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

Transcription

1

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

3 The 13th Nordic Steel Construction Conference (NSCC-2015) 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 (printed) ISBN (USB) ii

4 Preface The 13 th 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 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 The last time the conference was organized in Finland was in 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

5 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 Component method as a general tool for the design of joints under various loading conditions Jean-Pierre Jaspart Execution of steel structures - recent developments and future trend Bjørn Aasen Use of higher strength steel in construction, opportunities and obstacles Milan Veljkovic Fire design of steel structures with intumescent coating Peter Schaumann Energy-efficient solutions for steel structures case study of nearly zero-energy building Jyrki Kesti Plenary Session A Joint and column behaviour of slotted cold-formed steel studs Michael Andreassen, Jeppe Jönsson Steel solutions for enabling zero energy buildings Bernd Döring, Vitali Reger, Markus Kuhnhenne, Jyrki Kesti, Mark Lawson, Andrea Botti, Markus 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, Luboš Šabatka, Jaromír Kabeláč, Lukáš Gödrich, Marta Kurejková comparative evaluation of steel profiles in roof trusses Kristo Mela, Hilkka Ronni, Markku Heinisuo Plenary Session B Non-linear finite element modelling of steel-concrete-steel members in bending and shear, Marc Donnadieu, Ludovic Alexandru 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 Session 1A: Building Structures 1 Practical tubular truss optimization Jussi Jalkanen The impact of joint constraints on the optimal design of truss structures Roxane Van Mellaert, Geert Lombaert, Mattias Schevenels iv

6 Lateral buckling stress for H-shaped beams with continuous braces Yoshihiro Kimura, Yuki Yoshino Industrial Hall Constructions Nico Genge, Christian Remde, Klaus 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 Low cycle fatigue performance of integral bridge steel H-piles under seismic displacement reveals Murat Dicleli, Memduh Karalar System reliability analysis of steel railway bridge based on historic rolling stock records Gunnstein Frøseth, Anders 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, Murat Dicleli Session 1B: Building Structures 2 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, Kikuo 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 Effect of longitudinal stiffeners on the flanges to improve the low cycle fatigue performance of steel H-piles Memduh Karalar, Murat Dicleli Refined fatigue assessment of an existing steel bridge John Leander, Raid Karoumi Odins Bridge Kjeld Thomsen, Trygve Friedrichsen High-performance-steel girder of compact section Eiki Yamaguchi, Yuji Sugimura, Kenjiro Ohmichi Steel Bridge Technology used in Buildings Hans Exner v

7 Sundsvall Bridge Kjeld Thomsen, Helge Pedersen, Trygve Friedrichsen Session 1C: Building Structures 3 Aluminium deployment in bracing systems: Investigation of shear link application Evangelos Efthymiou, Vasileios G. Psomiadis, Alexios T. 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 Kimmo Ylinen, Wei Lu, Jari Puttonen Session 2C: Connections 1 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 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 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 vi

8 Different coating systems for the application in slipresistant connections N. Stranghöner, Nariman Afzali, J. Berg, M. Schiborr, A. Rudolf, S. 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 Session 1E: Composite Structures 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, Renata Obiala, Christoph Odenbreit 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 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 Innovative Construction of Student Residences Pedro Andrade, Milan Veljkovic, John Lundholm, Tim Heistermann 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 New developments in heavy plate production for modern steel construction Tobias Dr. Lehnert, Falko Dr. 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, Hanna Nilsson, Nicolas Neumann Tall ambitions onshore wind turbine tower - concepts for large hub heights Martin Jespersen, Mogens Nielsen, Ulrik Stottrup-Andersen vii

9 Session 2F: Connections 3 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 Behaviour of a novel column-splice joint, Finger Connection Pedro Andrade, M. Pavlović, C. Heistermann, M. Veljkovic, T. Heistermann 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 Bendability and microstructure of OPTIM 700 MC plus Vili Matias Kesti, Antti Juhani 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, Nicole Schillo, Markus 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 Marko Pavlovic, Panagiotis Manoleas, Milan Veljkovic, Efthymios Koltsakis Buckling observation of door openings for wind turbine towers Anh Tuan Tran, Milan Veljkovic, Carlos Rebelo, Luis Simões da Silva Session 2G: Stainless Steel & Connections Extension of the continuous strength method to the determination of shear resistance Najib George Saliba, Leroy Gardner 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 viii

10 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 BIM IN STRUCTURAL STEEL WORKFLOW Saku Järvinen Tekla Oy 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 models contain precise geometry and data needed to support the construction, fabrication, and procurement 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 otherwise 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 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

11 ~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 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 between two software packages. Also manufacturing calls for open, readable data. Figure 2. "Open BIM supports a transparent, open workflow, allowing project members to participate regardless of the software tools they use." It all comes down to enabling an end-result-friendly, building-quality-assuring, standardized cooperation process: getting all project parties to work together by using Open BIM. Development of the IFC standard has brought steel fabrication into the Open BIM workflow, which is the way the whole industry needs to go in pursuit of balanced and sustainable development. 2

12 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 machines 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 eliminating 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 receiving 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

13 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 layout 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

14 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 reportbased interfacing with ERP systems, improving information flow to production 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. Welding 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 involving robotics. Figure 5. PEMA robotic beam welding 5

15 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 production planning and management. Based on the Building information in the models these software 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 simulations 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

16 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 information 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 future 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 stakeholders 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 proprietary 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

17 4. Steel fabrication joins the BIM workflow While the industry was and still is trying to solve the building information needs with proprietary 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 fabrication. IFC has become the industry-standard neutral file format for exchanging Building Information Models between construction disciplines from structural steel to concrete, mechanical, electrical, 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 Construction) 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

18 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 Web technology enables more realistic distributed building information management 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 buildingrelated 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

19 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

20 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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 increasing and numerous initiatives have been taken by the European Commission to decrease the environmental impact and improve the resource efficiency and social performance in Europe. 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 every part of the organisation. Communication should be based on real achievements and greenwash 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 leisure 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 recycling 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 environment 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 compared. In the standard ISO LCA methodology was standardized and a number of issues dealt with but the LCA practitioners still had freedom to make studies in different ways making them difficult to compare. 11

21 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 similar rules for building products throughout Europe. This was the start of the technical committee 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 include civil engineering works and this standard will probably be ready by Some of the tools outlined in the standards will probably be necessary for companies to understand 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 products 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 abbreviations so a look at the lists at the end of the paper might help the reader. It is also important 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 separately 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 scenarios. Examples of scenarios are recurring maintenance, replacement of windows and refurbishment of the entire building. The standard includes necessary rules for service life, functional 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 databases exist. 12

22 Nordic Steel Construction Conference 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 according 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 After having it reviewed by a third party for compliance with the standard according to ISO 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 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

23 4 Nordic Steel Construction Conference 2015 The CEN/TC 350 standard EN could be seen as a super PCR for building materials that the certification schemes follow when developing their PCRs. The different building materials like concrete, timber and aluminium all have standards developed within the CEN system based on EN 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 support 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 module 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 CO 2 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 Gases, GHG. This will however, in the future, depend on how end of waste criteria are interpreted, 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 CO 2 storage. If module 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 incineration where the CO 2 is again emitted 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 before 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 series for environmental 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

24 Nordic Steel Construction Conference The EC puts strong pressure on CEN/TC 350 to add a number of impact categories to the environmental 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 toxicity 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 investigating 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 sustainability 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 industry participates in different pilot projects. The metals industries have for example contributed 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 early 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

25 6 Nordic Steel Construction Conference 2015 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. Rumours have it that there is an office building in London that was not possible to reach by bicycle but had bicycle racks since it gave inexpensive points in the rating system for sustainability 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 impacts like in the Stora Enso case above. With standards that cover environmental impacts together 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 standard 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 improvement 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 material 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

26 Nordic Steel Construction Conference a common EPD. Examples are steels that are covered by the same EPD but differ in alloy content, 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 scenarios. It is also possible that it affects the service life time. These properties could also 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 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 reduced 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

27 8 Nordic Steel Construction Conference 2015 Strengthened relations with employees, suppliers and existing and potential customers: 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 suggested 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 creating 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 consuming and costly. However, most of the data can be retrieved from internal systems Continuous improvement: Change takes time. Especially if many people are involved. 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 initiative. 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 existing 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 necessary and starting a must. In my mind there is no success possible in sheer compliance. Commitment 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 further 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

28 Nordic Steel Construction Conference This paper has referred to standards and they reflect in many ways the state of the art of application of methods and affect how these methods are used in the future. It is therefore important to understand the political dimension of standardisation. Standards are normally developed 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 present 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 expected 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 situation. 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 -- Principles 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 buildings - Calculation methods 19

29 10 Nordic Steel Construction Conference 2015 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] [2] [3] [4] [5] [6] [7] [8] Svenska Dagbladet Näringsliv (In Swedish) 20

30 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 COMPONENT METHOD AS A GENERAL TOOL FOR THE DESIGN OF JOINTS UNDER VARIOUS LOADING CONDITIONS Jean-Pierre Jaspart a and Jean-François Demonceau b 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 Eurocodes 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 procedures 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 composite 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

31 2 Nordic Steel Construction Conference 2015 bending, under static loading at room temperature. Its application in daily practice is facilitated 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 conformity 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 originality of the component method is to consider any joint as a set of individual basic components. 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. F b F b2 F b1 θ c z M b M b2 z z M b1 θ b F b = M / z b F b F = M / z F = M / z b2 b2 b1 b1 (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 compression, 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 individual 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 resistance characteristics of the whole joint. The assembly procedure is the step where the mechanical properties of the whole joint are derived from those of all the individual constituent components. That requires, according to the 22

32 Nordic Steel Construction Conference 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 Components Assembly Types of joints Ft,Ed Ft,Ed 2.1 Extending of the scope of the component method In EN [1] and EN [2], guidelines on how to apply the component method for the evaluation of the initial stiffness and the design moment resistance of steel and composite 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 ) 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 joints between open beam sections and tubular column sections. 23

33 4 Nordic Steel Construction Conference 2015 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 stiffness have been proposed ([5] to [8]). They aim at defining a rectangular cross section of concrete 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 significant number of composite joint configurations under a new specific loading (sagging moments). This demonstrates the flexibility and the adaptability of the component method. fck,actual 150 Z IPE HEB FRd,3 FRd,2 Sagging moment FRd,1 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 Ed 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 influence 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 acting in the joint remains lower than 5% of the axial design resistance N pl, Rd of the connected beam. Under this limit it is considered that the rotational response 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

34 Nordic Steel Construction Conference 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 applied to the whole joint but a new assembly procedure is required to cover the combined action 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 mechanical 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 extensional spring characterised by a non-linear F curve, where F and represent respectively 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 N z γ CWT CFB BT EPB BWT ϕ M N CWC BFC Column web panel in shear 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: 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, longitudinal stresses due to axial and bending forces present in the column and transversal Joint 25

35 6 Nordic Steel Construction Conference 2015 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 plastic 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 nonductile components. It is assumed in [1] that the deformation capacity around a bolt is sufficient if the design resistance of the plate-bolt assembly is lower or equal to 95% of the tension 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) N [kn] M24 HR ductile response non ductile response Row 1 Upper row=2 Row 3 Row 4 Row 5 Row 6 Lower row=7 HEB 400 S355 af = 14 mm aw = 8 mm 300X780X20 IPE 600 S355 α=20 S355 Fig. 4 - Bolted joint with numbering of force transfer rows M [knm] Fig.5 - Ductile and non-ductile M-N interaction 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 characterise 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 obtained by multiplying the corresponding properties derived at room temperature by ad-hoc reduction factors (respectively ke; 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 temperatureθ. 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

36 Nordic Steel Construction Conference the tests used as references, temperaturesθ i have been measured in all components i, (ii) the stiffness and resistance properties of the components i have been multiplied by ad-hoc reduction factors k θ and k θ and (iii) the assembly of the components has been finally achieves so E; i y ; i 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 typology 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 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 limits 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 developed 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. Indeed, 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

37 8 Nordic Steel Construction Conference 2015 ing conditions. In particular, it is necessary to know the post-yielding behaviour of the components accounting for the strain-hardening effects, their ultimate resistance, their deformations capacity but also the degradation of their strength and stiffness under the applied cycles associated to phenomena of oligo-cyclic fatigue. In addition, it is needed to know the unloading 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 numerical 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 combined 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 considered in the definition of the design loads and load combination cases. Under such exceptional actions, the structural elements and in particular the joints are generally subjected to loadings not initially foreseen through the ULS design. For instance, if the exceptional 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 occurs in the joints located just above the damaged column. Finally, the joints could be subjected to some dynamic effects if the column loss is for instance induced by an impact or an explosion. In Section 4, it has been shown how the M N plastic resistant curve of a joint can be predicted 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. However, 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

38 Nordic Steel Construction Conference 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 resistant 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 experimental results (Fig. 7). HOGGING MOMENTS M [kn] TENSION N [kn] SAGGING MOMENTS Analytical prediction_plastic resistance curve Analytical prediction_ultimate resistance curve Experimental results_test 1 Experimental results_test 2 Experimental results_test 3 Experimental results_test 4 Experimental results_test 5 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 capacity 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 composite 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 Eurocode 3: Design of steel structures Part 1-8: Design of joints. European committee for standardization, May [2] EN : 2004, Eurocode 4: Design of composite steel and concrete structures Part 1-1: General rules and rules for buildings, European committee for standardization, December

39 10 Nordic Steel Construction Conference 2015 [3] Heyman J. Structural analysis, a historical approach, Cambridge University Press, [4] Jaspart J.P. and Weynand K. Design hollow section joints using the component method to appear in the Proceedings in the 15 th International Symposium on Tubular Structures, Rio (Brazil), May 27-29, [5] Ferrario F. Analysis and modelling of the seismic behaviour of high ductility steelconcrete composite structures, PhD thesis presented at Trento University, [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 , [7] Demonceau, J.-F, Steel and composite building frames: sway response under conventional loading and development of membranar effects in beams further to an exceptional action, PhD thesis presented at Liège University, 2008 (freely downloadable at [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 , [9] Cerfontaine F. and Jaspart J.P. Resistance of joints submitted to combined axial force and bending, Proceedings of the Eurosteel 2005 Conference, Maastricht, [10] Cerfontaine F. Etude de l'interaction entre moment de flexion et effort normal dans les assemblages boulonnés, Université de Liège, PhD Thesis. [11] Simoes da Silva L., Santiago A. and Villa Real P. A component model for the behaviour of steel joints at elevated temperatures, Journal of constructional steel research, Vol. 57, pp , [12] EN Eurocode 3: Design of steel structures Part 1-2: General rules Structural fire design. European committee for standardization, April [13] Demonceau J.F., Haremza C., Jaspart J.P., Santiago A. and Simoes da Silva L. Composite joints under M-N at elevated temperatures, Proceedings of the Composite Construction VII conference, Palm Cove, Australia, [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, International Journal of Steel Structures, Vol. 9(4), pp , [15] Franssen J.M. SAFIR A thermal/structural program modelling structures under fire, Engineering Journal, AISC, pp , [16] Hoang V. L., Jaspart J.-P. and Demonceau J.-F. Hammer head beam solution for beamto-column joints in seismic resistant building frames, Journal of constructional steel research, Vol. 103, pp , [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 ). [18] Nogueiro, P., Simoes da Silva, L., Bento, R. and Simoes, R. Calibration of model parameters for the cyclic response of end-plate beam-to-column Steel-concrete composite joints, Steel and Composite Structures Journal, Vol. 9, N 1, pp , [19] Jaspart, J.-P. and Demonceau J.-F. Composite joints in robust building frames, Proceedings of the Composite Construction in Steel and Concrete Conference, Colorado, USA, July

40 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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 , is briefly described. Some problems that were discovered during the preparation of the standard are also included. Furthermore, EN 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 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 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 : 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

41 2 Nordic Steel Construction Conference 2015 Table 1 (continued) A (N) 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 coatings 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 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 structures, 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 implemented in 1999; where ISO 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 Communities, see the first version of Eurocode 3 [4]. These design codes were intended to establish 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 interference 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 management 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

42 Nordic Steel Construction Conference 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 pedestrian 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 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 : 1. The draft proposal of EN had no limitation on the carbon equivalent value, CEV. Welding experts of WG 3 pointed out that difficulty of writing the welding procedures (WPS) could occur prior to any material deliveries. It was proposed that ENV should include restrictions on maximum CEVvalues, 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 maximum 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 certification of steel workshops to be outside the scope ENV The discussion was settled by an informal Annex E in ENV : 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 inspection including weld acceptance criteria of steel structures subjected to static and dynamic actions; reference is made to ISO Mr. Lindewald (Finland) argued strongly in favour of the EN ISO 5817 prenv was issued including both the British and the ISO approaches. A large majority of CEN member countries voted for the ISO standard. 33

