Artificial Neural Network Modeling of Grinding of Ductile Cast Iron using Water Based SiO2 Nanocoolant

Abstract

This paper presents optimization of the grinding progress of ductile cast iron using water-based SiO2 nanocoolant. Conventional and water-based nanocoolant grinding was performed using a precision surface grinding machine. The study is aimed to investigate the effect of table speed and depth of cut on the surface roughness and material removal rate (MRR). Mathematical modeling is developed using the response surface method. An artificial neural network model is developed for predicting the surface roughness and MRR. Multi-layer perception and a batch back propagation algorithm are used. MLP is a gradient descent technique to minimize the error through a particular training pattern in which it adjusts the weight by a small amount at a time. From the experiment, the depth of cut is directly proportional to the surface roughness, but the table speed is inversely proportional to the surface roughness. The higher the value of the depth of cut, the lower the value of MRR, and vice versa for the table speed. It is concluded that the surface quality together with the material removal rate are the most affected by the depth of cut(s) and table speed.