximum displacements to target values and limiting (to negligible
values) the residual (permanent) deformations occurring after a seismic event. By appropriately combining
the alternative forms of energy dissipation (yielding, friction or viscous damping) in series and/or in parallel
together with the main source of re-centering capacity provided by unbonded post-tensioned tendons,
mechanical springs or Shape Memory Alloys (SMA) with superelastic behaviour, advanced high-
performance seismic resisting systems can be achieved. These are able to counteract the effects of both far

68
field and, more effectively, near field events characterized by a low number of cycles and high velocity
pulses.

(b)

(a)
Figure 3.9. (a) dissipative device assembled;(b) steel fuses inserted in the FSHD system

“Traditional” flag-shaped seismic resisting systems based on post-tensioning techniques were initially
developed for precast concrete buildings [Pampanin, 2005] and subsequently extended to steel frame
structures [Christopoulos et al.2002b], bridge piers [Hewes and Priestley, 1997; Palermo et al.2005a] and
more recently LVL (laminated veneer lumber) timber multi-storey buildings [Palermo et al. 2005a]. A
traditional flag-shape system combines the recentering capability from the unbonded post-tensioning cables
and the energy dissipating capability from additional energy dissipation hysteretic/yielding devices (either
internal or external). Figure 3.10 shows an idealized flag-shaped hysteretic behavior.

Figure 3.10. Hysteretic rule for ‘flag shaped’ devices

As a result, a properly designed flag-shaped system can achieve superior seismic performance when
compared with their traditional monolithic counterparts (i.e. elasto-plastic or Takeda hysteresis type of
behaviour), guaranteeing a limited maximum displacement and negligible residual (or permanent)
displacements. The importance of minimizing residual displacements for an adequate evaluation of seismic
performance has been recently highlighted in the literature [MacRae and Kawashima 1997; Pampanin et al.
2002]. For the above mentioned reasons, it was considered really important to investigate on the application
of such a dissipative systems to the analyzed building structures.

3.2. Set of designed structures (Definition of data-base of cross-sections)

Dimensions of members and actions, considered in the optimised design, are varied within suitable ranges
according to the statistical analysis previously illustrated. In order to comply with generic low-rise
commercial buildings, the design was carried out by considering structural sub-assemblages that may be re-
arranged to obtain complex configurations.
The following permanent vertical loads are considered in all cases:
- dead load of floor slabs = 2.85 kN/m2
(predalle panels or composite floors)
69
 - super-dead load = 1.80 kN/m2

Three different live loads are considered representing anthropic actions (frequently used in design) and
snow:
- standard live load = 5.00 kN/m2
- heavy live load = 8.00 kN/m2
1
- snow = 2.00 kN/m2

As for flooring systems, for the purpose of selecting a limited set of elements, their main dimensions (span
lengths) and the relevant tributary areas are evaluated by considering the scheme of figure 3.2 characterised
by the length of primary and secondary beams L1 and L2, respectively, and by the secondary beam spacing i.
The parameters are varied within the ranges L1 = 6 ÷ 24 m, L2 = 6 ÷ 12 m, i = 2 ÷ 4 m as better specified in
the following.
As for columns, one- and two-storey buildings are designed. In both the cases, two different storey heights
are considered, namely H = 6, 8 m for one-storey building and H = 4, 6 m for two-storey buildings. The three
loads already defined and the forces derived from the tributary areas multiplying L1 by L2 (see floors’
design) are considered for the definition of the resulting gravity loads.
For what concern braces and walls, a different approach is considered as they are designed with reference to
selected base shears and storey force distributions. The number of elements necessary to resist earthquakes
of given intensity is then evaluated with a suitable procedure.

3.2.1 Flooring systems

The set of flooring systems selected for the optimised design are summarised in table 3.1 which reports
geometrical data and loads considered. More than 100 flooring systems (primary and secondary elements)
have been designed; obviously, each optimised element may be considered individually when pre-designing
structures with different layouts provided that the tributary areas are almost the same (a bit less) of those
considered in the optimization.
3.2.2 Column elements
The set of columns selected for the design are summarised in table 3.2 More than 400 columns have been
designed. The same remarks made for flooring systems holds for columns, i.e., each optimised element may
be considered individually when pre-designing structures with different layouts provided that the tributary
areas are similar to those considered in the optimization.

3.2.3 Bracing systems

The set of braces designed for one- and two-storey buildings is reported in table 3.3. As already stated,
geometry of the braces is defined by varying the bay length (B) and the storey height (H) in order to be easily
accommodated into the flooring scheme considered. The link length (e) of eccentric braces is calculated in
order to optimize the structural solution and is not a given parameter. In addition to the geometric
parameters, the plan dimension in perpendicular direction (Ltransv) is considered in order to evaluate the
tributary area of the columns for the evaluation of the gravity loads.
The braces were designed for assigned base shear Vb assuming different distributions of the storey forces
according to storey masses and by assuming first mode linear shape (figure 3.11).

3.3 Optimization of structural modulus

Each main element was designed by optimising the structural weight. In the following sections, a brief
overview of the verifications carried out for the specific elements is provided.

3.3.1 Flooring systems with composite beams

Both serviceability (SLS) and ultimate (ULS) limit states have been considered by combining self weight,
dead and live loads according to the following formulas
Fd = 1.35Gk  1.50Qk ULS [3.1]

1
Snow loads depend actually on many data such as, geographical location of the site, site altitude, roof shape, heat flux
through the roof and exposure of the building. A complete definition of this action is not possible for the
PRECASTEEL purposes. The value has been selected by considering countries with moderate snows.
70
 Fd = Gk  Qk SLS (rare action) [3.2]
Steel grade S355 and concrete class C25/30, with the relevant partial safety factors JM = 1.10 and JC = 1.50,
are assumed in the design. Composite beams of flooring systems have been checked at construction and
service phases both for SLS and ULS.
2 2
0.62Vb Qk=2.00 kN/m 0.52Vb Qk=2.00 kN/m

2 2
Vb Qk=2.00 kN/m2 0.38Vb Qk=5.00 kN/m 0.48Vb Qk=8.00 kN/m

Vb Vb Vb
One storey Two-storey Two-storey
(seismic action with medium live load) (seismic action with heavy live load)

Figure 3.11. Definition of the storey forces

Typology L1 L2 i Qk
[m] [m] [m] [kN/m2]
Composite floors obtained with standard hot 6.00 6.00 2.67 2.00
rolled profiles (IPE, HE) 8.00 8.00 3.00 5.00
10.00 10.00 3.33 8.00
12.00 12.00 4.00
Primary truss girders and secondary composite 16.00 6.00 2.00 2.00
beams constructed with hot rolled profiles (HE) 20.00 8.00
24.00 10.00
12.00
Primary truss girders and secondary beams 16.00 6.00 2.00 2.00
constructed with cold formed profiles (ZKU, ZKUG, 20.00 8.00
rectangular) 24.00 10.00
12.00
Table 3.1. Set of designed flooring systems

Unpropped construction has been considered and, at the construction stage, Gk includes only beam self
weight and slab dead load (2.85 kN/m2) whereas Qk is not present. ULS has been verified checking shear and
bending resistance by considering the steel section only. Verification of the SLS at the construction phase
consisted in controlling that the vertical deflection Gv is not greater than L/200 considering only the steel part
of the girder.
At service conditions all loads are present. Resistance and stiffness have been evaluated with reference to the
composite section by considering complete connection and interaction between steel beam and concrete slab;
the plastic flexural resistance has been considered for the cross section in compliance with EC4. As for shear
resistance, this has been calculated with reference only to the shear area of the steel beam according to EC3.
For SLS, the following limits of the vertical deflection Gv have been assumed
Gv (Gk + Qk ) < L / 250 [3.3]
Gv (Qk ) < L / 350 [3.4]
In order to achieve the most economic solution, the cross section with the lowest self weight has been sought
among the IPE and HEA profiles that satisfy all the previous verifications.

71
 Profiles H Qk (*) Tributary area
2
[m] [kN/m ] [m2]
One-storey 6.00 2.00
6x6, 6x8, 6x10, 6x12, 8x8, 8x10,
HE, CHS, PEHE, buildings 8.00
8x12, 8x16, 10x10, 10x12, 10x16,
CFCHS, Rect.CF Two-storey 4.00 5.00
10x20, 12x12, 12x16, 12x20,
ZKUG CF buildings 5.00 8.00
12x24
6.00
HE = wide flange hot rolled profile; CHS = circular hollow section; PEHE = partially encased HE;
CFCHS = concrete filled CHS; Rect.CF = rectangular cold formed; ZKUG CF = double channel CF
(*)
For two-storey buildings, the snow load Qk = 2.00 kN is always considered on the roof and the value
in the table refers to the load at the first floor.
Table 3.2. Set of designed columns

B H Ltransv Vb
[m] [m] [m] [kN]
4.00, 5.00, 6.00, 500, 1000,
One-storey 8.00, 10.00,
6.00, 8.00 8.00, 10.00, 12.00, 1500, 2000,
buildings 12.00
Concentric 16.00, 20.00, 24.00 2500
braces 4.00, 5.00, 500, 1000,
Two-storey 4.00, 5.00, 4.00, 5.00, 6.00,
6.00, 8.00, 1500, 2000,
buildings 6.00 8.00, 10.00, 12.00
10.00, 12.00 2500
4.00, 5.00, 6.00,
One-storey 4.00, 5.00,
Eccentric 6.00, 8.00 8.00, 10.00, 12.00, 250, 450
buildings 6.00
braces with 16.00, 20.00, 24.00
one diagonal Two-storey 4.00, 5.00, 4.00, 5.00, 4.00, 5.00, 6.00,
250, 500
buildings 6.00 6.00 8.00, 10.00, 12.00
4.00, 5.00, 6.00,
Eccentric One-storey 8.00, 10.00, 250, 500,
6.00, 8.00 8.00, 10.00, 12.00,
braces with buildings 12.00 750
16.00, 20.00, 24.00
two
Two-storey 8.00, 10.00, 4.00, 5.00, 4.00, 5.00, 6.00, 500, 1000,
diagonals
buildings 12.00 6.00 8.00, 10.00, 12.00 1500
Table 3.3. Set of designed braces

3.3.2 Flooring systems with truss girders

Loads, materials and kind of verifications are the same of composite flooring systems. ZKU and ZKUG
sections (figure 3.12.a and 3.12.b) have been considered. The shear and moment resistances have been
calculated under the assumption of linear elastic stress distribution.
The truss girders have been designed for hot rolled sections as well as for cold formed sections. For hot
rolled sections, the HE series was used. For the cold formed sections, the ZKU and ZKUG (figure 3.12.a and
3.12.b) sections as used with the secondary beams have been considered. To cover stability issues, the
compressed bracings of the truss girders were charged with a parabolic imperfection of L/250 with L being
the length of the compressed member (figure 3.12.c).
It is considered that the upper and lower chords have constant cross section. Furthermore the same cross
section is used for all the vertical elements whereas for the diagonal elements the same cross section is
considered with exception of the end ones. All the elements are considered to be pinned although in the
reality the upper and lower chords are continuous.

3.3.3 Columns
Each element was checked according to EC3 and EC4 considering only the ultimate limit state (ULS) by
combining self weight, dead and live loads according to the formula [3.1].
Steel grade S355 and concrete class C25/30, with the relevant partial safety factors JM = 1.10 and JC = 1.50,
are assumed in the design.

72
Bare steel columns were checked against buckling by evaluating the capacity of pinned elements in
compliance with EC3. In order to achieve the most economic solution, the cross section with the lowest self
weight has been sought among the HE and CHS profiles that satisfy the verification. Class 4 cross sections
have not been used to avoid local buckling.
Partially encased (PEHE) and concrete filled (CFCHS) columns have been checked against buckling at
ultimate limit state by evaluating the capacity of pinned elements in compliance with EC4 by neglecting the
presence of reinforcements in the concrete component. In order to achieve the most economic solution, the
cross section with the lowest self weight has been sought among the HE and CHS profiles that satisfy the
verification. In order to avoid local buckling cross sections that do not fulfil the limitations 2
b/tf < 44HPEHE and d/t < 90H2 (CFCHS) have not been used.

(a) (b)

(c)
Figure 3.12. (a) cold formed profile ZKU; (b) cold formed profile ZKUG; (c) imperfection of truss bracing
members

3.3.4 Braces

3.3.4.1 Concentric braces

Braces have been checked against ULS by considering steel grade S355 with the partial safety factors
JM = 1.10.
The static scheme adopted (inverted V-diagonal with diagonal in compression) is not able to develop
significant ductility resources and has been classified as “non dissipative”. Braces have thus been designed
by considering a behaviour factor q = 1 and no capacity design rules have been applied. The elements under
compression (figure 3.13) have been checked against axial buckling according to EC3. The HE series has
been adopted for columns whereas IPE and HE series have been used for beams. HE and double-channel
cross sections were adopted for diagonal members. In order to achieve the most economic solution, the cross
sections with the lowest self weight has been sought among the profiles that satisfy the verification. Class 4
cross sections have not been used to avoid local buckling.

(a) (b)
Figure 3.13. Axial force in (a) one- and (b) two-storey concentric braces under horizontal forces
(red = compression)

2
EN1994-1 (6.7.1) H 235 f y

73
3.3.4.2 Eccentric braces with two diagonal members
The eccentric brace solution can reach high levels of ductility and the behaviour factor q = 6 has been
adopted according to EC8. Typical internal forces distributions under horizontal forces are shown in figure
3.14 for a two-storey brace. Links have been checked for resistance considering interaction of Nsd,E, Vsd,E and
Msd,E according to EC8. Link length e has been assigned in order to have “short” links that offer both high
level of ductility and stiffness. Link length has been considered variable in the range e = 800 mm÷1000 mm.
Only Class 1 sections have been used in order to obtain high ductility class (DCH). Since the contribution of
axial force Nsd,E is usually not very important, the link was sought among the IPE and HE series that fall into
Class 1 section.

Nsd,E Vsd,E Msd,E

Figure 3.14. Internal actions in two-storey eccentric braces with two diagonals

Section and length of the links have been designed in order to obtain the lowest value of the overstrength
ratio

:i 1,5V pl,link,i Vsd ,i , E [3.5]

where Vpl,link,i is the plastic shear resistance of the link. To achieve a global dissipative mechanism it has been
verified that the individual values of :i does not exceed the minimum value : by more than 25 %. The other
elements of the brace (columns and diagonals) were checked against axial buckling considering the capacity
design formula

N Rd t N Ed ,G  1,1J ov :N Ed , E [3.6]

where NRd is the buckling resistance of the column according to EC3, NEd,G is the axial force in the element
caused by vertical loads, Jov is an overstrength factor defined by National Annex to EC8 (it has been assumed
equal to 1.1), : is the minimum values of ratios evalueted for all the seismic links and NEd,G is the axial force
in the element caused by the design seismic actions.
The HE series has been adopted for columns by considering L0 to be equal to the storey height (H). HE and
duble-channel cross sections have been adopted for diagonal members. Class 4 cross sections have not been
used to avoid local buckling.

3.3.4.3 Eccentric braces with one diagonal member

As in the previous case, this eccentric brace solution can reach high levels of ductility and the behaviour
factor q = 6 has been adopted according to EC8.
This static scheme is hyperstatic due to the rigid connections at the joints, so the internal forces distributions
under horizontal forces depend on the relative stiffness of the brace members figure 3.15. An iterative
procedure has been carried out to achieve the optimum design of all the elements. As in the previous case,
links have been checked for resistance considering interaction of Nsd,E, Vsd,E and Msd,E. The link length e has
been chosen within the range e = 800 mm÷1000 mm in order to have “short” links.

74
 Nsd,E Vsd,E Msd,E

Figure 3.15. Internal actions in two-storey eccentric braces with one diagonal

Only Class 1 sections have been used in order to obtain high ductility class (DCH). Since the contribution of
axial force Nsd,E is usually not very important, the link was sought among the IPE and HE series that fall into
Class 1 section.
Section and length of the links have been designed in order to obtain the lowest value of the overstrength
ratio

:i 1,5V pl,link,i Vsd ,i , E [3.7]

where Vpl,link,i is the plastic shear resistance of the link. To achieve a global dissipative mechanism it has been
verified that the individual values of :i does not exceed the minimum value : by more than 25 %. The other
elements of the brace (columns and diagonals) were checked against buckling considering the capacity
design formula

N Rd (M Ed ; VEd ) t N Ed ,G  1,1J ov :N Ed , E [3.8]

where NRd is the buckling resistance of the element evaluated considering the interaction with shear (VEd) and
bending moment (MEd), NEd,G is the axial force in the element caused by vertical loads, Jov is an overstrength
factor defined by National Annex to EC8 (it has been assumed equal to 1.1), : is the minimum values of
ratios evaluated for all the seismic links and NEd,G is the axial force in the element caused by the design
seismic actions.
The HE series has been adopted for columns by considering L0 to be equal to the storey height (H). HE and
duble-channel cross sections have been adopted for diagonal members. Class 4 cross sections have not been
used to avoid local buckling.

3.3.4.4 Distribution of braces

As already mentioned, the braces have been designed to withstand assigned base shears Vb (see table 3.3).
Each brace can thus be used profitably to resist the seismic action corresponding to a tributary area which
depends on the earthquake intensities and on the site characteristics (design spectrum). By considering the
first vibration mode to be tentatively linear, consistently with the storey force distributions adopted in the
brace design, the fundamental period of the system constitute by the brace and its tributary area A may be
estimated from the expression

a ˜ Ma
T 2S A [3.9]
a ˜ Ka

where K is the translational stiffness matrix of the designed brace, M the mass matrix corresponding to a unit
area, and a is the vector

75
 ª1º
a «1 2» [3.10]
¬ ¼

The tributary area of the single brace may be evaluated by solving the following nonlinear equation obtained
by equating the assigned Vb to the base shear expected

Vb A
a ˜ Mr 
2
S
§
¨ 2S
a ˜ Ma ·
A ; q ¸¸ [3.11]
a ˜ Ma
d ¨ a ˜ Ka
© ¹

where Sd(T;q) is the suitable design spectrum.

Braces have been checked also with reference to DLS, by comparing inter storey drifts against the threshold
suggested by EC8 for buildings having non-structural elements that can withstand displacements without
undergoing brittle failures (0.005H).

theoretical link length e theoretical link length e

effective link length e’ effective link length e’

Figure 3.16. Effective link length

Since second order effects have not accounted for in the design, it has been also checked that

Ptot d r
- Vtot H
d 0,10 [3.12]

where T is the inter-storey drift sensitivity coefficient, Ptot is the total gravity load at and above the storey
considered in the seismic design situation, dr is the design inter-storey drift, Vtot is the total seismic storey
shear and H is the inter-storey height.
Finally, a last verification has been carried out regarding the link rotation capacity for eccentric braces by
checking that

Jp Bqd r d 0.08rad [3.13]

e' H

where B is the bay length, q is the behaviour factor, dr is the design interstorey drift, H is the interstorey
height and e’ is the effective link length defined in Figure 3.16.
Tributary areas of braces appearing in the PRECASTEEL data base have been evaluated by considering the
design spectra suggested by EC8 for soil Type B with PGAs equal to 0.32g (high seismicity), 0.16g (medium
seismicity) and 0.08g (low seismicity areas).

3.3.5 Application of cross-sections database to a defined set of case studies

In order to understand the range of applicability of the proposed database of cross-sections as defined by
UNICAM in WP3, several case studies were analyzed and discussed. In particular, considering the
variability of the accidental action (Qk), peak ground acceleration (a g) and geometry layout (Type of frame)
a set of 18 study cases was defined, each one labelled as reported in the following Table 3.4:

76
 ag = 0.32 (High) ag = 0.16 (Medium) ag = 0.08 (Low)
Qk=5kN/sm Qk=8kN/sm Qk=5kN/sm Qk=8kN/sm Qk=5kN/sm Qk=8kN/sm
Type 1 (CS1) CS1H5 CS1H8 CS1M5 CS1M8 CS1L5 CS1L8
Type 2 (CS2) CS2H5 CS2H8 CS2M5 CS2M8 CS2L5 CS2L8
Type 3 (CS3) CS3H5 CS3H8 CS3M5 CS3M8 CS3L5 CS3L8
Table 3.4 - Definition of analyzed case studies derived from the application of database cross-sections

After the cross-sections were extracted from data-base and applied, all frames were analyzed once again with
the aid of response spectrum analysis applied to three dimensional frame models; afterwards obtained results
were compared in terms of the resulting shear force at the column levels and of the obtained Ω factor for the
designed shear links.
After such a comprehensive analysis, two study cases were chosen and analyzed with the aid of incremental
dynamic analysis applied on planar models in the two main directions, in order to evaluate the actual seismic
performance.

3.3.5.1 Data-base main issues

At the concept design level, each structural system was defined in such a way to decuple the element load
bearing function for vertical and horizontal actions; as a consequence different sheets were included in the
cross-sections database. In particular, different sheets were defined for beam and column elements cross
sections under vertical loadings and for link-beam, column and diagonal elements being part of the
dissipative sub-structural scheme.
In such sheets each structural element was designed under specific issues consisting in considering a high
ductility class (q=6) in correspondence of each value of the peak ground acceleration and in fixing the length
of the designed links to a value of 80 cm, checking the geometric limitations in order to guarantee a shear
dissipation mode (short links).
Moreover, for what concerns the designed composite beams, the following geometric data were assumed
both for primary and secondary elements, independently of the spam length:
- beff = 700 mm for primary beams and beff =600 mm for secondary ones;
- hc = 120 mm (concrete slab thickness);
- offset between concrete slab and steel beam = 40 mm (Predalle thickness considered without a load-bearing
function).
A scheme of the cross sections for beams under vertical loadings is shown in Figures 3.17 a) and b).

600
700

120
120

40
40

IPE 400
IPE 400

(b)

Figure 3.17. Data base: a) and b) beam cross sections for primary and secondary elements; c) D-shaped
bracing system

For what concerns column cross sections under vertical loadings, different typologies were included in the
database, but in the analyzed set of frames only the partial encased solution was taken into account owing to
the higher feasibility of design solutions for what concerns structural connections. Finally, for what concerns
the dissipative system the static scheme is represented in Figure 3.17 c.
77
In the adopted static scheme for the dissipative system, brace elements are hinged both to the link-beam and
to the column elements which are continuous over two floors and hinged at the base ground; link-beam
elements are fixed to the adjoining column element at the shear link end while are hinged at the other;
column elements were oriented in the direction of higher inertia. After checking the geometric requirements
for the link-to-brace connections according to FEMA rules, it was decided for the geometry under
consideration, to adopt a D-shaped configuration for the brace system in order to obtain more suitable length
to height aspect ratios.
Different sheets were included into the data base for element cross-sections according to their load bearing
function. For beams under vertical loadings, several elementary schemes including primary and secondary
beams with different span length were taken into account, therefore once the geometry of the elements was
fixed, it was possible to derive their cross section in correspondence of the value assumed for accidental
loads. For what concerns column elements under vertical loadings, once their height was fixed it was
possible to enter the cross section table considering both the values of the area of influence and of the
accidental load.
For what concerns the structural elements included in the dissipative system, they were organized according
to beam length, column height and distance from an adjacent frame therefore once the geometry was fixed, it
was possible to enter a cross-section sub table according as a function of the value assumed by accidental
loads. Finally, the main parameter in order to obtain the most suitable cross sections was represented by the
area of influence assigned to each brace element, this value being a consequence of the assumed number of
brace in each direction and hence not only of structural factors but functional, esthetical and morphological
as well. Once the cross-sections were identified, it was possible to obtain the corresponding presumed value
of the design shear.
In the application of such a database to the defined set of frames, the number of bracing element in each
direction was considered fixed and equal to eight in order not to increase the number of cases too much.
The application of cross section database to all the defined set of frames is presented and discussed; basic
modulus for the identification of cross-sections for both vertical and horizontal actions were identified. In
particular, beam elements were chosen with reference to accidental actions, column elements were assigned
on the basis of the area of influence and finally brace cross-sections were assigned on the basis of their area
of influence.
For each study case, the extracted sub-tables for the assignment of cross-sections are shown in Appendix A.

3.3.5.2 Geometry layout of defined frames

The three different analyzed geometry layouts are presented; in each case, link beams and brace elements are
outlined in red, while beam elements sustaining vertical loadings in magenta.

3.3.5.2.1 Case Type 1 (CS1) – Geometry layout

The geometry layout of CS1 Type Frames is shown in Figure 3.18.

78
 Figure 3.18. Geometry layout for Type 1 (CS1) frame case studies

3.3.5.2.2 Case Type 2 (CS2) – Geometry layout

The geometry layout of CS2 Type Frames is shown in Figure 3.19. Frames and brace distribution, is the
same as Type 1 with the exception that beams have a longer span length in the longitudinal direction.

Figure 3.19. Geometry layout for Type 2 (CS2) frame case studies

3.3.5.2.3 Case Type 3 (CS3) – Geometry layout

The geometry layout of CS3 Type Frames is shown in Figure 3.20.

79
 6000
6000

4000
4000

6000 12000 6000 6000 12000 6000

5000
5000

6000 12000 12000 6000 12000 1200 6000

Figure 3.20. Geometry layout for Type 3 (CS3) frame case studies

3.3.5.3. Response spectrum analysis of the analysed set of frames

After the application of the cross-section database to defined set of frames, response spectrum analysis on
three dimensional frame models were performed for each obtained solution in order to check the obtained
accuracy in the designed of frames by making use of simplified design tools.
The main parameters taken into account in order to check and to compare the obtained results were the
obtained Ω factor and the actual value of the base shear force.
In the following pictures, the three dimensional models adopted for response spectrum analysis of the
analyzed set of frames for the three different geometry layouts are shown in Figures 3.21 – 3.23.

80
Figure 3.21. 3D model for response spectrum analysis for CS1 Type frames

Figure 3.22 . 3D model for response spectrum analysis for CS2 Type frames

81
 Figure 3.23 . 3D model for response spectrum analysis for CS3 Type frames

The obtained results are summarized in the following Table 3.5.

