An immersed boundary method for the discrete velocity model of the Boltzmann equation
Longqing Ge, Qingdong Cai, Yonghao Zhang, Tianbai Xiao · 2025
Read An immersed boundary method for the discrete velocity model of the Boltzmann equation by Longqing Ge, Qingdong Cai, Yonghao Zhang, Tianbai Xiao (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-2512-20252.txt.
Introduction
Longqing Ge, Qingdong Cai, Yonghao Zhang, and Tianbai Xiao simulate gas flows around solid bodies without fitting the grid to the body's shape. Their tool is an immersed boundary method built on the discrete velocity model of the Boltzmann equation. A Maxwell model of gas-surface interaction shapes the particle distributions in ghost cells near the surface.
Two improvements anchor the work. A compact upwind-weighted interpolation keeps the scheme stable for awkward geometries without wide stencils or normal projections. A cut-cell correction in velocity space repairs quadrature errors caused by surface discontinuities. Analysis confirms second-order accuracy and stability in both spaces. Numerical tests match body-fitted solvers while keeping the simplicity of a Cartesian grid.
Catalog record
| CREATOR | Longqing Ge, Qingdong Cai, Yonghao Zhang, Tianbai Xiao |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 78 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-20252
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
An immersed boundary method for the discrete velocity model of the Boltzmann equation Longqing Ge, Qingdong Cai, Yonghao Zhang, Tianbai Xiao Abstract Computational modeling and simulation of fluid-structure interactions constitute a fundamental cornerstone for advancing aerospace engineering endeavors. This paper addresses the notion and implementation of the immersed boundary method for the discrete velocity model of the Boltzmann equation. The method incorporates the Maxwell gas–surface interaction model into the construction of ghost-cell particle distribution functions, facilitating meticulous characterization of velocity slip and temperature jump effects within a Cartesian grid framework, which ultimately achieves accurate prediction of aerodynamic parameters. This study presents two principal advancements. First, an upwind-weighted compact interpolation strategy is developed in physical space, which ensures numerical stability and robustness for arbitrary geometries without relying on large stencils or normal-direction projections. Second, a cut-cell correction methodology is proposed in velocity space to address the degradation of quadrature accuracy caused by surface discontinuities. The resulting framework is equally applicable to both two- and three-dimensional problems without requiring any dimension-specific modifications. Rigorous analysis is provided to prove that the approach maintains second-order accuracy across both physical and velocity space, while ensuring robust numerical stability. Comprehensive numerical experiments demonstrate that the solution algorithm achieves the designed accuracy and delivers precise predictions comparable to body-conformal solvers, while retaining the simplicity, flexibility, and scalability of the Cartesian grid method. The proposed approach provides a unified and physically consistent immersed boundary framework for simulating dynamic interactions between non-equilibrium flows and structural components across a wide range of flow regimes. The open-source codes to reproduce the numerical results are available under the MIT license 1 1 1 https://github.com/CFDML/IBDVM-Paper . Table 1 : Nomenclature. Symbol Description IBM immersed boundary method DVM discrete velocity method t time variable \bm{x} space variable \bm{\xi} velocity variable (with its magnitude denoted as \xi ) \bm{c} peculiar velocity (with its magnitude denoted as c ) f particle distribution function \mathcal{Q} collision operator of kinetic equation g Maxwellian distribution function g^{S} Shakhov equilibrium distribution function \tau relaxation time \mu dynamic viscosity p pressure k Boltzmann constant \rho density n number density \mathbf{U} macroscopic velocity T temperature E energy \psi collision invariants \bm{W} macroscopic conservative variables (\rho,\rho\mathbf{U},\rho E) \bm{q} heat flux N_{V} number of collocation points \omega quadrature weight \phi arbitrary integrable function of \bm{\xi} \mathbf{\Omega} computational domain V volume of domain \mathbf{F}^{f} numerical flux for distribution function S area of cell face \bm{n} outward normal vector of surface \Theta Heaviside step function \bm{\sigma} spatial slope of distribution function f_{\mathrm{w}}^{*} surface distribution function considering wall effects f_{\mathrm{w}}^{i} incident particle distribution f_{\mathrm{w}}^{r} reflected particle distribution \alpha accommodation coefficient g_{\mathrm{w}} surface Maxwellian distribution \rho_{\mathrm{w}} density in g_{\mathrm{w}} \bm{U}_{\mathrm{w}} preconfigured wall velocity T_{\mathrm{w}} preconfigured wall temperature \xi_{n} projected particle velocity normal to the boundary gc, dc, ac ghost cell, donor cell, auxiliary cell N_{S} number of surface cells c.c. cut cell \mathcal{C} set of the cut cells’ indices \mathcal{G} subset of \mathcal{C} with \xi_{n}\leq 0 \mathcal{S} subset of \mathcal{C} with \xi_{n}>0 \lambda particle mean free