{"id":"2092ead6-2469-4eaf-a1ae-707dd0eba3ab","arxiv_id":"2507.04129","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A coupled OpenMC/NekRS model of the Molten Salt Fast Reactor adds transport of delayed neutron and decay heat precursors in the flowing fuel salt, giving preliminary temperature and concentration distributions.","lead":"This paper builds a computer model of a liquid-fuel nuclear reactor that couples neutron physics with flowing molten salt and tracks two types of radioactive precursor particles. It shows expected temperature and precursor patterns, but the model is not yet complete because the precursors do not feed back into the neutron calculation.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 2 s external-loop residence time used for recirculating DNPs/DHPs (Sec. 5.2) is inconsistent with the paper's stated 18 m³ circuit volume and 1882 kg/s flow, which imply ~20 s, so re-entering precursor concentrations are overestimated.","rationale":"The reader's CONDITIONAL verdict already identified the 2 s residence time as a weakly supported assumption. I agree that this is the most load-bearing concern, and I sharpen it with a concrete internal inconsistency: the paper's own stated total volume, core volume, mass flow rate, and density imply an external-loop residence time of about 19.7 s, not 2 s. This means the recirculating DNP/DHP inlet concentrations are overestimated, with consequences that propagate directly into the results the paper presents as its main contributions. The single-physics verifications in Section 6 are useful but do not exercise the recirculation boundary condition, so they do not mitigate this concern. I do not see this as grounds for outright rejection: the paper is explicitly preliminary, the model is transparently described, and the assumption is easily corrected or bounded by sensitivity analysis. The reader's conditional verdict therefore stands, though the condition is now concrete and numerical rather than a general request for sensitivity studies. If the proposed rerun with a ~20 s residence time changes the DNP/DHP results only slightly, the concern would be retired; if it changes them substantially, the 'reasonable behavior' claim must be revised. I also considered whether the OpenMC limitation (no delayed-neutron-source relocation) is more load-bearing, but the paper explicitly discloses this limitation and the central claim is about transport modeling, not feedback; the recirculation assumption is presented as a completed model and is internally inconsistent, making it the sharper point. No ad hominem is intended; the issue is purely with the argument and its numerical consistency.","tokens_in":7614,"tokens_out":5863,"duration_ms":61582,"concrete_test":"Compute the physical external-loop residence time from the paper's own parameters: V_total = 18 m³, V_core ≈ 8.95 m³ (cylinder 2.25 m high, 2.25 m diameter), Q = m_dot/ρ = 1882.12/4125.3 ≈ 0.456 m³/s, so τ_ext ≈ (18−8.95)/0.456 ≈ 19.7 s. Rerun the coupled Cardinal simulation with inlet concentrations C_in = C_out exp(−λ τ_ext) for all 6 DNP and 3 DHP groups, instead of the 2 s value, and also test a delayed recirculation model that preserves the wall-clock transit time. Compare the resulting re-entering precursor concentrations and the time-averaged DNP/DHP spatial distributions. If group-1 and group-2 DNP inlet concentrations change by more than a few percent, or if the group-1 DNP distribution visibly shifts, the presented 'reasonable behavior' results need revision; report this as a sensitivity table.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The weakest load-bearing point is the external-loop boundary condition for recirculating DNPs and DHPs. Section 5.2 states that when precursors reach the outlet, an average concentration is calculated and imposed at the inlet 'by a decay constant', assuming a residency time of 2 seconds in the pumps and heat exchangers. The paper's own parameters make this assumption internally inconsistent. Section 2 gives a total primary fuel volume of 18 m³, with half of the salt inside the core at any time; Table I gives a mass flow rate of 1882.12 kg/s and a fluid density of 4125.3 kg/m³. These imply a volumetric flow Q = 1882.12/4125.3 ≈ 0.456 m³/s and an external-loop volume of about 9 m³, so the mean external-loop residence time is approximately 9/0.456 ≈ 19.7 s, not 2 s. The 2 s value corresponds to only ~0.9 m³ of external volume. Consequently, the inlet precursor concentrations imposed in the model are too high by a factor of exp(−λ(19.7−2)) for each group. For DNP group 1 (λ = 0.012575 s⁻¹), this is about a 20% overestimate of the re-entering concentration; for shorter-lived groups the discrepancy is much larger. Since the central results are the DNP and DHP distributions, this boundary treatment directly shapes the fields claimed to show 'reasonable behavior'. No sensitivity analysis or physical justification is given, and the paper itself defers one to future work in Section 8. The separate limitation that OpenMC cannot move delayed neutron sources is explicitly acknowledged; the 2 s recirculation assumption, by contrast, is presented as a completed model component but is contradicted by the stated circuit volume and flow. This makes the quantitative precursor distributions unsupported, not merely unvalidated.