An Extension of the Localized Artificial Diffusivity Method for Immiscible and High Density Ratio Flows

Steven R. Brill, Britton J. Olson, Guillaume T. Bokman · 2025

Read An Extension of the Localized Artificial Diffusivity Method for Immiscible and High Density Ratio Flows by Steven R. Brill, Britton J. Olson, Guillaume T. Bokman (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-12680.txt.

Introduction

Steven Brill, Britton Olson and Guillaume Bokman extended in 2025 a numerical method for flows where two fluids meet and barely mix. The localized artificial diffusivity method steadies simulations of several materials, but it struggled when densities differed hugely or when fluids must stay unmixed. The authors traced the flaws to how the smoothing flux was built. Their fix aims the smoothing at volume fraction edges instead of mass fraction.

An added bulk diffusion term keeps pressure and temperature in balance across the boundary. A further term holds the interface at a finite thickness, as immiscible flows need. The design works for any number of materials at once. Tests on standard cases and on microbubble collapse in water matched measured data. The authors present the method as a robust tool for hard multi-fluid problems.

Catalog record

CREATORSteven R. Brill, Britton J. Olson, Guillaume T. Bokman
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE64 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-12680

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

An Extension of the Localized Artificial Diffusivity Method for Immiscible and High Density Ratio Flows Steven R. Brill, Britton J. Olson, Guillaume T. Bokman Abstract The localized artificial diffusivity (LAD) method is widely regarded as the preferred multi-material regularization scheme for the compact finite difference method, because it is conservative, easy to implement, and generally robust for a wide range of multi-material problems. However, traditional LAD methods face significant challenges when applied to flows with large density ratios and when maintaining thermodynamic equilibrium across material interfaces. These limitations arise from the formulation of the artificial diffusivity flux and the reliance on enthalpy diffusion for interface regularization. Additionally, traditional LAD methods struggle to ensure stability under large density ratio conditions, fail to maintain a finite interface thickness, and are therefore unsuitable for modeling immiscible interfaces. In this work, we discuss the origins of these issues in traditional LAD methods and propose modifications which enable the simulation of large density ratio and immiscible flows. The proposed method targets the artificial diffusion fluxes at gradients and ringing in the volume fraction, rather than the mass fraction in traditional methods, to consistently regularize large density ratio interfaces. Furthermore, the proposed method introduces an artificial bulk density diffusion term to enforce equilibrium conditions across interfaces. To address the challenge of modeling immiscible flows, a conservative diffuse interface term is incorporated into the formulation to ensure a finite interface thickness. Specific consideration is taken in the design of the method to ensure that these crucial properties are maintained for N N -material flows. The effectiveness of the proposed method is demonstrated through a series of canonical test cases, and its accuracy is validated by comparison with experimental data on micro-bubble collapse in water. These results highlight the method’s robustness and its ability to overcome the limitations of traditional LAD approaches. 1 Introduction Multi-material flows occur in a wide variety of applications such as combustion [ 1 ] , inertial confinement fusion [ 2 ] , targeted drug delivery [ 3 ] , and additive manufacturing [ 4 ] . These problems are often simulated with a single-fluid approach where a single set of governing equations treats all of the materials as separate components of a single fluid. Each component has its own material properties, equations of state, and potentially physical terms, such as material strength for solids. Accurately capturing the material interfaces introduces a number of computational challenges such as: accurately capturing large jumps in material properties, location of interfaces, and maintaining thermodynamic equilibrium at interfaces [ 5 ] . Simulations of multi-material flows are often performed on Eulerian grids, Lagrangian frames, or with arbitrary Lagrangian-Eulerian (ALE) methods. Lagrangian and ALE methods have shown successes for capturing complex multi-material mixing, but tend to be diffusive and suffer from poorly conditioned cells and mesh tangling [ 6 , 7 ] . In this work, we use Eulerian grids because they allow for the use of high-order finite difference methods, which can capture the small-scale features of turbulence without over-dissipation [ 8 ] . In the context of Eulerian methods, there are two main multi-material approaches: interface tracking and interface detection. Interface tracking involves explicitly evolving the location of the interface in methods such as the marker-in-cell method [ 9 ] and front-tracking methods [ 10 ] . These methods accurately capture interface shapes, but require complex treatments at the interface. Interface detection methods compute the location of the interface using the flow variables within the simulation, which allows for better capturing of topology changes and interface creation. Within interface detection methods, there are sharp and diffuse interface methods. Sharp interface methods capture the material interface as a single-point jump, which often requires special treatment to maintain numerical stability [ 11 ] . Additionally, methods such as the volume-of-fluid [ 12 ] and level-set methods [ 13 ] reconstruct the interface location for accuracy, but struggle to maintain exact conservation [ 14 ] . Diffuse interface methods spread the material interface over multiple grid points, which avoids numerical instabilities, but introduces complexity with the treatment of points with multiple materials [ 15 , 5 , 16 , 17 ] . In this work, we are using a diffuse interface approach due to its simplicity of interface representation and conservation properties. In this work, we are adopting a four-equation approach, where conservation of species mass, total momentum, and total energy are solved and points with multiple materials are assumed to instantaneously reach pressure and temperature equilibrium. In the literature, authors have studied the five [ 16 ] , six [ 15 ] , and seven-equation models [ 17 ] where pressure and temperature equilibrium are reached via relaxation terms and additional partial differential equations (PDEs) are solved for material volume fractions, internal energies, and velocities respectively. While these methods can be more expensive due to the additional PDEs, they have shown success for applications with large density ratios, disparate equations of state, and non-equilibrium effects [ 15 , 16 , 17 ] . A wide range of numerical methods from low-order finite volume [ 18 ] , finite difference [ 19 ] , high-order discontinuous Galerkin [ 20 ] , and high-order finite difference methods [ 21 ] have shown success for simulating multi-material flows with the diffuse interface method. In this work, we are choosing to use high-order compact finite difference method [ 8 ] due to excellent resolution of the small-scale features of turbulence. High-order methods introduce additional challenges in the context of multi-material flows because large differences in densities or material properties at interfaces can lead to Gibbs oscillations, which lead to unphysical values and code crashes [ 22 ] . In order to address this issue, Cook introduced artificial-fluid large-eddy simulation (AFLES), where artificial fluid properties are added to the diffusive fluxes to stabilize unphysical oscillations while maintaining the excellent spectral properties of the underlying numerical method [ 23 ] . Then Cook added an enthalpy diffusion term to the multi-material treatment in AFLES to avoid violations of the entropy conditions and unphysical temperature spikes caused by the artificial species diffusion [ 21 ] . The artificial diffusivity method of AFLES was developed further for curvilinear grid and to improve its properties for turbulence and coined localized artificial diffusivity (LAD) methods [ 24 , 25 , 26 , 27 , 22 ] . LAD methods have also been developed in the context of different underlying numerical schemes as well such as 2nd order finite difference methods and flux reconstruction [ 19 , 28 , 29 , 30 , 31 ] . One limitation of previous LAD methods is that they are unable to maintain equilibrium conditions at material interfaces without the adding additional PDEs and specifying specific equations of state [ 32 ] . Additionally, LAD methods struggle for large density ratios. Another weakness of LAD methods is that they continuously diffuse at the interface, so a finite interface thickness cannot be enforced. One common way to address this is through the phase-field method, where an additional PDE is solved to compute an additional flux term to enforce a constant interface thickness [ 33 ] . This method has seen success for a wide range

Read the complete text