Proceeding of 13th Nordic Steel 2015 Construction Conference PDF

Proceedings of
The 13th Nordic Steel Construction Conference (NSCC-2015)

23-25 September 2015, Tampere, Finland
Invited keynotes and extended abstracts
Edited by Markku Heinisuo & Jari Mäkinen
Tampere University of Technology. Department of Civil Engineering

Tampere 2015
i
The 13th Nordic Steel Construction Conference (NSCC-2015)
23-25 September 2015
Scientific committee
Prof. Markku Heinisuo, TUT - Chairman
Assoc. Prof. Jari Mäkinen, TUT - Vice Chairman
Dir. Björn Aasen, Norconsult, Norway
Prof. Michael Joachim Andreassen, Technical University of Denmark
Assoc. Prof. Jean-Marc Battini, KTH
Prof. Timo Björk, LUT
Prof. Ján Bujňák, University of Žilina
Prof. Dan Dubina, University of Timisoara
Prof. Dr.-ing. Markus Feldmann, RWTH Aachen University
Prof.-em. Torsten Höglund, KTH
Prof. Jeppe C Jönsson, Technical University of Denmark
Dr. Olli Kerokoski, TUT
Prof. Reijo Kouhia, TUT
Prof. Ove Lagerqvist, LTU
Prof. Mikko Malaska, TUT
Dr. Kristo Mela, TUT
Prof. David A. Nethercot, Imperial College
Prof. Jari Puttonen, Aalto University
Prof. Luis Simões da Silva, University of Coimbra
Prof. Milan Veljkovic, LTU
Local organizing committee

Assoc. Prof. Jari Mäkinen, TUT - Chairman
Prof. Markku Heinisuo, TUT
Dr. Kristo Mela, TUT
Dir. Janne Tähtikunnas, FCSA
Dir. Markku Leino, FCSA
Dir. Jouko Kouhi, FCSA
Dir. Veikko Numminen, FCSA
Hanna Grönman, FCSA
ISBN 978-952-15-3578-9 (printed)

ISBN 978-952-15-3579-6 (USB)
ii
Preface
The 13th Nordic Steel Construction Conference 2015 (NSCC-2015)
The Nordic Steel Construction Conference (NSCC) is a conference with proud traditions. The
conference was first held in Stockholm in 1970. Since then, the conference has circulated between
the Nordic countries and held approximately every three years. The last conference was held in Oslo
in 2012 and gathered scientists, representatives of steel manufacturers, steel wholesalers, contractors,
consultants, architects, etc. Finland is responsible for the event in 2015.
The last time the conference was organized in Finland was in 2001. The organizers were the Finnish
Constructional Steelwork Association (FCSA) together with Helsinki University of Technology.
Now the baton has once again been passed to Finland. This time the FCSA will be responsible for
the administrative part of the event while Tampere University of Technology (TUT) takes
responsibility of the scientific process. The Scientific Committee was composed of leading steel
professors and researchers from the Nordic countries and the rest of Europe.
For this year's conference we received 123 abstracts from all over the world, which have been
reduced to 96 presentations via the reviewing process. The editors are grateful to the members of the
scientific committee who carried out the reviews of the submitted full-length manuscripts. The
contributions cover most of the important issues for modern steel construction; building structures,
bridges, high strength and stainless steel, structural steel connections, fire, and sustainable
engineering and composite structures.
This conference proceeding contains 89 two-page extended abstracts and seven keynote papers by
invited keynote speakers: Dr.Eng. Björn Aasen, Rutger Gyllenram, Prof. Jean-Pierre Jaspart, Saku
Järvinen, Dr. Jyrki Kesti, Prof. Peter Schaumann, and Prof. Milan Veljkovic. In addition, a selection
of the NSCC-2015 full-length papers have been published in the journal "Steel Construction - Design
and Research", Volume 8 Issue 3. The other full-length papers appear in the accompanying USB
memory stick.
Sincere thanks go to all of the authors and participants for making the Nordic Steel 2015 a
stimulating conference.
Tampere, September 2015
Editors
iii
Table of contents
Keynote lectures
BIM in structural steel workflow
Saku Järvinen .......................................................................................................................1
Making sustainability activities a key to your success - from compliance to commitment
Rutger Gyllenram............................................................................................................... 11
Component method as a general tool for the design of joints under various loading
conditions
Jean-Pierre Jaspart ........................................................................................................... 21
Execution of steel structures - recent developments and future trend
Bjørn Aasen ....................................................................................................................... 31
Use of higher strength steel in construction, opportunities and obstacles
Milan Veljkovic ................................................................................................................... 41
Fire design of steel structures with intumescent coating
Peter Schaumann .............................................................................................................. 51
Energy-efficient solutions for steel structures – case study of nearly zero-energy building
Jyrki Kesti ........................................................................................................................... 61
Plenary Session A
Joint and column behaviour of slotted cold-formed steel studs
Michael Andreassen, Jeppe Jönsson ............................................................................... 71
Steel solutions for enabling zero energy buildings
Bernd Döring, Vitali Reger, Markus Kuhnhenne, Jyrki Kesti, Mark Lawson, Andrea Botti,
Markus Feldmann .............................................................................................................. 73
Plastic resistance of composite slabs in partial shear connection
Leopold Sokol, Anna Palisson .......................................................................................... 75
Future design procedure for structural connections is component based finite element
method
František Wald, Luboš Šabatka, Jaromír Kabeláč, Lukáš Gödrich, Marta Kurejková .... 77
comparative evaluation of steel profiles in roof trusses
Kristo Mela, Hilkka Ronni, Markku Heinisuo ..................................................................... 79
Plenary Session B
Non-linear finite element modelling of steel-concrete-steel members in bending and shear,
Marc Donnadieu, Ludovic Alexandru Fülöp...................................................................... 81
Assessment of existing steel bridge structures
Jan Bujnak ......................................................................................................................... 83
Local buckling behaviour of welded box sections made of high strength steel - comparison
of experiments with EC3 and general method
Nicole Schillo, Markus Feldmann ...................................................................................... 85
Sustainable design of buildings in steel and composite structures
Richard Stroetmann ........................................................................................................... 87
Steel construction excellence center
Jarmo Havula, Pekka Roivio, Markku Heinisuo ............................................................... 89
Session 1A: Building Structures 1
Practical tubular truss optimization
Jussi Jalkanen ................................................................................................................... 91
The impact of joint constraints on the optimal design of truss structures
Roxane Van Mellaert, Geert Lombaert, Mattias Schevenels ........................................... 93
iv
Lateral buckling stress for H-shaped beams with continuous braces
Yoshihiro Kimura, Yuki Yoshino ........................................................................................ 95
Industrial Hall Constructions
Nico Genge, Christian Remde, Klaus Weynand .............................................................. 97
Effect of end stiffener reinforcement on lateral torsional buckling behavior of H-shaped
beams with large depth-thickness ratio
Daiki Kubota, Kikuo Ikarashi ............................................................................................. 99
Session 2A: Bridges & Fatigue
Low cycle fatigue performance of integral bridge steel H-piles under seismic displacement
reveals
Murat Dicleli, Memduh Karalar ........................................................................................ 101
System reliability analysis of steel railway bridge based on historic rolling stock records
Gunnstein Frøseth, Anders Rönnquist ........................................................................... 103
Fatigue problems at riveted railway bridges investigation and rehabilitation methods
Hans Vagn Jensen .......................................................................................................... 105
On actual behaviour of continuous composite girder bridges and their conventional
modelling
Jaroslav Odrobiňák, Ján Bujňák ..................................................................................... 107
New cycle counting method for the assessment of low cycle fatigue in steel H-piles of
integral bridges
Memduh Karalar, Murat Dicleli ........................................................................................ 109
Session 1B: Building Structures 2
Resistance of eccentrically loaded beam-columns
Josef Vican, Peter Janik .................................................................................................. 111
Experiments on plate girders with a very slender web
Roland Abspoel................................................................................................................ 113
Experimental study into bending-shear interaction of rolled I-shaped sections
Rianne Dekker, H.H. Snijder, J. Maljaars ....................................................................... 115
Effect of neutral-axis position on the elastic buckling characteristics of continuous
composite beams
Daigo Shirai, Kikuo Ikarashi ............................................................................................ 117
Amplified seismic loads in steel moment frames
Bulent Akbas .................................................................................................................... 119
Design rules for slim-floor girders considering the composite behavior
Markus Schäfer ................................................................................................................ 121
Session 2B: Bridges
Effect of longitudinal stiffeners on the flanges to improve the low cycle fatigue performance
of steel H-piles
Memduh Karalar, Murat Dicleli ........................................................................................ 123
Refined fatigue assessment of an existing steel bridge
John Leander, Raid Karoumi .......................................................................................... 125
Odins Bridge
Kjeld Thomsen, Trygve Friedrichsen............................................................................... 127
High-performance-steel girder of compact section
Eiki Yamaguchi, Yuji Sugimura, Kenjiro Ohmichi ........................................................... 129
Steel Bridge Technology used in Buildings
Hans Exner ...................................................................................................................... 131
v
Sundsvall Bridge
Kjeld Thomsen, Helge Pedersen, Trygve Friedrichsen ............................................... 133
Session 1C: Building Structures 3
Aluminium deployment in bracing systems: Investigation of shear link application
Evangelos Efthymiou, Vasileios G. Psomiadis, Alexios T. Ampatzis ............................ 135
Design of Wind Turbine Structures based on a multivariate stochastic Approach
Frank Kemper, Markus Feldmann .................................................................................. 137
Time history simulation in seismic design
Peter Knoedel, Thomas Ummenhofer ............................................................................ 139
Steel composite dowels in cracked concrete
Martin Classen, Alexander Stark..................................................................................... 141
Cross-sectional capacity of compocite column by the two methods of en 1994-1-1
Kimmo Ylinen, Wei Lu, Jari Puttonen ............................................................................. 143
Session 2C: Connections 1
Beam-to-column joints subjected to impact loading
Erik L. Grimsmo, Arild H. Clausen, Arne Aalberg, Magnus Langseth ........................... 145
Design resistance of end-plate splices with hollow sections
Yvonne Steige, Klaus Weynand ..................................................................................... 147
Conception, analysis and design of a special joint for fixing lattice towers legs during testing
of transmission line tower
Fabio Paiva, Jorge Henriques, Rui C. Barros ................................................................ 149
Generalized block failure
Jeppe Jönsson ................................................................................................................. 151
FEM simulation of a tubular KT-joint
Karol Bzdawka, Jolanta Baczkiewicz .............................................................................. 153
Session 1D: Cold Formed Structures
Bearing capacity of cold-formed unlipped channels with restrained flanges - EOF and IOF
load cases
Mahen Mahendran, Balasubramaniam Janarthanan, Shanmuganathan Gunalan ...... 155
Elastic buckling of an I-beam with sandwich flanges
Krzysztof Magnucki, Piotr Paczos ................................................................................... 157
A numerical parametric study on the load carrying behaviour under bending of honeycomb
girders made of trapezoidal corrugated steel sheets
Tobias Petersen, Manuel Krahwinkel ............................................................................. 159
Elastic buckling characteristics of corrugated tank under fundamental load
Yoshifumi Yokoyama, Kikuo Ikarashi.............................................................................. 161
Buckling strength of light-gauge members with large openings
Atsushi Sato, Seiji Mori, Tetsuro Ono, Kazunori Fujihashi............................................. 163
Experimental and numerical investigations of the steel storage rack uprights
Zhong Ren, Xianzhong Zhao, Ru Qin............................................................................. 165
Experimental investigation on the behavior of perforated steel storage rack columns under
axial compression
Bassel El Kadi, Guven Kiymaz, Atakan Mangir.............................................................. 167
Session 2D: Connections 2
Monotonic behaviour of bolted T-stubs: a refined theoretical model for flange yirlding and
bolt fracture failure mode
Antonella Francavilla, Massimo Latour, Vincenzo Piluso, Gianvittorio Rizzano ........... 169
vi
Different coating systems for the application in slipresistant connections
N. Stranghöner, Nariman Afzali, J. Berg, M. Schiborr, A. Rudolf, S. Berger ................. 171
Influence of Different Testing Criteria on the Slip Factor of Slip-Resistant Connections
N. Stranghöner, N. Afzali, Jörn Berg, M. Schiborr, F. Bijlaard, N. Gresnigt, P. de Vries, R.
Glienke, A. Ebert .............................................................................................................. 173
Simplified model for connections of steel structures in OpenSees
Ricardo Costa, Filippo Gentili, Luis Simões da Silva ..................................................... 175
Design approach for stability check of members with hanging-profile connections
Dasu Liu ........................................................................................................................... 177
Reasons for Charles de Gaulle airport collapse
Toomas Kaljas ................................................................................................................. 179
Investigations on the behaviour of threaded and shank bolts under combined tension and
shear
Anja Renner, Jörg Lange ................................................................................................ 181
Session 1E: Composite Structures
Behavior improvement of pultruded frp beam-column bolted connections
Ossama El Hosseiny, Hassan Maaly, Saeed Ibrahim ................................................... 183
Material strength effect on the behaviour of steel-concrete composite elements
Janis Brauns .................................................................................................................... 185
Vibration response of USFB composite floors
Richard Kansinally, Konstantinos Tsavdaridis................................................................ 187
Analyses of the load bearing behaviour of deep-embedded concrete dowels, CoSFB
Matthias Braun, Renata Obiala, Christoph Odenbreit .................................................... 189
Session 2E: Fire Engineering & Building Structures
Evaluation of axial force impact on the flexibility of a steel beam-to-beam end-plate joint
subjected to bending when exposed to fire
Mariusz Maslak, Malgorzata Snela ................................................................................. 191
Fire design of CFST columns - Improvements required for Eurocode 4
Matti V. Leskela ............................................................................................................... 193
Calculation of steel temperature in open cross sections based on fire exposure from CFD
Joakim Sandström, Wickström Ulf .................................................................................. 195
Lateral torsional buckling resistance – a comparison of analytical and numerical models
Rebekka Ebel, Markus Knobloch .................................................................................... 197
Innovative Construction of Student Residences
Pedro Andrade, Milan Veljkovic, John Lundholm, Tim Heistermann ............................ 199
Session 1F: Sustainable Engineering
Fatigue life improvement of welded bridge details using high frequency mechanical impact
(HFMI) treatment
Poja Shams Hakimi, A. Mosiello, K. Kostakakis, M. Al-Emrani ..................................... 201
New developments in heavy plate production for modern steel construction
Tobias Dr. Lehnert, Falko Dr. Schröter ........................................................................... 203
Stainless steel, a sustainable material for sustainable structures
Anders Finnås, Camilla Kaplin ........................................................................................ 205
Dynamic responce of pipe rack steel structures subjected to explosion loads
Anton Stade Aarønæs, Hanna Nilsson, Nicolas Neumann ........................................... 207
Tall ambitions – onshore wind turbine tower - concepts for large hub heights
Martin Jespersen, Mogens Nielsen, Ulrik Stottrup-Andersen ........................................ 209
vii
Session 2F: Connections 3
Lateral stability of verandas by means of the glass panels
Maarten Fortan, Jesse De Clercq, Marc Meeus, Barbara Rossi ................................... 211
End Plate Connection for Rectangular Hollow Section in Bending
Arne Aalberg, Arne Martin Uhre, Per Kristian Larsen .................................................... 213
Structural Behaviour of a novel column-splice joint, Finger Connection
Pedro Andrade, M. Pavlović, C. Heistermann, M. Veljkovic, T. Heistermann ............... 215
Structural analysis models of steel trusses
Teemu Tiainen, Kristo Mela, Timo Jokinen, Markku Heinisuo....................................... 217
Buckling of members of welded tubular truss
Markku Heinisuo, Äli Haakana ........................................................................................ 219
Session 1G: High Strength Steel
Bendability and microstructure of OPTIM® 700 MC plus
Vili Matias Kesti, Antti Juhani Kaijalainen, Juho Mourujärvi, Raimo Ruoppa ................ 221
Experimental behaviour of tension plates with centre hole made from high strenght steel
Pál Turán, László Horváth ............................................................................................... 223
Derivation of strain requirements for high strength steel using Johnson Cook model
Simon Schaffrath, Nicole Schillo, Markus Feldmann ..................................................... 225
Buckling strength of HSS beams
Mark Andrew Bradford..................................................................................................... 227
True stress-strain relationship for finite element simulations of structural details under
diffuse necking
Petr Hradil, Asko Talja ..................................................................................................... 229
Calibration of the ductile damage material model parameters for a high strength steel
Marko Pavlovic, Panagiotis Manoleas, Milan Veljkovic, Efthymios Koltsakis................ 231
Buckling observation of door openings for wind turbine towers
Anh Tuan Tran, Milan Veljkovic, Carlos Rebelo, Luis Simões da Silva......................... 233
Session 2G: Stainless Steel & Connections
Extension of the continuous strength method to the determination of shear resistance
Najib George Saliba, Leroy Gardner............................................................................... 235
Stainless steel at slightly elevated temperatures
Hans L. Groth, Erik Schedin, Emma Jacobsen, Rita Lindström .................................... 237
New steel damper with displacement dependent recentering for seismic protection of
structures
Murat Dicleli, Ali Salem Milani ......................................................................................... 239
Fretting fatigue phenomenon in bolted high-strength steel plate connections
Olli-Pekka S. Hämäläinen, Timo J. Björk ........................................................................ 241
Comparison of relative volumes of different type of welds
Juha Kukkonen, Markku Heinisuo .................................................................................. 243
Investigation of cold formed steel beam to column bolted gusset plate connections
Žilvinas Bučmys, Alfonsas Daniūnas .............................................................................. 245
Resistance results for the crocodile connection
Panagiotis Manoleas, Kristoffer Öhman, Efthymios Koltsakis, Milan Veljkovic............. 247
viii
Nordic Steel Construction Conference 2015
Tampere, Finland
BIM IN STRUCTURAL STEEL WORKFLOW
Saku Järvinen
Tekla Oy
saku.jarvinen@tekla.com
+358 40 707 1622
Abstract: This paper will go through how BIM has evolved in the structural steel workflow
over the years. The main focus is on the recent developments especially in steel fabrication and
the use of IFC (Industry Foundation Classes) for fabrication. The future directions of BIM in
structural steel workflow are addressed at the end.
1. BIM in short
With BIM (Building Information Modeling) technology, one or more accurate virtual models
of a building are constructed digitally. They support design through its phases, allowing better
analysis and control than manual processes. When completed, these computer-generated mod-
els contain precise geometry and data needed to support the construction, fabrication, and pro-
curement activities through which the building is realized. (BIM Handbook; Eastman, Teicholz,
Sacks & Liston 2011)
BIM enables automation of the use of information – CAD (Computer Aided Design) already
automated the creation of information. From software, BIM asks for accuracy and a capability
to handle lots of information, and in practice also compatibility with other solutions as other-
wise achieving a collaborative workflow would be challenging at best.
BIM currently helps the construction industry around the world to achieve better workflows
and thus bottom lines. Using BIM allows making informed decisions early in the design and
construction process. Accurate, constructible models let designers try out solutions before
building. The builders and fabricators can better manage the risk of unforeseen costs and loss
of time.
1.1. The challenges of the construction industry
Worldwide vertical construction is a $3.5 trillion industry (excluding residential construction).

At least 20% of this is wasted:
 ~10% materials are wasted
 ~30% of construction is rework
1
 ~40% of jobsite work is unproductive
 ~40% of projects are over budget
 ~90% of projects are late
Ineffective communication, planning and collaboration are the sources of most problems.
Figure 1. Source: US Dept of Labor: Bureau of Labor Statistics (BLS). Production Price Index – PPI, Consumer
Price Index – CPI.
1.2. Workflow compatibility

Data compatibility is crucial to succeeding in modern construction. A few noteworthy players
in the field have made a strategic decision to support, develop and promote the Open BIM
concept, which greatly benefits the whole industry.
Construction project parties should be able to work together

smoothly regardless the tools they use. Open BIM is about
workflow-level compatibility, not just compatibility be-
tween two software packages. Also manufacturing calls for
open, readable data.
It all comes down to enabling an end-result-friendly, build-

ing-quality-assuring, standardized cooperation process: get-
ting all project parties to work together by using Open BIM.
Development of the IFC standard has brought steel fabrica-
tion into the Open BIM workflow, which is the way the
Figure 2. "Open BIM supports a trans- whole industry needs to go in pursuit of balanced and sus-
parent, open workflow, allowing pro- tainable development.
ject members to participate regardless
of the software tools they use."
2
2. Solving the initial problem – Automating drawing production and CNC
steel processing
The use and capabilities of 3D modelling evolved rapidly throughout the 1990s and into the
2000s as the technology became more affordable and desktop computers grew more powerful.
The software used within the steel industry was developed further so that the initial goal, to
produce drawings and reports from the 3D model automatically, was resolved.
Some structural steel fabricators were quick to realize these early benefits of BIM. They were
pushing the software vendors to develop an interface with CNC machines. At the same time
when the initial postprocessors were developed to feed the part geometry into the CNC ma-
chines the first standards (SDNF, DSTV-NC, KISS, CIS/2, etc.) were developing. Since the
late 1990s fabricators have created highly detailed and intelligent 3D models to produce shop
drawings, generate reports and extract CNC data to drive machines in the shop, aiming at elim-
inating common errors.
The 3D technology had evolved to resolve the automatic steel processing of individual parts
with CNC interfaces. The DSTV (Deutscher Stahlbau-Verband) NC standard is still the most
commonly used standard to automatically process single steel parts with CNC equipment.
The industry is now requiring more than the current standards were designed to provide.The
DSTV-NC based fabrication is a one-way information export channel from the information rich
3D models and leaves the steel fabrication outside the BIM workflow. Even if IFC models were
used by the project design phase participants to exchange information, fabrication was still re-
ceiving drawings and thousands of individual files.
Challenges of the current process:

 For MIS/Production planning, multiple files with multiple settings are required from
detailed model
 For every individual part (beam, clip angle, stiffener, etc.) NC files are required, which
results in hundreds or even thousands of files to manage
 NC files do not contain revision control mechanisms and must be managed carefully to
ensure that the latest file is used
 NC settings in detailing software must be adjusted according to the specific fabricator
set up, equipment type and process
 Status feedback from the shop floor/equipment to MIS and detailing systems is rare
3
Figure 3. AISC BIMsteel, Interoperability Initiatives for the Structural Steel Industry
3. Moving on – Assembly life cycle management

While the usage of BIM was growing in the early 2000s, a few innovative fabricators started to
envision BIM models being able to resolve additional challenges in the fabrication process. The
assembly phase had become the bottleneck in the process. It was a very laborious, manual and
costly phase and it began to be difficult to find skillful experienced personnel to manage the
fitting of assemblies.
The visionary fabricators understood that the CNC machines could be harnessed to do the lay-
out marking based on the information already available in the models. The problem was that
there was no interface for retrieving the information from the model. They worked together
with the leading suppliers of the CNC equipment suppliers to build proprietary interfaces for
getting the hierarchic assembly geometry with attributes from the Building Information Models.
These innovative tools could read the model or assembly data and literally draw the information
the fitter needed onto the beam or piece. This approach became quickly very common among
the CNC machine vendors, but they all had to develop their own interface and applications to
access the 3D models.
Figure 4. FICEP scribing technology for layout marking
4
Since the introduction of the automatic layout marking technology in 2004, the industry started
to envision more widely how the Building Information Model could be used throughout the
process to increase productivity. Similar applications were also developed to replace the report-
based interfacing with ERP systems, improving information flow to production.
3.1. Automation in steel processing and assembly phase

The constant pressure for higher productivity within the steel industry was driving the supplier
market to seek new innovations. At the same time, with the emergence of BIM, demand was
pushing the equipment vendors to innovate beyond the traditional, standalone CNC equipment
offering. The market was now demanding higher productivity and automation of the assembly
phase.
Automation of material handling during steel processing was one of the areas where major
productivity gains were achievable. The automation (software and equipment) secures material
availability to the CNC machines at all times, reducing the downtime previously faced in the
manual infeed and outfeed of the steel processed.
Robotic welding is becoming more common as cost of investment comes down and capabilities
increase. Welding robots need to be able to analyze weld paths, understand weld size and type
and know where two pieces connect. With robotic welding, the industry is entering a field that
is not as straight forward - from model to equipment - as the basic CNC steel processing where
the geometry in the model simply is replicated as the physical part by the CNC machines. Weld-
ing is a creature of a different kind. Welding is not always defined in the detailed model to a
level that meets the needs of the shop. The shop might also want to optimize the welds based
on various requirements and procedures.
So many different types of steel structures are fabricated that it is impossible to say if robotic
welding would be the most cost-efficient way to make all types and sizes of assemblies. The
implementation of robotic welding requires extensive research on the type of steel work. It
should not be considered as an out of the box solution that is suitable for all structural steel
applications by default.
In some cases the use of automatic layout marking, manual tack welding and using a robot only
for final welding can be most efficient. In other cases completely automatic robotic assembly
and welding lines have been found productive. Material tolerances also play a role when in-
volving robotics.
Figure 5. PEMA robotic beam welding
5
Automatic robotic assembly machines are already on the market. These machines can manage
the whole assembly fitting and welding process. They recognize parts, then pick, place and
weld them in the correct position within the assembly.
Figure 6. Zeman, automatic robotic steel beam assembler
Advanced steel fabrication specific software (MIS, PLM, etc.) is taking a bigger role in pro-
duction planning and management. Based on the Building information in the models these soft-
ware programs organize and simulate work on the shop floor to resolve bottlenecks; allowing
the fabricator to optimize the schedule before fabrication begins. The leading industry around
the world has moved to fully automated steel processing. They plan, schedule, and run simula-
tions of their work processes based on information coming from a Building Information Model
before work begins on the shop floor.
Figure 7. FICEP automatic steel fabrication and software
The project management tools allow fabricators to schedule, record (automatically from the
CNC machines or manually via bar code or RFID), and visualize a project’s progress, which
assemblies are on the shop floor and which are already on their way to the job site.
All these new innovations and technologies are creating various requirements for information
exchange. Some of the information required did not exist in the models at the needed level.
Welds and weld preparations were some of the major development areas within BIM software.
At the same time the proprietary interface applications were further developed to meet the needs
of the new innovations, and transfer the information between the model and the workshop.
6
3.2. Model usage expands
The role of the model had grown in the assembly phase. The next step was to start using the
model to resolve problems with logistics and scheduling. Forward thinking fabricators started
to include erection schedules in the model. The schedule on an assembly level was added to the
model objects and this schedule was then serving as the basis for production schedule planning.
The production offices found it more logical to use the visual 3D model to plan the logistics
down to the level of individual truck loads before passing the information on to MIS/PLM/ERP
systems for the actual production planning.
Figure 8. Example of the planned truck loads shown in different colours. The document is also delivered to the
site with the assemblies.
Since the amount of valuable information in the model had been increasing and the status in-
formation was automatically exchanged with ERP and MIS software, it became obvious that
the mobile usage of the model would help understanding the project status and deliverables.
Changes have always played a major role in building projects. As the models were already
cumulated with object level attribute information, also communication of past, current and fu-
ture changes was logical to store in the models. This innovative use of new technology lead to
the developing tools for also managing and communicating the changes with various stakehold-
ers and provided a much better understanding faster of the possible implications of changes,
even with the work ongoing at the shop floor or site.
All these activities were impossible with the existing industry standards. There was demand for
more model information and also the ability to return the information. Development of propri-
etary applications was the only solution for a wide range of vendors to enable accessing the
model and fulfil their information needs. This way they had a bidirectional interface to the
model with their application (SW or CNC).
The challenge that remained was the fact that all these technological innovations were not
openly communicating with the various tools on the market. They were relying on a limited
selection of tools that had only one to one, point to point exchange of the information.
7
4. Steel fabrication joins the BIM workflow
While the industry was and still is trying to solve the building information needs with proprie-
tary applications, it became clear to the leading vendors that the time had come to develop a
new standard that fulfils the broader needs of this millennium.
To drive the steel fabrication industry towards higher degree of automation and to take new
technologies more widely into use, the leading vendors decided in 2009 to use their knowledge
gained from their development works to define and create a new standard that would meet the
current and coming demands of the steel fabrication industry.
As IFC had already been developed for many years to meet the needs of the design community
for exchanging information, enhancing IFC to meet the needs of the steel industry was set as
the goal. This work included all fabrication-level information — not just coordination-level
information — allowing the steel fabricator’s model to become meaningful part of the entire
project workflow while maintaining the detailed information required specifically for fabrica-
tion.
IFC has become the industry-standard neutral file format for exchanging Building Information
Models between construction disciplines from structural steel to concrete, mechanical, electri-
cal, plumbing, fire protection, and more. Without IFC, even the relatively slow adoption of
model-based coordination in the construction industry would not have happened as it now has.
The commitment of the software vendors is further strengthened by a compatibility certification
process established by buildingSMART International. AISC (American Institute of Steel Con-
struction) is facilitating development of IFC for structural steel as part of its BIMsteel initiative.
The IFC based process:

• A single IFC file can be released from a detailed model
• MIS/Production Planning can extract the data they need from the IFC file and manage
it
• Machinery and other equipment can extract the data they need from the IFC file and
manage it
• Feedback, real-time or not, can be sent back to the detailed model using IFC
8
Figure 9. AISC BIMsteel, Interoperability Initiatives for the Structural Steel Industry
5. Future insight
While the development of the IFC fabrication view is still ongoing, the future directions are
already under research. IFC will resolve the transfer of the digital Building Information Model
and data, but one main obstacle remains. File-based information transfer is not supporting the
whole delivery chain in using the model information effectively. This means that IFC as such
does not provide an answer to how models couldn be used to make projects’ internal processes
and workflows more effective.
The productivity of each design discipline ha significantly improved. However similar

improvements hasnot happened to those different disciplines that use the building information,
which is why BIM's larger productivity potential has not yet been realized.
The way the building projects are organized is changing as the tools develop. Decentralized
project organisation on a global level is common nowadays. The software industry is constantly
developing tools to support the scattered building project organisations that are distributed
globally. New web technologies need to be adapted in order to support the construction industry
workflows.
5.1. Web technology enables more realistic distributed building information manage-
ment
Recent studies suggest a significant turn in the way BIM is considered to make the building and
construction industry work in a more efficient and automated way. Building information
modeling and management are no longer considered a process of centralizing all building-
related data to a single model, but rather a distributed, linked network of models that various
disciplines involved in the construction project publish for various purposes. This trend was
observed by DRUM, a 3-year joint research work package conducted by Finnish and US
universities, construction software providers, and construction companies.
BIM technologies usually appear to us as visual 3D models within the modeling tools, including
software products like Trimble SketchUp, ArchiCAD, Tekla Structures, or Autodesk Revit.
There are BIM tools for various disciplines within design and engineering, and using the models
created with these tools has become a norm for challenging construction projects.
9
5.2. Web technology replaces IFC, workflow scheme remains
With IFC, the whole model is opened and transferred within a project, which is a heavy process.
With web formats such as RDF, we can search for and retrieve single objects, like beams,
combine their data, and easily utilize it elsewhere. The data and content are in a format that the
designers and builders can understand without programming skills.
Linking through web technologies enables instance and object-level compatibility so that, for
example, the architects can ensure that the wall they designed matches the wall that the
structural designer created. They do not need to see the whole project model for that purpose.
It also allowsfollowing a single element's journey from fabrication to construction site in real
time, just like with a package tracking service on the web. Data is retrieved from different linked
systems but is available in one place.
Results from DRUM studies in Finland support the idea of keeping data records saved at
construction project parties' own computers and organizing the links between them using web
technologies as routers. Up-to-date building data stays distributed among its publishers and no
other parties are able to touch their formats, only to retrieve relevant information for their own
purposes.
10
Tampere, Finland
Making Sustainability Activities a Key to Your Success

From Compliance to Commitment
Rutger Gyllenram
Kobolde & Partners AB, / Swedish Steel Building Institute, SBI, SWEDEN
Abstract: The demands on companies regarding environmental and social performance is in-
creasing and numerous initiatives have been taken by the European Commission to decrease
the environmental impact and improve the resource efficiency and social performance in Eu-
rope. In order to turn these demands into an opportunity to benefit from the sustainability
work it is necessary to make it a central part of the business idea and allow it to penetrate eve-
ry part of the organisation. Communication should be based on real achievements and green-
wash avoided. Benefits may come from cost savings, increased product value, strengthened
value chains both upstream and downstream, co-worker loyalty and improved public relations.
1 Introduction
The general view of what is normal and acceptable changes over time. Half a century ago it
was considered normal for each country house to have its own dump in the backyard or in the
adjacent forest. It was also considered environmentally friendly to bring stones aboard a lei-
sure sailing boat to ensure that the plastic bags with household waste sunk properly and did
not stay floating in the reed. Smoke from the factory stacks was a sign of progress and dilution
was the solution.
After imposing laws on municipal garbage collection and a couple of decades with focus on
emission control in production facilities the attitude has changed completely. The consumer
behaviour could however still be summarised as “buy, wear and throw away” with huge city
dumps as a result. A number of directives from the EU stating producer responsibility for re-
cycling of electronic goods, vehicles and packaging paved the way for organized recycling and
here also the attitudes have changed and people have learned to separate metals from paper
and plastics before throwing it away.
With development of Life Cycle Assessment, LCA, the impact of a product on the environ-
ment during its entire life cycle from “cradle to grave” could be studied. A variety of methods
were developed with the consequence that results from different studies could not be com-
pared. In the standard ISO 14044 LCA methodology was standardized and a number of issues
dealt with but the LCA practitioners still had freedom to make studies in different ways mak-
ing them difficult to compare.
11
2 Nordic Steel Construction Conference 2015
In 2004 the European Commission, EC, gave the European Standardisation Organisation,
CEN, the mandate to develop a standard for environmental impacts from the production of
building products. The main objective was to support the EU inner market and to create simi-
lar rules for building products throughout Europe. This was the start of the technical commit-
tee CEN/TC 350 “Sustainability of Construction Works” [1] and development of a set of
standards for the building sector. The committee decided to extend the scope to work with all
aspects of sustainability on the building level; environmental, social and economic. The first
versions of these standards are now on the market. The scope has also been extended to in-
clude civil engineering works and this standard will probably be ready by 2018.
Some of the tools outlined in the standards will probably be necessary for companies to un-
derstand and use. This will be in order to comply with regulations or qualify for different
kinds of ratings or public procurement schemes. It could also be done to position their prod-
ucts against less environmentally conscious alternatives or as an answer to market demands
for information. This work will cost money.
In this plenary paper the question of how tomorrow’s successful steel building companies can
benefit from their sustainability work is discussed. It is full of references to standards and ab-
breviations so a look at the lists at the end of the paper might help the reader. It is also im-
portant to point out that the paper reflects my personal view of the development within EU
and personal experiences from work within European Standardisation. This has affected my
choice of examples and there are of course other views.
2 LCA as a Sustainability Tool in today’s Standards

2.1 The life cycle
The life cycle of a building consists of several stages shown in fig. 1; production of building
products, construction of the building, use of the building, and taking care of the building at
the end of life. After the life cycle, benefits or loads that emanate from this life cycle but occur
in another life cycle can be noted. An example may be a steel beam where the production is
reported in module A for one building can be reused in another building. The principle used in
the CEN/TC 350 work is that information from the different modules should be reported sepa-
rately and not summarised to a total.
An assessment can be of different types depending on its use. For a building material it must
include module A1-A3, “cradle to gate”, but can as an example include ABC, “cradle to
grave”, or ABCD which does not have a name in the standard but might be called cradle to
cradle.
When making an LCA for a building not yet produced the assessment is made based on sce-
narios. Examples of scenarios are recurring maintenance, replacement of windows and refur-
bishment of the entire building. The standard includes necessary rules for service life, func-
tional unit that is assessed etc but the developer has a rather big freedom to develop scenarios
and take different factors into account as will be discussed later.
Actual data for environmental impacts from the production process is preferred but database
data can be used for upstream and downstream processes and a number of commercial data-
bases exist.
12
Nordic Steel Construction Conference 2015 3
Fig. 1. Modularity for Building Assessment Information used by CEN/TC 350.
There is no limit of the number of environmental impacts that can be taken into account ac-
cording to CEN/TC 350 standards. These are however mandatory:
 Global warming
 Ozone depletion
 Acidification for soil and water
 Eutrophication (over fertilization of aquatic systems)
 Photochemical ozone creation
 Depletion of abiotic resources-elements
 Depletion of abiotic resources-fossil fuels.
Furthermore, data on resource use should be given:

 Use of renewable and non-renewable primary energy used for raw material or energy
 Use of secondary material
 Use of renewable and non-renewable secondary fuels
 Net use of fresh water.
2.2 Environmental Product Declarations (EPDs) and Product Category Rules (PCRs)
Any company can make an LCA for their products according to ISO 14044. After having it
reviewed by a third party for compliance with the standard according to ISO 14025 it can be
used in market communication and called an Environmental Product Declaration, EPD. The
freedom given by this standard has led to the practice of developing rules on how to apply ISO
14044 for different product categories in order to produce an EPD.
These rules, Product Category Rules or PCR, are administered by certification schemes like
“The International EPD System” often called “Environdec” managed by IVL in Sweden, Das
Institut Bauen und Umwelt, IBU, in Germany and many more. The different organisations are
now working together in the Eco-Platform to make their PCRs coherent so that an EPD based
on a PCR in one system is accepted in the other systems. [2]
13
The CEN/TC 350 standard EN 15804 could be seen as a “super PCR” for building materials
that the certification schemes follow when developing their PCRs. The different building ma-
terials like concrete, timber and aluminium all have standards developed within the CEN sys-
tem based on EN 15804 to serve as super PCRs for their respective material. In an initiative
from the Nordic Council of Ministers a PCR for constructional steel has been developed but it
has no official status.
2.3 The Battle of Materials

Steel, timber and concrete compete as materials for building frames. In standardisation work
this becomes obvious as the three industries have completely different interests. Steel together
with other metal industries emphasize the benefits of recyclability and as a consequence sup-
port the use of Module D and would prefer to have it mandatory. Industries like for example
concrete and mineral wool, where the value of recycling is low, would rather be without mod-
ule D.
Claiming the “polluters pay principle” concrete has advocated waste status for inputs like fly
ash, wood chips or used tires. They have also claimed allocation principles for a co-product
like blast furnace slag that are favourable for them. Another important issue for the concrete
industry is carbonisation, the CO2 uptake in concrete. By this way of counting it is possible to
claim that concrete can be produced and used with very small emissions of Green House Gas-
es, GHG. This will however, in the future, depend on how end of waste criteria are interpret-
ed, on how the steel industry treats slag and on scientific results for carbonatisation.
The most important factor for the forest industry seems to have been the CO2 storage. If mod-
ule D is not taken into account the GHG-emissions from wood is negative which is reasonable
if it is stored in eternity after its use. The normal recycling principle for wood is however in-
cineration where the CO2 is again emitted.
2.4. Social and Economic Factors

The standard for social methods today covers only the use stage and handles soft values like
apprehension of safety, accessibility and indoor climate. The standard for economic methods
covers the calculation of Life Cycle Cost, LCC.
Even if the CEN/TC 350 standards cover all three aspects of sustainability the standards are
not totally coherent and we will have to wait until after the first revision of the standards be-
fore this can be achieved. In the standard for civil engineering under way the three are treated
in the same standard which may facilitate a better integration.
3 Trends in Sustainability Demands and its Impact on Standards

3.1 Development Work in Standardization
The most important standard in environmental work is probably the ISO 14000-series for en-
vironmental management systems. In the imminent next revision of the standard the life cycle
thinking is emphasized and we might see LCA as a tool for continuous improvement of a
company’s products [3]. This is probably a very good application for LCA but will increase
the demands on LCA skills and awareness in the entire organisation.
14
The EC puts strong pressure on CEN/TC 350 to add a number of impact categories to the en-
vironmental standards. At present a technical report is being developed covering methods for:
 Human toxicity (cancer and non-cancer effects)
 Eco toxicity (terrestrial, freshwater and marine)
 Particulate matter formation, ionising radiation (human health and ecosystem health)
 Land use (occupation and transformation)
 Biodiversity and water scarcity.
Criticism against introduction of these impact categories are for example that methods for tox-
icity lack assessed data for metals and that they are not suited for LCA since they are based on
risk and not impact and for land use that it is local to its nature etc. What will happen in the
field of additional indicators will be decided in the near future.
In order to make the standards coherent another technical report is under development investi-
gating methods for social aspects on the modules A, C and D. The work is still in an early
stage but responsible sourcing of raw materials, noise and disturbance of ordinary life might
be topics that will be covered.
3.2 EC - PEF, Circular Economy and Construction Demolition Waste

As stated before the EC is very active in the sustainability field. The inner market and removal
of environmentally motivated trade barriers is a major goal as well as is boosting the sustaina-
bility work as such. The number of initiatives is impressive. Sometimes it seems that the
commission is moving a little too fast and needs to slow down. Not all initiatives succeed but
the direction is stable and the signal sent to governments, companies and people in general is
clear: we must decrease our environmental footprint.
The Product Environmental Footprint, PEF, aims at an EPD system for consumer products
within the EU [4]. It is developed by the EU Joint Research Centre, JRC in Ispra and the in-
dustry participates in different pilot projects. The metals industries have for example contrib-
uted with the development of a PCR for metal sheet. A result that can be expected from PEF
is that it becomes evident that LCA is a necessary tool for companies to use in order to show
the environmental properties of their products in the future.
A recent initiative covering the cyclic economy is just starting up. Where it will end is too ear-
ly to say but one might expect that renewable materials and material reuse and recycling with
very small losses in mass and function will be important. The Construction Demolition Waste,
CDW, work is more substantial [6]. So far, all member countries have reviewed amounts of
CDW and the recovery rate and published these in a set of reports. It is reasonable to believe
that this interest will strengthen the position of module D and eventually make it mandatory in
all EPDs. Another initiative that will affect the industry is the “Efficient Buildings” study to
develop a common EU framework of indicators to assess the environmental performance of
buildings.[7]
3.3 Greenwash Awareness

A recent trend is the public awareness of unfounded claims of sustainability from companies.
An example is “The Swedish Greenwash Price” funded by the Swedish government but
awarded by an NGO. An example of an award is the Swedish-Finnish forestry company Stora
Enso that “won” the price in 2012 after publicity concerning child labour and threatening bio-
15
diversity by turning rain forests into Eucalyptus farms which contradicted claims of being a
sustainable company. The Chief Executive Officer, CEO, now declares that the company will
change and from now on intend to be a good example to others [8]. Similar examples exist
from other industries.
Another example of claims that add ridicule to the issuer can be fetched from England. Ru-
mours have it that there is an office building in London that was not possible to reach by bicy-
cle but had bicycle racks since it gave “inexpensive” points in the rating system for sustaina-
bility that was used.
With increased knowledge about LCA and with EPDs as a common way to communicate, we
might see a much more agile consumer market in the future. A market that scrutinizes the
claims and questions assumptions in the search for greenwash in EPDs will lead to a more
careful use by issuing companies.
3.4 Towards Sustainability Product Declarations, SPDs

A final observation is that social issues today are mentioned together with environmental im-
pacts like in the Stora Enso case above. With standards that cover environmental impacts to-
gether with social and economic issues, it is not farfetched to assume that we in the future will
have something called Sustainability Product Declaration, SPD, as an alternative to EPD.
4 Success in the Steel Building Business

4.1 Sustainability Work in a Steel Construction Company
Although we do not know how the interpretation of Life Cycle thinking in the revised stand-
ard for environmental management, ISO 14001, will be, we can assume that it will have much
in common with the development of EPDs in most companies. One possibility is that the EPD
forms a base line that can be compared with the actual values obtained in the LCA from the
environmental management system.
Working with LCA methodology in continuous improvement will certainly take corporate
sustainability work to a new level but also raise some questions. One which is important for
steel construction is how to deal with variations in the ratio between steel from scrap and steel
from ore since replacement of efficiently produced virgin steel with inefficiently produced
steel from scrap would show a decrease in GHG emissions but would not represent an im-
provement in a wider sense.
Examples of areas for continuous improvement for a steel construction company are:
1. Sourcing of raw materials from a supply chain striving for best practice reported in
module A1-A3.
2. Transport to building site reported in module A4
3. Yield in raw material use reported in module A5
4. Yield in raw material recycling from the construction site including the quality of ma-
terial to recycling reported in module A5.
The first point requires data that to a large extent comes from steel producers which means
that they need EPDs for their products. An interesting question is how different steel qualities
should be treated since they have different environmental impacts but are reported together in
16
a common EPD. Examples are steels that are covered by the same EPD but differ in alloy con-
tent, working process or heat treatment resulting in different technical properties and different
environmental impacts.
4.2 Steel Constructions Competing with Other Construction Types

Steel has a number of properties that makes it different from other materials and part of the
success is to identify and communicate them with good arguments. The following three points
are examples of claims that a steel construction company can make when steel competes with
other materials.
 Construction and deconstruction with steel is fast and silent. The fact that steel is
relatively light in weight means less transport which should be noted in module B4.
Steel sections can be produced elsewhere and transported to the construction site. This
means that construction and deconstruction can be done with a low noise level and
with little disturbance to ordinary life which is an important social factor.
 Steel constructions are inorganic and not hygroscopic. This means that steel does
not rot and takes little time to dry up which are important properties in case of water
leakages and possible consequences of a more humid climate. In module B3 repair, B4
replacement and B5 refurbishment this information could be included in the LCA sce-
narios. It is also possible that it affects the service life time. These properties could al-
so be reflected in the social impact “indoor climate” especially considering airtight
houses.
 Steel is 100% recyclable to the same quality or better. This claim will probably
meet more understanding in the future according to the trends outlined above. It should
be noted in module D.
Making claims like this must be backed up by hard facts. It can be measurements made by the
company or results from research on these issues where the companies take part. The ISO
14000-standards ask for improvements. Is it possible to make the building process smoother
with less noise? Is it possible to further increase the buildings’ robustness against water by
changing accompanying materials, methods or design? And finally, is it possible to improve
the recycling rate and decrease quality and material losses in steel recycling?
The three characteristics above have that in common that they are typical for steel and won’t
be mentioned by companies working with competing materials. It is therefore an important
task to communicate the advantages in a way that customers or end users will start asking all
suppliers on the market for the information.
4.3 Benefits and Success Factors

The benefits from successful work improving environmental and social performance may be
manifold as is shown in the three examples below that result in improved revenues and/or re-
duced costs.
 Strengthening steel as an attractive construction material by setting the agenda:
Some of the strong arguments for steel presented above are not really effective since
they are not reported by other materials or requested by customers. An active role in
sustainability development gives the opportunity to set the agenda and create a demand
for information about for example disturbance, noise, robustness and recycling.
17
 Strengthened relations with employees, suppliers and existing and potential cus-
tomers: A company that take leadership in sustainability development becomes a
more attractive employer or business partner. The benefits come from the fact that it is
easier to attract skilled people and keep them, and commercial contacts can start from
existing relationships based on cooperation and trust which reduces communication
problems.
 Improved resource efficiency: Improving yields in production, lean design etc may
lead to decreased material use. This is perhaps the benefit that is easiest to measure.
It is possible to make a long list of factors that supports success and some examples are sug-
gested below.
 Credibility based on leadership: Credibility is perhaps the most important asset in a
change process where many people are involved. Commitment and leadership in creat-
ing supply chains and solutions with the lowest possible environmental impacts and
good social performance forms the basis for credibility and trust.
 Participation in-house, upstream and downstream: The entire organisation must be
involved and committed in the work together with all parts of the supply chain to the
end user.
 Efficiency in reporting systems: Collecting data for LCA and EPD work is time con-
suming and costly. However, most of the data can be retrieved from internal systems
 Continuous improvement: Change takes time. Especially if many people are in-
volved. A trick is to start simple and then improve and extend the scope and let the
work mature slowly. The most important is to start as soon as possible to take the ini-
tiative.
 Efficient communication channels: Finally, internal and external communication
channels must be developed to communicate sustainability information in a way that is
possible to understand for all concerned.
5 Discussion
The days are gone when people saw earth as something that could be consumed. Today we are
entering an era where companies are required to take responsibility for their products from
raw material to recycled material. There are also signs that we are about to leave a period
where companies can claim superiority in the field of sustainability by just applying immature
metrics on what they already do. We have to assume a common mind set where people expect
companies to have full knowledge of their supply chain and their environmental and social
performance and to take measures to continuously improve the situation.
Sustainability work is time consuming and costly even for very big companies. For smaller
companies it is probably only possible to take leadership if the work is developed within exist-
ing business processes. It must be handled in the same way as other management systems like
quality and working environment/safety. Having a goal or vision is crucial, skills are neces-
sary and starting a must. In my mind there is no success possible in sheer compliance. Com-
mitment and leadership is necessary in order to benefit from your sustainability work.
Some information needed for module B, C and D discussed above must be investigated fur-
ther and collected for use in LCA scenarios. This opens for joint research programs within the
steel construction community and in the entire construction field.
18
This paper has referred to standards and they reflect in many ways the state of the art of appli-
cation of methods and affect how these methods are used in the future. It is therefore im-
portant to understand the political dimension of standardisation. Standards are normally de-
veloped by scientists with strong influence from stakeholders who advocate interpretations of
the scientific methods that in different ways are favourable to the stakeholders’ interests. In a
process where the aim is to simplify and to reduce the freedom in how you produce an LCA
many opportunities to support such self-interest exist. The only long term remedy is to be pre-
sent in the work and support solutions that are fair and can be accepted by all parties, and of
course, to act from a position of trust and credibility.
6 Conclusions
This paper has concluded that a successful company working with steel construction is ex-
pected by its customers and end users to have full knowledge of their supply chain and their
environmental and social performance and to take measures to continuously improve the sit-
uation. Furthermore it has been concluded that steel has many advantages that will not show
in sustainability assessments unless this kind of information is asked for. This can only be
achieved if steel construction companies have the credibility that let them affect or even set
the agenda for sustainability discussions.
Standards mentioned in text

ISO 14001:2004 Environmental Management Systems
ISO 14044:2006 Environmental management -- Life cycle assessment -- Requirements and guidelines
ISO 14025:2006 Environmental labels and declarations -- Type III environmental declarations -- Prin-
ciples and procedures
EN 15804:2012 Sustainability of construction works - Environmental product declarations - Core
rules for the product category of construction products
EN 15978:2011 Sustainability of construction works — Assessment of environmental performance of
buildings — Calculation method
EN 16309:2014 Sustainability of construction works - Assessment of social performance of buildings -
Calculation methodology
EN 16627:2015 Sustainability of construction works. Assessment of economic performance of build-
ings - Calculation methods
19
Abbreviations used in text

CDW Construction Demolition Waste
CEN European Committee for Standardization
CEO Chief Executive Officer
EPD Environmental Product Declaration
GHG Green House Gases
IBU Institut Bauen und Umwelt e.V.
IVL IVL Swedish Environmental Research Institute
LCA Life Cycle Assessment
LCC Life Cycle Cost
NGO Non-Governmental Organisation
PCR Product Category Rules
PEF Product Environmental Footprint
SPD Sustainability Product Declaration
References
[1] http://portailgroupe.afnor.fr/public_espacenormalisation/CENTC350/index.html
[2] http://www.eco-platform.org/
[3] http://www.iso.org/iso/home/standards/management-standards/iso14000/iso14001_revision.htm
[4] http://ec.europa.eu/environment/eussd/smgp/product_footprint.htm
[5] http://ec.europa.eu/environment/circular-economy/index_en.htm
[6] http://ec.europa.eu/environment/waste/construction_demolition.htm
[7] http://susproc.jrc.ec.europa.eu/Efficient_Buildings/index.html
[8] http://www.svd.se/vdn-trakigt-och-skammigt-med-barnarbetet ,
Svenska Dagbladet Näringsliv 2015-07-21 (In Swedish)
20
Tampere, Finland
COMPONENT METHOD AS A GENERAL TOOL FOR THE DESIGN

OF JOINTS UNDER VARIOUS LOADING CONDITIONS
Jean-Pierre Jasparta and Jean-François Demonceaub

a,b
Liège University, Belgium
Abstract: In the Eurocodes [1, 2], the component method is used as a reference for the design
of joints in steel and composite structures. Its use enables a wide range of application as far as
the individual response of the constitutive components of the studied joint are known and the
so-called “assembly procedure of the components” is available. Nowadays the acquired
knowledge allows covering a large set of joint configurations where the joints are subjected to
bending mainly, as it is the case in the Eurocodes. In the present paper, a review is made of
recent developments making possible the application of the component method to various
loading situations, including fire, earthquake, impact or explosion.
1 Introduction
In section 2, the component method is briefly introduced as well the different possible ways to
extend its scope. In the next ones, the following recent developments are then commented:
composite joints under sagging moments, joints under combined bending moments and axial
forces, joints in fire, joints under cyclic loading and finally joints under exceptional events.
All correspond to an extending of the scope in terms of loading situations, knowing that Euro-
codes cover mainly joints under positive static bending moments and shear forces.
2 The component method

The characterisation of the response of the structural joints in terms of stiffness, resistance and
ductility is a key aspect for design purposes. From this point of view, three main approaches
may be followed: (i) experimental, (ii) numerical and (iii) analytical.
The only practical one for the designer is usually the analytical approach. Analytical proce-
dures enable a prediction of the joint response based on the knowledge of the mechanical and
geometrical properties of the so-called “joint components”. The general analytical procedure
termed component method is of particular interest as it applies to any type of steel or compo-
site joints, whatever the geometrical configuration or the type of member cross-sections.
The method is nowadays widely recognised, and particularly in the Eurocodes, as a general
and powerful procedure to evaluate the mechanical properties of joints subjected mainly to
21
bending, under static loading at room temperature. Its application in daily practice is facilitat-
ed through the use of design tools available to designers in the form of computer software,
design tables with standardised joints or simplified design procedures derived in full conform-
ity with [1] and [2].
2.1 Introduction to the component method

In the past, a joint has been generally considered as a whole and studied accordingly; the orig-
inality of the component method is to consider any joint as a set of individual basic compo-
nents. For the particular joint shown in Fig. 1 (joints with extended end-plate connections
mainly subject to bending), the relevant components (i.e. zones of transfer of internal forces)
are the following:
column web in compression;
beam flange and web in compression;
column web in tension;
column flange in bending;
bolts in tension;
end-plate in bending;
beam web in tension;
column web panel in shear.
Fb Fb2 Fb1
z z
θc Mb2 z
Mb Mb1
θb
Fb
Fb2 = Mb2 / z Fb1 = Mb1 / z
Fb = M b / z
(a) Single sided joint configuration (b) Double sided joint configuration
Fig. 1: Joints with end-plate connections
Each of these basic components possesses its own strength and stiffness either in tension,
compression or shear. But the column web, for instance, is subjected to coincident compres-
sion, tension and shear. This coexistence of several components within the same joint element
can obviously lead to stress interactions that are likely to decrease the resistance of the indi-
vidual basic components.
The application of the component method requires the following steps (see Table 1):
1. identification of the active components in the joint being considered;
2. evaluation of the stiffness and/or resistance characteristics for each individual basic
component;
3. assembly of all the constituent components and evaluation of the stiffness and/or re-
sistance characteristics of the whole joint.
The assembly procedure is the step where the mechanical properties of the whole joint are de-
rived from those of all the individual constituent components. That requires, according to the
22
static theorem (A.A. Gvozdev, 1938, translated in [3]), defining how the external forces acting
on the joint distribute into internal forces acting on the components in a way that satisfies
equilibrium and respects the component behaviour (in terms of resistance and ductility).
Table 1: Schematic illustration of the component method
Joints with I or H sections under static loading and

mainly subjected to bending
Types of joints
Components
Ft,Ed
Ft,Ed
Assembly
2.1 Extending of the scope of the component method

In EN 1993-1-8 [1] and EN 1994-1-1 [2], guidelines on how to apply the component method
for the evaluation of the initial stiffness and the design moment resistance of steel and compo-
site joints are provided; the aspects of ductility are also addressed. The combination of the
components proposed in [1] and [2] allows one to cover a wide range of joint configurations
and should be largely sufficient to satisfy the needs of practitioners (welded joints, bolted
joints with end-plates or cleats, various joint stiffening including transverse column stiffeners,
supplementary web plates, backing plates, column web plates and beam haunches). However
assembly design rules are provided for joints under static loading and mainly subjected to
bending moments and shear forces, but also for joints connecting profiles with open sections.
The extending of the application field of the component method may follow separate ways.
Amongst them: (i) increase the number of components for which design rules are provided to
the user and (ii) derive knowledge for the characterisation of the component and the assembly
of components in other loading conditions. Through (i), the field is extended to other joint
configurations. As an example, CIDECT is nowadays sponsoring a research project [4] so as
to convert the present available design rules for joints in tubular construction (chapter 7 of EN
1993-1-8) into a component format. Beyond the interest to refer to a single procedure for
joints into the Eurocodes, such a project will directly allow, for instance, to cover though EN
1993-1-8 joints between open beam sections and tubular column sections.
23
In the following sections of the paper, it is intended to investigate further the second way (ii)
and to present recent works in this domain.
3 Composite joints under sagging moment

The mechanical characterisation of composite joints subjected to hogging bending moment
may be achieved by means of Eurocode 4 [2]. However, insufficient information is provided
there to predict the behaviour of under sagging moments (Fig. 2). Indeed, even if most of the
activated components under such a loading can be characterised using [1] and [2], no rule is
available to characterise one of the activated components: the concrete slab in compression in
the vicinity of the column, i.e. where contact forces are transferred.
In recent researches, methods to characterise this component in terms of resistance and stiff-
ness have been proposed ([5] to [8]). They aim at defining a rectangular cross section of con-
crete participating to the joint resistance. In [7] and [8], the second author suggests to combine
two methods proposed respectively by Ferrario [5] and Liew [6], the combination of these two
methods reflecting in a more appropriate way how the concrete resists to the applied load in
the vicinity of the joint.
So, through the study of one single new component [7,8], it is possible to characterise a signif-
icant number of composite joint configurations under a new specific loading (sagging mo-
ments). This demonstrates the flexibility and the adaptability of the component method.
f ck,actual
Z
150
2970
IPE300 FRd,3
HEB260
2440
FRd,2
FRd,1
Sagging moment Load distribution in a composite connection under

sagging moment
Fig. 2: Composite joint subjected to sagging moment
4 Joints under combined bending moments and axial forces

In most of the cases, beam-to-column joints and beam splices are subjected to compression or
tension axial forces in addition to bending moments and shear forces. These ones have an in-
fluence on the rotational stiffness, moment resistance and rotational capacity of the joints.
And that is why in Part 1.8 of Eurocode 3 [1] the proposed field of application is limited to
joints in which the force N Ed acting in the joint remains lower than 5% of the axial design re-
sistance N pl , Rd of the connected beam. Under this limit it is considered that the rotational re-
sponse of the joints is not significantly influenced by the axial forces. It has however to be
stated that this value is a fully arbitrary one and is not at all scientifically justified. The 5%
rule covers beam-to-column joints and beam splices in multi-storey building frames, but usu-
24
ally not similar joints in pitched-roof industrial portal frames. Similarly column bases and
column splices transfer high axial forces and therefore do not fulfil the limiting 5% criterion.
Part 1.8 considers that the interaction resistance diagram is defined by the polygon assembling
the 4 points corresponding to positive and negative bending resistances in absence of axial
force to axial tension and compression resistances in absence of bending.
It should be noted that for such joints “under M − N ”, the principles of the component method
is still valid – as the behaviour of the components is independent on the type of loading ap-
plied to the whole joint – but a new assembly procedure is required to cover the combined ac-
tion of bending moments and axial forces. The main difficulty results from the variation of the
active components in the joints according to the relative importance of the bending moment
and axial force, and obviously according to the respective signs of the applied forces.
These items are addressed in [9] and [10] for the characterisation respectively of the resistance
and of the rotational stiffness. The analytical procedures presented there consider the mechan-
ical model shown in Fig. 3 to represent the behaviour of a joint submitted to both bending and
axial forces. In this model, each constitutive component of the joint is represented by an ex-
tensional spring characterised by a non-linear F − ∆ curve, where F and ∆ represent respective-
ly the force acting in the component and the related displacement. According to the definitions
[1], the joint is seen to be constituted of a connection subjected to bending moment and axial
force and a column web panel in shear.
M
M CWT CFB BT EPB BWT ϕ N
N γ
z
CWC BFC
Column web panel Connection

in shear
Joint
Fig. 3: Mechanical model as a reference used in the proposed analytical procedures
For M − N interaction, two particular effects have also to be considered:

− “Group effects”: these effects may occur in bolted connections and more especially
(Fig. 3) in constitutive plate components subjected to transverse bolt forces (endplates
in bending – EPB -, column flanges in bending - CFB …). There, where a bolt force is
applied, a yield plastic mechanism may develop in the plate component; if the distance
between bolts is high, separate yield lines will form in the plate component around the
bolts (individual bolt mechanisms), while a single yield plastic mechanism common to
several bolts may develop when the distance between the latter decreases (bolt group
mechanisms). Group effects also affect the resistance of following components (Fig.
3): column web in tension – CWT – and beam web in tension – BWT -.
− "Component interactions": Interaction effects between components may occur in the
column components where different types of stresses co-exist: shear stresses, longitu-
dinal stresses due to axial and bending forces present in the column and transversal
25
stresses due to the introduction of the load of the joint. These interactions may affect
the resistances of the related components.
In a first step, the behaviour of each of the constitutive joint components has been assumed in
[9] and [10] to be infinitely ductile. As a result, a full plastic redistribution of the internal
forces in the joint under M and N carried out on the basis of the so-called static theorem [3]
may be contemplated. The so-derived ductile resistance interaction diagram corresponds to a
plastic resistance surface; the actual applied bending moment and axial force in a connection
define a couple of values which should remain inside the M − N interaction diagram (see Fig.
5 for the joint reported in Fig. 4) so as to ensure the sufficient resistance of the studied joint.
However the ductility of some components is sometimes not sufficient to allow for a full plas-
tic redistribution of the internal forces in the joints. When a non-ductile component reaches its
deformation capacity, any additional deformation causes the brittle failure of that component
and consequently of the whole joint. Besides welds, bolts in tension are considered as non-
ductile components. It is assumed in [1] that the deformation capacity around a bolt is suffi-
cient if the design resistance of the “plate-bolt assembly” is lower or equal to 95% of the ten-
sion bolt resistance. Moreover, the “beam flange and web in compression” component (BFC)
may also be considered as non-very ductile when the beam cross-section becomes slender and
its resistance is limited by buckling phenomena (class 4 sections). In [9] and [10] the proposed
design procedure has been adapted to this particular aspect and validated through comparisons
with experimental test results; it is applied in Fig. 5 to the joint shown in Fig. 4, in which rows
3 and 6 are now assumed to exhibit a non-ductile behaviour (BT).
2400 N [kN]
ductile response
1600
M24 HR 8.8 non ductile response
Row 1
Upper row=2
800
Row 3
af = 14 mm
Row 4 IPE 600 S355
α=20° 0 M [kNm]
-660 -440 -220 0 220 440 660
Row 5
aw = 8 mm -800
Row 6
Lower row=7
-1600
300X780X20 S355
HEB 400 S355
-2400
Fig. 4 - Bolted joint with Fig.5 - Ductile and non-ductile M-N interaction
numbering of force transfer rows diagrams (including stress interactions)
5 Joints in fire
In [11], Da Silva et al. have first proposed to refer also to the component method to character-
ise the behaviour of steel joints at high temperature. In their study, they have demonstrated
that the rotational stiffness and the bending resistance of a structural joint may be simply ob-
tained by multiplying the corresponding properties derived at room temperature by ad-hoc re-
duction factors (respectively k E ;θ for stiffness and k y ;θ for resistance) evaluated according to
Eurocode 3 Part 1-2 [12]. These factors express the decrease of the steel Young modulus E
and yield strength f y at temperature θ . However this simple approach is limited to cases
where all the components are subjected to the same increase of temperature, what is not often
reflecting the reality. This is why in [13] the procedure has been improved as follows: (i) in
26
the tests used as references, temperatures θi have been measured in all components i , (ii) the
stiffness and resistance properties of the components i have been multiplied by ad-hoc reduc-
tion factors k E ;θi and k y ;θi and (iii) the assembly of the components has been finally achieves so
as to characterise the global response of the joint, duly attention being so paid to the variation
of temperatures in the joint. The procedure has been validated through comparisons with test
results on composite joints. So as to allow an easy application in practice, Demonceau et al.
[14] have decided to go one step further by replacing the measurement of actual temperatures
in the joint during laboratory tests by a thermal analysis achieved with SAFIR [15]. In some
cases, he has even suggested analytical expressions to determine the temperature of individual
components as a function of the time.
6 Joints under cyclic loading

