Prediction of Percentage Dilution in AISI 1020 Low Carbon Steel Welds Produced from Tungsten Inert Gas Welding

Abstract

The objective of this study is to predict the percentage dilution in AISI 1020 low carbon steel welds produced from tungsten inert gas welding using Artificial Neural Networking (ANN) approach. The regression plot showed R = 0.9992 as progress of training, R = 0.99865 as progress of validation and R = 0.85285 as progress of the training test. This led to overall correlation coefficient (R) of 0.90007 which signified that ANN is a robust tool for predicting the percentage of weld dilution. To test the reliability of the trained network, the ANN model was employed to predict its own value of percentage dilution using the same input parameters generated from the central composite design. Based on the observed and the predicted values of percentage dilution, a regression plot of outputs was thereafter generated, and r² value of 0.9876 was obtained which led to the conclusion that the trained network can be used to predict percentage dilution beyond the limit of experimentation. There was proximity in the results obtained, as both the experimental and predicted weld dilution fell between 44.5-71.55%. Hence, prediction adopted in this study can be applied in actual scenario without fear of inacuracies.