path R_{\mathrm{w}} boundary curvature radius \theta proportion of particle velocity exhibiting discontinuity \left[f\right] jump of distribution function E^{*} quadrature error h,b reduced distribution functions H,B reduced Maxwellian distributions H^{S},B^{S} reduced equilibrium distributions L characteristic length \gamma specific heat ratio C dimensionless sound speed C_{p} pressure coefficient 1 Introduction Gaseous flows demonstrate distinct behavioral patterns across different regimes. The Knudsen number, defined as the ratio of molecular mean free path to a characteristic length scale, serves as a fundamental parameter for distinguishing flow regimes based on the degree of dilution. Continuum mechanics, e.g., the Euler and Navier-Stokes equations, is rigorously applicable solely under the condition that the Knudsen number is asymptotically small. With the advancement of modern industries (e.g., aerospace technology and advanced chip packaging) towards extreme conditions, rarefied gas dynamics has become increasingly critical, with applications ranging from very-low-earth-orbit flight (when the mean free path is considerably large) llop2014very to micro-electromechanical system (when the scale of interest becomes small) gad2005mems . Various numerical methods, e.g., stochastic particle methods bird2013dsmc and deterministic solvers for the Boltzmann equation broadwell_study_1964 , have been developed to simulate the evolution of rarefied gases. While numerical methods employing body-conformal meshes remain predominant in flow simulations, the Cartesian grid methodology has been attracting growing attention. In this approach, the computational domain is discretized uniformly using structured grids aligned with Cartesian coordinates. By avoiding cumbersome (re)meshing for specific geometries, the simulation pipeline can be significantly simplified and automated. The regular, structured grids allow straightforward indexing, thus directly supporting higher-order stencil computations peng2009high , optimized memory access, and load balancing. The method is naturally compatible with convergence acceleration techniques, e.g., the multi-grid method trottenberg2001multigrid and adaptive mesh refinement plewa_adaptive_2005 . Among others, accurate computational modeling of the interaction between gases and immersed objects is a prerequisite for the faithful prediction of flow dynamics and aerodynamic parameters by Cartesian grid methods. Several approaches have been proposed to model the fluid-structure interaction using the Cartesian grid method. The cut-cell method embeds a curved boundary into the Cartesian mesh by trimming the cell that intersects the geometry into an irregular fluid cell, and computing its reduced volume, partial face areas, and centroids ingram2003developments . The governing equations are then applied in these new cells, and the boundary conditions can be directly enforced on the newly created cut edges. The technique is non-intrusive to the solution algorithm and thus easier to implement. It has been applied to particle methods (represented by the direct simulation Monte Carlo) zhang2012robust , jin2017new , plimpton2019direct and discrete velocity methods of the Boltzmann equation dechriste_cartesian_2016 in rarefied gas dynamics. The main challenges of the cut-cell method arise from the tiny, irregular cells created at the boundary. They enforce a severely restrictive time step (often requiring intricate cell-merging seo_sharp-interface_2011 or flux-redistribution schneiders2013accurate stabilization), demand elaborate pre-processing (calculating geometric parameters for newly created cells), and break the uniform Cartesian stencils near the wall (resulting in lower-order or one-sided interpolations). The immersed boundary method (IBM) provides an alternative model for solving fluid-structure coupling. It employs a fixed Cartesian mesh throughout the solution process, thereby avoiding the additional computational complexity caused by the grid cutting. Technically, this approach can be broadly divided into two categories. The first category incorporates the notion of introducing a continuous force term within the governing equations to emulate the influence of boundary effects. Since Peskin’s original method peskin_flow_1972 , a series of such methods have been proposed to solve continuum flows, typically represented by (but not limited to) the direct-forcing method uhlmann2005immersed , su2007immersed , yang2009smoothing and penalization method arquis1984conditions , angot1999penalization , kevlahan2001computation . This type of method is also known as the diffused interface method, as the immersed interface can be smeared across multiple meshes due to the use of the discrete delta function or mask function. The second category of methods turns to implicitly model boundary effects by modifying discretization stencils in the solution algorithm, thus eliminating this generally undesirable feature. Representative efforts include the immersed interface method leveque1994immersed , li2001immersed and the ghost-cell-based method mohd1997simulations , ye1999accurate , fadlun2000combined , gibou2002second , ghias2007sharp , han2025navier . A significant challenge inherent to Cartesian grid-based IBM lies in the fact that the number of grids increases at