Pith. sign in

REVIEW 3 major objections 4 minor 55 references

Thermodynamics-Informed Input Reparameterization for Neural Prediction of Real-Fluid Thermodynamic Properties in Supercritical Combustion

T0 review · 3 major / 4 minor · reviewed 2026-08-01 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Neural surrogates for real-fluid thermodynamics become substantially more accurate when their leading input coordinate is matched to the target property: replacing raw enthalpy with an ideal-gas temperature estimate for T and an ideal-gas d

desk verdict Controlled study shows a cheap thermodynamic input reparameterization helps on held-out data; the transfer claim needs better evidence about coverage of the unseen flame. read the letter →

arxiv 2607.19241 v1 pith:GE2TH2XB submitted 2026-07-21 cs.LG physics.comp-phphysics.flu-dyn

Haoze Zhang , Han Li , Ke Xiao , Yangchen Xu , Runze Mao , Zhi X. Chen This is my paper · ORCID
classification cs.LGphysics.comp-phphysics.flu-dyn
keywords supercriticalcombustionreal-fluidthermodynamicsneural-networksurrogateinputreparameterizationenthalpy-basedclosurePeng–Robinsonequationofstatecompressibilitycoefficienttarget-aligned
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper argues that the learnability of neural surrogates for real-fluid thermodynamic closure in supercritical combustion depends less on network capacity than on which thermodynamic coordinate is supplied as the primary input. In the enthalpy-based pressure-correction formulation, the closure maps (h, p, Y) to (T, ρ, ψ), but raw enthalpy is a poorly informative coordinate for density and compressibility because it conflates caloric and volumetric information. The authors propose target-aligned input reparameterization (TAIR): for the temperature network, replace h with T̃, the temperature obtained by inverting a constant-c_p ideal-gas enthalpy approximation; for the density and compressibility networks, use ρ̃ = p/(R_m T̃), the ideal-gas density. These algebraic transformations are free of iteration and preserve the solver-available state, leaving the networks to learn only the real-fluid departure from an ideal-gas baseline. On supercritical methane–oxygen counterflow flame data, TAIR reduces held-out RMSE by factors of about 1.5, 2.0, and 7.5 for T, ρ, and ψ, with even larger gains at an unseen strain rate, while target-inconsistent cross-reparameterization performs worse—evidence that the benefit comes from thermodynamically matched input design rather than generic preprocessing.

What carries the argument

The central object is the TAIR coordinate transformation: T̃ = T0 + (h − Σ_i Y_i h_f,i) / (Σ_i Y_i c_p,i) and ρ̃ = p/(R_m T̃), both computed algebraically from solver-available variables and species constants. The paper also derives the ideal-gas identity (∂ρ̃/∂p)_{h,Y} = ρ̃/p, which motivates supplying ρ̃ to the compressibility network: since pressure is retained as an input, the pair (ρ̃, p) carries the leading ideal-gas baseline for the isenthalpic density response that ψ represents. The transformation functions as a thermodynamic preconditioner, exposing the dominant ideal-gas dependence of each target and leaving the networks to model only the real-fluid departure from that baseline.

What would settle it

Compute the TAIR-versus-raw-input RMSE separately for test points where the compressibility factor Z deviates most from unity (strongly non-ideal states). If the TAIR advantage disappears or reverses on these points, the claim that TAIR guides networks to learn real-fluid departures would be limited to states near the ideal-gas baseline.

Watch

Extended reading notes

Core claim

The paper establishes that a simple, physically motivated reparameterization of the input coordinate can substantially improve neural prediction of real-fluid properties. For the temperature network, replacing raw enthalpy h with an estimated ideal-gas temperature T̃—obtained by setting h equal to a mixture enthalpy with constant heat capacities—reduces the regression to a near-monotonic, weakly nonlinear map. For the density and compressibility networks, replacing h with the ideal-gas density estimate ρ̃ = p/(R_m T̃) aligns the leading pressure and composition dependence of the target, making the learned map close to linear over most of the sampled range. The cross-reparameterization contro

Load-bearing premise

The training data are generated by random augmentation in (T, p, Y) space around flame states at 1000, 5000, and 10000 s⁻¹, and the paper assumes this augmented distribution covers the thermodynamic states encountered in the unseen 3000 s⁻¹ flame; if the augmentation bounds miss correlations among h, p, and Y that occur in that flame, the reported transfer improvements could be optimistic.

