Pith. sign in

REVIEW 3 major objections 5 minor 32 references

Scaling law for the slow flow of an unstable mechanical system coupled to a nonlinear energy sink

T0 review · 3 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

Pith's one-line read Near the left fold point of the critical manifold, the slow flow of a mechanically unstable system with a nonlinear energy sink reduces to the dynamic saddle-node normal form, giving a scaling law with exponents 1/3 and 2/3 and a corrected

desk verdict Real analytical result with no fitted parameters: the ε^{2/3} correction to the mitigation limit is credible, but the center-manifold tangent-space approximation leaves the quantitative constants unproved. read the letter →

arxiv 2512.14943 v1 pith:5CD7FJVL submitted 2025-12-16 nlin.CD physics.app-phphysics.flu-dyn

classification nlin.CDphysics.app-phphysics.flu-dyn MSC 34E1534C2370K50
keywords nonlinearenergysinkslowflowcriticalmanifoldfoldpointdynamicsaddle-nodebifurcationAiryfunctionscalinglawaeroelasticwing
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 claims that when a primary structure with one unstable mode is coupled to a nonlinear energy sink, the slow dynamics near a fold point of its critical manifold can be reduced to the normal form of a dynamic saddle-node bifurcation. Solving this normal form yields an exact scaling law: the distance between the critical manifold and the actual trajectory scales with the mass-ratio parameter epsilon to the 1/3 and 2/3 powers, not with the first power. This law predicts where the trajectory leaves the manifold (the jump point) and where it lands (the arrival point), and from it the paper derives a corrected NES mitigation limit that depends on epsilon^{2/3}. If true, this replaces the zeroth-order prediction, which ignores epsilon entirely and is inaccurate even at small epsilon.

What carries the argument

The central object is the reduction of the slow flow to the normal form of the dynamic saddle-node bifurcation. Near a fold point, the center manifold theorem is used to eliminate one fast variable via a tangent-space approximation, leaving a single equation with a slowly varying bifurcation parameter. The exact solution of this normal form is expressed through Airy functions and their zeros, which determine the epsilon^{2/3} location of the departure and arrival points and the epsilon^{2/3} correction to the mitigation limit.

What would settle it

Compute the first jump ordinate r_jump from direct numerical integration of the slow flow (38) for several values of epsilon, e.g., 0.001, 0.005, 0.02, and 0.1, at fixed mu and alpha, and plot r_jump - r_LF against epsilon^{2/3}. If the points do not fall on a straight line through the origin with slope 2.33810 a1 (f_LF a2)^{2/3} / a2, or if the intercept is not r_LF, the claimed scaling law fails.

Watch

Extended reading notes

Core claim

The central claim is that, in a neighborhood of the left fold point of the critical manifold, the slow flow of an unstable mechanical system coupled to an NES is described by the normal form epsilon q' = q^2 + v, where v is a slowly varying parameter. This dynamic saddle-node normal form has an exact solution in terms of Airy functions: q(v) = epsilon^{1/3} Ai'(-epsilon^{-2/3} v)/Ai(-epsilon^{-2/3} v). Consequently, the jump point and arrival point of a relaxation cycle are set by the first zeros of the Airy function derivative and the Airy function respectively, producing the fractional exponents 1/3 and 2/3. The paper then uses this scaling law to correct the mitigation limit, replacing th

Load-bearing premise

The quantitative constants in the scaling law and mitigation-limit correction rely on three unchecked steps: replacing the center manifold by its tangent plane, keeping only the Airy branch that satisfies the initial condition at negative infinity, and assuming the limit value r_infinity has already been reached when the trajectory lands on the right attracting branch.

