{"id":"9318291e-a76e-4786-a967-0b3c0776fede","arxiv_id":"2611.17023","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"A settling flexible fiber in a hexagonal pillar array either zigzags straight down or drifts sideways at a fixed angle near 0.36 times the array's structural angle, depending on short-time fiber-pillar scattering.","lead":"This paper simulates a flexible fiber sinking under gravity through a hexagonal array of fixed pillars in a viscous fluid. It finds two long-term behaviors: a straight zigzag path with no sideways drift, and a skew path with a steady sideways drift, set by how the fiber bounces off individual pillars.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek/deepseek-v4-flash","headline":"The minimal model is not yet shown to predict migration modes: it is validated outside its stated regime and the single-collision map is never iterated into a mode-level prediction.","rationale":"The direct simulation evidence for two migration modes and for a correlation between short-time scattering patterns and long-time modes is internally consistent and reasonably supported by the collision-by-collision tracking in Fig. 6 and the orientation-robustness checks in Appendix B. The soft spot is the paper's additional claim that this long-time behavior can be predicted by the minimal model. That claim requires three things: validity of the model in the regime where it is applied, a closed update rule connecting successive collisions, and a quantitative derivation or test of the event-to-mode rule. The paper supplies validation at only one parameter point outside the stated d/L>>1 and Be<<1 regime, with initial conditions taken from the simulations being explained, and it supplies only a qualitative mapping from scattering sequences to modes. The reader's weakest_assumption identifies the d/L mismatch; my concern overlaps with it but places more weight on the missing mode-level predictive loop, since even a perfect single-collision map would not establish the headline prediction without an iteration scheme. This is a conditional-supporting concern rather than a refutation: the simulation observations can stand on their own, and a closed iterated-map test could resolve the issue. Because the reader's verdict is already CONDITIONAL for reasons that include this gap, no verdict adjustment is needed.","tokens_in":20361,"tokens_out":13287,"duration_ms":91712,"concrete_test":"Close the single-collision model of Eqs. (4)-(11) into a return map: after solving Eq. (11), set the next initial orientation to the escape angle from the current collision and set the next R_O(0) from the lattice geometry and updated fiber position, without feeding simulation outputs back in. Run this map at the current spacing (d/L about 1.5) and at one wider spacing with d/L>=2.5, and compare the predicted asymptotic In/Out sequence and mode label against the simulated labels for the (Be, xi) cases in Figs. 3 and 6. If the map cannot be closed, requires simulation-derived inputs, or misclassifies any zigzag/displacement case, the claimed prediction of migration modes fails; the two-mode simulation observation would remain credible, but the predictive-model claim would need to be downgraded.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing concern is not the simulation evidence but the theoretical prediction attached to it. Section IV.B.2 frames the model as valid for Be<<1 and d/L>>1 (b171), yet it is tested only at Be=10, xi=2.36, where d/L is roughly 1.5, with initial conditions theta_0 and R_O(0) taken from the simulations being explained. More importantly, the model outputs single-collision contact points via the constrained problem in Eq. (11), not migration modes. The step from a sequence of In/Out scattering events to a long-time mode is stated qualitatively (all-In implies zigzag; cyclic In/Out implies displacement) and is never implemented as an iterated map over successive pillars. Fig. 7(c)-(d) is a one-collision phase diagram; nothing shows that iterating it reproduces the simulated mode labels across the (Be, xi) grid. The abstract's claim that long-time behavior can be predicted using a theoretical model with minimal ingredients is therefore not yet supported by the presented analysis.","agreement_with_reader":"partial"},"referee_report":null,"author_rebuttal":null,"desk_editor":{"model":"deepseek/deepseek-v4-flash","letter":"Colleague,\n\nThis paper reports a clean numerical finding: a flexible fiber sedimenting through a hexagonal pillar array falls into one of two long-time modes—zigzag (no net lateral displacement) or displacement (net lateral drift at a constant angle ~0.36 phi_s). The mode diagram in (Be, xi) space is directly supported by simulations, and the short-time scattering classification (In/Out events) is documented with collision-by-collision tracking. The constant migration angle for all displacement cases is a genuinely new observation and, if it holds up, gives a simple design rule for gravity-driven DLD sorting of elongated particles. The simulations themselves appear competently done (RPY hydrodynamics, resolved contact mechanics, GPU-accelerated).