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a multiphysics model of the Molten Salt Fast Reactor (MSFR) built in the Cardinal framework, coupling OpenMC for neutronics with NekRS for thermal-hydraulics and for the transport of delayed neutron precursors (DNPs) and decay heat precursors (DHPs). The DNP and DHP fields are solved as passive scalars with decay and source terms, and the external loop is represented by a simplified outlet-to-inlet boundary condition that imposes a decayed average concentration with an assumed 2 s residence time. The authors report single-physics verification of OpenMC against Serpent and of a Nek5000 k-tau model against OpenFOAM, followed by preliminary coupled results for temperature, heat source, velocity, and precursor distributions. The paper explicitly acknowledges several limitations, including the absence of DNP feedback in OpenMC, constant fluid properties, and a simplified external loop.","tokens_in":8004,"tokens_out":7615,"duration_ms":81572,"significance":"If the limitations are properly resolved, the work is a useful step toward high-fidelity MSFR multiphysics simulation, because it demonstrates a practical coupling of Monte Carlo neutronics with a spectral-element CFD code that includes precursor drift. The use of established open-source codes (OpenMC, NekRS, Cardinal) is a strength, and the single-physics verification exercises provide some confidence in the individual components. The paper is honest about its preliminary nature and about the missing delayed-neutron feedback, which is the main physics that would make the precursor transport relevant to the neutronics. However, the two load-bearing inconsistencies identified below (power normalization and the external-loop residence-time assumption) mean that the quantitative results, and the claim that the precursor distributions are 'reasonable', are not yet fully supported.","major_comments":[{"comment":"The stated reactor power is inconsistent with the power used in the coupled model. Section 2 states that the MSFR is a 3000 MWth reactor, but Table I lists P = 3.0e8 W = 300 MW. The mass flow rate (1882.12 kg/s), heat capacity (1593.9 J/kg.K), and inlet–outlet temperature difference (100 K) in Table I are mutually consistent with 300 MW (m_dot*Cp*Delta_T ≈ 3.0e8 W), so the model is evidently normalized to 300 MW rather than 3000 MWth. This factor-of-ten discrepancy affects the heat source magnitude and all thermal and precursor fields, so it must be corrected before the results can be compared with the MSFR reference design.","section":"Section 2 and Table I"},{"comment":"The external-loop residence time of 2 s is not justified and appears inconsistent with the circuit volume and flow rate. Section 2 states that the primary circuit contains 18 m^3 of salt with half inside the core at any time, and Table I gives a mass flow rate of 1882.12 kg/s and a density of 4125.3 kg/m^3, implying a volumetric flow of about 0.456 m^3/s and an external volume of roughly 9 m^3. The resulting mean external-loop transit time is about 20 s, far larger than 2 s. Even if the model already resolves part of the external loop, the 2 s value is not derived from the geometry, and no sensitivity analysis is provided. Because the inlet precursor concentration directly shapes the DNP and DHP distributions that are the central results, the authors should either correct the residence time to match the stated circuit volumes, provide a physical basis for the 2 s value, or add a sensitivity study over this parameter.","section":"Section 5.2"},{"comment":"The calculated multiplication factor of 1.0683 is significantly above critical, and the verification models give keff around 1.043. The paper does not discuss this excess reactivity or how a non-critical model represents the MSFR. While a k-eigenvalue solution still provides a fission source shape, the meaning of the coupled results for a system that is critical by design is unclear. The authors should either adjust the model to be closer to criticality or explicitly discuss why a keff of 1.068 is acceptable for their demonstration.","section":"Sections 5.1 and 7"}],"minor_comments":[{"comment":"The source term in the DHP transport equation uses the subscript i (beta_{h,i}) while the equation is indexed by l; it should be beta_{h,l} for consistency with the rest of the notation.","section":"Equation (8)"},{"comment":"The table header reads 'Constans' and should be 'Constants'.","section":"Table III"},{"comment":"The keyword 'Molte Salt Fast Reactor' contains a typo; it should read 'Molten Salt Fast