Kinetic lattice Boltzmann modeling of rarefied gas flow and heat transfer in the transition regime ($\mathrm{Kn}\gtrsim0.1$): Knudsen-layer resolution, parameter-independent kinetic boundary conditions, and tractable discrete-velocity sets

A peer-reviewed Comment by Li-Shi Luo (Phys. Rev. E 84, 048301 (2011)) on the thermal lattice Boltzmann (LB) simulations of Ghazanfarian and Abbassi (Phys. Rev. E 82, 026307 (2010)) identified a coherent open methodological problem: standard LB schemes with lattice-tied discrete velocities and simple slip closures are not genuine Boltzmann-level kinetic solvers, so their apparent capture of rarefied microchannel flow and heat transfer up to $\mathrm{Kn}\sim1$ can be mesh- and relaxation-parameter artifacts. The open problem is to construct an LB formulation for coupled rarefied flow and heat transfer through the transition regime up to $\mathrm{Kn}\sim1$ that jointly (i) resolves the Knudsen layer against Boltzmann or DSMC reference data, (ii) imposes velocity-slip and temperature-jump conditions at physically fixed wall locations with accommodation dependence and without mesh or relaxation-parameter dependence beyond the slip-flow regime where the multiple-relaxation-time 'magic parameter' already works, and (iii) uses discrete-velocity sets recovering the required higher-order tensorial moments at practical computational cost.

Background

The lattice Boltzmann method (LBM) is a mesoscopic fluid solver that evolves discrete populations $f_i$ streaming on a lattice with a small fixed set of discrete velocities $\mathbf{c}_i$ and relaxing toward local equilibrium (the lattice Boltzmann equation, LBE). For rarefied gases the standard collision operator is the BGK model (a single-relaxation-time approximation of the Boltzmann collision term, named after Bhatnagar, Gross and Krook) or the multiple-relaxation-time (MRT) model, in which different moment combinations of the populations relax at independently chosen rates. The Knudsen number $\mathrm{Kn}=\lambda/H$ — the ratio of the molecular mean free path $\lambda$ to the channel width $H$ — classifies gas-flow regimes: in slip flow ($\mathrm{Kn}\lesssim0.1$) the Navier–Stokes equations with velocity-slip and temperature-jump wall conditions are adequate, while in the transition regime ($0.1\lesssim\mathrm{Kn}\lesssim10$) kinetic theory is required. Within about one mean free path of a wall the distribution function is strongly out of equilibrium; this near-wall region is the Knudsen layer, and resolving it is the signature difficulty of transition-regime modeling. Direct simulation Monte Carlo (DSMC), which simulates molecular motion and collisions statistically, and numerical solution of the Boltzmann equation itself serve as reference data in this regime.

Gas flow and heat transfer in micro- and nanochannels (cooling of microelectromechanical systems, MEMS, and shale-gas transport) motivated several attempts to extend the LBM into these regimes. Ghazanfarian and Abbassi (2010) simulated shear-, pressure-, and mixed-driven rarefied flow with heat transfer in planar micro/nanochannels for $0.1\le\mathrm{Kn}\le1$ using a thermal LBM with diffuse-scattering boundary conditions (DSBC — a kinetic wall condition in which impinging molecules are re-emitted with a Maxwellian distribution at the wall temperature) and claimed validity up to $\mathrm{Kn}=1$. Luo's Comment (2011) argued that this claim is structurally unreliable, for three reasons. First, under the diffusive scaling ($\delta t\sim\delta x^2$) in which the LBE is consistent, a scheme with a fixed lattice-tied velocity set converges to a limit equation different from the Boltzmann equation, so the standard LBE is not a genuine kinetic solver and its validity is confined to the Navier–Stokes level; resolving Knudsen layers for $\mathrm{Kn}\gtrsim0.1$ requires more. Second, bounce-back and diffuse-scattering-type wall implementations place the effective wall location — and hence the computed slip and temperature jump — at positions that depend on the relaxation time and mesh resolution, rather than at the physical wall; the magnitudes of the imposed slip and temperature jump should instead be governed by the wall accommodation coefficient — the fraction of tangential momentum and energy that impinging molecules exchange with the wall before re-emission; in slip Poiseuille flow the MRT model with the 'magic-parameter' choice $s_q=8(2-s_v)/(8-s_v)$ removes this dependence, but no analogous parameter-independent construction was known for the transition regime or for coupled heat transfer. Third, the practical lattice velocity sets lack the isotropy and higher-order tensorial moments — velocity moments of the distribution function beyond the low-order ones that close the Navier–Stokes equations, which become dynamically relevant far from equilibrium — needed at high Kn, and velocity sets that recover them grow in size impractically fast. Consequently, 'Knudsen effects' reported by such simulations (for example nonmonotonic or reversed pressure deviations from the continuum prediction, and anomalously enhanced slip) can be numerical artifacts of the model and boundary treatment rather than physics. These criticisms define the methodological agenda this problem records.

