{"id":"ce9fa41e-252b-487b-884a-59cd93e1436c","arxiv_id":"2505.03987","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"Nonlinear MHD simulations show that Mercier-unstable Wendelstein configurations generate a tokamak-like dynamo early, but partial reconnection and multi-mode competition prevent sustained flux pumping, so current drive and marginal stability may be needed.","lead":"Simulations of two Mercier-unstable stellarator configurations show that low-n pressure-driven modes generate a dynamo effect similar to tokamak flux pumping, but in both cases the modes do not settle into a benign saturated state and instead cause crashes or mode competition. The results suggest that reproducing saturated low-n modes in stellarators may require current drive or control and configurations closer to marginal stability.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"W7-AS discrepancy may be an artifact of the assumed linear pressure profile; more peaked profiles could permit the sustained (1,2) mode the paper claims is absent.","rationale":"The reader's weakest assumption identifies the same load-bearing concern: the W7-AS negative result is built on a linear pressure profile and a hand-modified Er profile, and the paper itself acknowledges that the linear profile is likely incorrect and that peaked profiles could make the (1,2) mode easier to observe. This concern does not overturn the paper's carefully scoped claims about 'current simulations', and the W7-X-like analysis is explicitly presented as a strongly unstable, non-experimental test case; the authors also propose near-marginal configurations and current sources as future work, which is consistent with the stated conclusions. The paper also has genuine supporting evidence: linear cross-validation between JOREK and CASTOR3D, clear documentation of the artificial low-n initialisation in Section 2, and explicit limitation statements throughout. Therefore I would keep the reader's CONDITIONAL verdict rather than moving to rejection: the central simulation results are plausible, but the W7-AS extrapolation needs a profile-sensitivity test before the experimental discrepancy can be claimed as a robust physics finding.","tokens_in":18651,"tokens_out":5074,"duration_ms":51646,"concrete_test":"Repeat the W7-AS JOREK case with a peaked pressure profile p(s)=p0(1-s^alpha) for alpha in {1.5, 2, 3}, at the same total beta, same 1/2 resonance, same Spitzer resistivity, and the same modified Er profile and resolutions as Table 1, with ntor >= 30. The decisive outcome is whether a coherent (1,2) mode persists for several milliseconds with total stored energy maintained below 3.2 MW of heating. If it does, the claimed failure to reproduce W7-AS is a reconstruction artifact rather than a generic property of Mercier-unstable stellarators; if it still crashes or bursts, the negative result survives this test.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The W7-AS leg of the central negative claim depends on an equilibrium whose core pressure is prescribed as p(s)=p0(1-s) (Section 3.1 and Appendix C). The authors themselves state this assumption is 'likely incorrect given observations of peaked profiles in experimental diagnostics of similar discharges' (Section 4), and their own profile-maintained run (bottom row of Figure 8) shows weaker low-n MHD signatures when the linear profile is artificially held, whereas the discussion notes that more peaked profiles make low-n modes more prominent. If the experimental W7-AS pressure profile is instead peaked, the (1,2) mode may saturate into the coherent, sustained state seen in the experiment, which would remove the main experimental counterexample and weaken the conclusion that a current source and a single marginal instability are required to avoid crash dynamics. The hand-modified Er profile (enforced subsonic and zero at the boundary) adds a second uncontrolled input; either profile error could change which mode dominates. Because the paper's claims are explicitly scoped to 'current simulations', this is not an internal inconsistency, but the W7-AS comparison is the load-bearing experimental validation and it rests on an acknowledged reconstruction assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies nonlinear evolution of low-n Mercier-unstable stellarator configurations using the JOREK code, with the aim of determining whether benign nonlinear saturation and flux-pumping-like states can arise in stellarators as they do for quasi-interchange modes in tokamaks. Two cases are considered: a low-iota, high-mirror W7-X-like configuration, where a (2,3) interchange mode is found to produce a mean dynamo term but then undergo partial reconnection and a core temperature crash; and a W7-AS intermediate-beta equilibrium reconstructed from experiment, where low-n MHD activity is reproduced only transiently and a sustained (1,2) mode as seen in the experiment is not obtained. The authors argue that the missing ingredients for flux pumping in stellarators are a current source and a single marginal instability, and they discuss several modelling uncertainties, most prominently the assumed linear pressure profile in the W7-AS