{"id":"b66ff355-f815-43ab-9e3f-3142013a13f0","arxiv_id":"2505.17788","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Gyrotactic swimmers in oscillatory channel flow gain a tunable mean drift, enhanced axial and lateral dispersion, and species separation, while quasi-steady Taylor-dispersion closures fail at high Womersley numbers.","lead":"This paper uses computer simulations to study how tiny swimming algae spread out in a vertical tube when the fluid is pumped up and down in pulses. It finds that changing the pulse frequency can make the algae drift, spread, mix, or separate into different groups, and that a common equation-based shortcut fails at high frequencies.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The predicted drift and separation are sensitive to the wall reflection rule: at Wo=0.2 the Lagrangian and Eulerian models give opposite drift directions, so the central tuning curves depend on an unvalidated boundary condition.","rationale":"The paper is an honest, internally consistent numerical study: the Lagrangian code reproduces the passive-particle benchmark, the high-Wo drift limit matches the analytic von Mises value, and the authors explicitly delimit the Eulerian GTD breakdown. The reader's verdict of CONDITIONAL is appropriate. In stress-testing, the most load-bearing assumption is not the gyrotactic torque model itself (which is standard and qualitatively supported by the pilot experiments) but the wall reflection rule (10). The paper's own comparison in Section 4.4 reveals that at Wo=0.2 the Eulerian no-flux model produces downward drift while the Lagrangian specular-reflection model produces upward drift. Since the abstract's claims of drift and species separation are quantitative and the tuning curves in Fig. 4 are the central results, an unsupported boundary condition could change the sign of the predicted effect. A computational test varying the boundary rule is cheap and would settle whether the qualitative claims are robust. This sharpens, rather than replaces, the reader's weakest assumption: the microswimmer model is the carrier of the predictions, and the boundary condition is the specific part of that model most directly shown to alter outcomes.","tokens_in":16856,"tokens_out":16099,"duration_ms":114730,"concrete_test":"Re-run the Lagrangian simulations of Figs. 4a-c with alternative boundary conditions: (i) elastic specular reflection as in Eq. (10), (ii) reflection with randomized orientation drawn from the stationary von Mises distribution, (iii) wall-sliding with orientation unchanged, and (iv) a short wall-residence time. Track U_e for C. augustae and D. salina at Wo=0.106, 0.2, 0.383, and 5.0. If the sign of U_e and the existence of a U_e^s crossover are invariant across these alternatives, the boundary-condition concern is resolved; if the sign flips at low Wo, the central claim needs qualification or direct experimental measurement of cell wall interactions.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative results are produced by the individual-based model of Section 3.2 with the specular reflection rule (10). Section 4.4 and Fig. 5y show that at Wo=0.2 the Lagrangian model gives upward drift while the Eulerian no-flux model gives downward drift for the same parameters (PéR=2, βR=1). The authors attribute the difference to the boundary treatment and state that specular reflection is 'a better representation' of real wall interactions, citing [35], but no quantitative experimental test of the wall rule is provided. Because the abstract's headline claims (drift, species separation, and the tuning of dispersion) are carried by the sign and magnitude of U_e and U_e^s in Figs. 4a-c and inset A, a hidden sensitivity of these quantities to the unvalidated wall boundary condition is the most load-bearing assumption. If real cells adhere, slide, or reorient at walls differently, the predicted drift can reverse and the separation crossover Wo=0.383 may shift or vanish.