{"id":"442da22a-8fc5-443d-9fd0-ca9ba7e794dd","arxiv_id":"1908.07850","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Rigid millimeter fibers in strongly sheared Taylor-Couette turbulence align at a Reynolds-independent angle of -0.38π, and a simple-shear Jeffery model approximates the measured orientation distribution.","lead":"An experiment in a high-speed Taylor-Couette tank shows that millimetric plastic fibers in strongly turbulent, sheared water consistently tilt at about 68 degrees against the flow direction, no matter the rotation speed, fiber concentration, or position. The same tilt angle is roughly captured by the 103-year-old Jeffery equation for a particle in simple shear, suggesting that large fibers still feel local flow gradients.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Jeffery model peak (-0.27π) lies outside the measured range (-0.38π ± 0.05π); the 'explanation' is qualitative, not quantitative.","rationale":"The strongest claim has two components: a robust measurement and a point-particle explanation. The measurement appears solid: 64,000 images per case, consistent PDFs across Re, α, and position. The weak link is the explanation. The model peak is outside the measurement's uncertainty interval, and the missing 20° is attributed to effects that are not quantified. This is not an internal inconsistency—the model is explicitly approximate—but it means the abstract's 'can explain' overstates the quantitative support. Re-running the model with isotropic 3D initial conditions would test whether the projection and out-of-plane dynamics can close the gap. The reader's CONDITIONAL verdict already captures this; our stress test does not justify rejection because the empirical universal alignment and the qualitative shape agreement remain valuable.","tokens_in":10146,"tokens_out":8490,"duration_ms":92262,"concrete_test":"Run a 3D stochastic Jeffery simulation with initial orientations drawn uniformly on the unit sphere, a simple shear with the measured mean shear rate, and randomization every Cτ_l for C in [0.5, 4]. Project the resulting orientations onto the r-θ plane and compute the PDF of θp. If no C places the projected peak inside the measured 95% interval [-0.43π, -0.33π] while matching the PDF amplitude, the Jeffery mean-field explanation is quantitatively falsified; if some C does, the peak offset can be resolved by including out-of-plane randomization, and the claim survives.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The model comparison in Fig. 5 is the pivot of the explanatory claim. The paper reports a preferred orientation of -0.38π ± 0.05π, yet the Jeffery mean-field calculation gives a peak at about -0.27π. The offset (≈20°, 0.11π) is more than twice the quoted 9° uncertainty, so the model does not reproduce the headline number. The integration time Cτ_l is tuned (C chosen so the amplitude matches at about 2τ_l), not the peak, so the free parameter cannot fix this discrepancy. The manuscript attributes the shift qualitatively to inertia and turbulent fluctuations, but provides no model or measurement of that offset. Because the measured θp is the 2D projection of the fiber orientation onto the r-θ plane while the Jeffery integration is 3D, the comparison also presupposes that out-of-plane orientation is negligible; no evidence for this is given. The paper itself flags that Stk_r is not zero, so this is an acknowledged limitation, but it is exactly the load-bearing part of the claim that finite-size fibers retain point-particle-like orientation statistics. Thus the central assertion is empirically interesting, but the 'explained by Jeffery' part is only qualitative.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This experimental letter studies rigid millimetric fibers in a strongly sheared Taylor-Couette turbulent flow. Using high-speed imaging with 64,000 images per case, the authors measure the two-dimensional projection of the fiber orientation onto the radial-azimuthal plane and report a preferred orientation of −0.38π ± 0.05π (−68 ± 9°) relative to the mean azimuthal flow direction, with the result essentially independent of Reynolds number, fiber volume fraction, radial position, and axial position. They also measure fiber velocities and angular velocities, finding that the fibers follow the local azimuthal flow closely and that their angular velocity is strongly intermittent. To explain the preferred orientation, they integrate Jeffery's equation for ellipsoidal particles in a simple shear flow with the bulk mean shear rate, assuming that turbulence randomly reorients the fibers on a timescale of order τ_l. The resulting model PDF has a peak at approximately −0.27π, which is within about 15 to 18 degrees of the measured value, and the model amplitude is matched by choosing an integration time of about 2τ_l. The authors conclude that finite-sized anisotropic particles can