Editorial extensions

If this is right

  • The finite-epsilon jump and arrival points of the slow flow are determined by Airy zeros, so they can be computed once the parameters a1, a2, and f_LF are known, without simulating the full system.
  • The NES mitigation limit depends on the mass-ratio parameter epsilon through the 2/3 power, meaning even small epsilon produces corrections visible at order epsilon^{2/3} rather than order epsilon.
  • The optimal NES damping coefficient shifts by a term proportional to epsilon^{2/3}, so NES design can explicitly account for the finite mass of the absorber.
  • The method applies to a full aeroelastic wing model, where the corrected mitigation limit matches numerical simulations of the full order system better than the zeroth-order prediction for a range of epsilon values.
  • The zeroth-order mitigation limit is shown to be insufficient: it deviates from numerical results even at epsilon = 0.001, while the new prediction remains accurate up to epsilon = 0.02 and qualitatively acceptable at epsilon = 0.1.

Reading between the lines

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

  • The same normal-form reduction should apply at the right fold point, yielding an analogous scaling law with Airy-function constants; this is not developed in the paper but follows directly from the symmetry of the fold geometry.
  • The exact Airy solution of the normal form is not an asymptotic approximation, so the scaling law may remain valid beyond the perturbative regime, provided the tangent-space and single-branch assumptions hold.
  • A testable extension is to measure the actual jump ordinates from direct numerical simulations of the slow flow for several epsilon values and fit K_infinity; a consistent epsilon^{2/3} slope would confirm the law and calibrate the correction term.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper analyzes a mechanical system with one unstable mode coupled to a cubic nonlinear energy sink (NES). After modal reduction and complexification-averaging, the slow flow is a (2,1)-fast-slow system whose critical manifold has two fold points. The zeroth-order analysis recovers earlier predictions of the mitigation limit. The claimed novelty is a center-manifold reduction near the left fold to the normal form of a dynamic saddle-node bifurcation, an exact Airy-function solution of that normal form, and the resulting scaling law with 1/3 and 2/3 fractional powers of the mass-ratio parameter ε. From this scaling law the paper derives corrected jump/arrival points and an ε^{2/3} correction to the mitigation limit and to the optimal NES damping. The method is tested against direct numerical simulations of both the slow flow and the full aeroelastic wing model.

Significance. If established, the result is significant: it provides an explicit analytic description of the fold escape in a mechanical NES system, going beyond the usual zeroth-order analysis, and it makes falsifiable predictions involving Airy zeros and ε^{1/3}, ε^{2/3} scalings. A notable strength is that no parameter is fitted to the numerical results: the constants a1, a2, f_LF are evaluated from physical inputs, and 1.01879 and 2.33810 are tabulated Airy zeros. The numerical comparisons in Fig. 6 give nontrivial evidence that the predicted mitigation limit tracks the simulations, including for moderately large ε. The work is therefore a useful contribution to the NES literature, provided the three explicit approximations in Section 4 are properly justified or their error is quantified.

