Pith. sign in

REVIEW 43 references

Neural Differential Equations for Oscillatory Flows in Aeroelasticity Applied to Transonic Buffet

T0 review · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read The paper claims that a hybrid reduced-order model — a Rayleigh oscillator for the self-excited buffet, a pruned multi-input Volterra series for structural-motion memory, and a small neural-network correction — identified from a single forc

arxiv 2607.22402 v2 pith:FY4MKX7B submitted 2026-07-24 physics.flu-dyn physics.comp-ph

Michael Candon , Pier Marzocca , Earl Dowell This is my paper · ORCID
classification physics.flu-dynphysics.comp-ph
keywords transonicbuffetreduced-ordermodelingneuraldifferentialequationsRayleighoscillatorVolterraserieslimitcycleoscillationsfrequencylock-inmulti-inputsystemidentification
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

This paper claims that the expensive cycle of full-order CFD for transonic buffet aeroelasticity can be replaced by a hybrid reduced-order model trained from a single prescribed-motion CFD simulation. The model stacks three ingredients: a Rayleigh oscillator that generates the self-sustained buffet oscillation, a finite-memory multi-input Volterra series that carries direct and cross-modal aerodynamic memory, and a small neural-network correction. Coupled to a two-degree-of-freedom airfoil structure, it reproduces full-order predictions of aeroelastic stability, frequency lock-in, and limit-cycle amplitudes, and it is cheap enough to map large regions of parameter space. The maps lead to a physical conclusion: buffet-driven instabilities in this configuration are single-mode in nature, while modal coupling can suppress subcritical instability, smooth it into supercritical growth, or let one mode excite another.

What carries the argument

The central object is the physics-guided neural differential equation: each generalized aerodynamic force obeys an acceleration equation made of three additive terms. A Rayleigh oscillator supplies the autonomous buffet limit cycle and its saturation; a diagonally pruned finite-memory multi-input Volterra series represents direct and nonlinear cross-modal forcing from structural displacement/velocity histories; a compact feedforward neural network adds a residual correction. All parameters — oscillator coefficients, Volterra kernels, network weights — are optimized jointly by backpropagation through a forward-Euler rollout, so the network augments rather than replaces the physical backbone.

What would settle it

Run a full-order CFD aeroelastic simulation in the ROM-predicted subcritical regime (for example mode-2 frequency ratio 0.90, mode-1-to-mode-2 frequency ratio 0.81, zero structural damping) with a large initial modal velocity perturbation. If no coexisting low-amplitude buffet branch and large-amplitude LCO branch appear, or if the LCO does not persist down to very small fluid-to-structural mass ratio, then the ROM's extrapolation from forced-motion training to free response is unsupported.

Watch

Extended reading notes

Core claim

The central claim is that one prescribed-motion CFD simulation, in which all retained structural modes are excited simultaneously with orthogonal band-limited signals, carries enough information to identify a multi-input neural differential equation ROM for transonic buffet. In the reported best variant — first-order diagonally pruned Volterra memory in heave velocity and pitch, plus a 16-hidden-unit neural correction — the ROM reaches a cross-validation NRMSD of 2.75% and, when coupled to a typical-section structural model, reproduces full-order predictions of LCO amplitude and frequency across frequency-ratio sweeps, including the onset and upper boundary of lock-in and the effect of struc

Load-bearing premise

The load-bearing premise is that one prescribed-motion CFD run with all modes excited simultaneously reveals enough about direct and cross-modal aerodynamic forces that the model, trained on forced motion, can be trusted to predict free aeroelastic response outside the training envelope — and the multi-mode regime maps are produced by the ROM without full-order confirmation (Sections 2.2.4 and 4.2.4).

Editorial extensions

If this is right

  • A single forced-motion CFD run can yield a reusable aerodynamic ROM for a given flow condition, making wide sweeps over structural frequency, damping, and mass ratio computationally feasible.
  • The ROM reproduces the known buffet lock-in asymmetry — pitch-only instability above a frequency ratio of unity and heave-only instability below it — and reveals that a pitch-dominated mixed mode can become unstable below unity at large static unbalance.
  • The response maps imply that buffet-driven instabilities in this configuration are single-mode in nature: modal coupling can suppress a subcritical instability, smooth it into a supercritical branch, or allow one mode to drive another, but no classical two-mode flutter appears in the explored parameter space.
  • The predicted supercritical, subcritical, and transitional regimes, with modal amplitude ratios of order one, provide concrete points where the ROM's extrapolation can be tested by full-order CFD or experiment.

