In this research, AISI 1040 steel whose hardness is 46 HRc was processed in CNC lathe. Taguchi L16
experiment design was created based on cutting speed, feed rate and depth of cut of which is consisted of four levels. As a result of these experiments, average surface roughness (Ra) values were measured. Multiple regression models for measured Ra values were created by using MINITAB 14 program. The closest results to the experiment results in regression models created for Ra were obtained with the quadratic regression model with the 99.8% coefficient of determination. With the regression models created, it was determined that the
most effective parameters are the feed rate parameters. From mathematical equations created in the result of the experiments carried out, it was determined that the quadratic regression equation is approximately 90% correct.