{"id":"a322e641-ae0a-4ec4-af4f-7ab7e4e786ca","arxiv_id":"2507.04760","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For 3D compressible liquid crystal flows with density-dependent viscosity, global strong solutions exist for large initial density and small director gradient, with no smallness condition on velocity.","lead":"This paper proves that a 3D compressible liquid crystal flow has a unique global strong solution when the initial density is high, the initial director gradient is small, and viscosity grows faster than linearly with density. The result removes the usual smallness condition on velocity, which is a notable step for a well-posedness question in nonlinear partial differential equations.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 3.7's density equation (3.81) has the wrong sign on ∇P; as written it cannot produce the dissipation in (3.82), so the density bounds that close the bootstrap are not justified.","rationale":"The reader's verdict identified Lemma 2.1 as the weakest assumption. I agree that local existence is not proved and is a genuine gap, but the paper's own text contains a more acute, checkable inconsistency in the density estimates. Equation (3.81) is the starting point of Lemma 3.7, which provides E_{ρ,2}, E_{ρ,3} and ultimately E_{ρ,1}. The sign of ∇P in (3.81) is opposite to what follows from the definition of F. If taken literally, the 'dissipation' in (3.82) would be a growth term, invalidating the Gronwall argument. Since the bootstrap Proposition 3.1 needs the improved density bounds (3.9) to run the continuation, and Theorem 1.1's density bound (1.12) is one of its conclusions, this is load-bearing. The proposed check is a short algebraic re-derivation and can settle the issue immediately. I keep the reader's CONDITIONAL verdict because both issues are fixable in principle, but the sign inconsistency is the more decisive one for the proof as written. I therefore partially agree with the reader: local existence is a concern, but the density-estimate sign error is the single most load-bearing spot.","tokens_in":23547,"tokens_out":25179,"duration_ms":231236,"concrete_test":"Independently re-derive (3.81) from (1.1)_1 and F=(2µ1+µ2)divu+P(ρ)-P(¯ρ), and record whether the ∇P coefficient is +ρ/(2µ1+µ2) or -ρ/(2µ1+µ2). Then re-derive (3.82) and check that the dissipative term has the correct sign and the exponent γ-α, not γ. If the correct sign is negative and the exponent is γ-α, the error is typographical and Lemma 3.7 can be repaired; if the sign is positive or the exponent is wrong, the density-gradient bound is not available and the proof of Proposition 3.1 does not close.","verdict_should_be":"UNCHANGED","load_bearing_attack":"From (1.1)_1, ρ_t+div(ρu)=0, differentiation gives ∇ρ_t+u·∇^2ρ+∇u·∇ρ+∇ρ divu+ρ∇divu=0. With F=(2µ1+µ2)divu+P(ρ)-P(¯ρ) in (3.35), one has divu=(F-P+Pbar)/(2µ1+µ2), hence ∇divu=(∇F-∇P)/(2µ1+µ2). Therefore the last term in the density equation is +ρ(∇F-∇P)/(2µ1+µ2), so the coefficient of ∇P is minus, not plus as printed in (3.81). Multiplying the correct equation by |∇ρ|^{q-2}∇ρ and integrating yields the positive term aγ/(2µ+λ)∫ρ^{γ-α}|∇ρ|^q only after moving -ρ∇P to the left side; with the printed '+' sign, the same manipulation produces a positive source on the right and no dissipation. Lemma 3.7 uses exactly this dissipation to derive (3.83)-(3.87) and then (3.92), which gives E_{ρ,1}(T) ≤ ¯ρ/4. Without it, the improved density bounds in Proposition 3.1 fail and the continuation in Section 4 has no restarting bounds. Even if this is a one-character typo, the manuscript as submitted contains an algebraically inconsistent key estimate and must be corrected and rechecked before the global theorem is supported.