In Eurocode 8, Part 1-1, and in particular in Chapters 6 and 7 dealing respectively with the
seismic design of steel and composite structures, it is clearly stated that the use of partial
strength joints is permitted but the number of requirements to be respected for this joint typol-
ogy are such that it is nowadays nearly compulsory to perform experimental tests to check
when these requirements are fulfilled; this fact is confirmed in Eurocode 8, Part 1-1, in the
clause (6) of Section 6.5.5.
Accordingly, the use of partial strength joints in structures prone to seismic actions is very
limited and, as a consequence, the practitioners have to design full strength joints taking into
account the possible overstrength effects, which leads to expensive joint solutions and so lim-
its the competiveness of steel structures compared to other structural solutions. An example of
such optimised full strength joint solution is presented in Fig. 6 [16]; this solution was devel-
oped in the framework of a recent RFCS project entitled HSS-SERF (High Strength Steel in
SEismic Resistant building Frames) [17].
Fig. 6: Full strength optimised beam-to-column joint solution
The component method could be a solution to overcome “full-strength” this obligation. In-
deed, the component method has the potential to predict the response of joints under cyclic
loading but for that, it is required to know the behaviour of each component under such load-
27
ing conditions. In particular, it is necessary to know the post-yielding behaviour of the com-
ponents accounting for the strain-hardening effects, their ultimate resistance, their defor-
mations capacity but also the degradation of their strength and stiffness under the applied cy-
cles associated to phenomena of oligo-cyclic fatigue. In addition, it is needed to know the un-
loading behaviour of each component as such unloading may occur during the cyclic loading
imposed by the seismic action.
Investigations have been recently conducted at the University of Coimbra [18] with very
promising results. The proposed model consists in a numerical implementation of a hysteretic
model able to simulate a generic steel or steel-composite joint behaviour. The use of a numer-
ical implementation is here required as the evolution of the loads in each component at each
step of the applied loading has to be known in order to be able to detect the strength/stiffness
degradation and the moment at which the maximum deformation capacity of a component is
reached. The proposed model in [18] is still under development/improvement nowadays, in
particular through contributions to the European RFCS project EQUALJOINTS (European
pre-QUALified steel JOINTS).
7 Joints under exceptional events

A structure should be designed to behave properly under service loads (at SLS) and to resist
design factored loads (at ULS). The type and the intensity of the loads to be considered in the
design process may depend on different factors such as: the intended use of the structure (type
of variable loads…), the location (wind action, level of seismic risk…) and even the risk of
accidental loading (explosion, impact, flood…). In practice, these individual loads are com-
bined so as to finally derive the relevant load combination cases. In this process, the risk of an
exceptional (and therefore totally unexpected) event leading to other accidental loads than
those already taken into consideration in the design process in itself is not at all covered. This
is a quite critical situation in which the structural integrity should be ensured, i.e. the global
structure should remain globally stable even if one part of it is destroyed by the exceptional
event (explosion, impact, fire as a consequence of an earthquake …). In conclusion, structural
integrity is required when the structure is subjected to exceptional actions not explicitly con-
sidered in the definition of the design loads and load combination cases.
Under such exceptional actions, the structural elements and in particular the joints are general-
ly subjected to loadings not initially foreseen through the ULS design. For instance, if the ex-
ceptional event “loss of a column” is considered, the joints will experience high tying forces
after the loss of a column, as a result of the development of membrane forces in the beams
located just above the damaged or destroyed column while these joints are initially designed
to transfer shear forces and hogging bending moments. Moreover a reversal of moments oc-
curs in the joints located just above the damaged column. Finally, the joints could be subject-
ed to some dynamic effects if the column loss is for instance induced by an impact or an ex-
plosion.
In Section 4, it has been shown how the M − N plastic resistant curve of a joint can be predict-
ed through the use of the component method. Of course, the methodology presented there can
be of help to predict the behaviour of the joints when subjected to exceptional events. Howev-
er, it has to be pointed out that, when considering the behaviour of structures subjected to such
event, the main objective is to ensure that the building will remain globally stable and so it
can be accepted to go a step further in the resistance of the structural elements in comparison
28
to what is imposed as limits for ULS. Accordingly, in addition to the prediction of the plastic
resistance of the joints under M − N , it is also important to be able to predict the ultimate re-
sistant curve, i.e. to be able to predict the M − N combinations under which the joints fail. In
[19], it is explained how the M-N plastic and ultimate resistance curves can be predicted for a
composite joints and how the proposed model has been validated through comparisons to ex-
perimental results (Fig. 7).
70 500
60
HOGGING MOMENTS
50
40
30
20
10
TENSION
N [kN]
M [kN]
0
0 50 100 150 200 250 300 350 400 450 500 550 600 650 700
-10
-20 Analytical prediction_plastic resistance curve

SAGGING MOMENTS
-30 Analytical prediction_ultimate resistance curve
Experimental results_TEST 1
-40
-50
-60
-70 Experimental results_TEST 5
-80
Fig. 7: M − N plastic and ultimate resistance curves for a composite joint
In parallel to the prediction of the resistance curves, another key issue to be considered when
predicting/estimating the robustness of a structure is the prediction of the deformation capaci-
ty or the ductility of a joint. Research efforts are still required in this field. Finally, another
aspect to be dealt with when considering the behaviour of joints under exceptional events is
the possible dynamic effects which can be induced by these events, dynamic effects which can
be associated to strain rate effects in some joint components. This aspect is presently under
investigation at the University of Liège in the framework of a RFCS European project entitled
ROBUSTIMPACT.
8 Conclusions
1. The component method is a general procedure for the characterisation of the mechanical
properties of structural joints nowadays recommended in Eurocodes for steel and com-
posite structures.
2. In the aforementioned standards, it is used mainly for joints in bending and shear under
static loading.
3. In the present paper, it is shown how its scope of application may be easily extended to
various other loading situations including fire, earthquake or exceptional events.
References
[1] EN 1993-1-8. Eurocode 3: Design of steel structures – Part 1-8: Design of joints. Euro-
pean committee for standardization, May 2005.
[2] EN 1994-1-1: 2004, “Eurocode 4: Design of composite steel and concrete structures –
Part 1-1: General rules and rules for buildings”, European committee for standardiza-
tion, December 2005.
29
[3] Heyman J. “Structural analysis, a historical approach”, Cambridge University Press,

1996.
[4] Jaspart J.P. and Weynand K. “Design hollow section joints using the component meth-
od” to appear in the “Proceedings in the 15th International Symposium on Tubular
Structures”, Rio (Brazil), May 27-29, 2015.
[5] Ferrario F. “Analysis and modelling of the seismic behaviour of high ductility steel-
concrete composite structures”, PhD thesis presented at Trento University, 2004.
[6] Liew R.J.Y., Teo T.H. and Shanmugam N.E. “Composite joints subject to reversal of
loading – Part 2: analytical assessments”, Journal of Constructional Steel Research, pp.
247-268, 2004.
[7] Demonceau, J.-F, “Steel and composite building frames: sway response under conven-
tional loading and development of membranar effects in beams further to an exceptional
action”, PhD thesis presented at Liège University, 2008 (freely downloadable at
http://orbi.ulg.ac.be/handle/2268/2740).
[8] Demonceau, J.-F, Jaspart, J.-P., Klinkhammer, R., Weynand, J.-P., Labory, F. and Cajot,
L-G. “Recent developments on composite connections”, Steel Construction – Design
and Research Journal, pp. 71-76, 2008.
[9] Cerfontaine F. and Jaspart J.P. “Resistance of joints submitted to combined axial force
and bending”, Proceedings of the Eurosteel 2005 Conference, Maastricht, 2005.
[10] Cerfontaine F. “Etude de l'interaction entre moment de flexion et effort normal dans les
assemblages boulonnés”, Université de Liège, 2003. PhD Thesis.
[11] Simoes da Silva L., Santiago A. and Villa Real P. “A component model for the behav-
iour of steel joints at elevated temperatures”, Journal of constructional steel research,
Vol. 57, pp. 1169-1195, 2011.
[12] EN 1993-1-2. Eurocode 3: Design of steel structures – Part 1-2: General rules – Struc-
tural fire design. European committee for standardization, April 2005.
[13] Demonceau J.F., Haremza C., Jaspart J.P., Santiago A. and Simoes da Silva L. “Compo-
site joints under M-N at elevated temperatures”, Proceedings of the Composite Con-
struction VII conference, Palm Cove, Australia, 2013.
[14] Demonceau J.F., Hanus F., Jaspart J.P. and Franssen J.M. “Behaviour of single-sided
composite joints at room temperature and in case of fire after an earthquake”, Interna-
tional Journal of Steel Structures, Vol. 9(4), pp. 329-342, 2009.
[15] Franssen J.M. “SAFIR A thermal/structural program modelling structures under fire”,
Engineering Journal, AISC, pp. 143-158, 2005.
[16] Hoang V. L., Jaspart J.-P. and Demonceau J.-F. “Hammer head beam solution for beam-
to-column joints in seismic resistant building frames”, Journal of constructional steel
research, Vol. 103, pp. 49-60, 2014.
[17] Dubina D. et al. “High strength steel in seismic resistant building frames (HSS-SERF)”,
Final report of the HSS-SERF RFCS project, European Commission, 2015 (ISBN 978-
92-79-44081-6).
[18] Nogueiro, P., Simoes da Silva, L., Bento, R. and Simoes, R. “Calibration of model pa-
rameters for the cyclic response of end-plate beam-to-column Steel-concrete composite
joints”, Steel and Composite Structures Journal, Vol. 9, N°1, pp. 39-58, 2009.
[19] Jaspart, J.-P. and Demonceau J.-F. “Composite joints in robust building frames”, Pro-
ceedings of the Composite Construction in Steel and Concrete Conference, Colorado,
USA, July 2008.
30
Tampere, Finland
EXECUTION OF STEEL STRUCTURES

RECENT DEVELOPMENTS AND FUTURE TREND
Bjørn Aasen a
a
Norconsult AS, Sandvika, Norway
Abstract: The development of European rules for fabrication and erection of steel structures,
EN 1090-2, is briefly described. Some problems that were discovered during the preparation
of the standard are also included. Furthermore, EN 1090-2 has become a bench mark for
preparation of a new ISO standard for the execution of structural steelwork
1 Introduction
Life is lived forwards but is understood backwards.
Søren Kierkegaard
The revised prEN 1090-2 marks the end of 25 years of standardization of execution rules for
steel structures. The scope states that this EN standard specifies requirements for fabrication
and erection of any type and shape of steel structures including structures subjected to fatigue
or seismic actions. The standard applies to structures designed according to the relevant part
of EN 1993, but also for structures designed according to other design rules.
Table 1 shows that prEN 1090-2 has 12 chapters and 13 annexes. In this paper only a few
points are considered that may shed light on developments and future work.
Table 1: Table of contents of prEN 1090-2: 2014

1 Scope
2 Normative references
3 Terms and definitions
4 Specifications and documentation
5 Constituent products
6 Preparation and assembly
7 Welding
8 Mechanical fastening
9 Erection
10 Surface treatment
11 Geometrical tolerances
12 Inspection, testing and correction
31
Table 1 (continued)
(N)
A Additional information, list of options and requirements
related to the execution classes
B(I) Check-list for the content of a quality plan
C(N) Geometrical tolerances
D(I) Procedure for checking capability of thermal cutting processes
E(I) Welded joints in hollow sections
F(I) Guidance on the selection of weld classes
G(N) Corrosion protection
H(N) Test to determine slip factor
J(N) Calibration test for preloaded bolts under site conditions
K(I) Procedure for checking loss of preload for thick surface coat-
ings
L(I) Resin injection bolts
M(I) Guide to flow diagram for development and use of a WPS
N(N) Sequential method for fasteners inspection
Notes:
(N) means a normative annex
(I) means an informative annex
2 Development of international standards for execution of steel structures

2.1 Model standard for fabrication and erection of steel structures
Following a Norwegian initiative Technical Committee ISO/TC 167 was established in 1977.
The main purpose was to prepare model standards for steel and aluminium structures. The
standards should comprise requirements for design, fabrication and erection of metallic struc-
tures, together with materials, structural components and connections. Thus, code writers
could use these ISO standards as a basis for own national standards.
Two standards for steel structures were prepared by ISO/TC 167 before a standby was im-
plemented in 1999; where ISO 10721-2 comprises fabrication and erection of steel structures.
The standard was prepared by Subcommittee 2 with British chairmanship and secretariat.
2.2 Previous ENV-standard for execution of steel structures

Technical Committee CEN/TC135 was established in 1987 by another Norwegian initiative.
The original terms of reference was to provide a European design standard for steel structures
as an alternative to Eurocode 3. The Eurocodes were commissioned by the European Com-
munities, see the first version of Eurocode 3 [4]. These design codes were intended to estab-
lish a set of common rules as an alternative to differing rules in force in the various member
states. However, this new approach was considered by the CEN members as an undue inter-
ference in their business.
To avoid a conflict of interest between the Commission and CEN it was accepted that Norway
should have the chairmanship and the secretariat. The experience with Norwegian manage-
ment of ISO TC167 was also crucial. However, due to involvement by Professors Brozzetti
(France), Sedlacek (Germany) and Stark (the Netherlands) the purpose of TC 135 was
changed from standardization of design rules to execution rules for steel structures. Their
commitment had an impact on the lay-out of ENV 1090 “Execution of steel structures” which
is similar to that of Eurocode 3:
32
Part 1: 1996 “General rules and rules for buildings”;

Part 2: 1997 “Rules for cold formed thin gauge members and sheeting”;
Part 3: 1997 “Supplementary rules for high strength steels”;
Part 4: 1997 “Supplementary rules for hollow section lattice structures”;
Part 5: 1998 “Supplementary rules for bridges and plated structures;
Part 6: 2000 ”Supplementary rules for stainless steel”.
The aim of this approach was to avoid duplication of execution rules in the various parts, but
the main disadvantage was fragmentation of requirements. For example fabrication of a pe-
destrian bridge made of hollow section members would call up three different parts such as
Parts 1, 4 and 5.
At the kick-off meeting of TC 135 that was held in Oslo in December 1988, it was decided
that four different working groups should prepare ENV 1090-1. The provisional annexes of
Eurocode 3 comprising rules for fabrication, bolted connections, welded connections and
erection were transferred to the working groups. A coordination group was set up to monitor
the work and progress. In this group the professors as previously mentioned, contributed by
their ability of stringent and logical thinking. Together with the members of working groups
TC 135 became an efficient forum for exchanging experiences about steel structures.
At the beginning progress in preparing Part 1 was good. Several problems occurred during the
discussions in WG 3 “Welding”, which caused delay of ENV 1090-1:
1. The draft proposal of EN 10025 had no limitation on the carbon equivalent value,
CEV. Welding experts of WG 3 pointed out that difficulty of writing the welding pro-
cedures (WPS) could occur prior to any material deliveries.
It was proposed that ENV 1090-11 should include restrictions on maximum CEV-
values, a proposal that was rejected by the European Committee for Iron and Steel
Standardization, i.e. ECISS/TC10.
The problem was solved when ECISS published an addendum EN 10025, where max-
imum CEV-values were given as an option. In the current version of this EN standard
maximum CEV-values have become mandatory.
2. The Construction Product Directive (CPD) approved by EU in December 1989 created

a lot of discussions within WG 3. The majority of the members considered certifica-
tion of steel workshops to be outside the scope ENV 1090-1.
The discussion was settled by an informal Annex E in ENV 1090-1: Guidelines for
welding coordination. Here, the competence of a welding coordinator was depending
on the level of quality requirements for welding according to EN 729 (now EN ISO
3834).
3. Different views on weld acceptance criteria for weld imperfections caused the biggest
discussions. Dr. Ogle (UK) had developed a comprehensive system for weld inspec-
tion including weld acceptance criteria of steel structures subjected to static and dy-
namic actions; reference is made to ISO 10721-2. Mr. Lindewald (Finland) argued
strongly in favour of the EN ISO 5817
prENV 1090-1 was issued including both the British and the ISO approaches. A large
majority of CEN member countries voted for the ISO standard.
33
There was some tension between CEN/TC 121 “Welding” and CEN/TC 135 because the first
committee prepared a standard for arc welding of ferritic steels, i.e. EN 1011. But in practice
there was no conflict of interests because Mr. Allen (UK), was the Convener for EN 1011 and
also member of WG3.
2.3 Present EN-standard for execution of steel structures

Eurocode 3 was transferred from the EU Commission to CEN and published in 1993 as a Eu-
ropean Prestandard ENV 1993-1-1. Some year later work was launched with financial support
from EU and EFTA to convert this design standard into four different EN standards, i.e.
EN 1993-1-1, EN 1993-1-8, EN 1993-1-9 and EN 1993-1-10. Thus, it became necessary to
convert ENV 1090 by merging Parts 1 - 6 together in one standard, namely EN 1090-2.
In order to comply with the CPD a standard for conformity assessment of structural compo-
nents was prepared, i.e. EN 1090-1. The hierarchy for design, execution and CE-making of
steel structures is shown in Fig. 1.
Fig.1: Design and execution of steel structures

NOTE: CE-marking according to EN 1090-1 is restricted to fabrication of steel structures
In order to have a single volume for execution of all types of steel structures a differentiation
of the reliability of completed works or structural components was required. Execution clas-
ses (EXC1 – EXC4) were introduced and defined as a classified set of requirements specified
for the execution of the works as a whole, of an individual component or a detail of a compo-
nent. The execution requirements are progressively increasing from EXC1 up to EXC4, where
EXC4 should be applied to structures with extreme consequences of structural failure.
34
The procedure for determination of an execution class includes three steps:
1. the required reliability class or consequence class according to EN 1990,

2. the type of structure, component or detail according to the relevant part of EN 1993,
3. the type of loading for which the structure, component or detail is designed.
Guidance for the determination of execution classes is given in EN 1090-2: Annex B. This
annex was made informative, pending on new guidelines to be included in Eurocode 3. An
amendment to EN 1993-1-1: 2005/A1:2014 was published and should at now have been im-
plemented in design practice. The selection of execution class is given in Table 2.
Table 2: Choice of execution class (EXC) according to EN 1993-1-1 [5]

Type of loading
Reliability Class (RC) or
Static, quasi-static Fatigue b
Consequence Class (CC)
or or
(EN 1990)
seismic DCL a seismic DCM or DCH a
RC3or CC3 EXC3 EXC3 c
RC2 or CC2 EXC2 EXC3
RC1 or CC1 EXC1 EXC2
Notes
a. Seismic classes according to EN 1998-1: Low = DC; Medium = DCM; High = DCH
b. See EN 1993-1-9
c. EXC4 may be specified
There is still a concern that EN 1090-2 may lead to increased costs for 'normal' steel struc-
tures. If EXC1 is selected for the whole steel structure, then EN 1993-1-1: 2005/A1:2014 rec-
ommends that EXC2 should be applied to some types of welded components such as:
 components manufactured from structural steel S355 and above

 components essential to structural integrity that are assembled by site welding
After the approval of EN 1090-2 in 2008 several errors, ambiguities and shortcomings were
discovered and discussed. An amendment was approved by CEN in 2011 and the updated
version the standard was published as EN 1090-2: 2008+A1:2011. Unfortunately, some dis-
cussions continued regarding:
 the acceptance criteria for weld imperfections,

 lack of available EN-standards for bolts and nuts in non-preloaded connections,
 the maximum initial out-of-straightness of compression members.
Mr. Måseide (Norway) [16] has reported that the acceptance criteria for weld imperfections
according to EN ISO 5817 are more onerous than those given in the Norwegian offshore
standard M101. In Table 3 a comparison of the quality levels is made between ISO and
NORSOK regarding the internal weld imperfections for butt welds. The NORSOK standards
are developed by the Norwegian petroleum industry to ensure adequate safety, value adding
and cost effectiveness for petroleum industry developments and operations. Thus, it is a para-
dox that offshore structures have more relaxed requirements than 'normal' steel structures.
However, one reason could be that costly repairs of welds may ultimately impair the welded
components of offshore structures, while another reason could be the greater extent of NDT
offshore platforms than for building structures.
35
Table 3: Comparison of quality levels between EN ISO 5817 and NORSOK M-101
for internal weld imperfections in butt welds
Type of imperfection ISO 5817 NORSOK
ISO 6520-1 B C D A&B C, D & E
Slag inclusions h ≤ 2 mm h ≤ 3 mm h ≤ 4 mm h ≤ 6 mm h ≤ 6 mm
No. 301 l ≤ 25 mm l ≤ 50 mm l ≤ 75 mm l ≤ 50 mm l ≤ 100 mm
Lack of fusion Not Not h ≤ 4 mm l ≤ 25 mm l ≤ 50 mm

No. 401 permitted permitted
Single pore d ≤ 3 mm d ≤ 4 mm d ≤ 5 mm d ≤ 6 mm
No. 2011
Regarding bolting assemblies for non-preloaded bolted connections EN 1090-2 refers only to
EN 15048. As an alternative the more expensive high strength bolts and nuts according to
EN 14399 may be used. EN 15048 is a harmonized standard that requires both bolts and nuts
to be tested together and supplied by one producer, while EN ISO 898-1 and EN ISO 20898-2
describe independent testing of bolt and nuts with no corrosion protection, respectively.
Unfortunately, the manufacturers did not produce bolting assemblies according to EN 15048.
Therefore bolting assemblies according to outdated DIN standards were still specified, at least
in Norway. But from the 1st of July 2014 CE marked bolting assemblies have also become
mandatory forcing the bolt manufacturers to provide products with Declaration of Perfor-
mance (DoP) according to the CPR which has replaced the CPD. Besides, the revised ver-
sions of EN 15048 have been submitted for public review. According to prEN 15048-2 the
well-known ISO bolts and nuts have received first priority. These are good news for all par-
ties involved in design, specification and execution of 'normal' steel structures!
The most controversial changes in EN 1090-2 appears to be the increase of out-of-straightness

limit Δ for welded compression members from ±L/1000 to ±L/750, where L is the member
length. The European buckling curves a – c which was published in 1970, is based on the as-
sumption of a maximum geometrical deviation of 1/1000. The main concern is what reduction
in the level of safety may be expected when compression members are designed according to
Eurocode 3 and fabricated in accordance with the tolerances of EN 1090-2.
Dr. Taras (Austria) [15] has carried out a large number of FEM simulations of imperfect steel
columns including random distributions of geometrical and material imperfections. The con-
clusion is that the increased tolerance limit for compression members according to EN 1090-2
cannot be justified by the current buckling formulas of Eurocode 3. Furthermore, the modified
tolerance for initial-out-of straightness could decrease the column buckling strength of about
5 %. This loss of safety may be counteracted by increasing the γM1-value. However, an inter-
mediate solution could be to specify class 2 of functional manufacturing tolerance, see EN
1090-2.
2.3 Revised EN standard for execution of steel structures

As mentioned in the Introduction the revised EN 1090-2 has been submitted for public re-
view. Now it is important to gather most of the experiences that have been obtained during
36
practical use of the standard, see [13] and [14]. The revised EN 1090-2 is expected to be ap-
proved at the end of 2016 or at the beginning of 2017.
Upon initiative from European industry a stand-alone execution standard for cold-formed thin
gauge members and sheeting has been prepared and issued for public review, i.e. prEN 1090-
4.
3 Future developments
3.1 Globalization of structural steel fabrication requires ISO standards
Comparing the normative references as listed in ENV 1090-1 (1996), EN 1090-2 (2008/2011)
and prEN 1090-2 (2014), one discovers that a large number of EN standards that have been
replaced by ISO standards. This development applies to the areas that will influence fabrica-
tion of steel structures such as:
 structural steels, see Table 4,

 welding, see Table 5,
 NDT-methods,
 mechanical fasteners,
 preparation and surface treatment.
The conversion from EN standards to ISO standards is an outcome of the Vienna Agreement
which was approved in 2001 [17]. The statistics regarding CEN in relation to ISO show that:
 approximately 33 % of all EN standards are identical to ISO standards,

 33 % of EN standards are on similar topics,
 33 % of EN standards are without ISO counterparts.
The revised ISO 630 standards for structural steels, see Table 4, were approved in 2013 and
2014. The similarity between the various parts of ISO 630 and EN 10025 is striking. Time
will show when ISO 630 is included in future European design and execution standards for
steel structures.
A more obvious trend is the transition from EN to ISO standards within welding and NDT-
methods. The fabrication of steel structures for bridges, buildings, towers and masts is in-
creasingly carried out in the Far East and exported to Europe. As shown in Table 5 the re-
placement of welding standards is complete.
As previously mentioned ISO offers a complete set of standards for bots, nuts, washers and
corrosion protection for non-preloaded connections. The situation remains unsolved for bolt-
ing assemblies suitable for preloading because American, Japanese and European engineers
do not yet agree.
37
Table 4: Comparison of standards for structural steels

Structural steels ENV 1090-1 prEN 1090- ISO/wd1
2 17607
Technical delivery terms - EN 10025-1 ISO 630-1
Non-alloy structural steels EN 10025 EN 10025-2 ISO 630-2
Normalized weldable fine grain structural steels EN 10113-2 EN 10025-3
Thermomechanical rolled weldable fine grain EN 10113-3 EN 10025-4 ISO 630-3
structural steels
Structural steels with improved atmospheric EN 10155 EN 10025-5 ISO 630-4
corrosion resistance
High yield strength structural steels in the - EN 10025-6 ISO 630-5
quenched and tempered condition
Seismic-improved structural steels - - ISO 630-6
Hot finished hollow sections of structural steels EN 10210-1 EN 10210-1 ISO 12633-1
Cold formed hollow sections of structural steels EN 10219-1 EN 10219-1 ISO 10799-1
Types of inspection documents EN 10204 EN 10204 ISO 10474
Note: 1. ISO/wd is a working draft of ISO/TC 167/WG 3
Table 5: Comparison of standards for welding structural steels

Welding of structural steels ENV 1090-1 prEN 1090-2 ISO/wd1 17607
Qualification of welders. Fusion welding EN 287-1 EN ISO 9606-1 ISO 9606-1
Specification and approval of welding procedures for metallic materials

Welding procedure specification EN 288-2 EN ISO 15609- ISO 15609-1
1
Welding procedure tests EN 288-3 EN ISO 15614- ISO 15614-1
1
Approval using approved welding consuma- EN 288-5 EN ISO 15610 ISO 15610
bles
Approval related to previous experience EN 288-6 EN ISO 15611 ISO 15611
Approval by a standard welding procedure EN 288-7 EN ISO 15612 ISO 15612
Approval by a pre-production welding test. EN 288-8 EN ISO 15613 ISO 15613
Approval testing of welding personnel for EN 1418 EN ISO 14732 ISO 14732
fully mechanized and automatic welding
Welding coordination EN 719 EN ISO 14731 ISO 14731
Quality requirements for welding – Fusion welding of metallic parts

Guidelines for selection and use EN 729-1 EN ISO 3834-1 ISO 3834-1
Comprehensive quality requirements EN 729-2 EN ISO 3834-2 ISO 3834-2
Standard quality requirements EN 729-3 EN ISO 3834-3 ISO 3834-3
Elementary quality requirements EN 729-4 EN ISO 3834-4 ISO 3834-4
Welding – Recommendations for welding of metallic materials

General guidance for arc welding EN 1011-1 EN 1011-1 ISO/TR2 17671-
1
Arc welding of ferritic steels EN 1011-2 EN 1011-2 ISO/TR2 17671-
2
Notes:
1. ISO/wd is a working draft of ISO/TC 167/WG 3
2. ISO/TR is an ISO technical report
38
3.2 Preparation of a revised ISO standard for execution of steel structures

ISO/TC167 was reactivated in 2010 by an American initiative for the following reason: “Cur-
rently, there are only national and regional (EN) standards that address steel structures. While
ISO 10721-1:1997 and ISO 10721-2:1999 still remain in the ISO database, these standards
are outdated and not widely used or implemented. With the upturn in global trade, especially
where steel structures are fabricated in many different geographical locations and then ex-
ported to others, an ISO solution is urgently needed to ensure uniformity of product and safe-
ty,” see [11].
Norway has retained the chairmanship and secretariat for ISO/TC 167, while the drafting
work is prepared by ISO/TC 167/WG3 with Mr. Sindel (USA) as the Convener and the Amer-
ican Welding Society (AWS) as secretariat.
Table 6 shows how the work is carried out and presented. Here, EN 1090-2 is the bench mark.
When the ISO draft is expected to be complete within two years, this document will show the
relationship between the EN and the revised ISO standard. Thus, CEN/TC 135 will get valua-
ble input for the future work of updating EN 1090-2.
Table 6: Contiguous working draft showing the relation between EN 1090-2 and NP 17607
Note: Status 2014-04-14
4 Conclusions
Substantial progress has been made to prepare a European standard for the execution of steel
structures. EN 1090-2 has been translated to the majority of official EU languages, promoting
the steel workshops to deliver steel structures and components cross borders within the Single
European Market.
EN 1090-2 has been considered as efficient basis for developing an new ISO standard for ex-
ecution of steel structures.
39
References
EN standards
Note: For EN 10025 and EN 1011, reference is made to Tables 4 and 5.
[1] ENV 1090-1: 1996 “Execution of steel structures. Part 1: General rules and rules for
buildings.”
[2] EN 1090-2: 2008+A1:2011 “Execution of steel structures and aluminium structures.
Part 2: Technical requirements for steel structures.”
[3] prEN 1090-2: 2015 “Execution of and aluminium steel structures. Part 2: Technical re-
quirements for steel structures.”
[4] Eurocode 3: 1984 “Common unified rules for steel structures.” EUR 8849, Commission
of the European Communities, Brussels 1984
[5] EN 1993-1-1: 2005/A1:2014 “Eurocode 3: Design of steel structures – Part 1-1: General
rules and rules for buildings.”
[6] EN 1990: 2002 “Eurocode — Basis of structural design”
[7] prEN 15048-2: 2014 “Non-preloaded structural bolting assemblies - Part 2: Fitness for
purpose test”
[8] EN ISO 5817: 2003 “Welding - Fusion-welded joints in steel, nickel, titanium and their
alloys. Quality levels for imperfections.”
ISO standards and working drafts

[9] ISO 10721-2: 1999 “Steel structures. Part 2: Fabrication and erection.”
[10] ISO/WD 17607 “Steel structures – Fabrication and erection.” Working draft of ISO/TC
167/WG3
[11] “NP 17607 – Provisional report of voting”, ISO/TC 167/WG3
Other references
[12] M-101:2011 “Structural steel fabrication.” NORSOK, Oslo
[13] Schmidt, H. et al. “Ausführung von Stahlbauten. Kommentare zu DIN EN 1090-1 und
DIN EN 1090-2.“ Beuth, Ernst und Sohn, Berlin 2012
[14] “Tillverkning, montering och kontroll. Handbok för tillämpning av SS-EN 1090-2.” (in
Swedish), SBI Pub. No. 182, Stockholm 2010
[15] Taras, A. “Column straightness requirements and stability design.” Proceedings Euro-
steel, Budapest 2011
[16] Måseide, M. “Norwegian proposals for amendments.” Oslo 2011
[17] “Agreement on Technical Cooperation between ISO and CEN (Vienna Agreement)”
VA codified – Version 3.3 2001, http://boss.cen.eu/ref/Vienna_Agreement.pdf
Abbreviations
CEN the European Committee for Standardization
CPD the Construction Products Directive (has been replaced by the CPR)
CPR the Construction Products Regulation
ECISS the European Committee for Iron and Steel Standardization
ISO the International Organization for Standardization
NORSOK Norwegian Standards for petroleum industrial activities, developments and op-
erations
40
Tampere, Finland
Fire design of steel structures with intumescent coatings
P. Schaumanna, F. Tabelingb, W. Weisheimc

a,c
Institute for Steel Construction, Leibniz University Hannover, e-mail: schaumann@stahl.uni-hannover.de
b
SHL Engineering Consulting, Hannover, e-mail: tabeling@shl-ing.com
Abstract: In this paper two different approaches for the prediction of the temperature of steel
structures with intumescent coatings are presented. The main objective is to provide the user
with a method to design the resistance of coated steel structures in fire. For this purpose, first-
ly a short overview of the experimental investigations is given, which are necessary to obtain
the required thermal material properties of intumescent coatings. Secondly, a two-dimensional
numerical model of a coated I-section profile is presented taking into account the foaming
process of the intumescent coating explicitly. Finally, a simplified approach based on Fouri-
er’s law is introduced as alternative to predict the temperatures of the coated profile in fire.
1 Introduction
Whenever filigree steel constructions are requested in combination with fire resistance re-
quirements, intumescent coatings (IC) are often used. The application of IC combines both the
maintenance of the filigree appearance of the steel construction and the required fire re-
sistance. Therefore, IC are mostly applied on roof constructions of sports arenas, industrial
halls and atriums as shown in Fig. 1.
Trapezoidal steel sheet Trussed girder with an IC
Reinforced column
Fig. 1: Roof construction of a sports arena composed of trussed girders with an intumescent
coating and adjacent trapezoidal steel sheets
Intumescent coatings are designed to build up a char structure when being exposed to fire.
Therefore, the thermal protection of IC is strongly related to the expansion process. To enable
the user to design the fire resistance of steel structures with intumescent coatings, us-
er-friendly methods for the prediction of the steel temperatures are aimed. Suitable methods
have been provided in the last ten years for example by di Blasi [1] and Staggs [2]. Although
di Blasi and Staggs developed mathematical and numerical models for the description of
physical processes, taking place within the char structure, or the effective thermal conductivi-
ty of IC, none of the methods include the foaming process of the IC explicitly. Therefore, a
51
numerical approach for a two-dimensional temperature field simulation as well as a simplified

approach for the prediction of the temperature of steel profiles with IC is presented in the fol-
lowing. Both, the numerical approach and the simplified approach are based on experimental
investigations on material properties of IC, performed by the authors.
2 State of the art

Since harmonised rules for the design of structural members are introduced in Europe, steel
structures with intumescent coatings can be designed according to EN 1993-1-2 [3]. The de-
sign rules are formulated for both protected and unprotected structures. In general, the user is
allowed to determine the fire resistance of steel structures by using simplified and advanced
computational models as well as by performing tests. The simplified calculation models in-
clude the determination of the fire resistance of structural members under tension, compres-
sion and bending actions such as tensile rods, columns and girders. In addition, the sufficient
fire resistance of structural members, which are not prone to stability problems, can be also
determined on the basis of a critical steel temperature. The evidence of the sufficient fire re-
sistance is proved, when the temperature of the investigated structural member is lower than
the calculated critical temperature. In order to apply this form of structural design, mathemati-
cal models for the prediction of the temperature of protected and unprotected structural mem-
bers based on Fourier’s law are given. The approach for protected members is only valid for
fire protective product with constant thicknesses. Therefore, no valid method for the determi-
nation of the temperature of steel structures with IC is available yet. Until now, the contribu-
tion of IC to the fire resistance of structural members can only be quantified by experimental
investigations according to test methods of EN 13381-4 [4]. Consequently, a more user-
friendly method for the determination of the thermal protective properties of IC is required.
3 Experimental investigations
In order to develop user-friendly methods for the prediction of the temperature of steel struc-
tures with IC, the authors performed experimental investigations on waterborne, solvent-
borne and epoxy resin-based IC. The investigations included the digital measurement of the
expansion factor (cf. Fig. 2a and 2b) as well as the heat capacity with a simultaneously meas-
urement of mass loss of IC using the differential scanning calorimetry (DSC) device illustrat-
ed in Fig. 2c. For the measurement of the expansion factor small steel plates (20 x 20 x 1 mm)
were coated with IC and arranged within an electric furnace.
Ambient temperature Temperature of IC about

Thermo- Scale 500 °C
couples
Test specimen Char structure
a) Initial configuration b) Expansion at 500 °C c) Measuring device for DSC

Fig. 2: Test setup for the experimental investigations on thermal material properties of IC
The test specimen where exposed to standard fire according to ISO-834. The expansion of the
IC was recorded by a digital camera using a scale. In order to formulate the expansion factor
52
as a function of temperature, the temperature of the char structure was measured using ther-
mocouples. Parts of the results are shown in Fig. 3. A detailed description of the test setup and
the evaluation of the results are presented in Tabeling [5] and Mensinger and Schaumann
et al. [6].
4 Numerical approach
The numerical simulation is performed using the finite element software Abaqus [7] in a fully
coupled thermal-stress analysis. To simulate the thermal behaviour of IC, a fully coupled
thermal-stress analysis is strictly necessary due to the fact, that the mechanical foaming pro-
cess of IC is influenced by the rising temperature field of the coating and vice versa. Moreo-
ver, a large-displacement formulation is used to describe the foaming process, in which the
elements are formulated in the current configuration using the current nodal position. In this
manner, the volumetric growth of the IC during the foaming process is considered within the
numerical approach. Since IC undergo a negative growth at temperatures higher than
ca. 600 °C, shrinkage processes are taken into account explicitly as well, thus considering the
high temperature behaviour of IC close to reality. To take the foaming and shrinkage process-
es of IC into account, a thermal expansion coefficient αT according to Eq. (1) is implemented
in Abaqus [7] considering logarithmic strains.
ln  di d0  ln  d0   d0  ln  
T    (1)
i   0 i   0 i   0
with: T : Thermal expansion coefficient (1/K)
d0 : Initial thickness of IC (mm)
di : Thickness of IC at time increment i (mm)
0 : Initial temperature of IC (°C)
i : Temperature of IC at time increment i (°C)
: Thermal expansion factor (-)
Equation (1) shows, that the thermal expansion coefficient αT depends on the temperature de-
velopment within the IC and a thermal expansion factor α. This factor is defined as the ratio
between the foam thickness of the IC at time increment i and the initial dry film thickness of
the IC. Therefore, the expansion factor is of major importance for the description of the foam-
ing process in Abaqus [7]. In order to determine the expansion factor, a large number of small
scale tests has been performed, as introduced in section 3. The measured expansion factor α is
shown in Fig. 3a as a function of temperature.
Since the expansion factor contains the volumetric change of IC, the factor is also elementary
for the development of additional material properties of IC, which are needed in a fully cou-
pled thermal-stress analysis. With regard to the mathematical formulation of the thermal con-
ductivity of IC based on the approaches of di Blasi [1] and Staggs [2], the porosity ψ of IC
can be described according to Eq. (2) as a function of the thermal expansion factor.
 1
 ( )  (2)

with:  ( ) : Porosity (-)
: Thermal expansion factor (-)
53
Regarding IC the porosity ψ describes the ratio between the volume of gas, which is trapped
within the pores, and the volume of the whole char structure. The mathematical evaluation of
Eq. (2) is depicted in Fig. 3a as a function of temperature.
Based on the above defined expression of porosity, the equivalent thermal conductivity λeq of
IC can be calculated according to Eq. (3) based on the approach of di Blasi [1] with regard to
the assumption of Staggs [2] concerning the thermal radiation within the pores.
eq ( )    ( p  4    IC 3  d p )  (1  )  IC (3)
with: eq : Equivalent thermal conductivity (W/m·K)

p : Thermal conductivity of the trapped gas within the pores (W/m·K)
IC : Thermal conductivity of IC at room temperature conditions (W/m·K)
: Porosity (-)
: Stefan-Boltzmann-Constant (W/m²·K4)
 IC : Temperature of IC (K)
dp : Diameter of the pores (m)
It is assumed, that the convection within the pores of IC has got a negligible effect on the
overall heat transfer within the char structure. However, the equivalent thermal conductivity is
strongly controlled by the thermal conductivity of the trapped gas within the pores, the ther-
mal conductivity of the char, which is assumed to be equivalent to the thermal conductivity of
IC at room temperature conditions, and by the radiation, which occurs within the pores. Based
on experimental investigations of Tabeling [5] on waterborne, solvent-borne and epoxy resin-
based IC the thermal conductivity of IC at room temperature conditions is assumed as
λIC = 0.45 W/m·K. The diameter of the pores, which effects the radiative amount of heat
transfer within the char structure, is set to dp = 1.2 mm. The equivalent thermal conductivity
of IC is shown in Fig. 3b as a function of temperature, taking the temperature-dependant po-
rosity ψ from Fig. 3a into account.
50 1.0 1.2 12000
Heat capacity C (kJ/m³·K)
Conductivity λ (W/m·K)
Expansion factor α (-)
40 0.8 1.0 10000

Porosity y ψ (-)
ψ α 0.8 8000
30 0.6 C
0.6 6000
20 0.4 λ
0.4 4000
10 0.2
0.2 2000
0 0.0 0.0 0
0 200 400 600 800 1000 0 200 400 600 800 1000
Temperature of IC (°C) Temperature of IC (°C)
a) Expansion factor and porosity b) Thermal conductivity and heat capacity
Fig. 3: Thermal material properties of IC
In order to formulate a material model for IC close to reality, the heat capacity of IC had to be
derived. Therefore, the data of DSC analyses, carried out by Tabeling [5], were evaluated and
translated to the heat capacity graph shown in Fig. 3b. The heat capacity takes into account
54
the change of density, resulted from the rise of temperature, explicitly. Therefore, the density
of IC can be assumed during the finite element simulation as constant ρIC = 1,400 kg/m³.
In order to perform a fully coupled thermal-stress analysis, mechanical material properties of
IC are needed as well. Therefore, the authors assumed the mechanical behaviour of IC as line-
ar elastic, using a young’s modulus of E = 1.0 N/mm² and a poisson’s ratio of ε = 0.0.
Based on this material model for IC, the designer is enabled to perform simulations on the
high temperature behaviour of IC, taking the foaming process of IC into account explicitly.
As an example, a finite element model of an I-section profile is modelled according to the ex-
perimental investigations of Tabeling [5] and Mensinger and Schaumann et al. [6]. Within
these experimental investigations a 1,100 mm long IPE 200 profile was exposed to standard
fire according to ISO-834 for 60 minutes. The test specimen was protected by a solvent-borne
IC with an averaged dry film thickness (dft) of 700 μm to achieve the fire resistance class of
R30. To ensure a continuous measurement of the steel temperatures, several thermocouples
were arranged in multiple rows on the upper and bottom flange as well as on the web. In
Fig. 4a two thermocouples (TC 1 and TC 2) are chosen to point out the temperature develop-
ment of the test specimen.
According to the test setup a two-dimensional finite element model of the coated IPE 200 pro-
file is developed. The model is discretised by CPE4T elements, enabling a two-dimensional
fully coupled thermal-stress analysis. The IC is discretised much finer than the steel profile,
thus enabling the coating to expand in layers and to cause a nonlinear temperature field within
the IC. In addition, the expansion of the IC is assumed to proceed strictly orthogonal to the
coated surface as shown in Fig. 4b. The material properties of steel are set according to
EN 1993-1-2 [3], whereas the thermal and mechanical properties of the IC are described by
the material model presented above.
The fire exposure according to ISO-834 is modelled with thermal coefficients for the emissiv-
ity of the IC εIC = 0.8 according to Tabeling [5] and convection αc = 25 W/m²K according to
EN 1991-1-2 [8].
IPE 200
IPE 200
TC 1
IC IC
TC 2
a) Test specimen b) Two-dimensional finite element model

Fig. 4: Coated I-section profile IPE 200 with a dft of 700 μm
The simulated temperature fields of the coated I-section profile are shown for different points
in time in Fig. 5. After 10 minutes of fire exposure a noticeable growth of the foam thickness
occurs, resulting in a temperature gradient of 343 K (IC: 641 °C, I-section profile: 298 °C)
between the fire exposed surface of the IC and the steel profile. During this time the maxi-
mum foam thickness amounts to 7.4 mm.
The thickness of the IC grow with further fire exposure, thus reaching the maximum after
22 minutes. At this moment in time the coating thickness exhibits a value of 20.9 mm. The
55
corresponding temperature gradient between the fire exposed surface of the IC and the steel
profile amounts to 339 K (IC: 791 °C, I-section profile: 452 °C). Since shrinkage processes of
the IC are considered as well, the coating thickness underlies a negative growth during the
remaining fire exposure. For that reason, the coating thickness amounts to 19.3 mm at the end
of the simulation. At this time the protective effect of the coating results in a temperature gra-
dient of only 63 K.
a) 10 minutes b) 22 minutes c) 60 minutes

Fig. 5: Simulated temperature fields of the coated I-section profile IPE 200 with a dft of 700 μm
In order to verify the predicted temperatures of the finite element model, the results obtained
by the numerical model are validated against the test data of Tabeling [5]. Therefore the com-
parison between the predicted and the measured steel temperatures of the IPE 200 profile are
discussed in the following. In Fig. 6 both, the measured and the simulated temperatures of the
profile are illustrated. It is apparent, that the web temperature as well as the upper flange tem-
perature is slightly overestimated by the finite element model in the first ten minutes, taking
the measured furnace temperature as a basis for fire exposure. Nevertheless, the difference
between the predicted and the measured temperature decreases with ongoing fire exposure.
After a fire exposure of 30 minutes the predicted upper flange temperature (576 °C) deviates
from the measured flange temperature (593 °C) only by 17 K. Also the deviation of 39 K (ex-
periment: 613 °C, simulation: 574 °C) between the web temperatures lies within tolerable lim-
its.
1000
Furnace temperature
Temperature (°C)
800
600
ISO-834
400
200
Experiment Simulation
0
0 10 20 30 40 50 60
Time (min)
Fig. 6: Comparison between the measured and simulated steel temperatures of I-section profile IPE
200 (dft: 700 μm)
56
Hence, the results of the conducted validation underline distinctively, that the new developed
material model for IC, which takes the foaming process of the coatings into account explicit-
ly, leads to reliable and very promising results for the prediction of two-dimensional tempera-
ture distributions of steel structures.
5 Simplified approach
The numerical simulation of the heating behaviour of the coated I-section profile IPE 200 pre-
sented in section 4 proves distinctively, that the new developed material model for IC leads to
very promising results. Nevertheless, the modelling process of the investigated structure, in-
cluding the temperature-dependant foaming process of the coating, is high sophisticated.
Therefore, the authors aimed to develop a simplified approach for the prediction of tempera-
tures of coated steel structures. The simplified approach is based on the assumption that the
temperature gradient inside the char structure of IC should be at each point in time linear as
shown in Fig. 7. Moreover, within the simplified approach the heat capacity of the IC is ne-
glected due to the thin dry film thickness.
θg,i
Thickness (mm)
IC
θa,i
Steel plate Temperature (°C)
Fig. 7: Assumed temperature gradient inside the char structure of IC
Based on this assumptions the simplified approach can be derived from the Fourier’s law
according to Eq. (4), where the heat flux is defined as a function of the temperature gradient
between the fire exposed surface of the coating and the steel surface as well as of the thermal
conductivity and the thickness of the IC.
 g , i   a ,i
q  IC ,i  (4)
d IC ,i
with: q: Heat flux (W/m²)
 g ,i : Gas temperature according to ISO-834 at time increment i (°C)
 a ,i : Steel temperature at time increment i (°C)
d IC ,i : Thickness of IC at time increment i (m)
IC ,i : Thermal conductivity of IC at time increment i (W/m·K)
Whenever a pysical body, described by its mass and its specific heat capacity, undergoes an
increase in temperature, the supplied thermal energy can be calculated according to Eq. (5).
Q  a,i  ma  ca (5)
with: Q: Supplied thermal energy (J)

 a ,i : Change in steel temperature at time increment i (K)
ma : Mass of the steel profile (kg)
ca : Specific heat capacity (J/kg·K)
57
In addition, the supplied thermal energy of Eq. (5) can be also described as heat flux due to
thermal conduction through the fire exposed surface A within a time increment ∆t according
to Eq. (6).
Q  q  A  t (6)
with: Q: Supplied thermal energy (J)
q: Heat flux (W/m²)
A: Fire exposed surface (m²)
t : Time increment (s)
The mass of a physical body, which is needed in Eq. (5), can be determined according to
Eq. (7). Therefore the density and the volume of the physical body should be known.
ma  a V (7)
with: ma : Mass of the steel profile (kg)
a : Density of steel (kg/m³)
V: Volume of the steel profile (m³)
Based on the Eq. (4) – (7) the temperature change of a coated steel profile can be calculated
incrementally using Eq. (8). The established calculation rule is similar to the equation for the
calculation of the temperature of steel profiles protected with boards or plaster, given in
EN 1993-1-2 [3]. Therefore, the authors indicate that the established calculation rule is only
valid, if the fire exposure corresponds to ISO-834. Moreover, for solving Eq. (8) the time in-
crement ∆t should not exceed the limit of 30 seconds.
A A
IC ,i  ( g ,i   a ,i )  ( g ,i   a ,i ) 
 a ,i  V  t  V  t (8)
d IC ,i  a  ca RIC ( )  d IC ,0  a  ca
with: RIC ( ) : Thermal insulation resistance of IC (m·K/W)

d IC ,0 : Initial thickness of IC (m)
With regard to Eq. (8) the incremental temperature rise of a coated steel profile is dependent
on the profile factor A/V and the heat capacity ρa·ca of the profile as well as on the thermal
conductivity and the variable thickness of the coating. The factor dIC,i/λIC,i indicates a standard
for the thermal insulation resistance RIC of the coating. In spite of calculating the insulation
resistance incrementally by using the expansion factor and the thermal conductivity of intu-
mescent coatings (cf. Fig. 3), the authors developed two approximation rules given in Eq. (9)
and (10).
200
RIC ( )  2,13  0,0284 IC ,i
for 20C   IC ,i  520C (9)
1  110250  e
RIC ( )  19  5200  e
0,0066 IC ,i
for 520C   IC ,i  1000C (10)
with:  IC ,i : Mean temperature of IC at time increment i (°C)

 g , i   a ,i
 IC ,i  (
)
2
The graph of the thermal insulation resistance is depicted in Fig. 8a. Therein the first range is
described by Eq. (9), whereas the second range is characterized by Eq. (10).
In order to verify the simplified approach, the heating behaviour of the I-section profile
IPE 200 is calculated and compared to the test data of Tabeling [5] as well as to the numerical
58
results from Fig. 6. The material properties of steel are set according to EN 1993-1-2 [3],
whereas the profile factor is calculated to A/V = 269 m-1 for a circumferential fire exposure
referring to ISO-834.
In Fig. 8b the measured, the simulated and the simplified calculated temperatures of the pro-
file are illustrated. Although the simplified calculated graph indicates a faster temperature rise
in the first ten minutes, the temperature profile of the I-section is predicted close to reality.
After 30 minutes of fire exposure time the simplified approach overestimates the measured
temperature by 7 K (simplified: 595 °C, measured: 588 °C). However, the simulated tempera-
ture (542 °C) is even overestimated by 53 K. Nevertheless, after 60 minutes of fire exposure
time all three approaches exhibit nearly the same steel temperatures (measured: 889 °C, sim-
plified: 886 °C, simulated: 883 °C).
250 1000
Range 1 Range 2 Furnace temperature
Insul. resist. RIC (mK/W)
200 800
Temperature (°C)
150 600
20°C ≤ θIC ≤ 520°C
Simulation
100 400
Simplified approach
50 200
520°C < θIC ≤ 1000°C Experiment
0 0
0 200 400 600 800 1000 0 10 20 30 40 50 60
Temperature of IC (°C) Time (min)
a) Thermal insulation resistance of IC b) Comparison of steel temperatures
Fig. 8: Input parameter and results of the simplified approach
To ensure a sufficient reliability of the simplified approach, besides the IPE 200 profile fur-
ther 10 profiles composed of IPE, HEA, HEB and HEM profiles were investigated. The pro-
files were chosen to cover the typical range of the profile factor as shown in Fig. 9a, thus re-
sulting in different dry film thicknesses of the coating (160 μm - 780 μm). As basis of com-
parison the heating behaviour of the coated profiles was simulated using the numerical ap-
proach presented in section 4.
300 1000 + 15 %
Simplified approach (°C)
Overestimation
Profile factor A/V (m-1)
250 800 60 minutes - 15 %

200
600
150 30 minutes
IPE 400
100 HEA
HEB 200
50
HEM Underestimation
0 0
0 2 4 6 8 10 12 0 200 400 600 800 1000
Profile number (-) Numerical approach (°C)
a) A/V-factor of investigated steel profiles b) Compariosn of steel temperatures
Fig. 9: Verification of the simplified approach
59
The comparison of the steel temperatures in Fig. 9b shows, that the simplified approach al-
ways overestimates the simulated steel temperature for the chosen examples. Nevertheless, for
both points in time (30 and 60 minutes) the simplified calculated temperatures do not exceed
the simulated temperatures by 15 %. As a consequence, the simplified approach offers a
promising alternative to the numerical simulation. For that reason both approaches provide
innovative methods to predict the temperature of coated steel structures, thus ensuring a relia-
ble fire design of steel structures with intumescent coatings.
6 Conclusions
In this paper two different approaches for the prediction of the temperature of steel structures
with intumescent coatings are presented. The main objective is to provide the user with a
method to design the resistance of coated steel structures in fire. Therefore, a two-dimensional
numerical model of a coated I-section profile is presented taking into account the foaming
process of the intumescent coating explicitly. Besides the numerical model, a simplified ap-
proach based on Fourier’s law is introduced as alternative as well. For validation purposes the
results of the numerical and the simplified approach are compared to own test data of a coated
I-section profile IPE 200. Since the results of both approaches show great accordance to the
measured steel temperature with only little deviation, the presented approaches constitute
promising methods for the prediction of the temperature of coated steel structures. Based on
the temperature values, obtained from the numerical or the simplified approach, the user is
enabled to design the resistance of any steel structure with intumescent coatings.
Acknowledgments
The results of this paper have been produced during the German research
project ‘Optimierter Einsatz intumeszierender Anstriche im Stahlbau’
(IGF 17200 N) from DASt. The project was funded by the ‘Federal Ministry
for Economic Affairs and Energy’ via AiF. The authors express their deep
gratitude.
References
[1] Di Blasi, C.; Branca, C.: Mathematical Model for the Nonsteady Decomposition of Intumescent
Coatings. American Institute of Chemical Engineers Journal Bd. 47. Wiley, 2001, P.2359–2370.
[2] Staggs, J. E. J.: Thermal conductivity estimates of intumescent chars by direct numerical simu-
lation. Fire Safety Journal Bd. 45. Wiley, 2010, P. 228–237.
[3] EN 1993-1-2: Eurocode 3: Design of steel structures – Part 1-2: General rules – Structural fire
design. Brussels, Belgium: Comité Europeén de Normalisation, 2010.
[4] EN 13381-4: Test methods for determining the contribution to the fire resistance of structural
members – Part 4: Applied passive protection to steel members. Brussels, Belgium: Comité Eu-
ropeén de Normalisation, 2013.
[5] Tabeling, F.: High temperature behaviour of intumescent coating on steel constructions (in
German: Zum Hochtemperaturverhalten dämmschichtbildender Brandschutz-systeme im
Stahlbau). Hannover, Leibniz Universitaet Hannover, Phd-Thesis, 2014.
[6] Mensinger, M.; Schaumann, P.; Kraus, P.; Tabeling, F.: Optimised applications of intumescent
coatings on steel elements (in German: Optimierter Einsatz intumeszierender Anstriche im
Stahlbau); Deutscher Ausschuß für Stahlbau e.V. 2014. Research report.
[7] ABAQUS: Abaqus/Standard Version 6.10. Pawtucket: Hibbit, Karlsson & Sorensen, Inc. 2011.
[8] EN 1991-1-2: Eurocode 1: Actions on structures – Part 1-2: General actions – Actions on struc-
tures exposed to fire. Brussels, Belgium: Comité Europeén de Normalisation, 2010.
60
Tampere, Finland
23–25 September 2015
ENERGY-EFFICIENT SOLUTIONS FOR STEEL STRUCTURES –

CASE STUDY OF NEARLY ZERO-ENERGY BUILDING
Jyrki Kesti
Ruukki Construction Oy
Abstract: Finland’s first nearly zero-energy single-storey commercial/industrial building was

completed in the spring of 2015 in Hämeenlinna. Built on the campus of Häme University of
Applied Sciences (HAMK), the building is used for research, development and teaching pur-
poses by the university, Ruukki Construction and HAMK’s Sheet Metal Centre. The purpose
of the construction project was to show that a building exceeding today’s strict energy-
efficiency requirements by over 50 per cent can be built at a profit. Construction of the build-
ing, totalling approximately 1,500 m2, started in May 2014. The project was executed by
HAMK and Ruukki Construction.
1 Introduction
Energy efficiency has risen to the same level as construction quality and cost efficiency to
become one of the most important factors guiding construction projects. Investments in ener-
gy efficiency have already been made, particularly in residential and office construction. The
guiding principle behind product development at Ruukki Construction has been to develop
products and construction solutions that improve business, industry and logistics through en-
ergy efficiency. In Ruukki’s concept, buildings must be designed and executed as complete
entities – not split up into subareas. This approach is almost contrary to present-day construc-
tion, in which design and build are split up across several parties without the overall entity
being properly managed. The execution of commercial, industrial and logistics buildings
should also be managed by, for instance, so-called alliance agreements, in which the parties
involved are bound to share responsibility for executing buildings in accordance with custom-
er requirements (in addition to technical cooperation).
More complex construction requirements, such as cost efficiency, quality, energy efficiency
and environmental friendliness, also underline the need to plan and manage entities as a
whole. Finland’s first nearly zero-energy “big-box” type single-storey building (nZEB) for
commercial, logistic or industrial use was designed and constructed to meet an objective: to
be a building with an economical lifecycle that saves energy and uses existing renewable en-
ergy sources (See Fig. 1). The new structure was designed and executed to enable economic
use of the building and optimization of construction solutions. Optimization means selecting
solutions based on investment outlays, additional usage costs and future savings. A well insu-
61
lated and airtight envelope in the building’s walls and roof enables savings in energy re-
quirements, and the use of solar power, day-lighting and geothermal energy harness renewa-
ble energy for use in the building. The building’s economic performance was estimated by
comparing the investment costs and life-cycle costs of a reference building and an nZEB
building. The reference building level was agreed with the customer and designers.
Fig. 1: Completed nearly zero-energy building in Hämeenlinna, Finland
2 Technical solutions for the nearly zero-energy building

2.1 Building envelope
The shell of the building walls and roof are highly significant for its energy efficiency. For
this reason, the outer walls of the building are fitted with Ruukki’s energy panel system, with
ultra airtight panels and carefully designed and executed seals between the panels, plinth,
roof, windows and doors. The energy panels are based on sandwich panels with an insulating
layer between two thin steel sheets. Five colours of energy panels totalling 1,520 m2 were
used in the building. For the first time, module moulding of various colours was incorporated
between the panels, producing extensive variety in the façade. The thickness of the insulation
in both the wall and the corner panels is 230 mm, with a U-value of 0.16 W/m2K. In the west
gable of the building, a graphic image has been attached to the façade. In this energy-efficient
building, as shown in Fig.2, a cloudy sky pattern illustrates the building’s positive take on the
environment. The building’s roof incorporates a new type of prefabricated PIR roof elements
with a U-value of 0.12 W/m2K. The measured airtightness of the entire building was as low as
q50 = 0.76 m3/h,m2.
62
Fig. 2: Energy panels with decorative printing.
2.2 Utilization of day-lighting

The sizes and directionality of the building’s windows are optimized for energy efficiency.
The large windows are aimed south and west. The need for artificial lighting is reduced by the
windows, due to their directionality and surface area. Traditional large windows bring light in
– but also conduct heat out. In this building, the glass windows facing south have been re-
placed by cell windows made of polycarbonate (Fig. 3). These “daylight” windows isolate
heat well – the warm rays of the sun during the summer do not heat the premises. During pe-
riods of bright daylight, light coming through traditional windows in indoor areas causes
glare. Instead of this, daylight windows distribute light into the premises in a pleasantly even
manner without glare, and blinds are not needed. The building is equipped with a day-lighting
control to reduce artificial lighting. The U-value of these day-lighting windows is approxi-
mately 0.8 W/m2K. The north wall of the building incorporates Ruukki Construction energy
panel system clear windows, forming a dense structure with the panels.
Fig. 3: Installation of day-lighting windows
63
2.3 Heating, cooling and ventilation systems

New types of radiation-based heating and cooling profiles developed by Ruukki Construction
are installed in the building. The profiles are affixed to the underside of the roof element as
shown in Fig 4. The radiation they generate either cools or heats the interior, depending on
the season and the desired indoor temperature of the building. Radiant profiles work with a
low temperature difference to the ambient air, allowing the heat pump installed in the building
to perform well. The new control system reduces temperature variations on each floor in the
building, thereby considerably increasing usage comfort and improving well-being at work
and productivity. A new type of indoor heating and cooling system also reduces energy con-
sumption compared to air-heating systems. A ventilation airflow is now required only for the
influx and removal of fresh air – not actually for heating the premises. The mechanical venti-
lation machine is equipped with an 80% heat recovery system.
Fig. 4: Ruukki radiant profile and installation
2.4 Renewable heating energy system

Geothermal energy is utilized for the building’s heating and cooling requirements. In total, 64
Ruukki Construction energy piles of 11m in length under the floor and columns are incorpo-
rated in the foundation to use renewable energy to heat the building. The energy pile system is
based on steel foundation piles, heat-collecting pipes installed in the piles, connecting pipes
via manifolds to the heat pump, and heat-transfer liquid. Fig. 5 shows the heat collector pipes
installed. Furthermore, two conventional heat wells of 200 m in length were installed for heat-
ing and free cooling of the building. The heat pump capacity is 35 kW.
Fig. 5: Heat collector pipes installed in the floor slab piles
64
A total of 24m2 of Ruukki Classic solar collectors are installed on the roof of the building’s
technical area,. The Classic solar system integrates fully with the roof, as shown in Fig. 6. So-
lar collectors accumulate thermal energy from the sun and transfer it to the soil through the
energy piles. The soil is charged whenever there is heating potential available – even in Janu-
ary, thanks to the very low temperature level in the ground.
Fig. 6: Roof-integrated solar heat collectors
The soil acts as a seasonal thermal reserve, much like a battery. Beneath the building, a clay
layer extends to a depth of 11 metres. Clay has a greater thermal storage capacity than, for
example, gravel. In the winter, the piles transfer energy from the soil to heat the building.
During the summer, the energy pile loop is closed from the heat pump, and is charged by the
solar collectors only. Cooling of the building is via the deep heat wells by free cooling utiliz-
ing the low temperature of the ground rock. The principle of the system is shown in Fig 7.
Fig. 7: Ground energy system
65
2.4 Building-integrated solar electricity solution

Solar power is also used in the outer walls of the building. Ruukki Construction’s on-wall so-
lar panels, which generate electricity from the light of the sun for the building’s network, are
installed on its southern façade. A total of 61 m2 PV (Photovoltaic) panels with total peak
power of 10 kW are incorporated in the wall (see Fig. 8).In addition to the panels, the entity
includes mounting systems and electronic components as well as grid connections.
Fig. 8: Building-integrated solar PV panels installed on the southern facade
2.5 Monitoring of the building

The building is equipped with a large number of energy meters and other measuring devices
to ensure extensive monitoring and ascertain the real energy performance of the building. In
particular, the energy pile and solar heat systems are monitored carefully to study the soil be-
haviour in the long run. Some building elements are also equipped with thermal-moisture sen-
sors in order to monitor their condition throughout their life-cycle. Furthermore, the amount
of snow on the roof is also monitored with a Ruukki Smart Roof application based on strain
gauge measuring of the metal roof structure.
3 Building simulations
The energy efficiency of the reference building and the nZEB were determined by energy
simulations with the IDA ICE 4.5 program [1]. The simulations were performed by Ruukki
and Tallinn Technical University. The reference building represents the current normal, al-
ready very energy-efficient, construction custom. A simulation model included a model of the
building with the structures and technical systems as well as the energy piles and heat wells.
The initial data for the reference case and the nearly zero-energy case are given in Table 1.
66
Table 1: Initial data for simulated cases

