{"id":"8e7e3bf9-1e96-4d13-a7c5-76dc5bf26a3b","arxiv_id":"2506.11961","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"For an oscillating pressure source in shallow water, wave-driven thrust is maximized by radiating waves only in the aft direction, yielding thrust equal to the injected power at every drift speed.","lead":"A mathematical study shows the best way for an oscillating surface pressure source to push itself across shallow water is to emit waves only behind it, making every unit of injected power become thrust at any cruising speed. This gives a general design rule for wave-driven propulsion and a power-modulation recipe to accelerate from rest to speeds beyond the wave speed.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The bounded-power η=1 result is model-internal; the more load-bearing gap is the unverified quasi-static use of periodic optimal solutions to claim acceleration from rest to supercritical speeds, since Eq. (7.4) is a steady balance with no equation for dv*/dt.","rationale":"The reader's weakest_assumption lists both the linear 1D shallow-water idealization and the quasi-static acceleration assumption in Section 7. I agree with the first concern but regard it as a stated modelling limitation rather than a load-bearing flaw, since the paper transparently frames its contribution as a template and the prefactor discrepancy is acknowledged in Section 2. The quasi-static assumption is more load-bearing because it is the only bridge from the static optimal-control results to the claimed physical outcome (accelerating from rest to supercritical velocities) and it is asserted rather than derived. The paper itself flags the condition Ü << ωc in Sections 7 and 8, but never writes the momentum equation for the accelerating body, never defines the relevant timescales, and never quantifies the error of the quasi-static approximation. The internal algebra of Sections 3-6 is consistent and numerically validated, so the central mathematical finding (F_T = δ for bounded power in a fixed frame) is not in question. The concern is specifically that the extension in Section 7 to time-varying δ mixes a terminal-velocity balance with an acceleration claim. A concrete check is to write the body momentum equation with v(t) and δ(t) and compare the integrated trajectory against the algebraic v*(δ). This does not require rejecting the paper; it requires an additional timescale assumption that the current text does not provide. I therefore keep the reader's CONDITIONAL verdict, with the condition being the explicit and tested quasi-static validity of the Section 7 acceleration recipe.","tokens_in":23380,"tokens_out":2130,"duration_ms":69940,"concrete_test":"Derive and simulate the coupled quasi-static system on a finite domain: take the optimal bounded-power source Q(x,t;δ(t)) from (7.5) with δ(t) ramping from a small value to the supercritical value over a time T_ramp, and integrate the body equation m dv/dt = F_T(v,δ(t)) - (1/2)C_D ρ L v^2, with m the body mass per unit width and F_T(v,δ) = δ (dimensionless scaling) over the ramp. Compare the trajectory v(t) to the algebraic v*(δ(t)) from (7.4). Repeat for ramp times T_ramp spanning the wave period 2π/ω and the momentum relaxation time m/(ρ L^2 c). If the lag |v-v*| remains small for T_ramp >> m/(ρ L^2 c) but not for shorter ramps, the quasi-static claim needs that timescale to be stated as an explicit assumption. If no parameter choice yields small lag, Eq. (7.4) does not support the claimed acceleration recipe.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central physical claim in Section 7 is that slowly modulating the power bound δ accelerates the body from rest to supercritical velocities, with v* = sqrt(2Hδ/(L C_D)) at every instant. This requires that the periodic optimal-control solutions derived for constant v remain valid at each instant of an accelerating trajectory. The paper states the condition Ü << ωc (Section 7: 'A dimensional acceleration such that Ü << ωc, permits the use of the previous results'), but no acceleration equation is written and no error bound is derived. The quasi-static approximation has two distinct parts that are conflated. First, the wave field must be periodic in a frame whose velocity is changing; a Galilean transformation only maps between inertial frames, so the ODE (4.3) and the Sommerfeld conditions (4.5) are not rigorously valid when v = v(t). Second, even if the wave field is quasi-static, the body momentum balance is dv*/dt = (F_T(v,δ) - F_D(v))/m, not the algebraic condition F_T = F_D used in (7.4). Setting F_T = F_D only defines the terminal velocity for constant δ; it does not by itself show that a time-varying δ tracks the instantaneous terminal velocity. If δ changes faster than the body's momentum relaxation time, the actual v will lag v*(δ), and the claimed 'modulation recipe' requires an extra timescale separation m/(ρ L^2 c) that is not stated. The self-referential limitation in Section 8 ('A small acceleration Ü << ωc is required to maintain the periodic assumption') is exactly the condition that is neither formalised nor tested in Section 7, so the acceleration claim rests on an unverified quasi-static assumption.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies optimal wave-driven propulsion in a 1D shallow-water model where a periodically oscillating pressure source acts as the body. The authors derive expressions for time-averaged thrust and power directly from the wave equation, then solve two constrained optimal-control problems (bounded norm and bounded power) for the start-up, subcritical, critical, and supercritical velocity regimes. The main analytical result is that in the bounded-power case, the optimal thrust equals the power bound (F_T = delta), yielding efficiency eta = 1 for all v > 0. Using a quadratic drag law, the paper derives a cruising velocity v* = sqrt(2H delta/(L C_D)) and argues that slow modulation of the power bound accelerates the body from rest to supercritical velocities. Numerical validation with Ipopt/JuMP is reported for the optimality conditions and thrust values.","tokens_in":23695,"tokens_out":3927,"duration_ms":58753,"significance":"If the results stand, the paper provides a clean analytical template for optimal wave-driven propulsion in a simplified setting, with self-contained derivations of thrust and power from the wave equation rather than imported formulas. The numerical validation is a genuine strength: the analytical optima are checked against an independent discretized optimization, with relative errors mostly below 0.1%, and the authors provide a repository for the code. The bounded-power result that an aft-travelling wave achieves F_T = delta and eta = 1 across all velocity regimes is elegant and intuitively appealing. However, the quantitative reach is limited by the acknowledged idealizations (shallow water, no surface tension, no viscosity, 1D), and the prefactor discrepancy with the classical radiation-stress formula (1.1) means the dimensional cruising-velocity prediction (7.4) should be treated as model-internal unless that factor is resolved.","major_comments":[{"comment":"The claim that modulating the power bound delta accelerates the body from rest to supercritical velocities is not supported by the analysis. Equation (7.4) is an algebraic balance F_T = F_D, which defines the terminal cruising velocity for a fixed delta, not an equation of motion. To demonstrate acceleration, the manuscript must state and use the body momentum balance m dv/dt = F_T(v,delta) - F_D(v) and show that a slowly varying delta(t) tracks the instantaneous equilibrium v*(delta(t)). This requires an additional timescale separation, essentially m/(rho L^2 c) being much longer than the modulation timescale; the stated condition \\ddot{U} << omega c does not by itself ensure that the body velocity follows v*(delta). Without this step, the central demonstration in Section 7 is incomplete.","section":"Section 7, Eq. (7.4)"},{"comment":"The quasi-static use of the periodic optimal solutions during acceleration needs a rigorous error bound or a multiple-timescale derivation. The Galilean transformation (4.2) maps between inertial frames and is only valid for constant v; when v = v(t), the wave equation in the moving frame contains additional terms involving dv/dt, and the Sommerfeld boundary conditions (4.5) are not exact. The condition \\ddot{U} << omega c is asserted but never derived from the transformed PDE. A concrete test would be to substitute v(t) = v0 + epsilon t into the original wave equation and estimate the size of the neglected terms relative to the leading-order optimal solution. As written, the paper does not establish that the periodic optimal solutions remain valid instantaneously during the proposed acceleration protocol.","section":"Section 7, Eqs. (4.3) and (4.5)"},{"comment":"The prefactor discrepancy between the derived thrust (2.10) and the classical radiation-stress formula (1.1) is