{"id":"a3a1b953-7b55-4584-b871-25471ceb690c","arxiv_id":"2608.08140","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For a CutFEM with Lagrange multipliers, the authors prove k-th order convergence for fluid velocity and solid displacement and (k-1)-th order for pressure in a fully coupled fluid-structure interaction problem with a moving interface.","lead":"The paper proves that a numerical method letting a solid move through a fixed fluid grid converges at the expected rate for fluid-structure interaction, and claims this is the first such proof for a deforming interface. A generalist should read it because these methods are attractive for heart valves, parachutes, and other moving-body simulations, where remeshing is costly.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed O(h^{2k}) rate depends on the sketched cut-domain Ritz projection estimates in Lemma 9.1, whose omitted proofs could hide a loss of one order in k.","rationale":"The reader's weakest assumption identified the smoothness/regularity conditions (C1)-(C3) as the main risk. Those conditions are standard for a priori error analysis and do not by themselves threaten the internal validity of the proof: if the exact solution is smooth on [0,T], the theorem is a conditional statement of exactly that kind. The more serious issue is that a load-bearing technical lemma, Lemma 9.1, is only sketched, and the final O(h^{2k}+H^{2k}) rate depends on the uniformity of the hidden estimates in the cut geometry. This does not invalidate the paper's central claim, but it does justify the reader's conditional verdict: the claim is plausible and the numerical experiments support the expected rates, yet the proof should be completed or independently checked before full acceptance. I therefore recommend no change to the reader's verdict.","tokens_in":102178,"tokens_out":5385,"duration_ms":63515,"concrete_test":"Write out the complete proof of Lemma 9.1, tracking all constants uniformly in t∈[0,T], in h and H, and in the position of Γ[XH] in the background mesh, and then re-derive Lemma 9.2 and inequality (8.132) using only those stated estimates; if any step uses h^{k-1} in place of h^k without a compensating factor, the theorem's rate is not yet proven.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central estimate (8.133)-(8.134) is closed in Section 8.6 via Lemma 9.2, and Lemma 9.2's Part 7 relies on the Ritz projection bounds (9.2a)-(9.2c) to control the consistency remainder R*_{h,H}(eu, ep, ∂ted, eσ). Lemma 9.1 is load-bearing: (9.2a) and (9.2b) are asserted with the phrase 'details are omitted', and (9.2c) is obtained by a duality argument that invokes the moving-domain Stokes regularity (3.11)-(3.12) plus a chain of inverse and trace estimates. The time-derivative estimate (9.2b) gives only h^{k-1} in the H1-type norm, so any non-uniformity in the cut position, or any hidden condition on the H2×H1 regularity constant as Γ(t) moves through the background mesh, could propagate through (9.30)-(9.36) and (8.132) and degrade the final bound from O(h^{2k}+H^{2k}) to O(h^{2k-2}+H^{2k}). This is a genuine gap in the written proof: the rate of Theorem 3.1 is not independently established without Lemma 9.1.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a semidiscrete immersed CutFEM for a fully coupled fluid--structure interaction problem with a deforming interface, using a Lagrange multiplier to enforce the kinematic coupling and a global ghost-penalty stabilization. The main result, Theorem 3.1, claims an a priori error bound of order k for the fluid velocity and solid displacement and order k-1 for the pressure, for finite elements of degree k>=3, under smoothness assumptions on the exact solution and a mesh proportionality condition. The proof is organized around an interpolated exact configuration, geometric transport identities, consistency estimates, and a continuation argument. The paper also reports numerical experiments with k=2 and k=3.","tokens_in":102440,"tokens_out":9372,"duration_ms":88574,"significance":"If the stated convergence result is correct, this is a significant contribution: it would be the first rigorous error analysis of an unfitted finite element method for a fully coupled FSI system with a genuinely deforming interface. The paper introduces a systematic framework for handling the coupling between interface geometry error and fluid/solid errors, and it explicitly formulates the required uniform geometric, trace, inverse, and inf-sup estimates on moving cut domains. The claimed result is concrete and falsifiable, and the numerical experiments, although not covered by the theorem for k=2, are consistent with the