Reference nZEB
Wall, U-value 0.17 0.16
Roof, U-value 0.09 0.12
Window, U-value 1.0 1.0
Floor EPS 150, ʎ=0.034 EPS 150, ʎ=0.034
Infiltration q50 4 m3/ m2 h 0.87 m3/ m2*
AHU Heat Recovery 50% 80%
Lights 15 W/ m2 10 W/ m2 (LED)**
Domestic Hot Water 68 l/ m2 a 68 l/ m2 a
People 30 30
Fresh air (SFP=2.0) 1.5(2***) l/sm2 1.5
Temp. set points 18°C /25°C 18°C /25°C
Heating system Air-heating Radiant heating
Energy source District heating Ground heat pump
Cooling EER 2.5 free cooling
Schedule 8–17 weekdays 8–17 weekdays
*initial value used in calculations, measured value 0.76
**includes only lighting system, day-lighting control not considered
***overall rate with air-circulation in air-heating case
The final results for the energy demand and the delivered energy are shown in Tables 2 and 3.
A comparison of delivered energies of the cases is given in Fig. 9. The results show that it is
possible to halve the total energy use of a building with smart design and solutions.
Table 2: Energy demand and delivered energy of the reference case

Energy demand Delivered energy
kWh kWh/m2 kWh kWh/m2
Heating 88272 59.0 88272 59.0
Cooling 8502 5.7 3401 2.3
Fans electricity (SFP=2.0) 18498 12.4 18498 12.4
Pumps electricity 79 0.1 79 0.1
Lighting 51178 34.2 51178 34.2
DHW 5918 4.0 5918 4.0
Total electricity: 73156 49
Total distr. heat: 94190 63
Table 3: Energy requirements and energy inputs of the reference case

Energy demand Delivered energy
kWh kWh/m2 kWh kWh/m2
Top-up heating 1998 1.3 1998 1.3
Heat pump 56500 37.8 13320 8.2
Cooling 8502 5.7 0 0
Fans electricity (SFP=2.0) 14302 9.6 14302 9.6
Pumps electricity 6254 4.2 6254 4.2
Lighting 34109 22.8 34109 22.8
DHW 5918 4.0 5918 4.0
Total electricity: 74901 50
67
Fig 9: Comparison of deliverd energies of the reference case and nZEB case
Furthermore, the annual yield of the building-integrated solar PV panels is approximately

7000 kWh/a. Thus the need for delivered energy decreases further by approximately 5
kWh/m2, corresponding to a decrease in a primary energy use of approximately 10%.
As yet there are no official requirements for nearly zero-energy levels in Finland, but the
proposals made in the national “FInZEB” project [2] indicate that this building would clearly
meet future targets.
4 Economic feasibility studies

Economic feasibility studies and comparisons between the two cases were carried out by an
independent Finnish third-party, FIRA Oy. The economic calculations took into consideration
all investment costs that differed in the two cases as well as future energy savings due to im-
proved energy efficiency. The net present values of the future savings were determined based
on a 6% interest rate and a 4% increase in the energy price. The initial price of electricity was
85 €/MWh and 65 €/MWh for district heating. A 33% residual value for the investments was
used in the calculations. All prices are excluding VAT. The results are shown in Fig. 10. As
Fig. 10 shows, the nearly zero-energy solution is economically reasonable, with a payback
period of around 10 years. Also, it should be noted that the real extra investments of the nZEB
solution are only about 2% of the total construction costs. Solar PV installations are not in-
cluded in the studies, because their impact is easy to separate from the overall building energy
performance. The separately calculated payback for solar PV installations is approximately 15
to 20 years.
68
Fig 10: Net present values for the nearly zero-energy building compared to the reference
building
5 Conclusions
The following main conclusions may be drawn:
1. The execution of nearly zero-energy buildings should be managed by, for instance, so-
called alliance agreements, in which the parties involved are bound to share responsibil-
ity for producing buildings in accordance with customer requirements.
2. The nearly zero-energy building here is technically feasible if the energy demand of the
building is reduced and renewable energy sources are utilized in a smart way.
3. Nearly zero-energy buildings can be constructed in a Nordic climate in a cost-efficient
way.
4. The extra costs of energy efficiency may be very low if the building is optimized as a
whole – not via sub-optimizing.
Acknowledgments
The contributions of all the project partners are gratefully acknowledged.
Notation
AHU Air Handling Unit
DHW Domestic Hot Water
EER Energy Efficiency Ratio
HR Heat Recovery of the ventilation system
q50 Air-tightness of the entire building [m3/(h·m3]
SFP Seasonal Factor of Performance
U-value Heat conductivity [W/(m2·K)]

References
[1] EQUA Simulation AB, “IDA Indoor Climate and Energy V 4.5.” [Online]. Available at
http://www.equasolutions.co.uk/en/software/idaice
[2] FInZEB, project for nearly zero-energy definitions in Finland. www.finzeb.fi
69
70
Tampere, Finland
JOINT AND COLUMN BEHAVIOUR OF SLOTTED COLD-FORMED

STEEL STUDS
Michael J. Andreassena and Jeppe Jönssonb

a,b
Technical University of Denmark, Department of Civil Engineering,
Brovej, Building 118, DK-2800 Kgs. Lyngby, Denmark
*
Author for contact. Tel.: +45 45251814, E-mail: mican@byg.dtu.dk
Extended abstract
Slotted cold-formed steel studs are used in load bearing external plasterboard walls. The cold-
formed steel studs in these walls are supported by and joined to track profiles at the bottom
and top level. The slots in the web of the studs considerably change the transverse bending
and shear stiffness of the web. This has influence on the local and distortional buckling be-
haviour of the stud and thus also on the behaviour of the joints. A small number of papers on
the strength of load-bearing slotted cold-formed steel members such as [1,2,3,4] have been
published. Furthermore only few papers have been published regarding the joints between
slotted load-bearing cold-formed steel members such as [5].
In order to observe the behaviour of the studs and joints between slotted cold-formed steel
members, this paper presents several tests of studs and joints used for external bearing walls
in compression (without influence of the plasterboards). The experiments are performed in
cooperation with a manufacturer and include a joint design made and used by this manufac-
turer. The studs are C-profiles and the tracks are U-profiles, such that the studs fit into the U-
profile, i.e. track profile. Eight different test series are performed, having different column
lengths, thicknesses, and are both assembled with and without web stiffeners to see the influ-
ence of these on the behaviour and load capacity. Relative short column lengths are used in
order to be able to investigate the behaviour of the joints and not only column buckling. The
studs and tracks are connected to the tracks using four self-tapping screws per joint, two in
each side of the flange.
The conclusion regarding the maximum load capacity is that there is no effect of using the
special web stiffener. Regarding the failure modes the observed behaviour is that almost the
same failure modes occurs with and without joint stiffeners. Even for relatively short and thin
columns, with a length of 1000 mm and a thickness of 1.0 mm and 0.7 mm, respectively, the
failure configuration is an interactional global-distortional stability failure of the flanges close
to the middle of the column. This is also the case for columns with a length of 500 mm, a
thickness of 1.0 mm and without web stiffeners.
71
Having a column length of 500 mm with a thickness of 1.0 mm and with web stiffeners as
well as a column length of 500 mm with a thickness of 0.7 mm without and with web stiffen-
ers, the failure configurations is a combination of an end failure and an interactional global-
distortional stability failure close to the middle of the column. Different local end failures are
shown in Fig. 1 without web stiffeners and in Fig. 2 with web stiffeners.
Fig. 1: End failures, without web stiffener.
Fig. 2: End failures, with web stiffener.
References
[1] Thöyrä T. “Strength of slotted steel studs”, PhD thesis, Royal Institute of Technology,
Department of Structural Engineering, Stockholm, Sweden, 2001.
[2] Höglund T. “Beräkning av slitsad tunnplåtsregel”, Report 42, Royal Institute of Tech-
nology, Department of Structural Engineering, Stockholm, Sweden, 1999.
[3] Borglund J., Jonsson J. “Bärförmåga för slitsade stålreglar”, M.Sc. thesis 84, Royal In-
stitute of Technology, Department of Structural Engineering, Stockholm, Sweden,
1997.
[4] Kesti J., Mäkeläinen P. “Compression behavior of perforated steel wall studs”, Light-
Weight Steel and Aluminium Structures, 123-130, Finland, 1999.
[5] Costa M.M. “Support Strength of walls with slotted studs”, M.Sc. thesis 127, Royal In-
stitute of Technology, Department of Structural Engineering, Stockholm, Sweden,
1999.
72
Tampere, Finland
STEEL SOLUTIONS FOR ENABLING ZERO ENERGY BUILDINGS
Bernd Döringa,*, Vitali Regerb, Markus Kuhnhennec, Jyrki Kestid, Mark Lawsone and Andrea
Bottif
a
FH Aachen University of Applied Sciences
b,c
RWTH Aachen University, Institute of Steel Construction
d
Ruukki Construction Oy
e,f
University of Surrey, Department of Civil and Environmental Engineering
*
Author for contact. Tel.: +49 241 6009-51213; E-mail: doering@fh-aachen.de
Abstract: The European Directive on the Energy Performance of Buildings (EPBD) obliges
the member states to ensure that by 31 December 2020, all new buildings are nearly zero-
energy buildings (nZEB). This paper presents solutions for steel intensive commercial build-
ings that achieve this requirement. Several key components such as façades, floor systems,
steel piles for ground energy storage were investigated in detail by numerous numerical simu-
lations and practical tests of selected options. Furthermore, options for a whole building,
which fulfil the approach of a "zero energy building", were identified for different European
climates by performing a parametric study using a thermal building simulation tool.
1 Introduction
The objective of the EU-funded project ZEMUSIC (Zero energy solutions for multifunctional
steel intensive commercial buildings), which is the background of this paper, is to address
ways in which significant energy reductions can be made in the operation phase of multi-
storey commercial buildings with solutions using steel elements. The zero energy approach
requires a combination of energy conservation and energy generation techniques. The focus
will be on systems where the building fabric and structure participates actively in the energy
balance of the building, and therefore reduces the building’s energy demand. The commercial
target is to gain market share for steel intensive low and zero energy solutions.
2 Concept of reference building

The investigations of the energy performance of a multi-storey office building were made us-
ing a real building as a reference plan form. The building has six storeys and offers a net floor
area of 9400 m². This building was ‘virtually’ placed in three different European locations:
North (Helsinki), west/maritime (London) and south (Bucharest). For the thermal simulations,
weather data-sets on an hourly base were taken for these three sites.
This geometrical model was combined with different build-ups of facades (U-Value, transpar-
ency), different deck systems and with various solutions regarding heating and cooling using
a heat pump (reversible). Furthermore, this building was equipped with PV-elements on the
73
roof and the façade to achieve a zero-energy building by balancing energy use and generation
over the year.
3 Integrated flooring systems

The structural system is based on 17 m span fabricated beams that span across the building
and are placed at 7.5 m to 8.1 m spacing along the building.
The 200 mm deep double layer flooring system spans up to 7 m between secondary beams
and is provided with 100 mm concrete topping. Certain ribs in the decking system are con-
crete filled and reinforced as a series of T-sections. The double layer deck system (fig. 1) of-
fers a number of options for improving the energy efficiency of an office building
Fig. 1: Role of double layer flooring element for service integration
The performance of the double layer flooring element with integrated radiant heating / cool-
ing was tested numerically and by measurement of test specimen in a test rig.
4 Steel piles for ground heat storage

In the energy pile a vertical heat exchanger system is installed inside a steel pile. The heat ex-
changer is usually made of PE, the inner construction could be coaxial, single U or double U
system. After the pipes have been placed to a designed position the inside of the pile is grout-
ed. The energy pile system works in similar way as a traditional ground heat well and can be
utilized for heating with help of heat pump and for cooling either with free ground cooling or
with help of a chiller. An advantage of the energy pile system is a good thermal storage ca-
pacity under the building due to large number of piles near each others.
5 Whole building energy performance

For the final determination of energy consumption numerous simulations were performed.
The challenge was the optimization of the energy supply system for the nearly zero energy
solution, at which a pre-selection of the components was made.
The goal "nZEB" was reached by an integral strategy: Optimization of the building envelope,
introduction of heat storage (building fabric and ground), efficient HVAC-solutions and on-
site energy generation by PV and solar-thermal collectors. The simulations showed, that the
energy demand (only electricity) can be reduced to about 50 kWh/m²a, at which the demand
for heating, cooling and ventilation amounts only about 10 kWh/m²a (for all investigated cli-
mates). On the other hand, the energy productions is about 15 to 20 kWh/m², so the energy
demand for HVAC can be covered, and a substantial part of the total energy demand can be
covered be renewables, hence the nZEB-approach is fulfilled.
74
Tampere, Finland
PLASTIC RESISTANCE OF COMPOSITE SLABS

IN PARTIAL SHEAR CONNECTION
Anna Palisson
SOKOL-PALISSON Consultants, Paris, France
Leopold SOKOL
SOKOL Consultants, Guyancourt, France
EXTENDED ABSTRACT
The use of profiled steel sheeting for concrete slabs started in the United States, in the 1920s, used at
the beginning as shuttering, without taking into account of any collaboration with concrete.
The first calculation method for composite slabs, considering the steel-concrete collaboration, was an
empirical, so-called "m-k" method, formulated at the end of the 1960s and refined at the beginning of
the 1970s.
The partial shear connection method of composite slabs design was developed at the end of the 1970s.
Since then, much theoretical and experimental research has been performed on this concept. Based on
this researches, especially the one carried out at TNO -IBBC in the Netherland, the partial connection
method has been added to the EN 1994-1-1, in parallel to the formerly existing m-k method.
The advantage of the partial connection method relies on the fact that it is based on a rational
mechanical behaviour model.
However, some arbitrary approximations were used for the calculation model adopted in the EN 1994-
1-1, concerning the behaviour of the sheeting acting as sagging reinforcement for the slab.
In particular it concerns the contribution of the flexion resistance of the sheeting to the global flexion
resistance of the slab.
The resistance moment of the composite section according to this model is decomposed into two
terms: MRd = Mcr + Mpr
The first term in this equation is calculated using an arbitrary approximate expression for the lever arm
z of the internal forces.
The second term in this equation is defined according to the arbitrary adopted approximate expression
that assigns an approximate bilinear function to the interaction between the tension and the plastic
resistance moment of the sheeting
This law of discontinuous behaviour is not justifiable from the point of view of mechanical behaviour.
The plastic resistance moment of the sheeting is calculated as for section Class 1 within the meaning
of article 5.5.2.1 of the EN 1993-1-1, having a full capacity of forming a plastic hinge without strength
reduction.
This assumption is not realistic considering that it concerns the thin-walled sections.
As observed during the testing in the ultimate stage of loading, the collapse of the profile does not
occur by a plastic yielding of the section but by a local buckling of the compressed walls.
In this paper, firstly is presented the verification of composite slab resistance according to the partial
connection method of EN 1994-1-1, with some critical observations.
75
Then is developed an original formulation for the plastic bending resistance of composite slabs, which
is more accurate than that given in EN 1994-1-1 as:
- The formulation EN 1994-1-1 does not take into account that the thin-walled steel decking do not
have the ability to form a perfect plastic hinge without reduction of resistance of compressed section
part. In order to keep the main line of analysis used for the EN 1994-1-1 behaviour model, it is
similarly assumed that the stress diagram in the components of the section (steel and concrete) has a
rectangular block form, however with a reduced resistance in compression zone of sheeting by mean
of a reduction factor  < 1.
- The proposed formulation provides a more accurate solution for the lever arm of internal forces,
avoiding the use of arbitrary approximation adopted in EN 1994-1-1.
The presented analysis shows that the bending resistance of composite slabs defined according to the
current version of the EN 1994-1-1 is overestimated for small values of the connection degree (i.e.
near of the support). This may affect the safety especially where concentrated loads are applied near
the support.
Finally, it should be noted that the proposed modification improve some particular aspects of the
partial connection method model adopted in the current version EN 1994-1-1, without questioning the
general principle of its behaviour model.
The aim of the present paper is to improve this approach on the basis of more advanced equilibrium
model, avoiding the above mentioned arbitrary approximations.
The theoretical investigations are completed with experimental tests.
As a result of this paper, an improvement of the EN 1994-1-1 formulation for partial connection
method is proposed.
76
Tampere, Finland
FUTURE DESIGN PROCEDURE FOR STRUCTURAL CONNECTIONS

IS
COMPONENT BASED FINITE ELEMENT METHOD
František Walda*, Luboš Šabatkab, Jaromír Kabeláčb, Lukáš Gödricha and Marta Kurejkováa
a
Czech Technical University in Prague
b
IDEA RS
*
Author for contact. Tel.: +420 224 354 757; E-mail: frantisek.wald@fsv.cvut.cz
Abstract: This paper introduces component based finite element model (CBFEM) which is
a new method to analyse and design connections of steel structures with features demonstrat-
ed here on a portal frame eaves bolted connection. The connection in CBFEM procedure is
analysed by FEM method. The proper behaviour of components is treated by introducing a
components representing well is behaviour in term of initial stiffness, ultimate resistance and
deformation capacity, of bolts, welds etc. As for another FEM design procedure a special care
needs to be given the Validation and Verification procedures which is demonstrated
on example of portal frame eaves welded connection.
1 Introduction
Four decades ago computational analysis of structural connection was treated by some re-
searchers as a non-scientific matter. Two decades later it was already a widely accepted addi-
tion or even extension of experimental and theoretical work. Today computational analysis,
in particular computational mechanics and fluid dynamics, is commonly used as an indispen-
sable design tool and a catalyst of many relevant research fields. The recommendation for
design by advanced modelling in structural steel is already hidden but ready to be used in
Chapter 5 and Annex C of EN 1993-1-5:2005. Development of modern general-purpose soft-
ware and decreasing cost of computational resources facilitate this trend. As the computa-
tional tools become more readily available and easier to use, even to relatively inexperienced
engineers, more scepticism and scrutiny should to be employed when judging one’s computa-
tional analysis. The only way to prove correctness of simulated results is through a methodi-
cal verification and validation process. Without it the analysis is meaningless and cannot be
used for making any decisions also in connection design. In the case when the analysed event
is too complex or overly expensive to test experimentally, hierarchical validation is recom-
mended.
2 Connection behaviour by CBFEM
The advantages of FEM analyses of steel plates may be documented on behaviour of a well-
designed portal frame eaves moment bolted connection developed based on US best practice
and applied in good European practice represented by British and German design books. The
77
rafter of cross section IPE 400 column is connected to column HEA 320 by end plate 20 mm
on full depth of connection by 12 bolts M20 8.8. The rafter 1200 mm long is cut from the IPE
400. The stiffeners are designed from P20. Material S355. The results of analyses show in
Fig. 1 the development of plastic zines in connection by CBFEM analyses, from first yielding
under the tensile bolt, through development of full plasticity in the column web panel in shear
till reaching the 5 % strain in panel.
b)
a)
d)
c)
Fig. 1: Development of plastic zines in connection by CBFEM analyses, from firs yielding
under the tensile bolt a), through development of full plasticity in the column web panel load-
ed in shear, till reaching the 5 % strain in panel d)
5 Conclusion
As the global analyses of steel structures is today carried out by FEM and all the traditional
procedures are not used any more. Very soon will be designed the connections by advanced
procedures, like the component based finite element method instead of today used curve fit-
ting for hollow sections joints and component method for open sections joints.
Acknowledgments
The work was prepared under the R&D project MERLION supported by Technology Agency
of the Czech Republic, project No. TA02010159.
78
Tampere, Finland
COMPARATIVE EVALUATION OF STEEL PROFILES IN ROOF

TRUSSES
Kristo Mela, Hilkka Ronni and Markku Heinisuo
Tampere University of Technology, Department of Civil Engineering, Tampere, Finland
Abstract: In this study, utilization of different steel profiles in roof trusses is considered. Cold-
formed rectangular hollow sections are compared with a combination of hot-rolled I-sections
and cold-formed circular hollow sections. The comparison is carried out by performing topol-
ogy optimization for given span, truss height and load. In topology optimization, the optimum
number of members and their connectivity is determined along with member cross-sections.
This allows each cross-section type to be utilized most efficiently, enabling a fair comparison.
Minimum cost trusses are computed. The cost of the profiles and surface treatment are included
in the cost function.
1 Introduction
One of the key decisions in truss design is the choice of member profile shapes. Depending on
the application, some shapes are more economical than others, but it is difficult to know a priori,
which profile shapes yield the most economical truss. Moreover, the topology of the truss has
a great impact on the economy, and the optimum topology depends on the profile alternatives.
Thus, for different profile shapes, the optimum topologies can be different.
In this study, the economy of two groups of profile shapes is evaluated for roof truss with typical
span and load. The first group consists of hot-rolled I-profiles (HEA/HEB) of S460 steel and
cold-formed circular hollow sections (CHS) of S355 steel. The second group consists of cold-
formed square hollow sections (SHS) of S420 and S355 steels. For both groups, the higher steel
grade is used in the chords, whereas the braces are made of the lower grade steel. Topology
optimization is performed for both profile groups. The objective is to minimize the cost of the
truss. The cost function takes into account the manufacturing of the profiles, surface treatment
(blasting and painting), and quality control.
2 Problem description
The ground structure approach for topology optimization [1] is employed and applied to the
design domain of the roof truss, shown in Fig. 1. Three different spans are considered, and the
height of the truss is varied along with the span. The magnitude of the uniform load is the same
for all spans.
79
1 : 20
h
α
L
Fig. 1: Design domain of the roof trusses.
For each span, the optimum topology is determined for three different ground structures. Mem-
ber strength and joint geometry rules of Eurocode 3 [2, 3] are included in the problem as con-
straints. The resulting optimization problem is a mixed-integer linear programming problem [4],
where all discrete variables are binary. This problem can be solved to global optimum using a
contemporary desktop machine.
3 Results
The computations show that the optimum topology depends on the span, profile selection and
the ground structure. In all solutions, high (over 0.9) utilization ratios are obtained for many
members of the truss, which indicates efficient use of material.
Trusses with HEA/HEB chords and CHS braces yield slightly more economical solutions than
SHS trusses. However, the differences in cost become smaller as the span is increased. As the
cost function employed is rather coarse, it can be concluded that for a different (more detailed)
cost function and a wider range of steel grades, SHS profiles might prove to be more economical.
Thus, it can only be recommended that an optimization procedure as described here is routinely
carried out in order to determine the most economical solutions.
References
[1] Dorn W., Gomory R., Greenberg M. ”Automatic design of optimal structures”. Journal
de Mécanique, 3, 25–52, 1964.
[2] EN 1993–1–1, Eurocode 3: Design of Steel Structures. Part 1-1: General rules and rules
for buildings. CEN, 2005.
[3] EN 1993–1–8, Eurocode 3: Design of Steel Structures. Part 1-8: Design of joints. CEN,
2005.
[4] Rasmussen M., Stolpe M. ”Global optimization of discrete truss topology design prob-
lems using a parallel cut-and-branch method”. Computers and Structures, 86, 1527–1538,
2008.
80
Tampere, Finland
NON-LINEAR FINITE ELEMENT MODELLING OF STEEL-

CONCRETE-STEEL MEMBERS IN BENDING AND SHEAR
Marc Donnadieua, Ludovic Fülöpb

a
Institut Français de Mécanique Avancée, IFMA
b
VTT Technical Research Centre of Finland
Abstract: Steel-Concrete-Steel (SC) construction comprises two steel plates with concrete
infill. SC has structural advantages and leads to faster construction compared to classical
reinforced concrete solutions, and advantage especially important in the industrial sector dif-
ficult to build environments. In this study we developed a general purpose modelling tool for
evaluating the bending and shear strength of SC members. We noted that the numerical
modelling techniques found in the technical literature cannot usually predict the general
behaviour of SC members. Hence this work aimed to develop a single finite element model
(FEM), which is able to predict all failure modes relevant to SC.
1 Introduction
This work aims to develop an FE model which is able to predict all types of failure modes in
SC structures: steel plate yielding, steel plate buckling, de-bonding of tensile sheet driven by
stud shear failure, concrete crushing and vertical or horizontal shear failure.
Methodologically, we started from a very simple model and then improved it gradually as the
shortcomings of the simpler models were revealed. During this process, it has been shown
that some modelling solutions are inconsistent, while others are strictly calibrated for
particular failure modes observed during the tests confirming the model results. This makes it
impossible to apply the modeling methods for a beam experiencing other modes of failure,
like in the case of a parametric study.
The outcome of the study is a FEM methodology calibrated against a broad range of
experimental results, which permits prediction of the behaviour of SC members in bending
and shear.
2 Proposed FEM modeling for SC

All available materials properties were taken from papers describing the SC tests [1,2] to be
able to compare the FEM predictions with the experimental results.
In order to describe accurately the concrete behaviour damaged plasticity (DP) model was
used [3,4]. DP is a continuum damage model based on concrete plasticity. It requires the defi-
nition of the two uni-axial failure mechanisms, tensile cracking and compressive crushing of
81
the concrete. The evolution of the failure surface is controlled by two hardening variables,
both equivalent plastic strain, in tension and compression, provided in tabular format in
ABAQUS. The tension behaviour was defined according to Wang and Hsu [3]. Finally,
material properties were converted in true values, and concrete damage d c defined to range
from zero for undamaged material to one for the total loss of load-bearing capacity. Steel was
modelled using a tri-linear stress–strain curve.
Steel plates were modelled with shell elements (S4R), in order to allow plate buckling to be
estimated, while concrete has been modelled with 3D solid elements (C3D8R). Studs and tie
bars are defined with entire geometry, and holes were created in concrete in order to host
studs and ties. This was necessary, because simpler modeling techniques with steel studs and
ties embedded in the concrete are insufficient to take account of the concrete-steel interaction
or dowel effect. One of these models is illustrated in Fig. 1.
Fig. 1: Geometry of the proposed model
Concerning interactions, the general contact definition ”hard” was used in normal direction.
Stud-to-plate welds and the chemical bound between studs/ties and concrete were modelled
with tie constraints.
3 Conclusions
The FE model developed, is able to predict a broad range of failure modes relevant to SC
structures with reasonable accuracy. For instance, the behavior of several of the test reported
by Oduyemi and Wright [1], failing in bending (B2, B4, D1) and the beam SP1-5 tested by
Varma et al [2] failing in vertical shear were captured by the model.
The modeling technique described were used for developing a general purpose PYTHON
based plug-in in ABAQUS in order to allow users to generate SC beam geometries quickly,
and run different configuration SC models efficiently.
References
[1] Oduyemi TOS, Wright HD. “An experimental investigation into the behaviour of dou-
ble-skin sandwich beams,” Journal of Constructional Steel Research, 14(3), 197-220,
1989.
[2] Varma AH, Sener KC, Zhang K, Coogler K, Malushte SR, “Out-of-plane shear behav-
ior of SC composite structures” in Transactions SMiRT 21, 6–11, 2011
[3] Wang T, Hsu TTC. “Nonlinear finite element analysis of concrete structures using new
constitutive models”, Computers and Structures, 79(32), 2781–2791, 2001
82
Tampere, Finland
ASSESSMENT OF EXISTING STEEL BRIDGE STRUCTURES
Jan Bujnaka,* and Richard Hlinkab

a,b
University of Zilina, Faculty of Civil Engineering, Slovakia
*
Jan Bujnak. Tel.: +421 41 5135650; E-mail: jan.bujnak@fstav.uniza.sk
Abstract: Appropriate controlling and maintenance of building structures during their service
lifecycle represent actions of equivalent significance as suitable design and realisation quality.
The assessment of the reliability of an existing structure producing the evidence that it will
function safely over a specified residual service life should be also a continuous activity to
ensure the security of the public. The present practice is illustrated in the paper by case studies
on existing bridges under exploitation.
1 Bridge inspection, maintenance

With respect to construction industry, four principal activities can be distinguished, as re-
search and development, designing, construction and management. During design and execu-
tion stage, a building is obviously properly supervised. But, inspections, maintenance and
structural management provide quality assurance for constructions and play either a very im-
portant role during entire and long-time exploitation. Over the past three decades, the con-
struction inspection program evolved into one of the most-sophisticated management systems.
Even the evaluation of existing bridges under operation is a continuous activity to ensure the
safety of the public. First of all, because bridges were built gradually in different time periods,
thus they were designed according to time-knowledge and live-load, which reflected the level
of transport technique.
2 Bridge superstructure rating

The most important parameter ensuring the reliable service of a bridge during its service life is
the load carrying capacity. It is defined as a maximum momentary weight of vehicles, which
can pass through the bridge under certain conditions. According to type of idealised vehicle
moving load on highways, roads and local communications, three types of load carrying ca-
pacities can be determined. Normally, vehicles can cross a bridge without special limits in
number, location at the road pavement and speed. Corresponding normal load carrying capaci-
ty is given as the weight Vn of one of six identical lorries representing the load model. Obvi-
ously only the exclusive load carrying capacity is interesting in the case of current bridge fail-
ures. It is given by maximum momentary weight Ve of a four-axle vehicle. Except of this vari-
able action is no moving load at the bridge. Sometimes, an industry would like to transport
83
heavy machinery from one location to another site. This heavy haul trailer would weigh much
more than the design vehicles and thus the bridge owner may need to determine the extraordi-
nary live-load-carrying capacity of the bridge. In relevant structural analysis, actual geometric
parameters of elements, real material properties and the most probable bridge behaviour and
current conditions should be considered. The data are provided by technical diagnostic and
investigation reports. Load capacity of road or highway bridge limited by critical section can
be determined from equation
 G .EG  . Q .E Vi   f yd (1)
The partial safety factors G and Q are given in standards, EG means the effect of all dead load
and EVi ) effect of variable action, produced by relevant load model. The dynamic coeffi-
cient  depends on structural element span. Yield strength fyd is output of material standard
tests. The latest approaches for assessment of railway bridges have been incorporated into the
guideline established by our department. The specification is based on the limit state concept
and live load is taken in accordance with the ideal train scheme UIC - 71. Load carrying ca-
pacity Z of critical section can be determined from the modified condition
 G .EG  Z .. Q .E UIC   f yd (2)
In this relation Z is ration of remaining resistance to the theoretical requirements for the UIC –
71 train. Depending on actual load, vehicles in operation are classified by railway authorities
in eight effect categories. For practical assessment, the ratio of a given vehicle to UIC – 71
train effects is of interest and denoted λUIC. The passage of certain group of railway vehicles
can be allowed, if the loading bridge carrying capacity is greater than maximum vehicle effect
λUIC. Thus, the following criterion should be satisfied
 .Z  UIC (3)
The factor can take into account real dynamic actions.
The evaluation provides necessary information on the actual bridge structural conditions.
However, when a bridge is found to have inadequate capacity for legal vehicles, engineers
need to look at several alternatives prior to closing the bridge to the public. Some of the pos-
sible remedial measures are imposing speed limits, reducing vehicular traffic, limiting for ve-
hicle weight, restricting the vehicles to certain lanes, recommending possible small repairs to
improve the problem.
Using selected case studies, establishing the live load-carrying capacity and the bridge rating
is illustrated in more details. First the modified behaviour of a road truss bridge of Pratt con-
figuration due to vertical post imposed imperfection is studied and retrofitting validated. Im-
portance of exeptional inspections might confirm a collapse by buckling of the upper chords
of a temporary road bridge. For achieving a lighter, more economical structure, the case of
temporary bridge river crossing can prove that damage-tolerant design can be preferred. A
failure of lower flange of a railway overpass was developed by an important impact of a trailer
to the bridge structure. The reduced load carrying capacity was insufficient for ensuring the
reliable service of the damage bridge for the effect of engines operating on the track and re-
placement of superstructure was necessary. The comprehensive railway truss bridge evalua-
tions of Warren configuration after nearly forty years of exploitation are assumed to conclude
this issue. Refined investigations of stress distribution by the advanced transformation model
include also field testing for the bridge behaviour verification. With the real stress range spec-
trums, damage accumulations were obtained applying Palmgren-Miner damage rule and re-
maining fatigue life was estimated.
84
Tampere, Finland
LOCAL BUCKLING BEHAVIOUR OF WELDED BOX SECTIONS MADE OF HIGH

STRENGTH STEEL
COMPARISON OF EXPERIMENTS WITH EC3 AND GENERAL METHOD
Nicole Schillo* and Markus Feldmann
Institute of Steel Structures, RWTH Aachen University, Germany

*
Tel.: +49 (0) 241 - 80 25823; E-mail: n.schillo@stb.rwth-aachen.de
Abstract: Within the RFCS funded research project “RUOSTE” it was aimed to study the ef-
fects of high strength steel properties on local buckling. The respective parameters were inves-
tigated on 34 stub column specimens. These were made of S500M, S700M and S960Q, with a
varying non dimensional local slenderness between 0.64 up to 1.55. Extensive imperfection
measurements were undertaken and analysed. The specimens were then used as stub column
tests to investigate the local buckling behaviour. The results were first compared with the re-
sistance curve of the current Eurocode (EC), which showed to be rather optimistic, especially
towards the slender range. Secondly, the results were compared with the general method, which
uses an equivalent imperfection approach.
1. Introduction
Previous research on local buckling of high strength steels showed an apparently optimistic
prediction of resistance according to EN 1993 Part 1‐5 (e.g. [1]). However, similar results could
be shown for mild steels [2] and the overall amount of tests on high strength steel can be con-
sidered low.
To increase the existing results and informativeness, stub column tests with varying load sce-
narios and slenderness were conducted. Tests included S960 specimens aiming at an inclusion
of these steel grades in further EC developments.
A modified general method approach is introduced in this paper, allowing for simplification by
using the global buckling curves from EN 1993 Part 1‐1. The improved resistance due to local
buckling behaviour is taken into account by an effective imperfection factor and is based com-
pletely on gross-cross-section values. A similar approach has been used in assessing interaction
of lateral torsional buckling and local buckling on beams with varying cross-section [3].
2. Comparison of effective width approach and general method approach

The in the experiments achieved ultimate load 𝐹𝑢 is scaled to the respective resistance model
𝐸𝐶 (acc. to Eurocode) and 𝐺𝑀 (modified general method). The ratio of the experimental results
(re) over the theoretical results (rt), plotted over the slenderness, is depicted in Fig. 1. The ef-
fective width approach tends to underestimates the resistance of plates against local buckling
in the stocky range and overestimates the resistance with increasing slenderness.
85
The modified general method has a lower standard deviation from the experimental results
across the investigated slenderness range. Although the results are more conservative in the
slender range than EC, it is aimed in ongoing research to achieve higher re/rt- ratios.
1.20
GM
1.15
EC
1.10
1.05
re/rt
1.00
0.95
0.90
0.85
0.80
0.60 0.80 1.00 1.20 1.40 1.60 1.80
76
Fig. 1: Experimental results scaled to resistance modells GM and EC
3. Conclusions
In this paper, three topics were approached: first the material properties of high strength steel
compared to the requirements given in EN 1993 Part 1‐12. Second, the comparison of local
buckling behaviour in comparison with EN 1993 Part 1‐5 and third the introduction of a modi-
fied general method approach to simplify the design for local buckling.
1. The material property requirements given EN 1993 Part 1‐12 were fulfilled by S500 and
S700 specimens, which are included in the code. Applying the same requirements on
S960 material shows insufficient strain capacity at ultimate strength. However, the per-
formance in the investigated field of local buckling was not affected.
2. The effective width approach tends to overestimate the resistance of plates against local
buckling with increasing slenderness. This could be already observed in other studys,
including mild steels and are thus not a special problem of high strength steels.
3. The herein introduced modified general method aims to combine the design check of
global and local buckling. This leads to reduced efforts concerning the calculation (one
design check instead of a separate global and local calculation).
4. The return of a local buckling check to the European buckling curves leads to a check
on the gross-cross-section. Thus, an extensive calculation process on an effective
cross-section including calculation of a shift of the centroid can be neglected.
5. The results shown in this paper are considered as a first step towards a simplified design
check. However, further research in scope of the RUOSTE-project is underway to im-
prove the safeness of resistance prediction.
References
[1] “LIFTHIGH”, RFCS project Contract No 7210 PR/379, 2002-2005.
[2] Dwight, J.B., Moxham, K.E.: “Welded steel plates in compression”, The Structural En-
gineer, Vol. 47, No.2, February 1969.
[3] Schillo N., Feldmann M.: “Interaction of local buckling and lateral torsional buckling of
T-shaped cantilevering beams”, Thin-walled structures, Vol. 81, p.236-241, 2014.
86
Tampere, Finland
Sustainable design of buildings in steel and composite structures
Richard Stroetmann a, Christine Podgorski b

a, b
Institute of Steel and Timber Construction, Technische Universität Dresden, Germany
1 Introduction
The demographic change and the growing awareness of sustainability are examples of chang-
ing social conditions, which affect the user requirements of buildings. Resource efficiency,
recyclability, life cycle costs and conservation of value, even under changing property condi-
tions, are increasingly the focus of planning. From the very first, the structural system of the
building is of great importance. The position of columns and walls determines the spatial pos-
sibilities and allows different flexibility of the floor layout or may restrict it. Diverse technical
and formation requirements for the structural system arise from the geometry and the different
functions of a building. Taking into account different usage scenarios over the life cycle of a
building, appropriate attention to the structural system and the finishes are to be provided (e.g.
the use of flexible partition systems) to allow changes of use with low-order conversion and
short interruptions. From the requirements of the building and the specifications for the facade
and column grids, the conditions and design parameters for the structural system arise such as
the spans of slabs and beams, floor-to-floor heights, live and additional dead loads, arrange-
ments for fire protection, the design of components and spaces for services. For this purpose,
suitable construction systems and components as well as design principles should be selected
with sustainability in mind.
2 Assessment of sustainability of structural system

To assess the ecological quality, the life cycle assessment (LCA) of a building, construction
system or component is performed. Based on life cycle inventory analysis and life cycle im-
pact assessment, auxiliary variables are used, such as the ecological quality of BNB or DGNB
in Germany, in which the weighted environmental impacts are summarised. For a compara-
tive assessment during a building component and system optimisation, it is necessary to de-
fine target and limit values. With these values, it is possible to assess the quality of a proposed
solution. By using factors of relevance, the degree of fulfilment can be determined subse-
quently in accordance with the systems of DGNB and BNB. In the context of the research
project P881 [1], a corresponding assessment procedure was developed.
The assessment of the economic quality of buildings includes the building-related life cycle
costs (LCC) and the ability for market and tertiary use. Flexible designs are characterized by
absence of columns or at least optimally positioned inner columns taking into account various
usage scenarios. Floor space efficiency and conversion feasibility are ensured by appropriate
floor plan layouts, floor heights and building access considering relevant usage scenarios.
87
In Fig. 1 (left), the construction height, the total mass (slabs and beams) per square metre
floor is shown for various distances a of the beams and its total length L=L1+L2 as well as for
concrete and composite slabs. The composite beams are designed as two-span beams with a
support at 4.80 m (corresponding to the room depth of a cellular office). With increasing dis-
tance of the beams, the construction height, the total mass and the masses of the reinforcement
increase, while the mass of the steel section per square metre decreases. In Fig. 1 (right), the
ecological degree of fulfilment and the costs for composite floor systems with various dis-
tances of beams are shown. From the charts, it is clear that the slightly higher mass for a beam
distance of 3.6 m in comparison to 2.4 m can be compensated, and this is the favourable op-
tion according to the environmental performance.
Fig. 1: Comparison of masses, construction heights, ecological performances and costs for compo-
site floors for different distances of the beams (S355) and slab types (see [1], [2])
4 Conclusions
In the paper the assessment of the ecological and economical quality of buildings by the ger-
men rating systems DGNB and BNB and its adaption for bearing structures are shown. By
examples for parametric studies the optimisation of slabs, ceiling systems, columns and its
combinations are presented. The construction sector is responsible for around 50 % of the re-
source consumption and environmental impact. Significant savings are possible by a suitable
choice of materials and construction types. Ecology and costs are not in conflict but can be
optimised together.
References
[1] Mensinger M, Stroetmann R, Eisele J, Feldmann M, Lingnau V, Zink J et al. “Nachhal-
tigkeit von Stahl- und Verbundkonstruktionen bei Büro- und Verwaltungsgebäuden”,
Final Report, AiF project No. 373 ZBG (2014), Düsseldorf, 2015.
[2] Stroetmann R, Podgorski C. “Zur Nachhaltigkeit von Stahl- und Verbundkonstruktionen
bei Büro- und Verwaltungsgebäuden – Tragkonstruktionen Teil 1”, Stahlbau, 83(4),
245-256, Ernst & Sohn, 2014.
88
Tampere, Finland
STEEL CONSTRUCTION EXCELLENCE CENTER
Jarmo Havulaa, Pekka Roiviob and Markku Heinisuoc

a
Hämeenlinna University of Applied Sciences, Hämeenlinna, Finland
b
Ruukki Construction, Hämeenlinna, Finland
c
Tampere University of Technology, Tampere, Finland
Abstract: The concept of Steel Construction Excellence Center (SCEC) of Hämeenlinna is

described. An example of a case study which deals with the fabrication and response of welded
high strength steel (HSS) tubular joints is shortly presented.
1 Introduction
The action of SCEC is based on contract between City of Hämeenlinna, SSAB and three levels
of educational institutes: Tavastia Vocational College (Tavastia), Häme Univesity of Applied
Sciences (HAMK) and Tampere University of Technology (TUT). The mission is to enhance
the competitiveness of steel construction companies in the Hämeenlinna region. Other compa-
nies are invited to participate in the actions of SCEC and this will be realized via the educa-
tional institutes and their facilities for experimental research. Figure 1 shows the structure of
the SCEC network.
Figure 1: Structure of the SCEC network.
89
The first actions of this network have been completed with industrial and other partners. These
are dealing with education and research. Basic funding is provided by the City of Hämeenlinna,
Tavastia, HAMK and SSAB, but considerable portion is coming from individual projects.
SCEC has been proven to be an effective concept. For Ruukki Construction it has i.e. signifi-
cantly shortened R&D project lead-times. Fast prototyping and testing is a shortcut to best so-
lutions. As an example of realized projects could be mentioned the new load bearing sheet
T130, which is optimized for Nordic loads. Due to optimal shape and embossed patterns on
flanges, the material usage was successfully minimized. Load bearing capacity was determined
by full scale test program. Other examples are i.e. sandwich panel tests, fastener capacity tests
and capacity determinations.
2 Example of a case study
One example of a case study deals with fabrication and response of high strength steel (HSS)
structures. The scope was in HSS welded tubular T-joints. The issues studied were:
 Cutting of tubes using different techniques, such as laser cutting and sawing. Tubes
were made of different steel grades with different sizes in Ruukki;
 Welding time measurements where welds were made using different welding technolo-
gies such as robot and manual welding with butt- and fillet welds. This was completed
in Tavastia, Kemppi and HAMK;
 Defining experimentally the initial stiffness and the moment resistance of welded T-
joints with steel grades from S420 to HSS Optim 700 MH, done in HAMK.
The analyses for the validation of theoretical models for welding cost estimation and mechani-
cal response will continue later at TUT. The results of the case study have been presented at
the IIW conference, Helsinki 2015 [1].
3 Conclusions
Based on the experience gained so far, SCEC seems to be a very well working concept. This
kind of network provides “one counter” possibility for the companies to get both educational
and research services. The co-operation between three levels of educational institutes is a novel
approach. By these means both the permanent staff and students must be involved in the same
project and gain experience of different tasks needed in the industry. The City of Hämeenlinna,
together with other partners, has found an original way to ensure the welfare of steel intensive
companies. It is believed that also other organizations get ideas of this network for planning
the future of their operations.
References
[1] J. Havula, H. Myllymäki, I. Sorsa, J. Haapio, M. Heinisuo, Experimental research of weld-

ed tubular HSS T-Joints, welding times and moment resistances (2015).
90
Tampere, Finland
PRACTICAL TUBULAR TRUSS OPTIMIZATION
Jussi Jalkanen
Sweco Structures, Finland
Introduction
Tubular trusses have become popular in the design of steel structures such as roofing trusses
and frameworks. The selection of commercially available structural hollow sections is large
and they have excellent mechanical properties. These profiles have high bending and torsional
rigidity compared to their weight and they are suitable for compressed members.
In the design of tubular trusses the next natural step is to move from analysis to optimization.
Structural optimization offers a systematic way to go further than the traditional analysis of a
few candidate structures that were selected based on designer’s experience and intuition.
The aim of this paper is to offer a practical engineer’s view to tubular truss optimization. This
study is a continuation to previous structural optimization research which was done in
Tampere University of Technology. After the academic research author has had an
opportunity to work with steel structures in a consultant office.
Practical demands in tubular truss optimization

The practical needs of steel structure project set certain demands for the design of tubular
trusses in a consultant office. Some of these demands are due to the way projects proceed and
some due to the code of practise and manufacturing. These demands have to be fulfilled also
in optimization or it cannot be considered as a realistic tool.
As the most important single practical demand it can be mentioned that the result of optimiza-
tion has to fulfill all the necessary design rules presented in the code of practice. If some de-
sign rule, which are taken from eurocode 3 in this study, is violated structure is not accepta-
ble.
The low total cost is usually the ultimate target in steel structure optimization. The price of
steel is more or less the same for all workshops. Manufacturing cost varies strongly between
different workshops even in the same country depending on manufacturing machines. The
cost of final steel structure correlates rather well with the mass of structure although the min-
imum cost truss is not usually the same as the minimum mass truss.
91
In practical design work it is not necessary to find the actual optimal solution in the tubular
truss optimization problem. It is enough if a good or a better than the so far best known solu-
tion is found. In a typical design case the outer dimensions of building as well as the free
height inside the building are already fixed based on architect’s work. Load carrying struc-
tures like roof truss and columns have to fit inside a given area. Even though steel designer
does not necessarily know which workshop will manufacture trusses there are some common
demands for the workshop friendly design of tubular trusses.
Tabu search
In the example case tabu search is chosen as the optimization algorithm because this
algorithm enables the use of good initial guess. Usually designer can produce cost-effective
initial guess based on his/her experience. The other possibility to find it is by checking some
candidate solutions selected based on simple rules of thumb.
The example case

In the example problem the optimization of an industrial or warehouse building steel frame is
studied. Symmetrical frame consists of tubular truss made of square hollow sections (SHS)
and cantilever type I-beam columns (HEA). The material for the frame is steel S355. There
are load bearing steel sheets as the roof and the distance between frames is 6 m. Building lo-
cates in Southern Finland. Joints are gap type joints without extra strengthening plates. Ac-
cording to Finnish practice the structural fire design is not needed in this case.
The set of available SHS profiles consist of 37 different sizes from 50x5 to 300x12,5. All pro-
files belong to cross section classes 1 or 2 in compression and their wall thickness is at least 5
mm. For the columns the whole range from HEA100 to HEA400 is available.
In the current sizing optimization problem there are four different SHS profile sizes and col-
umn size chosen as the design variables. Beside these also the eccentricities in joints are se-
lected as the design variables. As the object function the mass of frame is chosen. Constraints
take care that steel frame fulfills all the needed rules given in eurocode 3 for profiles, joints
and displacements.
The minimum structure has been sought in two phases. At first the sizes of profiles are deter-
mined based on simple rules and checking the group of candidate solutions one by one. The
second phase in the tubular truss optimization is the use of tabu search with phase one initial
guess.
Based on result it can be noticed that tabu search improves the initial guess 246 kg i.e. it is 8,1
% lighter. Column size and eccentricities remain the same. The distribution of mass is such
that upper chord corresponds 53,4 %, lower chord 27,4 % and diagonals 19,2 %.
Conclusions
In the tubular truss design it is better to exploit the advantage of structural
optimization than to abide by the analysis of few candidate structures selected based on
designer’s experience and intuition.
92
Tampere, Finland
THE IMPACT OF JOINT CONSTRAINTS ON THE OPTIMAL DESIGN

OF TRUSS STRUCTURES
Roxane Van Mellaerta∗ , Geert Lombaertb and Mattias Schevenelsa

a Department of Architecture, Faculty of Engineering Science, KU Leuven, Belgium
b Department of Civil Engineering, Faculty of Engineering Science, KU Leuven, Belgium
Abstract: This paper proposes a method to account for joint constraints in the global discrete
size optimization of a steel truss structure. The design of a statically determinate N-type truss
girder is considered first without and then with the joint constraints specified in Eurocode 3. In
order to guarantee global optimality in both cases, the optimization problem is reformulated as
a mixed-integer linear program. In the first case, a design is obtained where the joints have to
be strengthened in a postprocessing step. In the second case, a design is obtained that satisfies
the joint constraints. The weight of this design is about 15% higher than in the first case. This
shows that the joint constraints have a significant impact on the optimal design.
1 Introduction
Real-world design problems are often governed by a large number of constraints and practical
issues. For a steel truss girder with welded joints, the usual displacement, member force, and
buckling constraints as formulated in Eurocode 3 are imposed. In addition, the member sections
must be chosen from a given section catalog, and the joints must obey certain geometrical rules
in order to ensure structural integrity and weldability, and mechanical rules in order to avoid
chord web, chord shear, and brace failure. Most existing design optimization algorithms cannot
take into account all these practical constraints. As a consequence, a manual postprocessing
step is required, where the optimized design is modified to satisfy the constraints which are not
considered during the optimization. This costs precious engineering time and may lead to a
suboptimal design or a design that no longer fulfills the stress and displacement constraints.
2 MILP-reformulation
The optimization method used in this paper has originally been proposed by Grossmann et al. [1]
for discrete size optimization problems and is extended by Rasmussen and Stolpe [2] and Mela
and Koski [3] for truss topology design problems. The optimization problem is reformulated as
a Mixed-Integer Linear Problem (MILP), which is solved with the branch-and-bound method
in order to achieve global optimality. This MILP is obtained by means of binary decision
variables and the Simultaneous ANalysis and Design (SAND) approach: the structural nodal
displacements and the member end forces are considered as additional design variables and the
equilibrium equations are enforced by means of additional equality constraints.
∗
Corresponding author. Tel.: +32 16 32 13 68. E-mail: roxane.vanmellaert@asro.kuleuven.be.
93
In order to ensure that all joint resistance constraints can be reformulated as linear constraints in
terms of the design variables, the scope of this paper is limited to statically determinate analysis
models. Since the member forces of statically determinate models do not depend on the sections
and remain constant in the optimization, they do not have to be considered as additional design
variables. As a consequence, the normal forces are not adopted as design variables and the
equilibrium constraints are dropped. The original MILP proposed for discrete size optimization
is thus simplified. The additional joint constraints - which would be quadratic if the member
forces are considered as design variables - can then be reformulated as mixed-integer linear
constraints. The design variables therefore consists of (1) binary decision variables which select
a section from a catalog for each member, (2) nodal displacement variables, and (3) joint gaps
when joint constraints are taken into account.
3 Results
Fig. 1: N-truss girder with HEA top chord sections, UPN bottom chord sections, and RHS braces.
In the first case the truss shown in figure 1 is optimized considering only displacement and
member constraints as formulated in part 1-1 of Eurocode 3. After verifying to what extent
the optimized design satisfies the constraints that are not explicitly considered in the MILP, it
is observed that the joint constraints are not satisfied. As a consequence, the joints need to be
strengthened. This can be done by either selecting different profiles, or by locally strengthening
the joints e.g. by means of stiffening plates. The first approach would lead to a suboptimal
result, as it is very difficult to determine which section should be made heavier. The second
approach would only have a limited impact on the weight of the truss, but the fabrication costs
would become much higher.
In the second case the joint constraints as formulated in part 1-8 of Eurocode 3 are also taken
into account. The results show that taking into account joint constraints during the optimization
has a significant impact on the optimized design: the weight of the obtained result is 15% higher
than in the case where joint constraints are not considered. If the joint constraints are accounted
for in a suboptimal way (e.g. by manually selecting heavier sections), the additional weight may
be even higher. Taking into account joint constraints during the optimization therefore leads to
a cost reduction at two levels: in terms of engineering cost (no manual postprocessing step is
needed) as well as fabrication cost (joint strengthening is avoided).
References
[1] I.E. Grossmann, V.T. Voudouris, and O. Ghattas. Mixed-integer linear programming refor-
mulations for some nonlinear discrete design optimization problems. In C.A. Floudas and
P.M. Pardalos, editors, Recent advances in global optimization, pages 478–512. 1992.
[2] M.H. Rasmussen and M. Stolpe. Global optimization of discrete truss topology design
problems using a parallel cut-and-branch method. Computers & Structures, 86(13):1527–
1538, 2008.
[3] K. Mela and J. Koski. Distributed loads in truss topology optimization. In Proceedings of
the 10th World Congress on Structural and Multidisciplinary Optimization, Orlando, USA,
2013.
94
Tampere, Finland
LATERAL BUCKLING STRESS FOR H-SHAPED BEAMS WITH

CONTINUOUS BRACES
Yoshihiro KIMURA* and Yuki YOSHINO**
* Professor，New Industry Creation Hatchery Center，Tohoku University，Dr. Eng.

** Assistant Professor, National Institute of Technology, Sendai College, Japan. Dr. Eng
The lateral buckling of H-shaped beams used in steel structures has been issued for
decades so that practical steel frame design codes including Japanese design code recommend
to connect lateral braces to the compressive flanges of H-shaped beams to prevent the lateral
buckling in [1],[2]. In practice, lateral braces are usually connected to the upper flanges of H-
shaped beams in the steel structures. Under a combination of dead loads and lateral seismic
loads, the bottom flange without any bracings could be subjected to axial compression, may
resulting in a beam undergoing lateral buckling deformation initiated by the flexural buckling
of the bottom flange.
When the continuous braces such as folded-roof plate are connected to the upper
flanges of H-shaped beams, they could restrain the lateral deformation of H-shaped beams.
However, folded-roof plates, categorized in non-structural members, are not considered as ef-
fective bracings in design. Therefore, their actual effects on prevention of the lateral buckling
of beams has not been clarified. In fact, it is shown that the folded-roof plates deformed due to
the lateral buckling of the beams in the previous reconnaissance reports of the earthquake, so
that it is considered that the plates carried the lateral force and the rotational moment occurred
by the lateral buckling.
Our previous research [3] (Kimura, Yoshino 2013) clarified the relation between the
lateral buckling strength of H-shaped beams and the demands of the lateral and rotational ri-
gidities for continuous braces when a beam is subjected to uniform moment distribution. The
elastic buckling strength of the H-shaped beams by formulating energy conservation equations
was estimated, considering web deformation of the beam section in addition to the lateral and
torsional deformation as shown Fig.1.
In this paper, the elastic lateral buckling load of H-shaped beams with continuous
braces are developed with the energy method, and the relation between the lateral and rota-
tional rigidities of braces and the lateral buckling strength is investigated as shown in Fig.1. In
the eigenvalue analyses, H-shaped beams consist of four node shell elements and the continu-
ous braces are replaced on the lateral and rotational springs as shown in Fig.2. The buckling
load from the energy method and analyses results is compared. Also this paper suggests the
required bracing rigidities to restrain the buckling deformation of H-shaped beams
Next, the elasto-plastic buckling strength is calculated by the large deformation anal-
yses, and is evaluated using the buckling curve for Japanese standard code with the modified
95
slenderness ratio of the yield strength to the elastic lateral buckling load for continuous braces
from energy method as shown in Fig. 3.
References
[1] Architectural Institute of Japan, AIJ (1998). Recommendation for Limit State Design of Steel
Structures. (in Japanese)
[2] Architectural Institute of Japan, AIJ (2005). Design Standard for Steel Structures. (in
Japanes )
[3] Kimura,Y. and Yoshino, Y. (2013). Effect of Lateral-Rotational Restraint and Strength of
Continuous Braces on Lateral Buckling Load for H-shaped Beams, AIJ. 78, 683, 193-201 (in
Japanese)
P1
P2
l x u1
k ku
1
w
x h o d
y
z tw tf u
P1 y 2
s
P2 b u2
(a)Lateral Buckling of H-Shaped beam (b) Bucking deformation of H-Shaped section

Figure 1 Lateral Buckling Deformation of H-Shaped Beam
with Continuous Braces on Upper Flange
M2
Lateral Springs ku l'
Rotational Springs k l' l'

4 Node Element
y
z d - tf ) t w
Af + (12 2
x l Pz1
Pz2
Pz3 Pz3 d1 d2 d3 d tw
M1 Pz2 6
Pz1
Figure 2 Numerical Analysis Model

σcr/σy
1.2
Recommendation for Limit State Deσign
1 of Steeλ Structureσ, AIJ
Deσign Standard for Steeλ Structureσ, AIJ
Tangent Line (σcr/σy=0.6)
0.8
No Bracing
0.6 Type A
No Rotationaλ Bracing
0.4 Type B
No Rotationaλ Bracing
0.2 Rotationaλ Bracing
0 λb
0 0.5 1 1.5 2
Figure 3 Lateral Buckling Stress of H-shaped Beams with Braces
96
Tampere, Finland
Industrial Hall Constructions
N. Gengea, C. Remdeb, K. Weynandc, J. Kuckd

a,b
Vallourec Deutschland GmbH, Düsseldorf, Germany
c,d
Feldmann + Weynand GmbH, Aachen, Germany
1 General Information
The worldwide demand for wide-span industrial buildings is constantly growing, especially in
terms of logistics centers and hangars, just to name a few. To meet this trend PREON box, a
modular construction system, has been developed. It is an in-house development including a
patented steel roof frame system that enables the economic realization of large spans up to
100 meters. PREON box allows combining standardized manufacturing with high flexibility
to respond to customer's needs. To make even the planning process more efficient, a software
tool called „PREON designer“ has been developed specifically for the design of this system.
The aim is to provide a software tool simplifying the design of hollow section structures.
2 System Specification
In order to design an industrial hall in an automatic way by means of a computer program, it is
essential that a detailed technical description of all components is available. Therefore, in a
first step, a so-called system specification has been developed. The system specification is a
kind of knowledge base that describes all technical details of the PREON box system like
geometry, generation and design algorithms as explained hereafter more in detail.
PREON box is a modular construction system. In contrast to other typical construction
systems, all structural elements, such as a complete girder or a connection detail, are
parameterized components. This means that a component, e.g. a plate, is not specified for
example by a particular fixed length, width and thickness, rather than by the parameters L, B
and T, examples are seen in Fig. 1 for a girder and a girder support. The advantage of this
approach is the fact that, on one side, all components are standardized and can therefore be
implemented as predefined types in the design software and, on the other side, the actual sizes
or dimensions are very flexible to be adapted to the needs of the client.
Because the content of the specification should be implemented in a design software, it is im-
portant that also all requirements and limitations for the use of each element are specified, e.g.
minimum and maximum girder spans, maximum crane loads, etc. In other words: the system
specification says exactly what is possible and what is not possible.
97
Reference line or plane
Fig. 1: Examples of parameterized components (PREON girder and girder support)
3 Design Tool
To run a design sequence the user has to perform three steps:
Firstly, some information related to the location of the site like its height above sea level,
wind and snow zone must be given. These data will be used to generate climatic actions like
wind loads and snow loads. Imposed loads and self-weight of the roofing and cladding is
provided by the user. Self-weight of the steel structure is considered automatically.
In a second step, the user specifies the global layout of the steel structure by means of number
of bays, bay height and width, distance of the frames, inclination of the roof, etc. Through
those data, the structure is defined by so-called reference lines or planes respectively. As an
example, a hall with two bays is shown in Fig. 2.
a) Reference lines b) Loadbearing frame

Fig. 2: PREON box structure
The last step will start with the automatic generation of the structure. Based on the initial
choice of elements evaluated in the pre-design, all individual members of the primary steel
structure are generated. Requirements for connection details are directly taken into account.
For example, for the generation of all lattice girders, the positions of braces are chosen in such
a way that all K joints will be, if possible, gap joints. The requirements for the size of the gaps
specified in EN 1993-1-8 are directly considered. The generated system for the example
shown in Fig. 2a can be seen in Fig. 2b.
Then the automatic design will be executed. In an iterative procedure, optimized sections will
be evaluated. Verification of the steel structure is made according to the Eurocodes.
When the design procedure is successfully terminated, the software generates a full design
note and a material list. Based on this bill of material, cost estimation is made. The paper pre-
sents the technical benefits of PREON box. Those lead to very economic design for wide–span
industrial buildings combined with a much higher flexibility and shorter planning/realization pro-
cess compared to traditional building solutions.
98
Tampere, Finland
EFFECT OF END STIFFENER REINFORCEMENT ON LATERAL

TORSIONAL BUCKLING BEHAVIOR OF H-SHAPED BEAMS WITH
LARGE DEPTH–THICKNESS RATIO
Daiki KUBOTAa, Kikuo IKARASHIa

a
Department of Architecture and Building Engineering, Tokyo Institute of Technology, Tokyo, Japan
Abstract: End stiffeners can potentially restrain both local plate buckling and lateral torsional
buckling. By extending the stiffening length, the plastic deformation capacity should also be
improved. Therefore, this study clarifies the behavior of lateral torsional buckling of H-
shaped beams with large depth–thickness ratio, reinforced with end stiffeners. In loading test,
the plastic deformation capacity was improved by installing stiffeners without lateral bracing.
Further improvement was achieved by extending the stiffening length. The effective coverage
and shape of stiffeners were evaluated in a numerical analysis using the finite element method.
1 Introduction
As commonly known, H-shaped beams with large depth–thickness ratio are prone to local
plate buckling, which can be restrained by installing reinforcement stiffeners at the beam ends.
This restraint also improves the plastic deformation capacity. In such a construct, the collapse
mode may change from local plate buckling to lateral torsional buckling. However the end
stiffeners can potentially restrain both local plate buckling and lateral torsional buckling with
increasing stiffening length. In addition, sufficient plastic deformation capacity should be ob-
tainable without lateral bracing. Therefore, this study elucidates the basic lateral torsional
buckling behavior of H-shaped beams with large depth–thickness ratio and reinforced with
end stiffeners. Furthermore, the effects of the stiffener parameters on the plastic deformation
capacity of the beams are numerically investigated by a finite element method (FEM).
2 Loading test of H-shaped beams reinforced with stiffeners

The H-shaped beams in the monotonic loading test were selected for their sufficiently thick
flange and very thin web (Fig. 1). The loading form is a cantilever, which is half of the H-
shaped beams under anti-symmetric moment. To improve the plastic deformation capacity
without raising the maximum strength, horizontal stiffeners were welded to the web neutral
axis. The length of the horizontal stiffeners was 375 mm (equaling the specimen depth) and
1000 mm (half the specimen length). The results were compared among this specimen type
and three other types. The plastic deformation capacity was improved by the stiffener rein-
forcements. At longer stiffener lengths, the plastic deformation capacity was further improved
to that of a beam with fixed lateral displacement only (Fig. 2).
99
L=2000(mm) Q/Qp
1.5
Horizontal Stiffener θ (Only lateral displacement
tf N-n S-1000 N-f was fixed)
Vertical Stiffener
d Q
1.0
S-375
N-n
End Plate
PL-9.0×40 Stiffener
B Ls=375 or 1000(mm) (Both Sides) 0.5 N-n
bs=40(mm)
N-f
375(mm)
S-375
9(mm)
: Maximum load
S-1000
ts L=2000mm
0.0
0 2 4 6 8 θ/θp
Stiffener Reinforcement B' S-375, S-1000 100(mm)
B-B' section Fig. 2: Non-dimensional load–
Fig. 1: Loading models deformation relationship
Q/Qp 5R 5R
1.5
750mm bs/ts=16 bs/ts=4.4 θ max
500mm 1000mm R= −1
4 bs/ts=10 bs/ts=1.6 4 θp
bs/ts=7.2
1.0 375mm
1500mm 3 3
2bs ts
Ls=375mm δ=
2 2 dtw
Ls=500mm
0.5
Ls=750mm No Stiffener No Stiffener
Ls=1000mm 1 1
δ=0.44 bs/ts=4.44
Ls=1500mm : Maximum load
0.0 0 0
0 2 4 6 8 θ/θp 0.0 0.1 0.2 0.3 0.4 0.5 0.6 δ 0.0 0.1 0.2 0.3 0.4 0.5 0.6 δ
Fig. 3: Non-dimensional load– a) Ls=500(mm) b) Ls=1000(mm)
deformation relationship for Fig. 4: Relationship between cross-sectional area ratio (δ)
various stiffener lengths and width–thickness ratio of stiffener
3 Lateral torsional buckling behavior