Reading between the lines

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

  • The same architecture should transfer to other self-excited oscillatory flows — vortex-induced vibration, galloping, non-synchronous turbomachinery vibrations — where a Rayleigh-oscillator backbone is already standard; the Volterra memory and neural correction would be the parts needing re-identification.
  • Because the paper trains in physical heave/pitch coordinates and then projects onto modal bases, the identified aerodynamic operator may be reusable for different structural mode shapes within the validated frequency and amplitude envelope; the paper does not demonstrate that reuse beyond the reported cases.
  • The regime maps in the coupled-mode section are produced by the ROM alone; the most decisive next test is a full-order CFD run at the ω1/ω2 = 0.80 transition to see whether the abrupt supercritical-to-subcritical switch is real.
  • The absence of classical coupled-mode flutter should be read as a statement about this airfoil and parameter range, not a general law; the same tool could be used to search for genuinely coupled-mode instabilities in configurations with closer modal frequencies or different mode shapes.
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.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The model is deliberately parameter-rich: fitted Rayleigh, Volterra, and NN coefficients plus hand-chosen hyperparameters. The central aeroelastic prediction is not directly fitted (no aeroelastic training data), but the physical-backbone assumptions are all carried in from prior literature or chosen for tractability.

free parameters (6)
  • Rayleigh growth/saturation coefficients epsilon_i = not reported
    Fitted jointly with Volterra and NN parameters via backpropagation over rollout (Eq. 15); govern linear growth and saturation of each generalized aerodynamic force.
  • Rayleigh amplitude normalization alpha_i = defined via Q_i,ref from data
    Set by Eq. (4) using the observed generalized-force oscillation amplitude Q_i,ref from the CFD data.
  • Rayleigh autonomous frequency omega_F = buffet frequency
    Identified from the unforced buffet portion of the CFD trajectory (Sec. 2.2.3) or fitted; sets the fluid limit-cycle frequency.
  • Volterra kernel coefficients H (including A_ij) = not reported
    Fitted end-to-end; include direct and cross-modal memory kernels; N_L=50 lags, orders 1-3 tested.
  • Neural-network weights/biases W,b = not reported
    Fitted jointly via backprop; 16 hidden units per layer; L2 regularization with unspecified lambda.
  • Hyperparameters (N_L=50, p=1..3, hidden units=16, excitation band, dt=0.1) = N_L=50, HU=16
    Chosen by hand based on authors' prior work; not optimized systematically in this paper.
assumptions (5)
  • domain assumption URANS with SST k-omega and curvature correction accurately resolves transonic buffet for OAT15A at M=0.73
    All training and FOM reference data come from this CFD model (Sec. 3.1); if the CFD is wrong, ROM agreement with FOM does not imply physical accuracy.
  • ad hoc to paper The diagonal pruning of multi-input Volterra kernels retains the essential nonlinear coupling
    Eq. (6) keeps only k1=...=kp to reduce cost; no convergence or error analysis with respect to full kernels is provided.
  • domain assumption The neural correction trained on prescribed motion generalizes to free aeroelastic response
    The ROM is trained on forced-motion rollouts (Sec. 2.2.4) then applied as an aeroelastic predictor; there is no proof or independent dataset demonstrating this transfer.
  • domain assumption Rayleigh oscillator is a sufficient autonomous model for buffet shock oscillation
    Used to generate self-sustained oscillations (Sec. 2.2.1); based on prior oscillator models, not derived from flow physics.
  • domain assumption Simultaneous orthogonal band-limited excitation of all modes yields identifiable direct and cross-modal kernels
    Training uses one CFD run with all modes excited together (Sec. 2.2.4); the authors do not report identifiability or uniqueness tests.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Neural Differential Equations for Oscillatory Flows in Aeroelasticity Applied to Transonic Buffet." pith.science (2026). https://pith.science/paper/FY4MKX7B

@misc{pith2026260722402,
  author       = {Pith},
  title        = {Pith review of: Neural Differential Equations for Oscillatory Flows in Aeroelasticity Applied to Transonic Buffet},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FY4MKX7B}},
  note         = {Machine review of arXiv:2607.22402}
}
read the original abstract

