Genetic algorithm– principal component analysis -driven Multiobjective optimization of wire electric discharge machining for nitinol shape memory alloy

Nitinol shape memory alloys are being used in different applications including aerospace, actuators, biomedical, and sensors. The Nitinol (NiTi) alloys are very difficult to machine, and the conventional machining of these NiTi shape memory alloy is subjected to various challenges. Therefore, unconventional machining process like wire electric discharge machining is used to study the machining behaviour of NiTi alloy. This paper emphasizes the effect of input parameters such as pulse on time, pulse off time and servo voltage on the output responses i.e., material removal rate and surface roughness. Design of experiments (DOE) is employed to structure the experiment. Two different Multiobjective approaches, Genetic Algorithm and Principal Component Analysis were used to identify the best input parameters that gives more material removal rate with reduced surface roughness. The optimal sets of input parameters are generated using the pareto front in genetic alogoritim. The surface topography examination confirms the greater surface roughenss for higher values of pulse on time. The recast layer has also been examined for the samples consisting lowest and highest material removal rate.