Skip to content
All apps

Double pendulum — butterfly effect (3D)

About this app

Up to 60 double pendulums start almost identically — their glowing trails form a 3D curtain that visibly tears into chaos. Perturbation Δθ₀ down to 10⁻⁷° adjustable, a divergence plot shows exponential growth (Lyapunov); at small angles everything stays permanently ordered.

Chaos from two hinges

A double pendulum is two rods hinged end to end. Its equations of motion are fully deterministic, yet at large swings its behaviour is chaotic: two starting states that differ only slightly separate exponentially, roughly as Δ(t) ≈ Δ₀·e^(λt). The growth rate λ is called the Lyapunov exponent. This sensitive dependence on initial conditions is often called the butterfly effect. At small angles the same pendulum behaves almost linearly and stays predictable.

What you see

A 3D scene shows 5 to 60 double pendulums side by side. Neighbours differ in their starting angle only by the offset Δθ₀. Their glowing trails first form a single curtain that tears apart after a while. The divergence panel measures the angle difference between the front and back pendulum and plots log₁₀ Δ against time. A rising straight line in this chart means exponential growth; the time until separation is shown.

Things to try

Compare the two presets: with “Chaos (135°)” the curtain tears after a short time, with “Gentle (28°)” all pendulums stay in step. Reduce the offset step by step from 1° to 10⁻⁷°: separation comes later, but it still comes as long as the starting angle is large. In the chart, notice that each factor of ten less offset only delays the split by roughly the same amount of time.

Frequently asked questions

Why is a double pendulum chaotic?
A double pendulum is a nonlinear system with two coupled degrees of freedom. At large swings the two arms exchange energy irregularly, and tiny differences in the starting state grow exponentially. Its motion can therefore only be predicted for a short time, even though the equations are known exactly.
What is the butterfly effect?
The butterfly effect describes the sensitive dependence of chaotic systems on initial conditions. A tiny change at the start can lead to a completely different course after some time. The term goes back to the meteorologist Edward Lorenz.
What does the Lyapunov exponent measure?
The Lyapunov exponent λ tells you how fast two neighbouring trajectories of a system separate: their distance grows roughly like e^(λt). A positive λ means the system is chaotic. The prediction horizon then grows only logarithmically with measurement accuracy.
What can I try in this app?
You set the starting angle, the offset Δθ₀, the number of pendulums and the speed, and watch when the pendulums drift apart. The divergence panel plots the gap on a log scale over time. The 28° and 135° presets let you compare ordered and chaotic motion directly.

Subject: Chaos & Dynamics

More from Chaos & Dynamics

A VisuApp by heyprof: interactive, free in your browser, no sign-up. What is a VisuApp? · All apps