major comments (3)
  1. [§4.1, Eqs. (46)–(47)] The center manifold reduction is invoked, but then the manifold is replaced by its tangent, q_b = ℓ(q_a)=0, 'for sake of simplicity'. This truncation sets the coefficients a1 and a2, which enter the normal form (48) and hence the quantitative constants in Eqs. (55)–(57) and in the mitigation-limit correction (64). The manuscript gives no estimate of the error made by omitting the curvature of ℓ, nor a proof that ℓ contributes only at higher order than the retained q_a^2 and u terms. Since the central quantitative claims depend on a1 and a2, the authors should either justify the truncation by an explicit normal-form calculation or provide a numerical convergence study showing that the omitted terms do not change the predicted constants at the claimed accuracy.
  2. [§4.2, Eq. (52)] The text states: 'Assuming that x(−∞) = −√(−y), only the contribution of Ai(s) is kept (this is not proved here).' This is an explicit admission of an omitted proof at a load-bearing point: the choice of Ai instead of Bi determines the numerical constants in Eqs. (55)–(57) and in Eq. (64). The branch selection is in fact standard—Ai(s) is selected by matching to the attracting branch because Ai'/Ai → −√s as s→∞, whereas Bi'/Bi → +√s—but the paper should contain that argument rather than leaving the choice unproved.
  3. [§4.3, Eq. (58)] The new mitigation limit assumes that at the arrival point on the right attracting branch the limit value r∞ has already been reached. This is an uncontrolled approximation: the finite-ε jump has a nonzero duration during which r changes, and no error estimate is given. The numerical agreement in Fig. 5 is encouraging, but Eq. (58) is the basis of the central quantitative prediction (64), so the authors should state the expected order of the error and, ideally, verify it numerically over a range of ε and μ rather than for one parameter set.
minor comments (5)
  1. [Fig. 4 caption] The caption states 'ξ_h = 4' in one place while Eq. (71) fixes ξ_h = 5 and the text says 'ξ_x = 4 and ξ_φ = 8'. This inconsistency should be corrected.
  2. [Fig. 6 caption] The panel labels are garbled: '((a)ε=0.001, (b),ε=0.005, ε=0.02 and (d)ε=0.1' should read '(a) ε=0.001, (b) ε=0.005, (c) ε=0.02, (d) ε=0.1'.
  3. [§4.1, text near Eq. (46)] The assumption f(r,s,Δ)=f_LF is introduced without comment. It is probably legitimate at leading order near the fold, but it should be stated explicitly as part of the truncation so that the reader can track all neglected terms.
  4. [§4.3, first paragraph] Typo: 'right attracting par of M0' should be 'right attracting part of M0'.
  5. [Eq. (40)] The sentence 'using Eq. (40b)' should probably refer to both (40a) and (40b); as written it is slightly confusing.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the 1/3-2/3 scaling law and Airy constants are derived from the slow flow via the standard dynamic saddle-node normal form, with no fitted parameters; self-citations are background.

full rationale

The derivation chain is self-contained and does not reduce any prediction to an input. Section 3 derives the slow flow (Eqs. 20-23), the critical manifold (Eq. 28), and the zeroth-order mitigation limit (Eqs. 35-36) analytically from the model parameters a, b, µ, α. Section 4 linearizes the fast subsystem at the left fold (Eqs. 39-41), applies the center manifold theorem to obtain the normal form (Eq. 48), solves it through the Airy equation (Eqs. 50-52) with tabulated Airy zeros, and thereby obtains the scaling law (Eq. 55) and the ε^{2/3} mitigation-limit correction (Eq. 64). No parameter is fitted to the numerical results shown in Section 5: the constants a1, a2, and f_LF come from the physical coefficients, the fractional exponents 1/3 and 2/3 come from the normal-form rescaling, and the constants 1.01879 and 2.33810 are tabulated Airy values. The numerical comparison is an external check against the full-order wing model (67) and the slow flow (38), not a tuning loop. The self-citations [22,23] are used for background classification and for the zeroth-order regime description; the new scaling law is justified in the paper text and by external dynamical-systems results [28,30,31], so those citations are not load-bearing. The acknowledged approximations - the tangent-space center manifold q_b = ℓ(q_a) = 0 in Sec. 4.1, the Airy-branch selection 'this is not proved here' in Sec. 4.2, and the r∞-arrival assumption in Sec. 4.3 - can affect the quantitative constants and are genuine correctness/robustness risks, but they are explicit modeling assumptions, not circular definitions or fitted inputs. No step exhibits the required reduction of a claimed output to an input by construction.

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

The derivation has no fitted free parameters; all coefficients are system data. Its burden is instead the three ad hoc approximations (tangent center manifold, Ai-branch choice, r∞ jump assumption) plus the usual averaging and modal-truncation assumptions. These are stated in the text but not justified to the level of error bounds.