Reactor'.","section":"Keywords"},{"comment":"Reference 10 (Cammi et al., 'Multiphysics analysis of RANS-based turbulent transport of solid fission products in the Molten Salt Fast Reactor,' Nuclear Engineering and Design, 391, 111739 (2022)) and Reference 12 (Di Ronco et al., same title and journal details, with 'et. at.') appear to be the same reference; one should be removed or replaced.","section":"References"},{"comment":"The phrase 'continuous inlet/outlet' is ambiguous; clarify whether the inlet and outlet are separate openings at different locations or a single annular opening.","section":"Section 5.2"}],"recommendation":"major_revision","confidential_remarks":"The power inconsistency (300 MW vs 3000 MWth) is the most serious issue and may simply be a typographical error in the text or the table, but it must be resolved. The residence-time assumption also warrants a sensitivity study. The paper is otherwise an honest, preliminary demonstration of a useful coupling capability, and the authors already acknowledge the key physics limitation (no DNP feedback in OpenMC). The scope is appropriate for a journal paper if the load-bearing issues are fixed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a legitimate step forward—the first Cardinal MSFR model with DNP/DHP transport—and the single-physics verification is solid. But the recirculation boundary condition is internally inconsistent. The paper's own numbers (18 m³ total primary volume, half in the core, 1882.12 kg/s flow, 4125.3 kg/m³ density) imply an external-loop volume of about 9 m³ and a mean residence time of roughly 20 s, not the 2 s used in Sec. 5.2. That is load-bearing, not minor.\n\nWhat's good: the DNP/DHP equations are standard, the OpenMC–NekRS coupling is clearly described, and the separate OpenMC-vs-Serpent and Nek5000-vs-OpenFOAM checks give some real confidence. The authors are also upfront that results are preliminary and that OpenMC cannot yet move delayed neutron sources. The qualitative physics—short-lived precursors decay near the fission source, long-lived ones circulate and re-enter—is sensible and well illustrated.\n\nThe soft spots: the 2 s residence time is the big one. The stress-test math holds up. For DNP group 1 (λ = 0.012575 s⁻¹), using 20 s instead of 2 s changes the re-entering concentration by about 20%; for shorter-lived groups the error is much larger. Since the central claimed results are the DNP/DHP distributions, this makes the quantitative fields unsupported. No sensitivity analysis is provided; the paper defers it to future work. Other simplifications—constant fluid properties, no pumps or heat exchangers, a 50 cm inlet/outlet separation—are reasonable for a first pass but should be more carefully flagged. There is also no code or data release, which would help reproducibility.\n\nThe paper deserves a serious referee: it is clear, honest, and the main flaw is fixable. But a referee should require that the residence-time assumption be reconciled with the stated circuit volume and flow, or at minimum that a sensitivity study bound its effect. As written, I would not trust the precursor numbers.\n\nRecommendation: send to peer review, but expect major revision. I would cite it only if I worked on MSFR modeling, which I don't.","headline":"A real but incremental step for MSFR multiphysics, undermined by an internally inconsistent 2 s external-loop residence time that makes the quantitative DNP/DHP fields unsupported.","tokens_in":8536,"tokens_out":2366,"would_cite":false,"duration_ms":24851,"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":"This paper develops a coupled neutronics–thermal-hydraulics model of the Molten Salt Fast Reactor that transports delayed-neutron and decay-heat precursors through the primary circuit, and reports that the resulting temperature, heat…","keywords":["Molten Salt Fast Reactor","MSFR","Cardinal","OpenMC","NekRS","delayed neutron precursors","decay heat precursors","multiphysics coupling"],"falsifier":"Run the same coupled simulation with the assumed external-loop residence time varied from 1 s to 4 s, or with a distribution of residence times instead of a single average, and check whether the time-averaged temperature, DNP, and DHP fields stay within the reported agreement; if the outlet temperature or the long-lived precursor distributions shift noticeably, the recirculation model is controlling the claimed behavior. A complementary experiment would measure precursor concentrations at the core outlet and inlet of a circulating loop and compare the attenuation and phase lag against the 2-second decay model.","tokens_in":7431,"feed_emoji":"⚛️","tokens_out":8105,"duration_ms":80814,"temperature":0.7,"pith_summary":"The paper tries to establish that a high-fidelity multiphysics model of the Molten Salt Fast Reactor can represent the movement of delayed-neutron precursors and decay-heat precursors inside the circulating fuel salt. This matters because in a liquid-fuel reactor