equilibrium.","tokens_in":18968,"tokens_out":2932,"duration_ms":30839,"significance":"If the results are robust, the paper makes a valuable contribution by showing that Mercier-unstable stellarators do not automatically settle into benign flux-pumping-like states, and by identifying concrete missing ingredients (current source, marginality) for such states. The study is among the first to compare stellarator interchange dynamics with tokamak flux pumping using a common nonlinear MHD framework with experimentally motivated parameters. Strengths include the explicit comparison of the linear (2,3) mode structure between JOREK and CASTOR3D, the use of experimental Spitzer resistivity and heating powers for W7-AS, and the candid, well-documented discussion of modelling limitations. The central negative result for W7-AS, however, rests on an equilibrium reconstruction with an acknowledged uncertain pressure profile and a hand-modified radial electric field, so the significance hinges on the sensitivity of the conclusions to these inputs.","major_comments":[{"comment":"The W7-AS equilibrium is constructed with the assumed pressure profile p(s)=p0(1-s), which the authors themselves state in Section 4 is 'likely incorrect given observations of peaked profiles in experimental diagnostics of similar discharges'. Since the central claim that no sustained (1,2) mode appears in simulations is based on this equilibrium, and since Section 3.2 reports that more peaked pressure profiles make low-n MHD signatures more prominent, the possibility that the experimental profile is peaked directly undermines the paper's main experimental counterexample. The authors should quantify this risk, for example by performing linear stability scans with CASTOR3D (as they themselves suggest is appropriate) or by running a nonlinear case with a more peaked initial profile, before concluding that the (1,2) mode cannot saturate.","section":"Section 3.1 and Appendix C"},{"comment":"The radial electric field used in the W7-AS simulations is modified by hand from the neoclassical calculation to enforce subsonic poloidal flow and zero boundary electric field, an inconsistency the authors acknowledge. This input directly controls which mode families are stabilised: the paper reports that without the modified E×B flow the high-n branch dominates, and with it the (5,10) mode dominates. Because the result is so sensitive to this profile, the paper should demonstrate how the nonlinear outcome changes under plausible variations of the Er profile (e.g., amplitude, radial transition location, boundary value). Without such a sensitivity scan, the claim that E×B flow shear is a necessary ingredient for matching the experiment remains unsupported.","section":"Section 3.1, Figure 7"},{"comment":"The W7-X-like case is initialised by freezing the Nf=0 and Nf=1 mode families for t<0.85 ms so that the low-n (2,3) mode can lead the dynamics. This is an artificial intervention that prevents the high-n branch from naturally competing during the linear phase, and the later observation of partial reconnection and crash may depend on this scheduling. The paper states that this is done to investigate the case where low-n modes lead, but it should be made explicit that the conclusion 'the MHD dynamo cannot be followed over a resistive timescale' (Section 4) applies only to this specific initialization protocol. A test relaxing the freeze time, or a demonstration that the crash is robust to the freeze interval, is needed to support the more general statement.","section":"Section 2.1 and 2.2"},{"comment":"The nonlinear dynamics in the W7-AS case are highly sensitive to the mode-family initialization schedule: with Nf=0 evolved longer the (5,10) mode dominates and causes a crash, while allowing Nf=1 to grow early leads to a dominant (1,2) structure at t=3 ms (bottom-right panel of Figure 9). Given this, the statement in Section 4 that 'a coherent sustained (1,2) mode like the one observed in the experiment has not been found in any of the simulations' is not fully transparent. The authors should clarify whether the t=3 ms (1,2) structure persists or subsequently breaks up, and should provide the time trace beyond 3 ms. If the mode is only transient, that should be stated explicitly with evidence; if it persists, the central claim needs revision.","section":"Section 3.2, Figure 9"}],"minor_comments":[{"comment":"The declaration reads 'The authors report not conflict of interest' and should be 'no conflict of interest'.","section":"Acknowledgements"},{"comment":"The notation s = sqrt(psi_t) is introduced in the text but the definition of psi_t itself is not given; a one-line definition of the normalised toroidal flux would help readers.","section":"Section 1"},{"comment":"The parallel momentum equation contains a time-derivative term involving partial B^2/partial t that is not discussed; a brief comment on why this term is retained in the reduced MHD ordering would improve readability.","section":"Appendix B, Equation B5"},{"comment":"The colour scale for the iota value in the Poincare plots is not defined in the caption; adding a colourbar or describing the range of colours would make the claimed