Professional Documents
Culture Documents
Kina25871enn 002
Kina25871enn 002
net/publication/274640991
CITATIONS READS
6 9,341
22 authors, including:
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
All content following this page was uploaded by George E. Varelis on 08 April 2015.
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
(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;
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
European Commission
B-1049 Brussels
European Commission
Final report
Directorate-General for Research and Innovation
ISBN 978-92-79-29011-4
doi:10.2777/5499
Printed in Luxembourg
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.
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
Romania
Romania
Poland
Poland
Hungary
Hungary
Steel
Finland Concrete
Finland 1
Timber
2 Spain
Spain Other
>=3
Portugal
Portugal
Italy
Greece
Greece
Germany
Germany
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.
(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.
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.
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.
17
Spain
Portugal
Light industrial activities
Warehousing
Heavy industrial activities
Greece
Germany
Romania 60%
Poland
50%
Hungary
Steel 40%
Finland Concrete
Timber 30%
Spain
Other
20%
Portugal
Italy 10%
Greece 0%
Germany
Steel Concrete Timber Other
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
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
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)
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
20
70%
Romania
60%
Poland
50%
Hungary
Finland 40%
beam
Spain truss
30%
Portugal
20%
Greece
10%
Germany
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
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)
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)
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)
Romania 100%
90%
80%
Spain 70%
60%
MR
50%
Portugal CB
40%
EB
30%
Greece 20%
10%
0%
Germany MR CB EB
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)
40,0%
Romania
35,0%
Hungary
30,0%
Finland
25,0%
<0.05
Spain 20,0%
0.05-0.15
5,0%
Greece
0,0%
Germany <0.05 0.05-0.15 0.15-0.25 >0.25
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)
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
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)
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)
26
Staff
Customer entrance
Sales area
Warehouse and
preparation
Customer services
Staff services
Customers
Car park
Staff
Tradesmen
Hypermarket and
minimarket
Shops
Information offices
Toilets
Restaurant
Warehouse
Customers
Car park
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.
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
Germany Germany
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)
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
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
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.
Spain
Romania
Portugal
Poland
Germany
Finland
Czech Rep.
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
Spain
Romania
Portugal
Poland
<0.05
Italy 0.05-0.15
Hungary 0.15-0.25
>0.25
Greece
Germany
Finland
Czech Rep.
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.
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 %.
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.
1800
1600
1400
1200
mill. eur
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.
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
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.
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
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 - prefabricated
(%): Czech R.
Steel- roof truss w ith any
type of columns (%):
Poland
Steel - portal frames (%):
Romania
Steel - cold-formed steel (%):
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.
0.8 1.2
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
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
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
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
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.
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:
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.
(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 )
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
2
2
15.00 20.00
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
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)
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
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
46
70.00
60.00
30.00
0.16 g
20.00 0.08 g
10.00
0.00
50 kN 100 kN 250 kN
ground
Resistance out-of-plane
Number of frame bays
Base Constraint
Frame distance
Type of girder
(out-of-plane)
acceleration
Maximum
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.
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.
ground
Resistance out-of-plane (1
Base Constraint
Frame distance
Type of girder
acceleration
Profile type
Maximum
of-plane)
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.
(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.
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.
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.
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.
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
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)
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)
e.g.Z150/2
T=6000 T=6000
(a)
(b)
Figure 2.22. Connection to purlins to the frame
57
Kt Ktf
h
Kcf
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
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.
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)
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.
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)
Resistance out-of-plane
Number of frame bays
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
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.
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.
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.
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
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)
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 3.5. Individuation of most pre-stressed wall element in the commercial building plan
67
(b)
(a)
Figure 3.7. (a) HDR-based device; (b) hysteretic device
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.
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.
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.
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.
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)
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
3.3.3 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
(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.
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
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.
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
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
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.
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
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
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).
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.
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.
78
Figure 3.18. Geometry layout for Type 1 (CS1) frame case studies
Figure 3.19. Geometry layout for Type 2 (CS2) frame case studies
79
6000
6000
4000
4000
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
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
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:
Figure 3.24. Planar frames position Figure 3.25. CS1H5 planar frame in X 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
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.
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
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
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
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
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.
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.
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.
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.
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.
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
Connections
longitudinal
joints web
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.
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 4.6. 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 4.9 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
(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
97
Figure 4.13. Load-deflection curves of test 6
98
4483
315 4078
258
B B B
451
Leading and B Side B B Sec. B-B Sec. A-A
contrast system panels
348
250
B B A
Sec. C-C Sec. B-B Sec. A-A
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
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.
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
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
400 kN IDRAULIC
1395
FSHD DISSIPATOR
ATTUATOR
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
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
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.
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]
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]
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/.
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.
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.
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.
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.
111
Figure 5.3. Cost module structure
112
Screen captures and work flow chart of industrial building
113
Figure 5.5. Step 1 of industrial building application
114
Screen captures and work flow chart of commercial building
115
Figure 5.8. Step 4 of commercial building application
116
Statistical module
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.
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
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)
PHASE 3 PHASE 2
NO 2nd order YES
criterion?
AUTO AUTO
3, 4 1, 2 3, 4 1, 2 3, 4 1, 2
Class Class Class
PHASE 4 PHASE 3
Ultimate multiplier
Maximum design
load
END
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.
report-mfr.txt
Figure 5.14. The first view of EV-Frame tool when started, showing the geometry tab
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
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?
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.
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
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.
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.
130
U
L D O
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
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.
132
Figure 6.8: Geometry according to EC 3 [image source: EN 1993-1-5]
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].
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]
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 .
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.
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+1114 (critical region, confined zone lc = 0.15·B = 0.90 m)
As,shear = 18/10cm
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:
Figure 6.14. Position of connective zones between wall and gravity structure
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+1220 (critical region, confined zone lc = 0.15·B = 0.90 m)
As,shear = 110/10cm
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
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+916
As,shear = 18/15cm
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
4500
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)
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.
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)
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.
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).
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.
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.
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.
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 ൌ ܭ௬ Τ݊, where n is the number of devices placed at the
reference floor level.
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
Figure 6.31 . Determination of estimated maximum displacement using intersection of capacity spectrum
with MADRS
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.
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:
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.
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
For sake of simplicity, the same curve obtained for the first floor level was used in correspondence of both
floors.
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.
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:
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.
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
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.
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
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]
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]
2,0E+06 2,0E+06
1,0E+06 1,0E+06
0,0E+00 0,0E+00
-1,0E+06 -1,0E+06
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
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
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)
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
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