Editorial extensions

If this is right

  • If TAIR's accuracy gains hold, neural thermodynamic surrogates can achieve the same closure error with smaller networks, freeing capacity for other parts of reacting-flow surrogates.
  • The accuracy improvement transfers to an unseen strain-rate flame within the augmented thermodynamic envelope, suggesting the method is useful for off-design operating conditions without retraining.
  • TAIR adds negligible computational overhead (about 8.8% over the raw-input network) while preserving a roughly 90-fold speedup over iterative real-fluid closure, making it directly compatible with expensive CFD solvers.
  • The target-matching principle is general: any target with an inexpensive analytical baseline could be reparameterized similarly, potentially extending to transport properties and other closure quantities.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper's results suggest a general 'baseline-preconditioning' design rule for neural surrogates: wherever a cheap physical approximation exists for a target, feed that approximation as an input so the network learns the residual. This likely applies beyond thermodynamics to chemical kinetics and turbulence-related regressions.
  • A testable extension would be to apply TAIR to other equations of state (e.g., SRK or multi-parameter) and other fuel–oxidizer pairs; the benefit should scale with how much of the target variance the ideal-gas baseline captures, and should shrink in strongly non-ideal near-critical regions.
  • The large improvement for ψ hints that finite-difference-derived quantities are especially hard to learn from raw enthalpy; since the paper does not enforce differential consistency between independently predicted ρ and ψ, a natural next step is to use the relation (∂ρ/∂p)_h,Y = ψ as a regularization or consistency loss.
  • The transfer test is a priori only; running TAIR surrogates inline in the pressure-correction solver would test whether the accuracy advantage survives error accumulation, pressure–density feedback, and excursions outside the training envelope.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper proposes target-aligned input reparameterization (TAIR) for neural surrogates of real-fluid thermodynamic closure in supercritical combustion. Instead of feeding raw enthalpy h to three MLPs predicting temperature T, density ρ, and compressibility ψ, TAIR uses an ideal-gas estimated temperature T̃ for the temperature network and an ideal-gas density ρ̃ for the density and compressibility networks, while retaining pressure p and composition Y. These transformations are explicit, use only solver-available state and species constants, and are invertible at fixed p and Y, so no information is lost. On a 1.2M-sample augmented database from supercritical methane–oxygen counterflow flames at 100 bar, TAIR reduces held-out RMSE by factors of about 1.5, 2.0, and 7.5 for T, ρ, and ψ relative to a raw-input baseline, and by factors of about 3.6, 14.5, and 6.0 on an unseen a=3000 s⁻¹ flame. Target-inconsistent cross-reparameterization controls perform worse, supporting the claim that the benefit comes from target-matched input design rather than generic preprocessing. The paper also reports a closure-level speed-up of about 89× relative to the iterative Cantera reference.

Significance. The methodological idea is simple, thermodynamically motivated, and potentially general: if a target property admits a cheap ideal-gas or reduced-order approximation, that approximation can serve as an input coordinate, leaving the network to learn only the departure. The controlled comparison—identical architecture, data, and training across raw-input, TAIR, and cross-reparameterization configurations—is well designed and directly isolates the effect of input reparameterization. The explicit statement of invertibility and the use of target-inconsistent controls are commendable and rule out the most obvious circularity concerns. If the quantitative claims are confirmed with full specification of the data-generation protocol and uncertainty quantification, this would be a useful contribution to neural thermodynamic surrogates for reacting-flow simulations.