local iota modification visible to the reader.","section":"Figure 3"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid simulation study with an honest appraisal of its limitations, but the main experimental validation leg (W7-AS) depends on an equilibrium pressure profile that the authors themselves consider likely incorrect and on a hand-modified Er profile. The sensitivity of the conclusions to these inputs should be addressed before publication. The W7-X-like case, while clearly labelled as not directly experimental, is the more novel part and is reasonably well supported; however, the claims about the dynamo and crash are conditional on the specific initialization schedule. I would not reject the paper, but the central claims need either stronger sensitivity studies or a more careful scope limitation."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth a read if you work on stellarator MHD stability or flux pumping. The paper does two things. First, it shows a clear (2,3) interchange-generated dynamo with the same (0,0) electric-field signature as tokamak flux pumping, but then shows the dynamo is not sustained because overlapping modes cause partial reconnection and a core temperature crash. Second, it revisits the W7-AS saturated (1,2) mode with new simulations that include E×B flow. The flow suppresses the high-n branch, but a sustained (1,2) mode still does not appear; instead the (5,10) mode dominates. That is an honest negative result, and the authors are careful to say it is a statement about current simulations, not a general no-go.\n\nCredit where due: the linear mode structure is cross-checked between JOREK and CASTOR3D; the W7-AS runs use Spitzer resistivity, neoclassical Er (with a disclosed modification), and experimental heating. The discussion of what is missing — a current source and a single marginal instability — is clearly labeled as hypothesis. The paper also does a service by making the flux-pumping theory explicit for a zero-net-current stellarator in appendix A.\n\nThe soft spots are mostly acknowledged in the text. The W7-AS equilibrium uses p(s)=p0(1-s), which the authors themselves call likely incorrect; they even note that a more peaked profile makes low-n modes more prominent. So the stress-test concern about the W7-AS leg is real, but it is the authors' concern too, and they do not overclaim. The bigger worry is that the W7-X-like result depends on an initialization schedule that freezes Nf=0/1 modes for 0.85 ms so the low-n mode leads. That is necessary given the numerical difficulty, but it means the crash is a property of that particular nonlinear path. The high viscosity and the absence of code or input files also limit reproducibility. None of this breaks the central claim; it just means the extrapolation to real devices is thinner than the headline suggests.\n\nWho should read it: people doing nonlinear MHD in stellarators, especially those hoping to operate beyond Mercier. It is not a definitive resolution, but it is a genuinely useful data point and a fair statement of where the modeling stands. I would send it to review — it deserves a serious referee — and I'd expect the discussion of current sources to be useful in future design work. If I were reviewing, I'd ask for the W7-AS pressure-profile sensitivity to be quantified, even in a limited way, before publication.","headline":"A careful, honest JOREK study of Mercier-unstable stellarators that reproduces the tokamak-like dynamo but finds crash dynamics rather than sustained flux pumping; the W7-AS comparison is credible if you accept the acknowledged profile assumptions.","tokens_in":19534,"tokens_out":2015,"would_cite":true,"duration_ms":20134,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["52.30.-q","52.35.Py","52.55.Hc"],"model":"deepseek-v4-flash","headline":"Mercier-unstable stellarators do not automatically settle into the benign flux-pumping state seen in tokamaks.","keywords":["Mercier criterion","stellarator","flux pumping","interchange instability","quasi-interchange mode","partial reconnection","nonlinear MHD","W7-AS"],"falsifier":"A concrete test would be a nonlinear simulation of the W7-X-like configuration with parameters moved close to marginal stability and with a small current-drive source added: if the dynamo then persists over a resistive timescale and the core temperature does not crash, the paper's proposed route to stellarator flux pumping is supported, whereas a persisting crash would falsify the claim that missing current sources and marginality explain the absence of a benign state.","tokens_in":18459,"feed_emoji":"🌀","tokens_out":13205,"duration_ms":118806,"temperature":0.7,"pith_summary":"Mercier's criterion is usually treated as a hard stability limit in stellarator design, but experiments have sometimes operated past it. This paper tests whether the benign nonlinear state that tokamaks reach through flux pumping—where an MHD dynamo redistributes current and keeps the plasma stable—also appears in Mercier-unstable Wendelstein stellarators. In a low-$\\iota$, high-mirror W7-X-like configuration, the unstable $(2,3)$ interchange initially produces a dynamo resembling tokamak flux pumping, but