Based on the loading test results, this section evaluated the effect of stiffener reinforcements
on the buckling behavior. Restraining the local plate buckling by stiffener reinforcements
shifted the collapse mode to lateral torsional buckling. Moreover, the improved plastic defor-
mation capacity at longer stiffener lengths was attributed to delayed increase of the lateral
displacement. The lateral torsional buckling was effectively restrained by reinforcement using
a closed cross section with a large torsional rigidity.
4 Parametric study using FEM analysis

This section examines the appropriate coverage and shape of the stiffeners for restraining lat-
eral torsional buckling, while considering the rigidity of stiffeners. The evaluation was con-
ducted by FEM analysis. High plastic deformation capacity (R) was achieved by reinforcing
half of the cantilever (from the beam-ends) with stiffeners. Except for relatively small width–
thickness ratios of the stiffener, the optimal cross-sectional area ratio of stiffener to web (δ)
was approximately 0.4 (Figs. 3 and 4).
5 Conclusions
The basic behavior of lateral torsional buckling of reinforced H-shaped beams with large
depth–thickness ratio was clarified in a loading test. Numerical FEM analysis confirmed that
the plastic deformation capacity can be improved to that of a beam with fixed lateral dis-
placement only. For this purpose, the stiffeners should cover half of the beam, from the fixed
end to the beam inflection point, and the cross-sectional area ratio of stiffener to web should
equal approximately 0.4, unless the width–thickness ratio of the stiffener is relatively small.
100
Tampere, Finland
LOW CYCLE FATIGUE PERFORMANCE OF INTEGRAL BRIDGE STEEL H-

PILES UNDER SEISMIC DISPLACEMENT REVEALS
Murat Dicleli a, Memduh Karalar b,*

a,b
Department of Engineering Sciences, Middle East Technical University
*
Author for contact. Tel.: +90 (312) 210-4451; E-mail: mdicleli@metu.edu.tr,
Abstract
Under the effect of medium and large intensity ground motions, the seismically-induced lat-
eral cyclic displacements in steel H-piles of integral bridges (IBs) could be considerable. As a
result, the piles may experience cyclic plastic deformations following a major earthquake.
This may result in the reduction of their service life due to low-cycle fatigue effects. Accord-
ingly, in this study, low cycle fatigue in integral bridge steel H-piles is investigated under
seismic effects. For this purpose, a two-span integral bridge is considered. Three dimensional
nonlinear structural models of the IB including dynamic soil-bridge interaction effects are
built. Then, time history analyses of the IB models are conducted using a set of ground mo-
tions with various intensities representing small, medium and large intensity earthquakes. In
the analyses, the effect of various properties such as soil stiffness, pile size and orientation are
considered. The magnitude of cyclic displacements of steel H piles are then determined from
the analyses results. In addition, using the existing data from experimental tests of steel H-
piles, a fatigue damage model is formulated. This fatigue damage model is used together with
the cyclic displacement obtained from seismic analyses to determine the remaining service life
of IBs under cyclic displacement due to thermal effects. The fatigue damage analyses results
revealed that the calculated cumulative fatigue damage indices in the steel H-piles induced by
seismic loadings are negligible.
Introduction
An integral bridge (IB) is one in which the continuous superstructure, the abutments and the
single row of flexible piles supporting the abutments are built monolithically to form a rigid
frame structure. The most common types of piles used at the abutments are steel H-piles. The
seasonal and daily temperature changes result in imposition of cyclic horizontal displacements
on the continuous bridge deck of integral bridges and thus on the steel H-piles supporting the
abutments. As a result, the piles may experience cyclic plastic deformations. These plastic de-
formations causes low cycle fatigue in steel H-piles of IBs. The service life of the IBs highly
depends on these low-cycle fatigue effects due to temperature changes. In addition, under the
effect of medium and large intensity ground motions, the seismically-induced lateral cyclic
101
displacements in steel H-piles of integral bridges (IBs) could be considerable. Modern IBs are
known to have performed well in recent earthquakes due to the increased redundancy, larger
damping resulting from cyclic soil-pile-structure interaction, smaller displacements and elimi-
nation of unseating potential. The monolithic construction of IBs also provides better transfer
of seismic loads to the backfill and pile foundations. However, similar to their performance
under thermal effects, the seismic performance of IBs may depend on abutment height and
thickness, pile size and orientation, backfill compaction level as well as stiffness of the foun-
dation soil. A comprehensive seismic research study on IBs has not been conducted yet to
provide clear suggestions for the configuration and geometric detailing of IB structural com-
ponents as well as appropriate backfill and foundation soil properties to enhance their seismic
performance. In the last decades, many research studies have been conducted on the perfor-
mance of IBs under thermal loads, live load distribution among components of IBs and soil-
structure interaction effects in IBs (Dicleli 2005, Erhan and Dicleli 2009, Kalaycı et. al. 2012).
However, research studies concerning the seismically induced low cycle fatigue effects in
steel H-piles of IBs does not exist in the literature. Accordingly, this research study is aimed at
experimentally and analytically investigation the effect of various structural and geotechnical
properties and parameters on the seismic performance of IBs.
Conclusions
Followings are the conclusions deduced from this parametric study: The effect of axial load is
observed to have a significant effect on the low cycle fatigue performance of steel H-piles in
two ways: (i) when the pile is subjected to moderate strain amplitudes (five times the yield
strain), the presence of axial load is observed to enhance the low cycle fatigue life of the pile.
This mainly due to the fact that, the presence of axial load decreases the amplitude of the ten-
sile strain that results in cracking of the material (ii) when the pile is subjected to larger strain
amplitudes (10 times the yield strain), the presence of axial load is observed to decrease the
low cycle fatigue life of the pile. This is mainly due to local buckling of the flange under the
effect of compressive stresses from the axial load and high compressive strains due to the ef-
fect of bending. Local buckling increases the local curvature and strains. This locally accel-
erates the cracking of the material. Furthermore, IBs with shorter and thinner abutments sup-
ported on larger steel H-piles oriented to bend about their strong axis and driven in softer
foundation soil is expected to exhibit a better seismic performance and negligible low cycle
fatigue effects in steel H-piles.
References
[1] Dicleli, M. (2005). “Integral abutment-backfill behavior on sand soil—Pushover analy-
sis approach.” J. Bridge Eng. , 10 (3 ), 354–364.
[2] Dicleli M, Erhan S (2011) Live load distribution formulas for single-span prestressed
concrete integral abutment bridge girders. Journal of Bridge Engineering 14(6):472–486
[3] Kalayci E, Civjan SA, Brena SF (2012) Parametric study on the thermal response of
curved integral abutment bridges. Engineering Structures, 43: 129-138.
102
Tampere, Finland
SYSTEM RELIABILITY ANALYSIS OF STEEL RAILWAY BRIDGE

BASED ON HISTORIC ROLLING STOCK RECORDS
Gunnstein Thomas Frøseth, Anders Rönnquist
Norwegian University of Science and Technology
1 Introduction
Riveted steel open deck truss bridges are typical for medium and longer spans in the Norwegian
railway network and fatigue is frequently the limiting state for remaining service life of steel
bridges. Due to the high partial factors reflecting the large amount of uncertainty associated with
both fatigue resistance and railway load models, deterministic assessment methods prove overly
conservative. Probabilistic concepts have been applied in fatigue assessment of steel railway
bridges by several researchers, e.g [1, 2], to obtain refined estimates of the remaining service
life. The focus of these studies have been on the reliability of single components. Assessment
of the system reliability is scarce, among the exceptions are [3] which considers the system
reliability of a typical riveted stringer-to-crossgirder connection.
In this paper, a preliminary case study is presented where the current system reliability index
against fatigue of a open deck riveted steel truss railway bridge is considered.
2 Description of the Case Study

Tallerås railway bridge was put in service in 1912, and is located on the railway between Otta
and Dombås station on the line Dovrebanen at KM 327.630. The bridge is a riveted steel
open deck truss bridge with width 5.0m and a main span of 52.0m. Stringers, crossgirders and
verticals have short influence lines which make them susceptible to fatigue. Verticals on axes 2
and 4 do not carry appreciable load, the study is therefore limited to the verticals at axes 1, 3
and 5 and all stringers and crossgirders. Furthermore, only primary actions are considered and
deterioration effects on fatigue resistance is neglected.
3 Fatigue Assessment and Reliability Analysis

The fatigue assessment is conducted with SN-curves. Fatigue load spectra for periods 1912 −
1960, 1960 − 1980, 1980 − 2013 are obtained through finite element simulations of trains pass-
ing the bridge. The trains are sampled from rolling stock in a similar method to [1] by con-
sidering the geometry and relative number of each vehicle, axle load distribution, dynamic
amplification and train configuration. Modeling uncertainty is included as the ratio between
103
actual and calculated stress. Damage is evaluated by Miner’s rule. The limit state function for
a component becomes
Xmm m

g(X) = ∆ − ∑ qi nQ,i Seq,i ≤0 (1)
C i
where ∆ is the damage at failure, Xm is the modeling uncertainty, C and m are the intercept
and slope parameters of the SN-curve, qi nQ,i is the number of cycles and Seq,i is the equivalent
fatigue stress for period i. The bridge is considered a series system and the system reliability is
assessed by the first order series bounds and first order reliability method.
4 Results and Conclusions
Figure 1: The figure shows the projected reliability of the structure under assumption that the
traffic grows at an annual rate of 2%.
The current component and system reliability of the considered failure modes was found
satisfactory at target reliability β = 2.3. The first order series bounds yielded wide estimates
on remaining service life, and correlation between failure modes should be considered in future
work. Reliability was found to be strongly dependent on failure of stringers at boundaries of
the structure, closer investigation of boundary conditions are in order. Future work should
include alternative failure modes due to deformation induced stresses and deterioration models
on fatigue resistance.
References
[1] Daniel H. Tobias and Douglas A. Foutch. Reliability-Based Method for Fatigue Evaluation
of Railway Bridges. Journal of Bridge Engineering, 2(2):53–60, May 1997.
[2] Alessio Pipinato and Claudio Modena. Structural Analysis and Fatigue Reliability Assess-
ment of the Paderno Bridge. Practice Periodical on Structural Design and Construction,
15(2):109–124, May 2010.
[3] Boulent M. Imam, Marios K Chryssanthopoulos, and Dan M Frangopol. Fatigue System
Reliability Analysis of Riveted Railway Bridge Connections. Structure and Infrastructure
Engineering, 8(10):967–984, 2012.
104
Tampere, Finland
FATIGUE PROBLEMS IN RIVETED RAILWAY BRIDGES -
INVESTIGATION AND REHABILITATION METHODS
H. Vagn Jensen Chief Consultant, M.Sc., Ramboll and Claus Pedersen, Project Director,
M.Sc., Ph.D., Ramboll
*
Author for contact. Tel.: +45 51616129; E-mail: hvj@ramboll.dk
Abstract
This paper summarizes investigation methods to detect fatigue cracks at riveted bridges.
During recent years three different types of serious fatigue cracks have appeared at the two
major 75 year old Danish railway bridges, Masnedsund Bridge and Storstroem Bridge. The
reasons for the fatigue cracks have in all cases been determined to be poor detailing of riveted
joint connections, either too stiff or flexible joint connections or unintended notches.
None of the cracks have been easily accessible for regular visual inspection. They appeared at
elements either with very difficult access or in primary webs hidden behind joint splice plates.
The experiences show that the most efficient approach to detect such cracks is a combination
of fatigue analyses for identification of possible weak details, visual inspections and the NDT
(Non Destructive Testing) methods normally used for weld inspection.
By combining photo records and NDT methods the extent of the cracks has been determined
and criteria and methods for rehabilitation have been developed. NDT using digital X-ray film
has turned out to be an efficient method for detecting cracks at web plates covered by splice
plates. However, ultrasonic testing has also been efficient, provided the structure allows for
access of probes, and X-ray films for the same area are available for calibration.
Even though several hundred joints seem to be identical, the presence and length of the crack
is highly unpredictable. This leads to the conclusion that limited spot testing should be con-
sidered with care.
For both bridges the necessary rehabilitation and strengthening has been performed. In some
cases minor repair was carried out, but in the most serious case the bridge was closed immedi-
ately for railway traffic for more than a month. After these incidents, both bridges are moni-
tored regularly and will be kept in service until planned replacement in 2020.
105
106
Tampere, Finland
ON ACTUAL BEHAVIOUR OF CONTINUOUS COMPOSITE GIRDER

BRIDGES AND THEIR CONVENTIONAL MODELLING
Jaroslav Odrobiňáka, Ján Bujňákb

a,b
Faculty of Civil Engineering, University of Žilina, Slovakia
Abstract: An experimental verification of flexural behaviour of a composite steel-concrete

girder-bridge is presented. Focus is taken on the changes in stiffness of continuous superstruc-
ture above piers. The concepts of simplified modelling of the concrete part in the regions of
hogging moments are discussed in the research.
1 Contents
The research deals with a four-span continuous road bridge across a highway with the spans
of 18 + 22 + 22 + 18 metres, Fig. 1. Because of an angular crossover and arch curvatures of
side road approaches, theoretical spans of left and right main girders are not equal.
Fig. 1: The bridge from the bottom view; “pier” cross section and strain gauges arrangement
The deflections of both main girders in the middle of each span and the settlement of all bear-
ings were monitored. Moreover, strains in the targeted bridge “pier” cross section above the
fourth support were observed (see Fig. 1). Four lorries with the average gross-vehicle weight
of 42.76 tons were arranged into two positions (load cases LC3 and LC4).
A spatial numerical FEM model was developed. To compare the results with the linear global
analysis a simplified modelling of the concrete part in the regions of hogging moment was
introduced by three alternatives: “cracked”, “uncracked” and “reduced”, where reduced stiff-
ness was applied to normal stiffness of the slab, while flexural stiffness of the slab stayed un-
changed. A small part of comparison of the numerically obtained values with those observed
during measurement is shown in Fig. 2.
107
2 Results and conclusions

-6 LC3
Left girder Right girder
-4
Deflection in 3rd span [mm]
-2 LC - 4 900 900
0
Height of steel girder [mm]

750 750
2
left girder right girder 600 600
4
6 450 450
8 300 300
10
LC - 3 150 150
12
14 0 0
10
20
10
20
-40
-30
-20
-10
-40
-30
-20
-10
0
0
16
measured cracked reduced uncracked

a) Mid-span deflections of the 3rd span b) Stresses in [MPa] through the height of steel girders in
“pier” cross-section above third pier
Fig. 6: Comparison of the results
It could be stated that within the three analyzed models, the “uncracked” analysis gave the
results, which are closest to the observed ones. Analyses of the other two models with re-
duced stiffness above piers produced higher differences. These differences are even more
visible, when strains in concrete or reinforcement are compared. However, in the phase of
bridge design it is necessary to ensure the safe determination of the bridge response to action. In
that case, the stiffness reduction in the hogging regions shell be given by the corresponding codes
on conservative side to fulfil the requirement of the safe design.
Acknowledgments
The paper presents results of works supported by Scientific Grant Agency of the Slovak Re-
public under the project No. 1/0583/14 and by the Slovak Research and Development Agency
under the contract No. APVV-0106-11.
References
[1] Moravčík M, Bahleda F. “Static load of the composite arch bridge”, Civil and Environ-
mental Engineering, 7(1), 35-41, 2011.
[2] Gocál J, Hlinka R, Jošt J, Bahleda F. “Experimental Analysis of Stiffness of the Riveted
Steel Railway Bridge Deck Members’ Joints”, Civil and Environmental Engineering, 10(2),
105-109, 2014.
[3] Odrobiňák J, Vičan J. “Behaviour analysis of composite motorway bridge during proof-load
test”, Proceedings of the 5th International Conference Concrete and Concrete Structures,
(EDIS - University of Zilina), Žilina, Slovakia, 401-408, 2009.
[4] Odrobiňák J, Vičan J, Bujňák J. “Verification of composite steel-concrete bridge behav-
iour”, Procedia Engineering (Pub.: Elsevier), 65, 440-446, 2013.
[5] Odrobiňák J. “Verification of Flexural Behavior and Simplified Modeling of Steel-Concrete
Composite Bridge”, Transactions of the VŠB – Technical University of Ostrava, Civil Engi-
neering Series (Pub.: De Gruyter), 14(1), 67-74, 2014.
[6] Bujňák J, Odrobiňák J. “Cracking Of Concrete Deck in Composite Structures”, Proceed-
ings of the 4th International Conference Eurosteel 2005, vol. B (Eds.: B. Hoffmeister, O.
Hechler, Pub.: Verlag-Mainz), Maastricht, The Netherlands, 4.2.15-4.2.22, 2005.
[7] Bujňák J, Odrobiňák J. “On design of composite beams with concrete cracking”, COST C12
Final Conference Proceedings, Innsbruck, Austria, 441-446, 2005.
108
Tampere, Finland
NEW CYCLE COUNTING METHOD FOR THE ASSESSMENT OF LOW CYCLE

FATIGUE IN STEEL H-PILES OF INTEGRAL BRIDGES
Memduh Karalar a,*, Murat Dicleli b

a,b
*
Author for contact. Tel.: +90 (312) 210-2395; E-mail: memduh@metu.edu.tr,
Abstract
Several cycle counting methods exist in the literature for the study of fatigue dam-age gener-
ated in structures. However, these methods do not take into consideration the prima-ry small
amplitude and secondary small amplitude strain cycles. Thus, a new cycle counting method is
developed. The developed cycle counting method is then used to estimate the num-ber of
large, primary and secondary small amplitude strain cycles. Then, an equation is devel-oped to
estimate the fatigue life of integral bridge steel H-piles. It is observed that secondary small
amplitude cycles do not have a very significant effect on the low cycle fatigue life of steel H-
piles. Experimental studies are conducted to verify the analytical results.
Introductıon
The daily and seasonal temperature changes result in imposition of cyclic horizontal dis-
placements on the continuous bridge deck of integral bridges and thus on the abutments, back-
fill soil, steel H-piles, and cycle control joints at the ends of the approach slabs. Due to these
seasonal temperature changes the abutments are pushed against the approach fill and then
pulled away, causing lateral deflections at the tops of the piles that support the bridge as ob-
served from Fig.1 (French et al. 2004). The magnitude of these cyclic displacements is a func-
tion of the level of temperature variation, type of the superstructure material and the length of
the bridge. As the length of the integral bridges gets longer, the temperature-induced cyclic
displacements and forces in steel H-piles components may become larger as well. This may
result in the reduction of their service life due to low-cycle fatigue effects (Dicleli & Albhaisi
2003, Arsoy et al. 2004).
Figure 1. Lateral deflections at the tops of the piles
109
In this study, the field measurements obtained for integral bridges are used to determine the
amplitude and the number of temperature induced cycles on steel H-piles in integral bridges.
Using the obtained measurements, the number of large strain cycles per year due to seasonal
temperature changes and the number and relative amplitude (relative to the amplitude of large
displacement/strain cycles, i.e. β=small strain cycle amplitude / large strain cycles amplitude)
of small strain cycles per year due to daily or weekly temperature changes are determined.
Additionally, the number of small cycles (secondary cycles) between the maximum and min-
imum cycle above and/or under the large strain is counted. Using the available data on the
number and amplitude of temperature induced displacement-strain cycles, a new cycle count-
ing method is developed to determine the number and amplitude of large and small displace-
ment/strain cycles (small strain cycles are composed of primary and secondary strain cycles).
Then, a new equation is obtained to determine a dis-placement/strain cycle amplitude repre-
sentative of a number of small amplitude cycles (primary and secondary) existing in a typical
temperature induced displacement/strain history in steel H-piles of integral bridges.
It is found that, the secondary strain cycles have a negligible effect on low cycle fa-
tigue life of steel H piles in integral bridges.
Conclusions
Field test results of several integral bridges in the US are studied to better understand the ef-
fect of thermal fluctuations on integral bridges. Then, the field measurements obtained for in-
tegral bridges are used to determine the amplitude and the number of temperature induced cy-
cles on steel H-piles in integral bridges. Using the obtained measurements, the amplitude of
large strain cycles and the number and relative amplitude, β of small strain cycles per year due
to daily or weekly temperature changes are determined. Additionally, the number of secondary
small cycles between the maximum and minimum cycle above and/or under the large strain
cycles is counted. Using the available data on the number and amplitude of temperature in-
duced displacement or strain cycles, a new cycle counting method is developed. Consequent-
ly, it seems that small amplitude cycles do not have a very significant effect on the low cycle
fatigue life of steel H-piles (difference ranges between 2% and 9%). To verify new cycle
counting method’s results, experimental test set up is prepared. Tests are also conducted to
investigate the effect of small amplitude strain cycles combined with large amplitude strain
cycles on the low cycle fatigue performance of the steel H-piles. First, the effect of small am-
plitude strain cycles (ratio of small to large strain amplitude is taken as 0.30) is investigated
using an HP220x57 steel section without the presence of axial load. It is observed that the
small amplitude strain cycles do not have a significant effect on the low cycle fatigue perfor-
mance of the steel H-piles. This confirms the earlier analytical observations.
REFERENCES
[1] Arsoy S., Duncan J.M., Barker R.M. 2004. Behavior of a se-miintegral bridge abutment
under static and temperature-induced cyclic loading. Journal of Bridge Engineering, Vol.9,
No. 2
[1] Dicleli M, Albhaisi SM.2003. Maximum lengths of integral abutment bridges based on
the strength of abutments and the performance of steel H-piles under cyclic thermal load-ing.
BU-CEC-03-01, Department of Civil Engineering and Construction, Bradley University, Peo-
ria, IL.
[1] French C., Huang J., Shield C. 2004. Behavior of concrete in-tegral abutment bridges.
Final Report.
110
Tampere, Finland
RESISTANCE OF ECCENTRICALLY LOADED BEAM-COLUMNS
Josef Vicana,, Peter Janikb

a,b
University of Zilina, Faculty of Civil Engineering, Department of Structures and Bridges
Abstract: The paper presents the results of the experimental and numerical analyses of the
resistance of the pinned-fixed beam-column subjected to eccentrically acting axial compres-
sive force compared to standard approaches to the beam-column resistance verification.
1 Introduction
The verification of the beam-column resistance subjected to the axial force in combination
with the bending moments is a very complex task from the viewpoint of design practice. The
standard approach to beam-column resistance assessment according to EN 1993-1-1 is based
on the simplified model of substitute member whose accuracy and suitability is controversial
in the special cases of member load effects and its boundary conditions. Therefore, the paper
presents results of experimental and numerical analyses of the determination of the beam-
column resistance compared to the standard verification of beam-column resistance.
2 Theoretical background
Generally, the bending and torsional moments' equilibrium of the pin-ended beam-column ini-
tially imperfect about both axes subjected to the compressive axial force N and bending mo-
ments My, Mz, could be described by means of the system of three differential equations. The
exact solution in the analytical form should be only found, when the axial force N is acting
eccentrically with respect to both cross-sectional axes and the eccentricities are constant
within the length of the beam-column, so that bending moments My, Mz have constant shapes.
3 Experimental investigation
The main objective of the experimental investigation was to determine the actual ultimate re-
sistance of tested pinned-fixed beam-columns subjected to the eccentrically acting compres-
sive axial force causing the end bending moments and also to verify the correctness of stan-
dard approach to the assessment of beam-column resistance. Four sets of beam-column sam-
ples, designated as A, B, C and D according to the type and eccentricity magnitude, were in-
vestigated. Each set of samples comprised three beam-columns of 1400 mm long, made of
IPE 120 introducing member relative slenderness of λ = 0.82 . Beam-columns of set A were
tested with the zero eccentricity of axial force, samples of set B had the preliminary measured
eccentricity of ey = 32.2 mm in the y-axis direction, samples of set C had the eccentricity of ez
= 52.2 mm measured in the z-axis direction and beam-columns of set D had the both type of
111
eccentricities of axial force, i.e. ey = 32.2 mm and ez = 55.2 mm measured before testing. The
actual geometrical and material characteristics of the beam-columns were determined and
evaluated statistically. The initial bow imperfections in direction of both axes were measured
by means of geodetic method. The strains and lateral deflections in chosen beam-columns lo-
cations were monitored using gauges 6/120 LY11 (HBM) and potentiometer sensors of de-
formations TR50 recorded by means of Spider 8. Results of the experimentally determined
beam-column ultimate resistances are presented in Table 1.
Table 1: Ultimate resistances Nexp of tested members

Nexp Nexp Nexp Nexp
Designation Designation Designation Designation
(kN) (kN) (kN) (kN)
A1 346.0 B1 133.0 C1 221.0 D1 107.0
A2 345.0 B2 149.0 C2 220.0 D2 114.0
A3 341.0 B3 148.0 C3 211.0 D3 112.0
Average 344.0 Average 143.3 Average 217.3 Average 111.0
4 Numerical analysis
Numerical models of tested members were developed in the working environment of the
software Ansys -Workbench using actual geometric characteristics and actual eccentricities of
tested members by means of the 3D finite elements Solid 186 and Solid 187, enabling geo-
metrically and materially nonlinear analysis with imperfections. Effects of residual stresses
were taken into account by means of equivalent geometric imperfection in accordance with
EN 1993-1-1. The bilinear material model with actual yield strength fy = 300 MPa and the
nominal value of the Young’s elasticity modulus E = 210 GPa was used to approximate mate-
rial behavior. Results of numerically calculated ultimate resistances Nnum of tested members
A1, B1, C1 and D1 compared to experimentally determined ones Nexp and compared to ap-
proach according to EN 1993-1-1 using method 1 (N1) and 2 (N2) are presented in Table 2.
Table 2: Comparison of experimentally and numerically determined resistances to the standard ones
Nexp Nnum Nnum/ Nexp N1 N2 N1/Nexp. N2/ Nexp
Designation
[kN] [kN] [kN] [kN] [%] [%]
A1 323.00 346.16 1.000 295.50 295.50 0.915 0.915
B1 133.00 133.81 1.006 126.00 131.00 0.985 0.895
C1 221.00 226.88 1.027 214.00 196.00 0.968 0.887
D1 106.60 111.4 1.045 122.00 110.00 1.144 1.032
5 Conclusions
From the results analysis the following conclusions could be done:
1. comparison proved very good correspondence of the results of numerical calculations
with the results of experimental tests;
2. the developed numerical models could be used for further parametric studies to obtain
more information about actual behavior of those complicated structural members;
3. the standard approach designated as method 1 also proved relatively good compliance
due to its more precise but also more complicated formulation;
4. the experimental and numerical analyses confirmed the correctness of theory related to
importance of the point with the maximum effect of the second-order theory.
112
Tampere, Finland
EXPERIMENTS ON PLATE GIRDERS WITH A VERY SLENDER WEB
Roland Abspoela
a
Delft University of Technology, The Netherlands
1 Introduction
This paper reports about the experiments carried out in the Stevin II Laboratory of Delft Uni-
versity of Technology in the framework of the PhD-research by the author on his investiga-
tions on the distribution of a certain amount of steel over flanges and web of an I-shaped dou-
ble symmetric plate girder to achieve the maximum bending moment resistance. The lever
arm between both flanges is limited by the phenomenon of flange induced buckling as deter-
mined by Basler [1] and given in EN1993-1-5 [2]. The experiments show that the maximum
web slenderness is not actually based on this phenomenon and the maximum web slenderness
can be increased enormously and so the bending moment resistance increases.
In this study the ultimate bending moment resistance of a fabricated plate girder, given a cer-
tain weight per unit length, is the main topic for optimization. Using higher steel grades, ap-
plying most material in the flanges and increasing the lever arm between both flanges are the
main possibilities to maximize the bending moment resistance of a plate girder under pure
bending. In case of a constant cross-sectional area, by increasing the lever arm, more material
is placed in the web, reducing the left over material for the flanges. The lever arm can also be
increased by increasing the web height and decreasing the web thickness. This process is re-
stricted by ending up with a practical thickness of the web, to make welding and also handling
of the plate girder possible.
2 Description of the test girders

The design of the test specimens is based on the ratio of area   Aw Atf and the web slender-
ness  w  hw tw , related on the dimensions of the plate girder Aw  hw  t w and Atf  btf  ttf ,
see Fig. 1 for the symbols. The web slenderness  w of interest lies in-between 400 and 800,
related to the ratio of areas  which lies in-between ½ and 2 according to Basler [1]. Based
on Basler’s formula for the maximum web slenderness  w.max , see Eq.(1), the web slenderness
 w lies in-between 360 and 720 for S235 assuming a residual stress level of  r  f y. f 2 .
2 Aw E2
 w.max    (1)
24  1   2  Atf f y .tf   f y .tf   r 
Because of the shift of the neutral axis, due to the use of the effective width method for the
class 4 web, the maximum web slenderness can be higher, namely in-between 400 and 800.
The slenderness of the flanges are based on the cross section classification according to
EN1993-1-1 [3]. In Abspoel [4] to [7] the design of the scaled test girders is described.
113
btf
tw
bbf
Fig. 1: Cross section of the test girders
3 Conclusions
Based on the 10 laboratory test girders the following is concluded:
1. All test girders, except test girder 10 fail by yielding of this flange;
2. The actual web slenderness’s of all test girders, except of test girders 1 and 5, are higher
than the maximum web slenderness’s based on EN1993-1-5 [2], but the test girders did
not fail by flange induced buckling;
3. The bending moment resistance M u of all test girders, except of test girder 10, are higher
than the effective bending moment resistance M eff based on EN1993-1-5 [2];
4. Test girder 10 fails far before the effective bending moment resistance M eff is reached,
because of the huge initial imperfections caused by the fabrication process;
5. The webs do not buckle at the maximum load and act much stiffer. They support the com-
pressive flange much better than based on column buckling of the web;
References
[1] Basler, K. “Strength of plate girders”, PhD dissertation, Lehigh University Bethlehem,
1959.
[2] Eurocode 1993-1-5: Eurocode 3: Design of steel structures – Part 1-5. Plated structur-
al elements, European Committee for Standardization, Brussels, November 2006.
[3] Eurocode 1993-1-1: Eurocode 3. Design of steel structures – Part 1-1: General rules
and rules for buildings, European Committee for Standardization, Brussels, January
2006.
[4] Abspoel, R. “The maximum web slenderness of plate girders”. Proceedings of the 5th
European conference on steel and composite structures (Eds. R. Öfner, D. Beg, R.
Greiner and H. Ünterweger), Timisoara, Romania, 1185-1190, 2008.
[5] Abspoel, R. “Optimising plate girder design”. Proceedings of the 11th Nordic Steel Con-
struction Conference (Eds. , Malmö, Sweden, September, 2009.
[6] Abspoel, R. and Bijlaard, F.S.K. “Optimizing of plate girders”, Steel Construction, 7,
116-225, 2014.
[7] Abspoel, R. “The maximum bending moment resistance of plate girders”. Proceedings
of the 7th European conference on steel and composite structures (Eds. Raffaele Lan-
dolfo and Frederico M. Mazzolani), Naples, Italy, 2014
114
Tampere, Finland
EXPERIMENTAL STUDY INTO BENDING-SHEAR INTERACTION OF

ROLLED I-SHAPED SECTIONS
R.W.A. Dekkera, H.H. Snijdera and J. Maljaarsa,b

a
Eindhoven University of Technology, the Netherlands
b
TNO, Delft, the Netherlands
Abstract: The bending-shear resistance of steel cross-sections is covered in EN 1993-1-1 [1],

taking presence of shear into account by a reduced yield stress for the shear area. An experi-
mental investigation on bending-shear interaction of rolled HE280A beams in S235 and S355
was performed by means of 3-point bending tests. Two criteria were considered for the evalu-
ation of the experimental results, being the load at which complete yielding of the cross-section
occurred and the ultimate resistance, i.e. the top of the load-displacement curve. Experimental
results complied with the design rule when the criterion of the ultimate resistance was chosen.
Yielding as failure criterion generated unconservative results for shorter beams.
Conclusions
In EN 1993-1-1, bending-shear interaction is taken into account by reducing the yield stress for
the shear area. This results in the reduced plastic resistance moment allowing for the shear force
MV. This paper presents the experimental test results of HEA280 beams in strong axis bending
and gives a comparison with the EN 1993-1-1 design rule.
The bending-shear interaction graph of Fig.1 shows the substantial differences between three
design standards for a HE280A section in S235 and S355. This figure indicates the influence
caused by the use of different shear areas. The middle graph of Fig. 1 presents the test results
in a non-dimensional bending-shear interaction graph with the 3 design rules. The shear utili-
zation ratio V/Vpl (V being the shear force and Vpl the plastic shear resistance) is plotted on the
horizontal axis and the bending utilization ratio M/Mpl (Mpl being the plastic moment resistance)
on the vertical axis.
Fig. 1: M-V graph comparing EN, DIN and NEN (left and middle); Fexp-Ftheorygraph (right)
115
Fourteen HEA280 beams were tested in 3-point bending, see Fig. 2 (left), of which 7 in steel
grade S235J0+M and 7 in S355J2+M. Different spans (L) – 1083, 1400, 1790, 2426, 3630 mm
– were used to invoke a range of utilization ratios V/Vpl, namely 1.00, 0.83, 0.67, 0.50, 0.33.
Table 1: Test results of HEA280 beams in S235 and S355

Fy Vy My Using fyf, fyw, Aw, fyw, Av Fu Vu Mu Using fyf, fyw, Aw, fyw, Av
(kN) (kN) (kNm) Vy/Vpl My/Mpl (kN) (kN) (kNm) Vu/Vpl Mu/Mpl
A1 411 206 373 0.33 1.15 443 222 402 0.35 1.24
A2a 594 297 360 0.47 1.11 664 332 403 0.53 1.24
A2b 615 308 373 0.49 1.15 659 330 400 0.52 1.23
A3 754 377 377 0.60 1.04 896 448 401 0.71 1.23
A4a 789 395 276 0.63 0.85 1047 524 366 0.83 1.13
A4b 773 387 271 0.61 0.83 1083 542 379 0.86 1.16
A5 839 420 227 0.67 0.70 1219 610 330 0.97 1.01
B1 525 263 476 0.40 1.19 528 264 479 0.40 1.19
B2a 737 368 447 0.56 1.12 794 397 481 0.60 1.20
B2b 744 372 451 0.57 1.13 729 365 442 0.56 1.10
B3 891 445 398 0.68 1.00 1067 534 478 0.81 1.19
B4a 918 459 321 0.70 0.80 1237 618 433 0.94 1.08
B4b 938 469 328 0.71 0.82 1266 633 443 0.97 1.10
B5 963 482 261 0.73 0.65 1356 678 367 1.03 0.92
Fig. 2: Left: test set-up and measurement positions, right: load-displacement graph for beam A4b
All tests failed in bending-shear interaction. In the shorter specimens shear deformations dom-
inated, while in the longer specimens mainly bending deformations were observed. The load at
which the complete section yielded (Fy) and the ultimate load (Fu) were determined (Fig. 2
right) and listed in Table 1 for all specimens. For longer specimens the difference between yield
or ultimate load is not significant. All standards considered provide conservative resistances for
these longer specimens. This is in contrast to the short specimens where strain hardening results
in a significant difference between Fu and Fy. The standards are conservative in case Fu is con-
sidered, but unconservative if Fy is considered as resistance for short beams (Fig. 1 right). The
EN 1993-1-1 design rule complies best with Fu test results.
Acknowledgments
This research has received funding from the European community’s Research Fund for Coal
and Steel (RFCS) under grant agreement no. RFSR-CT-2013-00023. The specimens were pro-
vided by ArcelorMittal and tested at the Pieter van Musschenbroek Laboratory, TU/e.
116
Tampere, Finland
EFFECT OF NEUTRAL-AXIS POSITION

ON THE ELASTIC BUCKLING CHARACTERISTICS
OF CONTINUOUS COMPOSITE BEAMS
Daigo SHIRAIa, Kikuo IKARASHIa

a
Abstract This study uses theoretical analysis to show the relationship between the neutral-axis
position and the elastic buckling characteristics of H-shaped beams used as composite beams.
It also describes an estimation of the elastic buckling strength of composite beams. We found
that the web local buckling strength under bending stress is especially subject to the neutral-
axis position and that the elastic buckling strength of composite beams may decrease under
negative bending. Furthermore, we found that the effect of the neutral-axis position on the
buckling strength is roughly determined by the width–thickness ratio of the web and flange.
1 Introduction
Although many studies have been conducted on the buckling characteristics of composite
beams, few studies examine both restraint on the upper flange and change of the neutral-axis
position by theoretical method. Moreover, composite-beam problems involve numerous factors,
and the buckling characteristics are not sufficiently understood. Therefore, it is important to
clarify the effect of neutral-axis position on the buckling characteristics of H-shaped beams in
an ideal condition where some of the factors are not present. This study demonstrates the
relationship between the neutral-axis position and elastic buckling characteristics.
2 Outline of elastic buckling analysis

Fig. 1 shows the analytical model.  is a parameter
of the neutral-axis position and is constant in the
axial direction. The analysis objects are both ends
fixed beams under negative bending. Analysis is
conducted by energy method in consideration of
continuous restraint on the upper flange and
neutral-axis position. Elastic buckling coefficient
Fig. 1: Theoretical analytical model K is an indicator of the elastic buckling strength.
3 Effect of the neutral-axis position to elastic buckling characteristics

Fig. 2 shows the relationship between K and the aspect ratio w. According to Fig. 2, K
decreases by moving upward of neutral axis. Fig. 3 shows the buckling waves for four values
117
25 K
H-150×75×5×7
 = 0.5
 = 2.0
20 0.6
0.7
0.8
15 0.9
1.0
10
Web local buckling
Flange local buckling
Lateral-torsional buckling
5
0 5 10 15 20 25 30 35 w
Fig. 2: Effect of neutral axis on buckling strength Fig. 3: Buckling waves
of w, shown by ▽, , and ▼ symbols in Fig. 2. According to Figs. 2 and 3, the behavior of
lateral-torsional buckling and local buckling caused by shear stress is hardly affected by . In
contrast, the decrease of local buckling caused by bending stress is large. The bending stress
distribution varies markedly with upward movement of the neutral axis, which is different from
the shear stress distribution. However, the lateral-torsional buckling strength is hardly affected
by the change of the stress distribution in the web and is determined by the amount of the
bending stress at the lower flange. As a result, it is important to clarify the effect of the neutral-
axis position on the behavior of local buckling caused mainly by bending stress.
4 Effect of the neutral-axis position to elastic local buckling characteristics

Local buckling analysis is conducted. As a result, the kd
0.0
relationship between  and the decrease of K in local Section which
flange local buckling
buckling caused by bending stress is determined solely -0.5 web local buckling
by the cross-sectional shape of the H-shaped beam. is caused when  is equal to 0.5
Furthermore, as compared to flange local buckling, the
-1.0
behavior of web local buckling is more affected by . d  kd (  0.5)  1
Fig. 4 shows the relationship between kd and cross- K  d  ( K ) 0.5
-1.5
sectional shape in the realistic range of  from 0.5 to
0.75. Here, kd is an indicator of the degree of decrease
-2.0
of K. The smaller kd is, the larger the amount of  ; 0.5 to 0.75
decrease of K is. The elastic buckling strength of w = 30
-2.5
composite beams can be derived using the equations in 0 2 4 6 8 10 12 14 b16w t w18
Fig. 4 and approximate formulas for the buckling b f tf 
strength of a pure H-shaped beam.
Fig. 4: Raletionship between kd and
cross-sectional shape
5 Conclusions
The elastic buckling strength of composite beams was theoretically derived in consideration of
changes in the neutral-axis position. When maximum bending stress was used as an indicator
of the elastic buckling strength, the strength decreased by moving upward of neutral axis. The
strength of web local buckling caused by bending stress is especially subject to the neutral-axis
position. Finally, the effect of the neutral-axis position on the elastic local buckling strength
was roughly determined by the width–thickness ratio of the web and flange.
118
Tampere, Finland
AMPLIFIED SEISMIC LOADS IN STEEL MOMENT FRAMES
Bora Aksara, Selcuk Dogrua, Bulent Akbasa*, Jay Shenb, Onur Sekerb, and Rou Wenc
a
Gebze Technical University, Department of Earthquake and Structural Engineering, Gebze, Turkey
b
Iowa State University, of Civil, Construction and Environmental Engineering, Ames, IA, USA
*Bulent Akbas. Tel.: +90 262 6051651; E-mail: akbasb@gyte.edu.tr
During the 1994 Northridge Earthquake, many buildings with modern steel moment
resisting frames (SMRFs) suffered from connection failures. Similar damages occurred one
year later, in the 1995 Kobe earthquake in Japan. The unexpected seismic response of SMRFs
resulted in comprehensive analytical and theoretical investigations and major changes in steel
building design have been implemented consequently. One of the requirements in the
subsequent seismic design codes is the stability check of the columns. Column yielding in a
seismic force resisting systems (SFRSs) is not the desired damage mode and might result in
column rupture or global buckling and threaten life safety.
Seismic codes require that column stability should be checked under amplified seismic loads.
Seismic design procedure introduced in ASCE 7-10 (2010) defines some coefficients such as
response modification factor (R values), deflection amplification factors (Cd values), and
system overstrength factors (Ωo values). ASCE 7-10 (2010) acknowledges that structures will
be loaded beyond their elastic range during strong ground motions.Seismic axial loads in
columns (tension or compression) that might be induced during strong ground motions have
substantial impact on the stability of columns.. Axial tensile forces can cause substantial
demands especially on column splices. In tall frames, tensile axial force can have a significant
impact on the column splices even under moderate ground motions. The effect of tensile
forces combined with peak bending moment has been investigated in detail in Shen et al.
(2010) and Akbas et al. (2013). The overwhelming majority of this study focuses on
exploring the seismic axial compression loads for columns in SMRFs under strong ground
motions. For this purpose, the increase in axial loads in low-, medium-, and high-rise SMRFs
are investigated at the maximum lateral load level, Vmax, and the corresponding lateral
displacement. Nonlinear dynamic time history analyses are conducted on three SMRFs with
4-, 9-, and 20-stories under a set of strong ground motions.
This study focuses on exploring the seismic axial loads for columns in SMRFs under strong
ground motions. For this purpose, the increase in axial loads in low-, medium-, and high-rise
SMRFs are investigated at the maximum lateral load level and the corresponding lateral
displacement. The results are presented in terms of plastic hinge rotations, average system
overstrength factors (Ωo) of all columns in the frames under the selected ground motions, the
distribution of Ωo in the individual columns in the frame, and axial load levels in columns.
The first three gives the amplification levels in the axial compression loads, but nothing about
119
the column stability and reserved capacity. The main outcomes of this study can be
summarized as follows:
1. The maximum average o in the individual columns in each story of the low-rise
frames varies between 1.75 and 2.30. For medium-rise frames, the maximum average
o values remains below 2.1.
2. The maximum average o in the individual columns in each story of the high-rise
frames varies between 1.30 and 2.60.
3. The maximum o occurs in the exterior columns in low-, medium-, and high-rise
frames. The o factor increases in upper stories.
4. Under the combination of D+0.25L, for low-rise and medium-rise frames the average
axial load levels are less than 6% and under strong ground motion, the maximum axial
load level remains below 0.20 and 0.35.
5. For the high-rise frames, the average axial load levels are less than 15% for the
combination of D+0.25L and under strong ground motion, the maximum axial load
level can get as high as 0.954.
120
Tampere, Finland
Design rules for slim-floor girders considering the composite behaviour
Univ.-Doz. Dr.-Ing. Markus Schäfera

a
Research Unit in Engineering Sciences
UNIVERSITY of LUXEMBURG
Tel.: +352 466 644 5533; E-mail: markus.schaefer@uni.lu
Abstract: Due to the demand for sustainable buildings and slim constructions, composite
slim-floor systems become more important. The present European Codes do not include com-
plete design rules for slim-floor beams. Therefore, the objective of the recent research work
was concentrated on the development of additional regulations, considering ultimate limit
state, serviceability limit state and fire design. Apart from traditional sections consisting of
rolled profiles also floor-slabs with integrated steel-box sections have been considered.
1 Introduction
The integration of steel-profiles in flat concrete slabs allows innovative composite ceiling sys-
tems. They provide optimal conditions for the installation of technical building service. The
efficiency of composite slim-floors slabs results from high ratio of prefabrication and the as-
sociated reduction of erection costs. The low self-weight of the structure allows small cross-
sectional dimensions for the slab and the following components up to the foundations. In ad-
dition, by the low slab height results a reduction of the construction volume, this leads to sav-
ings in facade and maintenance costs. Furthermore, the integration of the steel-section in the
concrete slab leads to a favourable ratio of the flamed steel surface to the volume of the sec-
tion, hence a high fire resistance period can be achieved without additional measures. Recent
developments of slim-floor beams are focused on activating also the concrete chord to reach a
composite behaviour and high bearing capacity.
Fig. 1: Slim-Floor beams with composite bearing behaviour
2 Resistance in ultimate limit state

In final state the composite action between steel and concrete results from ductile shear con-
nectors, realized by headed studs. Equally, the composite behaviour can be ensured by open-
121
ings arranged in the web or upper flange, work concurrently as concrete dowels. The determi-
nation of moment resistance follows the rules for composite structures, respectively whereas
in many cases the strain limited design is relevant and additionally the influences from trans-
verse bending in the bottom flange have to be considered.
Fig. 2: Strain limited and full plastic moment resistance
Because of the sheathing of the concrete by the steel cover in case of concrete-filled box-
sections a hybrid truss model with a compression strut in the concrete is developed, increas-
ing the shear resistance of this systems.
Fig. 3: Vertical shear resistance, hybrid truss model
By reason of the additional stresses in the web (Fig. 3), the interaction between moment and
vertical shear forces according to EN 1994-1-1 is not sufficient for the design. The tension
stresses from the hybrid truss model are to consider for the MV-interaction. In case of open-
ings in the web, also the influence of the secondary bending moment is to respect.
3 Design in fire situation

In case of fire, the directly flamed bottom flange can be substituted by reinforcement bars and
a high fire resistance can be realized without any additional activities. For the fire-analysis, a
design-method is extracted according to EN 1994-1-2. Based on a transient analysis thermal
analytic functions are developed to describe the temperature distribution in the cross-section.
4 Design in serviceability limit state

The crack behaviour of the concrete slab has already a significant impact on the girder defor-
mation. The concrete flange adopts a remarkable part of the bending-moment. Compared to
general treatment of common composite girders, the neglecting of this effect can lead to an
unrealistic camber of the beam. Therefore, an approximation procedure is derived, consider-
ing the resilience of composite connectors and the crack behaviour on the structure defor-
mation.
References
[1] Schäfer M. „Zum Tragverhalten von Flachdecken mit integrierten hohlkastenförmigen
Stahlprofilen“, Dissertation, Institut für Konstruktiven Ingenieurbau, Heft 8, Bergische
Universität Wuppertal, 2007.
122
Tampere, Finland
EFFECT OF LONGITUDINAL STIFFENERS ON THE FLANGES TO IMPROVE

THE LOW CYCLE FATIGUE PERFORMANCE OF STEEL H- PILES
Memduh Karalar a,*, Murat Dicleli b

a,b
*
Author for contact. Tel.: +90 (312) 210-2395; E-mail: memduh@metu.edu.tr,
Abstract
In this study, the effect of stiffeners on the low cycle fatigue life of steel H-piles in integral
bridges is investigated to prevent local buckling occurring at high strain amplitudes. For this
purpose, experimental testing of a number of regular and stiffened steel H-piles under cyclic
displacement reversals is conducted. Then, to compare experiment results with the finite
element model for the HP220x57 steel specimen, finite element model is constructed similarly
according to actual HP220x57 steel specimen in the test set up.
Introduction
Each daily variation in temperature completes a cycle of expansion and contraction and the
cycles repeat over time as shown in Figure-1. The maximum expansion occurs during summer
days while the maximum contraction forms during winter nights. The extreme lateral dis-
placements of integral bridges are controlled by these extreme temperature changes. In inte-
gral bridges, when the steel H-piles at the abutments are subjected to large strain amplitudes
due to thermal induced displacements, the presence of axial load, depending on the flexural
strain level, is experimentally observed to decrease the low cycle fatigue life of the pile. This
is mainly due to local buckling of the flanges of the steel H-pile under the effect of compres-
sive strains from the axial load and high compressive plastic strains due to the effect of bend-
ing.
Detail-A
(a) (b )
Fig. 1. (a) Pile displacement due to thermal changes, (a) Thermal Expansion, (b) Thermal contraction
123
Local buckling increases the local curvature and strains. This locally accelerates the cracking
of the material resulting in earlier fracture of the material and associated reduction in the low
cycle fatigue life of the steel H-piles. To prevent this local buckling occurring at high strain
amplitudes, longitudinal stiffeners may be placed at the tip of the flanges of the steel H-pile.
To investigate the effect of these stiffeners on the low cycle fatigue life of steel H-piles, exper-
imental testing of a number of regular and stiffened steel H-piles under cyclic displacement
reversals (simulating thermal induced displacements in the steel H-piles) are conducted as
shown in Figure-2. It is observed that stiffening the flanges of the steel H-pile in the region of
maximum flexural strains significantly improved the low cycle fatigue life of steel H-piles
used in integral bridges.
Stiffener
Axial
load
Lateral Load
HP 220x57
Figure 2. Longitudinal stiffeners and Test set up
Conclusions
Due to these seasonal temperature changes the abutments are pushed against the approach fill
and then pulled away, causing lateral deflections and buckling at the tops of the piles that
support the bridge. Therefore, low-cycle fatigue may occur in piles of long integral bridges.
these local buckling increases the local curvature and strains in the steel H piles of integral
bridges. This locally accelerates the cracking of the material resulting in earlier fracture of the
material and associated reduction in the low cycle fatigue life of the steel H-piles. To prevent
this local buckling occurring at high strain amplitudes due to thermal induced displacements
in the steel H-piles, longitudinal stiffeners is placed at the tip of the flanges of the steel H-pile
in this study. To investigate the effect of these stiffeners on the low cycle fatigue life of steel
H-piles, experimental testing of a number of regular and stiffened steel H-piles under cyclic
displacement reversals (simulating thermal induced displacements in the steel H-piles) are
also conducted. these tests are conducted on piles with stiffeners welded to the flanges to
delay local buckling and hence, to improve the low cycle fatigue performance of the piles
under axial load and large amplitude strains. It is observed that stiffening the flanges of the H-
pile in the region of maximum flexural strain enhanced its low cycle fatigue life more than
20%.
124
Tampere, Finland
REFINED FATIGUE ASSESSMENT OF AN EXISTING STEEL BRIDGE
John Leandera, Raid Karoumia

a
Division of Structural Engineering and Bridges, KTH Royal Institute of Technology
Abstract: This paper treats the fatigue assessment of existing steel bridges for road traffic. It
is focused on the estimation of the load effect. The deterministic assessment methods suggest-
ed in governing codes are reviewed and a comparison is performed against a reliability-based
assessment. The latter enables a consideration of reduced uncertainties from measurements of
the real load effect. The Vårby Bridge in Sweden, a steel–concrete composite bridge south of
Stockholm, is used as a case study. The results show a considerable increase in fatigue life
with the use of measurements and a reliability-based assessment. Another conclusion is that
the load models in the Eurocode give an unjustified conservative result.
1 Introduction
The conventional assessment methods following the Eurocodes are reviewed. The basic fea-
tures of the methods are summarized including the resistance and the load. By long term
measurements, a representative stress range spectrum for the instrumented detail can be rec-
orded and used in the fatigue assessment. A reliability-based model is used to incorporate the
measured stresses. It enables an assessment against a target reliability and a consideration of
uncertainties related to the model and the measured response.
2 Results
The results presented are calculated for the Vårby Bridge south of Stockholm in Sweden. It is
composed by two parallel steel–concrete composite bridges carrying the highway E4 between
Stockholm and Södertälje. Both bridges have the same design and are continuous bridges in
six spans with the total length of 255 meters. The span lengths are 38 m and 44 m for the end
spans and the intermediate spans, respectively. Each cross-section is built up by two I shaped
steel beams and a concrete deck.
The bridge were instrumented with strain gauges and measurements were performed at the
end of June and beginning of July 2009. The total duration corresponds to about three days.
The short duration of the measurements makes the estimation of a representative traffic vol-
ume highly uncertain. Two stress range spectra from the measurements are shown in Fig.
1(a). Gauge 8, located on the beam carrying the larger portion of the load in the slow lane, has
the greatest number of cycles for all stress ranges.
125
2000 14 ΔσC = 125 MPa

Gauge 8 12 ΔσC = 80 MPa
1500 Gauge 4 10 ΔσC = 40 MPa
8
β
nE
1000
6
4 β = 3.1
500
2 β = 2.3
0 0
0 10 20 30 40 50 20 60 100 140 180
Δσ/MPa T/years
(a) Stress range spectrum. (b) Reliability index
Fig. 1: Results for the measured response.
A reliability index β is calculated using the first order reliability method (FORM). The result
is shown in Fig. 1(b) as the reliability index over time. Two reliability levels are indicated in
the figure, β = 3.1 and β = 2.3. These are the target reliability indices stated in ISO 13822 for
fatigue assessment of existing structures. The higher value is suggested for non inspectable
components and the lower value for inspectable components. All results are summarized in
Table 1.
Table 2: Results of the different assessment methods.
Method Traffic Fatigue life
ΔσC = 125 MPa ΔσC = 80 MPa ΔσC = 40 MPa
Deterministic FLM3 ∞ 50 < 20
FLM4 ∞ 67 < 20
Real ∞ > 200 < 20
FORM, β = 3.1 Real > 200 > 200 24
FORM, β = 2.3 Real > 200 > 200 41
3 Conclusions
The following conclusions are based on the fatigue assessment of a specific section of the
Vårby Bridge in Stockholm, Sweden. The assessment is performed with different methods
and for three different connection details frequently occurring in steel bridges.
1. The load models from the Eurocode gives a conservative estimate of the fatigue life in
comparison to the life determined for the measured response. For detail category 80,
the fatigue life is estimated to 50 and 67 years for FLM3 and FLM4, respectively. A
deterministic assessment based on measured response from real traffic gives a fatigue
life longer than 200 years.
2. A reliability-based assessment using measured response increases the estimated fa-
tigue life even further.
3. The reliability-based assessment shows that the partial safety factors used in the de-
terministic verification are appropriate also for measured response.
4. For the Vårby Bridge, the fatigue load model 3 in the Eurocode limits the service life
to 50 years for detail category 80. The reliability-based assessment gives a reliability
index of about β = 6 for the same service life which is significantly higher than the
suggested target reliabilities in ISO 13822.
126
Tampere, Finland
ODINS BRIDGE
Kjeld Thomsen, MSc. CEO
ISC Consulting Engineers A/S, Oster Alle 31, 2100 Copenhagen
Abstract: The longest swing bridge in Europe is now in operation.
In September 2009 ISC Consulting Engineers A/S won the design competition for a 900 m
long bridge connection comprising a bridge crossing the 80 m wide Odense navigation canal
and on the western side of the canal an approximately 540 m approach bridge and dam joining
the main circular road.
The key part of this connection is the 194 m long swing bridge. The main structure is de-
signed in steel as a twin box girder bridge with orthotropic steel deck with a 3 m clearance
between the box girders. The centre span of the bridge is 120 m and the side spans are 37 m
each. The bridge carries two lanes in each direction as well as pedestrian and bicycle paths of
4 meters width. The bridge has been designed for 100 year lifetime.
The two main supports for the swing parts are pulled back from the embankment, ensuring
that no parts of the bridge would be located outside the shores of the navigation canal in the
open condition.
The main concrete supports carrying the permanent bearings and the bearings for rotation
have a diameter of 12.0 m.
The superstructure is a monoplane triangular shaped support structure with top level 20 me-
ters above the road surface.
The bearing system for the bridge is a new invention, based on sliding bearings for rotation of
the bridge. 2 permanent bearings supports the swing section in service and 2 other bearings
are jacked up under the bridge to secure balance during rotation.
127
The swing bridge consists of two equal rotating parts which in the closed position are linked
to the end abutments with hydraulic operated shear bars. In the centre of the bridge shear bars
are provided as well.
The design started in September 2009. Construction started in beginning of 2010 and the
bridge was inaugurated June 15th 2014. The total cost is DDK 400 mill. The Bridge has been
awarded the Danish IABSE structure prize 2013.
128
Tampere, Finland
HIGH-PERFORMANCE-STEEL GIRDER OF COMPACT SECTION
E. Yamaguchi, Y. Sugimura, K. Ohmichi
Department of Civil Engineering, Kyushu Institute of Technology, Japan
Abstract: The maximum width-to-thickness ratios for compact I-shaped sections are first
studied for three girders: the SM490Y girder, the SBHS500 girder and the hybrid girder.
SM490Y is a conventional steel and SBHS500 is a high-performance steel. The hybrid girder
consists of a SM490Y web and SBHS500 flanges. The three girders with compact sections
are then designed for a given plastic moment. The result shows that the SBHS500 girder can
be the lightest. The cost study is also conducted and the hybrid girder turns out to be the most
competitive at the current steel price.
1 Introduction
In 2008, new steels, SBHS500 and SBHS700, were registered in Japanese Industrial Stand-
ards (JIS). Because of their high yield strengths and various advantages such as good welda-
bility, SBHS500 and SBHS700 are called high-performance steels.
Focusing on the high yield strength, the present study explores the effective use of SBHS500.
To this end, the maximum width-to-thickness ratios for compact I-shaped sections are first
obtained by nonlinear analysis. Based on those maximum width-to-thickness ratios, three I-
section girders are studied, two of which have homogeneous sections and one of which has a
hybrid section. The homogeneous section is made of either SM490Y or SBHS500, and the
hybrid section is of SM490Y for a web and SBHS500 for flanges, where SM490Y is a con-
ventional steel given in JIS. The optimum compact sections are then designed for the three
girders under a given plastic moment Mp. The results are compared and discussed.
2 Analysis Models
Fig. 1 presents the girder model to be analyzed. The cross section is I-shaped and doubly
symmetric.
129
bf
tf
M M
tw
bw
tf
bf
5000 mm
Fig. 1: Steel girder model
Table 1: Comparison of costs of three girders

Price Ratio (SBHS500/SM490Y)
C ＜ 1.08 ＜ 1.23 ＜ 1.50 ＜
o Low SBHS500 SBHS500 Hybrid Hybrid Hybrid Hybrid SM490Y
s Middle Hybrid Hybrid SBHS500 SBHS500 SM490Y SM490Y Hybrid
t High SM490Y SM490Y SM490Y SM490Y SBHS500 SBHS500 SBHS500
3 Compact sections (maximum Width-to-thickness Ratio)

The maximum width-to-thickness ratio is obtained numerically. Large difference between the
girders is noted. The cross sections whose width-to-thickness parameters lie below the maxi-
mum width-to-thickness ratio curve are compact.
4 Comparison of three girders

Under the condition of Mp = 5.0 x 1010 mm, the compact cross section having the smallest
cross-sectional area is obtained for each of the three girders. The smallest area of the three is
attained by the SBHS500 girder. The weight of the SBHS500 girder can be 19% less than that
of the SM490Y girder.
For the selection of the optimum girder, the cost is also an important factor. Table 1 presents
the comparison of the three girders in terms of the cost. The optimum girder varies, depending
on the price ratio of SBHS500/SM490Y. Since the current price of SBHS500 is about 33%
higher than that of SM490Y, the hybrid girder is considered the most competitive at present.
5 Concluding remarks
The maximum width-to-thickness ratios for the compact section are obtained for three girders.
Using those ratios, the optimum compact sections that attain the same Mp are designed for the
three girders. While the SBHS500 girder is found the lightest, the most economical girder
varies, depending on the price ratio of SBHS500 to SM490Y. At the current steel price, the
hybrid girder turns out to be the most competitive: its cost is 14% and 7% lower than those of
the SBH500 and SM490Y girders, respectively.
Acknowledgments
Financial support from the Japan Iron and Steel Federation for the present research is grate-
fully acknowledged.
130
Tampere, Finland
STEEL BRIDGE TECHNOLOGY USED IN BUILDING PROJECTS
Hans Exner
Msc. Eng., Ph. D. in structural engineering, Ramboll Denmark

Tel +45 5161 6558. E-mail he@ramboll.dk
Abstract:
Steel bridges have for many years been built with optimized closed boxes of welded steel
plates. A welded steel box is a strong and rigid structural element that can resist bending and
twisting forces. It is highly adabtible to various geometric arrangements. The technology has
now been used in buildings with architectural and functional challenges and demands for an
adaptable, strong, stiff and slender structure. This innovative unification of bridge and building
principles had not been realized before on a similar scale.
The major structural expressive element of the rewarded Opera Building in Copenhagen is the
roof that cantilevers horizontally over the front plaza. The main roof structure is a steel box
that has high bending and torsional strength and thereby reaches out to its far corners, with an
economical use of material and a minimum weight. The stiffness of the box structure ensures
against wind-induced vibrations as well as visible deflections. The arrangement of the outdoor
box and the indoor truss beams ensure unconstrained movements with low stresses at varying
outdoor temperatures.
The new Maritime Museum in Helsingør has a unique architecture. It is constructed below
grade around a former ship yard dock, which is left open to the sky and spanned by a series of
bridges, forming part of the new museum building. The box structures have been fully adapted
to the architectural geometry. The fully glazed facades without bracing were only possible with
the steel boxes. The facades and connections have been designed with respect of the deflec-
tions. Unpleasant vibrations have been avoided by damping in the structures and facades.
The new Central Library in Helsinki being in the design process now, includes in the 150 m
long building a bridge arch spanning 100 m and carrying the upper floors, including a public
balcony at level three cantilevering 14 m outside the building or 18 m horizontally perpendicu-
lar to the main arch. The arches have steel box section. The building has large glass facades.
Deflection and vibration aspects are being carefully analyzed.
A steel box plate structure can be adapted to the available space and to unusual support and
spanning arrangements creating plate bending forces in two directions as well as torsion. It
saves weight as the plates take stresses in several directions. However, a building has other
demands than a bridge. Therefore, nothing can just be copied, so the building projects have
been developed with careful consideration of any relevant aspect.
131
The Opera in Copenhagen has an extreme 43 m cantilevered roof, part of which is indoor while
other parts have outdoor temperatures. The layout and temperature movements had to be care-
fully designed in order to avoid high stresses.
Fig. 1: Opera Copenhagen. Fig. 2: Danish Maritime Museum
The Danish Maritime Museum in Helsingør has an open transparent architecture with exhibi-
tion and auditorium rooms on slender steel box structures spanning freely across a former ship
building dock. The architecture called for a structure that could be adapted to the special ge-
ometry and still be slender, strong and stiff. The design ensured sufficient movements in joints
of window panes and installation components under the deflections. Potential unpleasant vibra-
tions were avoided by damping.
In the table below, the design situations are compared, and some major focus items are shown.
Table 1: Properties
Steel bridges Opera Copenha- Maritime Museum
gen
Corrosion protection Minor internal protection
Geometric adaption Slender roof. Slopes and angles. Canti-

Large cantilever levered support arms. Fa-
cades totally glazed
Strength Important Important Important
Stiffness Potential vibration. Potential vibration. Vibration prevented by

