ABSTRACT
Now a day’s zirconia toughened alumina have advance applications in manufacturing sectors as a cutting tool due to its superior properties compared to alumina ceramics. The optimisation of process parameter plays an important role in order to get better performance. Hence this experiment, has attempted to optimise the process parameter for machining of AISI4340 steel to evaluate the quality surface at different set of conditions. In this analysis ANN has been developed to predict the quality of surface. Feedback loop has been used to eliminate higher errors. Co-precipitation route has been used to prepare the well homogeneous powders of ZTA. The developed powders are shaped and sized according to the standard of ISO SNUN 120408 (ISO) for square inserts through powder metallurgy route. Cutting speed, feed rate and depth of cut have been selected as input variables to carry out the experiments whereas surface roughness has been considered as output of network at said corresponding conditions of machining. The observed experimental results are trained according to artificial neural network (ANN). Feed forward back propagation algorithm has been used for modelling to assess the surface morphology. The values of both training and testing for the convergence of mean square error are found in good agreement. The developed models are validated with experimental data signifies that the developed model are suitable for predication purpose. At last, the optimisation of process parameter is carried out on Design Expert software using Box – Behnken design approach. The optimum value obtained for minimum surface roughness with a desirability of 99.9983%.