major comments (3)
  1. [§II C, §III D, Abstract] The unseen-strain-rate transfer factors (3.6×, 14.5×, 6.0×) are load-bearing for the generalization claim and depend entirely on the assertion that the a=3000 s⁻¹ flame lies 'within the augmented thermodynamic envelope.' However, the augmentation procedure is not specified quantitatively: no bounds on temperature and pressure, no power-law exponent for species perturbation, and no acceptance/rejection thresholds are given. Fig. 3 does not overlay the a=3000 flame, and no coverage measure (e.g., local density of augmented samples on the unseen flame manifold, nearest-neighbor distances, or kernel density ratio) is provided. As reported, the experiment is not reproducible, and the improvement factors could reflect out-of-distribution artifact rather than a TAIR-specific benefit. Please specify all augmentation constants and provide quantitative coverage diagnostics for the a=3000 state dis
  2. [§II A, Eq. (10)–(11); §II C] Two numerical constants required for exact reproducibility are missing. First, the relative pressure perturbation ε used in the reference ψ evaluation is never given; because ψ is a supervised target, this value directly affects the reported RMSE numbers. Second, the augmentation 'prescribed bounds' and 'common power-law exponent' are referenced but not defined. These are not cosmetic details: the ψ target and the training distribution are determined by them. The authors should report ε and the full augmentation parameter set, including the ranges and the exponent, as part of the experimental setup.
  3. [§III C, §III D, §II D] All RMSE values are single-run point estimates with no repeated-seed statistics. Given stochastic optimization (random initialization, data shuffling, augmentation randomness) and the single data partition, the improvement factors—especially the 14.5× density improvement on the unseen flame—need uncertainty quantification to establish that the rankings are robust. I recommend reporting mean ± standard deviation (or minima/confidence intervals) over at least 5–10 independent training runs per configuration, with the same seeds across configurations for paired comparisons.
minor comments (4)
  1. [§II C / Fig. 3] Fig. 3 shows only 2D projections with point clouds; the local density of augmented samples is not visible. A density plot or histogram overlay would help assess coverage, especially for the unseen flame.
  2. [§III D] The transfer result is reported at a single time instant (t = 7×10⁻⁴ s). Since the flame is transient, a single snapshot may not represent the full manifold. Reporting error statistics over multiple snapshots or time-averaged fields would strengthen the transfer claim.
  3. [§II B, Eq. (14)] The notation for the constant heat capacity at T0 is introduced as c̊_p,i in the text but the equation appears to use a slightly different symbol. Please unify notation.
  4. [§IV B / Fig. 9] The sensitivity analysis selects 'one representative database state in each temperature bin,' but the binning procedure is not described (number of bins, selection rule). Please specify.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: TAIR is a parameter-free input reparameterization, and the reported accuracy gains are empirical comparisons against external Cantera reference targets.

full rationale

The derivation chain is self-contained. The transformed coordinates are explicit algebraic functions of the solver-available state: T-tilde in Eq. (15) is computed from h, Y, and fixed species constants (T0, h_f,i, c_p,i), while rho-tilde in Eq. (17) is computed from p, T-tilde, and Rm. No coefficient is fitted to the target values T, rho, or psi, and the paper explicitly states that the transformations 'use only solver-available variables and species constants.' At fixed p and Y, the map h <-> T-tilde is invertible (Sec. IV A), so TAIR does not smuggle in target information; it changes the regression coordinates, not the information content. The reference targets come from the Cantera Peng-Robinson closure, an external code, so the RMSE reductions are empirical benchmark outcomes rather than identities. The only self-citations (DeepFlame solver [26], augmentation procedure [31]) are used as tools or data-generation choices, not as load-bearing theorems that force the reparameterization; both the held-out and unseen-strain-rate claims are evaluated by actual predictions against reference states. The paper explicitly discloses limitations: the unseen-strain-rate test is 'within the augmented thermodynamic envelope' and 'does not, by itself, establish extrapolation beyond that envelope or stable inline coupling,' and it notes that no differential consistency is enforced. Even if the augmentation-envelope coverage is under-specified, that is a reproducibility/correctness concern, not a circularity in the derivation.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

The paper's central empirical claim rests on a reference real-fluid closure, a chosen coordinate transformation, and a data-generation distribution. The transformation uses standard thermochemical constants; the main hand-chosen inputs are the augmentation bounds, the reference pressure-perturbation ε, the network architecture, and the trained weights. No new physical entities are introduced.