Shape optimized Wind response ana- stiffness and damping in
against wind load. lysed. the facades. Flexible joints
Dampers. for facades and ventilation
Transport and erec- Ship or other. Ship. Site welding Ship. Site welding
tion Site welding
Differential tempera- Important Very important Important
ture
Material consump- Self weight is the Good Less important
tion major bridge load.
Had it not been for the steel box technology, the 3D design tools and the assertive cooperation
between architects and engineers these innovative buildings had not been built as they actually
were. Thus, the technology is a necessary tool of realizing the architectural visions.
132
Tampere, Finland
SUNDSVALL BRIDGE
Kjeld Thomsen1
Helge Skov Pedersen2
1
MSc CEO ISC Consulting Engineers A/S, Oster Alle 31, 2100 Copenhagen
2
MSc Chief Engineer ISC Consulting Engineers A/S, Oster Alle 31, 2100 Copenhagen
Abstract: The longest steel bridge to be constructed in Scandinavia since the inauguration of
the Øresunds link between Denmark and Sweden in 2000.
The Sundsvall Bridge will be crossing the Sundsvall Fjord with a total length of 1,420 m as
part of the new eastern highway in Sweden. The contract won in a European competition, by
PBM Joint Venture with ISC Consulting Engineers A/S as designers for the bridge super-
structure.
The key part of the connection has a total length of 2.1 km including approach spans, elevated
concrete abutments and the 1.42 km long steel bridge. The bridge is designed as a continuous
girder bridge with 11spans, a center span of 170 m and side spans of 88 m.
The bridge concept related to geometry is quite outstanding. The steel superstructure is curved
in the horizontal plane as well as in the vertical plane and one side slope transverse. The
bridge width varies from 26.2 m in the center span up to 36.9 m at the southern abutment and
38.8 m at the northern abutment.
The bridge superstructure is designed as a closed box girder bridge with orthotropic bridge
deck, wing shaped with tapered under flanges towards the edges. The height of the box girder
is 6.5 m in the center span reduced to app. 3.5 m at the abutments. The box girder is provided
with lattice diaphragms in distance app. 4 m and further with longitudinal webs as plate gird-
ers at the third points of the cross sections. None of the 400 bulk heads have equal geometry
which requires thorough accuracy in the construction phase of the steel structure. The bridge
133
has a free height at the center span of 33 m in the navigation channel of 90 m. The foundation
level is around 30 m below the water surface as a thick layer of mud and soft soil covers the
seabed.
It is anticipated that mass dampers have to be provided in all spans. The box girder spans are
supported via a V shaped support to the concrete pillars. The V shaped supports consists of
tubular members with a maximum diameter of 2 m.
The fabrication of the steel superstructure comprising approx. 22,000 ton was carried out by
Max Bögl in Munich in Germany and transported by barge on German waterways to Rotter-
dam and further on to Stettin, Poland. The bridge elements was assembled to full bridge spans
of approximately 2,000 ton. The 88 -170 m long bridge sections were loaded on barges and
sailed to the final destination in Sundsvall. The work commenced in 2011 and was terminated
in 2014. The total cost is approx. 1.5 bill. SKr.
134
Tampere, Finland
ALUMINIUM DEPLOYMENT IN BRACING SYSTEMS:

INVESTIGATION OF SHEAR LINK APPLICATION
Evangelos Efthymiou1, Vasileios G. Psomiadis e Alexios T. Ampatzis
Institute of Metal Structures, Department of Civil Engineering, Faculty of Engineering

Aristotle University of Thessaloniki, University Campus, GR 54124, Thessaloniki, Greece
Abstract
In the aftermath of several severe earthquakes, intense research activity has been carried out
in the last decades to advance seismic design studies as well as investigating different braced
framed design configurations, towards achieving more effective hysteretic behaviour. Moreo-
ver, alternatively to conventional approaches, a new seismic design trend is nowadays based
on controlling and limiting as much as possible the dynamic effects on the structural elements
produced by earthquakes. In order to exploit the dissipative action as much as possible, the
use of ductile metals with limited yielding strength is needed. For this aim, the adoption of
low yield strength (LYS) steel application has been proposed, i.e. in shear panel utilizations.
Considering that low yield strength steel is less available on the world market, the use of pure
aluminium as metal material to build shear panels has been introduced, as well as an alumini-
um beam shear-link was primarily developed.
Within this framework, the deployment of aluminium in seismic engineering application can
provide an effective choice, being a ductile material as well as demonstrating excellent metal-
lic yielding behaviour. The present paper investigates the application of aluminium shear link
in eccentric braced frames (EBFs) and evaluates its effectiveness in view of seismic response.
In the framework of the study, a prototype of a hybrid eccentric braced system is designed
according to contemporary Eurocode 3 provision, comprising of steel elements and alumini-
um shear link. For the purposes of the herein presented work, finite element techniques are
employed and nonlinear analysis is implemented. A comparison with an all steel solution is
conducted and useful conclusions on the effectiveness of the proposed utilization are high-
lighted. In wider context, the study aims to contribute to enrichment of the knowledge basis
regarding the use of aluminium in earthquake engineering, identifying its feasibility and po-
tential in such applications.
The proposed hybrid EBF is designed according to EC3 and EC8 with the respective capaci-
ty-overstrength provisions and thus the central buckling of the diagonal braces is prevented
throughout the lateral loading process. Three aluminium alloys are chosen based on their duc-
tility features for the subsequent analyses: the early yielding and ductile 5154A, the highly
ductile 6061 T4 and the 6063 T6. The EBF system is modeled and meshed in ANSYS (Fig.1).
The connectivity between primary structural parts is accomplished by contact elements. Cy-
clic loading is imposed and the hysteretic behavior of the aluminium link is observed stable
1
Corresponding author: Tel.: +302310995656; E-mail: vefth@civil.auth.gr
135
with well-shaped loops until failure (Fig.2). Analysis showed that proposed EBF model with
aluminium link attracts less base shear both in monotonic and cyclic loading. As indicated by
the stable and repeatable hysteresis loops produced from cyclic loading, the suggested ap-
proach exhibits reliable behavior in the inelastic range. The link under examination using al-
uminium alloys, which post elastic behavior is characterized by a semantic strain hardening
feature, as well as by its bounded ductility, dissipated large amounts of energy effectively and
reliably even at large strains. Links such as the one studied have also the advantage of being
relatively easily replaced when severely damaged by a major earthquake and can also be de-
ployed in existing frame designs as a retrofit device.
Fig. 1: Model of the proposed hybrid bracing system
Fig. 2: Base shear vs top horizontal displacement (cycling loading)

References
[1] Mazzolani FM. Innovative Steel Structures for Seismic Protection of Buildings, Prin
2001, Polimetrica, Milan, 2006.
[2] Mistakidis ES, De Matteis G, Formisano A, Mazzolani F M. Low yield metal shear
panels as a alternative for the seismic upgrading of concrete structures, Advances in En-
gineering Software, 38, pp. 626–636 , 2007.
[3] Rai DC, Wallace, BJ. “Aluminium Shear-links for enhanced seismic resistance”, Earth-
quake Engineering & Structural Dynamics, 27(4), 315-342, 1998.
[4] Mazzolani FM. Aluminium Alloy Structures, 2nd Edition, E&FN SPON, Chapman &
Hall, London, 1995.
136
Tampere, Finland
DESIGN OF WIND TURBINE STRUCTURES BASED ON

A MULTIVARIATE STOCHASTIC APPROACH
F. H. Kemper a,* and M. Feldmann a

a
Institute of Steel Construction, RWTH Aachen University
*
Tel.: +49 241 80 25218; E-mail: kemper@stb.rwth-aachen.de
Abstract
The structural design of blades and the machinery of wind turbines are nowadays based on
detailed transient dynamic calculations using multi-body systems (MBS) - which are without
doubt indispensable for the construction of the machinery. But with respect to the design of
the supporting structure, i.e. the tower and its foundation, the MBS models are unnecessary
complex and the obtained results can therefore often only be taken in a simplified manner.
With this load input, the design of the structure is carried out afterwards by means of individ-
ual calculations. As the dimensioning of the structural components influences the overall dy-
namic behaviour of the wind turbine with respect to the aeroelastic load components and the
inherent dynamic amplifications induced by the rotor, the design has to be iterative. With re-
gard to the number of load cases and the time consumption of MBS simulations the structure
is often not fully optimised, due to necessary simplifications.
In the field of wind engineering it is common practice to consider the complex stochastic wind
load process on flexible structures by means of stochastic methods. An example of this meth-
odology is the gust response factor approach given in Eurocode 1. Besides this simplified ap-
proach which is formulated for a single degree of freedom system, it is also possible to con-
sider arbitrary multi-degree of freedom systems as long as the structural dynamic behaviour
can be treated linear. With application to wind turbines, this approach enables the necessary
description of the correlated multivariate wind field and the formulation of the structural dy-
namic behaviour of the response using complex spectral matrices. The main advantage is the
algorithm is the allowance for a quick analysis of the load-response chain and to consider the
structural interactions between the main components implicitly.
With the presented strategy, a design tool is introduced which especially focusses on the sup-
porting structures of wind turbines taking into account the global dynamic behaviour and ne-
glecting machinery details. Due to its faster usability, it might turn out as a useful tool for an
optimized design of wind towers and foundations.
137
138
Tampere, Finland
TIME HISTORY SIMULATION IN SEISMIC DESIGN
Peter Knoedela,* and Thomas Ummenhoferb
KIT Steel & Lightweight Structures, Karlsruhe Institute of Technology, Germany

*
Author for contact. Tel.: +49 721 608 46976; E-mail: peter.knoedel@kit.edu
Abstract
In seismic design of steel structures different levels of modelling of the structural system are
possible. In the simplest case horizontal substitute quasi-static-forces are determined accord-
ing to EC8. This procedure requires the smallest effort by the designer. Maximum workload
is required when setting up a dynamic FE model which includes geometric and material non-
linearities and which is driven by recorded or artificial time-displacement-histories at the
foundation.
Fig. 1: Dynamic Amplification

a) Theory – comparison of absolute and relative displacements
b) Results of present numerical study
In order to reduce the amount of workload and computer time a constant-amplitude harmonic
drive can be used with a duration of 10 seconds, which would be in agreement with EC8
3.2.3.1.2 (3). Typically the frequency of the drive is chosen to be coincident with the lowest
natural frequency of the structure, as far as industrial buildings with not too many storeys are
concerned.
139
In a numerical study we investigated an plane single storey frame under elastic and plastic
conditions. The amplifications found are plotted over the normalised driving frequency in Fig.
1.
The main conclusions are:

1. Plastic steel structures do not exhibit a pronounced resonance peak as known with elas-
tic steel structures.
2. Still, missing the resonance frequency by 1 % might give results which are by 20 % un-
safe.
3. A Duffing-type jump phenomenon could not be observed.
4. No advice on detuning can be given since we found variations to both sides of the elas-
tic natural frequency. Thus we recommend to have several runs close to the resonance
in order to check the sensitivity of the system and to find the peak of the amplification
function.
5. “Ordinary” engineering decisions on how much accuracy is needed with respect to the
length of the period might be wrong for this type of problem. It is evident that limiting
the error of the period to 1 % is not a sensible measure if the response of the structure
varies by 20 % around ±1 % of the resonance.
6. It is underlined again by this findings that FE analyses need a very thorough documen-
tation which includes the verification procedure. Otherwise the results should not be
considered trustworthy.
140
Tampere, Finland
STEEL COMPOSITE DOWELS IN CRACKED CONCRETE
M. Classena, A. Starka
a
Institute of Structural Concrete, RWTH Aachen
1 Introduction
Steel composite dowels are efficient, innovative shear connectors consisting of interlocking
steel and concrete dowels. These dowels can be used to transfer shear forces in filigree steel
composite beams (Fig. 1). If the concrete slab of a composite beam is exposed to tensile stresses
in the longitudinal direction, transversal cracking occurs. For example, cracking can arise in the
region of the interior supports of continuous beams or in the concrete tensile chord of integrated
steel composite floor slabs [1]. EN 1994-1-1 as well as the technical approval for composite
dowels [2] neglect the impact of transversal cracking on the shear capacity of connectors, as in
the few known tests with longitudinal tension [3], no significant decrease in shear capacity has
been observed. However, the small database of usable experiments does not allow for a defini-
tive assessment of the cracking impact. Hence, the present paper deals with comprehensive
experimental investigations on this issue.
Fig. 1: Filigree steel composite beam and composite dowels
2 Shear tests
The shear tests were split into two classes: tests under longitudinal compression in the concrete
slab (SD), and tests under tensile stress (SZ). While the test set-up for compression shear tests
conforms to EN 1994-1-1, specifications for shear tests with tensile stress are lacking in the
codes. The chosen test set-up for series SZ is shown in Fig. 2. Here, the shear force was applied
by hydraulic cylinder, which was placed between the specimen and an anchor plate fixing lon-
gitudinal reinforcement bars. By activating the cylinder, shear forces occur in the composite
141
connection while tensile forces are introduced into the rebars and into the slab yielding to trans-
versal concrete cracking. In total, the paper comprises 12 shear tests on steel composite dowels
in transversely cracked and uncracked concrete slabs with different arrangements of reinforce-
ment.
Fig. 2: Shear tests with tensile stress and transversal cracking in the concrete slab (dimensions in cm)
2 Results
Where composite dowels fail due to concrete pry-out, transversal concrete cracking leads to a
significant reduction of shear capacity between 22% and 33% depending on the chosen rein-
forcements. Fig. 3 shows the dowel characteristics and the concrete crack patterns of specimens
with compressive vs. tensile stress. Obviously, the transversal cracks induce a detachment of
the concrete pry-out cone and a limitation of the pry-out cone’s length lout,tension, which approx-
imately complies with the average spacing between adjoining transversal cracks srm. The con-
siderable interference between the composite dowels’ shear behavior and transversal cracking
can be decisive for a number of case studies (e.g. interior supports of continuous composite
beams) and should be considered in design. Furthermore, it was found, that stirrups have crucial
impact on the concrete confinement in the vicinity of the composite dowel and prevent a split-
ting failure of the concrete slab.
Fig. 3: Comparison of dowel characteristics and crack patterns for different longitudinal stress states
References
[1] Hegger, J. et al.: Multifunctional composite slab system with integrated building ser-
vices. STAHLBAU Vol. 83, Iss. 7, pp. 452–460, July 2014.
[2] Allgemeine bauaufsichtliche Zulassung Verbunddübelleisten, Z-26.4-56, 13. Deutsches
Institut für Bautechnik, 2013.
[3] Wurzer, O.: Zur Tragfähigkeit von Betondübel, Dissertation am Institut für Konstrukti-
ven Ingenieurbau, Universität der Bundeswehr München, 1998.
142
Tampere, Finland
CROSS-SECTIONAL CAPACITY OF COMPOSITE COLUMN BY THE

TWO METHODS OF EN 1994-1-1
Kimmo Ylinen, Wei Lu and Jari Puttonen

Department of Civil and Structural Engineering, School of Engineering, Aalto University
Abstract: The options of EN 1994-1-1 [1] for designing composite columns are a simplified
method and general method. The simplified method has been calibrated by test results and its
scope is limited by the range of results available, whereas the general method is more a set of
principles than a design method and its scope is unlimited. However, the scope of simplified
method covers a substantial share of practical applications. Therefore, the simplified method
is most commonly used. It may be assumed that the simplified method leads to conservative
design whereas the general method optimizes the use of materials. In this paper, both methods
are compared in calculating the cross-sectional capacities of concrete-encased steel I-sections.
Comparisons are carried out by a computer code programmed by a non-commercial GNU Oc-
tave language [2].
The material of concrete is modelled with a parabola-rectangle type of stress-strain relation-

ship, and the steel is with an elastic-plastic stress-strain relationship without strain-hardening.
Eccentricities of loading about both major-axis and minor-axis of cross-section are consid-
ered. In comparisons, material grades of both concrete and steel, dimensions of the steel I-
profiles, and the number of reinforcement bars were varied within the scope of the simplified
method but second-order effects were neglected. The total number of 27 separate cases are
studied.
The cross-sectional interaction diagrams for compressive axial force and bending moment
about both major (My) and minor (Mz) axis for the cases with extreme differences between
general simplified methods and general methods are presented in Fig. 1 and Fig. 2, respec-
tively. The solid line is the interaction curve calculated with the general method; dark dotted
line and light dotted line represent the simplified method with aM = 0,8 and aM = 1,0, respec-
tively. The results reveal that in respect of cross-sectional capacities the simplified method
gave up to 30 % larger values than the general method. These overestimations are observed
especially for the combinations of low concrete strength and high steel grades with heavy
steel profiles bending about minor axis. On the other hand, the studies also reveal that the
simplified method can underestimate more than 20 % the cross-sectional capacities in the case
of the combinations of high strength of concrete and high steel grades with lighter steel pro-
files for both axes bending. The results indicate that future improvements of simplified meth-
od are necessary for efficient, effective and safe design solutions.
143
Fig. 1: Comparisons of the capacity curves in modified Case 15 to give extreme ∆max.
Fig. 2: Comparisons of the capacity curves in modified Case 6 to give extreme ∆min.
Notation
b, bs Width of the concrete section and steel profile
fad, fyd Design value of the yield strength of the structural steel and reinforcement bars
fcd Characteristic value of the compressive strength of the concrete
h, hs Height of the concrete section and steel profile
tf,, tw Thickness of the flange and web of the steel profile
M, N Bending moment and Axial force
aM Reduction factor for MRd in simplified method
d Steel contribution ratio according to EN 1994-1-1
D Difference between two design methods
r Reinforcement ratio
References
[1] European Committee for Standardization. EN1994-1-1, Eurocode 4: Design of compo-
site steel and concrete structures. Part 1-1: General rules and rules for buildings, 2006.
[2] GNU Octave, http://www.octave.org (e), 2014.
144
Tampere, Finland
BEAM-TO-COLUMN JOINTS SUBJECTED TO IMPACT LOADING
Erik L. Grimsmo, Arild H. Clausen, Arne Aalberg, and Magnus Langseth
Structural Impact Laboratory (SIMLab), Centre for Research-based Innovation and Department of Structural
Engineering, Norwegian University of Science and Technology, NO-7491 Trondheim, Norway.
*
Corresponding author. Email address: erik.l.grimsmo@ntnu.no. Phone number: +47 73595263.
Abstract: A double-sided beam-to-column joint configuration has been tested in impact load
conditions. The test specimens consisted of H-section beams and columns that were joined by
end-plate connections and high-strength bolts. The results show that the general behaviour
was the same as in comparable quasi-static tests. However, for the dynamic tests, the inertia
of the test specimen imposed deformation modes that caused significant shearing action,
which is an effect that could lead to shear failure of the joint. Thus, it can be unsafe to assume
that joints in impact load conditions behave as in static load conditions.
1 Introduction
Several events, such as accidental explosions and impact, can impose severe impulsive load-
ing to beam-to-column joints. There is a limited amount of experimental data available in the
open literature on the topic of steel joints subjected to extreme dynamic, non-cyclical loading.
An experimental program therefore commenced at the Norwegian University of Science and
Technology, where a double-sided beam-to-column joint configuration was tested in quasi-
static and impact load conditions. The test set-ups and some results have been presented and
discussed in detail by Grimsmo et al [1]. The companion paper to this extended abstract ex-
plains briefly the test set-up for the dynamic tests, and presents results obtained with one type
of test specimen from the experimental programme. The focus is mainly on the shearing ac-
tion that was observed in the dynamic tests.
2 Experimental programme
As shown in Fig. 1, the test specimens consisted of H-sections, two beams and one column,
which were joined by bolted end-plate connections. The 12 mm thick end-plates were welded
to the beams with a continuous fillet weld, and both the end-plate and profile material was
grade S355 steel. Six M16 bolts with grade 8.8 were used in each connection. Fig. 1 also indi-
cates the loading the test specimens were subjected to; a force was applied to the column,
while the tip of the beams was fixed in the direction of the force. This induced tension in the
two uppermost bolt rows of the test specimen in Fig. 1. In the dynamic tests, the force was
applied by an impact through the use of a pendulum accelerator. The column of the test spec-
imens were impacted by a trolley with a mass of 726.7 kg and an initial velocity that ranged
between 8 and 12 m/s in the various tests.
145
a) Elevation view b) Section (A-A) view

Fig. 1: Dimensions, loading, and boundary conditions of the test specimen.
3 Results
Fig. 2a displays a picture frame obtained from a high-speed camera at the instant the column
of the test specimen had displaced about 70 mm (approx. elapsed time: 10 ms). In this photo,
the column is oriented horizontally, and the beams vertically. The bending deformation of the
end-plate is clear, and one of the bolts has fractured since the head of the bolt is not in contact
with the end-plate. The bolts fractured in tension for both the quasi-static and dynamic tests,
see Fig. 2b. However, the bolt originating from the dynamic test has experienced shear de-
formation, which was caused by that the column flange slid relatively to the end-plate in the
beginning of the test. This occurred due to inertia effects, and was thus only observed for the
dynamic tests.
a) Deformed joint at 70 mm displacement of b) Fractured bolts coming from a quasi-static

column, obtained from high-speed camera test (left) and a dynamic test (right)
Fig. 2: Deformed joint and bolts.
4 Conclusions
The joints behaved in a preferable manner in the sense that the joints failed in the same way in
the quasi-static and dynamic tests. However, due to the shear deformation of the bolts in the
dynamic tests, it can be argued that shear failure of the joints is possible by for instance in-
creasing the mass of the beams. In design practice, one should be aware of that extreme im-
pulsive loading imposes other deformation modes to the joints, which might lead to other
failure modes compared to what is found for static load conditions.
References
[1] Grimsmo E.L., Clausen A.H., Langseth M., and Aalberg A., An Experimental Study of
Static and Dynamic Behaviour of Bolted End-plate Joints of Steel. Submitted for
publication, 2015.
146
Tampere, Finland
Design resistance of end plate splices with hollow sections
Yvonne Steigea and Klaus Weynanda

a
Feldmann + Weynand GmbH, Aachen, Germany
Abstract: The paper presents a design approach to calculate rectangular hollow section
(RHS) splices (bolted end-plate connections) under tension forces or bending moments in ac-
cordance with EN 1993-1-8. Based on models that exist in literature an Eurocode conform
model is presented by using the component method. The original model, based on experi-
mental and numerical investigations, uses a three dimensional yield line model to predict the
tension resistance of bolted splices with hollow sections considering the joint as a whole. The
adapted model is fully compatible with EN 1993-1-8. Moreover, it also allows to predict the
design moment resistance of such RHS splices.
1 Background
EN 1993 Part 1.8 contains application rules for the evaluation of the resistance of end plate
connections with open profiles (Chapter 6) by means of the component method. Furthermore,
the standard provides rules to calculate the design resistance of welded hollow section joints
in lattice girders (Chapter 7). However, there are no explicit application rules or design for-
mulas for bolted end plate joints with hollow sections.
Bolted end plate joints are used for example as chord splices in lattice girders under normal
loading conditions. Typical bolt patterns in RHS splices are bolts on two opposite sides or
bolts placed on four sides of the hollow section. The hollow section is connected with the end
plate by a one sided fillet weld around the perimeter of the section, which should not exceed
the resistance of the connected end plate or section.
For joints with bolts on all four sides of the connected hollow section, no information is avail-
able on how to determine the effective length of the effective T-stub for the “corner bolts”.
Therefore, a three dimensional yield line model, which can be found in literature, is used to
determine the effective length for the corner bolts. Supplementing the effective length in
EN 1993-1-8 with this effective length it is possible to calculate the design tension and mo-
ment resistance for RHS splices.
2 Resistance model
The design resistance of end plate connections with open profiles under bending moment can
be calculated according to EN 1993-1-8 based on the component method approach. This
method could be applied to end plate connections with RHS as well. The only difference is
the calculation of the resistance of the component end plate in bending with the T-stub model.
147
The relevant resistance results from the minimum of the resistances of the three failure modes.
Mode 1 and 2 are calculated with the plastic moment of the T-stub flange, which depends on
the effective lengths.
The effective lengths of two-sided splices can be calculated in accordance with EN 1993-1-8.
Additionally, for four-sided connections, an effective length part for the corner bolts is devel-
oped. This is based on a three dimensional yield line model derived from literature, which in-
dicates a formula for the total design resistance of the joint. To calculate the effective length,
this resistance is set equal to the resistance of a half T-stub and added to the effective length
for outer bolt rows, which are also used for two-sided connections.
2.1 Tension Resistance

The design tension resistance results from the component end plate in bending FT,Rd and beam
web in tension FBWT,Rd.
For the determination of FT,Rd, the resistance of the individual connection side has to be calcu-
lated. Therefore a modified T-stub model is introduced. This model does not take into account
the area between the two webs of the hollow section. As described in EN 1993-1-8, one bolt
row includes two bolts. Then the resistances of the individual connection sides are added.
2.2 Moment Resistance

The determination of design moment resistance of the joint Mj,Rd is explained only for bend-
ing around the y-direction. The calculation for z-direction can be derived accordingly.
Mj,Rd results from the effective tension resistance Ftr,Rd of the individual bolt rows r multiplied
with the distance of the bolt row to the center of compression hr.
It is assumed that the center of compression lies in the beam flange of the hollow section.
For the calculation it is necessary to distinguish between the two different bolt patterns (four-
and two-sided). The design moment resistance of two-sided can be calculated according to
EN 1993-1-8.
The first row from four-sided connections, called external row, has to be considered separate-
ly. In EN 1993-1-8, one bolt row consists of two bolts, but in the case of RHS splice the ex-
ternal row can also have just one or even more than two bolts. Therefore, the resistance of the
external row is calculated with a rotated T-stub as presented in EN 1993-1-8, but taking into
account the fact that the number of bolts, here n, is not set to a certain value.
The effective resistance of the individual bolt rows is the minimum of the components end
plate in bending and beam web in tension. The subsequent procedure is calculated according
to EN 1993-1-8.
3 Conclusion
The presented model provides a calculation method for joints with RHS joints with bolted end
plates according to EN 1993-1-8. Comparison with experimental results from literature shows
that the model predicts safe sided resistances. The resistance of the component end plate in
bending is calculated with the T-stub model with two bolts. It would be useful to apply a half
T-stub for bolted joints with RHS, so that the individual sides of the connection can be repre-
sented. This approach would also correspond to the component method and could thus be eas-
ier applied to other connections types.
148
Tampere, Finland
CONCEPTION, ANALYSIS AND DESIGN OF A SPECIAL JOINT FOR

FIXING LATTICE TOWERS LEGS DURING TESTING OF TRANS-
MISSION LINE TOWERS
Fábio Paiva, Jorge Henriques and Rui C. Barros*
Faculty of Engineering of the University of Porto, Civil Engineering Dept, Structural Division, Portugal
*
Author for contact. Tel.: +351 22 508 1882; Mobile: +351 91 438 5405; E-mail: rcb@fe.up.pt
Abstract: Tower testing is extremely important in the transmission line industry and may be
performed for many reasons. The paper describes a special joint that allows the connection of
lattice tower legs to a universal base in Portugal (Metalogalva Group, Trofa). The main
considerations used during the modelling, analysis and design of these connections are
described in detail. For that purposes a numerical model (with solid finite elements) was
created in Solidworks, using simplified connectors to simulate the behaviour of bolts between
the components; contact interaction (without penetration) was also considered in the analysis.
Overall, the special joint designed satisfies all requisites demanded for the universal base and
consequently for the testing station to be constructed in Portugal.
1 Introduction
In a traditional proof test, the test is set up to verify the design conditions, only statics loads
are applied, the support has level fixed foundations, and the restraints at the load points are
the same as in the design model. This kind of test will verify the adequacy of the members
and their connections to withstand design loads specified for that structure as an individual
entity under controlled conditions. Proof tests provide information on support behaviour
under load, fit-up verification, actions on the structure in deflected positions, adequacy of
connections, and other benefits [1].Since the transmission tower industry has unique design
codes, the validity of major design assumptions and the correctness of the overall design can
be verified. Another advantage of testing is the complete assembly of the tower that provides
an excellent check of the fabrication details. A successful test also provides a level of
confidence with the computer model used to design the tested tower and its combination of
bodies and legs [2].
2 Conception/Description of the universal base

The design of the special joint that allows the connection of lattice tower legs to the universal
base in Portugal (Metalogalva Group, Trofa) is umbilical linked to the universal base to be
developed.
The conception of the universal base (Fig. 1) attempted to comprise three main group of
structures to test in the “tower station”: towers, H-frames and poles. In simplified way the
149
universal base (dimension in plant 20*20 m2) can be divided in two parts, an internal strip
destined to poles and frames structures and the whole base to tower structures (with 3,4 or 5
legs) recurring to other special elements if need to advance with the test. For the testing of
poles a joint in ring format (up to 6 m diameter) allow different pole base sizes to be tested.
In this paper only the joint related with the connection of the tower leg to the universal base is
addressed.
Fig. 1: Description of the Universal base of the testing station

3 Conclusions
An overview of the universal base main components is described, regarding the connections
of the testing structures. Detail depiction of the special joint that allows the connection of
lattice tower leg to the universal base is given.
The solid finite element model was sufficient accurate for the purposes the analyses
undertaken and provide sufficient confidence in the design of the joint (plates and bolts). The
contact simulation assumed between the different joint components, tried to guarantee
primarily a safe design and then to approximate the joint behaviour to real service conditions.
Consequently the special joint described in this paper satisfies all the initial pre-conditions
established at the beginning of the design process.
Acknowledgments
This work was co-participated by funds from the project “VHSSPOLES-Very High Strength
Steel Poles” (Faculty of Engineering of the University of Porto, reference 21518).
References
[1] Paiva F, Henriques J, Barros R.C. Review of Transmission Tower Testing Stations
around the World. Modern Building Materials, Structures and Techniques, 2013.
[2] Nuño J, Miller M, Kempner L. Historical Perspective of Full-Scale Latticed Steel
Transmission Tower Testing. Electrical Transmission and substation structures, 2012
150
Tampere, Finland
GENERALIZED BLOCK FAILURE
Jeppe Jönsson
Technical University of Denmark, DTU Civil Engineering
Extended abstract
Theoretical and experimental work leading towards generalized capacity methods including
the combined influence of normal force, shear force and moment on the block tearing capacity
of gusset and fin plate connections will be presented and exemplified in this paper for the cas-
es shown in Fig. 1.
Block Tension Block Shear Large Eccentricity
Where does it act?

Fig. 1: Gusset and fin plate connections and relevant blocks for block tearing.
The presented work focuses on the development of a few simple block failure capacity formu-
las and a set of relevant interaction formulas with a format related to those already in use for
cross section analysis in the Eurocodes. The practical formulations are most efficiently based
on a simple generalization of the current method using very simple stress distributions along
the yield lines surrounding the block in combination with a simplified yield condition to be
fulfilled along these lines. A formal yield stress of fm=(fy+fu)/2 corresponding to the mean
value of the yield stress and the ultimate stress is introduced, since the straining of the yield
lines varies along each line and strain hardening will commence before the yield mechanism
has fully formed. Some adequate observations will be discussed in connection with a short
review of relevant literature.
Three relatively simple block tearing situations for a C-cut-out are shown in Fig. 2. In the pa-
per an L-cut-out is also treated. It is assumed that the holes are just inside the block and the
outer gross dimensions are given by hg and bg. The thickness of the plate is denoted by t. The
related net lengths hn and bn are found by deducting the diameter of all the holes along each
length. Furthermore it is assumed that the normal stresses are acting on a reduced thickness
corresponding to the net to gross area ratio, i.e. th  t hn hg and tb  t bn bg along the respec-
151
tive lines. This assumption leads to a relatively simple method for determination of the ca-
pacities also including the effect of the holes with respect to the normal stress distribution.
Fig. 2: Normal and shear stress distributions for block tearing forces and moment in a C-cut-out.
The three basic block tearing capacities found for a C-cut-out can for example be found as:
 b   h  b h 
N R  tf m  2 g  hn  , VR  tf m  2bn  g  , M R  thg f m  g  n  (1)
 3   3  3 4
By scaling the three basic stress distributions in Fig. 2 by N N R , V VR and M M R respec-
tively and checking the formal yield condition along all yield lines shows us that the follow-
ing interaction formula needs to be fulfilled:
2 2
 N M  V 
     1 (2)
 N R M R   VR 
Furthermore the paper discusses a rigid plastic upper bound method – approximated with re-
spect to bolt holes - and an experimental investigation for a large eccentricity connection.
F
Contra weight
F Test plate
Fig. 3: Experimental test setup and photograph of block failure found in experiment.
The experimental test setup is illustrated and the related block failure experiments on the bolt-
ed connections are reported. The test setup is illustrated to the left in Fig. 3 and to the right a
photograph of block tearing found in the experiment is shown.
To conclude block tearing is generalized to include connection force interaction by a few

simplifying assumptions. A theoretical plasticity based upper bound method is used to verify
the magnitude of the capacities and the interaction formula. Furthermore a small experimental
investigation is reported for a C-cut-out.
152
Tampere, Finland
FEM SIMULATION OF A TUBULAR KT-JOINT
Jolanta Bączkiewicza, Karol Bzdawkaa

a
Poznan University of Technology
*
Author for contact. Tel.: +48 692 856 486; E-mail: karol.bzdawka@put.poznan.pl
Abstract: Taking into account real joint stiffness in tubular trusses allows for reduction of
member buckling length. Presented research investigates rotational stiffness of a single KT-
joint of a SHS truss using FEM. The joint is modelled as it is manufactured – gap joint with
vertical attached to the tensile diagonal. In-plane rotational stiffness has been determined. Gap
sizes between the diagonals and between the chord and the vertical pale were found to influ-
ence the stiffness of individual connections. As a result KT-joint mustn’t be treated as sepa-
rate K and Y-joints even if the gaps are maximum allowed by the Eurocode.
1 Introduction
Eurocode 3 [1] and former Dutch code NEN6770 give contradictory requirements regarding
determining the buckling length for lattice truss members. Inspired by that Boel [2] developed
a method for calculating the buckling length factor K with the use of two parameters β and γ
that describe the geometry of the joint. Boel proposed a beam model where the diagonals are
attached to the outer face of the hollow-section profile. The diagonals are connected with the
beam representing the chord via a short rigid link. At the chord-diagonal interface a spring is
assumed that has the same in-plane and out-of-plane rotational stiffness as the full 3D joint.
From this model the buckling lengths for chord and diagonal could be determined. Haakana
[3] took his method further and confronted it with experimental results [4] that allowed for
developing better material model. Haakana [3] used it to widen the applicability of Boel’s
method to joints of unidentical sections and angles, and non-minimal gaps. She found that
changing the size of the bigger diagonal when the smaller was loaded yielded great difference
in the joints stiffness, while changing the size of the smaller diagonal when the bigger was
loaded had minor effect on its stiffness. Research presented in this paper takes this approach
further to determine the in-plane rotational stiffness of a KT-joint.
2 Considered joint
Presented study considers a KT-joint shown in Fig. 1 in order to determine the influence of
the pale on the diagonal-chord joint’s in-plane stiffness (springs C1,C2 and C3 – Fig. 1) and if
it is influenced by the size of gaps between the diagonals g1,2 and between the pale and chord
g2,3. The sections were 150x150x5 for the chord, 120x120x4 for the diagonals and 80x80x3
153
for the pale, with 4 and 3 mm throat thickness welds for diagonal-chord and pale-diagonal
connections, respectively. The gap sizes were minimum and maximum allowed by the code
[], giving a total of four combinations.
LP
80x80x3
LBL LBR
beam elements
120x120x4
120x120x4
45° g2,3 45°
g1,2 C3
C1 rigid links C2
150x150x5
Figure 1: View of the considered joint (left) and its beam model (right)
5 Results
Joint stiffnesses obtained from the analyses for each individual load and each spring connec-
tion are presented in Table 1. It was found that the pale influences the stiffnesses of both di-
agonals, but especially the one to which it is attached. The pale was found to have only a mi-
nor effect on the stiffness of the diagonal to which it is not directly attached. Stiffness of the
diagonal with pale was found to be 2-3 times greater than that of the diagonal without pale –
depending on the gap sizes.
Main conclusion is that KT-joint with eccentrically located pale cannot be considered as sepa-
rate K and Y joints even if the gap sizes are maximal. More tests are needed with varying
cross sections and angles to determine the buckling length factors in this type of truss.
Table 1: Individual connection stiffnesses obtained from Abaqus analysis. Given in [kNm/rad].
g 1,2 - g 2,3 min-min max-min min-max max-max
LBL, C1 1416 966 1225 805
LBR, C2 2986 2781 2618 2438
LP, C2 3159 2860 2945 2709
LP, C3 573 560 443 442
Acknowledgement
SSAB is acknowledged for funding the research.
References
[1] EN 1993-1-8, Eurocode 3: Design of steel structures - Part 1-8: Design of joints, European
Committee on Standardization, Brussels, 2005.
[2] Boel HD., Buckling Length Factors of Hollow Section Members in Lattice Girders, MSc thesis,
Dept. of Architecture Building and Planning, Eindhoven University of Technology, The Nether-
lands, 2010.
[3] Haakana Ä., In-Plane Buckling and semi-rigid joints of tubular high strength steel trusses, MSc
thesis, Faculty of Business and Built Environment, Tampere University of Technology, Finland,
2014.
[4] Tuominen, N., Björk, T. Ultimate Capacity of Welded Joints Made of High Strength Steel
CFRHS. EUROSTEEL 2014, Naples, Italy, September 10-12, 2014, Laboratory of Steel Struc-
tures, Lappeenranta University of Technology.
154
Tampere, Finland
BEARING CAPACITY OF COLD-FORMED UNLIPPED CHANNELS WITH

RESTRAINED FLANGES - EOF AND IOF LOAD CASES
Balasubramaniam Janarthanan1, Shanmuganathan Gunalan2 and Mahen Mahendran3

1, 2, 3
Queensland University of Technology, Brisbane, QLD 4000, Australia
Abstract: Cold-formed steel sections have been developed as more economical building
solutions to the alternative heavier hot-rolled steel sections in the commercial and residential
industries. Cold-formed steel unlipped channel sections are commonly used as bearers and
joists in floor systems. However they suffer from bearing failures when subjected to
concentrated loads or reactions. The bearing capacity and failure modes of cold-formed steel
sections mainly depend on the loading types, locations and connection types. Four load cases
are listed in the currently available standards (AISI S100, AS/NZS 4600 and Eurocode 3 Part
1.3), namely, end-one-flange (EOF), interior-one-flange (IOF), end-two-flange (ETF) and
interior-two-flange (ITF), based on failure locations and loading types. A unified bearing
capacity equation with different bearing coefficients is given in the current AISI S100 and
AS/NZS 4600 specifications to predict the bearing capacity of cold-formed steel channel
sections. However, relevant bearing coefficients are not available for fastened cold-formed
steel unlipped channel sections (‘fastened support - unstiffened flange’ condition). Eurocode
3 Part 1.3 does not distinguish between fastened and unfastened support conditions. Hence an
experimental study consisting of 28 tests was conducted in this research to assess the bearing
behaviour and capacities of cold-formed steel unlipped channels with their flange fastened to
supports under one flange loading (EOF and IOF load cases).
Cold-formed unlipped channel sections with their depth in the range of 100 mm to 230 mm,
flange width in the range of 51 mm to 76 mm and their thickness in the range of 1.5 mm to 6
mm were chosen in the experimental study. The nominal yield strength of all the sections was
450 MPa, however, tensile coupon tests were performed to determine the accurate
mechanical properties. The specimen length was taken as three times the section depth plus
the bearing plate lengths based on the recently updated AISI standard method. Test
specimens were constructed using two unlipped channel sections placed facing each other in
a box-beam arrangement as mentioned in the AISI standard test method. The top and bottom
flanges of the sections were interconnected using angles at quarter points of the length of
specimens to make the test arrangement laterally and torsionally stable. There different
bearing plates (50 mm, 100 mm and 150 mm) were used in this experiment study. These
bearing plates were fastened to the flanges of the channel sections using Grade 8.8 M12 bolts
with a washer with 28 mm nominal outer diameter and 2.5 mm thickness as specified in the
Australian standard AS 1252 to simulate fastened support conditions. The webs of the
155
sections were stiffened at midspan for the EOF load case and at the ends for the IOF load
case to ensure that the failure would occur at intended locations. The load was applied using
displacement control method. The ultimate failure load, the load versus vertical displacement
at the loading point of the section, and the load versus lateral deflection of webs at failure
locations were recorded from the experimental study.
The bearing capacities of tested unlipped channels sections were predicted using the currently
available design rules (AISI S100, AS/NZS 4600 and Eurocode 3 Part 1.3) and were
compared with the experimental capacities. In this study, AISI S100 and AS/NZS 4600
design rules were used with two sets of available bearing coefficients (‘unfastened support -
unstiffened flange’ and ‘fastened support - stiffened flange’ conditions). The comparison
showed that ‘unfastened support-unstiffened flange’ coefficients generally underestimated the
bearing capacity for EOF load case while inconsistently for IOF load case. The predictions
using ‘fastened support-stiffened flange’ coefficients were found to be unconservative for
both EOF and IOF load cases. Eurocode 3 Part 1.3 design rules underestimated the bearing
capacities by 100 percent on average for EOF load case although they predicted the bearing
capacities well for IOF load case. Overall, the predictions by the current design rules were
found to be inconsistent. Hence improved equations were proposed in the format of current
AISI S100 design rule to determine the bearing capacities of cold-formed steel unlipped
channels based on the ultimate bearing capacity results from this study.
In addition to this, a new design rule was also proposed based on the direct strength method
(DSM) to predict the bearing capacities of fastened unlipped channel sections under EOF and
IOF load cases. The DSM method uses the elastic buckling load (Pcr) and yield load (Py) of
the section. The elastic buckling loads (Pcr) of the tested sections with fastened support
conditions were obtained from finite element analyses. This paper presents the details of this
investigation on the bearing capacities of cold-formed unlipped channel sections with
restrained flanges, and the results including the improved design rules.
Keywords: Cold-formed steel structures, Unlipped channel sections, Bearing capacity, EOF
and IOF load cases, Direct strength method.
Corresponding Author: Mahen Mahendran, Professor of Structural Engineering, Science and

Engineering Faculty, Queensland University of Technology, Brisbane,
m.mahendran@qut.edu.au, Ph: +61731382543, Fax: +617 3138 1170.
156
Tampere, Finland
ELASTIC BUCKLING OF AN I-BEAM WITH SANDWICH FLANGES
Krzysztof Magnuckia, Piotr Paczosb

a,b
Institute of Applied Mechanics, Poznan University of Technology, Poznan, Poland
Abstract: The subject of the paper is a thin-walled I-beam with sandwich flanges. The beam
is composed of cold-formed parts joint by fusion welding. The channel beams make the main
part of the beam – the web and inner faces of the flanges. The paper is devoted to a theoretical
study of elastic buckling of the I-beam under pure bending with the use of the Finite Strip
Method (FSM) and Finite Element Method (FEM). Results of the study are presented in Ta-
bles and in Figures.
1 Specification of the I-beam with sandwich flanges

The subject of the paper is a thin-walled I-beam with sandwich flanges. The beam is com-
posed of cold-formed parts, namely two channel beams, two trapezoidally corrugated cores of
the flanges, and two flat sheets (Fig.1).
Fig. 1: Scheme of the cross-section of the I-beam with sandwich flanges
Numerical investigation is realized for the following sizes of the cross-section: depth of the
beam D  220 mm , thickness of the flanges t f  10 mm , thickness of the sheets t  1mm ,
width of the beam b  140 mm , and the total area of the cross-section A  1445.6 mm2 .
2 Numerical study - Finite Strip Method and Finite Element Method

The simply supported I-beam with sandwich flanges is under pure bending. Numerical calcu-
lation is realized with the use of the system CUFSM. The local and global states are deter-
mined for selected length of the beam. Local buckling of the beam: the value of the load factor
of the critical load f floc  15.0805 for the length of the beam Lb  300 mm , therefore the
157
half-wavelength Lw  300 mm . Global buckling of the beam – lateral buckling: the value of
the load factor of the critical load f f glob  2.4771 for the length of the beam Lb  4000 mm ,
therefore the half-wavelength Lw  4000 mm (Fig.2).
Fig. 2: The local and global buckled shape (CUFSM v3.12)
Numerical calculations have been also realized with the use of FEM method. A numerical
buckling analysis was done using ANSYS software. Shapes of local and global buckling of
the I-beam with sandwich flanges – FEM analysis is shown in Figure 3.
Fig. 3: The local and global buckled shape (FEM Ansys Workbench 12)
Basing on the FEM analysis the load factors of the critical load, and the mode of buckling in
the elastic range has been determined. The comparison of the results obtained from the FSM
and FEM investigations in linear cases is presented in Table 2.
Table 2: Values of the load factors of the critical loads

I-beam ff(loc) ff(glob)
FSM 15.0805 2.4771
FEM 15.132 2.644
Acknowledgments
The research was conducted within the framework of Statutory Activities in 2015.
158
Tampere, Finland
A Numerical Parametric Study on the Load Carrying Behaviour under

Bending of Honeycomb Girders made of Trapezoidal Corrugated Steel
Sheets
Tobias Petersena,*, Manuel Krahwinkela

a
HafenCity University Hamburg, Innovative Bauweisen und Baukonstruktion, Germany
*
Author for contact. Tel.: +49 40 42827 5344; E-mail: tobias.petersen@hcu-hamburg.de
Extended Abstract
In the past, trapezoidal corrugated steel sheets (TCSS) have been established as an efficient
way of cladding the facades and roofs of industrial steel buildings. The trapezoidal geometry
of this cold-formed profile generates the stiffness, which is necessary to reach a span of up to
10 m. The height of these profiles has the greatest influence on the stiffness and therefore on
the reachable span as well. The production process has reached a maximum with a height of
up to 200 mm. Hence, it needs to find other solutions to increase the span of these profiles.
A honeycomb girder (HCG) is defined by a mirror symmetric composition of two TCSSs.

Mechanical fasteners such as screws or rivets connect them. This connection creates a honey-
comb-like profile, which permits double height and thereby explicitly improves the stiffness
compared to a single TCSS. The load carrying behaviour of this construction under bending is
affected by the geometric nonlinear behaviour of the single sheeting and the nonlinear behav-
iour of the connection under shear.
a) TCSS profile Hoesch T-160.1 b) HCG made of 2 TCSSs Hoesch T-160.1

Fig. 1: Example of a HCG arrangement
A verified numerical model, which was introduced in former publications [6,7], acts as the
basis for a fundamental parametric study to investigate the construction type HCG. All in all
360 different HCG constellations were investigated due to their load carrying behaviour under
bending by varying in 2 different profile types, 4 thicknesses, 5 mechanical fasteners and 9
159
different spans between 7 to 15 m. Figure 2 exemplifies the results for HCGs made of two
different mechanical fasteners, a non cutting self-drilling screw and blind rivet, in combina-
tion with a T-150.1 TCSS. The diagrams show the maximum characteristic surface load in
dependence on span of a simple beam system. The value of qk,max includes all necessary safety
factors, is reduced by the dead load of HCGs and makes it therefore comparable with the
characteristic value of snow or wind loads. Additionally to the results for the HCG systems
equivalent TCSS systems were calculated and evaluated in the same manner. The comparison
shows the potential of HCG systems with span of 10 m and higher.
T-150-1 / SMS01Z 4.8x20 T-150-1 / PolyGrip 6.4x15
5,0 5,0
qk,max [kN/m2]
qk,max [kN/m2]
4,0 4,0
3,0 3,0
2,0 2,0
1,0 1,0
0 0
5 7 9 11 13 15 17 5 7 9 11 13 15 17
Span [m] Span [m]
tN=0.88 mm tN=1.00 mm tN=1.25 mm tN=1.50 mm
tN=0.88 mm tN=1.00 mm tN=1.25 mm tN=1.50 mm

Fig. 2: Example of a HCG arrangement
Conclusions
1. For 360 HCG constellations the maximum characteristic surface qk,max load was deter-
mined by using numerical simulation software.
2. The mechanical fasteners Hilti SMS01Z 4.8x20, Hilti SMD01Z 5.5x19 and Gesipa Pol-
yGrip 6.4x15 generated the highest characteristic surface loads. Further investigations
should focus on these fastener types.
3. In the future studies the equidistant arrangement of the fasteners should be replaced by
an arrangement, which is more affine to the shear forces in the gap between upper and
lower TCSS. Therefore the early failure of the connection might be retarded and the
maximum characteristic surface load increased.
4. Investigations on continuous beam systems of HCGs promise higher values for their
maximum characteristic surface load. Therefore it is imperative to develop adjoining
techniques for HCG systems in support areas.
References
[6] Petersen T., Krahwinkel M. “The Honeycomb Girder – A Comparison between Labora-
tory Tests and Simulation”, 7th European Conference on Steel and Composite Structures
(Eds.: R. Landolfo, F. M. Mazzolani), Naples, Italy, 469-470, 2014.
[7] Petersen T., Krahwinkel M. “Der Wabenträger”, 19. DASt-Forschungskolloquium (Eds.:
Deutscher Ausschuss für Stahlbau DASt) Hannover, Germany, 34-39, 2014.
160
Tampere, Finland
Elastic Buckling Characteristics of Corrugated Tank

under Fundamental Load
Yoshifumi YOKOYAMAa, Kikuo IKARASHIb

a,b
Abstract: Corrugated tanks used as containers comprise corrugated steel plates, and thus are
expected to have higher buckling strength against shear and higher out-plane strength than
normal cylinder. However, corrugated tanks have not been sufficiently studied to confirm this,
and their physical characteristics have not clarified. The purposes of this study are to clarify
the physical characteristics of corrugated tanks and to establish a design method considering
those characteristics. In this study, the elastic buckling characteristics of corrugated tanks un-
der fundamental loading conditions are clarified by FEM analysis.
1 Introduction
Corrugated tanks that comprise corrugated steel plates are commonly used as containers.
They have geometric characteristics that provide higher buckling strength against shear and
out-plane strength than flat steel plates of the same scale. However, physical characteristics of
corrugated tanks have never been sufficiently clarified, and thus an efficient design method
based on those characteristics has not been established. Therefore, this study intends to clarify
the elastic buckling characteristics of corrugated tanks under fundamental load.
2 Outline of FEM analysis

type-N
直径 d type-V type-H
The definitions of analytical models and geo- d
metric parameters are shown in Fig. 1. The mod- 板厚 t
els are categorized into three groups: type-V,
type-H and type-N. The elastic buckling charac- 高さ h
teristics of these models are examined under axi- h
al compression and bending shear.
shellシェル要素
element
3 Elastic buckling characteristics of rigid 剛体要素
body a波高 a 円筒半径
b
r
corrugated tanks r t 板厚 t
2 (y ) 波高 a
5 (θy) a
3.1 Axial compression The elastic buckling 6 (θz)

波長 r  波角度
O 円筒半径
characteristics of types V, H and N are exam- 3 (z) 1 (x) t 板厚 t r
4 (θx) b
円筒中心軸
ined under axial compression acting on the models. 円筒中心
When models are subjected to axial compression, Fig. 1: Geometric definition of models
161
200  [N/mm ]
2
N H(=300) Axial
V(=300) H(=200) compression
type-V is shaped into very fine waves and the buckling stress of V(=200) H(=100)
V(=100)
type-V is smaller than that of type-N in the entire analytical 150
range of this study. As seen in Fig. 2, the buckling stress of type-

V reduces with increasing height of corrugation a and length of 100
corrugation . The buckling stress of type-H is also shown in

Fig. 2. The buckling stress of type-H has the minimum value 50
with changing a. In cases where a model has large a and  the
elastic buckling stress of type-H is larger than that of type-N. 0
d=8595mm h=10500mm t=4.5mm
0 20 40 60 a[mm]
Fig. 2: Influence of height of
3.2 Bending shear corrugation a
The elastic buckling characteristics of types V, H and N are
examined under transverse end load or antisymmetric moment.
When types H and N are subjected to bending shear, the mod-
els are shaped into only shear mode in the entire analytical range
of this study. In contrast, type-V is shaped into either shear,
bending, or shear and bending buckling mode indicated in Fig. 3.
(a) Shear mode (b) Bending mode
The buckling modes of type-V are shown in Fig.5, where A re- Fig. 3: Elastic buckling mode
fers to shear mode and B refers to bending mode. As seen in
 [N/mm ] 2
Fig.4, the elastic buckling stress of type-H under either loading 30 Transverse end load / Antisymmetric moment
condition is almost the same. In models that have large a and N N
H(=300)
the elastic buckling stress of type-H is larger than that of H(=300)
H(=200) H(=200)
type-N. The influence of a on the elastic buckling stress of type- 20 H(=100) H(=100)
V in models subjected to transverse end load and antisymmetric

moment is indicated in Figs. 5(a) and (b). In either loading con-
10
dition in the range of small a, type-V is shaped into shear mode
and the elastic buckling stress increases with increasing a. How-
ever, in the range of large a type-V is shaped into bending mode 0
d=8595mm h=10500mm t=4.5mm
and the elastic buckling stress decreases with increasing a. 0 20 40 60 a [mm]
Therefore, the elastic buckling Fig. 4: Buckling stress of type-H

stress is at its maximum value at 60  [N/mm ] 60  [N/mm ]
2 2
Transverse end load Antisymmetric

the point that the buckling mode N moment
AB B
B
B
changes. In models under either V(=300)
A
V(=200) B
loading condition, the elastic buck- 40 V(=100) 40
B
B
B
ling stress of type-V is higher than A
A
B B B B A
that of type-N in the entire analyti- A
A
B A
A
B B B B
B B
cal range. In models subjected to 20 A B
B B
20 A
A
A B
antisymmetric moment, the elastic B
B B B B B
A
buckling stress of type-V is much

d=8595mm h=10500mm t=4.5mm d=8595mm h=10500mm t=4.5mm
higher than that of type-N. 0
0 20 40 60 a [mm]
0
0 20 40 60 a [mm]
(a)Transverse end load (b)Antisymmetric moment
4 Conclusion Fig. 5: Buckling stress of type-V
1) The buckling stress of type-V in models subjected to axial compression is smaller than that
of type-N. In contrast, type-V is advantageous under bending shear.
2) In either loading condition considered in this study the buckling stress of type-H has the
minimum value with increasing a. With increasing buckling stress increases.
This study shows above results. However, to practicalize them, it would be necessary to study
about post-buckling behavior the effect of initial imperfections.
162
Tampere, Finland
BUCKLING STRENGTH OF LIGHT-GAUGE MEMBERS WITH

LARGE OPENINGS
Atsushi SATOa, Seiji MORIb, Tetsuro ONOc , and Kazunori FUKIHASHId

a, b
Nagoya Institute of Technology
c
Sugiyama Jogakuen University
d
NS Hi-Parts Corporation
*
Author for contact. Tel.: +81 (0)52 735 5200; E-mail: sato.atsushi@nitech.ac.jp
Abstract
Steel Framed House (SFH) is one of the structural system that utilizes light-gauge where the
thickness of the steel sheet is less than 2.3mm. To use the ceiling cavity more efficiently and
maximize the living space of SFH, light-gauge beams with openings at web are often used.
This opening can be used for electrical wiring and/or pluming. Current design code provides
the formula which can evaluate the opening effect to the shear strength of the beam [1].
However, the formula is only a function of the opening ratio which was based on simple
opening tests conducted in 1980’s. Light-gauge is easily formed by press; therefore, it is easy
to form an additional lip at the edge of the opening for reinforcement purpose. The lip formed
at the edge of the opening (hereinafter called burring) is expected to delay the shear buckling
of the web and increase the strength. The advantage of the burring (shape effect) must be in-
cluded in the future design formula.
The purpose of this study is to clarify the effect of web opening which have lip at the opening
edge. Full-scale testing and numerical simulation were conducted to propose a new design
formula that can evaluate opening and burring effects.
To evaluate the effect of opening at the beam web of light gauge, three series of specimens
are prepared (Fig. 1). One is full web (Fig. 1.a), second is simple opening (Fig. 1.b), and the
last is opening with burring (Fig. 1.c). Fig. 1.d shows the shape of the burring.
a) Full web b) Simple opening c) Burring opening d) Burring

Fig. 1: Shape of openings at beam web
163
Fig.2 shows test and FEM results. In all figures, current strength reduction factor Eq. (2) is
also shown. As shown in the figures, opening will reduce the shear strength of the beam;
larger opening will have larger strength reduction. It is clearly shown that burring, which was
formed for reinforcement purpose, will increase the shear strength of the beam. Moreover,
test results shows that current design formula is too conservative.
SO(Analysis) SO(Analysis) SO(Analysis)

1.0 1.0 1.0
BO(Analysis) BO(Analysis) BO(Analysis)
(Test) (Test) (Test)
0.8 0.8 0.8
anaQcr / anaQcr0
anaQcr / anaQcr0
anaQcr / anaQcr0
0.6 0.6 0.6
0.4 0.4 0.4

Eq.(2) Eq.(2) Eq.(2)
0.2 0.2 0.2
0.0 0.0 0.0

0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
a/h a/h a/h
a) t = 0.8mm b) t = 1.2mm c) t = 1.6mm

Fig. 2: Strenth reduction due to opening (Test and FEM)
Strength reduction factor for simple opening can be evaluated by following formula.
s qs  ana Qs ana Q0  1.097  1.126( a / h)  1.0 (1)
Effect of burring can be evaluated by following formula.
ana Qb ana 
Q s  7827(t / h) 2  0.389 ( a / h) 2  1.097  (2)
Consequently, strength reduction factor of burring opening can be expressed as follow:
s qb  {(7827(t / h) 2  0.389 )( a / h) 2  1.097 }  1.097  1.126( a / h )  1.0 (3)
Fig. 3 shows the validity of the proposed formula. The calculated errors are less than 10%.
Fig. 3: Comparison between calculated results with test and analysis results
Conclusions
To clarify the effect of web opening to the shear strength of the beam. Full-scale testing and
Numerical simulation was conducted in this study. Following results were found.
1. Cold-formed lip at the edge of opening (burring) provided reinforcement; shear strength
due to opening was improved and buckling strength increased;
2. New design formula which can evaluate the opening in a high accuracy was proposed.
References
[1] The Japanese Iron and Steel Federation, Manual for Light Gauge Building Design,
JISF, 2006.
164
Tampere, Finland
EXPERIMENTAL AND NUMERICAL INVESTIGATIONS OF THE

STEEL STORAGE RACK UPRIGHTS
Chong Rena, Xianzhong Zhaoa,* and Ru Qina

a
Department of Building Engineering, Tongji University, Shanghai, 200092, China.
*
Author for contact. Tel.: +86-21 65982926; E-mail: x.zhao@tongji.edu.cn
Extended abstract
Cold-formed steel perforated uprights are usually used as main structural members in steel
storage racks. The perforated uprights have arrays of holes along the length, which are al-
lowed the beam to be connected at variable heights and the bracing to be bolted to form the
frames. The effects of perforations on the strength of members have been performed by many
researchers, but a definitive analytical solution for pallet rack members has not yet been estab-
lished. Over the decades, several researchers have investigated the behaviour of perforated
member by using experimental method. The type of holes in most researches is different with
perforations on the upright of racks, the influence on the performance between large web
holes and the perforations systematically located at web and flanges is significant different.
Additionally, some papers only focused on the short length uprights. However, in nowadays
intermediate and long length uprights are widely used in structures of steel storage rack sys-
tems and are also necessary to study. The perforated sections and non-standard restraint condi-
tions make the numerical analysis too complicated to be used in the design of storage rack
structures. Therefore, current design codes of steel storage rack are based on test procedures.
For the stability investigations, since the 1970s there has been substantial research activity in
the field of cold-formed structures which led to a numerous published work on the local, dis-
tortional and lateral-torsional buckling of the cold-formed steel sections, and research on in-
teractions involving distortional buckling has been carried out very recently. The significant
influence of distortional buckling has been proved, and the interaction of distortional/global
buckling on the perforated members is considered to be sufficiently important to warrant fur-
ther investigation.
In this paper an experimental investigation of the behaviour of steel storage rack uprights sub-
jected to axial compression is presented. Perforated and non-perforated steel rack uprights, in
various cross-sections and with different lengths (67 specimens in total) were carried out us-
ing axial compression test, which is to determine the load bearing capacity of rack uprights.
The bidirectional hinged joint bases were fitted at the upper crosshead and lower actuator to
simulate simply supported boundary conditions. The comparisons of ultimate loads between
perforated and non-perforated uprights are shown and failure modes are demonstrated in detail.
The comprehensive data of the ultimate loads recorded from the compression test are provided
in the test report of Tongji University. Only part of test data is presented in this paper, which
demonstrates the comparisons of load carrying capacities and failure modes between perforat-
165
ed and non-perforated members. According to the buckling failure modes, specimens are clas-
sified into three categories of buckling failure: distortional buckling (DB) specimens, flexural-
torsional buckling (FTB) or flexural buckling specimens (FB) and interaction of distortion-
al/flexural (or flexural-torsional) buckling (DB+FB or FTB). It can be found in this paper that
compare with non-perforated members, the ultimate loads of perforated members are reduced
and a few failure modes are altered, which indicate that perforations have significant influence
on the performance of steel uprights and also have impact on the buckling failure modes.
In this paper, finite element analysis is employed and validated by the results obtained from
the experiment. The modified Riks method built in ABAQUS is used. The four node shell el-
ement of reduced integration scheme built in ABAQUS is employed to carry out the static
nonlinearity analyses. The element used is a thin, shear flexible, isometric quadrilateral shell
element. The load-displacement response curve for each member analysed is computed, and
the limit load is determined from the peak point of the load-displacement curve. The devel-
oped finite element model is verified against the experimental results, and the figures reveal
the good agreement of load-displacement curves between the numerical and experimental re-
sults. Moreover, the comparisons of experimental and numerical failure modes of specimens
at ultimate load are also in very good agreement and are presented in this paper.
The Direct Strength Method (DSM) is also conducted to compare with test and numerical re-
sults in this paper. The slenderness of the upright is represented by λ = (Py/Pcr)0.5, where Pcr is
the critical load of buckling which is calculated using the first eigenvalue of buckling from
FEA. The identical cross-section dimensions, thickness and length of members are compared,
the only difference is whether the member has perforations. It can be found the predictions of
distortional buckling failures and interaction of distortional/flexural (or flexural-torsional)
buckling failures of non-perforated members obtained from FEA show good agreement with
test results and the distortional buckling curve of DSM. However, the conservative predictions
of distortional buckling curve of DSM is demonstrated when the perforated members have
distortional buckling or interaction of distortional/flexural (or flexural-torsional) buckling.
This paper also reveals the predictions of flexural-torsional buckling are conservative for both
perforated and non-perforated members when compare with flexural-torsional buckling curve
of DSM. The reason for this is explained by that the interaction of global buckling and materi-
al yield may have significant influence on the performance of perforated and non-perforated
members. Due to influence of perforations, some distortional buckling failures of non-
perforated members are turned into interaction of distortional/flexural (or flexural-torsional)
buckling failures of perforated members. These evidences above imply the perforations have
significant influence not only on the performance but also on the buckling failure mode of
steel storage rack uprights.
166
Tampere, Finland
EXPERIMENTAL INVESTIGATION ON THE BEHAVIOR OF

PERFORATED STEEL STORAGE RACK COLUMNS UNDER AXIAL
COMPRESSION
Bassel EL KADIa,, Guven KIYMAZb* and Atakan MANGIRc

a,c
Fatih University, Department of Civil Engineering
b
Antalya International University, Department of Civil Engineering
*
Author for contact. Tel.:+90 532 608 1581; E-mail: guven.kiymaz@antalya.edu.tr
Extended Abstract
The present study is focused on the behavior and design of perforated steel storage rack
columns under axial compression. These columns may exhibit different types of behavior and
levels of strength owing to their peculiar features including their complex cross-section forms
and perforations along the member. In the present codes of practice, the design of these
columns is carried out using analytical formulas within which experimentally determined pa-
rameters are used.
In the present study, an experimental program was carried out to verify the accuracy of a re-
cently proposed design approach that has the potential to eliminate the need for design by test-
ing. The proposed approach includes modifications in the Direct Strength Method (DSM) to
include the effects of perforations (the so-called reduced thickness approach). The elastic
buckling parameters of the studied members needed for strength calculations were obtained
by using the CUFSM and CUTWP programs.
The experimental study included axial compression tests on members of different lengths. The
cross-section geometry and dimensions were kept constant. The abovementioned design ap-
proach was used to estimate the load carrying capacity of the tested columns. A comparison
between the experimental and the design approach results is presented. It was found out that
experimental results compare very well with the design approach estimations.
The main conclusions of the experimental study that was applied on columns of the same
cross-section but of different lengths, can be summarized as follows
• One of the best ways to determine the effective center of gravity of a steel storage rack
column is carrying out experimental tests, by changing the location of the load appli-
cation point along the symmetry plane of the cross-section and investigating the posi-
tion that gives the maximum failure load, i.e. the position of the effective center of
gravity.
167
• As the column length increases, the stiffness and the ultimate failure load decrease (
see Fig.1)
• The displacement levels corresponding to the ultimate failure load are close. (Fig.1)
• Local buckling mode, although might have occurred, was not visually detected even in
the stub column tests.
• It is observed that the distortional buckling was the dominating buckling mode for all
of the tests as shown in Fig. 2. For columns of length 1250 mm and 1100 mm, small
effects of flexural buckling was noticed.
The main conclusions of the analytical study can be summarized as follows

