Monodisperse gas-solid mixtures with intense interphase interaction in two-fluid Smoothed Particle Hydrodynamics

Nicolay Snytnikov; Olga Stoyanovskaya; Tatiana Glushko; Valery Snytnikov

REVIEW 3 major objections 3 minor 17 references

Monodisperse gas-solid mixtures with intense interphase interaction in two-fluid Smoothed Particle Hydrodynamics

T0 review · 3 major / 3 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read Cell-averaged drag lets two-fluid SPH capture strongly coupled gas-dust dynamics at coarse resolution.

desk verdict A useful numerical comparison of three TFSPH drag schemes, but the paper's asymptotic-preserving claim for IDIC outruns the evidence: two test problems, no error control for the cell averaging, and parameter tables that do not match the figures. read the letter →

arxiv 1908.01682 v2 pith:FQSAEE4S submitted 2019-08-05 physics.comp-ph cs.NAmath.NA

Olga Stoyanovskaya , Tatiana Glushko , Valery Snytnikov , Nicolay Snytnikov This is my paper · ORCID

classification physics.comp-phcs.NAmath.NA

keywords gas-solidmixturestwo-fluidsmoothedparticlehydrodynamicsdragforcestiffrelaxationtermasymptoticpreservingschemeparticle-in-cellEpsteinDustyshocktestproblem

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

The paper compares three ways of computing the drag force between gas and dust in two-fluid Smoothed Particle Hydrodynamics, focusing on monodisperse mixtures where the velocity relaxation time $t_{\rm stop}$ is much shorter than the dynamical time. In that regime, a classical explicit pairwise drag scheme and a semi-implicit interpolation-based scheme both develop strong dissipation unless the smoothing length obeys $h < c_s t_{\rm stop}$, which forces very fine spatial resolution. The paper describes and tests a third scheme, 'drag in cell' (IDIC), in which the drag force is computed from cell-averaged gas and dust velocities; it argues that this scheme is asymptotic preserving, so the coupled gas-dust solution is captured even when the timestep $\tau$ and smoothing length $h$ are much larger than the drag scale. On the Dustywave and Dustyshock tests with intense drag ($K=500$, $\rho_d/\rho_g=1$), IDIC reproduces the reference solutions at coarse resolution while the other two methods do not. If the claim holds, strongly coupled gas-dust simulations can run at ordinary SPH resolution, which matters for planet formation and industrial particle-flow applications.

What carries the argument

The central object is the 'drag in cell' (IDIC) approximation. The domain is partitioned into disjoint Euler cells; within a cell containing $N$ gas particles of mass $m_g$ and $L$ dust particles of mass $m_d$, one defines the mass ratio $\varepsilon^* = m_d L/(m_g N)$, the effective drag factor $K^* = \rho_d^*/t_{\rm stop}^*$, and the cell-averaged velocities $v_* = \sum_a v_a/N$ and $u_* = \sum_j u_j/L$. Drag on a gas particle uses $v_a - u_*$, drag on a dust particle uses $v_* - u_j$, and the drag factor and densities are evaluated on the previous time layer while the relative velocity is evaluated on the new one. This semi-implicit per-cell treatment of the relaxation term removes the $\tau < t_{\rm stop}$ stability restriction and, according to the comparison, the $h < c_s t_{\rm stop}$ resolution restriction; the first-order-in-time version conserves momentum cell by cell and can be advanced by a direct update. In the stiff limit $t_{\rm stop}\to 0$, the scheme is designed to reproduce the fully coupled dusty-gas dynamics without resolving the relaxation scale.

What would settle it

Run IDIC on a Dustyshock-style setup over a range of shock strengths and dust-to-gas ratios while keeping the Euler cells fixed, and compare coarse-cell results against a converged high-resolution reference; if the error grows systematically with shock strength or with dust loading, the asymptotic-preserving property does not extend beyond the tested range. A second concrete check is a problem with nonlinear drag, for example $t_{\rm stop}$ depending on relative speed, where the cell-averaged velocity may not be the correct representative velocity; a systematic bias there would mark the linear-drag boundary of the method.

