Deep Neural Network와 Support Vector Regression을 이용한 인콜로이 825의 고온변형 연구

Abstract

Incology825 is a nickel-based alloy with good corrosion resistance as well as heat resistance. For effective application of Incoloy825, the flow behavior of the Incoloy 825 during hot deformation needs to be studied and the efficient flow model developed. In this study, hot deformation test was carried out under the temperature of 925℃, 1050℃ and 1150℃ and strain rate of 0.003/s and 3/s. Deep neural network and support vector regression algorithm was thereafter used to model the flow behavior of Incoloy825. The research found that both algorithms describes the flow stress of incoloy825 well. Comparison between deep neural network and support vector regression was done.