assumptions (5)
  • domain assumption Applicability of the center manifold theorem to reduce the fast-slow system (24) near the left fold point, where the fast Jacobian has eigenvalues 0 and -µ.
    Used in Section 4.1 to pass from (38b)-(38c) to the reduced normal form (48); requires the eigenvalue structure computed in Eq. (39) and the fold condition µ<1/√3.
  • ad hoc to paper The tangent-space approximation q_b = ℓ(q_a) = 0 is sufficient for the quantitative coefficients a1, a2 of the normal form.
    Section 4.1: 'For sake of simplicity one chooses the tangent space approximation'; the actual center manifold is not computed, so the curvature correction to a1,a2 is unknown.
  • ad hoc to paper The solution branch is selected by assuming q_a(-∞) → -√(-v); hence only the Ai contribution is retained.
    Section 4.2: 'Assuming that x(−∞) = −√−y, only the contribution of Ai(s) is kept (this is not proved here)'; this fixes which Airy function solves the boundary-layer matching.
  • domain assumption Stable modal coordinates q_n (n≥2) of the primary structure are negligible; their coupling is O(ε).
    Section 2.2: based on the diagonalized system and previous works [12,23,24]; neglects nonlinear modal interactions and internal resonances of the primary structure.
  • ad hoc to paper When the trajectory reaches the right attracting branch of M0, the limit value r∞ of the local scaling law has already been reached.
    Section 4.3: 'one can assume that when the trajectory reaches the right attracting part of M0, the limit value r∞ has already been reached'; used to construct Eq. (58) for the mitigation limit.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Scaling law for the slow flow of an unstable mechanical system coupled to a nonlinear energy sink." pith.science (2026). https://pith.science/paper/5CD7FJVL

@misc{pith2026251214943,
  author       = {Pith},
  title        = {Pith review of: Scaling law for the slow flow of an unstable mechanical system coupled to a nonlinear energy sink},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5CD7FJVL}},
  note         = {Machine review of arXiv:2512.14943}
}
read the original abstract

In this paper one first shows that the slow flow of a mechanical system with one unstable mode coupled to a Nonlinear Energy Sink (NES) can be reduced, in the neighborhood of a fold point of its critical manifold, to a normal form of the dynamic saddle-node bifurcation. This allows us to then obtain a scaling law for the slow flow dynamics and to improve the accuracy of the theoretical prediction of the mitigation limit of the NES previously obtained as part of a zeroth-order approximation. For that purpose, the governing equations of the coupled system are first simplified using a reduced-order model for the primary structure by keeping only its unstable modal coordinates. The slow flow is then derived by means of the complexification-averaging method and, by the presence of a small perturbation parameter related to the mass ratio between the NES and the primary structure, it appears as a fast-slow system. The center manifold theorem is finally used to obtain the reduced form of the slow flow which is solved analytically leading to the scaling law. The latter reveals a nontrivial dependence with respect to the small perturbation parameter of the slow flow dynamics near the fold point, involving the fractional exponents 1/3 and 2/3. Finally, a new theoretical prediction of the mitigation limit is deduced from the scaling law. In the end, the proposed methodology is exemplified and validated numerically using an aeroelastic aircraft wing model coupled to one NES.

Figures

Figures reproduced from arXiv: 2512.14943 by the authors.