acknowledged but then set aside. This is acceptable for the dimensionless optimization results, but it is load-bearing for the dimensional cruising-velocity prediction in Section 7: if the physically correct thrust is the classical 3/4 rho g [h^2] result rather than the 1/2 [k|h|^2] version, then v* in (7.4) changes by a factor sqrt(3/2). The manuscript should either incorporate the correct prefactor into the dimensional force balance or explicitly state that (7.4) is only a model-internal estimate, not a quantitative physical prediction.","section":"Section 2, Eqs. (1.1), (2.10), and (7.2)-(7.4)"}],"minor_comments":[{"comment":"The text reports a relative error of 0.003% for 'lambda|D_L|^2' but the surrounding equations and text elsewhere refer to lambda^2|D_L|^2; please check the exponent and make the notation consistent.","section":"Section 3.1, Eq. (3.5)"},{"comment":"The step-function ansatz is presented as representative, but the manuscript does not explain how it was chosen among the infinite family of optimal sources; a brief remark that the optimality conditions are underdetermined would clarify that the shown solution is one example, not a unique optimum.","section":"Section 3.2, Eq. (3.17)"},{"comment":"The resonance condition is written for all integers n in Z\\{0}, but the subcritical resonances are only physical when 0 < n < l/pi, while supercritical resonances require n < 0; the statement 'for all possible n' is slightly misleading and could be rephrased to specify the allowed ranges.","section":"Section 6, Eq. (6.7)"},{"comment":"The caption for Figure 8 does not explain the colour scheme beyond 'blue' and 'red'; it would help to state explicitly that the first and second halves of the period are plotted in blue and red, respectively.","section":"Section 7, Fig. 8"}],"recommendation":"major_revision","confidential_remarks":"The core optimal-control derivations are solid and well validated numerically, but the paper's advertised acceleration result in Section 7 lacks a dynamical model and relies on an unproven quasi-static assumption. If the authors can add an equation of motion and a timescale-separation argument, the revised manuscript would be publishable. The prefactor issue with (1.1) is not fatal but should be resolved or explicitly bracketed as model-internal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read this one if you care about wave-driven propulsion or optimal control of radiation-stress thrust. The central result is clean and the derivations are honest. The bounded-power optimum—pure aft wave, F_T = delta, efficiency 1 for every v—is genuinely new as far as I can tell, and the resonant zeros in the bounded-norm case are a nice bonus. The authors derive thrust and power from the wave equation rather than importing them, and the Ipopt/JuMP numerics match the analytics to 0.01% or better, with a public repo. That is solid, reproducible work.\n\nThe soft spots are real but mostly addressable. The 3/2 prefactor discrepancy with the classic Longuet-Higgins & Stewart formula is dismissed with a sentence; it changes the dimensional cruising velocity in Section 7, so it should at least be given a footnote. Global optimality in the bounded-norm case is inferred from the variational necessary conditions plus numerics; that's acceptable for a physics paper but a referee could ask for a convexity argument.\n\nThe load-bearing gap is Section 7. The paper says a small acceleration U_ddot << omega c permits the quasi-static use of the periodic optima, but no equation of motion for the body's drift velocity is written and no error bound is derived. The force balance F_T = F_D gives the terminal velocity for constant delta; it does not by itself show that a time-varying delta tracks that terminal velocity. Whether the modulation recipe works in practice depends on a timescale separation (momentum relaxation vs. period) that is never stated. The authors call Section 7 a 'demonstration of a simple theory,' so it is not an overclaim, but the acceleration claim is incomplete.