Problem Statement

Construct a lattice Boltzmann formulation for rarefied gas flow with heat transfer in micro/nanochannels that is trustworthy through the transition regime up to $\mathrm{Kn}\sim1$ — removing Luo's structural objection that the standard LBE with lattice-tied velocities is not a genuine kinetic solver — by jointly satisfying three requirements:

1. Kinetic fidelity: the scheme must approximate Boltzmann-level behavior rather than Navier–Stokes-plus-slip, resolving the Knudsen-layer structure of velocity and temperature profiles for $\mathrm{Kn}\gtrsim0.1$ up to $\mathrm{Kn}\sim1$ in coupled flow and heat transfer, validated against Boltzmann-equation solutions or DSMC reference data.

2. Kinetic boundary conditions: velocity-slip and temperature-jump conditions must be imposed at physically fixed wall locations with explicit accommodation-coefficient dependence, with the effective wall location independent of mesh resolution and of the relaxation parameters — generalized beyond slip Poiseuille flow (where the MRT magic-parameter construction already achieves it) to the transition regime and to coupled heat transfer.

3. Tractable discrete velocities: the discrete-velocity sets (or an equivalent formulation) must recover the higher-order tensorial moments required at high Kn at computational cost practical for engineering simulation, rather than velocity sets whose size grows impractically with Kn.

These three requirements are facets of one objective, so a solution must discharge them jointly within a single formulation, not separately on disjoint benchmark cases: it discharges them when it reproduces the transition-regime velocity and temperature profiles of requirement 1 against the cited Boltzmann-equation or DSMC reference data across the stated Knudsen-number range — which thereby fixes the 'required' higher-order moments of requirement 3 — while the wall conditions of requirement 2 hold their effective location invariant under mesh refinement and relaxation-parameter change, and the velocity set of requirement 3 remains cheaper than direct molecular simulation (DSMC or direct Boltzmann solution) of the same case.

