{"id":"5bff676f-f639-4709-8337-bafd734b8728","arxiv_id":"1908.03943","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"3D simulations show that stress overshoot and transient shear banding in jammed soft solids are robust across different damping models and boundary conditions, with sample age controlling the effect.","lead":"This paper uses 3D computer simulations of soft jammed spheres to show that a stress overshoot and transient shear banding during shear start-up occur regardless of the type of damping or boundary conditions used. A generalist reader might care because this suggests these yielding phenomena are real material physics, not numerical artifacts, which matters for processing soft solids like pastes and gels.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 'large enough samples' qualifier in the central claim is never tested: a single 97,556-particle cubic cell of side ~42a cannot establish that transient banding is an asymptotic, size-independent feature.","rationale":"The reader's weakest_assumption identifies the same gap I consider most load-bearing: the single system size is implicitly assumed to be in the asymptotic regime, and 'large enough samples' is asserted without a finite-size study. I did not focus on the second part of the reader's weakest_assumption, the Q≈1 proxy for overdamped dynamics, because the paper's own Q-sweep (Q=0.5 to 100, Figs. 9 and 11) provides partial support and the finite-size issue is more directly tied to the literal wording of the abstract. The paper otherwise does a genuine protocol comparison: DPD versus Stokes-like drag, Lees-Edwards versus walls, different DPD cutoffs, and different cooling rates. These are independent checks within the paper and strengthen the claim that the phenomena are not artifacts of one particular numerical implementation. However, the central claim is explicitly qualified by system size, and that qualifier is never operationalized or tested. The verdict CONDITIONAL is therefore appropriate, and I recommend no change to it.","tokens_in":15750,"tokens_out":4029,"duration_ms":48076,"concrete_test":"Run the LEBC1 Q=1, gamma_dot=1e-4 startup protocol on the same well-annealed preparation (Gamma=5e-4) for at least three system sizes at fixed volume fraction, e.g., N≈12,000, 97,556, and 780,000 (Ly≈21a, 42a, and 84a with matching Lx and Lz), using three independent initial configurations per size. For each run, compute (i) overshoot stress and overshoot strain, (ii) a quantitative band metric, e.g., the maximum deviation of the local shear rate from affine or the width of the region with negative velocity slope, and (iii) the strain at which the velocity profile returns to homogeneous. If any of these quantities changes systematically with Ly beyond sample-to-sample scatter, or if the band is absent at the largest size, the 'large enough samples' qualifier fails; if the results collapse, the finite-size concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is conditional on 'large enough samples,' but the paper never tests this condition. Section II states that all simulations use 10^5 (97,556) particles, and no finite-size variation is reported anywhere; the system is one approximately cubic cell of side ~42a. The transient banding shown in Figs. 6-8 is a single band spanning the gradient direction (y from about -20a to 20a), with periodic or wall boundary conditions at the cell edges. In this geometry, the observed instability may reflect the finite cell: a band of intrinsic width comparable to Ly can be stabilized by periodic images, or wall nucleation can dominate the entire sample. Either behavior would not persist in a genuinely 'large enough' system where bands nucleate and persist locally without spanning the cell. Because the abstract's qualifier is doing real work, the absence of any size extrapolation means the data support the weaker claim 'in this N, this Ly, with these protocols' rather than the stated asymptotic robustness. The qualitatively different wall behavior in Section IV (bands always nucleate near walls and take longer to homogenize) shows boundary effects are not negligible, making the missing size study more consequential, not less. A finite-size study is therefore the single most load-bearing gap between the evidence and the stated claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports three-dimensional molecular dynamics simulations of a polydisperse, athermal soft-sphere jammed solid (volume fraction about 0.70) sheared at finite rates, with the aim of identifying which features of the shear start-up transient are intrinsic to the material rather than numerical artifacts. The authors prepare samples with different cooling rates, characterize their linear viscoelastic and structural properties, and then compare three shearing protocols: Lees-Edwards boundary conditions with DPD-type pairwise drag, Lees-Edwards boundary conditions with a Stokes-like single-particle drag, and a wall-confined geometry. For a well-annealed sample at a low shear rate, all protocols show a stress overshoot, a transient flow instability that develops near yielding, a back-flow region in the velocity profile, and eventual recovery of homogeneous flow; these features are accompanied by a positive first normal stress difference and changes in the fraction of icosahedrally coordinated particles. The paper also varies the damping strength via an inertial quality