Selected, extended paper from the SDSS 2019 special session ECCS/TC8 - Structural Stability
This paper discusses the simplified method for the lateral-torsional buckling design of I-shaped beams with discrete lateral restraints. This popular method, also adopted in Eurocode 3, Part 1-1:2005, consists of verifying the resistance of the compressed “flange” of the beam subjected to major-axis bending. In this case the compressed flange is considered to be composed of the flange itself plus 1/3 of the compressed part of the web. This paper summarizes the analytical derivation of the simplified method and its key assumptions. It then presents inconsistencies arising if the simplified method is compared with the more generally applicable design methods for lateral-torsional buckling given in Eurocode 3, Part 1-1:2005. In addition, the design methods are compared with numerical simulations to reveal the lack of conservatism in the simplified method in some cases. The final part of the paper discusses how the simplified method may be modified in order to ensure a sufficient level of safety.

Selected, extended paper from the SDSS 2019 special session ECCS/TC8 - Structural Stability
The cross-sectional strength of square (SHS) and rectangular (RHS) hollow sections loaded in compression and various degrees of uniaxial or biaxial bending are governed by local instabilities in the elastic or plastic range. Common design checks for cross-sectional strength, e.g. those found in the Eurocodes, regularly penalize these sections through a conservative omission of various mechanical effects and a categorization of cross-sections into distinct classes with corresponding, markedly different, design rules. This leads to discontinuities and inaccuracies in the strength representation. Such conservatism is particularly detrimental to the introduction of high-strength steel hollow sections, which often fall into the semi-compact and slender cross-section classes for which local buckling is more relevant. This paper discusses the results of extensive research work carried out during the RFCS research project HOLLOSSTAB. In this project, new design rules were developed for the cross-sectional and member design checks of hollow sections with various shapes and slenderness ratios, termed the “Generalized Slenderness-based Resistance Method - GSRM”. This paper summarizes the experimental and numerical campaign carried out within HOLLOSSTAB and describes the new GSRM design rules and their background for the case of the cross-sectional strength of SHS and RHS.

Selected, extended paper from the SDSS 2019 special session ECCS/TC8 - Structural Stability
This paper presents the development and assessment of an innovative cross-section design method for structural steel circular and elliptical hollow sections (CHS and EHS) - the generalised slenderness-based resistance method (GSRM). A numerical simulation programme was first conducted to expand the data pool for CHS and EHS. Finite element (FE) models were established, validated against existing test data and then utilised for parametric studies, where a total of over 3 700 cross-section resistance data were numerically generated. The development of the GSRM for CHS and EHS is then presented. Key design parameters, including the reference resistances and the generalised local slenderness, are initially defined. The general design procedure is subsequently introduced. Two design alternatives for CHS and EHS - a strength-based approach, and a deformation-based approach based on the continuous strength method (CSM), are developed and presented. Finally, the proposed GSRM is assessed using the previously collected test results and freshly generated FE data, where excellent accuracy and consistency in the resistance predictions are clearly revealed for all loading scenarios. Subsequent reliability analyses demonstrate that the current partial safety factor used in EN 1993-1-1 can be applied to the GSRM, achieving an appropriate level of reliability.

Kurita-Albrecht Best Technical Paper Award - Steel at 12th Japan German Bridge Symposium, September 2018 in Munich
The bridge across the Schorgast Valley is based on a competition-winning design. It comprises an overhead structure consisting of six pylons each with 10 stay cables and a slender composite cross-section with a torsion-resistant steel box as the bridge superstructure.
The jury concluded that the competition submission was a self-assured accomplishment integrating harmoniously into the sensitive landscape of the Schorgast Valley. The substructure solution shows great transparency, emphasizing the significance of the superstructure. All in all, the design is a convincing contribution to the region's building culture characterized by innovative spirit and creative power.
The structural analysis of the cross-girders and the supporting frame structures is explained in this article. To this end, the section forces from the global model were applied to a local finite element model.