Ai [m2] Adatabase [m2] Ωmin Vx,max [kN] Vy,max [kN]

Vdatabase [kN] %Ai [m2] %V [kN]
H5 225 234,72 1,670 402 450 400 4,32 6,103286
M5 225 234,72 3,134 215 243 200 4,32 12,66376
L5 225 234,72 5,645 121 139 100 4,32 23,07692
CS1
H8 225 226,50 1,814 454 505 500 0,67 4,275287
M8 225 304,03 3,417 239 271 300 35,12 17,64706
L8 225 449,81 6,155 200 99,92
H5 279 361,41 1,580 480 500 500 29,54 2,040816
M5 279 478,70 2,891 250 258 300 71,58 18,11024
L5 279 465,79 5,015 133 130 500 66,95 280,2281
CS2
H8 DATABASE IS NOT APPLICABLE
M8 279 304,02 2,786 281 292 300 8,97 4,712042
L8 279 449,81 4,786 151 154 200 61,22 31,14754
H5 198 255,67 1,750 413 463 400 29,13 8,675799
M5 198 273,07 3,384 216 245 200 37,91 13,2321
L5 198 273,07 6,085 122 140 100 37,91 23,66412
CS3
H8 198 233,10 1,833 485 502 500 17,72 1,317123
M8 198 340,48 3,459 260 269 300 71,96 13,42155
L8 198 372,88 6,191 200 88,32
Table 3.5. Response Spectrum Analyses performed on 3D model set of frames

Some important conclusions can be obtained from response spectrum analyses performed on 3D models of
the predefined set of frames which can be summarized in the following points:
‚ Values of Ωmin increase (as aexpected approximatively with linear variation) with the decreasing of
PGA value (due to the fact that the q factor is fixed)
‚ When applying the database for low values of PGA the limitation on the 25% of Ωmin is not fulfilled
most of the times, moreover even if database is applicable it’s not possible to have confidence in the
obtained results

82
 ‚ The more the area of influence of bracing elements is closer to the value used to enter in the database
the more the obtained value of design shear is closer to the value provided in the cross-section data
base itself (the number of bracing elements and hence geometry layout of the frame have to be
checked carefully)
‚ The values of Ωmin increase with the increasing of the applied live loads
‚ Depending on the values of the area of influence it was not always possible to apply a data base
solution for the given geometry layout (see CS2H8 frame type)
‚ Some additional remarks can be made on the application of cross-section database to frame
structure can be:

‚ The application of database cross-sections provided really satisfactory results in correspondence of

high values of PGA both in terms of Ωmin and design shear, therefore its use is high recommendable
in such cases
‚ When entering the sub-table providing cross-sections for bracing and link beam elements attention
should be paid in choosing an influence area (and hence a number of bracing elements) differing
from actual value not more than 30% in order to obtain an accurate estimate of the resulting shear
and to avoid an over-sizing of element cross-sections (see CS2L5 frame type)

3.3.5.4. Incremental Dynamic Analysis

Incremental Dynamic Analyses were performed with the aid of OPENSEES numerical program.

3.3.5.4.1 Planar frame models

Incremental Dynamic Analyses were performed on a set of planar frames, extracted from 3D models; in
particular, two planar frames were chosen for each case study selected for refined analyses, namely CS1H5
and CS2H5.
The analysed planar frames are shown in Figures 3.24 – 3.29:
5000
5000

6000 6000 12000 6000 6000

Figure 3.24. Planar frames position Figure 3.25. CS1H5 planar frame in X direction

Figure 3.26 . CS1H5 planar frame in Y direction

83
 5000
5000
6000 6000 12000 6000 6000

Figure 3.27. Planar frames position Figure 3.28. CS2H5 planar frame in X direction

5000

5000

6000 12000 12000 6000 12000 12000 6000

Figure 3.29. CS2H5 planar frame in Y direction

The 2D models were extracted from the 3D models taking mass nodes and loads in way to represent a
coherent structural behavior. In particular for both 3D models the mass of each storey was divided into the
number of brace-frames and also divided into the number of 2D frame column elements; mass and loads
were applied on the node at top of every columns for each storey.
Beams and column elements were modeled with non linear elements and cross sections are discretized by
fiber models which were created by a script procedure; material models behavior used for steel and concrete
were “Steel02” and “Concrete01”, respectively the first consisting in a uniaxial Giuffrè-Menegotto-Pinto
steel material like behavior with isotropic strain hardening while the second in an uniaxial Kent-Scott-Park
concrete material like behaviour with degraded linear unloading/reloading stiffness and no tensile strength.

3.3.5.4.2 The adopted artificial accelerograms

The suite of ground motion adopted for Incremental Dynamic Analyses consisted of seven artificial
accelerogram compatible with the design spectrum and having the same peak ground acceleration set equal
to 0,32 g; the artificial accelerogram were generated with the aid of SIMQKE software with a record time
equal to 20 seconds. The set of artificial accelerogram is shown in Figure 3.30:
Acc. 1 Acc. 2
0,4 0,4

0,3 0,3

0,2 0,2
0,1 0,1
PGA [g] 0 PGA [g] 0
-0,1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0,1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
-0,2 -0,2
-0,3 -0,3
-0,4 -0,4
sec. sec.

Figure 3.30. Set of artificial accelerograms adopted for incremental dynamic analyses – 2 of the 7 time
histories

3.3.5.4.3 Obtained capacity curves and shear link behaviour

In the following pictures the results obtained from incremental dynamic analysis performed on planar frames
are shown in terms of capacity curves and rotation of the shear links, as plotted in Figure 3.31, where two
graphs as examples are reported.

84
 [rad/1000]
200
900 link3_acc1_max
link3_acc2_max
800 150 link3_acc3_max
link3_acc4_max
700 100 link3_acc5_max
600 link3_acc6_max
acc_1 50 link3_acc7_max
Vb 500 acc_2 link3_acc1_min
[KN] 400 link3_acc2_min
acc_3 0 link3_acc3_min
acc_4
300 link3_acc4_min
acc_5 -50 link3_acc5_min
200 acc_6 link3_acc6_min
acc_7 -100 link3_acc7_min
100
0,08 rad
0 -150
-0,08 rad

0,00 0,05 0,10 0,15 0,20 0,25 0,0 0,2 0,4 0,6 0,8 1,0 1,2 1,4 1,6 1,8 2,0
Drift [m] acc. scale factor
(a) (b)
Figure 3.31. Capacity curve for CS1H5 case study: a) X planar frame; b) rotation of Link 3 – second floor
level

3.3.6 Prefabricated walls

The design of prefabricated wall solutions, to be considered as alternative to steel bracing systems, suitable
for the implementation of simplified and quite automated sizing procedure was made according to the
following hypothesis concerning the idealization of the structural behaviour, load levels and geometrical
parameters.

3.3.6.1 Calculation hypothesis

‚ simplified static schemes, obtained by extracting substructures with lower complexity but still able
to describe the behaviour of the whole structure;
‚ floor systems, columns and walls are designed separately considering vertical loads for the first two
and horizontal actions (seismic and wind actions) for the third;
‚ linear elastic analyses;
‚ static seismic analyses to pre-design ductile walls (ULS), simplified dynamic seismic analyses to
give an estimate about influence area/wall (considering lumped masses for each storey);
‚ overturning vibration modes are avoided by technical joints (gaps between architectural modules)
and a symmetrical disposition of the walls;
‚ ductile walls are uncoupled (i.e. C or L plan shapes for staircases);
‚ shear wall deformation is taken into account through a refined wall stiffness model (Timoshenko
model);
‚ limitations imposed by Eurocodes are considered in order to obtain structural performances
consistent with standard provisions.

3.3.6.2 Load hypothesis

‚ maximum base shear horizontal loads (earthquake, wind) for a single r.c. wall:
‚ Vb = 500 - 1000 - 1500 - 2000 kN;
‚ roof/floor dead load:
‚ G1k = 2.85 kN/m2 (floor systems self weight, i.e. Predalle or composite floor);
‚ super-dead load:
‚ G2k = 1.80 kN/m2;
‚ heavy live load:
‚ Qk = 8.00 kN/m2 (commercial activities);
‚ standard live load:
‚ Qk = 5.00 kN/m2 (commercial activities);
‚ light live load:
‚ Qk = 2.00 kN/m2 (snow);
‚ seismicity level (PGA, peak ground acceleration):
‚ ag = 0.08 g (low seismicity);
‚ ag = 0.16 g (medium seismicity);
‚ ag = 0.32 g (high seismicity);

85
 ‚ r.c. walls are designed to resist both seismic and wind actions, assuming four different distributions
of the storey forces (distributions A, B, C, D, explained in figure 3.32). In the case of wind, the base
shear were distributed so that the force applied at the first storey is twice the one applied at the roof
level; in the case of seismic actions, by assuming the first vibration mode to be linear, the base shear
was distributed according to the following formulas:
§ M1 ˜ H ·
F1 Vb ¨¨ ¸¸ [3.14]
© M1 ˜ H  2 ˜ M 2 ˜ H ¹
§ 2˜ M 2 ˜ H ·
F2 Vb ¨¨ ¸¸ [3.15]
© M1 ˜ H  2˜ M 2 ˜ H ¹

3.3.6.3 Geometrical hypothesis:

‚ wall inter-storey height:
‚ H = 6 - 8 m (one-storey buildings);
‚ H = 4 - 5 - 6 m (two-storey buildings);
‚ wall width:
‚ B = 8 - 10 - 12 m (one-storey buildings);
‚ B = 4 - 5 - 6 m (two-storey buildings);
‚ wall thickness:
‚ s = 0.20 - 0.25 - 0.30 - 0.35 - 0.40 m.

B
0.33 V b

H
b

0.66 V
H

H
V b

(a) V b

(b)
B B

2 2
2.00 kN/m 2.00 kN/m
0.62 V b
0.52 V b
H
H

2 2
5.00 kN/m 8.00 kN/m
0.38 V b
0.48 V b
H
H

V b V b

(c) (d)
Figure 3.32. Distribution of horizontal forces: (a) seismic and wind action; (b) wind action; (c) seismic
action; (d) seismic action

In order to obtain the minimum number of seismic-resistant walls, able to withstand assigned base shears Vb
and given a specific commercial building area, the following procedure is adopted.
The first vibration mode is assumed to be linear, consistently with the previous storey force distributions
adopted, and may be expressed as:

ª1 i º
aT «¬ n ... n ... 1»¼ i = 1, ..., n (storey number) [3.16]

86
Being K the translational stiffness matrix of the walls and M the mass matrix corresponding to a unit area,
the fundamental period of the system can be estimated from the expression:
a ˜ Ma
T 2S ˜ ˜ A [3.17]
a ˜ Ka
where A is the unknown wall influence area.
The influence area A of the single wall may be evaluated by solving the following nonlinear equation,
obtained by equating the assigned Vb to the base shear expected:

Vb A˜
a ˜ Mr 
2
˜ S d T  [3.18]
a ˜ Ma
where Sd is the design spectrum. This approach is valid only for type “A”, “C” and “D” distribution of static
forces (figure 3.32), in which base shear is due to seismic actions; for distribution type “B” the forces are
originated by wind and the analysis of the influence area A of the single wall is unpredictable because it
requires knowing the exact form of the building and the surfaces exposed to wind. The obtained results are
referred to the design spectra suggested by Eurocode 8 for soil Type B.
Then, to sum up, the input steps that a designer has to follow are:
1) Fixing the maximum base shear horizontal loads (earthquake, wind) for a single r.c. wall.
2) Choosing the intensity of live load applied to CBs structures (heavy, standard, light live load).
3) Defining the site seismicity level (PGA, peak ground acceleration).
4) Assuming a distribution of the storey forces (A, B, C or D) with reference to the numbers of the
stories and the nature of actions (wind or seismic).
5) Fixing the ductility class of the structure (DCH and DCM for dissipative structures, DCL for r.c.
structures that does not dissipate energy under cyclic loads generated by an earthquake).
6) Estimation of behaviour factor (q) of the structure with reference to ductility class and geometrical
properties of the r.c. wall (storey height, width, thickness).
Considering all the previous hypothesis and applying the simplified design procedure, according to
Eurocodes standards, has been possible to build a database (an abstract, i.e., in tab. 1) in which the final
software will be able to provide designer a variety of feasible solutions for r.c. wall systems and the seismic
surface of a Commercial Building that this type of r.c. bracing system is able to sustain.

Figure 3.33. Simplified approach to estimate the influence area A of a single wall

87
 Table 3.6. Input data for the selection of a single Precast r.c. wall

Table 3.7. Output data for the selection of a single Precast r.c. wall

3.3.6.4. R.C. Walls as alternative bracing systems

The definition of a pre-design procedure for Commercial Buildings that is able to provide designers some
solutions about r.c. walls as alternative bracing systems, is based on some hypothesis concerning the
idealization of the structural behaviour, load levels and geometrical parameters.
It is worth noticing and it is important to underline the following aspects about the adoption of r.c. walls as
bracing systems of a Commercial Building:
ƒ Reinforced concrete walls solutions (figure 3.34a) are usually competitive towards concentric steel
bracing systems due to their capacity of dissipating seismic energy, although their fundamental
vibration periods are usually shorter. The above consideration is not true, almost always, comparing
performances of r.c. walls solutions towards eccentric steel bracing systems, because these ones have
a much bigger capacity of energy dissipation due to cyclic actions (in particular, seismic actions).
ƒ The main hypothesis of decoupling vertical and horizontal loads in CBs structures must be strictly
respected for steel bracing systems, especially for eccentric ones; in fact, these types of bracing
systems are able to dissipate a much bigger energy due to cyclic actions if they support only lateral
loads, because vertical loads (gravity or live loads) could compromise the optimal behaviour of the
seismic link. Vice versa, r.c. walls are able to support also vertical load during an earthquake,
without compromise their capacity of dissipating seismic energy and this make in evidence a wider
versatility of employing these typology of bracing systems for general structures; moreover, this
property of r.c. walls allows designers to reduce the number of steel members (columns, beams and
joints, positioned around bracing systems to sustain vertical loads) and then to bring down the total
cost estimation of the building.
ƒ Pre-design procedure of r.c. walls as bracing systems is easy, intuitive and fast, also in the definition
of the structural details suggested and prescribed by Eurocodes (Eurocode 8 for details about critical
regions and confined zones of the walls for ductility classes DCH and DCM, Eurocode 2 for details
about low dissipative r.c. structures or isolated structures by specific devices).
ƒ Connections between steel structure (beams, columns) and r.c. walls as bracing systems could be
realized in an easy way by chemical or mechanical anchors, after the erection of the r.c. structure.
Another way to realize this kind of joints is the classical bolted connection, shaped and included in
the formwork before the concrete pour. If the main structure is isolated by dissipative devices, at
each floor there are specific steel joints to prevent hammering between wall systems and floor
structures.
ƒ There are other advantages of choosing r.c. walls as bracing systems, for example the rapidity in the
phases of building and assembling, possibility to create openings (i.e. windows or doors) along the

88
 height of the r.c. wall, as well as the opportunity to link each other two walls by coupling r.c. beams,
creating a seismic-resistant system characterized by a higher level of dissipating capabilities.
ƒ Before casting concrete inside the r.c. wall, the precast element must be propped. Props must be
anchored in a plate able to support their compression and tension stresses. In correspondence of the
corner it is necessary to fix the double-slab walls with a steel plate opportunely shaped or with other
props (see Figure 3.34b). The cast should be made with a speed of 50cm/hour in order to avoid that
lateral pressure would be greater than 2500 daN/m2. Moreover, the cast must be done in different
time, according to the design of lattice girders.
ƒ The main limit of the developed methodology, to give an estimation about the number and typology
of r.c. walls as alternative bracing systems, concern the assumption of hypothesis about defined
geometries and loads, but one of the scope of this research project was to provide designer an
instrument easy and fast to make an approximate evaluation about feasibility of a project regarding
Commercial and Industrial Buildings.

a) b)
Figure 3.34. Operations of assembling precast r.c. wall.

89
4. Experimental testing
In the present chapter two experimental campaign are presented. First regards the testing of an improved
solution employing light gauge steel profiles and ribbed steel sheeting for realizing light and high resistance
main girder for IB solutions. The second experimental campaign is related to the testing of a novel
dissipative device with re-centering capabilities and steel hysteretic behavior; this experimental campaign
was carried out on prototypes specially designed with a complex and long feasibility and structural study.
The scope of this novel dissipative device is to work in series with the prefabricated r.c. walls used as
alternative bracing systems in the CB configuration. The matrix of planned tests has been modified more
focusing the attention on the corrugated web systems for main girder of IB and on a novel dissipative devices
endowed with hysteretic dissipation fuses and re-centering capabilities. The tests executed on reinforced
concrete prefabricated shear walls employing electro-welded lattice girders have not been reported.

4.1. Experimental assessment of light gauge steel girders for industrial buildings
The experimental tests conducted at RWTH Aachen aimed at the investigation of the corrugated web girder
(CWG) for industrial buildings developed during the analysis of results derived from the performance
assessment on IB configuration, which are described in chapter 2. The tests included 4-point-bending-tests as
well as tests of the connection girder-column.

4.1.1 4-point-bending-tests
In order to gain information concerning the load-displacement characteristics of the CWG itself, four 4-
point-bending tests were conducted, figure 4.1. Total length of the specimens was 6.00 m, load introduction
was arranged at one third and two thirds of the total length. The vertical spacing between the centres of
gravity of the CL-profiles was 1.05 m.

Figure 4.1. 4-point-bending tests

As it is favourable to introduce transverse loads directly into the web, load introduction and bearings were
realised by use of hollow profiles that were embedded in the trapezoidal sheet metal and that were connected
to the adjacent flange by means of blind rivets (Goebel GmbH Go-Lock 6.4 mm). In order to guarantee a
symmetric load introduction with regard to the web, load distributing plates were located at the opposite side
of the trapezoidal web, figure 4.1 and figure 4.2. In order to conserve the cross sectional shape of the CL-
shaped flanges during the tests, the cross section was stabilised by means of plates at the bearings and at the
load introduction, figure 4.3.b. However, these plates were left out in tests no. 3 and 4 proving that such
devices are not necessary for an adequate performance of the flanges.

91
Longitudinal joints of the trapezoidal sheeting were necessary every 750 mm. For these joints the same blind
rivets as for the load introduction were used. In figure 4.2.b a longitudinal joint can be seen next to the load
distributing plate.

(a) (b) (c)

Figure 4.2. Load introduction

Lateral support was applied at the bearings and at two locations near the load introductions in order to
prevent failure due to stability effects, figure 4.3.a.

(a) (b)
Figure 4.3. a) Boundary conditions and b) plates for cross-sectional stability

Table 4.1. gives a survey of the test matrix. Tests 1-3 served to figure out the most suitable way to connect
the flanges to the web. Therefore, three different fasteners were tested. The arrangement of the fasteners can
be taken from table 4.2.

Test Web Web Flange Flange Connection Supporting Lateral

no. thickness thickness tF web-flange plates support
1 TKH HP 107 1.5 mm CL 150-30 3.0 mm Blind rivets yes yes
2 TKH HP 107 1.5 mm CL 150-30 3.0 mm M 10 yes yes
3 TKH HP 107 1.5 mm CL 150-20 2.0 mm M6 no yes
4 TKH HP 107 1.5 mm CL 150-40 4.0 mm M 10 no no
Table 4.1. 4-point-bending tests

In the first test, blind rivets with a diameter of 6.4 mm of type ‘Go-Lock’, fabricated by company Goebel
GmbH, Ø = 6.4 mm, grip = 2-15 mm, were used to fix the flanges to the web. In each flange of the
trapezoidal sheeting, 4 fasteners were applied. The same connectors were used for the longitudinal joints of
the HP 107 profiles in all tests. It was necessary to drill holes of diameter 6.7 mm before the blind rivets
could be put in place.
92
Specimen 2 and 4 featured bolted connections M 10 12.9. As open CL-shaped flanges were used, access to
the bolts from two sides was possible so that assembly of the bolts was feasible. In each flange of the
trapezoidal sheeting 2 fasteners were applied.
For specimen 3, 4 bolts M 6 12.9 each corrugation were used. It was intended to find out if with this
configuration any advantages could be gained with regard to resistance against crinkling of the trapezoidal
sheet metal adjacent the connectors.
In test 4, the lateral supports were left out in order to get information about the girder’s susceptibility to
lateral torsional buckling.

Test no. Æ 1 2 3 4

Connections
web-flange

4 Blind rivets 2 M 10 12.9 4 M 6 12.9 2 M 10 12.9

6.4 mm

Connections
longitudinal
joints web

Blind rivets 6.4 mm

Table 4.2. Arrangement of fasteners

In all tests, deflection at mid-span as well as applied force was recorded. In test 4, lateral deflection was
recorded at 4 locations along the upper and lower flange in order to detect torsional movement of the girder.

4.1.2 Results of 4-point-bending-tests

The load-displacement curves of tests 1-4 are depicted in figure 4.4. The respective maximum forces are
listed in Table 4.3.. All specimens except specimen 3 tolerated a transverse force of beyond FB = 162 kN
(corresponding ultimate load Fult = 2*162 = 322 kN) which was determined to be the level of force at which
failure of the web would occur.

Test 1 2 3 4
Fult [kN] 288.5 412.8 360.9 411.6
Table 4.3. 4-point-bending-tests: Maximum forces

93
 Figure 4.4. 4-point-bending-tests: Load vs. displacement

It can be taken from the curves that in all cases except test 3 the initial linear elastic range was followed by a
distinct range of constant force and increasing deflection. In this range, the corrugations of the web
underwent plastic torsional deformation, Figure 4.5, and subsequent local crippling in the area of the
connectors, which finally led to failure of the fasteners due to peeling of the sheet metal, figure 4.6.

Figure 4.5. Torsional deformation of corrugation (here: specimen 3)

Figure 4.6. Local crinkling and peeling of the bolts (here: specimen 2)

Nevertheless, as can be seen in figure 4.6, the connection did not exhibit progressive failure after one
fastener was peeled out, but the connection was still able to transfer shear forces. Accordingly, the load-
deflection-curves of figure 4.4 do not show abrupt failures except for test 3 where failure occurred due to
buckling of the upper CL-flange in the area of load introduction, figure 4.7 a) and b). Although the level of
maximum force is only 13 % below tests 2 and 4, the abrupt failure due to local buckling of the flange of
2 mm thickness is a non-favourable failure mode. The local deformations of the 2 mm thick CL 150-20
profiles in the contact area between trapezoidal sheeting and flange, figure 4.7.c, is another indicator that the
configurations using CL-150 profiles profiles of at least 3 mm thickness (tests 2 and 4, respectively) are
superior. In addition, the larger washer of the 2 M 10 bolts (tests 2 and 4) used in tests 2 and 4 provided a
better peeling resistance than the smaller washers of the 4 M 6 bolts applied in test 3.
94
 a) b) c)
Figure 4.7. Test 3 – local damages after testing

It can be taken from figure 4.4 that specimen 1 where blind rivets were used provided not only a flattish
inclination of the load-deflection-curve compared to specimen 2 but also reached only 70 % of the
corresponding maximum load. This is due to the smaller peeling resistance of the blind rivets. As slippage of
the rivets was observed at a quite early stage of the test, figure 4.8, at least the products used in the present
test series do not seem applicable in web-to-flange connections of CWG. However, the longitudinal joints of
the web featuring blind rivets proved adequate in all tests.

Figure 4.8. Slippage of blind rivets, specimen 1

Figure 4.9 shows the deformation of specimen 2 near the ultimate load. One can clearly identify the
predominant shear deformation in the part between bearing and load introduction. Although failure of several
bolts had already occurred, the connection web-flange withstood a noticeable amount of distortion of the
trapezoidal sheeting. The longitudinal joints of the web as well as the load introductions are still intact at this
stage. It can be taken from the figure that no pure global shear buckling mode appeared across several
corrugations but a rather distinct inversely arranged torsional deformation of the corrugations. It is assumed
that this was due to the thicker material (1.5 mm instead of 0.75 mm) and the higher web of the trapezoidal
sheeting (107 mm instead of 50 mm) compared to the specimens.
In test 4 no lateral supports were arranged to avoid global buckling. Calculation of the critical moment with
regard to lateral torsional buckling (LTB) and consideration of the imperfections given in EC 3 showed that
for the actual statical system LTB would occur prior to local failure of the web. However, no distinct lateral
movement of any parts of the CWG could be recorded during the test, and the failure mode resembled the
one described for specimen 2. Although only one test addressing LTB was performed, this result indicates
that respective design rules for CWG featuring welded web-to-flange connections may be applicable for the
present solution.

95
 Figure 4.9. Deformation of specimen 2

4.1.2.1 Comments on experimental evidences

It could be shown in the first test series that 2 bolts M 10 in each flange of the trapezoidal sheet metal offer a
reliable connection between web and flanges capable of load redistribution during failure. It could also be
shown that the application of CL profiles of at least 3 mm thickness leads to a more ductile failure mode
compared to the 2 mm solution. The configurations applied in specimen 2 and specimen 4 reached a level of
ultimate force that exceeds the calculated ultimate buckling resistance of the web. The load-deflection curves
of these tests show a distinct plateau of even force and increasing deflection. One test was performed that
gives indication that rules for the determination of the critical moment with regard to lateral torsional
buckling can be transferred to the present solution.

4.1.3 Testing of girder-to-column connections

The second test series included two tests for the connection of the girder to the frame column. In the first
test, statical loading was applied, while in the second test the structure was subject to cyclic loading in order
to assess the tolerance of the connections to this kind of loading. The connection between CWG and column
is described in detail in chapter 2.

(a) (b)
Figure 4.10. (a) Testing of frame corner; (b) Load introduction and lateral support

The test setup is depicted in figure 4.10.a. Due to limited height of the testing rig it was necessary to incline
the lower part of the column at an angle of 90 °. With the inclined column and the arrangement of bearings
chosen the distribution of moment and transverse force of the real frame could be simulated in the
experiments. The deflection was monitored at the location w1 indicated in figure 4.10.a.

96
As shown in figure 4.10.a and figure 4.10.b load was introduced by means of a frame that was capable of
applying forces in both upward and downward direction. Lateral support was provided in the axis of the
column, figure 4.10.b

4.1.4 Results of tests

The test setup underwent a notable amount of slippage due to the arrangement of bearings. However, as can
be seen in figure 4.11, this slippage can easily be identified in the load-deflection-curve. During test 5, it was
necessary to apply a distance piece between specimen and hydraulic jack at a displacement of about 83 mm
in order to allow for further increase of load. The maximum load applied of F ult = 206 kN exceeds the design
buckling strength of the web of FB = 162 kN.
Failure occurred due to crinkling of the CWG’s web adjacent to the connection web-flange and subsequent
peeling of the trapezoidal sheeting, figure 4.12. In this way, the results of the 4-point-bending tests could be
reproduced. The connection girder-column itself did not show any damage after the test and proved
adequate.
In test 6, cyclic deformation was applied in three steps corresponding to load levels of ±100 kN, ±150 kN,
and ±200 kN. The load-deflection curves are reproduced in figure 4.13. In the same figure, the deflection
versus time diagram of load level 3 is given. Similar to test 5, distance pieces had to be applied each load
cycle leading to the slightly bumpy appearance of this curve. The slippage of the test rig of about ±10 mm is
not favourable but can easily be identified in the plots. The frame corner failed at a load of Fult = 203 kN.
Compared to test 5, only a neglectable decrease of ultimate force of 1.5 % due to cyclic loading can be
observed.