free parameters (4)
  • MLP weights and biases = trained on 1.0e6 samples, 1500 epochs
    All three property networks are fitted to the augmented database; the reported RMSE comparisons are evaluated at these fitted weights.
  • Augmentation bounds and power-law exponent = not reported
    Section II C: 'temperature and pressure are randomly varied within prescribed bounds' and species fractions are perturbed with a common power-law exponent; these hand-chosen values define the 'augmented thermodynamic envelope' used for the transfer claim.
  • Reference pressure perturbation epsilon = not reported
    Section II A, Eq. (10): the reference compressibility ψ is computed with a first-order backward finite difference using p' = p(1−ε); the value of ε affects the reference ψ targets.
  • Network hyperparameters (64-32-16 widths, batch 1024, epochs 1500, cosine LR) = chosen by hand
    Fixed across configurations, but these choices affect attainable accuracy and therefore the magnitude of the reported improvement factors.
assumptions (6)
  • domain assumption Peng–Robinson EoS with classical one-fluid mixing rules is an adequate reference for the sampled supercritical states.
    Section II A, Eq. (8); all reference targets and speed-up claims are evaluated against this EoS.
  • domain assumption Cantera's reference thermodynamic implementation is correct for the sampled states.
    Section II A: 'reference states... evaluated using the Cantera Peng–Robinson implementation' and used as the supervised targets.
  • ad hoc to paper Constant-cp ideal-gas enthalpy inversion at T0 yields a useful, information-preserving input coordinate.
    Eqs. (14)–(15); this is the paper's design choice—an idealization chosen because it is algebraically invertible and cheap to evaluate.
  • ad hoc to paper Ideal-gas density estimate ρ̃ = p/(Rm T̃) carries the leading ideal-gas dependence of ρ and ψ.
    Eqs. (17)–(18); used to select the density/compressibility input coordinate.
  • domain assumption Random augmentation in (T,p,Y) produces a training distribution whose augmented envelope covers the unseen strain-rate flame manifold.
    Section II C and Section III D; the transfer claim depends on the a=3000 s⁻¹ flame lying within the augmented thermodynamic envelope.
  • standard math At fixed p and Y, T̃ is an invertible affine function of h, and ρ̃ is determined by T̃, p, Y.
    Section IV A: the invertibility ensures all input configurations contain the same state information, so observed gains are due to preconditioning rather than added information.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Thermodynamics-Informed Input Reparameterization for Neural Prediction of Real-Fluid Thermodynamic Properties in Supercritical Combustion." pith.science (2026). https://pith.science/paper/GE2TH2XB

@misc{pith2026260719241,
  author       = {Pith},
  title        = {Pith review of: Thermodynamics-Informed Input Reparameterization for Neural Prediction of Real-Fluid Thermodynamic Properties in Supercritical Combustion},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GE2TH2XB}},
  note         = {Machine review of arXiv:2607.19241}
}
abstract

Real-fluid thermodynamic property evaluation is a major computational cost in supercritical combustion simulations. In the enthalpy-based pressure-correction formulation, the closure evaluates temperature T, density $\rho$, and compressibility coefficient $\psi$ from the solver state (h,p,Y) through enthalpy-temperature inversion and repeated real-fluid equation-of-state evaluations. Neural-network surrogates offer fixed-cost inference, but direct mapping from (h,p,Y) to $(T,\rho,\psi)$ must capture the enthalpy-temperature relation and non-ideal equation-of-state response, resulting in a complex regression problem. This work introduces a thermodynamics-informed input reparameterization strategy, termed target-aligned input reparameterization (TAIR). TAIR replaces the raw enthalpy coordinate of each property network with a target-matched thermodynamic coordinate: the temperature network uses a temperature estimate obtained by inverting a constant-$c_p$ ideal-gas mixture enthalpy approximation, whereas the density and compressibility networks use an ideal-gas density estimate. These algebraic transformations use only solver-available variables and species constants, guiding the networks to learn real-fluid departures from ideal-gas baselines rather than reconstructing the full closure from raw enthalpy. The method is assessed using supercritical methane-oxygen counterflow flame data against a raw-input baseline and target-inconsistent cross-reparameterization controls. TAIR reduces held-out RMSE by factors of about 1.5, 2.0, and 7.5 for T, $\rho$, and $\psi$, respectively. For an unseen strain-rate flame within the augmented thermodynamic envelope, the corresponding factors are 3.6, 14.5, and 6.0. The target-inconsistent controls perform worse, indicating that the gains arise from thermodynamically matched input design rather than generic preprocessing.