\n\nThe soft spots are real but not fatal. The biggest one is the gap between the theoretical model and the paper's claim that it 'predicts' the migration mode. Section IV.B.2 presents a single-collision model that computes contact points given initial conditions taken from the simulations it is meant to explain. The model is validated at one parameter point (Be=10, xi=2.36) that violates its own stated assumption d/L >> 1 (here d/L ~ 1.5). The step from single-collision outcomes to long-time mode is qualitative: all-In implies zigzag, cyclic In/Out implies displacement. The model is never iterated over successive pillars to produce a mode prediction independently of the full simulation. The paper acknowledges the d/L >> 1 assumption (b171) and the limited validation, so this is not a hidden flaw, but the abstract's claim that 'long-time behavior can be predicted using a theoretical model with minimal ingredients' overstates what is actually shown.\n\nOther concerns are minor. The contact-point validation uses theta_0 and R_O(0) from the same runs being explained—partial circularity, but the authors present it as a consistency check, not a blind test. No code or data are shipped, which limits reproducibility but is common for this subfield. The parameter sweep is limited to one lattice geometry (hexagonal, fixed d/omega, n=2).\n\nWho is this for? Researchers working on DLD sorting of elongated particles or on fiber transport in porous media. The simulation phenomenology is solid and useful. The theoretical model is a helpful sketch but not yet a predictive tool.\n\nRecommendation: I would send this to peer review. The simulation results are worth publishing, and the model can be tightened in revision. The authors should be asked to either (a) iterate the model over multiple pillars to show it reproduces the mode diagram, or (b) tone down the predictive claim and present the model as a qualitative rationalization. Either path is acceptable. The paper is not ready for a high-profile journal as is, but it deserves a serious referee.","headline":"Clear numerical phenomenology of fiber modes in a pillar array; the theoretical model is a sketch, not a predictor, and the paper's own stated assumptions limit its reach.","tokens_in":21277,"tokens_out":677,"duration_ms":9812},"pith_extraction":{"msc":["76D07","76M28","74F10"],"pacs":["47.15.G-","47.63.mf","47.57.E-"],"model":"deepseek/deepseek-v4-flash","headline":"Flexible fibers settling through pillar arrays fall into two migration modes set by initial collisions","keywords":["flexible fibers","sedimentation","pillar arrays","deterministic lateral displacement","zigzag mode","displacement mode","Stokes flow","fiber-pillar collisions"],"falsifier":"An experiment or simulation in which a fiber with Be = 10, ξ = 2.36 is released with an initial horizontal offset such that its first collision is an Out-scattering event rather than the two consecutive In-scattering events observed; if the fiber still enters displacement mode with ϕ ≈ 0.36 ϕ_s, the claim that the mode is set by the first few collisions is supported; if the mode changes, the deterministic link between short-time scattering and long-time mode would be falsified.","tokens_in":21702,"feed_emoji":"🧬","tokens_out":1643,"duration_ms":12131,"temperature":0.7,"pith_summary":"This paper uses numerical simulations to show that a single flexible fiber settling under gravity through a periodic array of rigid pillars will, after a long time, follow one of two trajectories: a zigzag mode with no net lateral drift, or a displacement mode with a constant lateral migration angle. Which mode the fiber enters is determined not by slow long-term dynamics but by the first few collisions with pillars. The authors identify two elementary scattering events—In-scattering and Out-scattering—and show that the cyclic alternation between them produces the displacement mode, while repeated In-scattering produces the zigzag mode. They also propose a minimal theoretical model, based on the fiber's settling trajectory in an obstacle-free fluid and the geometry of the array, that predicts which scattering events will occur for a given fiber stiffness and length, provided