Physical Pendulums & Torsion · Guide · glidepathscience.org
GuidePhysical Pendulums & Torsion
Contents
This Guide explains the ideas and equations behind every Walkthrough step, Concept Challenge question and Predict round, in the order a textbook takes them. Part 1 prepares the Intro track and Part 2 the Advanced track; each takes about twelve minutes to read. Each section ends with a worked example that is not one of the sim’s own problems.
Part 1 · The essentials
Math you’ll need: algebra, square roots and the sine of an angle; one optional box uses an integral.
A pendulum whose length is hidden in its shape
A simple pendulum is a small bob on a light string: all its mass sits at one distance $L$ from the pivot, and for small swings its period is $T = 2π\sqrt{L/g}$. Almost nothing that swings in the real world is like that. A swinging leg, a door on its hinges, a gate, a church bell, a bat held at the handle and a clock’s heavy rod all have their mass spread out, some of it close to the pivot and some far from it. Such a rigid body, swinging about a fixed pivot, is a physical pendulum (also called a compound pendulum).
Its period depends on how the mass is spread, and on where the pivot is, in ways that surprise people. A meter stick hung from one end swings faster than a bob on a meter of string. Moving the pivot toward the middle first makes it swing faster still, then slower and slower without limit. A hoop and a disk of the same size and mass swing at different rates. This Guide builds the one equation that explains all of it, then turns to a second kind of pendulum, the torsion pendulum, which twists on a wire instead of swinging under gravity.
Where gravity pulls: the center of mass
Gravity pulls on every piece of a body, and each piece has its own lever arm about the pivot. Added up, those torques are exactly the torque of the whole weight $mg$ acting at one point, the center of mass. That is what makes the center of mass so useful: a stick, a disk or an oddly shaped bar can be treated, for gravity’s torque, as if all its mass were there.
Call $d$ the distance from the pivot to the center of mass, and θ the angle the line from the pivot to the center of mass makes with the vertical. The weight acts straight down, so its lever arm is the horizontal distance $d\sinθ$, and gravity’s torque about the pivot is
$τ = −mgd\sinθ$
The minus sign says the torque always turns the body back toward θ = 0, where the center of mass hangs straight below the pivot: it is a restoring torque. Straight down, the lever arm is zero and so is the torque; the body is in equilibrium.
Inertia about the pivot, and the period
How strongly a body resists being set turning depends on its rotational inertia $I$ about the axis, the sum of each piece’s mass times the square of its distance from the axis. Tables give $I_{cm}$, the inertia about an axis through the center of mass: a thin rod of length $L$ has $mL^2/12$ about its middle; a uniform disk of radius $R$ turning in its own plane has $\tfrac12mR^2$; a thin hoop has $mR^2$, all its mass at the rim. About any parallel axis a distance $d$ away, the parallel-axis theorem adds one term:
$I = I_{cm} + md^2$
The added term is never negative, so the center of mass is the axis of least inertia. For a pivot near the end of a stick, the $md^2$ term is usually the larger one: most of the inertia comes from the center of mass being carried round the pivot.
Newton’s second law for rotation, $Iα = τ$, now reads $Iα = −mgd\sinθ$. For small swings, $\sinθ ≈ θ$ in radians, so the torque is $−mgd\,θ$: proportional to the angle, just as a spring’s force is proportional to its stretch. The motion is simple harmonic, with $mgd$ playing the spring constant and $I$ the mass, and the period is
$T = 2π\sqrt{I/mgd}$
The mass cancels, because both $I$ and $mgd$ are proportional to it: an aluminum stick and a steel stick of the same shape keep the same time. What the period depends on is the shape and the pivot.
Moving the pivot, and the equivalent length
Drill a row of holes along a meter stick and hang it from each. From the end, moving inward, the period first gets shorter: $I$ falls, because the $md^2$ term shrinks quickly. But the torque $mgd$ shrinks too, and near the center it wins. With the pivot at the center of mass there is no restoring torque at all, and the stick does not swing back. In between there is a fastest hole: for a uniform stick it is 21.1 cm from the end, where the period is 1.52 s for a 1.00 m stick.
