Search

A comprehensive experimental and numerical study on flexural buckling behaviour of welded stainless steel box-section beam-columns is presented in this paper. A total of ten test specimens were prepared and tested subjected to eccentric compression. The cross-section and length of the test specimens were carefully designed by referring to existing design specifications to be insusceptible to local buckling. The full experimental load versus deformation and load versus strain curves were obtained. Prior to member testing, both material tests and geometric imperfection measurements were conducted and reported in detail. Finite element (FE) models were developed and validated against the obtained beam-column test results afterwards, which were utilised to perform parametric studies on the flexural buckling behaviour, generating sufficient numerical data points. Based on the obtained test and FE results, the design methods provided in EN 1993-1-4+A1 and AISC DG 27 were assessed. It was shown that both EN 1993-1-4+A1 and AISC DG 27 provided generally conservative and scattered predictions for the buckling resistance of welded stainless steel box-section beam-columns. Moreover, a newly revised proposed interaction factor was proposed in the framework of EN 1993-1-4+A1, and strength predictions with reduced scatter were therefore achieved.

The objective of this work is to develop design guidance for existing hollow steel columns that are retrofitted by infilling concrete into the tubes. The primary challenge is the unknown effect of the existing preload on the steel columns prior to the concrete infill. Composite performance and buckling behaviour of circular hollow section (CHS) steel columns strengthened by infilling concrete under preload was experimentally and numerically investigated in this study. A total of 34 CHS steel columns were tested under pin-ended boundary conditions, and the overall buckling failure modes and corresponding ultimate buckling resistances were recorded. Prior to the member testing, material properties of the steel columns and the infilled concrete were attained. By means of the finite element (FE) software package ABAQUS, elaborate FE models for the CHS columns strengthened by infilling concrete were developed and validated against the obtained test results, which were further verified with other available test data. Using the validated FE models, systematic parametric studies were conducted to examine the influences of the major factors affecting the ultimate capacities of the CHS columns strengthened by infilling concrete, including preload ratios, steel and concrete strengths, steel ratios, initial global imperfections, eccentricity ratios and column slenderness ratios. The obtained test and numerical results were therefore utilised to develop design criteria for predicting the overall buckling resistance of CHS steel columns strengthened by infilling concrete. In view of the difficulty of determining the key parameters in practice and the uncertainty of the strengthening process, a new simplified design coefficient was proposed, taking into account the influence of the preload. It has been demonstrated that accurate and reasonable strength predictions can be provided by the proposed design method.

This paper complements previous research conducted by the authors on stainless steel bolted joints and reports the first systematic study on the structural behaviour of stainless steel beam-to-tubular column joints with the use of Hollo-bolts. Six full scale specimens including two flush plate joints, two top and seat cleat connections, and two top, seat and web cleat connections of single-sided beam-to- hollow section column joints have been tested and the results are reported in detail. All specimens were in Grade EN 1.4301 austenitic stainless steel and were designed in such a way that a variety of failure modes could be obtained including plastification of the tubular column face, of the connecting angle cleats or the flush plates. For each specimen the moment-rotation response, the evolution of strains in critical locations of the connections and the separation between the column face and the connected beam was monitored and is reported herein. In parallel with the tests, advanced nonlinear FE models were developed. Following validation against the obtained experimental results, parametric studies were conducted to study the effect of key parameters such as, material grade, column face thickness, end plate/angle cleat thickness and bolt spacing on the stiffness, strength and ductility of the joints. Based on both the experimental and numerical results, it was verified that the connections displayed excellent ductility and attained loads much higher than the ones predicted by design standards for carbon steel joints.

