Pith. sign in

REVIEW 4 major objections 5 minor 30 references

Knudsen boundary layer equations with incoming boundary condition: full range of cutoff collision kernels and Mach numbers of the far field

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Knudsen layers with incoming data solved for all Mach numbers

desk verdict Solid technical machinery, but the damping-removal step in both linear and nonlinear theorems is asserted rather than proved; the paper is not self-contained. read the letter →

arxiv 2501.04035 v1 pith:UPNR3SBO submitted 2025-01-02 math.AP

Ning Jiang , Yi-Long Luo , Yulong Wu , Tong Yang This is my paper · ORCID
classification math.AP MSC 35Q2076P0535F3035B4535A0135A02
keywords KnudsenlayerequationincomingboundaryconditionBoltzmanncutoffcollisionkernelexponentialdecayexistenceanduniquenessMachnumberartificialdamping
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proves that the nonlinear Knudsen layer equation with an incoming (absorbing) boundary condition has a unique solution near a prescribed far-field Maxwellian, for every angular cutoff collision kernel in the full range $-3<\gamma\leq 1$ and for every Mach number of the far field. The result matters because Knudsen layers control the boundary conditions of fluid equations derived from the Boltzmann equation; for incoming data these are pressure-jump conditions whose coefficients come from solving this layer problem. Previous existence results were restricted to particular kernels or to nondegenerate Mach numbers, whereas the paper's classification gives the count of solvability conditions uniformly: 0, 1, 4, or 5 depending on the position of the Mach number relative to $\pm 1$. As a corollary, the solution decays like $\exp\{-c x^{2/(3-\gamma)}-c|v|^2\}$ in the $L^\infty$ framework.

What carries the argument

The carrying object is the $(x,v)$-mixed weight $\sigma(x,v)$ of (1.24), which grows like $(\delta x+l)^{2/(3-\gamma)}$ at large $x$ and interpolates to $(\delta x+l)/(1+|v-u|)^{1-\gamma}+3|v-u|^2$ inside the velocity-dependent layer, so that $|v_3|\sigma_x\lesssim \nu(v)$ and $\sigma_x\lesssim(\delta x+l)^{-\Theta}$ with $\Theta=(1-\gamma)/(3-\gamma)$. This weight makes the transport term $v_3\partial_x(e^{\hbar\sigma}f)$ balance the collision frequency $\nu(v)$, turning the equation into a damped problem for $f_\sigma=e^{\hbar\sigma}f$. The complementary mechanism is the artificial damping term $-\bar{\alpha}(\delta x+l)^{-\Theta/2}P^+v_3g-\bar{\beta}(\delta x+l)^{-\Theta/2}P^0g$, built from the entropy-flux projections $P^+,P^0,P$; these projections encode the signs of the diagonal entries $P(\psi_i,\psi_i)$ and supply coercivity on the null space of the linearized operator $L$. Removing the damping by imposing $P^+v_3f_1=P^0f_1=0$ yields the Mach-number-dependent solvability conditions.

What would settle it

Test the quoted inequality (2.11) of Lemma 2.5 for a soft potential such as $\gamma=-2$ with data supported in the singular region $|v_3|<1$; a violation of that inequality would remove the weighted $L^2$ control at the core of Lemma 3.4 and break the proof of Theorem 1.2. A numerical check of this estimate across the kernel range would settle whether the method is sound.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1.2: for parameters satisfying (PH), if the normalized source and boundary data lie in the vanishing-sources set VSS and the smallness quantity $\varsigma$ of (1.52) is below a threshold $\varsigma_0$, then the nonlinear Knudsen layer problem (1.1) has a unique solution $F(x,v)$ near the far Maxwellian $M(v)$, with $E^\infty(e^{\hbar\sigma}(F-M)/\sqrt{M})\leq C\varsigma$. The proof first establishes a linear theory (Theorem 1.1) for the problem (KL) by solving a damped version in a finite slab, taking the slab length to infinity, and then showing that the damping terms vanish exactly under finitely many solvability conditions. The same iteration closes the nonlinear problem using a bilinear estimate for the quadratic collision term $\Gamma(f,g)$. The resulting pointwise bound is $|f(x,v)|\lesssim e^{-c(\delta x+l)^{2/(3-\gamma)}}e^{-c|v|^2}$ for some $c>0$.

