Achieving high precision and productivity in laser machining of Ti6Al4V alloy: A comprehensive study using a n-predictor polynomial regression model and PSO algorithm

Abstract

Ti-6Al-4V, the Titanium alloy, has significant utilizations in aerospace, automotive, and marine sectors for its low density and high strength at elevated temperature. But its chemical activity and low thermal conductivity inhibits its machining by conventional method. Nd: YAG laser beam machining (LBM) finds extensive use in rapid and precise cutting of Ti6Al4V. This study has examined the influences of various LBM machining variables, including laser power, gas pressure and stand-off distance, in cutting 5 mm thick Ti-6Al-4V plate. In assessing the effectiveness and performance of the LBM process, three response functions—surface roughness, angle of kerf, and material removal rate—have been designated. From the experimental data, different regression models have been established to estimate these response functions in terms of the machining parameters. Based on R² score and RMSE, multi-dimensional polynomial regression is decided as the most suitable regression model. Subsequently, the Particle Swarm Optimization technique has been applied to identify the optimal machining parameters for reducing angle of kerf and surface roughness, while increasing material removal rate. Three individual single-objective functions underwent optimization, along with a multi-objective function. Furthermore, experimental verification was conducted for the optimal input parameters in the single-objective as well as the multi-objective optimization scenarios, resulting in an accuracy of 97% and 98%, respectively. Such a close agreement emphasizes the accuracy of the developed regression model as well as it signifies the reliability and efficacy of the optimization technique.