Figure 4.11. Load-deflection curve of specimen 5 - Maximum force: 206 kN

Figure 4.12. Failure of specimen 5

97
 Figure 4.13. Load-deflection curves of test 6

4.2. Experimental characterization of a novel hysteretic dissipative device

The second experimental campaign executed inside Precasteel research project was devoted to the
mechanical characterization of a original dissipative devices with re-centering capabilities developed with
the cooperation of another research project. This dissipative system, defined as Flag Shaped Hysteretic
Devices – FSHD –, was conceived to work in parallel with the prefabricated reinforced concrete wall as a
high performance and high technology seismic resistant system alternative to the steel EBF cantilever braces
for the gravity structure of the CB solutions.

4.2.1. Prototype description

Dissipative system FSHD, currently under patent, is a dissipative self-centring system completely made of
steel and made up of:
‚ an external case;
‚ an internal sliding frame;
‚ a piston used for the introduction of the external load;
‚ 2 anchor plates;
‚ a dissipative elements system;
‚ 2 prestressing cables.
External case
The external case is made up mainly of 2 sheets 10 mm thick linked as shown in figure 4.14. On one end the
case has a perforated element that allow the connection, by means of a pin, to the external structure. Within
the case four sheets are welded. They are used as leading system for the sliding frame and as contrast system
for the anchor plates. The case is equipped with side panels that shall avoid buckling phenomena due to the
external compression.

98
 4483

315 4078
258
B B B

451
Leading and B Side B B Sec. B-B Sec. A-A
contrast system panels

Figure 4.14. Global view and sections of external case

Internal sliding frame

The internal frame is realized with a couple of square hollow element 70x8.3 and 3100mm long. Both
element, at both ends, are welded with rectangular hollow elements 160x80x10 and 190mm long.
3500
B B A

348
250
B B A
Sec. C-C Sec. B-B Sec. A-A

Figure 4.15. Global view and sections of internal sliding frame

As shown in figure 4.16, the 2 elements are linked by:

‚ 2 plates, with a thickness respectively of 50mm and 70mm. As shown in Figure 5.3, both plates have
4 rectangular openings and 2 welded sheets necessary for the insertion and the joint of the dissipative
elements. The 70mm thick plate has also a circular opening necessary to the insertion of the piston.
144 50 123 123 70 144
268

Figure 4.16. Connecting plates

‚ A 12mm thick sheet necessary for the anchorage of the piston.

‚ Intermediate elements that allow the insertion of the piston and that reduce the free length.

Piston
As shown in figure 4.17, it is obtained by a circular hollow element Ф88.9x3.2. It is jointed at one end to the
internal frame by bolts and it has on the other end a perforated plates necessary to join the piston, by means
of a pin, to the external structure.

260 3257
200
100

3717

Figure 4.17 Piston

99
Anchor plates
As shown in figure 4.18, they are obtained by a plate 50mm and 70mm thick respectively. Both plates have 4
rectangular openings and 2 welded sheets necessary for the insertion and the joint of the dissipative elements.
Both plates have also a couple of circular opening for the insertion of the prestressing cables. The 70mm
thick plate has also a circular opening necessary to the insertion of the piston.

176 50 123 176 70 123

428

428
Figure 4.18. Anchor plates

Dissipative elements
They are obtained by dog bone shaped sheet and jointed by friction bolts to the anchor plate and to the
internal frame. The dissipative elements are equipped with a system that avoid the lateral buckling during the
compression phase, as shown in figure 4.19.

Prestressing cables
Open spiral strands equipped with adjustable cylindrical socket with threaded rod provided by Redaelli
Tecna Spa were used, as shown in figure 4.20.

500
170
40

10

(a) (b)
Figure 4.19. a) Dissipative element b) buckling restraining system

Figure 4.20. Pre-stressing cable

100
4.2.2. Test setup
Low cycle fatigue tests on the self-centring dissipator were conducted in the "Laboratorio Ufficiale per le
Esperienze dei Materiali da Costruzione" of the Civil Engineering Department at the University of Pisa. The
general test setup is shown in figure 4.21.

REACTION WALL

1110 2550 900 4666

400 kN IDRAULIC
1395

FSHD DISSIPATOR
ATTUATOR

Figure 4.21. General test setup

As load system has been used a 40 tons hydraulic jack, equipped with a load cell and a displacement
transducer. The hydraulic jack, placed horizontally at an height of 1395mm, has been connected at one end
to the reaction wall and on the other end to a steel structure that assure the vertical support but allow the
horizontal movement of the jack. To the same structure the dissipating device has been linked by a pin joint,
as shown in figure 4.22.

a) b)
Figure 4.22. a) Connection between hydraulic jack and dissipating device b) Connection between dissipative
device and the fixed structure

The other end of the dissipative device has been linked, by a pin joint, to a steel structure realized and
restrained in a way that prevent horizontal and vertical movement as shown in figure 4.22.b.

Gauge system
In order to measure displacement, strain and load, 8 LVDT (Linear Variable Differential Transformer)
sensors, 20 strain gauges and the jack internal load cell. All these sensors were connected to a National
Instrument Data Acquisition System. Sensors position are shown in figure 4.23. LVDT

101
 LVDT Displacement
sensor

Strain gauge

12 11

F 10 9 D 8 B
8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8

7 6
8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8

E 5 4 C 3 A Ext.1
2 1
Ext.2 Front side
11 12
B 13 14 15 F
8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8

16 17
8.8 8.8 8.8 8.8 8.8 8.8 8.8 8.8

A 18 19 20 E
Ext.1
1 2 Ext.2
Rear side

Figure 4.23.Sensor position

4.2.3. Testing procedure

Short testing procedure suggested by ECCS was used. In this procedure monotonic displacement increase
tests are not foreseen and only the low cycle fatigue test is carried out using a step of displacement
sufficiently small to ensure that at least four levels of displacement are reached before the yielding
displacement.
For the execution of the lab test an initial displacement step of 0.1mm has been used until the displacement
level reaches 0.5mm. Reached this value, the displacement step becomes equal to 0.5mm and for every
displacement level, 3 cycles are performed as schematically shown in figure 4.24. The testing displacement
rate has been fixed equal to 3mm/min.

Displacement History
8
Displacement [mm]

Time [s]
0 500 1000 1500 2000 2500
-2

-4

-6

-8

Figure 4.24. Displacement history used for the short testing procedure

4.2.4. Results
In figure 4.25, results of the last 11 loading cycles of the low-cycle fatigue test on the dissipating device are
shown. The geometrical and mechanical properties of the pre-stressing and dissipative element used are
summarized in table 4.4.

102
 Force - Displacement
150.00

Force [kN]
100.00

50.00

0.00
-10.00 -5.00 0.00 5.00 10.00 15.00

-50.00
Displacement [mm]

-100.00

-150.00

-200.00

-250.00

Figure 4.25. Force. Internal Frame displacement curve

It can be taken from figure 4.25 that, in every cycle, the residual displacement level is lower than 0.5mm and
so the dissipating device has an effective self-centring capacity. It also can be noted that the device shows a
stable hysteresis loops for every displacement level reached during the test, assuring a constant level of
energy dissipation.
The stability of hysteresis loops also during the unloading phase were assured by the presence of the
dissipative element buckling restraining system. In fact during this phase the dissipative elements are
subjected to a compression action that yield the elements. Thanks to the buckling restraining system it has
been possible to plasticize the dissipative element in compression without the presence of a global lateral
buckling, as shown in figure 4.26.

Figure 4.26. Local buckling of the dissipative element.

Number of elements 8
Cross sectional area 40 mm2
Dissipative element
Yielding stress (mean value) 240 N/mm2
Reduced beam section length 170 mm
Number of elements 2
Outside diameter 12 mm
Prestressing element Design yielding stress 1670 N/mm2
Total length 3500 mm
Prestress ratio (initial stress/yielding stress) 0.4
Table 4.4. Dissipative and prestressing element geometrical and mechanical characteristics
103
The different behaviour in tension and in compression can be attributed to the excessive transversal
deformation, happened during the test, of one of the welded sheet within the external case and the subsequent
loss of an anchor plate contrast as shown in figure 4.27. This contrast loss caused a different stiffness of the
dissipating device in tension and in compression, but did not compromise the self-centring capacity of the
dissipating device.

Figure 4.27. Loss of contact between the anchor Figure 4.28. C-formed element used to assure the
plate and the welded sheet contrast

Currently the problem has been solved with a C-shaped element jointed to the above mentioned welded sheet
that provide a larger contrast surface, as shown in figure 4.28. Other experimental tests were carried out
modifying internal mechanical properties of FSHD components in order to define dissipative devices suitable
for the application to the case study 1 for the CB solution with prefabricated r.c. wall. The modification of
steel fuse geometry and the section and pre-stressing rate of the post-tensioned cable allow to define different
FSHD with different yielding level, dissipated energy (i.e. area of cycle), maximum elongation and
hardening ratio, see figure 4.29.
Axial Force [kN]

Axial Force [kN]

1000 600
800
400
600
400 200
200
0 0
-200
-200
-400
-600 -400
-800
-600
-1000
-40 -30 -20 -10 0 10 20 30 40
-60 -40 -20 0 20 40 60
Displacement [mm] Displacement [mm]
Axial Force [kN]

Axial Force [kN]

800 1000
600 800
600
400
400
200
200
0 0
-200 -200
-400
-400
-600
-600
-800
-800 -1000
-30 -20 -10 0 10 20 30 -20 -10 0 10 20
Displacement [mm] Displacement [mm]
Figure 4.29. Different Flag shaped hysteresis obtained varying steel fuses section and pre-stressing rate

104
5 Software development
In this chapter two applicative tools derived from all the work developed inside Precasteel are presented. The
first tools is named Precasteel Web Application 2.0 and it is a software working on internet platform that
support engineers and designers in the use and application into day-to-day practice of the structural solutions
studied inside the research project. The second tool is a software for the structural optimization of one-storey
industrial building using genetic algorithm procedure implemented using VB macro inside an excel
worksheet. These two software are freely distributed on the Precasteel web-site http://riv-
precasteel.rivagroup.com/precasteel/.

5.1 Precasteel Software scope

The main objective of the application is to provide an integrated tool for automatic assessment of proposed
solutions: designed, optimized and parameterized in the previous research work for both applications:
industrial and commercial buildings.
More specifically, the final aim of the program is to provide a graphical interface that, starting from a finite
set of parameterized and optimized set of data, supplies the drawings and estimation of the cost of the
structure, floors, roof and connections for industrial and commercial buildings.
The statement “integrated tool” means that the program provides drawings encompassing: main structure,
secondary elements, roofing, floors, connections and cost estimation. This is the novelty of the programming
part of this research project, and is in fact an achievement in itself. The software produces the drawings of
the main and secondary structural elements as well as the general drawings and templates of the roofing
system, floor system, secondary elements, partition and external walls, beam to column connections and
column bases. In addition, an automated procedure for cost estimation and statistical case analysis has been
also implemented in the package.
The software application relies on a large database from which all the possible building solutions are selected
and retrieved. This database contains all the commercial and industrial cases that have been designed,
analyzed and optimized in this research project. It is important to clarify that the integrated software does not
include specific structural element analysis or design.

5.1.1 Workflow of the software

The application has been develop as a web application. The application guides the user through a variety of
input options until he arrives at the design, industrial or commercial building, that better fits his necessities.
Both applications are divided in seven steps where the user choose parameters like number of spans, type of
cross sections, type of wall and roof system, etc. Due to the different nature of the commercial and industrial
buildings the steps are not the same for each one of them. The web application allows the user to go one or
several steps back at any time of the input action to modify the different options.
The main interface of both applications is divided in three parts: step menu, step panel and selections panel
(figure 5.1):
‚ The step menu is located at the top of the page. This menu informs the user about the steps that have
been completed and how many remain to be completed. The user can click any of the completed
steps to go back and select a different option.
‚ The step panel located at the middle of the page shows all the options the user has to select in order
to complete the definition of the element associated with this step.
‚ The selections panel at the bottom of the page shows all the selections made by the user at this
moment. The user can click any of them to go back and modify its value.
Due to the amount of information available in the database, the application provides comprehensive bar
charts to easy the user selection process (figure 5.2). These bar charts compare, in terms of weight and cost
(also number of elements required) the solutions available in the database that match the criteria chosen by
the user.
Once the user has made all the selections the software provides (as download documents) all the information
available for the chosen typology. At this step, the cost estimation module can be used to obtain the final
price of the building including: structure, wall, slab and roof systems (figure 5.3).

5.2 Industrial building

All the key input options provided to the user (length of the span, load conditions, etc.) depend on the
solutions stored in the database. The input options shown below have been defined according to the

105
information developed in the other work packages of this project. However, the database can be easily
adapted to store new cases depending on other user necessities.
The main advantage of using a pre-analyzed solutions database is that the user gets instant access to all the
available solutions. However, the number of choices is finite and step wise. If one wants an intermediate case
between two given options then it will have to be analysed by the user, and added (if desired) to the database,
thus the database of this computer tool may be enlarged with time. At this stage, the set parameters for the
industrial building database are the following:
‚ The distance between frames is 6 m.
‚ The roof peach is 15%.
‚ The lateral bracing system has been studied for a module of 6 frames.
‚ The wind velocity is 30 m/s.
These values were set at the beginning of the project during the statistical analysis of most common national
typologies.
Input. The seven input steps that are required to define the industrial building are summarized as follows
(see figure 5.4 for a graphical scheme of the workflow):
Step 1. Select the number of spans: 1 or 2 (see figure 5.5).
Step 2. Select the length of the span. Five different span lengths are available: 16 m, 20 m, 24 m, 27 m, 28 m,
30 m and 32 m (see figure 5.6).
Step 3. Define the seismic action. Three options are provided to the user: low (0.08g), medium (0.16g) and
high (0.32g) peak ground acceleration. These values correspond to ground Type A of the Eurocode 8.
Step 4. Then, the web application provides solutions stored at the database that better fits the user input. At
this step, each solution could be one of the next typologies:
1. Frame with hot-rolled sections.
2. Frame with welded-tapered sections.
3. Frame with truss girders.
4. Frame with light-gauge sections.
5. Frame with folded-web sections.
At this stage the program presents to the user the different options available for each family of typologies.
Also, it shows the value of the snow load: 0.75 kN/m2 or 1.5 kN/m2, and the value of the crane load: 0 kN, 50
kN, 100 kN and 250 kN. With this information the user can easily figure out if the selected case fits his loads
demands.
Step 5. Select the roofing system from the different solutions that are provided.
Step 6. Select the cladding system from the different solutions that are provided.
Step 7. Define the location for the building and the distance from the suppliers. This information is used in
the cost estimation analysis module.
Output. The program provides the information available in the database for the predesigned and optimized
solution that the user has selected. This information is divided into several sections:
1. The geometrical and loads conditions of the frame selected.
2. The structural layout: cross-sections of each element of the frame and bracing system, and restraint
conditions (in-plane and out-of-plane).
3. The cladding system: main properties of wall and roof cladding system are shown.
4. Data available. This section provides all the technical drawings and information that completely
define the selected industrial building. These documents are:
‚ Structure drawings including: front, top and side view as a .dxf file.
‚ Details of the connections: column base, apex and eaves connections as a .dxf file.
‚ Cladding information for wall and roof system as a .pdf file.
‚ Text summary of the main characteristics of the industrial building including cost estimation in a .rtf
file.
5. Cost estimation. This section has been designed as an interactive module inside the main application.
It is divided in two parts: the spreadsheet and pie chart (see figure 5.3). The spreadsheet is divided in two
blocks: cost of the structure and cost of the cladding system. The cost of the structure is divided in material,
transportation, assembly and fire protection. The cost of the cladding system is divided in material,
transportation and assembly. Any parameter may be modified at this stage to check a new cost estimation.
The pie chart summarizes all the cost of the industrial building, and helps to identify, in an easy way, the
most expensive parts.

106
As mentioned above the software application relies on a large database of industrial buildings from which
possible solutions are selected and retrieved. This database contains all the cases that have been designed,
analyzed and optimized in this research project. Therefore, if at a particular step the user wants to consider
different loading conditions, then he should choose one of the given solutions and perform a structural
analysis of the selected frame including the user specified loads. If the solution is acceptable then the
drawings and information provided by the application can be used, otherwise, the user will have to modify
the graphical output accordingly.

5.3 Commercial building

As in the case of industrial buildings all the key input options available to the user (length of the span, load
conditions, etc.) depend on the solutions stored in the database.
Input. The seven steps that composed input data for commercial buildings are summarized as follows (see
figure 5.7 for a graphical scheme of the workflow):
Step 1. Select the length for both directions of the module (primary and secondary beam lengths) and the
number of spans for both directions. The options available for primary beams are: 6 m, 8 m, 10 m, 12 m, 16
m, 20 m and 24 m. The options available for secondary beams are: 6 m, 8 m, 10 m and 12 m. The number of
spans is currently limited to 25.
Step 2. Choose the number of stories and the height: 1 or 2 stories. For 1-storey solution two heights are
available: 6 m or 8 m. For 2-storey solution the height of both floors can be 4 m, 5 m or 6 m.
Step 3. Define the value of the live load and the seismic action. For the live load the following values are
available: heavy live load (8.00 kN/m2), standard live load (5.00 kN/m2) and snow load (2.00 kN/m2). The
seismic options provided to the user are: low (0.08g), medium (0.16g) and high (0.32g). These values of
peak ground acceleration are set for ground Type A of the Eurocode 8.
Step 4. The application provides solutions stored at the database that better fit the user input in terms of the
columns, the beams and the bracing system (see figure 5.8). The structural typologies available for each
element are:
‚ Columns cross-section: CHS (circular hollow section), CICHS (concrete in-filled circular hollow
section), RHS (rectangular hollow section), ZKUG, HE, PEHE (partially encased HE).
‚ Beam cross section (primary and secondary): IPE, HE, trusses with hot rolled profiles and trusses
with cold formed profiles.
‚ Bracing system (both directions): concentric with HE, eccentric with one diagonal and HE, eccentric
with two diagonals and HE, and concrete wall.
Step 5. Select the wall cladding system.
Step 6. Choose the horizontal cladding system for each storey.
Step 7. Define the building location and the distance from the suppliers. This information it is used in the
cost estimation analysis module.
Output. The program provides the information available in the database for the predesigned and optimized
solution that the user has selected. This information is divided into several sections:
1. The geometrical and loads conditions of the commercial building.
2. The structural layout: cross sections of each element of the frame and bracing system.
3. Data available. This section provides all the technical drawings and information to completely define
the commercial building selected. These documents comprise:
‚ Structure drawings including: front, top and side view as a .dxf file.
‚ Details of the connections as a .dxf file.
‚ Cladding information for wall and roof system as a .pdf file.
‚ Text summary of the main characteristics of the commercial building, including cost estimation, as a
.rtf file.
4. Cost estimation. This section has been designed as an interactive module inside the main application.
It is divided in two parts: the spreadsheet and pie chart (see figure 5.3). The spreadsheet is divided in two
blocks: cost of the structure and cost of the cladding system. The cost of the structure is divided in material,
transportation, assembly and fire protection. The cost of the cladding system is divided in material,
transportation and assembly. Any parameter may be modified at this stage to check a new cost estimation.
The pie chart summarizes all the cost of the commercial building, and helps to identify, in an easy way, the
most expensive parts.

107
The software application relies on a large database of commercial buildings from which the possible
solutions are selected and retrieved. This database contains all the cases that have been designed, analyzed
and optimized in this research project. Therefore, if at a particular step the user wants to consider different
loading conditions, then he should choose one of the given solutions and perform a structural analysis of the
selected frame including the user specified loads. If the solution is acceptable then the drawings and
information provided by the application can be used, otherwise, the user will have to modify the graphical
output accordingly.

5.4 Technical issues concerning the software

Internet technology has been adopted as the platform for the development of this software. This technology
provides an online application, focused on the server, so the maintenance of the resource is easier and users
will always have up to date information. Moreover, the dissemination is faster compared to a stand-alone
application. All the templates and tools are standard and system independent.

5.4.1 Server side technologies

The server side application has been developed using several technologies that are commonly known as
LAMP architecture. This architecture has become popular in the web industry as a way of deploying web
applications. LAMP is composed of (L) Linux operating system, (A) Apache as web server, (M) MyQSL as
database storage and (P) PHP as server side scripting language.
The core of the software has being developed with PHP (Hypertext Preprocessor), a server side scripting
language for producing dynamic web pages. PHP generally runs on a web server, taking PHP code as its
input and creating web pages as output. In this application, PHP manages the application, the input data from
the user, the output data provided to the user and the connection with the database.
The database, which stores all the designed and parameterized solutions, has been implemented in MySQL
(my standard query language). MySQL is a relational database management system (RDBMS) that runs as a
server providing multi-user access to a number of databases.

5.4.2 Client side technologies

Client application has been developed with three different technologies, which complement each other:
Extensible Hypertext Mark-up Language (XHTML), Cascading Style Sheets (CSS) and Asynchronous
JavaScript And XML (AJAX).
The Extensible Hypertext Markup Language, or XHTML, is a mark-up language that has the same depth of
expression as HTML, but also conforms to XML syntax. XHTML is used to define the structure of each web
page rendered by the browser.
Cascading Style Sheets, or CSS, is a style sheet language used to describe the presentation (that is, the look
and formatting) of a document written in a markup language. Its most common application is to style web
pages written in HTML and XHTML. CSS has been used for the look of the interface.
Asynchronous JavaScript And XML, or Ajax, is a group of interrelated web development techniques used to
create interactive web applications or rich Internet applications. With Ajax, web applications can retrieve
data from the server asynchronously in the background without interfering with the display and behaviour of
the existing page. The use of Ajax has led to an increase in interactive animation on web pages. Ajax has
been used in this project to make the interface more user friendly, and similar to those provided by desktop
applications.

5.4.3 Web application resources structure

A large number of files have been written to develop the web application proposed in this project. Currently
the application is divided in three parts:
1. Database of predefined solutions.
2. Source code of server framework.
3. Source code of client application.
As stated before, the database has been defined in MySQL and has been divided in several tables that store
all the information required by the application. The pre-analyzed solutions for Industrial and Commercial
buildings are also stored there. Tables stored in the server are grouped in three classes:
‚ Industrial building tables: “IFrames”, “ICosts” and “ICladdings” that store all the information related
to pre-analyzed cases for industrial buildings.

108
 ‚ Commercial building tables: “CBeams”, “CBracings”, “CColumns”, “CCladdings” and “CCosts”
that store all the information related to predesigned cases for commercial buildings.
‚ Application table: “sessions” that stores all the options made by each user during the current session.
‚ The server framework has been fully developed using PHP. The 25 PHP files developed have been
grouped in three categories:
‚ Commercial building web application controllers (stored in /phpFW/Comercial folder). The files
from Comercial1.php to Comercial8.php handle each step of the application providing the user the
options available for selection. The file “PrecasteelC.php” contains all the scripts necessaries to
handle the communication with client application (browser). The file “PrecasteelCGetFile.php”
handles all the information that the user can download for the options selected, and “DrawFloor.php”
renders the floor and side drawings of the commercial building. These are the only drawings that are
rendered at runtime by the server.
‚ Industrial building web application controllers (stored in /phpFW/Industrial folder). Files from
Industrial1.php to Industrial8.php handle each step of the application providing the user the options
available for selection. The file “PrecasteelI.php” contains all the scripts necessaries to handle
communication with client application (browser), and “PrecasteelIGetFile.php” handles all the
information that the user can download for the options selected.
‚ Common application controllers (stored in /phpFW folder). These scripts define the model-view-
controller logic of the application. Some of the tasks managed by these scripts are: track options
made by the user during current session and define the general scheme of both applications. All of
this has been implemented in: “Precasteel.php”, “PrecasteelLogin.php” and “PrecasteelSession.php”
files.
‚ Client application has been developed with javascript and AJAX to improve the user interaction. It is
also grouped in three categories stored in “/js” folder:
‚ Common application controller. This is the underlying core of the client application. It is composed
of several scripts that improve user experience and handle all the communications between the
industrial and commercial web application and the server. These scripts are coded and included in
the following files: “json2.js”, “PrecasteelLib.js”, “PrecasteelXHR.js”, and “PrecasteelMain.js”.
‚ Industrial building web application controller. These scripts handle all the specific client-server
communication related to industrial options. All this code is located in “PrecasteelI.js” file.
‚ Commercial building web application controller. These scripts handle all the specific client-server
communication related to commercial options. All this code is located in “PrecasteelC.js” file.
‚ Additional folders and files have been used to store all the resources required by the application.
This information is stored in three folders:
‚ Resources for commercial building. The “/Commercial” folder contains: all the drawings to define
each commercial building selected, all the information of the wall and floor cladding systems, master
file for cost estimation definition and all the images used at each step of the commercial application.
‚ Resources for industrial building. The “/Industrial” folder contains: all the drawings to define each
industrial building selected, all the information of the wall and roof cladding systems, master file for
cost estimation definition and all the images used at each step of the industrial application.
‚ Resources for the interface of the application. “Precasteel.css” located in “/css” folder defines all the
rules about how the browser should render the application. The folder “/img” contains all the images
use by the interface which are not specific for each application.

5.4.4 Statistical case analysis

In order to make the most of the technologies implemented in this research an additional module has been
developed. At the end of the last step of each application the web server stores all the information related to
the user selections in the database. Then this module provides the host of the application all the main
information relative to the user selections (see figure 5.9). This information is quite useful to identify user
patterns such as: the most frequent modules, number of spans on both directions, typology of the structure
and values of loads, etc.
This statistical information is provided as pie charts in two containers: industrial and commercial. The initial
information comes from all the users that have accessed the application during the last month and have
arrived to the last step. Making use of one of the two-drop boxes, the host can filter this information by

109
country and perform market analysis. This module is located in “PrecasteelSta.php” file, in the root level of
the web application.

