REVIEW 4 major objections 5 minor 1 references
Improvement of Solid-Fluid Interaction Scheme in Lattice Boltzmann Immiscible Pseudopotential Models
T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash
Pith's one-line read A local adhesion parameter lets lattice Boltzmann wetting simulations reach contact angles from 10 to 165 degrees.
desk verdict A useful local-adhesion fix for immiscible pseudopotential LBM, with solid benchmarks, but the curved-boundary copying rule is under-tested and the strongest claims outrun the evidence. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing object is the local adhesion parameter $G_{w,1}(\rho_e)$, built from the normalized effective density $\rho_e=(\rho_1-\rho_2)/(\rho_1+\rho_2)$. For $|\rho_e|>\delta$ the parameter saturates to the bulk value $G_{w,1}$ or to $G_{w,2}=G_{12}$; inside the interface, quadratic interpolation functions $g_1(\rho_e)$ and $g_2(\rho_e)$ smooth the transition, with $\delta=0.8$ chosen to match an interface width of about five lattice spacings at $G_{12}=0.17$. The second mechanism is the eighth-order isotropic force stencil, whose two-layer interaction range leaves second-layer nodes undefined near walls; the paper fills those nodes by copying the information from the geometrically corresponding first-layer node, which is what makes curved boundaries work in this scheme.
What would settle it
Run the static-droplet-on-a-cylinder benchmark at the same physical settings with cylinder radii of 35, 70, and 140 lattice units, and record the measured contact angle plus the minimum fluid density near the wall. If the contact angle drifts with radius or the mass-transfer layer reappears at the smallest radius, then the copy rule, not the local adhesion parameter, is responsible for the curved-surface results.
Extended reading notes
Core claim
This paper claims that the failure mode of the original Martys–Chen solid-fluid interaction—an unphysical layer through which one fluid appears to dissolve into the other near the wall—is caused by applying a uniform adhesion parameter regardless of local phase composition. The improved scheme defines an effective density $\rho_e=(\rho_1-\rho_2)/(\rho_1+\rho_2)$ and chooses $G_{w,1}$ from a piecewise function of $\rho_e$, with a transition width $\delta$ matched to the diffuse interface width. That makes the wall force respond locally to which fluid dominates and to how close the node is to the interface, so the interface stays sharp at the boundary. The paper reports that the scheme attains static contact angles from about 10 to 165 degrees, eliminates the mass-transfer layer, cuts average spurious currents by roughly half relative to the original scheme, and conserves mass. It also introduces a boundary reconstruction for the eighth-order isotropic interaction stencil: the second-layer fluid-node values are copied from the corresponding first-layer nodes, allowing curved walls to be handled without ghost nodes.
Load-bearing premise
The curved-boundary results stand on the assumption that the flow is locally uniform enough that copying first-layer node values to second-layer nodes two lattice spacings away is accurate; the paper reports no grid-convergence or curvature-sensitivity test of this rule.
Editorial extensions
If this is right
- Contact angles below about 30 degrees and above about 120 degrees become accessible without the mass-transfer artifact, so strongly hydrophilic and strongly hydrophobic microfluidic and porous-media flows can be simulated.
- Average spurious currents near the wall drop by roughly half in hydrophilic regimes and by about a third in hydrophobic regimes, improving numerical stability for body-force-driven flows.
- The scheme retains the simple single-parameter calibration of the original method, since only the strength coefficient $\lambda = G_{w,1}/G_{w,2}$ needs to be tuned.
- The second-layer copying rule extends the eighth-order isotropic stencil to curved no-slip surfaces, making the improved scheme usable on cylinders and spheres without ghost nodes.
- Quantitative benchmarks—velocity profiles, finger length, width, slip distance in displacement, and droplet height-to-diameter ratios—stay within about 5% of analytic or reference results.
Reading between the lines
- Editorial inference: the same local-adhesion construction could be ported to other diffuse-interface lattice Boltzmann variants by replacing the constant wall coupling with a function of the local order parameter.
- Editorial inference: the claimed complete elimination of the mass-transfer layer is demonstrated at one interface width ($\delta=0.8$, $G_{12}=0.17$); retuning $\delta$ for other surface tensions or density ratios would be needed to confirm universality.
- Editorial inference: the copying rule for second-layer nodes is plausible but untested for high curvature; a grid-convergence study varying cylinder radius would show whether the boundary reconstruction itself introduces a curvature-dependent error.