Figure 1
Figure 1. Typical example of the critical manifold in the (s, r)-plane given by Eq. (28a) for µ = 0.25 and α = 5. manifold M0. The two scalars s D and s U, which are the horizontal projection of the fold points on the critical manifold, are defined by H  s RF = H  s D  and H  s LF = H  s U  giving s D = 2 √ 2 3 √ α r 1 − q 1 − 3µ2 and s U = 2 √ 2 3 √ α r 1 + q 1 − 3µ2. (31) 3.3 Fixed points and fold singularities of t… view at source ↗
Figure 2
Figure 2. Illustration of the normal form of the dynamic saddle-node bifurcation. Result of the numerical integration of Eq. (48) with initial condition (qa(0) = −1, v(0) = −0.5) (red dashed line) compared to the analytical scaling law q ⋆ a (y) given by (52) (blue line, the dashed parts are the horizontal asymptotes of q ⋆ a (y) corresponding to the zeros Airy function) for ϵ = 0.01. The first zero and the first singularity … view at source ↗
Figure 3
Figure 3. Sketch of the two DOFs aircraft wing coupled to one NES. z and ˜φ are respectively the heave and the angle of attack (pitch) of the wing and y is the displacement of the NES. A is the aerodynamic center, B the elastic axis, G the center of gravity of the aircraft wing. e is the location aerodynamic center A measured from B (positive ahead of B). Kz and Kφ are the linear heave and pitch stiffnesses respectively where… view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Direct numerical integration of the wing-NES system (67) (in blue) depicting the several response regimes described in Section 3.4 with, in addition to (70) and (71), ϵ = 0.01, ζh = 0.3, ξh = 4, ξx = 4 and ξφ = 8. Four values of the reduced speed are used, namely: (a) …
Figure 5
Figure 5. Figure 5: Illustration of the proposed analytical procedure. (a) The critical manifold M0 (27) (black line) superimposed to the direct numerical integration of the slow flow (38) (red line) and its scaling law near the left fold point given by Eq. (55) (blue line, the dashed par…
Figure 6
Figure 6. Figure 6: Theoretical mitigation limits: the one obtained within the zeroth-order approximation (36) (blue line) and those derived from the scaling law (solving numerically Eq. (58) (green line) and the one given by Eq. (64) (orange line)) compared with mitigation limits measure…
Figure 7
Figure 7. Figure 7: The physical bifurcation parameter Θ with respect to the generalized bifurcation parameter ρ. As in [PITH_FULL_IMAGE:figures/full_fig_p026_7.png]
Figure 8
Figure 8. Figure 8: The theoretical optimal value of the NES damping coefficient with respect to the small perturba￾tion parameter ϵ. The one obtained from the zeroth-order approximation µ opt 0 given by Eq. (37) (blue line) is compared to those derived from the scaling law (55): the one …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

32 extracted references · 12 canonical work pages

  1. [1]

    Gendelman, L

    O. Gendelman, L. I. Manevitch, a. F. Vakakis, R. M’Closkey, Energy Pumping in Nonlinear Mechanical Oscillators: Part I—Dynamics of the Underlying Hamiltonian Systems, Journal of Applied Mechanics 68 (1) (2001) 34.doi:10.1115/1.1345524

  2. [2]

    Vakakis, O

    A. Vakakis, O. Gendelman, Energy pumping in nonlinear mechanical oscillators: Part II - Resonance capture, Journal of Applied Mechanics 68 (2001) 42–48

  3. [3]

    A. F. Vakakis, O. Gendelman, L. A. Bergman, D. M. McFarland, G. Kerschen, Y. Lee, Nonlinear Targeted Energy Transfer in Mechanical and Structural Systems, no. 156, Springer, 2009.doi:10. 1007/978-1-4020-9130-8

  4. [4]

    Z. Lu, Z. Wang, Y. Zhou, X. Lu, Nonlinear dissipative devices in structural vibration control: A review, Journal of Sound and Vibration 423 (2018) 18–49

  5. [5]

    Y. S. Lee, A. F. Vakakis, L. A. Bergman, D. M. McFarland, Suppression of limit cycle oscillations in the van der Pol oscillator by means of passive non-linear energy sinks, Structural Control and Health Monitoring 13 (1) (2006) 41–75.doi:10.1002/stc.143

  6. [6]

    O. V. Gendelman, T. Bar, Bifurcations of self-excitation regimes in a Van der Pol oscillator with a nonlinear energy sink, Physica D 239 (3-4) (2010) 220–229.doi:10.1016/j.physd.2009.10.020

  7. [7]

    Domany, O

    E. Domany, O. Gendelman, Dynamic responses and mitigation of limit cycle oscillations in van der pol-duffing oscillator with nonlinear energy sink, Journal of Sound and Vibration 332 (2013) 5489–5507

  8. [9]

    Y. S. Lee, A. F. Vakakis, L. A. Bergman, D. M. McFarland, G. Kerschen, Suppression aeroelastic instability using broadband passive targeted energy transfers, part 2: Experiments, AIAA Journal 45 (3) (2007) 2391–2400.doi:10.2514/1.24062