110
5.4.5 Screen captures of the application

Figure 5.1. Structure of the web application interface, similar for both buildings.

Figure 5.2. Bar chart to facilitate user selections.

111
Figure 5.3. Cost module structure

112
Screen captures and work flow chart of industrial building

Figure 5.4. Flow chart of industrial building application

113
Figure 5.5. Step 1 of industrial building application

Figure 5.6. Step 2 of industrial building application

114
Screen captures and work flow chart of commercial building

Figure 5.7. Flow chart of commercial building application

115
Figure 5.8. Step 4 of commercial building application

116
Statistical module

Figure 5.9. Statistical case analysis

5.5 Software for IB optimization

Optimization of structures is a difficult engineering task. In the everyday design practice, optimization is
carried out integrated into the design process. It is self understood that the design engineer tries to provide an
optimized solution for the design task at hand. As a result, the quality, in the sense of optimal or not, of the
resulting engineering structures depend on the expertise of the designer. More systematic optimization is
usually not carried out, because of lack of expertise or time pressure during the design process. One of the
solutions for overcoming both this factors is to give the designer easy to use and efficient software tools for
optimising structures. In this work two such tools, AP-Frame and EV-Frame have been developed, both
using genetic algorithms (GA) as optimisation method, figure 5.10.

117
 Figure 5.10. The basic diagram of two optimization tools developed at VTT

The tools were coded with a special focus on their performance because it was needed to calculate over 400
thousands finite element models in order to produce all results delivered to the project. In case of 3D shell
models, we used the advantage of parallel processing on 11 computers with four CPU units each. However,
as the average computational performance rapidly increased in recent years we can expect that this kind of
optimization task will be performed by designers on their personal workstations in the near future. Figure
5.11 demonstrates one of the shorter runs (HR frames were simple in comparison with WT frames and
usually a few hundred calculations produced the optimal solution). Figure 5.11 presents the chosen
configurations, in terms of (1) how feasible they are to fulfil the design objectives and their (2) weight. At
the right hand of the graph, with read, there are unfeasible solutions violating some design requirements. As
it can be seen both feasibility and weight is decreasing towards the right of the graph, the aim of the
algorithm is to choose a still feasible but also light frame. Figure 5.11.b presents the evolution of the fitness
of frames on the generation by generation basis in the GA. In this example, the “elite frame”, the optimised
solution was found already in the 9th generation, the rest of the search up to the 40th generation did not result
in a better configuration.

AP-Frame: Python script for Abaqus CAE

Python analytical script is a software module designed to perform the design calculations for a single frame
configuration, or of a pre-defined set of frames, using global non-linear analysis (GMNLA) or the general
method (GMA). The module can be used to evaluate the load bearing capacity of frames with known
geometry, or integrated into an optimization procedure, figure 5.12.
The AP-Frame script was coded in Python, in order to easily implement Abaqus/CAE commands for
creating the numerical model, and evaluating the results. The method for exporting frame drawings uses
XML specific language. Parts of the code also contain UNIX or Windows specific system commands. The
module was tested with Abaqus version 6.9, using Abaqus/CAE license for modelling and result
interpretation; and up to 4 Abaqus/Standard licences for calculations.

118
 1600 frames in 40 generations (520 calculated) - sorted by
feasibility

90000 9
80000 Feasibility 8
70000 Weight 7
60000 6
Feasibility

Weight (t)
50000 5
40000 4
30000 3
20000 2
10000 1
0 0
-10000 -1
HE450B IPE550

HE320B IPE500

HE600B IPE300
HE400B IPE300

HE340B IPE270

HE400B IPE180
HE500B IPE750x137

HE800A IPE550

HE450A IPE360

HE360A IPE220
HE550A IPE750x223

HE260A IPE750x197

HE260A IPE750x137
HE1000A IPE750x223

HE340AA IPE600
HE340AA IPE550

HE300AA IPE500

HE280AA IPE400

HE340AA IPE270

HE140AA IPE270
HE600AA IPE750x137

HE400AA IPE750x197

HE360AA IPE750x147

(a)
7
6.75
6.5
6.25 Weight avg.
6
5.75 Fitness avg.
5.5 Elite fitness
5.25
5
Weight (t)

4.75
4.5
4.25
4
3.75
3.5
3.25
3
2.75
2.5
2.25
2
1
3
5
7
9
11
13
15
17
19
21
23
25
27
29
31
33
35
37

39

Generation
(b)
Figure 5.11. HR frame optimization results for persistent/transient design situation, axis-to-axis span 16 m,
height 6 m, roof angle 15%, haunch length 1/11 of the span, distance between frames 6 m, pinned supports,
dead load 380 N/m2, wind load 30 m/s, terrain type 2, steel S275 with strain hardening, design method
General Method, tournament selection, simulated binary crossover, polynomial mutation

According to the load settings the script automatically selects if vertical or horizontal loads are the
incremental loads (i.e. leading action). If snow is leading load, then the structure is preloaded with other
accompanying loads (e.g. wind, crane load), and the snow load is gradually increased until failure. If e.g.
earthquake is the leading load, then the structure is preloaded with the fraction of the snow load, and
earthquake load is increased until failure.
In order to achieve the best performance when calculating multiple frames, several jobs can be submitted by
AP-Frame at the same time, figure 5.13. The computer memory is redistributed according to the number of
currently opened processes. When all jobs are submitted, AP-Frame switches to a CONTROL LOOP where
is monitoring the progress of calculation and terminating jobs if necessary. In order to increase the speed of
calculation AP-Frame includes method to monitor and terminate a non-linear analysis when enough data is
collected. The load decrease is usually recognized by three consequent descending steps.
Standard monitoring of running jobs implemented in Abaqus/CAE could not be used because they were not
stable when monitoring several jobs at the same time. The control method is checking for existence of status
119
file (e.g. “Job-1.sta”) and lock file (e.g. “Job-1.lck”) which indicates that the calculation is running. Load
step increments are obtained by reading the contents of status file. In UNIX based systems simple deleting of
all job files terminates the running analysis immediately. On the other hand, the running “standard.exe”
process has to be terminated in Windows. Both methods proved to be stable for most of the calculations. In
case of unexpected problems (e.g. licence errors, database is corrupted, etc.), the AP-Frame automatically
restarts the calculation. When no results are obtained within a given time-out limit, the whole analytical part
is restarted.

BEGIN

Seismic (seismic, dead, 0.2x snow) Standard (dead, snow, wind)

Loading
PHASE -1 (EQ)

LA
Column stifness
(Static, wire)

Method
GENERAL GLOBAL
METHOD ANALYSIS
PHASE 1 PHASE 1
LBA out-of-plane LBA
Critical multiplier
(Buckling, shell) (Buckling, shell)

PHASE 2
1st ORDER 2nd ORDER
Nonlinearity

AUTO

LBA in-plane
LBA
(Buckling, wire)

Critical multiplier Perturbed shape Perturbed shape

PHASE 3 PHASE 2
NO 2nd order YES
criterion?

ELASTIC PLASTIC ELASTIC PLASTIC ELASTIC PLASTIC

Plasticity Plasticity Plasticity

AUTO AUTO

classification classification classification

3, 4 1, 2 3, 4 1, 2 3, 4 1, 2
Class Class Class

LA MNA GNIA-y GMNIA-y GNIA GMNIA

(Static, wire) (Riks, wire) (Riks, wire) (Riks, wire) (Riks, shell) (Riks, shell)

PHASE 4 PHASE 3

Ultimate multiplier

Maximum design
load

END

Figure 5.12. ABAQUS script AP-Frame

120
EV-Frame – the VBA script using Microsoft Excel
EV-Frame was developed in VBA under Microsoft Excel, to cover the same inputs as AP-Frame, figure
5.13. EV-Frame uses 1st order elastic calculation to determine internal forces, and buckling reduction factors
for performing design checks. The frame is modelled with elastic beam elements only (even though sway
imperfections are taken into account). EV_Frame is much faster then AP-Frame, which is based on 2nd order
nonlinear calculations.
In our study, EV-Frame was mainly used for benchmarking and testing convergence of optimization
algorithms. Its level of conservativeness in the design is higher than that of AP-Frame. Since it is
conservative in design (i.e. it predicts lower load bearing capacity & produces heavier frames), it is not
perfect to optimize frames. The structural analysis part of EV-Frame is modification of open-source finite
element solver under the GNU General Public Licence.

Read the input file

BEGIN input.txt
into memory

next process next process

no all processes yes

Process loop begin Control loop begin
running?

yes Release the yes job

closed?
process finished?
no
no
yes
running? yes
descending
Terminate the job
no results?

Evaluate last results no

and submit a new job
Control loop end

yes Write yes

last phase? last frame?
results
no no

next phase next frame Close process

output.txt
database.txt
report.txt
Process loop end

report-mfr.txt

yes any process no

Write overall report END
open?

Figure 5.13. ABAQUS script parallel processing

Design method algorithms

AP_Frame and EV_Frame implement processes for portal frames design in fundamental and seismic load
situations, using the limit states conditions of EN 1993 and EN 1998, figure 5.14. Both allow automatic
generation of computational models and automatic result evaluation. The results of the design checks are
expressed as a minimum load amplifier of the design loads to reach the ultimate (ULS) or serviceability
(SLS) limit state criteria. The vertical serviceability limit corresponds to an apex deflection of span/200,
whereas the horizontal serviceability limit is height/100, derived in a simplified way from EN 1998.
The design methods follow EN 1993, where the sway imperfections are applied (if needed) to the model
prior to calculation of in-plane amplifier which decides about the order of analysis. The optimization tools
are designed to work mainly with slender frames, where elastic analysis should be performed. Sway
121
imperfection is applied by modifying the model geometry, instead of the commonly used load amplification
method. I.e. the Positions of all nodes are shifted horizontally in order to achieve calculated global initial
sway imperfection at eaves height. There is no check whether sway imperfections could be neglected and
they are applied automatically.

Figure 5.14. The first view of EV-Frame tool when started, showing the geometry tab

Method 1: Global nonlinear analysis (GNLA)

The most computationally expensive method implemented in AP-Frame is the numerical calculation of 3D
shell model of the frame (gray boxes in the flow chart in figure 5.15). No special checks for out-of-plane
stability are needed, because the calculation is materially and geometrically nonlinear and it is taking into
account the appropriate initial bow imperfection. GMNIA is already the second-order analysis and therefore
no further checks of in-plane critical amplifier are needed. Imperfections are inserted from a preliminary
buckling analysis, using the first positive buckling shape scaled to the EN 1993 recommended amplitude (i.e.
using Table 5.1 from EN 1993).
Method 2: General method analysis (GMA)
This method takes into account out-of-plane stability with a global reduction factor. The out-of-plane critical
load is calculated using a 3D shell model of the frame, taking into account cross-section variation and
support condition. Method 2 has the advantage of faster calculation compared to Method 1. The
computational advantage results from the use of the simplified 2D beam model for the nonlinear part of the
analysis.
Method 3: EN 1993 interaction formulae
Linear structural analysis together with design using cross-sectional checks to express limit state conditions
are the commonly used methods described in EN 1993. However, not all of the EN 1993 rules are applicable
to elements with variable cross-section and eccentric lateral restraints, as portal frames are. In order to
increase accuracy, additional theoretical considerations were implemented in the calculation. Even though
the method is the most computationally effective, it is more conservative especially in its out-of-plane
stability approach where no lateral supports of compressed flanges are used by default.
Optimization methods
In optimization terms, the portal frame problem is both nonlinear and discrete, which causes difficulties for
classical direct and gradient-based optimization methods. Portal frame optimization is more suited for
genetic algorithms (GAs). GA’s have also been successfully applied in the field of structural optimization
with several advantages, such as possibility of parallel computing, easy handling of multiple variables, and
straightforward coding practice.
Genetic algorithm is a procedure which tries to mimic the natural evolution process. After an initial
population is created and analysed, the fitness for purpose of each individual is evaluated. Then a new
population is created by favouring the fittest individuals and by combining the properties of the population
members using genetic operators, such as crossover and mutation. GA proceeds iteratively towards the
optimal solution by creating a new population using the properties of the previous one. Elitism ensures that

122
the best individual/solution is kept during all of the genetic operations. While no proof of convergence of
GA’s to an absolute optimum exists (at least not in finite time), good results are found in a reasonable time.
The real-coded genetic algorithm (RCGA), used for our problem, can handle discrete and real variable types
easily. In EV-Fame, besides the RCGA procedure, the particle swarm optimization (PSO) has also been
implemented and tested. With RCGA, the coding-decoding characterizing binary-coded GAs is also avoided,
therefore it is closer to handling engineering quantities in a simpler data format.

Elastic analysis

Start
EC3 §5.2.1(3) METHODS
METHOD 3 1 AND 2 ARE
AUTOMATICALLY
SECOND ORDER
YES Dcr ≥ 10 NO

EC3 §5.2.2(5)B
Dcr ≥ 3
DESIGNED AS
CLASS 3 OR
LOWER Increase actions
YES
on structure

EC3 §5.5 Class 4

YES Geometry
section? NO
YES OK for amplified
loads?
Determine
NO Not feasible in
effective CS
properties Method 3
NO
Determine design
resistance of CS Second order by
analysis model
EC3 §6.2.2.5 EC3 §5.3.2(3)
METHOD 2
Design A B
EC3 §6.2 resistance NO METHOD 1
OK? Apply in-plane
bow imperfections
Apply all bow
YES imperfections

Determine buckling NOT

Calculate ultimate
load amplifier Dult,k
Revise
resistance of CS FEASIBLE

Calulate critical
Buckling
Second order
global analysis
out-of-plane Dcr,op
EC3 §6.3.3(4) amplifier
resistance NO
OK?

Design General
resistance NO NO method
YES OK? check?

YES YES EC3 §6.3.4(1)

Stop

Figure 5.15. Elastic analysis algorithm with Method 1 (GMNIA) – blue, Method 2 (GM) – green and Method
3 (Cross-sectional checks) – yellow

Optimization literature provides a large catalogue of different selection, crossover and mutation operators
that can be combined to create a GA suited for the problem at hand. In this report, the well known simulated
binary crossover (SBX) and parameter based polynomial mutation operator are utilized. The crossover
operator has a self-adapting behaviour, which favours creating children near to parents, when the parents are
123
near to each other in the variable space. The basic behaviour of the genetic algorithm is enhanced by two
methods developed to improve the steel portal frame optimization:
‚ a local search method is creating individuals very similar to the current elite individual found, which
ensures that the local optimum is found with high probability;
‚ the diversity of the population is maintained by using so called diversifying operator, which
introduces new genetic material to the population preventing premature convergence to the local
optimum.

5
1
4.5 2
Frame mass[tons]

3
4
4
3.5 5

2.5

2
0 10 20 30 40 50
Generation
Figure 5.16. Evolution of the elite frame fitness in five optimisations of the same frame configuration with
the GA

In terms of optimization objective function, it has been opted to concentrate on easily measurable
performance parameters, such as weight, with the flexibility to expand results to more financially oriented
targets (e.g. price). The algorithm can be used with all of the design methods and the flowchart of the
optimization combined with design methods is presented in figure 5.15.
AP-Frame contains the ABAQUS based design methods and optimization, whereas EV-Frame refers to the
design and optimization developed in a Microsoft Excel workbook. In both cases, the GA runs for
predetermined number of generations, and the best configuration found is given as an output. Figure 4.16
presents five example optimisation runs of the same frame configuration, which shows how the weight of the
elite individual typically decreases during the genetic algorithm optimisation. In this example, eight variables
were optimized in a welded-tapered frame with population size 20 and maximum 50 generations. The
number of variables can be larger as some parameters treated as constants can become variables themselves.
In the procedure limits/constraints based on the physically available values, can also be set for the variables
e.g. certain plate thicknesses, material grades etc.).
As it can be seen in figure 5.16, it is typical that the weight of the elite frame in the population decreases
sharply during the first generations, and then the search focuses around the elite individual refining the
solution. The slow decrease of weight usually corresponds to instances when the procedure is finding better
solutions around an optimum typology, white sudden drop in fitness occurs when the algorithm finds a better
typology, meaning that many variables can change drastically. In other words, the alternative typologies
correspond to regions of local minimum/optimum in the search space, and the sudden drop occurs when the
GA finds better region of local minimum, with fundamentally differently configured frames. In fact, GA’s
have this advantage over gradient based optimisation methods, not to be trapped searching around a local
minimum.
Implementation of optimization methods
The optimization process was developed around AP-Frame script to optimize portal frames using the genetic
algorithms, figure 5.17. Optimization starts with a randomly generated population of portal frame
configurations, based on the constants and range of variables defined for that problem e.g. span=20m,
height=6m. The individuals of this population are analysed using one of the design methods (Method1,
Method2, Method3) and their feasibility is evaluated. Negative feasibility of an individual means that the
124
frame fails one or more design check, feasibility a bit over 0 means that the frame/individual has just passed
all design checks, and large positivity feasibility means a frame/individual with large reserves in passing the
design checks. Fitness is then assigned to each individual/frame based on its feasibility and target objective
i.e. small weight. Generally, a more complex objective function can be used (e.g. price) if it can be defined.
In case of this project weight has been used as a simple measure of good performance. However, design and
optimisation procedures have been prepared for use with more general objective functions as they are
reporting several physical properties of the frame (e.g. weight, length of welds, surface area for painting
etc.). Unfortunately, the definition of a price based objective function was not possible, due to the large
diversity of targeted market countries.

Figure 5.17 Flowchart of the GA optimization process

In the next step, based on the fitness of individuals in the initial population, a new population is created using
genetic operators: selection, crossover and mutation. Well accepted GA operators have been used in this
study, , with some modifications implemented both to operators and to the GA procedure, in order to better
fit the portal frame problem.
This new modified population is evaluated and genetic operations are performed again. The algorithm
proceeds iteratively towards optimal solution while always storing the best solution found. Each population
of portal frames is sent to evaluation by writing the input.txt, and after the completion of the structural
analysis part, the output is read from output file. Therefore, form the point of view of the genetic solver, the
output file contains the input for the GA, while in the input file the output is written after each GA step.

125
 Figure 5.18. Detailed flowchart of the GA optimization

Optimization algorithm for EV-Frame tool was implemented in Visual Basic for Applications, figure 5.18,
and objective function is calculated using the frame calculation sheets. In this case, given the freedom to use
the calculation sheets, it is possible to allow the user to define a “custom fit” objective. General optimization
parameters affect both Genetic algorithm (GA) and Particle swarm (PSO) optimization methods. Advanced
options are available for users understanding the basics of the optimization methods.
The most important optimization parameters are the population (generation) size and maximum number of
generations. These two parameters affect the optimization time and performance. In general, longer runs of
the algorithm, i.e. larger population size and/or maximum number of generations, are more likely to find
good results.
Other optimization parameters also affect the optimization and some gains can be achieved by adjusting
these parameters. Furthermore, if the user is familiar with GA or PSO, additional options are available in
“PSO”, “GA” and “GAW” worksheets. For example, multiple runs of the genetic algorithm with the same
parameters can be performed by setting “Number of optimization runs” to desired number. The elite value
found in each consecutive run will be saved in “GA_runs” or “GA_W_runs” sheet. However, default
GA/PSO parameters should give satisfactory performance.

126
Figure 5.19. View of optimization tab in the EV-Frame tool.

127
 6. Design guidelines and final design of applicative examples

6.1 Design Guidelines for the IB modulus

The proposed module for the industrial buildings consists of five single/double span frames which are
repeated in the out-of-plane direction having a constant distance between them. (Figures 6.1 and 6.2)

Figure 6.1. Basic module for the one bay industrial building configuration.

Figure 6.2. Basic module for the two bay industrial building configuration.

The distance between the consecutive frames in the out-of-plane direction is 6 m. In the in-plane direction,
the moment resisting frames withstand the horizontal actions. In the out-of-plane direction a Bracing System
for every five consecutive frames is responsible to carry all the horizontal forces. The value used for the
roofing slope is 15% which corresponds to 8.53°. Steel material S275 is used for the hot-rolled profiles and
welded-tapered sections while the FeE 350G steel grade is used for the light-gauge elements.

6.1.1 Cross-sections used

The IB modulus is characterized by fixed profile series for each structural member according to the
following assignation:
‚ HEA profiles are used for the columns of the structure.
‚ IPE profiles are used for the beams of the structure.
‚ CHS profiles are used for the bracing system of the structure.
‚ IPE 140 every 2 m is used for the purlins that support the roof cladding in case of low snow loading (0.75
kN/m2).
‚ IPE 180 every 2 m is used for the purlins that support the roof cladding in case of high snow loading (1.5
kN/m2)
‚ IPE 120 every 1.5 m is used for all the purlins that support the side cladding.
Moreover, in some configurations the haunches are used to strengthen the beam cross-section only at the
points where the beams are connected with the columns of the structure and to resist high bending moments
in the corner area. The height of the haunch cross-section at the end of the beam element can be assumed as

129
 1.5 – 2 times the height of the beam cross-section that is used for the rest part of the beam. The other cross-
sectional characteristics (tf, b, bw, etc.) are taken same as the characteristics of the steel profile used for the
beam. The length of the haunch is taken between L/5 – L/6, where L is the length of the half frame opening
in the in-plane direction.
In the solutions employing light-gauge steel members, the profiles used for the realization of the light-gauge
steel solutions belonged to the KONTI 2B and KONTI C library (Figure 6.3)

KONTI 2B KONTI C
Figure 6.3. Cold-formed section profile adopted for IB modulus.

Similarly to the solution with hot-rolled profiles, also in this case the purlins spacing and the profile selection
were made in order to guarantee an adequate safety level; in particular the following profiles are used:
‚ KO-2KB 250X15 is used for the purlins that support the roof cladding in case of low snow loading (0.75
kN/m2)
‚ KO-2KB 250X20 is used for the purlins that support the roof cladding in case of high snow loading (1.5
kN/m2)
Haunched solutions for the light-gauge steel solutions are not adopted.
The sections used for the realization of the trussed beam are double L profiles (Figure 6.4).

Figure 6.4. Typical section for the adopted trussed girder solution.

6.1.2 Welded-tapered
Welded-tapered frames are generally composed of welded I-profiles with variable height. Column is tapered
and there is also a haunched part of the beam near the corner. Cross-sections were selected with the
following restrictions:
‚ Cross-section class is max. 3 and the local buckling failure is not critical in the welded-tapered frames.
‚ Width of the flange, flange thickness and web thickness are the same in column and rafter.
‚ Column height at the base is always the same as frame width.
‚ Haunch height is at least 1,5 times bigger than the minimum height of the beam or the column.

6.1.3 Trussed solutions

The elements used for the realization of the trussed beam are grouped in four groups (Upper, Lower,
Diagonal, Ortho) as shown in the next scheme. Two different profiles are used for groups U, L, D and O
(Figure 6.5).

130
 U

L D O

Figure 6.5. ID scheme for element in trussed girder solution.

6.1.4 Welded tapered solutions

The welded tapered solutions were designed considering the in-plane support of frame as pinned; therefore
the smallest possible cross-section was used at the column base. There are no lateral supports of inner flange
in modeled frames and the cross-sections are only supported 120 mm from the upper flange simulating
purlins and side rails. Crane loads were not modeled for welded-tapered frames because their sensitivity to
the lateral loads (Figure 6.6). All results were calculated using General Method (EN 1993-1-1) elastic
analysis with out-of-plane critical loads from the buckling analysis of 3D shell model. Geometry of welded-
tapered frames including the haunch length is optimized using Genetic Algorithm.

Lb

ab D
ac

Lh
tw
y
H

h
Lc

S/2
tf
0,5 m

b
z
Figure 6.6. Geometrical and structural scheme of welded-tapered frame considered in the Precasteel database

6.1.5 Type and intensity of actions

The structures contained in the Precasteel database were designed assuming the loads intensity reported in
the table 6.1. It is important to remind that the use of database allows to obtain solutions completely verified
and optimized for those combinations of loads.
In particular, it is worth reminding that the crane load was considered only in those structural configurations
(i.e. geometry) that more frequently from statistical analysis showed crane load presence.

Seismicity (p.g.a.) High Medium Low

0.32 g 0.16 g 0.08 g
Snow loads High Low
1.5 kN/m2 0.75 kN/m2
Wind load
Average EU reference velocity value: 30 m/s
Crane loads
0 kN 50 kN 100 kN 250 kN
Cladding
0.15 kN/m2
Table 6.1. External actions considered in the structural design
131
6.2 Corrugated web girder: Design guidance
The general layout of the corrugated web girders (CWG) with bolted web-to-flange connections as
introduced in chapters 2 and 4, respectively, is depicted figure 6.7. Standard details that have to be checked
in design include connections between web and flange (here: 2 M10 12.9 each corrugation), longitudinal
joints of the web (here: blind rivets), and load introduction in case concentrated loads are present (here:
hollow profile along the depth of the girder, connected to the web by means of blind rivets and load
distributing plate).

Figure 6.7. Views of proposed layout of CWG

Design methods for CWG with welded web-to-flange connection under static loading are available in EN
1993-1-5, Annex D. The given design checks have been applied to the CWG with bolted web-to-flange
connection depicted in figure 6.7. Experimental testing showed that once the connection web-to-flange is
capable of transferring the respective shear flow this procedure is suitable also for this type of girder.
In the following, the design rules according to EN 1993-1-5 are outlined and a method for the determination
of the CWG’s deflection is given.

6.2.1 Design of the web

The geometric variables and denominations are adopted from EN 1993-1-5, see Figure 6.8. The shear
resistance VRd of the web is determined as
f yw
VRd Fc hwt w [6.1]
J M1 3
where
fyw yield strength
JM1 Partial safety factor

132
 Figure 6.8: Geometry according to EC 3 [image source: EN 1993-1-5]

The reduction factor Fc is determined as

Fc min( F c,l , F c, g ) [6.2]
where index ‘l’ stands for ‘local buckling’ and index ‘g’ stands for ‘global buckling’ of the web. The
reduction factors can be calculated by evaluation of the formulae given in EN 1993-1-5.

