This work presents the Lagrangian entropic lattice Boltzmann method (LELBM), a novel framework for supersonic lattice Boltzmann method that surpasses traditional restrictions on velocity, temperature, viscosity, and memory, resulting in stable and efficient Courant-free LBM formulation, while still maintaining explicit time stepping. A regularized central moment collision method coupled with a double-distribution function (DDF) that considers internal degrees of freedom enabled the precise control of shear viscosity, bulk viscosity, and thermal conductivity. Lagrangian acoustic stencils (LAS), a generalized multispeed shifted lattice construction method, is introduced to adapt to arbitrary local velocity and temperature. Additionally, entropic population reconstruction (EPR) has been used to obtain entropy-satisfying post-collision population while achieving both positivity and conservation of moments. Memory requirements are drastically reduced from a polynomial order of temperature to a constant by employing moment streaming (MS) algorithm. The unified LELBM framework is validated through comprehensive benchmark problems. In one dimension, the model is evaluated through Sod's shock tube, Lax problem, and Shu-Osher wave. In two dimensions, 2D Riemann problem, double Mach reflection, oblique shock, supersonic flow past a circular cylinder, and supersonic flow past a NACA0012 airfoil have been tested to assess the robustness and accuracy of the proposed algorithm.
Introduction
The lattice Boltzmann method (LBM), developed in the late 1980s[1], is based on lattice gas automata (LGA) from the 1970s[2, 3, 4]. Unlike LGA, LBM streams particle populations to simulate the discrete Boltzmann equation. The conventional lattice Boltzmann method is characterized by three processes: integration, collision, and streaming[5]. The LBM algorithm is simple and easy to implement, since all processes except streaming are local, which is very favorable for the highly parallelized computation, demanded for modern CFD solvers. However, several limitations hinder its practical application. These include stability issues in velocity, temperature, and viscosity[6, 7, 8]. Additionally, because traditional LBM stores all populations, its memory demand is significantly higher than conventional continuum methods[9].
The essence of the lattice Boltzmann method lies in the Bhatnagar-Gross-Krook (BGK) collision model[10], which simplified the nonlinear interaction of particles to a linear decay rate of non-equilibrium dynamics[11]. The most traditional collision model is the so-called single relaxation time model (SRT), which first constructs the equilibrium distribution using Hermite expansion[12] and conducts a collision based on the relaxation time τ to the raw population to adjust the viscosity. After the introduction of the multiple relaxation time model (MRT)[13, 14, 15], the collision process can be broadly classified into two categories: population-based and moment-based. The population-based method explicitly performs collision on the population, while enhancing the stability by either regularization[16, 17, 18, 19, 20] or entropic stabilization[21, 22, 23, 24, 25, 26]. The moment-based method conducts collision in moment space and conducts back-transformation to reconstruct the population that satisfies the post-collision moments, while also relaxing higher-order moments to equilibrium. Developed from the traditional raw moment method, the advanced method conducts collision in different moment space such as central moments[27, 28, 29, 30], or cumulants[31, 32, 33].
Considering that the mean velocity of the gas particles is proportional to the square root of temperature[34], higher-order velocity stencils are required to simulate high-temperature flow. The first multispeed lattice was introduced by Alexander et al.[35] to simulate thermohydrodynamic flows using the lattice Boltzmann method. Although originally developed for subsonic thermal flows, it was later extended to handle supersonic flows[36, 37, 38]. However, this approach consequently increased the number of velocity stencils even for the case of supersonic flow under low temperature. To address this issue, Sun[39, 40, 41, 42] proposed the shift of velocity stencils by bulk velocity and reconstructing stencils in a relative frame. But this method has long been forgotten until Frapolli[23] revisiting the shifted lattice for their supersonic entropic LBM framework.
The earliest energy models in the LBM involved the direct integration of second-order moments[43]. However, this approach was limited to monatomic gases and lacked the flexibility in adjusting key thermodynamic properties, such as thermal conductivity and specific heat. From the Chapman-Enskog expansion, it becomes clear that using a single collision time is insufficient to fully capture the fluid properties described by the Navier-Stokes-Fourier (NSF) equations, particularly because third-order moments are closely linked to heat flux[44, 45]. To address this limitation, He et al.[46] introduced the double-distribution function (DDF) model, which incorporates a second distribution function for internal energy to solve the energy equation alongside the kinetic equations. Still, this model required finite differencing to capture the effect of viscous heating. Developing from the internal energy model, Guo[47] proposed a modified approach about the total energy, enabling the accurate modeling of viscous coupling without finite differentiation. In parallel, Nie[48] introduced the DDF method for the internal degree of freedom, which is now the common approach for modern supersonic LBM models. Besides the complete LBM framework for the energy equation, hybrid LBM (HLBM)[20, 49, 50, 51] instead adopted the traditional finite-volume method for energy, while still using LBM for the mass and momentum equations.
Modern supersonic lattice Boltzmann methods can broadly be classified into four distinct approaches. While the hybrid recursive regularization LBM (HRR-LBM) solves the energy equation using a traditional continuum method, this method also attempts to improve the stability of LBM formulation through regularization process[20, 52, 53, 54]. While HRR-LBM constructs populations for cell-centered velocity stencils, semi-Lagrangian LBM (SLLBM) reconstructs a population on a compact, non-cell-conforming velocity stencils for each lattice, to enhance stability under compressible flow conditions[55, 56, 57, 58, 59]. As a more "physical" approach, the entropic LBM (ELBM)[21, 22, 23] dynamically adjusts the collision frequencies to achieve local maximum entropy in each cell in accordance with the Boltzmann H-theorem[60]. Besides the traditional Hermite polynomial-based population reconstruction approach, numerical equilibria LBM (NE-LBM) constructs populations through the application of Lagrange multipliers to satisfy the maximum entropy principle[61, 62, 63]. This approach results in an exponential-shaped distribution function that preserves both positivity and conservation of properties while also enhancing the numerical stability.
