Residual stresses of square hollow sections with comparable width-to-wall thickness ratios and effects of cold forming on high-strength steel columns

Loading...
Thumbnail Image

URL

Journal Title

Journal ISSN

Volume Title

Insinööritieteiden korkeakoulu | Master's thesis

Department

Mcode

Language

en

Pages

127 + 17

Series

Abstract

This study examines the influence of profile wall thickness and steel grade on the distribution of residual stresses in cold-formed square tubes. Residual stresses were derived from the strains measured in the profiles with comparable width-to-wall thickness ratios (150x10 mm and 40x2.5 mm) made of steel grades S355J2H/S420MH and S700MLH. Measurements employed a sectioning technique, considering both longitudinal and transverse strains. The method of direct measurement of the longitudinal residual strains at a distance of 1.4 times the side length from the end was proven to be reliable for the larger profile. For the transverse stress definition, a single splitting of the ring was sufficient for strain measurement of the larger profile, but with the smaller one, the ring was necessary to cut into pieces. A comparison between the stresses derived indicated that the thickness of the steel plate significantly affects the distribution of residual stress. The thick walled profile produced greater residual stresses and more complex stress distributions. The high strength of steel slightly reduced residual stresses in the SHS150x10 and produced generally more visible effects on the SHS40x2.5. Based on the measurements, the patterns of longitudinal bending residual stresses were developed. Studies on the accuracy of the finite element analysis to describe the effect of cold forming on stub column behavior emphasized the necessity to incorporate strength enhancements and initial geometric imperfections into numerical models. Using pure nominal material properties reduced column strength by 15-20%, and initial geometric imperfections advanced local buckling, reducing the effective strength of the column by up to 3%. The effect of longitudinal residual stresses on the accuracy of the stub column was negligible. Additionally, the accuracy of intermediate column analysis revealed that the stiffness of response is sensitive to the method used for modelling residual stresses across the wall thickness, and the ultimate load is sensitive to modelling of strength enhancements and initial global geometric imperfections.

Description

Supervisor

Puttonen, Jari

Thesis advisor

Lu, Wei
Saremi, Pooya

Other note

Citation