43 4 Nordic Steel Construction Conference 2015 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 But in practice there was no conflict of interests because Mr. Allen (UK), was the Convener for EN 1011 and also member of WG Present EN-standard for execution of steel structures Eurocode 3 was transferred from the EU Commission to CEN and published in 1993 as a European Prestandard ENV 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 , EN , EN and EN Thus, it became necessary to convert ENV 1090 by merging Parts 1-6 together in one standard, namely EN In order to comply with the CPD a standard for conformity assessment of structural components was prepared, i.e. EN 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 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 classes (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 component. 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

44 Nordic Steel Construction Conference 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 : Annex B. This annex was made informative, pending on new guidelines to be included in Eurocode 3. An amendment to EN : 2005/A1:2014 was published and should at now have been implemented in design practice. The selection of execution class is given in Table 2. Table 2: Choice of execution class (EXC) according to EN [5] Type of loading Static, quasi-static or Reliability Class (RC) or Consequence Class (CC) (EN 1990) Fatigue b or seismic DCM or DCH a seismic DCL a RC3or CC3 EXC3 EXC3 c RC2 or CC2 EXC2 EXC3 RC1 or CC1 EXC1 EXC2 Notes a. Seismic classes according to EN : Low = DC; Medium = DCM; High = DCH b. See EN c. EXC4 may be specified There is still a concern that EN may lead to increased costs for 'normal' steel structures. If EXC1 is selected for the whole steel structure, then EN : 2005/A1:2014 recommends 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 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 : 2008+A1:2011. Unfortunately, some discussions 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 paradox 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

45 6 Nordic Steel Construction Conference 2015 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 B C D A & B C, D & E Slag inclusions No. 301 h 2 mm l 25 mm h 3 mm l 50 mm h 4 mm l 75 mm h 6 mm l 50 mm h 6 mm l 100 mm Lack of fusion No. 401 Not permitted Not permitted h 4 mm l 25 mm l 50 mm Single pore No d 3 mm d 4 mm d 5 mm d 6 mm Regarding bolting assemblies for non-preloaded bolted connections EN refers only to EN As an alternative the more expensive high strength bolts and nuts according to EN may be used. EN is a harmonized standard that requires both bolts and nuts to be tested together and supplied by one producer, while EN ISO and EN ISO describe independent testing of bolt and nuts with no corrosion protection, respectively. Unfortunately, the manufacturers did not produce bolting assemblies according to EN Therefore bolting assemblies according to outdated DIN standards were still specified, at least in Norway. But from the 1 st of July 2014 CE marked bolting assemblies have also become mandatory forcing the bolt manufacturers to provide products with Declaration of Performance (DoP) according to the CPR which has replaced the CPD. Besides, the revised versions of EN have been submitted for public review. According to pren the well-known ISO bolts and nuts have received first priority. These are good news for all parties involved in design, specification and execution of 'normal' steel structures! The most controversial changes in EN 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 assumption 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 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 conclusion is that the increased tolerance limit for compression members according to EN 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 γ M1 -value. However, an intermediate solution could be to specify class 2 of functional manufacturing tolerance, see EN Revised EN standard for execution of steel structures As mentioned in the Introduction the revised EN has been submitted for public review. Now it is important to gather most of the experiences that have been obtained during 36

46 Nordic Steel Construction Conference practical use of the standard, see [13] and [14]. The revised EN is expected to be approved at the end of 2016 or at the beginning of 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 Future developments 3.1 Globalization of structural steel fabrication requires ISO standards Comparing the normative references as listed in ENV (1996), EN (2008/2011) and pren (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 fabrication 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 The similarity between the various parts of ISO 630 and EN 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 NDTmethods. The fabrication of steel structures for bridges, buildings, towers and masts is increasingly carried out in the Far East and exported to Europe. As shown in Table 5 the replacement 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 bolting assemblies suitable for preloading because American, Japanese and European engineers do not yet agree. 37

47 8 Nordic Steel Construction Conference 2015 Table 4: Comparison of standards for structural steels Structural steels ENV pren ISO/wd Technical delivery terms - EN ISO Non-alloy structural steels EN EN ISO Normalized weldable fine grain structural steels EN EN Thermomechanical rolled weldable fine grain EN EN ISO structural steels Structural steels with improved atmospheric EN EN ISO corrosion resistance High yield strength structural steels in the - EN ISO quenched and tempered condition Seismic-improved structural steels - - ISO Hot finished hollow sections of structural steels EN EN ISO Cold formed hollow sections of structural steels EN EN ISO Types of inspection documents EN EN ISO 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 pren ISO/wd Qualification of welders. Fusion welding EN EN ISO ISO Specification and approval of welding procedures for metallic materials Welding procedure specification EN EN ISO ISO Welding procedure tests EN EN ISO ISO Approval using approved welding consumables EN EN ISO ISO Approval related to previous experience EN EN ISO ISO Approval by a standard welding procedure EN EN ISO ISO Approval by a pre-production welding test. EN EN ISO ISO Approval testing of welding personnel for fully mechanized and automatic welding EN 1418 EN ISO ISO Welding coordination EN 719 EN ISO ISO Quality requirements for welding Fusion welding of metallic parts Guidelines for selection and use EN EN ISO ISO Comprehensive quality requirements EN EN ISO ISO Standard quality requirements EN EN ISO ISO Elementary quality requirements EN EN ISO ISO Welding Recommendations for welding of metallic materials General guidance for arc welding EN EN ISO/TR Arc welding of ferritic steels EN EN ISO/TR Notes: 1. ISO/wd is a working draft of ISO/TC 167/WG 3 2. ISO/TR is an ISO technical report 38

48 Nordic Steel Construction Conference 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: Currently, there are only national and regional (EN) standards that address steel structures. While ISO :1997 and ISO :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 exported to others, an ISO solution is urgently needed to ensure uniformity of product and safety, 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 American Welding Society (AWS) as secretariat. Table 6 shows how the work is carried out and presented. Here, EN 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 valuable input for the future work of updating EN Table 6: Contiguous working draft showing the relation between EN and NP Note: Status Conclusions Substantial progress has been made to prepare a European standard for the execution of steel structures. EN 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 has been considered as efficient basis for developing an new ISO standard for execution of steel structures. 39

49 10 Nordic Steel Construction Conference 2015 References EN standards Note: For EN and EN 1011, reference is made to Tables 4 and 5. [1] ENV : 1996 Execution of steel structures. Part 1: General rules and rules for buildings. [2] EN : 2008+A1:2011 Execution of steel structures and aluminium structures. Part 2: Technical requirements for steel structures. [3] pren : 2015 Execution of and aluminium steel structures. Part 2: Technical requirements for steel structures. [4] Eurocode 3: 1984 Common unified rules for steel structures. EUR 8849, Commission of the European Communities, Brussels 1984 [5] EN : 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 : 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 : 1999 Steel structures. Part 2: Fabrication and erection. [10] ISO/WD Steel structures Fabrication and erection. Working draft of ISO/TC 167/WG3 [11] NP 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 und DIN EN Beuth, Ernst und Sohn, Berlin 2012 [14] Tillverkning, montering och kontroll. Handbok för tillämpning av SS-EN (in Swedish), SBI Pub. No. 182, Stockholm 2010 [15] Taras, A. Column straightness requirements and stability design. Proceedings Eurosteel, 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 , 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 operations 40

50 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 Fire design of steel structures with intumescent coatings P. Schaumann a, F. Tabeling b, W. Weisheim c a,c Institute for Steel Construction, Leibniz University Hannover, [email protected] b SHL Engineering Consulting, Hannover, [email protected] 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, firstly 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 Fourier 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 requirements, 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 resistance. 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, user-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 conductivity of IC, none of the methods include the foaming process of the IC explicitly. Therefore, a 51

51 2 Nordic Steel Construction Conference 2015 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 following. 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 [3]. The design 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 include the determination of the fire resistance of structural members under tension, compression 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 resistance 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, mathematical models for the prediction of the temperature of protected and unprotected structural members 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 determination of the temperature of steel structures with IC is available yet. Until now, the contribution of IC to the fire resistance of structural members can only be quantified by experimental investigations according to test methods of EN [4]. Consequently, a more userfriendly 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 structures with IC, the authors performed experimental investigations on waterborne, solventborne 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 measurement of mass loss of IC using the differential scanning calorimetry (DSC) device illustrated 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 Thermocouples Scale Temperature of IC about 500 C 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

52 Nordic Steel Construction Conference as a function of temperature, the temperature of the char structure was measured using thermocouples. 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 process of IC is influenced by the rising temperature field of the coating and vice versa. Moreover, 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 processes of IC into account, a thermal expansion coefficient α T according to Eq. (1) is implemented in Abaqus [7] considering logarithmic strains. d d d d ln ln ln i T i 0 i 0 i 0 (1) with: T : Thermal expansion coefficient (1/K) d 0 : Initial thickness of IC (mm) d i : 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 α T depends on the temperature development 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 foaming 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 coupled thermal-stress analysis. With regard to the mathematical formulation of the thermal conductivity 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 ( ) with: ( ): Porosity (-) : Thermal expansion factor (-) (2) 53

53 4 Nordic Steel Construction Conference 2015 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 λ eq 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. 3 eq ( ) ( p 4 IC 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 K 4 ) : Temperature of IC (K) IC d : p 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 thermal 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 resinbased IC the thermal conductivity of IC at room temperature conditions is assumed as λ IC = 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 d p = 1.2 mm. The equivalent thermal conductivity of IC is shown in Fig. 3b as a function of temperature, taking the temperature-dependant porosity ψ from Fig. 3a into account. Expansion factor α (-) ψ Temperature of IC ( C) α Porosity y ψ (-) a) Expansion factor and porosity b) Thermal conductivity and heat capacity Fig. 3: Thermal material properties of IC Temperature of IC ( C) 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 Conductivity λ (W/m K) C λ Heat capacity C (kj/m³ K) 54

54 Nordic Steel Construction Conference 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 ρ IC = 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 linear 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 experimental 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 development of the test specimen. According to the test setup a two-dimensional finite element model of the coated IPE 200 profile 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 [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 emissivity of the IC ε IC = 0.8 according to Tabeling [5] and convection α c = 25 W/m²K according to EN [8]. IPE 200 TC 1 TC 2 IC IPE 200 IC 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 maximum 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

55 6 Nordic Steel Construction Conference 2015 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 gradient 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 comparison 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 temperature 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 (experiment: 613 C, simulation: 574 C) between the web temperatures lies within tolerable limits. Temperature ( C) Furnace temperature ISO-834 Experiment Simulation Time (min) Fig. 6: Comparison between the measured and simulated steel temperatures of I-section profile IPE 200 (dft: 700 μm) 56

56 Nordic Steel Construction Conference 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 explicitly, leads to reliable and very promising results for the prediction of two-dimensional temperature distributions of steel structures. 5 Simplified approach The numerical simulation of the heating behaviour of the coated I-section profile IPE 200 presented 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, including the temperature-dependant foaming process of the coating, is high sophisticated. Therefore, the authors aimed to develop a simplified approach for the prediction of temperatures 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 neglected due to the thin dry film thickness. 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. q g, i a, i IC, i (4) dic, i with: q : Heat flux (W/m²) gi, : Gas temperature according to ISO-834 at time increment i ( C) ai, : Steel temperature at time increment i ( C) d : Thickness of IC at time increment i (m) IC, i 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 m c (5) a, i a a with: Q : Supplied thermal energy (J) : Change in steel temperature at time increment i (K) ai, m : a c : a IC Steel plate Thickness (mm) θ a,i Temperature ( C) Mass of the steel profile (kg) Specific heat capacity (J/kg K) θ g,i 57

57 8 Nordic Steel Construction Conference 2015 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. with: m a m a : a : V : V (7) a Mass of the steel profile (kg) Density of steel (kg/m³) 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 [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 increment t should not exceed the limit of 30 seconds. A A IC, i ( g, i a, i ) ( g, i a, i ), V ai t V t (8) d c R ( ) d c IC, i a a IC IC,0 a a with: R ( ): Thermal insulation resistance of IC (m K/W) IC d : Initial thickness of IC (m) IC,0 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 ρ a c a of the profile as well as on the thermal conductivity and the variable thickness of the coating. The factor d IC,i /λ IC,i indicates a standard for the thermal insulation resistance R IC of the coating. In spite of calculating the insulation resistance incrementally by using the expansion factor and the thermal conductivity of intumescent coatings (cf. Fig. 3), the authors developed two approximation rules given in Eq. (9) and (10). 200 R IC ( ) 2,13 for 20 C 0,0284 IC, i 520 C (9) IC, i e RIC 0,0066 IC, i e for, ( ) C 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 IC i 58

58 Nordic Steel Construction Conference results from Fig. 6. The material properties of steel are set according to EN [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 profile 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 temperature (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, simplified: 886 C, simulated: 883 C). Insul. resist. R IC (mk/w) Range 1 20 C θ IC 520 C Range C < θ IC 1000 C Temperature of IC ( C) 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 further 10 profiles composed of IPE, HEA, HEB and HEM profiles were investigated. The profiles were chosen to cover the typical range of the profile factor as shown in Fig. 9a, thus resulting in different dry film thicknesses of the coating (160 μm μm). As basis of comparison the heating behaviour of the coated profiles was simulated using the numerical approach presented in section 4. Profile factor A/V (m -1 ) IPE HEA HEB HEM Profile number (-) Temperature ( C) Simplified approach ( C) 1000 a) A/V-factor of investigated steel profiles b) Compariosn of steel temperatures Fig. 9: Verification of the simplified approach Furnace temperature Experiment Simulation Simplified approach Time (min) Overestimation 60 minutes 30 minutes + 15 % Underestimation - 15 % Numerical approach ( C) 59

59 10 Nordic Steel Construction Conference 2015 The comparison of the steel temperatures in Fig. 9b shows, that the simplified approach always 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 reliable 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 approach 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 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 [2] Staggs, J. E. J.: Thermal conductivity estimates of intumescent chars by direct numerical simulation. Fire Safety Journal Bd. 45. Wiley, 2010, P [3] EN : Eurocode 3: Design of steel structures Part 1-2: General rules Structural fire design. Brussels, Belgium: Comité Europeén de Normalisation, [4] EN : Test methods for determining the contribution to the fire resistance of structural members Part 4: Applied passive protection to steel members. Brussels, Belgium: Comité Europeén de Normalisation, [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, [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 Research report. [7] ABAQUS: Abaqus/Standard Version Pawtucket: Hibbit, Karlsson & Sorensen, Inc [8] EN : Eurocode 1: Actions on structures Part 1-2: General actions Actions on structures exposed to fire. Brussels, Belgium: Comité Europeén de Normalisation,

60 Nordic Steel Construction Conference 2015 Tampere, Finland 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 purposes 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 energyefficiency requirements by over 50 per cent can be built at a profit. Construction of the building, totalling approximately 1,500 m 2, started in May 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 energy 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 energy 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 construction, 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 customer 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 energy 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

61 2 Nordic Steel Construction Conference 2015 lated and airtight envelope in the building s walls and roof enables savings in energy requirements, and the use of solar power, day-lighting and geothermal energy harness renewable 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 m 2 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/m 2 K. 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/m 2 K. The measured airtightness of the entire building was as low as q 50 = 0.76 m 3 /h,m 2. 62

62 Nordic Steel Construction Conference 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 replaced 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 periods 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 approximately 0.8 W/m 2 K. 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

63 4 Nordic Steel Construction Conference 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 consumption 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 ventilation 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 incorporated 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 heating 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

64 Nordic Steel Construction Conference A total of 24m 2 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. Solar 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 January, 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 utilizing the low temperature of the ground rock. The principle of the system is shown in Fig 7. Fig. 7: Ground energy system 65

65 6 Nordic Steel Construction Conference Building-integrated solar electricity solution Solar power is also used in the outer walls of the building. Ruukki Construction s on-wall solar 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 m 2 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 behaviour in the long run. Some building elements are also equipped with thermal-moisture sensors 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, already 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

66 Nordic Steel Construction Conference Table 1: Initial data for simulated cases Reference nzeb Wall, U-value Roof, U-value Window, U-value Floor EPS 150, ʎ=0.034 EPS 150, ʎ=0.034 Infiltration q50 4 m 3 / m 2 h 0.87 m 3 / m 2 * AHU Heat Recovery 50% 80% Lights 15 W/ m 2 10 W/ m 2 (LED)** Domestic Hot Water 68 l/ m 2 a 68 l/ m 2 a People 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 Cooling Fans electricity (SFP=2.0) Pumps electricity Lighting DHW Total electricity: Total distr. heat: Table 3: Energy requirements and energy inputs of the reference case Energy demand Delivered energy kwh kwh/m2 kwh kwh/m2 Top-up heating Heat pump Cooling Fans electricity (SFP=2.0) Pumps electricity Lighting DHW Total electricity:

67 8 Nordic Steel Construction Conference 2015 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/m 2, 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 improved 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 included 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

68 Nordic Steel Construction Conference 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, socalled alliance agreements, in which the parties involved are bound to share responsibility 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 q 50 Air-tightness of the entire building [m 3 /(h m 3 ] SFP Seasonal Factor of Performance U-value Heat conductivity [W/(m 2 K)] References [1] EQUA Simulation AB, IDA Indoor Climate and Energy V 4.5. [Online]. Available at [2] FInZEB, project for nearly zero-energy definitions in Finland. 69