As an even more sophisticated approach, the particle on demand method combined with the discrete unified gas kinetic scheme (PonD/DUGKS) and entropic stabilization has recently been developed[64, 65]. The characteristic of PonD is its shift from an absolute to a relative reference frame, while the moments are also scaled to an optimal local temperature. Consequently, the populations are constructed through Hermite expansion in that shift-scaled frame, then mapped back to the absolute frame using Lagrange polynomials. DUGKS[66, 67, 68], in turn, employs a finite-volume formulation that explicitly computes population fluxes across cell interfaces, improving spatial resolution. By combining with entropic stabilization, the PonD/DUGKS framework proved its robustness under extreme conditions, such as Mach 30 astrophysical jets.
Despite numerous unique formulations proposed over the past decades to simulate supersonic flows using the lattice Boltzmann method, however, to the best of the author’s knowledge, no universal numerical scheme has been yet developed that incorporates all benefits of previous methods. This paper presents a novel formulation, the Lagrangian entropic lattice Boltzmann method (LELLBM), which unifies multiple approaches, including the philosophy of moment-matching of MRT model, the double distribution function (DDF), shifted multispeed lattice, and entropic formulations for stability. Additionally, this work adopted the moment expression of memory storage, inspired by the work of Valero-Lara[69], to reduce the memory demand bottleneck. The paper is structured into six main sections. Section II introduces the LELBM formulation, providing a detailed theoretical framework that bridges various techniques developed in previous research and provides algorithmic flowchart for better understanding in practical application. Section III introduces the boundary condition techniques applied for the evaluation of the present model. Section IV verifies the theory of the present framework, from fluid properties to computational cost. Section V evaluates both the accuracy and robustness of the present model on various canonical to extreme compressible flow problems. Finally, Section VI concludes the paper by summarizing the key findings and contributions, while also listing the required improvements for practical implementations.
Lagrangian Entropic Lattice Boltzmann Method
Theoretical Background
Lattice Boltzmann method is an automaton-based method that, rather than solving partial differential equations about macroscopic variables as in traditional continuum methods, solves discretized Boltzmann equation in a less "math-oriented" approach. The Boltzmann equation is written as below:
∂t∂f+ξα∂xα∂f+ρFα∂ξα∂f=Ω(f).(1)
Here, f is the continuous distribution function of the particle in 2D+1 dimensions, defined in both D dimensions of position and momentum, with an additional dimension from time. ξ denotes particle velocity vector, and F is the external forces. Ω is the collision kernel that results in macroscopic phenomena such as hydrodynamic stress or heat conduction[70]. With BGK collision model[10], we can write the collision kernel in a simplified form of
Ω(f)=−τ1(f−feq),(2)
where τ is the collision time that determines the decay rate of non-equilibrium contribution on distribution function. Here, feq is the equilibrium distribution, which is known to follow the Boltzmann distribution.
feq(ρ,ξ,u,RT)=(2πRT)D/2ρexp(−2RT∣ξ−u∣2).(3)
From the above distribution ρ is the macroscopic mass density of fluid in position of interest, ξ is the microscopic velocity of particles, u for the macroscopic bulk velocity, and RT is the temperature with gas constant. Now, if we conduct temporal discretization on Eq. (1) in a total derivative form and separating the equilibrium and non-equilibrium distributions from the raw distribution function f, we obtain
where fneq is the non-equilibrium distribution function which is defined as
fneq=f−feq.(5)
As explained in Appendix C, ω in Eq. (4) is the implicit collision frequency arising from the implicit time stepping. Since f is continuous both in position and momentum, the distribution is discretized according to the velocity stencils ci.
Equation (6) is the automaton algorithm in which the population fi defined by x and ci is landing at position x+ciΔt after Δt. Since an LBM algorithm in general streams the population to exactly the cell center in each time step, the streaming algorithm shows a lattice-like network. The algorithm of the traditional method can be simply expressed as Figure 1. See[5, 43] for details on LBM algorithm.
Figure 1. Traditional LBM algorithm. The numbering indicate the cascaded operation procedure and arrow direction indicates the transfer of memory.
In the fields of LBM, the distinction between the continuous distribution function f and the population (or the discrete distribution function) fi has long been ambiguous. However, their dimensional units of f and fi are [ρU−D] and [ρ], respectively, where [ρ] and [U] are the dimensions of density and velocity, and D is the number of translational DoF. Since LBM generally expects the population to be located at the lattice center, their relationship can be expressed as
fi(x,ci)=∭ΔCif(x,ξ)dξ,(7)
where ΔCi is the translational volume of speed for the specified ith population we are interested in, which is centered at ci.