- Editorial inference: the operating-range comparison across schemes is made on static equilibrium droplets; dynamic contact angles under strong flow or high capillary number may reveal larger differences than the static benchmarks.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript proposes an improved solid-fluid interaction scheme for immiscible pseudopotential lattice Boltzmann simulations. Instead of a constant adhesion parameter, Eqs. (11) and (12) define a local interaction parameter based on the normalized effective density, with piecewise interpolation controlled by a threshold δ = 0.8. The scheme is combined with an eighth-order isotropic interaction stencil and a boundary treatment that copies first-layer fluid-node values to missing second-layer nodes at curved interfaces (Sec. II.D). The proposed scheme is tested on four benchmarks: co-current Poiseuille and Couette flow, static contact angles on cylindrical and spherical surfaces, droplet dynamics past a square obstacle in a microchannel, and immiscible displacement in a channel and in a porous-media-like geometry. The central claims are that the scheme eliminates the unphysical mass-transfer layer near solid boundaries, extends the accessible contact-angle range to roughly 10°–165°, and reduces spurious currents relative to the original scheme.
Significance. If the claims are substantiated, this is a useful contribution to pseudopotential LBM practice: the scheme is simple, uses a single calibration parameter, and is benchmarked against analytical velocity profiles (relative error below 1.84%), against Kang et al.'s displacement data (errors below 5%), and against a dedicated experiment (mean absolute deviation 2.1°). The manuscript also provides implementation details in an appendix and a source-code link. The main weakness is that the headline claims are measured on curved geometries where the boundary treatment relies on a copying rule that is not subjected to any sensitivity or convergence test; the significance is therefore conditional on additional evidence.
major comments (4)
- [Section II.D, Fig. 2(b3), Table I] The curved-boundary copying rule is load-bearing for the paper's central claims. With the eighth-order stencil, the interaction range needs data on second-layer nodes outside the fluid, and the paper fills those nodes by copying information from corresponding first-layer nodes. This is a local-uniformity assumption over two lattice spacings, and it is used in all curved-surface benchmarks (Secs. III.B and III.E). The manuscript gives no grid-convergence study, no variation of cylinder or sphere radius at fixed resolution, and no comparison with an independent boundary closure such as ghost nodes or interpolation. Because the contact-angle range (10–165°) and the claimed elimination of the mass-transfer layer are measured on curved geometries, the absence of a sensitivity test is load-bearing; either add such tests or qualify the headline claims.
- [Section III.B, Eq. (11) and the paragraph defining δ] The threshold δ = 0.8 is hand-set by matching an interface width of about five lattice grids, but no δ-sensitivity study is reported. Since δ controls the width of the piecewise interpolation interval for the local adhesion parameter, the reported contact-angle range and the suppression of the mass-transfer layer could depend on this choice. A short sweep of δ at a fixed λ, or a stated criterion for selecting δ, is needed to show that the results are robust.
- [Section III.B, Fig. 5(b)] The correlation θ = -13.76λ² + 94.62λ + 9.47 is presented as a predictive tool, but it is a polynomial fit to the authors' own simulation data. The paper should state explicitly that this is an empirical fit, report its fit quality and range of validity, and avoid giving the impression that the 10–165° range is a parameter-free consequence of the scheme. This matters because Table III uses the range as a headline comparison.
- [Section III.B, Fig. 5(a), and Section IV] The claim of 'complete elimination' of unphysical mass-transfer layers is supported only by visual inspection of density contour plots. No quantitative metric (such as thickness or amplitude of the dissolved-density layer, or a mass-flux measure) is defined, and Fig. 8 shows localized spurious-current maxima at the contact line. Please define a measurable criterion for the mass-transfer layer and quantify the reduction, or soften the claim to the demonstrated level of suppression.
minor comments (5)
- [Throughout] There are numerous typographical errors and garbled equations (for example, Eq. (1) has missing superscripts, Eq. (11) contains unclear definitions of the s and t coefficients, and the text includes 'boudnary', 'V olume of Fluid', and 'k ∈ 11, 2}'). A careful proofread is needed.
- [Section III.E] The text states a 'solid cylinder of radius R=50' for the spherical-surface test; this should read 'sphere'.
- [Table III] Table III lists Li's scheme as also achieving a 10–165° contact-angle range, so the improved scheme's claimed extension of the operating range is relative to the original scheme only; the text should make this comparison explicit.
- [Abstract and Section IV] The abstract and introduction state a 33% reduction in average spurious currents, while the conclusion states 40–50% in hydrophilic regimes and 30–40% in hydrophobic regimes; these numbers should be reconciled or the conditions for each figure stated.
- [Section III.B, Fig. 7] The experimental comparison is limited to flat substrates, whereas the numerical benchmarks are on curved surfaces; the paper should note explicitly that the experimental validation does not exercise the curved-boundary copying rule.
Circularity Check
No significant circularity; only the lambda-contact-angle calibration fit is over-labelled 'predictive', and the curved-boundary copying rule is an unvalidated assumption rather than a circular step.
-
fitted input called prediction
[Section III.B, Fig. 5(b), paragraph beginning 'To facilitate practical implementation...']