stated rates. However, the proof contains load-bearing gaps that, as written, prevent the claimed rates from being fully established.","major_comments":[{"comment":"The Ritz projection estimates (9.2a) and (9.2b) are central to the proof: they are used in Lemma 9.2, in the estimates of (9.15a), (9.22), (9.39)-(9.40), and in the construction of the extension w in Part 8. The proof of Lemma 9.1, however, states 'The details are omitted here' for both estimates. Since this is a cut-domain Ritz projection on a moving interface with ghost penalties, these estimates are not routine off-the-shelf results, and the time-derivative bound (9.2b) only gives h^{k-1} in the H^1-type norm. A loss of one order in h at this stage would propagate through (9.30)-(9.36) and (8.132) and degrade the final bound in Theorem 3.1 from O(h^{2k}) to O(h^{2k-2}). The authors must supply the full proof of (9.2a)-(9.2b) or otherwise justify that no order loss occurs.","section":"Section 9, Lemma 9.1, Eqs. (9.2a)-(9.2b)"},{"comment":"The intermediate energy estimates contain lower-order terms that appear incompatible with the claimed O(h^{2k}) rate. For example, (8.94)-(8.95) include C(h^{2k-2}+H^{-1}h^{2k}+h^{-1}H^{2k}+H^{2k-1}), which with h~H is dominated by h^{2k-2}. This lower-order term is carried into the estimate of (4.14h) in (8.121), which introduces h^{2k-3} coefficients. The final estimate (8.132), however, states C(h^{2k}+H^{2k}) without explaining how these lower-order terms are absorbed or cancelled. Since h^{2k-2} > h^{2k} for small h, the written estimates do not establish the convergence rate claimed in Theorem 3.1. The authors should either revise the estimates to obtain the optimal rate or state the weaker rate that the current estimates actually yield.","section":"Section 8.5 and Section 8.6, Eqs. (8.94), (8.95), (8.121), (8.132)"},{"comment":"The continuation argument in Section 5 recovers the smallness assumptions (5.2) from the 'final' estimates (8.133)-(8.134). If, as noted above, the actual estimates only give O(h^{2k-2}) for the H^1-type errors, then the inverse-inequality step in (5.5) would need to be revisited. For k=3, the L-infinity bound on e_d would be O(h^{3/2}) rather than O(h^{5/2}), which is still small enough for (5.2a), but the W^{1,∞} bound in (5.6) would be O(h^{1/2}) in three dimensions instead of O(h^{3/2}); this would violate the required O(H^{1/4}) bound only for very small h, so the continuation may still close, but the stated rates in Theorem 3.1 would not follow. The proof needs to make the dependence of the constants on the final rate explicit.","section":"Section 5 and Section 8.6, Eqs. (5.2)-(5.9), (8.132)-(8.134)"}],"minor_comments":[{"comment":"The title contains a typo: 'INTERF ACE' should be 'INTERFACE'.","section":"Title"},{"comment":"The passage to the limit epsilon_h -> 0 is deferred with 'the details are omitted'; since this is not central to the main convergence claim, it is acceptable, but the omission should be noted more explicitly in the text.","section":"Section 3, Remark 3.4"},{"comment":"The numbering (4.8a)-(4.8l) is dense and makes the consistency remainder difficult to follow; a table or a more descriptive decomposition would improve readability.","section":"Section 4, Eq. (4.8)"},{"comment":"The numerical experiment uses E_sigma^{2,0} in the L^2(bGamma_H) norm rather than the H^{-1/2} norm of Theorem 3.1; the text acknowledges this, but the reader should be reminded that the displayed rates for the traction are not directly comparable to the theorem.","section":"Section 10"},{"comment":"The local well-posedness argument relies on the positivity of epsilon_h to turn the system into ODEs; the role of epsilon_h in the time-derivative ghost penalty should be stated even more explicitly in the main text, as it is a nonstandard term.","section":"Appendix A"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses a genuinely open problem and the overall architecture of the proof is plausible and careful. However, the two issues in the major comments—the omitted proofs in Lemma 9.1 and the unexplained lower-order terms in the energy estimates—are load-bearing for the claimed convergence rates. The first is a completeness issue; the second, if not a typographical artefact, could mean the theorem only holds with a reduced rate. Both are potentially fixable within the manuscript's scope, which is why I recommend major revision rather than rejection. The editor may wish to ask the authors to state explicitly which of the displayed rates (e.g., h^{2k} versus h^{2k-2}) are actually proved, and to supply the missing details for Lemma 9.1."