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper investigates the dispersion, drift, mixing, and separation of gyrotactic microswimmers in a two-dimensional vertical channel driven by an oscillatory pressure gradient, i.e., Womersley flow. The authors present qualitative experiments with D. salina, an individual-based Lagrangian model (Eqs. 7-8) with specular wall reflection (Eq. 10), and an Eulerian advection-diffusion model with a generalized Taylor dispersion (GTD) closure developed in Appendix A.2. Comparing these approaches, they find that the Womersley number tunes the mean excess drift U_e, the effective axial diffusivity D_e, and the lateral mixing time t_mix; that different species can be separated near a crossover Wo approximately equal to 0.383; and that the Eulerian GTD closure breaks down when Wo^2 Sc is of order one or larger.","tokens_in":17032,"tokens_out":5846,"duration_ms":52845,"significance":"The central message, that oscillatory shear provides a tunable control knob for dispersion, mixing, and species separation of active suspensions, is potentially useful for bioreactor and cell-separation applications. The numerical work has credible internal checks: passive-particle simulations reproduce Lee et al.'s dispersion results (Fig. 2a), and the high-Wo drift limit approaches the analytic von Mises value -0.7281 for lambda = 2.2 (Appendix A.2 and Figs. 4a-c). The two-dimensional GTD closure is derived in closed form and checked against asymptotic limits. However, the headline quantitative predictions (drift sign and magnitude, separation crossover) are carried by the Lagrangian model, whose wall-reflection rule is not validated independently, and the experiments are only qualitative motivation. The contribution is significant if the model is taken as a predictive tool, but the claims need to be framed or tested more carefully.","major_comments":[{"comment":"At Wo = 0.2 with PéR = 2 and βR = 1, the Lagrangian and Eulerian models predict opposite signs of the excess drift U_e, and the authors attribute this to the wall treatment, stating that specular reflection (Eq. 10) is 'a better representation' of real boundaries. However, no quantitative experimental test or independent measurement of cell-wall interactions is provided for the species and flows considered. Because the sign and magnitude of U_e in Figs. 4a-c and the separation crossover Wo ≈ 0.383 in inset A depend on how cells accumulate at the walls, the central claims rest on an unvalidated boundary condition. I request either a sensitivity study (e.g., partial absorption, slip, or reorientation at walls) or an explicit treatment of the wall rule as a modeling assumption whose consequences are mapped, with the claims softened accordingly.","section":"§4.4, Eq. (10), Fig. 5y"},{"comment":"The experimental section is qualitative: only two Wo values are shown, and there are no measured drift velocities, dispersion coefficients, mixing times, or quantitative comparisons with the simulation parameters in Table 1. The abstract and conclusion present drift, separation, and tuning as findings of the study; given that the experiments only motivate the models, these results should be described as predictions of the individual-based model unless quantitative validation is added. This does not invalidate the theoretical contribution, but the wording should not imply experimental confirmation.","section":"§2, Figs. 1b-m"},{"comment":"The quantitative claims in Fig. 4—the location of the drift maximum at Wo = 0.106, the sign change near Wo = 0.481, the separation crossover Wo ≈ 0.383, and the local extrema in D_e and t_mix—are based on single runs of 5000 particles with no error bars or convergence tests. Given the visible stochastic noise in the curves, I ask the authors to add uncertainty estimates (e.g., multiple seeds or bootstrapped confidence intervals) or to restrict claims to features that are robust across such estimates.","section":"§4.3, Fig. 4"}],"minor_comments":[{"comment":"The text reads 'the Womesley number, Wo'; this should be 'Womersley number.'","section":"Conclusion, p. 12"},{"comment":"The sentence 'the non-dimensional reorientation rate is dr = 1/Wo^2 Sc' should read 'rotational diffusivity' rather than 'reorientation rate,' since dr is the dimensionless rotational diffusivity.","section":"Appendix A.1"},{"comment":"The phrase 'Strongly gravitactic particles do not accumulate in the centre and at walls as much and, therefore, have smaller drift' is ambiguous; please specify with respect to which species or parameter values this comparison is made.","section":"§4.3, first bullet"},{"comment":"The definitions of V and tm are introduced quickly; please define them explicitly before using them in the formula for De*.","section":"§4.1, Eq. (15)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of the journal and the numerical framework is sound overall. The main risk is overclaiming: the wall-reflection sensitivity needs to be addressed before publication. If the authors add a small boundary-condition robustness study and tone down the experimental-validation language, I would support publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing you should know: this is the first paper I've seen that puts gyrotactic active particles into an oscillatory Womersley flow, and the core simulation results are internally consistent and honestly labeled. The combination is genuinely new—prior work did passive tracers in oscillatory flow and active particles in steady flow, but not both. The paper also delivers a new 2D generalized Taylor dispersion closure (Appendix A.2), which is real math and can be used elsewhere.