A comparison of three models of a whirling string as it becomes chaotic
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Nonlinear Sciences. The day in nonlinear science: chaos, patterns, networks and adaptation.
Kai: We're looking at the paper: "A comparison of three models of a whirling string as it becomes chaotic".
Mira: Here's the gist of the paper. A comparison of three models of a whirling string as it becomes chaotic investigates how different mathematical approximations affect the description of nonlinear string motion as it approaches chaos The geometrically exact, two-mode Kirchhoff–Carrier, and averaged equations for a whirling string are integrated under similar conditions to compare their behavior along the route to chaos from a Hopf bifurcation through period doublings to chaotic attractors organized by a Shil’nikov homoclinic orbit Model Comparison and Agreement The three models show the same route to chaos, and that the averaged equations shift local bifurcations by an amount of order ε, consistent with averaging theory for the Hopf bifurcation and period doublings The geometrically exact and two-mode models agree with each other to within 0.002 at the local bifurcations for ε ≤ 0.01 The averaged equations place each local bifurcation 0.7ε to 1.2ε higher in the scaled detuning Bifurcation Shifts and Sensitivity The averaging places each local bifurcation 0.7ε to 1.2ε higher in the scaled detuning The geometric exact and two-mode models agree with each other to within 0.002 at the local bifurcations for ε ≤ 0.01 Global bifurcations can be shifted by several times ε, and the size of this error is set by how steeply the splitting function of the homoclinic loop crosses zero The homoclinic bifurcation is the most sensitive to modeling approximations Global Bifurcation Analysis The primary homoclinic orbit is found by following the unstable manifold of the upper planar saddle focus The splitting function β(∆) marks the homoclinic orbit where it crosses zero The Shil’nikov ratio, which decides whether chaos accompanies the homoclinic orbit, agrees among the three models to within 1.5% for ε ≤ 0.01 The averaged equations agree with the two string models in the sequence of bifurcations on the route to chaos and in the Shil’nikov ratio of the saddle focus that organizes it For local bifurcations, the coefficient of ε is of order one At ε = 0.02, the full string’s shifts are 70 to 80% of the two-mode string’s, and the difference is mostly one of amplitude, not of ε Model Limitations and Future Directions The study offers an example of what can be expected when a reduced model is used near a route to chaos The natural next step is a theory of the O(ε) shifts, such as second-order averaging or spectral submanifolds The comparison could be made for other distributed systems whose motion is confined to a few modes Data Availability and Acknowledgements The run data that support the findings of this study (the brute-force scans, the growth-rate runs and the homoclinic-orbit runs for the three models) and the MATLAB code for the three models, the scans and the analysis are deposited at Zenodo The author used Claude (Anthropic) to document and test computer code, structure simulation sequences, and check text of paper The research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors The author declares that he has no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper References
1: R.
Kai: J.
Mira: Hanson, J.
Kai: M.
Paper summary: Kai: So we've talked about the setup for this paper, which is comparing three models of a whirling string as it becomes chaotic. To recap, the core idea is testing how different mathematical approximations affect our description of nonlinear string motion as it approaches chaos.
Mira: They are comparing the geometrically exact string equations with a two-mode model called Kirchhoff–Carrier and then the averaged equations for the slow amplitudes of those two modes. The thesis is that they integrate all three under similar conditions along this route to chaos, which starts with a Hopf bifurcation of the steady whirl and passes through period doublings to chaotic attractors organized by a Shil’nikov homoclinic orbit.
Lev: So it's not just about solving the equations; it's about how these different mathematical tools—exact, reduced, and averaged—handle the transition from simple whirling motion into something truly chaotic.
Kai: Right. The paper claims that qualitatively, all three models agree on the sequence of bifurcations, what kinds of attractors appear in those transitions, and even their return maps and Lyapunov exponents. But they also have to be specific about how far apart these models drift quantitatively as the driving frequency changes.
Mira: They found that averaging places each local bifurcation zero point seven epsilon to one point two epsilon higher in the scaled detuning, which they say is consistent with averaging theory for the Hopf bifurcation and period doublings <ref:2611.06567#pg1,averaging places each local bifurcation 0>.
Lev: And they also noted that when comparing the exact and two-mode models at those local bifurcations, they agree within zero point zero zero two at those points when epsilon is less than or equal to zero point zero one. But those averaged equations are shifted by a factor of order one in terms of epsilon for the local bifurcations.
