Prediction and Optimization of Surface Roughness and Kerf Width in AWJM Using Soft Computing Tools

Recent years have observed tremendous growth in the evolution of tougher and more complex materials that are harder to cut. The manufacturing industry is in a constant dilemma of making trade-offs between time, cost, and quality. Abrasive Water Jet Machining (AWJM) has been ever-increasing in popularity as a non-conventional machining process and has diverse applications in machining and cutting of materials that are traditionally difficult to machine. The cutting operation in AWJM is performed using a thin jet of high-pressure water that consists of tiny abrasive particles, and the cutting takes place through erosion. The process parameters considered in this study are traverse rate, standoff distance, and abrasive flow rate. A full-factorial DOE was conducted to help build a predictive model for the surface roughness and kerf width for the Titanium grade 5 alloy (Ti–6Al–4V) using a soft computing tool known as adaptive neuro-fuzzy inference system (ANFIS). Furthermore, the optimal set of process parameters was found using a multi-objective genetic algorithm.