"To facilitate practical implementation, we also established a predictive correlation between contact angles and solid-fluid interaction strength parameters, 213.76 94.62 9.47θ λ λ= − ⋅ + ⋅ + , as shown in Fig. 5(b)."
The quadratic is obtained by fitting the simulated static contact angles shown in Fig. 5(b) for the λ values the authors chose; it is therefore a calibration curve, not an independent prediction. Any later simulation that sets λ from this curve for a target contact angle (e.g., the 30.2° case in Sec. III.C) will reproduce the fit's own data. The paper's central claims — mass-transfer-layer elimination, extended range, and spurious-current reduction — are direct observations of the scheme and do not depend on the fit, so this is a minor overstatement rather than the core of the paper.
full rationale
The proposed solid-fluid interaction scheme is an explicitly constructed ansatz (Eqs. 10-12), not a derivation from first principles, so the relevant question is whether its validation is circular. The central claims are checked against independent evidence: analytical velocity profiles for Poiseuille and Couette flow (Sec. III.A, relative errors below 1.84%), experimental flat-substrate contact angles (Sec. III.B, mean absolute deviation 2.1°), and prior numerical results from Kang et al. and Li's scheme (Secs. III.D and III.E). These benchmarks are external to the fitted λ-θ curve, so the main results are not true by construction. No load-bearing self-citation was found: references to Li et al. and Huang et al. are to external prior work, not to the present authors. The main unquantified risk is the second-layer copying rule at curved boundaries (Sec. II.D, Fig. 2(b3)): the paper states 'curved boundaries could not be validated experimentally under the present conditions' and gives no grid-convergence or curvature-sensitivity test, so the cylindrical and spherical contact-angle results rest on a local-uniformity assumption. That is a robustness/correctness gap, not a circularity. The only circular element is the over-calling of a fitted calibration curve as 'predictive,' which is minor and does not infect the paper's central comparisons.
Assumptions & free parameters
free parameters (3)
- delta (effective-density threshold) =
0.8
- G12 (fluid-fluid interaction strength) =
0.17
- theta-lambda calibration polynomial coefficients =
-13.76, 94.62, 9.47
assumptions (5)
- standard math Standard single-relaxation-time LBM with the Guo forcing scheme.
- domain assumption Immiscible pseudopotential model with a diffuse interface and 'dissolved' density.
- domain assumption Near-unity density ratio is sufficient for the intended applications.
- ad hoc to paper Copying first-layer node data to second-layer nodes is valid at curved boundaries.
- ad hoc to paper The piecewise interpolation functions and threshold delta are appropriate.
Cite this review
Pith. "Pith review of Improvement of Solid-Fluid Interaction Scheme in Lattice Boltzmann Immiscible Pseudopotential Models." pith.science (2026). https://pith.science/paper/JY7YOBNQ
@misc{pith2026250522256,
author = {Pith},
title = {Pith review of: Improvement of Solid-Fluid Interaction Scheme in Lattice Boltzmann Immiscible Pseudopotential Models},
year = {2026},
howpublished = {\url{https://pith.science/paper/JY7YOBNQ}},
note = {Machine review of arXiv:2505.22256}
}
read the original abstract
The pseudopotential model within the Lattice Boltzmann Method (LBM) framework has emerged as a prominent approach in computational fluid dynamics due to its dual strengths in physical intuitiveness and computational tractability. However, when modeling wettability phenomenon, existing solid-fluid interaction schemes exhibit persistent challenges in multi-component immiscible fluid systems, notably manifested through spurious velocity artifacts and unphysical mass-transfer boundary layers. This study presents an improved interaction scheme that preserves implementation simplicity while effectively mitigating these numerical artifacts. Furthermore, leveraging the enhanced isotropy characteristics of eighth-order discrete schemes, we develop a novel boundary treatment methodology addressing second-layer lattice data reconstruction at complex interfaces. To verify the universality of the proposed optimization scheme, four benchmark scenarios, including static contact angle measurement on cylindrical surface, droplet dynamics through confined geometries, immiscible displacement processes, and co-current flow in microchannels, are simulated to demonstrate the proposed scheme's capability. The results show that the improved scheme can well simulate various complex immiscible multiphase flows.
Figures
Figures from the paper (1 more)
Reference graph
Works this paper leans on
-
[1]
1 H. Deng, K. Jiao, Y . Hou, J.W. Park, and Q. Du, “A lattice Boltzmann model for multi-component two-phase gas-liquid flow with realistic fluid properties,” Int. J. Heat Mass Transf. 128, 536–549 (2019). 2 S. Khajepor, J. Cui, M. Dewar, and B. Chen, “A study of wall boundary conditions in pseudopotential lattice Boltzmann models,” Comput. Fluids 193, 103...
work page 2019
Reviewed August 7, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.