Modeling and Optimization of Surface Residual Stress Profiles in Milling of Aluminum 7075-T6 Alloy

Abstract

Abstract Aluminum 7075-T6 alloy has been widely employed in aviation, transport, and automobile applications due to its remarkable properties, while a lot of residual stresses can be generated in the machined surface and subsurface during the machining process. The machining parameters have significant effects on the formation of residual stress, it’s important to predict the residual stress distribution with the cutting parameters and optimize the machining parameters to acquire the desirable residual profiles. Although many efforts of current studies have been paid to the prediction of residual stress profiles in different materials and machining processes, however, few works focused on residual stress in-depth profiles in the machining of 7075-T6 aluminum alloy, and the optimization of cutting parameters for required residual stress profile has also rarely been reported as well. Therefore, this study proposed an integrated prediction model, which combines exponential decay cosine function (EDC), particle swarm optimization (PSO), and back propagation neural network (BP), to predict the in-depth residual stress profile of the machined surface in milling of 7075-T6 aluminum alloy. Furthermore, according to the predicted residual stress profile, the key features for describing the residual stress profile include the surface residual stress (SRS), maximum compressive residual stress (MCRS), depth of maximum compressive residual stress (DMCS), and depth of residual stress (DRS), were identified and analyzed. And a multiple objectives optimization was conducted based on the predicted residual stress profile features, where Kriging-based models were employed to establish the relationships between machining parameters and each objective (SRS, MCRS, and MRR i.e. material removal rate). Finally, a two-stage optimization strategy integrating NSGA-III, MOPSO, and TOPSIS algorithms, was used to address the multi-objective optimization model to obtain the expected residual stress profile and MRR. This work can provide some practical guidance for industrial production in machining 7075-T6 aluminum alloy.