6.2.2 Design of the flanges

Design of the flanges is governed by the aspired moment resistance of the CWG. According to EN 1993-1-5,
the moment resistance can be calculated as follows:
­ b2t2 f yf ,r § t  t · b1t1 f yf ,r § t  t · b1t1 Ff yf ,r § t  t ·½
M Rd min ® ¨ hw  1 2 ¸; ¨ hw  1 2 ¸; ¨ hw  1 2 ¸¾ [6.3]
¯ JM0 © 2 ¹ JM0 © 2 ¹ J M1 © 2 ¹¿
In equation 10, the first term represents the yielding resistance of the tensile flange, and the second term is
the yielding resistance of the compression flange. The third term represents the buckling resistance of the
compression flange. Evaluation of these formulae is carried out according to EN 1993-1-5. It has to be
considered that the CL-profiles used in the present campaign are categorised cross sectional class 4, thus
effective values for cross sectional resistances have to be used in design.

6.2.3 Design of connections web-flange

As will be discussed in detail in the following chapter, the consideration of shear deformations of the web is
of importance when the deflection of the CWG is determined. However, the connections web-to-flange can
be designed under assumption of a rigid cross section without consideration of shear deformations. This
approach is safe sided with regard to the shear flow between web and flange. As described in chapter 4, it is
generally favourable to oversize the web-flange connection as this leads to a more ductile failure mode.
Accordingly, the connectors are designed to resist the shear flow
VEd ˜ S y ,F
T [6.4]
I yy
where Sy,F is the statical moment of the flange.
If it is assumed that Iyy is dominated by the flanges,
I yy ACLh 2 [6.5]
where
ACL = effective cross sectional area of one CL-profile
h = spacing between the centres of gravity of the CL-profiles
then equation 11 can be expressed as
VEd
T [6.6]
h
133
However, in order to reduce peeling effects and to allow for redistribution of force during failure, it is
favourable to oversize the connection. For example, for the specimens tested in the experimental campaign
described in chapter 4 the shear resistance of the connection was oversized approximately 25 %. A suitable
arrangement of bolts should consider the torsional deformations of the corrugations causing peeling effects
of the connection. As described in chapter 2, at least two fasteners per corrugation should be used in order to
fixate the flanges of the trapezoidal sheeting to the CL-profiles.

6.2.4 Calculation of deflection

As described in chapter 4, the deformation of the CWG was governed by shear deformations of the web.
Accordingly, these deformations have to be taken into account in the design. Figure 6.9 shows the load-
deflection-curve of specimen 2 at a level of transverse force of VEd = 100 kN. It can be seen that if the
bending stiffness of the girder is assumed to be equal to I rigid ACL ˜ h , i.e. no shear deformation is taken
2

into account, the corresponding deflection of wrigid = 0.28 cm shows a significant difference to the result of
the experiment of wexp = 10.9 cm. However, the calculation method described in the following is able to
reproduce the deflection measured in the experiment, leading to a value of w calc = 10.8 cm. A detailed
description can be found in [1].

Figure 6.9: Deflection of specimen 2 at transverse force VEd = 100 kN

6.2.4.1 Basic principles

The calculation is based on the definition of the cross-sectional displacement
vx ( x, s) Z (s) ˜ U c( x) [6.7]
where x indicates the beam’s coordinate and s represents the cross-sectional coordinates in the sense of a
contour coordiante. Figure 6.10 illustrates the kinematic scheme between warping function, here Z(s) = -z,
and weighting function, here U’(x) = w’(x) on the example of the conventional technical bending theory.

Figure 6.10: Warping- and weighting function

134
By differentiation and the assumption of a linear-elastic material, expressions for normal and shear strains as
well as for stresses can be obtained:
H ( x, s) Z ( s) ˜ U cc( x) V ( x, s) E ˜ Z ( s) ˜ U cc( x)
[6.8]
J ( x, s) Z ( s) ˜ U c( x) W ( x, s) G ˜ Z ( s) ˜ U c( x)
To avoid the presence of an unknown resulting warping function, a series expansion over a set of arbitrary
warping functions k (s) is carried out:
u ( x, s) k
¦
Z ( s)˜k U c( x)
k
[6.9]

In the expression for the elastic potential of the bridge body

3 ³ (V ˜ H  W ˜ J )dV  ³ p ˜ v dO
V O
[6.10]

where p stands for external forces and v is the displacement complement, the stresses and strains can be
substituted by warping- and weighting functions. The postulation that in an equilibrium state the variation of
the elastic potential shall be equal to zero then leads to
l
§ ·
G3 ¦¦ ³ ¨¨ E U ccG U cc³ i k i
Z k Z dA G iU cG kU c³ i Z k Z dA G kU p ¸¸dx 0 . [6.11]
k i x 0 © A A ¹
The execution of the variation for the unknown weighting functions kU(x) leads, after partial integration, to a
set of differential equations which can be summarised in a matrix equation
EC0U cccc  GC1U cc p [6.12]
where
³ Z k Z dA ³ Z k Z dA
i ,k i i ,k i
C0 C1 [6.13]
A A
are resistance matrices that are mobilised by the cross-sectional warping functions used. As the warping
functions are arbitrary, these resistance matrices can be obtained just after the warping functions are chosen.
In general, the resistance matrices C0 and C1 have members on the main diagonal as well as on other
positions, i.e. the differential equation system represents a coupled system in general. To eliminate the
members outside the main diagonals a principal axis transformation is carried out by solving the general
eigenvalue problem
GC1  OEC0  r 0 . [6.14]
A matrix K, which contains the eigenvectors r as solutions of equation 18, can then be used to ‘rotate’ the
resistance matrices and to obtain a diagonalised equation system with new warping functions as generalised
coordinates of the equation system
~ ~ ~~ ~
EC0U cccc  GC1U cc p [6.15]
where
~ ~
C0 K tr C0 K C1 K tr C1 K
[6.16]
are the resistance matrices of the orthogonal warping functions. The equation system can now be solved line-
wise, each line representing an independent linear differential equation. The generalised load vector is an
outcome of the transformation of equation 16 using principal axis rotation: ~
p K tr p .

6.2.4.2 Appliance to CWG

The calculation method described above can be applied to CWG by using three warping functions in the
series decomposition of equation 13. The first one corresponds to normal force, the second one to bending of
the rigid cross section without consideration of shear deformations, while the third one represents a shear
deformation of the web of the CWG, Figure 6.11. The choice of Z3 exhibiting a zero crossing at the girder’s
center of gravity simplifies the problem as Z1 and Z3 are decoupled.

135
 Figure 6.11: Choice of warping functions

The members of the resistance matrices of equation 6.14 are determined as follows:
4ACL 0 0
C0 = 0 4(I0,CL+ACLh²/4) 4*1/2*ACLh/2
0 4*1/2*ACLh/2 4* (1/2)²
*ACL
0 0 0
GC1 = 0 0 0
0 0 hG ²
where I0,CL is the moment of inertia of a single CL-profile and G is the shear modulus of the trapezoidal
sheeting given by the manufacturer.
Solution of the general eigenvalue problem of equation 18 leads to diagonal stiffness matrices. In this case
where Z3 is independent from Z1 with regard to warping resistance, the transformation matrix K contains
only one entry outside the main diagonal:
1 0 0
K= 0 1 K2,3
0 0 1

The resulting deflection of the girder is determined according to Ures(x) = 2U(x)+K2,3* 3U(x), Figure 6.9.

6.3 Design guidance of precast r.c. walls in CBs

In the present paragraph some applicative examples about the design and employment of precast reinforced
concrete walls to CB solutions are presented; this guide served as a final support for the preliminary sizing
software developed inside the project.
The examples, presented hereafter, are characterized respectively by different seismic design hypothesis:
dissipative or not dissipative walls; walls coupled with dissipative devices.

6.3.1 Study case 1 – Dissipative r.c. walls

This is a two-storey Commercial Building, covering an area for each floor (figure 6.12) of:
ACBs = D x L = 36 x 50 = 1800 m2
First of all, the intensity of live loads applied to CBs structures are:
Qk1 = 5 kN/m2 (standard live load, on the first floor)
2
Qk2 = 2 kN/m (snow live load, on the second floor)
The site seismicity level (PGA, peak ground acceleration) is:
ag = 0.32 g (high seismicity area)
The distribution of the storey forces, with reference to the numbers of the stories and the nature of actions
(seismic), is type “C”.
Ductility class of the structure assumed is “DCH” for dissipative structures under cyclic loads generated by
an earthquake.
A correct estimation of behaviour factor “q” of the structure depends on ductility class and geometrical
properties of the r.c. wall (storey height, width, thickness):
H = 5.00 m (interstorey height)
136
B = 6.00 m (width of the r.c. wall)
s = 0.25 m (thickness of the r.c. wall)
then, according to Eurocode 8 provisions, the behaviour factor is assumed:
q = 3.56 (behaviour factor)
Refining the range database about shear horizontal loads (earthquake, wind) every 250 kN, the best fitting of
a dissipative r.c. wall system is reached for:
Vb = 1250 kN (base shear for a single r.c. wall)
that is related to a wall influence area of:
Awall = 444 m2
Finally, the number of r.c. dissipative walls for each floor and direction (X, Y) is:
nwall = ACBs / Awall = 1800 / 444 ≈ 4

FIRST FLOOR SECOND FLOOR

Figure 6.12: Study case n°1 (dissipative r.c. walls).

Here are reported briefly, according to Eurocode 2 and Eurocode 8 suggestions and provisions, all the main
structural design and verifications (ULS) about r.c. wall systems.
Material properties:
concrete C30/37 (Rck=37 MPa; fck=30 MPa; fcd=15.94 MPa)
steel B450C (fyk=450 MPa; ftk=540 MPa; fyd=391 MPa)
Geometrical properties:
s = 0.25 m (r.c. wall resistant thickness)
B = 6.00 m (width of the wall)
H = 5.00 m (interstorey height of the walls)
Actions and verifications (ULS) on r.c. wall system:
Bending moment → MEd = (0.38·1250)·5.00 + (0.62·1250)·10.00 = 10125 kNm ≤ MRd = 11524 kNm
Shear force → VEd = 1.5·Vb = 1875 kN ≤ VRd = min (VRsd, VRcd) = 1889 kN
The above ULS verifications are referred to the detailed reinforcing bars and stirrups related to the following
technical drawing (Figure 6.13), with:
As,bending = 11+11‡14 (critical region, confined zone lc = 0.15·B = 0.90 m)
As,shear = 1‡8/10cm

137
 Figure 6.13: Typical technical drawing for r.c. wall.

As said in the past paragraphs, about connections between steel structure (beams, columns) and r.c. walls,
there are two possibilities:

6.3.1.1 Connection decoupling horizontal and vertical loads

In this case we need an additional auxiliary beam, that transfers gravity loads towards the main steel
columns, we can connect our bracing system to the steel frame decoupling vertical and horizontal loads
(Figure 6.14).

Figure 6.14. Position of connective zones between wall and gravity structure

6.3.1.2 Connection for both horizontal and vertical loads

In this case we don’t need an additional auxiliary beam and we can connect directly our bracing system to
the steel frame; then, it is possible for the walls support even vertical loads without compromise their seismic
behaviour (Figure 6.15).

Figure 6.15. Connection between r.c. wall and steel beams of the floor system

138
6.3.2 Study case n°2 - Dissipative r.c. walls
This is a two-storey Commercial Building, covering an area for each floor (Figure 6.16) of:
ACBs = D x L = 36 x 66 = 2376 m2
First of all, the intensity of live loads applied to CBs structures are:
Qk1 = 5 kN/m2 (standard live load, on the first floor)
Qk2 = 2 kN/m2 (snow live load, on the second floor)
The site seismicity level (PGA, peak ground acceleration) is:
ag = 0.32 g (high seismicity area)
The distribution of the storey forces, with reference to the numbers of the stories and the nature of actions
(seismic), is type “C”.
Ductility class of the structure assumed is “DCH” for dissipative structures under cyclic loads generated by
an earthquake.
A correct estimation of behaviour factor “q” of the structure depends on ductility class and geometrical
properties of the r.c. wall (storey height, width, thickness):
H = 5.00 m (interstorey height)
B = 6.00 m (width of the r.c. wall)
s = 0.25 m (thickness of the r.c. wall)
then, according to Eurocode 8 provisions, the behaviour factor is assumed:
q = 3.56 (behaviour factor)
Refining the range database about shear horizontal loads (earthquake, wind) every 250 kN, the best fitting of
a dissipative r.c. wall system is reached for:
Vb = 1750 kN (base shear for a single r.c. wall)
that is related to a wall influence area of:
Awall = 618 m2
Finally, the number of r.c. dissipative walls for each floor and direction (X, Y) is:
nwall = ACBs / Awall = 2376 / 618 ≈ 4
Here are reported briefly, according to Eurocode 2 and Eurocode 8 suggestions and provisions, all the main
structural design and verifications (ULS) about r.c. wall systems.
Material properties:
concrete C30/37 (Rck=37 MPa; fck=30 MPa; fcd=15.94 MPa)
steel B450C (fyk=450 MPa; ftk=540 MPa; fyd=391 MPa)
Geometrical properties:
s = 0.25 m (r.c. wall resistant thickness)
B = 6.00 m (width of the wall)
H = 5.00 m (interstorey height of the walls)
Actions and verifications (ULS) on r.c. wall system:
Bending moment → MEd = (0.38·1750)·5.00 + (0.62·1750)·10.00 = 14175 kNm ≤ MRd = 15384 kNm
Shear force → VEd = 1.5·Vb = 2625 kN ≤ VRd = min (VRsd, VRcd) = 2699 kN
The above ULS verifications are referred to the detailed reinforcing bars and stirrups related to the previous
technical drawing (Figure 6.13), with:
As,bending = 12+12‡20 (critical region, confined zone lc = 0.15·B = 0.90 m)
As,shear = 1‡10/10cm

139
 6000
FIRST FLOOR SECOND FLOOR

6000
12000

12000
12000

12000
6000

6000
12000

12000
12000

12000
6000

6000

6000 6000 12000 6000 6000 6000 6000 12000 6000 6000

Figure 6.16: Study case n°2 (dissipative r.c. walls).

6.3.3 Study case N°1 (Dissipative devices + Elastic r.c. walls)

In this paragraph, we examine study case n°1 considering that r.c. walls must maintain an elastic response
towards cyclic horizontal actions (seismic) and the largest part of energy due to earthquake has to be
dissipated into specific HDR devices.
This is a two-storey Commercial Building, covering an area for each floor (figure 6.12) of:
ACBs = D x L = 36 x 50 = 1800 m2
First of all, the intensity of live loads applied to CBs structures are:
Qk1 = 5 kN/m2 (standard live load, on the first floor)
Qk2 = 2 kN/m2 (snow live load, on the second floor)
The site seismicity level (PGA, peak ground acceleration) is:
ag = 0.32 g (high seismicity area)
The distribution of the storey forces, with reference to the numbers of the stories and the nature of actions
(seismic), is type “C”.
Ductility class of the structure assumed is “DCL” for non dissipative r.c. structures under cyclic loads
generated by an earthquake.
Geometrical properties of the r.c. wall (storey height, width, thickness) are:
H = 5.00 m (interstorey height)
B = 4.50 m (width of the r.c. wall)
s = 0.25 m (thickness of the r.c. wall)
The behaviour factor, in this case, is assumed:
q = 1.00 (behaviour factor, elastic structure)
Horizontal forces at the base of each floor and for both directions (X, Y) are:
FX,1 = 280 kN

140
FX,2 = 360 kN
FY,1 = 320 kN
FY,2 = 400 kN
In the case of the r.c. wall in direction Y (more stressed than direction X), here are reported briefly,
according to Eurocode 2 suggestions and provisions, all the main structural design and verifications (ULS)
about r.c. wall systems.
Material properties:
concrete C30/37 (Rck=37 MPa; fck=30 MPa; fcd=15.94 MPa)
steel B450C (fyk=450 MPa; ftk=540 MPa; fyd=391 MPa)
Geometrical properties:
s = 0.25 m (r.c. wall resistant thickness)
B = 4.50 m (width of the wall)
H = 5.00 m (interstorey height of the walls)
Actions and verifications (ULS) on r.c. wall system:
Bending moment → MEd = 320·5.00 + 400·10.00 = 5600 kNm ≤ MRd = 5770 kNm
Shear force → VEd = Vb = 320 + 400 = 720 kN ≤ VRd = min (VRsd, VRcd) = 944 kN
The above ULS verifications are referred to the detailed reinforcing bars and stirrups related to the next
technical drawing (Figure 6.17), with:
As,bending = 9+9‡16
As,shear = 1‡8/15cm
Stirrups 1Ø8/15 Stirrups 1Ø8/15
0

0
10

10
Stirrups 1Ø8/15 Stirrups 1Ø8/15

Ties Ties
0

0
10

10

18Ø16 vertical Lattice girder 1Ø8/15 horizontal 1Ø8/20 vertical 18Ø16 vertical

40
250

330
40

Precast concrete slab Electrowelded mesh

4500

Figure 6.17: Technical drawing for r.c. wall (study case n°1, dissipative devices coupled with r.c. walls).

6.4. General guidelines for commercial buildings

Commercial centres provide the proper environment for selling and storing goods, and often accommodate
offices and leisure activities. Their spaces are planned to best house business service requirements as well as
the products to be stored and handled. Designers focus on making the spaces functional and efficient, while
providing a safe and comfortable environment for customers and workers. Building image and aesthetics,
landscaping, customer and worker safety as well as comfort, become important issues in competitive real
estate markets. Structures for low-rise commercial buildings should comply with these issues and the idea of
prefabricating the components is attractive. Nevertheless, setting up tables or tools for the quick pre-design
of elements is possible only for regular and repetitive structural layouts.
Usual dimensions and organization of spaces of commercial buildings are such that the structure can actually
be arranged according to regular grids for a large part of the building. Nevertheless, particular arrangements
or special elements are often used to improve the aesthetical value of the building or create suggestive
spaces. In most cases, only facades are interested by architectonic inventions and the large part of the
structure may be easily prefabricated (Figure 6.18).

141
 Figure 6.18. Venusio Shopping Center (www.promozioneacciaio.it)

Structures for low-rise commercial buildings proposed in PRECASTEEL are conceived to comply with these
issues. The structure is thought to be arranged according to regular grids and is obtained by coupling a
gravity structure with lateral resisting elements so that each component may be easily optimised: the gravity
structure has to withstand vertical actions whereas the lateral resisting elements have to resist all horizontal
actions (wind and earthquakes) and stabilize the whole system against geometrical effects due to vertical
loads.
The behaviour of the whole system is subordinate to the existence of in-plane stiff diaphragms that connect
the gravity structure to the lateral resisting elements. These may be constituted by composite slabs realised
with collaborating steel sheeting or by reinforced concrete slabs cast on slim prefabricated r.c. slabs.

6.4.1. Gravity structure

The gravity structure is constituted by columns and horizontal decks. Columns are supposed to be pinned at
the base and beams of the deck are supposed to be pinned at both their ends (Figure 6.19b). Consequently the
former are subjected to axial forces and very moderate bending and shear forces, whereas the latter behave as
simply supported elements and are subjected to bending and shear actions. The assumption of pinned
elements conforms to the column base connections and to the beam-to column and secondary-to-primary
beam connections made simple (e.g., double web-angle connection) Figure 6.19c,d to reduce manufacture
costs. The continuity of reinforcements (wire fabrics and transverse reinforcements at the shear connection of
the primary beam) is neglected in secondary beams, thus leading to a safer solution. In the primary beams a
suitable gap between slab and columns should be realised in order to avoid undesired composite actions at
the beam-to-column connections.
To facilitate the workmanship and speed-up the erection, unpropped construction should be privileged;
pre-cambered beams can be profitable used to control the deflections due to permanent loads instead of
considering larger cross sections. Full interaction between beams and the top slab may be achieved thanks to
a suitable shear connection. The presence of the slab is extremely beneficial for the behaviour of the beam
that works under sagging moments: due to the presence of the concrete slab, the neutral axis crosses the
beam web very close to the top flange that is also restrained against the lateral buckling; consequently, cross
sections usually fall into Class 1 and Class 2 (EC4) permitting to fully exploit the plastic resistance of the
materials.
For larger spans it may be necessary to adopt trusses instead of solid beams to reduce the structural weight.
The trusses may be optimised by considering different profiles for their various elements. In the
PRECASTEEL project different solutions have been evaluated such as composite elements obtained with hot
rolled profiles, cold formed beams (ZKU, ZKUG, rectangular), and trusses constructed with hot rolled
profiles and cold formed profiles. To facilitate the transport and speed-up the erection, long trusses should be
divided at least into two sub-assemblages by using welded joints realised in the shop with automatic systems,
and bolted joints for the site erection (Figure 6.19).

142
 (a) (b)

(c) (d)
Figure 6.19 – Gravity structure: (a) flooring system regular grid; (b) pinned beams and columns; (c) possible
connection between CHS and composite beams; (d) possible connection between hot rolled columns and
composite beams

Figure 6.20. Joints and connections for truss obtained with ZKU and SKUG cold formed
143
For the beams, both ultimate (ULS) and serviceability (SLS) performances have to be checked. The former
consists in usual verifications of bending and shear capacity under service loads whereas lateral flexural-
torsional buckling is not expected for the scheme adopted. However, this verification may become important
during construction unless profiled steel sheeting are suitably riveted to the primary and secondary beams (or
stud connectors are welded through the sheet) constituting a valid lateral constraint. As for serviceability
limit states, it is important to control that deflection is not greater than L/250, considering rare actions, and
L/350 under life loads only in order to guarantee aesthetics and comfort.
When truss girders are used, stability verifications become important and must be carried out with suitable
methods that take into account imperfections directly or indirectly (stability curves).
In the PRECASTEEL project a parametric analysis was carried out optimising flooring systems obtained
with composite solutions or trusses. Costs of the cases designed were evaluated using prices from Central,
Eastern and Southern Europe to give useful indications on the more suitable system (Figure 6.21). It has been
found that costs of flooring systems are strongly influenced by the structural grid dimensions. If particular
architectonic needs are not present, the cases with minor span lengths resulted to be the most affordable. It is
worth to note that large spans for floors subjected to live loads other than snow, in which truss girders should
be used, are not convenient compared to solutions with smaller spans. However, very large spans (over 20
m) are usually considered only for roofs.
Columns may be obtained with bare hot rolled (usually HE profiles) and cold formed profiles or with
Circular Hollow Sections. Partially encased composite columns and concrete filled hollow sections may be
profitably used due to their enhanced load bearing capacity and reduced costs for fire proofing as compared
to the bare profiles. The advantage of considering CHS columns relies on their ability to be used not only for
the rectangular grid considered but also in more complex patterns which may be not rare for commercial
buildings due to architectonic requirements.
Due to the scheme adopted for the gravity structure, columns may be designed by considering loads applied
on the tributary areas descending from the beam patterns considered for the flooring systems (Figure 6.20a).
Due to the beam to column connection and to the presence of horizontal diaphragms and seismic braces (or
shear walls), the structure is non-sway.
160.00
300.00
140.00
250.00
120.00 L1 L2
L1
L2
costs (€/sm)

100.00 200.00
costs (€/sm)

80.00
150.00
60.00
100.00
40.00 Live loads Live loads
8.00 kN/sm 8.00 kN/sm
20.00 5.00 kN/sm 50.00
5.00 kN/sm
2.00 kN/sm 2.00 kN/sm
0.00 0.00
0 25 50 75 100 125 150 50 100 150 200 250 300
L1
L1 xx L2 (m2)
L2 (sm) L1
L1xx L2 (m2)
L2 (sm)

Figure 6.21. Flooring systems - cost per square meter vs. area of the plan grid (data from Central Europe)

For the columns the performance at ultimate limit state has to be checked considering buckling phenomena.
For this purpose the verifications have to be carried out accounting for local imperfections or by considering
stability curves using the design slenderness evaluated against the critical axial force of the column with the
buckling length equal to the storey height. Local buckling of webs, flanges or hollow section walls, should
be avoided by restricting the profile set to Class 3.
In the PRECASTEEL project a parametric analysis was carried out also optimising columns and evaluating
their costs (Figure 6.22). It has been found that the cost of the columns is only slightly influenced by the grid
dimensions and reduces when increasing beam lengths. However, the cost of the columns is less than one
tenth of the cost of the flooring system and an optimum grid dimension that minimises the cost of the whole
144
gravity structure (columns and flooring systems) does not exist. The most economic solution for the gravity
structure still remains that with the smaller span length of the flooring systems. The parametric analysis also
demonstrated that solutions with CHS are characterised by the minor material consumption and minor costs
among the other profiles considered. Besides, composite solutions (partially encased hot rolled profiles and
concrete filled profiles) are the best solutions to be adopted.

32.00

28.00 2 kN/sm HE
6 PEHE
24.00 CHS
costs (€/sm)

20.00 CFCHS
cf rect.
16.00
cf ZKUG
12.00

8.00

4.00

0.00
0 50 100 150 200 250 300
2
Tributary
influencearea
area(m )
(sm)
28.00
HE 2 kN/sm
24.00 PEHE 6 8 kN/sm
20.00 CHS
6
CFCHS
costs (€/sm)

16.00

12.00

8.00

4.00

0.00
0 50 100 150 200 250 300
Tributary area (m2)

Figure 6.22. Columns - cost per square meter vs. tributary area (data from Central Europe)

6.4.2 Lateral resisting elements

Lateral resisting elements play a key role in the stability of the structure. In the case of commercial buildings
these may be constituted by steel concentric or eccentric braces or by reinforced concrete shear walls to
which the structure may be rigidly connected or interfaced by dissipative devices. Moment resisting frames
are usually not suitable since they are characterised by high lateral deformability and high cost of beam-to-
column connections.
Braces may be easily accommodated at the external walls or at staircases and facilities areas. In the case of
reinforced concrete elements, these may be constituted by real cores where staircases and elevators may be
profitable housed.
The geometry of the braces may be defined differently from the span lengths of flooring systems because
specific columns may be arranged without modifying the behaviour of the gravity structure.
For what concern steel braces, solutions with hot rolled open profiles are the most suitable as they permits
simple connections among the components. Furthermore, chevron configurations are preferable to cross
braces as they leave the possibility of creating openings in the walls that may be quite important for
commercial buildings.
Steel concentric braces may be designed for ductility but this seems to be not suitable in the case of
commercial buildings. Due to the capacity design, beams of chevron braces suffer for bending resulting in

145
deep cross sections. By considering a behaviour factor q = 1, no capacity design rules have to be applied
provided that diagonals are checked against buckling. As usual, all connections should be designed to
withstand forces grater than those acting in the connected elements. In these cases, the lateral deformability
is very low and the performance under moderate earthquake is not a problem.