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the Cauchy problem for the three-dimensional compressible simplified Ericksen-Leslie system with density-dependent viscosities $\\mu_i(\\rho)=\\mu_i\\rho^\\alpha$. The main result, Theorem 1.1, asserts that for $\\alpha>1$, $\\gamma>1$, $\\alpha>(\\gamma+1)/2$, $\\alpha\\ge \\gamma-1$, and initial data satisfying (1.10), there exist thresholds $\\Lambda_0$ and $\\varepsilon_0$ such that if the far-field density $\\bar\\rho$ is sufficiently large and $\\|\\nabla d_0\\|_{L^3}$ is sufficiently small, then the system admits a unique global strong solution with no smallness assumption on the initial velocity. The proof is a bootstrap argument organized in Section 3: estimates for the director (Lemmas 3.1--3.3), for the velocity (Lemmas 3.4--3.6), and for the density (Lemma 3.7), followed by the continuation argument in Section 4.","tokens_in":23825,"tokens_out":18013,"duration_ms":169915,"significance":"If the theorem is correct, it would be the first global strong solution result for three-dimensional compressible liquid crystal flows with density-dependent viscosity and large velocity. The underlying mechanism---that for $\\alpha>1$ large density makes the effective Reynolds number small regardless of the flow speed---is natural and the bootstrap structure is coherent. The paper also gives explicit dependence of the thresholds on the initial data and attempts to keep the director smallness in the critical space $L^3$. However, the proof as written contains load-bearing algebraic errors, most seriously in the density estimate of Lemma 3.7, and several definitions and identities in Lemma 3.4 are not self-consistent. No machine-checked proofs are supplied; the claims rest on hand-written estimates that must be corrected and re-verified.","major_comments":[{"comment":"The gradient of the continuity equation is computed with the wrong sign. From (1.1)_1 one obtains $\\nabla\\rho_t+u\\cdot\\nabla^2\\rho+\\nabla u\\cdot\\nabla\\rho+\\nabla\\rho\\,\\mathrm{div}\\,u+\\rho\\nabla\\,\\mathrm{div}\\,u=0$. With $F=(2\\mu_1+\\mu_2)\\mathrm{div}\\,u+P(\\rho)-P(\\bar\\rho)$ as in (3.35), the relation $\\nabla\\,\\mathrm{div}\\,u=(\\nabla F-\\nabla P)/(2\\mu_1+\\mu_2)$ holds, so the last term in (3.81) should contain $\\nabla F-\\nabla P$, not $\\nabla F+\\nabla P$. With the printed plus sign, multiplication by $|\\nabla\\rho|^{q-2}\\nabla\\rho$ and integration produce a positive source $a\\gamma/(2\\mu+\\lambda)\\int\\rho^{\\gamma-\\alpha}|\\nabla\\rho|^q\\,dx$ instead of the dissipation used in (3.82). Since (3.83)--(3.87), (3.92), and the improved density bounds in Proposition 3.1 all rely on this dissipation, the sign error is load-bearing and the estimate must be corrected before the continuation argument can be accepted.","section":"Lemma 3.7, Eq. (3.81)"},{"comment":"Local well-posedness is stated without proof and attributed to [37] via \"similar arguments\". Because the viscosities here are $\\mu_i(\\rho)=\\mu_i\\rho^\\alpha$ with $\\alpha>1$ and the continuation argument in Section 4 restarts the solution from the improved bounds (3.9), the transfer of the local existence result to this exact system is a load-bearing premise. The manuscript should either prove Lemma 2.1 or give a precise reference that explicitly covers the present system, including the density-dependent viscosity and the director equation. In addition, Section 4 refers to \"Theorem 2.1\", but the stated result is Lemma 2.1.","section":"Lemma 2.1 / Section 4"},{"comment":"The interpolation estimate for $E_{\\rho,1}$ is written with the wrong power of $\\bar\\rho$. By the Gagliardo-Nirenberg inequality, $\\|\\rho-\\bar\\rho\\|_{L^\\infty}\\le C\\|\\rho-\\bar\\rho\\|_{L^2}^{2(q-3)/(5q-6)}\\|\\nabla\\rho\\|_{L^q}^{3q/(5q-6)}$. Using (3.11) gives $\\|\\rho-\\bar\\rho\\|_{L^2}\\le C\\bar\\rho^{(3-\\gamma)/2}$, so the exponent of $\\bar\\rho$ should contain the factor $(3-\\gamma)/2$ multiplying $2(q-3)/(5q-6)$. The printed expression omits this factor, and consequently the