{"id":"6234bb89-344d-4f95-a9e8-17532ae45168","arxiv_id":"2502.05522","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"In 2D viscoelastic channel turbulence, the Reynolds stress is negative and dominated by first- and third-quadrant velocity fluctuations tied to inclined polymer sheet structures.","lead":"In a two-dimensional computer model of polymer flow, the turbulence state called elasto-inertial turbulence shows an unusual feature: the Reynolds stress is negative, so velocity fluctuations reduce rather than increase drag. The finding could change how drag-reducing polymer flows are modeled and gives cheap 2D simulations a measurable fingerprint.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Negative Reynolds stress claim lacks error bars and convergence checks; the sign of a small quantity could be a sampling or resolution artifact.","rationale":"The reader's weakest assumption correctly identifies the most fragile point: the absence of demonstrated statistical convergence, resolution adequacy, and domain-size sufficiency for the reported negative Reynolds stress. My stress-test pass did not find an internal inconsistency in the equations or a more fundamental logical flaw; the qualitative mechanism (Q1/Q3 dominance associated with inclined polymer sheets) is coherent with the reported quadrant statistics and flow visualizations. However, the central observational claim is quantitatively fragile because tau_R is small relative to the other stress contributions, and a small mean from a finite simulation in a periodic 2D box can be biased by large-scale intermittency. The additional high-Wi FENE-P saturation concern reinforces the need for resolution checks rather than replacing the reader's concern. Since the reader's verdict is already CONDITIONAL and explicitly conditioned on exactly this kind of quantified convergence evidence, my analysis does not change the verdict. I would, however, make acceptance explicitly contingent on the outcome of the proposed convergence and statistical-significance test: if the negative tau_R is not robust to that test, the paper's headline contribution should be withdrawn or substantially revised.","tokens_in":18286,"tokens_out":4159,"duration_ms":47314,"concrete_test":"Recompute the Wi=200 (and Wi=40) case under three variations: (i) doubled grid resolution 2048x608 with the same 20h domain; (ii) doubled streamwise domain Lx=40h at the original resolution; (iii) a run at least ten times longer at the original settings. For each run, compute tau_R(y) from block averages with bootstrap 95% confidence intervals and repeat the quadrant analysis. If the negative tau_R profile and Q1/Q3 dominance persist in all three variations with confidence intervals excluding zero, the anomalous-stress claim is supported; if the sign or magnitude changes materially, the claim is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim—that 2D EIT exhibits an anomalous negative Reynolds stress tau_R = -u'v' (Section 3.1, Figure 3c)—depends on the sign and magnitude of a deliberately small quantity. The text states that tau_R is 'significantly lower than the other stresses,' yet no uncertainty quantification is provided: no block-averaging over independent time windows, no confidence intervals, and no convergence tests with grid resolution, time step, or domain length. The grid (1024x304), time step, and domain (20h) are inherited from Zhang et al. (2024) without repeating grid-independence or domain-size checks for the present Wi range, and the statistical averaging time is not reported. In 2D turbulence, low-frequency large-scale intermittency can bias a finite-time mean of u'v', so a small negative mean could easily be a sampling artifact. In addition, at Wi=200 the mean polymer extension approaches 90% of L, meaning the flow is close to the FENE-P finite-extensibility limit; if the thin, highly extended polymer sheets are under-resolved, the stress balance and quadrant statistics could be distorted. If the negativity of tau_R does not survive these checks, the quadrant interpretation and the broader conclusion about the 'objective existence' of the 2D nature of EIT lose their factual foundation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports two-dimensional (2D) direct numerical simulations of FENE-P viscoelastic channel flow at Re=2000, beta=0.9, L=100, for Wi from 2 to 200, and characterizes the resulting elasto-inertial turbulence (EIT). The central claims are that, in contrast to 3D EIT and inertial turbulence, 2D EIT exhibits a negative Reynolds stress tau_R=-u'v'; that this negative stress arises from dominance of first- and third-quadrant velocity fluctuations associated with inclined polymer sheet-like structures; and that energy budgets, pressure redistribution, and spectral power laws of 2D EIT are sufficiently similar to those of 3D EIT to establish the 'objective existence of the 2D nature of EIT.' The paper also reports a mean velocity profile