Load-bearing premise

The proof leans on several weighted bounds for the collision operator and for the quadratic nonlinearity that are stated without proof and taken from other papers; the existence theorem stands or falls with those deferred estimates.

Editorial extensions

If this is right

  • If Theorem 1.2 is correct, the nonlinear layer problem has a unique small solution for every far-field Mach number, with the solvability count given by 0, 1, 4, or 5 according to the position of $M_\infty$ relative to $-1$ and $1$.
  • The solution satisfies $|f(x,v)|\lesssim e^{-c(\delta x+l)^{2/(3-\gamma)}}e^{-c|v|^2}$, so the layer is exponentially thin in $x$ with a kernel-dependent exponent and Gaussian decay in velocity.
  • The linear theory (Theorem 1.1) characterizes VSS as a $C^1$ manifold of codimension $\#\{I^+\cup I^0\}$, giving a complete classification of admissible incoming boundary data.
  • The theorem provides the layer solution needed to compute pressure-jump coefficients in acoustic and compressible Euler limits with absorbing boundary conditions, now for the full range of cutoff kernels.
  • The artificial-damping iteration is constructive enough that the same scheme yields existence on slabs and then passes to the half-space, so the result can serve as a template for other half-space kinetic boundary problems.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural next step is to convert the linear decay proven here into nonlinear stability of the layer, extending previous stability results that covered only special kernels or restricted Mach numbers.
  • The same damping-projection mechanism should apply to Maxwell reflection boundary conditions with arbitrary accommodation coefficients, unifying the two complementary boundary-layer theories at the kinetic level.
  • The decay exponent $2/(3-\gamma)$ is plausibly optimal, since it matches the transport-collision balance $|v_3|\lesssim\nu(v)|v_3|^{1-\gamma}$; numerical experiments could test whether softer potentials indeed produce thicker layers with this exponent.
  • The codimension jump at $M_\infty=\pm 1$ suggests that linearization near these critical Mach numbers may require refined asymptotic expansions beyond a single far-field Maxwellian.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper studies the nonlinear Knudsen layer equation in a half-space with incoming boundary condition, perturbing around a far-field Maxwellian with arbitrary Mach number. It introduces an (x,v)-mixed weight and an artificial damping term, proves uniform a priori estimates on a finite-slab approximating problem, passes A→∞ to obtain a damped half-space problem, and then uses an iterative fixed-point argument to claim existence and uniqueness for the nonlinear problem under a vanishing-source-set condition, with exponential pointwise decay e^{-c x^{2/(3-γ)}-c|v|^2}. The main results are Theorem 1.1 for the linear problem and Theorem 1.2 for the nonlinear problem, covering -3<γ≤1 and all M∞∈R.

Significance. If completed, the result would be a genuine extension of prior boundary-layer existence results: the full cutoff range -3<γ≤1 and all Mach numbers with incoming data have not been treated together before, and the mixed-weight machinery is nontrivial. The paper provides a detailed architecture of weighted energy estimates and uses a plausible finite-slab approximation strategy. The proof is not self-contained, however: several families of operator estimates and the key nonlinear estimate are quoted or deferred, and the nonlinear solvability step is asserted rather than proved. With those gaps filled, this would be a useful contribution to the kinetic boundary-layer literature.

