Composite Construction in Steel and Concrete V

The use of hybrid girders with yield strength for the web steel lower than that used for the flanges leads to a material cost saving in comparison to homogeneous girders. This paper investigates the effect of the yielded zone in the web, at its junction with the flange, on the static strength and the fatigue resistance of hybrid girders in bending. Nine fatigue tests on girders of 2 to 5 meter span with main details usually encountered in small and medium span bridges were realised and analysed. Crack growth for selected tests was simulated with a boundary-element program using linear-elastic fracture mechanics. The distances between measured restlines were evaluated by photogrammetric procedures and assigned to the recorded number of cycles. Some design recommendations in compliance with EN versions of Eurocodes concerning the static strength and the fatigue behaviour are presented.

In May 2003 a new set of codes, the DIN-Reports, were introduced by the German authority. The paper deals with DIN-Report 104 for Composite Bridges. The new bridge code is mainly based on the ENV-Version of Eurocode 4 Part 2 and also takes into account new aspects coming from drafting the final EN-Versions of the Eurocodes. The paper gives an overview of the essential design rules for the ultimate, fatigue and serviceability limit states of composite bridges and explains the interaction between the development of the DIN-Report 104 and the development of the EN-Version of Eurocode 4 Part 2.

Launching is a well-known, elegant and proven method for the erection of steel structures such as bridges. However, when the structure to be erected is of a particularly large span, with large and slender section dimensions, special attention must be given to the manner in which the support reactions are introduced into the thin girder webs. This contribution presents the considerations given and the particular actions taken in respect of the launch of the "Vaux Viaduct" over two large spans of 130 m. The monitoring of the support reactions in real time, along with the capability of correcting them, facilitated the rapid and safe erection of the steel portion of the viaduct structure. Web stresses and deformations were also measured in order to complement the support reaction data. Measurements taken during the introduction of the longitudinal prestressing in the concrete slab and temperature measurements taken during hydration of the concrete are also reported.

This paper presents an experimental and analytical study to examine the partial-interaction behavior of steel-concrete composite bridge deck consists of corrugated steel plate, longitudinal stiffeners, studs, and concrete. To examine the behavior of steel-concrete interface, push-out tests were performed for nine specimens. Then, static flexural tests for ten steel-concrete composite decks and nonlinear analysis were performed to analyze the partial-interaction behavior of steel-concrete composite deck. As the results of push-out tests, the degrees of interaction had significant influence on the ultimate load but the degree of interaction on the relative slip corresponding to the ultimate load was not so influencial. The nonlinear analysis connecting modeling of the steel-concrete interaction by interface element with the push-out test showed that it could simulate, with high precision, complicating interface behavior. The ratios of ultimate load depending on the degree of interaction of steel-concrete composite deck tended to decrease as the ratios of shear length and effective depth increased. For the steel-concrete composite deck in this paper, the critical ratio to distinguish flexural failure from slip failure was shown as approximately 5.0.

The objective of this paper is to analyse the possible use of "uncracked" elastic global analysis to design continuous beams of composite bridges at ultimate limit state provided to complete this analysis by appropriate bending moment redistributions. The pre-design of the investigated beams is performed almost automatically by optimising the cross-section resistances in both hogging and sagging bending zones. Using a non-linear finite element model, several numerical simulations are performed for three-span beams when the variable loads increase up to the ultimate limit state. The aspect of moment redistribution is pointed out comparing the hogging moments at the internal supports with those calculated using an uncracked linear analysis. The so-found percentage of redistribution depends clearly on the ratio between the adjacent spans.

For the construction of open-topped steel box girder bridges, prefabricated concrete slab could offer several advantages over cast-in-situ deck including good quality control, fast construction, and elimination of the formwork for concrete slab casting. However, precast decks with unreinforced joints should be designed to prevent the initiation of cracking at the joints, because the performance of the joint is especially crucial for the integrity of a structural system. In this paper, experimental results on a 20m-20m continuous composite bridge model with precast decks are presented. Internal tendons and external tendons were used to prevent cracking at the joints. Judging from the inelastic behavior of the bridge model, external prestressing in negative moment regions could offer better performance in terms of cracking and flexural stiffness. Simple plastic analysis was carried out to estimate the ultimate flexural capacity of the continuous bridge, and the incremental prestress of external cables after cracking was estimated.

In concrete slabs of composite structures tension forces may occur, exceeding the tension forces in reinforced or prestressed structures. Deck slabs of composite bridges are often completely cracked, even under service loading. Those cracks have to transmit the shear forces induced by concentrated loads such as wheel loads. Laboratory experiments have been carried out by the Institute of Structural Design at the University of Stuttgart, to thoroughly study the shear resistance of composite slabs cracked by tension stresses. A numerical model on the basis of the finite element method has been verified by the test results and used to investigate the influence of different parameters on the shear strength. A new concept has been developed modelling the shear failure mechanism of slabs in tension. The results yielded a design equation rendering realistic values for the shear capacity of reinforced concrete slabs in tension.

Concrete cracking has implications, not only for durability, but also for the mechanical behaviour of steel-concrete composite bridges. This paper presents a study of this influence on mechanical behaviour under non-monotonic service loading using the results of tests on composite beams and also by means of a theoretical approach. Three cracked composite beams, each measuring 7.2 m in length, were tested and a non-linear numerical model was developed. The model was used to evaluate the differences between the real behaviour of continuous composite bridges and the behaviour determined with a simple model that does not consider the influence of the cracked concrete. The tests and the non-linear model demonstrate that tension stiffening of cracked concrete is counteracted, under non-monotonic service loading, by the plastic behaviour of the concrete-reinforcement bond. Finally, design proposals have been developed which can be used by practising engineers.