The shear behaviour of stainless steel plate girders investigated in this study is related to the introduction of diagonal stiffeners. Four plate girders with or without diagonal stiffeners were fabricated by welding hot-rolled stainless steel plates of different thicknesses. All the plate girders were tested to failure subject to shear loading. The critical shear buckling stress, ultimate resistance and post-peak response were recorded and carefully analysed. Elaborate finite element (FE) models were developed by means of the FE software package ABAQUS and were validated against the test results. A comprehensive parametric analysis of key parameters including the web aspect ratio and slenderness, the flexural stiffness of diagonal stiffeners and the material properties was further conducted, reflecting the influence of these parameters on structural responses. The critical shear buckling stresses obtained from eigenvalue buckling analysis were compared with theoretical predictions of the elastic buckling stress, and a new calculation approach for the shear buckling coefficient of diagonally stiffened web panel that can account for the flexural stiffness of diagonal stiffeners and the effective restraint from flanges was proposed. Based on the obtained ultimate shear resistances, the existing design methods were assessed; and a new design proposal within the framework of EN 1993-1-4 + A1 for predicting the shear resistance of diagonally stiffened stainless steel plate girders was developed. In this novel method, the shear contributions from diagonal stiffeners and the shear buckling coefficients for diagonally stiffened web panels were revised. The reliability of the new proposal was also confirmed by further statistical analysis.

A comprehensive experimental study on structural behaviour of stainless steel bolted T-stub connections is presented in this paper. A total of 27 stainless steel bolted T-stubs with various geometric configurations were fabricated from hot-rolled stainless steel plates and assembled with stainless steel bolts. Two stainless steel grades − austenitic EN 1.4301 and duplex EN 1.4462, and two other types of stainless steel bolts − A4-70 and A4-80, were introduced in the experimental programme. Tensile coupon tests were performed to determine the material properties of the stainless steel plates and bolts. The bolted T-stub specimens were tested under monotonic loading, and ultimate resistances, plastic deformation capacities and failure modes were obtained. Based on the experimental results, the existing design methods for predicting tension resistances of the bolted T-stub connections, including design provisions in EN 1993-1-8, the continuous strength method (CSM), AISC manual and JGJ 82 and other design formulae for T-stubs with four bolts per row, were all evaluated. It was indicated that all the existing design methods provided generally conservative predictions for stainless steel bolted T-stub connections.

Research on stainless steel structures has primarily focused on the structural response of individual members, whilst the response of joints has received far less attention to date. This paper reports for the first time full-scale tests on stainless steel beam-to-column joints, subjected to static monotonic loads, whilst the companion paper reports numerical studies on similar connection typologies to the ones studied herein. The joint configurations tested include one flush and one extended end plate connection, two top and seated cleat connections, and two top, seated and web cleat connections of single-sided beam-to-column joints. All connected members and connecting parts including bolts, angle cleats and end plates are in Grade EN 1.4301 stainless steel. The full moment-rotation characteristics were recorded for each test and the experimentally derived stiffness and moment resistance for each joint was compared to the codified provisions of EN1993-1-8. It was verified that the connections displayed excellent ductility and attained loads much higher than the ones predicted by design standards for carbon steel joints.

A comprehensive experimental study on the structural behaviour of stainless steel T-stubs in Grade EN 1.4003 ferritic stainless steel and employing A4–80 stainless steel bolts is reported in this paper. A total of 17 ferritic stainless steel T-stubs of various geometric configurations including both single and double bolt rows were tested under monotonic tension. The focus of the tests was on T-stubs failing predominantly in mode 1. Auxiliary tensile tests were performed on coupons extracted from the T-stub flanges in three directions thereby enabling the material anisotropy to be quantified. The ultimate resistance, plastic deformation capacity and failure mode of each T-stub were obtained and are reported herein. The obtained results are utilised to assess the accuracy of existing design rules, which, contrary to austenitic and duplex stainless steels, seem to offer a good prediction of the experimentally determined plastic resistance. Finally, the applicability of recently proposed predictive equations accounting for membrane actions in carbon steel T-stubs to ferritic stainless steel T-stubs has also been assessed.

The behaviour of stainless steel plate girders subjected to combined bending and shear was experimentally studied in this paper. Both tensile and compressive material properties of the two adopted stainless steel alloys, including austenitic grade EN 1.4301 and duplex grade EN 1.4462, were determined by standard coupon tests. The three-dimensional (3D) optical scanning technology was introduced to acquire an accurate distribution of initial geometric imperfections for each plate girder specimen. A total of six plate girders were fabricated by hot-rolled stainless steel plates, and were tested to failure under combined bending and shear. In-depth analyses of the critical buckling characteristics, the ultimate resistances and the collapse behaviour of the tested specimens were all presented. The obtained ultimate resistances were further employed to assess the existing moment and shear (M-V) interaction design methods in EN 1993-1-5, GB 50017-2017, ANSI/AISC 360-16 and SEI/ASCE 8-02, and the design proposal presented by Jáger et al. It has been found that most of the existing codified M-V interaction formulae can be applicable for both carbon steel and stainless steel plate girders, yet they lead to relatively conservative predictions for stainless steel plate girders, except that the design method in ANSI/AISC 360-16 provides apparently unsafe strength predictions.