• The conventional method for calculating the column nominal strength, using the Di-
rect Strength Method without taking into consideration the effect of perforations along
the length of the member, gave imprecise results, (the error discrepancies between 7 to
21%) compared to the experimental test results as shown in Figs. 3-4.
• The different alternatives of the recently proposed approach show a good accuracy
compared to the experimental results. Alternative 1 and Alternative 2 gave more accu-
rate results (the error discrepancies between 0.5 to 4%) than Alternative 3 and Alterna-
tive 4 (the error discrepancies between 0.5 to 11.5%) as shown in Figs. 3-4.
Generally, results obtained from the experimental and analytical studies compare very well.
This indicates the validity of the recently proposed approach for predicting the ultimate
strength of steel storage rack columns with perforations along their length.
250 C500
225
200 C650
175
C800
150
125
Load (kN)
C950
100
C1100
75
50 C1250
25
0
0 1 2 3 4 5
Displacement (mm)
Fig. 1 Fig. 2
250 Alt 3 and
250 Alt 1 and
240 Alt 2 240 Alt 4
230 Put ave 230 Put ave
220 220
210 Test1 210 Test1
Pu (kN)
Pu (kN)
200 200
190 Test2 190 Test2
180 180
170 Test3 Test3
170
160 160
150 Test4 Test4
150
350 500 650 800 950 1100 1250 1400 350 500 650 800 950 1100 1250 1400
Pn non Pn non
Length (mm) per Length (mm) per
Fig. 3 Fig. 4
168
Tampere, Finland
MONOTONIC BEHAVIOUR OF BOLTED T-STUBS: A REFINED

THEORETICAL MODEL FOR FLANGE YIELDING AND BOLT
FRACTURE FAILURE MODE
Antonella B. Francavillaa*, Massimo Latoura, Vincenzo Pilusoa and Gianvittorio Rizzanoa

a
University of Salerno
*
Author for contact. Tel.: 328 7482186; E-mail: afrancavilla@unisa.it
1 Introduction
The prediction of the behaviour of beam-to-column connections can be obtained by means of
the so-called “component method”, largely used in research studies and currently codified in
Eurocode 3. To date, the EC3 still provides some drawbacks especially dealing with the pre-
diction of the ductility supply and the prediction of the cyclic behaviour. In fact, even though
some authors have already investigated some aspects related to prediction of the ductility sup-
ply [1,2] and cyclic behaviour of connections [3,4] past experimental and theoretical research-
es have often focused their attention mainly on predicting stiffness and resistance of joints. In
such connection typologies, usually, the most important components, such as the column
flange or the end plate in bending, are modelled by means of equivalent T-stubs, i.e. two equal
T-shaped elements connected through the flanges by means of one or more bolt rows. There-
fore, in order to propose a theoretical approach for predicting the whole force-displacement
response up to failure of bolted T-stubs a new refined model is presented, taking properly into
account the existing literature.
2 General description of the proposed model

In this paper, the mechanical model proposed aims to define the T-stub behaviour up to failure
accounting for the following effect: the contact forces are considered applied in a point in be-
tween the tip of the plate and the edge of the bolt head, the bolt forces are considered uniform-
ly distributed under the bolt head, the failure of the T-stubs is modelled by checking the ulti-
mate strain on the basic materials composing the plate and the bolt, the compatibility condi-
tion between the displacements of the plate and the uplift of the bolt is taken into account and
the displacements of the T-stub are evaluated step-by-step as the sum of the elastic and plastic
part.
2.1 Force-Displacement Curve of the T-stub

As far as the kinematic mechanism is defined and the mathematical laws to be used in order to
evaluate the rotations of the plastic hinges are given, it is possible to calculate the ultimate
displacement of the bolted T-stub. For a fixed value of the bending moment Mj acting in cor-
respondence of the T-stub web, the known parameters are five: the force of the T-stub (F), the
169
prying force (Q), the value of the distributed load corresponding to the action provided by the
bolt head (q), the ratio between the bending moment acting at the bolt line and that arising at
T-stub web (ψ) and the location of the prying forces in the contact zone (n*). The values of
these parameters can be obtained by considering five equations: three equilibrium equations
(one translational equilibrium, two rotational equilibrium), and two compatibility equations.
3 Comparison with experimental results

The theoretical model has been validated by means of a comparison with experimental tests
carried out at Material and Structure Laboratory of the Department of Civil Engineering of
Salerno University [5]. The comparison with experimental evidence shows a satisfactory
agreement in term of ductility and resistance between the theoretical model and the experi-
mental results (Fig.1).
Fig. 1: Comparison between experimental results and theoretical predictions
The comparisons of the model with the experimental tests shows a very satisfactory agreement
in terms of shape of the force-displacement curve and in terms of prediction of the ductility
supply.
4 Conclusions
In this paper a theoretical model for predicting the whole force-displacement curve of the T-
stubs has been presented. The comparison with experimental tests carried out by the same au-
thors at laboratory on materials and structures of Salerno University has shown a good accura-
cy of the model. The obtained results are really encouraging about the possibility of predicting
accurately the ductility supply of T-stub by means of a theoretical approach.
References
[1] Girão Coelho, A.M., da Silva, L.S. & Bijlaard, F.S.K., 2004. Characterization of the
nonlinear behavior of single bolted T-stub connections, Proc., 5th Int. Workshop on
Connection in Steel Struct., Amsterdam, 53-64.
[2] Beg, D., Zupancic, E. & Vayas, I., 2004. On the Rotation Capacity of Moment Connec-
tions. Journal of Constructional Steel Research, Volume 60, pp. 601-620.
[3] Iannone, F., Latour, M., Piluso, V. & Rizzano, G., 2011. Experimental Analysis of Bolt-
ed Steel Beam-to-Column Connections: Component Identification. Journal of Earth-
quake Engineering, 15(2), pp. 214-244.
[4] Latour, M., Piluso, V. & Rizzano , G., 2011b. Cyclic Modeling of Bolted Beam-to-
Column Connections: Component Approach. Journal of Earthquake Engineering, 15(4),
pp. 537-563.
[5] Piluso V., Faella C. & Rizzano G., 2001. Ultimate behaviour of bolted T-Stubs. II:
Model validation, Journal of Structural Engineering ASCE, 127 (6):694-704.
170
Tampere, Finland
DIFFERENT COATING SYSTEMS FOR THE APPLICATION IN SLIP-

RESISTANT CONNECTIONS
Natalie Stranghönera, Nariman Afzalib, Jörn Bergc, Markus Schiborrd,

Andrea Rudolfe, Susanne Bergerf
a,b,c,d
University of Duisburg-Essen, Institute for Metal and Lightweight Structures, Essen, Germany
e,f
Institute for Corrosion Protection Dresden GmbH, Dresden, Germany
Abstract: Exposure to the environment increases the vulnerability of bolted steel connec-
tions. The common way to protect steel surfaces of slip-resistant connections against corro-
sion is to cover the faying surfaces with a protective layer. In the frame of this paper, the in-
fluence of different surface treatments (alkali-zinc silicate coating (ASI-Zn-coating) and
ethyl-zinc silicate coating (ESI-Zn-coating)) on the load bearing capacity of slip-resistant
connections are presented considering different thicknesses of the coating materials and dif-
ferent types of testing procedures (EN 1090-2 and TL/TP-KOR-Stahlbauten).
1 Introduction
Slip-resistant connections are required, when deformations in bolted connections must be lim-
ited to pre-defined values either for serviceability or ultimate limit reasons. A common coat-
ing system for slip-resistant connections is an alkali-zinc silicate (ASI-Zn)-coating. In the
frame of the presented research activities different ethyl-zinc silicate (ESI-Zn)-coatings have
been investigated as an alternative to an ASI-Zn-coating. The experimental testing has mainly
been carried out on the basis of the EN 1090-2-testing procedure. Some comparative tests
have also been performed on the basis of TL/TP-KOR-Stahlbauten.
In total nine different slip test series have been examined considering static slip factor tests
and creep tests, see Table 1; results of extended creep tests are not presented:
 two test series for an ASI-Zn-coating (50 m nominal dry film thickness (NDFT) accord-
ing to ISO 19840), one relying on the TL/TP-KOR-Stahlbauten testing procedure and one
relying on the Annex G, EN 1090-2 testing procedure and
 seven test series for five different customary and one especially produced ESI-Zn-coatings
using the Annex G, EN 1090-2 testing procedure. One of the customary ESI-Zn-coatings
has been tested twice considering two coating thicknesses (50 m and 80 m NDFT acc.
to ISO 19840). All other tests were conducted with coating thicknesses of 50 m NDFT.
One ESI-Zn-coating has been especially for the aim of achieving high slip factors.
171
Table 1: Test specimens and slip factors considering static tests only and static + creep tests
Series NDFT Number of µnom,mean µinit,mean µactual,mean V (µactual) 1)
ID ISO 19840 test results
[µm] nstat/nstat+cree (stat/stat+creep) (stat/stat+creep) (stat/stat+creep) (stat/stat+creep)
[-] [-] [-] [%]
p
[-]
ASI-Zn-Coating
ASI-E2) 8/10 0.70/0.68 0.69/0.68 0.80/0.79 2.3/3.5
50
ASI-T3) 10/- 0.81/- 0.81/- 0.89/- 3.6/-
ESI-Zn-Coating
ESI-I2) 8/10 0.58/0.59 0.57/0.58 0.63/0.65 2.1/4.8
50
ESI-II2) 8/10 0.43/0.44 0.43/0.44 0.45/0.47 3.2/7.2
ESI-IIa2) 80 8/10 0.49/0.49 0.49/0.49 0.53/0.54 4.0/5.5
ESI-III2) 8/10 0.52/0.52 0.52/0.52 0.57/0.57 2.5/4.4
ESI-IV2) 8/10 0.67/0.67 0.67/0.67 0.77/0.77 3.0/4.5
50
ESI-V2) 8/10 0.55/0.56 0.56/0.56 0.61/0.63 2.9/5.8
ESI-VI2) 8/10 0.54/0.54 0.54/0.54 0.59/0.59 3.6/6.5
1)
Coefficient of variation for µactual  2) Level of Preload: 110 kN  3)
Level of Preload:100 kN
Comparing the different ESI-Zn-coatings (50 m NDFT acc. to ISO 19840) it becomes obvi-
ous that the results show a relatively large scatter with a slip factor range of 0.47  actual,mean
 0.77. Compared to the ESI-Zn-coatings, the ASI-Zn-coating shows the highest slip load and
slip factor with actual,mean = 0.79. Nevertheless, ESI-IV reaches nearly the same value as ASI-
E. Considering the scattering of the measured ESI-Zn-coating thicknesses it becomes not
clear, whether the scatter of the slip factor test results are caused by the different coating reci-
pes or by the different real coating thicknesses. This is still examined. Comparing the results
for ESI-II and ESI-IIa (same coating recipe, different coating thicknesses), it can be seen that
the tendency seems to be confirmed that with higher coating thicknesses, higher slip factors
might be achieved. It has to be mentioned that due to the fact that all creep tests clearly failed,
extended creep tests are necessary in order to be able to formulate reliable slip factor recom-
mendations for the investigated coating systems. Furthermore, slip factors for design purposes
have to consider preload losses due to the creep of the coating, transversal contraction, fatigue
etc. Further investigations are carried out.
3 Conclusions
The mean slip factors achieved on static tests are relatively high with values greater than 0.5
in all cases without one ESI-Zn-coating. Due to the fact that in all cases the creep tests failed,
extended creep tests will be performed in future investigations, which surely lead to a de-
crease of the slip factor. Nevertheless, it became obvious that a great scatter of slip factors is
achieved for the investigated customary and special ESI-Zn-coatings. Up to now, it did not
become clear, whether the scatter of the slip factor test results are caused by the different
coating recipes or by the different coating thicknesses. As the coating thickness might have a
significant influence on the slip factor, this effect is still examined. To take benefit of higher
slip factors, it is highly recommended for the coating suppliers or for special industrial pro-
jects, to perform individual slip factor tests in order to determine the actual design slip factor.
Performing slip factor tests, special care has to be spent on the testing procedure (measuring
the preload in the bolts, positioning of the displacement transducers for the determination of
the slip, influences of the tightening procedure etc.).
172
Tampere, Finland
INFLUENCE OF DIFFERENT TESTING CRITERIA ON THE SLIP

FACTOR OF SLIP-RESISTANT CONNECTIONS
Natalie Stranghönera, Nariman Afzalib, Jörn Bergc, Markus Schiborrd, Frans Bijlaarde,
Nol Gresnigtf, Peter de Vriesg, Ralf Glienkeh, Andreas Eberti
a,b,c,d
University of Duisburg-Essen, Institute for Metal and Lightweight Structures, Essen, Germany
e,f,g
Delft University of Technology, Stevin II Laboratory, Netherlands
h,i
Fraunhofer-Anwendungszentrum Großstrukturen in der Produktionstechnik, AGP, Rostock, Germany
Abstract: Slip of slip-resistant connections has to be prevented either for serviceability or

ultimate limit state reasons. EN 1090-2 specifies slip factors for often used surface conditions.
For deviating conditions, slip factors have to be determined experimentally according to An-
nex G of EN 1090-2. The practice shows, that the slip test procedure according to Annex G is
not clear in detail. For instance, the slip load has been defined as the load corresponding to a
slip of 0.15 mm. Furthermore, static tests have to be performed using a “normal speed” which
is not clearly specified. For these and other reasons, a comprehensive investigation is going
on to be carried out to resolve these problems. First results are presented in this contribution.
1 Introduction
The slip resistance of bolted slip-resistant connections is influenced by different parameters
such as the condition of the faying surfaces, the preload level of the bolts, the geometry of the
structural details etc. Slip factors for some specified surface conditions are given in EN 1090-
2 or can be found in literature. For those surface conditions which have not been considered
in EN 1090-2 or if higher slip factors are required, slip factor tests should be performed ac-
cording to Annex G of EN 1090-2. However, the practice has shown, that the current slip test
procedure according to Annex G is not clear in detail.
In the frame of the RFCS-research project SIROCO, experimental investigations have been
carried out regarding the questions of (1) methods which ensure a measurement of the preload
in the bolts with sufficient accuracy, (2) the influence of the position of the displacement
transducers on the slip factor, (3) the influence of the tensile loading velocity on the slip re-
sistance behaviour and (4) the slip-criterion.
Slip factor tests were carried with test specimens according to the M20-bolt-geometry of EN
1090-2. In total, three surface conditions were considered (1) grit-blasted Sa 2 ½ (GB), (2)
173
grit-blasted Sa 2 ½ + alcali-zinc silicate (ASI-Zn)-coating (ASI) and (3) grit-blasted Sa 2 ½ +

zinc-spray-metalized-coating (Zn-SM).
The following results have been achieved so far:
(1) The deviations between the measurement by instrumented bolts with implanted strain
gauges (SG) and especially prepared load cells (LC) are negligible small with a maxi-
mum deviation of 1.3 %. Furthermore, the mean values of the losses of preload during
testing were detected to approximately 9 % for GB-I and 7 % for ASI-I. As in static slip
factor tests the main part of the loss of preload is caused by transversal contraction, the
preload losses correspond to the level of the slip load.
(2) Using LCs leads to a relatively large clamping length of the bolts which influences the
loss of preload and consequently the level of the slip load. Evaluating the slip factor con-
sidering the nominal preload in the bolts without taking into account the large clamping
length, might lead to an overestimation of the slip factor because the preload losses de-
crease and the slip load increases with increasing clamping length.
(3) The positioning of the displacement transducers (LVDTs) to measure the slip is of great
importance. They have to be positioned in the centre of the upper resp. lower part of the
specimen otherwise the elongation of the plates is implicitly measured as well which
might lead to larger slip deformations and herewith to lower slip loads at the slip criterion
of 0.15 mm.
(4) Within the variation of the investigated time frames the slip factor was tested for, the in-
fluence of the test duration on the test results is small (less than +/- 5%). For GB a ten-
dency of a slightly higher slip factor for longer test durations can be observed, for Zn-SM
the opposite can be stated and for ASI-Zn no effect of the test duration is found. For the
majority of the tests a deformation controlled loading method was used. Additionally, for
Zn-SM two load controlled tests were executed as well. These tests yield to a significant-
ly higher slip factor than was found in the deformation controlled tests.
(5) Annex G of EN 1090-2 gives a fixed value of 0.15 mm displacement at which the slip
load has to be determined. However, this displacement does not always describe “the
point of slip”. In order to define a slip deformation criterion to be used in slip tests, the
question which has to be answered first is: how much slip can be allowed in a “stiff”
connection? What is the deformation limit? As a reliability analysis with regard to the de-
formation limit under static and fatigue loading has not been carried out up to now, this
will be performed in the frame of SIROCO. At this stage it can be concluded that in prin-
ciple a variable determination of the slip criterion should be applied according to ECCS-
TC 10 or RCSC. For special industrial projects, more flexible slip criteria must be appli-
cable to allow an economic design of special steel structures.
3 Conclusions
To realize cost effective slip-resistant connections, the slip factor has to be determined by ex-
periments very carefully. In the frame of the present contribution it could be verified that the
testing procedure of EN 1090-2, Annex G has to be specified more precisely e. g. regarding
the load velocity, the positioning of the displacement transducers, the devices for the meas-
urement of the preload of the bolts and the slip criterion.
174
Tampere, Finland
SIMPLIFIED MODEL FOR CONNECTIONS OF STEEL STRUCTURES

IN OPENSEES
R. Costaa, F. Gentilib and L. Simões da Silvac

a
Department of Civil Engineering, University of Coimbra, Portugal
b
filippo.gentili@uc.pt, +351 239797254, ISISE, Department of Civil Engineering, University of Coimbra,
Portugal,
c
ISISE, Department of Civil Engineering, University of Coimbra, Portugal
Abstract: According to a modern approach, based on the so-called component method, Euro-
code allows characterizing the moment-rotation curve of semi-rigid connections. This paper
deals with the formulation of a simplified mechanical model composed of extensional springs
and rigid links, for the characterization of a cruciform configuration where the left and right
connection are modelled by two separate moment-rotation curves and the web panel by one
additional moment-rotation curve. Two macro-elements are described, covering nodes con-
necting beams with the same and with different beam depths. The developed FEM elements
were implemented in OpenSees and were validated with some benchmark examples.
1 Introduction
Nowadays, developments in structural analysis and the increased capacity of personal com-
puters allow for robust and rigorous analyses including the semi-rigid behaviour of connec-
tions without increased burdens on the user. The continuous search for economical solutions
require the accurate modelling of the of beam-to-column joints in structural analysis and the
component method is recognized as an effective procedure to account for it. In this paper, the
formulation of two FEM macro-elements developed based on two mechanical models suitable
for symmetric and asymmetric steel beam-to-column internal joints are presented and their
implementation in the Open System for Earthquake Engineering Simulation (OpenSees) is
explained with reference to some case studies.
3 Formulation of beam-to-column joint models

The rigorous analysis of beam-to-column joints requires to distinguish separate sources of
deformability, namely those due to the column web panel and those due to the connection.
In the model represented in Fig. 1 (on left) the shear force in the column web can be assumed
constant and, accordingly, a single shear panel was considered (SP). On the other hand, in
case the depth of the beams connected by a beam-to-column joint is different, the shear stress
distribution in the column web can no longer be assumed constant [1], leading to the model
represented in Fig. 1 (on right) with a double shear panel (DP).
175
8
9
node 3
7
(external)
8 11
9 5
0
node 3 12 2
(external) 7
5
10
dbR/2
6
5 3
0
2 7 14
11
dbL/2
11
11
5 12 node 6 4
12 6 node 2
dbR/2
6
db /2
node 7 (internal) (external)

6 3 13
7 14 (internal)
13 10 
node 4 node 5
10 node 2 4
node 4 node 5 node 6 (external) (internal)
15
db /2
(internal) (internal) (external) 1

(external)
dbL/2

8
9
4 1 
0
8
node 1 10
4
0
(external) 1 9
2 3 node 1
(external) 1
0 dc/2 dc/2 0 2 3
0 dc / 2 dc / 2 0
Fig. 1: Single panel (SP) (on left) and Double panel (DP) (on right)nbeam-to-column joint model.
2 Case study
In order to assess the Single (SP) and Douple (DP) model, the behaviour of frame of Fig. 2
has been studied under the indicated loading. Consider pinned column base joints, and full-
strength rigid joints at the external nodes. Three strategies were considered for the modeling
of the internal node: assuming rigid behavior (Case 2a); disregarding the asymmetric joint,
using a Single Panel element were db is assumed equal to 220 mm (Case 2b); and applying the
Double Panel element (Case 2c). Table 1 compares the results in terms of bending moment
(absolute values) for relevant cross-sections for the 3 cases. The results from Case 2c are tak-
en as reference, since the DP model represents the most appropriate modelling strategy.
p1= 15kN/m Table 1: Comparison between bending moment

p2= 5kN/m Case Case Case
2a 2b 2c
1 3e 3d 5
[kNm] [%] [kNm] [%] [kNm]
IPE 400 IPE 220
3p
M1 77.8 -12.4 75.8 -9.6 69.2
M2 163.1 -0.7 164.7 -1.8 161.9
10m
M3e 117.7 8.1 116.1 9.4 128.1
HEA HEB HEA
d
M3
240 400 240 52.0 -8.4 53.1 -10.9 47.9
p
M3 76.8 16.1 74.5 18.5 91.5
M4 11.6 -0.2 10.9 6.6 11.6
12m 8m
M5 0.8 79.5 0.2 94.5 3.7
Fig. 2: Geometry, sections, loads of case study
Comparing the results for the DP model (case 2c) and the other cases in terms of bending
moment (Table 10) shows significant differences either when comparing to rigid modelling of
the node or when disregarding the asymmetric node.
References
[1] Jordão S, Simões da Silva L, Simões R. “ Behaviour of welded beam-to-column joints
with beams of unequal depth”, Journal of Constructional Steel Research 91: 42-59, 2013.
176
Tampere, Finland
DESIGN APPROACH FOR STABILITY CHECK OF MEMBERS WITH

HANGING-PROFILE CONNECTIONS
Dasu Liua,*, Jörg Langeb

a,b
Technische Universität Darmstadt, Institute of Steel Structures and Materials Mechanics, Germany
*
Author for contact. Tel.: 0049 6151 162345; E-mail: liu@stahlbau.tu-darmstadt.de
Abstract: The “hanging-profile” is a commonly used type of joint in steel structures for
horizontal bracing systems in large building projects, where large compression forces are
transferred. This kind of joint is characterized by the connection of a bracing beam to gusset
plates only at the top flange. This type of structure is usually simplified as a single-span beam
with so-called fork bearings at the ends. Based on an extensive parametric study using finite-
element-models calibrated on a serial of 1:1 scale tests, a new design approach was developed
for the lateral torsional buckling check of members with this kind of connections.
1 Introduction
In large building projects, such as power plants, the “hanging-profile” is a very commonly used
joint for diagonals of horizontal bracing systems. As illustrated in Fig. 1 only the top flange of
the bracing beam is connected to the gusset plates.
b) Angels α,β and γ for the geometry

a) A typical hanging-profile conenction
variation of gusset plates
Fig. 1: Hanging profile connection
This type of structure is usually designed and verified as a single-span member with so-called
end fork bearings, which strictly restrain all the in plane cross section deformations
(translations, rotation about the longitudinal axis and cross section deformation). Due to the
177
overestimation of the system stiffness in the connection areas under this assumption, basic
engineering with this conventional approach has been considered generally not safe. This study
attempts to shed light on the calculation of this type of structures and develop a new design
method derived from Eurocode for lateral torsional buckling verification of here discussed
cases.
2 New design approach and comparison

A serial of tests on HEA 200 beams with system lengths from 3 m to 6 m and variation of gusset
plate geometries (α−β−γ-combinations, see Fig. 1 b) were conducted. Based on the test results,
an FE model with the geometric imperfections, residual stresses and nonlinear material
behaviour was established for a parametric study.
This extensive parametric study covered the commonly used HEA, HEB und HEM profiles,
different gusset plate shapes and two steel grades (S235 & S355). Based on the results of the
parametric study, new design formulae were developed. These formulae were further analysed
with the semi-probabilistic method according to EN 1990 Annex D, so that a safety factor (γM1
= 1.1) could be calculated to meet the standard safety level. The new approach has the following
format:
, = , ∙ ⁄ (1)
With:
= + − ∙ ̅ (2)
= 0,5 1 + ∙ ̅
− ̅ + ∙ ̅ (3)
̅ = , ⁄ (4)
, = , , ⁄ 0,5 ∙ ℎ ∙ (1 − 0,5 ∙ ) + , , ⁄ , (5)
with: = −2∙ ∙ ⁄ ≤ 0,5
∗ ∗) ∗)
= ( ∙ + − ( ∗ ∙ + ∗) −4 ∗ ∙ ∗ ∙( − ∗) 2( − (6)
with: ∗
= ∙ ⁄ ; ∗
= + ∙ ⁄ ; ∗
= ℎ² ∙ ⁄(8 )
The coefficients are:
Table 1: Coefficients
S235 S355
Profil α λ β α λ β
HEA 0,75 0,15 0,85 0,70 0,15 0,85
HEB 0,55 0,10 0,95 0,50 0,10 0,95
HEM 0,35 0,10 1,00 0,30 0,10 1,00
3 Conclusions
From this study the following main conclusions can be drawn:
1. In comparison with the new approach, the conventional approach under the fork bearing
assumption cannot predict the carrying capacity of structures with hanging-profile
connections correctly. The error band is up to 20%;
2. On one hand, since the bracing member is partially fixed about the major bending axis by
the gusset plates, higher ultimate loads (compared with single span member) can be
achieved thanks to the higher buckling loads. On the other hand, the gusset plates can also
effect the carrying load of the entire structure disadvantageously due to the much lower
bending and torsional stiffness compared with the beam. The second effect dominates
especially in systems with low slenderness, where the gusset plates behave in a wider
range plastically and therefore suffer under more stiffness lost.
178
Tampere, Finland
REASONS FOR CHARLES DE GAULLE AIRPORT COLLAPSE
Toomas Kaljas
Rak Tek Solutions Oy
*
Author for contact. Tel.: +358 445575835 E-mail: toomas.kaljas@rakteksolutions.fi
1. Abstract
In the early morning hours of May 23, 2004, passengers in Terminal 2E at the Charles de
Gaulle Airport in Paris partially collapsed resulting in several fatalities. Structural failure was
caused by multiple reasons, all contributing to failure. Similar structures have been success-
fully erected and built around the world. One famous and comparable structure is the Berlin
main railway station. After investigations it becomes clear that Charles de Gaulle Airport
lacks suitable and effective geometry, which is present in Berlin Railway station.
The aim of this paper is to compare the similarities and differences between de Gaulle airport
and Berlin main railway station externally reinforced elliptical portal frames.
Both Charles de Gaulle airport and Berlin main railway station are externally reinforced with
tension bars –tendons. Both structures are working elliptical frames with hinge support condi-
tions. However, the structures look similar, but have substantial differences.
In Berlin main railway station the external reinforcement follows the shape of moment dia-
gram and is placed at the tension side of bending. External reinforcement is also attached to
inner compressive side through diagonals, which provide proper shear stiffness and allow ten-
sile and compressive side to work together as one composite member.
The Charles de Gaulle airport external reinforcement had inadequate shape of external tensile
reinforcement and no suitable shear stiffness and strength between tensile and compressive
side disabling the composite effect
179
2. Conclusions
1. Based on numerical analysis of Charles de Gaulle airport terminal 2E geometry and

photographic evidence, it is clear that the external reinforcement had been chosen
based on appealing architecture, not based on solid engineering judgement. Ambitious
geometry could not have been rigorously analysed and designed, since the failed struc-
ture obviously had lacked important design aspects, like proper geometry and suitable
shear stiffness
2. Similar externally reinforced curved frame had been successfully designed and build
in Berlin main railway station. This structure clearly demonstrates all the good design
features for such frames. Berlin main railway station has also redundancy due to clear-
ly designed shear stiffness between external reinforcement and internal compressive
frame allowing the external reinforcement- internal steel arch work like a composite
member under un-symmetrical loading conditions and under extreme loading.
3. Charles de Gaule airport terminal 2E failed on locations, where concrete shells had
been penetrated by passenger tunnels. Such penetrations worked like stress concentra-
tors, but are not by itself the reasons for collapse. Concrete structures, just like steel
have redistribution capacity. With suitable and careful design, even opening among
concrete shells are not catastrophic.
4. Independent of main compressive member material, steel or concrete, good design is
achievable. For tension members steel is suitable material.
References
[1] Building Collapse Cases/Charles de Gaulle Airport

http://matdl.org/failurecases/Building_Collapse_Cases/Charles_de_Gaulle_Airport.ht
ml
[2] Berlin main railway

http://en.wikipedia.org/wiki/Berlin_Hauptbahnhof
[3] EN 1992-1-1:2004
[4] EN 1993-1-1
[5] RFEM 5.04 (https://www.dlubal.com/en/rfem-5xx.aspx)
180
Nord
dic Steel Co nstruction Conference
C 2015
Tam
mpere, Finlland
23-255 Septemberr 2015
IN
NVESTIG
GATIONS ON THE
E BEHAV
VIOUR OF THREA
ADED AN
ND
S
SHANK BOLTS
B UNDER
U C
COMBINE
ED TENS
SION AND
D SHEAR
R
Anja Rennnera and Jö

örg Langeb
a,b
Instittute for Steel Structures
S andd Materials Mechanics,
M TU
U Darmstadt, G
Germany
Abstracct: To estim mate the acttual behavioour of comb bined loaded bolts a seeries of tests on dif-
ferent tyypes of boltts was cond ducted. By nnow all testt series have been finisshed and th he results
are not as expectedd. The threaaded bolts shhow a nearlly ideal plasstic behavioour that fits with the
previouusly known quadratic ru ule. The ressults of testts on all bolts with shaank in shearr plane -
indepenndent of thee bolt gradee - are overeestimated by b the quadrratic approaach. The reeason for
this uneexpected beehaviour is the t subject of a curren nt large scalle parametrric numericaal study.
The firsst results off this investtigation andd some first attempts fo or an interppretation aree subject
of this ppaper.
1 Testts on boltts
More thhan 150 bollts of a sizee M 20, bollt grades 4.6, 8.8 and 10.9 and w with either thread or
shank inn the shear plane
p were tested undeer different tension-to-s
t shear load raatios.
2 Num
merical model
m
A numeerical model was createed using AN NSYS Workbench to get g a better view what happens
in a boltt when loadded simultan
neously withh tension an nd shear.
Fig. 1 shows a com mparison off the real tessted bolts and the nummerical simuulation undeer two of
the testeed angles. The
T numericcal model iss able to sim mulate the deeformation well.
Fig.. 1: Bolts graade 4.6 after testing comppared to defo

ormation plotts of numericcal simulatioon under
different angles
a of loaad applicationn (0° would equal pure teension, 90° ppure shear)
181
2 Nordic
No Steel Constructioon Conferen
nce 2015
3 Resu
ults
In Fig. 2 for bolts with
w shank in shear plaane the meaan values of all test ressults and th he results
of the nnumerical annalysis are summarize d. The ultim mate load ofo the testedd bolds wass divided
into its normal andd its shear coomponent. T These valuees were relaated to the ppure tension n or pure
shear reesistance, which
w were determined
d by tests on
n bolts of th
he same chaarge. The an nalytical
results ccan also be related to th
heir pure tennsion and sh
hear load caapacities.
The testt and the annalytical resu
ults can noww be compaared to the bi-linear
b lim
miting functiion from
the EN 1993-1-8, the t so-called Eurocodee 3, and the quadratic interaction
i rrules of thee old Eu-
ropean sstandards.
Fig.. 2: Test and analytical reesults for bollts with shan

nk in shear pllane, related to their puree tension
and shear loadbearing g resistance
4 Inteerpretatioon
Fig. 2 shhows that thhe analyticaal results liee clearly und
der the quaddratic functtion as the results
r of
the boltt tests. The image of thet deformeed bolt (Fig g. 3) gives a first ideaa why the quadratic
q
functionn is not reacched. The quadratic
q asssumption goes
g back to
o the linearr-elastic relaation be-
tween nnormal and shear
s stress, and forcess are relatedd to the undeeformed oriiginal cross section.
Fig. 3: Deformed
d bolt with leeft over cross section to transfer
t norm
mal force
Lookingg at the defoormed bolt (Fig. 3) shoows that thee area left ov
ver to transffer the norm
mal force
is by faar smaller thhan the orig
gin cross secction. It might be moree appropriaate to relate the nor-
mal forcce componeent to the ten nsion load ccapacity of the reducedd cross sectiion.
182
Tampere, Finland
BEHAVIOR IMPROVEMENT OF PULTRUDED FRP

BEAM - COLUMN BOLTED CONNECTIONS
Ossama M. El Hosseinya, Hassan M. Maalyb, Said A. Ibrahimc

a,b,c
Structural Engineering department, Zagazig University, Sharkia, Egypt
dr.ossama@bect.net Tel. 002102 22131501
Abstract: Bolted joints for its advantages of the easy assembly and disassembly are often
preferred in many composite joining applications. In this paper, a finite element model using
ANSYS software is developed to investigate the behavior of connection between pultruded
fiber reinforced polymer (PFRP) elements i.e beam and column. It was concluded that stiff-
ening lower connecting angle at beam - column connection could have better effect on con-
nection capacity than stiffening the upper connecting angle. stiffening upper and lower angles
have a great effect on improving moment rotation capacity and failure load of the connec-
tions.
1 Introduction
During last few decades, Pultruded Fiber Reinforced Polymer (PFRP) composites have been
progressively used in corrosive environments. Standard pultruded (thin-walled) FRP profiles
may have the same cross-sectional shape as conventional steelwork [1]. Technical infor-
mation on the pultrusion process, and PFRP shapes themselves, is given in two manufactur-
ers’ Design Manuals [2,3]. Although FRP shapes might be similar to steel sections in shape,
their structural behavior is different [2,3]. Bolted joints are commonly used to connect beams
and columns in a FRP framed structure [4]. Behavior of a joint is represented by moment-
rotation characteristics relating the moment transmitted by the joint to the relative rotation be-
tween the connected members. One of the pioneering studies was that of Mosallam [5] pre-
senting results of a comprehensive theoretical and experimental program to evaluate both the
short-and long-term behavior of PFRP portal frames subjected to both quasi-static and sus-
tained loading . A numerical FEM developed in this paper and validated based on a test pro-
grams already described [5]. A parametric study is executed for the behavior of upper, lower
and web cleat angles connecting beam to column using the developed FEM. Different cleat
angles gusset plate stiffener locations are studied and the different relations (failure load,
moment rotation capacity and stress concentration) are presented.
183
2 Parametric study
Four full-size connections were modelled using FEM (ANSYS 15.0) ,where the connections
consist of two wide-flange 8 x 8 x 3/8 vinyl ester standard pultruded sections (Pultex™ 1625
manufactured by Creative Pultrusion, Inc.)[2] , 6 x 6 x 1/2 pultruded angles in a conventional
top, bottom (seat) and 3 x 3 x 1/4 web clip angle. Connection (a) stiffening the connecting
angle by adding top and bottom triangular gusset plates (FSC) (b) stiffening the connecting
angle by adding top triangular gusset plate (USC) (c) stiffening the connecting angle by add-
ing bottom triangular gusset plate (LSC) (d) no stiffeners (NSC) . The angles were bolted to
both the beam and column sections using steel bolts M16 grade 4.6 and 3 mm thickness
washer. Fig. 1 shows a good agreement comparison between the test results of previous ex-
perimental work [5] and the present developed finite element analysis. Stiffening the tension
side angle of the beam - column connection (USC) increase the moment capacity by about 5%
and decrease the rotation by about 34% compared to NSC model. Stiffening the compression
side angle (LSC) increase the moment capacity by about 15% and decrease the rotation by
about 28% compared to NSC model. Stiffening both sides (FSC) increase the moment capaci-
ty by about 46% and decrease the rotation by about 30% compared to NSC model.
15
FEM
EXP [7]
Moment(KN.m)
10
USC
NSC
5 FSC
LSC
0
0 5 10 15 20 25 30 35 40 45 50
Rotation(mrad)
Fig. 3: Moment – Rotation relation

3 Conclusions
Stiffening lower connecting angle at beam column connection is more efficient than
stiffening the upper one. Stiffening upper and lower connecting angles shows a significant
improvement in connection behavior. Furthermore, it was found that drilling holes in the
composite material should be taken into consideration to avoid creating cracks due to stress
concentration.
References
[1] Bank LC. Composites for construction - Structural design with FRP materials. John
Wiley & Sons, New Jersey, 2006.
[2] The new and improved Pultex pultrusion design manual. Creative Pultrusions Inc., Al-
um Bank, PA. (www.creativepultrusions.com/library.html) (April 5, 2014).
[3] Strongwell design manual. Strongwell, Bristol, VA. (www.strongwell.com.
[4] TURVEY, G. J. and COOPER, C., “Review of tests on bolted joints between pultruded
GRP profiles”, Structures and Buildings, Vol.157, No. 3, 2004, pp. 211-233.
[5] L.C.Bank, A.M.ASCE, A.S. Mosallam and H.E.Gonsior. (1990)"Beam – To – Column
Connections for Pultruded FRP Structure ", NSF grant no. MSM-8896265.
184
185
186
Tampere, Finland
VIBRATION RESPONSE OF USFB COMPOSITE FLOORS
Richard Kansinallya and Konstantinos Daniel Tsavdaridisb

a,b
Institute for Resilient Infrastructure, School of Civil Engineering, University of Leeds, Leeds, LS2 9JT, UK
Abstract
USFB Flooring System
Nowadays, structural engineers are faced with new challenges such as innovating flooring so-
lutions that minimise construction cost while simultaneously allow for optimum space utilisa-
tion within certain constraints. As a result, slender (e.g. slim or shallow) floors are created
leading to the issue of unwanted floor vibrations, which many engineers today are not too
cognisant of (Mello et al., 2008). By slim or shallow floor, it is implied that the depth of the
concrete slab is located within the flanges of the steel beam as opposed to the traditional SCC
flooring systems in which the slab is supported by the top flange of the steel beam (Fig. 1a).
In this way, it is evident that it is possible to have a reduction in the structural depth which
translates into cumulative savings in the floor-to-floor height in medium to high-rise struc-
tures (Fig. 1a). One way to construct these slim floors, Ultra-Shallow Floor Beams (USFBs)
are made by fabricating welded or rolled steel sections to make an asymmetrical I-section that
results in a wider bottom flange. This is done to provide sufficient bearing distances for the
steel decking or the precast concrete units.
a) Cross Section of the USFB b) Structural Plan of Slab

Fig. 1: Structural Model of the Composite Slab
As the demand for lightweight structures with clear floor spans increases, long spanning SCC
floors are encouraged through the use of perforated steel beams (e.g. USFB), which also allow
for possible service integration within the floor depth (Tsavdaridis et al., 2013). Even though
a lightweight flooring solution that is capable of accommodating long spans is achieved
through the use of the USFB, the reduction in floor depth results in a flexible structure that
becomes sensitive to excessive vibrations.
187
Structural Model
The SCC flooring system investigated comprised of a perforated steel USFB which was treat-
ed with both pinned and fixed supports for the ease of comparison. In addition, the concrete
depth varied in order to draw their influence on the vibration response of the novel composite
slab. The length of the USFB model remained constant at 7.4m while the slab span also varied
between 2.5m and 4.5m (Fig. 1b). The structural properties of the perforated USFB were de-
veloped by combining two sections; namely the 305x127x37UB and 254x254x73UC. Prima-
ry beams were the 305x127x42UB. The structural concrete depth varied throughout the anal-
yses, and 1.2mm thickness was adopted for the decking, assuming a 210ComFlor steel deck-
ing properties.
FE Model of the USFB System

Modelling of this new composite floor type was done by using popular commercial software
ABAQUS CAE V.6.10. A two-bay floor arrangement consisted of secondary beams
210x127/254x55 ACB and supported by 305x127x42UB primary beams. In order to practi-
cally represent construction procedures, structural coping (notching) of 62.2mm was applied
to the secondary USFBs.
Parametric Study
Modal analysis was carried out to primarily extract the natural frequencies and to assess how
this dynamic response changes with concrete depth and boundary conditions alterations. The
vibration mode shapes were also examined. This investigation was performed on five differ-
ent floor spans (7.4mx5m, 7.4mx6m, 7.4mx7m, 7.4mx8m, 7.4mx9m) so as to develop the
rational limits about the vertical plane for such USFB floors. In the second stage, a linear per-
turbation steady state modal dynamic analysis was conducted to assess the acceleration per-
formance of this novel composite floor under a human induced load model suggested by Mur-
ray et al. (2003) and Mello et al. (2008). Throughout this study, a notional damping of 3%
was utilised in accordance with Bachmann et al., 1995.
Conclusions
The main conclusions of the research were:
Less participation of increased mass in earlier vibration modes.
Slabs with fixed supports yield higher natural frequencies and are preferable.
Increasing slab spans reduced the natural frequencies.
Potential for the use of composite slabs with USFBs as frequencies were higher than
minimum floor frequency of 3Hz.
References
Bachmann, H. et al., 1995. Vibration Problems in Structures: Practical Guidelines. Basel: Birkhauser
Verlag Basel.
Mello, A. de A.V. et al., 2008. Modal Analysis of Orthotrophic Composite Floor Slabs with Profiled
Steel Decs. Latin American Journal of Solids and Structures, Volume 5, pp. 47-73.
Murray, T.M. et al., 2003. Floor Vibrations due to Human Activity. United States of America:
American Institute of Steel Construction.
Tsavdaridis, K.D. et al., 2013. Experimental and Computational Study of Vertical Shear Behaviour of
Partially Encased Perforated Steel Beams. Engineering Structures, Vol. 56, pp. 805-822.
188
Tampere, Finland
ANALYSES OF THE LOAD BEARING BEHAVIOUR OF DEEP-

EMBEDDED CONCRETE DOWELS, CoSFB
Matthias BRAUNa,*, Renata OBIALAb and Christoph ODENBREITb

a
ArcelorMittal Europe – Long Products, L-4009 Esch-sur-Alzette, Luxembourg
b
University of Luxembourg, Chair of Steel and Façade Engineering, L-1359 Luxembourg
*
Author for contact. Tel.: +352 53 13 30 20; E-mail: mathias.braun@arcelormittal.com
Abstract: In this paper the development of CoSFB-Betondübel is presented. CoSFB-

Betondübel is a deep-embedded concrete dowel connecting an in-situ concrete with a steel
section assuring a composite action and allowing for composite beam design. The load bear-
ing behaviour and parameters influencing this behaviour were determined through experimen-
tal tests. Special focus was given to the influence of the ratio of the resistance of the concrete
dowel to the concrete compression class. Further investigations will be performed via FE-
analysis in ABAQUS. 3D models with nonlinear material and geometry were prepared and
validation was undertaken. In addition, a realized application example for CoSFB is shown.
1 Deep-embedded concrete dowels – “CoSFB-Betondübel”

Concrete dowels consist of circular openings in the web of a hot rolled steel section, rein-
forcement bars crossing the openings and concrete infill (Fig. 1), Eurosteel [1].
concrete dowel in-situ concrete
dowel reinforcement
slab element
CoSFB
Fig. 1: CoSFB - Composite Slim-Floor Beam [1]

The so called “CoSFB-Betondübel” is a concrete dowel placed in the chamber of a hot rolled
section [2]. Because concrete in the chamber is restrained by the lower and the upper flange
and by the web of the steel section, failure of the concrete due to an expansion towards a free
edge cannot occur [3].
2 Test Campaign
The global behaviour of the composite slim-floor beam was investigated with two shear beam
tests and two beam tests. All tests showed a very ductile behaviour with large deformations
until failure. The whole width of the concrete slab of the specimen was activated. Also a full
composite action between the steel section and the concrete chord could be proved.
189
To investigate the characteristics of the shear connection 27 push-out tests with varied pa-
rameters were performed. All performed push-out tests reached a slip > 6mm and can be clas-
sified as ductile in accordance with EN 1994-1-1. Failure occurred due to an exceeding of the
maximum elongation of the dowel reinforcement. The test results are described in detail in
[4]. The application and the design of “CoSFB-Betondübel” in ULS and SLS are covered by a
National Technical Approval [2].
3 Numerical Analysis
Numerical simulation of the composite system described before is a highly nonlinear problem
with material and contact nonlinearities and large displacements. For solving such a complex
problem a commercial FE-package ABAQUS [5] was used. Elasto-plastic material model was
applied to all the steel parts. Concrete was modelled using well recognised Concrete Damaged
Plasticity (CDP) model.
Fig. 2: Boundary conditions and mesh of the FE model

The aim of this ongoing research is to develop a model suitable to simulate behaviour of the
CoSFB system, especially local behaviour and failure modes of the shear connection. The
model will be validated with the results of the experimental tests, where different sections for
the beam and different concrete grades were used. The authors of the paper are at the very be-
ginning of the process. Once the model is validated and is fully reliable, a parametric study
will be performed. The parameters to be verified are mainly: diameter of the dowel, diameter
of the hole in the web, strength of materials (steel and concrete).
4 Conclusions & Outlook

The CoSFB-Betondübel enables to combine benefits coming from slim-floor construction
with composite construction. The presented results show that application of this shear connec-
tion significantly increases the load-bearing capacity. It is important that all these can be
achieved with easy fabrication and erection. With the completed validation of the numerical
model and intense parametrical study it will be possible to extend the application scope be-
yond the experimentally tested range of parameters and to develop optimized design rules.
References
[1] M. Braun, R. Obiala, Chr. Odenbreit, O. Hechler: CoSFB – Design and Application of a new Generation of Slim-Floor
Construction. 7th European Conference on Steel and Composite Structures (EUROSTEEL). Naples, Italy, 2014.
[2] Deutsches Institut für Bautechnik: Allgemeine bauaufsichtliche Zulassung – ArcelorMittal Belval & Differdange S.A.,
CoSFB-Betondübel. Zulassungsnummer Nr. Z-26.4-59, Berlin 2014.
[3] M. Braun, O. Hechler, R. Obiala: Untersuchungen zur Verbundtragwirkung von Betondübeln – Anwendung von tief-
liegenden Betondübeln bei Slim-Floor-Konstruktionen (CoSFB). Stahlbau 83 (2014), Issue 4, p. 302 – 308. Germany.
[4] M. Braun et al.: Experimentelle Untersuchungen von Slim-Floor-Trägern in Verbundbauweise – Anwendung von tief-
liegenden Betondübeln bei Slim-Floor-Konstruktionen - CoSFB. Stahlbau 83 (2014), Issue 10, p. 741 – 749. Germany.
[5] Abaqus 6.11 Online Documentation, © Dassault Systèmes, 2011
190
Tampere, Finland
EVALUATION OF AXIAL FORCE IMPACT ON THE FLEXIBILITY

OF A STEEL BEAM-TO-BEAM END-PLATE JOINT SUBJECTED TO
BENDING WHEN EXPOSED TO FIRE
Mariusz Maślaka, Małgorzata Snelab

a
Cracow University of Technology, Cracow, Poland
b
Lublin University of Technology, Lublin, Poland
Abstract: In the article the axial force influence on structural response of a steel, beam-to-
beam end-plate joint subjected to bending under fire conditions is estimated and discussed in
detail. The proposed calculation algorithm is based on the suitable generalization of the clas-
sical component method. In such a formal model, illustrated with the attached numerical ex-
ample, the effect of different reactions of the joint components being the parts of the joined
beams and of the bolts connecting these members to the considered fire exposure is taken into
account in the computational procedure. Finally, the modified relations between the bending
moment applied to the joint and the rotation of such a joint are specified.
1 Introduction
The aim of the presented paper is to demonstrate that the consideration of the bending mo-
ment – axial force interaction can play an important role when the resistance and the flexibil-
ity of a steel, beam-to-beam end-plate joint subjected to bending are assessed, especially in the
case when such a joint is exposed to fire. The effects of this interaction are visualized by the
appropriate modification of the relations between the bending moment applied to the joint and
the accompanying joint rotation. The detailed analysis deals with the semi-rigid joint shown in
Fig. 1. Consequences of both the bending – tension and the bending – compression interac-
tions are presented, compared and discussed in detail. In the first step of the analysis the joint
resistance is assessed including the impact of the axial force. Subsequently, the influence of
such an axial force on the considered joint rotation is estimated. Finally, the resistance of the
particular joint components is identified assuming that they are exposed to a fully developed
fire.
Fig. 1 The steel beam-to-beam end-plate joint considered in the example.
191
2 Obtained results and conclusions
Taking into account the increasing joint temperature under fire conditions, the effect of the
axial force impact, evaluated previously without considering fire influence, should be added to
the different effects being a consequence of the weakening of the mechanical properties of the
heated structural steel of which the considered joint components are made. The most im-
portant in such circumstances seem to be the appropriate reduction of the steel yield limit as
well as the inevitable decrease of the steels’ longitudinal elasticity modulus. Conclusively, the
suitable joint flexibility relations, between the bending moment applied to the joint and the
rotation of such a joint are significantly modified as shown in Fig. 2.
Fig. 2. Reduction of the considered joint resistance and of the accompanying joint stiffness under fire
conditions. Left - the comparison of the results obtained assuming the occurrence of the bending –
tension and the bending – compression interactions. Right - the consequences of the bending – tension
interaction calculated assuming 0% , 30% and 70% shares of the axial force.
Selected references
[1] Cerfontaine F, Jaspart JP. “Analytical study of the interaction between bending and axial
force in bolted joints”, Proceedings of the 3rd European Conference on Steel Structures
“Eurosteel”, Coimbra, Portugal, September 19-20, 2002, 997-1006.
[2] De Lima LRO, da Silva LS, da S. Vellasco PCG, de Andrade SAL. “Experimental analy-
sis of extended end-plate beam-to-column joints under bending and axial force”, Pro-
ceedings of the 3rd European Conference on Steel Structures “Eurosteel”, Coimbra, Por-
tugal, September 19-20, 2002, 1121-1130.
[3] Sokol Z, Wald F, Delabre V, Muzeau JP, Švarc M. “Design of end-plate joints subject to
moment and normal force”, Proceedings of the 3rd European Conference on Steel Struc-
tures “Eurosteel”, Coimbra, Portugal, September 19-20, 2002, 1219-1228.
[4] Urbonas K, Daniūnas A. “Component method extension to steel beam-to-beam and beam-
to-column knee joints under bending and axial forces”, Journal of Civil Engineering and
Management, 3, (2005), 217-224.
[5] Del Savio AA, Nethercot DA, Vellasco PCGS, Andrade SAL, Martha LF. “Generalised
component-based model for beam-to-column connections including axial versus moment
interaction”, Journal of Constructional Steel Research, 65, (2009), 1876-1895.
[6] Daniūnas A, Urbonas K. “Influence of the semi-rigid bolted steel joints on the frame be-
haviour”, Journal of Civil Engineering and Management, 16 (2), 2010, 237-241.
192
Tampere, Finland
FIRE DESIGN OF CFST COLUMNS

Improvements required for Eurocode 4
Matti V. LESKELA
RAKOSPER, FINLAND
Tel.: +358 40 5915530; E-mail: matti.v.leskela@dnainternet.net
Abstract
The inconsistencies and irregularities in the fire design rules of EN 1994-1-2 for CFST col-
umns have been known for years, and the method of Annex H is based on odd principles to-
tally different from those given in section 4.3.5.1 of EN 1994-1-2 which should be applicable
for all types of the composite column in the scope of the code, according to which the axial
resistance is evaluated as the buckling load taking into account the effective bending stiffness
and effective length of the fire exposed column. The bending effects should also be consid-
ered, although it seems that the information of Fig. 4.6 of the section has erroneously been
understood to mean that all fire exposed columns of the braced frames turn into axially loaded
ones without any eccentricity effects. In fact, the Figure only shows how the transverse de-
flection profile and effective length of a continuous column are changing during the fire ex-
posure, but at the same time there is a phenomenon that the column in the fire exposed storey
does not receive bending moments from the storeys above and below it. Moreover, it is obvi-
ous that bending moments are not distributed from the heated storey to the unheated ones.
However, it is a misunderstanding that the fire exposed column unconditionally turns into a
purely axially loaded one. Therefore it is required to evaluate the axial load bearing capacity
and its reduction due to possible bending effects.
Conclusions
The temperature distribution in the concrete cross-section of a fire exposed CFST column is
highly variable whereas in the steel hollow section temperature variation is quite limited. To
simplify the task of treating the mechanical properties of the concrete section, an equivalent
temperature for the whole concrete section can be specified such that its employment in every
element of the heated cross-section will yield a compressive resistance Nfi,c,R and/or a nomi-
nal bending stiffness (EI)c, that are similar to the values from the analysis based on variable
temperatures, i.e. c,equ = max{c,A,c,I} where the first temperature in the parentheses refers
to Nfi,c,R and the second one to (EI)c,. Numerical analyses indicate that typically c,A and c,I
are quite close to each other and the maximum of the two can be used for determining both
Nfi,c,R and (EI)c,. For Nfi,a,R and (EI)a,, the design temperature a may be defined as an aver-
age value of the steel section and the design temperature for the reinforcement, s may be
defined from the concrete temperature in the location of the bars.
193
Adaptation factors, called as reduction factors in EN 1994-1-2, are required for evaluating the
effective bending stiffness of the column section from the nominal values evaluated according
to the design temperatures of the material components. The values of the adaptation factors
depend on the definition of the buckling curve, which is specified as curve c in the present EN
1994-1-2, independent of the cross-section geometry. However, it has been shown that the
buckling curve defined for the column in ambient temperature design can as well be used as
the basis of defining the factors [1], or a specific fire buckling curve can be developed [2].
The effect of bending moments in a fire exposed CFST column is best analyzed using the de-
sign eccentricity of the load, efi = Mfi,Ed/Nfi,Ed. The design requirement is Nfi,Ed  Nfi,Rd,,
with Nfi,Rd, = Nfi,Rd and   1 to be defined in relation to efi. Taking into account the mo-
ment magnification due to second order effects,  may be resolved from
  
   0.5   fi   fi  4 /  (  ) 
2
 1 N fi,pl ,Rd  N fi,c,Rd (1)

 fi  1   e fi
  (  ) M fi,pl ,Rd
The plastic bending resistance Mfi,pl,Rd may be based on the three design temperatures, a, s
and c,equ and the respective design strengths, and with the help of these all the methods es-
tablished for Mpl,Rd in the ambient temperature design are available.
Acknowledgments
The author’s work for developing a simple fire design method for CFST columns based on
section 4.3.5.1 of EN 1994-1-2 has financially been supported by Peikko Group Oy. The
support is gratefully acknowledged.
References
[1] Espinós A, Numerical analysis of the fire resistance of circular and elliptical slender
concrete filled tubular columns. Doctoral Thesis, Universidad Politecnica de Valencia,
Department of Construction Engineering and Civil Engineering Projects, Spain 2012
[2] Renaud C, “Improvement and Extension of the Simple Calculation Method for Fire Re-
sistance of Unprotected Concrete Filled Hollow Columns” CIDECT Research Project
15Q, Final report, CTICM, France 2004
[3] Espinós A, Romero, ML and Hospitaler A, “Advanced model for predicting the fire re-
sistance for concrete filled tubular columns” Journal of Constructional Steel Research
66(8-9), 1030-1046, 2010
[4] Bergmann M und Hanswille G, “Näherungverfahren für die Brandbemessung von
Hohlprofilverbundstützen” Stahlbau, 81(12), 833-839, 2012
[5] Zhao B, “Slenderness limit for composite columns with concrete filled hollow sections
under fire situation” CTICM Report SRI – 10/83 – BZ/NB, France 2010
194
Tampere, Finland
CALCULATION OF STEEL TEMPERATURE IN OPEN CROSS

SECTIONS BASED ON FIRE EXPOSURE FROM CFD
Joakim Sandströma,b, Ulf Wickströma

a
Luleå University of Technology
b
Brandskyddslaget AB
Abstract: Evaluation of steel temperature for small and complex structural elements directly
in FDS introduces local effects which can lead to over prediction of the solid temperatures.
The solid temperature calculation in FDS is based on a one dimensional assumption and can-
not handle all the aspects of heat loss due to conduction. FDS is therefore likely to over pre-
dict the temperature in, for example, the web in open cross sections. In this paper, this issue is
demonstrated and handled with by the use of shadow effects in FE analysis. Two different
methods handling the local effects are presented. The different methods show different levels
of accuracy presenting a more complete method for thermal response calculations based on
numerical calculations of experimental data.
1 Introduction
In striving to achieve optimal structural fire safety design, the use of field models increase in
popularity. FDS (Fire Dynamics Simulator) is an open source CFD code developed
specifically for fire driven flows [1]. The surface temperatures of an exposed structural
element may then be calculated directly in the CFD code for various points. This
approximation is, however, not recommendable for complex element as the solid temperature
is calculated in one dimension in FDS. Two dimensional effects are ignored introducing
progressive errors due to local effects. As an option, data in the form of fire exposure from the
CFD calculations may be coupled to thermal response calculations performed with finite
element codes like TASEF [2].
This paper focus on the calculation of surface temperatures directly in FDS compared to
temperatures calculated with TASEF to demonstrate the numerical error introduced as a
consequence of local effects due to the one dimensional assumption. The paper also presents
and evaluates two methods for one way coupling with regards to shadow effects avoiding the
local effects introduced by the one dimensional assumption.
195
2 Method
The fire compartment studied consists of a room with the properties and dimensions of an
ISO room corner test [8]. A propane burner is located at one of the far end corners with an
elevation of 0,65 m yielding a constant effect of 450 kW. In this room, a beam supports the
ceiling slab of concrete. The dimensions and reference temperature points of the steel cross
section is shown in Fig. 1.
Upper flange temperature

Web temperature
200
10
Lower flange temperature
10
200
a) Reference points for thermal calculations b) dimensions of steel cross section

Fig. 1: Dimensions of the cross section supporting the concrete slab.
The steel temperature is calculated directly in FDS for the three reference points. This result is
then compared to finite element calculations using three different methods to assign the
boundary conditions. The first method assigns the adiabatic surface temperature calculated in
FDS for each surface in the FE-analysis. The second method assumes a radiating surface with
temperature 𝑇𝑟 covering the space between the flanges creating a void. The convective heat
transfer inside the void comes from the caluculated gas temperature. The third method uses
the same concept of a radiating plate covering the space between the flanges. In this method,
the plate is assigned the temperature 𝑇𝐴𝑆𝑇 and also heats the air inside the void.
3 Results
The temperature is calculated for the upper and lower flange and the centre of the web, see
Table 1.
Table 1: Maximum steel temperature for the different reference points.