A modular concept is presented for short span composite bridges that is based on simple to place prefabricated steel modules. No steel work is required on site. It allows multi-span multi lane elevated roads. Various concrete deck types can be used, from prefabricated over hybrid to cast-in-place solutions. Reinforcing and formwork is greatly simplified by the absence of concrete cross-beams. Hogging moments can be adjusted by the choice of reinforcement allowing for an easy optimization of the moment distribution. The simple moment connections are very efficient but need further investigations with respect to fatigue. A gap between cross beams of adjacent modules requires a gasket which principle is well established but needs detailed design. Because it greatly simplifies the concrete work and moves all steel work into the shop, this concept is superior in economy over other solutions and competitive to pre-stressed concrete for short-span and elevated road bridges.

Since the introduction of mechanical forms of shear connection in the 1950s to provide the composite action in steel-concrete beams, engineers typically assume full-interaction behaviour because of its simplicity and conservativeness with respect to the fatigue design of the shear connection. However, as the need to assess bridges increases for increased allowable live loads, remedial work or extended lives, so too does the need for improved analytical assessment techniques. This paper presents the simplified tiered assessment technique for continuous composite steel-concrete beams quantifying the decrease in the stress range resisted by the shear connection and the increase in the flexural stress allowing for partial-interaction. The advantages of this technique are its relative simplicity and applicability to continuous beams of any number of spans and variable span lengths, cross-section and connector distribution.

Composite bridges are an important alternative compared to concrete bridges due to considerable advantages regarding the design, construction time, durability and expenses. In this paper a survey is given with particular view on composite bridge application for small and medium spans. For this purpose, standardised bridge designs with a high grade of full- and partial depth prefabricated elements have been developed covering different carriageways and lanes. The European research project "Composite Bridge Design for Small and Medium Spans" has been conducted to investigate economic designs with special regard to sufficient robustness for execution, durability and fatigue resistance by using prefabricated solutions. A particular feature of this new design is an easy buildable concept so that it is attractive for being executed by the main contractor without problems on site. Test results of this research are presented covering hybrid girders, special Push-Out tests, conventional and dry joints in slabs, simplified connection techniques for erection and shear buckling behaviour of high strength steels.

The Sandwich Plate System (SPS) comprising steel faceplates acting compositely with a compact polyurethane core can replace steel, precast and reinforced concrete in deck-like structures. Its flexural stiffness and strength are tailored to meet particular static and dynamic requirements. Local plate buckling is precluded. The SPS is lighter, performs better and is cost competitive.

This paper presents results of tests and FEM analysis of newly proposed bare type concrete-filled steel tube (CFT) column bases, reinforced by central reinforcing bars. Strength and stiffness are discussed based on the results of tests and FEM analysis of column bases subjected to combined axial tension or compression, bending and shear.

Wall sandwich panels are used in the façades of buildings. Due to various necessities it may be necessary to cut openings which reduce its load-carrying capacity. A calculation model is presented for this case, which considers the load transfer over the longitudinal joint into neighboring elements. For the modeling of the panel interconnection the rigidity of the longitudinal joint, the rigidity in element transverse direction and the torsional rigidity are necessary. The global load model of the panel interconnection is based on a beam theory model and provides displacement and general loads. A further local model provides the stress distribution in the cross-section of the opening using the FE method. Comparisons between theoretical and experimental investigations are pointed out and the failure modes of a panel interconnection are demonstrated.

The paper describes experiments which were made at the Faculty of Civil Engineering of Brno University of Technology. Short- and long-term experiments are dealt with. The short-term experiments serve as a base for design of strengthening. The long-term experiment investigates the influence of time on the strengthened elements and on the changes in their behaviour. The short-term experiment presents series of the tests of concrete specimens, on which distinct lengths of CFRP glued reinforcement were applied. These specimens demonstrate the behaviour of the reinforcing element under tension and the adhesion between individual layers in the reinforcement area. Strains of the reinforcement are recorded together with the mode of failure of the specimens. For the long-term experiments failed RC floor panels were used. The panels were subsequently strengthened with several different types of reinforcement (CFRP strips, carbon and glass fabric). Already reinforced floor panels were stored and exposed to the long-term loading at different levels. Another factor affecting the behaviour of these panels was climatic conditions. The results of the experiments have shown that the deformations and deflections caused by the temperature changes have a reversible character. The evaluation of the measured values is given.

Design data for the longitudinal shear resistance of composite slabs, with associated safe-load tables, are provided by manufacturers of profiled sheeting, for use in the m-k design method. The data are usually based on analyses of tests, that may not be strictly in accordance with either of the Eurocode methods, as given in the final version of EN 1994-1-1 (Eurocode 4). Design of composite slabs based on the use of sheeting outside its country of origin may in future require demonstration of conformity to Eurocodes. Using the assumptions of the more recent partial-interaction method of Eurocode 4, a mechanical model for the m-k method is derived, and checked against test data. It enables resistances to longitudinal shear to be deduced from existing test results, for slabs with thickness, shear span, or properties of materials that differ from the properties of the slabs tested. This should greatly reduce the need for further testing.

The verifications that are required for the design of composite slabs (load carrying capacity, serviceability) are long and complicated. A new simplified method to calculate the moment-curvature relationship at the critical cross-section has been developed, based on the results of pull-out shear tests and a simple calculation model. Several inter-laboratory test series have been carried out to improve the pull-out test specimen and to show the repeatability and the reproducibility of the results. The tests have also enabled the operating procedures to be precisely defined. A new design approach is proposed, based on the definition of a partial connection diagram, determined through parametric analysis. Diagram proposals have been established that give the bending resistance of simply supported composite slabs as a function of span length, slab depth and type of profiled sheeting. This approach can be used for a simple, safe and economical verification by calculation of the ultimate limit state of any composite slab presenting a ductile shear connection behaviour.

