Fluid-Solid Coupling in Kinetic Two-Phase Flow Simulation

Wei Li, Mathieu DesbrunACM Trans. Graphics 2023 · OpenReview

TL;DR

This paper introduces a more robust and efficient kinetic two-phase flow simulator that significantly improves fluid-solid coupling, enabling realistic simulations of complex interactions like splashing and bubbling. By refining interface handling and boundary conditions, the method overcomes previous limitations that led to simulation instability, opening up possibilities for simulating a wider range of real-world phenomena.

What Problem Does This Paper Address?

Simulating the dynamic interplay between fluids and solid objects is crucial for realistic visual effects and scientific understanding. While advancements in fluid simulation, particularly using kinetic solvers like the Lattice Boltzmann Method (LBM), have enabled impressive multiphase flow simulations (e.g., water and air) at high Reynolds numbers, accurately coupling these with moving solid objects has remained a significant challenge. Previous methods often suffer from numerical instabilities, especially with fast-moving or complex objects, leading to spurious oscillations and simulation blow-ups. This paper aims to bridge this gap by developing a more stable and versatile fluid-solid coupling technique for kinetic two-phase flow simulations.

Methods

The authors build upon an existing phase-field and velocity-distribution based LBM solver for multiphase flows. Their key contributions lie in several numerical improvements:

  • Hybrid Moving Bounce-Back (HMBB) Scheme: They adapt the moving bounce-back (MBB) method for fluid-solid coupling. This scheme, which handles momentum exchange between the fluid and moving solids, is enhanced by using a smoothed pressure filter to mitigate spurious oscillations near boundaries.
  • Improved Interface Normal Evaluation: Accurate calculation of the interface normal is critical for interfacial forces. The paper proposes a novel scheme that combines two existing methods to achieve a more isotropic and artifact-free gradient, crucial for stability, especially in dynamic scenarios.
  • Finer Resolution for Phase Field: To further improve interface accuracy, the phase field is computed on a grid that is twice as fine as the fluid grid. This staggered grid approach is handled efficiently through interpolation and averaging techniques.
  • Cut-Cell Updates: For dynamic solids, the simulation needs to track which grid cells are occupied by the solid (cut-cells). The paper details a robust method for updating these "dead" and "fresh" cells to maintain simulation integrity without requiring explicit refilling.
  • Soft-Start Initialization: To handle situations with fast-moving solids that can otherwise cause transient compression waves and instabilities, a soft-start initialization strategy is employed for the first 20 frames.

Datasets

Not discussed in the paper. The paper focuses on developing and validating its simulation methodology through various challenging scenarios, often compared to real-world experiments or previous simulation results.

Key Insights and Ideas

  • The "sweet spot" for interface accuracy: The authors demonstrate that simply increasing the phase field grid resolution isn't enough; the way the interface normal is calculated is paramount for stability, especially in dynamic fluid-solid interactions.
  • Combining strengths for robustness: By blending different approaches for interface normal calculation, they achieve a result that is both accurate in fluid-solid interaction regions and smooth in bulk fluid regions, avoiding artifacts that plagued previous methods.
  • Staggered grids are manageable: The paper shows that using different grid resolutions for the fluid and phase fields, while seemingly complex, can be handled efficiently with careful interpolation and averaging, leading to improved accuracy without excessive computational overhead.
  • Pressure oscillations are the enemy: The core of their stability improvements revolves around mitigating spurious pressure oscillations near solid boundaries, which are a common failure point in previous fluid-solid coupling methods.

Key Results

  • Enhanced Stability: The proposed method significantly improves the stability of fluid-solid coupling, allowing for simulations of fast-moving and complex objects that previously led to blow-ups. This is demonstrated through examples like a car in a flood and a coin skipping on water.
  • More Realistic Dynamics: The simulations exhibit finer turbulent details and more accurate bubble formation and splashing compared to previous work, as shown in comparisons with [Li et al. 2022].
  • Efficiency Gains: Despite using a finer grid for the phase field, the authors report a significant reduction in memory usage (around 54% less than a comparable method) and improved simulation speed (around 4.5 times faster in one example).
  • Broader Applicability: The solver can handle a wide range of scenarios, including thin and thick objects, objects of varying densities, and high Reynolds number flows, with a density ratio of 800 between air and water.

Limitations and Future Work

  • Mass Preservation: The current "dead cut-cell" treatment is efficient but does not guarantee exact mass preservation. The authors suggest that adding Lagrangian particles could address this.
  • Deformable Solids: The current work focuses on rigid bodies. Coupling with deformable solids is a potential area for future research, as it might require additional efforts to prevent blow-ups.
  • Rigid Body Solver Accuracy: The authors note that many real-time rigid body solvers have limitations in collision treatment accuracy, which can impact the stability of coupled fluid-solid simulations. Improving the coupling to accommodate these flaws without resorting to monolithic solvers is a future direction.

Summary updated

Abstract

Real-life flows exhibit complex and visually appealing behaviors such as bubbling, splashing, glugging and wetting that simulation techniques in graphics have attempted to capture for years. While early approaches were not capable of reproducing multiphase flow phenomena due to their excessive numerical viscosity and low accuracy, kinetic solvers based on the lattice Boltzmann method have recently demonstrated the ability to simulate water-air interaction at high Reynolds numbers in a massively-parallel fashion. However, robust and accurate handling of fluid-solid coupling has remained elusive: be it for CG or CFD solvers, as soon as the motion of immersed objects is too fast or too sudden, pressures near boundaries and interfacial forces exhibit spurious oscillations leading to blowups. Built upon a phase-field and velocity-distribution based lattice-Boltzmann solver for multiphase flows, this paper spells out a series of numerical improvements in momentum exchange, interfacial forces, and two-way coupling to drastically reduce these typical artifacts, thus significantly expanding the types of fluid-solid coupling that we can efficiently simulate. We highlight the numerical benefits of our solver through various challenging simulation results, including comparisons to previous work and real footage.
BibTeX
@article{li2023fluid,
  author = {Wei Li and Mathieu Desbrun},
  title = {{Fluid-Solid Coupling in Kinetic Two-Phase Flow Simulation}},
  journal = {ACM Trans. Graphics},
  year = {2023},
  url = {https://openreview.net/forum?id=Sxsz1D81Jfk},
}
Figure