Nominated for the Bernt Johansson Outstanding Paper Awards at Nordic Steel 2019
The growing popularity of high-strength steel grades leads to the need for more accurate design specifications, with the aim of fully exploiting the material benefits and creating economic advantages. According to Eurocode 3, the maximum rotational capacity of a section is limited and linked to the definition of cross-sectional classes. For class 1, the rotational limit &thgr; is assumed to be “infinite”, whereas it drops significantly for classes 2 and 3 to a maximum rotation of &phgr;pl and &phgr;el, respectively. In reality, despite their lower hardening capacity and ultimate strains, high-strength steel sections exhibit a non-negligible rotational capacity that exceeds these code predictions, which were developed for mild steel and with a level of analysis in mind that is suitable for hand calculations. In order to achieve greater use of high-strength steel sections, it is thus very important to understand and be able to predict a section's deformation and rotational capacity more accurately, with the aim of implementing the findings in tools for advanced, FEM-based design by analysis (DbA) approaches. As an initial step in this direction, this paper shows the results from numerical calculations of the rotational capacity of HSS rectangular hollow sections. The numerical results were calibrated against laboratory tests. Consequently, different rectangular cross-section dimensions and variations of the steel grade and thickness were chosen and analysed in the ABAQUS finite element software.

Nominated for the Bernt Johansson Outstanding Paper Awards at Nordic Steel 2019
The current design rules of EN 1994-1-1 covering headed stud shear connectors for composite beams with profiled sheeting lead, in some cases, to an overestimation of the load bearing capacity. Owing to their empirical nature, these equations are not able to capture the real behaviour of the connector. Therefore, the load bearing mechanisms of the shear connection are identified in this work with the support of experimental and numerical results. According to the static system proposed, the concrete rib is modelled as a system of diagonal struts in combination with the stud in bending. It was observed that at 1-4 mm slip, a “strut and beam” mechanism prevails, where the resistance of the connector depends on the activation of the plastic hinges in the stud and on the capacity of the diagonal strut in front of it. By increasing the slip (approx. 4-10 mm), the surrounding concrete gradually crushes, while the tensile stresses at the edge of the rib reach the tensile strength of the material. Because of this loss of rotational stiffness, the bending moment in the stud decreases and the upper plastic hinge gradually moves towards the slab. At higher displacements (approx. 20-40 mm), high tensile forces develop in the stud due to non-linear geometric effects and the load is carried through a “strut and tie” resistance mechanism, provided that the embedment of the stud is sufficient to prevent the rotation of the rib. As the slip increases further, failure occurs either in the form of concrete pull-out or stud rupture.

Nominated for the Bernt Johansson Outstanding Paper Awards at Nordic Steel 2019
Most steel moment frames are designed so that the beams act as structural fuses during earthquake loading. This approach, while good for saving lives, has poor resilience because it is difficult to replace damaged beams in a building. A repairable steel moment-frame connection has been developed that may be attractive to engineers and building contractors. The connection has a fuse plate bolted to the beam bottom flange which experiences shear yielding during severe earthquakes, protecting both the column and the beam. After an earthquake, the fuse plates can be removed and replaced to unlock residual frame deformations and prepare the building for another event.
Full-scale experiments were conducted to validate the performance of this repairable connection. Two tests were performed with a W690 × 125 beam and W530 × 196 column. The tests demonstrated rotation capacities of at least 0.05 rad, which exceeds AISC requirements for special moment frames. The fuse plate was effective in preventing beam damage in each test, and the second test demonstrated that the connection could be repaired by replacing the fuse plate.

Nominated for the Bernt Johansson Outstanding Paper Awards at Nordic Steel 2019
This paper presents a Two-Stage-Model for determining the fatigue life of steel structures. The model follows the distinction between crack initiation and crack propagation according to the phenomenological background. The first stage uses the local strain-life approach to predict the number of cycles for crack initiation Ni considering the cyclic material behaviour as well as the load-time history. The second stage calculates the crack propagation based on a linear-elastic fracture mechanics approach. Using X-FEM in combination with cyclic loading, an arbitrary, solution-dependent crack path is achieved to determine the number of cycles for propagation Np. The Two-Stage-Model was applied to butt welds to determine their total fatigue life Nf in terms of a proof-of-concept study. The study investigated the ability of the model to consider the influence of the weld geometry, i.e. plate thickness and weld imperfections (excess and vertical offset). The results are presented by means of S-N curves for initiation life, propagation life and total fatigue life. A comparison with experimental results provided in the literature shows that the Two-Stage-Model can reflect the weld quality of butt welds in terms of fatigue life.