Two composite Slim-Floor slabs for different ground floor designs are presented to investigate load-deformation- and vibration-behaviour on optimised cross-sections. The Slim-Floor slabs are developed for various Korean high-rise buildings in the frame of a Korean-German research project. Various Slim-Floor systems are evaluated in respect of bearing resistance- and serviceability demands, cost effectiveness and special Korean requirements. Favourite construction techniques are optimised to minimise the construction height of the cross section. The test programme is presented showing the bearing resistance and serviceability of the slab systems. Particular emphasis is drawn to the dynamic properties and the vibration behaviour due to human activities. All tests are complimented by FE-models including sensitivity analyses and parameter studies. Final conclusions are given to transfer of the research results to the design and execution of high-rise buildings.

Precast concrete floors are widely used in building construction, but there is little detailed design guidance on their application in steel-framed buildings. Traditionally the steel beams have been designed to support the precast slabs on their top flange. However, there are an increasing number of composite frames and slim floor constructions where the precast slabs are designed to interact structurally with the steel frame. Composite action can be developed by welded shear connectors attached to the steel beams and by transverse reinforcement; however, this form of construction is currently outside the provisions of the current codes of practice. This paper discusses some of the particular issues that affect this form of construction, and presents design guidance using the Eurocode methodology.

The first test evidence in Finland that the hollow core slabs supported on beams suffer from a considerable reduction of the load-bearing capacity when compared to the maximum capacity of the slabs supported on walls, dates back to the beginning of the previous decade. During the past 12 years, the behavior of the slabs has been intensively studied both experimentally and theoretically. The design code provisions were renewed in 2002 – 2003 and are now in the format of the structural Eurocodes. The provisions rely on the `composite beam model' and take into account the behavior of the hollow core slab ends as a part of a composite structure. The paper overviews the contents of the code, explains its safety concept and the background of the provisions in the light of the test evidence.

In the region of web openings in composite beams the concrete slab carries a large amount of the shear force. The aim of a research project at the Technical University of Kaiserslautern is to investigate the shear-bearing capacity of the concrete slab in combination with the shear connectors. Seventeen full-scale tests on composite beams with large web openings under high vertical shear forces have been conducted and four different failure modes were examined. The tensile forces in the studs in the region of the opening showed that the three studs closest to the high moment end of the opening transmit the biggest part of the vertical shear force from the steel web into the concrete slab and that a considerable amount of the total vertical shear force is transmitted over the opening with the help of the headed studs since they act as a kind of shear reinforcement.

For high rise buildings a combination of flat concrete slabs and composite columns is becoming increasingly common and leads to very economic structures. Where composite columns with small dimensions are used, often the resistance to punching shear is relevant for the slab thickness. In order to avoid localized thickening of the slab, especially in case of small dimensions of the columns, additional reinforcement systems with double headed studs often used. Another possibility is the use of structural steel shear heads integrated in the slab. The paper deals with the development of a new composite shear head system. Test results, Finite-Element studies and a simplified design model are presented for this new type of composite shear head system.

Secondary composite beams incorporating trapezoidal profiled steel sheeting laid transverse to the longitudinal axis of the steel beam can exhibit unwanted brittle failure modes with low stud strength and poor ductility. Tests in Australia have identified a special type of longitudinal shear failure occurring in the concrete flange over the top of the sheeting ribs. As a consequence, it is mandatory in Australian Standard AS 2327.1 "Simply-Supported Composite Beams" to provide a waveform reinforcing component to prevent this shear failure. Further tests have confirmed the occurrence of brittle concrete pull-out failure, which is particularly prevalent when pairs of studs instead of single studs are used in each pan. The same waveform reinforcing component can be used effectively to prevent pull-out failure, which results in increased stud strengths and much improved ductility.

In 1989, the Steel Construction Institute (SCI) produced a design guide that presents simplified rules for designers to calculate the serviceability response of composite floors from occupant-induced vibrations. However, advancements in composite construction have led to much longer, slender floors than was originally considered in the development of this design guide. Furthermore, due to the fact that much more stringent requirements are now demanded in specialist applications (e.g., hospital operating theatres, etc.), design's for composite floor systems are now becoming limited by vibration serviceability criteria. Over the last 6-years the SCI have been undertaking in situ vibration tests on a wider range of floor systems and types than was considered in the calibration of the original SCI design guide. This paper presents the results from these tests and compares the measurements with the SCI publication and AISC/CISC design guidance. Following this review, a simplified design method for estimating the response of composite floors is proposed, and is compared with test results.

The design of edge beams in slim floors using precast hollow core slabs (PCHS) is governed by torsion. Three large scale tests were carried out at the Darmstadt University of Technology and their results will be reported. The test specimens measured 6 m by 6.5 m and were loaded in 36 points. A finite element analysis has been made to assess the behaviour of the beams and to develop the basis for a simple design procedure. This procedure is compared to the FEM and the test results.

In recent years the application of cellular beams is more and more requested by architects and designers. Not only their aesthetical appearance as visible steel structures and economic design are popular, also the possibility to pass service integrations through the openings is given. In the scope of the ECSC project 7210-PR-315 "Large web openings for service integration in composite floors" the performance of beams with web openings was studied experimentally and theoretically. The aim of this paper is to present the investigation on beams with multiple regular web openings with special focus on different failure mechanisms. The test program comprehended large scale tests with the following objectives: (1) Composite action close to support, (2) Behavior of elongated openings, (3) Load introduction via secondary beams, (4) Asymmetric design, and (5) Stiffening of web openings. On the basis of the test results and theoretical investigations including numerical simulations, design recommendations were established.