that shifts downward with Wi and appears to approach a logarithmic asymptote u+=6 ln y+ + 1.5, and spectra with exponents such as -11/3, -19/6, and -13/3.","tokens_in":18480,"tokens_out":3388,"duration_ms":38561,"significance":"If the negative Reynolds stress is robust, the finding is significant for the physics of EIT because it contradicts the usual positive sign of tau_R in inertial turbulence and in reported 3D EIT/DRT states, and it suggests a distinct momentum transport mechanism in 2D EIT. The paper also provides a useful systematic data set over a wide Wi range and applies standard diagnostic tools (stress balance, quadrant analysis, TKE/TEE budgets, pressure decomposition, spectra) that are appropriate for characterizing the flow. The in-house solver uses a tensor-based interpolation method that is intended to preserve the invariants and positive definiteness of the conformation tensor, which is a strength if the associated validation is supplied. However, the central observation rests on single-run statistics without uncertainty quantification or resolution/domain convergence checks, and several asymptotic and spectral claims are made by visual inspection rather than quantitative fitting.","major_comments":[{"comment":"The central claim that tau_R=-u'v' is negative in 2D EIT is not supported by any statistical uncertainty quantification. The paper states only that tau_R is 'significantly lower than the other stresses,' but the reported magnitude is small, and the sign of a small time-averaged quantity is precisely what can be biased by finite sampling, large-scale intermittency, or incomplete convergence. The averaging time is not reported, no block averaging or confidence intervals are given, and the grid (1024x304), time step, and domain length (20h) are adopted from prior work without repeating grid-independence, time-step, or domain-size checks for the present Wi range. This is particularly concerning at Wi=200, where the mean polymer extension approaches 90% of L, so under-resolution of highly extended polymer sheets could distort the stress balance and quadrant statistics. To make the anomalous Reynolds stress claim load-bearing, the authors need to report convergence diagnostics and error bars on tau_R and on the quadrant contributions, including tests at higher resolution, smaller time step, longer domain, and longer averaging windows.","section":"§2.3, §3.1, Fig. 3(c)"},{"comment":"The statement that the mean velocity profile 'ultimately converging to a distinct asymptotic regime (u+=6 ln y+ +1.5) when Wi=200' is based on a single Wi value and appears to be a visual fit to the profile. No fitting procedure, uncertainty, or additional higher-Wi case is shown, even though the text mentions Wi=1000 data that are not presented. Since the asymptotic regime is used to support the interpretation of 2D EIT as a distinct regime with no MDR-like convergence, the claim needs either a quantitative fit with confidence bounds or additional data at higher Wi demonstrating that the profile has stopped changing.","section":"§3.1, Fig. 1"},{"comment":"The spectral power-law exponents (-11/3, -19/6, -13/3) are key evidence for the proposed dynamic regime, but the manuscript does not explain how the exponents are estimated. The text refers to 'converging power-law decay' and 'the converging power-law decay of 2D EIT is k^{-11/3} within k>10,' yet no fits are drawn in Fig. 13, no fitting ranges are defined by a reproducible criterion, and no uncertainties are provided. Because the exponents are used both to compare 2D and 3D EIT and to connect EIT with elastic turbulence, the analysis should include a quantitative fitting procedure and error estimates, or the claims should be weakened to qualitative statements.","section":"§3.5, Fig. 13"}],"minor_comments":[{"comment":"The phrase 'remain scare' should be 'remain scarce.'","section":"Abstract"},{"comment":"The notation (·) for ensemble averaging is undefined; specify that averages are taken over time and the homogeneous streamwise direction (and possibly over realizations).","section":"Eq. (3.1)"},{"comment":"The adjective 'imcompressible' in the first sentence is a typo for 'incompressible.'","section":"§2.1"},{"comment":"The phrase 'ejection and sweep motions motions' contains a duplicated word.","section":"§3.2"},{"comment":"The reference list gives 'Warholic, M. D., Massah, H., & Hanratty, T. J. 2021' but the text cites Warholic et al. (1999) for the MDR-state Reynolds stress measurement; the year and entry should be corrected for consistency.","section":"References"},{"comment":"The labels in the probability-density contours and the superimposed vector fields are difficult to read at the printed size; please enlarge panels, define the color scale, and state explicitly which arrows correspond to Q1 and Q3 motions.","section":"Figs. 6 and 7"}],"recommendation":"major_revision","confidential_remarks":"The paper is squarely within the scope of JFM and addresses an interesting question, but the central claim of negative Reynolds stress needs stronger numerical evidence. In addition to uncertainty quantification, I would encourage the editor to ask the authors to state whether the simulations used for statistics are the same runs as in Zhang et al. (2024) or new runs, and to clarify the amount of time averaging in wall units. The interpretation of the quadrant analysis would also be more convincing if the authors quantified the convergence of Q1/Q3 dominance over independent time blocks."