major comments (4)
  1. [§5, iteration scheme (5.5) and condition (5.15)] The proof of Theorem 1.2 shows only that the Cauchy limit f of the iteration (5.5) solves the damped problem, and then asserts that f solves the nonlinear problem (1.50) if and only if P+v3Iγ(f~b)=P0Iγ(f~b)=0. No argument is given that the constructed limit satisfies this condition under the hypothesis (H/√M,Fb/√M)∈VSS from (1.22), which is a linear condition on the data. For j∈I+ the projected equation for a_j=(ψ_j,v3f1) is a_j'=-α(δx+l)^{-Θ}a_j, so a_j(0)=0 is a nontrivial restriction on the outgoing trace; the text does not derive it from the hypotheses. Since (5.15) is the only mechanism for removing the artificial damping, the theorem as stated is not established.
  2. [§4.3, Eqs. (4.42)–(4.43)] The sign in the second-order equation for y=(Xj,v3f1) is inconsistent. From (4.42), d/dx(ψj,v3f1)=-β(δx+l)^{-Θ}(Xj,v3f1)(ψj,ψj)/(Xj,LXj), and from the preceding line d/dx(Xj,v3f1)+(ψj,v3f1)=0. Combining these gives y''=+β(δx+l)^{-Θ}(ψj,ψj)/(Xj,LXj)y, not the displayed negative sign. The sign is load-bearing for the 'Freezing Point Method' conclusion and for the codimension count, and the argument is not supplied.
  3. [§2.3, Lemmas 2.3–2.5; §5, Lemma 5.1; §3, Lemma 3.7] Several estimates that are load-bearing for both theorems are not proved in this manuscript. Lemmas 2.3–2.5 are quoted from the unpublished preprint [14] (Sections 8.1–8.3), Lemma 3.7 is deferred to [4], and Lemma 5.1, the trilinear estimate A∞(e^{ℏσ}Γ(f,g))≤C E∞(e^{ℏσ}f)E∞(e^{ℏσ}g), is stated without proof with only 'similar arguments in Lemma 3 of [24]'. Since the existence and uniqueness claims collapse if any of these estimates has unstated hypotheses or is wrong, the proof is not self-contained; the statements and hypotheses of the quoted lemmas should be reproduced or verified, and Lemma 5.1 should be proved or given a precise derivation.
  4. [Theorem 1.1 and Table 2; §4.3] The codimension statement for VSS is asserted rather than proved. After (4.43), the text states that condition (4.43) 'defines a codimension #{I+∪I0} subset of boundary data', and Table 2 lists the counts, but no argument shows that the map fb↦(P+Iℏ(fb), P0Iℏ(fb)) has full rank or that the count is exactly #{I+∪I0} in each Mach regime. The same issue appears in Table 3 for the nonlinear theorem. This codimension claim is part of the main results (Remark 1.1 and Theorem 1.2) and therefore needs a proof.
minor comments (5)
  1. [Abstract] The abstract contains several typos: 'tahe' should be 'the', 'bsed' should be 'based', and 'weihgt' should be 'weight'.
  2. [§4.3, Eq. (4.36)] The equation 'v3P0f=PS' in (4.36) appears to be missing the x-derivative; compare with (1.55) and (4.37), where the correct form is v3∂xP0f=PS.
  3. [§1.4.2, Eq. (1.52)] In Theorem 1.2, the quantity ς is defined using Dℏ(H/√M) without the superscript ∞ that appears in D∞ℏ in (1.44); the notation should be made consistent.
  4. [§5, Lemma 5.1] The reference to 'Lemma 3 of [24]' should be made precise, since the nonlinear estimate in (5.2)-(5.3) is central and the 'similar arguments' are not detailed.
  5. [§3, Lemma 3.7] The proof of Lemma 3.7 is omitted entirely with 'The proof is similar to Lemma 3.1 of [4]'; stating the precise relationship to [4] would help the reader verify the hypotheses on α,β,ℏ.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular reduction of the main theorems; the proof has deferred self-cited estimates and an unverified nonlinear solvability step, but these are dependencies and gaps, not equivalence-by-construction.

full rationale

The central claims are not forced by their assumptions through a circular chain. Theorem 1.2 assumes the linear vanishing-sources condition (H/√M, Fb/√M) ∈ VSS defined in (1.22), while the conclusion is the existence of a nonlinear solution; the iteration scheme (5.5), the nonlinear estimate (5.2)–(5.3), and the Cauchy argument (5.10)–(5.14) are substantive content rather than restatements of the hypothesis. The load-bearing operator lemmas 2.3–2.5 are quoted from the authors' own companion preprint [14] ('See [14] Section 8.1/8.2/8.3'), and Lemma 5.1 is deferred to 'similar arguments in Lemma 3 of [24]' with details omitted; these are serious external dependencies and correctness risks, but they are independent technical estimates about K and Γ, not reformulations of the Knudsen-layer existence theorem, so they do not constitute circularity under the strict definition. The main internal issue is a gap rather than circularity: after constructing the damped iteration limit, the text asserts at (5.15) that 'f solves the nonlinear problem (1.50) if and only if P+v3Iγ(f̃b) = P0Iγ(f̃b) = 0' but never verifies this condition for the constructed limit, and the analogous linear removal of damping invokes a 'Freezing Point Method from [14]' without details. A missing proof is not a circular reduction. Finally, the definition (1.22) makes the bare existence part of Theorem 1.1 definitional, but that theorem's substantive content lies in the a priori estimates, uniqueness, and codimension count, which are not supplied by the definition. Overall the paper should be scored low for circularity while noting that the proof is not self-contained.