Watch

Extended reading notes

Core claim

The central claim is that IDIC, whose drag term is evaluated with Eulerian cells, is asymptotic preserving for monodisperse gas-dust mixtures with linear drag. Inside each cell the gas velocity acting on dust is replaced by the cell mean $v_*$ and the dust velocity acting on gas by the cell mean $u_*$; the drag factor and densities are taken from the previous time layer and the relative velocity from the new one. The paper shows that this first-order semi-implicit treatment conserves momentum cell by cell and allows a direct update of $u^{n+1}$ and $v^{n+1}$. In the two test problems, IDIC keeps the $L_2$ error in dust velocity far below that of the two fully Lagrangian schemes when the smoothing length is raised from $h=0.001$ to $h=0.025$. The summary states that IDIC permits a timestep and smoothing length independent of drag intensity, whereas the classical pairwise and semi-implicit interpolation schemes require fine spatial resolution for intense interphase interaction.

Load-bearing premise

The load-bearing premise is that a single cell-averaged gas velocity, with drag coefficient and densities taken from the previous time level, faithfully represents the gas seen by dust inside that cell even when the cell contains steep velocity or density gradients; the paper's evidence for this rests on two test problems rather than a general error bound.

Editorial extensions

If this is right

Strongly coupled gas-dust mixtures can be simulated with SPH resolution set by the gas flow features rather than by the dust stopping length.
The timestep can be chosen from the Courant condition for the gas instead of the much more restrictive $\tau < t_{\rm stop}$ condition.
Momentum is conserved per Euler cell, so the coarse-cell drag treatment does not introduce spurious momentum sources or sinks.
The comparison gives a concrete warning that the two fully Lagrangian drag formulations should be used for intense interphase interaction only when $h < c_s t_{\rm stop}$ is satisfied.
For monodisperse solids with Epstein or Stokes drag, the method opens a practical simulation route for planet formation and particle-flow engineering.

Reading between the lines

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

A cell-averaged drag law acts as a spatial filter: if the relative velocity varies within a cell, the scheme responds to the mean rather than the local slip, so problems with sub-cell dust-gas velocity structure may need intra-cell velocity-gradient corrections.
The asymptotic-preserving property is demonstrated only for linear drag with $t_{\rm stop}$ independent of relative velocity; the same cell-averaged implicit idea could plausibly extend to nonlinear drag laws, but the averaging error would then need separate analysis.
In the limit $t_{\rm stop}\to 0$ the mixture should behave as a single fluid, and IDIC should reduce naturally to the pure-gas SPH update; verifying this limit in the code would be a direct check of the asymptotic-preserving claim.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

What you should know: this is a short conference paper that compares three ways to compute drag in two-fluid SPH on the Dustywave and Dustyshock benchmarks. The IDIC method itself is not new, the authors say it was introduced in their 2018 Astronomy and Computing paper. The new content is the head-to-head L2-error comparison with MK and ISPH. On the two tests, IDIC clearly wins: at h = 0.01–0.02, its L2 errors are 20–100 times smaller than MK or ISPH, and the figures agree with the tables. That is a real, useful result for people working on strongly coupled gas-dust SPH.

The soft spots are in the interpretation. The Summary says IDIC is 'asymptotic preserving' and lets time step and smoothing length be independent of drag intensity. That is a general claim, but the evidence is two test problems with no formal AP analysis, no systematic scan of t_stop, and no convergence study in the Euler cell size. In fact the Euler cell size is never reported in Section 2.3 or the test descriptions, so you cannot tell whether the good behavior comes from the smoothing length h or from the cell averaging. The cell-mean drag in Eqs. (19)–(21) replaces local velocities with cell averages, and there is no error bound in terms of t_stop, h, or cell size. So the 'asymptotic preserving' label is more than the paper justifies.

