Surface Wear Rate Prediction in Reinforced AA2618 MMC by Employing Soft Computing Techniques

As a result of the enhancement of mechanical and thermal properties, Metal matrix Composites (MMC) has shown enormous potentials as a likely material for various aerospace and automotive applications. Thermomechanical wear properties of Aluminium Alloy (AA 2618) is reinforced with Silicon Nitride (Si3N4), Aluminium Nitride (AlN) and Zirconium Boride (ZrB2) have been investigated in the proposed work. For the fabricated MMC, the prediction of wear features can be developed by the regression model and employing Statistical and Data Analysis named Minitab. The predicted response was found to be a linear with the actual responses. Utilizing the Support Vector Machine (SVM) and Artificial Neural Networks (ANN), prediction of the wear characteristics is performed with the developed regression model. The statistical performance parameter through ANN exhibits the minimization of the Mean Absolute Error (MAE) when compared to other approaches. The % of MAE through the soft computing techniques such as ANN, SVM, LR are 10%, 26% and 29% respectively. Predicted values obtained by means of the proposed approaches through the optimized ANN model are having sustained consistency on par with the conventional values.