Optimization of machining parameters of A16061 composite to minimize the surface roughness – modelling using RSM and ANN

Abstract

<span style="font-size:9.0pt;mso-bidi-font-size: 10.0pt;color:#131313" lang="EN-US">The <span style="font-size:9.0pt; mso-bidi-font-size:10.0pt" lang="EN-US">A16061/SiC_p metal matrix composites are most frequently used for automobile and aerospace applications and the end milling operations are performed to achieve the better surface roughness of these composites. It is difficult to achieve required surface roughness on these composites due to their hardness. By considering the spindle speed (<i style="mso-bidi-font-style:normal">s</i>), feed rate (<i style="mso-bidi-font-style:normal">f</i>), depth of cut (<i style="mso-bidi-font-style: normal">d</i>) and nose radius (<i style="mso-bidi-font-style:normal">r</i>) and as predominant parameters and are optimized to achieve required surface roughness. In this regard, a versatile prediction model is required to determine the surface roughness of the composite considering the effect of machining parameters. In this research work, the response surface method (RSM) and an artificial neural network (ANN) based prediction models are developed to determine the surface roughness (<i style="mso-bidi-font-style: normal">R</i>_{<span style="mso-bidi-font-style:italic">a</span>}) of A16061/SiC_p and the performance of the RSM and ANN models are compared with experimental results for their effectiveness. The genetic algorithm (GA) based optimization of machining parameters for the RSM and ANN models are also carried out to minimize surface roughness. The results of the GA optimal parameters are analyzed for the convergence of various crossover and mutation probabilities and also to find the better prediction model.

</span></span>