Kai: So the key finding here is that while the averaged equations give a good general path, they introduce a predictable shift in where you see the local bifurcation compared to what you get from the more detailed two-mode model.
Mira: And globally, they found that global bifurcations can be shifted by several times epsilon, and this error size depends on how steeply the splitting function of the homoclinic loop crosses zero.
Lev: That tells me that for those large, global changes, you need to be much more careful about your modeling choice because the sensitivity is tied to that specific geometric feature of the system.
Kai: So they establish a hierarchy: local bifurcations are shifted by order one in epsilon, while global bifurcations are shifted by several times epsilon, and it all depends on the specific approximation you choose. That sets up a clear warning for using reduced models near these kinds of transitions.
Mira: It gives us confidence that the averaged equations aren't completely off track for the main sequence of events, even if they aren't perfectly matching on every small detail. This is helpful because it confirms the general physics of the transition to chaos is captured by those simplified equations.
Lev: For a real hardware run, knowing that these shifts are proportional to epsilon gives us a way to estimate how much uncertainty we have when we use an averaged model versus trying to run the full string equation.
Conclusion: Kai: So wrapping up this comparison of three models of a whirling string as it becomes chaotic. The paper compares the geometrically exact string equations, the two-mode Kirchhoff–Carrier model, and the averaged equations.
Mira: They confirmed that qualitatively, all three models show the same route to chaos: Hopf bifurcation leading to period doublings and then chaotic attractors organized by a Shil’nikov homoclinic orbit. But quantitatively, they showed how those different approximations introduce predictable shifts in where local bifurcations occur based on epsilon.
Lev: It seems the main implication is that using averaged equations gives you a good general path, but you have to account for those systematic shifts—zero point seven epsilon to one point two epsilon higher in scaled detuning for the local bifurcations, which we see consistently across the models when looking at those steps <ref:2611.06567#pg1>.
Kai: And globally, those shifts are bigger and depend on how steep the splitting function crosses zero, meaning you need to be cautious about how much you trust a reduced model when it comes to large-scale changes in behavior.
Mira: The authors are essentially saying that for understanding the general physics of this transition, these simplified equations are useful tools, but you have to know exactly what assumptions they're making about the dynamics.
Lev: For us building things, it means if we want high fidelity, we need to look at those epsilon terms carefully when comparing our simple models to the full string dynamics.
Kai: And for the broader field, it shows that when you're dealing with these kinds of nonlinear systems near a route to chaos, you can get a very solid general picture from reduced models, as long as you quantify the expected errors based on your chosen approximation level.
Mira: It validates using these specific averaged equations as a useful framework for studying how string dynamics move into chaos, which is an important step in understanding more complex distributed systems confined to just a few modes.
Nicholas B. Tufillaro
Aqualytics
nlin.CD, math.DS, physics.class-ph
Submitted: 2026-11-04
Updated: 2026-11-04
Comments: 38 pages, 7 figures, 4 tables. Run data and MATLAB code: doi:10.5281/zenodo.23931253. Submitted to Journal of Sound and Vibration
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Importance score: 80/100
The gist: A comparison of three models of a whirling string as it becomes chaotic investigates how different mathematical approximations affect the description of nonlinear string motion as it approaches chaos…
Key concepts
- Geometric Exact Model
- This is a highly detailed mathematical description of a whirling string that accounts for all its physical complexities. It serves as a benchmark against simpler models to check how much accuracy is lost when simplifying the equations.
- Averaged Equations
- These are simplified mathematical approximations derived from the more complex models. They are useful because they show the general behavior near certain points, but they introduce small errors, which are quantified by an order ε.
- Homoclinic Orbit
- This is a specific path in the mathematical system where a trajectory starts and returns to its original point after following an unstable path. It is crucial because it often marks the boundary between regular motion and chaotic behavior in these systems.
