Data-Driven Approach for Investigation of Irradiation Hardening Behavior of RAFM Steel

Abstract

Knowledge reasoning plays an important role in applications such as the human relation web and semantic search. However, how to use this method to solve materials problems is still a challenge. Defects and damage induced by neutron irradiation significantly affect the service performance of materials. Reduced Activation Ferritic/Martensitic (RAFM) steel is a very promising candidate for application in fusion reactor cladding. Understanding irradiation hardening effects in RAFM steel is one of the critical issues. Some experimental data of RAFM steel under irradiation are collected to construct a data set. The relationship between yield strength variation after irradiation and elements and irradiation conditions is trained by the machine learning method. The influence of irradiation condition and alloy elements on the hardening behavior of RAFM steel was explored, and some optimal alloy elements composition was also recommended. This work will give some direction for RAFM steel research.