The parameter reporting is sloppy. Table 1 gives h = 0.01, τ = 0.00025, while the caption of Fig. 1 says h = 0.01 uses τ = 0.001; Table 2 and Fig. 2 similarly disagree about τ. The Dustyshock t_stop = 0.00025 in the figure caption is not obviously consistent with K = 500 and ρd/ρg = 1. These inconsistencies are minor in the sense that the qualitative conclusion probably survives, but they make the quantitative errors hard to trust. No code or data is shipped, so the numbers cannot be checked independently.

Credit where earned: the comparison is against analytic solutions, so it is not circular. The IDIC momentum-conservation property was shown in the authors' earlier paper. For a conference proceedings, this is a legitimate extension, but the central claim needs to be toned down to 'IDIC behaves robustly on these two stiff test problems.' If this were submitted as a journal paper, I would send it to review with a request for a formal AP argument or a cell-size convergence study, and for corrected parameter reporting. As it stands, it is a useful, citable data point, but not a proof of asymptotic preservation.

Referee Report

3 major / 3 minor

Summary. The paper addresses two-fluid Smoothed Particle Hydrodynamics (TFSPH) simulations of monodisperse gas-solid mixtures with intense interphase interaction, i.e., drag relaxation times t_stop much smaller than the dynamical time. It compares three drag-discretization strategies: the explicit Monaghan–Kocharyan (MK) scheme, a semi-implicit interpolated SPH (ISPH) scheme, and the authors' previously introduced SPH-IDIC ('drag in cell') scheme, in which the drag force is evaluated using piecewise-constant Euler-cell-averaged velocities. The manuscript presents Dustywave (isothermal sound wave) and Dustyshock (Sod shock tube) test problems with reference solutions, reports L2 errors for the three methods at different smoothing lengths and time steps, and concludes that IDIC is asymptotic preserving and permits smoothing length and time step to be chosen independently of drag intensity, while MK and ISPH require h < c_s t_stop. The numerical evidence shows dramatically smaller IDIC errors in the reported cases, but the manuscript does not supply an asymptotic-preserving proof, a systematic parameter scan, or a cell-size convergence study, and the reported test parameters contain internal inconsistencies.

Significance. If the central claim is correct, the result is practically significant: it would remove the severe resolution constraint h < c_s t_stop for monodisperse gas-dust mixtures with linear drag, allowing affordable TFSPH simulations of stiff momentum exchange in astrophysical and engineering applications. The paper's positive features include the use of independent analytic reference solutions for both benchmark problems, explicit error tables, and a clear statement of the mechanism (cell-averaged implicit drag) underlying the proposed method. However, the significance is currently conditional on a missing asymptotic analysis or resolution study: the two tests, as presented, cannot separate genuine asymptotic-preserving behavior from favorable grid-smoothing effects, especially because the Euler-cell size is never specified.

major comments (3)

[Section 2.3, Eqs. (19)–(21)] The Euler-cell decomposition is introduced only abstractly ('disjoint volumes' with N gas and L dust particles); the actual cell construction, cell size, and any dependence of the results on cell resolution are never reported. Since the IDIC scheme replaces the gas velocity with the cell mean v* and the drag factor and densities with previous-time-layer cell quantities, the cell size is a discretization parameter that directly controls the error. Without a statement of the cell size used in Figures 1–2 and Tables 1–2, or a cell-size convergence study, the numerical results cannot establish that h, rather than the Euler-cell size, is the resolution that controls the asymptotic error. This omission is load-bearing for the paper's central claim.
[Summary and Section 3] The claim that IDIC is 'asymptotic preserving' is not supported by a proof or by an asymptotic-resolution study. The paper shows two test problems at fixed values of t_stop, h, and tau; there is no epsilon-asymptotic analysis as t_stop -> 0, no systematic variation of t_stop over several orders of magnitude, and no scan of the final time t_stop relative to t. The observed small IDIC errors at the reported parameters are consistent with the claim but do not demonstrate asymptotic preservation, which requires showing that the numerical solution converges to the equilibrium (zero-relative-velocity) solution at a controlled rate independent of the stiff parameter. The authors should either add a formal consistency/AP argument for Eqs. (19)–(21) or provide a resolved convergence study in t_stop, h, and cell size.
[Tables 1–2 and Figures 1–2] The reported test parameters are internally inconsistent, which undermines the quantitative comparison. With the definitions in Eqs. (3) and (7), K = 500 and rho_d/rho_g = 1 imply t_stop = 0.002 (taking rho_g = 1), yet the Dustyshock description states t_stop = 0.00025. In addition, Table 1 lists h = 0.01 with tau = 0.00025 and h = 0.02 with tau = 0.001, while the Figure 1 caption states that for h = 0.025 and h = 0.01 the time step is tau = 0.001; Table 2 lists h = 0.01 with tau = 0.000025 and h = 0.02 with tau = 0.0001, while the Figure 2 caption states tau = 0.001 for h = 0.01 and tau = 0.0001 for h = 0.001. These inconsistencies must be resolved before the error values can be trusted as evidence for the AP claim.

