[Paper Review] General Phase-Field Model with Stability Requirements on Interfaces in $N$-Dimensional Phase-Field Space

Evgeny Pogorelov, Julia Kundin+1 more|arXiv (Cornell University)|Apr 24, 2013

Solidification and crystal growth phenomena3 citations

TL;DR

This paper presents a generalized N-dimensional phase-field model that ensures thermodynamic stability and eliminates ghost phases through Lagrange multiplier-based constraint enforcement and carefully constructed model functions. It enables accurate simulation of multi-phase microstructures, including four-phase reactions in Al-Cu-Ni alloys, with controlled nucleation via thermal noise and validated adherence to Young’s law for contact angles.

ABSTRACT

In this paper a general multi-phase-field model is presented which is an extension and modification of the model proposed by Folch and Plapp for three phase fields [R. Folch and M. Plapp, Phys. Rev. E 72 011602 (2005)] to the arbitrary number of phases. In the model a physical constraint requiring that the sum of all phase fields in the system is equal to one is resolved by the method of Lagrange multipliers. Namely, the thermodynamic driving force is reduced to its projection on the plane of the constraint. The general model functions in a $N$-dimensional phase field space were derived which justify the requirements for the stability of the total free energy functional on dual interfaces and hence the absence of "ghost" phases. Furthermore, the case of the different interface energies and mobility parameters on the individual interfaces is resolved in a comprehensive manner. It is shown that the static equilibrium for three or four phases fulfils Young's law for contact angles with high accuracy. Also the model is verified by the quantitative simulation of the solidification in an Al-Cu-Ni alloy in the case of the four-phase transformation reaction. We found the way to control the character of new phase nucleation using additional terms in free energy functional.

Motivation & Objective

• To develop a general phase-field model for N-phase systems that maintains thermodynamic stability and avoids ghost phases.
• To extend the Folch-Plapp three-phase model to arbitrary numbers of phases using Lagrange multipliers for the sum-to-one constraint.
• To incorporate different interface energies and mobilities for each interface, enabling realistic kinetic behavior.
• To verify the model’s ability to reproduce Young’s law for contact angles and control nucleation via thermal noise.
• To apply the model to quantitative simulation of four-phase peritectic-like reactions in multicomponent Al-Cu-Ni alloys.

Proposed method

• The model enforces the physical constraint that the sum of all phase fields equals one using Lagrange multipliers, projecting the thermodynamic driving force onto the constraint plane.
• Stable and flat model functions are constructed in N-dimensional phase-field space to ensure stability on both interfaces and triple junctions.
• The free energy functional includes terms for interface energy barriers and chemical driving forces, with anisotropy in interface energy and mobility parameters explicitly modeled.
• A thin-interface asymptotic analysis is applied to derive the phase-field evolution equations, ensuring consistency with standard phase-field models.
• Thermal noise is introduced into the phase-field evolution equation to trigger and control heterogeneous nucleation of new phases.
• The model is validated through numerical simulations of microstructure evolution in a 2D Al-Cu-Ni alloy under isothermal conditions.

Experimental results

Research questions

• RQ1 Can a general N-phase-field model be formulated that ensures stability and avoids ghost phases in multi-phase systems?
• RQ2 How can different interface energies and mobilities be consistently incorporated into a multi-phase model with a sum-to-one constraint?
• RQ3 To what extent does the model reproduce Young’s law for contact angles at triple junctions?
• RQ4 Can thermal noise in the phase-field equation be used to control the nucleation of new phases in a physically consistent manner?
• RQ5 How accurately does the model simulate complex four-phase transformation reactions in multicomponent alloys?

Key findings

• The model successfully enforces the sum-to-one constraint on phase fields via Lagrange multipliers, ensuring thermodynamic consistency and eliminating ghost phases.
• The model accurately reproduces Young’s law for contact angles, with deviations observed only in 3D simulations, indicating strong agreement with analytical predictions.
• Thermal noise in the phase-field evolution equation enables controlled, physically realistic nucleation of new phases at favorable energetic conditions, such as on solid/liquid interfaces.
• In simulations of Al-Cu-Ni alloy solidification, new β- and γ-phases nucleate heterogeneously on the boundaries of the liquid and α-phase, leading to lamellar microstructure formation.
• Without thermal noise, one phase overgrows the other; with noise, nucleation occurs immediately at favorable sites, enabling stable and uniform lamellar growth.
• The model allows tuning of the nucleation barrier through additional terms, including suppression of spurious nucleation at triple junctions.