Lateral cyclic testing and backbone curve development of high-strength steel built-up box columns under axial compression

Abstract

The AISC 341 (2016) has a more stringent width-to-thickness (b/t) limit for highly ductile hollow box columns (HBCs) than the AIJ (2010) or Taiwan Code (2010), resulting in significant thickness difference in design. Moreover, the cyclic backbone curves based on ASCE 41 (2013) and NIST (2017) underestimate the post-buckling flexural strength of HBCs, particularly in high axial compression force. This paper presents test results of six full-scale, built-up HBCs using SM 570 M steel with the actual yield strength of 460–530 MPa. The lateral cyclic behavior of built-up columns was studied in terms of different b/t ratios, magnitudes of axial compression forces, and lateral drift histories (i.e., cyclically symmetric versus near-fault displacement histories). The built-up box columns were 290–400 mm wide with b/t ratios from 11 to 21, 4000 mm high, and tested laterally with both ends fixed after under a constant axial compression force, 2591–7935 kN. The HBC specimens which were designed based on the highly ductile member requirement in AISC 341 (2016), even under an axial compression force (=40%Py), performed satisfactorily at 4% drift and experienced flange and web fracture at 5% drift. However, the HBC specimens that satisfied the most compact b/t requirement in AIJ (2010) or Taiwan code (2010) did not perform well at 4% drift, losing the axial capacity after significant column local buckling and shortening. The gathered test data, supported by more test data in this work, were analyzed by a multiple regression method to obtain empirical formulations for predicting the maximum column moment, plastic rotation and post-yield hardening parameters. The proposed formulation reasonably predicts the first-cycle envelope curves of built-up HBCs, improving prediction results based on ASCE 41 (2013) and NIST (2017).