factor Q, varies the DPD cutoff, and compares different sample ages. The central claim, stated in the abstract, is that the stress overshoot and transient shear banding are robust features for overdamped systems in large enough samples, independent of the specific drag and boundary conditions.","tokens_in":15864,"tokens_out":3679,"duration_ms":44905,"significance":"If the central claim holds, the paper makes a useful contribution to the debate on shear banding in soft jammed solids: it provides evidence that transient banding during yielding is not an artifact of a particular thermostat, drag model, or boundary condition, and it extends observations to three dimensions at finite shear rates. The systematic comparison of two drag implementations and wall versus Lees-Edwards boundary conditions is a genuine strength, as is the explicit study of sample age through different cooling protocols. The paper also offers falsifiable predictions, namely that the overshoot magnitude and the banding phenomenology should persist under changes of damping and boundary conditions at low rates. However, the evidence presented is largely qualitative and based on a single system size and, in the key figures, on a single initial configuration; the quantitative support for the 'large enough samples' and 'overdamped' qualifiers is not yet at the level claimed in the abstract.","major_comments":[{"comment":"","section":"Section II; abstract"},{"comment":"","section":"Section IV; Figs. 6-8"},{"comment":"","section":"Section III; Section V"},{"comment":"","section":"Section V; Fig. 11"}],"minor_comments":[{"comment":"","section":"Section VI; Fig. 12"},{"comment":"","section":"Fig. 5 caption"},{"comment":"","section":"Fig. 3 caption"},{"comment":"","section":"Throughout"},{"comment":"","section":"Section IV"},{"comment":"","section":"Section II"}],"recommendation":"major_revision","confidential_remarks":"The manuscript appears to be part of a larger research program, with a companion letter [64] cited for the dependence of band persistence on sample age. Please check with the authors that the present paper does not duplicate material from [64] and that the overlap is clearly disclosed. The paper would also benefit from a data-availability statement, especially since it emphasizes robustness; making the initial configurations and analysis scripts available would substantially increase the value of the study. In my view, the finite-size issue is the key obstacle: without a size study, the abstract's 'large enough samples' claim is not supported by the evidence reported here."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know this paper is a solid, careful 3D simulation study of transient shear banding in jammed soft spheres. The new thing is the systematic protocol comparison: two drag formulations (DPD and Stokes-like) and two boundary conditions (Lees-Edwards and walls), all at finite shear rates, in 3D. The authors show that the stress overshoot and the accompanying transient band appear across all protocols, which is the right way to argue these are not numerical artifacts. That claim, for the system size and parameter range tested, is well supported.\n\nWhat it does well: the load curves and velocity profiles are clearly presented, the distinction between protocol-dependent details and robust features is exactly the right question, and the check on the damping coefficient (Q from 0.5 to 100) strengthens the case that the phenomenon isn't an artifact of the specific thermostat. The paper also connects the banding to the evolution of local icosahedral packing and normal stress differences, which lends some microscopic texture. The methods are detailed enough to reproduce, though the modified LAMMPS code is not released.\n\nThe soft spots are real but not fatal. The biggest one is the qualifier 'large enough samples.' The paper simulates only one system size (97,556 particles, box side ~42a) and never tests whether the banding persists for larger systems. The stress-test note is right: the bands shown are localized in the gradient direction, but the cell could still influence nucleation or band width, and the wall protocol shows that boundary effects are nontrivial. So the abstract's 'large enough samples' is doing work that the evidence doesn't yet cover. A finite-size study, even two more sizes, would close the gap.\n\nSecond, Fig. 11 compares overshoot magnitude across protocols without error bars, even though independent samples exist. The apparent collapse could be partly noise. Third, the banding is characterized qualitatively; there is no defined metric (e.g., local shear rate variance) or statistical threshold. These are minor-to-moderate issues; they don't overturn the central observation.\n\nThe Q≈1 approximation for the overdamped limit is taken from the literature, but the Q-sweep shows banding even for underdamped cases, so the overdamped-specific claim is not directly probed. That's a smaller concern.\n\nAll told, this paper deserves a serious referee. The claim is somewhat broader than the evidence, and the missing size study is the main thing I'd want addressed. But the core finding is almost certainly correct, and the protocol comparison is a useful contribution. I'd cite it in my own work on soft glassy rheology.