\n\nOverall: the paper is worth a serious referee. The math is self-contained, the numerics are convincing, and the template could be useful beyond shallow water. The referee should push on the quasi-static assumption and the prefactor. If those are tightened, this is a solid JFM-type paper.","headline":"Clean optimal-control template for wave-driven thrust, with a strong bounded-power result; the acceleration claim in Section 7 needs a formal quasi-static justification.","tokens_in":24291,"tokens_out":2262,"would_cite":true,"duration_ms":29579,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["76B15","49J20"],"pacs":[],"model":"deepseek-v4-flash","headline":"Waves alone can propel a vessel at perfect efficiency","keywords":["wave-driven propulsion","radiation stress","shallow water waves","optimal control","bounded power","self-generated waves","cruising velocity","Froude number"],"falsifier":"Measure the thrust and the injected power of a real oscillating pressure source on a shallow water tank while resolving the fore and aft wave amplitudes. If the maximum achievable thrust-to-power ratio falls below 1 once surface tension, viscosity, and 2D losses are present, or if the optimal source cannot be made to emit zero wave ahead, the efficiency-1 claim fails.","tokens_in":23100,"feed_emoji":"🌊","tokens_out":5103,"duration_ms":58926,"temperature":0.7,"pith_summary":"The paper asks how an oscillating body floating on shallow water should move its pressure field to get the most forward thrust from the waves it generates itself. In the physically motivated case where the time-averaged power injected by the body is bounded, the answer is simple: emit a wave only in the aft direction and none ahead. Under that condition the dimensionless thrust equals the power bound, $\\bar F_T = \\delta$, so the efficiency defined as thrust divided by power is exactly 1 for every drift velocity $v \\ge 0$, in subcritical, critical, and supercritical regimes alike. The paper then couples this optimal thrust to a simple quadratic drag law to show that slowly increasing the power bound accelerates the body from rest to supercritical cruising speeds, with cruising velocity $v_* = \\sqrt{2H\\delta/(L C_D)}$.","feed_headline":"Waves alone can propel a vessel at perfect efficiency","feed_subtitle":"Optimal thrust comes from a purely aft wave; ramping up power accelerates the body from rest to supercritical speeds.","key_machinery":"The load-bearing object is the radiation-stress thrust identity $\\bar F_T = \\langle \\hat Q, \\hat h' \\rangle = -\\tfrac12 [k|\\hat h|^2]_-^+$, which equates the thrust to the time-averaged difference of fore-aft wave amplitude squared, derived by multiplying the wave equation by $\\hat h'$ and integrating. The optimal-control mechanism is the variational calculus with the power bound $\\langle \\hat Q, i\\hat h\\rangle - v\\langle \\hat Q, \\hat h'\\rangle \\le \\delta$, whose Euler-Lagrange condition reduces to a first-order eigenvalue ODE with eigenvalues $\\lambda = \\pm 1$, selecting a single Doppler-shifted travelling wave. This is what forces the optimal source to radiate only aft, and it is what converts the optimisation into a linear constraint problem with infinitely many solutions, e.g., step-function pressure distributions.","core_discovery":"The central discovery is that, in a linear 1D shallow-water model, the thrust from self-generated waves is the fore-aft difference of wave amplitude squared (radiation stress), and when the injected power is bounded the optimal pressure source is a purely aft-travelling wave. This yields dimensionless thrust equal to the power bound, $\\bar F_T = \\delta$, and formal efficiency $\\eta = 1$ for all drift velocities $v \\ge 0$, because no power is wasted on a forward wave. The optimum holds across the subcritical, critical, and supercritical regimes, defined by $v = U/c$ with $c = \\sqrt{gH}$, and in the supercritical case both waves travel leftward in the body frame yet the aft choice still achieves $\\bar F_T = \\delta$.","pith_inferences":["The 'no forward wave' principle suggests a general design rule for wave-propelled craft: regardless of speed, shape and oscillate the body so that all emitted wave energy is directed opposite to the direction of travel.","The same optimal-control template, with a Galilean transform and Doppler-shifted wavenumbers, might be applied to other linear wave-bearing media; a plausible inference is that efficiency-1 optimality extends to any energy-conserving linear wave equation without dissipation.","Adding surface tension, viscosity, or finite-depth dispersion would likely break the exact efficiency-1 result, so a testable extension is to compute the optimal fore-aft wave balance for gravity-capillary waves and check how much the aft-only rule changes."],"forward_implications":["In the bounded-power case, the optimal time-averaged thrust is exactly equal to the power bound, $\\bar F_T = \\delta$, at