{"id":"46cd0655-5daa-4a69-9ee0-8d97844717c8","arxiv_id":"2608.05787","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A lattice Boltzmann model for three-phase dielectric fluid flows is derived from the Onsager variational principle and validated against analytical benchmarks.","lead":"Three-phase electrohydrodynamic flows, where charged droplets move in electric fields, get a new simulation model built on the Onsager variational principle. The authors also build a lattice Boltzmann solver and test it on benchmarks like droplet deformation and coalescence, aiming at a thermodynamically consistent approach for engineering applications.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The chemical potential in Eq. (2.27c) is not the variational derivative of the free energy in Eq. (2.18), so the claimed Onsager-based Cahn-Hilliard dynamics do not follow from the stated free energy.","rationale":"The reader's formally named weakest assumption is the ad hoc charge-diffusion energy F_q = (λ/2)∫q² in Eq. (2.12). That is a legitimate modeling concern, but it is not the most load-bearing point: even if F_q were physically justified, the derivation still fails at an elementary algebraic step. The chemical potential used in the Cahn-Hilliard equation and in the LB solver is not the variational derivative of the stated free energy. Eq. (2.20) is the paper's own equilibrium condition obtained from variation, and it contains the derivative form φ_i(φ_i-1/2)(φ_i-1) plus a -∇²φ_i gradient term. Eq. (2.27c) replaces this with φ_i²(1-φ_i)² and +|∇φ_i|², which is not equivalent. Since the numerical method in Section 3.1 is built directly on Eq. (2.27c), the actual simulated dynamics are not guaranteed to descend the claimed free energy, so the central thermodynamic-consistency result is not supported as written. This is checkable by direct differentiation and is independent of any dispute about the charge-diffusion term. I agree with the reader's overall CONDITIONAL assessment because the defect appears repairable - likely a typographical or transcription error, since the benchmarks would probably not agree if the code used the wrong chemical potential - but the manuscript must state the correct μ_i and confirm the LB implementation uses it. I therefore leave the verdict unchanged while strengthening the specific technical basis for the condition.","tokens_in":21357,"tokens_out":11705,"duration_ms":108716,"concrete_test":"Re-derive the variational derivative of F in Eq. (2.18) term by term and compare it with Eq. (2.27c). In particular, compute δ/δφ_i[Σ (6/D)λ_k φ_k²(1-φ_k)² + (3/8)Dλ_k|∇φ_k|²] together with the electric term. If the resulting μ_i^var differs from Eq. (2.27c) - as it does for the bulk and gradient terms - replace Eq. (2.27c) with the correct μ_i^var, rerun the Section 4.2 static compound-droplet convergence test and the Section 4.4 deformation benchmarks, and check whether the reported errors and analytical agreement change. If the results change materially, the thermodynamic-consistency claim and the validation require revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that Eqs. (2.27) are strictly derived from the free energy (2.18) via the Onsager variational principle, so that the Cahn-Hilliard equation (2.27c) is thermodynamically consistent. For this to hold, the μ_i in (2.27c) must be the variational derivative of F. Differentiating (2.18) with respect to φ_i gives μ_i^var = (24/D)λ_i φ_i(φ_i-1/2)(φ_i-1) - (3/4)Dλ_i ∇²φ_i - (1/2)ε'(φ_i)|E|², the same expression that appears as the equilibrium condition in Eq. (2.20). Eq. (2.27c) instead states μ_i = (24/D)λ_i φ_i²(1-φ_i)² + (3/4)Dλ_i|∇φ_i|² - (1/2)ε'(φ_i)|E|². The bulk term is four times the bulk energy density rather than its derivative, and the gradient term has the wrong sign and is an energy density rather than a Laplacian. The LB solver in Section 3.1 uses this μ_i, so the numerical model solves a different Cahn-Hilliard equation than the one implied by the free energy. This is a direct break in the derivation chain, independent of the ad hoc charge-diffusion term. The omission of the -1/2∇(α/σq²) term between (2.26) and (2.27b) is a further inconsistency, though it could be absorbed into pressure; the chemical potential mismatch cannot.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper proposes a phase-field lattice Boltzmann model for three-phase dielectric electrohydrodynamic (EHD) flows, with the central claim that the governing equations are derived strictly from Onsager's variational principle and therefore thermodynamically consistent. The total free energy combines ternary Cahn–Hilliard, electrostatic, kinetic, and charge-diffusion contributions; minimization of the Rayleighian yields constitutive relations for the phase fluxes, conduction current, and body force. The authors then develop LB solvers for the Cahn–Hilliard, Navier–Stokes, electric-potential, and charge-transport equations, and validate the framework against analytical solutions for electro-osmotic flow, static compound droplets, liquid-lens spreading, and