these precursors are born in the core and then carried through pumps and heat exchangers before returning, so where delayed neutrons and decay heat are released depends on the flow. The authors couple Monte Carlo neutronics with spectral-element CFD inside the Cardinal framework, solving convection–diffusion equations for six delayed-neutron and three decay-heat precursor groups. They report that the computed temperature, fission heat source, velocity, and precursor distributions show reasonable behavior: short-lived precursors decay near their birth sites, while long-lived ones survive the loop and re-enter the core. The model explicitly does not yet feed the drifted delayed-neutron sources back into the neutronics calculation, a limitation the authors flag as ongoing work.","feed_headline":"MSFR simulation tracks delayed-neutron precursors through the loop","feed_subtitle":"A coupled neutronics-CFD model adds drifting precursors and decay heat to the molten-salt fast reactor.","key_machinery":"The load-bearing object is the passive-scalar treatment of precursors inside NekRS: each DNP group $c_i$ and DHP group $d_l$ satisfies a convection–diffusion equation with a decay term and a source proportional to the fission source, Equations (7) and (8), with the volumetric heat source re-expressed through decay-heat groups in Equation (9). The recirculation condition closes the loop: at the outlet an average concentration per group is computed, multiplied by a decay factor for a 2-second residence time in pumps and heat exchangers, and imposed at the inlet. The coupling that makes this a multiphysics model is the Picard iteration in time managed by Cardinal, which transfers $q'''_{\\mathrm{fis}}$ and $R_{\\nu f}$ from OpenMC to NekRS and returns fuel temperature and density to OpenMC for the next neutronics update.","core_discovery":"The paper's central claim is that the MSFR can be modelled in Cardinal with neutronic–thermal-hydraulic feedback and with transport of delayed neutron precursors (DNPs) and decay heat precursors (DHPs), and that the resulting fields are physically reasonable. OpenMC supplies the fission heat source $q'''_{\\mathrm{fis}}$ and nu-fission rate $R_{\\nu f}$; NekRS solves the incompressible RANS $k$–$\\tau$ equations plus convection–diffusion equations for the precursor groups, with decay and source terms. The DNP/DHP transport equation is the key addition over earlier Cardinal MSFR work. For the external loop, the outlet concentration of each precursor group is averaged, decayed over an assumed 2-second residence time, and imposed at the inlet. The reported results show the expected ordering by decay constant: the shortest-lived DNP group tracks the nu-fission reaction rate, while the longest-lived group spreads through the circuit and re-enters the core; DHP groups show the analogous behavior. The paper is explicit that OpenMC cannot currently modify the location or intensity of delayed neutron sources, so DNP feedback is not included.","pith_inferences":["The fixed 2-second residence time is likely the largest source of model uncertainty for long-lived precursor groups; a simple sensitivity sweep would show whether the 'reasonable behavior' claim is robust to that choice.","Once delayed-neutron source positions can be adjusted in OpenMC, the same machinery should predict a drift-induced spatial shift in delayed neutron emission and hence a reactivity effect; the sign and size of that effect are not addressed in the paper.","The passive-scalar formulation should extend directly to other fission products or soluble species, so the same coupling could be used to estimate fission-product plate-out, corrosion species transport, or source terms in the primary loop.","The ability to track decay-heat precursors with the flow suggests that decay heat after shutdown may be distributed along the primary circuit rather than concentrated in the core, with implications for decay-heat removal strategies."],"forward_implications":["The model gives a way to compute where delayed neutrons and decay heat are actually deposited in the MSFR, rather than assuming they appear at the fission site.","The reported DNP distributions imply that spatial precursor drift can change the effective delayed-neutron distribution enough to matter for transient analysis, motivating the planned OpenMC extension.","The temperature field identifies stagnant regions in the core where the fuel salt is trapped and heated, pointing to locations that would need attention in a safety or thermal design review.","The verified single-physics components provide a baseline that can be reused when the coupled model is extended."],"supporting_citations":[{"why":"Earlier Cardinal MSFR model without precursor drift; this work extends it.","marker":"[2]"},{"why":"Deterministic MSFR transient model supplying DHP group constants and a comparison baseline.","marker":"[3]"},{"why":"Defines MSFR fuel composition, density, and reference