Haunches can be formed as part of a solid slab to increase composite beam depth or they can occur directly in primary composite beams incorporating open rib trapezoidal steel decks with the concrete ribs, often narrow, laid parallel to the longitudinal axis of the steel beam. In solid slabs, haunches can be designed to provide full shear connector strength. In composite slabs, the geometry of the concrete ribs is limited by the deck geometry. This can result in a reduction in stud connector strength below that for a solid slab and is taken into account in design codes by the use of stud strength reduction factors. Brittle failure modes have been observed in recent testing on specimens incorporating trapezoidal steel decks. The reductions in stud strength were significant and could be more severe than those predicted by design codes. Two types of reinforcing components have been developed to control the brittle failure modes and to provide increased stud strengths and improved stud ductility.

Non-ductile shear connections in composite flexural members are defined as ones where the connection shear forces cannot effectively re-distribute to a plastic resistance value prior to developing a certain maximum end slip. Typical non-ductile connections are those based on the bond of the concrete, such as the grouted joints between the decking elements and the beam in the shallow floor construction. The paper discusses the effects of the non-ductile shear connections, including the zip-flyer effect and its consequences in the load-deflection behavior of various types of beams. The influence of the beam span, relative connector density and end slip are considered. Typical structural responses, as analysed by the non-linear finite element method are reported.

Aim of the project is to measure deflections of standard steel concrete composite beams, used in office and industrial buildings, assess and analyse the reasons for those deformations and develop simple design aids for a realistic calculation of the deflections. This research is necessary, since experience shows that the calculated results do frequently not match the deflections measured on site. This can reduce the serviceability of a structure and lead to high rectification costs. It should be possible to take structural influences on the deflection, e.g. details of supports as well as imperfections during production of the beams and erection into account while dimensioning the beam, to get realistic results. We focus on the differences between calculations and reality by analysing the measured deflections of beams in structures under construction. The factors that lead to the differences will be pointed out and the reasons for them will be discussed.

Composite girders with reduced height and slim floor girders form an attractive alternative to common reinforced concrete beams. However, their economical design suffers from assumptions concerning the effective width of the concrete chord that are not reflecting the real behaviour. This paper reports on experiments of six test girders with large concrete chords as well as on analytical and numerical evaluations of the effective width for such sections. First results promise an advantageous possibility to increase the assumed ratios of the effective width for the design of these sections in serviceability.

This paper describes the design of the structure of a multi-story industrial building. The selected structural system is composed of steel underspanned beams, prefabricated concrete elements and a layer of cast-in-place concrete. The 12.5 m long principal beams are spaced every 2.5 m and supported on 7.5 m support beams of similar construction. The upper chords of the principal beams, which are triangular in shape and include a ribbed reinforcing bar, are integrated within the thickness of the concrete slab. Composite behaviour is attained through adhesion of the reinforcing bar. No formwork is required in the assembly. The technical and economical advantages of the structural system result from innovative solutions adopted to satisfy the modularity of the project, the need for large spaces free of columns, the high dynamic service loads and the density of the technical installations. A dynamic analysis showed the sensitivity of the composite floors to vibrations. A cracking analysis showed that such a system consisting of steel beams embedded in the concrete resulted in a high probability of appearance of large cracks concentrated at the locations of the discontinuities. Tests conducted on a prototype permitted the verification of the various hypotheses related to the resistance, rigidity, cracking and dynamic behaviour of the composite structural system.

This paper presents the results of an extensive experimental program designed to consider the effects of partial shear connection for beams and joints subjected to hogging moments. The experimental program elucidates some important concepts and highlights the benefits that may be attributable when the use of partial shear connection is invoked. In particular the experiments illustrate that the effect of partial shear connection provides increased rotational capability without a significant compromise in strength. This is due to the ability of the beams to exhibit partial interactive behavior and this may be able to be utilized when trying to justify plastic analysis for continuous and semi-continuous composite beam systems. The paper concludes with some suggestions for further research.

Composite structures made of structural steel and concrete maximize the advantages of the two components. Using high strength materials and partially concrete encased steel profiles is becoming more and more popular and thus also more interesting for composite construction. Therefore, the "German Research Foundation" DFG has supported the research project "Load Bearing Behavior of Concrete Encased Composite Beams using High Strength Steel and Concrete". The main aspect was to evaluate the load carrying behavior of concrete encased steel profiles with and without longitudinal shear connection. The most remarkable results regarding the load bearing behavior of high strength encased composite beams are presented in the following. Furthermore, the results of three-dimensional finite element analysis describing the interaction between concrete encasement (web encasement) and steel profile are introduced and the load bearing mechanisms are described.

The current codified requirements for the minimum number of shear connectors in composite beams are based on the span and in some cases the ratio of flange area. The backgrounds to these limits are studies on the slip in the connectors for uniform sections under uniform distributed loads. A method for calculating the slip using elastic properties for the beam and rigid-plastic properties for the studs has been developed by others. This method has been developed to include typical stress strain material curves and elastic deformation of the connector. The results of the method will be compared with other published values. The method will also be used to make preliminary studies of the effect of variation in certain parameters. Variations that will be considered include, ratio of construction to total load, loads less than the beam capacity, no connectors at the ends of the beam, tapered sections and asymmetric point loads.

This paper deals with the ductility (rotation capacity) and moment resistance of the negative moment region of continuous composite beams with thin endplate connections. This study of 16 cantilever beam tests, reveals that the most important parameters in causing premature fracture of the reinforcement or limited connection rotation are thick endplates, small ratios of reinforcement force to steel beam force and large ratios of negative moment resistance at first yield to moment resistance of the connection at first yield. The level of the plastic neutral axis is used to explain the influence of these parameters. Design guidelines are provided to avoid reinforcement fracture and improve connection rotation.