Show all 32 references
  1. [10]

    Y. S. Lee, A. F. Vakakis, D. M. Bergman, L. A. McFarland, G. Kerschen, Enhancing the robustness of aeroelastic instability suppression using multi-degree-of-freedom nonlinear energy sinks, AIAA Journal 46 (6) (2008) 1371–1394

  2. [11]

    Manevitch, Complex representation of dynamics of coupled nonlinear oscillators, in: L

    L. Manevitch, Complex representation of dynamics of coupled nonlinear oscillators, in: L. Uvarova, A. Arinstein, A. Latyshev (Eds.), Mathematical Models of Non-Linear Excitations, Transfer, Dynamics, and Control in Condensed Systems and Other Media, Springer US, 1999, pp. 269–...

  3. [12]

    Gendelman, A

    O. Gendelman, A. Vakakis, L. Bergman, D. McFarland, Asymptotic analysis of passive nonlinear sup- pression of aeroelastic instabilities of a rigid wing in subsonic flow, SIAM Journal on Applied Mathe- matics 70 (5) (2010) 1655–1677.doi:10.1137/090754819. 27 Scaling law for the...

  4. [13]

    Luongo, D

    A. Luongo, D. Zulli, Aeroelastic instability analysis of nes-controlled systems via a mixed multiple scale/harmonic balance method, Journal of Vibration and Control 20 (13) (2014) 1985–1998.doi: 10.1177/1077546313480542

  5. [14]

    Vaurigaud, L

    B. Vaurigaud, L. Manevitch, C.-H. Lamarque, Passive control of aeroelastic instability in a long span bridge model prone to couple dflutter using targeted energy transfer, Journal of Sound and Vibration 330 (2011) 2580–2595

  6. [15]

    R. K. R. Tumkur, E. Domany, O. V. Gendelman, A. Masud, L. A. Bergman, A. F. Vakakis, Reduced- order model for laminar vortex-induced vibration of a rigid circular cylinder with an internal nonlinear absorber, Communications in Nonlinear Science and Numerical Simulation 18 (7) ...

  7. [16]

    H. L. Dai, A. Abdelkefi, L. Wang, Vortex-induced vibrations mitigation through a nonlinear energy sink, Communications in Nonlinear Science and Numerical Simulation 42 (2017) 22–36.doi:10.1016/ j.cnsns.2016.05.014

  8. [17]

    Bergeot, S

    B. Bergeot, S. Bellizzi, B. Cochelin, Analysis of steady-state response regimes of a helicopter ground resonance model including a non-linear energy sink attachment, International Journal of Non-Linear Mechanics 78 (2016) 72 – 89.doi:http://dx.doi.org/10.1016/j.ijnonlinmec.2015.10.006

  9. [18]

    Bergeot, S

    B. Bergeot, S. Bellizzi, B. Cochelin, Passive suppression of helicopter ground resonance using nonlinear energy sinks attached on the helicopter blades, Journal of Sound and Vibration 392 (2017) 41–55. doi:10.1016/j.jsv.2016.12.039

  10. [19]

    Gourc, S

    E. Gourc, S. Seguy, G. Michon, A. Berlioz, Chatter Control in Turning Process with a Nonlinear Energy Sink, Advanced Materials Research 698 (2013) 89–98.doi:10.4028/www.scientific.net/AMR.698.89

  11. [20]

    Nankali, Y

    A. Nankali, Y. S. Lee, T. Kalmar-Nagy, Targeted energy transfers for suppressing regenerative machine tool vibrations, Journal of Computational and Nonlinear Dynamics Transactions of the ASME 12 (2017) 011010–1 –011010–11

  12. [21]

    Bergeot, S

    B. Bergeot, S. Berger, S. Bellizzi, Mode coupling instability mitigation in friction systems by means of nonlinear energy sinks : numerical highlighting and local stability analysis, Journal of Vibration and Control 24 (15) (2017) 3487–3511.doi:10.1177/1077546317707101