\n\nWhat it does well: the Lagrangian simulations reproduce the passive-particle benchmark of Lee et al., and the high-Wo drift limit matches the analytic von Mises prediction of -0.7281. That gives me confidence the numerics are not off in la-la land. The central finding—that oscillatory flows can induce net drift, enhance axial and lateral dispersion, and separate species by motility—is supported by the simulations. The paper is also candid about the GTD Eulerian breakdown when Wo^2 Sc ~ 1. The experiments in Section 2 are explicitly preliminary and only qualitative motivation; the paper doesn't oversell them.\n\nSoft spots: the stress-test concern about the wall reflection rule is legitimate, though it's not hidden. At Wo=0.2, the Lagrangian model (specular reflection) gives upward drift while the Eulerian no-flux model gives downward drift for the same parameters. The authors attribute the difference to boundary treatment and argue specular reflection is better, citing [35], but there's no quantitative experimental test of that rule. Since the headline tuning curves in Figs. 4a-c come from the Lagrangian model, a different wall interaction—sliding, adhesion, or reorientation—could shift or reverse the predicted drift and move the separation crossover at Wo=0.383. That's the most load-bearing assumption in the paper, and it would benefit from explicit sensitivity analysis or experimental validation. Also, there are no error bars on the Langevin drift and dispersion values, and no code/data release, so the quantitative predictions are hard to independently reproduce.\n\nOverall: the central argument holds up as a theoretical/numerical study. The paper is for applied mathematicians and fluid dynamicists working on active suspension dispersion, and for anyone designing oscillatory-flow bioreactors, though the latter should treat the numbers as indicative rather than final. It deserves a serious referee. I'd send it to peer review, but I'd ask for a sensitivity analysis on boundary conditions and a clearer statement about what's robust and what's model-dependent.","headline":"Genuinely new combination of gyrotactic dispersion and oscillatory flow; internally consistent numerics, but the headline drift predictions lean on an unvalidated wall reflection rule.","tokens_in":17590,"tokens_out":2961,"would_cite":true,"duration_ms":23288,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["76Z05","92C17"],"pacs":[],"model":"deepseek-v4-flash","headline":"Oscillatory flows can be tuned to control drift, dispersion, mixing, and separation of gyrotactic swimmers.","keywords":["pulsatile flows","biofluid mechanics","Taylor dispersion","active suspensions","gyrotaxis","Womersley number","particle separation","mixing"],"falsifier":"If a dilute suspension of gyrotactic algae in a vertical tube under a zero-net-flux oscillatory flow shows no centre-of-mass drift that varies with the Womersley number, or if the predicted near-coincident drift of two species at the crossover frequency does not occur, the drift and separation mechanism is falsified.","tokens_in":16602,"feed_emoji":"🌊","tokens_out":9580,"duration_ms":76310,"temperature":0.7,"pith_summary":"The paper argues that a purely oscillatory pressure-driven flow in a vertical channel—Womersley flow—can be used to fine-tune the dispersion of a suspension of gyrotactic swimmers, the bottom-heavy cells that reorient under the competing action of viscous and gravitational torques. Using Lagrangian simulations of non-interacting point swimmers, it finds that the oscillation frequency controls not only axial and lateral effective diffusivity but also a non-zero mean drift, which passive tracers never exhibit under a zero-net-flux oscillation. It further reports that cells initially confined to the left and right halves of the channel can be mixed, and that species with different gyrotactic strengths acquire different drift speeds, enabling non-invasive separation. The paper also claims that the usual Eulerian closure based on generalised Taylor dispersion breaks down when the oscillation period is comparable to the cell reorientation time, roughly when $\\mathrm{Wo}^2\\,\\mathrm{Sc}\\gtrsim 1$. This matters because the same mechanism could mix or separate microorganisms