69 70

70 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 JOINT AND COLUMN BEHAVIOUR OF SLOTTED COLD-FORMED STEEL STUDS Michael J. Andreassen a and Jeppe Jönsson b a,b Technical University of Denmark, Department of Civil Engineering, Brovej, Building 118, DK-2800 Kgs. Lyngby, Denmark * Author for contact. Tel.: , [email protected] Extended abstract Slotted cold-formed steel studs are used in load bearing external plasterboard walls. The coldformed 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 behaviour 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 manufacturer. 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 influence 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

71 2 Nordic Steel Construction Conference 2015 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 stiffeners, the failure configurations is a combination of an end failure and an interactional globaldistortional 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. References Fig. 2: End failures, with web stiffener. [1] Thöyrä T. Strength of slotted steel studs, PhD thesis, Royal Institute of Technology, Department of Structural Engineering, Stockholm, Sweden, [2] Höglund T. Beräkning av slitsad tunnplåtsregel, Report 42, Royal Institute of Technology, Department of Structural Engineering, Stockholm, Sweden, [3] Borglund J., Jonsson J. Bärförmåga för slitsade stålreglar, M.Sc. thesis 84, Royal Institute of Technology, Department of Structural Engineering, Stockholm, Sweden, [4] Kesti J., Mäkeläinen P. Compression behavior of perforated steel wall studs, Light- Weight Steel and Aluminium Structures, , Finland, [5] Costa M.M. Support Strength of walls with slotted studs, M.Sc. thesis 127, Royal Institute of Technology, Department of Structural Engineering, Stockholm, Sweden,

72 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 STEEL SOLUTIONS FOR ENABLING ZERO ENERGY BUILDINGS Bernd Döring a,*, Vitali Reger b, Markus Kuhnhenne c, Jyrki Kesti d, Mark Lawson e and Andrea Botti f 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.: ; [email protected] 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 zeroenergy buildings (nzeb). This paper presents solutions for steel intensive commercial buildings 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 simulations 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 multistorey 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 using 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, transparency), 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

73 2 Nordic Steel Construction Conference 2015 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 concrete filled and reinforced as a series of T-sections. The double layer deck system (fig. 1) offers 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 / cooling 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 exchanger 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 grouted. 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 capacity 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 onsite 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 climates). 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

74 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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 , 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 , 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: M Rd = M cr + M pr 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 of the EN , 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 , with some critical observations. 75

75 2 Nordic Steel Construction Conference 2015 Then is developed an original formulation for the plastic bending resistance of composite slabs, which is more accurate than that given in EN as: - The formulation EN 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 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 The presented analysis shows that the bending resistance of composite slabs defined according to the current version of the EN 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 , 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 formulation for partial connection method is proposed. 76

76 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 FUTURE DESIGN PROCEDURE FOR STRUCTURAL CONNECTIONS IS COMPONENT BASED FINITE ELEMENT METHOD František Wald a*, Luboš Šabatka b, Jaromír Kabeláč b, Lukáš Gödrich a and Marta Kurejková a a Czech Technical University in Prague b IDEA RS * Author for contact. Tel.: ; [email protected] 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 demonstrated 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 researchers as a non-scientific matter. Two decades later it was already a widely accepted addition or even extension of experimental and theoretical work. Today computational analysis, in particular computational mechanics and fluid dynamics, is commonly used as an indispensable 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 :2005. Development of modern general-purpose software and decreasing cost of computational resources facilitate this trend. As the computational 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 computational analysis. The only way to prove correctness of simulated results is through a methodical 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 recommended. 2 Connection behaviour by CBFEM The advantages of FEM analyses of steel plates may be documented on behaviour of a welldesigned 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

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 M 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. a) b) 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 loaded 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 fitting 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. TA d) 78

78 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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. Coldformed 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 topology 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 coldformed 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

79 2 Nordic Steel Construction Conference 2015 q h 1 : 20 α L Fig. 1: Design domain of the roof trusses. For each span, the optimum topology is determined for three different ground structures. Member strength and joint geometry rules of Eurocode 3 [2, 3] are included in the problem as constraints. 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, [2] EN , Eurocode 3: Design of Steel Structures. Part 1-1: General rules and rules for buildings. CEN, [3] EN , Eurocode 3: Design of Steel Structures. Part 1-8: Design of joints. CEN, [4] Rasmussen M., Stolpe M. Global optimization of discrete truss topology design problems using a parallel cut-and-branch method. Computers and Structures, 86, ,

80 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 NON-LINEAR FINITE ELEMENT MODELLING OF STEEL- CONCRETE-STEEL MEMBERS IN BENDING AND SHEAR Marc Donnadieu a, Ludovic Fülöp b 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 difficult 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 definition of the two uni-axial failure mechanisms, tensile cracking and compressive crushing of 81

81 2 Nordic Steel Construction Conference 2015 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 double-skin sandwich beams, Journal of Constructional Steel Research, 14(3), , [2] Varma AH, Sener KC, Zhang K, Coogler K, Malushte SR, Out-of-plane shear behavior 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), ,

82 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 ASSESSMENT OF EXISTING STEEL BRIDGE STRUCTURES Jan Bujnak a,* and Richard Hlinka b a,b University of Zilina, Faculty of Civil Engineering, Slovakia * Jan Bujnak. Tel.: ; [email protected] 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 research and development, designing, construction and management. During design and execution stage, a building is obviously properly supervised. But, inspections, maintenance and structural management provide quality assurance for constructions and play either a very important role during entire and long-time exploitation. Over the past three decades, the construction 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 capacities 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 capacity is given as the weight Vn of one of six identical lorries representing the load model. Obviously only the exclusive load carrying capacity is interesting in the case of current bridge failures. It is given by maximum momentary weight Ve of a four-axle vehicle. Except of this variable action is no moving load at the bridge. Sometimes, an industry would like to transport 83

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 extraordinary 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 coefficient 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 Load carrying capacity 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 (3) UIC 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 possible remedial measures are imposing speed limits, reducing vehicular traffic, limiting for vehicle 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 configuration due to vertical post imposed imperfection is studied and retrofitting validated. Importance 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 replacement of superstructure was necessary. The comprehensive railway truss bridge evaluations 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 spectrums, damage accumulations were obtained applying Palmgren-Miner damage rule and remaining fatigue life was estimated. 84

84 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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) ; [email protected] Abstract: Within the RFCS funded research project RUOSTE it was aimed to study the effects of high strength steel properties on local buckling. The respective parameters were investigated 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 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 resistance 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 considered low. To increase the existing results and informativeness, stub column tests with varying load scenarios 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 completely 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 F u 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 effective width approach tends to underestimates the resistance of plates against local buckling in the stocky range and overestimates the resistance with increasing slenderness. 85

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 GM EC r e /r t Conclusions Fig. 1: Experimental results scaled to resistance modells GM and EC In this paper, three topics were approached: first the material properties of high strength steel compared to the requirements given in EN 1993 Part Second, the comparison of local buckling behaviour in comparison with EN 1993 Part 1 5 and third the introduction of a modified 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 performance 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 improve the safeness of resistance prediction. References [1] LIFTHIGH, RFCS project Contract No 7210 PR/379, [2] Dwight, J.B., Moxham, K.E.: Welded steel plates in compression, The Structural Engineer, Vol. 47, No.2, February [3] Schillo N., Feldmann M.: Interaction of local buckling and lateral torsional buckling of T-shaped cantilevering beams, Thin-walled structures, Vol. 81, p ,

86 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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 changing social conditions, which affect the user requirements of buildings. Resource efficiency, recyclability, life cycle costs and conservation of value, even under changing property conditions, 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 possibilities 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, arrangements 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 impact 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 comparative assessment during a building component and system optimisation, it is necessary to define 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 subsequently 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

87 2 Nordic Steel Construction Conference 2015 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=L 1 +L 2 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 distance 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 distances 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 option according to the environmental performance. Fig. 1: Comparison of masses, construction heights, ecological performances and costs for composite 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 germen 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 resource 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. Nachhaltigkeit von Stahl- und Verbundkonstruktionen bei Büro- und Verwaltungsgebäuden, Final Report, AiF project No. 373 ZBG (2014), Düsseldorf, [2] Stroetmann R, Podgorski C. Zur Nachhaltigkeit von Stahl- und Verbundkonstruktionen bei Büro- und Verwaltungsgebäuden Tragkonstruktionen Teil 1, Stahlbau, 83(4), , Ernst & Sohn,

88 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 STEEL CONSTRUCTION EXCELLENCE CENTER Jarmo Havula a, Pekka Roivio b and Markku Heinisuo c 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 companies are invited to participate in the actions of SCEC and this will be realized via the educational institutes and their facilities for experimental research. Figure 1 shows the structure of the SCEC network. Figure 1: Structure of the SCEC network. 89

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. significantly shortened R&D project lead-times. Fast prototyping and testing is a shortcut to best solutions. 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 technologies 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 mechanical 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 welded tubular HSS T-Joints, welding times and moment resistances (2015). 90

90 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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 optimization has to fulfill all the necessary design rules presented in the code of practice. If some design rule, which are taken from eurocode 3 in this study, is violated structure is not acceptable. 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 minimum cost truss is not usually the same as the minimum mass truss. 91

91 2 Nordic Steel Construction Conference 2015 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 solution 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 structures 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 locates in Southern Finland. Joints are gap type joints without extra strengthening plates. According 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 profiles 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 column size chosen as the design variables. Beside these also the eccentricities in joints are selected 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 determined 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

92 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 THE IMPACT OF JOINT CONSTRAINTS ON THE OPTIMAL DESIGN OF TRUSS STRUCTURES Roxane Van Mellaert a, Geert Lombaert b and Mattias Schevenels a 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.: [email protected]. 93

93 2 Nordic Steel Construction Conference 2015 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 reformulations for some nonlinear discrete design optimization problems. In C.A. Floudas and P.M. Pardalos, editors, Recent advances in global optimization, pages [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): , [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,

94 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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 effective 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 rigidities 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 rotational 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 continuous 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 analyses, and is evaluated using the buckling curve for Japanese standard code with the modified 95

95 2 Nordic Steel Construction Conference 2015 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, (in Japanese) P 1 P 2 l x z h o y d tw t f u P y 2 1 s P 2 b u 2 (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 k k u x u w 1 1 Lateral Springs ku l' M2 Rotational Springs k l' l' y z x M1 l Pz1 P z2 Pz3 Pz3 P z2 P z1 4 Node Element t A ( 12 d f f ) tw d1 d2 d3 6 d tw Figure 2 Numerical Analysis Model cr / y Recommendation for Limit State Deσign of Steeλ Structureσ, AIJ Deσign Standard for Steeλ Structureσ, AIJ Tangent Line ( cr / y =0.6) No Bracing Type A No Rotationaλ Bracing Type B No Rotationaλ Bracing Rotationaλ Bracing b Figure 3 Lateral Buckling Stress of H-shaped Beams with Braces 96

96 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 Industrial Hall Constructions N. Genge a, C. Remde b, K. Weynand c, J. Kuck d 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 important 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

97 2 Nordic Steel Construction Conference Design Tool Reference line or plane Fig. 1: Examples of parameterized components (PREON girder and girder support) 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 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 presents 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 process compared to traditional building solutions. 98

98 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 EFFECT OF END STIFFENER REINFORCEMENT ON LATERAL TORSIONAL BUCKLING BEHAVIOR OF H-SHAPED BEAMS WITH LARGE DEPTH THICKNESS RATIO Daiki KUBOTA a, Kikuo IKARASHI a 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 obtainable 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 reinforcements. At longer stiffener lengths, the plastic deformation capacity was further improved to that of a beam with fixed lateral displacement only (Fig. 2). 99

99 Nordic Steel Construction Conference 2015 Horizontal Stiffener Vertical Stiffener Stiffener Reinforcement Q/Q 1.5 p mm 375mm 1500mm L s =375mm L s =500mm L s =750mm Fig. 1: Loading models 750mm 1000mm L s =1000mm δ=0.44 b s /t s =4.44 L s =1500mm : Maximum load θ/θ p 5 R b s /t s =16 b s /t s =10 b s /t s =7.2 No Stiffener Q/Q 1.5 p 1.0 b s /t s =4.4 b s /t s = N-n S-375 (Only lateral displacement S-1000 N-f was fixed) 0.5 N-n N-f : Maximum load S-375 S-1000 L=2000mm θ/θ p Fig. 2: Non-dimensional load deformation relationship Fig. 3: Non-dimensional load a) L s =500(mm) b) L s =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 deformation 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 lateral torsional buckling, while considering the rigidity of stiffeners. The evaluation was conducted 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 End Plate L=2000(mm) N-n B Ls=375 or 1000(mm) B' ts θ S-375, S-1000 tf PL Stiffener (Both Sides) bs=40(mm) 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 displacement 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. d 375(mm) 9(mm) Q 100(mm) B-B' section δ 5 R R = θ max θ p 1 No Stiffener δ = 2b st s dt w δ 100

100 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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) ; [email protected], Abstract Under the effect of medium and large intensity ground motions, the seismically-induced lateral 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. Accordingly, 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 motions 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 deformations 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

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 elimination 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 foundation 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 components 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 performance of IBs under thermal loads, live load distribution among components of IBs and soilstructure 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 tensile 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 effect of bending. Local buckling increases the local curvature and strains. This locally accelerates the cracking of the material. Furthermore, IBs with shorter and thinner abutments supported 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 analysis approach. J. Bridge Eng., 10 (3 ), [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): [3] Kalayci E, Civjan SA, Brena SF (2012) Parametric study on the thermal response of curved integral abutment bridges. Engineering Structures, 43:

102 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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 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 , , are obtained through finite element simulations of trains passing the bridge. The trains are sampled from rolling stock in a similar method to [1] by considering 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

103 2 Nordic Steel Construction Conference 2015 actual and calculated stress. Damage is evaluated by Miner s rule. The limit state function for a component becomes g(x) = X m m C [ qi n Q,i S m ] eq,i 0 (1) i where is the damage at failure, X m is the modeling uncertainty, C and m are the intercept and slope parameters of the SN-curve, q i n Q,i is the number of cycles and S eq,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 [2] Alessio Pipinato and Claudio Modena. Structural Analysis and Fatigue Reliability Assessment of the Paderno Bridge. Practice Periodical on Structural Design and Construction, 15(2): , May [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): ,

104 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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.: ; [email protected] 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 considered 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 immediately for railway traffic for more than a month. After these incidents, both bridges are monitored regularly and will be kept in service until planned replacement in

105 106

106 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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 superstructure 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 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 bearings 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 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 stiffness was applied to normal stiffness of the slab, while flexural stiffness of the slab stayed unchanged. A small part of comparison of the numerically obtained values with those observed during measurement is shown in Fig

107 2 Nordic Steel Construction Conference Results and conclusions Deflection in 3rd span [mm] left girder LC - 4 right girder LC - 3 Heightof steel girder [mm] Left girder LC Right girder 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 reduced 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 Republic under the project No. 1/0583/14 and by the Slovak Research and Development Agency under the contract No. APVV References [1] Moravčík M, Bahleda F. Static load of the composite arch bridge, Civil and Environmental Engineering, 7(1), 35-41, [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), , [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, , [4] Odrobiňák J, Vičan J, Bujňák J. Verification of composite steel-concrete bridge behaviour, Procedia Engineering (Pub.: Elsevier), 65, , [5] Odrobiňák J. Verification of Flexural Behavior and Simplified Modeling of Steel-Concrete Composite Bridge, Transactions of the VŠB Technical University of Ostrava, Civil Engineering Series (Pub.: De Gruyter), 14(1), 67-74, [6] Bujňák J, Odrobiňák J. Cracking Of Concrete Deck in Composite Structures, Proceedings of the 4th International Conference Eurosteel 2005, vol. B (Eds.: B. Hoffmeister, O. Hechler, Pub.: Verlag-Mainz), Maastricht, The Netherlands, , [7] Bujňák J, Odrobiňák J. On design of composite beams with concrete cracking, COST C12 Final Conference Proceedings, Innsbruck, Austria, ,

108 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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 Department of Engineering Sciences, Middle East Technical University * Author for contact. Tel.: +90 (312) ; [email protected], Abstract Several cycle counting methods exist in the literature for the study of fatigue dam-age generated 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 displacements on the continuous bridge deck of integral bridges and thus on the abutments, backfill 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 observed from Fig.1 (French et al. 2004). The magnitude of these cyclic displacements is a function 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

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 minimum 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 counting method is developed to determine the number and amplitude of large and small displacement/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 representative 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 fatigue 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 effect of thermal fluctuations on integral bridges. Then, 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 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 induced displacement or strain cycles, a new cycle counting method is developed. Consequently, 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 amplitude 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 performance of the steel H-piles. This confirms the earlier analytical observations. REFERENCES [1] Arsoy S., Duncan J.M., Barker R.M 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 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, Peoria, IL. [1] French C., Huang J., Shield C Behavior of concrete in-tegral abutment bridges. Final Report. 110