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper claims the first rigorous convergence analysis of an unfitted finite element method for a fully coupled Navier-Stokes/linear-elasticity FSI problem with a genuinely deforming interface. That claim looks right: the literature review is careful, and I don't know of a competing proof. The paper does a lot well: the error equation (4.14) is derived cleanly, the continuation argument in Section 5 is coherent, and the geometric perturbation machinery (intermediate interfaces, transport formulas, norm equivalences) is systematically developed. The numerical experiments are honest, including k=2 results that go beyond the theorem.\n\nThe soft spot is real and it is where the stress-test lands. Lemma 9.1 is load-bearing: the Ritz projection estimates (9.2a) and (9.2b) are asserted with 'the details are omitted', and (9.2c) is only sketched. These feed directly into Lemma 9.2 and hence into the final O(h^{2k}+H^{2k}) rate. The time-derivative estimate (9.2b) is only h^{k-1} in the H1 norm, and the duality argument that would lift it to L2 relies on the moving-domain Stokes regularity (3.11)-(3.12). If a hidden loss of order lurks there, the rate degrades to h^{2k-2}+H^{2k}. That would change the theorem's content. I don't see a fatal error in the sketch, and the estimate is plausible for a cut-domain Ritz projection, but this is precisely the kind of lemma that needs a complete proof in a paper whose whole contribution is the proof.\n\nThe regularity assumptions are also strong: the exact solution must be smooth up to T and the Stokes problem must have uniform H2xH1 regularity for every t. For Navier-Stokes coupled to elasticity, such solutions are known only locally in time. The authors are transparent about this, and the theorem is honest as a conditional result. The k>=3 restriction is flagged as technical, and the experiments suggest k=2 works, but the paper does not prove it.\n\nWho is this for? Specialists in unfitted FEM and FSI analysis. A serious referee should engage with it, not desk-reject. But the referee should demand the omitted proofs of Lemma 9.1 and Lemma E.2 before signing off on the rate. If those details check out, this is a major reference. As written, the claim is conditional on a sketched lemma.","headline":"First convergence proof for unfitted FSI with a deforming interface; serious and likely correct, but the central Ritz projection lemma is sketched and the claimed rate is not yet fully supported.","tokens_in":102948,"tokens_out":3003,"would_cite":true,"duration_ms":31603,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["65M60","65M15","74F10","76D05"],"pacs":[],"model":"deepseek-v4-flash","headline":"A rigorous error bound is proved for a cut finite element method applied to fully coupled fluid–structure interaction with a moving interface.","keywords":["CutFEM","unfitted finite element method","fluid–structure interaction","moving interface","a priori error analysis","Lagrange multiplier","ghost penalty","Navier–Stokes"],"falsifier":"Run the Section 10 radially oscillating annulus test with $k=2$ on a sequence of refined meshes: if the $H^1$ velocity error decays strictly slower than $h^2$, the theorem's conclusion (or its conjectured extension to $k=2$) would be false. More generally, construct an FSI solution whose fluid domain develops a re-entrant corner at the interface at some time, so the uniform Stokes regularity in Assumption (C1) fails; the bound (3.14) should then break, showing the smoothness hypothesis is load-bearing.","tokens_in":101943,"feed_emoji":"🌊","tokens_out":6296,"duration_ms":57841,"temperature":0.7,"pith_summary":"This paper proves, for the first time, a rigorous a priori error estimate for an unfitted (cut) finite element method applied to a fully coupled fluid–structure interaction problem in which the interface between the fluid and the solid is part of the unknown solution. The method couples the incompressible Navier–Stokes equations to a linearly elastic solid through a Lagrange multiplier that enforces the kinematic and traction conditions on the moving interface. Under smoothness assumptions on the exact solution, the paper shows that with finite elements of degree $k\\geq 3$, the fluid velocity and solid displacement converge at order $k$ in their natural energy norms, while the pressure converges at order $k-1$. The significance is that it closes the two-way feedback loop between the error in the interface position and the error in the fluid solution, which is the main obstacle in analyzing unfitted methods for deforming interfaces.","feed_headline":"CutFEM for deforming fluid–solid interfaces proven convergent","feed_subtitle":"First