compound-droplet deformation in electric fields. The validated framework is applied to double-droplet coalescence/separation under electric fields and to combined EHD and shear flow.","tokens_in":21719,"tokens_out":13771,"duration_ms":123441,"significance":"If the derivation were correct, the model would be a valuable contribution because it addresses three-phase thermodynamics and surface charge convection within a single variational framework, and the LB implementation offers a potentially scalable numerical route. The paper's strengths include the breadth of benchmark comparisons against external analytical solutions and the reduction-consistency construction of the mobility matrix. However, the manuscript contains a central mathematical inconsistency in the chemical potential used in the final governing equations and LB solver, which breaks the claimed strict derivation. In addition, the charge-diffusion term is introduced ad hoc and is not calibrated, and a key validation benchmark does not exercise the proposed charge-transport mechanism. The contribution is therefore conditional on correcting the variational derivation and re-validating the numerical results.","major_comments":[{"comment":"Eq. (2.27c) does not follow from the free energy (2.18). The variational derivative of (2.18) with respect to φ_i is μ_i^{var} = (24/D)λ_i φ_i(φ_i−1/2)(φ_i−1) − (3/4)Dλ_i ∇²φ_i − (1/2)ε′(φ_i)|E|², which is also the expression appearing in the equilibrium condition (2.20). Eq. (2.27c) instead states μ_i = (24/D)λ_i φ_i²(1−φ_i)² + (3/4)Dλ_i|∇φ_i|² − (1/2)ε′(φ_i)|E|². The bulk term is four times the double-well energy density rather than its derivative, and the gradient term has the wrong sign and is a gradient energy density rather than a Laplacian. Since the LB solver in §3.1 uses this μ_i, the discretized Cahn–Hilliard system is not the variational model derived from (2.18). This is a direct break in the derivation chain and undermines the paper's central claim of strict thermodynamic consistency.","section":"§2.2, Eqs. (2.18), (2.20), (2.27c)"},{"comment":"The constitutive relation (2.26) for the body force F includes the term −(1/2)∇(α/σ q²), but the momentum equation (2.27b) omits this term. Because the term is a pure gradient, it can be absorbed into the pressure by redefining p; however, the manuscript does not state this, so the derivation as written is internally inconsistent. If the absorption is intended, it should be stated explicitly together with the corresponding redefinition of pressure.","section":"§2.2, Eqs. (2.26) and (2.27b)"},{"comment":"Eq. (2.12) introduces the charge diffusion free energy F_q = (λ/2)∫q² dΩ with an unspecified coefficient λ. This term is not derived from any microscopic or continuum argument, and the resulting charge diffusivity α = σλ in Eq. (2.27e) is never calibrated or reported in the numerical sections. The abstract's claim that the model is derived 'strictly' from the Onsager principle 'without requiring a priori assumptions' is therefore overstated: the thermodynamic consistency of the charge transport equation depends on this hand-inserted energy contribution. The authors should either provide a physical derivation or calibration for F_q, or substantially temper the claim.","section":"§2.2, Eq. (2.12) and §4"},{"comment":"Eq. (2.20) is obtained by minimizing F with respect to φ_i as if the three order parameters were independent, but the model is defined under the constraint (2.5), φ_1+φ_2+φ_3=1. The correct equilibrium condition requires a Lagrange multiplier for this constraint, or an explicit statement that the common part of the μ_i is irrelevant because the mobility matrix has zero row sums. As written, μ_i ≡ const for each phase is not the constrained minimizer of (2.18). This issue propagates into the chemical potentials and Cahn–Hilliard fluxes in Eqs. (2.26)–(2.27c) and should be clarified.","section":"§2.2, Eq. (2.20)"},{"comment":"The validation does not actually exercise the surface-charge-convection mechanism claimed in the abstract. The electro-osmotic benchmark in §4.1 solves the Nernst–Planck equation (4.2) for ionic concentrations rather than the proposed charge transport model (2.27e)/(3.17)–(3.21). The compound-droplet deformation benchmarks in §4.4 compare against small-deformation analytical solutions (4.9)–(4.11), which are based on the leaky-dielectric approximation and do not include charge convection; those comparisons therefore cannot establish that the new convection terms are correctly captured. A benchmark with a finite electric Reynolds number, or an explicit comparison of charge transport with convection, is needed.","section":"§4.1 and §4.4"}],"minor_comments":[{"comment":"The notation '∇2ϕ)' should be '∇²φ_i'; the subscript and closing parenthesis are missing in the printed equation.","section":"§2.2, Eq. (2.19)"},{"comment":"The global relative errors in Table 1 are internally inconsistent: at δx=1/512, Err(φ1)=8.99×10⁻² is larger than at δx=1/256 (2.76×10⁻²), and the reported rates do not follow from the tabled values. For