110 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 RESISTANCE OF ECCENTRICALLY LOADED BEAM-COLUMNS Josef Vican a,, Peter Janik b 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 compressive 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 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 beamcolumn 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 initially imperfect about both axes subjected to the compressive axial force N and bending moments M y, M z, 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 M y, M z have constant shapes. 3 Experimental investigation The main objective of the experimental investigation was to determine the actual ultimate resistance of tested pinned-fixed beam-columns subjected to the eccentrically acting compressive axial force causing the end bending moments and also to verify the correctness of standard approach to the assessment of beam-column resistance. Four sets of beam-column samples, designated as A, B, C and D according to the type and eccentricity magnitude, were investigated. Each set of samples comprised three beam-columns of 1400 mm long, made of IPE 120 introducing member relative slenderness of λ = Beam-columns of set A were tested with the zero eccentricity of axial force, samples of set B had the preliminary measured eccentricity of e y = 32.2 mm in the y-axis direction, samples of set C had the eccentricity of e z = 52.2 mm measured in the z-axis direction and beam-columns of set D had the both type of 111

111 2 Nordic Steel Construction Conference 2015 eccentricities of axial force, i.e. e y = 32.2 mm and e z = 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 locations were monitored using gauges 6/120 LY11 (HBM) and potentiometer sensors of deformations 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 N exp of tested members Designation N exp N Designation exp N Designation exp N Designation exp (kn) (kn) (kn) (kn) A B C D A B C D A B C D Average Average Average Average 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 geometrically and materially nonlinear analysis with imperfections. Effects of residual stresses were taken into account by means of equivalent geometric imperfection in accordance with EN The bilinear material model with actual yield strength f y = 300 MPa and the nominal value of the Young s elasticity modulus E = 210 GPa was used to approximate material behavior. Results of numerically calculated ultimate resistances N num of tested members A1, B1, C1 and D1 compared to experimentally determined ones N exp and compared to approach according to EN using method 1 (N 1 ) and 2 (N 2 ) are presented in Table 2. Table 2: Comparison of experimentally and numerically determined resistances to the standard ones Designation N exp N num N num / N exp N 1 N 2 N 1 /N exp. N 2 / N exp [kn] [kn] [kn] [kn] [%] [%] A B C D 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

112 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 EXPERIMENTS ON PLATE GIRDERS WITH A VERY SLENDER WEB Roland Abspoel a a Delft University of Technology, The Netherlands 1 Introduction This paper reports about the experiments carried out in the Stevin II Laboratory of Delft University of Technology in the framework of the PhD-research by the author on his investigations on the distribution of a certain amount of steel over flanges and web of an I-shaped double 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 determined by Basler [1] and given in EN [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 certain weight per unit length, is the main topic for optimization. Using higher steel grades, applying 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 restricted 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 slenderness w hw tw, related on the dimensions of the plate girder Aw hw t and w Atf btf t, tf 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 lies in-between 360 and 720 for S235 assuming a residual stress level of f. 2. w w w.max 2 A Atf fytf. fytf. 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 EN [3]. In Abspoel [4] to [7] the design of the scaled test girders is described. E r y f (1) 113

113 2 Nordic Steel Construction Conference 2015 b tf t w 3 Conclusions b bf Fig. 1: Cross section of the test girders 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 EN [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 EN [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 compressive flange much better than based on column buckling of the web; References [1] Basler, K. Strength of plate girders, PhD dissertation, Lehigh University Bethlehem, [2] Eurocode : Eurocode 3: Design of steel structures Part 1-5. Plated structural elements, European Committee for Standardization, Brussels, November [3] Eurocode : Eurocode 3. Design of steel structures Part 1-1: General rules and rules for buildings, European Committee for Standardization, Brussels, January [4] Abspoel, R. The maximum web slenderness of plate girders. Proceedings of the 5 th European conference on steel and composite structures (Eds. R. Öfner, D. Beg, R. Greiner and H. Ünterweger), Timisoara, Romania, , [5] Abspoel, R. Optimising plate girder design. Proceedings of the 11 th Nordic Steel Construction Conference (Eds., Malmö, Sweden, September, [6] Abspoel, R. and Bijlaard, F.S.K. Optimizing of plate girders, Steel Construction, 7, , [7] Abspoel, R. The maximum bending moment resistance of plate girders. Proceedings of the 7 th European conference on steel and composite structures (Eds. Raffaele Landolfo and Frederico M. Mazzolani), Naples, Italy,

114 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 EXPERIMENTAL STUDY INTO BENDING-SHEAR INTERACTION OF ROLLED I-SHAPED SECTIONS R.W.A. Dekker a, H.H. Snijder a and J. Maljaars a,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 [1], taking presence of shear into account by a reduced yield stress for the shear area. An experimental 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 evaluation 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 , 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 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 utilization 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); F exp-f theorygraph (right) 115

115 2 Nordic Steel Construction Conference 2015 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, 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 A A2a A2b A A4a A4b A B B2a B2b B B4a B4b B 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 dominated, 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 considered, but unconservative if Fy is considered as resistance for short beams (Fig. 1 right). The EN 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 The specimens were provided by ArcelorMittal and tested at the Pieter van Musschenbroek Laboratory, TU/e. 116

116 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 EFFECT OF NEUTRAL-AXIS POSITION ON THE ELASTIC BUCKLING CHARACTERISTICS OF CONTINUOUS COMPOSITE BEAMS Daigo SHIRAI a, Kikuo IKARASHI a a Department of Architecture and Building Engineering, Tokyo Institute of Technology, Tokyo, Japan 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 neutralaxis 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: Theoretical analytical model 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 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

117 25 20 K = H = Web local buckling Flange local buckling Lateral-torsional buckling 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 neutralaxis 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 relationship between and the decrease of K in local buckling caused by bending stress is determined solely by the cross-sectional shape of the H-shaped beam. Furthermore, as compared to flange local buckling, the behavior of web local buckling is more affected by. Fig. 4 shows the relationship between kd and crosssectional shape in the realistic range of from 0.5 to Here, kd is an indicator of the degree of decrease of K. The smaller kd is, the larger the amount of decrease of K is. The elastic buckling strength of composite beams can be derived using the equations in Fig. 4 and approximate formulas for the buckling strength of a pure H-shaped beam. 5 Conclusions kd ; 0.5 to 0.75 w = b16 w tw18 b f t f Fig. 4: Raletionship between k d and cross-sectional shape 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 Section which flange local buckling web local buckling is caused when is equal to 0.5 d kd ( 0.5) 1 K d ( K )

118 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 AMPLIFIED SEISMIC LOADS IN STEEL MOMENT FRAMES Bora Aksar a, Selcuk Dogru a, Bulent Akbas a*, Jay Shen b, Onur Seker b, and Rou Wen c 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.: ; [email protected] 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 (C d 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, V max, 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

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 For medium-rise frames, the maximum average o values remains below The maximum average o in the individual columns in each story of the high-rise frames varies between 1.30 and 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 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

120 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 Design rules for slim-floor girders considering the composite behaviour Univ.-Doz. Dr.-Ing. Markus Schäfer a a Research Unit in Engineering Sciences UNIVERSITY of LUXEMBURG Tel.: ; [email protected] 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 complete 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 systems. 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 associated reduction of erection costs. The low self-weight of the structure allows small crosssectional dimensions for the slab and the following components up to the foundations. In addition, by the low slab height results a reduction of the construction volume, this leads to savings 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 section, 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 connectors, realized by headed studs. Equally, the composite behaviour can be ensured by open- 121

121 2 Nordic Steel Construction Conference 2015 ings arranged in the web or upper flange, work concurrently as concrete dowels. The determination 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 transverse 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 boxsections a hybrid truss model with a compression strut in the concrete is developed, increasing 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 is not sufficient for the design. The tension stresses from the hybrid truss model are to consider for the MV-interaction. In case of openings 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 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 deformation. 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, considering the resilience of composite connectors and the crack behaviour on the structure deformation. References [1] Schäfer M. Zum Tragverhalten von Flachdecken mit integrierten hohlkastenförmigen Stahlprofilen, Dissertation, Institut für Konstruktiven Ingenieurbau, Heft 8, Bergische Universität Wuppertal,

122 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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 Department of Engineering Sciences, Middle East Technical University * Author for contact. Tel.: +90 (312) ; [email protected], 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 displacements of integral bridges are controlled by these extreme temperature changes. In integral 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 compressive strains from the axial load and high compressive plastic strains due to the effect of bending. Detail-A (a) (b) Fig. 1. (a) Pile displacement due to thermal changes, (a) Thermal Expansion, (b) Thermal contraction 123

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, 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 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 HP 220x57 Lateral Load Conclusions Figure 2. Longitudinal stiffeners and Test set up 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

124 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 REFINED FATIGUE ASSESSMENT OF AN EXISTING STEEL BRIDGE John Leander a, Raid Karoumi a 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 suggested 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 features 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 recorded 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 The total duration corresponds to about three days. The short duration of the measurements makes the estimation of a representative traffic volume 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

125 2 Nordic Steel Construction Conference 2015 n E Gauge 8 Gauge Δσ/MPa (a) Stress range spectrum. β T/years (b) Reliability index Δσ C= 125 MPa Δσ C= 80 MPa Δσ = 40 MPa C β = 3.1 β = 2.3 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 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 Δσ C = 125 MPa Δσ C = 80 MPa Δσ C = 40 MPa Deterministic FLM3 50 < 20 FLM4 67 < 20 Real > 200 < 20 FORM, β = 3.1 Real > 200 > FORM, β = 2.3 Real > 200 > 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 fatigue life even further. 3. The reliability-based assessment shows that the partial safety factors used in the deterministic 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

126 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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 designed 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 meters 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

127 Nordic Steel Construction Conference 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 Construction started in beginning of 2010 and the bridge was inaugurated June 15 th The total cost is DDK 400 mill. The Bridge has been awarded the Danish IABSE structure prize

128 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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 Standards (JIS). Because of their high yield strengths and various advantages such as good weldability, 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 conventional 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

129 2 Nordic Steel Construction Conference 2015 b f M M t f b w t w t f 5000 mm Fig. 1: Steel girder model b f C o s t Table 1: Comparison of costs of three girders Price Ratio (SBHS500/SM490Y) < 1.08 < 1.23 < 1.50 < Low SBHS500 SBHS500 Hybrid Hybrid Hybrid Hybrid SM490Y Middle Hybrid Hybrid SBHS500 SBHS500 SM490Y SM490Y Hybrid 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 maximum width-to-thickness ratio curve are compact. 4 Comparison of three girders Under the condition of Mp = 5.0 x 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 gratefully acknowledged. 130

130 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 STEEL BRIDGE TECHNOLOGY USED IN BUILDING PROJECTS Hans Exner Msc. Eng., Ph. D. in structural engineering, Ramboll Denmark Tel 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 deflections. 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 perpendicular 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

131 2 Nordic Steel Construction Conference 2015 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 carefully 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 exhibition 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 geometry 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 vibrations were avoided by damping. In the table below, the design situations are compared, and some major focus items are shown. Corrosion protection Steel bridges Minor internal protection Table 1: Properties Opera Copenhagen Maritime Museum Geometric adaption Slender roof. Large cantilever Strength Important Important Important Slopes and angles. Cantilevered support arms. Facades totally glazed Stiffness Transport and erection Differential temperature Material consumption Potential vibration. Shape optimized against wind load. Dampers. Ship or other. Site welding Potential vibration. Wind response analysed. Ship. Site welding Vibration prevented by stiffness and damping in the facades. Flexible joints for facades and ventilation Ship. Site welding Important Very important Important Self weight is the major bridge load. Good Less important 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

132 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 SUNDSVALL BRIDGE Kjeld Thomsen 1 Helge Skov Pedersen 2 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 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 superstructure. 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 girders 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

133 2 Nordic Steel Construction Conference 2015 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 Rotterdam and further on to Stettin, Poland. The bridge elements was assembled to full bridge spans of approximately 2,000 ton. The 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 The total cost is approx. 1.5 bill. SKr. 134

134 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 ALUMINIUM DEPLOYMENT IN BRACING SYSTEMS: INVESTIGATION OF SHEAR LINK APPLICATION Abstract Evangelos Efthymiou 1, 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 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. Moreover, 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 aluminium 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 metallic 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 aluminium 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 highlighted. 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 potential in such applications. The proposed hybrid EBF is designed according to EC3 and EC8 with the respective capacity-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 ductility 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. Cyclic loading is imposed and the hysteretic behavior of the aluminium link is observed stable 1 Corresponding author: Tel.: ; [email protected] 135

135 2 Nordic Steel Construction Conference 2015 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 approach exhibits reliable behavior in the inelastic range. The link under examination using aluminium 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 deployed in existing frame designs as a retrofit device. Fig. 1: Model of the proposed hybrid bracing system References Fig. 2: Base shear vs top horizontal displacement (cycling loading) [1] Mazzolani FM. Innovative Steel Structures for Seismic Protection of Buildings, Prin 2001, Polimetrica, Milan, [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 Engineering Software, 38, pp , [3] Rai DC, Wallace, BJ. Aluminium Shear-links for enhanced seismic resistance, Earthquake Engineering & Structural Dynamics, 27(4), , [4] Mazzolani FM. Aluminium Alloy Structures, 2nd Edition, E&FN SPON, Chapman & Hall, London,

136 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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.: ; [email protected] 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 individual calculations. As the dimensioning of the structural components influences the overall dynamic 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 regard 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 methodology is the gust response factor approach given in Eurocode 1. Besides this simplified approach which is formulated for a single degree of freedom system, it is also possible to consider 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 dynamic 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 supporting structures of wind turbines taking into account the global dynamic behaviour and neglecting 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

137 138

138 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 TIME HISTORY SIMULATION IN SEISMIC DESIGN Peter Knoedel a,* and Thomas Ummenhofer b KIT Steel & Lightweight Structures, Karlsruhe Institute of Technology, Germany * Author for contact. Tel.: ; [email protected] 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 according 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 nonlinearities 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 EC (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

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 elastic steel structures. 2. Still, missing the resonance frequency by 1 % might give results which are by 20 % unsafe. 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 elastic 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 documentation which includes the verification procedure. Otherwise the results should not be considered trustworthy. 140

140 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 STEEL COMPOSITE DOWELS IN CRACKED CONCRETE M. Classen a, A. Stark a 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 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 definitive 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 , 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 longitudinal reinforcement bars. By activating the cylinder, shear forces occur in the composite 141

141 connection while tensile forces are introduced into the rebars and into the slab yielding to transversal 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 reinforcement. 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 reinforcements. 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 approximately complies with the average spacing between adjoining transversal cracks srm. The considerable 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 splitting 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 services. STAHLBAU Vol. 83, Iss. 7, pp , July [2] Allgemeine bauaufsichtliche Zulassung Verbunddübelleisten, Z , 13. Deutsches Institut für Bautechnik, [3] Wurzer, O.: Zur Tragfähigkeit von Betondübel, Dissertation am Institut für Konstruktiven Ingenieurbau, Universität der Bundeswehr München,

142 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 CROSS-SECTIONAL CAPACITY OF COMPOSITE COLUMN BY THE TWO METHODS OF EN Kimmo Ylinen, Wei Lu and Jari Puttonen Department of Civil and Structural Engineering, School of Engineering, Aalto University Abstract: The options of EN [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 Octave language [2]. The material of concrete is modelled with a parabola-rectangle type of stress-strain relationship, 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 considered. 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 (M y ) and minor (M z ) axis for the cases with extreme differences between general simplified methods and general methods are presented in Fig. 1 and Fig. 2, respectively. The solid line is the interaction curve calculated with the general method; dark dotted line and light dotted line represent the simplified method with a M = 0,8 and a M = 1,0, respectively. 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 profiles for both axes bending. The results indicate that future improvements of simplified method are necessary for efficient, effective and safe design solutions. 143

143 2 Nordic Steel Construction Conference 2015 Fig. 1: Comparisons of the capacity curves in modified Case 15 to give extreme max. Notation Fig. 2: Comparisons of the capacity curves in modified Case 6 to give extreme min. b, b s Width of the concrete section and steel profile f ad, f yd Design value of the yield strength of the structural steel and reinforcement bars f cd Characteristic value of the compressive strength of the concrete h, h s Height of the concrete section and steel profile t f,, t w Thickness of the flange and web of the steel profile M, N Bending moment and Axial force a M Reduction factor for M Rd in simplified method d Steel contribution ratio according to EN D Difference between two design methods r Reinforcement ratio References [1] European Committee for Standardization. EN , Eurocode 4: Design of composite steel and concrete structures. Part 1-1: General rules and rules for buildings, [2] GNU Octave, (e),

144 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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. address: [email protected]. Phone number: 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 loading 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 quasistatic 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 explains 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 action 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 indicates 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 specimens were impacted by a trolley with a mass of kg and an initial velocity that ranged between 8 and 12 m/s in the various tests. 145