When drafting EN1994-1-1 the simplified design method for composite columns in ENV1994-1-1 was changed using second order analysis and equivalent initial bow imperfections taking into account the effects of residual stresses and geometrical imperfections. These equivalent geometrical imperfections were determined for different types of cross sections, for steel grades S235 (f_y = 235 N/mm²) and S355 (f_y = 355 N/mm²) and concrete grades from C30 (f_ck = 30 N/mm²) to C60 (f_ck = 60 N/mm²). In addition, the method was checked for the use of high strength steel, so as to determine whether the basis of the simplified design method and the full plastic interaction between normal force and bending moment can also be applied. The paper deals with the necessary modifications of the simplified design method for the use of high strength steel, the classification into buckling curves, and the background of the determination of the equivalent initial bow imperfections. Some comments for special sections for composite columns are also given.

This paper summarizes research in which the damage evolution of rectangular concrete-filled steel tube (RCFT) members, connections, and frames, was documented using results from experimental tests from around the world. Two types of damage indices were then defined and used for each type of RCFT component to quantify damage. Based on the quantified values of damage, each local damage level identified for the RCFT members was mapped onto appropriate performance levels including Immediate Occupancy, Life Safety, Near Collapse, and Collapse Prevention, to identify how this information may be used in the context of a performance-based design methodology. The experimental data recorded in the database was also used in developing an analysis model to simulate RCFT beam-columns for use in assessing seismic demand and capacity in RCFT structures. A mixed fiber-based finite element formulation was developed allowing geometrically and materially nonlinear analysis RCFTs and good correlation with the experiments was attained.

Based on experimental research work new design models for the load introduction in composite columns with concrete filled hollow sections were developed taking into account the effects of partially loaded areas and the effects of the restraint by the hollow section. The presented models are the basis for the design rules of composite columns in prEN 1994-1-1. It is verified by tests that for the load introduction reinforcement can be taken into account for the resistance of the column, even where the reinforcement is not welded or in direct contact to end plates.

Concrete filled steel tubes (CFTs) are efficient members in structural applications including bridges, buildings and piled foundations. Their design has been based on theory and tests on columns under load applied axially or at relatively small eccentricities. For loading at large eccentricities or loading in pure flexure in the beam situation, there has been relatively little research performed and design has been based on extrapolation from the column situation. Therefore, an extensive research program on the flexural behavior of CFTs has been undertaken at the University of Western Sydney with a particular emphasis on thin-walled steel tubes that can provide an economical form of construction. Numerical studies have been used to investigate the influence of imperfections and plasticity on the behavior of bare steel tubes and this work was extended to examine the effect of the concrete infill on the behavior, particularly local buckling of the steel tube. A series of flexural tests has been carried out on full-scale tube specimens, both filled and unfilled. Accurate measurements of the imperfections were taken using an innovative method allowing the complete imperfect shape of the tube to be determined. This was needed for the analytical studies. The results have allowed the influence of the concrete infill on the flexural stiffness and strength of the tubes to be established. The test results and the analytical studies are being used to develop a design procedure that accounts for imperfections, local buckling, concrete strength, steel strength and the interaction between the tube and the infill.

The purpose of this research is to develop new type of frames made of reusable members. The frame consists of three parts: circular CFT (concrete-filled tube) columns, wide flange steel beams and beam-to-column connections that are made of circular tube with wide flange shape brackets. These parts are all prefabricated in the shop. When we reuse the structural members of the frame, all members must be taken apart easily. For this reason, beams are connected to beam-to-column connection by high-strength bolts and splicing plates, and columns are just set into the tube of beam-to-column connection. The gap between two tubes (CFT column and beam-to-column connection) is filled by high strength mortar. In order to clarify the elastic-plastic characteristics of this frame, he T-shape frame models consisting of CFT column and beam-to-column connection were tested under the constant axial load and cyclic horizontal load.

Concrete filled tubular (CFT) frames have been used in high-rise buildings in Japan since the early 1990s, and their use is increasing. However, there is a need to seek more cost-effective connection details, and this is especially true for square tubes. To meet this need, a relatively simple and cost-saving solution, in which a connection is stiffened with two interior T-shaped diaphragms, was investigated. Based on the obtained test results and observations, strength models and design formulas were derived using a limit analysis technique. Comparison with the test results showed that the proposed formulas provided accurate strength estimates.

This paper is concerned with the behaviour of beam — column connections of steel — concrete composite beams with precast hollow core slabs. Experiments were carried out to investigate the joint rotation characteristics and ultimate moment capacity of these connections. Details of the test specimens, instrumentation, test set-up and test procedures are described. Results obtained for the connection moment capacity, rotation capacity and failure modes are presented. It is found that through proper design and detailing, these simple steel connections display the characteristics of a semi-rigid connection with very little extra cost.

High ductile composite frame structures can represent a competitive solution in seismic-prone areas if they are endowed with efficient structural solutions. In detail, the introduction of dissipative partial-strength beam-to-column composite joints can lead to the formation of global mechanisms characterized by large inelastic rotations and therefore great amounts of dissipated energy. Monotonic and cyclic tests were conducted on subassemblies at the University of Pisa. The inelastic response of joints was divided mainly between shear yielding of the column web panel and inelastic bending of the extended end plate and column flange. Also yielding of the rebars and concrete crushing of the slab against the column flange took place. In order to guarantee a ductile behavior of partial-strength beam-to-column composite joints, it is needed a suitable capacity design of joint's mechanical components. Nonetheless, mechanical models of composite joints under seismic actions are not fully defined yet and many doubts still exist mainly with regard to strut and tie mechanisms in the reinforced concrete slab governing the force transfer between beam and column. In the paper, mechanical models of interior and exterior beam-to-column joints are presented on the basis of experimental results obtained in monotonic tests, through the analysis and modeling of each component part governing their structural behavior. A preliminary validation of the models is presented and discussed.