the mode does not saturate: overlapping instabilities cause partial reconnection and a core temperature crash within about a millisecond. In reconstructions of W7-AS discharges where a saturated $(1,2)$ mode was observed experimentally, the same simulation approach fails to produce a sustained coherent $(1,2)$ mode even with background flow shear, although the $(1,2)$ mode appears transiently and the plasma can be held at experimental heating power. The authors conclude that a benign flux-pumping state in a stellarator is not automatic; it may require a current source to counterbalance the dynamo and a mode spectrum with a single marginal instability.","feed_headline":"Stellarator cores crash where tokamaks find flux pumping","feed_subtitle":"The dynamo fades into reconnection and no sustained (1,2) mode appears; current drive and marginality may be missing.","key_machinery":"The central object is the MHD dynamo: the $(0,0)$ component of the $\\boldsymbol{v}\\times\\mathbf{B}$ electric field induced by an unstable interchange mode, which in tokamak flux-pumping scenarios redistributes the current profile and keeps the safety factor near a rational value. The simulations are reduced viscoresistive MHD runs that include helical mode coupling and fluid compressibility, which the paper shows are necessary to capture the low-$n$ interchange, and they add a background $\\mathbf{E}\\times\\mathbf{B}$ flow to test shear stabilisation of high-$n$ modes. A second load-bearing concept is partial reconnection: when several interchange instabilities overlap, current sheets break the nested flux surfaces locally, altering the local $\\iota$ and ejecting core heat, which cuts off the dynamo before it can self-organise.","core_discovery":"On the paper's own terms, the central claim is that Mercier-unstable Wendelstein configurations do not automatically enter the quasi-stationary, dynamo-sustained state seen in flux-pumping tokamaks. Mercier's criterion—the local stability condition for pressure-driven interchange modes in a toroidal plasma—is violated across most of the W7-X-like volume, and the linearly dominant $(2,3)$ interchange creates an $m=0$, $n=0$ dynamo electric field with the same character as the $(1,1)$ quasi-interchange in tokamaks. But the dynamo is not sustained on the resistive timescale: after about $1.15$ ms, current-sheet formation and reconnection partially reconnect the core, the local $\\iota$ is modified toward the $2/3$ rational value, and the core temperature crashes despite heating comparable to experimental power. In the W7-AS case, at the experimental Spitzer resistivity and with an equilibrium $\\mathbf{E}\\times\\mathbf{B}$ flow added to suppress high-$n$ modes, the simulations show multiple competing modes and a dominant $(5,10)$ structure rather than the sustained $(1,2)$ mode seen experimentally; a coherent $(1,2)$ mode appears only transiently, depending on initialization, and the stored thermal energy can be maintained when flows are included. The paper argues that the missing ingredients are a current source or current-control mechanism, which in tokamaks counterbalances the dynamo, and operation near marginal stability so that only a single coupled instability is present.","pith_inferences":["Editorial inference: if a current source is the missing ingredient, then quasi-axisymmetric stellarators, with their substantial bootstrap current, are more natural candidates for a stellarator flux-pumping regime than quasi-isodynamic ones; the paper notes this connection but does not test it.","Editorial inference: scanning the W7-AS reconstruction toward more peaked pressure profiles or adding a co-injected current source is the most direct test of whether a sustained $(1,2)$ mode emerges; the paper itself notes that peaked profiles strengthen the low-$n$ signature.","Editorial inference: the paper's dichotomy implies that a Mercier-unstable stellarator with many overlapping unstable helicities will tend to crash, while a configuration with a single marginally unstable coupled mode should saturate; this could be checked by a linear scan over rotational-transform and pressure profiles.","Editorial inference: if no current source is available, the tokamak flux-pumping picture transfers to stellarators only in modified form, with Pfirsch-Schlüter or bootstrap currents playing the counterbalancing role; whether those currents suffice remains open."],"forward_implications":["Removing the Mercier constraint in stellarator design does not by itself guarantee a benign saturated state; in the W7-X-like simulation the core temperature drops abruptly on a roughly millisecond timescale.","A stellarator flux-pumping state is likely to require an external or internal current source to counterbalance the dynamo, since the net toroidal current in a quasi-isodynamic device is negligibly small.","Equilibrium $\\mathbf{E}\\times\\mathbf{B}$ flow shear suppresses high-$n$ modes and can allow the plasma to be maintained at experimentally relevant heating power, but it does not automatically recover the experimental $(1,2)$ mode structure.","The sustained $(1,2)$ mode observed in W7-AS was not reproduced; the mode is transient