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

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. regression plot showed R = 0.9992 as progress training, 0.99865 validation and 0.85285 training test. This led overall correlation coefficient (R) 0.90007 which signified that ANN a robust tool for predicting weld dilution. To test reliability trained network, model was employed its own value same input parameters generated central composite design. Based on observed predicted values dilution, outputs thereafter generated, r2 0.9876 obtained conclusion network can be used beyond limit experimentation. There proximity results obtained, both experimental fell between 44.5-71.55%. Hence, prediction adopted applied actual scenario without fear inacuracies.