the lattice spacing is much larger than the fiber length. If correct, the work provides a rationale for designing gravity-based deterministic lateral displacement devices that sort fibers by their mechanical properties without requiring external flow.","feed_headline":"Two migration modes set by first collisions in fiber-pillar arrays","feed_subtitle":"A flexible fiber settling through a hexagonal lattice either zigzags straight or drifts laterally at a constant angle, determined by…","key_machinery":"The central objects are two scattering event types—In-scattering (contact angle and contact point on opposite sides of the pillar) and Out-scattering (contact angle and contact point on the same side)—which together form a discrete dynamical system describing the fiber's path through the array. The key identity is that the migration angle in the displacement mode is ϕ ≈ 0.36 ϕ_s, independent of Be and ξ, implying that the array's structural angle alone sets the lateral drift once the displacement mode is triggered. The theoretical model uses the obstacle-free settling trajectory (Eq. 10) to compute contact points, assuming d/L ≫ 1 so that each collision involves only one pillar.","core_discovery":"The paper establishes that the long-time trajectory of a flexible fiber sedimenting through a hexagonal array of pillars is classifiable into two distinct modes—zigzag (no average lateral displacement, migration angle ≈ 0) and displacement (net lateral drift with a constant migration angle ϕ ≈ 0.36ϕ_s, independent of fiber stiffness and length in the explored range). Crucially, the mode is determined by the short-time scattering dynamics: the fiber's first few collisions with pillars set up a repeating pattern of either purely In-scattering events (leading to zigzag) or a cyclic alternation of In- and Out-scattering events (leading to displacement). The authors further demonstrate that the escape angle from one collision becomes the initial angle for the next, and that this mapping can be predicted from a theoretical model combining the obstacle-free settling trajectory with the array geometry, under the assumption d/L ≫ 1.","pith_inferences":["The result that ϕ ≈ 0.36 ϕ_s is independent of Be and ξ suggests that the displacement mode corresponds to a geometric resonance between the fiber's post-collision drift and the lattice periodicity; a similar constant-angle property might hold for other lattice types (e.g., square arrays) but would need separate investigation.","If the assumption d/L ≫ 1 is violated (as it is for the longest fibers studied, where L/d ≈ 1.78), multi-pillar contacts may alter the scattering map; the paper's theoretical model would need extension to handle such cases, but the numerical results show that zigzag mode can still occur for ξ = 6.28, suggesting that the mode classification may be robust even when the assumption breaks.","The out-of-plane motion observed for Be = 100, ξ = 2.36 (reorientation in the yz-plane) implies that 2D models of fiber transport through pillar arrays may miss important three-dimensional effects; future experimental designs should account for potential fiber rotation out of the intended sorting plane."],"forward_implications":["A gravity-driven DLD device can sort flexible fibers by stiffness and length without external flow: stiff fibers (low Be) enter displacement mode and drift laterally, while flexible fibers (high Be) zigzag straight through.","The migration angle in displacement mode is constant (ϕ ≈ 0.36 ϕ_s) for all fiber parameters tested, meaning the device's sorting angle is set solely by the array geometry once the mode is triggered.","The mode (zigzag vs displacement) is determined within the first few collisions, so short arrays may suffice for classification, potentially enabling compact sorting devices.","The theoretical model provides a design tool: given fiber properties (Be, ξ) and array geometry (d, ϕ_s), one can predict whether the fiber will be displaced or not, without running full simulations."],"supporting_citations":[{"why":"Previous study by the same authors on flexible fiber settling against a single obstacle; provides the foundation for understanding gliding vs trapping and the lateral displacement magnitude used here.","marker":"[20]"},{"why":"Li et al. 2013, sedimentation of flexible filaments; supplies the theoretical model for the obstacle-free settling trajectory (Eq. 10) that the paper adapts to predict scattering events.","marker":"[18]"},{"why":"Marchetti et al. 2018, deformation of a flexible fiber settling in a quiescent fluid; provides the equilibrium shapes and the elastogravitational number Be used to characterize fiber