and the outcome depends on initialization and on whether the background profiles are maintained.","Partial reconnection from overlapping interchange instabilities is a plausible core-crash mechanism for Mercier-unstable stellarators, analogous to the multiple-interchange sawtooth model in tokamaks."],"supporting_citations":[{"why":"Defines the stability criterion whose hard enforcement the paper questions and whose violation defines the unstable equilibria studied.","marker":"Mercier 1964"},{"why":"Provides the original experimental evidence that magnetic-flux pumping sustains stationary high-performance tokamak plasmas.","marker":"Petty et al. 2009"},{"why":"Supplies the flux-pumping theory and the MHD dynamo mechanism in nonlinear simulations against which the stellarator results are compared.","marker":"Krebs et al. 2017"},{"why":"Gives the partial-reconnection, multiple-interchange crash model used to interpret the stellarator core temperature crash.","marker":"Jardin et al. 2020"},{"why":"Source of the W7-AS experimental observation of a saturated $(1,2)$ mode that the simulations are designed to reproduce.","marker":"Weller et al. 2002"},{"why":"Baseline W7-AS nonlinear MHD setup and previous partially successful validation study that this work extends.","marker":"Ramasamy et al. 2024"},{"why":"Quantitative simulation of flux pumping in ASDEX Upgrade that anchors the tokamak comparison and the single-marginal-instability interpretation.","marker":"Zhang et al. 2025"},{"why":"Documents the NBI-driven currents and current-control requirements in W7-AS that motivate the missing-current-source hypothesis.","marker":"Geiger et al. 2003"}],"fun_headline_variants":["Mercier-unstable stellarators fail to sustain dynamo; tokamaks pump","Stellarator reconnection wins over tokamak-style flux pumping","No sustained dynamo in Mercier-unstable stellarators, unlike tokamaks","Tokamak flux pumping not reproduced in Mercier-unstable stellarators"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper's central negative result for W7-AS rests on an equilibrium reconstruction that assumes a linear pressure profile $p(s)=p_0(1-s)$ and a hand-adjusted radial electric field; if the true profiles are more peaked, the $(1,2)$ mode might saturate as in the experiment and the proposed explanation for the missing flux-pumping state would not apply.","fun_headline_variants_meta":{"raw":{"variants":["Mercier-unstable stellarators fail to sustain dynamo; tokamaks pump","Stellarator reconnection wins over tokamak-style flux pumping","No sustained dynamo in Mercier-unstable stellarators, unlike tokamaks","Tokamak flux pumping not reproduced in Mercier-unstable stellarators"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000802,"raw_usage":{"total_tokens":3573,"prompt_tokens":1038,"completion_tokens":2535,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":654,"completion_tokens_details":{"reasoning_tokens":2453}},"tokens_in":654,"tokens_out":2535,"duration_ms":20832,"temperature":1.0,"reasoning_tokens":2453,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T23:40:33.736346+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete test would be a nonlinear simulation of the W7-X-like configuration with parameters moved close to marginal stability and with a small current-drive source added: if the dynamo then persists over a resistive timescale and the core temperature does not crash, the paper's proposed route to stellarator flux pumping is supported, whereas a persisting crash would falsify the claim that missing current sources and marginality explain the absence of a benign state.","supporting_citations":[{"cited_title":"Nuclear Fusion 4 (3), 213","cited_arxiv_id":null,"evidence_quote":"Defines the stability criterion whose hard enforcement the paper questions and whose violation defines the unstable equilibria studied."},{"cited_title":"Physical review letters 102 (4), 045005","cited_arxiv_id":null,"evidence_quote":"Provides the original experimental evidence that magnetic-flux pumping sustains stationary high-performance tokamak plasmas."},{"cited_title":"Physics of Plasmas 24 (10)","cited_arxiv_id":null,"evidence_quote":"Supplies the flux-pumping theory and the MHD dynamo mechanism in nonlinear simulations against which the stellarator results are compared."},{"cited_title":"Physics of Plasmas 27 (3)","cited_arxiv_id":null,"evidence_quote":"Gives the partial-reconnection, multiple-interchange crash model used to interpret the stellarator core temperature crash."},{"cited_title":"In 19^ th IAEA fusion energy conference, Lyon (France), 14-19 Oct 2002\\/","cited_arxiv_id":null,"evidence_quote":"Source of the W7-AS experimental observation of a saturated $(1,2)$ mode that the simulations are designed to reproduce."},{"cited_title":", Aleynikova, K","cited_arxiv_id":null,"evidence_quote":"Baseline W7-AS nonlinear MHD setup and previous partially successful validation study that this work extends."},{"cited_title":"Nuclear Fusion","cited_arxiv_id":null,"evidence_quote":"Quantitative simulation of flux pumping in ASDEX Upgrade that anchors the tokamak comparison and the single-marginal-instability interpretation."}],"review_version":1}