A Modified BGK Collision Operator for Exact Conservation in Numerical Solutions of Boltzmann-BGK Fusion Energy Division, Oak Ridge National Laboratory, Oak Ridge, TN, USA

Vienna B. Rossmanith · 2025

Read A Modified BGK Collision Operator for Exact Conservation in Numerical Solutions of Boltzmann-BGK Fusion Energy Division, Oak Ridge National Laboratory, Oak Ridge, TN, USA by Vienna B. Rossmanith (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-2511-20664.txt.

Introduction

Vienna B. Rossmanith, of Oak Ridge National Laboratory, develops in 2025 a gas-simulation method that keeps physics honest. The method solves the Boltzmann-BGK equation, which tracks a distribution of particle speeds and positions instead of following every molecule. Many solvers quietly leak mass, momentum, or energy, especially when the speed range is cut off. Rossmanith set out to remove those leaks entirely.

Her scheme splits the work into a transport part, handled by a third-order Lax-Wendroff method, and a collision part handled by a second-order TR-BDF method. The new ingredient multiplies the target Maxwell-Boltzmann distribution by a quadratic Hermite polynomial, a correction curve whose coefficients are picked so the totals come out exact. On test runs, mass, momentum, and energy were conserved up to machine precision, the smallest error a computer can represent. The method ships as Java code with Python plotting tools.

Catalog record

CREATORVienna B. Rossmanith
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE33 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2511-20664

