Modeling Cutting Forces in High-Speed Turning using Artificial Neural Networks

Abstract

Cutting forces are very important variables in machining performance because they affect surface roughness, cutting tool life, and energy consumption. Reducing electrical energy consumption in manufacturing processes not only provides economic benefits to manufacturers but also improves their environmental performance. Many factors, such as cutting tool material, cutting speed, and machining time, have an impact on cutting forces and energy consumption. Recently, many studies have investigated the energy consumption of machine tools; however, only a few have examined high-speed turning of plain carbon steel. This paper seeks to analyze the effects of cutting tool materials and cutting speed on cutting forces and Specific Energy Consumption (SEC) during dry high-speed turning of AISI 1045 steel. For this purpose, cutting forces were experimentally measured and compared with estimates of predictive models developed using polynomial regression and artificial neural networks. The resulting models were evaluated based on two performance metrics: coefficient of determination and root mean square error. According to the results, the polynomial models did not reach 70 % in the representation of the variability of the data. The cutting speed and machining time associated with the highest and lowest SEC of CT5015-P10 and GC4225-P25 inserts were calculated. The lowest SEC values of these cutting tools were obtained at a medium cutting speed. Also, the SEC of the GC4225 insert was found to be higher than that of the CT5015 tool.