DEVELOPMENT OF PREDICTIVE CUTTING FORCE MODELS IN TURNING USING DIFFERENT EXPERIMENTAL DESIGNS

Abstract

Cutting forces are a very important indicator of the machining process and play a significant role in defining the machinability of the material, tool wear, vibrations, temperature in the cutting zone, process accuracy, modeling, monitoring and management of the machining process. The modeling of cutting forces is relatively difficult due to a number of influencing factors and the lack of knowledge of the interaction between them. The paper discusses and compares the application of different experimental designs (full factorial design, Box–Behnken and Taguchi’s design) for the development of the main force prediction models of varying mathematical forms. Dry longitudinal single-pass turning of not heat-treated non-alloyed and low-alloyed steels with carbon content higher than 0.55% was considered. Main cutting force estimates were acquired by varying feed rate, depth of cut and rake angle and by the application of the Walter machining calculator. Initially, linear, quasi-linear, power and quadratic models were developed and compared. In addition, based on dimensional analysis (DA), an approach that reduces the number of needed experimental trials, a cutting force prediction model was proposed.