example, log2(2.76×10⁻²/8.99×10⁻²) is negative, not 1.62, and log2(3.40×10⁻²/1.03×10⁻³) is about 5.0, not 1.73. Please correct the table and recompute the convergence rates.","section":"§4.2, Table 1"},{"comment":"Eq. (3.6) ends with a stray ' , .' and the formula for F_i^j has an extra closing bracket; these LaTeX/punctuation errors should be cleaned up.","section":"§3.1, Eq. (3.6)"},{"comment":"The label 'Prensent' in Fig. 3 should be 'Present'.","section":"§4.2, Fig. 3"},{"comment":"The phrase 'computer techology' should be 'computer technology', and there are several other typographical errors throughout the introduction that should be corrected.","section":"§1"},{"comment":"The symbols Γ, ζ, Π, and Λ_A−H are said to be documented in Ref. [13], but the paper should state at least their physical meaning or range to make the comparison self-contained for readers.","section":"§4.4, Eqs. (4.9)–(4.11)"}],"recommendation":"major_revision","confidential_remarks":"The chemical potential inconsistency in §2.2 is the key obstacle. It is fixable in principle, but if the LB code actually uses the printed μ_i from Eq. (2.27c), the benchmark results will need to be regenerated after correcting the model, because the solved Cahn–Hilliard equation differs from the one implied by the free energy. I also note that the coalescence/separation simulations in §4.5 are not compared with experimental or independent reference data, so they are illustrative rather than validated. The paper fits the journal's scope, but the 'strictly derived' claim needs to be made accurate before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Candid take: this paper is a promising but broken-on-submission attempt. The Onsager route for three-phase EHD with an LB solver is genuinely new, and the benchmarks against electroosmotic flow, liquid lens, and compound droplet deformation are the right tests. But the derivation chain does not close as written, and one table is self-contradictory.\n\nThe new material is real: the reduction-consistent mobility matrix of Eq. (2.28), the charge-diffusion contribution, and the first LB implementation of the Onsager-based three-phase model. The paper also honestly cites Ref. 29 for the existing thermodynamically consistent three-phase EHD model and Ref. 43 for their own two-phase LB work. That framing is accurate.\n\nNow the soft spots, in order of severity. (1) The chemical potential in Eq. (2.27c) is not the variational derivative of the free energy (2.18). Equation (2.20) gives the correct derivative: (24/D)λ_i φ_i(φ_i−1/2)(φ_i−1) − (3/4)Dλ_i∇²φ_i − (1/2)ε'|E|². Equation (2.27c) instead has the bulk double-well energy density plus a positive gradient energy density. Unless the LB solver secretly uses the correct form, which the text never states, the numerical scheme solves a different Cahn-Hilliard equation than the one derived. That breaks the 'strictly derived' claim. (2) The body force in Eq. (2.26) contains −1/2∇(α/σ q²), which disappears from Eq. (2.27b) without comment. It might be absorbed into pressure, but the text should say so. (3) Table 1 is internally inconsistent: Err(ϕ1) at δx=1/512 is 8.99e−2, larger than at δx=1/256 (2.76e−2), yet the listed rate is 1.62. Either the table is misprinted or the claimed second-order convergence is not supported. (4) The charge diffusion energy F_q = (λ/2)∫q² dΩ is ad hoc; no physical derivation or calibration is given, and the thermodynamic-consistency narrative leans on it.\n\nIf these are typos, the paper could be repaired. If not, the model as written is not the model implemented. Either way, a serious referee can help the authors sort it out. I would send it to peer review rather than desk reject, because the architecture and benchmark design are worth the effort. But it should not be accepted in this form.","headline":"Promising three-phase EHD/LB framework, but the Onsager derivation breaks in Eq. (2.27c) and the convergence table is inconsistent—worth a careful revision, not desk rejection.","tokens_in":22269,"tokens_out":3290,"would_cite":false,"duration_ms":29759,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["76T30","76M28","49S05"],"pacs":[],"model":"deepseek-v4-flash","headline":"One variational principle yields consistent three-phase EHD equations.","keywords":["Lattice Boltzmann method","electrohydrodynamics","three-phase flows","Onsager variational principle","phase-field method","surface charge convection","compound droplet","Cahn-Hilliard equation"],"falsifier":"A decisive check would be to compute the total free energy rate along a closed-domain simulation trajectory under Eq. (2.27); if $\\dot{F} > 0$ ever occurs, the discrete solution contradicts the claimed dissipation guarantee. Alternatively, at high electric Reynolds number where surface charge convection is strong, comparing the predicted surface charge distribution and flow direction with high-fidelity experimental or direct numerical data would settle whether convection is captured without ad hoc assumptions.","tokens_in":21094,"feed_emoji":"⚡","tokens_out":6738,"duration_ms":63545,"temperature":0.7,"pith_summary":"This