stiffness.","marker":"[19]"},{"why":"Li et al. 2026, microfluidic band-pass filter for flexible fiber separation; demonstrates size selectivity in flow-driven pillar arrays, which the present work extends to gravity-driven settling.","marker":"[15]"},{"why":"Chakrabarti et al. 2020, transport of semiflexible polymers in periodic post arrays; shows trapping, gliding, and vaulting modes in flow, providing context for the scattering classification.","marker":"[14]"}],"fun_headline_variants":[],"cache_read_input_tokens":0,"weakest_assumption_plain":"The theoretical predictor assumes d/L ≫ 1 (lattice spacing much larger than fiber length), so the fiber never touches more than one pillar during a collision; the model is only validated at Be = 10, ξ = 2.36 where d/L ≈ 1.5 (O(1) rather than ≫ 1), and for the longest fibers (ξ = 6.28) the fiber is longer than the lattice spacing, making multi-pillar contact plausible.","fun_headline_variants_meta":{"raw":{"variants":[]},"model":"deepseek/deepseek-v4-flash","effort":"low","cost_usd":7.4e-05,"raw_usage":{"cost":7.37282e-05,"is_byok":false,"cost_details":{"upstream_inference_cost":7.37282e-05,"upstream_inference_prompt_cost":7.33642e-05,"upstream_inference_completions_cost":3.64e-07},"total_tokens":1013,"prompt_tokens":1011,"completion_tokens":2,"prompt_tokens_details":{"audio_tokens":0,"video_tokens":0,"cached_tokens":256,"cache_write_tokens":0},"completion_tokens_details":{"audio_tokens":0,"image_tokens":0,"reasoning_tokens":0}},"tokens_in":1011,"tokens_out":2,"cost_source":"openrouter_usage","duration_ms":3381,"image_pages":[],"temperature":1.0,"billed_attempts":1,"reasoning_tokens":0,"transport_provider":"openrouter","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-11-13T07:29:43.715331+00:00","model_set":{"reader":"deepseek/deepseek-v4-flash"},"pipeline_version":"pith-pipeline@v1.4.1-daily-deepseek-or-medium","input_kind":null,"input_chars":null,"input_total_chars":null,"passes_missing":null,"published_standing":null,"published_confidence":null,"objection_audit":null,"run_record":null,"basis":{"kind":"not_recorded","source":null,"chars_given":null,"chars_total":null},"receipt":null,"falsifier":"An experiment or simulation in which a fiber with Be = 10, ξ = 2.36 is released with an initial horizontal offset such that its first collision is an Out-scattering event rather than the two consecutive In-scattering events observed; if the fiber still enters displacement mode with ϕ ≈ 0.36 ϕ_s, the claim that the mode is set by the first few collisions is supported; if the mode changes, the deterministic link between short-time scattering and long-time mode would be falsified.","supporting_citations":[{"cited_title":"Obstacle-induced lateral dispersion and nontrivial trapping of flexible fibers settling in a viscous fluid, Physical Review Fluids8, 044303","cited_arxiv_id":null,"evidence_quote":"Previous study by the same authors on flexible fiber settling against a single obstacle; provides the foundation for understanding gliding vs trapping and the lateral displacement magnitude used here."},{"cited_title":"The sedimentation of flexible filaments, Journal of Fluid Mechanics 735, 705","cited_arxiv_id":null,"evidence_quote":"Li et al. 2013, sedimentation of flexible filaments; supplies the theoretical model for the obstacle-free settling trajectory (Eq. 10) that the paper adapts to predict scattering events."},{"cited_title":"´E. Guazzelli, and C. Duprat, Deformation of a flexible fiber settling in a quiescent viscous fluid, Physical Review Fluids3, 104102","cited_arxiv_id":null,"evidence_quote":"Marchetti et al. 2018, deformation of a flexible fiber settling in a quiescent fluid; provides the equilibrium shapes and the elastogravitational number Be used to characterize fiber stiffness."},{"cited_title":"A microfluidic band-pass filter for flexible fiber separation, Proceedings of the National Academy of Sciences123, e2520537123","cited_arxiv_id":null,"evidence_quote":"Li et al. 2026, microfluidic band-pass filter for flexible fiber separation; demonstrates size selectivity in flow-driven pillar arrays, which the present work extends to gravity-driven settling."},{"cited_title":"Trapping, gliding, vaulting: Transport of semiflexible polymers in periodic post arrays, Soft Matter16, 5534","cited_arxiv_id":null,"evidence_quote":"Chakrabarti et al. 2020, transport of semiflexible polymers in periodic post arrays; shows trapping, gliding, and vaulting modes in flow, providing context for the scattering classification."}],"lane_accuracy":{"status":"not yet measured","lane":{"model":"deepseek/deepseek-v4-flash","input_kind":null},"url":"https://pith.science/accuracy"},"review_version":1}