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe one thing you should know: the paper claims a negative Reynolds stress in 2D elasto-inertial turbulence, growing with Weissenberg number and attributed to Q1/Q3 quadrant dominance. That's a new and physically interesting claim, but the evidence base is thinner than the headline suggests. The rest of the paper—systematic statistics, energy budgets, pressure decomposition, spectra—is a useful contribution in its own right.\n\nWhat's new: 2D EIT was already known (Sid et al. 2018, Zhu et al. 2021), but the negative Reynolds stress and the Q1/Q3 interpretation are not in the prior literature. The paper also provides the first compiled 2D energy budgets and a Wi-sweep of the spectral behavior. The internal consistency is good: the stress balance in Eq. (3.1) is satisfied, and the plots support the qualitative story.\n\nWhat's soft: the central claim is the sign of a small quantity, and there are no error bars, no block-averaging statistics, no convergence checks, and no reported averaging time. In 2D turbulence, large-scale intermittency can bias a finite-time mean of u'v', so a negative value could be a sampling artifact. The grid and domain are inherited from the authors' prior paper, which is fine, but the present Wi range extends to 200 with polymer extension approaching 90% of L. A grid-refinement study at the highest Wi would address the worry about under-resolved polymer sheets. The spectral exponents and the u+ log-law fits are eye-fitted; those are secondary. The conclusion about 'objective existence of the 2D nature of EIT' is overstated—2D DNS is 2D by construction, and similarities to 3D budgets don't prove that 3D EIT is fundamentally 2D. That sentence should be softened.\n\nWho it's for: anyone working on viscoelastic turbulence, drag reduction, or elasto-inertial instability. The negative Reynolds stress, if it survives scrutiny, would be a distinctive fingerprint of 2D EIT and would sharpen the distinction from the 3D MDR state.\n\nRecommendation: send it to review. A serious referee should ask for uncertainty quantification and a grid check before the central claim is accepted, but the paper is coherent, new, and worth the refereeing effort. I'd cite it once the convergence evidence is in.","headline":"A plausible but under-verified negative Reynolds stress claim in 2D EIT, wrapped in a solid statistical characterization that deserves a referee's time.","tokens_in":19099,"tokens_out":4788,"would_cite":true,"duration_ms":63001,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"In two-dimensional elasto-inertial turbulence of a viscoelastic channel flow, the Reynolds stress is negative over much of the channel and grows in magnitude with the Weissenberg number, opposite to inertial turbulence.","keywords":["elasto-inertial turbulence","two-dimensional turbulence","viscoelastic channel flow","FENE-P model","Reynolds stress","quadrant analysis","polymer extension sheets","turbulent kinetic energy budget"],"falsifier":"Double the grid resolution and the domain length at $Wi=100$ and $200$ and recompute $\\tau_R(y)$ over a much longer averaging window; if the negative Reynolds stress disappears, changes sign, or shifts substantially, the claim fails. An independent check is whether the joint probability density of $(u',v')$ keeps its oblique Q1/Q3 elongation under these changes.","tokens_in":18034,"feed_emoji":"🌊","tokens_out":8763,"duration_ms":77041,"temperature":0.7,"pith_summary":"This paper tries to establish that two-dimensional elasto-inertial turbulence (EIT) in a viscoelastic channel flow has a genuinely anomalous momentum balance: the Reynolds stress $\\tau_R=-u'v'$ is negative across most of the channel and grows in magnitude as the Weissenberg number increases, exactly opposite to inertial turbulence and to reported three-dimensional EIT. The authors attribute the reversal to the dominance of first- and third-quadrant velocity fluctuations tied to inclined polymer sheet-like extension structures, and they show that elasticity strengthens EIT in 2D whereas it progressively suppresses inertial turbulence in 3D. If correct, the result would define a clean two-dimensional regime in which elasticity alone sustains turbulence, and it would support the claim that the familiar three-dimensional maximum-drag-reduction state is not pure EIT, since its Reynolds stress stays non-negative. This matters