Local and post-local buckling of fabricated high-strength steel and composite columns

https://doi.org/10.1016/j.jcsr.2018.12.004Get rights and content

Highlights

  • Experimental results on the local buckling slenderness limits for high-strength steel plates were presented.

  • A set of sixteen tests were conducted on both hollow steel and steel-concrete composite sections.

  • A numerical model was developed to predict the local buckling and post-local buckling response of box and I-section columns.

  • Yield slenderness limits were obtained from the experimental and numerical investigation.

  • Slenderness limits of hollow and composite high-strength steel sections were compared with existing codes of practice.

Abstract

High-strength structural steel plates are being increasingly used as composite columns in tall buildings, bridges and large infrastructure. The presence of concrete infill in these composite sections enhances their local buckling strength, and thus very slender steel plates can be used in their fabrication. This paper presents the results of an experimental study and numerical investigation of the local buckling slenderness limits for high-strength steel plates. A set of sixteen tests were conducted on both hollow steel and steel-concrete composite sections to explore their local and post-local buckling behaviour under axial compression. A numerical model which accounts for the effects of residual stresses and initial geometric imperfections was developed to predict the local buckling and post-local buckling response of box and I-section columns. This model was verified against the test results. Yield slenderness limits obtained from numerical results were compared with existing codes of practice for both hollow steel and composite sections incorporating high-strength steel plates.

Introduction

By filling concrete into steel sections, steel-concrete composite columns develop greater axial loading resistance, favourable ductility performance, larger energy absorption capacity and better fire resistance compared with conventional structural steel columns and reinforced concrete columns [[1], [2], [3]]. Therefore, composite columns become a popular choice in the design of heavily loaded compressive members. The use of high-strength steel in composite columns can achieve greater benefits as the material provides greater strength to cross-section ratio [4,5]. However, as the strength of the steel increases, the buckling characteristic becomes a dominant issue that reduces the loading capacity [6,7]. The slenderness limits for both local and global buckling also become significantly important as they define the compactness/slenderness of the sections of members. In this paper, the local and post-local buckling behaviour of hollow and steel-concrete composite sections was studied with the presence of high-strength steel plates.

Uy [8] conducted an experimental and analytical study on concrete-filled steel tubular (CFST) composite columns employing mild structural steel. Local and post-local buckling behaviour of box sections was investigated. A comparison between the experimental results and Australian Standard AS 4100 [9] proved that the existing approach was conservative in ultimate capacity estimations. Bridge and O'Shea [10] studied hollow and composite columns with sectional slenderness ranging from 34 to 130, where geometric imperfections and residual stresses were considered. The same conclusion was obtained that AS 4100 [9] was capable of being used in the design of hollow and composite columns fabricated by mild structural steel. Uy [[11], [12], [13]] conducted an extensive set of tests to study the local buckling behaviour of hollow and composite sections with mild and high-strength steels. A theoretical model was proposed to determine the effective widths and the strength of hollow and composite sections. Extensive strength tests of stub concrete-filled columns were conducted by Sakino et al. [14]. Box and circular sections with a wide range of structural steel types and sectional slenderness were included in the tests. According to the test results, axial load capacity reduction factors for slender hollow and composite sections were proposed. More recently, Lee et al. [15] and Ma et al. [16] reported two sets of extensive experimental programs, where the high-strength steel box sections were tested under different loading conditions. The slenderness limit for these high-strength steel plates was investigated accordingly.

