A Cluster-Based Computational Thermodynamics Framework with Intrinsic Chemical Short-Range Order: New Developments and Implications for Precipitate Nucleation Kinetics

Incorporating chemical short-range order (CSRO) into CALPHAD has long remained a major challenge in computational thermodynamics. Conventional CALPHAD frameworks lack the ability to explicitly characterize CSRO, while cluster-based methods such as the cluster variation method (CVM) incur prohibitive computational cost when applied to multicomponent systems. In this talk, we present our newly developed CVM–CALPHAD framework, which achieves an effective balance between flexibility, accuracy, and computational efficiency. By applying the Fowler–Yang–Li (FYL) transform to the CVM formalism, the computational cost is significantly reduced without sacrificing thermodynamic rigor [1]. Furthermore, physics-based parameterization is employed to explicitly account for non-configurational free energy contributions, enabling resolution of vibrational, elastic, and electronic effects.
Recent applications of the CVM–CALPHAD framework will be demonstrated, including CSRO phase diagram calculations for the ternary Cu–Au–Ag system and thermodynamic modeling of the Ni–Cr system exhibiting both FCC and BCC phases. The role of CSRO as a precursor to precipitate nucleation will be discussed. In particular, we highlight the critical importance of incorporating CSRO to accurately evaluate coherent interphase boundary energies and to provide physically reliable input parameters for nucleation kinetics models in multicomponent alloys.