Abstract
Alternative schemes for the prediction of liquidus temperatures have been tested against a large compilation of experimental measurements for steels ranging from low carbon to stainless steels. The best agreement overall was obtained with thermodynamic solutions using the IDS (model of interdendritic solidification) database, although certain empirical equations from the literature were adequate for low alloyed steels. Version L of the commercial ThermoCalc thermodynamics package was outperformed for some steel types by the empirical equations, prompting a reassessment of the Fe–Cr–Ni, Fe–Cr–Si, and Fe–Ni–Si systems used in the IDS database. Additionally, regression equations were determined from several thousand predictions computed by the IDS thermodynamic approach, and these performed almost as well as the proper thermodynamic approach at a fraction of the central processing unit time, also outperforming the earlier approaches.
Get full access to this article
View all access options for this article.