In addition to the studies into slenderness limits for high-strength steel sections, extensive research has been performed over the past few decades to study the strength of composite columns with high-strength steel plates. The reduction in terms of loading capacity of these high-performance composite columns was investigated. Liew et al. [17] further studied the axial and flexural capacity of CFST columns utilising high-strength materials. The test results confirmed the applicability of Eurocode 4 [18] in the strength estimation for such types of CFST columns. Khan et al. [19,20] investigated stub and slender concrete-filled columns fabricated by high-strength steel. Sectional and global slenderness limits of concrete-filled columns were proposed. Tao et al. [21] conducted a comprehensive numerical study with finite element analysis on composite stub columns. A theoretical stress-strain model of confined concrete was proposed to simulate the concrete behaviour in different stages. Extensive parametric studies were carried out to explore the sensitivity of each parameter specified in ABAQUS. An accurate and comprehensive numerical model was developed and verified against more than 200 test results of box and circular sections. Thai et al. [22] developed a reliable non-linear finite element model for concrete-filled steel box columns with consideration of geometric imperfections and residual stresses. The parametric studies showed that the Eurocode 4 [18] and AS/NZS 5100.6 [23] could be used for the design of high-strength concrete-filled columns. Aslani et al. [24] carried out an experimental and numerical investigation of the axial load capacity of stub CFST columns. Based on the experimental and numerical results, AS/NZS 5100.6 [23] gave a most accurate prediction of ultimate strength.

According to the literature mentioned above, it can be found that the research on sectional slenderness limits of high-strength steel is still insufficient. To optimise the benefits of high-strength materials in composite columns, it is critical to identify the corresponding sectional slenderness limits and prevent the elastic local buckling of steel sections. In this paper, an experimental study and numerical investigation of the local buckling behaviour of both hollow and composite stub columns was carried out. A set of sixteen tests under axial compression were conducted to explore their local buckling behaviour. A numerical model accounting for geometric imperfections and residual stresses was developed to predict the local and post-local buckling behaviour. The accuracy of the proposed model was verified against the obtained test results. Sectional slenderness limits obtained from numerical results for hollow and composite sections were compared with existing codes of practice.

Section snippets

Specimen preparation

The aim of this experimental study was to determine the slenderness limits for both high-strength steel box and I-section columns with and without concrete infill. The test specimens were fabricated from nominally 5 mm thickness BISALLOY80 high-strength steel plates with a nominal yield stress of 690 MPa. Steel plates of 5 mm thickness were initially tack-welded into specific sections followed by longitudinal fillet welding along the full length of the columns. Bracing inside the columns was

Experimental results and discussion

This section presents the measured results of the test specimens, in particular, the results that are of major interest are the load-axial shortening behaviour, load-axial strain relationships, buckling behaviour and failure modes. For comparison purposes, this section is further divided into two parts that present the results for the box and I-section columns, respectively.

Modelling description

An accurate finite element model was developed using ABAQUS/Explicit to simulate the local buckling and post-local buckling response as well as the failure modes of the tested specimens. Unlike the implicit method, the explicit method can easily overcome numerical convergence issues encountered in the implicit method due to large deformation and contact problems as in the case of composite columns. The accuracy of the explicit solution is dependent on the loading rate used. In the explicit

Numerical results and discussion

The sectional slenderness limits and effective widths were studied using the verified numerical model by simulating steel and composite columns with compact and slender sections under various width to thickness ratios. In this study, 5 mm thick structural steel plates of Grade 690 were examined. The nondimensional sectional slenderness λe as defined in Eq. (1) was taken from 15 to 120 for box sections and 15–60 for I-sectionsλe=btfy250where b and t respectively are the width and thickness of

Conclusions

This paper presented the experimental results of 16 hollow and composite columns fabricated with high-strength steel plates. The local buckling behaviour and ultimate capacity under axial compression were investigated. An explicit numerical model was developed and verified against the experimental results. Initial imperfections and residual stresses were also considered in the present model. A series of parametric studies on steel and composite columns with various width to thickness ratios (b/t

Acknowledgements

The research described in this paper is financially supported by the Australian Research Council (ARC) under its Discovery Scheme (Project No: DP170100001) and La Trobe University. The financial support is gratefully acknowledged. The assistance from Sina Kazemzadeh Azad in preparing and conducting experimental tests is also gratefully acknowledged. Assistance from Dr. Mohanad Mursi and staff of CASE at the J.W. Roderick Laboratory of the University of Sydney are also gratefully acknowledged.

References (39)

Cited by (60)

View all citing articles on Scopus
View full text