because it gives a quantitative statistical signature that distinguishes elasto-inertial from inertial turbulence and offers a benchmark for experimental and numerical identification of EIT.","feed_headline":"2D elasto-inertial turbulence flips Reynolds stress negative","feed_subtitle":"Polymer sheets reorder velocity fluctuations so that the Reynolds stress resists the flow instead of driving it.","key_machinery":"The central object is the quadrant-resolved Reynolds stress inside the stress balance $\\tau_R+\\tau_e+\\tau_v=\\tau_{\\mathrm{total}}$, resolved by splitting velocity fluctuations into four quadrants: Q1 (outward high-speed motion), Q2 (low-speed ejection), Q3 (inward low-speed motion), and Q4 (high-speed sweep). The key identity is that the sign of $\\tau_R$ is set by which quadrant pair dominates, and the paper shows that growing $Wi$ suppresses Q2/Q4 and enhances Q1/Q3, the latter aligned with the inclined polymer sheet-like extension structures; this inversion is what makes $\\tau_R$ negative. The dynamical machinery is completed by the TKE and elastic-energy budgets $P_k-\\varepsilon_k-G=0$ and $P_e-\\varepsilon_e+G=0$, where the transfer term $-G$ from polymer elastic energy to turbulent kinetic energy is the dominant production channel, and by a pressure decomposition into rapid, slow, polymer, and Stokes parts that shows the polymer pressure dominating wall-normal redistribution at high $Wi$. A tensor-based interpolation scheme that preserves the conformation tensor's positive definiteness allows the simulations to reach the high-$Wi$ regime without artificial diffusion.","core_discovery":"Using direct numerical simulations of two-dimensional plane Poiseuille flow of a FENE-P (finitely extensible nonlinear elastic-Peterlin) viscoelastic fluid at $Re=2000$ and $Wi$ from 10 to 200, the authors find that increasing elasticity intensifies EIT rather than suppressing inertial turbulence: mean velocity profiles deviate further from the laminar state and converge to an asymptotic $u^+=6\\ln y^+ + 1.5$ that is neither the Newtonian log law nor the Virk maximum-drag-reduction asymptote. The load-bearing statistical result is the stress balance $\\tau_R+\\tau_e+\\tau_v=\\tau_{\\mathrm{total}}$: the elastic stress $\\tau_e$ grows monotonically with $Wi$ and peaks between $y=0.2$ and $0.3$, while the Reynolds stress $\\tau_R$ is consistently negative, grows in magnitude with $Wi$, and saturates at high $Wi$. Quadrant analysis attributes the negative sign to the prevalence of Q1 and Q3 motions arranged along the polymer extension sheets inclined from the near-wall region toward the channel centre, with the usual Q2/Q4 ejection-sweep events of inertial turbulence suppressed. The energy budgets show that turbulent kinetic energy is produced almost entirely by the transfer from polymer elastic energy ($-G$), not by the Reynolds-stress production term $P_k$, which is negative, and the polymer contribution to pressure-strain redistribution dominates at high $Wi$. The authors take these 2D/3D budget similarities as evidence that 2D EIT is the objective, purely elastic-inertial state that coexists with residual dynamics in the 3D drag-reducing regime.","pith_inferences":["If the negative Reynolds stress survives independent grid refinement and longer averaging, a natural next test is its dependence on polymer extensibility $L$ and viscosity ratio $\\beta$; the paper itself leaves the parametric sensitivity of the 2D asymptotic state open.","The Q1/Q3 dominance suggests a measurable experimental signature in quasi-2D or thin-film viscoelastic flows: the joint probability density of $(u',v')$ should become obliquely elongated along the Q1/Q3 diagonal, unlike the Q2/Q4 elongation seen in inertial wall turbulence.","The spectral exponents between $-11/3$ and $-14/3$ could serve as a quantitative diagnostic for identifying pure EIT in experimental velocity or concentration spectra, independent of the sign of the Reynolds stress.","Generalizing the stress-balance sign test to three dimensions, one could search for local regions of negative Reynolds stress near polymer sheets in 3D simulations; their absence would confirm that residual inertial dynamics distinguish 3D MDR from pure EIT."],"forward_implications":["Reynolds stress contributes negatively to flow resistance in 2D EIT, so drag is carried almost entirely by viscous and elastic stresses and the standard inertial-turbulence momentum-transport picture does not apply there.","Velocity profiles in 2D EIT converge to an asymptotic form $u^+\\approx 6\\ln y^+ + 1.5$, distinct from both the Newtonian log law and the Virk MDR asymptote, marking 2D EIT as a separate regime.","Because $P_k$ is negative while $-G$ dominates TKE production, Reynolds-stress production cannot sustain 2D EIT; polymer elastic energy is the energy source, with polymer pressure dominating redistribution at high $Wi$.","The TKE spectrum follows $k^{-\\alpha}$ with $\\alpha>3$, converging