Abstract:

In order to the desired surface quality in the parts manufactured in the reverses, the choice of the appropriate cutting parameters (cutting speed, progress speed and cutting depth) is very important. The purpose of this study is to experimentally reveal how the surface smoothness is affected depending on the size of the work piece (diameter and output length) in addition to the cutting parameters, and to present a model that predicts the surface smoothness according to the given parameters. Progress speed, cutting speed, cutting depth and different output lengths are taken as variable parameters and two different-dimensional work parts are turning in the CNC turning machine. The experiments are designed according to the full factor experiment design method, and for 4 different parameters 3 levels, two different dimensions (for each diameter) 34 = 81 experiments were performed. Using the results obtained from the experiments, models that predict surface smoothness with the Adaptive Neural Network-Based Bully Output System (ANFIS) have been developed and the best model for each diameter has been selected. Created models will be a guide for the correct and appropriate selection of processing parameters in the industry to obtain the desired surface quality in the manufacturing processes.

Keywords: