A Method for Analytical Solutions in the Lattice Boltzmann Method

Jordan Larson, Alexander J. Wagner · 2025

Read A Method for Analytical Solutions in the Lattice Boltzmann Method by Jordan Larson, Alexander J. Wagner (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-2505-24170.txt.

Introduction

Jordan Larson and Alexander Wagner find exact solutions for a popular fluid simulation method in this 2025 paper. The lattice Boltzmann method mimics fluids on a grid, and exact answers help researchers probe how the method behaves and where its boundaries hold. Their trick starts from a known flow shape and works backward to the solution.

They test the method on Couette flow, a simple shear flow, both aligned with the grid and tilted to it. With one standard setup the exact solution works in both positions. With an entropy-based setup it works when aligned, within practical settings. Tilted flow breaks that setup's momentum balance, a flaw the authors expose and discuss.

Catalog record

CREATORJordan Larson, Alexander J. Wagner
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE52 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-24170

Browse the full library · About the Tower of Babel

Full text

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

A Method for Analytical Solutions in the Lattice Boltzmann Method Jordan Larson, Alexander J. Wagner Abstract Analytical solutions to the lattice Boltzmann Equation make it possible to study the method itself, explore the properties of its collision operator, and identify implementations of boundary conditions. In this paper, we propose a method to find analytical solutions where the macroscopic flow profile is known. We test this method on bulk Couette flow aligned and inclined to the simulation lattice with the quadratic and entropic equilibrium distributions. Our method indeed provides an analytical solution to these flows when using the quadratic distribution. When the flow is aligned to the lattice, our method provides an analytical solution using the entropic distribution for practical relaxation times and shear rates. We show that a small even order truncation of the formal solution is optimal for accuracy-compute-time trade-off. In the inclined case, our method does not conserve momentum, by a small relative error, when using the entropic distribution. We also discover that entropic lattice Boltzmann method is not compatible with the angled Couette flow. We discuss the application of our method to more complicated flows. I Introduction Lattice Boltzmann methods have emerged as a powerful tool to simulate fluids. These methods were originally based on averaging lattice gas methods and therefore unconditionally stable Higuera and Succi 1989 . Lattice Boltzmann methods became even more versatile by breaking the link to lattice gases by allowing for arbitrary local equilibrium distributions Higuera and Jiménez 1989 ; Qian et al. 1992 . One advantage of this method is the relative ease with which complex geometries can be implemented, although sub-lattice resolution for boundary conditions remain an area of current research Ginzburg et al. 2023 . We define an analytical solution to be a solution of both the lattice Boltzmann Equation and the Navier-Stokes Equations. It is highly desirable to find such solutions to known flow profiles. Such knowledge would prove the lattice Boltzmann method recovers the analytical solutions directly or would guide development on novel lattice Boltzmann methods that can. Analytical solutions have previously been derived for some specific flow profiles. Zou et al. Zou et al. 1995 were the first to derive analytical solutions for Poiseuille and Couette flows aligned to the lattice using a quadratic equilibrium distribution and BGK collision operator. We show that our solution applied to the Couette flow recovers their result (and we identify a misprint in Zou et al. 1995 ). Ginzburg et al. Ginzburg et al. 2023 analyzed Couette flow (particularly in the more complex version including a normal injection leading to an exponential velocity profile) and found analytical solutions for an inclined Couette flow in a two relaxation time scheme. This analysis shows the utility of analytical solutions to guide the development of better boundary conditions. The derived solutions, however, are only exact for some special “magic values” for the relaxation times and the specific equilibrium distribution considered. Our approach is somewhat different. Here we present a formal analytical solutions for an arbitrary equilibrium distribution but only for a single relaxation time scheme, and only for flows for which the flow solution is know. This is similar in spirit to the work of Házi Házi 2003 . They presented a method to check the consistency of the lattice Boltzmann equation with a known analytical flow description, but only for full relaxation to local equilibrium. The basis for this idea is that local density and velocity completely define the local equilibrium distribution. Somewhat surprisingly, Házi did not recover the known analytical solutions but instead used his results to analyze presumed error terms. These discrepancies were due to the fact that he used non-standard forcing terms for the lattice Boltzmann method, leading to the errors he observed. In contrast, our analysis is valid for arbitrary relaxation times. Ansumali et al. Ansumali et al. 2007 also highlighted the importance of analytical solutions by considering a time and space continuous lattice Boltzmann formulation. They found analytical solutions for non-inclined Couette flows in arbitrary Knudsen numbers. They showed that a next nearest neighbor approximation was insufficient to fully recover boundary conditions at high Knudsen numbers. Johnson et al. Johnson et al. 2024 found an analytical solution for the bounded Couette flow using special boundary conditions of Maxwell-type. They compared their solution to molecular dynamics simulations and found excellent agreement. Johnson et al. 2024 exhibits the utility of analytical solution to validate lattice Boltzmann methods. Although the reviewed literature used the popular quadratic equilibrium distribution to derive their analytical solutions, the entropic equilibrium distribution is another important choice. This is derived by minimizing an H-functional Ansumali et al. 2003 and also results from the lattice Boltzmann limit of an integer lattice Blommel and Wagner 2018 . Entropic lattice Boltzmann has multiple advantages, as illustrated by Frapolli et al. Frapolli et al. 2020 , such as its ability to handle higher mach-number flows and its greater stability. Thus, in addition to considering the standard quadratic equilibrium distribution, we extend the analytical solutions to entropic lattice Boltzmann. We identify a formal analytical solution of the lattice Boltzmann method when the macroscopic flow fields are known. We then examine this formal solution to determine its consistency with the lattice Boltzmann method. The first application of the new approach is on a Couette flow that can be rotated with respect to the lattice by an arbitrary angle. In this case, we show that the analytical solution for a quadratic equilibrium distribution is consistent with the lattice Boltzmann Equation. However, the solution for the entropic distribution is more complicated. Machine-accuracy agreement is observed for practical relaxation times and shear rates when the Couette flow is aligned with the lattice. When the flow is inclined, we observe deviations between the analytical flow profile and the entropic lattice Boltzmann method. This paper observes the following outline. First, we introduce the lattice Boltzmann method. Second, we explain the formal the formal analytical solution. Third, we introduce the aligned Couette flow, and we apply the formal analytical solution to this problem for both the polynomial and entropic local equilibrium distributions. Fourth, we extend the analysis to the inclined Couette flow. Finally, we conclude by discussing potential applications of this approach to more complex flow situations. II Lattice Boltzmann Method In the lattice Boltzmann method, probability densities f_{i}(\mathbf{x},t) , i\in I for some finite index set I , are assigned to each lattice node \mathbf{x} at time t which have velocity vector \mathbf{v}_{i} . We require that \mathbf{x}+\mathbf{v}_{i}\Delta t to be another node. When lattice Boltzmann methods are derived from underlying lattice gas methods, the f_{i} represent the expectation value for particle occupation numbers for particles moving from lattice cell x-v_{i} at time t-\Delta t to lattice site x at time t . The f_{i}(x,t) densities evolve according to the lattice Boltzmann Equation et al. 2017 : f_{i}(\mathbf{x}+\mathbf{v}_{i}\Delta t,t+\Delta t)=f_{i}(\mathbf{x},t)+\Omega_{i}, (1) where \Omega_{i} is the collision operator. The simplest collision operator in the lattice Boltzmann method, proposed by Qian et al. 1992 , is \Omega_{i}=\frac{\Delta t}{\tau}\left[f^{\text{eq}}_{i}(\rho,\mathbf{u})-f_{i}(\mathbf{x},t)\right]. (2) Collisions bring the f_{i} densities closer to a local equilibrium distribution f_{i}^{\text{eq}} , which only depends on the locally conserved \rho and \mathbf{u} . Relaxation time \tau controls the rate of

Read the complete text