Neural Network based Predictive Analysis of Surface Roughness using Bayesian Regularization in Turning of Monel K500

Hybrid machining techniques are becoming more popular in the processing of difficult-to-cut materials because of the inherent benefits they provide over individual machining processes provided by the combined use of two or more machining machines. This study investigates the transformation of Monel K500 into a superalloy using a neural network-based Bayesian regularization algorithm. The performance parameter was average surface roughness (Ra), while the control parameters were spindle speed, depth of cut, and feed rate. The modeling and prediction abilities of the proposed neural network were compared to experimental trials using the coefficient of determination. Typically, there is a 1.97% difference between actual values and predictions made using neural network-based modeling. The neural network's predictions of surface roughness levels closely match the actual experimental results.