MODELLING AND OPTIMIZATION OF GRINDING AND WHEEL PARAMETERS IN CYLINDRICAL LONGITUDINAL MACHINING OF TOOL STEEL

Abstract

In this study, the effects of ten input factors (wheel speed, workpiece speed, feed, depth of cut, wheel diameter, wheel width, wheel abrasive material, wheel grain size, wheel hardness and wheel structure) on three response variables (dimensional deviation, surface roughness, and material removal rate) were investigated during cylindrical longitudinal grinding of tool steel. The experimental design was constructed using a custom design of experiments based on the I-optimality criterion, enabling efficient and statistically reliable evaluation of the effects of a large number of input factors. Based on the experimental results, regression analysis was performed and process regression models were developed. All input factors, as well as certain interaction effects, had a significant influence on dimensional deviation and surface roughness. Using the developed models, a multi-objective optimization was carried out to simultaneously improve dimensional accuracy, surface quality, and process productivity. Model validation and verification of the optimal solutions were conducted through additional confirmation experiments. The low confirmation errors obtained indicate the adequacy of the modelling and optimization procedures.