On the consistency of pseudo-potential lattice Boltzmann methods

Luiz Eduardo Czelusniak, Tim Niklas Bingert, Mathias J. Krause, Stephan Simonis · 2025

Read On the consistency of pseudo-potential lattice Boltzmann methods by Luiz Eduardo Czelusniak, Tim Niklas Bingert, Mathias J. Krause, Stephan Simonis (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-2504-15018.txt.

Introduction

Luiz Czelusniak, Tim Bingert, Mathias Krause, and Stephan Simonis put a popular multiphase flow method, in a 2025 preprint, on firm mathematical footing. The pseudo-potential lattice Boltzmann method simulates fluids that split into liquid and vapor. Users long tuned its knobs by trial and error. The authors derive the exact equation the method solves under diffusive scaling. That derivation ties each knob to a physical trait: phase densities, interface thickness, and surface tension. Users can now set properties in real units from the start.

The approach ships inside the OpenLB software with a unit converter for multiphase work. Benchmarks use a refrigerant as the working fluid, with all properties in physical units. A flat interface, a still droplet, and flow between plates all behave as theory predicts. The method converges at second order to the derived equation. The authors call it ready for accurate, user-friendly multiphase simulation.

Catalog record

CREATORLuiz Eduardo Czelusniak, Tim Niklas Bingert, Mathias J. Krause, Stephan Simonis
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE45 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-15018