definition of $\\Lambda_4$ in (3.93) does not follow from the displayed estimate. Please correct the exponent and recompute the threshold.","section":"Lemma 3.7, Eq. (3.92)"},{"comment":"The notation and algebraic identities in this lemma are not self-consistent. The symbol $\\lambda$ is used for the second viscosity coefficient, although $\\lambda$ already denotes the director diffusion coefficient in (1.1)_3. The definition of $H$ in (3.34) has signs that do not match the divergence of the momentum equation as written, and (3.36) states $-\\Delta F=\\mathrm{div}\\,H$, which is not consistent with the definition $F=(2\\mu_1+\\mu_2)\\mathrm{div}\\,u+P-P(\\bar\\rho)$ in (3.35). Since the $L^p$ estimates (3.38)--(3.45) are used repeatedly in Lemmas 3.5--3.7, please clarify the notation, correct the identities, and verify that the stated estimates remain valid with the corrected definitions.","section":"Lemma 3.4, Eqs. (3.34)-(3.37)"}],"minor_comments":[{"comment":"The factors $1/p$ and the exponent $p-2$ in (3.82) should be $1/q$ and $q-2$; the notation switches between $p$ and $q$ within the same display.","section":"Eq. (3.82)"},{"comment":"The term $\\|G(\\rho_0)\\|_{L^1}^\\gamma$ should presumably be $\\|G(\\rho_0)\\|_{L^1}$; the exponent $\\gamma$ is unexplained and is not consistent with the energy inequality being proved.","section":"Eq. (3.11)"},{"comment":"In (4.2) and (4.4), the expression $\\sup \\|\\nabla u_0\\|^2$ should be $\\sup \\|\\nabla u(t)\\|^2$, and the time interval in the definition of $T_1^*$ should be $[0,T]$ rather than $[0,T_1]$.","section":"Section 4, Eqs. (4.2)-(4.4)"},{"comment":"The notation $C([0,T;H^3])$ and $L^2(0,T;H^4)$ should be written as $C([0,T];H^3)$ and $L^2(0,T;H^4)$; moreover, the theorem states existence on $(0,\\infty)$ but (1.13) is written with a generic $T$.","section":"Theorem 1.1, Eq. (1.13)"},{"comment":"The bound $\\int_0^T\\|\\nabla u\\|^4\\,dt\\le C\\|\\nabla u_0\\|^2\\bar\\rho^{1-\\alpha}$ is not directly justified by the preceding line; it should be derived from the $E_{u,1}$ assumption in (3.8) together with the energy estimate (3.11), or the argument should be expanded.","section":"Lemma 3.1, Eq. (3.18)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript appears to be an early preprint: it contains many typos, an unproved local-existence lemma, and a genuine sign error in the key density estimate. The sign error in Lemma 3.7 is likely repairable, and the overall strategy is plausible, but the proof must be carefully corrected and rechecked before the global theorem is supported. The novelty claim is also contingent on the companion works [16] and [27] not already covering the same mechanism."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear Colleague,\n\nThe headline: the paper proves the first global strong solution for 3D compressible liquid crystal flows with density-dependent viscosity and no velocity smallness, under large background density and small L3 gradient of the initial director. That is a real step beyond the 2D result [48] and the Navier-Stokes result [16], and the underpinning idea—that α>1 makes the effective Reynolds number small when the density is large—is sound.\n\nWhat is new is the theorem and the bootstrap structure. The energy estimates in Lemmas 3.1–3.6 are standard but carefully organized, and the high-density absorption is handled with the right powers of ¯ρ. I see no circularity and no fitted parameters; the proof is self-contained modulo standard inequalities and the deferred local-existence lemma.\n\nThe soft spots are real, and one is load-bearing. In Lemma 3.7, equation (3.81) writes\n\n∇ρ_t + u·∇²ρ + ∇u·∇ρ + ∇ρ divu + (1/(2µ+λ)) ρ(∇F + ∇P) = 0.