Your Cart

Your cart is empty

Home  >  Volume 48

43. APPLICATION OF ARTIFICIAL NEURAL NETWORK TO PREDICT THE HEAT INPUT OF TIG MILD STEEL WELDS by Uwoghiren F.O. and Achebo J.I. Volume 48 (Sept. & Nov., 2018 Issue), pp343 –346
Sale price: $5.00
Quantity:

APPLICATION OF ARTIFICIAL NEURAL NETWORK TO PREDICT THE HEAT INPUT OF TIG MILD STEEL WELDS

Uwoghiren F.O. and Achebo J.I. 

Department of Production Engineering, University of Benin, Benin City, Nigeria.

Abstract

With the rapidly changing scenario in the manufacturing industry, the optimization of process parameters is essential for a manufacturing unit. Heat input affects the quality of the weld. The aim of this study is to apply  expert systems such as artificial neural network to predict weld metal heat input of low carbon steel using the tungsten inert gas welding process in order to produce a better weldment. Mild steel plate was cut into dimension 60mm x 40mm x 10mm with a power hack saw, grinded and cleaned before the welding process. The experimental data was divided into three parts, which is training 60%,validating25% and testing 15%. The network output has a Momentum gain of 1.0e-16 and an error value of 1.2134e-6 at epoch 9 is an evidence of a network with strong capacity to predict the heat input. A Coefficient of determination (r2) values of 0.9974  was employed to draw a conclusion that the the trained network can be used to predict the heat input beyond the limit of experimentation 

click here to download full Abstract
Close
Loading...