120.00
HS
6
100.00 MS
6
LS

80.00
costs (€/sm)

60.00

40.00

20.00

0.00
0 1000 2000 3000 4000 5000 6000
2
Tributary
area area offor
of influence thethe
single
singlebrace (m(sm)
brace )
24.00
HS
6
20.00 MS
6 LS

16.00
costs (€/sm)

12.00

8.00

4.00

0.00
0 4000 8000 12000 16000 20000 24000
2
Tributary
area area for
of influence of the
thesingle
single brace
brace (m )
(sm)

Figure 6.23. Braces - cost per square meter vs. tributary area (data from Central Europe)

Eccentric systems may be properly designed for high ductility by considering high behaviour factors (q # 6).
Chevron configurations with two diagonals or with only one diagonal may be used depending on
architectonical requirements. Short links should be preferred to long or intermediate links to have good
performances against moderate earthquakes that may damage non structural elements.
Only Class 1 sections have to be used for the dissipating links; proper capacity design has to be carried out
for the other elements and the connections to achieve a global dissipative mechanism. The rotation capacity
of the links has also to be checked to verify whether the element local ductility comply with the global
displacements. It is worth to note that in the case of one-diagonal brace the link-to-column connection is
particularly complex due to the local detailing that has to ensure the transmission of high moments and shear
forces.
The structure has to be checked also with reference to the damage limit state (DLS) by comparing interstorey
146
drifts against suitable thresholds depending on the displacement capacity of the non-structural elements. It is
also important to check the whole structure against second order effects that, in the case of braced frames can
be easily limited.
Similarly to flooring systems and columns, a parametric optimization was carried out for braces in the
framework of PRECASTEEL project. Their costs have also been evaluated and compared by considering
different seismic levels (Figure 6.24). It was found that eccentric braces are definitely competitive compared
to concentric braces and that the best solution in term of material consumption is the one with few braces.
However, the last information should be considered carefully since placing few braces does not usually
comply with other important issues related to the capacity of horizontal diaphragms and the difficulties of
connecting them to the braces.

6.4.3 HDR Based Dissipative devices: guidelines and examples

HDR-based devices may be profitable used in structural systems in which reinforced concrete walls are used
as lateral resisting elements. This permits achieving the energy absorption necessary to resist high-intensity
earthquakes and to control the vibrations of the building due to the wind or to low-intensity earthquakes.
Dissipation devices have to be designed case-by-case and an optimization study, similar to the one
performed for the other elements, cannot be carried out. Nevertheless, by exploiting the possible linearization
previously presented (based on the definition of secant stiffness and equivalent damping for the HDR-based
devices), a simple design procedure based on the Direct-displacement-Based Design (DDBD) may be used to
determine the total thickness (tr) and area (Ar) of the rubber layers of each dissipative device, and the number
of devices (nd) to be implemented in the building.

Figure 6.24. Simplified modelling for designing HDR-based dissipative devices

With reference to the two-storey building, by assuming the wall to be stiff and the columns to be hinged at
each floor, the structure may be reduced to the couple of 1-DoF systems depicted in Figure 6.24. The design
procedure may be divided into the following four steps:
Step 1
The same values of the design displacements ud at the first and second floors are chosen according to the
desired performance threshold (e.g. 0.005 times H to avoid damage of non structural elements at the first
elevation). The thickness of rubber layers of the device may be promptly evaluated by imposing the limit
strain for the rubber Jd (at ultimate it is usually assumed to be 200%)
ud
ts [6.17]
Jd
Step 2
The period T0 of the two 1-DoF systems is evaluated by imposing iteratively the condition
S De [ s T0 ,ud ,T0  ud
[6.18]
where SDe is the design displacement response spectrum.
Step 3
The total area of the rubber layer of the devices at the i-th floor (As,tot,i) is calculated by the equation
4S 2 tr
As ,tot ,i Mi
T02
Gs T0 , J d  [6.19]
where Gs(T0, Jd) is the rubber shear modulus.
Step 4
147
Dimensions of each damper at the i-th floor may be evaluated by fixing the total number of devices nd (or
vice versa).

Despite its simplicity, the procedure revealed to be effective. Time-integration analyses demonstrated that
the displacements achieved at the two levels are almost the same even if the columns are really continuous at
the first floor. Of course, application of dissipative devices requires detailed verifications and the procedure
proposed can be used only for pre-designing purposes.
As already stated, the use of HDR-based devices for structural systems is very promising in controlling the
structural response under actions like wind and earthquakes falling within a wide range of intensities. The
behaviour of a device depends on the total thickness and area of the rubber layers as well as on the
mechanical properties of the material.
Dissipative devices may be profitably used when the lateral resisting elements are r.c. shear walls. The basic
idea is to exploit the presence of r.c. cores, where stair cases, elevators or other facilities may be placed, to
constitute the reaction structure to which the dissipative devices are connected (Figure 6.25). In this way
HDR-based devices undergo deformation, and consequently dissipate energy, under the lateral storey
displacements in case of horizontal loading.

(a) (b)

Figure 6.25. Implementation of HDR-based devices: (a) position of devices in a low rise building; (b)
possible details for the connection of devices to the r.c. wall and to the floor slab

Dissipation devices have to be designed case-by-case by exploiting linearised models (based on the
definition of secant stiffness and equivalent damping for the HDR-based devices). Simple procedures
consisting of imposing suitable lateral displacements (Direct Displacement-Based Design) may be easily
carried out to determine the total thickness and area of the rubber layers of each dissipator, and the number
of devices to be implemented in the building. Nevertheless, it is worth to remember that the application of
dissipative devices requires detailed verifications preferably carried out by considering a suitable set of
accelerograms.
As for construction problems, the implementation of dissipators in the structure requires preferably to
consider vertical columns close to the reinforced concrete cores (Figure 6.25a) to avoid that vertical loads
may reflect badly on the behaviour of the dissipators. The connections have to be developed in order to avoid
undesired eccentricities of reactions with respect to the slab by using devices placed in pair as shown in
Figure 6.25b. Other important issues are related to the fire protection of the devices that must be placed in
protected places that must guarantee their periodic inspection. The devices replacement must also be
guaranteed by considering suitable connection devices embedded in the concrete. Finally, even if the
displacements are of the order in some centimetres, suitable gaps must be ensured between the gravity
structure and the lateral resisting structures with particular attention to utilities that have not to suffer for
relative displacements of the structures.

148
6.4.4 FSHD-based dissipative devices – guidelines and examples
Another typology of dissipative device behaving according to a ‘flag-shaped’ hysteretic curve and possessing
self-centering abilities (FSHD) was taken into account. In order to apply such a device to the predefined
frame geometry layout, a specific design procedure was suitably developed basing on the application of the
capacity spectrum method for the computation of the required equivalent damping in order to obtain a
predetermined maximum displacement (target displacement or performance point).

6.4.4.1. Description of the structural system and adopted design issues

For the purpose of dimensioning the FSHD hysteretic devices, the three dimensional structures of the
analyzed low-rise commercial buildings were replaced with the corresponding planar frames bearing the
horizontal actions in the two main earthquake directions.
The structural system under consideration was made up with three main structural components which can be
summarized as the followings:
‚ the steel concrete composite vertical frame which is composed by hinged beams and continuous
steel concrete composite columns with pinned base connections;
‚ the reinforced concrete shear wall which is realized with a suitably designed prefabrication system
and is fully fixed at the base;
‚ the flag-shaped hysteretic devices (FSHD) which are realized with two co-axial steel tubes placed
horizontally at each floor level and connected to the shear wall from one side and to the vertical
frame to the other
A schematic picture of the adopted structural system and its components is shown in Figure 6.26.

Figure 6.26. Frame equipped with FSHD: schematic of the adopted structural system and of its components

In order to understand the behavior of such a structural system, some considerations on the different
components are needed. First of all, this system differs from the so called frame-wall typology since the
latter is composed by a continuous vertical frame possessing its own lateral stiffness and thus being able to
sustain a predetermined part of the design shear force, while in present case the vertical frame has no lateral
stiffness because of the adopted restraining conditions.
For what concerns the reinforced concrete shear walls, they play an important role depending on both
geometry and mechanical characteristics; in Eurocode2, the following definitions can be found:
‚ Ductile wall: it is defined as a base fixed wall so that its relative rotation with respect to the whole
structural system is inhibited; it is designed in order to dissipate the seismic input energy in a
flexural mode so that a plastic hinge is allowed to form just above the fixed base.
‚ Large lightly reinforced wall: this definition applies for shear walls with big cross sections
dimensions and complying with the followings aspect ratios: ݈௪ ൐ ͶǤͲ݉ or ݈௪ ൐ ʹΤ͵ ‫ܪ‬, whichever
is the smaller, where ݈௪ is the wall length and H is the total wall height. In this case, the wall has
none or limited inelastic behavior, therefore no formation of plastic hinges is expected
Basing on such definitions, and in the framework of a displacement based design procedure, some
preliminary hypothesis on the behavior of the adopted structural walls were needed, since their behavior was
important in order to understand the seismic response of the entire structural system.
The most simple assumption that could be made was to consider the wall as infinitely rigid; if so, it can be
represented as a fixed point in the system thus having no influence on the frame displacement profile. This is
a theoretical assumption which can present some feasibility in real construction practice depending on the
shear wall characteristics; moreover, it can offer great advantages for what concerns the schematization of
the structural system in view of obtaining its capacity curve. In this case in fact, the whole structure may be
reduced to a simple 2DOF system with the double option of modeling the ‘equivalent’ column element as a
rigid frame or as a flexible frame. In both cases, the system results easily solvable with hand calculations.
149
However, in the framework of the definitions given above, the design choice adopted in present analyses was
to consider the large lightly reinforced walls category only, this meaning that the sole non-linearity of the
structural system was concentrated in the adopted FSHD devices while all the other components were
deemed to remain in the elastic range.
Bearing in mind these considerations, the composite frame was designed according to the static combination
of actions to sustain vertical loads only. On the other hand, vertical shear walls were designed to sustain the
maximum value of the calculated design shear force at ULS. Finally, the adopted FSHD devices were
designed with the aim of limiting the horizontal displacements of the frame in such a way to comply with the
fixed performance objectives, this resulting in the determination of their equivalent damping parameter.

6.4.4.2 Description of case studies

The selection of case studies for the application of FSHD hysteretic devices, was made within the framework
of a previously analyzed set of frames equipped with dissipative eccentric bracings.
With reference to Table 6.2, CS1H5 and CS2H5 case studies were chosen.

ag = 0.32 (High) ag = 0.16 (Medium) ag = 0.08 (Low)

Qk=5kN/sm Qk=8kN/sm Qk=5kN/sm Qk=8kN/sm Qk=5kN/sm Qk=8kN/sm
Type 1 (CS1) CS1H5 CS1H8 CS1M5 CS1M8 CS1L5 CS1L8
Type 2 (CS2) CS2H5 CS2H8 CS2M5 CS2M8 CS2L5 CS2L8
Type 3 (CS3) CS3H5 CS3H8 CS3M5 CS3M8 CS3L5 CS3L8
Table 6.2. Definition of analyzed case studies derived from the application of database cross-sections

The geometry layout for such frames was the same as for the eccentric braced solution, however some
modification were needed in order to replace the steel eccentric bracings with the system made by the
coupling of shear walls with FSHD hysteretic devices. In particular, the coupling steel beams adopted in
such a way to release the bracing system from the applied vertical loads were not needed anymore so that
they were removed. Moreover, the beam-column cross-sections of the portal frames designed for the seismic
combination of actions, were replaced by the ones designed for the static combination of actions (as defined
in the cross-sections data base). As a design choice, shear walls were placed in the outer perimeter of the
frame so that no element of the original vertical frame was removed this helping both in a clear separation of
the load bearing functions and in a simplification of the joint typologies used to connect FSHD with both the
shear walls and the vertical frame.
Moreover, a further design parameter to be discussed consisted in the number of shear walls to be placed in
each direction, this clearly having direct consequences on the dimensioning of both the shear walls
themselves and the hysteretic devices.
For the considered case studies, after determining the amount of design base shear corresponding to the PGA
value selected for the ultimate limit state (PGA=0,32 g), it was decided to place two shear walls in each
direction, as it is shown in Figure 6.27 for CS1H5 case study:

150
 (a)

(b) (c)
Figure 6.27. CS1H5 case study – shear wall disposition: a) in plane; b) X-elevation; Y-elevation

The same shear wall distribution was adopted for CS2H5 case study, as well.

6.4.4.3. Design of frames equipped with FSHD: preliminary considerations

The proposed design procedure relies on the fundamental principles of displacement based design and on the
assumed structural performance objectives. In this framework, the core of the procedure consists in the
identification of the fundamental characteristics of the equivalent SDOF structure which are based on
geometry, mass distribution and displacement profile of the reference structural system. In particular, the
characteristic displacement (referred to as the design displacement) represents the displacement of the
equivalent SDOF system that will ensure the structural and non structural deformation limits are not
exceeded. These values are set on the bases of the evaluated seismic risk for the analyzed site, a Type 1
elastic spectrum was adopted with reference to Type B soil.

6.4.4.3.1 Definition of performance objectives

In this case, three performance levels were defined, each one corresponding to a determined PGA value and
to a defined limit of the inter storey drift ratio; the adopted values are summarized in Table 6.3.
Level 1 corresponds to serviceability limit state meaning that only non significant damage can be expected,
and any necessary repairs can be carried out but without affecting the normal usage of the building structure.
Level 2 intensity represents the damage control state, meaning that damage should be economically
reparable. Level 3 intensity represents the collapse prevention limit state meaning that the performance
criteria is to require no collapse; usually in this case no limits are set for non-structural drift; however, in
present analyses the 5% drift limit ratio was set in correspondence of the ultimate limit state (ULS) PGA
design value in order to check if the system possess some extra structural resource available in
correspondence of exceptional events corresponding to PGA=0,46 g for the analyzed site. Such drift limits
151
were set in accordance with displacement based design issues found in ‘A model code for the Displacement-
Based Seismic Design of structures’ by Calvi.

Level 1 Level 2 Level 3

PGA = 0,059g PGA = 0,086g PGA = 0,320g

Drift Limit
Brittle non structural elements 0,004 0,025 0,05
Ductile non structural elements 0,007 0,025 0,05
Non structural elements detailed to sustain
0,01 0,025 0,05
building displacement
Table 6.3 Drift limits for different design performance levels

6.4.4.3.2 Description of the numerical model adopted for pushover analyses

With reference to the adopted static structural scheme and to the above mentioned assumptions on the
behaviour of the structural components, a simplified equivalent 2DOF model was adopted, instead of the
complete MDOF one, thus sparing computational time, as it is shown in Figure 6.28:

Figure 6.28. Numerical models of planar frames: a) MDOF system; b) equivalent 2DOF system

The FSHD hysteretic devices (in red) were modelled with multi-linear elastic spring elements able to
reproduce the monotonic behaviour, as it is required for non linear elastic analyses. Finally, beam-column
elements were modelled with linear elastic two-joint frame elements.
The simplified 2DOF equivalent model was obtained with the following considerations based also on the
symmetry of the structural system. Since the vertical frame had no lateral stiffness and all beam-column
elements were acting in parallel, given the rigid floor assumption, it was possible to eliminate the whole
vertical frame system by simply substituting it with a single point representing the assigned seismic mass at
each floor level; a sliding support was provided in order to account for the horizontal displacements at each
level. The assigned number of FSHD devices acting in parallel at each floor level, was replaced by a single
‘equivalent’ one and in the same way the shear walls were replaced by an ‘equivalent’ one whose
characteristics in terms of gross cross-sectional area, moment of inertia, shear area were assigned to a single
frame element fully fixed at the base. The FSHD hysteretic devices were once again modelled with multi-
linear elastic spring elements; the resulting simplified numerical model is shown in Figure .
The simplified 2DOF model was adopted to determine the capacity curves of the whole structural system
with the aid of the numerical program SAP2000, as stated at the step 2) of the proposed design procedure
described in the following section.

6.4.4.4. The proposed four-steps design procedure

The proposed design procedure for the dimensioning of Flag Shaped Hysteretic Devices (FSHD) can be
summarized in the following steps:
1. Definition of FSHD monotonic curves
2. Determination of the system capacity curve
3. Determination of the structural performance point (PP)
4. Definition of FSHD cyclic curves (determination of the ߚ factor defining the amplitude of the
hysteretic cycle)
152
6.4.4.4.1 Determination of FSHD monotonic curves
The first step of the proposed procedure consisted in the definition of the monotonic curves to be assigned to
the FSHD posed at different floor levels; this was made with reference to the equivalent 2DOF structural
system defined in section 3.2 above and therefore the obtained data refers to an ‘equivalent’ FSHD device
including all the devices working in parallel at a fixed floor level.
The schematic of a FSHD monotonic curve, is represented in Figure 6.29:

rK 0 E Fy
Fy

K0

'
A

Figure 6.29. ‘Flag-shaped’ hysteretic curve: the main parameters characterizing cyclic behaviour

To define the monotonic curve, six parameters were needed and only four of them had to be set in an
independent way; of course, the choice of the four parameters to be assigned was completely arbitrary this
depending only on the design purposes. The six parameters are:
‚ ‫ܨ‬௬ , the applied force at yielding
‚ ‫ܭ‬௢ , the elastic stiffness
‚ ݀௬ ,the device displacement at yielding
‚ r ,the hardening ratio
‚ ‫ܨ‬௨ ,the applied force at the ultimate limit state
‚ ݀௨ ,the device displacement at the ultimate limit state
In present analyses, displacements at both yielding and ultimate limit state were fixed with reference to the
defined performance drift limit ratios, the applied force at yielding ‫ܨ‬௬ was determined with reference to the
Level 1 PGA, corresponding to serviceability limit state, while the hardening ratio r was defined according to
engineering judgment and basing on considerations on mechanical properties of the adopted materials for the
FSHD components.
To determine the seismic design action, an estimate of the fundamental structural period was made by
making use of the simplified expression in given Eurocode8 by:
ܶଵ ൌ ‫ ܪܥ‬ଷΤସ [6.20]
where C = 0,075 is a correction factor for r.c. structures and H is the total height of the building.
Afterwards, it was possible to determine the total base shear force ‫ܨ‬௕ , with the following:
‫ܨ‬௕ ൌ ܵௗ ሺܶଵ ሻ݉ߣ [6.21]
ሺܶ ሻ
where: ܵௗ ଵ ൌ ʹǡͷܵܽ௚ is the ordinate of the design spectrum in correspondence of the fundamental period
ܶଵ , m is the total seismic mass of the system and ߣ is a correction coefficient setting to 0,85 if two times the
fundamental period ܶଵ is less than the corner period at plateau ܶ஼ and setting to 1, otherwise.
After the total base shear force ‫ܨ‬௕ was determined, it had to be distributed on the two floor levels, this
depending on the definition of the design yield displacement profile; to do this, different options are
available. In particular, of the two simple following assumptions can be made:
1) displacement profile varying linearly with the building height; in this case, the imposed force at the ݅ ௧௛
floor level can be determined by the following:
௠௭
‫ܨ‬௜ ൌ ‫ܨ‬௕ σ೙ ೔ ೔ [6.22]
೔ ௠೔ ௭೔
where ݉௜ and ‫ݖ‬௜ are the floor mass and height for the i-th floor level.
2) displacement profile varying proportionally to floor masses; in this case, the imposed force at the ݅ ௧௛ floor
level can be determined by the following:
௠ఋ
‫ܨ‬௜ ൌ ‫ܨ‬௕ σ೙ ೔ ೔ [6.23]
೔ ఋ
௠೔ ೔
where ݉௜ and ߜ௜ are the floor mass and the imposed displacement at the i-th floor level.
153
Once the yield profile of displacements is defined, the elastic stiffness for the ‘equivalent’ FSHD device at
each floor level is automatically determined as ‫ܭ‬௘௤௜ ൌ ‫ܨ‬௜ Τ݀௬௜ ; afterwards, it is sufficient to fix the hardening
ratio r and the ultimate displacement ݀௨ according to the assumptions made in the introduction and the
monotonic curve is defined. The curve of the isolated dissipative device can be deduced from the equivalent
one considering that the devices at each level act together as parallel springs, these being subjected to the
same displacements. Therefore, under the simplifying assumption of having the same kind of device at the
single level we can define ‫ܭ‬௜ ൌ ‫ܭ‬௘௤௜ Τ݊ and ‫ܨ‬௬௜ ൌ ‫ܭ‬௘௤௬௜ Τ݊, where n is the number of devices placed at the
reference floor level.

6.4.4.4.2 Determination of the capacity curve

The capacity curve of a structural system is conventionally defined as the plot of the total base shear force
against the top horizontal displacement under assumed sets of horizontal loads increasing monotonically
until collapse of the structural system.
In present analyses, the capacity curve of the system is determined with reference to the simple numerical
model described in section x, under the following assumptions:
‚ The deformation of the structure under applied vertical loads is not taken into account
‚ The applied set of horizontal loads is proportional to floor masses at each level
Once the capacity curve of the 4DOF structural system is obtained, it is possible to transform it into the
capacity curve of the equivalent SDOF system scaling it by the structural participation factor Γ, defined as:
௠ ఝ ା௠ ఝ
Ȟ ൌ భ మభ మ మ మ [6.24]
௠ ఝ ା௠ ఝ
భ భ మ మ
where ߮ ൌ ሾ߮ଵ Ǣ ߮ଶ ሿ represents the vector of displacements of the frame for the first fundamental mode of
vibration.
In the same way, the equivalent mass me of the SDOF system can be determined with the following
expression:
ሺ௠ ఝ ା௠ ఝ ሻమ
݉௘ ൌ ሺ௠ భ భమ మ మ మ ሻ [6.25]
ఝ ା௠ ఝ
భ భ మ మ
Once the capacity curve of the equivalent system is defined, the yielding point ൫‫ܨ‬௬‫ כ‬Ǣ ݀௬‫ כ‬൯can
be identified so
that it is possible to go through the determination of the structural performance point defined in terms of the
maximum (target) displacement.

6.4.4.4.3 Determination of the structural performance point (PP)

When determining the performance point of a structural system, it is common practice to operate in the so
called AD (acceleration – displacement) format; therefore, the first thing was to transform both demand
(elastic spectrum) and capacity curve in the acceleration displacement format.
For the capacity curve it is sufficient to scale the ordinate values for the equivalent mass ݉௘ of the SDOF
system, as defined above while the demand curve was obtained by determining displacements from
corresponding periods. Finally, the structural performance point of the system was determined basing on the
capacity spectrum method.
The conventional capacity spectrum method uses the secant period as the effective linear period in
determining the maximum displacement. This assumption results in the maximum displacement occurring at
the intersection of the capacity curve for the structure and a demand curve for the effective damping in
ADRS format.
Equivalent linearization procedures, normally require the use of spectral reduction factors to adjust an initial
response spectrum to the appropriate level of the effective damping expression ߚ௘௙௙ , by making use of the
following:
ௌ
ሺܵ௔ ሻఉ ൌ ೌ೚ [6.26]
஻൫ఉ ೐೑೑ ൯
where ‫ܤ‬൫ߚ௘௙௙ ൯ can be determined by the following approximate expression:
ସ
‫ܤ‬ൌ [6.27]
ቀହǤ଺ି௟௡ఉ೐೑೑ ሺ௜௡Ψሻቁ
Both the effective period ܶ௘௙௙ and the effective viscous damping ߚ௘௙௙ are taken as the optimal equivalent
linear parameters. The approximate equations for the effective viscous damping value, expressed as a
percentage of critical damping have been optimized for the application to any capacity curve, independent of
hysteretic model type or alpha value adopted in the study and are given in FEMA440 by the following
expressions:
154
1.0 <ߤ ൏ ͶǤͲ ߚ௘௙௙ ൌ ͶǤͻሺߤ െ ͳሻଶ െ ͳǤͳሺߤ െ ͳሻଷ ൅ ߚ௢
if 4.0 <ߤ ൏ ͸Ǥͷ ߚ௘௙௙ ൌ ͳͶǤͲ ൅ ͲǤ͵ʹሺߤ െ ͳሻ ൅ ߚ௢ [6.28]
଴Ǥ଺ସሺఓିଵሻିଵ ் ଶ
if ߤ ൒ ͸Ǥͷ ߚ௘௙௙ ൌ ͳͻ ቂ ሾ଴Ǥ଺ସሺఓିଵሻሿమ ቃ ቀ ೐೑೑ ቁ ൅ ߚ௢
்బ
Under the same assumptions, approximate equations are provided for the effective period ܶ௘௙௙ :
݂݅1.0 <ߤ ൏ ͶǤͲ ܶ௘௙௙ ൌ  ሼͲǤʹͲሺߤ െ ͳሻଶ െ ͲǤͲ͵ͺሺߤ െ ͳሻଷ ൅ ͳሽܶ଴
if 4.0 ≤ߤ ൏ ͸Ǥͷ ܶ௘௙௙ ൌ  ሾͲǤʹͺ ൅ ͲǤͳ͵ሺߤ െ ͳሻ ൅ ͳሿܶ଴ [6.29]
ሺఓିଵሻ
if ߤ ൒ ͸Ǥͷ ܶ௘௙௙ ൌ ൜ͲǤͺͻ ൤ට െ ͳ൨ ൅ ͳൠ ܶ଴
ଵା଴Ǥ଴ହሺఓିଶሻ
Since both the effective period ܶ௘௙௙ and the effective damping ߚ௘௙௙ are functions of the ductility demand,
the calculation of the maximum displacement using equivalent linearization is not direct and requires an
iterative or graphical solution procedure. In FEMA440, three alternative procedures are proposed, each of
them requiring the same initial steps listed below:
1. Selection of a spectral representation of the ground motion of interest with an initial damping ߚ௢
(normally 5%)
2. Conversion of the selected spectrum to an acceleration displacement response spectrum format this
representing the initial ADRS demand
3. Generation of capacity curve for the structure (normally resulting in a relationship between base
shear and roof displacement) and conversion into ADRS format for equivalent linearization
procedures
4. Selection of an initial performance point ൫ܽ௣௜ Ǣ ݀௣௜ ൯, maximum acceleration and displacement
5. Development of a bilinear representation of the capacity spectrum through the definition of the
initial period ܶ଴ , yield displacement ݀௬ and yield acceleration ܽ௬
6. For the bilinear representation of the above step, calculation of the post elastic stiffness ߙ and of the
ductility ߤ, can be made by using the following expressions:
ೌ೛೔ష 
ቆ೏ ష೏ ቇ
೛೔ ೤
ߙൌ ೌ೤ [6.30]
൬ ൰
೏೤
ௗ೛೔
ߤ ൌ ௗ [6.31]
೤
7. Determination of the corresponding values of the effective damping ߚ௘௙௙ and of the effective period
ܶ௘௙௙ according to the defined expressions.
8. After this step, three different options are available to determine a specific solution; in present
analyses Procedure B, defined as Intersection with MADRS was adopted. In this case, the
performance point is defined as the intersection of the capacity spectrum with the modified ADRS
(MADRS) which is generated by modifying the ADRS for different values of the effective damping,
determined according to expression defined above.