145 2 Nordic Steel Construction Conference Results a) Elevation view b) Section (A-A) view Fig. 1: Dimensions, loading, and boundary conditions of the test specimen. 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 deformation, 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 increasing the mass of the beams. In design practice, one should be aware of that extreme impulsive 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,

146 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 Design resistance of end plate splices with hollow sections Yvonne Steige a and Klaus Weynand a 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 accordance with EN Based on models that exist in literature an Eurocode conform model is presented by using the component method. The original model, based on experimental 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 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 formulas 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 available 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 with this effective length it is possible to calculate the design tension and moment 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 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

147 2 Nordic Steel Construction Conference 2015 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 Additionally, for four-sided connections, an effective length part for the corner bolts is developed. This is based on a three dimensional yield line model derived from literature, which indicates 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 F T,Rd and beam web in tension F BWT,Rd. For the determination of F T,Rd, the resistance of the individual connection side has to be calculated. 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 , 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 M j,rd is explained only for bending around the y-direction. The calculation for z-direction can be derived accordingly. M j,rd results from the effective tension resistance F tr,rd of the individual bolt rows r multiplied with the distance of the bolt row to the center of compression h r. 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 (fourand two-sided). The design moment resistance of two-sided can be calculated according to EN The first row from four-sided connections, called external row, has to be considered separately. In EN , one bolt row consists of two bolts, but in the case of RHS splice the external 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 , 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 Conclusion The presented model provides a calculation method for joints with RHS joints with bolted end plates according to EN 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 represented. This approach would also correspond to the component method and could thus be easier applied to other connections types. 148

148 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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.: ; Mobile: ; [email protected] 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

149 2 Nordic Steel Construction Conference 2015 universal base (dimension in plant 20*20 m 2 ) 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. 3 Conclusions Fig. 1: Description of the Universal base of the testing station 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, [2] Nuño J, Miller M, Kempner L. Historical Perspective of Full-Scale Latticed Steel Transmission Tower Testing. Electrical Transmission and substation structures,

150 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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 cases 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 formulas 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 f m =(f y +f u )/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 paper 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 h g and b g. The thickness of the plate is denoted by t. The related net lengths h n and b n 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. t h th n h g and t b tb n b g along the respec- 151

151 2 Nordic Steel Construction Conference 2015 tive lines. This assumption leads to a relatively simple method for determination of the capacities 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: bg hg bg h n NR tfm 2 hn, VR tfm 2 bn, MR thg fm (1) By scaling the three basic stress distributions in Fig. 2 by N N R, VV R and M MR respectively and checking the formal yield condition along all yield lines shows us that the following interaction formula needs to be fulfilled: 2 2 N M V 1 (2) NR MR VR Furthermore the paper discusses a rigid plastic upper bound method approximated with respect to bolt holes - and an experimental investigation for a large eccentricity connection. F a 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 bolted 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

152 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 FEM SIMULATION OF A TUBULAR KT-JOINT Jolanta Bączkiewicz a, Karol Bzdawka a a Poznan University of Technology * Author for contact. Tel.: ; [email protected] 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 KTjoint 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 influence the stiffness of individual connections. As a result KT-joint mustn t be treated as separate 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 C 1,C 2 and C 3 Fig. 1) and if it is influenced by the size of gaps between the diagonals g 1,2 and between the pale and chord g 2,3. The sections were 150x150x5 for the chord, 120x120x4 for the diagonals and 80x80x3 153

153 2 Nordic Steel Construction Conference 2015 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. LBL 80x80x3 LP LBR beam elements 120x120x4 120x120x g2,3 g1,2 C3 C1 rigid links C2 5 Results 150x150x5 Figure 1: View of the considered joint (left) and its beam model (right) Joint stiffnesses obtained from the analyses for each individual load and each spring connection are presented in Table 1. It was found that the pale influences the stiffnesses of both diagonals, but especially the one to which it is attached. The pale was found to have only a minor 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 separate 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, C LBR, C LP, C LP, C Acknowledgement SSAB is acknowledged for funding the research. References [1] EN , Eurocode 3: Design of steel structures - Part 1-8: Design of joints, European Committee on Standardization, Brussels, [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 Netherlands, [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, [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 Structures, Lappeenranta University of Technology. 154

154 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 BEARING CAPACITY OF COLD-FORMED UNLIPPED CHANNELS WITH RESTRAINED FLANGES - EOF AND IOF LOAD CASES Balasubramaniam Janarthanan 1, Shanmuganathan Gunalan 2 and Mahen Mahendran 3 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

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 (P cr ) and yield load (P y ) of the section. The elastic buckling loads (P cr ) 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, [email protected], Ph: , Fax:

156 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 ELASTIC BUCKLING OF AN I-BEAM WITH SANDWICH FLANGES Krzysztof Magnucki a, Piotr Paczos b 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 Tables 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 composed 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 220mm, thickness of the flanges t 10mm, thickness of the sheets t 1mm, width of the beam b 140mm, and the total area of the cross-section f 2 A mm. 2 Numerical study - Finite Strip Method and Finite Element Method The simply supported I-beam with sandwich flanges is under pure bending. Numerical calculation is realized with the use of the system CUFSM. The local and global states are determined for selected length of the beam. Local buckling of the beam: the value of the load factor loc of the critical load f for the length of the beam L 300mm, therefore the f b 157

157 2 Nordic Steel Construction Conference 2015 half-wavelength L w 300mm. Global buckling of the beam lateral buckling: the value of glob the load factor of the critical load f for the length of the beam L 4000mm, therefore the half-wavelength f L w 4000mm (Fig.2). b 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. Acknowledgments Table 2: Values of the load factors of the critical loads I-beam (loc) f f (glob) f f FSM FEM The research was conducted within the framework of Statutory Activities in

158 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 A Numerical Parametric Study on the Load Carrying Behaviour under Bending of Honeycomb Girders made of Trapezoidal Corrugated Steel Sheets Tobias Petersen a,*, Manuel Krahwinkel a a HafenCity University Hamburg, Innovative Bauweisen und Baukonstruktion, Germany * Author for contact. Tel.: ; [email protected] 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 honeycomb-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 behaviour of the connection under shear. a) TCSS profile Hoesch T b) HCG made of 2 TCSSs Hoesch T 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

159 2 Nordic Steel Construction Conference 2015 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 combination with a T TCSS. The diagrams show the maximum characteristic surface load in dependence on span of a simple beam system. The value of q k,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 / SMS01Z 4.8x20 T / PolyGrip 6.4x15 q k,max [kn/m 2 ] 5,0 4,0 3,0 2,0 1, Span [m] t N =0.88 mm t N =1.00 mm q k,max [kn/m 2 ] 5,0 4,0 3,0 2,0 1, Span [m] t N =1.25 mm t N =1.50 mm t N =0.88 mm t N =1.00 mm t N =1.25 mm Fig. 2: Example of a HCG arrangement t N =1.50 mm Conclusions The main conclusions are: 1. For 360 HCG constellations the maximum characteristic surface q k,max load was determined by using numerical simulation software. 2. The mechanical fasteners Hilti SMS01Z 4.8x20, Hilti SMD01Z 5.5x19 and Gesipa PolyGrip 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 Laboratory Tests and Simulation, 7 th European Conference on Steel and Composite Structures (Eds.: R. Landolfo, F. M. Mazzolani), Naples, Italy, , [7] Petersen T., Krahwinkel M. Der Wabenträger, 19. DASt-Forschungskolloquium (Eds.: Deutscher Ausschuss für Stahlbau DASt) Hannover, Germany, 34-39,

160 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 Elastic Buckling Characteristics of Corrugated Tank under Fundamental Load Yoshifumi YOKOYAMA a, Kikuo IKARASHI b a,b Department of Architecture and Building Engineering, Tokyo Institute of Technology, Tokyo, Japan 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 under 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 The definitions of analytical models and geometric parameters are shown in Fig. 1. The models are categorized into three groups: type-v, type-h and type-n. The elastic buckling characteristics of these models are examined under axial compression and bending shear. 3 Elastic buckling characteristics of corrugated tanks 3.1 Axial compression The elastic buckling characteristics of types V, H and N are examined under axial compression acting on the models. When models are subjected to axial compression, 高さ h h 3 (z) type-n 直径 d type-v type-h d 板厚 t shell element rigid 剛体要素 body a 波高 a 円筒半径 r r 2 (y) 波高 5 (θy) a r 波長 1 (x) t 6 (θz) O 4 (θx) シェル要素 b 板厚 t 円筒中心軸円筒中心 Fig. 1: Geometric definition of models a b t 板厚 t 波角度円筒半径 r 161

161 2 Nordic Steel Construction Conference 2015 type-v is shaped into very fine waves and the buckling stress of type-v is smaller than that of type-n in the entire analytical 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 corrugation. The buckling stress of type-h is also shown in Fig. 2. The buckling stress of type-h has the minimum value 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. 3.2 Bending shear 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 models 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. The buckling modes of type-v are shown in Fig.5, where A refers to shear mode and B refers to bending mode. As seen in Fig.4, the elastic buckling stress of type-h under either loading condition is almost the same. In models that have large a and the elastic buckling stress of type-h is larger than that of type-n. The influence of a on the elastic buckling stress of type- V in models subjected to transverse end load and antisymmetric moment is indicated in Figs. 5(a) and (b). In either loading condition in the range of small a, type-v is shaped into shear mode and the elastic buckling stress increases with increasing a. However, in the range of large a type-v is shaped into bending mode and the elastic buckling stress decreases with increasing a. Therefore, the elastic buckling stress is at its maximum value at the point that the buckling mode changes. In models under either loading condition, the elastic buckling stress of type-v is higher than that of type-n in the entire analytical range. In models subjected to antisymmetric moment, the elastic buckling stress of type-v is much higher than that of type-n. 4 Conclusion [N/mm 60 2 ] Transverse end load A A B A A B B B B N V( =300) V( =200) V( =100) [N/mm 60 2 ] Antisymmetric moment d=8595mm h=10500mm t=4.5mm d=8595mm h=10500mm t=4.5mm a [mm] a [mm] (a)transverse end load (b)antisymmetric moment 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. B B B B B B B B B B B B [N/mm ] Fig. 2: Influence of height of corrugation a (a) Shear mode (b) Bending mode Fig. 3: Elastic buckling mode A A A N V( =300) V( =200) V( =100) d=8595mm h=10500mm t=4.5mm a[mm] [N/mm 30 2 ] a [mm] Fig. 4: Buckling stress of type-h A A A A A A H( =300) H( =200) H( =100) B B AB B B B Axial compression Transverse end load / Antisymmetric moment N H( =300) H( =200) H( =100) N H( =300) H( =200) H( =100) d=8595mm h=10500mm t=4.5mm B B B B B B 162

162 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 BUCKLING STRENGTH OF LIGHT-GAUGE MEMBERS WITH LARGE OPENINGS Atsushi SATO a, Seiji MORI b, Tetsuro ONO c, and Kazunori FUKIHASHI d a, b Nagoya Institute of Technology c Sugiyama Jogakuen University d NS Hi-Parts Corporation * Author for contact. Tel.: +81 (0) ; [email protected] 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 included 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

163 2 Nordic Steel Construction Conference 2015 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. anaqcr/ anaqcr SO(Analysis) BO(Analysis) (Test) Eq.(2) anaqcr/ anaqcr SO(Analysis) BO(Analysis) (Test) Eq.(2) anaqcr/ anaqcr SO(Analysis) BO(Analysis) (Test) Eq.(2) 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 anaqs anaq ( a / h) 1.0 (1) Effect of burring can be evaluated by following formula. 2 2 ana Qb anaqs 7827( t / h) ( a / h) (2) Consequently, strength reduction factor of burring opening can be expressed as follow: 2 2 s q b {(7827( t / h) 0.389)( a / h) 1.097} ( a / h) 1. 0 (3) Fig. 3 shows the validity of the proposed formula. The calculated errors are less than 10%. Conclusions Fig. 3: Comparison between calculated results with test and analysis results 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,

164 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 EXPERIMENTAL AND NUMERICAL INVESTIGATIONS OF THE STEEL STORAGE RACK UPRIGHTS Chong Ren a, Xianzhong Zhao a,* and Ru Qin a a Department of Building Engineering, Tongji University, Shanghai, , China. * Author for contact. Tel.: ; [email protected] 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 allowed 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 established. 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 systems and are also necessary to study. The perforated sections and non-standard restraint conditions 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, distortional and lateral-torsional buckling of the cold-formed steel sections, and research on interactions 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 further investigation. In this paper an experimental investigation of the behaviour of steel storage rack uprights subjected 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 using 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

165 2 Nordic Steel Construction Conference 2015 ed and non-perforated members. According to the buckling failure modes, specimens are classified into three categories of buckling failure: distortional buckling (DB) specimens, flexuraltorsional buckling (FTB) or flexural buckling specimens (FB) and interaction of distortional/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 element 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 developed 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 results. 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 results in this paper. The slenderness of the upright is represented by λ = (P y /P cr ) 0.5, where P cr 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 material yield may have significant influence on the performance of perforated and non-perforated members. Due to influence of perforations, some distortional buckling failures of nonperforated 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

166 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 EXPERIMENTAL INVESTIGATION ON THE BEHAVIOR OF PERFORATED STEEL STORAGE RACK COLUMNS UNDER AXIAL COMPRESSION Bassel EL KADI a,, Guven KIYMAZ b* and Atakan MANGIR c a,c Fatih University, Department of Civil Engineering b Antalya International University, Department of Civil Engineering * Author for contact. Tel.: ; [email protected] 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 parameters are used. In the present study, an experimental program was carried out to verify the accuracy of a recently proposed design approach that has the potential to eliminate the need for design by testing. 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 approach 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 application point along the symmetry plane of the cross-section and investigating the position that gives the maximum failure load, i.e. the position of the effective center of gravity. 167

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 Direct 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 The different alternatives of the recently proposed approach show a good accuracy compared to the experimental results. Alternative 1 and Alternative 2 gave more accurate results (the error discrepancies between 0.5 to 4%) than Alternative 3 and Alternative 4 (the error discrepancies between 0.5 to 11.5%) as shown in Figs 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. Load (kn) C500 C650 C800 C950 C1100 C1250 Displacement (mm) Fig. 1 Fig. 2 Pu (kn) Length (mm) Alt 1 and Alt 2 Put ave Test1 Test2 Test3 Test4 Pn non per Pu (kn) Length (mm) Alt 3 and Alt 4 Put ave Test1 Test2 Test3 Test4 Pn non per Fig. 3 Fig

168 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 MONOTONIC BEHAVIOUR OF BOLTED T-STUBS: A REFINED THEORETICAL MODEL FOR FLANGE YIELDING AND BOLT FRACTURE FAILURE MODE Antonella B. Francavilla a*, Massimo Latour a, Vincenzo Piluso a and Gianvittorio Rizzano a a University of Salerno * Author for contact. Tel.: ; [email protected] 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 prediction 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 supply [1,2] and cyclic behaviour of connections [3,4] past experimental and theoretical researches 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. Therefore, 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 between the tip of the plate and the edge of the bolt head, the bolt forces are considered uniformly distributed under the bolt head, the failure of the T-stubs is modelled by checking the ultimate strain on the basic materials composing the plate and the bolt, the compatibility condition 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 correspondence of the T-stub web, the known parameters are five: the force of the T-stub (F), the 169

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 experimental 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 authors at laboratory on materials and structures of Salerno University has shown a good accuracy 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., Characterization of the nonlinear behavior of single bolted T-stub connections, Proc., 5th Int. Workshop on Connection in Steel Struct., Amsterdam, [2] Beg, D., Zupancic, E. & Vayas, I., On the Rotation Capacity of Moment Connections. Journal of Constructional Steel Research, Volume 60, pp [3] Iannone, F., Latour, M., Piluso, V. & Rizzano, G., Experimental Analysis of Bolted Steel Beam-to-Column Connections: Component Identification. Journal of Earthquake Engineering, 15(2), pp [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 [5] Piluso V., Faella C. & Rizzano G., Ultimate behaviour of bolted T-Stubs. II: Model validation, Journal of Structural Engineering ASCE, 127 (6):

170 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 DIFFERENT COATING SYSTEMS FOR THE APPLICATION IN SLIP- RESISTANT CONNECTIONS Natalie Stranghöner a, Nariman Afzali b, Jörn Berg c, Markus Schiborr d, Andrea Rudolf e, Susanne Berger f 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 connections. The common way to protect steel surfaces of slip-resistant connections against corrosion is to cover the faying surfaces with a protective layer. In the frame of this paper, the influence 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 different types of testing procedures (EN and TL/TP-KOR-Stahlbauten). 1 Introduction Slip-resistant connections are required, when deformations in bolted connections must be limited to pre-defined values either for serviceability or ultimate limit reasons. A common coating 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 testing procedure. Some comparative tests have also been performed on the basis of TL/TP-KOR-Stahlbauten. 2 Experimental investigations 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) according to ISO 19840), one relying on the TL/TP-KOR-Stahlbauten testing procedure and one relying on the Annex G, EN testing procedure and seven test series for five different customary and one especially produced ESI-Zn-coatings using the Annex G, EN 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