Browse the full library · About the Tower of Babel

Full text

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

A Modified BGK Collision Operator for Exact Conservation in Numerical Solutions of Boltzmann-BGK Fusion Energy Division, Oak Ridge National Laboratory, Oak Ridge, TN, USA Vienna B. Rossmanith Abstract Ideal gases can be modeled by the Boltzmann equation from statistical physics. Instead of trying to track the position and velocity of a large number of gas molecules, it is possible to describe the particles with a particle distribution function. The Boltzmann equation provides the rule for evolving the distribution function over time, allowing one to simulate the gas dynamics. In this work, we develop a novel numerical method for solving the 1D1V Boltzmann-BGK equation. Several important ingredients are combined to create an accurate, efficient, robust numerical method valid in all flow regimes. First, we make use of operator splitting to create separate transport and collision sub-steps, each of which is easier to discretize than the whole system; second, we introduce a third-order accurate Lax-Wendroff-type scheme for the transport sub-step; third, we make use of the second-order L-stable TR-BDF method for the collision sub-step. The final component we introduce is a novel treatment of the collision sub-step to guarantee that the total mass, momentum, and energy are conserved even in the presence of a truncated velocity range and quadrature errors in computing moments. The key idea is to multiply the Maxwell-Boltzmann distribution in the collision sub-step by a quadratic Hermite polynomial, where the coefficients in the polynomial are chosen to ensure exact conservation. The resulting scheme is verified on piecewise constant initial data with periodic boundary conditions; exact conservation up to machine precision is demonstrated for mass, momentum, and energy. The resulting method is implemented in a freely available Java code with Python plotting routines. \newsiamremark remarkRemark \newsiamremark hypothesisHypothesis \newsiamthm claimClaim \newsiamremark factFact \dedication Project advisor: Preeti Sar 1 Introduction A gas is a collection of unbound molecules; in the absence of external forces, each molecule moves with some velocity until it collides with another molecule. The collisions act to drive the gas towards a thermodynamic equilibrium. Mathematically, one could consider a range of models, including (1) molecular dynamics, which is valid at very small scales, where each molecule in some small portion of the gas is tracked; (2) kinetic Boltzmann equations, which is valid at intermediate scales, where a particle distribution function replaces individual molecules; and (3) fluid models such the Navier-Stokes equation, which are valid at large scales, where macroscopic quantities such as mass, momentum, and energy are tracked (e.g., see [ 12 ] ). The focus of this work is the kinetic Boltzmann equation; in particular, we consider the 1D1V Boltzmann-BGK equation [ 2 ] , in which the full Boltzmann collision operator is replaced by a far simpler collision operator that nonetheless captures key dynamics. While there exist several approaches to solving the Boltzmann-BGK equation, we consider in this work so-called direct kinetic methods, rather than stochastic methods such as Direct Simulation Monte-Carlo (DSMC) (e.g., see [ 10 ] ). In a direct kinetic approach, various numerical discretizations can be applied on a computational mesh that covers a space/velocity domain that is truncated in velocity. The goal of this work is to develop a finite difference method for the one-dimensional Boltzmann-BGK model of gas dynamics that is similar in spirit to existing approaches (e.g., Pieraccini and Puppo [ 11 ] , Li et al. [ 8 ] , and Sar [ 14 ] ), but with the additional feature that mass, momentum, and energy are exactly conserved. Several important ingredients are combined to create an accurate, efficient, robust numerical method valid in all flow regimes. First, we make use of operator splitting to create separate transport and collision sub-steps, each of which is easier to discretize than the whole system; second, we introduce a third-order accurate Lax-Wendroff-type scheme for the transport sub-step; third, we make use of the second-order L-stable TR-BDF method for the collision sub-step. The final component we introduce is a novel treatment of the collision sub-step to guarantee that the total mass, momentum, and energy are conserved even in a truncated velocity range and quadrature errors in computing moments. The key idea is to multiply the Maxwell-Boltzmann distribution in the collision sub-step by a quadratic Hermite polynomial, where the coefficients in the polynomial are chosen to ensure exact conservation. The remainder of this paper is organized as follows. In § 2 we briefly describe the 1D1V Boltzmann-BGK equation. In § 3 we develop the main components of the proposed numerical method using operator splitting and finite differences. The scheme described in § 3 does not exactly conserve mass, momentum, and energy. Therefore, in § 4 we develop a modified collision operator that is able to recover exact conservation; in this same section, we present the final proposed numerical method in two short algorithms. A numerical example illustrating the benefits of the proposed modification is provided in 5 . We conclude in §sec:conclusion. 2 Boltzmann-BGK A particle distribution function (PDF) of time, position, and velocity can be used to describe the movement of molecules in a gas. In a one-dimensional system, the PDF is a function of the following form: f(t,x,v):\mathbb{R}_{\geq 0}\times\mathbb{R}\times\mathbb{R}\mapsto\mathbb{R}_{\geq 0}, (1) where t\in\mathbb{R}_{\geq 0} is time, x\in\mathbb{R} is the space coordinate, and v\in\mathbb{R} is the velocity coordinate. The total number of particles that can be found in some spatial region x\in\left[x_{1},x_{2}\right] with some range of velocities v\in\left[v_{1},v_{2}\right] can be computed by integrating the PDF over these values: \text{total number of particles in}\,\left[x_{1},x_{2}\right]\times\left[v_{1},v_{2}\right]=\int_{x_{1}}^{x_{2}}\int_{v_{1}}^{v_{2}}f\left(t,x,v\right)\,dv\,dx. (2) According to gas kinetic theory, the PDF satisfies the Boltzmann equation, which in one dimension can be written as follows: \frac{\partial f}{\partial t}+v\,\frac{\partial f}{\partial x}=\frac{1}{\varepsilon}C(f), (3) which consists of two key pieces: (1) the motion of particles (i.e., gas molecules) according to their velocity, v , and the collision of particles as modeled by the collision operator, C(f) . The parameter \varepsilon>0 is called the Knudsen number and is defined as the ratio of the mean-free path (i.e., how far a particle travels on average before undergoing a collision) to an important characteristic length scale in the problem (i.e., size of the domain). Small \varepsilon>0 means high collisionality, while large \varepsilon>0 means low collisionality. The net effect of the collision operator on the dynamics is that it drives the gas towards a thermodynamic equilibrium , which in one dimension is given by the following distribution: {\mathcal{M}}\left(t,x,v\right)=\frac{\rho(t,x)}{\sqrt{2\pi\,T(t,x)}}\exp\left[-\frac{\left(v-u(t,x)\right)^{2}}{2T(t,x)}\right]; (4) this distribution is referred to as the Maxwell-Boltzmann distribution and depends on three macroscopic (or fluid) variables: Mass density: \displaystyle\quad\rho(t,x)=\int_{-\infty}^{\infty}f(t,x,v)\,dv=\int_{-\infty}^{\infty}{\mathcal{M}}(t,x,v)\,dv, (5) Fluid velocity: \displaystyle\quad u(t,x)=\frac{1}{\rho(t,x)}\int_{-\infty}^{\infty}vf(t,x,v)\,dv=\frac{1}{\rho(t,x)}\int_{-\infty}^{\infty}v{\mathcal{M}}(t,x,v)\,dv, (6) \displaystyle\begin{split}\text{Temperature:}&\quad T(t,x)=\frac{1}{\rho(t,x)}\int_{-\infty}^{\infty}{\bigl(v-u(t,x)\bigr)^{2}}f(t,x,v)\,dv\\ &\hskip 42.67912pt=\frac{1}{\rho(t,x)}\int_{-\infty}^{\infty}{\bigl(v-u(t,x)\bigr)^{2}}{\mathcal{M}}(t,x,v)\,dv.\end{split} (7) These variables are also related to two other important quantities: Momentum: \displaystyle\quad m(t,x)=\int_{-\infty}^{\infty}vf(t,x,v)\,dv=\int_{-\infty}^{\infty}v{\mathcal{M}}(t,x,v)\,dv, (8) Energy: \displaystyle\quad{\mathcal{E}}(t,x)=\int_{-\infty}^{\infty}v^{2}\,f(t,x,v)\,dv=\int_{-\infty}^{\infty}v^{2}\,{\mathcal{M}}(t,x,v)\,dv, (9) where m=\rho u\qquad\text{and}\qquad{\mathcal{E}}=\rho\left(u^{2}+T\right). (10) One difficulty in solving equation ( 3 ) is that the collision operator, C(f) , can be quite complicated and involve computational expensive integration over phase space. To reduce this complexity, a standard approach is to replace the full Boltzmann operator by something simpler; a commonly used model is the Bhatnagar, Gross, and Krook (BGK) [ 2 ] approximation, which mimics the net effect of the Boltzmann collision operator, namely driving the distribution function, f , towards its thermodynamic equilibrium: ( 4 ). The Boltzmann-BGK equation can be written as follows:

Read the complete text