every drift velocity, so propulsive efficiency is 1 across subcritical, critical, and supercritical motion.","The optimal pressure source radiates no wave ahead of the body: all injected power goes into the aft wave, and this remains true for every value of the drift velocity.","Modulating the power bound $\\delta$ changes the cruising velocity according to $v_* = \\sqrt{2H\\delta/(L C_D)}$, so a slowly accelerating body can be driven from rest to supercritical speeds by increasing the injected power.","In the bounded-norm case, there are resonant velocities where the forward wave vanishes and efficiency reaches 1, but away from these resonances efficiency drops, so the bounded-power case is the one that guarantees optimal efficiency continuously.","There are infinitely many optimal pressure distributions satisfying the bounded-power conditions, including simple step functions, so the optimal strategy is a family rather than a unique waveform."],"supporting_citations":[{"why":"Supplies the radiation-stress mechanism that identifies thrust with the fore-aft difference in wave amplitude squared.","marker":"Longuet-Higgins & Stewart 1964"},{"why":"Applies radiation stress to a self-propelling craft, the wave-driven propulsion scenario the paper optimises.","marker":"Longuet-Higgins 1977"},{"why":"Provides the wave-driven propulsion model and the drag-law comparison used to define the cruising velocity.","marker":"Benham et al. 2024"},{"why":"The numerical optimisation framework (JuMP/Ipopt) used to validate the analytical optimality conditions.","marker":"Lubin et al. 2023"},{"why":"Models the body as a moving pressure region on the surface, the source representation adopted throughout the paper.","marker":"Doctors & Sharma 1972"}],"fun_headline_variants":["Self-generated waves achieve perfect propulsion efficiency","Aft wave yields 100% efficiency in wave-driven propulsion","Optimal wave propulsion from self-generated aft wave only","Perfect wave-driven propulsion: rearward wave is optimal","Wave propulsion reaches maximal thrust with aft wave"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The central claim rests on the linear 1D shallow-water wave model with no surface tension, viscosity, or 2D effects, and the acceleration recipe further assumes that the speed changes slowly enough for the periodic optimal solutions to hold at every instant.","fun_headline_variants_meta":{"raw":{"variants":["Self-generated waves achieve perfect propulsion efficiency","Aft wave yields 100% efficiency in wave-driven propulsion","Optimal wave propulsion from self-generated aft wave only","Perfect wave-driven propulsion: rearward wave is optimal","Wave propulsion reaches maximal thrust with aft wave"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000189,"raw_usage":{"total_tokens":1337,"prompt_tokens":948,"completion_tokens":389,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":564,"completion_tokens_details":{"reasoning_tokens":316}},"tokens_in":564,"tokens_out":389,"duration_ms":5266,"temperature":1.0,"reasoning_tokens":316,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T01:01:24.912001+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the thrust and the injected power of a real oscillating pressure source on a shallow water tank while resolving the fore and aft wave amplitudes. If the maximum achievable thrust-to-power ratio falls below 1 once surface tension, viscosity, and 2D losses are present, or if the optimal source cannot be made to emit zero wave ahead, the efficiency-1 claim fails.","supporting_citations":[{"cited_title":"& Stewart, R.W","cited_arxiv_id":null,"evidence_quote":"Supplies the radiation-stress mechanism that identifies thrust with the fore-aft difference in wave amplitude squared."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Applies radiation stress to a self-propelling craft, the wave-driven propulsion scenario the paper optimises."},{"cited_title":", Devauchelle, O","cited_arxiv_id":null,"evidence_quote":"Provides the wave-driven propulsion model and the drag-law comparison used to define the cruising velocity."},{"cited_title":", Dowson, O","cited_arxiv_id":null,"evidence_quote":"The numerical optimisation framework (JuMP/Ipopt) used to validate the analytical optimality conditions."},{"cited_title":"& Sharma, S.D","cited_arxiv_id":null,"evidence_quote":"Models the body as a moving pressure region on the surface, the source representation adopted throughout the paper."}],"review_version":1}