The modeling and process analysis of resistance spot welding on galvanized steel sheets used in car body manufacturing

In this study, the resistance spot welding (RSW) process of the galvanized interstitial free (IF) steel sheets and galvanized bake hardenable (BH) steel sheets, used in the manufacturing of car bodies, has been modeled and optimized. The quality measure of a resistance spot welding joint is estimated from the tensile–shear strength. Furthermore, four important process parameters, namely welding current (WC), welding time (WT), electrode force (EF), and holding time (HT) are considered as the factors influencing the quality of the joints. In order to develop an accurate relationship between the process inputs (4-component vectors) and the response output (tensile–shears strength) at first a linear regression model was utilized but the residuals analysis revealed a non-linear behavior. Therefore, an artificial neural network (ANN) was proposed because the ANNs are capable of mapping the non-linear systems. A back propagation neural network model was developed to analyze RSW process and the interaction effects of the parameters. In the second phase of this research, Genetic Algorithm with the fitness function based on an ANN model was employed as an optimization procedure for determining a set of process parameters; as a result, the maximum joint strength was obtained. Optimization results showed high compatibility with the actual experimental data.