Terminology
Summary
A comparison of three models of a whirling string as it becomes chaotic investigates how different mathematical approximations affect the description of nonlinear string motion as it approaches chaos
The geometrically exact, two-mode Kirchhoff–Carrier, and averaged equations for a whirling string are integrated under similar conditions to compare their behavior along the route to chaos from a Hopf bifurcation through period doublings to chaotic attractors organized by a Shil’nikov homoclinic orbit
Model Comparison and Agreement
The three models show the same route to chaos, and that the averaged equations shift local bifurcations by an amount of order ε, consistent with averaging theory for the Hopf bifurcation and period doublings The geometrically exact and two-mode models agree with each other to within 0.002 at the local bifurcations for ε ≤ 0. p. 0 p. 001 The averaged equations place each local bifurcation 0. p. 37ε to 1.2ε higher in the scaled detuning
Bifurcation Shifts and Sensitivity
The averaging places each local bifurcation 0.7ε to 1.2ε higher in the scaled detuning The geometric exact and two-mode models agree with each other to within 0. p. 0002 at the local bifurcations for ε ≤ 0.01 Global bifurcations can be shifted by several times ε, and the size of this error is set by how steeply the splitting function of the homoclinic loop crosses zero The homoclinic bifurcation is the most sensitive to modeling approximations
Global Bifurcation Analysis
The primary homoclinic orbit is found by following the unstable manifold of the upper planar saddle focus The splitting function β(∆) marks the homoclinic orbit where it crosses zero The Shil’nikov ratio, which decides whether chaos accompanies the homoclinic orbit, agrees among the three models to within 1.5% for ε ≤ 0.01
Conclusion on Model Accuracy
The averaged equations agree with the two string models in the sequence of bifurcations on the route to chaos and in the Shil’nikov ratio of the saddle focus that organizes it For local bifurcations, the coefficient of ε is of order one At ε = 0.02, the full string’s shifts are 70 to 80% of the two-mode string’s, and the difference is mostly one of amplitude, not of ε
Model Limitations and Future Directions
The study offers an example of what can be expected when a reduced model is used near a route to chaos The natural next step is a theory of the O(ε) shifts, such as second-order averaging or spectral submanifolds The comparison could be made for other distributed systems whose motion is confined to a few modes
Data Availability and Acknowledgements
The run data that support the findings of this study (the brute-force scans, the growth-rate runs and the homoclinic-orbit runs for the three models) and the MATLAB code for the three models, the scans and the analysis are deposited at Zenodo The author used Claude (Anthropic) to document and test computer code, structure simulation sequences, and check text of paper The research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors The author declares that he has no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper
References
[1] R. J. Hanson, J. M. Anderson, H. K. Macomber, Measurements of nonlinear effects in a driven vibrating wire, J. p. 1 Acoust. Soc. Am 96 (3) (1994) 1549–1556
[2] T. C. A. Molteno, N. B. Tufillaro, Torus doubling and chaotic string vibrations: Experimental results, J. p. 1 Sound Vib 137 (2) (1990) 327–330
[3] T. C. Molteno, N. B. Tufillaro, An experimental investigation into the dynamics of a string, Am J Phys 72 (9) (2004) 1157–1169
[4] G. R. Kirchhoff, Vorlesungen über mathematische Physik, B. p. 0 G. Teubner, Leipzig, 1883, three volumes
[5] G. F. Carrier, On the non-linear vibration problem of the elastic string, Q Appl. p. 2 Math 3 (2) (1945) 157–165
[6] G. S. S. Murthy, B. S. Ramakrishna, Nonlinear character of resonance in stretched strings, J Acoust. p. 1 Soc. Am 38 (3) (1965) 461–471
[7] J. W. Miles, Stability of forced oscillations of a vibrating string, J Acoust. p. 1 Soc. Am 38 (5) (1965) 855–861
[8] R. Narasimha, Non-linear vibration of an elastic string, J Sound Vib 8 (1) (1968) 134–146
[9] P. M. Morse, K. U. Ingard, Theoretical Acoustics, McGraw-Hill, New York, 1968
