Towards nonlinear thermohydrodynamic simulations via the Onsager-Regularized Lattice Boltzmann Method

Anirudh Jonnalagadda, Amit Agrawal, Atul Sharma, Walter Rocchia, Sauro Succi · 2025

Read Towards nonlinear thermohydrodynamic simulations via the Onsager-Regularized Lattice Boltzmann Method by Anirudh Jonnalagadda, Amit Agrawal, Atul Sharma, Walter Rocchia, Sauro Succi (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-2501-16000.txt.

Introduction

Anirudh Jonnalagadda and co-authors strengthen a fluid simulation method in a 2025 paper. Lattice Boltzmann methods simulate fluids on grids, but the simplest nine-cell grid distorts flow because it lacks directional balance. A regularized variant promises to fix those errors without bolt-on corrections.

The team gives a general theory of the method for two standard equilibrium choices, proving the recovered flow equations are far more accurate than the basic collision model. On the standard grid the scheme self-adjusts its viscosity to cancel the imbalance. Numerical tests on a decaying swirl and a shock tube confirm the gains. The work grounds a generic framework for nonlinear heat-flow simulation.

Catalog record

CREATORAnirudh Jonnalagadda, Amit Agrawal, Atul Sharma, Walter Rocchia, Sauro Succi
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE39 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-16000