Self-excited aerodynamic flows arise across a broad range of systems and can drive nonlinear fluid-structure interactions and aeroelastic instabilities that are challenging and computationally expensive to predict. This paper presents a physics-guided neural differential equation (DE) reduced order model (ROM) combining a nonlinear fluid oscillator, a finite-memory multi-input Volterra series, and a compact neural network correction. The multi-input aerodynamic formulation is generalized to m structural modes, capturing direct and nonlinear cross-modal coupling. The model is identified from a single prescribed-motion CFD simulation with simultaneous excitation of all retained structural modes, and is then coupled with the structural equations of motion for efficient aeroelastic prediction. Applied to transonic buffet over the ONERA OAT15A airfoil, the time-marching ROM predicts aeroelastic stability, frequency lock-in, and limit cycle amplitudes in good agreement with full-order reference solutions. The ROM is used to provide substantial new insight into buffet-induced aeroelastic instabilities involving more than one structural mode.

Figures

Figures reproduced from arXiv: 2607.22402 by the authors.

Figure 1
Figure 1. Computational grid(47,400 cells)  m mxαb mxαb Iα  h¨ α¨  +  chh chα chα cαα h˙ α˙  +  kh 0 0 kα  h α  =  L Mc/4  (18) Here, m is the sectional mass, Iα is the sectional moment of inertia about the elastic axis, and xαb is the offset between the center of mass and elastic axis. The quantities ωh and ωα are the heave and pitch natural frequencies, while ζh and ζα are the corresponding structural damping … view at source ↗
Figure 2
Figure 2. Validation loss histories for the three model variants during training. [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Time histories for a cross-validation case comparing the FOM (CFD) and R-V1-NN solutions. [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (14 more)
Figure 4
Figure 4. Figure 4: Comparison of FOM and ROM predictions of LCO amplitude and frequency versus natural frequency [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]
Figure 5
Figure 5. Figure 5: Comparison of FOM and ROM (R-V1-NN) predictions of LCO amplitude and frequency versus natural [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 6
Figure 6. Figure 6: Lissajous curves of aerodynamic moment, CM, versus pitch angle, α, comparing the FOM and ROM (R-V1-NN) for representative natural frequency ratios [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]
Figure 7
Figure 7. Figure 7: Comparison of FOM and ROM (R-V1-NN) predictions of pitch response time histories for representative [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]
Figure 8
Figure 8. Figure 8: Cross-validation of the multi-input ROM under simultaneous prescribed heave and pitch excitation, [PITH_FULL_IMAGE:figures/full_fig_p012_8.png]
Figure 9
Figure 9. Figure 9: Pure-heave instability with fˆα = 1.5, xα = 0, and ζh = ζα = 0.005, showing the LCO amplitudes, frequency, and heave-to-pitch amplitude ratio. 13 [PITH_FULL_IMAGE:figures/full_fig_p013_9.png]
Figure 10
Figure 10. Figure 10: Pure-pitch instability with fˆh = 0.6, xα = 0, and ζh = ζα = 0.005, showing the LCO amplitudes, frequency, and heave-to-pitch amplitude ratio. 4.2.3 Single-Degree-of-Freedom Instabilities of Coupled Heave-Pitch Modes Next, the center of gravity is shifted from xCG = x…
Figure 11
Figure 11. Figure 11: Mode 2 instability with fˆ1 = 0.515, xα = 0.43, and ζ1 = ζ2 = 0.00, showing the LCO amplitudes, frequency, and amplitude ratios [PITH_FULL_IMAGE:figures/full_fig_p015_11.png]
Figure 12
Figure 12. Figure 12: LCO frequency ratio as a function of the mode 2 natural frequency and static unbalance with [PITH_FULL_IMAGE:figures/full_fig_p015_12.png]
Figure 13
Figure 13. Figure 13: LCO amplitudes as functions of the mode 2 natural frequency and static unbalance in the (a) physical [PITH_FULL_IMAGE:figures/full_fig_p016_13.png]
Figure 14
Figure 14. Figure 14: LCO amplitude, frequency, and amplitude ratio as a function of fluid-to-structural mass ratio for [PITH_FULL_IMAGE:figures/full_fig_p017_14.png]
Figure 15
Figure 15. Figure 15: LCO amplitude, frequency, and amplitude ratio as a function of fluid-to-structural mass ratio for [PITH_FULL_IMAGE:figures/full_fig_p018_15.png]
Figure 16
Figure 16. Figure 16: LCO amplitude with the other mode frozen for the supercritical and subcritical regimes with [PITH_FULL_IMAGE:figures/full_fig_p019_16.png]
Figure 17
Figure 17. Figure 17: Supercritical and subcritical regions as a function of fluid-to-structural mass ratio for various mode 2 [PITH_FULL_IMAGE:figures/full_fig_p020_17.png]

