Analysis and design of concrete-filled steel tubular members considering size effect
Concrete-filled steel tubular (CFST) members have been widely used in buildings and bridges to better utilize the benefits of both concrete and steel materials. In CFST members, steel provides confinement and ductility to concrete, while concrete delays the local buckling and provides stiffness to the steel tube. Extensive experimental and theoretical studies have been conducted on the behaviour of CFST columns and beams, but little attention has been paid to the influence of size effect. Recently, size effect has been clearly observed in CFST stub columns with different diameter-to-thickness ratios and confinement factors. Many design codes such as EC4, GB, AIJ, AS, and AISC have been developed based on test results of small-size specimens, and specified limits on material strength and section slenderness. However, the influence of size effect has not been incorporated in the existing codes.
A number of finite element (FE) models are available in the literature to simulate the behaviour of CFST members. But size effect has been seldom considered in previous FE analysis, thus existing models may not be suitable for simulating real CFST members with a size above 300 mm. Meanwhile, hundreds of tests have been conducted on CFST stub columns, and those test results are widely used to calibrate key parameters in the material models for FE analysis. In reality, tests on CFST stub columns can be conducted via two methods, i.e., load-controlled method and displacement-controlled method. Although the selection of a test method might have negligible influence on the ascending branch of the load-deformation curve, it indeed could significantly affect the descending branch of the load-deformation curve. It is envisaged that a more reliable descending branch can be obtained using the displacement-based method compared to the load-controlled method. However, in developing previous FE models, no differentiation has been made between test data generated from the two different test procedures. This could lead to the development of inaccurate FE models.
The main aim of this research is to develop an accurate and versatile finite element model for CFST members to consider the size effect. The quality of test data of CFST stub columns will be considered in calibrating key material parameters of the FE model. The developed material models will be further implemented in the FE analysis of CFST beams, slender columns and beam-columns. Parametric analysis will be conducted using the refined FE model to study the influence of different important factors. Finally, the prediction accuracy of existing design codes will be evaluated. Based on the evaluation, more accurate design equations will be proposed for CFST stub columns, beams, slender columns and beam-columns, where the influence of size effect will be considered.
Research Project Supervisor/s:
Professor Zhong Tao (Principal), Doctor Md Kamrul Hassan and Doctor Utsab Katwal
Bulbul Ahmed is currently conducting his 2nd year as a PhD candidate at the School of Engineering, Western Sydney University, with a research focus in Structural Engineering.