Figures

Figures reproduced from arXiv: 2607.19241 by the authors.

Figure 1
Figure 1. schematically summarizes this cell-local workflow. Temperature recovery and density evaluation are shown se￾quentially for clarity, although they are numerically cou￾pled because each temperature iteration requires an EoS state evaluation. Evaluation of ψ further requires repeat￾ing the temperature–density closure at the perturbed pressure. The reference thermodynamic states underlying Eqs. (4)–(11) are evaluated us… view at source ↗
Figure 2
Figure 2. FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 4
Figure 4. FIG. 4 [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figures from the paper (5 more)
Figure 5
Figure 5. Figure 5: FIG. 5 [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6 [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: compares the network predictions with the reference values on the held-out test subset of the augmented database. All three configurations produce predictions close to the line of exact agreement, but their errors differ substantially when [PITH_FULL_IMAGE:figures/ful…
Figure 8
Figure 8. Figure 8: FIG. 8 [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]
Figure 9
Figure 9. Figure 9: FIG. 9 [PITH_FULL_IMAGE:figures/full_fig_p011_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 23 canonical work pages

  1. [1]

    Neural Networks , volume =

    Yarotsky, Dmitry , title =. Neural Networks , volume =. 2017 , doi =

    2017

  2. [2]

    and Bengio, Yoshua and Courville, Aaron , title =

    Rahaman, Nasim and Baratin, Aristide and Arpit, Devansh and Draxler, Felix and Lin, Min and Hamprecht, Fred A. and Bengio, Yoshua and Courville, Aaron , title =. Proceedings of the 36th International Conference on Machine Learning , series =

  3. [3]

    J. C. Oefelein. Mixing and combustion of cryogenic oxygen-hydrogen shear-coaxial jet flames at supercritical pressure. Combust. Sci. Technol. 2006. doi:10.1080/00102200500325322

    work page doi:10.1080/00102200500325322 2006

  4. [4]

    Liquid Rocket Engine Combustion Instability. 1995. doi:10.2514/4.866371

    work page doi:10.2514/4.866371 1995

  5. [5]

    Mayer and A

    W. Mayer and A. Schik and M. Sch \"a ffler and H. Tamura. Injection and mixing processes in high-pressure liquid oxygen/gaseous hydrogen rocket combustors. J. Propul. Power. 2000. doi:10.2514/2.5647

    work page doi:10.2514/2.5647

  6. [6]

    J. Bellan. Supercritical (and subcritical) fluid behavior and modeling: drops, streams, shear and mixing layers, jets and sprays. Prog. Energy Combust. Sci. 2000. doi:10.1016/S0360-1285(00)00008-3

    work page doi:10.1016/s0360-1285(00)00008-3 2000

  7. [7]

    Lacaze and J

    G. Lacaze and J. C. Oefelein. A non-premixed combustion model based on flame structure analysis at supercritical pressures. Combust. Flame. 2012. doi:10.1016/j.combustflame.2012.02.003

    work page doi:10.1016/j.combustflame.2012.02.003 2012

  8. [8]

    P. C. Ma and Y. Lv and M. Ihme. An entropy-stable hybrid scheme for simulations of transcritical real-fluid flows. J. Comput. Phys. 2017. doi:10.1016/j.jcp.2017.03.022

    doi:10.1016/j.jcp.2017.03.022 2017

Show all 55 references
  1. [9]

    Peng and D

    D.-Y. Peng and D. B. Robinson. A new two-constant equation of state. Ind. Eng. Chem. Fundam. 1976. doi:10.1021/i160057a011

    1976 doi

  2. [10]

    G. Soave. Equilibrium constants from a modified Redlich-Kwong equation of state. Chem. Eng. Sci. 1972

    1972

  3. [11]

    B. E. Poling and J. M. Prausnitz and J. P. O'Connell. The Properties of Gases and Liquids. 2001

    2001

  4. [12]