ΔCi=α=1∏DΔciα.(8)
Thus, the population fi has an approximation relation with the distribution function f by
fi(x,ci)≈f(x,ci)ΔCi.(9)
Additionally, for the LELBM framework, the population itself follows the cell-averaging concept of the finite-volume method. The volume-integrated population ⟨fi⟩ is expressed as
⟨fi⟩(x,ci)=∭ΔV∭ΔCif(x,ξ)dξdV,(10)
or approximately
⟨fi⟩(x,ci)≈fi(x,ci)ΔV≈f(x,ci)ΔCiΔV.(11)
Here, ΔV denotes the physical volume of the cell at position x. The dimensional unit of ⟨fi⟩ is [ρLD], corresponding to the mass fraction distributed throughout the velocity space with the mean particle velocity of ci for each cell, within the control volume of ΔV. For notational consistency with the traditional LBM framework, ⟨fi⟩ is simply expressed as fi throughout the remainder of this paper. These dimensional formulations are unnecessary in the standard lattice unit setting, where ΔCi=1, Δx=1, ΔV=1, Δt=1, which is also adopted in this work. Although the rigorous definitions introduced in this section are not required for the current uniform lattice implementation, this background knowledge is necessary for understanding key concepts in supersonic LBM models.
Additionally, the definition of moment is required for the present framework. The raw moment of an arbitrary continuous distribution function f can be calculated as
Mαβγ=∭−∞∞ξxαξyβξzγfdξ.(12)
Defining ξxαξyβξzγ as the moment basis function TN(ξ), we can define arbitrary Nth-order raw moments in the form of
MN=∭−∞∞TN(ξ)fdξ.(13)
Central moments are evaluated in a similar approach, but with a shifted population velocity of ξ−u.
MN=∭−∞∞TN(ξ−u)fdξ.(14)
Moments of an arbitrary discrete distribution function can be obtained by taking a summation instead. TN is now the discrete moment basis function, which is in matrix form.
MN=i∑TN(ci)fi,(15)
MN=i∑TN(ci−u)fi.(16)
The dimensions of both continuous and discrete moments are [ρUN]. However, since LBM conducts all operations in volume-integrated form, the normalization and scaling processes are involved throughout the algorithm.
Regularized Central Moment Collision
With the introduction of the multiple relaxation time (MRT) model[13, 14, 15], moment-based collision generally follows a three-step process consisting of moment transformation, moment collision, and inverse transformation:
M=Tf,(17)
M∗=Meq+(I−Ω)Mneq,(18)
f∗=T−1M∗.(19)
Here, f and M are the vector of population and moments. T is the moment basis matrix, I is the identity matrix, and Ω is the matrix of collision frequencies. Traditionally, MRT models have been applied mainly to athermal flows with fixed DdQ3D lattices. This was due to the sporadic emergence of a negative post-collision population f∗ arising from direct inversion of the moment basis matrix T in higher-order lattice. As a result, simulations on multispeed lattices often led to nonphysical solutions.
However, as proved in Appendices B and C, assigning distinct collision frequencies to a different order of moments allows independent control of fluid properties. When combined with a positivity-preserving reconstruction scheme, moment-based collision models can be extended to simulate thermal flows with flexible adjustment of thermohydrodynamic properties. This section provides the procedure for the moment-based collision model tailored for thermal LBM. The corresponding reconstruction strategy is addressed in § 2 D.
Founded upon the concept of cascaded digital lattice automata[27], central moment method performs collisions on moments evaluated in a relative reference frame[29, 30]. This approach effectively decoupled the momentum and energy in their moments, and was found to enhance stability even in a high Reynolds number flow. The regularized central moment collision method extends the athermal framework of the central moment collision model to thermal flows.
References
[1]G. R. McNamara and G. Zanetti, “Use of the Boltzmann equation to simulate lattice-gas automata,” Physical Review Letters 61, 2332–2335 (1988).pubmed.ncbi.nlm.nih.gov/10039085
[2]J. Hardy, Y. Pomeau, and O. Pazzis, “Time evolution of a two-dimensional classical lattice system,” Physical Review Letters 31, 276–279 (1973).DOI
[3]U. Frisch, D. d’Humières, B. Hasslacher, P. Lallemand, Y. Pomeau, and J. Rivet, “Lattice gas hydrodynamics in two and three dimensions,” Complex Systems 1, 649–707 (1987).
[4]U. Frisch, B. Hasslacher, and Y. Pomeau, “Lattice-gas automata for the Navier-Stokes equation,” Physical Review Letters 56, 1505–1508 (1986).pubmed.ncbi.nlm.nih.gov/10032689
[5]T. Krüger, H. Kusumaatmaja, A. Kuzmin, O. Shardt, G. Silva, and E. M. Viggen, The Lattice Boltzmann Method (Springer, Cham, 2017).
[6]J. Chao, R. Mei, and W. Shyy, “Effect of variable viscosity on the accuracy of lattice Boltzmann method,” in APS Division of Fluid Dynamics Meeting Abstracts (2004) p. KL.009.
[7]J. D. Sterling and S. Chen, “Stability analysis of lattice Boltzmann methods,” Journal of Computational Physics 123, 196–206 (1996).arxiv.org/abs/comp-gas/9306001
[8]X. B. Nie, X. Shan, and H. Chen, “Galilean invariance of lattice Boltzmann models,” Europhysics Letters 81, 34005 (2008).DOI
[9]M. Matyka and M. Dziewonski, “Memory-efficient lattice Boltzmann method for low Reynolds number flows,” Computer Physics Communications 267, 108044 (2021).DOI
[10]P. L. Bhatnagar, E. P. Gross, and M. Krook, “A model for collision processes in gases. I. small amplitude processes in charged and neutral one-component systems,” Physical Review 94, 511–525 (1954).