\n\nRecommendation: send it to peer review, with a request for a finite-size check and error bars on the rate-dependent comparison.","headline":"A careful 3D simulation study that robustly shows transient banding across protocols, but the 'large enough samples' qualifier is never tested.","tokens_in":16504,"tokens_out":2419,"would_cite":true,"duration_ms":24804,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that stress overshoot and transient shear banding are robust features of yielding overdamped jammed solids, not artifacts of the simulation setup.","keywords":["shear banding","stress overshoot","soft jammed solids","transient rheology","yielding","dissipative particle dynamics","Lees-Edwards boundary conditions","molecular dynamics simulation"],"falsifier":"Run the same well-annealed sample preparation and low-rate shear with the shear-gradient dimension doubled, tripled, or increased by an order of magnitude while keeping the thickness-to-particle-size ratio and shear rate fixed. If the overshoot magnitude shrinks, the transient band becomes wall-localized, or the banding disappears at larger system sizes, then the claim that banding is intrinsic would be falsified. A complementary check would be to simulate at $Q$ well below 1, such as $Q = 0.01$, to directly probe the overdamped limit rather than relying on $Q \\approx 1$ as equivalent.","tokens_in":15406,"feed_emoji":"🧪","tokens_out":5792,"duration_ms":57972,"temperature":0.7,"pith_summary":"The paper sets out to decide whether the stress overshoot and transient shear banding seen when well-aged jammed solids start to yield are genuine material behavior or numerical artifacts. By running 3D simulations of a polydisperse soft-sphere model at finite shear rates, with two different damping schemes and two different ways of applying shear, the study finds the same qualitative sequence in every case: linear loading, an overshoot, a spatially localized band of fast flow alongside nearly stuck material, and eventual return to homogeneous flow. The conclusion is that, for sufficiently large and well-annealed samples at low shear rates, the overshoot and transient banding are robust features of the overdamped dynamics of jammed solids, while the quantitative details of the load curve do depend on the drag and boundary choices. Poorly annealed samples do not show the overshoot, which ties the phenomenon to sample age and frozen-in stresses. This matters because it separates what is intrinsic to yielding from what is imposed by the simulation setup, and it supports looking for the same phenomenology in experiments.","feed_headline":"Stress overshoot and shear bands are no numerical artifact","feed_subtitle":"3D simulations show the transient flow localization persists across damping models and with periodic or wall boundaries.","key_machinery":"The central machinery is a set of 3D athermal simulations of about 97,556 polydisperse soft spheres interacting through a truncated and shifted Lennard-Jones potential at a volume fraction near 0.70. Shear is imposed either through Lees-Edwards periodic boundary conditions or by confining the sample between two frozen walls; dissipation is either pairwise through dissipative particle dynamics or particle-wise through a Stokes-like drag. The inertial quality factor $Q = \\tau_{\\text{damp}}/\\tau_{\\text{vib}}$, the ratio of the drag time to the vibrational time, quantifies how overdamped the dynamics are, and the paper uses $Q \\approx 1$ as its overdamped working point, with additional tests spanning $Q = 0.5$ to $Q = 100$. The robustness argument rests on tracking velocity profiles across the shear-gradient direction in successive strain windows: homogeneous flow before the overshoot, a band with back-flow during the stress decay, and return to homogeneous flow in steady state.","core_discovery":"The paper claims that in well-aged, deeply jammed polydisperse soft-sphere solids under start-up shear, the stress overshoot and the accompanying transient shear banding are intrinsic consequences of the material's yielding rather than numerical artifacts. It supports this by repeating the same deformation with two boundary-condition schemes (Lees-Edwards periodic and wall-confined) and with different dissipation implementations (pairwise dissipative-particle-dynamics drag, Stokes-like free-draining drag, and pairwise drag with a transverse contribution), and by varying the damping strength over two decades of an inertial quality factor. In every protocol at low shear rate, the load curve develops an overshoot, the velocity profile develops a banded region during the stress decay, and the band eventually disappears as flow becomes homogeneous. The abstract's stated caveat is that this robustness holds for large enough samples, and the phenomenon is controlled by sample age, since poorly annealed samples show no overshoot.","pith_inferences":["If the robustness claim holds, transient shear banding during start-up could serve as a non-invasive diagnostic of sample age in jammed soft solids: the presence and height of the overshoot read out how deeply annealed the glass is.","The protocol dependence of the post-overshoot decay suggests that transient start-up rheology may encode information about the viscous dissipation mechanism itself, not just the yield stress, which is an implication the paper does not develop.","A direct testable extension would be to compare the simulated velocity profiles with 3D particle-tracking rheology on dense emulsions or microgels at matched Weissenberg numbers, looking for the same back-flow