Nominated for the Bernt Johansson Outstanding Paper Awards at Nordic Steel 2019
A state-of-the-art framework for the design of jacket foundations for offshore wind turbines is presented here. The article illustrates how an efficient, yet highly accurate, structural model can be achieved through a combination of a comprehensive beam finite element model and detailed shell/solid finite element models, where the former suffices for representing the overall global behaviour and the latter improves the representation of non-trivial structural details such as the transition piece and tubular joints. This approach allows the verification of standard structural elements according to various design codes to be automated on beam element level, while the fatigue performance via the hot-spot method as well as detailed stress and buckling analyses of relevant joints and details can be performed based on the detailed shell/solid finite element models readily available for stiffness representation. The computational framework ensures full consistency between the local and global description levels and an efficient cloud interface facilitates the thousands of design load cases that typically require consideration in order that a full detailed design can be simulated within a reasonable timeframe.

Nominated for the Bernt Johansson Outstanding Paper Awards at Nordic Steel 2019
Over the past few decades, the cold-formed steel industry has developed a manufacturing technique for producing self-supporting arches from trapezoidal steel sheets. The press-forming procedure is based on introducing transverse corrugations into the main direction of the flat profile in order to bend it. The main problem is that these indentations, which are essential to curve the profile, change the effective properties of the original steel sheet. Currently, there is no design code or standardized test procedure that can be used to obtain the effective properties of the corrugated profile. This research project analyses the behaviour of the corrugated profile when it is subjected to pure compression and compares it with the behaviour of the flat profile.

Nominated for the Bernt Johansson Outstanding Paper Awards at Nordic Steel 2019
The “column web panel in shear” is known to be a key component in the design of steel and steel-concrete composite joints as it can provide a reserve of ductility at the joint when activated and appropriately designed. Therefore, its behaviour has been studied for years and was thought to have been fully understood. However, some recent research projects have demonstrated that, in many cases, the simple analytical model proposed in Eurocode 3, Part 1-8, significantly overestimates the actual resistance of this component. In this context, this paper will look at the first results of investigations into that problem conducted at Liège University. In particular, beam-to-column welded joints have been studied in order to: i) highlight the aforementioned problem through comparisons between existing experimental results and Eurocode 3 predictions, ii) develop a sophisticated finite element model using the Abaqus© software, iii) validate this FE model using existing experimental results and iv) develop an extensive parametric study in order to highlight the key parameters governing the resistance capacity of the component studied. Based on the investigations conducted, the final goal consists of providing a new analytical formulation that is able to predict more accurately the shear strength of the column web panel.

Industrial storage operators use not only new pallet racking systems, but also old ones for which certification documents possibly do not exist. In order to check the carrying capacity of such racks, in addition to component tests, full-scale experimental tests were carried out on complete structures to help the development of reliable numerical models. The racks were subjected to vertical and pushover loads in the laboratory using an improved version of the base plates to achieve fixed support conditions. Vertical loading was imposed by filling tanks with water, an operation that was activated and controlled through an innovative hydraulic network specifically designed for this purpose. During the vertical loading tests, a moderate earthquake took place near the experimental facility and this influenced the rack behaviour positively. The spine brace was too flexible in the out-of-plane direction and did not participate in the resistance to lateral forces. Distortional buckling of the columns was observed at high lateral loads. The experimental tests allowed the calculation of conservative values for the behaviour factor q.

When ascertaining the effective widths of class 1 and 2 composite girders, it is claimed that at ULS, Eurocode 4 procedures give highly conservative results. A new approach to the problem is proposed in which the strain limitation values are introduced after deriving a new formula (graphically presented) with the intention of replacing the existing one. It turns out that the new effective width is significantly larger. Such a result leads to some advantages in reinforcement distribution in the support areas of continuous girders. It might also lead to some savings in reinforcing and structural steel. This is demonstrated by an example. Another example shows a situation in which the rough application of Eurocode 4 may produce a design with a lack of reliability for some of the bridge superstructure elements. The results obtained may also be useful for situations where a structure with composite girders already in service has to be redesigned for increased loads.

The Eurocodes (EC) represent European norms (EN) for the design and construction of building and bridge structures. The development of these European design codes is carried out on behalf of the European Union (EU) by the “European Committee for Standardization - CEN”. The Eurocodes have been in use for some years now. The motivation behind the development of the Eurocodes is the harmonization of the planning and design process for the structural components and construction elements of buildings and bridges all over Europe. Based on the experience gained through the use of the Eurocode conditions in recent years and many comments from industry, a revision of all parts of the Eurocode is currently in progress. The aim is to adapt the current design standards to the state of the art and research, to simplify the application and to harmonize the Eurocodes with each other. Finally, these modifications will be transferred to the second generation of European standards. An overview of the development phases and project teams is presented in this paper, accompanied by an insight into current standardization work and the technical discussions, mainly focusing on EN 1994-1-1.