Assumptions & free parameters 6 free parameters · 6 assumptions · 0 invented entities

The paper introduces a mixed weight σ and a damping term, both of which are technical devices rather than new physical entities. The main external inputs are the Grad cutoff kernel assumption and the borrowed operator estimates from the authors' prior preprint [14].

free parameters (6)
  • δ = sufficiently small
    Chosen in (PH); controls the mixed weight σ_x estimates; central to the proof.
  • ℏ = sufficiently small
    Weight exponent in f_σ=e^{ℏσ}f; smallness used throughout the a priori estimates.
  • ϑ = sufficiently small
    Velocity Gaussian weight exponent in w_{β,ϑ}; smallness required for Lemmas 2.3-2.5.
  • l = sufficiently large
    Shift in the mixed weight σ; large l makes σ_x small and helps coercivity.
  • α = 0<α<μ_γ
    Singular weight exponent z_{-α}(v); constrained by Lemma 2.5.
  • β = β ≥ β_γ + (1-γ)/2 + max{0,-γ}
    Algebraic velocity weight exponent in w_{β,ϑ}; chosen large enough for closure.
assumptions (6)
  • domain assumption Grad cutoff kernel: b(ω,u)=b̃(θ)|u|^γ, with 0≤b̃(θ)≤b̃0|cosθ| and -3<γ≤1
    Stated in (1.5)-(1.6); all results are confined to this class.
  • domain assumption Incoming boundary condition and far-field Maxwellian
    Problem (1.1); incoming data specified on v3>0 and far-field limit fixed.
  • domain assumption (S,f_b) ∈ VSS
    The theorems assume the source and boundary data lie in the vanishing sources set; the proof relies on this to get zero far-field limit.
  • domain assumption Smallness condition ς≤ς0 for nonlinear problem
    Theorem 1.2 is local; the iteration closes only for small sources.
  • ad hoc to paper Weighted K-estimates (Lemmas 2.3-2.5)
    Imported from the authors' companion paper [14], not proved here.
  • ad hoc to paper Nonlinear trilinear estimate (Lemma 5.1)
    Stated without proof, deferred to [24].

how reviews work

0 comments
Cite this review

Pith. "Pith review of Knudsen boundary layer equations with incoming boundary condition: full range of cutoff collision kernels and Mach numbers of the far field." pith.science (2026). https://pith.science/paper/UPNR3SBO

@misc{pith2026250104035,
  author       = {Pith},
  title        = {Pith review of: Knudsen boundary layer equations with incoming boundary condition: full range of cutoff collision kernels and Mach numbers of the far field},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UPNR3SBO}},
  note         = {Machine review of arXiv:2501.04035}
}
abstract

This paper establishes tahe existence and uniqueness of the nonlinear Knudsen layer equation with incoming boundary conditions. It is well-known that the solvability conditions of the problem vary with the Mach number of the far Maxwellian $\mathcal{M}^\infty$. We consider full ranges of cutoff collision kernels (i.e., $- 3 < \gamma \leq 1$) and all the Mach numbers of the far field in the $L^\infty_{x,v}$ framework. Additionally, the solution exhibits exponential decay $\exp \{- c x^\frac{2}{3 - \gamma} - c |v|^2 \}$ for some $c > 0$. To address the general angular cutoff collision kernel, we introduce a $(x,v)$-mixed weight $\sigma$. The proof is essentially bsed on adding an artificial damping term.

Figures

Figures reproduced from arXiv: 2501.04035 by the authors.