[10] G. V. Anand, Large-amplitude damped free vibration of a stretched string, J Acoust. p. 2 Soc. Am 45 (5) (1969) 1089–1096
[11] G. Tagata, Harmonically forced, finite amplitude vibration of a string, J Sound Vib 51 (4) (1977) 483–492
[12] S. S. Antman, The equations for large vibrations of strings, Am Math Mon 87 (5) (1980) 359–370
[13] J. W. Miles, Resonant, nonplanar motion of a stretched string, J Acoust. p. 2 Soc. Am 75 (5) (1984) 1505–1510
[14] J. M. Johnson, A K Bajaj, Amplitude modulated and chaotic dynamics in resonant motion of strings, J Sound Vib 128 (1) (1989) 87–107
[15] N. B. Tufillaro, Nonlinear and chaotic string vibrations, Am J Phys 57 (5) (1989) 408–414
[16] O. O’Reilly, P. p. 1 J. Holmes, Non-linear non-planar and non-periodic vibrations of a string, J Sound Vib 153 (3) (1992) 413–435
[17] A. Mielke, P. Holmes, O. O’Reilly, Cascades of homoclinic orbits to and chaos near a Hamiltonian saddle-center, J Dyn Differ Equ 4 (1) (1992) 95–126
[18] A. K. Bajaj, J. M. Johnson, On the amplitude dynamics and crisis in resonant motion of stretched strings, Philos Trans R Soc Lond A 338 (1649) (1992) 1–41
[19] O. O’Reilly, Global bifurcations in the forced vibration of a damped string, Int J Non-Linear Mech 28 (3) (1993) 337–351
[20] A. Watzky, Non-linear three-dimensional large-amplitude damped free vibration of a stiff elastic stretched string, J Sound Vib 153 (1) (1992) 125–142
[21] H. P. W. Gottlieb, Non-linear, non-planar transverse free vibrations of a constant-tension string, J Sound Vib 191 (4) (1996) 563–575
[22] M. B. Rubin, O. Gottlieb, Numerical solutions of forced vibration and whirling of a non-linear string using the theory of a Cosserat point, J Sound Vib 197 (1) (1996) 85–101
[23] P. V. Bayly, K. D. Murphy, Coupling between dissimilar modes in an asymmetrically forced string, J Acoust Soc Am 103 (6) (1998) 3362–3369
[24] M. J. Leamy, O. Gottlieb, Internal resonances in whirling strings involving longitudinal dynamics and material non-linearities, J Sound Vib 236 (4) (2000) 683–703
[25] S. Bilbao, Conservative numerical methods for nonlinear strings, J Acoust Soc Am 118 (5) (2005) 3316–3327
[26] S. Bilbao, Conservative numerical methods for nonlinear string dynamics: Non-planar vibration, Tech rep. p. 2, University of Edinburgh, arXiv 1905. p. 009235 (2006)
[27] M. Ducceschi, S. p. 2 Bilbao, Simulation of the geometrically exact nonlinear string via energy quadratisation, J Sound Vib 534 (2022) 117021
[28] S. Bilbao, M. Ducceschi, Models of musical string vibration, Acoust Sci Technol 44 (3) (2023) 194–209
[29] R. Russo, Non-iterative numerical simulation techniques for nonlinear string vibration in musical acoustics, Ph. p. 2D thesis, University of Bologna (2025)
[30] P. Suchde, A S Vasudeva Murthy, On the wave equations of Kirchhoff–Narasimha and Carrier, in Applied Analysis in Biological and Physical Sciences, Springer Proceedings in Mathematics & Statistics, Springer New Delhi 2016
[31] P. Suchde, A S Vasudeva Murthy, The tangled history of the string equation, Bh¯avan¯a 1 (2), https://bhavana. p. 1org.in/the-tangled-history-of-the-string-equation/ (2017)
[32] P. Holmes, A nonlinear oscillator with a strange attractor, Philos Trans R Soc Lond A 292 (1394) (1979) 419–448
[33] F. C. Moon, P. J. Holmes, A magnetoelastic strange attractor, J Sound Vib 65 (2) (1979) 275–296
[34] J. P. Cusumano, F C Moon, Chaotic non-planar vibrations of the thin elastica, Part I: Experimental observation of planar instability, J Sound Vib 179 (2) (1995) 185–208
[35] J. P. Cusumano, F C Moon, Chaotic non-planar vibrations of the thin elastica, Part II: Derivation and analysis of a low-dimensional model, J Sound Vib 179 (2) (1995) 209–226
[36] M. Pakdemirli, S. A Nayfeh, A H Nayfeh, Analysis of one-to-one autoparametric resonances in cables—discretization vs. p. 6 direct treatment, Nonlinear Dyn 8 (1) (1995) 65–83
[37] A. H. Nayfeh, W. Lacarbonara, On the discretization of distributedparameter systems with quadratic and cubic nonlinearities, Nonlinear Dyn 13 (3) (1997) 203–220
[38] G. Rega, Nonlinear vibrations of suspended cables—Part I: Modeling and analysis, Appl Mech Rev 57 (6) (2004) 443–478
[39] C. Touzé, M. Amabili, Nonlinear normal modes for damped geometrically nonlinear systems: Application to reduced-order modelling of harmonically forced structures, J Sound Vib 298 (4–5) (2006) 958–981