temperatures used in the model.","marker":"[5]"},{"why":"OpenMC Monte Carlo code that computes k-eigenvalue, fission heat source, and nu-fission rate.","marker":"[6]"},{"why":"NekRS spectral-element CFD code that solves flow, energy, and passive-scalar precursor transport.","marker":"[7]"},{"why":"Cardinal coupling framework that transfers fields between OpenMC and NekRS via MOOSE.","marker":"[9]"},{"why":"RANS-based turbulent transport of fission products in the MSFR; basis for the passive-scalar precursor equations.","marker":"[10]"},{"why":"Six-group delayed neutron precursor constants (decay constants and fractions).","marker":"[11]"},{"why":"Nek5000/OpenFOAM 2D verification case for the RANS turbulence model.","marker":"[12]"}],"fun_headline_variants":["MSFR multiphysics model tracks delayed-neutron precursors","Coupled neutronics-CFD captures MSFR precursor drift","MSFR model adds precursor transport, but no DNP feedback yet","OpenMC-NekRS MSFR model follows decay heat precursors","MSFR simulation with drifting precursors and decay heat"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The model's predictions for the recirculating precursor fields rest on the assumption that all precursors that leave the core can be replaced by a single uniform concentration at the inlet after a fixed 2-second residence time in the pumps and heat exchangers, with no sensitivity study or measured residence-time distribution to justify that value.","fun_headline_variants_meta":{"raw":{"variants":["MSFR multiphysics model tracks delayed-neutron precursors","Coupled neutronics-CFD captures MSFR precursor drift","MSFR model adds precursor transport, but no DNP feedback yet","OpenMC-NekRS MSFR model follows decay heat precursors","MSFR simulation with drifting precursors and decay heat"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000587,"raw_usage":{"total_tokens":2772,"prompt_tokens":973,"completion_tokens":1799,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":589,"completion_tokens_details":{"reasoning_tokens":1716}},"tokens_in":589,"tokens_out":1799,"duration_ms":13655,"temperature":1.0,"reasoning_tokens":1716,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T19:53:43.866763+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same coupled simulation with the assumed external-loop residence time varied from 1 s to 4 s, or with a distribution of residence times instead of a single average, and check whether the time-averaged temperature, DNP, and DHP fields stay within the reported agreement; if the outlet temperature or the long-lived precursor distributions shift noticeably, the recirculation model is controlling the claimed behavior. A complementary experiment would measure precursor concentrations at the core outlet and inlet of a circulating loop and compare the attenuation and phase lag against the 2-second decay model.","supporting_citations":[{"cited_title":"Monte Carlo multiphysics simulation on adaptive unstructured mesh geometry","cited_arxiv_id":null,"evidence_quote":"Earlier Cardinal MSFR model without precursor drift; this work extends it."},{"cited_title":"Modelling and Analysis of the MSFR Transient Behavior,","cited_arxiv_id":null,"evidence_quote":"Deterministic MSFR transient model supplying DHP group constants and a comparison baseline."},{"cited_title":"Neutronic Benchmark of the Molten Salt Fast Reactor in the Frame of the EVOL and MARS Collaborative Projects,","cited_arxiv_id":null,"evidence_quote":"Defines MSFR fuel composition, density, and reference temperatures used in the model."},{"cited_title":"OpenMC: A State -of-the-Art Monte Carlo Code for Research and Development,","cited_arxiv_id":null,"evidence_quote":"OpenMC Monte Carlo code that computes k-eigenvalue, fission heat source, and nu-fission rate."},{"cited_title":"NekRS, a GPU -accelerated Spectral Element Navier –Stokes Solver ,","cited_arxiv_id":null,"evidence_quote":"NekRS spectral-element CFD code that solves flow, energy, and passive-scalar precursor transport."},{"cited_title":"Coupled Monte Carlo and Thermal -Fluid Modeling of High Temperature Gas Reactors Using Cardinal,","cited_arxiv_id":null,"evidence_quote":"Cardinal coupling framework that transfers fields between OpenMC and NekRS via MOOSE."},{"cited_title":"Multiphysics analysis of RANS-based turbulent transport of solid fission products in the Molten Salt Fast Reactor,","cited_arxiv_id":null,"evidence_quote":"RANS-based turbulent transport of fission products in the MSFR; basis for the passive-scalar precursor equations."},{"cited_title":"Nuclear Reactor Analysis","cited_arxiv_id":null,"evidence_quote":"Six-group delayed neutron precursor constants (decay constants and fractions)."},{"cited_title":"Multiphysics Analysis of RANS -Based Turbulent Transport of Solid Fission Products in the Molten Salt Fast Reactor,","cited_arxiv_id":null,"evidence_quote":"Nek5000/OpenFOAM 2D verification case for the RANS turbulence model."}],"review_version":1}