\n\nBut from ρ_t + div(ρu)=0 and F=(2µ1+µ2)divu + P−P(¯ρ), the last term must be ρ(∇F − ∇P)/(2µ1+µ2). The printed plus sign turns the dissipation in (3.82) into a positive source: the term aγ/(2µ+λ)∫ρ^{γ−α}|∇ρ|^q, which is the entire mechanism for (3.87) and for E_{ρ,1}(T)≤¯ρ/4, is not justified. This is exactly the estimate that closes the bootstrap and lets the continuation argument restart. Even if it is a one-character typo, the manuscript as submitted is algebraically inconsistent at a key point, and the proof needs to be corrected and checked again.\n\nTwo further issues, less severe. Lemma 2.1, local existence for the density-dependent viscosity system with α>1, is stated without proof and attributed by “similar arguments” to [37]; since the whole continuation rests on restarting from it, that transfer should be made explicit. And the notation is sloppy: 2µ+λ in Section 3 conflates the viscosity coefficient µ2 with the director diffusion λ, which will confuse any reader.\n\nWho this is for: researchers working on global well-posedness of compressible flows with degenerate viscosities. The theorem is significant enough to warrant a serious referee, but not in its current form. Send it to peer review, and ask the referee to verify the corrected density equation and the local-existence transfer.","headline":"The theorem is new and the bootstrap idea is sound, but Lemma 3.7 contains a load-bearing sign error in the density equation that breaks the closing estimate as written.","tokens_in":24361,"tokens_out":4416,"would_cite":false,"duration_ms":43220,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35Q35","76N10","35A01","35B65"],"pacs":[],"model":"deepseek-v4-flash","headline":"For 3D compressible liquid crystals with density-dependent viscosity, a unique global strong solution exists for arbitrarily large initial velocity whenever the background density is large and the director gradient is small.","keywords":["compressible liquid crystal flows","density-dependent viscosity","global strong solutions","large velocity","Ericksen-Leslie system","a priori estimates","Cauchy problem"],"falsifier":"Run the local well-posedness argument cited as [37] on the density-dependent stress tensor $\\operatorname{div}(\\rho^\\alpha(2\\mu_1 Du+\\mu_2\\operatorname{div}u\\, I_3))$ and look for admissible data satisfying (1.10)–(1.11) for which no local strong solution with the regularity (2.1) exists, or for which the norm $\\|\\nabla u\\|_{H^1}$ becomes unbounded in finite time before the bootstrap can be restarted; any such example would falsify Lemma 2.1 and hence Theorem 1.1.","tokens_in":23331,"feed_emoji":"🌊","tokens_out":16353,"duration_ms":153225,"temperature":0.7,"pith_summary":"The paper proves a global existence theorem for the three-dimensional simplified Ericksen–Leslie system for compressible liquid crystal flows, with viscosity coefficients $\\mu_i(\\rho)=\\mu_i\\rho^\\alpha$ growing as powers of the density with $\\alpha>1$. It shows that for any admissible initial velocity, however large, a unique global strong solution exists provided the background density $\\bar\\rho$ is sufficiently large and the $L^3$ norm of the gradient of the initial director field is sufficiently small. This is, according to the paper, the first global strong result for three-dimensional compressible liquid crystal flows that imposes no smallness on the initial velocity. The mechanism is that the effective Reynolds number $\\rho u L/\\mu_i(\\rho)=\\rho^{1-\\alpha}uL$ becomes small when $\\alpha>1$ and the density is large, so density-dependent dissipation controls fast flows regardless of their speed. The theorem is proved by a bootstrap of a priori estimates in which large-density constants absorb the size of the velocity data.","feed_headline":"Large background density tames 3D liquid-crystal flow at any speed","feed_subtitle":"First 3D strong result with no smallness on velocity; only the director's L³ gradient must be small.","key_machinery":"The load-bearing mechanism is the observation that the effective Reynolds number $Re=\\rho