Application of Numerical and Computational Based Models for Modelling the Effects of the Electrode Density in Mild Steel TIG Welding Process

Abstract

The well known arc welding technique of Tungsten Inert Gas (TIG) welding is widely used to join thin pieces of practically all ferrous and non ferrous materials. However, there have been many improvements made to the TIG welding process as a result of the increased interest of industries in using the method for joining components with mid thick sections. In the study, emphasis was placed on the density of the TIG electrode being welded with mild steel in which the studies employed 100 pieces of coupons made of mild steel that were 80 x 40 x 10 (mm) in size. Using 5 specimens each time, the experiment was done 20 times. The plates' edges were machined and bevelled before being welded using tungsten inert gas welding equipment. The Response Surface Methodology (RSM) and the Artificial Neural Network (ANN) were used to ascertain and optimize the electrode density of the welded specimen. The RSM model generated a numerically ideal solution with the following values: 200.72A current, 20V voltage, 2.40mm wire diameter, and 20m/s wire feed speed, resulting in an electrode density of 6511.24kgm/s2. With a desirability value of 93.9%, the design expert determined that this solution was the best option. In the ANN, 70% of the data was used for training, 15 % was used for validating and the last 15% for the actual test. From the results obtained a regression plot that displays the relationship between the input factors and the desired outcome was produced with R2 values of 0.84831. The ANN is selected as the better predictive model over the RSM because the ANN output fits closer to the experimental than that of RSM. Thus, the approaches effectively optimized and predicted the electrode density.