    S. L. Brunton and B. R. Noack and P. Koumoutsakos. Machine learning for fluid mechanics. Annu. Rev. Fluid Mech. 2020

    2020

  5. [13]

    Yang and Y

    S. Yang and Y. Li and X. Wang and U. Unnikrishnan and V. Yang and W. Sun. Comparison of Tabulation and Correlated Dynamic Evaluation of Real Fluid Properties for Supercritical Mixing. 53rd AIAA/SAE/ASEE Joint Propulsion Conference. 2017. doi:10.2514/6.2017-4858

    2017 doi

  6. [14]

    Wang and H

    X. Wang and H. Huo and U. Unnikrishnan and V. Yang. A systematic approach to high-fidelity modeling and efficient simulation of supercritical fluid mixing and combustion. Combust. Flame. 2018. doi:10.1016/j.combustflame.2018.04.030

    2018 doi

  7. [15]

    Kawai and H

    S. Kawai and H. Terashima and H. Negishi. A robust and accurate numerical method for transcritical turbulent flows at supercritical pressure with an arbitrary equation of state. J. Comput. Phys. 2015. doi:10.1016/j.jcp.2015.07.047

    2015 doi

  8. [16]

    M. L. Huber and E. W. Lemmon and I. H. Bell and M. O. McLinden. The NIST REFPROP database for highly accurate properties of industrially important fluids. Ind. Eng. Chem. Res. 2022. doi:10.1021/acs.iecr.2c01427

    2022 doi

  9. [17]

    Liu and J

    Z. Liu and J. Liang and Y. Pan. Construction of thermodynamic properties look-up table with block-structured adaptive mesh refinement method. J. Thermophys. Heat Transfer. 2014. doi:10.2514/1.T4273

    2014 doi

  10. [18]

    Guideline on the Fast Calculation of Steam and Water Properties with the Spline-Based Table Look-Up Method ( SBTL )

    International Association for the Properties of Water and Steam. Guideline on the Fast Calculation of Steam and Water Properties with the Spline-Based Table Look-Up Method ( SBTL ). 2015

    2015

  11. [19]

    Alfaro-Isac and S

    C. Alfaro-Isac and S. Izquierdo-Estallo and J. Sierra-Pallares. Reduced-order modelling of equations of state using tensor decomposition for robust, accurate and efficient property calculation in high-pressure fluid flow simulations. J. Supercrit. Fluids. 2020. doi:10.1016/j.s...

    2020

  12. [20]

    Koukouvinis and C

    P. Koukouvinis and C. Rodriguez and J. Hwang and I. Karathanassis and M. Gavaises and L. Pickett. Machine learning and transcritical sprays: A demonstration study of their potential in ECN Spray-A. Int. J. Engine Res. 2022. doi:10.1177/14680874211020292

    2022 doi

  13. [21]

    D. N. Nguyen and C. S. Yoo. An OpenFOAM -based solver for modeling low Mach number turbulent flows at high pressure with real-fluid effects. Comput. Phys. Commun. 2025. doi:10.1016/j.cpc.2025.109600

    2025

  14. [22]

    o ll and T. Hitz and C. M \

    F. F \"o ll and T. Hitz and C. M \"u ller and C.-D. Munz and M. Dumbser. On the use of tabulated equations of state for multi-phase simulations in the homogeneous equilibrium limit. Shock Waves. 2019. doi:10.1007/s00193-019-00896-1

    2019 doi

  15. [23]

    D. T. Banuti. A critical assessment of adaptive tabulation for fluid properties using neural networks. AIAA Scitech 2021 Forum. 2021. doi:10.2514/6.2021-0927

    2021 doi

  16. [24]

    Roach and G.-M

    L. Roach and G.-M. Rignanese and A. Erriguible and C. Aymonier. Applications of machine learning in supercritical fluids research. J. Supercrit. Fluids. 2023. doi:10.1016/j.supflu.2023.106051

    2023

  17. [25]

    Zhu and E

    K. Zhu and E. A. M \"u ller. Generating a Machine-Learned Equation of State for Fluid Properties. J. Phys. Chem. B. 2020. doi:10.1021/acs.jpcb.0c05806

    2020 doi

  18. [26]