Browse the full library · About the Tower of Babel

Full text

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

On the consistency of pseudo-potential lattice Boltzmann methods Luiz Eduardo Czelusniak, Tim Niklas Bingert, Mathias J. Krause, Stephan Simonis Abstract We derive the partial differential equation (PDE) to which the pseudo-potential lattice Boltzmann method (P-LBM) converges under diffusive scaling, providing a rigorous basis for its consistency analysis. By establishing a direct link between the method’s parameters and physical properties—such as phase densities, interface thickness, and surface tension—we develop a framework that enables users to specify fluid properties directly in SI units, eliminating the need for empirical parameter tuning. This allows the simulation of problems with predefined physical properties, ensuring a direct and physically meaningful parametrization. The proposed approach is implemented in OpenLB, featuring a dedicated unit converter for multiphase problems. To validate the method, we perform benchmark tests—including planar interface, static droplet, Galilean invariance, and two-phase flow between parallel plates—using R134a as the working fluid, with all properties specified in physical units. The results demonstrate that the method achieves second-order convergence to the identified PDE, confirming its numerical consistency. These findings highlight the robustness and practicality of the P-LBM, paving the way for accurate and user-friendly simulations of complex multiphase systems with well-defined physical properties. Preprint: AIP/123-QED I Introduction The lattice Boltzmann method (LBM) has emerged as a promising tool for simulating multiphase systems [ 1 , 2 , 3 , 4 , 5 ] , primarily due to its high computational efficiency, particularly in parallel simulations using multicore CPUs [ 6 , 7 , 8 ] and GPUs [ 9 , 10 ] . Several multiphase approaches exist within the LBM framework, including the pseudo-potential [ 11 , 12 ] , free-energy [ 13 , 14 ] , phase-field [ 15 , 16 , 17 ] , and color-gradient methods [ 18 , 19 ] . Among these, the pseudo-potential method (P-LBM) is particularly useful for high-performance simulations [ 20 , 21 ] due to its simplicity and efficiency [ 22 ] . The interface naturally emerges from the interparticle forces acting on the fluid [ 23 ] , eliminating the need for interface capturing or tracking methods. However, many simulations using the P-LBM are conducted using lattice units without establishing a clear correspondence between the simulation and a real physical condition [ 24 ] . Baakeem, Bawazeer, and Mohamad [ 25 ] argue that the conversion or mapping of parameters between the macroscopic (or physical) scale and the mesoscopic (or lattice) scale is not a straightforward procedure for the P-LBM. Therefore, some authors have focused on the dimensional analysis of this method [ 25 , 26 , 24 , 27 , 28 ] . As a result, different strategies for mapping parameters between the macroscopic and mesoscopic scales were proposed. Despite the progress, some points still make P-LBM challenging to apply to real physical problems. First, the method relies on specifying the properties of the equation of state (EOS), whereas the phase-field method allows for the direct imposition of surface tension and interface thickness, making it more practical to use. Another difficulty mentioned by Czelusniak et al. [ 22 ] is that the method’s parameters, necessary to impose the desired surface tension and ensure thermodynamic consistency, must be found empirically through prior simulations such as the Young–Laplace test. To address these issues, we revisited the topic of P-LBM consistency. Through our investigation, we derived the form of the macroscopic equation to which the P-LBM converges under diffusive scaling. This understanding allowed us to directly relate the fluid’s physical properties—such as phase densities, interface thickness, and surface tension—to the P-LBM parameters. Consequently, one of the method’s major limitations was addressed: Users no longer need to fine-tune specific parameters, but can instead define desired fluid properties directly, greatly simplifying its application. To implement and validate these developments, we utilized OpenLB [ 29 ] based on version 1.7 [ 30 ] . This software is a powerful tool for parallel lattice Boltzmann simulations that is employed for example in sub-grid particulate flows [ 31 ] , fully resolved particle flows [ 32 , 8 ] , turbulence simulations [ 33 , 34 ] , optimization [ 35 , 36 ] , sub-grid multiphase flows [ 37 ] and fully resolved multiphase flows [ 38 ] . OpenLB employs unit converters, enabling users to specify physical variables directly in physical units that are then transformed into lattice units by the converter. We extended this framework by developing a dedicated converter for multiphase problems. This enhancement allows users to define fluid properties not only directly but also in physical units, making the method significantly more user-friendly and versatile. To validate our approach, we selected four benchmarks with analytical solutions for diffuse interface problems: planar interface, static droplet, Galilean invariance, and two-phase flow between parallel plates. All tests are performed in SI units using R134a as a working fluid. Each problem was solved using meshes with varying resolutions, and we demonstrated that the method achieves second-order convergence in all cases, confirming its consistency. The remainder of this paper is organized as follows. In Section II , we introduce the theoretical foundations of the pseudo-potential lattice Boltzmann method (P-LBM). Section III focuses on the derivation of the partial differential equation (PDE) approximated by the P-LBM under diffusive scaling. In Section IV , we present the methodology for directly relating the method parameters to physical properties. Section V describes the implementation of the proposed framework in OpenLB, including the development of a dedicated unit converter for multiphase problems. Section VI presents the results of the benchmark tests, validating the consistency of the method. Finally, Section VII concludes the paper with a summary of the main findings. II Pseudo-potential LBM In this work, we apply the lattice Boltzmann equation (LBE) with the single-relaxation-time (SRT or BGK) collision operator [ 39 ] . To establish the dimensions of each variable, we define the fundamental dimensions of mass ( M ), length ( L ), and time ( T ). We denote the dimension of a variable \phi as [\phi] . The LBE is expressed as: f_{i}(t+\Delta t,\bm{x}+\bm{c}_{i}\Delta t)-f_{i}(t,\bm{x})=-\frac{\Delta t}{\tau}(f_{i}-f_{i}^{eq})+\Delta tF_{i}^{\prime}, (1) where f_{i} ( [f_{i}]=ML^{-3} ) are the particle distribution functions associated with particle velocities \bm{c}_{i} , and f_{i}^{eq} are the local equilibrium distribution functions (EDF). The BGK collision operator is governed by the relaxation time \tau ( [\tau]=T ). The variables t and \bm{x} are the time and space coordinates, respectively. The velocity set used is the regular two-dimensional nine velocities scheme (D2Q9): \bm{c}_{i}=\begin{cases}(0,0),~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i=0,\\ (c,0),(0,c),(-c,0),(0,-c),~~~~~~i=1,...,4,\\ (c,c),(-c,c),(-c,-c),(c,-c),~i=5,...,8,\\ \end{cases} (2) where c is the lattice speed defined as c=\Delta x/\Delta t . The parameters \Delta x and \Delta t are the space and time steps. The specific form of f_{i}^{eq} ( [f_{i}^{eq}]=ML^{-3} ) varies depending on the particular method employed. In this work, we use the form: f_{i}^{eq}=w_{i}\rho\left(1+\frac{c_{i\alpha}}{c_{s}^{2}}u_{\alpha}+\frac{c_{i\alpha}c_{i\beta}-c_{s}^{2}\delta_{\alpha\beta}}{2c_{s}^{4}}u_{\alpha}u_{\beta},\right), (3) where w_{i} (dimensionless) are the weights associated with each velocity \bm{c}_{i} . For the D2Q9 scheme, these weights are given as w_{0}=4/9 , w_{1,2,3,4}=1/9 , and w_{5,6,7,8}=1/36 . The parameter c_{s} , referred to as the lattice sound speed, is defined as c/\sqrt{3} . The variable \rho represents the density, and u_{\alpha} denotes the fluid velocity. The term F_{i}^{\prime} on the right-hand side of Eq. ( 1 ) represents the

Read the complete text