in bioreactors without invasive forces.","feed_headline":"Oscillating flow steers, mixes, and separates swimming algae","feed_subtitle":"Tuning the Womersley number controls drift and dispersion of gyrotactic algae, enabling passive bioreactor mixing.","key_machinery":"The argument is carried by an individual-based microswimmer model, Eqs. (7)--(8): each cell is a point particle advected by the oscillatory channel flow plus a constant swimming velocity $V_s'\\hat{\\mathbf{p}}$, where the orientation angle $\\theta$ evolves under gyrotactic torque $(1/2B')\\cos\\theta$, vorticity advection $\\omega_z/2$, and rotational Brownian noise $\\sqrt{2d_r'}\\,dW_t$, with specular reflection at the walls. The control parameter is the Womersley number $\\mathrm{Wo}=R'\\sqrt{\\Omega'/\\nu'}$, which fixes the phase lag and shape of the velocity profile in Eq. (4). For the Eulerian comparison, the paper derives two-dimensional closed-form generalised-Taylor-dispersion expressions for the mean swimming direction $\\mathbf{q}$ and diffusion tensor $\\mathbf{D}$ as functions of shear rate, and feeds them into the advection--diffusion equation (11); the boundary criterion for that closure is $\\mathrm{Wo}^2\\,\\mathrm{Sc}\\sim 1$.","core_discovery":"On the paper's own terms, the central discovery is that oscillatory flows give a suspension of gyrotactic swimmers a net drift whose sign and magnitude depend on the Womersley number, and that this same knob tunes axial dispersion and cross-channel mixing. At small-to-intermediate $\\mathrm{Wo}$, the cells repeatedly alternate between centre-focused downwelling states, where they sample the fast core of the flow, and wall-spread upwelling states; the asymmetry of these two states in time and orientation produces drift, which for strongly gyrotactic algae peaks near $\\mathrm{Wo}=0.106$ and changes sign near $\\mathrm{Wo}=0.481$. At large $\\mathrm{Wo}$, the oscillatory velocity profile flattens into plug-like motion, advection can no longer organise the population, and the cells revert to their intrinsic upward swimming bias, giving a drift of about $-0.73$ in the dimensionless units used. The paper further maintains that the Eulerian generalised-Taylor-dispersion description, with its quasi-steady mean orientation and diffusion tensor, agrees with the individual-based model only in a window of $\\mathrm{Wo}$ and loses validity once $\\mathrm{Wo}^2\\,\\mathrm{Sc}\\gtrsim 1$, because the flow then changes on the same time scale as cell reorientation.","pith_inferences":["If the model transfers to real bioreactor suspensions, pulsing the flow rate rather than changing the mean flow becomes a non-invasive control that could keep cells suspended and mixed while a zero net flux protects fragile cells from pump damage.","The predicted sign change of the excess drift with $\\mathrm{Wo}$ implies a direct experimental calibration: tracking the cell cloud's centre of mass over a frequency sweep should show drift reversing near $\\mathrm{Wo}\\approx 0.48$ for strongly gyrotactic algae, a sharp test of the model.","The claimed breakdown of the quasi-steady Eulerian closure at $\\mathrm{Wo}^2\\,\\mathrm{Sc}\\gtrsim 1$ suggests that future continuum models should carry orientation as an explicit degree of freedom in the fast-oscillation regime, rather than averaging it out."],"forward_implications":["Oscillatory flows with zero mean flux can move gyrotactic cells net up or down depending on frequency; the drift peak occurs near $\\mathrm{Wo}=0.106$ for the base parameters.","Axial dispersion is largest at small $\\mathrm{Wo}$ and falls off roughly as $\\mathrm{Wo}^{-4}$ for large swimming Péclet number; the drift sign change coincides with a small local enhancement of dispersion.","Cross-channel mixing of initially right-half particles speeds up sharply around $\\mathrm{Wo}\\sim 1$, giving a practical mixing window.","Species with different gyrotactic biases separate in oscillatory flow; the relative drift between two model algae vanishes at a crossover Womersley number, analogous to gas-separation crossover frequencies.","Eulerian simulations using generalised Taylor dispersion should not be used for $\\mathrm{Wo}^2\\,\\mathrm{Sc}\\gtrsim 1$; there the quasi-steady orientation averaging breaks, and Lagrangian or full orientation-resolved descriptions are needed."],"supporting_citations":[{"why":"Defines Taylor dispersion, the passive-tracer baseline that active suspensions are compared against.","marker":"[1]"},{"why":"Supplies the gyrotactic torque balance that governs cell orientation in the Lagrangian model.","marker":"[3]"},{"why":"Establishes that biased