Experimental methodology and predictive tools in austenitic stainless steel with different ferrite content welds

Abstract

The ferrite content of austenitic stainless steel welds seriously affects to perform stainless steel welds. Keeping the ferrite content at the lower limit can reduce the thermal crack sensitivity of the weld. At high temperature, high ferrite content will lead to embrittlement tendency of weld. The traditional determination methods mainly include magnetic method and metallographic method. Magnetic method is often used as a method to obtain data for predicting ferrite content. However, magnetic method has high requirements on sample size and surface finish. Metallographic methods are usually used as a supplement to magnetic methods because of their destructive and other disadvantages. Accurate prediction of ferrite content can reduce measurement work and facilitate the selection of parameters before welding. The most widely used magnetic method is WRC-1992 diagram, but the effect of welding parameters is not reflected. The neural network method improves the prediction accuracy, but it still has the limitation of insufficient consideration of welding parameters. This article reviews the present status in this aspect, Additionally, some challenges and future outlooks of the existing measurement and prediction methods are summarised.