Figure 1
Figure 1. Derivation of uniform bounds for the approximate equation (Ap-eq). Here we denote by gσ = e ~σ g. reflection boundary condition g(A, v)|v3<0 = g(A, RAv) at x = A was imposed. In this paper, we employ the incoming data approximation at x = A as in [14], i.e., imposing the boundary condition g(A, v)|v3<0 = ϕA(v), where ϕA(v) is any fixed function with B(e ~σ(A,·)ϕA) < ∞ which means ϕA(v) → 0 as A → +∞. The rationality… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 28 canonical work pages

  1. [14]

    Jiang and Y.-L

    N. Jiang and Y.-L. Luo. Knudsen boundary layer equation s for full ranges of cutoff collision kernels: Maxwell reflection boundary with all accommodation coefficie nts in [0,1]. arXiv:2407.02852 [math.AP] , 2024

    arXiv 2024

  2. [24]

    R. M. Strain and Y. Guo. Exponential decay for soft poten tials near Maxwellian. Arch. Ration. Mech. Anal., 187 (2008), no. 2, 287-339

    work page 2008

  3. [4]

    Chen, T.-P

    C.-C. Chen, T.-P. Liu and T. Yang. Existence of boundary l ayer solutions to the Boltzmann equation. Anal. Appl. (Singap.) 2 (2004), no. 4, 337-363

    work page 2004

  4. [1]

    K. Aoki, C. Baranger, M. Hattori, S. Kosuge, G. Martal` o, J. Mathiaud and L. Mieussens. Slip boundary conditions for the compressible Navier-Stokes equations. J. Stat. Phys. , 169 (2017), no. 4, 744-781

    work page 2017

  5. [2]

    K. Aoki, F. Golse and N. Jiang, Derivation of acoustic sys tem with pressure jump boundary condition from linearized Boltzmann equation with absorbing boundar y. Preprint 2024

    work page 2024

  6. [3]

    Bardos, R

    C. Bardos, R. E. Caflisch and B. Nicolaenko. The Milne and K ramers problems for the Boltzmann equation of a hard sphere gas. Comm. Pure Appl. Math. , 39 (1986), no. 3, 323-352

    work page 1986

  7. [5]

    Coron, F

    F. Coron, F. Golse and C. Sulem. A classification of well-p osed kinetic layer problems. Comm. Pure Appl. Math., 41 (1988), no. 4, 409-435

    work page 1988

  8. [6]

    F. Golse. Analysis of the boundary layer equation in the k inetic theory of gases. Bull. Inst. Math. Acad. Sin. (N.S.) 3 (2008), no. 1, 211-242. 40 NING JIANG, YI-LONG LUO, YULONG WU, AND TONG YANG

    work page 2008

Show all 30 references
  1. [7]

    Golse, B

    F. Golse, B. Perthame and C. Sulem. On a boundary layer pro blem for the nonlinear Boltzmann equation. Arch. Ration. Mech. Anal. , 103 (1988), no. 1, 81-96

    1988

  2. [8]

    Golse and F

    F. Golse and F. Poupaud. Stationary solutions of the line arized Boltzmann equation in a half-space. Math. Methods Appl. Sci. 11 (1989), no. 4, 483-502

    1989

  3. [9]

    Guo, F.M

    Y. Guo, F.M. Huang and Y. Wang. Hilbert expansion of the Bo ltzmann equation with specular boundary condition in half-space. Arch. Ration. Mech. Anal. 241 (2021), no. 1, 231-309

    2021

  4. [10]

    H. Grad. Asymptotic theory of the Boltzmann equation. I I. Rarefied Gas Dynamics (Proc. 3rd Internat. Sympos., Palais de l’UNESCO, Paris, 1962), Vol. I (1963) pp. 26-59, Academic Press, New York

    1963

  5. [11]

    L.-B. He, N. Jiang and Y.L. Wu. Boundary layer solution o f the Boltzmann equation for Maxwell reflection boundary condition. 2024, preprint

    2024

  6. [12]

    Huang, Z.-H

    F.M. Huang, Z.-H. Jiang and Y. Wang. Boundary layer solu tion of the Boltzmann equation for specular boundary condition. Acta Math. Appl. Sin. Engl. Ser. , 39 (2023), no. 1, 65-94

