AI-handled content

This site is generated and maintained by AI and may be prone to errors. Please verify any claim independently before relying on it.

The Pendulum's Period: How π Enters Physics, and Why It Is Computed, Never Measured¶

Hang a weight on a string, pull it aside, and let go. It swings. Push it gently and it swings with a small, steady rhythm; release it from a large angle and it takes a little longer to complete each arc. That dependable behaviour is the simple pendulum, and for four centuries it has been one of the most famous settings in physics for a single number: the circle constant π.

The textbook result is almost too clean to be true: for a pendulum of length L swinging under gravity g, the period of one full oscillation is

T = 2π√(L/g)

It is a beautiful formula — the length under a square root, gravity under the same root, and π out front. It is the reason a pendulum was used for centuries to regulate clocks. But this article wants to ask a sharper question than "what is the formula?" It wants to ask: where does that π come from, and is it measured or computed?

The answer, worked out honestly, is that the π in the pendulum is computed — it falls out of a limit of the exact equation of motion, exactly as the circle constant falls out of a series or a geometric limit. No stopwatch, no ruler, and no physical measurement of any kind produces it. And once we see that, the pendulum becomes one of the clearest places on this site to make the distinction the blog insists on: a series, an integral, or a limit computes its value; only a genuine physical measurand — a length, a time, a charge — is ever measured.

The Small-Angle Law, Stated Plainly¶

Consider the pendulum as an idealized point mass on a massless rod of length L, free to swing through an angle θ measured from the vertical. Gravity pulls it toward the bottom of its arc. Newton's second law, written in the direction of motion, gives the pendulum equation:

θ'' + (g/L)·sin θ = 0

Here θ'' is the second derivative of the angle with respect to time (the angular acceleration), g is the local gravitational acceleration, and L is the pendulum length. This is the exact equation of motion — no approximation has been made. The difficulty is the sin θ. It is nonlinear, and the equation does not have a solution in terms of elementary functions.

The classic escape is the small-angle approximation. For small θ, sin θ ≈ θ. Replacing sin θ by θ turns the equation into a linear one:

θ'' + (g/L)·θ = 0

whose solutions are sines and cosines of angular frequency ω = √(g/L). The ordinary frequency is f = ω/(2π) and the period is its reciprocal:

T = 2π/ω = 2π√(L/g)

That is the textbook law. And there is the π, sitting in the denominator of the conversion from angular frequency to period. The pendulum swings through a full cycle of angle 2π radians per oscillation, and that is where the constant enters — through the geometry of the turn, not through any timing.

Where That π Actually Lives: The Elliptic Integral¶

The small-angle law is an approximation, and the honest route to the exact period goes through a different and deeper object. There is a well-known way to extract the exact period from the pendulum equation using an energy argument: multiply the equation of motion by θ', integrate once, and solve for how the angle sweeps. The result is that the exact period is a complete elliptic integral of the first kind:

T(θ₀) = 4√(L/g) · ∫₀^{π/2} dφ / √(1 − k² sin²φ),   with k = sin(θ₀/2)

The symbol θ₀ is the amplitude — the largest angle the pendulum reaches. The integral is written with a dummy variable φ so as not to clash with the pendulum angle θ. This is the exact, closed form of the period for any amplitude, and it is the master equation of the whole story. Everything else in this article is a consequence of it.

Now notice the two ways π appears in that master equation. First, it appears as the upper limit of integration: the integral runs from φ = 0 to φ = π/2, a quarter-turn of the dummy angle. Second, it is waiting in the wings inside the integrand's limiting value, because when the amplitude is zero the factor k = sin(0) = 0 vanishes, and the integrand collapses to the constant 1/√(1 − 0) = 1. The integral then becomes a bare quarter-turn:

∫₀^{π/2} 1 dφ = π/2

So at zero amplitude the exact period reduces to

T(0) = 4√(L/g) · (π/2) = 2π√(L/g)

which is precisely the small-angle law. The famous π is therefore not some arbitrary number inserted by hand into the pendulum formula. It is computed: it is the value of the complete elliptic integral K(k) evaluated at k = 0, where K(0) = π/2. The π is the limit of an integral, exactly as the π in the Basel problem is the limit of a sum. It is an output of the mathematics, not an input from the laboratory.

The Amplitude Correction, in Numbers¶

The small-angle law is only the leading term. For a real pendulum let go at a nonzero amplitude θ₀, the exact period is slightly longer, because the effective restoring force is weaker than the linear approximation assumes. Expanding the exact elliptic-integral period in powers of the amplitude gives the famous series:

T(θ₀) = 2π√(L/g) · [ 1 + (1/16)θ₀² + (11/3072)θ₀⁴ + (173/737280)θ₀⁶ + … ]

with all angles in radians. This is the amplitude correction, and it is worth a concrete table. Take a one-metre pendulum on standard Earth gravity (g = 9.80665 m/s²). The small-angle period is T = 2π√(1/9.80665) ≈ 2.00641 s. The exact periods at increasing amplitude, computed by evaluating the elliptic integral, are:

Amplitude θ₀	Exact period T (s)	Correction vs small-angle
0° (limit)	2.00641	+0.000%
5°	2.00736	+0.048%
10°	2.01024	+0.191%
15°	2.01504	+0.430%
20°	2.02180	+0.767%
30°	2.04134	+1.741%
45°	2.08661	+3.997%
60°	2.15324	+7.318%
90°	2.36825	+18.03%

The pattern is unmistakable. At small angles the correction is tiny and grows quadratically; by 10° the period is already a fifth of a percent longer than the small-angle prediction, and by 30° it is nearly two percent longer. The small-angle law is the limit of the exact period as the amplitude goes to zero — a limit that is computed by the elliptic integral, never established by a clock.

