Thermodynamic prediction of martensitic transformation temperature in Fe-C-X (X=Ni, Mn, Si, Cr) systems with dilatational coefficient model

Abstract

Listen

Translate article

• Linear expansion during martensitic transformation is analyzed by machine learning. • Chemical driving forces are calculated more exactly by revising magnetic parameter. • M s prediction model is extended to multicomponent Fe-C-X ( X =Ni, Mn, Si, Cr) system. • The prediction error is reduced to 5.6% by dilatational coefficient model. Martensitic transformation is significant to strengthen steels, but its thermodynamic prediction is restricted to simple systems due to lacking multicomponent interaction parameters. The driving forces of martensitic transformation can be divided into chemical and non-chemical driving forces. The magnetic parameters are carefully optimized because it affects the magnetic Gibbs free energy of austenite and ferrite, and have big impact on the chemical driving force. The dilatational strain energy provides major contribution to non-chemical driving force, thus the integrated-models for dilatational coefficient are constructed in a wide composition and temperature range based on the experimental dilatational data. It expands the scope of application of thermodynamic model and improved prediction accuracy of martensitic transformation temperature ( M s ). The prediction error reaches 5.6% for Fe-C- X ( X =Ni, Mn, Si, Cr) and 6.5% for Fe-C-Mn-Si- X ( X =Cr, Ni) steels.