rigorous error analysis of an unfitted FEM for fully coupled FSI with a moving interface.","key_machinery":"The central technical object is the interpolated configuration $X^*_H = I + d^*_H$, the finite-element interpolant of the exact solid motion, and the families of intermediate interfaces $\\Gamma[X^\\theta_H]$, $\\Gamma[X^{*,\\theta}_H]$, and $\\Gamma[X^{\\#,\\theta}_H]$ that connect the exact, interpolated, and numerical interfaces. Transport formulas with respect to the auxiliary parameter $\\theta$ turn geometric perturbations into integrals involving the displacement error $e_d = d_H - d^*_H$, which is the key to quantifying two-way coupling. Equally important is the projected kinematic mismatch $P_{\\Gamma[X_H]} e_u - \\partial_t e_d \\circ X_H^{-1}$ on the numerical interface, where $P_{\\Gamma[X_H]}$ is the $L^2$ projection onto the interface multiplier space; estimating this object transfers control between the fluid velocity error and the solid velocity error. Uniform ghost-penalty, trace, inverse, and inf-sup estimates on the moving cut domains supply the mesh-robustness needed throughout, and a continuation-in-time argument converts the required Lipschitz bounds on the numerical deformation into consequences of the energy estimates themselves.","core_discovery":"Theorem 3.1 is the paper's central claim. For a semidiscrete CutFEM with ghost-penalty stabilization, a Lagrange multiplier on the reference interface, and finite elements of degree $k\\geq 3$, it establishes the error bound $$\\max_{t\\in[0,T]}\\big(\\|u_h-u\\|^2_{$L^{2}$(\\Omega_f[X_H])}+\\|d_H-d\\circ\\Phi_H\\|^2_{$H^{1}$(\\hat\\$\\Omega$^s_H)}+\\|\\partial_t d_H-\\partial_t d\\circ\\Phi_H\\|^2_{$L^{2}$(\\hat\\$\\Omega$^s_H)}\\big)+\\int_0^T\\|u_h-u\\|^2_{$H^{1}$(\\Omega_f[X_H])}\\,dt+$H^{2}$\\int_0^T\\big(\\|p_h-p\\|^2_{$L^{2}$(\\Omega_f[X_H])}+\\|\\hat\\sigma_H-\\hat\\$\\sigma$\\circ\\Phi_H\\|^2_{$H^{{-1/2}}$(\\hat\\Gamma_H)}\\big)\\,dt\\le C($h^{{2k}}$+$H^{{2k}}$),$$ provided the exact solution is smooth and the moving-domain Stokes problem has uniform $H^2\\times H^1$ regularity. The proof treats the fluid equations on the numerically deformed domain, so the discrete interface, its normal, and the pulled-back traction all depend on the solid displacement error; conversely, the solid error is driven by the fluid traction and kinematic mismatch. The authors close this loop by introducing an interpolated configuration, establishing uniform norm equivalences and trace/inverse estimates on the moving cut domains, and using a continuation-in-time argument to upgrade a priori smallness assumptions into unconditional convergence on the whole smoothness interval.","pith_inferences":["The interpolated-configuration and transport-formula machinery is likely to transfer to other moving-interface coupled problems—two-phase flow, free-boundary problems, or FSI with hyperelastic solids—where the same type of geometry-error feedback appears, provided the corresponding regularity assumptions hold.","Because the theorem is conditional on global smoothness of the FSI solution, and such smoothness is only known locally in time for the Navier–Stokes–elasticity system, the practical reach of the result is a short-time error bound; extending it to long-time or nonsmooth regimes would require either global regularity results or a different (e.g., weak-solution) error framework.","The observed $(k+1)$-th order convergence in $L^\\infty(L^2)$-type norms in the numerical tests is not covered by the analysis; a sharper duality argument might recover these rates and would be a natural next step."],"forward_implications":["For any FSI solution satisfying the smoothness assumptions and for polynomial degree $k\\ge 3$, the semidiscrete CutFEM converges at the optimal rate $O(h^k+H^k)$ in the energy norms, and this rate is achieved uniformly on the whole time interval $[0,T]$.","The error constant is independent of the stabilization parameter $\\epsilon_h\\in(0,h]$, so the analysis covers the limit $\\epsilon_h\\to 0$ and hence the pure ghost-penalty formulation without time-derivative stabilization.","The theorem is the first rigorous convergence result for an unfitted FEM with a genuinely deforming interface; it opens the door to analyzing other immersed FSI formulations, including Nitsche-type and distributed-Lagrange-multiplier variants, by the same interpolated-configuration strategy.","The restriction $k\\ge 3$ is technical: the numerical experiments in Section 10 show clean second-order convergence for $k=2$, suggesting the theorem should extend once a direct $W^{1,p}$ error estimate for the cut-domain Ritz projection is