Predicting the Influence of Process Parameters on Depth of HAZ Using Artificial Neural Network on Shielded Metal Arc Welded AISI 1018 Low Carbon Steel Joints

An artificial neural network model was executed after being developed by training a program in C++ by utilizing welding variables including input parameters like electrode angle, welding current, welding speed and welding voltage and output parameters like depth of HAZ. Experimental data were utilized to model neural network based on back propagation algorithm to predict the effects of welding parameters on weld bead geometry factors. It has been noticed that an accurately trained artificial neural network model can be easily and efficiently utilized for predicting the optimum values of depth of heat affected zone.