[40] J. H. Curry, J. R. Herring, J. Loncaric, S. A. Orszag, Order and disorder in two- and three-dimensional Bénard convection, J Fluid Mech 147 (1984) 1–38
[41] D. Armbruster, J. Guckenheimer, P Holmes, Kuramoto–Sivashinsky dynamics on the center–unstable manifold, SIAM J Appl Math 49 (3) (1989) 676–691
[42] C. Touzé, O. Thomas, M. Amabili, Transition to chaotic vibrations for harmonically forced perfect and imperfect circular plates, Int J NonLinear Mech 46 (1) (2011) 234–246
[43] J. M. T. Thompson, H. p. 4 B. Stewart, Nonlinear Dynamics and Chaos, 2nd Edition, Wiley, Chichester, 2002
[44] A H Nayfeh, B Balachandran, Applied Nonlinear Dynamics Analytical Computational and Experimental Methods Wiley New York 1995
[45] F. C. Moon, Chaotic and Fractal Dynamics An Introduction for Applied Scientists and Engineers Wiley New York 1992
[46] S. H. Strogatz, Nonlinear Dynamics and Chaos, 2nd Edition Westview Press Boulder, 2015
[47] J. Guckenheimer, P Holmes, Nonlinear Oscillations Dynamical Systems Bifurcations of Vector Fields Springer New York 1983
[48] S. Wiggins, Introduction to Applied Nonlinear Dynamical Systems and Chaos, 2nd Edition Springer New York 2003
[49] P. Glendinning, Stability, Instability and Chaos An Introduction to the Theory of Nonlinear Differential Equations Cambridge University Press Cambridge 1994
[50] Y. A. Kuznetsov, Elements of Applied Bifurcation Theory, 3rd Edition Springer New York 2004
[51] L. P. Shil’nikov, A case of the existence of a denumerable set of periodic motions, Sov Math Dokl 6 (1965) 163–166
[52] P. Glendinning, C Sparrow, Local and global behavior near homoclinic orbits, J Stat Phys 35 (5–6) (1984) 645–696
[53] P. Gaspard, R. Kapral, G. Nicolis, Bifurcation phenomena near homoclinic systems: A two-parameter analysis, J Stat Phys 35 (5–6) (1984) 697–727
[54] C. Grebogi, E. Ott, J. A. Yorke, Crises, sudden changes in chaotic attractors, and transient chaos Physica D 7 (1–3) (1983) 181–200
[55] A H Nayfeh, D. T. Mook, Nonlinear Oscillations, Wiley New York 1979
[56] J. A. Sanders, F Verhulst, J Murdock, Averaging Methods in Nonlinear Dynamical Systems 2nd Edition Springer New York 2007
[57] J. A. Elliott, Intrinsic nonlinear effects in vibrating strings Am J Phys 48 (6) (1980) 478–480
[58] N. B. Tufillaro, T Abbott, J Reilly, An Experimental Approach to Nonlinear Dynamics and Chaos Addison-Wesley Redwood City CA 1992 <ref:2611. p. 2
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Improved AI can accurately predict where local bifurcations of a reduced model will occur by estimating their shift from averaged values as
the Hopf bifurcation of the string is within order epsilon of that of the averaged flow.
-
Improved AI can estimate global bifurcation shifts by using
the reciprocal of its slope
to calculate how much a global event like the primary homoclinic orbit moves based on the splitting function's zero crossing. -
Improved AI can reliably determine whether chaos accompanies a homoclinic orbit by calculating
the Shil’nikov ratio,
which in all three models agrees to within 1.5% for epsilon less than or equal to 0.01 and 2% at epsilon approximately 0.02. -
Improved AI can distinguish between amplitude effects and averaging errors by comparing the shift of bifurcation points across different damping levels, noting that
at the largest damping the full string’s shifts are smaller than the two-mode string’s by 20 to 30%.
-
Improved AI can identify that while local bifurcations are sensitive to epsilon,
the shift of a global bifurcation is larger where the splitting function of the homoclinic loop crosses zero with a small slope.
Sources
Related papers
- PRISM: Unifying multiphysics optothermal transport and identifiability-guided inference for nanoscale thermal metrology
- Quantifying The Complex Spatiotemporal Chaos of Cardiac Fibrillation in Ionic Models Across Parameter Regimes
- Modal interactions in strongly nonlinear coupled oscillators with closely spaced modes and nonclassical damping distribution
- An active Lorentz gas: walking droplets in periodic media
- Stochastic-layer width as an order parameter for the integrability-to-chaos transition in the bouncer model
- When more data become less informative: Finite-precision periodicization and collapse of forecast-error Lyapunov estimates