    Li and T

    Y. Li and T. Zhang and S. Sun. Acceleration of the NVT Flash Calculation for Multicomponent Mixtures Using Deep Neural Network Models. Ind. Eng. Chem. Res. 2019. doi:10.1021/acs.iecr.9b00527

    2019 doi

  19. [27]

    Sahranavardfard and D

    N. Sahranavardfard and D. Aubagnac-Karkar and G. Costante and F. N. Z. Rahantamialisoa and C. Habchi and M. Battistoni. Computation of real-fluid thermophysical properties using a neural network approach implemented in OpenFOAM. Fluids. 2024. doi:10.3390/fluids9030056

    2024 doi

  20. [28]

    Shahsavan Markadeh and A

    R. Shahsavan Markadeh and A. Arabkhalaj and M. Birouk. On the density prediction of multicomponent hydrocarbon mixtures at near- and supercritical conditions. Fluid Phase Equilibria. 2022. doi:10.1016/j.fluid.2022.113560

    2022

  21. [29]

    M. R. Dobbelaere and Y. Ureel and F. H. Vermeire and L. Tomme and C. V. Stevens and K. M. Van Geem. Machine learning for physicochemical property prediction of complex hydrocarbon mixtures. Ind. Eng. Chem. Res. 2022. doi:10.1021/acs.iecr.2c00442

    2022 doi

  22. [30]

    J. H. Ferziger and M. Peri \'c and R. L. Street. Computational Methods for Fluid Dynamics. 2019

    2019

  23. [31]

    Monnier and G

    F. Monnier and G. Ribert and L. Duhem-Duvilla. CH4/O2 supercritical flame structure and simulation. Proc. Combust. Inst. 2024. doi:10.1016/j.proci.2024.105514

    2024

  24. [32]

    Chung and M

    T.-H. Chung and M. Ajlan and L. L. Lee and K. E. Starling. Generalized multiparameter correlation for nonpolar and polar fluid transport properties. Ind. Eng. Chem. Res. 1988

    1988

  25. [33]

    Bengio and A

    Y. Bengio and A. Courville and P. Vincent. Representation learning: A review and new perspectives. IEEE Trans. Pattern Anal. Mach. Intell. 2013

    2013

  26. [34]

    Wan and M

    T. Wan and M. Zhou and P. Zhao and X. Wang. Challenges in the modeling and simulation of turbulent supercritical fluid flows and heat transfer. Propulsion and Energy. 2025. doi:10.1007/s44270-024-00005-3

    2025 doi

  27. [35]

    P. J. Milan and J.-P. Hickey and X. Wang and V. Yang. Deep-learning accelerated calculation of real-fluid properties in numerical simulation of complex flowfields. J. Comput. Phys. 2021. doi:10.1016/j.jcp.2021.110567

    2021

  28. [36]

    Xu and Y

    J. Xu and Y. Xu and Z. Weng and Y. Cai and R. Mao and R. Yang and Z. X. Chen. Detailed simulation of LOX/GCH4 flame-vortex interaction in supercritical Taylor--Green flows with machine learning. 2023. doi:10.48550/arXiv.2312.04830

    2023 doi arXiv

  29. [37]

    Cai and R

    Y. Cai and R. Yang and H. Li and J. Xu and K. Xiao and Z. X. Chen and H. Wang. Efficient machine learning method for supercritical combustion: Predicting real-fluid properties and chemical ODEs. Aerosp. Sci. Technol. 2025. doi:10.1016/j.ast.2025.110034

    2025

  30. [38]

    Ding and T

    T. Ding and T. Readshaw and S. Rigopoulos and W. P. Jones. Machine learning tabulation of thermochemistry in turbulent combustion: An approach based on hybrid flamelet/random data and multiple multilayer perceptrons. Combust. Flame. 2021. doi:10.1016/j.combustflame.2021.111493

    2021

  31. [39]

    Li and R

    H. Li and R. Yang and Y. Xu and M. Zhang and R. Mao and Z. X. Chen. Comprehensive deep learning for combustion chemistry integration: Multi-fuel generalization and a posteriori validation in reacting flow. Phys. Fluids. 2025. doi:10.1063/5.0248582