Welded tubular joints are widely used for the mast arm pole connection in signal support structures. In order to study the fatigue behavior of such welded tubular joints, static loading and fatigue tests were carried out on several full-scale experiments. In order to understand the local behavior and perform parametric study of critical factors such as initial crack location, finite element (FE) models are subsequently built and the numerical simulation results are validated with the experimental test results. Stress intensity factors (SIFs) at the deepest point of fatigue cracks are calculated by the FE method and an empirical equation are examined by comparison. It is found that the numerical simulation results are in good agreement with the experimental test results. The effect of initial crack location on the SIF and the fatigue life are also investigated. The Bowness equation will result in higher SIF value and more conservative fatigue life prediction for the concerned welded tubular joints. The SIFs at the deepest point decrease and fatigue life increases with the increasing initial crack angle.

The shear buckling behaviour of welded stainless steel plate girders with transverse stiffeners has been experimentally and numerically investigated in this paper. A total of seven plate girders with rigid/non-rigid end posts were fabricated from hot-rolled stainless steel plates, and each of the plate girders was subjected to a concentrated load at mid-span. The shear buckling characteristics and postbuckling behaviour were observed. Prior to testing, the initial local geometric imperfections and the material properties were accurately measured. The critical shear buckling strengths of the web plates were determined from the recorded surface strains and out-of-plane deflections, which were further compared with theoretically predicted values from elastic and inelastic assumptions. By using the general finite element (FE) software package ABAQUS, elaborated FE models were developed and validated against the obtained test results and other available existing test data. Upon validation of the FE models, parametric studies were subsequently carried out to explore the influences of initial local geometric imperfections, web aspect ratios, end post conditions, and material properties over a wide range of web slendernesses. The obtained test/FE results, together with other existing test data, were summarised to evaluate the current codified provisions in GB 50017, EN 1993-1-5, EN 1993-1-4, EN 1993-1-4+A1 and the design proposals from Estrada et al. Based on the test and numerical results, an alternative design approach that could account for the material non-linearity and rigid and non-rigid end posts has been proposed, which provides accurate and reasonable strength predictions for stainless steel plate girder with transverse stiffeners.

Building upon a previous investigation, this study reports a total of 13 experimental tests on austenitic and duplex stainless steel T-stubs subject to monotonic loading. The structural behaviour of the tested T-stub specimens, including load versus displacement (F-Δ) curves and corresponding failure modes, was obtained and is reported herein. The experimental tests were replicated by finite element (FE) analysis, and upon validation of numerical models, a comprehensive parametric study including 168 FE models was conducted to investigate the effects of key parameters such as material grade, bolt preloading, bolt diameter and flange thickness on the structural response. Based on both experimental and numerical results, the suitability of the design provisions for the determination of the plastic resistance specified in EN 1993-1-8, proposed extensions thereof by Demonceau et al. to cover T-stubs with four bolts per row for stainless steel T-stubs as well as the design method codified in Chinese code JGJ 82 were assessed. Novel design methods for the determination of the initial stiffness and the plastic resistance of stainless steel T-stubs, accounting explicitly for the observed structural response and the pronounced material strain hardening were developed. The proposed design methods lead to improved and more consistent capacity predictions and their adoption in design standards is recommended herein.

This paper presents twenty-eight stub columns tests on stainless steel built-up sections. The test specimens, including I-sections, square hollow sections (SHS) and rectangular hollow sections (RHS), were fabricated by shielded metal arc welding (SMAW) from hot-rolled plates of nominal thicknesses 6 mm and 10 mm. The twenty-eight stub columns, of two stainless steel alloys (austenitic EN 1.4301 and duplex EN 1.4462), were tested in pure axial compression. Both tensile and compressive material properties were obtained by means of coupon tests in three directions – longitudinal, diagonal and transverse to the rolling direction. Geometric imperfection measurements for each specimen were conducted by means of a calibrated electric guideway, and residual stress distributions in the built-up sections were determined by means of the sectioning method. The test strengths were used to evaluate the design strengths predicted by EN 1993-1-4, the Continuous Strength Method (CSM) and the direct strength method (DSM). It was demonstrated that the predicted strengths from EN 1993-1-4 provisions were generally conservative, while both the CSM and the DSM predicted values were closer to the test strengths.