retain point-particle-like signatures of the local turbulent flow.","tokens_in":10385,"tokens_out":3376,"duration_ms":35259,"significance":"The empirical result is significant: if correct, it shows a robust, geometry-independent preferential alignment of large fibers in high-Reynolds-number turbulence, with a 40% contrast between the most and least probable orientations, and it supports the practical use of simplified point-particle modeling for orientation statistics. The strengths of the paper are its careful and extensive measurements, the consistency of the orientation PDF across parameter variations, the direct measurement of fiber velocities showing agreement with the mean flow, and the documentation of strongly intermittent angular velocities with high kurtosis. However, the explanatory part is currently qualitative rather than quantitative: the Jeffery-model peak lies outside the quoted experimental uncertainty, the integration-time constant C is tuned to match the PDF amplitude, and the model does not account for rotational inertia or three-dimensional projection effects.","major_comments":[{"comment":"The measured headline orientation is −0.38π ± 0.05π, while the Jeffery-integration peak is reported as approximately −0.27π. The offset of roughly 0.11π (about 20°) is more than twice the quoted ±0.05π uncertainty, so the model does not quantitatively reproduce the central measurement. The text attributes the shift to inertial lag and turbulent fluctuations, but no calculation, model, or separate measurement quantifies either effect. Because the abstract and conclusions claim that Jeffery's equation explains the preferential alignment, this discrepancy is load-bearing; please either provide a quantitative account of the shift or reframe the claim as a qualitative shape comparison.","section":"Section 'In order to understand the preferential alignment...' and Fig. 5"},{"comment":"The stochastic mean-field model rests on the premise that each fiber sees a simple shear with the bulk mean shear rate and that turbulence only randomly resets the orientation every interval of order τ_l. This premise is not derived from the measurements or from an independent model of the fiber's Lagrangian velocity-gradient history. Moreover, the constant C is chosen so that C = 2τ_l matches the PDF amplitude; because the same parameter does not control the model peak, the partial agreement cannot be separated from this tuning. Please justify C from a measured Lagrangian correlation time or show explicitly that the predicted peak is insensitive to C.","section":"Section 'In order to understand the preferential alignment...', integration over t in [0, C τ_l]"},{"comment":"The measured θ_p is the two-dimensional projection of the fiber orientation onto the r-θ plane, whereas the Jeffery integration is performed in full three-dimensional orientation space. The comparison in Fig. 5 therefore presupposes that out-of-plane orientation and rotation are negligible. Since the fibers are free to rotate in all directions and the flow has secondary axial and radial velocities, this assumption needs support: for example, an estimate or measurement of the out-of-plane polar-angle distribution, or a projection of the three-dimensional model PDF onto the measurement plane, should be provided.","section":"Fig. 1C,D and Fig. 5"},{"comment":"The manuscript acknowledges that the fibers are not in the strictly inertialess limit, noting that Stk_r is of order 0.1 Stk_p, but it does not provide a value or a model for Stk_r under the present conditions. Since the unexplained peak shift is attributed to small inertia, a quantitative statement of the rotational Stokes number or a simple inertial correction would be needed to make the point-particle claim defensible. Without this, the statement that finite-size fibers still retain point-particle-like orientation statistics remains an assertion rather than a demonstrated result.","section":"Paragraph containing Stk_p and Stk_r"}],"minor_comments":[{"comment":"The name 'Jefferey's equation' is misspelled in the abstract; the correct spelling 'Jeffery' is used in the body and should be used consistently.","section":"Abstract"},{"comment":"There are typographical errors in the text, including 'an strongly sheared turbulent flow' and 'prefered alignment'; these should be corrected.","section":"Summary, second paragraph"},{"comment":"The sentence says the equations are 'duplicated here'; this should read 'reproduced here'.","section":"Eq. (1) preceding text"},{"comment":"The legend uses 'Re' while the text and other figures use 'Re_i'; please standardize the notation.","section":"Fig. 6"},{"comment":"The caption states 'A representation of the fiber alignment is shown at the top of the figure,' but the schematic is small and its relation to θ_p could be clarified for readability.","section":"Fig. 3 caption"}],"recommendation":"major_revision","confidential_remarks":"This is a well-executed experimental study with a compelling empirical result and clean data. The main risk is overclaiming explanatory power: the Jeffery-model comparison has a peak offset of about 20°, the integration-time factor is tuned, and the projection from 3D to 2D is not quantified. These are fixable within the manuscript's scope by reframing the conclusion as a qualitative agreement and reporting the necessary caveats, or by adding a more complete model. I would not recommend rejection; the measurement itself is solid and of broad interest. No concerns about citation practice or novelty disclosure."