uL/\\mu_i(\\rho)=\\rho^{1-\\alpha}uL$ is small when $\\alpha>1$ and the density $\\rho$ is large, so dissipation dominates convection independently of how large the velocity is. The proof implements this through a priori estimates built on the effective viscous flux $F=(2\\mu_1+\\mu_2)\\operatorname{div}u+P(\\rho)-P(\\bar\\rho)$ and the vorticity $w=\\operatorname{curl}u$, which turn the momentum equation into an elliptic system whose right-hand sides can be bounded by powers of $\\bar\\rho$; Gronwall-type arguments then close the estimates. The smallness of $\\|\\nabla d_0\\|_{L^3}$ is the critical-space condition that keeps the supercritical nonlinearity $|\\nabla d|^2d$ and the constraint $|d|=1$ under control.","core_discovery":"The central claim is Theorem 1.1: under $\\alpha>1$, $\\gamma>1$, $\\alpha>(\\gamma+1)/2$ and $\\alpha\\ge \\gamma-1$, with initial data satisfying $\\rho_0\\in[3\\bar\\rho/4,5\\bar\\rho/4]$, $\\rho_0-\\bar\\rho\\in D^{1,2}\\cap D^{1,q}$, $G(\\rho_0)\\in L^1$, $u_0\\in H^2$ and $d_0-e\\in H^3$, there exist constants $\\Lambda_0$ and $\\varepsilon_0$ such that if $\\bar\\rho\\ge\\Lambda_0$ and $\\|\\nabla d_0\\|_{L^3}\\le\\varepsilon_0$, the Cauchy problem (1.1)–(1.7) admits a unique global strong solution on $\\mathbb{R}^3\\times(0,\\infty)$. The solution keeps the density in $[2\\bar\\rho/3,4\\bar\\rho/3]$ and has the regularity listed in (1.13). The paper presents this as the first global strong result for three-dimensional compressible liquid crystal flows without smallness of the velocity.","pith_inferences":["The same Reynolds-number suppression could plausibly remove velocity smallness in neighbouring coupled systems with power-law density-dependent viscosities, such as compressible magnetohydrodynamics or heat-conducting fluids, once analogous effective-flux and vorticity estimates are closed; the paper does not treat those systems.","The authors state that the director smallness cannot be removed by their approach because of the supercritical term $|\\nabla d|^2d$ and the constraint $|d|=1$; a testable alternative would be to replace the $L^3$ smallness by a weaker critical Besov smallness or by a geometric condition such as $d_0$ lying in a hemisphere, as used in the incompressible weak-solution theory.","A quantitative version of the theorem would track how $\\Lambda_0$ grows with $\\|u_0\\|_{H^2}$; the paper only asserts existence of such a constant, so the exact trade-off between velocity size and required background density is left open."],"forward_implications":["For any fixed admissible initial velocity $u_0\\in H^2$, however large, the theorem gives global well-posedness once the background density $\\bar\\rho$ is chosen large enough.","Because $\\Lambda_0$ depends on $\\|u_0\\|_{H^2}$, the largeness of $\\bar\\rho$ is chosen after the velocity is fixed; for a given $\\bar\\rho$ only velocities up to a certain $H^2$ size are admitted.","Taking $d_0\\equiv e$ reduces the system to the isentropic compressible Navier–Stokes equations with density-dependent viscosity, so the theorem covers large-velocity global strong solutions for that subsystem in the same parameter range.","The only smallness imposed on the director is the critical-space norm $\\|\\nabla d_0\\|_{L^3}$; higher derivatives of the director and the whole velocity can be large.","The density stays between $2\\bar\\rho/3$ and $4\\bar\\rho/3$ for all time, so no vacuum or density concentration develops, even for initial data with arbitrarily large velocity."],"supporting_citations":[{"why":"Supplies the local strong-existence theory that Lemma 2.1 adapts to the density-dependent viscosity system; the global continuation argument restarts from it.","marker":"[37]"},{"why":"Proves the corresponding global large-strong-solution theorem for the isentropic compressible Navier–Stokes equations with density-dependent viscosities, whose