    2025 doi

  32. [40]

    Readshaw and T

    T. Readshaw and T. Ding and S. Rigopoulos and W. P. Jones. Modeling of turbulent flames with the large eddy simulation–probability density function (LES–PDF) approach, stochastic fields, and artificial neural networks. Phys. Fluids. 2021. doi:10.1063/5.0041122

    2021 doi

  33. [41]

    Mao and M

    R. Mao and M. Lin and Y. Zhang and T. Zhang and Z.-Q. J. Xu and Z. X. Chen. DeepFlame : A deep learning empowered open-source platform for reacting flow simulations. Comput. Phys. Commun. 2023. doi:10.1016/j.cpc.2023.108842

    2023

  34. [42]

    D. G. Goodwin and H. K. Moffat and I. Schoegl and R. L. Speth and B. W. Weber. Cantera: An object-oriented software toolkit for chemical kinetics, thermodynamics, and transport processes. 2023. doi:10.5281/zenodo.8137090

    2023 doi

  35. [43]

    Hendrycks and K

    D. Hendrycks and K. Gimpel. Gaussian error linear units ( GELU s). 2016. doi:10.48550/arXiv.1606.08415

    2016 doi arXiv

  36. [44]

    D. P. Kingma and J. Ba. Adam: A method for stochastic optimization. 3rd International Conference on Learning Representations ( ICLR 2015). 2015

    2015

  37. [45]

    Loshchilov and F

    I. Loshchilov and F. Hutter. SGDR : Stochastic gradient descent with warm restarts. 5th International Conference on Learning Representations ( ICLR 2017). 2017

    2017

  38. [46]

    Hornik and M

    K. Hornik and M. Stinchcombe and H. White. Multilayer feedforward networks are universal approximators. Neural Networks. 1989

    1989

  39. [47]

    D. T. Banuti. Crossing the Widom -line -- Supercritical pseudo-boiling. J. Supercrit. Fluids. 2015. doi:10.1016/j.supflu.2014.12.019

    2015 doi

  40. [48]

    V. Yang. Modeling of supercritical vaporization, mixing, and combustion processes in liquid-fueled propulsion systems. Proc. Combust. Inst. 2000. doi:10.1016/S0082-0784(00)80299-4

    doi

  41. [49]

    Ribert and N

    G. Ribert and N. Zong and V. Yang and L. Pons and N. Darabiha and S. Candel. Counterflow diffusion flames of general fluids: Oxygen/hydrogen mixtures. Combust. Flame. 2008. doi:10.1016/j.combustflame.2008.04.023

    2008 doi

  42. [50]

    Schmitt and L

    T. Schmitt and L. Selle and A. Ruiz and B. Cuenot. Large-eddy simulation of supercritical-pressure round jets. AIAA J. 2010. doi:10.2514/1.J050288

    2010 doi

  43. [51]

    G. E. Karniadakis and I. G. Kevrekidis and L. Lu and P. Perdikaris and S. Wang and L. Yang. Physics-informed machine learning. Nat. Rev. Phys. 2021. doi:10.1038/s42254-021-00314-5

    2021 doi

  44. [52]

    A. C. Hindmarsh and P. N. Brown and K. E. Grant and S. L. Lee and R. Serban and D. E. Shumaker and C. S. Woodward. SUNDIALS : Suite of nonlinear and differential/algebraic equation solvers. ACM Trans. Math. Softw. 2005. doi:10.1145/1089014.1089020

    2005

  45. [53]

    Combustion and Flame , volume =

    Monnier, Florian and Ribert, Guillaume , title =. Combustion and Flame , volume =. 2022 , issn =. doi:10.1016/j.combustflame.2021.111735 , url =

    2022

  46. [54]

    Sokolić, Jure and Giryes, Raja and Sapiro, Guillermo and Rodrigues, Miguel R. D. , journal=. Robust Large Margin Deep Neural Networks , year=

  47. [55]

    Proceedings of the 31st International Conference on Machine Learning , pages =

    Learning Polynomials with Neural Networks , author =. Proceedings of the 31st International Conference on Machine Learning , pages =. 2014 , editor =

    2014

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.