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"What you should know: this is a solid experimental letter whose main result is new and well supported. Rigid millimetric fibers in high-Reynolds Taylor–Couette flow align at roughly -0.38π relative to the mean flow, and that angle holds across Reynolds numbers, concentrations, and positions. The measurement looks careful: 64,000 images per case, fibers distributed homogeneously, and fiber velocities matching the flow. The orientation PDFs and the strongly intermittent rotation rates are credible and useful.\n\nThe empirical claim is the heart of the paper and it deserves credit. The paper also does a nice job estimating Stokes numbers at the fiber length scale, showing that the fibers are only mildly inertial despite being tens of Kolmogorov lengths long.\n\nWhere it gets softer is the explanation. The Jeffery mean-field model, with random reorientation every ~τ_l, produces a peak at about -0.27π. That sits outside the quoted ±0.05π uncertainty on the measured -0.38π. The paper attributes the offset to a slight inertial lag and to turbulent fluctuations, but that is a qualitative statement, not a quantitative derivation. The integration time C τ_l is chosen with C=2 to match the PDF amplitude, so part of the agreement is built in, and the peak offset is not fixed by that choice. There's also a projection issue: the measured θp is the 2D projection onto the r-θ plane, while the Jeffery calculation is fully 3D. The comparison assumes out-of-plane rotation is negligible, and that is not demonstrated. These are not fatal to the measurement, but they mean the phrase 'explain the preferential alignment using Jeffery's equation' is overreach. A fair summary would be: a simplified Jeffery model with a fitted decorrelation time reproduces the shape of the orientation distribution and gets within about 15 degrees of the preferred angle, with the residual shift attributed to inertia and turbulence.\n\nThe reference list is appropriate, and the self-citations are to the group's own apparatus and related particle work, which is legitimate. The model section is transparent about the C choice and notes that the rotational Stokes number is not zero, so the limitations are acknowledged even if not resolved.\n\nWho is this for? Experimentalists and modelers working on anisotropic particles in turbulence. The measured alignment angle is a clean, citable dataset even if the modeling remains unfinished. I would send this to peer review. A referee should press on the quantitative claim of the model and the projection correction, but the experimental result is solid and worth publishing.","headline":"A robust new experimental measurement of fiber alignment in strongly sheared turbulence, packaged with a Jeffery-based model that is honest but only qualitatively successful.","tokens_in":10917,"tokens_out":1721,"would_cite":true,"duration_ms":19581,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["47.27.-i","47.55.Kf"],"model":"deepseek-v4-flash","headline":"Millimetric fibers in turbulent shear flow align at a fixed −68° angle.","keywords":["fiber orientation","Taylor-Couette turbulence","Jeffery equation","point-particle approximation","anisotropic particles","rotational intermittency","Stokes number","sheared turbulence"],"falsifier":"Measure the full three-dimensional orientation of the fibers with stereoscopic imaging: if out-of-plane angles are large, or if the peak of the orientation PDF shifts with a change in the mean shear rate by more than the Jeffery-based prediction allows, then the simple-shear-plus-random-reset model is not the right mechanism.","tokens_in":9910,"feed_emoji":"🌀","tokens_out":4971,"duration_ms":44985,"temperature":0.7,"pith_summary":"This paper reports that rigid millimetric fibers suspended in strongly turbulent Taylor–Couette flow do not tumble randomly: they spend far more time oriented at about $\\theta_p = -0.38\\pi$ ($-68^\\circ$) to the mean azimuthal flow than at any other angle, a preference that survives across all tested Reynolds numbers, fiber concentrations, and radial and axial positions. The authors argue this is surprising because the fibers are large and inertial, with lengths tens of Kolmogorov lengths, yet they follow the