paper tries to show that the coupled equations for three immiscible dielectric fluids under an electric field can be derived by minimizing a single variational objective, so that thermodynamic consistency and surface charge convection are built into the model rather than added by hand. It then seeks to demonstrate that these equations can be solved accurately by a lattice Boltzmann method that tracks interfaces and charge transport on a mesoscopic grid. If correct, the result is a numerical framework for three-phase electrohydrodynamic flows that respects energy dissipation and can handle compound droplets, coalescence, separation, and combined electric-shear effects without phenomenological closure assumptions. This matters because existing three-phase EHD simulations mostly couple the fields phenomenologically and often omit charge convection, which becomes significant at high electric Reynolds numbers.","feed_headline":"One variational principle yields consistent three-phase EHD equations.","feed_subtitle":"A lattice Boltzmann solver built from the same derivation captures surface charge and reproduces droplet benchmarks.","key_machinery":"The load-bearing object is the Rayleighian $\\mathcal{R} = \\dot{F} + \\Phi_F$, where $F$ is the total free energy of Eq. (2.6) (chemical, electrostatic, kinetic, and charge-diffusion contributions) and $\\Phi_F$ is the dissipation function of Eq. (2.24). Minimizing $\\mathcal{R}$ with respect to the independent fluxes $u$, $J_{\\phi_i}$, and $J_D$ yields the constitutive relations (2.26) and closes the governing system (2.27). The charge-diffusion term $F_q = \\frac{\\lambda}{2}\\int_\\Omega q^2\\,d\\Omega$ is what endows the charge transport equation with the effective diffusion coefficient $\\alpha$; without it, the Nernst-Planck-type equation would lack the $\\nabla(\\alpha\\nabla q)$ contribution. The numerical machinery is a lattice Boltzmann scheme with four distribution functions, using second-order isotropic finite differences for gradients and a permittivity interpolation that keeps the electric body force regular across diffuse interfaces.","core_discovery":"The central discovery is a closed system of governing equations for three-phase dielectric EHD flows obtained by minimizing a Rayleighian with respect to the fluid velocity, the phase-field fluxes, and the conduction current. That minimization produces the constitutive relations of Eq. (2.26): the body force as a sum of chemical-potential, Coulomb, and dielectric-gradient forces, the phase fluxes as an Onsager mobility matrix applied to chemical-potential gradients, and the conduction current augmented by a charge-diffusion contribution. Substituting these relations into the Navier-Stokes, Cahn-Hilliard, Poisson, and charge-transport equations yields the coupled system (2.27), which the paper argues intrinsically satisfies the second law and reproduces the two-phase EHD model when one phase vanishes. The accompanying lattice Boltzmann solver uses four distribution functions to solve the Cahn-Hilliard, hydrodynamic, electrostatic, and Nernst-Planck subproblems, and the paper reports that it matches analytical electroosmotic profiles, equilibrium compound-droplet shapes, lens spreading geometries, and small-deformation compound-droplet deformation factors.","pith_inferences":["If the charge-diffusion energy term is accepted as a genuine free-energy contribution, the coefficient $\\lambda$ (equivalently the effective charge diffusivity $\\alpha$) becomes a free parameter that the paper leaves uncalibrated; a natural extension is to fit it against measured transient charge relaxation in three-phase droplets.","Because the derivation is variational rather than tied to a particular discretization, the same Onsager model could in principle be solved by finite-difference, finite-element, or spectral methods; the lattice Boltzmann solver is one numerical vehicle, not the only one.","The construction suggests a broader design rule: any additional dissipative or conservative mechanism in a multiphase EHD system could be added by inserting its free-energy contribution and dissipation rate into the Rayleighian, provided a reduction-consistent mobility matrix is maintained, although the paper does not pursue that extension."],"forward_implications":["The governing system (2.27) is closed without phenomenological coupling assumptions, so surface charge convection is represented explicitly and should remain valid at arbitrary electric Reynolds numbers.","The reduction-consistency property means that whenever one phase vanishes, the model collapses to the established two-phase EHD phase-field equations, providing a path for benchmark comparisons across models.","The lattice Boltzmann solver reproduces analytical electroosmotic velocity and potential profiles, equilibrium compound droplet shapes, and lens contact angles and spreading lengths to within a few percent.","For compound droplets in a uniform field, the model predicts three distinct deformation regimes, matching