[11]F. J. Higuera and J. Jiménez, “Boltzmann approach to lattice gas simulations,” Europhysics Letters 9, 663 (1989).DOI
[12]X. Shan and X. He, “Discretization of the velocity space in the solution of the Boltzmann equation,” Physical Review Letters 80, 65–68 (1998).DOI
[13]P. Lallemand and L. Luo, “Theory of the lattice Boltzmann method: Dispersion, dissipation, isotropy, Galilean invariance, and stability,” Physical Review E 61, 6546–6562 (2000).
[14]D. D’Humières, I. Ginzburg, M. Krafczyk, P. Lallemand, and L. Luo, “Multiple-relaxation-time lattice Boltzmann models in three dimensions,” Philosophical Transactions. Series A, Mathematical, Physical, and Engineering Sciences 360, 437–451 (2002).pubmed.ncbi.nlm.nih.gov/16214687
[15]P. Asinari, “Multiple-relaxation-time lattice Boltzmann scheme for homogeneous mixture flows with external force,” Physical Review E 77, 56706 (2008).pubmed.ncbi.nlm.nih.gov/18643194
[16]J. Latt and B. Chopard, “Lattice Boltzmann method with regularized pre-collision distribution functions,” Mathematics and Computers in Simulation 72, 165–168 (2006).DOI
[17]O. Malaspinas, “Increasing stability and accuracy of the lattice Boltzmann scheme: Recursivity and regularization,” arXiv preprint arXiv:1505.06900 (2015).arxiv.org/abs/1505.06900
[18]C. Coreixas, G. Wissocq, G. Puigt, J.-F. Boussuges, and P. Sagaut, “Recursive regularization step for high-order lattice Boltzmann methods,” Physical Review E 96, 33306 (2017).DOI
[19]J. Jacob, O. Malaspinas, and P. Sagaut, “A new hybrid recursive regularised Bhatnagar–Gross–Krook collision model for Lattice Boltzmann method-based large eddy simulation,” Journal of Turbulence 19, 1–26 (2018).DOI
[20]Y. Feng, P. Boivin, J. Jacob, and P. Sagaut, “Hybrid recursive regularized thermal lattice Boltzmann model for high subsonic compressible flows,” Physics 146, 282–300 (1998).
[21]N. Frapolli, S. Chikatamarla, and I. Karlin, “Entropic lattice Boltzmann model for compressible flows,” Physical Review E 92, 061301(R) (2015).DOI
[22]N. Frapolli, S. Chikatamarla, and I. Karlin, “Entropic lattice Boltzmann model for gas dynamics: Theory, boundary conditions, and implementation,” Physical Review E 93, 063302 (2016).DOI
[23]N. Frapolli, Entropic lattice Boltzmann models for thermal and compressible flows, Ph.D. thesis, ETH Zürich (2017).
[24]N. Frapolli, S. Chikatamarla, and I. Karlin, “Theory, analysis, and applications of the entropic lattice Boltzmann model for compressible flows,” Entropy 22, 370 (2020).
[25]S. A. Hosseini, M. Atif, S. Ansumali, and I. Karlin, “Entropic lattice Boltzmann methods: A review,” Computers & Fluids 259, 105884 (2023).
[26]S. A. Hosseini, P. Boivin, D. Thévenin, and I. Karlin, “Lattice Boltzmann methods for combustion applications,” Progress in Energy and Combustion Science 102, 101140 (2024).
[27]M. C. Geier, Ab Initio Derivation of the Cascaded Lattice Boltzmann Automaton, Ph.D. thesis, University of Freiburg (2006), available from Freidok: https://freidok.uni-freiburg.de/data/2860.
[28]M. Geier, A. Greiner, and J. G. Korvink, “Cascaded digital lattice Boltzmann automata for high Reynolds number flow,” Physical Review E 73, 66705 (2006).
[29]K. N. Premnath and S. Banerjee, “On the three-dimensional central moment lattice Boltzmann method,” Journal of Statistical Physics 143, 747–794 (2011).
[30]A. D. Rosis, R. Huang, and C. Correixas, “Universal formulation of central-moments-based lattice Boltzmann method with external forcing for the simulation of multiphysics phenomena,” Physics of Fluids 31, 117102 (2019).
[31]M. Geier, M. Schönherr, A. Pasquali, and M. Krafczyk, “The cumulant lattice Boltzmann equation in three dimensions: Theory and validation,” Computers & Mathematics with Applications 70, 507–547 (2015).
[32]M. Geier, A. Pasquali, and M. Schönherr, “Parametrization of the cumulant lattice Boltzmann method for fourth order accurate diffusion part I: Derivation and validation,” Journal of Computational Physics 348, 862–888 (2017).
[33]M. Geier, A. Pasquali, and M. Schönherr, “Parametrization of the cumulant lattice Boltzmann method for fourth order accurate diffusion part II: Application to flow around a sphere at drag crisis,” Journal of Computational Physics 348, 889–898 (2017).
[34]S. Salinas, Introduction to Statistical Physics (Springer, New York, 2001).DOI
[35]F. J. Alexander, H. Chen, S. Chen, and G. D. Doolen, “Lattice Boltzmann model for compressible fluids,” Physical Review A 46, 1967–1970 (1992).pubmed.ncbi.nlm.nih.gov/9908331
[36]X. Shan, X. Yuan, and H. Chen, “Kinetic theory representation of hydrodynamics: a way beyond the Navier–Stokes equation,” Journal of Fluid Mechanics 550, 413–441 (2006).
