Weight analysis of influencing factors on hot cracking susceptibility of multi-alloy steel based on solidification micro-segregation model

Abstract

Aiming at the problems of complex and diverse influencing factors and insufficient weight analysis in evaluating the hot cracking susceptibility of multi-alloy steel by solidification micro-segregation model, the brittleness temperature range (BTR) and Kou criterion were used as the evaluation indexes for hot cracking susceptibility, and the optimal number of principal components was determined by partial least squares regression (PLSR). Based on the optimal number of principal components, the interrelationships between factors, principal components and dependent variables were elucidated, and the weight of influencing factors of hot cracking susceptibility were analyzed. The results show that the factor weights affected the evaluation of hot cracking susceptibility were, in order, the initial concentration of components, the liquidus correlation coefficient, and the equilibrium distribution coefficient. Their contributions to BTR were 0.43, 0.38, and 0.16, respectively, and to Kou criterion were 0.42, 0.38, and 0.17, respectively. PLSR had good predictive performance for BTR and Kou criterion, with R2 values of 0.898 and 0.975, respectively, effectively capturing the relationship between factors and dependent variables.

Keywords

• Multi-alloyed steels
• Micro-segregation model
• Hot cracking susceptibility
• Partial least squares regression
• Factor weight