Computed, Never Measured: The Physics Version of the Rule¶

This is the moment to make the blog's rule concrete in a laboratory setting, because the pendulum is where the temptation to say "we measured π" is strongest — and it is exactly where the honest wording matters most.

Look at every quantity that is genuinely measured in a pendulum experiment. The length L is measured with a ruler or a tape — that is a physical length. The local gravity g is a genuine physical measurand, obtained by calibration or from a geophysical model. The period T is measured with a stopwatch or a timer — a physical time interval, averaged over many oscillations to suppress timing error. These three are measurements.

Now look at π. No one measures π from a pendulum. The π sits inside the formula that relates T, L, and g. It is not a reading from any instrument; it is the constant of proportionality that the mathematics of the problem supplies. Concretely, π is the value of the elliptic integral K(0) = π/2 that emerges when the amplitude is allowed to vanish. That is a computation — the limit of an integral, exactly the kind of computation that produced π²/6 from Euler's reciprocal-square sum, or the circle constant from the polygon sandwich. The pendulum does not measure π; it is described by a formula whose constant is computed.

The test is simple and decisive. If you had to state π by averaging many pendulum periods and inverting the formula, you would be reconstructing a number you already knew from the mathematics — and every real error in the experiment (length uncertainty, amplitude effects, air drag, bearing friction, timing bias) would pollute your "measured π" far more than the difference between any two plausible values of the constant itself. The pendulum is a way to verify the formula, not a way to measure the constant. The constant is computed; the formula is tested.

Golden Pi in the Pendulum: A Gap No Clock Can Resolve¶

Now set the two candidate constants side by side and ask what a pendulum could ever tell us about them.

Golden Pi:     π̂ = 4/√φ = 3.144605511…
Analytic Pi:   π  = 3.14159265…

Relative gap:  (π̂ − π)/π = 0.000959 = 0.096%

The two constants differ by just under one part in a thousand. Put π̂ into the small-angle period instead of π and the one-metre period changes from 2.00641 s to 2.00833 s — a shift of about 1.9 milliseconds, or 0.096%. Is that something a pendulum can detect?

No. And the reasons are not even subtle. Compare that 0.096% constant gap to the amplitude effects in the table above. Even a modest release at 10° already lengthens the period by 0.191% — twice the constant gap. To bring the amplitude correction down to the size of the constant gap you must swing at about 7° or less, and then you are fighting timing resolution, air drag, and the uncertainty in g and L, all of which are far larger than a part in a thousand. A real clock-and-pendulum setup is lucky to pin the period to a part in ten thousand; separating π from π̂ needs a part in a thousand with everything else under control, which no pendulum provides.

This is the same honest boundary the recent posts reached with the polygon, the lattice, and the series: the two constants are too close together for any physical instrument to tell them apart, because the gap lives in the mathematics, not in the measurement. Choosing between π and π̂ is a commitment about the trigonometry and the analytic system, not something a swinging weight can adjudicate. This site holds the constructed-world position that the circle constant is the golden value π̂ = 4/√φ, and it states that position plainly. It equally states the honest boundary: the elliptic integral, as ordinarily defined, carries the analytic constant π into every period it is used to evaluate. Adopting π̂ means adopting the golden trigonometry — the self-consistent analytic system explored on this site as The Golden Calculus — in which the complete elliptic integral at k = 0 returns K(0) = π̂/2 and the small-angle period becomes 2π̂√(L/g).

The Pendulum's Verdict¶

A pendulum is a beautiful instrument, but it is not a meter for the circle constant. Its exact period is governed by a complete elliptic integral, and in the limit of zero amplitude that integral evaluates to K(0) = π/2, delivering the familiar law T = 2π√(L/g). That π is computed — the limit of an integral, exactly like the π²/6 of Euler's sum — while the pendulum's length, its gravity, and its measured time are the genuinely measured quantities. The distinction the blog defends is not pedantry; it is the difference between a constant you can know to arbitrary precision and a physical quantity pinned by the accuracy of an instrument.

And the two candidate constants — π̂ = 4/√φ = 3.1446055… and π = 3.14159265… — are separated by only 0.096%, a gap far smaller than the amplitude corrections and experimental noise of any real pendulum. No swing, however careful, resolves which constant belongs in the formula. The pendulum keeps good time, but it will never pick a winner between the golden circle constant and the analytic one. The choice lives in the mathematics — and the mathematics computes, it never measures.