Numerical simulation and optimization in pulsed Nd: YAG laser welding of Hastelloy C-276 through Taguchi method and artificial neural network

Abstract

The effect of process parameters viz., welding speed, pulse energy and frequency on the aspect ratio and the peak temperature in the numerical simulation of Nd: YAG laser welding of Hastelloy C-276 is presented. The experimental conditions are designed based on Taguchi L9 orthogonal array and the simulation is performed in SYSWELD 16, a 3 D software package. The optimized parameters for attaining a maximum aspect ratio and a minimum peak temperature (responses) were determined by computing the Signal-to-Noise Ratio (SNR). Mathematical model based on regression analysis was developed for predicting the responses using statistical software Minitab-16. Further, the simulation results are correlated with the regression and artificial neural network models. The deviation is estimated to be less than 10% indicating their effectiveness in predicting the responses. The experimental macrograph for the optimized condition is concurrent with the numerical simulation.