Process parameter modeling and multi-response optimization of wire electrical discharge machining NiTi shape memory alloy

Abstract

In the present work, the effects of different wire electrical discharge machining (WEDM) process parameters on nickel titanium shape memory alloy (NiTi-SMA) properties were investigated. Based on the orthogonal experimental design of the Taguchi method, the influence of five process parameters including peak current, discharge frequency, wire tension, flushing pressure, wire speed on the cutting speed (CS) and kerf width (KW) was analyzed. After 27 experimental runs, a multiple regression (MLR) model and a back-propagation neural network (BPNN) model were established to predict the CS and KW properties under different input conditions of process parameters. Utilize the programming language R for statistical testing of models and determination of important control factors. The Bat Algorithm (BA) searches for the best combination of process parameters. The experimental results and the proposed optimization method show that the prediction errors of the two combined optimization methods of MLR-BA and BPNN-BA are controlled by ± 2%. It has been proved that these two combined optimization methods are practical tools for optimizing the processing parameters of WEDM. Statistical analysis and observation results show that the peak current is the main parameter affecting the machining efficiency and surface quality of WEDM.