Recently, SBHS (Steels for Bridge High-Performance Structure) has been introduced in Japan because of its remarkable mechanical properties. The main features of SBHS are high yield strength, improved toughness and weldability, which are superior to conventional steel materials at the corresponding tensile strength grades. In this study, these material properties of SBHS were compared with conventional steel materials through experimental studies as well as the Japanese design specifications. The main focus of this paper is the evaluation of the local buckling strength of two SBHS700 stub columns, each with cruciform section through compression tests. The load-displacement relationship, strain bifurcation related to plate buckling and out-of-plane deformation are discussed. Two test specimens with different width-to-thickness ratio parameters RR are compared. As the results indicate, for the specimen having a larger RR, an abrupt decrease in load was observed after the peak load, whereas the decrease in load for the specimen with a smaller RR proceeded gradually after the peak load with a much larger vertical displacement. In terms of local buckling strength, it was found to be possible to evaluate the SBHS700 stub column with cruciform section based on the existing ultimate strength curves.

This paper presents the development of the method for predicting the reusability of building components and whole structures and a pilot study of three different single-storey steel buildings in Finland. The method enables various building parts and products to be classified through a procedure to calculate their reusability index. These values can be further used to produce a single reusability indicator for the whole end-of-life scenario (e.g. complete or partial building reuse). The aggregated result of the whole building may be very useful when planning demolition or reconstruction works as well as when assessing the environmental impact of the new buildings. The main new thing explored in this paper is the possibility of integrating the economyaspect of the recovered components in the reusability assessment. Indeed, the reusability index developed earlier was purely based on the technical requirements of recovering, remanufacturing and reusing components. Hence, it was a technical reusability indicator. The possibility ofincluding information on the prospective marketability of the recovered component is a valuable extension to this method.

This paper focuses on the behaviour of connections with stiffened top and seat angles. There is a discussion of nine statically loaded full-scale tests carried out in three groups for different thicknesses of top and seat angle stiffeners. The purpose was to provide the necessary data to develop Eurocode 3. The maximum bending moment increased with the increase in stiffener thickness from 6 to 10 mm. Moreover, the plastic flexural resistance and bending moment capacity increased as the length of top and seat angles decreased. Furthermore, the plastic region of the seat angle section parallel to beam was greater than the vertical section of the angle, and the dissipated energy depended on the lengths of the angles.

Dedicated to Dipl.-Ing. Dr.-Ing. E.h. Reiner Saul on the occasion of his 80th birthday
This paper describes the selection of box girders for cable-supported railroad bridges from the point view of structural behaviour, fabrication, transportation and erection. There is a review of the use of wide single box girders and vented box girders with dual and triple separate boxes. The number of cable planes and the different transverse locations of pylon legs and main girders are also studied. A cable-stayed bridge with a main span of 520 m and a triple separate box girder accommodating four high-speed railway tracks and a six-lane carriageway is researched as a case study. The result reveals that the triple separate box girder is very suitable for long-span cable-supported railroad bridges with multiple railway tracks and traffic lanes.

Connections in steel structures play an essential role in the structural stability of the entire construction. This is also true in the event of a fire. The difficulty in designing connections for fire lies in the fact that not only material properties, but also loads applied to the connections change depending on temperature. In addition, the temperature-dependent decrease in the tensile strength of high-strength bolts does not coincide with the decrease in shear strength. So far, the influence of combined tension and shear loading on high-strength bolts has not been studied in detail.
This project included carrying out interaction tests on high-strength bolts of property class 10.9 (fu = 1000 N/mm2). Post-fire performance was examined in addition to observing the loadbearing behaviour during fire. The examinations were completed with additional material tests. The results were compared with earlier research on the temperature-dependent loadbearing capacity of high-strength bolts during and after temperature loading as well as on the loadbearing behaviour of bolts in steel structures under combined tension and shear loads. This investigation contributes to a better understanding of the loadbearing capacity of high-strength bolts under combined loading during and after a fire.