[37]S. S. Chikatamarla and I. V. Karlin, “Lattices for the lattice Boltzmann method,” Physical Review E 79, 46701 (2009).pubmed.ncbi.nlm.nih.gov/19518374
[38]F. Chen, A. Xu, G. Zhang, Y. Li, and S. Succi, “Multiple-relaxation-time lattice Boltzmann approach to compressible flows with flexible specific-heat ratio and Prandtl number,” Europhys. Lett. 90, 54003 (2010).
[39]C. Sun, “Lattice-Boltzmann models for high speed flows,” Physical Review E 58, 7283–7287 (1998).DOI
[40]C. Sun, “Adaptive lattice Boltzmann model for compressible flows: Viscous and conductive properties,” Physical Review E 61, 2645–2653 (2000).
[41]C. Sun and A. T. Hsu, “Three-dimensional lattice Boltzmann model for compressible flows,” Physical Review E 68, 16303 (2003).pubmed.ncbi.nlm.nih.gov/12935242
[42]C. Sun and A. Hsu, “Multi-level lattice Boltzmann model on square lattice for compressible flows,” Computers & Fluids 33, 1363–1385 (2004).
[43]S. Succi, The Lattice Boltzmann Equation: For Complex States of Flowing Matter (Oxford University Press, Oxford, 2018).
[44]M. Teixeira, Continuum Limit of Lattice Gas Fluid Dynamics, Ph.D. thesis, Massachusetts Institute of Technology (1992).
[45]G. R. McNamara, A. L. Garcia, and B. J. Alder, “Stabilization of thermal lattice Boltzmann models,” Journal of Statistical Physics 81, 395–408 (1995).DOI
[46]X. He, S. Chen, and G. Doolen, “A novel thermal model for the lattice Boltzmann method in incompressible limit,” Journal of Computational Physics 146, 282–300 (1998).DOI
[47]Z. Guo, C. Zheng, B. Shi, and T. Zhao, “Thermal lattice Boltzmann equation for low Mach number flows: Decoupling model,” Physical review. E, Statistical, nonlinear, and soft matter physics 75, 036704 (2007).pubmed.ncbi.nlm.nih.gov/17500823
[48]X. Nie, X. Shan, and H. Chen, “Thermal lattice Boltzmann model for gases with internal degrees of freedom,” Physical review. E, Statistical, nonlinear, and soft matter physics 77, 035701 (2008).
[49]A. Mezhrab, M. Bouzidi, and P. Lallemand, “Hybrid lattice-Boltzmann finite-difference simulation of convective flows,” Computers & Fluids 33, 623–641 (2004).
[50]Z. Li and Y. Zhang, “Hybrid lattice Boltzmann and finite volume method for natural convection,” Journal of Thermophysics and Heat Transfer 28, 68–77 (2014).DOI
[51]A. Nee, “Hybrid lattice Boltzmann—Finite difference formulation for combined heat transfer problems by 3D natural convection and surface thermal radiation,” International Journal of Mechanical Sciences 173, 105447 (2020).
[52]S. Guo, Y. Feng, and P. Sagaut, “Improved standard thermal lattice Boltzmann model with hybrid recursive regularization for compressible laminar and turbulent flows,” Physics of Fluids 32, 126108 (2020).
[53]F. Renard, Y. Feng, J.-F. Boussuge, and P. Sagaut, “Improved compressible hybrid lattice Boltzmann method on standard lattice for subsonic and supersonic flows,” Computers & Fluids 219, 104867 (2021).DOI
[54]S. Guo and Y. Feng, “Hybrid compressible lattice Boltzmann method for supersonic flows with strong discontinuities,” Physics of Fluids 36, 076129 (2024).
[55]A. Krämer, K. Küllmer, D. Reith, W. Joppich, and H. Foysi, “Semi-Lagrangian off-lattice Boltzmann method for weakly compressible flows,” Physical Review E 95, 023305 (2017).DOI
[56]D. Wilde, A. Krämer, D. Reith, and H. Foysi, “Semi-Lagrangian lattice Boltzmann method for compressible flows,” Physical Review E 101, 53306 (2020).DOI
[57]D. Wilde, A. Krämer, M. Bedrunka, D. Reith, and H. Foysi, “Cubature rules for weakly and fully compressible off-lattice Boltzmann methods,” Journal of Computational Science 51, 101355 (2021).
[58]D. Wilde, A. Krämer, D. Reith, and H. Foysi, “High-order semi-Lagrangian kinetic scheme for compressible turbulence,” Physical Review E 104, 25301 (2021).DOI
[59]P. Spelten, D. Wilde, M. C. Bedrunka, D. Reith, and H. Foysi, “Supersonic shear and wall-bounded flows with body-fitted meshes using the semi-Lagrangian lattice Boltzmann method: Boundary schemes and applications,” (2024), arXiv:2412.09051 [physics.flu-dyn].
[60]L. Boltzmann, “Further Studies on the Thermal Equilibrium of Gas Molecules,” in The Kinetic Theory of Gases (Imperial College Press, 2003) pp. 262–349.
[61]J. Latt, C. Coreixas, J. Bény, and A. Parmigiani, “Efficient supersonic flow simulations using lattice Boltzmann methods based on numerical equilibria,” Philos. Trans. R. Soc. A 378, 20190559 (2020).
