Soft computing models based prediction of cutting speed and surface roughness in wire electro-discharge machining of tungsten carbide cobalt composite

Abstract

In the present study, a second order multi-variable regression model and a feed-forward back-propagation neural network (BPNN) model have been developed to correlate the input process parameters, such as pulse on-time, pulse off-time, peak current, and capacitance with the performance measures namely, cutting speed and surface roughness while wire electro-discharge machining (WEDM) of tungsten carbide-cobalt (WC-Co) composite material. From a large number of neural network architectures, 4-11-2 has been found to be the optimal one, which can predict cutting speed and surface roughness with 3.29% overall mean prediction error. The multivariable regression model yields an overall mean prediction error of 6.02%. Both the models have been used to study the effect of input parameters on the cutting speed and surface roughness, and finally to corroborate them with those of the experimental results. Scanning electron micrographs reveal that at higher energy level the machined surface is characterized by several microcracks and loosely bound solidified WC grains.