Prediction of Impact Energy of TIG Mild Steel Welds Using ANN

Abstract

The present trend in the fabrication industries is the use of automated welding processes to obtain high production rates and high quality output.TIG welding, happens to be the best welding method employed in  the manufacturing industry. one of the problem facing the fabrication industry is the control of the process input parameters to obtain a good welded joint . however it is essential to establish the relationship between process parameters and weld quality output to predict and control weld bead quality .The aim of this study is to predict the impact energy of TIG mild steel welds using ANN.In this study, twenty experimental runs were carried out, each experimental run comprising the current, voltage and gas flow rate, the TIG welding process was used to join two pieces of mild steel plates measuring 60 x40 x10 mm , the impact energy was measured respectively. Thereafter the data collected from the experimental results was analysed with the ANN. The experimental results for the impact energy was analyzed with the Artificial Neural Networks. The overall R-value is shown to be 98.7%.  The best validation performance is 0.48429 and occurred at epoch five (5). The coefficient of correlation for training shows of 99.9%closeness ,99.4% for validation and 89.8% for  testing respectively.