minor comments (3)

[Throughout] The text contains numerous typographical errors and inconsistent capitalization, including 'bahaiviour', 'extreemly', 'assosiated', 'gravitaion', 'dymanics', and 'euler'; a careful proofreading pass is needed.
[Section 3, Figure captions] The captions of Figures 1 and 2 describe smoothing lengths and time steps that do not match the corresponding rows of Tables 1 and 2; the captions and tables should be reconciled so that each reported error is unambiguously associated with a single (h, tau) pair.
[Section 2.3, Eqs. (19)–(21)] It would help the reader to state explicitly whether the Euler cells are uniform, how their size is chosen relative to h, and whether the cell decomposition is recomputed each time step or held fixed.

Circularity Check

0 steps flagged · score 0.0 of 10

No construction-level circularity: the IDIC claim is benchmarked against analytic Dustywave and Sod solutions, not against its own inputs.

full rationale

The paper's central claim is that the IDIC method preserves the asymptotic solution for intense interphase drag while allowing timestep and smoothing length independent of drag intensity. The evidence for this claim is presented in Section 3 as numerical comparisons against the analytic Dustywave solution and the Sod shock-tube reference solution, both of which are external benchmarks. No parameter in the IDIC scheme is fitted to these benchmarks, and the asymptotic-preserving property is not defined in terms of the test outcomes. The methods MK, ISPH, and IDIC are each implemented from their published formulations, and the reported errors are computed against independent reference solutions. The main self-citation is the use of the authors' prior work [9] for the IDIC scheme, the test-problem setup, and the statement of momentum conservation for the scheme. This is a normal extension of prior work rather than a load-bearing citation: the present paper supplies its own numerical demonstration of the asymptotic-preserving behavior, and the benchmarks themselves are analytic and externally grounded. The absence of a formal asymptotic-preserving consistency proof or a reported Euler-cell size is a correctness or completeness concern, not a circularity concern, because the claim is tested rather than assumed. Therefore no circular step can be exhibited, and the appropriate score is 0.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

No fitted constants appear in the method; the K=500 drag test value and CFL=0.1 are chosen test parameters, not free parameters of the derivation. The Euler cells are a computational decomposition, not a new physical entity.

assumptions (4)

domain assumption The two-fluid continuity and momentum equations (1)-(2) with drag f_drag = rho_d (v-u)/t_stop describe the gas-solid mixture.
Standard gas-solid mixture model taken from [1,10]; not derived in this paper.
domain assumption t_stop is independent of relative velocity, corresponding to Epstein and Stokes regimes.
Stated in the Section 2 preamble; the SPH-IDIC implicit update relies on this linearity, and the asymptotic-preserving claim is not demonstrated for nonlinear drag.
ad hoc to paper Within each Euler cell, gas velocity is constant and equal to the cell average v*, with drag factor and densities taken from the previous time layer.
Introduced in Section 2.3 as the defining approximation of IDIC; this is the load-bearing modeling choice that makes the scheme cheap and implicit, and it is not justified by a cell-gradient error estimate.
domain assumption The reference solutions for Dustywave and Dustyshock are accurate.
The paper relies on its own [9] and standard analytic or shock-tube solutions; no independent derivation of the reference solutions is included.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Monodisperse gas-solid mixtures with intense interphase interaction in two-fluid Smoothed Particle Hydrodynamics." pith.science (2026). https://pith.science/paper/FQSAEE4S

