Prediction of residual stress fields

What is it about?

Surface residual stresses are critical parameters for evaluating the surface quality and can influence many mechanical properties of solids. These stresses inevitably arise in almost all engineering components during manufacturing. However, most experimental and finite element approaches cannot obtain complete surface residual stress field in a mechanical part. In this study, we proposed a predictive method to determine surface stress fields, depending on residual stresses being self-equilibrating. The effectiveness of the approach is verified using a numerical surface of a beam example with ideal measurements and a casting-milling surface with experimental data. Using the proposed method, surface residual stress fields can be obtained from using stresses of a limited number of points including boundary points to solve governing equations via a Fourier series bivariate polynomial as an Airy stress function with the Tikhonov regularization method. Our method does not require simulations of the residual stress generation process. This method is suitable for complex engineering parts where the manufacturing process is difficult to recreate in detail. The predicted stress field can be imported into a finite element solver as initial stresses to promote the design, manufacturing, and assessment of mechanical components.

Featured Image

Why is it important?

In this work, we have presented an approach to predict the complete surface residual stress distribution from stresses of partial points. Mathematical model, identification of unknown parameters, and computational procedure are detail presented. The effectiveness of the approach is verified using two examples: one numerical example with simple ideal measurements and another experimental example with complex experimental data. The results show that surface residual stress fields can be obtained from stresses of partial points to solve the governing equations via a Fourier series bivariate polynomial as an Airy stress function. The solution to the mathematical model is simple and the entire process avoids the randomness associated with the selection of test points. The residual stress can therefore be acquired at any point. A key feature of the proposed method is that our method does not require simulations of the residual stress generation process. Thus, it is suitable for complex engineering parts where the manufacturing process is difficult to recreate in detail. What is more, the predicted stress field can be imported into a finite element solver as initial stresses to promote the design, manufacturing, and assessment, mechanical analysis of mechanical components.