Abstract
Homoclinic bifurcation theory is discussed, beginning with the Lorenz equations as an example, and the corresponding two-dimensional Poincaré map and its one-dimensional approximation are derived. Symbolic dynamics and the Smale horseshoe are described, and then Shilnikov bifurcations for saddle-focus homoclinic bifurcations are analysed. The final section generalises the results to n-dimensional flows, and mentions the corresponding results for infinite-dimensional (partial differential equation) flows, without detail. The possible relation to fluid turbulence is mentioned.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
We can in fact assume that these points are exact by making a further change of variable, so that the stable manifold is locally given by the \((x_{2}, x_{3})\) plane, and the unstable manifold is the \(x_{1}\) axis. The procedure is similar to the centre manifold construction. But if the box is small (\(|c_{i}| \ll 1)\), the error is small anyway, and the formal construction involved is identical.
- 2.
An affine map is simply a linear, inhomogeneous map.
- 3.
In fact, this can be formally derived by simply linearising the flow map from \(S'_\pm \) to S.
- 4.
A Cantor set in an interval is one which is closed and uncountable but has no interior points and no isolated points.
- 5.
Essentially, this means that points in the set can be written as a coordinate pair \((x_u, x_s)\), with \(x_u\in M_u\) and \(x_s\in M_s\).
- 6.
Because we are solving ordinary differential equations to produce the Poincaré map; if we are solving partial differential equations, we can still manage but it requires some artistry.
- 7.
The orbit approaches every point in I arbitrarily closely.
- 8.
Fundamental matrices were in effect defined for linear systems of ordinary differential equations at (3.114).
- 9.
The one \(\propto te^{tD_0}\).
- 10.
This effectively places v on the (outgoing) face S of the box near 0, although it may be simply thought of as a convenient transformation.
- 11.
Remember that D is diagonal.
- 12.
In this context, the co-dimension refers to the number of independent parameters whose selection is necessary to secure the bifurcation.
- 13.
See Sect. 4.5.
- 14.
A finite domain can be effectively infinite if the smallest relevant dynamical length scale is much smaller than the domain size; this commonly occurs when the dynamical forcing parameter (e.g. the Reynolds number in shear flows, or the Rayleigh number in convection) is large.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this chapter
Cite this chapter
Fowler, A., McGuinness, M. (2019). Homoclinic Bifurcations. In: Chaos. Springer, Cham. https://doi.org/10.1007/978-3-030-32538-1_4
Download citation
DOI: https://doi.org/10.1007/978-3-030-32538-1_4
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-32537-4
Online ISBN: 978-3-030-32538-1
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)