@misc{pith2026190801682,
  author       = {Pith},
  title        = {Pith review of: Monodisperse gas-solid mixtures with intense interphase interaction in two-fluid Smoothed Particle Hydrodynamics},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FQSAEE4S}},
  note         = {Machine review of arXiv:1908.01682}
}

abstract

Simulations of gas-solid mixtures are used in many scientific and industrial applications. Two-Fluid Smoothed Particle Hydrodynamics (TFSPH) is an approach when gas and solids are simulated with different sets of particles interacting via drag force. Several methods are developed for computing drag force between gas and solid grains for TFSPH. Computationally challenging are simulations of gas-dust mixtures with intense intephase interaction, when velocity relaxation time $t_{\rm stop}$ is much smaller than dynamical time of the problem. In explicit schemes the time step $\tau$ must be less than $t_{\rm stop}$, that leads to high computational costs. Moreover, it is known that for stiff problems both grid-based and particle methods may require unaffordably detailed resolution to capture the asymptotical bahaiviour of the solution. To address this problem we developed fast and robust method for computing stiff and mild drag force in gas solid-mixtures based on the ideas of Particle-in-Cell approach. In the paper we compare the results of new and previously developed methods on test problems

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

17 extracted references · 17 canonical work pages

[1]

Saito, M

T. Saito, M. Marumoto, K. Takayama. Numerical investigations of shock waves in gas-particle mixtures. Evaluation of numerical methods for dusty-gas shock wave phenomena. Shock Waves 13: 299–322 (2003)

work page 2003
[2]

Jin, C.D

S. Jin, C.D. Livermore. Numerical Schemes for Hyperbolic Conservation Laws with Stiﬀ Relaxation Terms. Journal of Computational Physics 126, 449–467 (1996)

work page 1996
[3]

S. Jin. Asymptotic preserving schemes for multiscale kinetic and hyperbolic equa- tions: a review. Riv. Mat. Univ. Parma, 3, 177–216 (2012) 8 Olga P. Stoyanovskaya, Tatiana A. Glushko, Valeriy N. Snytnikov and Nikolay V. Snytnikov

work page 2012
[4]

G. Albi, G. Dimarco, L. Pareschi. Implicit-Explicit multistep methods for hyperbolic systems with multiscale relaxation. arXiv e-prints: 1904.03865

work page Pith review arXiv 1904
[5]

O. P. Stoyanovskaya, V. N. Snytnikov, E. I. Vorobyov. Analysis of methods for com- puting the trajectories of dust particles in a gas-dust circumstellar disk. Astron.Rep., Vol. 61, No. 12:1044–1060 (2017)

work page 2017
[6]

Degond, F

P. Degond, F. Deluzet. Asymptotic-Preserving methods and multiscale models for plasma physics. Journal of Computational Physics. 336, 429–457 (2017)

work page 2017
[7]

Laibe, D

G. Laibe, D. J. Price. DUSTYBOX and DUSTYWAVE: two test problems for numer- ical simulations of two-ﬂuid astrophysical dust-gas mixtures. MNRAS, 418:1491–1497 (2011)

work page 2011
[8]

Lor´ en-Aguilar, M

P. Lor´ en-Aguilar, M. R. Bate. Two-ﬂuid dust and gas mixtures in smoothed particle hydrodynamics: a semi-implicit approach. MNRAS, 443:927–945 (2014)

work page 2014

Show all 17 references

[9]

O. P. Stoyanovskaya, T. A. Glushko, N. V. Snytnikov, V. N. Snytnikov, Two-Fluid Dusty Gas in Smoothed Particle Hydrodynamics: Fast and Implicit Algorithm for Stiﬀ Linear Drag. Astronomy and Computing, 25:25–37 (2018)

2018
[10]