Browse the full library · About the Tower of Babel

Full text

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

Towards nonlinear thermohydrodynamic simulations via the Onsager-Regularized Lattice Boltzmann Method Anirudh Jonnalagadda, Amit Agrawal, Atul Sharma, Walter Rocchia, Sauro Succi Abstract This work presents a generalized, assumption-free, and stencil-independent theoretical analyses of the recently proposed Onsager-Regularized (OReg) lattice Boltzmann (LB) method [Jonnalagadda et al. , Phys. Rev. E 104, 015313 (2021) ] and demonstrates its ability to mitigate spurious errors associated with the insufficient isotropy of standard first-neighbor lattices without the inclusion of any external correction terms. The hydrodynamic limit recovered by the OReg scheme is derived for two equilibrium distribution functions, namely the so-called thermal guided equilibrium and the popular second order polynomial equilibrium, to show that the OReg scheme yields macroscopic dynamics that are 𝒪 ⁡ ( u ) \mathcal{O}(u) times more accurate than that of the bare BGK collision model. Specifically, we show that, with the guided equilibrium on the D2Q9 standard lattice, the OReg scheme inherently compensates for the insufficient lattice isotropy of the standard D2Q9 lattice by automatically adjusting the lattice viscosity, yielding 𝒪 ⁡ ( u 4 ) \mathcal{O}(u^{4}) and 𝒪 ⁡ ( u 2 ) \mathcal{O}(u^{2}) accurate kinetic models when operated at the reference and arbitrary temperatures respectively. Further, we also show that the OReg scheme presents an 𝒪 ⁡ ( u 3 ) \mathcal{O}(u^{3}) accurate kinetic model for the D2Q9 lattice when used with the second order polynomial equilibrium formulation. Thereafter, the accuracy of the OReg-guided-equilibrium kinetic model is numerically demonstrated for quasi-one-dimensional simulations of the rotated decaying shear wave and isothermal shocktube problems. The present work lays the theoretical foundation of a generic framework which can enable fully local, correction-free, nonlinear thermohydrodynamic LB simulations on standard lattices, thereby facilitating scalable simulations of physically challenging fluid flows. I Introduction The evolution of a Lattice Boltzmann (LB) system [ 1 , 2 , 3 , 4 , 5 , 6 , 7 ] emulates the Boltzmann transport equation in D -dimensions through microscopic operations on discrete populations, f_{i} , in a lattice-discretized velocity space having N velocities c_{i_{\alpha}} , i\in\left\{1,\cdots,N\right\} . In the single relaxation framework, the lattice-BGK update is defined as: \displaystyle f_{i}(x_{\alpha}+c_{i_{\alpha}}\Delta t, \displaystyle t+\Delta t)-f_{i}(x_{\alpha},t)=-\frac{1}{\tau}f_{i}^{(neq)}(x_{\alpha},t) \displaystyle=-\frac{1}{\tau}\left(f_{i}(x_{\alpha},t)-f_{i}^{(eq)}(x_{\alpha},t)\right), (1) where, \tau is the relaxation time, and f_{i}^{(eq)} , f_{i}^{(neq)}=\Big(f_{i}-f_{i}^{(eq)}\Big) are the equilibrium and non-equilibrium contributions of f_{i} respectively. In such a discrete framework, exactly recovering Navier-Stokes-Fourier (NSF) macrodynamics places explicit constraints on both the equilibrium and non-equilibrium contributions of f_{i} . Specifically, the discrete equilibrium is required to recover the following Maxwell-Boltzmann (MB) moments: \Big\langle f_{i}^{(eq)},\big\{1,c_{i_{\alpha}},c_{i}^{2}/2\big\}\Big\rangle=\left\{\rho,\rho u_{\alpha},E\right\} (2a) \displaystyle\Pi_{\alpha\beta}^{(eq)}=\Pi_{\alpha\beta}^{\text{MB}} \displaystyle=p\delta_{\alpha\beta}+\rho u_{\alpha}u_{\beta} (2b) \displaystyle\vskip-8.00003ptq_{\alpha}^{(eq)}=q_{\alpha}^{\text{MB}} \displaystyle=u_{\alpha}(E+p) (2c) where \left\langle\varphi_{i},\psi_{i}\right\rangle=\sum\limits_{i}\varphi_{i}\psi_{i} , the quantities \rho , \rho u_{\alpha} and E=\Pi_{\alpha\alpha}^{(eq)}/2=(Dp+\rho u^{2})/{2} are the mass, momentum, and energy density respectively, with p=\rho\theta being the thermodynamic pressure given through the ideal gas law, where \theta = RT is the reduced temperature, and \Pi_{\alpha\beta}^{(eq)}=\left\langle f_{i}^{(eq)},c_{i_{\alpha}}c_{i_{\beta}}\right\rangle and q_{\alpha}^{(eq)}=\left\langle f_{i}^{(eq)},\frac{1}{2}c_{i}^{2}c_{i_{\alpha}}\right\rangle are the equilibrium contributions of the pressure tensor and heat flux vector. Note that Eq. ( 2 ) corresponds to 8 and 13 linearly independent constraints on f_{i}^{(eq)} in 2- and 3-dimensions. Thus, one can completely describe the equilibrium state even on standard lattices ( Figure 1 ) lying within the first Brillouin zone. In contrast, on standard lattices, the description of the non-equilibrium populations, f_{i}^{(neq)} , and consequently the resulting out-of-equilibrium macrodynamics, requires more care. We recall from kinetic theory that f_{i}^{(neq)} must satisfy the so-called compatibility conditions (Eq. 3a ) and, should recover the NSF