Many recent earthquakes clearly illustrate the importance of local ground properties for the dynamic response of structures. The dynamic response of an engineering structure is influenced by the medium on which it is founded. On solid rock, a fixed-base structural response occurs which can be evaluated by subjecting the foundation to the free-field ground motion occurring in the absence of the structure. However, on deformable ground, a feedback loop exists. In other words, when a feedback loop exists, the structure responds to the dynamics of the soil, while the soil also responds to the dynamics of the structure. Structural response is then governed by the interplay between the characteristics of the ground, the structure and the input motion. This study involved a numerical investigation of the dynamic behaviour of unbraced steel frames resting on soft ground. Two types of mid-rise unbraced steel frame, including 5- and 15-storey buildings on a soft soil deposit, were selected and analysed under the influence of three different earthquake acceleration records. The above-mentioned frames were analysed under two different boundary conditions: i) fixed-base (no soil-structure interaction) and ii) flexible-base (considering soil-structure interaction). The results of the analyses in terms of structural forces and lateral displacements for the above-mentioned boundary conditions are compared and discussed.

Numerical investigations of compression members made of aluminium are presented and recommendations for reorganizing the buckling classes and curves are derived from these. Finally, the curves are compared with test results.

The year 2014 marked the 100th anniversary of August Wöhler's death. On that occasion the authors wrote a book about his scientific work, his work on the safety of components and structures in the emerging industrial society, his life and his significance for the historical development in the field of fatigue strength of components and structures [1]. Those investigations were accomplished in cooperation with the DVM (German Association for Materials Research and Testing). As national and international events have shown [2], that book is of great interest. Therefore, an English version suggested itself, not least because of the 200th anniversary of Wöhler's birth in 2019.
August Wöhler was an interesting and unique character, influenced by the 19th century Industrial Revolution, especially railway development. The spread of that new technology caused numerous problems. Fractures occurred on axles, wheels and rails at stresses below the component's static fracture strength. Through systematic tests, August Wöhler was able to prove that repeated stresses far below the static strength might lead to fractures. The present article gives a brief overview of Wöhler's life and achievements.

Kurita-Albrecht Best Scientific Paper Award - Steel at 12th Japan German Bridge Symposium, September 2018 in Munich
The temperature dependence of the mechanical behaviour of high damping rubber bearings (HDRBs) was investigated by way of cyclic loading tests at different ambient temperatures. As the number of loading cycles increased, so the difference between the ambient and internal temperatures of HDRBs increased due to self-heating, especially in subzero environments. It was found that the mechanical behaviour of HDRBs is governed by their internal temperatures, not by the ambient temperatures. A simple method for estimating the internal temperature is proposed. Previous cyclic loading tests results at different ambient temperatures have been re-examined based on the internal temperatures obtained from the proposed internal temperature estimation method. The temperature dependence of HDRBs is summarized in terms of the internal temperatures. The conventional temperature dependence based on the ambient temperature may underestimate the seismic response of bridges with HDRBs, especially for subzero environments.

Longitudinally stiffened plates are commonly used to increase the resistance to plate buckling. Closed cross-section stiffeners provide high values of the torsional stiffness and thus are very efficient to avoid their premature lateral torsional buckling. With the current software tools, it is easy to assess the critical stress for the elastic global mode, taking into account the beneficial effect of these high values of torsional stiffness for stiffeners with closed cross-sections. However, the torsional stiffness of longitudinal stiffeners is not considered in the analytical formula proposed by Annex A of EN 1993-1-5 for the assessment of the global buckling critical stress. A study is presented here where the effect of the torsional stiffness of longitudinal stiffeners on the plate buckling criterion of EN 1993-1-5 has been analysed. 216 configurations of stiffened plates were calculated through finite element GMNIA non linear analyses. Their compressive strength thus obtained were compared to their resistance to plate buckling as predicted by EN 1993-1-5, either by considering their torsional stiffness with EBPlate software or by completely ignoring it. The study clearly shows that EN 1993-1-5 leads to unsafe results when the torsional stiffness is taken into account in the calculations. Finally, it is concluded that the torsional stiffness of longitudinal stiffeners has to be ignored in the design of stiffened plates according to EN 1993-1-5.

Liner tray wall systems are widely used for industrial and commercial buildings. Up until now, the main type used was the conventional solution with a thin (3 mm) separation strip between liner tray and outer shell. In the meantime, various solutions exist on the market to improve the thermal performance of this wall system. On the one hand, this paper deals with numerical studies that show how these new solutions reduce the heat transfer coefficient of liner tray wall systems. On the other hand, it is about the related increase in the fixing distance s1 and its influence on the mechanical performance of liner tray wall systems. Extensive experimental investigations have been performed on liner trays with a directly attached outer façade within the scope of the European RFCS Research Project GRISPE. Practicable calculation methods have been derived based on existing regulations and methods. This paper depicts by way of excerpts the results for liner trays with a directly attached outer façade for fastener distances that are not or insufficiently covered in the standards.