Surface roughness prediction modelling for commercial dies using ANFIS, ANN and RSM

Abstract

Surface roughness of dies is considered as a vital quality characteristic. In this study, average surface roughness for a die material H13 has been measured after ball end milling operation. A design of experiment was prepared with response surface methodology (RSM). Forty-nine experiments have been conducted varying six different cutting parameters. This 49 data have been used for training purpose and further 25 testing data have been collected with random selection of input parameters. Better ANFIS model has been selected for minimum value of mean square error, which is constructed with two Gaussian membership functions (gauss2MF) for each input variables and a linear membership function for output. The selected ANFIS model has been compared with theoretical model, ANN and RSM. Comparison shows that the selected ANFIS model gives better result. Correlation test shows that only cutter axis inclination angle and radial depth of cut have positive correlation with surface roughness.