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 1 linked inside Pith

  1. [1]

    Levinski, D

    O. Levinski, D. Conser, C. Mouser, S. Koschel, R. Carrese, M. Candon, , P. Marzocca, An innova- tive high fidelity approach to individual aircraft tracking, AIAA Scitech Forum, AIAA 2020-1647, Orlando, FL (2020).doi:10.2514/6.2020-1647

    doi:10.2514/6.2020-1647 2020

  2. [2]

    Candon, M

    M. Candon, M. Esposito, H. M. Fayek, O. Levinski, S. Koschel, N. Joseph, R. Carrese, P. Marzocca, Advanced multi-input system identification for next generation aircraft loads monitoring using linear regression, neural networks and deep learning, Mechanical Systems and Signal Processing 171 (2022) 108809.doi:10.1016/j.ymssp.2022.108809

    arXiv 2022

  3. [3]

    N. F. Giannelis, G. A. Vio, O. Levinski, A review of recent developments in the understanding of transonic shock buffet, Progress in Aerospace Sciences 92 (2017) 39–84.doi:10.1016/j.paerosci. 2017.05.004

    doi:10.1016/j.paerosci 2017

  4. [4]

    D. E. Raveh, Numerical study of an oscillating airfoil in transonic buffeting flows, AIAA Journal 3 (47) (2009) 505–515.doi:10.2514/1.35237

    doi:10.2514/1.35237 2009

  5. [5]

    Hartmann, M

    A. Hartmann, M. Klaas, W. Schr¨ oder, Coupled airfoil heave/pitch oscillations at buffet flow, AIAA Journal 51 (7) (2013) 1542–1552.doi:10.2514/1.J051512

    doi:10.2514/1.j051512 2013

  6. [6]

    D. E. Raveh, E. H. Dowell, Aeroelastic responses of elastically suspended airfoil systems in transonic buffeting flow, AIAA Journal 52 (5) (2014) 926–934.doi:10.2514/1.J052185

    doi:10.2514/1.j052185 2014

  7. [7]

    J. Quan, C. Zhang, W. andGao, Z. Ye, Characteristic analysis of lock-in for an elastically suspended airfoil in transonic buffet flow, Chinese Journal of Aeronautics 29 (1) (2015) 129–143.doi:10. 1016/j.cja.2015.12.002

    2015

  8. [8]

    N. F. Giannelis, G. A. Vio, G. Dimitriadis, Dynamic interactions of a supercritical aerofoil in the presence of transonic shock buffet, Proceedings of the 27th International Conference on Noise and Vibration Engineering, Leuven, Belgium (2016)

    2016

Show all 43 references
  1. [9]

    C. Gao, W. Zhang, X. Li, Y. Liu, J. Quan, Z. Ye, Y. Jiang, Mechanism of frequency lock-in in transonic buffeting flow, Journal of Fluid Mechanics 818 (2017) 528–561.doi:10.1017/jfm.2017. 120

    2017 doi

  2. [10]

    Candon, P

    M. Candon, P. Marzocca, E. H. Dowell, Aeroelastic reduced-order model differential equations in transonic buffeting flow, AIAA Journal (In Press, Preprint Available) (2025).doi:10.48550/ arXiv.2510.22216

    2025 doi

  3. [11]

    Th´ emiot, V

    C. Th´ emiot, V. Brion, A. Lepage, Wind tunnel experiment on a pitch and plunge free airfoil under transonic buffet, 57th 3AF International Conference on Applied Aerodynamics, Bordeaux, France (2023)

    2023

  4. [12]

    Korth¨ auer, A

    T. Korth¨ auer, A. Accorinti, S. Scharnowski, C. J. K¨ ahler, Experimental investigation of transonic buffeting, frequency lock-in and their dependence on structural characteristics, Journal of Fluids and Structures 122 (2023) 103975.doi:10.1016/j.jfluidstructs.2023.103975

    2023

  5. [13]

    Candon, V

    M. Candon, V. Muscarello, W. Verhagen, P. Marzocca, O. Levinski, Aeroelastic response of an airfoil with structural freeplay in transonic buffeting flow, Journal of Fluids and Structures 140 (2025) 104445.doi:10.1016/j.jfluidstructs.2025.104445

    2025

  6. [14]