171 2 Nordic Steel Construction Conference 2015 Series ID Table 1: Test specimens and slip factors considering static tests only and static + creep tests NDFT Number of µ nom,mean µ init,mean µ actual,mean V (µ actual ) 1) ISO test results [µm] n stat /n stat+cree p [-] (stat/stat+creep) [-] (stat/stat+creep) [-] (stat/stat+creep) [-] (stat/stat+creep) [%] ASI-Zn-Coating ASI-E 2) 8/ / / / / ASI-T 3) 10/- 0.81/- 0.81/- 0.89/- 3.6/- ESI-Zn-Coating ESI-I 2) 8/ / / / / ESI-II 2) 8/ / / / /7.2 ESI-IIa 2) 80 8/ / / / /5.5 ESI-III 2) 8/ / / / /4.4 ESI-IV 2) 8/ / / / / ESI-V 2) 8/ / / / /5.8 ESI-VI 2) 8/ / / / /6.5 1) Coefficient of variation for µ actual 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 obvious that the results show a relatively large scatter with a slip factor range of 0.47 actual,mean Compared to the ESI-Zn-coatings, the ASI-Zn-coating shows the highest slip load and slip factor with actual,mean = 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 recipes 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 recommendations 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 decrease 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 projects, 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

172 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 INFLUENCE OF DIFFERENT TESTING CRITERIA ON THE SLIP FACTOR OF SLIP-RESISTANT CONNECTIONS Natalie Stranghöner a, Nariman Afzali b, Jörn Berg c, Markus Schiborr d, Frans Bijlaard e, Nol Gresnigt f, Peter de Vries g, Ralf Glienke h, Andreas Ebert i 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 specifies slip factors for often used surface conditions. For deviating conditions, slip factors have to be determined experimentally according to Annex G of EN 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 or can be found in literature. For those surface conditions which have not been considered in EN or if higher slip factors are required, slip factor tests should be performed according to Annex G of EN However, the practice has shown, that the current slip test procedure according to Annex G is not clear in detail. 2 Experimental investigations 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 resistance behaviour and (4) the slip-criterion. Slip factor tests were carried with test specimens according to the M20-bolt-geometry of EN In total, three surface conditions were considered (1) grit-blasted Sa 2 ½ (GB), (2) 173

173 2 Nordic Steel Construction Conference 2015 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 maximum 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 considering 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 decrease 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 influence of the test duration on the test results is small (less than +/- 5%). For GB a tendency 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 significantly higher slip factor than was found in the deformation controlled tests. (5) Annex G of EN 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 deformation 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 principle 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 applicable 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 experiments very carefully. In the frame of the present contribution it could be verified that the testing procedure of EN , Annex G has to be specified more precisely e. g. regarding the load velocity, the positioning of the displacement transducers, the devices for the measurement of the preload of the bolts and the slip criterion. 174

174 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 SIMPLIFIED MODEL FOR CONNECTIONS OF STEEL STRUCTURES IN OPENSEES R. Costa a, F. Gentili b and L. Simões da Silva c a Department of Civil Engineering, University of Coimbra, Portugal b [email protected], , 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, Eurocode 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 connecting 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 computers allow for robust and rigorous analyses including the semi-rigid behaviour of connections 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

175 2 Nordic Steel Construction Conference db /2 db / node 4 (external) node 5 (internal) 7 8 node 3 (external) node 1 (external) dc/2 dc/ node 6 (internal) node 2 (external) 4 dbl/2 0 dbl/ node 4 (external) 5 13 node 5 (internal) node 3 (external) node 7 (internal) 10 9 node 1 (external) dc / 2 dc / node 6 (internal) node 2 (external) Fig. 1: Single panel (SP) (on left) and Double panel (DP) (on right)nbeam-to-column joint model. 6 dbr/2 dbr/2 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 fullstrength 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 taken as reference, since the DP model represents the most appropriate modelling strategy. HEA 240 p 1 = 15kN/m 12m HEB 400 p 2 = 5kN/m 1 3 e 3 d 5 IPE 400 IPE p 8m HEA m Fig. 2: Geometry, sections, loads of case study Table 1: Comparison between bending moment Case 2a Case 2b Case 2c [knm] [%] [knm] [%] [knm] M M M 3 e M 3 d M 3 p M M 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,

176 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 DESIGN APPROACH FOR STABILITY CHECK OF MEMBERS WITH HANGING-PROFILE CONNECTIONS Dasu Liu a,*, Jörg Lange b a,b Technische Universität Darmstadt, Institute of Steel Structures and Materials Mechanics, Germany * Author for contact. Tel.: ; [email protected] 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 finiteelement-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

177 2 Nordic Steel Construction Conference 2015 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: = + =0, (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. (2) 178

178 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 REASONS FOR CHARLES DE GAULLE AIRPORT COLLAPSE Toomas Kaljas Rak Tek Solutions Oy * Author for contact. Tel.: [email protected] 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 successfully 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 conditions. However, the structures look similar, but have substantial differences. In Berlin main railway station the external reinforcement follows the shape of moment diagram 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 tensile 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

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 structure 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 clearly 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 concentrators, 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 ml [2] Berlin main railway [3] EN :2004 [4] EN [5] RFEM 5.04 (

180 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 INVESTIGATIONS ON THE BEHAVIOUR OF THREADED AND SHANK BOLTS UNDER COMBINED TENSION ANDD SHEAR Anja Renner a and Jörg Lange b a,b Institute for Steel Structures andd Materials Mechanics, TUU Darmstadt, Germany Abstract: To estimate the actual behaviour of combined loaded bolts a series of tests on dif- results ferent types of bolts was conducted. By now all testt series have been finished and the are not as expected. The threaded bolts show a nearly ideal plastic behaviour that fits with the previously known quadratic rule. The results of tests on all bolts with shank in shearr plane - independent of the bolt grade - are overestimated by the quadratic approach. The reason for this unexpected behaviour is the t subject of a current large scale parametric numerical study. The first results of this investigation andd some first attempts for an interpretation are subject of this paper. 1 Tests on bolts More than 150 bolts of a size M 20, bolt grades 4. 6, 8.8 and 10.9 and with either thread or shank in the shear plane were tested under different tension-to-s shear load ratios. 2 Numerical model A numerical model was created using ANSYS Workbench to get g a better view what happens in a bolt when loaded simultaneously withh tension and shear. Fig. 1 shows a comparison of the real tested bolts and the numerical simulation under two of the tested angles. The numerical model iss able to simulate the deformation well. Fig. 1: Bolts grade 4.6 after testing compared to deformation plots of numerical simulation under different angles of load applicationn (0 would equal pure tension, 90 pure shear) 181

181 2 Nordic Steel Construction Conference Results In Fig. 2 for bolts with shank in shear plane the mean values of all test results and the results of the numerical analysis are summarized. The ultimate load of o the testedd bolds wass divided into its normal and its shear component. These values were related to the pure tensionn or pure shear resistance, which were determined by tests on bolts of the same charge. The analytical results can also be related to their pure tension and shear load capacities. The testt and the analytical results can now be compared to the bi-linear b limiting function from the EN , the so-called Eurocodee 3, and the quadratic interaction i rules of the old Eu- ropean standards. Fig. 2: Test and analytical results for bolts with shank in shear plane, related to their pure tension and shear loadbearing resistance 4 Interpretation Fig. 2 shows that the analytical results liee clearly under the quadratic function as the results r of the bolt tests. The image of the t deformed bolt (Fig. 3) gives a first ideaa why the quadratic functionn is not reached. The quadratic assumption goes back too the linear-elastic relation besection. tween normal and shear stress, and forcess are related to the undeformed original cross Fig. 3: Deformed bolt with left over cross section to transfer t normal force Looking at the deformed bolt (Fig. 3) shows that the area left over to transfer the normal force is by far smaller than the origin cross section. It might be moree appropriate to relate the normal force component to the tension load capacity of the reducedd cross section. 182

182 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 BEHAVIOR IMPROVEMENT OF PULTRUDED FRP BEAM - COLUMN BOLTED CONNECTIONS Ossama M. El Hosseiny a, Hassan M. Maaly b, Said A. Ibrahim c a,b,c Structural Engineering department, Zagazig University, Sharkia, Egypt [email protected] Tel 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 stiffening lower connecting angle at beam - column connection could have better effect on connection 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 connections. 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 information on the pultrusion process, and PFRP shapes themselves, is given in two manufacturers 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 momentrotation characteristics relating the moment transmitted by the joint to the relative rotation between the connected members. One of the pioneering studies was that of Mosallam [5] presenting 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 sustained loading. A numerical FEM developed in this paper and validated based on a test programs 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

183 2 Nordic Steel Construction Conference 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 adding 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 experimental 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 capacity by about 46% and decrease the rotation by about 30% compared to NSC model. Moment(KN.m) FEM EXP [7] USC NSC FSC LSC Rotation(mrad) 3 Conclusions Fig. 3: Moment Rotation relation 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, [2] The new and improved Pultex pultrusion design manual. Creative Pultrusions Inc., Alum Bank, PA. ( (April 5, 2014). [3] Strongwell design manual. Strongwell, Bristol, VA. ( [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 [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

184 185

185 186

186 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 VIBRATION RESPONSE OF USFB COMPOSITE FLOORS Richard Kansinally a and Konstantinos Daniel Tsavdaridis b 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 solutions that minimise construction cost while simultaneously allow for optimum space utilisation 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 structures (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

187 2 Nordic Steel Construction Conference 2015 Structural Model The SCC flooring system investigated comprised of a perforated steel USFB which was treated 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 developed by combining two sections; namely the 305x127x37UB and 254x254x73UC. Primary beams were the 305x127x42UB. The structural concrete depth varied throughout the analyses, and 1.2mm thickness was adopted for the decking, assuming a 210ComFlor steel decking properties. FE Model of the USFB System Modelling of this new composite floor type was done by using popular commercial software ABAQUS CAE V A two-bay floor arrangement consisted of secondary beams 210x127/254x55 ACB and supported by 305x127x42UB primary beams. In order to practically 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 different 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 perturbation steady state modal dynamic analysis was conducted to assess the acceleration performance of this novel composite floor under a human induced load model suggested by Murray et al. (2003) and Mello et al. (2008). Throughout this study, a notional damping of 3% was utilised in accordance with Bachmann et al., 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., Vibration Problems in Structures: Practical Guidelines. Basel: Birkhauser Verlag Basel. Mello, A. de A.V. et al., Modal Analysis of Orthotrophic Composite Floor Slabs with Profiled Steel Decs. Latin American Journal of Solids and Structures, Volume 5, pp Murray, T.M. et al., Floor Vibrations due to Human Activity. United States of America: American Institute of Steel Construction. Tsavdaridis, K.D. et al., Experimental and Computational Study of Vertical Shear Behaviour of Partially Encased Perforated Steel Beams. Engineering Structures, Vol. 56, pp

188 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 ANALYSES OF THE LOAD BEARING BEHAVIOUR OF DEEP- EMBEDDED CONCRETE DOWELS, CoSFB Matthias BRAUN a, *, Renata OBIALA b and Christoph ODENBREIT b 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.: ; [email protected] 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 bearing behaviour and parameters influencing this behaviour were determined through experimental 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 FEanalysis 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, reinforcement bars crossing the openings and concrete infill (Fig. 1), Eurosteel [1]. c o n c r e te d o we l i n - s i tu c o n c r e te d o we l r e i n fo r c e me n t s l a b e l e me n t 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. C o S F B 189

189 2 Nordic Steel Construction Conference 2015 To investigate the characteristics of the shear connection 27 push-out tests with varied parameters were performed. All performed push-out tests reached a slip > 6mm and can be classified as ductile in accordance with EN 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 beginning 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 connection 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 beyond 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. 7 th European Conference on Steel and Composite Structures (EUROSTEEL). Naples, Italy, [2] Deutsches Institut für Bautechnik: Allgemeine bauaufsichtliche Zulassung ArcelorMittal Belval & Differdange S.A., CoSFB-Betondübel. Zulassungsnummer Nr. Z , Berlin [3] M. Braun, O. Hechler, R. Obiala: Untersuchungen zur Verbundtragwirkung von Betondübeln Anwendung von tiefliegenden Betondübeln bei Slim-Floor-Konstruktionen (CoSFB). Stahlbau 83 (2014), Issue 4, p Germany. [4] M. Braun et al.: Experimentelle Untersuchungen von Slim-Floor-Trägern in Verbundbauweise Anwendung von tiefliegenden Betondübeln bei Slim-Floor-Konstruktionen - CoSFB. Stahlbau 83 (2014), Issue 10, p Germany. [5] Abaqus 6.11 Online Documentation, Dassault Systèmes,

190 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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 Maślak a, Małgorzata Snela b 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-tobeam 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 classical component method. In such a formal model, illustrated with the attached numerical example, 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 moment axial force interaction can play an important role when the resistance and the flexibility 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 interactions 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

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 important 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, [2] De Lima LRO, da Silva LS, da S. Vellasco PCG, de Andrade SAL. Experimental analysis of extended end-plate beam-to-column joints under bending and axial force, Proceedings of the 3rd European Conference on Steel Structures Eurosteel, Coimbra, Portugal, September 19-20, 2002, [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 Structures Eurosteel, Coimbra, Portugal, September 19-20, 2002, [4] Urbonas K, Daniūnas A. Component method extension to steel beam-to-beam and beamto-column knee joints under bending and axial forces, Journal of Civil Engineering and Management, 3, (2005), [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), [6] Daniūnas A, Urbonas K. Influence of the semi-rigid bolted steel joints on the frame behaviour, Journal of Civil Engineering and Management, 16 (2), 2010,

192 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 FIRE DESIGN OF CFST COLUMNS Improvements required for Eurocode 4 Matti V. LESKELA RAKOSPER, FINLAND Tel.: ; [email protected] Abstract The inconsistencies and irregularities in the fire design rules of EN for CFST columns have been known for years, and the method of Annex H is based on odd principles totally different from those given in section of EN 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 considered, 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 deflection profile and effective length of a continuous column are changing during the fire exposure, 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 obvious 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 N fi,c,r and/or a nominal 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 N fi,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 N fi,c,r and (EI) c,. For N fi,a,r and (EI) a,, the design temperature a may be defined as an average 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

193 2 Nordic Steel Construction Conference 2015 Adaptation factors, called as reduction factors in EN , 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 , 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 design eccentricity of the load, e fi = M fi,ed /N fi,ed. The design requirement is N fi,ed N fi,rd,, with N fi,rd, = N fi,rd and 1 to be defined in relation to e fi. Taking into account the moment magnification due to second order effects, may be resolved from fi fi 4/ ( ) 1 Nfi,pl,Rd N (1) fi,c,rd fi 1 e fi ( ) M fi,pl,rd The plastic bending resistance M fi,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 established for M pl,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 of EN 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 Resistance 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 resistance for concrete filled tubular columns Journal of Constructional Steel Research 66(8-9), , 2010 [4] Bergmann M und Hanswille G, Näherungverfahren für die Brandbemessung von Hohlprofilverbundstützen Stahlbau, 81(12), , 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

194 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 CALCULATION OF STEEL TEMPERATURE IN OPEN CROSS SECTIONS BASED ON FIRE EXPOSURE FROM CFD Joakim Sandström a,b, Ulf Wickström a 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 cannot handle all the aspects of heat loss due to conduction. FDS is therefore likely to over predict 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

195 2 Nordic Steel Construction Conference 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 Lower flange temperature 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 T r 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 T AST and also heats the air inside the void 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 T r and T g 520 C 644 C 621 C 4 Shadow effects with T AST 552 C 639 C 618 C 4 Conclusions Fire exposure should be represented T r and T g. 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 T AST, 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

196 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 LATERAL TORSIONAL BUCKLING RESISTANCE A COMPARISON OF ANALYTICAL AND NUMERICAL MODELS R. Ebel a and M. Knobloch a 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 behavior of specified static systems subject to uniaxial bending. Design results of common simplified analytical models are compared to the results of a comprehensive numerical parametric 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 complex steel structures with regard to static system (e.g. continuous beams, cantilevers), boundary conditions, cross-sections etc. The paper focuses on simplified models using reduction factors and buckling curves. In addition to the model used in EN , the approach developed 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 numerical parametric study considered three different basic static systems (simple beam, cantilever beam and two-span beam) and two basic loading conditions (distributed loading and single loads). The study considered eigenmode-conform geometric imperfections (combination of initial deformations and rotations) and residual stress patterns considering typical distributions. The following failure criteria were considered for numerically determining the resistance of the steel beams: Cross-section failure at the supports and in the field, the eigenvalue 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 performing 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 with the factors χcal,v+ϑ derived from the numerical simulations. The static system, the cross sections and the slenderness ratio 197

197 2 Nordic Steel Construction Conference 2015 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 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 difference 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. Additionally, 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 disadvantage 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 model 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 moments. Possible approaches consider the Iy/IT-ratio and the Wy/Wz-ratio for example. 198

198 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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, Sweden 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 accommodations. 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 overview of the system within the execution process. Fig. 1: Snapshot of the assembly of one 3D module 199