swimmers disperse differently from tracers in pipe flow, motivating the active dispersion framework.","marker":"[5]"},{"why":"Provides the passive-particle oscillatory dispersion theory that the paper's simulations reproduce and extend.","marker":"[12]"},{"why":"Articulates the need for a full orientation-resolved description when flow varies on cell-reorientation time scales.","marker":"[26]"},{"why":"Supplies the population-level Eulerian model and closed-form orientation moments that the continuum comparison uses.","marker":"[27]"},{"why":"Forms the generalised Taylor dispersion foundation for computing mean swimming direction and diffusion tensor.","marker":"[29]"},{"why":"Is the benchmark whose passive oscillatory dispersion results validate the Lagrangian simulation method.","marker":"[34]"}],"fun_headline_variants":["Oscillatory flow fine-tunes drift and mixing of algae","Womersley flow controls swimming algae dispersion","Oscillating flows steer and separate motile cells","Oscillatory flow knobs control algae drift and separation","Oscillatory flows tune drift, mixing, and separation of algae"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The quantitative predictions hinge on the microswimmer equations (7)--(8): non-interacting point cells whose orientation obeys linear gyrotactic torque, vorticity advection, and rotational noise, and whose wall encounters are specular reflections; if real cells stick to walls, interact, or reorient differently, the predicted drift and separation speeds shift.","fun_headline_variants_meta":{"raw":{"variants":["Oscillatory flow fine-tunes drift and mixing of algae","Womersley flow controls swimming algae dispersion","Oscillating flows steer and separate motile cells","Oscillatory flow knobs control algae drift and separation","Oscillatory flows tune drift, mixing, and separation of algae"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000895,"raw_usage":{"total_tokens":3898,"prompt_tokens":1026,"completion_tokens":2872,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":642,"completion_tokens_details":{"reasoning_tokens":2788}},"tokens_in":642,"tokens_out":2872,"duration_ms":23565,"temperature":1.0,"reasoning_tokens":2788,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:41:50.985859+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"If a dilute suspension of gyrotactic algae in a vertical tube under a zero-net-flux oscillatory flow shows no centre-of-mass drift that varies with the Womersley number, or if the predicted near-coincident drift of two species at the crossover frequency does not occur, the drift and separation mechanism is falsified.","supporting_citations":[{"cited_title":"Dispersion of soluble matter in solvent flowing slowly through a tube","cited_arxiv_id":null,"evidence_quote":"Defines Taylor dispersion, the passive-tracer baseline that active suspensions are compared against."},{"cited_title":"Individual and collective fluid dynamics of swimming cells","cited_arxiv_id":null,"evidence_quote":"Supplies the gyrotactic torque balance that governs cell orientation in the Lagrangian model."},{"cited_title":"Dispersion of biased swimming micro-organisms in a fluid flowing through a tube","cited_arxiv_id":null,"evidence_quote":"Establishes that biased swimmers disperse differently from tracers in pipe flow, motivating the active dispersion framework."},{"cited_title":"On the longitudinal dispersion of passive contaminant in oscillatory flows in tubes","cited_arxiv_id":null,"evidence_quote":"Provides the passive-particle oscillatory dispersion theory that the paper's simulations reproduce and extend."},{"cited_title":"Foundation and challenges in modelling dilute active suspensions","cited_arxiv_id":null,"evidence_quote":"Articulates the need for a full orientation-resolved description when flow varies on cell-reorientation time scales."},{"cited_title":"Biased swimming cells do not disperse in pipes as tracers: a population model based on microscale behaviour","cited_arxiv_id":null,"evidence_quote":"Supplies the population-level Eulerian model and closed-form orientation moments that the continuum comparison uses."},{"cited_title":"Taylor dispersion of gyrotactic swimming micro-organisms in a linear flow","cited_arxiv_id":null,"evidence_quote":"Forms the generalised Taylor dispersion foundation for computing mean swimming direction and diffusion tensor."},{"cited_title":"Taylor dispersion in oscillatory flow in rectangular channels","cited_arxiv_id":null,"evidence_quote":"Is the benchmark whose passive oscillatory dispersion results validate the Lagrangian simulation method."}],"review_version":1}