The additional steps needed to complete the procedure for the determination of the maximum
displacement of the system can be summarized in the followings:

9. adjustment of the initial ADRS with the effective damping determined in previous step
10. multiplication of the acceleration ordinates only of the ADRS for ߚ௘௙௙ by the modification factor M
determined in correspondence of the effective period ܶ௘௙௙ to generate the modified acceleration-
displacement response spectrum (MADRS)
11. determination of the estimate of the maximum acceleration ܽ௜ and displacement ݀௜ as the
intersection of the MADRS with the capacity curve
12. comparison of the estimated maximum displacement with the initial (or previous) assumption,݀௣௜ ; if
it is in the acceptable tolerance, than the performance point corresponds to ሾܽ௜ Ǣ ݀௜ ሿ, otherwise such
procedure has to be repeated from point 5 on, adopting ሾܽ௜ Ǣ ݀௜ ሿ as new starting point.

The first three points of such iterative procedure were already covered in previous section, therefore the first
thing to do was to select the initial performance point. Generally, when in presence of structures possessing a
155
fundamental period higher than the corner period, the equal displacement rule applies, so that the starting
displacement can be assumed as the same of a structure behaving in the elastic range. However, for the low –
rise structures under consideration, the fundamental period is normally included in the plateau range,
therefore the equal displacement rule is not applicable.
Therefore, the starting performance point was determined following the procedure given in Eurocode8 for
structures in the short period range (having the fundamental period ܶ ൏ ܶ஼ ), as shown in Figure 6.30.

(a) (b)
Figure 6.30. Determination of starting performance point a) in the short period range; b) in medium and long
period range

The involved quantities are determined using the following expressions:

‫כ‬
௠ ‫ כ‬ௗ೤
ܶ௬‫ כ‬ൌ ʹߨට is the period of the equivalent SDOF structure
ி೤‫כ‬
் ‫כ‬ ଶ
‫כ‬
݀௘௧ ൌ ቀଶగቁ ܵ௘ is the target displacement of the structure with period ܶ ‫ כ‬and unlimited elastic behaviour
ௗ‫כ‬ ்
݀௧‫ כ‬ൌ ௤೐೟ ቀͳ ൅ ሺ‫ݍ‬௨ െ ͳሻ ்಴‫ כ‬ቁ ൒ ݀௘௧
‫כ‬
where ‫ݍ‬௨ is the ratio between the acceleration demands in the structure
ೠ
with unlimited elastic behaviour ܵ௘ and in the structure with limited strength ‫ܨ‬௬‫ כ‬Τ݉‫ כ‬and is defined as
ௌ೐ ௠‫כ‬
‫ݍ‬௨ ൌ ி೤‫כ‬
.
Once the starting point was determined, the abovementioned iterative procedure was developed until the
determination of the performance point, as shown in Figure 6.31.

Figure 6.31 . Determination of estimated maximum displacement using intersection of capacity spectrum
with MADRS

6.4.4.4.4 Definition of FSHD cyclic curves

The last step to complete the design procedure of FSHD dissipative devices consisted in the definition of
their cyclic curves, through the determination of the ߚ factor defining the amplitude of the hysteretic cycle.
The method adopted to determine such parameter relies on the basic assumption of equivalent linearization
techniques stating that the maximum inelastic deformation of a non linear SDOF system can be

156
approximated from the maximum deformation of a linear elastic SDOF system that has a period and a
damping ratio that are larger than the initial values of those for the non linear system. Moreover, the capacity
spectrum method of equivalent linearization assumes that the equivalent damping of the system is
proportional to the area enclosed by the capacity curve. The equivalent period, is assumed to be the secant
period at which the seismic ground motion, reduced for the equivalent damping, intersects the capacity curve
as shown in Figure 6.32.

Figure 6.32. Graphical representation of the capacity spectrum method of equivalent linearization

Within this framework, it was possible to determine the value to be assigned to the ߚ factor, by equating the
total energy dissipated by the system ‫ܧ‬஽ , represented by the area enclosed in the hysteretic cycle, to the
maximum strain energy ‫ܧ‬ௌ௢ , which under the given assumptions is given by the area of the triangle. Of
course, in present analyses, the involved quantities were determined with reference to the flag-shaped
hysteretic cycle.
After the hysteretic cycle was determined, the effective seismic performance of the structural system was
investigated by performing Incremental Dynamic Analyses; of course, if the obtained results were not
complying with the pre-defined performance objectives, the procedure was repeated modifying the shape of
FSHD monotonic curves, until performance objectives were met.

6.4.4.5. CS1H5 case study: application of FSHD design procedure and seismic performance evaluation
The results obtained from the application of the proposed design method described above to CS1H5 case
study are described in the following sections.

6.4.4.5.1 Definition of FSHD monotonic curves

The characteristics of the prefabricated reinforced concrete shear wall for the analyzed system are
summarized in Table 6.4.

H [m] lw [m] d [m] A [m2] As [m2] Jy [m4]

10 6 0,3 1,8 1,5 5,4
Table 6.4. Geometric and mechanic characteristics of the prefabricated shear wall

where: H is the total wall height; ݈௪ is the total wall length; ݀ is the wall depth; ‫ ܣ‬is the wall gross section
area; ‫ܣ‬௦ is the wall shear area and ‫ܬ‬௬ is the wall moment of inertia about y-axis.
The values of the seismic mass applied at each floor level are summarized in Table 6.5.

m1 [kNm/sec 2] m2 [kNm/sec 2] m1w [kNm/sec 2] m2w [kNm/sec 2] m1tot [kNm/sec 2] m2tot [kNm/sec 2]
570 499 22,5 11,25 615 521,5
Table 6.5. Values of the applied seismic mass at each floor level

157
The total base shear force ‫ܨ‬௕ determined in correspondence of the Level 1 of performance (PGA=0,059 g) is
given by the following Table 6.6:

C T1 [sec] Sd(T1) λ mtot [kNm/sec 2] Fb [kN]

0,075 0,422 1,736 0,85 1136,5 1677
Table 6.6. Determination of the base shear force in correspondence of the Serviceability Limit State

The assumed profile of displacement at yield with the obtained corresponding values of the yield force and
elastic stiffness for the equivalent FSHD at each floor level are summarized in Table 6.7.

δ1 [m] δ2 [m] F1 [kN] F2 [kN] KeqI [kN/m] KeqII [kN/m]

0,0028 0,0024 976 702 348485 295504
Table 6.7. Determination of yield force and elastic stiffness for equivalent FSHD at each floor level

In this case, brittle structural elements were considered for the identification of the maximum allowable drift
limit ratio; moreover, the hardening ratio r was set to 0,15.
The obtained monotonic curves for FSHD at the two floor levels are shown in Figure 6.33.

800 1000
800
600
600
400
400
200 200

0 0
-0,06 -0,04 -0,02 0 0,02 0,04 0,06 -0,06 -0,04 -0,02 -200 0 0,02 0,04 0,06
-200
-400
-400 -600
-600 -800

-800 -1000
I level II level

keqI 178571 kpl 26786 r keqII 148043 kpl 22206 r

dy 0,0028 du 0,05 0,15 dy 0,0024 du 0,05 0,15
Fy 500 Fu 1764 Fy 352 Fu 1409
Figure 6.33. Monotonic curve of the single FSHD device adopted for CS1H5 case study

For sake of simplicity, the same curve obtained for the first floor level was used in correspondence of both
floors.

6.4.4.5.2 Determination of the equivalent SDOF pushover curve

The values obtained for the modal participation factor Γ and equivalent mass ݉௘ of the SDOF system are
shown in Table 6.8, where ݀ଵ and ݀ଶ are the floor displacement of the 2DOF system in correspondence of
the fundamental mode of vibration.

d1 [m] d2 [m] φ1 φ2 Γ me [N sec2/m]

0,019 0,039 0,489 1 1,23 1011052
Table 6.8. Obtained values for the modal participation factor Γ and equivalent mass ݉௘ of the SDOF system

The pushover curve for the 2DOF system (in red) with the corresponding curve for the equivalent SDOF
system (in blue) are shown in Figure 6.34.

158
 10000000

9000000

8000000
Base shear [N] 7000000

6000000

5000000

4000000

3000000

2000000

1000000

0
0 0,01 0,02 0,03 0,04 0,05 0,06 0,07 0,08
Disp [m]

Figure 6.34. Pushover curve obtained for the equivalent SDOF system (in blue)

The characteristic values of the yielding point ൫‫ܨ‬௬‫ כ‬Ǣ ݀௬‫ כ‬൯ with the corresponding value of the first trial
performance point used to start the iterative procedure are shown in Table 6.9.

Fy * dy * T*=To [sec] det* [m] qu dt * [m]

1727993 0,004 0,303 0,007 1,755 0,009
Table 6.9. Characteristic yield point of the equivalent SDOF system and first trial performance point

6.4.4.5.3 Determination of the structural performance point (PP)

The plot of the obtained demand and capacity curve in the acceleration displacement format (AD) are shown
in Figure 6.35.

3,5

2,5

2
A

1,5

0,5

0
0 0,01 0,02 0,03 0,04 0,05 0,06 0,07 0,08
D [m]

Figure 6.35. Determination of the structural performance point through iterative procedure in AD format

The data obtained from the iterative procedure for the identification of the performance point of the system
are summarized in Table 6.10:

α μ βeff Teff B(βeff) (To/Tsec)^2 M Tsec x y

0,251601 2,244906 10,4717 0,375146 1,230268 0,58497754 0,894598 0,3966304 0,0082 2,1800
0,26 2,05 9,100145 0,356519 1,179346 0,62352 0,861199 0,3841762 0,0175 1,9800
0,04672 4,392561 18,44516 0,52209 1,489648 0,2637416 0,781191 0,5906991 0,0128 2,0800
Table 6.10. Determination of performance point PP and of the corresponding equivalent damping ߚ௘௙௙

159
6.4.4.5.4 Definition of FSHD cyclic curves
The data used for the determination of the ߚ factor defining the amplitude of the hysteretic cycle are
summarized in Table 6.11.

Atr1 Atr2 At3 βeff ED Eso βo β

0,005 0,002 0,000 9,100 0,010 0,010 0,041 0,70
Table 6.11. Determination of the β factor defining the amplitude of the hysteretic cycle

Finally, the obtained corresponding hysteretic curve in the acceleration displacement format (AD) for the
designed FSHD is shown in Figure 6.36.

2,5
2
1,5
1
0,5
0
-0,5
-1
-1,5
-2
-2,5
-0,01 -0,005 0 0,005 0,01

Figure 6.36. Determination of the cyclic curve of the FSHD

Such a curve was adopted to characterize the cyclic behaviour of FSHD devices when performing
incremental dynamic analyses (IDA) in order to check the effective seismic performance of the structural
system.

6.4.4.5.5. Evaluation of seismic performance: Incremental Dynamic Analysis (IDA)

Incremental Dynamic Analyses of the selected study case were performed with the aid of RUAUMOKO
numerical program in order to investigate the effective seismic performance; in particular, drift limits and
dissipated energy were checked.
The suite of ground motion adopted for Incremental Dynamic Analyses consisted of seven artificial
accelerogram compatible with the design spectrum and having the same peak ground acceleration set equal
to 0,32 g; the artificial accelerogram were generated with the aid of SIMQKE software with a recorded time
equal to 20 seconds. Moreover, each accelerogram was scaled of the necessary quantities in order to obtain
the range of exciting accelerations described above, thus obtaining a total of 35 input acceleration files.
Two of the 7 artificial accelerogram, adopted in the IDA, are shown in Figure 6.37:

acc 5 0,5 acc 6

0,5
0,4
0,4
0,3
0,3
0,2 0,2

0,1 0,1
PGA [g]
PGA [g]

0,0 0,0

-0,1 -0,1

-0,2 -0,2
-0,3 -0,3
-0,4 -0,4
-0,5 -0,5
0 5 10 15 20 0 5 10 15 20
sec sec

Figure 6.37. Set of artificial accelerogram adopted to perform Incremental Dynamic Analyses

160
The numerical model adopted for Incremental Dynamic Analyses was the same as the one adopted for non
linear static analyses but all 4DOF were considered; non linear spring elements were adopted to model
dissipative devices.
Data obtained from the set of seven artificial accelerogram were elaborated in such a way to consider for
each parameter the mean value with the corresponding 5% and 95% fractile; the values obtained in
correspondence of the acceleration scale factor equal to 1, correspond to the design PGA=0,32 set for ULS
verifications.
The values obtained for shear force at the base of the reinforced concrete wall are shown in Figure 6.38.
Wall Shear Force [N]

6,E+06
Shear force - max
Shear force - min
4,E+06

2,E+06

0,E+00

-2,E+06

-4,E+06

-6,E+06
0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6
acc scale factor
Figure 6.38. Obtained values for Shear Force at the base of precast r.c walls

The values obtained for axial force, axial elongation and dissipated energy in correspondence of each
dissipative device are shown in Figures from 6-39 to 6.41.
Axial elongation
Axial elongation

8,E-02 8,E-02

6,E-02 6,E-02
4,E-02 4,E-02
2,E-02
2,E-02
0,E+00
0,E+00
-2,E-02
-2,E-02
-4,E-02
-6,E-02 -4,E-02

-8,E-02 fshd3 max -6,E-02 fshd4 max

fshd3min fshd4min
-1,E-01 -8,E-02
0 0,25 0,5 0,75 1 1,25 1,5 0 0,25 0,5 0,75 1 1,25 1,5
acc scale factor acc scale factor
Axial elongation

Axial elongation

1,E-01 8,E-02
8,E-02
6,E-02
6,E-02
4,E-02
4,E-02
2,E-02
2,E-02
0,E+00
0,E+00
-2,E-02
-2,E-02
-4,E-02
-4,E-02
fshd7 max -6,E-02 fshd8 max
-6,E-02
fshd7min fshd8min
-8,E-02 -8,E-02
0 0,25 0,5 0,75 1 1,25 1,5 0 0,25 0,5 0,75 1 1,25 1,5
acc scale factor acc scale factor

Figure 6.39. Obtained values of Axial Elongation for FSHD devices at both levels

161
 Axial force [N]

Axial force [N]

3,0E+06 3,0E+06

2,0E+06 2,0E+06

1,0E+06 1,0E+06

0,0E+00 0,0E+00

-1,0E+06 -1,0E+06

-2,0E+06 -2,0E+06
fshd3 max fshd4 max
fshd3min fshd4min
-3,0E+06 -3,0E+06
0 0,25 0,5 0,75 1 1,25 1,5 0 0,25 0,5 0,75 1 1,25 1,5
acc scale factor acc scale factor
Axial force [N]

Axial force [N]

3,0E+06 3,0E+06

2,0E+06 2,0E+06

1,0E+06 1,0E+06

0,0E+00 0,0E+00

-1,0E+06 -1,0E+06

-2,0E+06 -2,0E+06 fshd8 max

fshd7 max
fshd7min fshd8min
-3,0E+06 -3,0E+06
0 0,25 0,5 0,75 1 1,25 1,5 0 0,25 0,5 0,75 1 1,25 1,5
acc scale factor acc scale factor

Figure 6.40. Obtained values of Axial Force for FSHD devices at both levels
Axial energy

Axial energy

6,0E+05 5,0E+05

5,0E+05
3,8E+05
4,0E+05

3,0E+05 2,5E+05

2,0E+05
1,3E+05
1,0E+05
fshd4
fshd3
0,0E+00
0,0E+00
0 0,25 0,5 0,75 1 1,25 1,5
0 0,25 0,5 0,75 1 1,25 1,5
acc scale factor acc scale factor
Axial energy

Axial energy

6,0E+05 5,0E+05

5,0E+05
3,8E+05
4,0E+05

3,0E+05 2,5E+05

2,0E+05
1,3E+05
1,0E+05
fshd7 fshd8
0,0E+00 0,0E+00
0 0,25 0,5 0,75 1 1,25 1,5 0 0,25 0,5 0,75 1 1,25 1,5
acc scale factor acc scale factor

Figure 6.41. Obtained values of Axial Energy for FSHD devices at both levels

Finally, the obtained values for relative displacements at both floor levels are shown in Figure 6.42; in each
case the limits posed both for SLS and ULS were fulfilled.

162
 Drift I floor level

Drift II floor level

1,2E-01 1,0E-02

8,0E-02
5,0E-03
4,0E-02

0,0E+00 0,0E+00

-4,0E-02
-5,0E-03
-8,0E-02
node4 max node5 max
node4 min node5 min
-1,2E-01 -1,0E-02
0 0,25 0,5 0,75 1 1,25 1,5 0 0,25 0,5 0,75 1 1,25 1,5
acc scale factor acc scale factor

Figure 6.42. Obtained values for drift at the two floor levels

The results obtained from the application of the proposed design procedure to the selected study case,
confirm that it can provide satisfactory results.

163
7. Results, conclusions and perspectives
The results of the research were more relevant and technologically advanced respect those declared in the
original research proposal. In fact, the automated software is now a web application completely free for all
practitioners and engineers, able to furnish a complete cost plan of the structure and general technical
drawings. Additionally, the testing programme firstly addressed to simple connections or details became a
more complex experimental study devoted to the structural assessment of a new and optimized structural
solution to be employed for IB and CB.
Moreover, all technological advanced or high performing solutions as CB with additional dissipative devices
for seismic protection or IB realized with web corrugated sheeting with improved connection details were
equipped also with design guidelines. The suitable combination of all these aspects with the original working
plan, see figure XIII, could have an high impact due to: utilization simplicity of the software, high quality of
information at disposal of the user (drawings, technical info, cost evaluation), suitable choice of geometries
and solutions requested by the market, open architecture of the PRECASTEEL software that can be
continuously updated with new technical solutions. Employment of light gauge steel members in seismic
areas and the application of dissipative devices based on different materials coupled with steel structures and
prefabricated are relevant technical aspects that, beyond PRECASTEEL conclusions, could be further
developed and extended in future researches. For this purpose, ILVA S.p.A. is actually defining a dedicate
web-site on which insert all complete products (deliverables) for a future dissemination action.

165
REFERENCES
‚ U. Peil, Trapezstegträger mit Steg-Gurt-Verbindung mit mechanischen Verbindungsmitteln, DAST
Forschungsbericht 3/1997
‚ K. Roik, J. Lindner, J. Carl, Biegetorsionsprobleme gerader dünnwandiger Stäbe
‚ Perälä, H. Nuuttila, and L. Fülöp, Collection of data on industrial and commercial building from:
Finland, Poland, the Czech Republic, Hungary and Romania. Espoo: VTT Technical Research Centre
of Finland, 2008, p. 37.
‚ L. Fülöp and P. Beaucaire, Advanced analysis of the performance of steel frames. Espoo: VTT
Technical Research Centre of Finland, 2009, p. 43.
‚ P. Hradil, M. Mielonen, and L. Fülöp, Optimization tools for steel portal frames - Software
documentation. Espoo: VTT Technical Research Centre of Finland, 2010, p. 33.
‚ P. Hradil, M. Mielonen, and L. Fülöp, Optimization tools for steel portal frames - Optimization results.
Espoo: VTT Technical Research Centre of Finland, 2011, p. 18.
‚ P. Hradil, M. Mielonen, and L. Fülöp, Optimization tools for steel portal frames - Effective modelling
of lateral supports. Espoo: VTT Technical Research Centre of Finland, 2011, p. 33.
‚ P. Hradil, M. Mielonen, and L. Fülöp, Optimization tools for steel portal frames - Design and
optimization methods. Espoo: VTT Technical Research Centre of Finland, 2010, p. 57.
‚ M. Mielonen, “Optimization of steel portal frames using genetic algorithms,” Master of Science, Aalto
University School of Science and Technology, 2010.
‚ Hradil P., Mielonen M., Fülöp L., VTT R 00524 10 Research report, Optimization tools for steel
portal frames – software documentation, Espoo, 2009
‚ K. Roik, G. Sedlacek: Biege- und Verdrehtheorie unter Berücksichtigung von Schubverformungen,
Die Bautechnik 1/1970, page 20ff
‚ AISC – American Institute of Steel Construction. (2005). Seismic Provisions for Structural Steel
Buildings. ANSI/AISC 341-05, AISC INC. Chicago, Illinois.
‚ CEN Technical Commission 250/SC8. (2005). UNI-EN1998-1-1: Eurocode 8 - Design of structures
for earthquake resistance. Part 1: General rules, seismic actions and rules for buildings. CEN,
Brussels, 2005.
‚ CEN Technical Commission 250/SC3. (2005a). UNI-EN1993-1-1: Eurocode 3 – Design of steel
structures. Part 1-1: General rules and rules for buildings. CEN, Brussels, 2005.
‚ CEN Technical Commission 250/SC4. (2005b). UNI-EN1994-1-1: Eurocode 4 – Design of composite
steel and concrete structures. Part 1-1: General rules and rules for buildings. CEN, Brussels, 2005.
‚ CEN Technical Commission 250. (2006). UNI-EN1990: Eurocode Basis of structural design. CEN,
Brussels, 2005.
‚ CEN Technical Commission 250 (2004). UNI-EN1991-1-1: Eurocode 1 – Actions on structures. Part
1-1: General actions - Densities, self-weight, imposed loads for buildings.
‚ CEN Technical Commission 250 (2004a). UNI-EN-1997-1-1: Eurocode 7 – Geotechcnial desing –
Part 1-1: General rules. CEN, Brussels, 2004.
‚ Vamvatsikos, D. and C. Allin Cornell, 2002, “Incremental Dynamic Analysis,” Earthquake
Engineering and Structural Dynamics, Vol. 31, Issue 3, pp. 491-514.
‚ ATC, 2007a, Guidelines for Seismic Performance Assessment of Buildings - 35% Draft, Report No.
ATC-58, prepared by the Applied Technology Council for the Federal Emergency Management
Agency, Washington, D.C.

167
Acronyms and Abbreviations

ADRS Acceleration Displacement Response Spectrum

AJAX Asynchronous Java script And XML
CB Commercial buildings
CF Census Form
CFCHS Concrete Filled Circular Hollow Section
CHS Circular Hollow Section
CSS Cascade Style Sheet
CWG Corrugated Wed Girder
CZR Czech Republic
D Diagonal
DCH Ductility Class High
DCL Ductility Class Low
DCM Ductility Class Medium
DDBD Direct Displacement Based Design
DLS Damage Limit State
DOF Degree Of Freedom
EB Eccentrical Bracing
EBF Eccentrically Braced Frame
EU European Union
FE Finite Element
FEM Finite Element Modelling
FEMA Federal Emergency Management Association
FSHD Flag Shaped Hysteretic Device
GA Genetic Algorithm
GMA General Method Analysis
GNLA Global Nonlinear Analysis
HDR High Dissipative Rubber
HDRB High Dissipative Rubber Bearings
HR Hot-Rolled
IB Industrial buildings
IDA Incremental Dynamic Analysis
IFM Interaction Formulae Method
L Lower
LAMP Linux, Apache, MySQL and PHP
LBA Lateral Buckling Analysis
LGS Light Gauge Steel
LRCB Low-Rise Commercial Buildings
LTB Lateral Torsional Buckling
LVDT Linear Variable Differential Transformer
LVL Laminated Veener Lumber
MADRS Modified Acceleration Displacement Response Spectrum
MDOF Multi Degree Of Freedom
MR Moment Resisting
O Orthogonal
PE Partially Encased
PGA Peak Ground Acceleration
PP Performance Point
PRC Prestressed Reinforced Concrete
PRCW Prefabricated Reinforced Concrete Walls
PSO Particle Swarm Optimization
RC Reinforced Concrete
RCGA Real-Code Genetic Algorithm
SBX Simulated Binary Crossover

169
SDOF Single Degree Of Freedom
SLS Serviceability Limit State
SMA Shape Memory Alloys
SSIB Single Story Industrial Building
U Upper
ULS Ultimate Limit State
VB Visual Basic
WT Welded Tapered
XHTML Extensible Hypertext Mark-up Language