[62]C. Coreixas and J. Latt, “Compressible lattice Boltzmann methods with adaptive velocity stencils: An interpolation-free formulation,” Physics of Fluids 32, 116102 (2020).
[63]K. Thyagarajan, C. Coreixas, and J. Latt, “Exponential distribution functions for positivity-preserving lattice Boltzmann schemes: Application to 2D compressible flow simulations,” Physics of Fluids 35, 126115 (2023).
[64]N. G. Kallikounis, B. Dorschner, and I. V. Karlin, “Particles on demand for flows with strong discontinuities,” Physical Review E 106, 10.1103/physreve.106.015301.DOI
[65]N. G. Kallikounis and I. V. Karlin, “Particles on demand method: Theoretical analysis, simplification techniques, and model extensions,” Physical Review E 109, 015304 (2024).
[66]Z. Guo, K. Xu, and R. Wang, “Discrete unified gas kinetic scheme for all Knudsen-number flows: low-speed isothermal case,” Physical Review E 88, 033305 (2013).
[67]Z. Guo, R. Wang, and K. Xu, “Discrete unified gas kinetic scheme for all Knudsen-number flows. II. Thermal compressible case,” Physical Review E 91, 033313 (2015).
[68]Z. Guo and K. Xu, “Progress of discrete unified gas-kinetic scheme for multiscale flows,” Advances in Aerodynamics 3, 6 (2021).DOI
[69]P. Valero-Lara, “Reducing memory requirements for large-size LBM simulations on GPUs,” Concurrency and Computation: Practice and Experi ence 29, e4221 (2017).
[70]H. Grad, “On the kinetic theory of rarefied gases,” Communications on Pure and Applied Mathematics 2, 331–407 (1949).DOI
[71]M. Watari and M. Tsutahara, “Supersonic flow simulations by a three-dimensional multispeed thermal model of the finite difference lattice Boltzmann method,” Physica A: Statistical Mechanics and its Applications 364, 129–144 (2006).DOI
[72]N. Frapolli, S. S. Chikatamarla, and I. V. Karlin, “Multispeed entropic lattice Boltzmann model for thermal flows,” Physical Review E 90, 43306 (2014).pubmed.ncbi.nlm.nih.gov/25375622
[73]E. W. Grafrarend, Linear and Nonlinear Models: Fixed Effects, Random Effects, and Mixed Models (Walter de Gruyter, Berlin, 2006).
[74]R. A. Brownlee, A. N. Gorban, and J. Levesley, “Stability and stabilization of the lattice Boltzmann method,” Physical Review E 75, 36711 (2007).pubmed.ncbi.nlm.nih.gov/17500830
[75]I. V. Karlin, A. N. Gorban, S. Succi, and V. Boffi, “Maximum entropy principle for lattice kinetic equations,” Physical Review Letters 81, 6–9 (1998).
[76]I. V. Karlin and S. Succi, “Equilibria for discrete kinetic equations,” Physical Review E 58, R4053–R4056 (1998).DOI
[77]S. Ansumali and I. V. Karlin, “Stabilization of the lattice Boltzmann method by the H theorem: A numerical test,” Physical Review E 62, 7999–8003 (2000).pubmed.ncbi.nlm.nih.gov/11138084
[78]S. Ansumali and I. V. Karlin, “Entropy function approach to the lattice Boltzmann method,” Journal of Statistical Physics 107, 291–308 (2002).DOI
[79]M. Boghosian, P. J. Love, P. V. Coveney, I. V. Karlin, S. Succi, and J. Yepez, “Galilean-invariant lattice-Boltzmann models with H theorem,” Physical Review E 68, 25103 (2003).DOI
[80]J. E. Broadwell, “Study of rarefied shear flow by the discrete velocity method,” Journal of Fluid Mechanics 19, 401–414 (1964).DOI
[81]L. Mieussens, “Convergence of a discrete-velocity model for the Boltzmann-BGK equation,” Computers & Mathematics with Applications 41, 83–96 (2001).DOI
[82]L. Mieussens, “Discrete velocity model and implicit scheme for the BGK equation of rarefied gas dynamics,” Mathematical Models and Methods in Applied Sciences 10, 1121–1149 (2000).DOI
[83]B. Dubroca and L. Mieussens, “A conservative and entropic discrete-velocity model for rarefied polyatomic gases,” ESAIM: Proc. 10, 127–139 (2001).
[84]P. L. Tallec and J.-P. Perlat, “Numerical Analysis of Levermore’s Moment System,” Research Report RR-3124 (INRIA, Projet M3N, 1997).
[85]P. Charrier, B. Dubroca, L. Mieussens, and R. Turpault, “Discrete-velocity models for numerical simulations in transitional regime for rarefied flows and radiative transfer,” in Transport in Transition Regimes, edited by N. B. Abdallah, I. M. Gamba, C. Ringhofer, A. Arnold, R. T. Glassley, P. Degond, and C. D. Levermore (Springer, New York, 2004) pp. 85–101.
[86]C. Baranger, J. Claudel, N. Hérouard, and L. Mieussens, “Locally refined discrete velocity grids for stationary rarefied flow simulations,” Journal of Computational Physics 257, 572–593 (2014).DOI
[87]S. Pressé, K. Ghosh, J. Lee, and K. A. Dill, “Principles of maximum entropy and maximum caliber in statistical physics,” Reviews of Modern Physics 85, 1115–1141 (2013).DOI
[88]B. Župančič and D. Kuić, “Relation between Boltzmann and Gibbs entropy and example with multinomial distribution,” Journal of Physics Communications 2, 045002 (2018).