    2023

  7. [13]

    Huang and Y

    F.M. Huang and Y. Wang. Boundary layer solution of the Bo ltzmann equation for diffusive reflection boundary conditions in half-space. SIAM J. Math. Anal. , 54 (2022), no. 3, 3480-3534

    2022

  8. [15]

    Jiang, Y.-L

    N. Jiang, Y.-L. Luo and S.J. Tang. Compressible Euler li mit from Boltzmann equation with complete diffusive boundary condition in half-space. arXiv:2104.11964 [math.AP] , 2021

    2021 arXiv

  9. [16]

    Jiang, Y.-L

    N. Jiang, Y.-L. Luo and S.J. Tang. Compressible Euler li mit from Boltzmann equation with Maxwell reflection boundary condition in half-space. arXiv:2101.11199 [math.AP] , 2021

    2021 arXiv

  10. [17]

    Jiang, Y.-L

    N. Jiang, Y.-L. Luo and S.J. Tang. Grad-Caflisch type dec ay estimates of pseudo-inverse of linearized Boltzmann operator and application to Hilbert expansion of compressible Euler scaling. arXiv:2206.02677 [math.AP], 2022

    2022 arXiv

  11. [18]

    Jiang and Y.L

    N. Jiang and Y.L. Wu, Acoustic limit for the Boltzmann eq uation with incoming boundary condition: renormalized solutions. Preprint 2024

    2024

  12. [19]

    Jiang and Y.L

    N. Jiang and Y.L. Wu, Formal derivation of the boundary c onditions for the compressible Navier-Stokes- Fourier system from Boltzmann equations with incoming and M axwell reflection conditions. Preprint 2024

    2024

  13. [20]

    Levermore and W.R

    C.D. Levermore and W.R. Sun. Compactness of the gain par ts of the linearized Boltzmann operator with weakly cutoff kernels. Kinet. Relat. Models 3 (2010), no. 2, 335-351

    2010

  14. [21]

    Q. Li, J.F. Lu and W.R. Sun. Half-space kinetic equation s with general boundary conditions. Math. Comp. 86 (2017), no. 305, 1269-1301

    2017

  15. [22]

    Y. Sone. Kinetic theory and fluid dynamics. Modeling and Simulation in Science, Engineering and Tech- nology. Birkh¨ auser Boston, Inc., Boston, MA, 2002

    2002

  16. [23]

    Y. Sone. Molecular gas dynamics. Theory, techniques, and applicati ons. Modeling and Simulation in Sci- ence, Engineering and Technology. Birkh¨ auser Boston, Inc., Boston, MA, 2007

    2007

  17. [25]

    S. Ukai, T. Yang and S.-H. Yu. Nonlinear boundary layers of the Boltzmann equation. I. Existence. Comm. Math. Phys. 236 (2003), no. 3, 373-393

    2003

  18. [26]

    S. Ukai, T. Yang and S.-H. Yu. Nonlinear stability of bou ndary layers of the Boltzmann equation. I. The case M∞ < −1. Comm. Math. Phys. 244 (2004), no. 1, 99-109

    2004

  19. [27]

    W.K. Wang, T. Yang and X.F. Yang. Nonlinear stability of boundary layers of the Boltzmann equation for cutoff hard potentials. J. Math. Phys. 47 (2006), no. 8, 083301, 15 pp

    2006

  20. [28]

    W.K. Wang, T. Yang and X.F. Yang. Existence of boundary l ayers to the Boltzmann equation with cutoff soft potentials. J. Math. Phys. 48 (2007), no. 7, 073304, 21 pp

    2007

  21. [29]

    X.F. Yang. Nonlinear stability of boundary layers for t he Boltzmann equation with cutoff soft potentials. J. Math. Anal. Appl. 345 (2008), no. 2, 941-953

    2008

  22. [30]

    X.F. Yang. The solutions for the boundary layer problem of Boltzmann equation in a half-space. J. Stat. Phys. 143 (2011), no. 1, 168-196. KNUDSEN LAYER EQUATION 41 (Ning Jiang) School of Mathematics and Statistics, Wuhan University, Wu han, 430072, P. R. China Email address : ...

    2011

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.