    Iovnovich, D

    M. Iovnovich, D. E. Raveh, D. Michaels, M. Adar, Computational study of transonic limit cycle oscillation phenomenon on F-16 fighter aircraft, Journal of Aircraft 54 (2) (2017) 783–793.doi: 10.2514/1.C033918

    2017 doi

  7. [15]

    R. T. Hartlen, I. G. Currie, Lift-oscillator model of vortex-induced vibration, Journal of the Engi- neering Mechanics Division 96 (5) (1970) 577–591.doi:10.1061/JMCEA3.0001276. 21

    1970 doi

  8. [16]

    W. A. Silva, Discrete-time linear and nonlinear aerodynamic impulse responses for efficient cfd analyses, Ph.D. thesis, Ph.D. Thesis, College of William & Mary, Williamsburg, V A (December 1997).doi:10.21220/s2-cw4r-jc50

    1997 doi

  9. [17]

    W. A. Silva, Identification of nonlinear aeroelastic systems based on the volterra theory: Progress and opportunities, Journal of Nonlinear Dynamics 39 (1-2) (January 2005).doi:10.1007/ s11071-005-1907-z

    2005

  10. [18]

    K. C. Hall, J. P. Thomas, E. H. Dowell, Proper orthogonal decomposition technique for transonic unsteady aerodynamic flows, AIAA Journal 38 (10) (2000) 1853–1862.doi:10.2514/2.867

    2000 doi

  11. [19]

    Fonzi, S

    N. Fonzi, S. L. Brunton, U. Fasel, Data-driven modeling for transonic aeroelastic analysis, Journal of Aircraft 61 (2) (2024) 625–637.doi:10.2514/1.C037409

    2024 doi

  12. [20]

    J. P. Thomas, E. H. Dowell, K. Hall, Modeling viscous transonic limit-cycle oscillation behavior using a harmonic balance approach, Journal of Aircraft 41 (6) (2004) 1266–1274.doi:10.2514/ 1.9839

    2004

  13. [21]

    Carlberg, C

    K. Carlberg, C. Farhat, J. Cortial, D. Amsallem, The gnat method for nonlinear model reduction: Effective implementation and application to computational fluid dynamics and turbulent flows, Journal of Computational Physics 242 (2013) 623–647.doi:10.1016/j.jcp.2013.02.028

    2013 doi

  14. [22]

    Sansica, J

    A. Sansica, J. C. Loiseau, M. Kanamori, A. Hashimoto, J.-C. Robinet, System identification of two- dimensional transonic buffet, AIAA Journal 60 (5) (2022) 3090–3106.doi:10.2514/1.J061001

    2022 doi

  15. [23]

    Q. Ma, C. Gao, D. Yang, W. Zhang, Unsteady aerodynamics identification of transonic buffet by incorporating shock position, Mechanical Systems and Signal Processing 236 (2025) 112995. doi:10.1016/j.ymssp.2025.112995

    2025

  16. [24]

    R. A. Skop, O. M. Griffin, A model for the vortex-excited resonant response of bluff cylinders, Journalof Sound and Vibration 27 (2) (1973) 225–233.doi:10.1016/0022-460X(73)90063-1

    1973 doi

  17. [25]

    E. H. Dowell, Non-linear oscillator models in bluff body aero-elasticity, Journal of Sound and Vibration 75 (2) (1981) 251–264.doi:10.1016/0022-460X(81)90343-6

    1981 doi

  18. [26]

    A. M. Marra, C. Mannini, G. Bartoli, Van der pol-type equation for modeling vortex-induced oscillations of bridge decks, Journal of Wind Engineering and Industrial Aerodynamics 99 (6-7) (2011) 776–785.doi:10.1016/j.jweia.2011.03.014

    2011 doi

  19. [27]

    Hollenbach, R

    R. Hollenbach, R. Kielb, K. Hall, Extending a van der pol-based reduced-order model for fluid–structure interaction applied to non-synchronous vibrations in turbomachinery, Journal of Turbomachinery 144 (3) (2021) 031006.doi:10.1115/1.4052405

    2021 doi

  20. [28]

    Balajewicz, F

    M. Balajewicz, F. Nitzsche, D. Feszty, Application of multi-input volterra theory to nonlinear multi-degree-of-freedom aerodynamic systems, AIAA Journal 48 (1) (2010) 56–62.doi:10.2514/ 1.38964

    2010

  21. [29]

    de Paula, F

    N. de Paula, F. Marquez, Multi-variable volterra kernels identification using time-delay neural networks: Application to unsteady aerodynamic loading, Nonlinear Dynamics 97 (2019) 767–780. doi:10.1007/s11071-019-05011-8

