An Interfacial Profile-Preserving Approach for Phase Field Modeling of Incompressible Two-Phase Flows

TL;DR

This paper introduces a novel phase field modeling approach that preserves interfacial profiles in incompressible two-phase flows, improving accuracy and mass conservation over traditional methods.

Contribution

It presents an interfacial profile-preserving formulation using a signed distance function, enhancing simulation accuracy and stability in phase field models.

Findings

Improved accuracy in simulating complex interfacial dynamics.

Enhanced mass conservation in two-phase flow simulations.

Validated effectiveness through multiple benchmark tests.

Abstract

In this paper, we introduce an interfacial profile-preserving approach for phase field modeling for simulating incompressible two-phase flows. While the advective Cahn-Hilliard equation effectively captures the topological evolution of complex interfacial structures, it tends to displace the fluid interface from its equilibrium state, impacting simulation accuracy. To tackle this challenge, we present an interfacial profile-preserving formulation that relies on a phase-field-related signed distance function, rather than the phase field function itself. It is solved iteratively to restore the equilibrium interface profile after each time step. This approach effectively minimizes discretization errors and enhances mass conservation accuracy for each phase. Our formulation is discretized using a second-order Total Variation Diminishing (TVD) Runge-Kutta method within iterations and a finite volume scheme in spatial discretization. We quantitatively compare our present profile-preserving method with the original method in terms of accuracy and convergence rate through simulations of a deforming drop in a single vortex and a rising bubble in quiescent fluid, and further validate the applicability through simulations of a two-dimensional contracting liquid filament, a drop impacting a deep liquid pool, and three-dimensional drop deformation in shear flow. Our results exhibit good agreement with analytical solutions, prior numerical results, and experimental data, demonstrating the effectiveness and accuracy of our proposed approach.