Residual stress model for hot-rolled steel I-sections

Residual stresses are self-equilibrating internal stresses that exist within structural members in the absence of external loads. In hot-rolled steel I-sections, these stresses develop primarily during manufacturing due to uneven cooling. Their distribution and amplitude can significantly influence the stability and ultimate strength of steel members; however, existing predictive models often fall short in terms of their accuracy. This limitation is addressed in the current paper by first reviewing existing models and compiling the most extensive residual stress database to date, comprising measurements on over 190 cross-sections from the global literature, and then proposing a new predictive model. Analysis of the collected residual stress measurements reveals several critical insights. The data show that (1) a bi-linear distribution best describes the residual stress pattern within the flanges, while the web stress distribution is most accurately represented by a parabolic distribution, (2) there is net tension in the flanges, balanced by net compression in the web in deeper I-sections, (3) the residual stress amplitudes are not dependent on the material yield strength and (4) in contrast to the existing ECCS model, there is a gradual variation in residual stress amplitude with cross-section height-to-width ( h / b ) ratio. Based on these findings, refined and experimentally validated predictive models in which the key stress amplitudes are a continuous function of h / b are proposed. Overall, the proposed model enhances the consistency, range of applicability and accuracy of the prediction of residual stresses in hot-rolled steel I-sections, which is essential for facilitating the growing use of advanced computational methods in the analysis and design of steel structures.