199 2 Nordic Steel Construction Conference 2015 The Frameup system introduces a new approach in terms of execution technique which consists 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 towers - 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 introduces 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 efficiency 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 components 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 extension 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, further research needs to be undertaken in order to optimize the lifting process, where a new feasibility 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

200 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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 performance 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 stresses. This paper gives a short presentation of the HFMI technology and examples of their application 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 presented. Also, a list of six examples of HFMI application for repairing existing bridges are given. Furthermore, feasibility assessments made on four different bridges of varying types are preformed to investigate the effects of post weld treatment and high strength steel. Also, a parametric study is carried out for simply supported railway bridges with varying span lengths. 201

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

202 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 New developments in heavy plate production for modern steel construction Dr. Tobias Lehnert a,* and Dr. Falko Schröter b a,b Dillinger Hütte, Werkstr. 1, Dillingen, Germany * Author for contact. Tel.: ; [email protected] 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 performance 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 fabricators. 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. Lately, 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 challenging 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 offshore 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 development 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

203 2 Nordic Steel Construction Conference 2015 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. thermomechanical 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

204 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 Stainless steel, a sustainable material for sustainable structures Anders Finnås *, Camilla Kaplin Outokumpu Stainless AB, Sweden * Author for contact. Tel.: +46 (0) ; [email protected] Abstract: A truly sustainable product needs to demonstrate that it can answer to all three aspects of sustainability: 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 approximately 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 economy. It has been demonstrated that the reduced painting required for stainless steel structures compared to carbon steel can reduce life cycle costs up to %. For the social dimension it is important that raw materials come from responsible sources and producers must act responsibly, such as committing to ISO 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 effectiveness. 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 performance 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 environments (salt or brackish water, de-icing salts). Initially most duplex bridges were in duplex 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

205 2 Nordic Steel Construction Conference 2012 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 conservation 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 migrating fish. The enhanced corrosion resistance of LDX 2404 is also a plus, since the waterways 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 enables engineers to choose the optimal solution for different service environments. 206

206 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 DYNAMIC RESPONSE OF PIPE RACK STEEL STRUCTURES SUBJECTED TO EXPLOSION LOADS Anton Stade Aarønæs a, Hanna Nilsson b,* and Nicolas Neumann c a,c Aker Solutions, Oslo, Norway b NCC Construction Sverige AB, Gothenburg, Sweden * Author for contact. Tel.: ; [email protected] 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 behaviour 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 parameters than the fundamental eigen period is therefore not taken into account, possibly leading 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 response 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

207 the effect is an increase of DAF. Based on this the conclusion is that calculations of multiplanar 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, Fig. 1: Results from sensitivity analysis as relationship between maximum DAFs (based on BS and OT) and eigen periods (T) 208

208 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 TALL AMBITIONS ONSHORE WIND TURBINE TOWER CONCEPTS FOR LARGE HUB HEIGHTS Martin Jespersen a *, Mogens G. Nielsen b, Ulrik Stottrup-Andersen c, 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 generated. 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 locations are already occupied or country geographical features are of a nature, so that new onshore 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 elevations to give a proper energy output and lower the turbulent loading. However the implementation of tall wind turbine towers is expensive and the traditional cylindrical 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 therefore 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 construction of tall towers and masts Ramboll have made several different conceptual and detailed designs within tall wind turbine towers. This among others includes a guyed wind turbine 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: [email protected], tel

209 2 Nordic Steel Construction Conference 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 engineers 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 or even timber): 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 supported 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 content, labour and local infrastructure are often to be considered. Fig. 1: The Guyed Wind Turbine tower developed by Ramboll Fig. 2: The K-Jacket tower developed by Salzgitter AG and Ramboll Each type of tower structure constitutes different dynamic systems, each significantly impacting the load generation on the tower. Forming an iterative process, design concepts must be closely evaluated with the loads generated. Wind turbine towers are primarily governed by fatigue (ideally a balance between fatigue and ultimate loading). The design of new tower concepts therefore largely entails the evaluation of stress hotspots 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 structural 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 years of service life with a minimum of inspection (especially of the tower structure) the towers must be designed and built to last 210

210 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 LATERAL STABILITY OF VERANDAS BY MEANS OF THE GLASS PANELS M. Fortan a, J. De Clercq, M. Meeus b, B. Rossi c 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 support 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 complete 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 relatively 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 increasingly 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 compression 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 described 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 procedure 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 support 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

211 2 Nordic Steel Construction Conference 2015 Fig. 1: Overview tests on frames Fig. 2: L connection Fig. 3: T connection Fig. 4: 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 (buckling) 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 concept described in this paper. Frames (-) Table 1: Options to resist 4.4kN with a maximum displacement of 5mm Size B x L Connection Support block (m x m) (knm/rad) (kn/mm) Displacement (mm) 1.1 x ,1 x ,1 x ,1 x ,2 x References [1] J. De Clercq, M. Fortan; De stabiliteit van veranda s ten gevolge van de glazen vulelementen; 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;

212 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 END PLATE CONNECTION FOR RECTANGULAR HOLLOW SEC- TION IN BENDING Arne Aalberg a*, Arne M. Uhre b and Per K. Larsen a 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. address: [email protected]. Phone number: 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 Classical yield line analysis is used to develop the expressions for the moment resistances. The test results are used for a discussion of the stiffness boundaries for joint classification for this particular connection. 1 Introduction a) Investigated detail, RHS 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 girders, 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 extended end plate welded to the section, and fastening by threaded bars or bolts. The design formulas for end plate connections given in EN cover typical beam-to-column connec- 213

213 2 Nordic Steel Construction Conference 2015 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 S j,ini S j,ini Yield mechanism Moment (knm) Yield mechanism Moment (knm) 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 loadingunloading 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 conservative. 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 S j, ini =560 knm/rad and S j,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 Conclusions The main conclusions are: 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

214 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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, Sweden 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 connections of modular buildings as part of the innovative construction method introduced in the research 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 (Decomposed) Profiles & plates: S355 Bolts: 10.9 M24 1 x Filler plate (300x192x[4,6,8] generating the Connection gap) 1 x Upper column (SHS 250x10) 8 Connection gap (4,6 and 8 mm) 1 x Finger plate (546x192x14) 3x Cover plates (192x56x6) 5 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

215 2 Nordic Steel Construction Conference 2015 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 fitting the bolts which are pre-installed in the lower column. Filler plates with different thicknesses (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 surface between the finger plate and the cover plates at the ultimate state. This phenomenon increases the slip-resistance of the connection, approximately twice compared to hand calculation models used in engineering practice (e.g. according to EN ). This increase cannot directly be accounted for in design due to the large slip that precedes the ultimate load. However, 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

216 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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, stateof-the-art models are typically very complex 3D models model with both material and geometrical 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 geometry 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

217 2 Nordic Steel Construction Conference 2015 Model 1 - Truss elements Model 2 - Chord as continuous beam Model 3 - Chord as continuous beam, rigid eccentricity element Model 4 - Rigid eccentricity elements for each brace Model 5 - Model 4 with rotational 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, , [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, , [3] Pasternak H, Krausche T, Launert B. Schweißen von Trägern mit dicken Blechen, Teil 1: Trägerfertigung unter Werksbedingungen Planung, Herstellung und Simulation. Bauingenieur, 89, 1 11, [4] Boel H. Buckling length factors of hollow section members in lattice girders. Master s thesis, Eindhoven university of technology,

218 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 BUCKLING OF MEMBERS OF WELDED TUBULAR TRUSS Markku Heinisuo a,* and Äli Haakana a a Tampere University of Technology, Tampere, Finland Abstract: The scope of the paper is tubular trusses with welded gap joints and flexural buckling 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 economics. In the structural design a very important design criteria is buckling resistance of the compressed 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 hollow section (where both chords and braces are square hollow sections, SHS) members: Lcr Chord : K (1) L sys L cr 1 : b Brace K (2) L sys Lbr, sys where b i is the width of the brace (i = 1,2) or chord (i = 0) and using the notations of EN for square hollow sections:

219 b b 2b 1 2 (3) b (4) 2t where t 0 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 exactly. 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-ofplane buckling length factors for the members of triangulated trusses made with welded gap K-joints, including both rectangular and circular hollow sections. The equations 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 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 stiffness 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, Master s Thesis, Eindhoven University of Technology, [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,

220 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 BENDABILITY AND MICROSTRUCTURE OF OPTIM 700 MC PLUS Vili Kesti a,*, Antti Kaijalainen b, Juho Mourujärvi b and Raimo Ruoppa c a SSAB, Finland b University of Oulu, Centre for Advanced Steels Research, Finland c Lapland University of Applied Sciences, Finland * Tel.: ; [email protected] 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 increasing. 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, max Si, max Mn, max P, max S, max Al, min CEV, typical YS, (MPa) TS, (MPa) El, (%) CV -60 C, (J/cm 2 ) * 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 technological properties of the steel are sufficiently good. Bending is one of the most important workshop 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 microstructure of 3 mm sheet consisted mainly of granular bainite (~73 %) together with quasipolygonal ferrite (~14 %), upper bainite (10%) and MA islands (3%). The corresponding microstructure of 6 mm sheet consisted mainly of granular bainite (~60 %) together with quasipolygonal 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 (d eff ) according to EBSD studies was found to be ~2 µm. 221

221 2 Nordic Steel Construction Conference 2015 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 microcracking 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, respectively. The minimum bending radii of Optim 700 MC are much better than the ISO EN standard requirements for 700MPa thermomechanically hot-rolled steels and conform 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 elongation. 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 elongation to fracture in a tensile test is not very high, the studied steel exhibits an excellent ability 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, during 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 thinning 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/cm 2, typically being over 100 J/cm 2. The minimum impact strength is also guaranteed in the transverse direction. T 28J temperatures were determined from these results using Wallin s formula. Ductile-brittle transition temperatures (T 28J ) are -110 C (longitudinal) and - 90 C (transverse). Excellent T 28J values can probably be attributed to the refinement of the coarsest grains (i.e. d 90% ): 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 propagation. 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 products. 222

222 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 EXPERIMENTAL BEHAVIOUR OF TENSION PLATES WITH CENTRE HOLE MADE FROM HIGH STRENGTH STEEL Pál Turán a, László Horváth b 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 [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 [2]. According to the present EN for sections with holes the design tension resistance (N t,rd ) should be taken as the smallest of: the plastic design resistance of the gross-section A f y N pl, Rd = γ M 0 (1) the ultimate design resistance of the net cross-section area at holes 0.9A net f N u u, Rd = (2) γ M 2 Before the present EN was published, in the last Working Group draft pren (stage 49), the design resistance of the net section was determined according to plastic resistance: Anet f y Nt, Rd = (3) γ M 0 In this report, as a new proposal the 0.9 reduction factor can be omitted, which is in accordance with [3]: Anet fu Nu, Rd = (4) γ M 2 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 d 0 varied from 8 to 40 mm in 8 mm steps. 223

223 2 Nordic Steel Construction Conference 2015 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. 3 Conclusions Fig. 1: Statistical evaluation with Eq. (3) 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 holeweakening 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 γ M2 =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 γ M0 =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, [2] EN 1993 Part 1-1: Eurocode 3: Design of steel structures Part 1-1: General rules and rules for buildings, [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, , [4] RUOSTE: Rules On High-Strength Steel; RFCS Project RFSR-CT [5] EN 1990: Eurocode Basis of structural design,

224 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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) ; [email protected] 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. Especially the strain requirements of EC are not reasonable in respect to actual steel properties. 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 :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 EC 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 [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 [1] and EN [2] assume the material to have these abilities, if three requirements are fulfilled: they concern 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] Net section resistance according to EC EN [1] as well as EN [2], reduce the net section resistance to 90 % by applying 0.9 Anet fu Nt, Rd (1) M 2 / M12 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

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 requirements 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-displacement curves are shown in Fig. 1. Force [kn] 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 experimental 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 d 16mm : A net *f u d 32mm : A net *f u d 0mm : A net *f u 4.4 % 4.6 % 4.5 % F u crack ini Displacement [mm] Displacement [mm] [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 and other toughness oriented rules in EN 1993, [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 , Force [kn] d 16mm : A net *f u d 32mm : A net *f u d 0mm : A net *f u 3.04 % 2.94 % 2.63 % F u crack ini 226

226 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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/mm 2, is finding increased use in practice, and as a result HSS has been included in traditional 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 flexuraltorsional buckling. Like a conventional mild steel beam which may also fail by flexuraltorsional 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

227 2 Nordic Steel Construction Conference 2015 (a) (b) Fig. 2: Typical elastic buckling eigenmodes with α = 0 5, (a) α z < 10 6 ; (b) α z > q u /q p 0.08 q u /q p numerical result 0 32/λ α = numerical result 0 32/λ α = Modified slenderness λ = (q p /q cr ) 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 produce 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 = 0 32 M M M, (1) b o p p in which M b is the bending strength, M p the fully plastic moment and M o 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 M o 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, , [2] Trahair NS, Bradford MA, Nethercot DA, Gardner L. The behaviour and design of steel structures to EC3, Fourth Edition, London: Taylor & Francis,

228 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 TRUE STRESS-STRAIN RELATIONSHIP FOR FINITE ELEMENT SIMULATIONS OF STRUCTURAL DETAILS UNDER DIFFUSE NECKING Petr Hradil a and Asko Talja b 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 recalculated 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 deformations. 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 tension (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 experimental 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

229 2 Nordic Steel Construction Conference 2015 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 specimens with holes 8 to 40 mm and predicted values base on first yield load f y A net, ultimate load P max and SMCS critical strain 4 Conclusions The main conclusions are: (1) Developed procedure for assessment of true stress-strain relation in Abaqus is powerful tool to effectively obtain good quality material model for simulations 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) Accurate 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 ( ) 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, , [2] Myers A.T., Deierlaine G.G, Kanvinde A. Testing and probabilistic simulation of ductile fracture initiation in structural steel components and weldments, Report No. 170, The John A. Blume Earthquake Engineering Center at Stanford University,

230 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 CALIBRATION OF THE DUCTILE DAMAGE MATERIAL MODEL PARAMETERS FOR A HIGH STRENGTH STEEL Marko Pavlovic a, Panagiotis Manoleas b, Milan Veljkovic c e Efthymios Koltsakis d a,b,c,d Luleå University of Technology, Sweden a University of Belgrade, Serbia (permanent position) ; [email protected]; 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 verification 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 explicit 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 [1] covers steel grades up to S460 and rules for higher steel grades, up to S700, are given in EN [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 calculations 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

231 2 Nordic Steel Construction Conference 2015 a) experiments (all specimens) b) FEA (CHT and CHT ) 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 procedure 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 conducted for any other geometry of the structure without having problems with extreme nonlinear 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 Design of steel structures, Part 1-1: General rules and rules for buildings, Brussels, Belgium: European Committee for Standardization, [2] EN 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, [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:

232 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 BUCKLING OBSERVATION OF DOOR OPENINGS FOR WIND TURBINE TOWERS Anh Tuan Tran a, Milan Veljkovic a, Carlos Rebelo b, Luis Simões da Silva b a Luleå University of Technology, Sweden [email protected], [email protected] b University of Coimbra, ISISE, Portugal [email protected], [email protected] 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 validation 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 research 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 specimens 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 compare 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

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 successfully recorded in the experiment. Good agreement between compression experiments monitored by the system and FE results are obtained. Result of the buckling development calibrated to load displacement curve is presented in Fig 2. Acknowledgment Fig. 2: Buckling development around opening. 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, [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 , [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 , [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, September [5] Abaqus , Simulia Dassault Systmes, [6] GOM Inspect V7.5, GOM mbh, [7] ARAMIS software version 6.3, GOM mbh,

234 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 Extension of the Continuous Strength Method to the Determination of Shear Resistance Najib Saliba a,* and Leroy Gardner b a University of Balamand, Lebanon b Imperial College London, UK * Author for contact. Tel.: ; [email protected] 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 γ csm /γ y, referred 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 csm but minimum u , for w < 0.83 (1) y w y y in which γ csm is the CSM limiting shear strain, γ y is the yield strain in shear, u is the ultimate shear strain of the material and w y cr is the web slenderness, where τ y is the yield strength in shear and τ cr 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 G sh ) material model in shear, given by Eq. (2). ( ) csm y Gsh csm y for γ csm γ y (2) 235

