Application of Bayesian neural network for modelling and prediction of ferrite number in austenitic stainless steel welds

As neural networks are extremely usefill in recognising patterns in complex data, Bayesian neural network analysis has been followed in the present work to reveal the influence compositional variations oil ferrite content for the austenitic stainless steel base compositions from the available database and to stud), the significance of individual elements for ferrite content in austenitic stainless steel welds based oil the optimised neural network model. The Bayesian neural network's predictions are accompanied by error bars and the significance of each input variable is automaticaly quantified in this type of analysis. A neural network model based oil Bayesian framework for ferrite prediction in austenitic stainless steel welds has been developed using the database which was used for generating the WRC-1992 diagram. The Bayesian framework uses a committee of models for generalisation rather than a single model. The best model was chosen based on the minimum in the test error and the maximum in the logarithmic predictive error. The optimised model call be used for predicting the ferrite number in austenitic stainless steel welds with a better accuracy than the constitution diagrams. Using this model, the influence of variations in the individual elements such as carbon, manganese, silicon, chromium, nickel, molybdenum, nitrogen, niobium; titanium, copper, vanadium, and cobalt on the ferrite number in austenitic stainless steel welds has been determined. It was found that the change in ferrite number is a nonlinear function of the variation in the concentration of the elements. Elements such as silicon, chromium, nickel, molybdenum, nitrogen, titanium, and vanadium were found to influence the ferrite number more significantly than the rest of the elements in austenitic stainless steel welds. Manganese was found to have less influence oil the ferrite number. Titanium was found to influence the ferrite number more significantly than niobium. This observation is new as the WRC-1992 diagram only considered the niobium content it) calculating the chromium equivalent.