    2019 doi

  22. [30]

    Candon, E

    M. Candon, E. Hale, M. Balajewicz, A. Delgado-Gutierez, P. Marzocca, Parameterization of nonlin- ear aeroelastic reduced order models via direct interpolation of taylor partial derivatives, Nonlinear Dynamics (2024) 1–22doi:10.1007/s11071-024-09976-z

    2024 doi

  23. [31]

    Candon, E

    M. Candon, E. Hale, A. Delgado-Guti´ errez, P. Marzocca, M. Balajewicz, Optimal sparsity in nonlinear nonparametric reduced order models for transonic aeroelastic systems, AIAA Journal 62 (10) (2024) 3841–3856.doi:10.2514/1.J064047. 22

    2024 doi

  24. [32]

    Candon, A

    M. Candon, A. Delgado-Gutierrez, P. Marzocca, M. Balajewicz, E. H. Dowell, Nonlinear aeroe- lastic reduced order modeling with optimized sparse multi-input volterra kernels, AIAA Journal (Available Online, 2025).doi:10.2514/1.J064780

    2025 doi

  25. [33]

    Candon, P

    M. Candon, P. Marzocca, E. H. Dowell, Sparse identification of aeroelastic reduced-order models in transonic buffeting flow, AIAA SciTech Forum, AIAA 2026-0812 (Orlando, FL 2026).doi: 10.2514/6.2026-0812

    2026 doi

  26. [34]

    S. L. Brunton, J. L. Proctor, N. Kutz, Discovering governing equations from data by sparse identi- fication of nonlinear dynamical systems, Proceedings of the National Academy of Sciences 113 (15) (2016) 3932–3937.doi:10.1073/pnas.1517384113

    2016 doi

  27. [35]

    R. T. Q. Chen, Y. Rubanova, J. Bettencourt, D. K. Duvenaud, Neural ordinary differential equa- tions, in: Advances in Neural Information Processing Systems, Vol. 31, 2018

    2018

  28. [36]

    Rackauckas, Y

    C. Rackauckas, Y. Ma, J. Martensen, C. Warner, K. Zubov, R. Supekar, D. Skinner, A. Ramad- han, A. Edelman, Universal differential equations for scientific machine learning, arXiv preprint arXiv:2001.04385 (2020).doi:10.48550/arXiv.2001.04385

    2001 doi arXiv

  29. [37]

    Kidger, J

    P. Kidger, J. Morrill, J. Foster, T. Lyons, Neural controlled differential equations for irregular time series, in: Advances in Neural Information Processing Systems, Vol. 33, 2020, pp. 6696–6707

    2020

  30. [38]

    Balajewicz, E

    M. Balajewicz, E. H. Dowell, Reduced-order modeling of flutter and limit-cycle oscillations using the sparse volterra series, Journal of Aircraft 49 (6) (2012) 1803–1802.doi:10.2514/1.C031637

    2012 doi

  31. [39]

    Parkinson, Phenomena and modelling of flow-induced vibrations of bluff bodies, Progress in Aerospace Sciences 26 (2) (1989) 169–224.doi:10.1016/0376-0421(89)90008-0

    G. Parkinson, Phenomena and modelling of flow-induced vibrations of bluff bodies, Progress in Aerospace Sciences 26 (2) (1989) 169–224.doi:10.1016/0376-0421(89)90008-0

    1989 doi

  32. [40]

    Jacquin, P

    L. Jacquin, P. Molton, S. Deck, B. Maury, D. Soulevant, Experimental study of shock oscillation over a transonic supercritical profile, AIAA Journal 92 (9) (2009) 1985–1994.doi:10.2514/1. 30190

    2009 doi

  33. [41]

    Workbench Users Guide Release 2024 R1, Ansys Academic Research, 2024

    2024

  34. [42]

    P. R. Spalart, M. Shur, On the sensitization of turbulence models to rotation and curvature, Aerospace Science and Technology 1 (5) (1997) 297–302.doi:10.1016/S1270-9638(97)90051-1

    1997 doi

  35. [43]

    Carrese, P

    R. Carrese, P. Marzocca, O. Levinski, N. Joseph, Initial investigation of supercritical airfoil dynamic response due to transonic buffet, 54 th AIAA Aerospace Sciences Meeting (AIAA-2016-1552, San Diego, CA, January 2016). 23

    2016

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.