Simple pendulum period at large angles: the exact formula in JavaScript
A simple pendulum's period is T = 2π√(L/g) only for small swings. At larger angles the true period is longer, and you can compute it exactly with the arithmetic-geometric mean (AGM) in a few lines of JavaScript. Below I compare the textbook formula, a short series and the exact value for a 1 m pendulum. (I build PractiSci, which has a free virtual pendulum lab.)
What is the period of a simple pendulum?
For small angles T₀ = 2π√(L/g). With L = 1 m and g = 9.81 m/s² that is 2.006 s. The formula assumes sin θ ≈ θ, which is only good for a few degrees.
How do I get the exact period at large angles?
The exact period is T = T₀ / AGM(1, cos(θ₀/2)). The AGM converges in a handful of steps:
const g = 9.81, L = 1.0;
const T0 = 2 * Math.PI * Math.sqrt(L / g);
function agm(a, b) {
while (Math.abs(a - b) > 1e-15) [a, b] = [(a + b) / 2, Math.sqrt(a * b)];
return a;
}
const exact = (deg) => T0 / agm(1, Math.cos((deg * Math.PI / 180) / 2));
const series = (deg) => {
const t = deg * Math.PI / 180;
return T0 * (1 + t * t / 16 + 11 * t ** 4 / 3072);
};
How wrong is the small-angle formula?
| Amplitude | Exact period (s) | Longer than T₀ by | Series (s) |
|---|---|---|---|
| 5° | 2.0070 | 0.05% | 2.0070 |
| 10° | 2.0099 | 0.19% | 2.0099 |
| 20° | 2.0215 | 0.77% | 2.0215 |
| 30° | 2.0410 | 1.74% | 2.0410 |
| 45° | 2.0863 | 4.00% | 2.0861 |
| 60° | 2.1529 | 7.32% | 2.1522 |
| 90° | 2.3678 | 18.03% | 2.3592 |
Up to about 10° the error is under 0.2%, which is why the school experiment uses small swings. At 90° the period is already 18% longer.
FAQ
Does the mass of the bob change the period?
No. The mass cancels out of the equation of motion, so only L and g matter.
Why do we measure 20 oscillations in the practical?
Timing 20 swings divides the reaction-time error of your stopwatch by 20.
Where can I try it?
In the free simple pendulum lab you change the length and the amplitude and time the swings. The oscillations topic guide explains the theory.
Comments
Join the discussionNo comments yet. Be the first to comment.