170
List of figures

Figure I. (a) typical beam-column connection with corbel and gutter beam; (b) collapsed precast r.c. frames
by earthquake – Turkey 1990; (c) typical Precast r.c. solution for low-rise commercial
building in Italy; (d) typical configuration of one-storey r.c. industrial building……….……7
Figure II. General flow-chart of the research project……………………………………………………..……8
Figure III. (a) most adopted materials for CB solutions; (b) number of storey for structures employed for
commercial activities………………………………………………………………….………9
Figure IV. (a) Number of bays in industrial buildings; (b) type of structural solutions………………….……9
Figure V. (a) Morphological general scheme for IB solutions; (b) Morphological general scheme for CB
solutions………………………………………………………………………………….……9
Figure VI. (a) IB solutions realized with light gauge steel members; (b) part of IB solution with corrugated
steel sheeting tested during the experimental programme……………………………….….11
Figure VII. (a) eccentric bracing system: shear link; (b) prefabricated reinforced concrete wall…………....12
Figure VIII. (a) prototype of FSHD system; (b) schematic mechanical behavior of FSHD………………....12
Figure IX. (a) HDRB system; (b) stable loops of HDRB at different maximum strains…………………......12
Figure X. Images of the web application containing PRECASTEEL deliverables ready-to-use: (a) selection
of roofing and cladding system for IB; (b) selection of columns type in CB adequate for the
vertical loading level; (c) complete cost analysis of the final structural modulus………..…13
Figure XI. Example of IB solution produced by web application – DXF drawing elements: (a) plan view; (b)
front view; (c) side view; (d) connections: apex and eave……………………………..……14
Figure XII. EV-tool: (a) Input window - fixing general frame geometry; (b) Input window – introduction of
loading information; (c) Input window – parameters for the optimization procedure of the
frame; (d) summary of optimization process results: steel sections; cost of the structure;
sustainability and environmental impact………………………………………………..…...14
Figure XIII. General organization of the project……………………………………………………….…….16
Figure 1.1. Main activities performed in the industrial buildings……………………………………..……..18
Figure 1.2. Type of structural system – a) survey country by country – b) European survey………..……...18
Figure 1.3. In-plane and out-of-plane direction in industrial building…………………………..……………18
Figure 1.4. Number of frame bays (in-plane direction – along the main resisting system)……..…………...19
Figure 1.5. Numbers of frames in the out-of-plane directions – (a) subdivision country by country – (b)
Aggregation of the data (European level)……………………………………………..…….19
Figure 1.6. Span length of each bay recognized during the investigation – (a) subdivision of the results
country by country – (b) Aggregation of the data (subdivision at European level)……..….19
Figure 1.7. Distance between two consecutive frames – (a) subdivision of the results country by country – (b)
Aggregation of the data (subdivision at European level)………………………………..….20
Figure 1.8. Height of the columns – (a) subdivision of the results country by country – (b) Aggregation of the
data (subdivision at European level)…………………………………………………..…….20
Figure 1.9. Type of girder employed in the main frame – (a) subdivision of the results country by country –
(b) Aggregation of the data (subdivision at European level)……………………….….……21
Figure 1.10. Slope of frame girder – (a) subdivision of the results country by country – (b) Aggregation of
the data (subdivision at European level)……………………………………………..……...21
Figure 1.11. Snow loading – (a) subdivision of the results country by country – (b) Aggregation of the data
(subdivision at European level)……………………………………………………..……….22
Figure 1.12. Live loading on roof system – (a) subdivision of the results country by country – (b)
Aggregation of the data (subdivision at European level)…………………………..………..22
Figure 1.13. Distribution of the crane capacity in the industrial buildings – (a) subdivision of the results
country by country – (b) Aggregation of the data (subdivision at European level)…..……..23
Figure 1.14. Resisting system of in-plane direction – (a) subdivision of the results country by country – (b)
Aggregation of the data (subdivision at European level)……………………………….…...23
Figure 1.15. Resisting system of out-of-plane direction – (a) subdivision of the results country by country –
(b) Aggregation of the data (subdivision at European level)………………………….…….24
Figure 1.16. Maximum peak ground acceleration – (a) subdivision of the results country by country – (b)
Aggregation of the data (subdivision at European level)……………………………………24

171
Figure 1.17. Wind load intensity – (a) subdivision of the results country by country – (b) Aggregation of the
data (subdivision at European level)………………………………………………..……….25
Figure 1.18. Type of roofing systems – (a) subdivision of the results country by country – (b) Aggregation of
the data (subdivision at European level)…………………………………………….………25
Figure 1.19. Type of side-cladding systems – (a) subdivision of the results country by country – (b)
Aggregation of the data (subdivision at European level)……………………………..……..26
Figure 1.20. Type of purlins – (a) subdivision of the results country by country – (b) Aggregation of the data
(subdivision at European level)…………………………………………………………..….26
Figure 1.21. Typical supermarket for food-stuff sales layout……………………………………………..….27
Figure 1.22. Typical department store layout…………………………………………………………..…….27
Figure 1.23. Material: (a) vertical elements; (b) horizontal elements (% from the number of buildings)…....29
Figure 1.24. Material: most adopted typologies…………………………………………………………...….29
Figure 1.25. Number of storeys (% from the number of buildings)……………………………………….....30
Figure 1.26. Storey height (% from the number of buildings)…………………………………………….....30
Figure 1.27. Lateral resisting structural systems (% from the number of buildings)……………………...….31
Figure 1.28. Lateral resisting structural systems (% from the number of buildings)……………………...….31
Figure 1.29. Span length: (a) solid elements; (b) truss elements (% from the number of buildings)……..….31
Figure 1.30. Seismic peak ground acceleration (% from the number of cases)…………………………..…..32
Figure 1.31. New industrial building volumes (€) is increasing in four countries. Source: Euroconstruct
6/2007…………………………………………………………………………………….….33
Figure 1.32. New commercial building volumes (€) can include also same kind of building space like storage
buildings. Euroconstruct 6/2007………………………………………………………….….33
Figure 1.33. New warehouses in sub-sectors in Finland…………………………………………………..….34
Figure 1.34. Share of different materials/structural typologies in case of SSIB…………………………..….35
Figure 1.35. (a) Total area and (b) and number of spans of SSIB’s……………………………………….…36
Figure 1.36. (a) Span and (b) height of SSIB’s……………………………………………………………....36
Figure 1.37. Share of different materials/structural typologies in case of LRCB…………………………….36
Figure 1.38. (a) ) Total area and (b) typical spans for LRC buildings………………………………………..37
Figure 1.39 (a) cost of the intumescent paint, €/m2, for REI60 transformed in €/kg; (b) cost of intumescent
paint, €/m2, for REI120 transformed in €/kg………………………………………………..38
Figure 1.40. Comparisons of costs between different structural systems in different markets: (a) CB
elements; (b) IB structural types – material costs…………………………………………...39
Figure 1.41. Comparisons of costs between different structural systems in different markets: (a) CB
elements; (b) IB structural types – production and assembling of structural elements……...40
Figure2.1: Basic module for single and double bay industrial building configuration……………………….43
Figure 2.2. (a) Truss girder scheme adopted in the preliminary design, (b) Beam that connects the heads of
the columns in the longitudinal direction……………………………………………………45
Figure 2.3. Comparison between assumed span length and steel consumption with variable snow loads and
variable seismic action (single span)………………………………………………………...45
Figure 2.4. Comparison between assumed span length and steel consumption with variable snow loads and
variable seismic action (double span)……………………………………………………….46
Figure 2.5. Comparison between assumed span length and steel consumption with variable snow loads,
variable seismic action and crane load influence (single span)……………………………...46
Figure 2.6. Influence of fixing condition at the column bases and PGA on steel consumption……………...46
Figure 2.7. Direct comparison between steel consumption and PGA………………………………………...46
Figure 2.8. Influence of crane load on steel consumption……………………………………………………47
Figure 2.9. Type of cold formed sections considered for the members sizing……………………………….47
Figure 2.10. Comparison between hot-rolled and cold-formed solutions…………………………………….48
Figure 2.11. Steel consumption for all designed tapered cases with 32 m span……………………………...49
Figure 2.12. (a) The developed model in SAP2000, (b) Imperfection load application - Deformed shape….50
Figure 2.13. (a) PushOver for seismic performance assessment; (b) loading of IB structure with static loads
at ULS – comparison between demand and capacity………………………………………..51
Figure 2.14. (a) Hot-rolled (HR) frames; (b) Welded tapered (WT) frames…………………………………52
Figure 2.15. LGS frame with back-to-back C elements and corner fixings………………………………….53
Figure 2.16. a) Initial profile; (b) and simplified shape for the frame analysis………………………………53

172
Figure 2.17. Typical LTB failure of a portal frame structure………………………………………………...53
Figure 2.18. Steps of the Global non-linear analysis…………………………………………………………54
Figure 2.19. Column and rafter names (WT frames)…………………………………………………………56
Figure 2.20.Steel consumption of single-span buildings…………………………………………………......57
Figure 2.21. Steel consumption of double-span buildings……………………………………………………57
Figure 2.22. Connection to purlins to the frame……………………………………………………………...57
Figure 2.23. Stiffness from purlin to frame…………………………………………………………………...58
Figure 2.24. Improved connection to purlins to the frame……………………………………………………58
Figure 2.25. Corner connection typologies for the WT frames………………………………………………59
Figure 2.26: General layout of proposed CWG………………………………………………………………61
Figure 2.27. Welding procedure………………………………………………………………………………61
Figure 2.28. ThyssenKruppHoesch HP 107…………………………………………………………………..62
Figure 2.29. Schrag CL-profiles………………………………………………………………………………63
Figure 2.30. Example of taylor made profiles for flanges……………………………………………………63
Figure 2.31. Two fasteners M 10 12.9 each rib………………………………………………………………63
Figure 2.32. Connection girder-column (see also figure 2.23)……………………………………………….64
Figure 3.1. Generic structural scheme for commercial buildings: (a) eccentric braces; (b) concentric braces;
(c) r.c. shear walls; (d) r.c. shear walls and dissipating devices……………………………..65
Figure 3.2. (a) Flooring system pattern; (b) geometry of truss girders; (c) double web-angle connections for
flooring systems; (d) bolted connections for trusses obtained with cold formed profiles…..66
Figure 3.3. (a) Columns tributary areas; (b) schemes adopted in the design…………………………………66
Figure 3.4. Braces considered in the project………………………………………………………………….67
Figure 3.5. Individuation of most pre-stressed wall element in the commercial building plan………………67
Figure 3.6. the precast double slab wall solution……………………………………………………………..67
Figure 3.7. (a) HDR-based device; (b) hysteretic device……………………………………………………..68
Figure 3.8. Stable loops of an HDR-based device: (a) different strain amplitudes; (b) different strain rates..68
Figure 3.9. (a) dissipative device assembled;(b) steel fuses inserted in the FSHD system…………………..69
Figure 3.10. Hysteretic rule for ‘flag shaped’ devices………………………………………………………..69
Figure 3.11. Definition of the storey forces…………………………………………………………………..71
Figure 3.12. (a) cold formed profile ZKU; (b) cold formed profile ZKUG; (c) imperfection of truss bracing
members……………………………………………………………………………………..73
Figure 3.13. Axial force in (a) one- and (b) two-storey concentric braces under horizontal forces
(red = compression)………………………………………………………………………….73
Figure 3.14. Internal actions in two-storey eccentric braces with two diagonals…………………………….74
Figure 3.15. Internal actions in two-storey eccentric braces with one diagonal……………………………...75
Figure 3.16. Effective link length…………………………………………………………………………….76
Figure 3.17. Data base: a) and b) beam cross sections for primary and secondary elements; c) D-shaped
bracing system……………………………………………………………………………….77
Figure 3.18. Geometry layout for Type 1 (CS1) frame case studies………………………………………….79
Figure 3.19. Geometry layout for Type 2 (CS2) frame case studies………………………………………….79
Figure 3.20. Geometry layout for Type 3 (CS3) frame case studies………………………………………….80
Figure 3.21. 3D model for response spectrum analysis for CS1 Type frames……………………………......81
Figure 3.22 . 3D model for response spectrum analysis for CS2 Type frames……………………………….81
Figure 3.23 . 3D model for response spectrum analysis for CS3 Type frames……………………………….82
Figure 3.24. Plane frames position……………………………………………………………………………83
Figure 3.25. CS1H5 planar frame in X direction……………………………………………………………..83
Figure 3.26 . CS1H5 planar frame in Y direction…………………………………………………………….83
Figure 3.27. Planar frames position…………………………………………………………………………..84
Figure 3.28. CS2H5 planar frame in X direction……………………………………………………………..84
Figure 3.29. CS2H5 planar frame in Y direction……………………………………………………………..84
Figure 3.30. Set of artificial accelerograms adopted for incremental dynamic analyses – 2 of the 7 time
histories……………………………………………………………………………………...84
Figure 3.31. Capacity curve for CS1H5 case study: a) X planar frame; b) rotation of Link 3 – second floor
level………………………………………………………………………………………….85

173
Figure 3.32. Distribution of horizontal forces: (a) seismic and wind action; (b) wind action; (c) seismic
action; (d) seismic action…………………………………………………………………….86
Figure 3.33. Simplified approach to estimate the influence area A of a single wall…………………………87
Figure 3.34. Operations of assembling precast r.c. wall………………………………………………….......89
Figure 4.1. 4-point-bending tests……………………………………………………………………………..91
Figure 4.2. Load introduction…………………………………………………………………………………92
Figure 4.3. a) Boundary conditions and b) plates for cross-sectional stability……………………………….92
Figure 4.4. 4-point-bending-tests: Load vs. displacement……………………………………………………94
Figure 4.5. Torsional deformation of corrugation (here: specimen 3)……………………………………......94
Figure 4.6. Local crinkling and peeling of the bolts (here: specimen 2)……………………………………..94
Figure 4.7. Test 3 – local damages after testing………………………………………………………………95
Figure 4.8. Slippage of blind rivets, specimen 1……………………………………………………………...95
Figure 4.9. Deformation of specimen 2………………………………………………………………………96
Figure 4.10. (a) Testing of frame corner; (b) Load introduction and lateral support…………………………96
Figure 4.11. Load-deflection curve of specimen 5 - Maximum force: 206 kN………………………………97
Figure 4.12. Failure of specimen 5……………………………………………………………………………97
Figure 4.13. Load-deflection curves of test 6………………………………………………………………...98
Figure 4.14. Global view and sections of external case………………………………………………………99
Figure 4.15. Global view and sections of internal sliding frame……………………………………………..99
Figure 4.16. Connecting plates………………………………………………………………………………..99
Figure 4.17 Piston …………………………………………………………………………………………..99
Figure 4.18. Anchor plates…………………………………………………………….…………………….100
Figure 4.19. a) Dissipative element b) buckling restraining system………………….…………………….100
Figure 4.20. Pre-stressing cable……………………………………………………….…………………….100
Figure 4.21. General test setup……………………………………………………….……………………..101
Figure 4.22. a) Connection between hydraulic jack and dissipating device b) Connection between dissipative
device and the fixed structure ………………………………………….…………………..101
Figure 4.23.Sensor position………………………………………………………………………………….102
Figure 4.24. Displacement history used for the short testing procedure……………………………………102
Figure 4.25. Force. Internal Frame displacement curve……………………………………………………..103
Figure 4.26. Local buckling of the dissipative element……………………………………………………..103
Figure 4.27. Loss of contact between the anchor plate and the welded sheet………………………………104
Figure 4.28. C-formed element used to assure the contrast…………………………………………………104
Figure 4.29. Different Flag shaped hysteresis obtained varying steel fuses section and pre-stressing rate..104
Figure 5.1. Structure of the web application interface, similar for both buildings………………………….111
Figure 5.2. Bar chart to facilitate user selections……………………………………………………………111
Figure 5.3. Cost module structure…………………………………………………………………….……..112
Figure 5.4. Flow chart of industrial building application……………………………………………………113
Figure 5.5. Step 1 of industrial building application……………………………………………….………..114
Figure 5.6. Step 2 of industrial building application………………………………………………….……..114
Figure 5.7. Flow chart of commercial building application…………………………………………………115
Figure 5.8. Step 4 of commercial building application………………………………………….…………..116
Figure 5.9. Statistical case analysis……………………………………………………………….…………117
Figure 5.10. The basic diagram of two optimization tools developed at VTT………………….…………..118
Figure 5.11. HR frame optimization results for persistent/transient design situation, axis-to-axis span 16 m,
height 6 m, roof angle 15%, haunch length 1/11 of the span, distance between frames 6 m,
pinned supports, dead load 380 N/m2, wind load 30 m/s, terrain type 2, steel S275 with
strain hardening, design method General Method, tournament selection, simulated binary
crossover, polynomial mutation……………………………………………………………119
Figure 5.12. ABAQUS script AP-Frame……………………………………………………………………120
Figure 5.13. ABAQUS script parallel processing………………………………………………….………..121
Figure 5.14. The first view of EV-Frame tool when started, showing the geometry tab………….………..122
Figure 5.15. Elastic analysis algorithm with Method 1 (GMNIA) – blue, Method 2 (GM) – green and Method
3 (Cross-sectional checks) – yellow………………………………………………….…….123

174
Figure 5.16. Evolution of the elite frame fitness in five optimisations of the same frame configuration with
the GA………………………………………………………………………………..…….124
Figure 5.17 Flowchart of the GA optimization process………………………………………………..……125
Figure 5.18. Detailed flowchart of the GA optimization………………………………………………..…..126
Figure 5.19. View of optimization tab in the EV-Frame tool………………………………………….……127
Figure 6.1. Basic module for the one bay industrial building configuration……………………………..…129
Figure 6.2. Basic module for the two bay industrial building configuration……………………………..…129
Figure 6.3. Cold-formed section profile adopted for IB modulus………………………………………..….130
Figure 6.4. Typical section for the adopted trussed girder solution………………………………………...130
Figure 6.5. ID scheme for element in trussed girder solution…………………………………………….…131
Figure 6.6. Geometrical and structural scheme of welded-tapered frame considered in the Precasteel
database…………………………………………………………………………………….131
Figure 6.7. Views of proposed layout of CWG……………………………………………………………..132
Figure 6.8: Geometry according to EC 3 [image source: EN 1993-1-5]……………………………….…..133
Figure 6.9: Deflection of specimen 2 at transverse force VEd = 100
kN………………………………………………………………………………..…………134
Figure 6.10: Warping- and weighting function………………….…………………………………………..134
Figure 6.11: Choice of warping functions…………………………………………………………………...136
Figure 6.12: Study case n°1 (dissipative r.c. walls)…………………………………………………………137
Figure 6.13: Typical technical drawing for r.c. wall………………………….……………………………..138
Figure 6.14. Position of connective zones between wall and gravity structure…………….……………….138
Figure 6.15. Connection between r.c. wall and steel beams of the floor system…………..………………..138
Figure 6.16: Study case n°2 (dissipative r.c. walls)…………………………………………………………140
Figure 6.17: Technical drawing for r.c. wall (study case n°1, dissipative devices coupled with r.c.
walls)……………………………………………………………………………………….141
Figure 6.18. Venusio Shopping Center (www.promozioneacciaio.it)........................................................142
Figure 6.19 – Gravity structure: (a) flooring system regular grid; (b) pinned beams and columns; (c) possible
connection between CHS and composite beams; (d) possible connection between hot rolled
columns and composite beams……………………………………………………………..143
Figure 6.20. Joints and connections for truss obtained with ZKU and SKUG cold formed…………….…..143
Figure 6.21. Flooring systems - cost per square meter vs. area of the plan grid (data from Central
Europe)……………………………………………………………………………………..144
Figure 6.22. Columns - cost per square meter vs. tributary area (data from Central Europe)………..……..145
Figure 6.23. Braces - cost per square meter vs. tributary area (data from Central Europe)………..………..146
Figure 6.24. Simplified modelling for designing HDR-based dissipative devices…………………….……147
Figure 6.25. Implementation of HDR-based devices: (a) position of devices in a low rise building; (b)
possible details for the connection of devices to the r.c. wall and to the floor slab……..…148
Figure 6.26. Frame equipped with FSHD: schematic of the adopted structural system and of its
components…………………………………………………………………………………149
Figure 6.27. CS1H5 case study – shear wall disposition: a) in plane; b) X-elevation; Y-elevation……..….151
Figure 6.28. Numerical models of planar frames: a) MDOF system; b) equivalent 2DOF system………....152
Figure 6.29. ‘Flag-shaped’ hysteretic curve: the main parameters characterizing cyclic behaviour…….….153
Figure 6.30. Determination of starting performance point a) in the short period range; b) in medium and long
period range…………………………………………………………………………….…..156
Figure 6.31 . Determination of estimated maximum displacement using intersection of capacity spectrum
with MADRS……………………………………………………………………….………156
Figure 6.32. Graphical representation of the capacity spectrum method of equivalent linearization….……157
Figure 6.33. Monotonic curve of the single FSHD device adopted for CS1H5 case study………..………..158
Figure 6.34. Pushover curve obtained for the equivalent SDOF system (in blue)……………….………….159
Figure 6.35. Determination of the structural performance point through iterative procedure in AD
format………………………………………………………………………………………159
Figure 6.36. Determination of the cyclic curve of the FSHD……………………………………….………160
Figure 6.37. Set of artificial accelerogram adopted to perform Incremental Dynamic Analyses……..…….160
Figure 6.38. Obtained values for Shear Force at the base of precast r.c walls………………………..…….161
Figure 6.39. Obtained values of Axial Elongation for FSHD devices at both levels………………….……161

175
Figure 6.40. Obtained values of Axial Force for FSHD devices at both levels……………….…………….162
Figure 6.41. Obtained values of Axial Energy for FSHD devices at both levels…………………...………162
Figure 6.42. Obtained values for drift at the two floor levels……………………………………...………..163

176
List of tables

Table I. Costs of columns for CB in Southern Europe (these costs include a lump-rate estimation of steel
working, bolts and general expenditure)…………………………………………….………..3
Table II. Costs of floor systems for CB in Southern Europe (these costs include a lump-rate estimation of
steel working, bolts and general expenditure)…………………………………………..…….3
Table 1.1 Contributions received by PRECASTEEL partners……………………………………………….17
Tab.1.2. Main structure and floor systems combinations…………………………………………….……....32
Table 1.3. Costs of CB elements for the Italian market…………………………………………………...….38
Table 1.4. Costs of CB elements for Romanian market……………………………………………...…….…38
Table 1.5. Costs of CB elements for German market………………………………………………...………39
Table 1.6. Costs IB modular structures for the three different markets…………………………….......…….39
Table 1.7. Abstract of data collection about some costs of roofing and flooring systems…………...…….…41
Table 1.8. Abstract of data collection – some costs of cladding systems…………………………...…….….42
Table 2.1. Assumed design value of acting loads…………………………………………………………….44
Table 2.2. Parameters assumed for designing light gauge steel solutions……………………………………47
Table 2.3. Profiles size obtained from structural design……………………………………………………...48
Table 2.4. Geometrical parameters of tapered solutions with 32 m spans. Note: Haunch IPE 600 * refer to the
haunched section used for original cases as described in paragraph 6.1………………….…49
Table 2.5. Geometric parameters and loads considered in the design of 32m tapered solutions………….….49
Table 2.6. The final selection of cross-sections for all the double-span cases examined…………………….52
Table 2.7. Component performance in case of LGS frame connection…………………………………...….60
Table 2.8. Selected case studies for the application of CWG………………………………………...………62
Table 2.9. CWG solutions……………………………………………………………………………...……..62
Table 3.1. Set of designed flooring systems…………………………………………………………..……...71
Table 3.2. Set of designed columns…………………………………………………………………..………72
Table 3.3. Set of designed braces……………………………………………………………………...……...72
Table 3.4 - Definition of analyzed case studies derived from the application of database cross-sections…...77
Table 3.5. Response Spectrum Analyses performed on 3D model set of frames………………………...…..82
Table 3.6. Input data for the selection of a single Precast r.c. wall……………………………………...……88
Table 3.7. Output data for the selection of a single Precast r.c. wall………………………………...……….88
Table 4.1. 4-point-bending tests………………………………………………………………………………92
Table 4.2. Arrangement of fasteners………………………………………………………………………….93
Table 4.3. 4-point-bending-tests: Maximum forces…………………………………………………..………93
Table 4.4. Dissipative and prestressing element geometrical and mechanical characteristics………...……103
Table 6.1. External actions considered in the structural design…………………………………..…………131
Table 6.2. Definition of analyzed case studies derived from the application of database cross-sections…..150
Table 6.3 Drift limits for different design performance levels…………………………………….………..152
Table 6.4. Geometric and mechanic characteristics of the prefabricated shear wall………………..………157
Table 6.5. Values of the applied seismic mass at each floor level………………………………..…………157
Table 6.6. Determination of the base shear force in correspondence of the Serviceability Limit State….…158
Table 6.7. Determination of yield force and elastic stiffness for equivalent FSHD at each floor level….…158
Table 6.8. Obtained values for the modal participation factor Γ and equivalent mass ୣ of the SDOF
system………………………………………………………………………………………158
Table 6.9. Characteristic yield point of the equivalent SDOF system and first trial performance point…...159
Table 6.10. Determination of performance point PP and of the corresponding equivalent damping Ⱦୣ୤୤ …..159
Table 6.11. Determination of the β factor defining the amplitude of the hysteretic cycle……………….…160

177
European Commission

EUR 25871 — Prefabricated steel structures for low-rise buildings in seismic areas (Precasteel)

Luxembourg: Publications Office of the European Union

2013 — 177 pp. — 21 × 29.7 cm

ISBN 978-92-79-29011-4
doi:10.2777/5499
 HOW TO OBTAIN EU PUBLICATIONS
Free publications:
• via EU Bookshop (http://bookshop.europa.eu);
• at the European Union’s representations or delegations. You can obtain their
contact details on the Internet (http://ec.europa.eu) or by sending a fax
to +352 2929-42758.
Priced publications:
• via EU Bookshop (http://bookshop.europa.eu).
Priced subscriptions (e.g. annual series of the Official Journal of the
European Union and reports of cases before the Court of Justice
of the European Union):
• via one of the sales agents of the Publications Office of the European Union
(http://publications.europa.eu/others/agents/index_en.htm).

EUROPEAN COMMISSION
Directorate-General for Research and Innovation
Directorate G — Industrial Technologies
Unit G.5 — Research Fund for Coal and Steel

E-mail: rtd-steel-coal@ec.europa.eu
RTD-PUBLICATIONS@ec.europa.eu

Contact: RFCS Publications

European Commission
B-1049 Brussels
 KI-NA-25871-EN-N

EU
Low-rise buildings are used for industrial and commercial activities. Most of
these buildings are built by prefabricated-concrete elements characterised by

Prefabricated steel
low efficiency of connections and a lower ductility performance when subjected
to earthquake loading.

The proposal aim is defining prefabricated steel solutions for single-storey and
low-rise buildings in earthquake-prone areas for industrial and commercial

structures for low-rise

activities. The selection of structural solutions is performed via statistical
analyses in different European countries about solutions requested by the
market. Industrial and commercial solutions are optimised in terms of structural
performance and of construction costs and two of those (technologically

Prefabricated steel structures for low-rise buildings in seismic areas (Precasteel)

buildings in seismic areas
advance and high-performing) are experimentally tested. The solution
selected for commercial activities is a dissipative device working in series with
prefabricated RC-Wall and pendulum steel-structure. The solution selected for
industrial activities is a warehousing system using cold-formed profiles acting
compositely with ribbed steel sheeting as girder web.

Automated integrated design software with cost-effectiveness analysis module

is realised to favour the use of proposed solutions into practice. Solutions
(Precasteel)
proposed by the software are optimised to meet three ‘criteria’:

(1) Prefabrication — structural members and connection types selected in order

to reduce on-site operations;

Pre-designed — structural solutions pre-designed and inserted in a

(2) 
database correlating member/structure with its maximum structural
performance; structural configuration for industrial/commercial buildings are
assembled according to the design inputs;

(3) Standardised — members and structural systems have static schemes

suitable for defining simplified designing procedures (transformed in
technical tables — Precasteel Software): database (catalogue) of solutions/
members may be enlarged enclosing more recent developments.

Studies and reports

EUR 25871

doi:10.2777/5499 Research and

Innovation EUR 25871 EN
View publication stats