235 The web shear resistance V bw, csm,rd is then given by Eq. (3), where Aw is the web area and γ M0 is a partial factor. V A τ (3) bw, csm,rd The CSM predictions of shear capacity were compared to the ultimate shear capacity of a series 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], improvements 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 interaction surfaces from the CSM and EN [9] is shown in Fig. 1. w csm M 0 V (kn) V b,csm,rd V b,rd Test result M-V interaction (EN ) M-V interaction (CSM) 0 M c,rd M csm,rd M (knm) Fig. 1: Moment-shear interaction diagrams according to EN [9] and CSM for plate girder I References [1] Gardner L. The continuous strength method, Proceedings of the Institution of Civil Engineers-Structures and Buildings, 161(3), , [2] Gardner L, Theofanous M. Discrete and continuous treatment of local buckling in stainless steel elements, Journal of Constructional Steel Research, 64(11), , [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), , [4] Saliba N, Gardner L. Cross-section stability of lean duplex stainless steel welded I- sections, Journal of Constructional Steel Research, 80, 1-14, [5] Saliba N. Structural behaviour of lean duplex stainless steel welded I-sections [PhD thesis], Department of Civil and Environmental Engineering, Imperial College London, [6] Afshan S, Gardner L. The continuous strength method for structural stainless steel design, Thin-Walled Structures, 68, 42-49, [7] Saliba N, Gardner L. Experimental study of the shear response of lean duplex stainless steel plate girders, Engineering Structures, 46, , [8] Saliba N, Real E, Gardner L. Shear design recommendations for stainless steel plate girders, Engineering Structures, 59, , [9] EN Eurocode 3: Design of steel structures - Part 1.4: General rules - Supplementary rules for stainless steel, CEN,

236 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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) ; [email protected] 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 C, where there is a significant drop in strength. Secondly, between 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, R p0.2, or tensile strength, Rm, the time dependency of the material will drastically decrease the strength as the temperature increases. 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 applications 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

237 Today different testing parameters and standards are used for the RT and the elevated temperature (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 austenitic grade, EN , and a high strength duplex stainless steel, EN (LDX 2101 ). It is shown that much of the drop in strength is highly and solely related to the testing parameters 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

238 NEW STEEL DAMPER WITH DISPLACEMENT DEPENDENT RECENTERING FOR SEISMIC PROTECTION OF STRUCTURES Murat Dicleli a 1, Ali Salem Milani b 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 Multidirectional 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 postelastic stiffness and very stable cyclic response. A design methodology for the device has also been completed. 1 [email protected], Phone: +90(312)

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 : T1 T2 2 1 : cos 2 cos 1 T2 T 1 : F 2 F 1 Lcos 2 Lcos 1 (a) (b) Fig. 2: (a) Working mechanism of MTHD responsible for geometric hardening; (b) MTHD response for different design hardening indices (HI=F max/f Y). (a) (b) Fig. 3: (a) 200kN, 120mm-capacity prototype MTHD, as tested at METU; (b) Sample cyclic response obtained from tests. 240

240 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 FRETTING FATIGUE PHENOMENON IN BOLTED HIGH-STRENGTH STEEL PLATE CONNECTIONS Olli-Pekka S. Hämäläinen a, Timo J. Björk b a,b Laboratory of Steel Structures, Lappeenranta University of Technology, Finland Correspondence: [email protected], tel 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 bolted 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 surfaces 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. Meanwhile the S960QC joints couldn t handle similar fatigue loadings and fractured after the fretting 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 million cycles, even though the loading was identical. The most common crack initi- 241

241 2 Nordic Steel Construction Conference 2015 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 uneven 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: according 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 fretting 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 σ f (fatigue strength coefficient) and b (fatigue strength exponent) 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 fretting 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

242 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 COMPARISON OF RELATIVE VOLUMES OF DIFFERENT TYPE OF WELDS Juha Kukkonen a, Markku Heinisuo b a Sweco Structures, Tampere, Finland b Tampere University of Technology, Tampere, Finland 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 transmitted 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 β w of Eurocodes 243

243 2 Nordic Steel Construction Conference 2015 Conclusions Fig. 4: T-joint, shear load, modified β w based on [1] Comparing the results which are calculated by using recommended values of correlation factors β w 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. Calculating 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 practically the same compared to steel grade S355 in the shear load and in the axial load. However, 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 Eurocodes. 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 , Stahlbau 81 (2012), Heft 4, : Ernst&Sohn 244

244 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 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 column gusset plate connections. It is important in normal exploitation period as in the construction sites during assembly period then not all the ties and connections are installed. The second 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. Displacement transducer Pinned support Vertical load Displacement Displacement transducer (2) transducer (1) Displacement transducer Pinned support a) The scheme of laboratory tests b) The specimen in laboratory Fig. 1: The specimen 245

245 2 Nordic Steel Construction Conference 2015 Conclusions The natural experiment results and analytical analysis according to Eurocode 3 of the connections of cold formed sections allow making such conclusions: 1. Eurocode 3 is accurate tool to calculate strength capacity of cold formed steel connections. Safety margin was from 3 % to 18,6 %. 2. Boundary conditions have huge impact to the strength of the connection. The test results showed that absence of lateral restrains resulted in 31 % decrease in strength capacity. 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 experimental value. This means that Eurocode 3 is not suitable for such connection calculations. Acknowledgments This work has been supported by the European Social Fund as part of the project Development and application of innovative research methods and solutions for traffic structures, vehicles and their flows, project code VP1-3.1-ŠMM-08-K References [1] Wong, M. F.; Chung, K. F Structural behaviour of bolted moment connections in coldformed steel beam-column sub-frames, Journal of constructional steel research Vol. 58, pp [2] Yu, W. K.; Chung, K. F.; Wong, M. F Analysis of bolted moment connections in cold formed steel beam-column sub-frames, Journal of constructional steel research, Vol. 61, pp [3] Sabbagh, A. B.; Petkovski, M.; Pilakoutas, K.; Mirghaderi, R Ductile moment resisting frames using cold formed steel sections: an analytical investigation, Journal of constructional steel research, Vol. 67, pp [4] Sabbagh, A. B.; Petkovski, M.; Pilakoutas, K.; Mirghaderi, R Development of coldformed steel elements for earthquake resistant moment frame buildings, Thin-walled structures Vol. 53, pp [5] Sabbagh, A. B; Petkovski, M.; Pilakoutas, K.; Mirghaderi, R Cyclic behaviour of bolted cold-formed steel moment connections: FE modelling including slip, Journal of constructional steel research, Vol. 80, pp [6] Bucmys, Z.; Sauciuvenas, G The behaviour of cold formed steel structure connections, Engineering structures and technologies, Vol. 5:3, pp [7] EN : 2005: Eurocode 3 - Design of steel structures - Part 1-8: Design of joints. Comité Européen de Normalisation, Brussels. [8] Urbonas, K.; Daniunas, A Behaviour of semi-rigid steel beam-to-beam joints under bending and axial forces, Journal of constructional steel research, Vol. 62, pp [9] Bucmys, Z.; Daniunas, A.; Rasiulis, K 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, ISBN pp

246 Nordic Steel Construction Conference 2015 Tampere, Finland September 2015 RESISTANCE RESULTS FOR THE CROCODILE CONNECTION Panagiotis Manoleas a, Kristoffer Öhman b, Efthymios Koltsakis c and Milan Veljkovic d 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 commonly 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 examples 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

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, Department of Civil engineering, University of Toronto, Final report to CIDECT on programme 8G, Aug [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 , Sep [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 , [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 , Apr [5] Edurne Nunez Moreno, Cyrill Tarbe, David Brown, and Adbul Malik, Joints in Steel Construction: Simple Joints to Eurocode 3, Reprint. SCI/BSCA, [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,

248 Author index Aalberg, Arne, 145, 213 Aarønæs, Anton, 207 Aasen, Bjørn, 31 Abspoel, Roland, 113 Afzali, Nariman, 171 Afzali, N., 173 Akbas, Bulent, 119 Al-Emrani, M., 201 Ampatzis, Alexios, 135 Andrade, Pedro, 199, 215 Andreassen, Michael, 71 Baczkiewicz, Jolanta, 153 Barros, Rui, 149 Berg, Jörn, 171, 173 Berger, S., 171 Bijlaard, F., 173 Björk, Timo J., 241 Botti, Andrea, 73 Bradford, Mark, 227 Braun, Matthias, 189 Brauns, Janis, 185 Bučmys, Žilvinas, 245 Bujňák, Ján, 83, 107 Bzdawka, Karol, 153 Classen, Martin, 141 Clausen, Arild, 145 Clercq De, Jesse, 211 Costa, Ricardo, 175 Daniūnas, Alfonsas, 245 Dekker, Rianne, 115 Dicleli, Murat, 101, 109, 123, 239 Donnadieu, Marc, 81 Döring, Bernd, 73 Ebel, Rebekka, 197 Ebert, A., 173 Efthymiou, Evangelos, 135 El Hosseiny, Ossama, 183 El Kadi, Bassel, 167 Exner, Hans, 131 Feldmann, Markus, 73, 85, 137, 225 Finnås, Anders, 205 Fortan, Maarten, 211 Francavilla, Antonella, 169 Friedrichsen, Trygve, 127, 133 Frøseth, Gunnstein, 103 Fujihashi, Kazunori, 163 Fülöp, Ludovic, 81 Gardner, Leroy, 235 Genge, Nico, 97 Gentili, Filippo, 175 Glienke, R., 173 Gresnigt, N., 173 Grimsmo, Erik, 145 Groth, Hans, 237 Gunalan, Shanmuganathan, 155 Gyllenram, Rutger, 11 Gödrich, Lukáš, 77 Haakana, Äli, 219 Hakimi, Poja Shams, 201 Havula, Jarmo, 89 Heinisuo, Markku, 79, 89, 217, 219, 243 Heistermann, C., 215 Heistermann, Tim, 199, 215 Henriques, Jorge, 149 Horváth, László, 223 Hradil, Petr, 229 Hämäläinen, Olli-Pekka, 241 Ibrahim, Saeed, 183 Ikarashi, Kikuo, 99, 117, 161 Jacobsen, Emma, 237 Jalkanen, Jussi, 91 Janarthanan, Balasubramaniam, 155 Janik, Peter, 111 Jaspart, Jean-Pierre, 21 Jensen, Hans Vagn, 105 Jespersen, Martin, 209 Jokinen, Timo, 217 Järvinen, Saku, 1 Jönsson, Jeppe, 71, 151 Kabeláč, Jaromír, 77 Kaijalainen, Antti, 221 Kaljas, Toomas, 179 Kansinally, Richard, 187 Kaplin, Camilla, 205 Karalar, Memduh, 101, 109, 123 Karoumi, Raid, 125 Kemper, Frank, 137 Kesti, Jyrki, 61, 73 Kesti, Vili, 221 Kimura, Yoshihiro,

249 Kiymaz, Guven, 167 Knobloch, Markus, 197 Knoedel, Peter, 139 Koltsakis, Efthymios, 231, 247 Kostakakis, K., 201 Krahwinkel, Manuel, 159 Kubota, Daiki, 99 Kuhnhenne, Markus, 73 Kukkonen, Juha, 243 Kurejková, Marta, 77 Lange, Jörg, 181 Langseth, Magnus, 145 Larsen, Hilmer, 127, 133 Larsen, Per, 213 Latour, Massimo, 169 Lawson, Mark, 73 Leander, John, 125 Lehnert, Tobias, 203 Leskela, Matti V., 193 Lindström, Rita, 237 Liu, Dasu, 177 Lombaert, Geert, 93 Lu, Wei, 143 Lundholm, John, 199 Maaly, Hassan, 183 Magnucki, Krzysztof, 157 Mahendran, Mahen, 155 Maljaars, J., 115 Mangir, Atakan, 167 Manoleas, Panagiotis, 231, 247 Maslak, Mariusz, 191 Meeus, Marc, 211 Mela, Kristo, 79, 217 Mellaert Van, Roxane, 93 Milani, Ali, 239 Mori, Seiji, 163 Mosiello, A., 201 Mourujärvi, Juho, 221 Neumann, Nicolas, 207 Nielsen, Mogens, 209 Nilsson, Hanna, 207 Obiala, Renata, 189 Odenbreit, Christoph, 189 Odrobiňák, Jaroslav, 107 Ohmichi, Kenjiro, 129 Ono, Tetsuro, 163 Paczos, Piotr, 157 Paiva, Fabio, 149 Palisson, Anna, 75 Pavlovic, Marko, 215, 231 Pedersen, Helge, 133 Petersen, Christian, 127 Petersen, Tobias, 159 Piluso, Vincenzo, 169 Psomiadis, Vasileios, 135 Puttonen, Jari, 143 Qin, Ru, 165 Rebelo, Carlos, 233 Reger, Vitali, 73 Remde, Christian, 97 Ren, Zhong, 165 Renner, Anja, 181 Rizzano, Gianvittorio, 169 Roivio, Pekka, 89 Ronni, Hilkka, 79 Rossi, Barbara, 211 Rudolf, A., 171 Ruoppa, Raimo, 221 Rönnquist, Anders, 103 Šabatka, Luboš, 77 Saliba, Najib, 235 Sandström, Joakim, 195 Sato, Atsushi, 163 Schaffrath, Simon, 225 Schaumann, Peter, 51 Schedin, Erik, 237 Schevenels, Mattias, 93 Schiborr, M., 171, 173 Schillo, Nicole, 85, 225 Schröter, Falko, 203 Schäfer, Markus, 121 Shirai, Daigo, 117 Simões da Silva, Luis, 175, 233 Snela, Malgorzata, 191 Snijder, H.H., 115 Sokol, Leopold, 75 Stark, Alexander, 141 Steige, Yvonne, 147 Stottrup-Andersen, Ulrik, 209 Stranghöner, N., 171, 173 Stroetmann, Richard, 87 Sugimura, Yuji, 129 Talja, Asko, 229 Thomsen, Kjeld, 127, 133 Tiainen, Teemu, 217 Tsavdaridis, Konstantinos, 187 Tuan Tran, Anh, 233 Turán, Pál, 223 Uhre, Arne, 213 Ulf, Wickström,

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,

251 Presentations Wednesday 23/Sep/2015 Keynote :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 , 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

252 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

253 Opening Location: Small Auditorium Keynote 1-2 Location: Small Auditorium Chair: Markku Heinisuo Lunch Keynote 3 Location: Small Auditorium Chair: Markku Heinisuo Plenary Session A Location: Small Auditorium Chair: Markku Heinisuo Coffee Keynote 4 Location: Small Auditorium Chair: Jari Mäkinen Plenary Session B Location: Small Auditorium Chair: Jari Mäkinen Ice Breaking Event Location: Plevna Brewery Pub & Restaurant Chair: Jari Mäkinen 10:10-10:30 10:30-12:00 12:00-13:00 13:00-13:45 13:45-15:00 15:00-15:30 15:30-16:15 16:15-17:30 19:00-21:00 Coffee Session 1B: Building Structures 2 Location: Sonaatti 1 Chair: Kristo Mela Lunch Keynote 6 Location: Sonaatti 1 Chair: František Wald Session 1C: Building Structures 3 Location: Sonaatti 1 Chair: František Wald Coffee Session 1D: Cold Formed Structures Location: Sonaatti 1 Chair: Jeppe Jönsson Conference Banquet Location: Hotel Tammer Chair: Markku Heinisuo 10:00-10:30 10:30-12:00 12:00-13:00 13:00-13:45 13:45-15:00 15:00-15:30 15:30-17:30 19:00-23:00 Session 3E: Workshop Location: Aaria Chair: Ludovic Fülöp Coffee Session 1F: Sustainable Engineering Location: Sonaatti 1 Chair: Mark Bradford Session 3F: Workshop Location: Aaria Chair: Ludovic Fülöp Lunch Session 1G: High Strength Steel Location: Sonaatti 1 Chair: Richard Stroetmann Coffee & Ending 9:00-10:00 10:00-10:30 10:30-12:00 10:30-12:30 12:00-13:00 13:00-15:00 15:00-15:30 ISBN (printed) ISBN (USB) Session 1E: Composite Structures Location: Sonaatti 1 Chair: Mikko Malaska Photo: Tampere-Hall Session 2G: Stainless Steel & Connections Location: Sonaatti 2 Chair: Jari Mäkinen Session 2F: Connections 3 Location: Sonaatti 2 Chair: Kristo Mela Session 2E: Fire Engineering & Building Structures Location: Sonaatti 2 Chair: Jean-Marc Battini Keynote 7 Location: Sonaatti 1 Chair: Mikko Malaska 8:00-8:45 Session 2D: Connections 2 Location: Sonaatti 2 Chair: Jari Mäkinen Session 2C: Connections 1 Location: Sonaatti 2 Chair: Michael J. Andreassen Session 2B: Bridges Location: Sonaatti 2 Chair: Jan Bujnak 8:45-10:00 Date: Friday, 25/Sep/2015 Session 1A: Building Structures 1 Location: Sonaatti 1 Chair: Richard Stroetmann 8:45-10:00 Session 2A: Bridges & Fatigue Location: Sonaatti 2 Chair: Jean-Marc Battini Keynote 5 Location: Sonaatti 1 Chair: Richard Stroetmann 8:00-8:45 Date: Thursday, 24/Sep/2015 Registration and Coffee 8:30-10:10 Date: Wednesday, 23/Sep/ th Nordic Steel Construction Conference (NSCC-2015) September 2015, Tampere Hall, Tampere, Finland Proceedings of The 13th Nordic Steel Construction Conference (NSCC-2015) Edited by Markku Heinisuo & Jari Mäkinen Invited keynotes and extended abstracts September 2015, Tampere, Finland The 13th Nordic Steel Construction Conference (NSCC-2015) Proceedings of

254 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