The present work has been an intricate collaboration between architecture and engineering. The basic idea envisaged the bridge assembly as a large minimalist sculpture set over the river. The system is formed by two very sturdy elements which are simple in form: (1) A large rectangular prismatic bar in weathering steel, serving as the deck; (2) A central element with a sharply tapered lenticular section, serving as the central pier and set orthogonally to the deck and parallel to the river current, with a stone external facing. In order to obtain the great slenderness demanded by the design, it was necessary to place a triple internal structural mechanism: (1) Elastic encasing cells at both abutments. (2) Rigid connection between pier and deck. (3) External prestressing at mid-span and over the central pier. The cross-section consists of two lateral box girder elements with double composite action connected by transversal composite trusses.

In the Netherlands the decision to use composite structures is not automatic. Engineers and contractors normally think in concrete or steel structures. However the last few years the use of composite structures has increased. The use of composite structures has several advantages compared to the traditional structures in steel or concrete. By presenting a number of characteristic projects the arguments for the use of composite construction will be discussed. Experiences gained in the design and execution of various composite structures will be reviewed and some important aspects will be discussed in more detail. Finally based on these experiences needs for further development of knowledge and design information will be given. In the Netherlands Eurocode 4 is used for the design of composite structures. To make design of composite structures more attractive for engineers, proposals will be presented to improve parts of the code and to add new subjects to the code.

The Nelson Mandela Bridge is a landmark quality structure suspended across a maze of railway tracks connecting Braamfontein with the suburb of Newtown in the city of Johannesburg. The bridge carries a 10.6m wide, two-way, single lane roadway with 1.5m wide pedestrian walkways both sides of the roadway. The site conditions required a 284m long cable stayed bridge with a 176m main span and side-spans of 42m and 66m, The resulting asymmetric cable stayed structure has an instantly recognisable profile, which will be associated with Johannesburg all over the world. The Nelson Mandela Bridge is an example of the structural use of both steel and concrete in composite construction. The composite construction is utilised both for the pylons and the main span deck and represents a significant departure from the reinforced concrete bridges common in Southern Africa. This paper will concentrate on the composite components of this bridge and discuss some practical design and construction aspects.

This paper briefly summarizes the results of an experimental research program involving the testing of push-out specimens with channel shear connectors. The test program consisted of three series, each with twelve push-out specimens. In each series, six specimens had solid concrete slabs and the other six specimens had concrete slabs incorporating wide-rib profiled metal deck with ribs parallel to the beam. The test parameters included the compressive strength of concrete, length and web thickness of the channel shear connector. The test results showed that, for a given length of channel, the concrete strength dictates the failure mode. In specimens with higher strength concrete, failure was caused by the fracture of the channel web. Crushing of the concrete adjacent to the channel web was the observed mode of failure in specimens with solid slabs when lower strength concrete was used. In most of the specimens with metal deck slabs, a concrete shear plane type of failure was observed.

Modern standards such as Eurocode 4 provide rules for determining the resistance of stud connectors welded within profiled steel decking. These rules are based on the results of push tests with specimens using both re-entrant and trapezoidal decks. This paper presents interim results from a push test study on specimens using modern trapezoidal deck types, which are supplemented with results from two full-scale beam tests and Finite Element analyses. From this study it is shown that: there is a need to standardize the push test when decking is employed; slip capacities lower than the Eurocode 4 recommendations do not significantly reduce the ductility and bending resistance of composite beams; and the artificial restraint provided by edge trim is generally beneficial in push specimens, and better represents the stud behaviour in composite beams.

Headed studs are widely used in composite bridges to provide longitudinal shear force transfer in the interface between concrete deck and steel beam. There is still no theoretical model available to describe the local failure mechanism and the resistance under the monotonic loading of the stud. The standard design of shear studs assumes that the static resistance is independent of fatigue. A test series on 12.5mm studs published by Oehlers however indicates a linear reduction of the static resistance with the number of cycles. In order to shed more light on this question a test series on 22mm studs was performed, which is the common size in bridges. The residual static resistance of shear studs after cyclic loading was measured using push tests according to Eurocode 4-1-1. The test program consisted of two series of tests. Five static tests were performed for establishing a reference static resistance and five residual strength tests were used to measure residual static resistance after cyclic preloading. In order to give an insight into the local failure mechanism under the monotonic loading a realistic 3D FE analysis of a single stud embedded in the concrete is presented. Results obtained from the FE model are compared with experiments and rather good agreement is obtained.

In consideration of structural, manufacturing and economic aspects of construction new efficient composite cross sections for buildings and bridges consisting of reinforced concrete slabs and structural steel beams lead to a horizontally lying arrangement of headed shear studs influenced by edge proximity. Compared to common vertical shear studs, the relatively small distance of horizontally lying shear studs from the edge of the slab leads to different bedding conditions of the connectors within the surrounding reinforced concrete material, which results in differing structural and fatigue behaviour of this innovative type of shear connection. In continuation of previous investigations about monotonic longitudinal shear, this paper deals with monotonic vertical shear, monotonic vertical and longitudinal shear, and cyclic longitudinal shear action. On the basis of experimental and theoretical investigations, design rules for future applications are presented.