at high $Wi$ to exponents such as $-11/3$, $-19/6$, and $-13/3$ at different wall distances, linking 2D EIT to elastic-turbulence spectral decay rather than inertial ranges.","The 2D/3D budget similarities imply that the 3D maximum-drag-reduction state is not pure EIT; the residual non-EIT dynamics in 3D are what keep its Reynolds stress non-negative."],"supporting_citations":[{"why":"Supplies the demonstration that sustained EIT exists in 2D DNS and the proposal that EIT has an inherently 2D nature.","marker":"Sid et al. (2018)"},{"why":"Introduces EIT and provides the experimental and 3D baseline where Reynolds stress is positive but minimal in the MDR state.","marker":"Samanta et al. (2013)"},{"why":"Establishes the mechanism of EIT, including polymer-stress energy transfer to TKE, the sheet-like structures, and the $k^{-11/3}$ spectral decay used for comparison.","marker":"Dubief et al. (2013)"},{"why":"Documents the non-convergent friction-factor behaviour of 2D EIT versus the 3D MDR asymptote, the key dimensional contrast extended here.","marker":"Zhu et al. (2021)"},{"why":"Provides the 2D domain-length criterion, grid, time step, and minimal-flow-unit analysis adopted for the present simulations.","marker":"Zhang et al. (2024)"},{"why":"Supplies the quadrant-analysis method used to decompose Reynolds stress into Q1-Q4 motions.","marker":"Lu et al. (1973)"},{"why":"Reports experimentally measured near-zero but non-negative Reynolds stress in drag-reducing flow, the baseline the negative 2D value is contrasted with.","marker":"Warholic et al. (1999)"},{"why":"Gives the rapid/slow/polymer pressure decomposition and pressure-strain analysis used in the redistribution section.","marker":"Terrapon et al. (2015)"},{"why":"Provides 3D simulation comparisons of stress profiles and the marginal-IT versus EIT distinction at MDR.","marker":"Wang et al. (2023)"}],"fun_headline_variants":["2D EIT: polymer sheets flip Reynolds stress negative","Elasticity strengthens 2D EIT, inverts Reynolds stress","Polymer extension drives negative Reynolds stress in 2D","2D viscoelastic turbulence: Reynolds stress resists flow","In 2D EIT, budgets match 3D but Reynolds stress inverts"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The central claim depends on the 2D simulations at $Wi$ up to 200 being fully developed, spatially resolved, and statistically converged; the paper uses one domain length and grid from prior work and gives no convergence or sampling diagnostics.","fun_headline_variants_meta":{"raw":{"variants":["2D EIT: polymer sheets flip Reynolds stress negative","Elasticity strengthens 2D EIT, inverts Reynolds stress","Polymer extension drives negative Reynolds stress in 2D","2D viscoelastic turbulence: Reynolds stress resists flow","In 2D EIT, budgets match 3D but Reynolds stress inverts"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000384,"raw_usage":{"total_tokens":2145,"prompt_tokens":1169,"completion_tokens":976,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":785,"completion_tokens_details":{"reasoning_tokens":885}},"tokens_in":785,"tokens_out":976,"duration_ms":10095,"temperature":1.0,"reasoning_tokens":885,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-08T18:59:28.708000+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Double the grid resolution and the domain length at $Wi=100$ and $200$ and recompute $\\tau_R(y)$ over a much longer averaging window; if the negative Reynolds stress disappears, changes sign, or shifts substantially, the claim fails. An independent check is whether the joint probability density of $(u',v')$ keeps its oblique Q1/Q3 elongation under these changes.","supporting_citations":[{"cited_title":"E., & Dubief, Y","cited_arxiv_id":null,"evidence_quote":"Supplies the demonstration that sustained EIT exists in 2D DNS and the proposal that EIT has an inherently 2D nature."},{"cited_title":"N., Wagner, C., & Hof, B","cited_arxiv_id":null,"evidence_quote":"Introduces EIT and provides the experimental and 3D baseline where Reynolds stress is positive but minimal in the MDR state."},{"cited_title":"E., & Julio, S","cited_arxiv_id":null,"evidence_quote":"Establishes the mechanism of EIT, including polymer-stress energy transfer to TKE, the sheet-like structures, and the $k^{-11/3}$ spectral decay used for comparison."},{"cited_title":"2021 Nonasymptotic elastoinertial turbulence for asymptotic drag reduction","cited_arxiv_id":null,"evidence_quote":"Documents the non-convergent friction-factor behaviour of 2D EIT versus the 3D MDR asymptote, the key dimensional contrast extended here."},{"cited_title":"N., CHENG, H","cited_arxiv_id":null,"evidence_quote":"Provides the 2D domain-length criterion, grid, time step, and minimal-flow-unit analysis adopted for the present simulations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides 3D simulation comparisons of stress profiles and the marginal-IT versus EIT distinction at MDR."}],"review_version":1}