J. J. Monaghan, A. Kocharyan. SPH simulation of multi-phase ﬂow. Computer Physics Communications, 87:225–235 (1995)

1995
[11]

S. T. Madisson, R. J. Humble, J. R. Murray. Building Planets with Dusty Gas. Bioastronomy 2002: Life Among the Stars, 213:231 (2004)

2004
[12]

Barriere-Fouchet, J

L. Barriere-Fouchet, J. F. Gonzalez, J. R. Murray, R. J. Humble, S. T. Maddi- son. Dust distribution in protoplanetary disks. Vertical settling and radial migration. AAp, 443:185–194 (2005)

2005
[13]

J. F. Gonzalez, G. Laibe, S. T. Maddison. Self-induced dust traps: overcoming planet formation barriers. MNRAS, 467:1984–1996 (2017)

2017
[14]

SPH method for two- ﬂuid modeling of particle–ﬂuid ﬂuidization

Xiong, Qingang and Deng, Lijuan and Wang, Wei and Ge, Wei. SPH method for two- ﬂuid modeling of particle–ﬂuid ﬂuidization. Chemical Engineering Science - CHEM ENG SCI, 66:1859–1865 (2011)

2011
[15]

R. A. Booth, D. Sijacki, C. J. Clarke. Smoothed particle hydrodynamics simulations of gas and dust mixtures. MNRAS, 452:3932–3947 (2015)

2015
[16]

W. K. M. Rice, G. Lodato, J. E. Pringle, P. J. Armitage, I. A. Bonnell. Accelerated planetesimal growth in self-gravitating protoplanetary discs. MNRAS, 355:543–552 (2004)

2004
[17]

M. J. Andrews, P. J. O’Rourke. The multiphase Particle-in-Cell (MP-PIC) method for dense particulate ﬂows. Int. J. of Multiphase Flow, 22:379–402 (1996) 9

1996

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Saito, M

T. Saito, M. Marumoto, K. Takayama. Numerical investigations of shock waves in gas-particle mixtures. Evaluation of numerical methods for dusty-gas shock wave phenomena. Shock Waves 13: 299–322 (2003)

work page 2003

[2] [2]

Jin, C.D

S. Jin, C.D. Livermore. Numerical Schemes for Hyperbolic Conservation Laws with Stiﬀ Relaxation Terms. Journal of Computational Physics 126, 449–467 (1996)

work page 1996

[3] [3]

S. Jin. Asymptotic preserving schemes for multiscale kinetic and hyperbolic equa- tions: a review. Riv. Mat. Univ. Parma, 3, 177–216 (2012) 8 Olga P. Stoyanovskaya, Tatiana A. Glushko, Valeriy N. Snytnikov and Nikolay V. Snytnikov

work page 2012

[4] [4]

G. Albi, G. Dimarco, L. Pareschi. Implicit-Explicit multistep methods for hyperbolic systems with multiscale relaxation. arXiv e-prints: 1904.03865

work page Pith review arXiv 1904

[5] [5]

O. P. Stoyanovskaya, V. N. Snytnikov, E. I. Vorobyov. Analysis of methods for com- puting the trajectories of dust particles in a gas-dust circumstellar disk. Astron.Rep., Vol. 61, No. 12:1044–1060 (2017)

work page 2017

[6] [6]

Degond, F

P. Degond, F. Deluzet. Asymptotic-Preserving methods and multiscale models for plasma physics. Journal of Computational Physics. 336, 429–457 (2017)

work page 2017

[7] [7]

Laibe, D

G. Laibe, D. J. Price. DUSTYBOX and DUSTYWAVE: two test problems for numer- ical simulations of two-ﬂuid astrophysical dust-gas mixtures. MNRAS, 418:1491–1497 (2011)

work page 2011

[8] [8]

Lor´ en-Aguilar, M

P. Lor´ en-Aguilar, M. R. Bate. Two-ﬂuid dust and gas mixtures in smoothed particle hydrodynamics: a semi-implicit approach. MNRAS, 443:927–945 (2014)

work page 2014