Regression Model for Prediction of Heat Generation and Peak Temperature During Friction Stir Welding of Steel

Abstract

AbstractFriction Stir Welding (FSW) is a popular and frequently used solid-state welding method because of its several advantages over traditional weldings, such as lower weight, shorter processing time, no fusion or melting and so on. Joining materials such as aluminium, magnesium, mild steel and stainless steel are welded using this approach. In this research, a regression model is developed to predict the heat generation rate and peak temperature during friction stir FSW welding of mild steel. In this model, based on the data available for power, heat generation and temperature in peak regression models were developed. The power required, heat generation per unit length and peak temperature predicted by thermal model (TM) and regression models (RM) were compared and it shows good correlation. The power required rate is proportional to the rotational speed (N) and welding speeds (V). The amount of heat rate produced, and the peak temperature reached is related to the RS and inversely proportional to the WS. Higher welding speeds and rotational speeds were found to generate the most heat.KeywordsMild steelRegression modelPower requiredHeat generationFriction stir weldingPeak temperature