[89]J. A. Shohat and J. D. Tamarkin, The problem of moments, Mathematical Surveys and Monographs (American Mathematical Society, Providence, RI, 1943).DOI
[90]N. I. Akhiezer, The classical moment problem and some related questions in analysis (Hafner Publishing Company, New York, 1965).
[91]M. G. Kreĭn and A. A. Nudel’man, The Markov moment problem and extremal problems, Translations of Mathematical Monographs, Vol. 50 (American Mathematical Society, Providence, RI, 1977).
[92]Mead, R. Lawrence, and N. Papanicolaou, “Maximum entropy in the problem of moments,” Journal of Mathematical Physics 25, 2404–2417 (1984).DOI
[93]M. J. Borwein and A. S. Lewis, “Convergence of best entropy estimates,” SIAM Journal on Optimization 1, 191–205 (1991).DOI
[94]C. E. Shannon, “A mathematical theory of communication,” The Bell System Technical Journal 27, 379–423 (1948).
[95]C. E. Shannon, “A mathematical theory of communication,” The Bell System Technical Journal 27, 623–656 (1948).
[96]G. H. Golub and C. F. Van Loan, Matrix Computations, 4th ed., Johns Hopkins Studies in the Mathematical Sciences (Johns Hopkins University Press, Baltimore, MD, 2013).
[97]P. J. Dellar, “Incompressible limits of lattice Boltzmann equations using multiple relaxation times,” Journal of Computational Physics 190, 351–370 (2003).DOI
[98]S. Ansumali, I. V. Karlin, and H. C. Öttinger, “Minimal entropic kinetic models for hydrodynamics,” Europhysics Letters 63, 798 (2003).DOI
[99]A. J. C. Ladd, “Numerical simulations of particulate suspensions via a discretized Boltzmann equation. Part 1. Theoretical foundation,” Journal of Fluid Mechanics 271, 285–309 (1994).DOI
[100]A. J. C. Ladd, “Numerical simulations of particulate suspensions via a discretized Boltzmann equation. Part 2. Numerical results,” Journal of Fluid Mechanics 271, 311–339 (1994).DOI
[101]A. J. C. Ladd and R. Verberg, “Lattice-Boltzmann simulations of particle-fluid suspensions,” Journal of Statistical Physics 104, 1191–1251 (2001).DOI
[102]A. W. Wagner and I. Pagonabarraga, “Lees–Edwards boundary conditions for lattice Boltzmann,” Journal of Statistical Physics 107, 521–537 (2002).DOI
[103]A. W. Lees and S. F. Edwards, “The computer study of transport processes under extreme conditions,” Journal of Physics C: Solid State Physics 5, 1921 (1972).DOI
[104]S. Chen, D. Martínez, and R. Mei, “On boundary conditions in lattice Boltzmann methods,” Physics of Fluids 8, 2527–2536 (1996).DOI
[105]R. Mei, L. Luo, and W. Shyy, “An accurate curved boundary treatment in the lattice Boltzmann method,” Journal of Computational Physics 155, 307–330 (1999).
[106]M. E. Bouzidi, M. Firdaouss, and P. Lallemand, “Momentum transfer of a Boltzmann-lattice fluid with boundaries,” Physics of Fluids 13, 3452–3459 (2001).DOI
[107]P. Lallemand and L. Luo, “Lattice Boltzmann method for moving boundaries,” Journal of Computational Physics 184, 406–421 (2003).DOI
[108]B. Dorschner, S. S. Chikatamarla, F. Bösch, and I. V. Karlin, “Grad’s approximation for moving and stationary walls in entropic lattice Boltzmann simulations,” Journal of Computational Physics 295, 340–354 (2015).DOI
[109]B. Dorschner, S. S. Chikatamarla, and I. V. Karlin, “Kinetic multirelaxation-time lattice Boltzmann method for moving and deforming geometries in three dimensions,” Physical Review E 95, 063306 (2017).DOI
[110]S. K. P. Sanjeevi, A. Zarghami, and J. T. Padding, “Choice of no-slip curved boundary condition for lattice Boltzmann simulations of high-Reynolds-number flows,” Physical Review E 97, 043305 (2018).DOI
[111]D. R. Noble, S. Chen, J. G. Georgiadis, and R. O. Buckius, “A consistent hydrodynamic boundary condition for the lattice Boltzmann method,” Physics of Fluids 7, 203–209 (1995).DOI
[112]T. Inamuro, M. Yoshino, and F. Ogino, “A non-slip boundary condition for lattice Boltzmann simulations,” Physics of Fluids 7, 2928–2930 (1995).DOI
[113]Q. Zou and X. He, “On pressure and velocity boundary conditions for the lattice Boltzmann BGK model,” Physics of Fluids 9, 1591–1598 (1997).
[114]Z. Guo, C. Zheng, and B. Shi, “Non-equilibrium extrapolation method for velocity and pressure boundary conditions in the lattice Boltzmann method,” Chinese Physics 11, 366 (2002).