References

1. Li-Shi Luo, "Comment on 'Heat transfer and fluid flow in microchannels and nanochannels at high Knudsen number using thermal lattice-Boltzmann method'", Physical Review E 84, 048301 (2011), DOI 10.1103/PhysRevE.84.048301, https://doi.org/10.1103/PhysRevE.84.048301
2. J. Ghazanfarian, A. Abbassi, "Heat transfer and fluid flow in microchannels and nanochannels at high Knudsen number using thermal lattice-Boltzmann method", Physical Review E 82, 026307 (2010), DOI 10.1103/PhysRevE.82.026307, https://doi.org/10.1103/PhysRevE.82.026307
3. J. Ghazanfarian, A. Abbassi, "Reply to 'Comment on "Heat transfer and fluid flow in microchannels and nanochannels at high Knudsen number using thermal lattice-Boltzmann method"'", Physical Review E 84, 048302 (2011), DOI 10.1103/PhysRevE.84.048302, https://doi.org/10.1103/PhysRevE.84.048302
4. Yong-Hao Zhang, Xiao-Jun Gu, Robert W. Barber, David R. Emerson, "Capturing Knudsen layer phenomena using a lattice Boltzmann model", Physical Review E 74, 046704 (2006), DOI 10.1103/PhysRevE.74.046704, https://doi.org/10.1103/PhysRevE.74.046704
5. Y.-H. Zhang, X.-J. Gu, R. W. Barber, D. R. Emerson, "Modelling thermal flow in the transition regime using a lattice Boltzmann approach", Europhysics Letters 77, 30003 (2007), DOI 10.1209/0295-5075/77/30003, https://doi.org/10.1209/0295-5075/77/30003
6. Santosh Ansumali, Iliya V. Karlin, "Kinetic boundary conditions in the lattice Boltzmann method", Physical Review E 66, 026311 (2002), DOI 10.1103/PhysRevE.66.026311, https://doi.org/10.1103/PhysRevE.66.026311
7. K. Suga, "Lattice Boltzmann methods for complex micro-flows: applicability and limitations for practical applications", Fluid Dynamics Research 45, 034501 (2013), DOI 10.1088/0169-5983/45/3/034501, https://doi.org/10.1088/0169-5983/45/3/034501
8. Junjian Wang, Li Chen, Qinjun Kang, Sheik S. Rahman, "The lattice Boltzmann method for isothermal micro-gaseous flow and its application in shale gas flow: A review", International Journal of Heat and Mass Transfer 95, 94–108 (2016), DOI 10.1016/j.ijheatmasstransfer.2015.12.009, https://doi.org/10.1016/j.ijheatmasstransfer.2015.12.009
9. Minoru Watari, "Is the Lattice Boltzmann Method Applicable to Rarefied Gas Flows? Comprehensive Evaluation of the Higher-Order Models", Journal of Fluids Engineering 138, 011202 (2016), DOI 10.1115/1.4031000, https://doi.org/10.1115/1.4031000
10. Zhaoli Guo, Kun Xu, Ruijie Wang, "Discrete unified gas kinetic scheme for all Knudsen number flows: Low-speed isothermal case", Physical Review E 88, 033305 (2013), DOI 10.1103/PhysRevE.88.033305, https://doi.org/10.1103/PhysRevE.88.033305
11. Zhaoli Guo, Ruijie Wang, Kun Xu, "Discrete unified gas kinetic scheme for all Knudsen number flows. II. Thermal compressible case", Physical Review E 91, 033313 (2015), DOI 10.1103/PhysRevE.91.033313, https://doi.org/10.1103/PhysRevE.91.033313
12. Zhaoli Guo, Kun Xu, "Progress of discrete unified gas-kinetic scheme for multiscale flows", Advances in Aerodynamics 3, 6 (2021), DOI 10.1186/s42774-020-00058-3, https://doi.org/10.1186/s42774-020-00058-3
13. Goncalo Silva, Viriato Semiao, "Consistent lattice Boltzmann modeling of low-speed isothermal flows at finite Knudsen numbers in slip-flow regime: Application to plane boundaries", Physical Review E 96, 013311 (2017), DOI 10.1103/PhysRevE.96.013311, https://doi.org/10.1103/PhysRevE.96.013311
14. Goncalo Silva, "Consistent lattice Boltzmann modeling of low-speed isothermal flows at finite Knudsen numbers in slip-flow regime. II. Application to curved boundaries", Physical Review E 98, 023302 (2018), DOI 10.1103/PhysRevE.98.023302, https://doi.org/10.1103/PhysRevE.98.023302
15. Haruka Yasuoka, Masayuki Kaneda, Kazuhiko Suga, "Thermal lattice Boltzmann method for complex microflows", Physical Review E 94, 013102 (2016), DOI 10.1103/PhysRevE.94.013102, https://doi.org/10.1103/PhysRevE.94.013102
16. Oleg Ilyin, "Gaussian Lattice Boltzmann method and its applications to rarefied flows", Physics of Fluids 32, 012007 (2020), DOI 10.1063/1.5126306, https://doi.org/10.1063/1.5126306
17. Xiaoyu Li, Zhi Ning, Ming Lü, "Microchannel Gas Flow in the Multi-Flow Regime Based on the Lattice Boltzmann Method", Entropy 26, 84 (2024), DOI 10.3390/e26010084, https://doi.org/10.3390/e26010084
18. Michel Ho, Jean-Michel Tucny, Sami Ammar, Sébastien Leclaire, Marcelo Reggio, Jean-Yves Trépanier, "Lattice Boltzmann Model for Rarefied Gaseous Mixture Flows in Three-Dimensional Porous Media Including Knudsen Diffusion", Fluids 9, 237 (2024), DOI 10.3390/fluids9100237, https://doi.org/10.3390/fluids9100237
19. Yongjia Wu, Ouyue Zhang, Qinggang Wang, Xinyi Yang, Donghao Zhao, Tingzhen Ming, "A Knudsen layer correction model based on the lattice Boltzmann method", Computers & Fluids 301, 106795 (2025), DOI 10.1016/j.compfluid.2025.106795, https://doi.org/10.1016/j.compfluid.2025.106795