The traditional method of designing a shear connection for composite structures loaded in flexure is to employ mechanical connectors; normally headed steel studs on top of the steel section. In new types of composite beam where most of the beam is embedded in the concrete, another method of arranging mechanical connection may be employed: the dowel action between the steel web and concrete appearing in the openings through the steel section. When developing applications for such connections, the shear behavior of the concrete dowel should be known. High initial stiffness is typical for concrete dowels, but the shear resistance will depend on the out-of-plane geometry of the opening. A total of 74 push-out specimens were tested in 2002 – 2003 to find out the shear-slip properties of the concrete dowel connection. Behavior of various edge profiles in the circular plate opening is studied. Various fracture mechanisms observed in the concrete dowels are reported in this paper.

The development of a new shear connection for steel-concrete composite bridges with full-depth precast prestressed concrete deck is presented in this paper. This connection, by "partial adherence", developed by Dauner, is particularly well suited for rapid bridge erection with high durability. Push-out tests and direct shear tests have been conducted to investigate the effect of several parameters on the shear strength of this new connection, as well as on its fatigue and creep behaviour. The main results of these tests are described. The objectives and the methodology of the theoretical research that is being currently conducted to model the behaviour of these promising innovative connections are presented.

The paper deals with new research results on the fatigue behaviour of embedded composite shear studs. Measurements of the slip-behaviour of real two-span composite girders with realistic cyclic loading allow conclusions to simulate the behaviour by so called cyclic displacement-controlled Push-Out-Tests. It can be shown that a displacement-controlled Push-Out-Test simulates far better the slip-behaviour of a composite girder than conventional load-controlled Push-Out-Tests. By using the test data from displacement-controlled Push-Out-Tests models are presented to describe the cyclic load-displacement behaviour. A slip-oriented Wöhler-line has been developed. In a second step the crack propagation of the studs during cyclic loading has been observed. Correlations between crack-propagation and number of load cycles can be derived. It can be shown, that fatigue of shear studs is dominated by crack propagation.

This paper describes a study on the minimum degree of partial shear connection for steel-to-concrete composite beams taking into account the non-linear load-slip behaviour of the shear connectors and further parameters. The minimum degree of shear connection for a large number of cases is determined based on the results from comprehensive nonlinear numerical simulations. The numerical model used is able to describe the post-ultimate limit state behaviour and thus allows for investigating the ductility of a composite beam. Based on the studies shown below the design rules given in Eurocode 4 for composite beams with partial shear connection (based on a characteristic slip of δ_uk = 6 mm) can be extended to characteristic slip values from 1 mm to 15 mm. While the values for the minimum degree of partial shear connection given in Eurocode 4 were confirmed for composite beams with two single loads, some differences were found for composite beams exposed to uniform load.

Calculations for the determination of the load bearing capacity of composite columns under exposure to fire will be presented. The calculations are a part of a running research project at the Darmstadt University of Technology. The main attention of the research project is given to concrete filled hollow sections with an embedded steel section used in high rise buildings because this application tends to cause the greatest problems due to the lack of codified rules. Firstly the temperature distribution over a cross section of a composite column accomplished with the FE-Program ANSYS will be explained. The influence of thermally induced stresses on the load bearing capacity will be illustrated. Thermally induced stresses can be assigned to structural imperfections and have to be considered for the determination of the load bearing capacity of composite columns under high temperatures. The geometrical imperfections and eccentric loading will be also included within the mechanical analysis.

In one of recent European research projects, the fire behaviour of steel and concrete composite structures has been investigated with help of real fire tests on full scale structures. The experimental evidence shows that fire protection of steel and concrete composite structure is not necessary under car fires if certain conditions are met. Further to this project, based on a 3D numerical modelling validated against the experimental results within this research project, a parametric study according to a specific fire safety engineering approach was performed in France with the objective of developing a practical design guide for open car parks using unprotected steel and concrete composite structures. The principle of the engineering procedure mentioned above, the results of this parametric study and the corresponding design guide will be presented and discussed in this paper.

Firstly a short presentation is given of an advanced finite element model specifically developed by the authors for simulating the mechanical behaviour and resistance of composite columns with hollow steel section exposed to fire. In particular, the model is able to take into account the more or less interaction resulting in slip at the steel-concrete interface and to consider columns with high buckling lengths (by means of 2^nd order geometrical analyses). This model was proved to be in good agreement with several fire tests carried out in France, Germany and Canada. The novel aspect of the paper is related to the effect of differential thermal stresses on the buckling resistance of axially loaded columns, which appears very significant when the percentage of reinforcement in the concrete core is low, the fire duration does not exceed R60 and the relative slenderness of column is high. The results of a larger parametric investigation are given in terms of appropriate buckling curves which are more unfavourable than the usual curve c specified in Eurocode 4 part 1.2. A few other results in relation with the same aspect are also shown in the case of composite columns in compression and bending.

Coupled core walls offer an efficient lateral load resisting system. Performance of coupled wall systems depends primarily on adequate stiffness and strength of coupling beams. Improvements in terms of constructability and performance are being explored as part of an ongoing study. A repairable or replaceable central "fuse" beam, where the inelastic deformations are concentrated, has been developed in an effort to protect the wall piers from local damage imparted by the coupling beams. A composite coupling beam utilizing an encased steel web plate is anticipated to facilitate construction of coupled core walls. Design methodologies for these systems were developed and evaluated. A summary of design methodologies along with measured responses is presented in this paper.

Three monotonic and nine cyclic tests have been performed at the National Institute of Applied Sciences (INSA) of Rennes-France on bolted beam-to-column steel and composite moment resisting joints. For composite joints the slab is composite with a profiled steel sheeting connected by means of welded headed studs to the steel beam. Both external and internal joints of frames are considered with two types of loading: symmetrical and unsymmetrical. The dissipation of plastic energy is made either inside the joint using thin end-plates, or outside the joint in the connected beam using bolted end-plates strengthened by adjacent haunches. The cyclic testing procedure recommended by the European Convention for Constructional Steelwork (ECCS) has been applied in order to simulate sever seismic actions. The experimental results are discussed in terms of rotational stiffness, moment resistance, rotational ductility and low-cycle fatigue with comparisons to analytical models given in Eurocodes 3 and 4.