small-deformation analytical solutions for each combination of permittivity and conductivity ratios.","Simulations of two compound droplets show electric-field-induced coalescence when the permittivity ratio exceeds the conductivity ratio and separation when the reverse holds, with outer shells and inner cores merging sequentially in the coalescence case."],"supporting_citations":[{"why":"Supplies the Onsager variational framework and the two-phase phase-field EHD model to which the present three-phase model reduces.","marker":"[30]"},{"why":"Provides the prior thermodynamically consistent three-phase EHD model based on energy dissipation, serving as the main comparison point for a variational alternative.","marker":"[29]"},{"why":"Establishes Onsager's reciprocal relations and the thermodynamic foundation on which the Rayleighian construction rests.","marker":"[42]"},{"why":"Gives the analytical small-deformation deformation factors and characteristic functions used to validate the compound droplet deformation benchmarks.","marker":"[13]"},{"why":"Supports the thermodynamically consistent two-phase phase-field lattice Boltzmann method that the present three-phase reduction-consistency claim extends.","marker":"[43]"},{"why":"Motivates the inclusion of surface charge convection in the charge transport equation and provides the reformulated charge transport form used in the solver.","marker":"[28]"},{"why":"Supplies the phase-field lattice Boltzmann equilibrium and forcing formulations adapted for the hydrodynamic and Cahn-Hilliard solvers.","marker":"[50]"}],"fun_headline_variants":["Onsager principle yields consistent 3-phase EHD equations","Single variational principle captures 3-phase EHD multiphysics","LBM for 3-phase EHD built on Onsager variational principle","Three-phase EHD equations from one Rayleighian minimization"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The thermodynamic-consistency claim rests on treating the added charge-diffusion term $F_q = \\frac{\\lambda}{2}\\int_\\Omega q^2\\,d\\Omega$ as a legitimate free-energy contribution, even though the paper does not derive or calibrate the coefficient $\\lambda$; if that term is not a true free energy, the 'strict' Onsager derivation no longer underpins the charge-transport equation.","fun_headline_variants_meta":{"raw":{"variants":["Onsager principle yields consistent 3-phase EHD equations","Single variational principle captures 3-phase EHD multiphysics","LBM for 3-phase EHD built on Onsager variational principle","Three-phase EHD equations from one Rayleighian minimization"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000231,"raw_usage":{"total_tokens":1506,"prompt_tokens":989,"completion_tokens":517,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":605,"completion_tokens_details":{"reasoning_tokens":445}},"tokens_in":605,"tokens_out":517,"duration_ms":4998,"temperature":1.0,"reasoning_tokens":445,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T23:30:18.771535+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A decisive check would be to compute the total free energy rate along a closed-domain simulation trajectory under Eq. (2.27); if $\\dot{F} > 0$ ever occurs, the discrete solution contradicts the claimed dissipation guarantee. Alternatively, at high electric Reynolds number where surface charge convection is strong, comparing the predicted surface charge distribution and flow direction with high-fidelity experimental or direct numerical data would settle whether convection is captured without ad hoc assumptions.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Onsager variational framework and the two-phase phase-field EHD model to which the present three-phase model reduces."},{"cited_title":"Zhang, F","cited_arxiv_id":null,"evidence_quote":"Provides the prior thermodynamically consistent three-phase EHD model based on energy dissipation, serving as the main comparison point for a variational alternative."},{"cited_title":"Onsager,Reciprocal relations in irreversible processes","cited_arxiv_id":null,"evidence_quote":"Establishes Onsager's reciprocal relations and the thermodynamic foundation on which the Rayleighian construction rests."},{"cited_title":"Behjatian andA","cited_arxiv_id":null,"evidence_quote":"Gives the analytical small-deformation deformation factors and characteristic functions used to validate the compound droplet deformation benchmarks."},{"cited_title":"Xiong, L","cited_arxiv_id":null,"evidence_quote":"Supports the thermodynamically consistent two-phase phase-field lattice Boltzmann method that the present three-phase reduction-consistency claim extends."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Motivates the inclusion of surface charge convection in the charge transport equation and provides the reformulated charge transport form used in the solver."},{"cited_title":"Liang, B","cited_arxiv_id":null,"evidence_quote":"Supplies the phase-field lattice Boltzmann equilibrium and forcing formulations adapted for the hydrodynamic and Cahn-Hilliard solvers."}],"review_version":1}