Preload in a cast-in-situ concrete-filled steel tube (CFST) structure is an important parameter that has been studied and considered in design and construction. However, the impact of a preload on fire resistance of CFST has not been taken into account in the existing design method of CFST under fire. In this paper, twelve CFST columns with and without preload in the steel tubes are experimentally studied to investigate their heating process, failure modes, thermal and structural responses. The test results show that preload of the steel tube has significant impact on the fire resistance of some of the columns. Further investigations on this important observation will be done numerically in a companion paper [24] to carry out a more comprehensive parametric study.

The local–overall interactive buckling behaviour of welded stainless steel I-section columns was experimentally and numerically examined in this study. A total of ten test specimens were fabricated from hot-rolled stainless steel plates and axially loaded between two pin-ended supports. The specimens failed by local–overall interactive buckling about the minor axis. Prior to the member testing, material properties, residual stresses and initial local and global geometric imperfections were all accurately determined. Detailed finite element (FE) models, capable of simulating the interactive buckling behaviour and predicting the ultimate capacity of welded stainless steel I-section columns, were validated against the obtained test results. The validated FE models were subsequently used to carry out systematic parametric studies, exploring the influences of the key input parameters, including the welding residual stresses, initial geometric imperfections, material properties and slenderness ratios. The generated test and numerical results were then used to assess the accuracy of a series of existing design methods: Eurocode 3 Part 1.4 and the design proposal of Rasmussen and Rondal, both of which employ the effective width concept, and the two separate design proposals of Becque et al. in the EN 1993-1-4 and AS/NZS 4673 formats and the proposal of Huang and Young, all of which are based on the direct strength method (DSM). Based upon the assembled data points, two separate design curves are proposed herein for austenitic and duplex stainless steels, which have been demonstrated to offer very accurate strength predictions for welded stainless steel I-section columns undergoing interactive buckling.

Residual stress magnitudes and distributions in structural stainless steel built-up sections have been comprehensively investigated in this study. A total of 18 test specimens were fabricated from hot-rolled stainless steel plates by means of shielded metal arc welding (SMAW). Two grades of stainless steel were considered, namely the austenitic grade EN 1.4301 and the duplex grade EN 1.4462. Using the sectioning method, the test specimens were divided into strips. The residual stresses were then computed by multiplying the strains relieved during sectioning by the measured Young׳s moduli determined from tensile and compressive coupon tests. Residual stress distributions were obtained for 10 I-sections, four square hollow sections (SHS) and four rectangular hollow sections (RHS). Peak tensile residual stresses reached around 80% and 60% of the material 0.2% proof stress for grades EN 1.4301 and EN 1.4462, respectively. Based upon the test data, simplified predictive models for residual stress distributions in stainless steel built-up I-sections and box sections were developed. Following comparisons with other available residual stress test data, the applicability of the proposed models was also extended to other stainless steel alloys. The proposed residual stress patterns are suitable for inclusion in future analytical models and numerical simulations of stainless steel built-up sections.

This paper aims to develop design methods for the shear strength of welded stainless steel I-shaped members. The current American AISC Design Guide 27 has been published for Structural Stainless Steel, and it refers to the specification ANSI/AISC 360-16 for structural steel buildings for the shear strength without considering tension field action (TFA). All the available test data on stainless steel plate girders were collected from literatures and employed to assess the provisions in the Design Guide 27, as well as the codified ones in Eurocode 3 Part 1.4. Based on the test results and obtained comparisons, two new proposals for unstiffened webs or webs with transverse stiffeners widely spaced (without TFA), and for interior webs with stiffeners spaced at 3hw or smaller (considering TFA) were developed and presented to match the format of the expressions in ANSI/AISC 360-16 for stainless steel webs under shear. Reliability analysis was further performed to calculate the Load and Resistance Factor Design (LRFD) resistance factors determined by the new proposals and to further justify the target factor in AISC Design Guide 27 for structural stainless steel.