ßÉßɱ¬ÁÏ

Topic

Design of direct- and indirect-formed rectangular hollow section beam-columns

Department of Civil Engineering

Date & location

• Friday, June 27, 2025

• 3:00 P.M.

• Virtual Defence

Reviewers

Supervisory Committee

• Dr. Min Sun, Department of Civil Engineering, ßÉßɱ¬ÁÏ (Supervisor)

• Dr. John Dower, School of Earth and Ocean Sciences, UVic (Member) 

External Examiner

• Dr. Kian Karimi, Department of Civil Engineering, British Columbia Institute of Technology 

Chair of Oral Examination

• Dr. Bernie Pauly, School of Nursing, UVic

   

Abstract

Hollow structural sections (HSS) are tubular steel profiles commonly produced by cold-forming for various engineering applications. Square and rectangular hollow sections—referred to as SHS and RHS, and collectively termed RHS hereinafter—are extensively used as structural members, including braces, beams, columns, and beam-columns. In North America, cold-formed RHS are primarily manufactured using two techniques: (1) direct-forming and (2) indirect-forming. In direct-forming, cold work is concentrated at the corners of RHS, while in indirect-forming, it is applied across the entire perimeter; thus, each technique results in different levels of residual stresses and distinct stress distributions across the cross-section, potentially influencing the mechanical behaviour of the members. Several studies by other researchers have indicated that, on average, direct-forming induces smaller and more localized residual stresses on the RHS cross-sections than indirect-forming. As a result, direct-formed RHS beams and columns have demonstrated relatively higher flexural and axial compressive capacities compared to their indirect-formed counterparts. However, the existing provisions of the Canadian and American steel design standards (CSA S16:19 and AISC 360-22) for RHS members do not account for mechanical differences resulting from different cold forming techniques (i.e., direct-forming and indirect-forming). Despite these findings, a comparative study on direct- and indirect-formed RHS beam-columns (i.e., members subjected to combined axial compression and flexure) is still lacking. The purpose of this research was to compare the mechanical behaviour and load-bearing capacity of cold-formed, untreated RHS beam-columns, made from different steel grades and fabricated using these two forming techniques. A comprehensive finite element (FE)-based parametric investigation was conducted on a total of 1040 RHS beam-columns in each of Chapters 1 and 2, with Chapter 1 focusing on stub members and Chapter 2 on long members, to evaluate their ultimate axial compressive and flexural load bearing capacities. In both chapters, several tensile coupon test data from previous experimental studies were collected and used to create a variety of steel material groups, covering a range of nominal yield strengths from 350 MPa to 700 MPa. To validate the FE modeling, 53 stub and 36 long beam-column models were generated to replicate actual specimens, boundary conditions, and loading configurations. The developed material groups were subsequently implemented in the FE models to define the corresponding material properties, without explicitly modeling residual stresses. Geometric and material nonlinearities, as well as initial geometric imperfections, were incorporated to enhance the accuracy of the analyses. The models were subjected to static axial compressive loading, applied either concentrically or with varying eccentricities, resulting in combined axial compression and flexure. Their load-deformation responses and ultimate load bearing capacities were then compared to experimental results, as presented in Chapters 1 and 2. In each chapter, a series of reliability analyses were conducted to assess the safety and adequacy of CSA S16:19 and AISC 360-22 in the design of RHS beam-columns. Based on the findings in Chapter 1 on stub members, the CSA S16:19 provisions were, in general, reasonably reliable, whereas those in AISC 360-22 were excessively conservative. As for the findings in Chapter 2 regarding long members, CSA S16:19 was found to be overly conservative for members with more slender sidewalls (Classes 3 and 4), yet less reliable for those with stockier sidewalls (Classes 1 and 2). Additionally, AISC 360-22 was observed to be slightly conservative for direct-formed members and comparatively less reliable for indirect-formed ones. According to both chapters, direct-formed members were mechanically advantageous over their indirect-formed counterparts, exhibiting higher ultimate load-bearing capacities. A set of modifications was proposed for the cross-section slenderness limits of direct-formed members in AISC 360-22, as well as for the coefficients of the axial compression-flexure interaction formulas for both direct-formed and indirect-formed members in both standards, to ensure appropriate reliability levels.