Two-Dimensional Flows - Main Theorem

TheoremPoincare-Bendixson Theorem

Let $\dot{\mathbf{x}} = \mathbf{F}(\mathbf{x})$ be a continuous-time dynamical system on $\mathbb{R}^2$ , and let $K \subset \mathbb{R}^2$ be a closed, bounded region that contains no fixed points. If a trajectory $\phi_t(\mathbf{x}_0)$ is confined to $K$ for all $t \geq 0$ (i.e., $\phi_t(\mathbf{x}_0) \in K$ for all $t \geq 0$ ), then one of the following holds:

The trajectory $\phi_t(\mathbf{x}_0)$ is a closed orbit (periodic)
The trajectory approaches a closed orbit as $t \to \infty$ (the $\omega$ -limit set is a periodic orbit)

More generally, if $K$ contains fixed points, the $\omega$ -limit set is either:

A single fixed point
A single closed orbit
A connected union of fixed points and trajectories connecting them (heteroclinic/homoclinic orbits)

This theorem completely characterizes the asymptotic behavior of planar flows.

The Poincare-Bendixson theorem is one of the most fundamental results in dynamical systems, providing a complete topological classification of long-term behavior in two dimensions. It guarantees that bounded trajectories in planar flows must exhibit relatively simple asymptotic dynamics: they either settle down to equilibria, approach periodic orbits, or follow connecting orbits between fixed points.

Remarkably, the theorem rules out chaotic behavior for planar autonomous systems. While one-dimensional maps can be chaotic and three-dimensional flows certainly exhibit chaos, two-dimensional flows occupy a middle ground where dynamics is rich enough for interesting behavior (limit cycles, heteroclinic connections) but constrained enough for complete classification.

TheoremBendixson-Dulac Criterion

Let $\dot{\mathbf{x}} = \mathbf{F}(\mathbf{x}) = (f(x,y), g(x,y))$ be a planar flow on a simply connected region $D \subset \mathbb{R}^2$ . If there exists a continuously differentiable function $\phi(x, y) > 0$ on $D$ such that:

$\nabla \cdot (\phi \mathbf{F}) = \frac{\partial (\phi f)}{\partial x} + \frac{\partial (\phi g)}{\partial y}$

has constant sign (not identically zero) on $D$ , then there are no closed orbits entirely contained in $D$ .

When $\phi \equiv 1$ , this reduces to the Bendixson criterion: if $\nabla \cdot \mathbf{F}$ has constant sign, there are no closed orbits.

The Bendixson-Dulac criterion provides a powerful method for ruling out limit cycles without solving the differential equations. By computing a divergence, one can determine whether periodic orbits exist in a given region. This is particularly useful in applications where one wants to prove that a system settles to equilibrium rather than oscillating.

ExampleNo Limit Cycles in Gradient Systems

A gradient system has the form $\dot{\mathbf{x}} = -\nabla V(\mathbf{x})$ for some potential function $V$ . For such systems:

$\nabla \cdot \mathbf{F} = -\nabla^2 V = -\left(\frac{\partial^2 V}{\partial x^2} + \frac{\partial^2 V}{\partial y^2}\right)$

Along any trajectory, $\frac{dV}{dt} = \nabla V \cdot \dot{\mathbf{x}} = -|\nabla V|^2 \leq 0$ , so $V$ is a Lyapunov function. Therefore:

Gradient systems have no closed orbits (since $V$ strictly decreases except at fixed points)
All trajectories approach fixed points as $t \to \infty$

This shows that oscillatory behavior requires non-conservative forces; purely conservative systems in 2D cannot sustain oscillations when starting from non-equilibrium states.

Remark

The Poincare-Bendixson theorem explains why autonomous chaos requires at least three dimensions. In 2D, the "room to maneuver" is insufficient for trajectories to exhibit sensitive dependence on initial conditions while remaining bounded. Three dimensions provide enough freedom for trajectories to fold and stretch in complex ways, creating strange attractors. This dimensional constraint is fundamental: it's not merely a limitation of current mathematical techniques but a genuine topological obstruction to chaos in planar flows.

These theorems provide both positive and negative results: Poincare-Bendixson tells us what can happen (approach to fixed points or limit cycles), while Bendixson-Dulac tells us what cannot happen (closed orbits when divergence has constant sign). Together, they form a powerful framework for completely characterizing planar dynamics.