Analytical modeling and multi-objective optimization algorithm for abrasive waterjet milling Ti6Al4V

Abrasive waterjet (AWJ) is a promising method for machining titanium alloy, which is widely used in the aerospace field, but the various process parameters of AWJ make it difficult to achieve a high machining quality. In this research, the main process parameters of AWJ, including the jet pressure, the abrasive flow rate, the stand-off distance, the jet angle, the traverse speed, and the feed rate, were all analyzed by considering their effects on the milling characteristics of Ti6Al4V alloy. Both single and interactive effects of the process parameters were studied, and regression models for predicting the milling depth h, the material erosion rate V˙\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\dot{V}$$\end{document}, and the X-directional roughness Rax were established. Furthermore, an ADM-MO-Jaya (adaptive decreasing method multi-objective Jaya) algorithm based on MO-Jaya was proposed to obtain the optimal process parameters, aiming for reaching the minimum Rax and the maximum h and V˙\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\dot{V}$$\end{document} at the same time. The results show that the correlation coefficients R2 of the models are all greater than 0.9, and model terms are relatively significant. The regression models of h, V˙\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\dot{V}$$\end{document}, and Rax are generally consistent with the overall trend of the experimental results, and the mean errors are 8.57%, 1.89%, and 10.58%, respectively. The operation efficiency of the ADM-MO-Jaya algorithm is 32% higher than that of the MO-Jaya, and the Pareto front is the most uniform and converges to a curve in the solution space without isolated points. The optimized set of 180 Pareto solutions can be directly selected by the operator for machining without complex process comparisons, which can guide the practical milling of titanium alloy by AWJ.