Surface roughness fuzzy inference system within the control simulation of end milling

Abstract

This paper presents a surface roughness control of end milling with associated simulation block diagram. The objective of the proposed surface roughness control is to assure the desired surface roughness by adjusting the cutting parameters and maintaining the cutting force constant. For simulation purposes an experimentally validated surface roughness control simulator is employed. Its structure combines genetic programming (GP), neural network (NN) and adaptive neuro fuzzy inference system (ANFIS) based models. Surface roughness control simulator simulates the surface roughness of the part by enabling the regulation of cutting force. The focus of this research is to develop a reliable method to predict surface roughness average during end milling process. An ANFIS is applied to predict the effect of cutting parameters (spindle speed, feed rate and axial/radial depth of cut) and cutting force signals on surface roughness. Machining experiments conducted using the proposed method indicate that using an appropriate cutting force signals, the surface roughness can be predicted within 3% of the actual surface roughness for various end-milling conditions. Simulation results are presented to confirm the efficiency of a control model.