fluid and keep point-particle-like orientation signatures. They show the alignment can be reproduced by integrating Jeffery's equation for ellipsoids in a simple shear whose rate is the bulk mean shear, with turbulent fluctuations restarting orientations every eddy-turnover time. A sympathetic reader would care because the result suggests simplified point-particle models can describe finite-sized anisotropic particles in turbulence, and because a single robust alignment angle may be exploited for flow sensing or fiber-laden industrial flows.","feed_headline":"Rigid fibers in turbulent shear lock onto a fixed tilt","feed_subtitle":"The same −68° tilt appears across all Reynolds numbers and positions, matching Jeffery's particle equation.","key_machinery":"The central object is Jeffery's equation for the orientation vector $p_i$ of an ellipsoidal particle in a viscous shear flow, repurposed as a stochastic mean-field process: every time interval of order $\\tau_\\ell$, the fiber's orientation is randomly reset, and between resets it evolves under a simple shear with the bulk mean shear rate $\\dot{\\gamma} = \\langle \\partial u_\\theta/\\partial r \\rangle$ over $\\tilde{r} \\in [0.25, 0.75]$. The aspect ratio $\\Lambda = 5.3$ enters through the shape factor $(\\Lambda^2 - 1)/(\\Lambda^2 + 1)$. This machinery converts turbulent fluctuations into a discrete reorientation process and yields an orientation PDF whose peak and shape can be compared with experiment; agreement is best for an integration time near $2\\tau_\\ell$. A supporting step is the revised Stokes number estimate: using a drag-corrected response time $\\tau_p$ and the fiber-scale eddy time $\\tau_\\ell$ gives $\\mathrm{Stk}_p \\approx 2$ to $3$, rather than the large values obtained with Kolmogorov scales, which explains why the fibers track the flow closely despite their size.","core_discovery":"The central claim is that finite-sized rigid fibers in high-Reynolds-number Taylor–Couette turbulence exhibit a statistically preferred orientation of $\\theta_p = -0.38\\pi \\pm 0.05\\pi$ with respect to the inner-cylinder wall, independent of $\\mathrm{Re}_i$ ($8.3\\times10^4$ to $2.5\\times10^5$), volume fraction ($0.025\\%$ to $0.100\\%$), radial bin, and axial position. The same PDF shape is found everywhere, with a 40% difference between the most and least probable orientations. The authors further claim that a stochastic mean-field model based on Jeffery's equation, using only the bulk mean shear rate and a reorientation time of order $\\tau_\\ell$, predicts a peak near $-0.27\\pi$, within about 15–18° of the measured value; they attribute the offset to inertia and to turbulence not captured by the model. They also find that the fiber angular velocity is Reynolds-number-independent and strongly intermittent, with kurtosis 34–40, which they interpret as evidence that even large fibers respond to local velocity gradients like small particles.","pith_inferences":["If the alignment angle is set mainly by the bulk shear direction, the same stochastic Jeffery recipe might predict preferred orientations in other shear-dominated turbulent geometries, with the angle set by the local mean velocity gradient.","The 15–18° offset between the measured peak and the model prediction is a testable handle: a systematic variation of fiber aspect ratio and density ratio should show whether the offset grows with particle inertia, as the paper qualitatively suggests.","Because the experiment measures a 2D projection, the true 3D orientation distribution could be broader; comparing these PDFs with direct numerical simulations of finite-size fibers in shear turbulence would settle how much out-of-plane rotation matters.","A practical consequence not pursued by the paper is that the stable, concentration-independent alignment angle could serve as a local flow-direction probe or as a constraint for rheological models of fiber suspensions."],"forward_implications":["Fiber orientation statistics in strongly sheared turbulence can be approximated without resolving the fiber's finite size: the bulk mean shear rate and one eddy time scale suffice.","The preferred angle $-0.38\\pi \\pm 0.05\\pi$ is a robust, geometry-specific signature independent of Reynolds number, fiber concentration, and position in the Taylor–Couette gap.","Finite-sized fibers with $\\mathrm{Stk}_p \\approx 2$ to $3$ still follow the local flow closely, so point-particle approaches may be extended to particles much larger than the Kolmogorov scale.","Fiber angular velocity intermittency (kurtosis 34–40) is far stronger than for spheres of similar size, consistent with the low rotational inertia of elongated bodies.","A single integration time of about $2\\tau_\\ell$ in the Jeffery-based model reproduces the measured orientation PDF shape, suggesting the fiber rotation is set by eddies comparable to the fiber