  13. [22]

    Bergeot, S

    B. Bergeot, S. Bellizzi, Asymptotic analysis of passive mitigation of dynamic instability using a nonlinear energy sink network, Nonlinear Dynamics 94 (2) (2018) 1501–1522.doi:10.1007/s11071-018-4438-0. URLhttp://link.springer.com/10.1007/s11071-018-4438-0

  14. [23]

    Bergeot, S

    B. Bergeot, S. Bellizzi, Steady-state regimes prediction of a multi-degree-of-freedom unstable dynamical system coupled to a set of nonlinear energy sinks, Mechanical Systems and Signal Processing 131 (2019) 728–750

  15. [24]

    Bergeot, S

    B. Bergeot, S. Bellizzi, S. Berger, Dynamic behavior analysis of a mechanical system with two unstable modes coupled to a single nonlinear energy sink, Communications in Nonlinear Science and Numerical Simulation 95 (2021) 105623.doi:10.1016/j.cnsns.2020.105623

  16. [25]

    Nayfeh, Perturbation Methods, Physics textbook, Wiley, 2008

    A. Nayfeh, Perturbation Methods, Physics textbook, Wiley, 2008. URLhttps://books.google.fr/books?id=eh6RmWZ51NIC

  17. [26]

    Jones, Geometric singular perturbation theory, in: R

    C. Jones, Geometric singular perturbation theory, in: R. Johnson (Ed.), Dynamical Systems, Vol. 1609 of Lecture Notes in Mathematics, Springer Berlin Heidelberg, 1995, pp. 44–118.doi:10.1007/BFb0095239

  18. [27]

    Fenichel, Geometric singular perturbation theory for ordinary differential equations, Journal of Dif- ferential Equations 98 (1979) 53–98

    N. Fenichel, Geometric singular perturbation theory for ordinary differential equations, Journal of Dif- ferential Equations 98 (1979) 53–98

  19. [28]

    Christian, Multiple Time Scale Dynamics, 1st Edition, Vol

    K. Christian, Multiple Time Scale Dynamics, 1st Edition, Vol. 191 of Applied Mathematical Sciences, Springer International Publishing, 2015. 28 B. Bergeot

  20. [29]

    Benoˆ ıt, J

    E. Benoˆ ıt, J. Callot, F. Diener, M. Diener, Chasse au canard (”duck hunting”), Collectanea Mathematica 32 (1-2) (1981) 37–119

  21. [30]

    Guckenheimer, P

    J. Guckenheimer, P. Holmes, Nonlinear Oscillations, Dynamical Systems, and Bifurcations of Vec- tor Fields, Applied Mathematical Sciences, Springer New York, 1983, Ch. 3.doi:10.1007/ 978-1-4612-1140-2. URLhttp://dx.doi.org/10.1007/978-1-4612-1140-2

  22. [31]

    Berlund, B

    N. Berlund, B. Gentz, Noise-Induced Phenomena in Slow-Fast Dynamical Systems, Probability and Its Applications, Springer-Verlag, 2006.doi:10.1007/1-84628-186-5. URLhttp://dx.doi.org/10.1007/1-84628-186-5

  23. [32]

    Wiggins, Introduction to Applied Nonlinear Dynamical Systems and Chaos, 2nd Edition, Texts in Applied Mathematics 2, Springer New York, 1990

    S. Wiggins, Introduction to Applied Nonlinear Dynamical Systems and Chaos, 2nd Edition, Texts in Applied Mathematics 2, Springer New York, 1990. URLhttp://gen.lib.rus.ec/book/index.php?md5=FC5B75B65E2F01A76E888A8082C8D63C

  24. [33]

    Abramowitz, I

    M. Abramowitz, I. Stegun, Handbook of Mathematical Functions with Formulas, Graphs, and Mathe- matical Tables, 9th Edition, Applied mathematics series, Wiley, 1972, Ch. Airy Functions. 29

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.