Dissimilar material welding and assessing reliability of super alloy for green and high efficiency thermal power plant

Abstract

This paper studies the prediction of fatigue and corrosion fatigue lives using neural network and accelerated life methods for dissimilar material weld between Alloy617 and 12Cr steel. First, dissimilar material welding between Alloy617 and 12Cr steel was performed using buttering technology. The fatigue and corrosion fatigue strengths, and electrochemical corrosion susceptibility of dissimilar material weld were assessed. After that, on the basis of obtained data, fatigue life and corrosion fatigue life of dissimilar material weld were predicted using the neural network and accelerated life test methods. The predicted results showed good agreement with the actual fatigue and corrosion fatigue lives. Especially, the results of the neural network prediction were more accurate than those of the accelerated life method.