Journal Published Online: 14 June 2019
Volume 8, Issue 1

Analysis of Artificial Neural Network for Predicting Erosive Wear of Nylon-12 Polymer

CODEN: MPCACD

Abstract

In this study, artificial neural network (ANN) is applied to predict the erosion rate of nylon-12 polymer to ensure the accuracy of soft computing. The ANN model was developed with 4 neurons in the input layer, 10 neurons in the hidden layer, and 1 neuron in the output layer. A backpropagation algorithm was utilized in a multilayered perception. The inputs include impact velocity (m/s), impingement angle (°), erodent size (μm), and stand-off distance (mm). Experimental data were used to predict the erosion rate in relation to the input parameters. The size of the erosive element of randomly shaped sand particles (silicon dioxide) is set between 300 and 600 μm, the impact velocity between 30 and 50 m/s, the impingement angle between 30° and 90°, and the stand-off distance between 15 and 25 mm. The consistency between the experimental and ANN model values, with an accuracy of 94.428 % and root mean square error of 9.729, signifies that the proposed ANN model is suitable for predicting the erosion rate of nylon-12 polymer. The prediction made using the ANN model was in good agreement with the experimental results. The ANN model can be used to estimate the maximum and total erosion rate of nylon-12 with high reliability. Therefore, this model can be applied for practical purposes.

Author Information

Shuvho, Bengir Ahmed
Department of Mechanical Engineering, Dhaka University of Engineering and Technology, Gazipur, Gazipur, Bangladesh
Chowdhury, Mohammad Asaduzzaman
Department of Mechanical Engineering, Dhaka University of Engineering and Technology, Gazipur, Gazipur, Bangladesh
Debnath, Uttam Kumar
Department of Mechanical Engineering, Dhaka University of Engineering and Technology, Gazipur, Gazipur, Bangladesh
Pages: 13
Price: $25.00
Related
Reprints and Permissions
Reprints and copyright permissions can be requested through the
Copyright Clearance Center
Details
Stock #: MPC20180164
ISSN: 2379-1365
DOI: 10.1520/MPC20180164