signature during the overshoot decay.","The reported variation with $Q$ hints that inertia softens the post-overshoot decay; at higher rates or lower damping one would expect the banding window to narrow, a prediction that could be probed by changing particle mass in simulations or solvent viscosity in experiments."],"forward_implications":["Low-rate start-up shear of a well-aged jammed solid should generically show a stress overshoot followed by transient flow localization, regardless of whether deformation is imposed uniformly or through walls.","The transient band's lifetime and the stress decay after the overshoot depend on the microscopic damping mechanism, so quantitative comparisons between simulations must account for how dissipation is implemented.","Because the band disappears in steady state, observing localization only during start-up does not require a non-monotonic constitutive flow curve.","Sample age is the controlling factor: fast-quenched samples lack the overshoot and therefore lack transient banding, so the aging state should be reported alongside rheological data.","The positive first normal stress difference that grows while the bands develop indicates that dilation accompanies the inhomogeneous flow, coupling transient banding to volume-change tendencies."],"supporting_citations":[{"why":"Supplies the frozen-wall shearing protocol used for the wall-based comparison.","marker":"[4]"},{"why":"Provides the framework of transient shear banding and the back-flow signature that the paper tests.","marker":"[5]"},{"why":"Provides the experimental reference for transient shear banding during start-up that motivates the claim of robustness.","marker":"[23]"},{"why":"Defines the truncated and shifted Lennard-Jones interaction used to model the soft repulsive spheres.","marker":"[35]"},{"why":"Justifies using $Q \\approx 1$ as an effective overdamped limit and provides the pairwise dissipative-particle-dynamics drag form.","marker":"[48]"},{"why":"Supports the mapping of damping strength to an inertial quality factor and the behavior expected in underdamped conditions.","marker":"[61]"},{"why":"Is the companion study the paper uses to connect sample age and the persistence of flow inhomogeneities.","marker":"[64]"}],"fun_headline_variants":["Shear bands and overshoot are real, not artifacts","Transient banding robust to damping and boundaries","Jammed solids: intrinsic overshoot and banding","Simulations confirm shear banding is physical","Yielding shows banding across all drag models"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The 'large enough samples' caveat is assumed rather than demonstrated: all reported simulations use a single system size of about 97,556 particles, so the robustness claim presumes that this size is already in the asymptotic regime.","fun_headline_variants_meta":{"raw":{"variants":["Shear bands and overshoot are real, not artifacts","Transient banding robust to damping and boundaries","Jammed solids: intrinsic overshoot and banding","Simulations confirm shear banding is physical","Yielding shows banding across all drag models"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000233,"raw_usage":{"total_tokens":1449,"prompt_tokens":857,"completion_tokens":592,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":473,"completion_tokens_details":{"reasoning_tokens":519}},"tokens_in":473,"tokens_out":592,"duration_ms":6222,"temperature":1.0,"reasoning_tokens":519,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:56:44.401346+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same well-annealed sample preparation and low-rate shear with the shear-gradient dimension doubled, tripled, or increased by an order of magnitude while keeping the thickness-to-particle-size ratio and shear rate fixed. If the overshoot magnitude shrinks, the transient band becomes wall-localized, or the banding disappears at larger system sizes, then the claim that banding is intrinsic would be falsified. A complementary check would be to simulate at $Q$ well below 1, such as $Q = 0.01$, to directly probe the overdamped limit rather than relying on $Q \\approx 1$ as equivalent.","supporting_citations":[{"cited_title":"Varnik, L","cited_arxiv_id":null,"evidence_quote":"Supplies the frozen-wall shearing protocol used for the wall-based comparison."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the framework of transient shear banding and the back-flow signature that the paper tests."},{"cited_title":"Divoux, D","cited_arxiv_id":null,"evidence_quote":"Provides the experimental reference for transient shear banding during start-up that motivates the claim of robustness."},{"cited_title":"Nicolas, J.-L","cited_arxiv_id":null,"evidence_quote":"Justifies using $Q \\approx 1$ as an effective overdamped limit and provides the pairwise dissipative-particle-dynamics drag form."},{"cited_title":"Permanent shear localization in dense disordered materials due to microscopic inertia","cited_arxiv_id":"1812.03948","evidence_quote":"Supports the mapping of damping strength to an inertial quality factor and the behavior expected in underdamped conditions."},{"cited_title":"Emergence and persistence of flow inhomogeneities in the yielding and fluidization of dense soft solids","cited_arxiv_id":"1709.08717","evidence_quote":"Is the companion study the paper uses to connect sample age and the persistence of flow inhomogeneities."}],"review_version":1}