[115]O. Malaspinas, B. Chopard, and J. Latt, “General regularized boundary condition for multi-speed lattice Boltzmann models,” Computers & Fluids 49, 29–35 (2011).DOI
[116]L. Xu, E. Serre, and P. Sagaut, “A theoretical analysis of mass leakage at boundaries within the lattice Boltzmann method,” Physics of Fluids 34, 065113 (2022).DOI
[117]H. C. Lee, S. Bawazeer, and A. A. Mohammad, “Boundary conditions for lattice Boltzmann method with multispeed lattices,” Computers & Fluids 162, 152–159 (2018).DOI
[118]F. Klass, A. Gabbana, and A. Bartel, “A non-equilibrium bounce-back boundary condition for thermal multispeed LBM,” Journal of Computational Science 53, 101364 (2021).DOI
[119]L. S. G. Kovásznay, “Turbulence in supersonic flow,” Journal of the Aeronautical Sciences 20, 657–674 (1953).DOI
[120]E. M. Viggen, “The lattice Boltzmann method in acoustics,” in Proceedings of the 39th Scandinavian Symposium on Physical Acoustics (Geilo, Norway, 2010).
[121]E. M. Viggen, The Lattice Boltzmann Method: Fundamentals and Acoustics, Ph.D. thesis, Norwegian University of Science and Technology (NTNU) (2014).
[122]L. E. Kinsler, A. R. Frey, A. B. Coppens, and J. V. Sanders, Fundamentals of Acoustics (Wiley, 2000).
[123]G. A. Sod, “A survey of several finite difference methods for systems of nonlinear hyperbolic conservation laws,” Journal of Computational Physics 27, 1–31 (1978).DOI
[124]P. D. Lax, “Weak solutions of nonlinear hyperbolic equations and their numerical computation,” Communications on Pure and Applied Mathematics 7, 159–193 (1954).DOI
[125]C. Shu and S. Osher, “Efficient implementation of essentially non-oscillatory shock-capturing schemes. II,” Journal of Computational Physics 83, 32–78 (1989).
[127]P. D. Lax and X. Liu, “Solution of two-dimensional Riemann problems of gas dynamics by positive schemes,” SIAM Journal on Scientific Computing 19, 319–340 (1998).DOI
[128]A. Kurganov and E. Tadmor, “Solution of two-dimensional Riemann problems for gas dynamics without Riemann problem solvers,” Numerical Methods for Partial Differential Equations 18, 584–608 (2002).DOI
[129]P. Woodward and P. Colella, “The numerical simulation of two-dimensional fluid flow with strong shocks,” Journal of Computational Physics 54, 115–173 (1984).DOI
[130]U. S. Vevek, B. Zang, and T. H. New, “On alternative setups of the double Mach reflection problem,” Journal of Scientific Computing 78, 1291–1303 (2019).DOI
[131]F. Zeng, Y. Shen, S. Liu, and L. Liu, “A high-performance fifth-order multistep WENO scheme,” International Journal for Numerical Methods in Fluids 91, 159–182 (2019).DOI
[132]J. E. A. John and T. G. Keith, Gas Dynamics (Pearson Prentice Hall, Harlow, UK, 2006).
[133]R. Qiu, Y. You, C. Zhu, and R. Chen, “Lattice Boltzmann simulation for high-speed compressible viscous flows with a boundary layer,” Applied Mathematical Modelling 48, 567–583 (2017).DOI
[134]M. Alperin, A Study of Detached Shock Waves in Two Dimensions, Ph.d. thesis, California Institute of Technology, Pasadena, CA (1950).DOI
[135]G. E. Kaattari, “Predicted Shock Envelopes about Two Types of Vehicles at Large Angles of Attack,” NASA Technical Note NASA-TN-D-860 (A-491) (NASA, Washington, DC, 1961).
[136]C. Kim, “Experimental studies of supersonic flow past a circular cylinder,” Journal of the Physical Society of Japan 11, 439–445 (1956).DOI
[137]J. Sinclair and X. Cui, “A theoretical approximation of the shock standoff distance for supersonic flows around a circular cylinder,” Physics of Fluids 29, 026102 (2017).DOI
[138]S. B. Q. Tran, F. Y. Leong, Q. T. Le, and D. V. Le, “Lattice Boltzmann method for high-Reynolds-number compressible flow,” Computers & Fluids 249, 105701 (2022).DOI
[139]M. de Tullio, P. D. Palma, G. Iaccarino, G. Pascazio, and M. Napolitano, “Immersed boundary technique for compressible flow simulations on semi-structured grids,” in Numerical Methods for Fluid Dynamics, Lecture Notes in Computational Science and Engineering (Springer, 2009) pp. 365–370.DOI
[140]C. Xu and Z. Ni, “Novel characteristics of wavy cylinder in supersonic turbulent flow,” European Journal of Mechanics – B/Fluids 67, 158–167 (2018).DOI
[141]M. Hafez and E. Wahba, “Simulations of viscous transonic flows over lifting airfoils and wings,” Computers & Fluids 36, 39–52 (2007).
[142]Y. Thornibert and D. Lagrava and O. Malaspinas and B. Chopard and C. Coreixas and J. de Santana Neto and R. Deiterding and J. Latt, “Local mesh refinement sensor for the lattice Boltzmann method,” Journal of Computational Science 64, 101864 (2022).
[144]S. Ubertini, P. Asinari, and S. Succi, “Three ways to lattice Boltzmann: A unified time-marching picture,” Physical Review E 81, 016311 (2010).pubmed.ncbi.nlm.nih.gov/20365464
[145]F. J. Higuera, “Lattice gas method based on the Chapman–Enskog expansion,” Physics of Fluids A: Fluid Dynamics 2, 1049–1051 (1990).DOI