This paper shows the feasibility of composite MR frame structures with partially encased columns and partial strength beam-to-column joints to provide strength and ductility. In detail, energy dissipation is concentrated both in column web panels which are not surrounded by concrete and in composite beam-to-column connections. A full-scale two-storey composite building was used to validate the system performance of composite MR frames with partial strength joints. The frame structure was subjected to pseudo-dynamic tests in order to simulate the structural response under ground motions. The ground motion for 10 % chance of exceedence in 10 years earthquake hazard caused minor damage while the one for 2 % chance of exceedence in 10 years earthquake hazard entailed column web panel yielding, connection yielding and plastic hinging at column base joints. An earthquake level chosen to approach the collapse limit state induced more damage and was accompanied by further column web panel yielding, connection yielding and inelastic phenomena at column base joints without local buckling. Successively, the structure was subjected to a final quasi-static cyclic test with interstory drift ratios over 4.6 %. Moreover, test offered additional opportunities to validate the performance of simulation FE models. Exploiting inelastic static pushover and time-history analysis procedures, behavior factors, design overstrength factors and the ductility demand of the structure were estimated.

This paper presents numerical and analytical studies carried out at Liège University, as part of a European research project, with the objective to investigate the in-plane behaviour of composite sway frames under static loading.

This paper evaluates and compares between the seismic behavior of hybrid coupled wall systems that are 12- and 18-stories high. System response is studied using validated, transient finite element simulations that account for the effects of concrete cracking, crushing, and steel yielding. Suites of analyses are conducted to investigate pertinent parameters including interstory drift ratios, wall rotations, roof displacements, dynamic base shear magnification, strain demands, and coupling beam plastic rotations. The results for both systems confirm that hybrid coupled walls are particularly well suited for use in regions of high seismic risk. Although the responses of both structures were essentially dominated by the first vibration mode, higher mode effects manifested themselves by causing substantial base shear magnification. The effect was somewhat higher in the taller structure.

Possibly, the most complicated and time-consuming task in experimental research is represented by the phase of data interpretation, especially for what concerns the relationship between the local data and the overall experimental behavior of the structures. One of the advantages in adopting the by-component modeling approach for the analysis and simulation of composite connections in structures is that it is possible to closely simulate the local behavior of a number of connecting members, and to put it directly in a relationship with the overall simulated behavior, thus highlighting patterns and providing explanations, which might not have been immediately discernible just by examining the recorded data separately. The paper discusses an experimental test involving two one-story, two-bay steel-concrete composite frames with partially-restrained connections, subjected to a series of loading/unloading cycles up to failure. The experimental data is interpreted by means of a mechanical model of the connections, specifically developed for the task at hand. The analysis of the structural behavior by means of the proposed modeling approach allows to closely relate the experimentally observed phenomena to the local behavior of single components, such as the connecting elements, the steel-concrete interface, the column web panel, etc, with an accuracy that would not else be possible. The by-component modeling approach demonstrates to be a very useful tool in data interpretation, especially for the case of partially-restrained connections, the behavior of which is generally very difficult to assess.

Coupled core walls offer an efficient lateral load resisting system. Performance of coupled wall systems depends primarily on ensuring that coupling beams provide adequate stiffness and strength. Current design guidelines, which are based on a strength-based design approach, often result in beams with unrealistic details. Previous studies have also demonstrated that coupled core walls behave differently from what is assumed in a traditional strength-based design approach. Therefore, currently available strength-based design methodology is inadequate to address a large class of coupled core wall systems, and alternative design approaches are necessary. In this paper, a rational approach based on a performance-based design method is demonstrated through the design and analysis of a prototype ten-story coupled core wall structure. Critical responses of the prototype structure at different limit states under a number of representative ground motions are compared in order to evaluate the adequacy of the performance-based design method.

This paper describes an experimental study that assesses the influences of infill and member slenderness on the performance of earthquake resistant hollow steel brace members. Two series of tests are described. In the first series, hollow and filled specimens were subjected to cyclic axial loading, and their hysteretic responses observed. Eventual failure of the stockier specimens occurred due to tensile rupture after initial fracture caused by local buckling in the post-buckled state. The effect of the infill on this mechanism is described. In the second experimental series, shaking table tests were carried out to determine the influence of the brace parameters on dynamic inelastic response. By normalising the hysteresis loops of the braces obtained from the shake table tests to the yield strength of the steel sections from quasi-static tests, the influence of the infill on the strength of brace is quantified, together with dynamic effects.

The seismic responses of two composite structures are presented in this paper, considering time-history dynamic analyses based on artificial accelerograms, which are consistent with the European C-soil spectrum. The structures have been pre-designed in a usual manner by elastic push-over analyses, assuming their location in a seismic zone of maximum intensity 0.35g. The dynamic analyses reveal the applicability of the simulated joints for the considered accelerograms, scaled in order to correspond to the design ground acceleration. The results are given in terms of required elasto-plastic rotations, maximum inter-storey drifts and behaviour factor q. In particular, these parameters are discussed with regard to recent criteria of ductility class specified in Eurocode 8. For the beams and columns, appropriate fiber finite elements are used. The beam-to column joint behaviour is integrated into the structural modeling by means of a sophisticated finite element that models with good accuracy the experimental behaviour of joints, tested previously at INSA of Rennes.