Method Lower Web Upper
1 – Directly in FDS 567°C 709°C 600°C
2 – Full AST approach 594°C 696°C 661°C
3 – Shadow effects with 𝑇𝑟 and 𝑇𝑔 520°C 644°C 621°C
4 – Shadow effects with 𝑇𝐴𝑆𝑇 552°C 639°C 618°C
4 Conclusions
Fire exposure should be represented 𝑇𝑟 and 𝑇𝑔 . This method, method 3, avoids local effects
introduced by the one dimensional assumption used in FDS. An alternative is to use the
simple approach using only 𝑇𝐴𝑆𝑇 , or method 4. The deviation in temperatures predicted in
methods 1 and 2 from method 3 and 4, is explained with the local effect introduced when
calculating solid temperatures only in one dimension.
For thermal calculations of small elements, especially with complex geometries such as open
cross sections, the accuracy of the temperature prediction can be improved by using FEM
temperature calculations based on CFD calculations of the thermal exposure, i.e. CFD to FEM
coupling.
196
Tampere, Finland
LATERAL TORSIONAL BUCKLING RESISTANCE – A COMPARISON

OF ANALYTICAL AND NUMERICAL MODELS
R. Ebela and M. Knoblocha

a
Institute of Steel, Lightweight and Composite Structures, Ruhr Universität Bochum
1 Introduction
This paper presents an analytical and numerical analysis of the lateral torsional buckling be-
havior of specified static systems subject to uniaxial bending. Design results of common sim-
plified analytical models are compared to the results of a comprehensive numerical paramet-
ric study using the finite element approach. These simplified models are easy to use, but have
difficulties in appropriately describing the lateral torsional buckling behavior of more com-
plex steel structures with regard to static system (e.g. continuous beams, cantilevers), bounda-
ry conditions, cross-sections etc. The paper focuses on simplified models using reduction fac-
tors and buckling curves. In addition to the model used in EN 1993-1-1, the approach devel-
oped at TU Graz in recent years and further developments are analyzed.
2 Numerical analysis and study for failure modes

A numerical study using the finite element approach was carried out for analyzing the lateral
torsional member buckling behavior of common steel sections in uniaxial bending. The nu-
merical parametric study considered three different basic static systems (simple beam, canti-
lever beam and two-span beam) and two basic loading conditions (distributed loading and
single loads). The study considered eigenmode-conform geometric imperfections (combina-
tion of initial deformations and rotations) and residual stress patterns considering typical dis-
tributions. The following failure criteria were considered for numerically determining the re-
sistance of the steel beams: Cross-section failure at the supports and in the field, the eigenval-
ue failure of the partially plastic system, the reaching of a limit rotation ϑ = 0.3 rad and the
eigenvalue failure of the elastic system. The type of failure of steel beams strongly depends
on the length of the beam and the non-dimensional slenderness ratio respectively.
3 Comparative study for the method using reduction factors

The load-bearing resistances obtained from the numerical simulations were used for perform-
ing a comparative study of simplified analytical models. The comparative study considered
the simplified European design method as well as a simplified model developed at TU Graz.
Fig. 1 compares results of the simplified European design method using reduction factors for
lateral torsional buckling χLT,mod according to EN 1993-1-1 with the factors χcal,v+ϑ derived
from the numerical simulations. The static system, the cross sections and the slenderness ratio
197
strongly affects the difference between the results according to the simplified method and the
numerical simulations. For instance, the slenderness ratio leads to differences of up to 25 %.
Fig. 1: Comparsion of reduction factors according to EN 1993-1-1 and numerical results for rolled
sections depending on system and loading
A further development of the simplified method for lateral torsional buckling using reduction
factors has been developed at TU Graz. The influence of the slenderness ratio on the differ-
ence between the simplified method and the results of the numerical simulations is smaller
compared to the simplified European design approach, particularly for two-span beams. Addi-
tionally, the simplified method is more consistent for different cross sections.
4 Further developments of simplified models using reduction factors

For reducing the profile-dependent scattering, however, simplified models that use more than
two curves are suitable. In particular, the influence of the shape of the cross sections on the
resistance should be represented by a set of different lateral torsional buckling curves. The
decrease factor αT is suitable to consider the influence of the shape of the cross-section on the
lateral-torsional buckling resistance. The use of the absolute length of the beam is a disad-
vantage of the approach using the decrease factor. The ratio Iy/IT is well suited for considering
the influence of the shape of the cross-section. The results of the numerical study were used to
determine imperfection coefficients required to obtain equal resistances with the simple mod-
el using reduction factors. The mean as well as the maximum values were used to determine a
simplified functional relationship between the imperfection coefficients and the ratio Iy/IT.
5 Conclusions
This paper has presented a comparative study on simplified models for the lateral torsional
buckling resistance. Simplified models using reduction factors are easy in use. However,
these models may lead to conservative and even unconservative design results. The selection
of the buckling curve based on the dimension ratio h/b of the cross-section does not lead to
suitable design results which can be adapted for all cases. Simplified models can be improved
by considering the influence of the shape of the cross-section on the reduction factor. These
approaches take account of the rotation of the beam and the ratio of the inertia torsion mo-
ments. Possible approaches consider the Iy/IT-ratio and the Wy/Wz-ratio for example.
198
Tampere, Finland
INNOVATIVE CONSTRUCTION OF STUDENT RESIDENCES
Pedro A P Andrade a, Milan Veljkovic b, John Lundholm c and Tim Heistermann d

a, b, d
Lulea University of Technology, Dept. of Civil, Environmental and Natural Resources Engineering, Swe-
den
c
Part Construction AB, Sweden
Abstract
Sweden has a strong demand on the construction of student accommodations and therefore
significant efforts have been taken towards an affordable and easy solution of the problem. A
concept combining these requirements may be based on the use of structural steel frames in
combination with prefabricated 3D modules fully equipped and suitable for student accom-
modations. Therefore, the need to investigate and develop a system suitable for an effective
assembly of student residences is considered in this paper, as part of an international project,
Optimization of the frames for effective assembling - FRAMEUP. The Fig. 1 reveals an over-
view of the system within the execution process.
Fig. 1: Snapshot of the assembly of one 3D module
199
The Frameup system introduces a new approach in terms of execution technique which con-
sists of the execution of a building starting from the roof to the 1st floor. The existence of a
lifting system constituted of a horizontal rigid frame - grid - in combination with lifting tow-
ers - pylons - permits the erection of the building, promoting each time the building is lifted, a
clearance of one-floor-height plus tolerances at the ground level. This creates room enough
for the assembly of the lower floor from below the previously assembled floor. The procedure
is repeated several times according to the number of floors until the 1st floor of the building,
the last floor of the execution sequence, is assembled. Moreover the Frameup system intro-
duces an innovation, the Frameup conveyor system, which streamlines the assembly process
so to move/slide the elements, as they come, directly from the lorry to their final position in
the building.
The development of the Frameup system benefits from a stepwise detailed 3D modeling and
structural analysis and design tools. However, when it comes to attest the reliability and effi-
ciency of the system, a full scale feasibility test is essential and it is performed on the majority
of the sequences of construction.
Conclusions
The present paper introduces an innovative construction method where the Lifting system and
its different components, together modular building, represent a major achievement towards
the reduction of time and consequent costs. Thus, the present paper introduces an innovative
construction method for modular buildings, where the Lifting system and its different compo-
nents represent a major achievement towards the time and costs savings. Therefore, based on
the developed construction method and its implementation in a full scale feasibility test, the
following conclusions are drawn:
1. The full scale feasibility test has demonstrated the validity of the Frameup system
since the majority of the sequences of construction were tested successfully.
2. The Lifting system is able to reduce the execution time because it takes the advantage
of prefabricated, modular elements and the 3D modules are directly assembled in their
final position by the originally designed conveyor system.
3. The risks and time losses associated with work at height and erection of construction
material do not exist, since the majority of the assembly work is performed at the
ground level. Quantity of the work to be performed in-situ is heavily reduced which
additionally improves safety and execution speed at the construction site.
4. The specific construction method introduced by the Frameup system has the advantage
of performing the whole work under protection of the building. This may allow the ex-
tension of the construction period with less cost, especially in places where climate is
an issue.
The combination of these factors proves the feasibility of the Frameup system. However, fur-
ther research needs to be undertaken in order to optimize the lifting process, where a new fea-
sibility test is intended to be performed before the construction of a Student Residence.
Acknowledgments
The paper has been elaborated within the framework of RFCS projects: Optimization of the
frames for effective assembling –”FRAMEUP”– RFS-PR-10121
200
Tampere, Finland
FATIGUE LIFE IMPROVEMENT OF WELDED BRIDGE DETAILS US-

ING HIGH FREQUENCY MECHANICAL IMPACT (HFMI) TREAT-
MENT
Poja Shams Hakimi a, Andrea Mosiello b, Konstantinos Kostakakis c, Mohammad Al-Emrani d

a,b,c,d
Department of Civil and Environmental Engineering, CTH, Gothenburg, Sweden
a
ELU Konsult AB
Extended Abstract
Post weld treatment (PWT) techniques are used as measures to enhance the fatigue perfor-
mance of steel and aluminium structures. These techniques have proven beneficial in various
applications such as submarine hulls, offshore wind platforms and cranes. High Frequency
Mechanical Impact (HFMI) treatment enhance the fatigue life of weldments by reducing the
notch stresses, hardening the metal surface and inducing compressive surface residual stress-
es. This paper gives a short presentation of the HFMI technology and examples of their appli-
cation in steel bridges. A feasibility assessment of four Swedish bridges is presented and a
parametric study on the potential of material saving with PWT on steel railway bridges is also
performed.
There are several different categories of PWT techniques, some of the most common being
burr grinding (BG), TIG dressing (TIG) and peening methods such as hammer and needle
peening. A more recent development in the field is the use of HFMI treatment. In general, all
PWT methods enhance the fatigue strength through two main mechanisms:
1. Smother geometric weld/plate transition in the area of weld toe, where fatigue cracks
are expected to initiate
2. Removal of surface weld defects (such as undercut) from the toe area.
The peening methods, such as HFMI treatment, give additional advantages by introducing
compressive residual stresses around the weld toe area. In as-welded details, the weld toe area
usually experiences considerable tensile residual stresses, which are unfavorable in terms of
fatigue. Thus, altering the residual stresses into favorable compressive stresses leads to higher
fatigue strength. In this paper, three examples of application of HFMI on new bridges are pre-
sented. Also, a list of six examples of HFMI application for repairing existing bridges are giv-
en.
Furthermore, feasibility assessments made on four different bridges of varying types are pre-
formed to investigate the effects of post weld treatment and high strength steel. Also, a para-
metric study is carried out for simply supported railway bridges with varying span lengths.
201
The utilization ratios in the ultimate, serviceability and fatigue limit states are calculated to
study which one that dominates at different span lengths and how the potential of material
saving may vary.
It is concluded that HFMI can yield substantial improvement of the fatigue strength of welded
structures, in particular when steel materials with higher yield strength than 355MPa are used.
This method gives the best results for high cycle fatigue due to that the S-N curves become
flatter, however, the risk of relaxation of the induced residual stresses must be considered
when there are potential of compressive overloads. For bridges, HFMI is an enabler for
implementation of steel grades above 355MPa which in turn can reduce the weight of the
structure further if the ultimate limit state governs.
The feasibility assessment shows that the benifits of PWT can be realized in the design of
both road and railway bridges. Substantial material reduction can be achieved by treating few
critical details in the bridge. If the ultimate limit state governs after treatment, implementing
steel grades above 355MPa gives additional weight and material reduction. For railway
bridges, which are considered in the parametric study, the material reduction varies between
30% for short-span bridges and 20% for bridges with spans of 30m, when the fatigue strength
is increased with three classes. No account is taken in this study for the change in slope of the
S-N curves of PW-treated details, which should give additional saving. In addition, if a fatigue
strength improvment of more than three classes can be realized, further reduction in material
can be obtained for bridges with spans less than 25m.
202
Tampere, Finland
New developments in heavy plate production for modern steel construction
Dr. Tobias Lehnerta,* and Dr. Falko Schröterb

a,b
Dillinger Hütte, Werkstr. 1, 66763 Dillingen, Germany
*
Author for contact. Tel.: +49 6838 47 2394; E-mail: tobias.lehnert@dillinger.biz
Extended Abstract
Small steps but big impact. Today’s steel construction is characterized by highest demands on
efficiency, quality and sustainability. Facing these challenges the steel construction fabricators
are longing for new material developments which can support their ambitions to either fasten
production or reduce the cost and environmental impact of a steel construction. Heavy steel
plates are nowadays one of the main input materials for steel constructions. Even though
modern heavy plate production is a far developed and well established industry, quality steel
producer invest consequently in research and development to further improve their high per-
formance steels. These smaller development steps can nevertheless lead to major advantages
for fabricators and designers in steel construction. This paper presents some of these recent
“small” steps in heavy plate production and demonstrates their potential benefits for the fabri-
cators. Thermomechanically rolling (TM), for example, is a well-known process used for
many years in linepipe industry. Over the past years such TM-plates have gained more and
more ground in the steel construction sector due to their superior processing properties. Late-
ly, ambitious requirements on even enhanced weldability for higher strength steel plates came
up in the steel market. By an onward improvement of the chemical composition these chal-
lenging demands can be served with TM-plates.
Beyond that, further expending the thickness of the primary material (so called slabs) allowed
rolling of the so far thickest TM-plates, with plate thickness up to 140 mm. Especially the off-
shore wind industry which needs thick and heavy plates with excellent weldability to allow
fast fabrication and to meet the designated efficiency goals will profit from this new develop-
ment in thermomechanical rolling.
In general, there are different ways a steel fabricator as well as an engineer can profit from
using such thermomechanically rolled steels:
1. Transition from “normalized” to “thermomechanically rolled” steels in the same
strength range (e.g. S355N  S355M)
2. Transition to higher strength while maintaining a very good weldability (e.g.
S355N  S460M)
203
The paper will further specify and name these benefits thermomechanically rolled plates can
offer and will give some examples of the usage of thermomechanically rolled plates in steel
construction.
A second recent development in this context is the combination of thermomechanically rolling
with weathering properties to overcome the existing problem of worse weldability for higher
strength weathering steel.
Weather-resistant  CET  (Weldability )

Higher Strength  CET  (Weldability )
This new steel grade opens up the possibility to exploit at once the beneficial effects which
arise from the usage of weathering steel, e.g. sustainability aspects as well as the ones coming
from higher strength steel, e.g. weight reduction, slender architecture etc.
The aim of this paper is to shortly present these and other modern steel concepts (e.g. ther-
momechanical rolling, weathering steels or longitudinally profiled plates).
Furthermore their benefits as well as their potential in reducing cost and energy consumption
in fabrication, assembly and transportation will be illuminated.
204
Tampere, Finland
Stainless steel, a sustainable material for sustainable structures
Anders Finnås*, Camilla Kaplin
Outokumpu Stainless AB, Sweden

*
Author for contact. Tel.: +46 (0)586 47 131 ; E-mail: anders.finnas@outokumpu.com
Abstract:
A truly sustainable product needs to demonstrate that it can answer to all three aspects of sus-
tainability: environmental, economic and social. Stainless steel has the potential to perform
excellently in all three areas. Environmental excellence starts with the fact that stainless steel
is recycled and Outokumpu products have a higher than average recycled content of approxi-
mately 85 % recycled material. The energy saved by recycling means that the carbon footprint
is reduced. Studies have indicated that raising the share of recycled input from 50 %, which is
world average recycled content of stainless steel to 85 % can result in as much as 21 % less
CO2 emissions
Since structures of stainless steel have long life-spans, stainless steel is often economical in a
life-cycle perspective and the minimized maintenance also contributes to a good overall econ-
omy. It has been demonstrated that the reduced painting required for stainless steel structures
compared to carbon steel can reduce life cycle costs up to 30-40 %.
For the social dimension it is important that raw materials come from responsible sources and
producers must act responsibly, such as committing to ISO 26000 on social responsibility.
Stainless steels are durable and particularly suitable in harsh bridge and water environments.
Duplex stainless steels offer corrosion resistance and high strength which makes them ideal
where corrosion protection and high load-bearing capability are required. The duplex grades
offer mechanical properties equal to or better than the common structural steels. New duplex
steel grades have been developed, and today there is a whole family of duplex grades, making
it easier to specify a suitable grade in terms of both corrosion resistance and cost effective-
ness. The fabrication properties of duplex stainless steels are very good, offering good welding
and forming properties at least equal those of the ordinary structural steels.
Over the last ten years, duplex stainless steels have started to be used in bridge structures and
there is a potential for much greater use, particularly as bridge designers seek high perfor-
mance materials with extended service life and lower maintenance requirements. Duplex
stainless steels are suitable wherever resistance to certain environmental conditions combine
with the need for high load-bearing capability (design strengths are typically 450 MPa). Their
full potential is reached in locations where the structure comes into contact with corrosive en-
vironments (salt or brackish water, de-icing salts). Initially most duplex bridges were in du-
plex grades 2205 (1.4462) or 2304 (1.4362) but lately the trend is to use LDX 2101 (1.4162)
and LDX 2404 (1.4662) depending of on the severity of the environments.
205
The Ljunga Bay Bridge in Sölvesborg, on the southeast coast of Sweden, is one example of
sustainable design. The bridge is a 760 m long bicycle and pedestrian bridge. The bridge has
three supporting 60-metre long arches made from LDX 2101duplex stainless steel. The key
reasons for selecting duplex steel for the bridge arches were the overall life-cycle costs and the
low environmental impact. It was found that the slightly higher initial costs compared to a
carbon steel bridge would pay off at the first re-painting The absence of repainting also
avoids paint residues polluting the bay waters with a negative impact on a nearby bird conser-
vation area.
All water regulation structures have long life span requirements as well as functional safety.
This requires durable materials that do not rust away and do not need extensive maintenance
and repair works. Duplex stainless steels offer such materials.
When sluice gates at Gårda Dämme in Gothenburg, on the Swedish west coast, were rebuilt
stainless steel was used. A long term sustainable and environmentally friendly solution for
river regulation and flooding protection from the sea was required. Duplex stainless steel was
selected to ensure long trouble-free regulation, an economical service life, and avoiding the
release of chemicals from repair work into the water that may damage the river and the mi-
grating fish. The enhanced corrosion resistance of LDX 2404 is also a plus, since the water-
ways at Gårda Dämme receive occasional inflows of salty sea water.
Duplex stainless steels are excellent structural materials for civil engineering structures with
long service life such as bridges and lock gates where sustainability and economical life-cycle
perspective and are the main requirements. The range of different duplex stainless grades en-
ables engineers to choose the optimal solution for different service environments.
206
Tampere, Finland
DYNAMIC RESPONSE OF PIPE RACK STEEL STRUCTURES

SUBJECTED TO EXPLOSION LOADS
Anton Stade Aarønæsa, Hanna Nilssonb,* and Nicolas Neumannc

a,c
Aker Solutions, Oslo, Norway
b
NCC Construction Sverige AB, Gothenburg, Sweden
*
Author for contact. Tel.: +46703578068; E-mail: hanna.nilsson@ncc.se
Abstract
When performing dynamic analyses on steel pipe racks, the method commonly used by the
structural engineer is to apply the static load multiplied by a dynamic amplification factor
(DAF). Thereby, a static analysis is sufficient to account for both static and dynamic behav-
iour in design. However, in lack of a recommended practice for multi degree of freedom
(MDOF) systems such as multi-planar lattice girder structures the DAF is normally obtained
from the theory of a single degree of freedom (SDOF) system. The possible effect of other pa-
rameters than the fundamental eigen period is therefore not taken into account, possibly lead-
ing to either non-conservative or unduly conservative designs.
In the present paper a parameter study of the dynamic behaviour of pipe rack steel structures
subjected to explosion loading has been performed. The pipe rack design is assumed to be a
multi-planar lattice girder consisting of rectangular or square hollow sections. Numerical
analysis with use of the finite element method is performed on a series of altogether 54 pipe
racks, varying parameters such as mass distribution and aspect ratio.
The study provides a deeper understanding of the dynamic response of multi-planar lattice
girder structures and forms the basis for a more accurate prediction. It is found that the re-
sponse of a MDOF system can vary significantly from that of a SDOF system, and to utilize
the maximum DAF from the theory of a SDOF system can lead to non-conservative results on
a MDOF system.
Moreover, it is found that the mass distribution is the studied parameter resulting in the most
significant effects on the dynamic behaviour, and by shifting the centre of mass upwards in a
structure the maximum DAF is reduced. This implies that an increase in mass, e.g. by adding
additional weight to the top of the structure, can reduce the maximum DAF resulting in a
more slender design, i.e. less use of steel and cost savings. However, the study shows that this
elevation of mass only results in a reduced DAF when DAF is calculated based on base shear
(BS); see Fig. 1. For DAF calculated based on over-turning moment (OT) on the other hand,
207
the effect is an increase of DAF. Based on this the conclusion is that calculations of multi-
planar lattice girder structures should conservatively be performed by multiplying the static
loads with a DAF based on OT. Details such as base plates, where the reaction force in the
plate is identical to the base shear, could account for a reduction relative to the difference in
DAF based on OT and BS.
To determine whether the DAF-curve from Biggs [1], which is based on a SDOF system, is
representative for MDOF systems, the results are contradictory. DAF-curves based on BS lie
below or close to the Biggs-curve while the DAF-curves based on OT are both below and
above it. Based on this it can be concluded that the dynamic amplification of structural details
influenced by BS can be designed based on the Biggs-curve while details influenced by OT
required a more time consuming approach.
References
[1] Biggs, JM. “Introduction to structural dynamics”, McGraw-Hill, 1964.
Fig. 1: Results from sensitivity analysis as relationship between maximum DAFs (based on BS
and OT) and eigen periods (T)
208
Tampere, Finland
TALL AMBITIONS – ONSHORE WIND TURBINE TOWER

CONCEPTS FOR LARGE HUB HEIGHTS
Martin Jespersena*, Mogens G. Nielsenb, Ulrik Stottrup-Andersenc,

a,b,c
Wind & Towers, Ramboll Energy, Hannemanns Allé 53, 2300 Copenhagen S, Denmark
Abstract: The production of wind power onshore is in many locations highly dependent on
the height of the turbine above the terrain, so the higher hub height the more power is generat-
ed. A large hub height is a challenge to the industry, as the traditional and popular cylindrical
steel tower is not economic or feasible for hub heights of more than 100m-120m. There is
therefore a focus in the wind power industry to develop new tower concepts that are economic
for large heights, taking into account all the various factors contributing to the total cost of
energy for the implemented turbine. This paper describes some promising developments of
such tower concepts.
1 Introduction
The wind power industry is constantly looking for cost reductions. In some areas the best loca-
tions are already occupied or country geographical features are of a nature, so that new on-
shore wind parks have to be placed at poor sites - for instance in wooden areas with rather
high turbulence and wind shear – therefore the turbine rotor must be placed at larger eleva-
tions to give a proper energy output and lower the turbulent loading.
However the implementation of tall wind turbine towers is expensive and the traditional cy-
lindrical steel towers has many problems onshore: steel consumption increases dramatically
with height, the diameter of the lower part of the tower becomes too large for transportation,
the installation of tower and turbine requires large mobile cranes, foundations becomes very
big and expensive, etc. The industry has been quite active to develop new tower concepts that
can bring down the cost for tall towers.
According to the industry the tall towers only account for approx. 5% of the total capacity
which is being installed annually (of the total capacity installed much less), the focus is there-
fore not only on providing cost reductions on the tower structure, but also a swift and effective
design development process primarily consisting of well-known and proven methodologies
and technology.
Through the experience gathered over 70 years of engagement in analysis, design and con-
struction of tall towers and masts Ramboll have made several different conceptual and de-
tailed designs within tall wind turbine towers. This among others includes a guyed wind tur-
bine tower for multiple MW turbines, innovated singlehandedly by Ramboll.
The presentation will treat the different tower designs in play, analysing their qualities and
challenges from a structural and practical point of view.
*Corresponding author: mtxj@ramboll.com, tel. +45 5161 6085

209
2 Tall onshore wind turbine tower concepts

Onshore wind turbines are faced with a number of challenges before commissioning. The
challenges are no-less as the hub height of the wind turbine is increased.
Challenges related to wind turbine towers intended for large hub heights (>100m) include:
 Substantial increase in material costs
 Transport restrictions – normal road transport limited to 4.2m in height in most countries.
 Crane (height) restrictions/availability/cost
 Access to site with heavy components
 Foundation forces
From the challenges presented here it is clear that
making it in the higher altitudes not only requires en-
gineers with ingenuity and a profound sense for steel
design, but also an awareness and basic understanding
of the entire supply-chain for the wind turbine tower
and the restrictions that this implies.
Looking into the solutions which fit to the challenges
faced, several different designs are in play, each of
them a perfect fit for the conditions and mechanisms,
predominant in the different markets of today.
In general there are 5 dominating types of steel tower
designs to be considered (excluding at least the same
number of designs in other materials such as concrete Fig. 1: The Guyed Wind Turbine
or even timber): tower developed by Ramboll
 Traditional cylindrical towers
 Segmented towers (Cylindrical towers formed from several
radial segments which are bolted together)
 Lattice towers (Traditional angle bar towers)
 Jacket hybrids (Part lattice – Part cylindrical tower)
 Guyed/stayed towers (Typically cylindrical towers support-
ed by guy wires or tubular stays)
Having faced all of the designs the distinctive properties for
each of them may be summarized and compared. Such studies
will typically be inconclusive or yield different conclusions
based on the market for which the towers are benchmarked.
Factors such as material availability, requirements to local con-
tent, labour and local infrastructure are often to be considered. Fig. 2: The K-Jacket
Each type of tower structure constitutes different dynamic sys- tower developed by
tems, each significantly impacting the load generation on the Salzgitter AG and
tower. Forming an iterative process, design concepts must be Ramboll
closely evaluated with the loads generated. Wind turbine tow-
ers are primarily governed by fatigue (ideally a balance between fatigue and ultimate load-
ing). The design of new tower concepts therefore largely entails the evaluation of stress hot-
spots and detail categories in the structural details of the tower, with the constant focus of
keeping it simple, minimizing fabrication and material usage.
Developing new tower concepts requires a constant attention and openness to the new struc-
tural products being developed or reintroduced, enabled by the large market potential, whilst
still scrutinizing specifications and evaluating shortcomings for each of them. With at least
20-25 years of service life with a minimum of inspection (especially of the tower structure)
the towers must be designed and built to last
210
Tampere, Finland
LATERAL STABILITY OF VERANDAS BY MEANS OF THE GLASS

PANELS
M. Fortana, J. De Clercq, M. Meeusb, B. Rossic

a,b,c
Department of engineering technology, University of Leuven, Belgium
Abstract: In recent years, verandas have become increasingly larger and are more often built
separately from the main building. Hence, the stability of windows submitted to horizontal
loading has become a problem. In this paper, a new fabrication concept is tested. It uses sup-
port blocks ensuring a uniformly distributed pressure zone contributing to a redistribution of
the horizontal load in the window’s plane and an increase of the whole system’s stiffness. On
the basis of separate tests on the connection elements, on the support blocks and on the com-
plete frames, the stiffness of a window was evaluated. The possibility of using this concept to
ensure the lateral stability of verandas was then investigated using a finite element model.
1 Introduction
Verandas are an important part of the Belgian building culture. In the past, verandas were rel-
atively small constructions made of three façades and a roof, attached to the main building.
These constructions were made by a limited number of specialized firms, and built on the
base of expertise more than on proper design rules. In the past decade, verandas became in-
creasingly larger and, more importantly, they became constructions independent of the main
building. Therefore, the overall stability against horizontal loading turned out to be a problem.
The size of verandas and the aesthetic aspects imply that new concepts are needed to ensure
the stability without visual changes. Support blocks distribute the weight of the glass onto the
aluminium frame. In this research [1] these wooden or plastic support blocks are placed in
each of the four corners. In this way, horizontal loads can be distributed through a compres-
sion diagonal in the glass panel. Therefore this research concentrates on instability of the glass
panel and the stiffness of the whole system composed of the aluminium frame, the support
blocks and the glass panel. By using a finite element model that is calibrated on results de-
scribed in [2], instability of the glass panel is negligible for the load range of verandas.
2 Test program and numerical analysis

The stiffness of the corners is determined by a test according to NBN EN 514 [3]. This proce-
dure is used for the L-connections and the T-connections, both non-reinforced as reinforced.
The support blocks were also tested separately by applying a concentrated force on the sup-
port block placed on the aluminium profile. The complete frames were tested in different
ways: empty or filled, with T- or L-connections and non-reinforced or reinforced.
211
Fig. 2: L– Fig. 3: T– Fig. 4:

Fig. 1: Overview tests on frames connection connection Profile
To analyse the results, two finite element models were made in SCIA Engineer: the first one
using the theoretical values of the stiffness and strength of the components and the second one
using the measured characteristics. With this model, the effect of different adjustments can be
investigated.
3 Conclusions
Glass panels have a relatively high compression resistance and, hence, the instability (buck-
ling) is the main restriction. However, this research shows that the critical buckling load of the
window panels is greater than the applied loads as a result of the wind.
The infill highly improves the resistance to horizontal loads. However, the stiffness of a
standard frame with a glass panel is rather limited with 84N/mm for T–connections and
101N/mm for L–connections. Therefore, the stiffness can nevertheless be increased by (1)
using several frames next to each other; (2) increasing the stiffness of the connections and/or
the support blocks. It was shown that improving the support blocks compressive behaviour
(by using closed sections) should be the first action. A veranda of 5m x 5m with a height of
2.5m, has to resist a wind load of 1.4kN/m² in Belgium. Therefore, the sidewall is loaded by a
concentrated load of 4.4kN. A maximum displacement of 5 mm has to be ensured to secure
the function of all components like doors and windows. Four non-reinforced T–frames are
clearly not stiff enough, as can be seen in the first row of Table 1. In this table, an overview is
provided with different options to secure the lateral stability of the veranda by using the con-
cept described in this paper.
Table 1: Options to resist 4.4kN with a maximum displacement of 5mm

Frames Size B x L Connection Support block Displacement
(-) (m x m) (kNm/rad) (kN/mm) (mm)
1.1 x 2 12 1.3 12.18
1,1 x 2 20 4 4.81
1,1 x 2 60 5 4.94
1,1 x 2 100 10 4.79
2,2 x 2 100 5 5.01
References
[1] J. De Clercq, M. Fortan; De stabiliteit van veranda’s ten gevolge van de glazen vulele-
menten; Master’s thesis; KU Leuven, Faculty of Engineering Technology; 2014
[2] F. Wellershof; Nutzung der Verglasung zur Aussteifung von Gebäudehüllen; 2006
[3] NBN EN 514: Unplasticized polyvinylchloride profiles for the fabrication of windows
and doors – Determination of the strength of welded corners and T-joints; 2000
212
Tampere, Finland
END PLATE CONNECTION FOR RECTANGULAR HOLLOW SEC-

TION IN BENDING
Arne Aalberga*, Arne M. Uhreb and Per K. Larsena

a
Department of Structural Engineering, Norwegian University of Science and Technology, Norway
b
Department of Structural Engineering, NTNU, and Construction company Gunvald Johansen, Bodø, Norway
*
Corresponding author. E-mail address: arne.aalberg@ntnu.no. Phone number: +47 73594624
Abstract: An extended end plate connection for rectangular hollow sections subjected to
bending moment is examined. The end plate extends on two sides of the RHS and has one
bolt on each side. The connection is commonly considered as ideally pinned for end-rotation
about the weak axis of the connection, i.e., about the axis through the two bolts, while it is
partial-rigid about the other main axis. The objectives of the present study are to obtain data
for both the initial bending stiffness and the capacity of the end plate connection for moment
about each of the axes, and to compare with predictions of the component method of EN
1993-1-8. Classical yield line analysis is used to develop the expressions for the moment re-
sistances. The test results are used for a discussion of the stiffness boundaries for joint classi-
fication for this particular connection.
1 Introduction
a) Investigated detail, RHS 80·80·4 b) End plate geometry, t=8 mm.
Fig. 1: Investigated end plate connection for RHS member.
Rectangular and quadratic hollow sections are widely used in steel frames, trusses and gird-
ers, and are extensively used as columns in buildings. A common design of the connection of
a hollow section to a concrete foundation, or to other structural elements, is to use an extend-
ed end plate welded to the section, and fastening by threaded bars or bolts. The design formu-
las for end plate connections given in EN 1993-1-8 cover typical beam-to-column connec-
213
tions and column bases for I-section members and gives design formulas for predicting both
the connection strength and stiffness. Earlier investigations pointed out complications when
applying the T-stub capacity models for RHS end plate cases, due to bending of the RHS wall
and a shift in the location of the hogging plastic line into the part of the end plate inside the
section.
2 Test program and results
8 6
Sj,ini Sj,ini Yield mechanism

6
Moment (kNm)
Moment (kNm)
4
4
Yield mechanism
2
2
0 0
0.00 0.02 0.04 0.06 0.08 0.10 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
Rotation (rad) Rotation (rad)
Fig. 2: Response curves: Left: Strong axis bending, Right: Weak axis bending
The connection was tested in two configurations; with bending moment about the strong axis
(Specimen A) and weak axis (Specimen B), resulting in the response curves in Fig. 2. The
initial bending stiffness of the specimens was defined from data from several loading-
unloading sequences performed at low load levels until initial plastification. Finally, both
specimens were loaded until plastic bending and large rotation occurred and the connection
was permanently deformed. No physical failures were observed. The deformation consisted of
significant bending of the end plates and a small bending of the bolts. The moment capacities
predicted by classical yield line models are indicated in the graphs in Fig. 2 and are conserva-
tive. The response curve extends significantly higher than the capacity prediction for the
strong axis case, probably due to the larger effect of developed membrane forces for this case.
The obtained values for the initial stiffness (Fig. 2) are Sj,ini=560 kNm/rad and Sj,ini=180
kNm/rad for the strong and weak axis, respectively, which indicate that the present joint
should be considered as semi-rigid according to EN 1993-1-8.
3 Conclusions
1. The investigated RHS end plate connection has a bending moment capacity which may
conservatively be assessed by yield line mechanism analysis, both for the strong axis
and weak axis bending case.
2. The connection behaves as semi-rigid for a bending moment about the strong axis
whereas it is close to pinned for a moment about the weak axis. For the connection to be
considered flexurally rigid for moment about the strong axis, an end plate thickness of
3.5 times the RHS wall thickness is necessary.
3. The definition of appropriate initial connection stiffness for elastic analysis should be
related to the utilization of the moment resistance of the connection, as the response
curve shows nonlinear behaviour even at low load levels.
4. A stiffness model should be developed for the weak axis bending case.
214
Tampere, Finland
STRUCTURAL BEHAVIOUR OF A NOVEL COLUMN - SPLICE JOINT

FINGER CONNECTION
Pedro A P Andrade a, Marko Pavlovic b, Christine Heistermann c, Milan Veljkovic d and Tim
Heistermann e
a, b, c, d,e
Lulea University of Technology, Dept. of Civil, Environmental and Natural Resources Engineering, Swe-
den
b
University of Belgrade, Faculty of Civil Engineering, Serbia
Abstract
The novel joint presented in this paper is a friction connection used for column-splice connec-
tions of modular buildings as part of the innovative construction method introduced in the re-
search project Optimization of frames for effective assembling - FRAMEUP. This type of
joint provides a quick assembly and can deal with misalignments by introducing a connection
gap. A filler and finger plate are welded to the upper part of the column to this end (see Fig.1)
Finger Connection 1 x Upper column

(SHS 250x10)
(Decomposed)
Profiles & plates: S355
Bolts: 10.9 M24
8
1 x Filler plate
(300x192x[4,6,8] generating Connection gap
the Connection gap) (4,6 and 8 mm)
1 x Finger plate
(546x192x14)
5
3x Cover plates
(192x56x6)
Column-splice gap
(5 mm)
9 x Bolt set
(1x bolt, 2x washers
1x nut)
1 x Lower column
(SHS 250x10)
Fig. 1: Detailed description of the Finger Connection
215
The gap between finger plates and lower column faces is closed during tightening of the bolts
and, thus, establishes a slip-resistant connection. The efficiency of the joint resistance based
on different connection gaps subjected to uniform compression is assessed.
The column-splice is composed of four slip-resistant connections, one at each side of the tube.
Each finger plate consists of three long slotted holes and is welded to the upper column face.
Long slotted holes are used to accommodate vertical misalignments and, therefore, allow fit-
ting the bolts which are pre-installed in the lower column. Filler plates with different thick-
nesses (4, 6 and 8 mm) welded between the finger plate and upper column face are used to
create a connection gap which allows balancing horizontal misalignments. The lower column
faces consist of each nine holes with no clearance in order to pre-fit the bolts in a workshop.
Thus, the assembling process on the construction site can be speeded up as once the lower
columns are in place all bolts can be tightened immediately.
Conclusions
The present novel column-splice connection is rather suitable for a quick assembly in-situ as
it adequately can compensate up to 8 mm misalignments in horizontal direction and up to 5
mm in vertical direction. Even bigger misalignments may be allowable but have not yet been
investigated.
The first bolt-row plays a major role in closing the connection gap, i.e. to bend the finger
plate towards the lower column face. This activation force has to be subtracted from the total
preload in the bolts to properly account for the slip resistance. For connection gaps between 4
mm and 8 mm, the activation force component varies from 18 to 45 % of the preloading force
in the two bolts of the first bolt-row. This effect leads to a reduction of the clamping force of
the connection of approximately 4 to 10 % in the considered cases.
Experimental and finite element results have indicated the presence of a second friction sur-
face between the finger plate and the cover plates at the ultimate state. This phenomenon in-
creases the slip-resistance of the connection, approximately twice compared to hand calcula-
tion models used in engineering practice (e.g. according to EN1993-1-8). This increase cannot
directly be accounted for in design due to the large slip that precedes the ultimate load. How-
ever, it constitutes an additional reserve of resistance which contributes to the safety of the
connection.
The novel joint in its full configuration with one finger connection on each column face and
no connection gap has the potential to generate a slip-resistance to an uniform load, within the
serviceability limit states limits, up to 2056 kN. For connection gaps between 4 to 8 mm, the
slip resistance is reduced to approximately 1924 kN to 1824 kN.
Acknowledgments
The paper has been elaborated within the framework of RFCS projects: Optimization of the
frames for effective assembling –”FRAMEUP”– RFS-PR-10121
216
Tampere, Finland
STRUCTURAL ANALYSIS MODELS OF STEEL TRUSSES
Teemu Tiainen, Kristo Mela, Timo Jokinen and Markku Heinisuo
Tampere University of Technology, Department of Civil Engineering, Tampere, Finland
Abstract:
Structural analysis with finite element method can be done generally using solid elements (3D),
shell elements (2D), beam or bar elements (1D) with or without springs (0D). In literature, state-
of-the-art models are typically very complex 3D models model with both material and geomet-
rical non-linearities taken into account. However, for an engineer in practice these methods are
too laborious and more simple linear models are typically used. In truss structures this means
beam or bar element models. Therefore, the focus of this paper is on 1D models used in the
structural analysis of a tubular steel truss. Special emphasis is put on usability of the models
with design code Eurocode 3.
1 Introduction
Structural analysis with the finite element method (FEM) can be done using 3D bricks, 2D shells,
1D beams including 0D springs. The same hierarchy holds for steel trusses. Despite their wide
use in research (among others [1, 2, 3]), 2D shell and 3D brick models are not typically used by
engineers in practice in truss design. In practice, design codes, like Eurocode 3, are used with
structural analysis done with 1D finite element model. Since a tubular truss can be modeled in
many different ways [4] with varying results it is not clear what kind of model should be used.
Therefore, in this paper five different 1D models of which one is complemented with 0D springs
are compared both in result accuracy aspects as well as the Eurocode 3 requirement aspects.
2 Truss analysis models

Models compared in the paper can be seen in Fig. 1. The variety goes from classical pin-jointed
model (Model 1) to exact geometry with real-life rotational stiffnesses in joints (Model 5). With
these models, a typical tubular steel truss with K gap joints were analyzed.
It was found that the axial forces are quite the same regardless of the model used even if geom-
etry is altered by omitting eccentricity (Models 1 and 2) or when taking into account the exact
geometry (models 3-5). Bending moment values can also be considered consistent between the
models where they are available.
217
Model 1 - Truss Model 2 - Chord as Model 3 - Chord as

elements continuous beam continuous beam,
rigid eccentricity
element
Model 4 - Rigid Model 5 - Model 4

eccentricity elements with rotational
for each brace springs
Fig. 1: Models considered.
When applying Eurocode design procedure for tubular trusses, several member and joint design
formulas are used. In member design, the bending moment diagram is needed and in joint design
forces at gap area are needed. Thus, some Eurocode formulas require internal forces that cannot
be obtained from all of the used models.
3 Conclusions
The compared models give fairly consistent results when assessing axial forces. The difference
becomes apparent when analysis results are used to carry out the design code calculations. It
seems that only Model 5 of compared models gives all the data that is required by the Eurocode
3 design procedure. Also, considerable additional bending due to eccentricity may occur even
if the Eurocode limit is not exceeded.
Acknowledgements
Financial support of FIMECC and Tekes are gratefully acknowledged.
References
[1] Al-Jabri K, Burgess I, Plank R. ”Spring-stiffness model for flexible end-plate bare-steel
joints in fire”. Journal of Constructional Steel Research, 61, 1672–1691, 2005.
[2] Perttola H, Heinisuo M. ”3D component method for base bolt joint”, Steel Structures:
Culture and Sustainability, Turkish Constructional Steelwork Association (TUSCA) (Eds.:
N. Yardimci, A. Aydöner, H. Gures, C. Yorgun), Istanbul, Turkey, 355–365, 2010.
[3] Pasternak H, Krausche T, Launert B. ”Schweißen von Trägern mit dicken Blechen, Teil 1:
Trägerfertigung unter Werksbedingungen – Planung, Herstellung und Simulation”. Bauin-
genieur, 89, 1–11, 2014.
[4] Boel H. ”Buckling length factors of hollow section members in lattice girders”. Master’s
thesis, Eindhoven university of technology, 2010.
218
Tampere, Finland
BUCKLING OF MEMBERS OF WELDED TUBULAR TRUSS
Markku Heinisuoa,* and Äli Haakanaa

a
Abstract: The scope of the paper is tubular trusses with welded gap joints and flexural buck-
ling of the members in the plane of the truss. Flexural buckling of the member is dependent
on the rotational stiffness of the joints between braces and chords. The stiffness is calculated
using non-linear 3D FEM models and the models are validated with experiments. The fillet
welds at the joints have considerable effects to the stiffness, even factor 2, especially when
using high strength steel with full strength welds. The results imply a potential to reduce the
total costs of the tubular trusses.
1 Introduction
Tubular welded trusses are widely used in constructions due to their appearance and econom-
ics. In the structural design a very important design criteria is buckling resistance of the com-
pressed members. In present design rules, such as Eurocodes, rather rough rules are given to
define the buckling lengths of the members, leading often to conservative and in some cases
to unsafe design [1]. The scope of the paper is flexural buckling of the members in the plane
of tubular triangulated trusses with welded gap joints. The flexural buckling of a member is
dependent on the rotational stiffness of the joints between braces and chords. It is supposed
that the truss will not buckle as a whole, but the critical modes are the buckling modes of the
members, and no interaction between buckling of members occurs.
2 Proposed approach
In [1] and [10] simple equations are given to define the buckling length factors for square hol-
low section (where both chords and braces are square hollow sections, SHS) members:
L
Chord : K  cr  1.25  0.6   0.9 (1)
Lsys
0.14
L  b 
Brace : K  cr  1.05  0.025   1   1.00 (2)
Lsys  
 Lbr , sys

where bi is the width of the brace (i = 1,2) or chord (i = 0) and using the notations of EN
1993-1-8 for square hollow sections:
219
b1  b2
 (3)
2b0
b
  0 (4)
2t 0
where t0 is the thickness of the chord. In this research the use of Eq. (1) and (2) is enlarged for
the trusses with different adjacent braces and for different chord sizes. The solution strategy
in this study was to
 Calculate the initial rotational stiffnesses of the joints using comprehensive 3D FEM
models with combinations of un-identical braces and combinations of un-identical
chords. Also, the fillet weld sizes were varied to observe the effect of the weld size to
the stiffness. It was supposed that the same weld was used all around the joint.;
 Validate the FEM model with experimental results available for HSS joints completed
in the RFCS project RUOSTE;
 Verify the rotations of the FEM model with the results of [1];
 Calculate the buckling factor K using Newmark’s equation;
 Compare the results of the Eq. (1) and (2) to see if they are safe for these new cases.
It can be concluded that:
 In present Eurocodes in buckling analysis the system length has not been defined ex-
actly. Buckling factor 0.75 has shown to be unsafe in some cases.
 In [1] and [10] are given simple and safe equations to define both in-plane and out-of-
plane buckling length factors for the members of triangulated trusses made with weld-
ed gap K-joints, including both rectangular and circular hollow sections. The equa-
tions were shown to be safe also in the cases which were studied in this study.
 When using full strength fillet welds the rotational stiffness increased a lot, compared
to cases without welds, or full penetration butt welds.
 The system lengths were defined exactly based on the local analysis model Variant 1,
which enables the use of semi-rigid joint models in the future. In many practical cases
the joints can be classified as rigid, based on EN 1993-1-8.
 This approach is safe and means in many cases the economical design.
The same approach will be used for other joint layouts, e.g. that shown in Fig. 1. In order to
use the method of semi-rigid joints in tubular truss optimization special meta-models will be
developed to define the rotational stiffness of the joints with Variant 1. Using rotational stiff-
ness at the joint then may develop interaction between buckling modes of adjacent members.
This is open question for tubular trusses.
References
[1] Boel H. “Buckling Length Factors of Hollow Section Members in Lattice Girders”, Mas-
ter’s Thesis, Eindhoven University of Technology, 2010.
[10] Snijder H, Boel J, Hoenderkamp J, Spoorenberg R. “Buckling length factors for welded
lattice girders with hollow section braces and chords”, Proceedings of Eurosteel, Budabest,
1881-1886. 2011.
220
Tampere, Finland
BENDABILITY AND MICROSTRUCTURE OF OPTIM® 700 MC PLUS
Vili Kestia,*, Antti Kaijalainenb, Juho Mourujärvib and Raimo Ruoppac

a
SSAB, Finland
b
University of Oulu, Centre for Advanced Steels Research, Finland
c
Lapland University of Applied Sciences, Finland
*
Tel.: +358503143694; E-mail: vili.kesti@ssab.com
Abstract: The use of ultra-high-strength steels (UHSS) in weight-critical constructions is an

effective way to save energy and minimize the carbon footprint of the application. At the
same time, the demands for structural design and reducing part manufacturing costs are in-
creasing. SSAB has employed the thermomechanical rolling and accelerated cooling process
(TM + ACC) to make a novel type of environmentally friendly ultra-high-strength strip steel
for structural applications. Chemical composition of SSAB`s Optim® 700 MC Plus strip steel
is shown in Table 1. The low carbon content guarantees high impact toughness after TM +
ACC leading to ferrite/ bainite microstructure thereby eliminating the need for tempering.
Low carbon content also ensures good weldability.
Table 1: Typical chemical composition and mechanical properties of Optim® 700 MC Plus.
C, Si, Mn, P, S, Al, CEV, YS, TS, El, CV -60°C,
max max max max max min typical (MPa) (MPa) (%) (J/cm2)
0.10 0.25 2.1 0.02 0.01 0.015 0.37 700 750-950 13 40
* In addition, niobium (Nb), vanadium (V), titanium (Ti), boron (B), molybdenum (Mo), nickel (Ni) or copper
(Cu) may be used as alloying elements either singly or in combination.
The advantages of ultra-high strengths (>700MPa) can only be fully realized if the technolog-
ical properties of the steel are sufficiently good. Bending is one of the most important work-
shop processes and a good bendability is essential for a structural steel. Small bending radii
are used to save space in steel structures and also to make constructions stiffer. It has been
found that special attention has to be paid to bending process when bending UHSS due to
their more limited deformability.
In this paper, the metallurgy and bendability of the proprietary strip steel Optim® 700 MC
Plus have been closely investigated. On the basis of SEM analysis and IQ-mapping, the mi-
crostructure of 3 mm sheet consisted mainly of granular bainite (~73 %) together with qua-
sipolygonal ferrite (~14 %), upper bainite (10%) and MA islands (3%). The corresponding
microstructure of 6 mm sheet consisted mainly of granular bainite (~60 %) together with qua-
sipolygonal ferrite (~18%), upper bainite (~17%) and MA islands (~5%). Both sheets consist
of mainly granular bainite and quasipolygonal ferrite without upper bainite at the ¼ depth.
The effective grain size (deff) according to EBSD studies was found to be ~2 µm.
221
The bendability of Optim® 700 MC Plus was found to be superior to conventional UHS steels
when minimum bending radiuses were investigated. The minimum bending radius is the
smallest radius that leads to bend surface without any signs of significant necking or cracking.
Smaller radii than minimum bending radius shouldn´t be used, as that can lead to microcrack-
ing and for example loss of fatigue resistance.
It was found that along rolling direction (RD) R=1t and transverse direction (TD) R=0.3t can
be achieved without any defects on bend surface up to 10 mm sheet thicknesses. When the
sheet thickness is over 10 mm, minimum bending radius to RD and TD are 1.5t and 0.5t, re-
spectively. The minimum bending radii of Optim® 700 MC are much better than the ISO EN
10149-2 standard requirements for 700MPa thermomechanically hot-rolled steels and con-
form even to requirements of S420MC despite the fact that Optim® 700MC Plus has smaller
total elongation A (min. 13%) than S420MC steels (min. 19%). According to Yamazaki et al.
bendability has no correlation with total elongation, but it is closely related to local elonga-
tion. Further studies have been made to understand the ability to withstand local strains during
bending. It was found that Optim® 700 MC Plus tolerates local strains up to 60 % before bend
surface shows signs of defects. Therefore, it can be concluded that even though the total elon-
gation to fracture in a tensile test is not very high, the studied steel exhibits an excellent abil-
ity to withstand local strains without cracking.
Springback, bending force together with strain hardening and thinning of the bend area were
investigated and new equation for bending force estimation is given. It was noticed that if too
small bending radiuses are used, some excessive thinning may occur. On the other hand, dur-
ing bending the surface area is strain hardening 10-20%. However, special attention should be
paid to obey the recommendations for minimum bending radiuses to prevent excessive thin-
ning and surface defects.
Attention shoud be also paid to material handling before bending as any scratches or dents
may localize strains on tensile surface or edges. Optim® 700 MC Plus is designed especially
for mechanical cutting. Traditionally S700 MC steels have poor mechanical cutting properties
as the quality of cut edge tends to be rough and also split. Therefore, it is recommended to
mechanically grind cut edges before bending to prevent edge cracking. However, Optim® 700
MC Plus has really good edge quality after shearing and it can be bent without grinding. Due
to excellent quality of sheared edges Optim® 700 MC Plus can be used in a very economical
way in the workshop as it does not require extra finishing before bending.
The impact toughness of Optim® 700 MC Plus is excellent as the minimum impact strength at
-60°C is 40 J/cm2, typically being over 100 J/cm2. The minimum impact strength is also guar-
anteed in the transverse direction. T28J temperatures were determined from these results using
Wallin’s formula. Ductile-brittle transition temperatures (T28J) are -110°C (longitudinal) and -
90°C (transverse). Excellent T28J values can probably be attributed to the refinement of the
coarsest grains (i.e. d90%): in the case of ferritic microstructures, at least, it has been shown
that the size of the largest grains rather than the mean grain size controls cleavage crack prop-
agation.
It was found that by optimizing chemical composition and process parameters, an excellent
combination of strength and ductility, in the thickness range from 3mm to 12mm, can be
achieved by control of the ferritic-bainitic microstructure. These properties make Optim® 700
MC Plus suitable for demanding structural applications such as lifting and transporting prod-
ucts.
222
Tampere, Finland
EXPERIMENTAL BEHAVIOUR OF TENSION PLATES WITH

CENTRE HOLE MADE FROM HIGH STRENGTH STEEL
Pál Turána, László Horváthb

a,b
Department of Structural Engineering, BME, Hungary
Abstract
High strength structural steel elements are becoming more widely available and are gaining
increasing use in practice. The latest standard which deals with the high strength steels (HSS)
is the EN 1993-1-12 [1]. This standard includes the steel grades up to S700 but in the practice
S1100 already been used. In this paper S700 and S960 steel grade are discussed. At this
moment the net section resistance design formulas for high strength steel structural elements
are the same as for the common used steel grades in EN 1993-1-1 [2].
According to the present EN 1993-1-12 for sections with holes the design tension resistance
(Nt,Rd) should be taken as the smallest of:
• the plastic design resistance of the gross-section
A⋅ fy
N pl , Rd =
γM0 (1)
• the ultimate design resistance of the net cross-section area at holes
0.9 Anet ⋅ f u
N u , Rd = (2)
γM2
Before the present EN 1993-1-12 was published, in the last Working Group draft prEN 1993-
1-12 (stage 49), the design resistance of the net section was determined according to plastic
resistance:
A ⋅f
Nt , Rd = net y (3)
γM0
In this report, as a new proposal the 0.9 reduction factor can be omitted, which is in
accordance with [3]:
A ⋅f
N u , Rd = net u (4)
γM2
The authors have conducted a large scale experiment program with centrally holed tension
specimens (wide plate). Two different steel grades from two different steel producers have
been investigated in the frame of the ongoing Rules On High Strength Steel (RUOSTE)
project [4], overall four different plate materials were used. Steel grades with nominal yield
strength 700 MPa and 960 MPa were used in the tests. The nominal plate thickness was t = 8
mm and the nominal plate width was b = 80 mm in every cases, the hole diameter d0 varied
from 8 to 40 mm in 8 mm steps.
223
Statistical evaluations have been conducted using the published and own experimental results
according to EN 1990 Annex D [7], a “best fit” model is shown in Fig 1.
Fig. 1: Statistical evaluation with Eq. (3)
3 Conclusions
The main conclusions are for the studied high-strength steel grades:
1. the plastic resistance of the gross section is not a governing failure mode, when hole-
weakening is used;
2. the design formula belongs to the failure of net section even the 0.9 factor is omitted
(Eq. (4)) is applicable with the standard value of γM2=1.25;
3. the design formula belongs to the failure of net section (Eq. (2)) is applicable but more
conservative than Eq. (4) for the high strength steel grades;
4. using the statistical evaluation according [8] cannot be confirmed that the formula for
the net section yielding (Eq. (3)) in the investigated cases is applicable with the standard
value of γM0=1.0 even tough when all results are on the safe side.
Acknowledgments
The research presented in this paper is based on the results of the RFCS-project RUOSTE.
The financial support of the RFCS is thereby gratefully acknowledged.
References
[1] EN 1993 Part 1-12: “Eurocode 3: Design of steel structures – Part 1-12: Additional
rules for the extension of EN 1993 up to steel grades S700”, 2007.
[2] EN 1993 Part 1-1: “Eurocode 3: Design of steel structures – Part 1-1: General rules and
rules for buildings”, 2005.
[3] Može, P., Beg, D., Lopatič, J. “Net cross-section design resistance and local ductility of
elements made of high strength steel”, Journal of Constructional Steel Research, Vol.
63, No. 11, 1431-1441, 2007.
[4] RUOSTE: Rules On High-Strength Steel; RFCS Project RFSR-CT-2012-00036
[5] EN 1990: “Eurocode – Basis of structural design”, 2002
224
Tampere, Finland
DERIVATION OF STRAIN REQUIREMENTS FOR HIGH STRENGTH STEEL USING

JOHNSON COOK MODEL
Simon Schaffrath*, Nicole Schillo and Markus Feldmann
Institute of Steel Structures, RWTH Aachen University, Germany

*
Tel.: +49 (0) 241 - 80 26243; E-mail: s.schaffrath@stb.rwth-aachen.de
Abstract: Despite the high potential for structural applications, the use of high strength steel is
still restricted due to code requirements, which to some extent not scientifically derived. Espe-
cially the strain requirements of EC3-1-12 are not reasonable in respect to actual steel proper-
ties. Therefore, within the scope of the RFCS-funded research project RUOSTE, wide plate
tests with different grade of damage were conducted on steel material S700MC and S960MC
(both meeting the requirements of EN 10149-2:2013). The results are evaluated within this
paper. Using geometric and material nonlinear analysis (GMNIA), a material resistance model
is derived which reflects the actual behaviour of wide plates under tension: the model allows
for a prediction of crack-initiation and the post-failure behaviour. In consequence, it allows for
the determination of realistic strain requirements based on the actual 3D-stress state. The results
show that the current strain requirements in EC3-1-12 are too conservative, and thus lower
values would be more appropriate. The application of realistic strain requirements would lead
to easier application of S700 steel and might be also promising to open the code for S960 grades.
1. Introduction
1.1. Ductility requirements in EC
The ultimate limit state design of EN 1993-1-1 [1] is based on engineering models, referring to
resistance models which are relate either to the yield strength fy or to the tensile strength fu.
These models require a certain ductility of the material to allow for the distribution of stresses
to develop full plastic capacity and overcome notch effects. EN 1993-1-1 [1] and EN 1993-1-
12 [2] assume the material to have these abilities, if three requirements are fulfilled: they con-
cern the yield to tensile strength ratio, the uniform elongation and elongation at fracture values
and shall ensure sufficient plastic deformation capacity. However, these values are derived from
simple coupon tests, and they do not reflect the realistic stress-strain state of other geometries
[3].
1.2. Net section resistance according to EC

EN 1993-1-1 [1] as well as EN 1993-1-12 [2], reduce the net section resistance to 90 % by
applying
0.9  Anet  fu
Nt , Rd  (1)
 M 2 / M 12
The factor 0.9 is based on fracture mechanics [4] and rather old tests. Previous research on
modern steels suggest that the 0.9 factor could be omitted [5]. However, these tests were done
225
on center holed test (CHT) specimens, while the 0.9-factor is derived also from notched tests,
which react more unfavourable due to the sharp notch.
2. Conclusions
Current requirements regarding S500 up to S700 steel grades are based rather on engineering
judgement than on scientific properties. This paper aims to show that although the code require-
ments might not be met by a certain steel, it is still able to reach full net section resistance.
A numerical, parametric study was conducted on wide plate tests, varying the material law of
two high strength steels and the amount of damage (i.e. hole diameter). The resulting load-dis-
S700 in Fig. 1.
placement curves are shown S960
600
d0mm : Anet *fu d0mm : Anet *fu
550 700
500
600
Force [kN]
Force [kN]
450 d16mm : Anet *fu d16mm : Anet *fu
400 4.4 % 500 3.04 %

4.6 % 2.94 %
350 d32mm : Anet *fu 4.5 % d32mm : Anet *fu 2.63 %
Fu 400 Fu
300
crack ini crack ini
250 300
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Displacement [mm] Displacement [mm]
Fig. 1: load-displacement curves with given parameters, S700 (left) and S960 (right)
It can be concluded, that:
1. The ultimate load of members in tension are depending on their failure mode, which
again is depending on material toughness and geometry properties. These properties
lead either to plastic instability in tension or material failure. While plastic instability
can be assessed directly by finite-element calculation, a Johnson-Cook based damage
mechanic model was applied to consider material failure in the numerical calculations.
2. The applied damage mechanic model showed very good compliance with the experi-
mental tests. But the results are sensitive to mesh properties and damage parameters.
3. Although the requirement with a min. value for εu and the fu/fy ratio indicate a necessary
hardening, the true hardening behaviour of the stress-strain-curve is still much more
important. Depending on the function of hardening, lower εu values than required by [2]
could lead to higher ultimate loads, fu respectively.
4. Experimental and numerical investigation suggest that omitting 0.9 in the net section
resistance calculation could be possible, provided a sufficient material toughness and no
sharp notches in the member.
References
[1] EN 1993 Part 1‐1: “Eurocode 3: Design of steel structures – Part 1‐1: General rules and
rules for buildings”, 2010 (DIN).
[2] EN 1993 Part 1‐12: “Eurocode 3: Design of steel structures – Part 1‐12: Additional
rules for the extension of EN 1993 up to steel grades S700”, 2010 (DIN).
[3] JRC Scientific and Technical Report: “Choice of steel material for the plastic design of
steel frames including seismic resistant structures”, unpublished.
[4] JRC Scientific and Technical Report: “Commentary and worked examples to EN 1993-
1-10 and other toughness oriented rules in EN 1993”, 2008.
[5] Može, P., Beg, D., Lopatič: “Net cross-section design resistance and local ductility of
elements made of high strength steel”, Journal of Constructional Steel Research, Vol.
63, pp. 1431-1441, 2007.
226
Tampere, Finland
BUCKLING STRENGTH OF HSS BEAMS
Mark A. Bradford & Huiyong Ban
Centre for Infrastructure Engineering and Safety, School of Civil and Environmental Engineering
UNSW Australia, UNSW Sydney, NSW 2052, Australia
High strength steel (HSS), being defined herein as having a yield strength greater than 450
N/mm2, is finding increased use in practice, and as a result HSS has been included in tradi-
tional “mild steel” design standards such as Eurocode 3 and the Australian AS4100 [1]. This
is as a result of advanced metallurgical technologies being able to produce HSS with reliable
performance and of welding technologies that allow for fabricated or built-up sections.
When used as a flexural member, a HSS element may reach a limit state governed by flexural-
torsional buckling. Like a conventional mild steel beam which may also fail by flexural-
torsional buckling, the limit state is an interaction between elastic buckling and yielding, but
unlike a mild steel beam, the yielding aspect is considerably different when HSS is used. This
is because the condition of a fully-yielded cross-section is not obtainable because of the lower
ductility of HSS and of its lack of a defined yield plateau, and also because residual stresses
are an important consideration. Nevertheless, the concept of a “beam curve” [2] is attractive
for engineers, with a prescriptive equation relating the bending capacity to the elastic buckling
stress (or moment) and the yield stress (or plastic moment).
This paper considers the flexural-torsional buckling of tapered HSS beams using ABAQUS
modelling, which has been shown elsewhere to be robust, accurate and efficient. Such a
member is shown in Fig. 1 and in the context of a half-through girder beam, its tension flange
is restrained elastically, as shown. This means that the web does not remain straight during
buckling, with Fig. 2 showing the buckling modes generated by ABAQUS.
Fig. 1: Buckling model
227
(a) (b)
Fig. 2: Typical elastic buckling eigenmodes with α = 0⋅5, (a) αz < 106; (b) αz > 107
0.1 0.1
0.08 qu/qp 0.08 qu/qp

0.06 numerical 0.06 numerical
result result
0.04 0.04
0⋅32/λ 0⋅32/λ
0.02 0.02
α = 0⋅5 α = 1⋅0
0 0
0 2 4 6 8 0 2 4 6 8
Modified slenderness λ = √(qp/qcr)
Fig. 3: Buckling strength curves
The effects of residual stresses are also included in the formulation. The beam curve concept
is addressed and it is shown that the familiar strength curve needs some modification, with
typical results being shown in Fig. 3. The form of this modification is quantified to as to pro-
duce much-needed design guidance for this rapidly evolving application of a new generation
of cost-efficient and low-weight steels. The beam curve proposed has the form
M b = 0 ⋅ 32 M o M p ≤ M p , (1)
in which Mb is the bending strength, Mp the fully plastic moment and Mo the elastic buckling
moment. It is shown further that except for highly tapered members, the buckling strength is
independent of the taper ratio and so Mo may be determined for a prismatic beam.
References
[1] Ban HY, Bradford, MA. “Flexural behaviour of composite beams with high strength
steel”, Engineering Structures, 56, 1130-1141, 2013.
[2] Trahair NS, Bradford MA, Nethercot DA, Gardner L. The behaviour and design of steel
structures to EC3, Fourth Edition, London: Taylor & Francis, 2008.
228
Tampere, Finland
TRUE STRESS-STRAIN RELATIONSHIP

FOR FINITE ELEMENT SIMULATIONS
OF STRUCTURAL DETAILS UNDER DIFFUSE NECKING
Petr Hradila and Asko Taljab

a,b
VTT Technical Research Centre of Finland
Abstract: The paper presents an automated numerical method for acquiring true stress-strain
relationship from the material test results of high-strength steels. The model beyond uniform
load is iterated to produce load-displacement relationship matching the experimental results re-
calculated by finite element method. We have used this approach to evaluate coupon tests of
high-strength grade S960. The results were validated against the tensile experiments of plates
with central hole.
1 Introduction
Steel structures are generally designed to the level of yield or ultimate strength of the material
in the cross-section. However, certain structural details tolerate relatively high strains in the
localized areas where the instability in tension called necking may occur. For instance the net
section resistance in tension can involve diffuse necking near the drilled holes. Such details are
critical especially for high strength steels where the ductile failure happens at relatively low de-
formations. To simulate such cases with finite element method (FEM), one needs the definition
of material plasticity in terms of true stress and true plastic strain relation also in the range of
necking. The material model would be most preferably obtained from the standardized coupon
tests.
2 True stress-strain curves for Abaqus models

Our iterative approach for the true stress-strain characterization of measured tensile test data
uses method introduced by ManSoo et al. [1] adapted to elastic-plastic materials with strain
hardening and fully automated to convert raw data from coupon testing to true stress-strain
curve readable by FEM solver. We have analysed test results from 16 coupon tests from S960
and the average material model was used to predict the load-displacement of centre hole ten-
sion (CHT) tests. The numerical results were compared to the real experiments from the same
material. As can be seen from Fig. 1 the predicted load-displacement is very close to the exper-
imental results. Moreover, we have estimated the failure of such specimens using SMCS failure
model [2] that was further simplified with the assumption that the highest possible triaxiality
would be 1 in the case of CHT details. The predicted failure occurs when the maximum
229
equivalent strain reaches the critical value. This simple approach is very convenient for quick
estimation of the ductile capacity of the studied details.
Fig. 7: Predicted load and displacement (thick lines) and measured values (thin lines) for CHT spec-
imens with holes 8 to 40 mm and predicted values base on first yield load fyAnet,
ultimate load Pmax and SMCS critical strain
4 Conclusions
The main conclusions are: (1) Developed procedure for assessment of true stress-strain rela-
tion in Abaqus is powerful tool to effectively obtain good quality material model for simula-
tions of structural details with large strains. (2) The resulting models are intended to be used
with the same finite element software and preferably the same meshing parameters. (3) Accu-
rate prediction of ductile failure in details with high stress concentration is usually beyond the
knowledge of common designer, but some estimation can be obtained using simple limits for
equivalent plastic strains.
Acknowledgments
The research leading to these results has received funding from Finnish Metals and Engineering
Competence Cluster’s (FIMECC) program BSA - Breakthrough steels and applications
(2014-2018) and its project Design beyond present codes – enabling efficient utilisation of new
materials. We would like to thank Ruukki Construction Oy for the experimental test results of
coupons and CHT specimens.
References
[1] ManSoo J., Jea G.E., Min C.L. “A new method for acquiring true stress–strain curves
over a large range of strains using a tensile test and finite element method”, Mechanics
of Materials, 40, 586–593, 2008.
[2] Myers A.T., Deierlaine G.G, Kanvinde A. Testing and probabilistic simulation of duc-
tile fracture initiation in structural steel components and weldments, Report No. 170,
The John A. Blume Earthquake Engineering Center at Stanford University, 2009.
230
Tampere, Finland
CALIBRATION OF THE DUCTILE DAMAGE MATERIAL MODEL

PARAMETERS FOR A HIGH STRENGTH STEEL
Marko Pavlovica, Panagiotis Manoleasb, Milan Veljkovicc e Efthymios Koltsakisd

a,b,c,d
Luleå University of Technology, Sweden
a
University of Belgrade, Serbia (permanent position) ; mark.pavel@gmail.com; +381641471748
Abstract: The on-going RUOSTE project aims to improve understanding of HSS by means of
tests and FEA, addressing issues of ductility and stability of structures made of HSS. Various
material models used in FEA are verified by tests. This paper presents calibration and verifica-
tion of ductile damage material model in Abaqus FE software package referring to series of
tensile test experiments on coupons and plate specimens with a single circular hole. Nominal
steel grades S700MC and S960Q are used. Damage initiation criterion and evolution law are
derived analysing localization of plasticity by coupon FEA. Quasi-static analysis using explic-
it dynamic solver is chosen in order to create the most realistic FEA of the specimens.
1 Introduction
Development of the steel production leads towards higher strength steels with grades higher
than S1100 in order to improve competitiveness of primarily mobile cranes. Implementation
of such steel is rather rare in construction sector. However, it is noticed that the design rules
given by Eurocodes are used even for other application than in construction, because of a lack
of better codes. EN 1993-1-1 [1] covers steel grades up to S460 and rules for higher steel
grades, up to S700, are given in EN 1993-1-12 [2]. Most of rules for high strength steel (HSS)
are adjusted from mild steel using the similar set of structural requirements
2 Ductile damage material model