bootstrap strategy this paper extends to the coupled director equation.","marker":"[16]"},{"why":"Gives global regular solutions under the stricter conditions α>1 and α≥2γ−1 with small density; the paper widens the admissible parameter range.","marker":"[45]"},{"why":"Provides the only earlier global strong result for compressible nematic liquid crystal flows with density-dependent viscosity, in two dimensions, which the present theorem complements in 3D.","marker":"[48]"},{"why":"Is the authors' companion result for three-dimensional inhomogeneous liquid crystal flows with density-dependent viscosity and large velocity, supplying the framework adapted here.","marker":"[27]"}],"fun_headline_variants":["First 3D compressible liquid-crystal strong solution without velocity smallness","Large density and small director gradient yield global strong 3D flow","No velocity smallness needed: global strong 3D liquid-crystal flow","Dense background, tiny director tilt: strong 3D flow at any speed"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Lemma 2.1, the local strong-existence theorem for this exact density-dependent viscosity system, is stated without proof and attributed by 'similar arguments' to a reference that is not shown to cover $\\mu_i(\\rho)=\\mu_i\\rho^\\alpha$ with $\\alpha>1$; if that transfer fails for some data satisfying (1.10), the global theorem has no starting point.","fun_headline_variants_meta":{"raw":{"variants":["First 3D compressible liquid-crystal strong solution without velocity smallness","Large density and small director gradient yield global strong 3D flow","No velocity smallness needed: global strong 3D liquid-crystal flow","Dense background, tiny director tilt: strong 3D flow at any speed"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00057,"raw_usage":{"total_tokens":2665,"prompt_tokens":884,"completion_tokens":1781,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":500,"completion_tokens_details":{"reasoning_tokens":1711}},"tokens_in":500,"tokens_out":1781,"duration_ms":15290,"temperature":1.0,"reasoning_tokens":1711,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T19:41:04.375803+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the local well-posedness argument cited as [37] on the density-dependent stress tensor $\\operatorname{div}(\\rho^\\alpha(2\\mu_1 Du+\\mu_2\\operatorname{div}u\\, I_3))$ and look for admissible data satisfying (1.10)–(1.11) for which no local strong solution with the regularity (2.1) exists, or for which the norm $\\|\\nabla u\\|_{H^1}$ becomes unbounded in finite time before the bootstrap can be restarted; any such example would falsify Lemma 2.1 and hence Theorem 1.1.","supporting_citations":[{"cited_title":"Ma, Classical solutions for the compressible liquid crystal flows with nonnegative initial densities, J","cited_arxiv_id":null,"evidence_quote":"Supplies the local strong-existence theory that Lemma 2.1 adapts to the density-dependent viscosity system; the global continuation argument restarts from it."},{"cited_title":"Xin and S","cited_arxiv_id":null,"evidence_quote":"Gives global regular solutions under the stricter conditions α>1 and α≥2γ−1 with small density; the paper widens the admissible parameter range."},{"cited_title":"Zhong and X","cited_arxiv_id":null,"evidence_quote":"Provides the only earlier global strong result for compressible nematic liquid crystal flows with density-dependent viscosity, in two dimensions, which the present theorem complements in 3D."},{"cited_title":"Global strong solution of the 3D inhomogeneous liquid crystal flows with density-dependent viscosity and large velocity","cited_arxiv_id":"2410.11881","evidence_quote":"Is the authors' companion result for three-dimensional inhomogeneous liquid crystal flows with density-dependent viscosity and large velocity, supplying the framework adapted here."}],"review_version":1}