Cutting Force and Surface Roughness Optimizations in End Milling of GFRP Composites Utilizing BPNN-Firefly Method

Abstract

The excessive cutting force that is generated in the end milling process of glass fiber-reinforced polymer (GFRP) composites can lower the surface quality. Hence, it is necessary to select the correct levels of end milling parameters to minimize the cutting force (CF) and surface roughness (SR). The parameters of the end milling process comprised the depth of cut (doc), spindle speed (n), and feeding speed (V_f). This study emphasized on the modeling and minimization of both CF and SR in the end milling of GFRP combo fabric by combining backpropagation neural network (BPNN) method and firefly algorithm (FA). The FA based BPNN was first performed to model the end-milling process and predict CF and SR. It was later also executed to obtain the best combination of end-milling parameter levels that would provide minimum CF and SR. The outcome of the confirmation experiments disclosed that the integration of BPNN and FA managed to accurately predict and substantially enhance the multi-objective characteristics.