In RUOSTE project HSS behaviour is investigated by comprehensive testing program and
FEA. The first step towards this goal is to verify tests by various material models. A material
model used in this paper is based on failure criterion defined under tri-axial stress state. The
main aim of the work is to validate the model before it is used for extensive numerical calcu-
lations of bolted connections. An easy to use and practical calibration procedure for ductile
damage material model in Abaqus, developed by Pavlovic et al. [3], is based only on results of
a coupon test. Parameters of ductile damage initiation criterion and damage evolution law are
derived analysing undamaged and damaged material response in the coupon test taking into
account localization of plasticity and assuming uniaxial stress state.
231
a) experiments (all specimens) b) FEA (CHT-960-0 and CHT-960-16)

Fig. 1: Comparison of experimental and FEA fractured modes of the plate specimens, S960Q.
3 Conclusions
Through this study advanced damage plasticity material model is successfully used to validate
results of tensile experiments using plated HSS specimens, it has been shown that:
1. Advanced damage plasticity material model and the presented calibration pro-
cedure can be successfully used for HSS specimens with relatively lower ultimate-to-yield
strength ratios. Results for ultimate forces obtained in FEA are up to 5 % higher compared to
experimental results. Displacements at fracture are predicted in FEA with up to 8 % accuracy.
2. Since the explicit dynamic solver is used, failure analysis is possible to be con-
ducted for any other geometry of the structure without having problems with extreme nonline-
ar behaviour and convergence of the results. For the failure analysis of a structure using this
material model, special attention must be given to the size and type of finite elements in the
model, since the damage evolution law depends on it.
3. Arbitrary mesh geometry is recommended for various failure analyses using the
presented damage material model in order to avoid the regular crack pattern driven by the
boundaries of the finite elements.
References
[1] EN-1993-1-1. Design of steel structures, Part 1-1: General rules and rules for buildings,
Brussels, Belgium: European Committee for Standardization, 2007.
[2] EN-1993-1-12. Design of steel structures, Part 1-12: Additional rules for the extension
of EN 1993 up to steel grades S700, Brussels, Belgium: European Committee for
Standardization, 2010.
[3] Pavlović M, Marković Z, Veljković M, Buđevac D. Bolted shear connectors vs. headed
studs behaviour in push-out tests. Journal of Constructional Steel Research
2013;88:134-149.
232
Tampere, Finland
BUCKLING OBSERVATION OF DOOR OPENINGS

FOR WIND TURBINE TOWERS
Anh Tuan Trana, Milan Veljkovica, Carlos Rebelob, Luis Simões da Silvab
a
Luleå University of Technology, Sweden
anhtuan.tran@ltu.se, milan.veljkovic@ltu.se
b
University of Coimbra, ISISE, Portugal
crebelo@dec.uc.pt, luisss@dec.uc.pt
Keywords: Buckling, FEA, Wind turbine tower, Door opening, High strength steel.
Abstract
Local buckling of the door opening is a part of wind turbine steel tower design verification.
Majority of FE studies of the door opening have been carried out without experimental valida-
tion of the results. This is not surprising because just few down-scale experiments have been
carried out. Therefore, an experimental program was established, within the HISTWIN 2 re-
search project, and the resistance of the door opening has been obtained. Compression tests,
using circular and polygonal down scale specimens with and without openings, have been
performed in the approximately scale 1:10. In addition to inductive sensors a part of the spec-
imens around the opening has been monitored by optical surface strain measurement system
to obtain a complete shape of deformation Fig 1. A software package has been used to com-
pare and analyse the images in order to gain insight into the local buckling development. The
captured regions have been prepared by painting contrast colours before the tests.
Fig. 1: Setup compression test with Aramis system.
233
Conclusions
The openings significantly influence on the critical loads. The critical loads of the circular
models and polygonal models are considerably decreased by 35.7% and 20.1% respectively
due to the presence of the opening. Critical loads of the circular models are higher than in the
polygonal models. The critical load of the circular model without opening is 26.7% higher
than the polygonal model without opening. Local buckling of door opening segment was suc-
cessfully recorded in the experiment. Good agreement between compression experiments
monitored by the system and FE results are obtained. Result of the buckling development cal-
ibrated to load – displacement curve is presented in Fig 2.
Fig. 2: Buckling development around opening.
Acknowledgment
The authors wish to thank the Research Fund for Coal and Steel for financially supporting the
research in this paper through the Research Project HISTWIN 2.
References
[1] A.T. Tran, M. Veljkovic, C. Rebelo, L. Simões da Silva, Resistance of door openings in towers
for wind turbines, 3rd South-East European Conference on Computational Mechanics an
ECCOMAS and IACM Special Interest Conference, Kos Island, Greece, 2013.
[2] C.A. Dimopoulos, C.J. Gantes, Experimental investigation of buckling of wind turbine tower
cylindrical shells with opening and stiffening under bending, Thin-Walled Structures, vol. 54, pp.
140-155, 2012.
[3] J.F. Jullien, A. Limam, Effects of openings of the buckling of cylindrical shells subjected to axial
compression, Thin-Walled Structures, vol. 31, p. 187–202, 1998.
[4] Tran A.T, Veljkovic M, Rebelo C, L. Simões da Silva. Influence of geometrical imperfections on
analyses of door openings in tubular steel towers for wind turbines. Proceeding of 7th European
conference on Steel and Composite Structures, Napoli, Italy, 10-12 September 2014.
[5] Abaqus 6.12-1, Simulia Dassault Systmes, 2012.
[6] GOM Inspect V7.5, GOM mbH, 2013.
[7] ARAMIS software version 6.3, GOM mbH, 2013.
234
Tampere, Finland
Extension of the Continuous Strength Method to the Determination of

Shear Resistance
Najib Salibaa,* and Leroy Gardnerb

a
University of Balamand, Lebanon
b
Imperial College London, UK
*
Author for contact. Tel.: 009613234591; E-mail: najib.saliba@balamand.edu.lb
Extended abstract
The continuous strength method (CSM) [1] is a recently developed deformation-based design
method for metallic structures. In this method, cross-section classification is replaced by a
normalised deformation capacity, which defines the maximum strain that a cross-section can
endure. This limiting strain is used in conjunction with an elastic, linear-hardening material
stress-strain model to determine cross-section capacity. To date, the CSM has been developed
for cross-section capacity under normal stresses (i.e. compression, bending and combined
loading), where it has been shown to offer more accurate predictions than current codified
methods [1-6]. In this paper, extension of the method to the determination of shear resistance
is described. An explanation of its development, the initial assumptions made and a summary
of the underpinning test and FE data [7, 8] are first provided.
The relationship between the normalized shear deformation capacity of the web γcsm/γy, re-
ferred to as the shear strain ratio, and the web slenderness  w was established on the basis of
experimental and numerical data, and is given by Eq. (1).
 csm 0.62   0.10 u 
 2.56 but csm  minimum 15,  for  w < 0.83 (1)
 y w y   
 y 
in which γcsm is the CSM limiting shear strain, γy is the yield strain in shear,  u is the ultimate
shear strain of the material and  w   y  cr is the web slenderness, where τy is the yield
strength in shear and τcr is the elastic shear buckling stress. Based on this shear deformation
capacity, the CSM limiting shear stress can be calculated from the elastic, linear-hardening (of
slope Gsh) material model in shear, given by Eq. (2).
 csm   y  Gsh ( csm   y ) for γcsm ≥ γy (2)
235
The web shear resistance Vbw,csm,Rd is then given by Eq. (3), where Aw is the web area and γM0
is a partial factor.
Vbw,csm,Rd  Aw τ csm  M 0 (3)
The CSM predictions of shear capacity were compared to the ultimate shear capacity of a se-
ries of tested stainless steel plate girders. On average, considering both the enhanced CSM
shear capacity proposed herein, and the CSM bending capacity from previous work [6], im-
provements in accuracy of prediction of 10% over existing methods was achieved, together
with a reduction in scatter of the prediction. A typical comparison of moment-shear interac-
tion surfaces from the CSM and EN 1993-1-4 [9] is shown in Fig. 1.
2400
Test result
M-V interaction (EN 1993-1-4)
2100
M-V interaction (CSM)
Vb,csm,Rd
1800
1500 Vb,Rd
V (kN)
1200
900
600
300
Mc,Rd Mcsm,Rd
0
0 300 600 900 1200 1500
M (kNm)
Fig. 1: Moment-shear interaction diagrams according to EN 1993-1-4 [9] and CSM for plate girder I-
600×200×12×10-1
References
[1] Gardner L. “The continuous strength method”, Proceedings of the Institution of Civil
Engineers-Structures and Buildings, 161(3), 127-133, 2008.
[2] Gardner L, Theofanous M. “Discrete and continuous treatment of local buckling in
stainless steel elements”, Journal of Constructional Steel Research, 64(11), 1207-1216,
2008.
[3] Gardner L, Wang F, Liew A. “Influence of strain hardening on the behaviour and design
of steel structures”, International Journal of Structural Stability and Dynamics, 11(5),
855-875, 2011.
[4] Saliba N, Gardner L. “Cross-section stability of lean duplex stainless steel welded I-
sections”, Journal of Constructional Steel Research, 80, 1-14, 2013.
[5] Saliba N. Structural behaviour of lean duplex stainless steel welded I-sections [PhD
thesis], Department of Civil and Environmental Engineering, Imperial College London,
2012.
[6] Afshan S, Gardner L. “The continuous strength method for structural stainless steel de-
sign”, Thin-Walled Structures, 68, 42-49, 2013.
[7] Saliba N, Gardner L. “Experimental study of the shear response of lean duplex stainless
steel plate girders”, Engineering Structures, 46, 375-391, 2013.
[8] Saliba N, Real E, Gardner L. “Shear design recommendations for stainless steel plate
girders”, Engineering Structures, 59, 220-228, 2014.
[9] EN 1993-1-4. “Eurocode 3: Design of steel structures - Part 1.4: General rules - Sup-
plementary rules for stainless steel”, CEN, 2006.
236
Tampere, Finland
STAINLESS STEEL AT SLIGHTLY ELEVATED TEMPERATURES
Hans L. Groth*, Erik Schedin, Emma Jakobsen and Rita Lindström
Outokumpu Stainless AB, Avesta Research Center, Avesta, Sweden

*
Author for contact. Tel.: +46 (0)226 81996; E-mail: hans.groth@outokumpu.com
Abstract: The mechanical properties at slightly elevated temperatures (outdoor in the sun)
will be discussed. It is known that the drop in strength is quite drastic when the temperature
increases over RT or 20 °C. This is a drawback of current design practice and leads to a too
conservative design.
Looking at the mechanical strength values in European standards for austenitic and duplex
stainless steels as function of the temperature (from room temperature and up) one can divide
the strength behaviour into four different regions. The first, from Room Temperature (RT,
about 20 °C) up to 100-200 °C, where there is a significant drop in strength. Secondly, be-
tween 200-550 °C the strength is relatively constant with only a minor drop. The third region
is above 550 °C, where the creep properties of the material play a more and more important
role. Even though we are considering the proof strength, Rp0.2, or tensile strength, Rm, the
time dependency of the material will drastically decrease the strength as the temperature in-
creases. The last region, below RT the strength increases slightly and in an almost linear way
as the temperature decreases. For most outdoor applications, however, the strength at RT is
used as design strength for the lower temperatures.
In present paper, only the first region with the significant drop in strength above RT will be
covered.
Many buildings, bridges, storage tanks and other infrastructure and industrial outdoor applica-
tions are designed for a maximum temperature of 40 to 60 °C in the warmer parts of the
world. The drastic drop in strength between RT and 40 to 60 °C has been noted by designers
and also by the standardization authorities in the building and construction sector. The drastic
drop in strength makes the stainless steels less attractive, as thicknesses and total weight will
be increased due to this.
This question has been raised by the Steel Construction Institute in UK and they have made a
proposal how to handle the “slightly elevated temperatures” that are quite common in the
building and construction sector.
237
Today different testing parameters and standards are used for the RT and the elevated temper-
ature (ET) tensile tests with wide range of possible testing parameters. Some recent tests will
be present and compared with traditional standardized data for a very common standard aus-
tenitic grade, EN 1.4307, and a high strength duplex stainless steel, EN 1.4162 (LDX 2101®).
It is shown that much of the drop in strength is highly and solely related to the testing parame-
ters and that "high temperature" in the traditional way of testing and testing procedures starts
already just above 20 °C. The present paper will discuss the reason for the drop in strength
and finally a proposal will be discussed how to handle the slightly elevated temperatures in the
design process and also how a fire design curves (strength retention curves) can be modified
for a more efficient design.
238
NEW STEEL DAMPER WITH DISPLACEMENT DEPENDENT
RECENTERING FOR SEISMIC PROTECTION OF STRUCTURES
Murat Diclelia 1, Ali Salem Milanib

a,b
Middle East Technical University, Department of Engineering Sciences
Abstract: In this paper, a summary of analytical and experimental studies into the behavior of
a new hysteretic damper, designed for seismic protection of structures is presented. The Multi-
directional Torsional Hysteretic Damper (MTHD) is a patented invention in which a
symmetrical arrangement of identical cylindrical steel elements is so configured as to yield in
torsion while the structure experiences planar movements due to earthquake shakings. The
device has gone through many stages of design refinement, multiple prototype verification
tests and development of design guide-lines and computer codes to facilitate its
implementation in practice.
Basic mechanisms and working principle of the MTHD

MTHD is designed to dissipate energy by torsionally-yielding cylindrical energy dissipaters,
named yielding cores. Eight of these identical yielding cores each attached to a torsion arm are
arranged in a symmetric configuration to create the MTHD device, as depicted in Figure 1-a,b.
To convert translational motion of the structure to twisting in the cylindrical cores, each arm is
coupled with a guiding rail which through a low-friction slider block guides the motion of the
arm. A distinguishing feature in force-displacement response of MTHD is the geometric
hardening behavior which is the outcome of translation-to-rotation motion conversion
mechanism in MTHD, as schematized in Figure 2-a. This mechanism also offers the
possibility of controlling the desired level of hardening in force-displacement response,
through adjustment of the arm length to maximum displacement ratio. Varying levels of
hardening obtained as such, leads to hysteresis loops of different shapes as shown in Figure 2-
b. A 200kN, 120mm-capacity version of the device was built and tested in UniBw/Munich and
also at METU, as shown in Figure 3-a. A typical force-displacement response loop obtained
from tests is given in Figure 3-b, which shows a very stable cyclic response with little
variation in force levels not exceeding %4.0 the mean value. MTHD is capable of reaching
high force and displacement capacities, shows high levels of damping, controllable post-
elastic stiffness and very stable cyclic response. A design methodology for the device has also
been completed.
1
mdicleli@metu.edu.tr, Phone: +90(312)2104451.
239
(a) (b)
(c)
Fig. 1: MTHD: (a) Isometric view showing the rail system and base device underneath; (b) side
view; (c) energy dissipation unit of MTHD: A yielding core, as attached to other components of the
device.
d 2  d 1 :  2  1 : cos 2  cos1 T2 T1
⇒  : F 2  F1
T1 ≤T2 L cos2 L cos1
(a) (b)
Fig. 2: (a) Working mechanism of MTHD responsible for geometric hardening; (b) MTHD
response for different design hardening indices (HI=Fmax/FY).
(a) (b)
Fig. 3: (a) 200kN, 120mm-capacity prototype MTHD, as tested at METU; (b) Sample cyclic
response obtained from tests.
240
Tampere, Finland
FRETTING FATIGUE PHENOMENON IN BOLTED HIGH-STRENGTH

STEEL PLATE CONNECTIONS
Olli-Pekka S. Hämäläinena, Timo J. Björkb

a,b
Laboratory of Steel Structures, Lappeenranta University of Technology, Finland
Correspondence: olli-pekka.hamalainen@lut.fi, tel. +358 40 7485803
Extended abstract
Welding of high-strength steels can be challenging. These materials set high requirements for
personnel as well as the equipment and especially in on-site welding these requirements are
very problematic to meet. Use of bolted connections instead of welded joints would simplify
the joining process and result in more economical as well as possibly more fatigue-resistant
solutions, since the welding process is detrimental to fatigue strength. The behavior of a bolt-
ed joint is otherwise fairly predictable, but sudden fatigue failures might occur because of a
phenomenon known as fretting – i.e. initiation and growth of microcracks between joined sur-
faces due to a combination of friction forces and local slipping.
Fretting fatigue of bolted joints is an interesting topic with many possible parameters affecting
the process. So far the scope of research in the field has very seldom included high-strength
steels – rather concentrating on aluminum or titanium alloys and coatings. In this research the
object was a double-lap bolted joint consisting of two middle plates, two connecting plates
and a total of four M16 fine thread bolts and nuts. Two test batches of plates were made – one
of regular S355 steel and one of high-strength S960QC steel. The joints were assembled and
fatigue tested in a servo hydraulic test machine. The observed fatigue lives of both batches
were compared to each other as well as to the results gained from an FE model. Analytical
calculations in order to find out the expected crack initiation life were made using the stress
data gained from FE model and Smith-Watson-Topper (SWT) parameter.
The results were at first sight very surprising. The S355 joints endured all the fatigue tests and
test specimens only fractured by fretting fatigue in the part where the specimen was connected
to the jaws of the test machine. The actual joints remained intact during the testing and after
disassembly only some fretting wear was present in plate surfaces – no fatigue cracks. Mean-
while the S960QC joints couldn’t handle similar fatigue loadings and fractured after the fret-
ting fatigue cracks had grown to a diameter of approximately 25 mm. While the S355 joints
lasted well over 5 million cycles without a failure, the S960QC joints failed after only
1.1…1.6 million cycles, even though the loading was identical. The most common crack initi-
241
ation area was in the proximity of the outer bolt hole – not in the centerline, but 5…15 mm to
either side.
FE analysis gave an explanation to this phenomenon. The coefficient of friction for untreated
S355 steel is around 0.3, but for S960QC it is roughly 0.5. This difference causes that in the
case of S355 the pretensioned bolts can’t develop enough friction in their neighborhood to
keep the joint together, but the friction is spread in a wider area. For this reason there are no
significant tangential stress concentrations near the edge of washer where the contact pressure
rapidly changes. However, in the case of S960QC the bolts can create enough friction in the
nearby areas to keep the joint together, and hence the tangential stress is significantly higher in
some areas near the bolt (and respectively remarkably lower in more remote areas). This une-
ven stress distribution shows as stress concentration peaks on both sides of the bolt hole, just
near the washer edges. This explains the fretting-sensitive behavior of S960QC joints.
Crack initiation and its significance in the fretting process is a slightly controversial topic: ac-
cording to some scientists it happens so rapidly that it can be neglected, while some claim that
it is so slow that it actually governs the whole fretting fatigue life. Since there were no signs of
propagated cracks in any other middle plates except the ones that had failed, though the fret-
ting conditions should be identical in both middle plates, the crack initiation life seemed to be
on the more significant side and therefore the most interesting part of analysis. The maximum
local tangential stress value in the critical spot was extracted from the FE model. Using this
value and some material parameters it was possible to calculate the expected crack initiation
life by SWT. This crack initiation life was calculated to be 1.36 million cycles for S960QC,
which is similar to the experimental results. The only thing hindering this result is that correct
values for material parameters σf’ (fatigue strength coefficient) and b (fatigue strength expo-
nent) are not yet published for S960QC so crude approximations had to be used. Because the
SWT parameter is highly volatile for these parameters the accuracy of the method in general is
likely not good enough to become a standardized procedure. Anyhow, all reasonable values
that were experimented predicted a long initiation life, so that combined with the experimental
results tends to lead in the direction that the initiation phase is an important part of fatigue
process.
Future work on the topic will include firstly clarifying whether the difference between the fret-
ting behavior of S355 and S960QC remains this large if the plate surfaces are machined to
same surface roughness. The analytical life prediction model must be tested with correct
S960QC material parameters when they will be found out. Also the FE model would benefit
from more dense mesh in the most interesting areas in terms of increased accuracy.
Keywords Fretting fatigue, high-strength steel, bolted joint, double-lap joint
242
Tampere, Finland
COMPARISON OF RELATIVE VOLUMES OF DIFFERENT TYPE OF

WELDS
Juha Kukkonen a, Markku Heinisuo b
a
Sweco Structures, Tampere, Finland
b
Introduction
The weld volumes with different steel grades are compared to the weld volumes of T-butt
weld volumes. In this comparison the sizes of the welded plates are changing in order to get
full strength welds. In order to make the comparison where the total force which is transmit-
ted at the joint is a constant, the scaling with respect of the resistance of the S355 was done,
ending to so called weld volume-strength-ratio.
Results
In Figs. 3 and 4 are given the results for the shear load of a T-joint. In the paper are given the
results for the axial load and all background of the study.
Fig. 3: T-joint, shear load, weld correlation factor βw of Eurocodes
243
Fig. 4: T-joint, shear load, modified βw based on [1]
Conclusions
Comparing the results which are calculated by using recommended values of correlation fac-
tors βw of Eurocodes can be concluded that the volume of the fillet weld increases strongly
with steel grades S420 to S690. The weld volume-strength-ratio is 1 for S690 steel grade, but
for S420, S460 and S550 this ratio is about 1.3 both for the shear load and the axial load. Cal-
culating the weld size with the modified correlation factor of [1] the volume of the fillet weld
increases strongly for steel grades S550 and S690 but the volume for steel grades S420 and
S460 it is decreasing. The weld volume-strength-ratio for steel grades S420 and S460 is prac-
tically the same compared to steel grade S355 in the shear load and in the axial load. Howev-
er, the weld volume-strength-ratio increases considerably for the steel grades S550 and S690
when using these modified parameters compared to the results using the parameters of Euro-
codes.
For steel grades S420 and S460 the modified correlation factors seems to be quite reasonable
and these values makes the use of steel grades S420 and S460 more competitive than using
present Eurocodes. Instead of this steel grades S550 and S690 seem to come not competitive
with these modified correlation factors.
In order to make the use of high strength steels competitive more research is needed to define
equally safe weld design rules and parameters no matter what the steel grade is.
References
[1] Stroetmann R, Deepe P, Rasche C, Kuhlmann U. ”Bemessung von Tragwerken aus

höherfesten Stählen bis S700 nach EN 1993-1-12”, Stahlbau 81 (2012), Heft 4, 332-
342: Ernst&Sohn
244
Tampere, Finland
INVESTIGATION OF COLD FORMED STEEL BEAM TO COLUMN

BOLTED GUSSET PLATE CONNECTIONS
Žilvinas Bučmys, Alfonsas Daniūnas
Vilnius Gediminas Technical University, Dept. Of Steel And Timber Structures, Lithuania
Abstract
Cold formed thin walled sections are widely used as bearing structures in construction sites
because of good cost to bearing capacity ratio, fast and easy erection. In most cases thin
walled sections are used as purlins, steel trusses and for light weight portal frames. It is simple
to connect cold formed sections using gusset plates and bolts.
The first task of this paper is to investigate the influence of lateral restraints to beam and col-
umn gusset plate connections. It is important in normal exploitation period as in the construc-
tion sites during assembly period then not all the ties and connections are installed. The sec-
ond task of the paper is to compare the experimental data of three specimens (Fig. 1) with the
analytical stiffness and strength calculations according to Eurocode 3.
Pinned support
Displacement
transducer
Vertical load
460
1200
1850
460
1 700
Displacement
Displacement transducer (2)
transducer (1)
Displacement
transducer
1000
Pinned support
a) The scheme of laboratory tests b) The specimen in laboratory
Fig. 1: The specimen
245
Conclusions
The natural experiment results and analytical analysis according to Eurocode 3 of the connec-
tions of cold formed sections allow making such conclusions:
1. Eurocode 3 is accurate tool to calculate strength capacity of cold formed steel connec-
tions. Safety margin was from 3 % to 18,6 %.
2. Boundary conditions have huge impact to the strength of the connection. The test re-
sults showed that absence of lateral restrains resulted in 31 % decrease in strength ca-
pacity.
3. Eurocode 3 may assign rigid or semi rigid connection as pinned connection, because it
does not take into account slipping due to bolt hole clearance. The Eurocode method
of describing the behaviour of connections as rigid, semi rigid or pined is not suitable
for such connections.
4. The stiffness of the connection is higher calculating with Eurocode 3 than experi-
mental value. This means that Eurocode 3 is not suitable for such connection calcula-
tions.
Acknowledgments
This work has been supported by the European Social Fund as part of the project “Develop-
ment and application of innovative research methods and solutions for traffic structures, vehi-
cles and their flows”, project code VP1-3.1-ŠMM-08-K-01-020.
References
[1] Wong, M. F.; Chung, K. F. 2002. Structural behaviour of bolted moment connections in cold-
formed steel beam-column sub-frames, Journal of constructional steel research Vol. 58, pp. 253-
274.
[2] Yu, W. K.; Chung, K. F.; Wong, M. F. 2005. Analysis of bolted moment connections in cold –
formed steel beam-column sub-frames, Journal of constructional steel research, Vol. 61, pp.
1332-1352.
[3] Sabbagh, A. B.; Petkovski, M.; Pilakoutas, K.; Mirghaderi, R. 2011. Ductile moment – resisting
frames using cold – formed steel sections: an analytical investigation, Journal of constructional
steel research, Vol. 67, pp. 634-646.
[4] Sabbagh, A. B.; Petkovski, M.; Pilakoutas, K.; Mirghaderi, R. 2012. Development of cold-
formed steel elements for earthquake resistant moment frame buildings, Thin-walled structures
Vol. 53, pp. 99-108.
[5] Sabbagh, A. B; Petkovski, M.; Pilakoutas, K.; Mirghaderi, R. 2013. Cyclic behaviour of bolted
cold-formed steel moment connections: FE modelling including slip, Journal of constructional
steel research, Vol. 80, pp. 100-108.
[6] Bucmys, Z.; Sauciuvenas, G. 2013. The behaviour of cold formed steel structure
connections, Engineering structures and technologies, Vol. 5:3, pp. 113-122.
[7] EN 1993-1-8: 2005: Eurocode 3 - Design of steel structures - Part 1-8: Design of joints. Comité
Européen de Normalisation, Brussels.
[8] Urbonas, K.; Daniunas, A. 2006. Behaviour of semi-rigid steel beam-to-beam joints under
bending and axial forces, Journal of constructional steel research, Vol. 62, pp. 1244-1249.
[9] Bucmys, Z.; Daniunas, A.; Rasiulis, K 2014. Investigation of cold formed steel connections :
experimental and numerical analysis of beam to column gusset plate connections, Eurosteel
2014: 7th European conferece on steel and composite structures, September 10-12, 2014,
Naples, Italy : abstracts book. Brussels : ECCS European Convention for Constructional
Steelwork, 2014. ISBN 9789291471218. pp. 417-418.
246
Tampere, Finland
RESISTANCE RESULTS FOR THE CROCODILE CONNECTION
Panagiotis Manoleasa, Kristoffer Öhmanb, Efthymios Koltsakisc and Milan Veljkovicd

a, b, c, d
Luleå Tekniska Universitet
Abstract: A novel type of connection for circular hollow sections (CHS) is investigated in the
LTU in the framework of the project High Strength Long Span Structures (HILONG). The so
called Crocodile Nose (CN) connection is an aesthetically improved alternative to the com-
monly used slotted-end CHS connection. The end of the CHS member is tapered and a pair of
inflected plates, welded on the cut faces, transfers the load. This research is focused on two
parameters: the bevelling angle and a stiffener. The program is completed through laboratory
tests and FEA.
1 The CN connection
Hollow sections possess excellent cross sectional properties and are commonly used in many
structural applications including bracing members and trusses. The appealing appearance of
CHS, together with its improved properties makes them a favourable choice of architects in
applications where the structural system is visible to the end-user.
The common case of slotted-end connections and its dominant failures have been extensively
studied in the context of stainless and high strength steel [1]–[4]. Design guidance and exam-
ples are also provided [5], [6].
The Crocodile Nose (CN) connection is studied and presented herein as an alternative to the
slotted-end connection. The CHS member has a tapered end from where a pair of plates,
welded on the CHS, is protruding offering space for bolts (Fig. 1). This distinctive geometry,
from which the connection grants its name, results in:
Fig. 1: Parts of the CN connection.

1. avoiding the aforementioned problems that occur in a slotted-end connection;
2. improving the shear-lag effect by reducing the CHS cross-sectional area;
3. a visually smooth transition from the member to the joint.
247
As shown in Fig. 1, the CN connection consists of four parts:
1. the connected CHS member;
2. the pair of inflected plates, welded on the tube and protruding outwards;
3. the gusset plate upon which the CN is connected;
4. a connecting piece welded on the inner faces of the inflected plates.
2 Methodology and conclusions

Four specimens of the CN connection were fabricated and tested. The test matrix was formed
taking into consideration 2 parameters: the tapering angle of the CHS and the contribution of
the connecting piece.
Table 1: Test specimen matrix

Nr. Tapering angle Connecting piece
CN 1 1:2 Without
CN 2 1:2 With
CN 3 1:1.5 Without
CN 4 1:1.5 With
Prior to the tests, a series of numerical analyses were ran using ABAQUS. The results of these
preliminary analyses helped in evaluating the magnitude and the gradient of the strain fields so
that proper strain gauges are used. An initial estimation for the strength was also possible.
The conclusions of this study can be summarised as:
1. The connecting piece can increase the resistance from 150% to 300%
2. Sharper tapering angles lead to more efficient utilization of the welds
3. 3 different failure modes were observed: Ductile progressive failure of the
welds, brittle crack of the entire length of the welds and rapture of the inflected plate.
References
[1] G. Martinez-Saucedo and J. A. Packer, “Slotted end connections to hollow sections,” De-
partment of Civil engineering, University of Toronto, Final report to CIDECT on pro-
gramme 8G, Aug. 2006.
[2] G. Kiymaz and E. Seckin, “Behavior and design of stainless steel tubular member welded
end connections,” Steel Compos. Struct., vol. 17, no. 3, pp. 253–269, Sep. 2014.
[3] G. Martinez-Saucedo and J. Packer, “Static Design Recommendations for Slotted End
HSS Connections in Tension,” J. Struct. Eng., vol. 135, no. 7, pp. 797–805, 2009.
[4] T. W. Ling, X. L. Zhao, R. Al-Mahaidi, and J. A. Packer, “Investigation of block shear
tear-out failure in gusset-plate welded connections in structural steel hollow sections and
very high strength tubes,” Eng. Struct., vol. 29, no. 4, pp. 469–482, Apr. 2007.
[5] Edurne Nunez Moreno, Cyrill Tarbe, David Brown, and Adbul Malik, Joints in Steel Con-
struction: Simple Joints to Eurocode 3, Reprint. SCI/BSCA, 2014.
[6] Wardenier J., Kurobane Y., J.A. Packer, G.J. van der Vegte, and X. -L. Zhao, Design
guide for circular hollow section (CHS) joint under predominantly static loading, 2nd ed.,
vol. 1. CIDECT, 2008.
248
Author index
Aalberg, Arne, 145, 213 Fujihashi, Kazunori, 163

Aarønæs, Anton, 207 Fülöp, Ludovic, 81
Aasen, Bjørn, 31 Gardner, Leroy, 235
Abspoel, Roland, 113 Genge, Nico, 97
Afzali, Nariman, 171 Gentili, Filippo, 175
Afzali, N., 173 Glienke, R., 173
Akbas, Bulent, 119 Gresnigt, N., 173
Al-Emrani, M., 201 Grimsmo, Erik, 145
Ampatzis, Alexios, 135 Groth, Hans, 237
Andrade, Pedro, 199, 215 Gunalan, Shanmuganathan, 155
Andreassen, Michael, 71 Gyllenram, Rutger, 11
Baczkiewicz, Jolanta, 153 Gödrich, Lukáš, 77
Barros, Rui, 149 Haakana, Äli, 219
Berg, Jörn, 171, 173 Hakimi, Poja Shams, 201
Berger, S., 171 Havula, Jarmo, 89
Bijlaard, F., 173 Heinisuo, Markku, 79, 89, 217, 219, 243
Björk, Timo J., 241 Heistermann, C., 215
Botti, Andrea, 73 Heistermann, Tim, 199, 215
Bradford, Mark, 227 Henriques, Jorge, 149
Braun, Matthias, 189 Horváth, László, 223
Brauns, Janis, 185 Hradil, Petr, 229
Bučmys, Žilvinas, 245 Hämäläinen, Olli-Pekka, 241
Bujňák, Ján, 83, 107 Ibrahim, Saeed, 183
Bzdawka, Karol, 153 Ikarashi, Kikuo, 99, 117, 161
Classen, Martin, 141 Jacobsen, Emma, 237
Clausen, Arild, 145 Jalkanen, Jussi, 91
Clercq De, Jesse, 211 Janarthanan, Balasubramaniam, 155
Costa, Ricardo, 175 Janik, Peter, 111
Daniūnas, Alfonsas, 245 Jaspart, Jean-Pierre, 21
Dekker, Rianne, 115 Jensen, Hans Vagn, 105
Dicleli, Murat, 101, 109, 123, 239 Jespersen, Martin, 209
Donnadieu, Marc, 81 Jokinen, Timo, 217
Döring, Bernd, 73 Järvinen, Saku, 1
Ebel, Rebekka, 197 Jönsson, Jeppe, 71, 151
Ebert, A., 173 Kabeláč, Jaromír, 77
Efthymiou, Evangelos, 135 Kaijalainen, Antti, 221
El Hosseiny, Ossama, 183 Kaljas, Toomas, 179
El Kadi, Bassel, 167 Kansinally, Richard, 187
Exner, Hans, 131 Kaplin, Camilla, 205
Feldmann, Markus, 73, 85, 137, 225 Karalar, Memduh, 101, 109, 123
Finnås, Anders, 205 Karoumi, Raid, 125
Fortan, Maarten, 211 Kemper, Frank, 137
Francavilla, Antonella, 169 Kesti, Jyrki, 61, 73
Friedrichsen, Trygve, 127, 133 Kesti, Vili, 221
Frøseth, Gunnstein, 103 Kimura, Yoshihiro, 95
249
Kiymaz, Guven, 167 Pedersen, Helge, 133
Knobloch, Markus, 197 Petersen, Christian, 127
Knoedel, Peter, 139 Petersen, Tobias, 159
Koltsakis, Efthymios, 231, 247 Piluso, Vincenzo, 169
Kostakakis, K., 201 Psomiadis, Vasileios, 135
Krahwinkel, Manuel, 159 Puttonen, Jari, 143
Kubota, Daiki, 99 Qin, Ru, 165
Kuhnhenne, Markus, 73 Rebelo, Carlos, 233
Kukkonen, Juha, 243 Reger, Vitali, 73
Kurejková, Marta, 77 Remde, Christian, 97
Lange, Jörg, 181 Ren, Zhong, 165
Langseth, Magnus, 145 Renner, Anja, 181
Larsen, Hilmer, 127, 133 Rizzano, Gianvittorio, 169
Larsen, Per, 213 Roivio, Pekka, 89
Latour, Massimo, 169 Ronni, Hilkka, 79
Lawson, Mark, 73 Rossi, Barbara, 211
Leander, John, 125 Rudolf, A., 171
Lehnert, Tobias, 203 Ruoppa, Raimo, 221
Leskela, Matti V., 193 Rönnquist, Anders, 103
Lindström, Rita, 237 Šabatka, Luboš, 77
Liu, Dasu, 177 Saliba, Najib, 235
Lombaert, Geert, 93 Sandström, Joakim, 195
Lu, Wei, 143 Sato, Atsushi, 163
Lundholm, John, 199 Schaffrath, Simon, 225
Maaly, Hassan, 183 Schaumann, Peter, 51
Magnucki, Krzysztof, 157 Schedin, Erik, 237
Mahendran, Mahen, 155 Schevenels, Mattias, 93
Maljaars, J., 115 Schiborr, M., 171, 173
Mangir, Atakan, 167 Schillo, Nicole, 85, 225
Manoleas, Panagiotis, 231, 247 Schröter, Falko, 203
Maslak, Mariusz, 191 Schäfer, Markus, 121
Meeus, Marc, 211 Shirai, Daigo, 117
Mela, Kristo, 79, 217 Simões da Silva, Luis, 175, 233
Mellaert Van, Roxane, 93 Snela, Malgorzata, 191
Milani, Ali, 239 Snijder, H.H., 115
Mori, Seiji, 163 Sokol, Leopold, 75
Mosiello, A., 201 Stark, Alexander, 141
Mourujärvi, Juho, 221 Steige, Yvonne, 147
Neumann, Nicolas, 207 Stottrup-Andersen, Ulrik, 209
Nielsen, Mogens, 209 Stranghöner, N., 171, 173
Nilsson, Hanna, 207 Stroetmann, Richard, 87
Obiala, Renata, 189 Sugimura, Yuji, 129
Odenbreit, Christoph, 189 Talja, Asko, 229
Odrobiňák, Jaroslav, 107 Thomsen, Kjeld, 127, 133
Ohmichi, Kenjiro, 129 Tiainen, Teemu, 217
Ono, Tetsuro, 163 Tsavdaridis, Konstantinos, 187
Paczos, Piotr, 157 Tuan Tran, Anh, 233
Paiva, Fabio, 149 Turán, Pál, 223
Palisson, Anna, 75 Uhre, Arne, 213
Pavlovic, Marko, 215, 231 Ulf, Wickström, 195
250
Ummenhofer, Thomas, 139
Wald, František, 77
Veljkovic, Milan, 41, 199, 215, 231, 233,
247
Weynand, Klaus, 97
Weynand, Klaus, 147
Vican, Josef, 111
Vries de, P., 173
Yamaguchi, Eiki, 129
Ylinen, Kimmo, 143
Yokoyama, Yoshifumi, 161
Yoshino, Yuki, 95
Zhao, Xianzhong, 165
Öhman, Kristoffer, 247
251
Presentations Wednesday 23/Sep/2015
Keynote 1-2 10:30 - 12:00 Session Chair Markku Heinisuo
• BIM in structural steel workflow, Saku Järvinen
• Making sustainability activities a key to your success - from compliance to commitment, Rutger Gyllenram
Keynote 3 13:00 - 13:45 Session Chair Markku Heinisuo

• Component method as a general tool for the design of joints under various loading conditions, Jean-Pierre Jaspart
Plenary Session A 13:45 - 15:00 Session Chair Markku Heinisuo

• Joint and column behaviour of slotted cold-formed steel studs, Michael Andreassen, Jeppe Jönsson
• Steel solutions for enabling zero energy buildings, Bernd Döring, Reger, Kuhnhenne, Kesti, Lawson, Botti, Feldmann
• Plastic resistance of composite slabs in partial shear connection, Leopold Sokol, Anna Palisson
• Future design procedure for structural connections is component based finite element method, František Wald, et. al.
• Comparative evaluation of steel profiles in roof trusses, Kristo Mela, Hilkka Ronni, Markku Heinisuo
Keynote 4 15:30 - 16:15 Session Chair Jari Mäkinen

• Execution of steel structures - recent developments and future trend, Bjørn Aasen
Plenary Session B 16:15 - 17:30 Session Chair Jari Mäkinen

• Non-linear finite element modelling of steel-concrete-steel members in bending and shear, Marc Donnadieu, Ludovic Fülöp
• Assessment of existing steel bridge structures, Jan Bujnak
• Local buckling behaviour of welded box sections made of high strength steel - comparison of experiments with EC3 and general
method, Nicole Schillo, Markus Feldmann
• Sustainable design of buildings in steel and composite structures, Richard Stroetmann
• Steel construction excellence center, Jarmo Havula, Pekka Roivio, Markku Heinisuo
Presentations Thursday 24/Sep/2015

Keynote 5 8:00 – 8:45 Session Chair: Richard Stroetmann
• Use of higher strength steel in construction, opportunities and obstacles, Milan Veljkovic
Session 1A: Building Structures 1 8:45 – 10:00 Session Chair: Richard Stroetmann
• Practical Tubular Truss Optimization, Jussi Jalkanen
• The impact of joint constraints on the optimal design of truss structures, Roxane Van Mellaert, G Lombaert, M Schevenels
• Lateral buckling stress for H-shaped beams with continuous braces, Yoshihiro Kimura, Yuki Yoshino
• Industrial Hall Constructions, Nico Genge, Christian Remde, Klaus Dr. Weynand
• Effect of end stiffener reinforcement on lateral torsional buckling behavior of H-shaped beams with large depth-thickness ratio,
Daiki Kubota, Kikuo Ikarashi
Session 2A: Bridges & Fatigue 8:45 – 10:00 Session Chair: Jean-Marc Battini
• Low cycle fatigue performance of integral bridge steel h-piles under seismic displacement reveals, Murat Dicleli, Karalar
• System reliability analysis of steel railway bridge based on historic rolling stock records, Gunnstein Frøseth, Rönnquist
• Fatigue problems at riveted railway bridges investigation and rehabilitation methods, Hans Vagn Jensen
• On actual behaviour of continuous composite girder bridges and their conventional modelling, Jaroslav Odrobiňák, Ján Bujňák
• New cycle counting method for the assessment of low cycle fatigue in steel H-piles of integral bridges, Memduh Karalar, Dicleli
Session 1B: Building Structures 2 10:30 – 12:00 Session Chair: Kristo Mela
• Resistance of eccentrically loaded beam-columns, Josef Vican, Peter Janik
• Experiments on plate girders with a very slender web, Roland Abspoel
• Experimental study into bending-shear interaction of rolled I-shaped sections, Rianne Dekker, H.H. Snijder, J. Maljaars
• Effect of neutral-axis position on the elastic buckling characteristics of continuous composite beams, Daigo Shirai, Ikarashi
• Amplified seismic loads in steel moment frames, Bulent Akbas
• Design rules for slim-floor girders considering the composite behavior, Markus Schäfer
Session 2B: Bridges 10:30 – 12:00 Session Chair: Jan Bujnak

• Effect of longitudinal stiffeners on the flanges to improve the low cycle fatigue performance of steel h- piles, Karalar, Dicleli
• Refined fatigue assessment of an existing steel bridge, John Leander, Raid Karoumi
• Odins Bridge, Kjeld Thomsen, Hilmer Larsen, Christian Petersen, Trygve Friedrichsen
• High-performance-steel girder of compact section, Eiki Yamaguchi, Yuji Sugimura, Kenjiro Ohmichi
• Steel Bridge Technology used in Buildings, Hans Exner
• Sundsvall Bridge, Kjeld Thomsen, Helge Pedersen, Hilmer Larsen, Trygve Friedrichsen
Keynote 6 13:00 – 13:45 Session Chair: František Wald

• Fire design of steel structures with intumescent coating, Peter Schaumann
Session 1C: Building Structures 3 13:45 – 15:00 Session Chair: František Wald
• Aluminium deployment in bracing systems: Investigation of shear link application, E Efthymiou, V Psomiadis, A Ampatzis
• Design of wind turbine structures based on a multivariate stochastic approach, Frank Kemper, Markus Feldmann
• Time history simulation in seismic design, Peter Knoedel, Thomas Ummenhofer
• Steel composite dowels in cracked concrete, Martin Classen, Alexander Stark
• Cross-sectional capacity of compocite column by the two methods of EN 1994-1-1, Kimmo Ylinen, Wei Lu, Jari Puttonen
Session 2C: Connections 13:45 – 15:00 Session Chair: Michael Joachim Andreassen
• Beam-to-column joints subjected to impact loading, Erik L. Grimsmo, Arild H. Clausen, Arne Aalberg, Magnus Langseth
• Design resistance of end-plate splices with hollow sections, Yvonne Steige, Klaus Weynand
• Conception, analysis and design of a special joint for fixing lattice towers legs during testing of transmission line tower, Fabio
Paiva, Jorge Henriques, Rui C. Barros
• Generalized Block Failure, Jeppe Jönsson
• FEM simulation of a tubular KT-joint, Karol Bzdawka, Jolanta Baczkiewicz
Session 1D: Cold Formed Structures 15:30 – 17:30 Session Chair: Jeppe Jönsson
• Bearing capacity of cold-formed unlipped channels with restrained flanges - EOF and IOF load cases, Mahen Mahendran,
Balasubramaniam Janarthanan, Shanmuganathan Gunalan
• Elastic buckling of an I-beam with sandwich flanges, Krzysztof Magnucki, Piotr Paczos
• A numerical parametric study on the load carrying behaviour under bending of honeycomb girders made of trapezoidal
corrugated steel sheets, Tobias Petersen, Manuel Krahwinkel
• Elastic Buckling Characteristics of Corrugated Tank under Fundamental Load, Yoshifumi Yokoyama, Kikuo Ikarashi
• Buckling strength of light-gauge members with large openings, Atsushi Sato, Seiji Mori, Tetsuro Ono, Kazunori Fujihashi
• Experimental and numerical investigations of the steel storage rack uprights, Zhong Ren, Xianzhong Zhao, Ru Qin
• Experimental investigation on the behavior of perforated steel storage rack columns under axial compression, Bassel El Kadi,
Guven Kiymaz, Atakan Mangir
Session 2D: Connections 2 15:30 – 17:30 Session Chair: Jari Mäkinen

• Monotonic behaviour of bolted T-stubs: a refined theoretical model for flange yirlding and bolt fracture failure mode, Antonella
Francavilla, Massimo Latour, Vincenzo Piluso, Gianvittorio Rizzano
• Different coating systems for the application in slipresistant connections, Stranghöner, Afzali, Berg, Schiborr, Rudolf, Berger
• Influence of different testing criteria on the slip factor of slip-resistant connections, N. Stranghöner, N. Afzali, Jörn Berg, M.
Schiborr, F. Bijlaard, N. Gresnigt, P. de Vries, R. Glienke, A. Ebert
• Simplified model for connections of steel structures in OpenSees, Ricardo Costa, Filippo Gentili, Luis Simões da Silva
• Design approach for stability check of members with hanging-profile connections, Dasu Liu
• Reasons for Charles de Gaulle airport collapse, Toomas Kaljas
• Investigations on the behaviour of threaded and shank bolts under combined tension and shear, Anja Renner, Jörg Lange
Presentations Friday 25/Sep/2015

Keynote 7 8:00 – 8:45 Session Chair: Mikko Malaska
• Energy-Efficient Solutions for Steel Structures – Case Study of Nearly Zero-Energy Building, Jyrki Kesti
Session 1E: Composite Structures 8:45 – 10:00 Session Chair: Mikko Malaska
• Behavior improvement of pultruded frp beam-column bolted connections, Ossama El Hosseiny, Hassan Maaly, Saeed Ibrahim
• Material strength effect on the behaviour of steel-concrete composite elements, Janis Brauns
• Vibration response of USFB composite floors, Richard Kansinally, Konstantinos Tsavdaridis
• Analyses of the load bearing behaviour of deep-embedded concrete dowels, CoSFB, Matthias Braun, R Obiala, C Odenbreit
Session 2E: Fire Engineering & Building Structures 8:45 – 10:00 Session Chair: J-M Battini
• Evaluation of axial force impact on the flexibility of a steel beam-to-beam end-plate joint subjected to bending when exposed to
fire, Mariusz Maslak, Malgorzata Snela
• Fire design of CFST columns - Improvements required for Eurocode 4, Matti V. Leskela
• Calculation of steel temperature in open cross sections based on fire exposure from CFD, Joakim Sandström, Wickström Ulf
• Lateral torsional buckling resistance – a comparison of analytical and numerical models, Rebekka Ebel, Markus Knobloch
Session 1F: Sustainable Engineering 10:30 – 12:00 Session Chair: Mark Bradford
• Fatigue life improvement of welded bridge details using high frequency mechanical impact (HFMI) treatment, Poja Shams
Hakimi, Andrea Mosiello, Konstantinos Kostakakis, Mohammad Al-Emrani.
• New developments in heavy plate production for modern steel construction, Tobias Lehnert, Falko Schröter
• Stainless steel, a sustainable material for sustainable structures, Anders Finnås, Camilla Kaplin
• Dynamic responce of pipe rack steel structures subjected to explosion loads, Anton Stade Aarønæs, H Nilsson, N Neumann
• Tall ambitions – onshore wind turbine tower - concepts for large hub heights, Martin Jespersen, M Nielsen, U Stottrup-Andersen
Session 2F: Connections 3 10:30 – 12:00 Session Chair: Kristo Mela

• Lateral stability of verandas by means of the glass panels, Maarten Fortan, Jesse De Clercq, Marc Meeus, Barbara Rossi
• End Plate Connection for Rectangular Hollow Section in Bending, Arne Aalberg, Arne Martin Uhre, Per Kristian Larsen
• Structural analysis models of steel trusses, Teemu Tiainen, Kristo Mela, Timo Jokinen, Markku Heinisuo
• Buckling of members of welded tubular truss, Markku Heinisuo, Äli Haakana
Session 1G: High Strength Steel 13:00 – 15:00 Session Chair: Richard Stroetmann
• Bendability and microstructure of OPTIM® 700 MC plus, Vili Matias Kesti, Antti Kaijalainen, Juho Mourujärvi, Raimo Ruoppa
• Experimental behaviour of tension plates with centre hole made from high strenght steel, Pál Turán, László Horváth
• Derivation of strain requirements for high strength steel using Johnson Cook model, Simon Schaffrath, N Schillo, M Feldmann
• Buckling strength of HSS beams, Mark Andrew Bradford
• True stress-strain relationship for finite element simulations of structural details under diffuse necking, Petr Hradil, Asko Talja
• Calibration of the ductile damage material model parameters for a high strength steel, Pavlovic, Manoleas, Veljkovic, Koltsakis
• Buckling observation of door openings for wind turbine towers, Anh Tuan Tran, M Veljkovic, C Rebelo, L Simões da Silva
Session 2G: Stainless Steel & Connections 13:00 – 15:00 Session Chair: Jari Mäkinen
• Extension of the continuous strength method to the determination of shear resistance, Najib George Saliba, Leroy Gardner 235
• Stainless steel at slightly elevated temperatures, Hans L. Groth, Erik Schedin, Emma Jacobsen, Rita Lindström
• New steel damper with displacement dependent recentering for seismic protection of structures, Murat Dicleli, Ali Salem Milani
• Fretting fatigue phenomenon in bolted high-strength steel plate connections, Olli-Pekka S. Hämäläinen, Timo J. Björk
• Comparison of relative volumes of different type of welds, Juha Kukkonen, Markku Heinisuo
• Investigation of cold formed steel beam to column bolted gusset plate connections, Žilvinas Bučmys, Alfonsas Daniūnas
• Resistance results for the crocodile connection, Panagiotis Manoleas, Kristoffer Öhman, Efthymios Koltsakis, Milan Veljkovic
13th Nordic Steel Construction Conference (NSCC-2015) Proceedings of
23-25 September 2015, Tampere Hall, Tampere, Finland
Date: Wednesday, 23/Sep/2015 The 13th Nordic Steel
8:30-10:10 Registration and Coffee
10:10-10:30 Opening Location: Small Auditorium
10:30-12:00 Keynote 1-2 Location: Small Auditorium Chair: Markku Heinisuo
Construction Conference
12:00-13:00 Lunch
13:00-13:45 Keynote 3 Location: Small Auditorium Chair: Markku Heinisuo (NSCC-2015)
13:45-15:00 Plenary Session A Location: Small Auditorium Chair: Markku Heinisuo
15:00-15:30 Coffee
15:30-16:15 Keynote 4 Location: Small Auditorium Chair: Jari Mäkinen
23-25 September 2015, Tampere, Finland
16:15-17:30 Plenary Session B Location: Small Auditorium Chair: Jari Mäkinen
19:00-21:00 Ice Breaking Event Location: Plevna Brewery Pub & Restaurant Chair: Jari Mäkinen
Date: Thursday, 24/Sep/2015

8:00-8:45 Keynote 5 Location: Sonaatti 1 Chair: Richard Stroetmann
8:45-10:00 Session 1A: Building Structures 1 Session 2A: Bridges & Fatigue
Location: Sonaatti 1 Chair: Richard Stroetmann Location: Sonaatti 2 Chair: Jean-Marc Battini
10:00-10:30 Coffee
10:30-12:00 Session 1B: Building Structures 2 Session 2B: Bridges
Location: Sonaatti 1 Chair: Kristo Mela Location: Sonaatti 2 Chair: Jan Bujnak
12:00-13:00 Lunch
13:00-13:45 Keynote 6 Location: Sonaatti 1 Chair: František Wald
13:45-15:00 Session 1C: Building Structures 3 Session 2C: Connections 1
Location: Sonaatti 1 Chair: František Wald Location: Sonaatti 2 Chair: Michael J. Andreassen
15:00-15:30 Coffee
15:30-17:30 Session 1D: Cold Formed Structures Session 2D: Connections 2
Location: Sonaatti 1 Chair: Jeppe Jönsson Location: Sonaatti 2 Chair: Jari Mäkinen
19:00-23:00 Conference Banquet Location: Hotel Tammer Chair: Markku Heinisuo
Date: Friday, 25/Sep/2015

8:00-8:45 Keynote 7 Location: Sonaatti 1 Chair: Mikko Malaska
8:45-10:00 Session 1E: Composite Structures Session 2E: Fire Engineering & Building Structures
Location: Sonaatti 1 Chair: Mikko Malaska Location: Sonaatti 2 Chair: Jean-Marc Battini
9:00-10:00 Session 3E: Workshop Location: Aaria Chair: Ludovic Fülöp

10:00-10:30 Coffee
10:30-12:00 Session 1F: Sustainable Engineering Session 2F: Connections 3
Location: Sonaatti 1 Chair: Mark Bradford Location: Sonaatti 2 Chair: Kristo Mela
10:30-12:30 Session 3F: Workshop Location: Aaria Chair: Ludovic Fülöp

12:00-13:00 Lunch
13:00-15:00 Session 1G: High Strength Steel Session 2G: Stainless Steel & Connections Invited keynotes and
Location: Sonaatti 1 Chair: Richard Stroetmann Location: Sonaatti 2 Chair: Jari Mäkinen
15:00-15:30 Coffee & Ending

extended abstracts
ISBN 978-952-15-3578-9 (printed)
Proceedings of The 13th Nordic Steel Construction Conference (NSCC-2015)
ISBN 978-952-15-3579-6 (USB) Edited by Markku Heinisuo & Jari Mäkinen

Photo: Tampere-Hall
• Conception, analysis and design of a special joint for fixing lattice towers legs during testing of transmission line tower, Fabio
Paiva, Jorge Henriques, Rui C. Barros
• Generalized Block Failure, Jeppe Jönsson
• FEM simulation of a tubular KT-joint, Karol Bzdawka, Jolanta Baczkiewicz
Session 1D: Cold Formed Structures 15:30 – 17:30 Session Chair: Jeppe Jönsson
• Bearing capacity of cold-formed unlipped channels with restrained flanges - EOF and IOF load cases, Mahen Mahendran,
Balasubramaniam Janarthanan, Shanmuganathan Gunalan
• Elastic buckling of an I-beam with sandwich flanges, Krzysztof Magnucki, Piotr Paczos
• A numerical parametric study on the load carrying behaviour under bending of honeycomb girders made of trapezoidal
corrugated steel sheets, Tobias Petersen, Manuel Krahwinkel
• Elastic Buckling Characteristics of Corrugated Tank under Fundamental Load, Yoshifumi Yokoyama, Kikuo Ikarashi
• Buckling strength of light-gauge members with large openings, Atsushi Sato, Seiji Mori, Tetsuro Ono, Kazunori Fujihashi
• Experimental and numerical investigations of the steel storage rack uprights, Zhong Ren, Xianzhong Zhao, Ru Qin
• Experimental investigation on the behavior of perforated steel storage rack columns under axial compression, Bassel El Kadi,
Guven Kiymaz, Atakan Mangir
Session 2D: Connections 2 15:30 – 17:30 Session Chair: Jari Mäkinen

• Monotonic behaviour of bolted T-stubs: a refined theoretical model for flange yirlding and bolt fracture failure mode, Antonella
Francavilla, Massimo Latour, Vincenzo Piluso, Gianvittorio Rizzano
• Different coating systems for the application in slipresistant connections, Stranghöner, Afzali, Berg, Schiborr, Rudolf, Berger
• Influence of different testing criteria on the slip factor of slip-resistant connections, N. Stranghöner, N. Afzali, Jörn Berg, M.
Schiborr, F. Bijlaard, N. Gresnigt, P. de Vries, R. Glienke, A. Ebert
• Simplified model for connections of steel structures in OpenSees, Ricardo Costa, Filippo Gentili, Luis Simões da Silva
• Design approach for stability check of members with hanging-profile connections, Dasu Liu
• Reasons for Charles de Gaulle airport collapse, Toomas Kaljas
• Investigations on the behaviour of threaded and shank bolts under combined tension and shear, Anja Renner, Jörg Lange
Presentations Friday 25/Sep/2015

Keynote 7 8:00 – 8:45 Session Chair: Mikko Malaska
• Energy-Efficient Solutions for Steel Structures – Case Study of Nearly Zero-Energy Building, Jyrki Kesti
Session 1E: Composite Structures 8:45 – 10:00 Session Chair: Mikko Malaska
• Behavior improvement of pultruded frp beam-column bolted connections, Ossama El Hosseiny, Hassan Maaly, Saeed Ibrahim
• Material strength effect on the behaviour of steel-concrete composite elements, Janis Brauns
• Vibration response of USFB composite floors, Richard Kansinally, Konstantinos Tsavdaridis
• Analyses of the load bearing behaviour of deep-embedded concrete dowels, CoSFB, Matthias Braun, R Obiala, C Odenbreit
Session 2E: Fire Engineering & Building Structures 8:45 – 10:00 Session Chair: J-M Battini
• Evaluation of axial force impact on the flexibility of a steel beam-to-beam end-plate joint subjected to bending when exposed to
fire, Mariusz Maslak, Malgorzata Snela
• Fire design of CFST columns - Improvements required for Eurocode 4, Matti V. Leskela
• Calculation of steel temperature in open cross sections based on fire exposure from CFD, Joakim Sandström, Wickström Ulf
• Lateral torsional buckling resistance – a comparison of analytical and numerical models, Rebekka Ebel, Markus Knobloch
Session 1F: Sustainable Engineering 10:30 – 12:00 Session Chair: Mark Bradford
• Fatigue life improvement of welded bridge details using high frequency mechanical impact (HFMI) treatment, Poja Shams
Hakimi, Andrea Mosiello, Konstantinos Kostakakis, Mohammad Al-Emrani.
• New developments in heavy plate production for modern steel construction, Tobias Lehnert, Falko Schröter
• Stainless steel, a sustainable material for sustainable structures, Anders Finnås, Camilla Kaplin
• Dynamic responce of pipe rack steel structures subjected to explosion loads, Anton Stade Aarønæs, H Nilsson, N Neumann
• Tall ambitions – onshore wind turbine tower - concepts for large hub heights, Martin Jespersen, M Nielsen, U Stottrup-Andersen
Session 2F: Connections 3 10:30 – 12:00 Session Chair: Kristo Mela

• Lateral stability of verandas by means of the glass panels, Maarten Fortan, Jesse De Clercq, Marc Meeus, Barbara Rossi
• End Plate Connection for Rectangular Hollow Section in Bending, Arne Aalberg, Arne Martin Uhre, Per Kristian Larsen
• Structural analysis models of steel trusses, Teemu Tiainen, Kristo Mela, Timo Jokinen, Markku Heinisuo
• Buckling of members of welded tubular truss, Markku Heinisuo, Äli Haakana
Session 1G: High Strength Steel 13:00 – 15:00 Session Chair: Richard Stroetmann
• Bendability and microstructure of OPTIM® 700 MC plus, Vili Matias Kesti, Antti Kaijalainen, Juho Mourujärvi, Raimo Ruoppa
• Experimental behaviour of tension plates with centre hole made from high strenght steel, Pál Turán, László Horváth
• Derivation of strain requirements for high strength steel using Johnson Cook model, Simon Schaffrath, N Schillo, M Feldmann
• Buckling strength of HSS beams, Mark Andrew Bradford
• True stress-strain relationship for finite element simulations of structural details under diffuse necking, Petr Hradil, Asko Talja
• Calibration of the ductile damage material model parameters for a high strength steel, Pavlovic, Manoleas, Veljkovic, Koltsakis
• Buckling observation of door openings for wind turbine towers, Anh Tuan Tran, M Veljkovic, C Rebelo, L Simões da Silva
Session 2G: Stainless Steel & Connections 13:00 – 15:00 Session Chair: Jari Mäkinen
• Extension of the continuous strength method to the determination of shear resistance, Najib George Saliba, Leroy Gardner 235
• Stainless steel at slightly elevated temperatures, Hans L. Groth, Erik Schedin, Emma Jacobsen, Rita Lindström
• New steel damper with displacement dependent recentering for seismic protection of structures, Murat Dicleli, Ali Salem Milani
• Fretting fatigue phenomenon in bolted high-strength steel plate connections, Olli-Pekka S. Hämäläinen, Timo J. Björk
• Comparison of relative volumes of different type of welds, Juha Kukkonen, Markku Heinisuo
• Investigation of cold formed steel beam to column bolted gusset plate connections, Žilvinas Bučmys, Alfonsas Daniūnas
• Resistance results for the crocodile connection, Panagiotis Manoleas, Kristoffer Öhman, Efthymios Koltsakis, Milan Veljkovic

Proceeding of 13th Nordic Steel 2015 Construction Conference PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Proceeding of 13th Nordic Steel 2015 Construction Conference PDF

Uploaded by

Copyright:

Available Formats

Proceedings of

The 13th Nordic Steel Construction Conference (NSCC-2015)

Edited by Markku Heinisuo & Jari Mäkinen

Tampere University of Technology. Department of Civil Engineering

Local organizing committee

ISBN 978-952-15-3578-9 (printed)

Tampere, September 2015

BIM IN STRUCTURAL STEEL WORKFLOW

1.1. The challenges of the construction industry

Worldwide vertical construction is a $3.5 trillion industry (excluding residential construction).

1.2. Workflow compatibility

Construction project parties should be able to work together

It all comes down to enabling an end-result-friendly, build-

Challenges of the current process:

3. Moving on – Assembly life cycle management

Figure 4. FICEP scribing technology for layout marking

3.1. Automation in steel processing and assembly phase

Figure 5. PEMA robotic beam welding

Figure 6. Zeman, automatic robotic steel beam assembler

Figure 7. FICEP automatic steel fabrication and software

The IFC based process:

The productivity of each design discipline ha significantly improved. However similar

Making Sustainability Activities a Key to Your Success

2 LCA as a Sustainability Tool in today’s Standards

Fig. 1. Modularity for Building Assessment Information used by CEN/TC 350.

Furthermore, data on resource use should be given:

2.3 The Battle of Materials

2.4. Social and Economic Factors

3 Trends in Sustainability Demands and its Impact on Standards

3.2 EC - PEF, Circular Economy and Construction Demolition Waste

3.3 Greenwash Awareness

3.4 Towards Sustainability Product Declarations, SPDs

4 Success in the Steel Building Business

4.2 Steel Constructions Competing with Other Construction Types

4.3 Benefits and Success Factors

Standards mentioned in text

Abbreviations used in text

COMPONENT METHOD AS A GENERAL TOOL FOR THE DESIGN

Jean-Pierre Jasparta and Jean-François Demonceaub

2 The component method

2.1 Introduction to the component method

Fig. 1: Joints with end-plate connections

Table 1: Schematic illustration of the component method

Joints with I or H sections under static loading and

2.1 Extending of the scope of the component method

3 Composite joints under sagging moment

Sagging moment Load distribution in a composite connection under

Fig. 2: Composite joint subjected to sagging moment

4 Joints under combined bending moments and axial forces

Column web panel Connection

Fig. 3: Mechanical model as a reference used in the proposed analytical procedures

For M − N interaction, two particular effects have also to be considered:

6 Joints under cyclic loading

Fig. 6: Full strength optimised beam-to-column joint solution

7 Joints under exceptional events

-20 Analytical prediction_plastic resistance curve

-30 Analytical prediction_ultimate resistance curve

Fig. 7: M − N plastic and ultimate resistance curves for a composite joint

[3] Heyman J. “Structural analysis, a historical approach”, Cambridge University Press,

EXECUTION OF STEEL STRUCTURES

Table 1: Table of contents of prEN 1090-2: 2014

2 Development of international standards for execution of steel structures

2.2 Previous ENV-standard for execution of steel structures