length."],"supporting_citations":[{"why":"Provides Jeffery's equation for ellipsoidal particle orientation, the basis of the stochastic mean-field model.","marker":"[14]"},{"why":"Describes the Twente turbulent Taylor–Couette (T3C) facility in which all experiments were performed.","marker":"[40]"},{"why":"Supplies the measured azimuthal velocity profile used to show fibers follow the flow and to extract the bulk mean shear rate.","marker":"[45]"},{"why":"Gives the modified viscous particle response time that enters the revised Stokes number estimate.","marker":"[48]"},{"why":"Defines the fiber-scale eddy turnover time $\\tau_\\ell$ used for the reorientation interval and the Stokes number.","marker":"[49]"},{"why":"Supports the claim that the rotational Stokes number is about one tenth of the translational Stokes number for elongated ellipsoids.","marker":"[22]"},{"why":"Provides the review context that anisotropic particles align with flow gradients, which the paper extends to finite-size fibers.","marker":"[11]"},{"why":"Fully resolved simulations of finite-size fibers in turbulent channel flow, the baseline that the point-particle claim contrasts with.","marker":"[26]"}],"fun_headline_variants":["Fibers in turbulent shear always tilt the same way","Universal fiber tilt in sheared turbulence: −68°","Jeffery's equation predicts fiber tilt in turbulence","Rigid fibers align at −68° in strong turbulent shear","Intermittent spinning and fixed tilt: fibers in turbulence"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The explanation assumes each fiber sees only the bulk mean shear while turbulent fluctuations act as an instantaneous randomizer every few eddy-turnover times, and that the measured in-plane angle captures the dominant rotation.","fun_headline_variants_meta":{"raw":{"variants":["Fibers in turbulent shear always tilt the same way","Universal fiber tilt in sheared turbulence: −68°","Jeffery's equation predicts fiber tilt in turbulence","Rigid fibers align at −68° in strong turbulent shear","Intermittent spinning and fixed tilt: fibers in turbulence"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000854,"raw_usage":{"total_tokens":3725,"prompt_tokens":971,"completion_tokens":2754,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":587,"completion_tokens_details":{"reasoning_tokens":2673}},"tokens_in":587,"tokens_out":2754,"duration_ms":20430,"temperature":1.0,"reasoning_tokens":2673,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:54:50.886858+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the full three-dimensional orientation of the fibers with stereoscopic imaging: if out-of-plane angles are large, or if the peak of the orientation PDF shifts with a change in the mean shear rate by more than the Jeffery-based prediction allows, then the simple-shear-plus-random-reset model is not the right mechanism.","supporting_citations":[{"cited_title":"The motion of ellipsoidal particles immersed in a viscous fluid,","cited_arxiv_id":null,"evidence_quote":"Provides Jeffery's equation for ellipsoidal particle orientation, the basis of the stochastic mean-field model."},{"cited_title":"The Twente turbulent Taylor-Couette (t3c) facility: strongly turbulent (multiphase) flow between independently rotating cylinders,","cited_arxiv_id":null,"evidence_quote":"Describes the Twente turbulent Taylor–Couette (T3C) facility in which all experiments were performed."},{"cited_title":"Logarithmic Boundary Layers in Strong Taylor–Couette Turbulence,","cited_arxiv_id":null,"evidence_quote":"Supplies the measured azimuthal velocity profile used to show fibers follow the flow and to extract the bulk mean shear rate."},{"cited_title":"Turbulent Transport of Material Particles: An Experimental Study of Finite Size Effects,","cited_arxiv_id":null,"evidence_quote":"Gives the modified viscous particle response time that enters the revised Stokes number estimate."},{"cited_title":"Motion of inertial particles with size larger than Kolmogorov scale in turbulent flows,","cited_arxiv_id":null,"evidence_quote":"Defines the fiber-scale eddy turnover time $\\tau_\\ell$ used for the reorientation interval and the Stokes number."},{"cited_title":"Rotation of Nonspherical Particles in Turbulent Channel Flow,","cited_arxiv_id":null,"evidence_quote":"Supports the claim that the rotational Stokes number is about one tenth of the translational Stokes number for elongated ellipsoids."},{"cited_title":"Anisotropic Particles in Turbulence,","cited_arxiv_id":null,"evidence_quote":"Provides the review context that anisotropic particles align with flow gradients, which the paper extends to finite-size fibers."},{"cited_title":"Simulation of finite-size fibers in turbulent channel flows,","cited_arxiv_id":null,"evidence_quote":"Fully resolved simulations of finite-size fibers in turbulent channel flow, the baseline that the point-particle claim contrasts with."}],"review_version":1}