Investigations into residual stresses in S690 cold-formed circular hollow sections due to transverse bending and longitudinal welding

This paper presents an experimental and numerical investigation into residual stresses of S690 cold-formed circular hollow sections (CFCHS) due to transverse bending and longitudinal welding. It is generally expected that adverse effects of residual stresses on both cross-section and member resistances in the S690 CFCHS are proportionally less pronounced, when compared with those in S355 CFCHS owing to increased yield strengths of the steels. Hence, there is a need to determine the distribution of residual stresses in the S690 CFCHS through a rational experimental and numerical investigation in order to provide accurate data for subsequent structural assessment on these sections. A total of four S690 CFCHS are fabricated with 6 mm thick plates with (i) transverse bending, and (ii) longitudinal welding. Surface temperature history at selected positions of these sections are measured with thermocouples during welding while surface residual stresses are measured with the sectioning method after welding. Owing to various practical constraints in measuring residual stresses of these sections accurately, a total of three coordinated finite element models are established in which their numerical results are integrated for rational analyses. The transverse bending process is simulated with two-dimensional models with plane-strain elements which undergo extensive plastic deformations to generate residual stresses after springback. The longitudinal welding process is simulated with two coupled three-dimensional models with solid elements to perform a sequentially-coupled thermomechanical analysis in the presence of those residual stresses due to transverse bending. Consequently, a rational distribution of the residual stresses due to both transverse bending and longitudinal welding in these sections are readily determined with these coordinated finite element models after careful calibration against measured data.