available."],"supporting_citations":[{"why":"Supplies the isoparametric interpolation theory and the geometry approximation map $\\hat\\Phi_H$ used to define the interpolated solid configuration.","marker":"[50]"},{"why":"Provides the norm-equivalence lemma for interpolated interfaces that underpins the geometric perturbation estimates in Section 6.","marker":"[37]"},{"why":"Gives norm equivalence for nearby surfaces, used to compare the exact, interpolated, and numerical interfaces.","marker":"[44]"},{"why":"Provides the ghost-penalty extension estimate (Lemma 5.2) that controls finite element functions on the extended fluid region.","marker":"[49]"},{"why":"Establishes the uniform trace inequalities and the unfitted inf-sup condition on which the pressure and traction stability estimates rest.","marker":"[40]"},{"why":"Supplies the boundedness of the $L^2$ projection on $H^s$ and $L^p$ surface spaces, needed for the interface multiplier estimates.","marker":"[53]"},{"why":"Provides the surface transport formulas used to differentiate integrals over the moving interfaces.","marker":"[33]"},{"why":"The analytical radially oscillating annulus solution used in the convergence tests of Section 10.","marker":"[73]"}],"fun_headline_variants":["First rigorous convergence proof for CutFEM in moving-interface FSI","CutFEM convergence proven for deforming fluid-solid interfaces","Rigorous error analysis: CutFEM for moving interfaces in FSI","CutFEM error bounds for FSI with deforming interface: order k","First unfitted FEM convergence proof for fully coupled FSI with moving interface"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof requires the exact fluid–structure solution to be smooth on the entire time interval and the moving-domain Stokes problem at each time to have uniform $H^2\\times H^1$ regularity; for the full incompressible Navier–Stokes/elasticity system, such global regularity is not known and is only expected for short times and restricted data.","fun_headline_variants_meta":{"raw":{"variants":["First rigorous convergence proof for CutFEM in moving-interface FSI","CutFEM convergence proven for deforming fluid-solid interfaces","Rigorous error analysis: CutFEM for moving interfaces in FSI","CutFEM error bounds for FSI with deforming interface: order k","First unfitted FEM convergence proof for fully coupled FSI with moving interface"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00071,"raw_usage":{"total_tokens":3235,"prompt_tokens":1021,"completion_tokens":2214,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":637,"completion_tokens_details":{"reasoning_tokens":2119}},"tokens_in":637,"tokens_out":2214,"duration_ms":15576,"temperature":1.0,"reasoning_tokens":2119,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T00:21:13.842775+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the Section 10 radially oscillating annulus test with $k=2$ on a sequence of refined meshes: if the $H^1$ velocity error decays strictly slower than $h^2$, the theorem's conclusion (or its conjectured extension to $k=2$) would be false. More generally, construct an FSI solution whose fluid domain develops a re-entrant corner at the interface at some time, so the uniform Stokes regularity in Assumption (C1) fails; the bound (3.14) should then break, showing the smoothness hypothesis is load-bearing.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the isoparametric interpolation theory and the geometry approximation map $\\hat\\Phi_H$ used to define the interpolated solid configuration."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the norm-equivalence lemma for interpolated interfaces that underpins the geometric perturbation estimates in Section 6."},{"cited_title":"Kov´ acs, B","cited_arxiv_id":null,"evidence_quote":"Gives norm equivalence for nearby surfaces, used to compare the exact, interpolated, and numerical interfaces."},{"cited_title":"Lehrenfeld and M","cited_arxiv_id":null,"evidence_quote":"Provides the ghost-penalty extension estimate (Lemma 5.2) that controls finite element functions on the extended fluid region."},{"cited_title":"Guzm´ an and M","cited_arxiv_id":null,"evidence_quote":"Establishes the uniform trace inequalities and the unfitted inf-sup condition on which the pressure and traction stability estimates rest."},{"cited_title":"Li and T","cited_arxiv_id":null,"evidence_quote":"Supplies the boundedness of the $L^2$ projection on $H^s$ and $L^p$ surface spaces, needed for the interface multiplier estimates."},{"cited_title":"Wan and O","cited_arxiv_id":null,"evidence_quote":"The analytical radially oscillating annulus solution used in the convergence tests of Section 10."}],"review_version":1}