Publication

Abstract : This work investigates surface roughness as a key indicator of machining performance, reflecting variations in cutting force, chip thickness, and chip compression ratio during the turning of Ti6Al4V. Experiments were performed by varying depth of cut , spindle speed , and feed rate , with surface roughness, cutting force, chip thickness, and chip compression ratio as responses. A Response Surface Methodology approach was used for experiment design, and Analysis of Variance identified significant factors influencing each response. Regression models developed from RSM accurately predicted experimental outcomes. Large dataset generated from these models was used to train an Adaptive Neuro-Fuzzy Inference System in MATLAB Simulink, forming an offline digital twin for process prediction. The ANFIS model was validated using 9 additional experiments, identifying optimal machining parameters—656.52 rpm SS, 0.3 mm DoC and 30 mm/min FR—demonstrating a reliable framework for predictive surface quality control and productivity improvement.