constitutive relations for the viscous stress tensor (Eq. ( 3b )) and heat flux vector (Eq. 3c )); \Big\langle f_{i}^{(neq)},\big\{1,c_{i_{\alpha}},c_{i}^{2}/2\big\}\Big\rangle=0, (3a) \displaystyle\vskip-10.00002pt\hskip-7.42493pt\Pi_{\alpha\beta}^{(neq)}=\Pi_{\alpha\beta}^{\text{NSF}} \displaystyle=-\mu\bigg(\partial_{\alpha}u_{\beta}+\partial_{\beta}u_{\alpha}-\frac{2}{D}\partial_{\chi}u_{\chi}\delta_{\alpha\beta}\bigg) (3b) \displaystyle q_{\alpha}^{(neq)}=q_{\alpha}^{\text{NSF}}=-\kappa\partial_{\alpha}\theta. (3c) The quantities \mu and \kappa are the dynamic viscosity and thermal conductivity respectively. (1, 1) (-1, -1) (a) D2Q9 (1, -1, 1) (-1, -1, 1) (1, -1, -1) (1, 1, -1) (-1, 1, -1) (b) D3Q27 Figure 1: Velocity space representations in two- and three-dimensions using the standard D2Q9 and D3Q27 lattices respectively. In order to evaluate Eq. ( 3a ), closed form expressions of f_{i}^{(neq)} are needed; while newer regularization schemes prescribe analytical expressions [ 8 ] , historically, f_{i}^{(neq)} is simply computed as f_{i}^{{(neq)}}=\Big(f_{i}-f_{i}^{(eq)}\Big) . However, since f_{i} is a computationally evaluated quantity, such a representation is not amenable for a theoretical evaluation of Eq. ( 3a ). Consequently, the following approach is adopted for enabling theoretical analyses: Firstly, the compatibility conditions given by Eq. ( 3a ) are assumed to hold. However, kinetic theory requires these conditions to be imposed on non-equilibrium descriptions of the distribution function [ 9 ] and, as evidenced by theoretical works on higher-order continuum transport equations [ 10 , 11 , 12 , 13 ] , such an assumption may not be valid even in the continuous phase space. Secondly, Eqs. ( 3b ) and ( 3c ) are evaluated by coupling \Pi_{\alpha\beta}^{(neq)}=\left\langle f_{i}^{(neq)},c_{i_{\alpha}}c_{i_{\beta}}\right\rangle and q_{\alpha}^{(neq)}=\left\langle f_{i}^{(neq)},\frac{1}{2}c_{i}^{2}c_{i_{\alpha}}\right\rangle to f_{i}^{(eq)} via higher-order equilibrium moments Q_{\alpha\beta\gamma}^{(eq)}=\left\langle f_{i}^{(eq)},c_{i_{\alpha}}c_{i_{\beta}}c_{i_{\gamma}}\right\rangle and R_{\alpha\beta\gamma\mu}^{(eq)}=\left\langle f_{i}^{(eq)},c_{i_{\alpha}}c_{i_{\beta}}c_{i_{\gamma}}c_{i_{\mu}}\right\rangle . However, this coupling to higher order equilibrium moments has severe repercussions in that spurious numerical errors are introduced on standard lattices due to insufficient lattice isotropy. In order to retain the computational advantage provided by LB on standard lattices, significant efforts have been made over the past three decades to eliminate these spurious errors [ 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 ] . The current state-of-the-art incorporates non-local correction terms into the lattice update to exactly recover the desired thermohydrodynamics [ 21 , 22 , 25 ] . However, it is noteworthy that this strategy is not only hard to generalize to different collision kernels and equilibrium representations, but is also detrimental to the parallel efficiency offered by LB on standard lattices. Furthermore, this performance sacrifice is exacerbated in numerical implementations that incorporate advanced meshing paradigms such as, e.g., grid refinement, needed for addressing complex problems involving fluid flow. In this work, we demonstrate that the recently proposed Onsager-Regularized (OReg) LB method [ 27 , 28 , 8 ] facilitates simulations of nonlinear NSF hydro-thermal transport phenomenon by alleviating the adverse standard lattice anisotropy effects without resorting to non-local corrections. Specifically, we first present a generalized, assumption-free, and stencil-independent theoretical analysis of the OReg scheme via the Chapman-Enskog (CE) procedure [ 5 , 6 ] that is valid for any equilibrium representation. We then employ the so-called \mathcal{O}(u^{4}) guided equilibrium [ 22 , 23 , 24 ] and the \mathcal{O}(u^{2}) polynomial equilibrium representation to obtain OReg kinetic models for the NSF equations on standard lattices. Thereafter, we present numerical results for canonical quasi-one-dimensional problems demonstrating the improved accuracy obtained with the OReg model with the guided equilibrium, and, finally, present our conclusions. II Onsager-Regularized LBM The OReg scheme uses the principles of linear irreversible thermodynamics and prescribes the non-equilibrium contribution of the distribution function at the NSF level in terms of viscous and thermal irreversible processes [ 29 , 30 ] that comply with Onsager’s Symmetry Principle. Here, for the sake of brevity, we restrict the discussion to isothermal flows of monatomic gases and direct the interested reader to Ref. [ 28 ] and the references therein for a more detailed exposition. In such a setting, the non-equilibrium thermodynamics representation of

Read the complete text