Every physical pendulum keeps time with some simple pendulum. Setting $2π\sqrt{L_{eq}/g}$ equal to $2π\sqrt{I/mgd}$ gives its equivalent length:
$L_{eq} = I/md$
Because $I$ is larger than $md^2$, $L_{eq}$ is always longer than $d$: the body swings more slowly than a bob at its center of mass would. For a stick hung from one end, $I = mL^2/3$ and $d = L/2$, so $L_{eq} = 2L/3$: shorter than the stick. The point $L_{eq}$ below the pivot, along the line through the center of mass, is the center of oscillation.
Disks and hoops
Hang a disk and a hoop of the same mass and radius on nails at their rims. Both have $d = R$, so gravity gives them the same torque. Their inertias about the nail differ: the disk has $\tfrac12mR^2 + mR^2 = \tfrac32mR^2$, the hoop $mR^2 + mR^2 = 2mR^2$, because the hoop’s mass is all at the rim, far from its center. So $L_{eq} = \tfrac32R$ for the disk and $2R$ for the hoop, and the hoop is slower by $\sqrt{4/3}$, about 15 %, whatever the two masses are. The mass cancels; the way it is spread does not.
The torsion pendulum
Hang a disk by its center from a thin wire and twist it. The wire twists back, with a torque proportional to the angle of twist for any twist the wire can take without permanent damage:
$τ = −κθ$
The torsion constant $κ$, in newton meters per radian, says how stiff the wire is. Newton’s second law gives $Iθ̈ = −κθ$, simple harmonic motion with no approximation at all, and the period is
$T = 2π\sqrt{I/κ}$
Two things stand out. There is no $g$: the same torsion pendulum keeps the same time on the Moon. And there is no angle: because the torque is exactly proportional to the twist, the period is the same at 10° or 170°. A gravity pendulum is different. Its torque, $mgd\sinθ$, falls short of $mgdθ$ at wide angles, so a wide swing is pulled back less than simple harmonic motion would need and takes longer: 7 % longer at 60°, 18 % at 90°.
Since $T ∝ \sqrt I$ on a given wire, the period tells shapes apart. A thin ring has twice the inertia of a disk of the same mass and radius, $mR^2$ against $\tfrac12mR^2$, so on the same wire its period is √2 times as long.
Part 2 · Going further
Math you’ll need: derivatives (to find a minimum), the equation of motion as a differential equation, energy conservation, and solving a formula with several terms.
The radius of gyration, the fastest pivot, conjugate pivots
Write the inertia about the center of mass as $I_{cm} = mk^2$. The radius of gyration $k$ is the distance from the center at which a single point of mass $m$ would have the same $I_{cm}$; a thin rod has $k = L/\sqrt{12}$, a disk $R/\sqrt2$, a hoop $R$. Then $I = m(k^2 + d^2)$, and the small-swing period becomes
$T = 2π\sqrt{(k^2 + d^2)/gd}$
which depends on $d$ only through $d + k^2/d$. Its derivative with respect to $d$ is $1 − k^2/d^2$, zero at $d = k$: the fastest pivot is one radius of gyration from the center of mass, and there $T_{min} = 2π\sqrt{2k/g}$. The bottom of the curve is flat, so a pivot near $d = k$ is nearly as fast. As $d$ → 0 the period grows as $1/\sqrt d$; for large $d$ it approaches a simple pendulum’s.
Because $d + k^2/d$ takes the same value at $d$ and at $k^2/d$, every period above the minimum belongs to two pivot distances, $d$ and $d′ = k^2/d$, on either side of the center of mass. These conjugate pivots are $d + d′ = L_{eq}$ apart: hang the body from its center of oscillation and it keeps the same time.
Kater’s reversible pendulum and Bessel’s formula
Measuring $g$ with a physical pendulum seems to need its inertia and its center of mass, both hard to know precisely. Henry Kater’s reversible pendulum (1817) avoids both. It has two knife edges facing each other, on opposite sides of the center of mass, and is adjusted until its periods about the two edges match. Then the edges are conjugate pivots, so their separation $ℓ$, which can be measured very precisely, is $L_{eq}$, and
$g = 4π^2ℓ/T^2$
The periods also match, trivially, when the two edges are the same distance from the center of mass; that mirror match gives the wrong $ℓ$, so a reversible pendulum is built with its center of mass well off the middle. In the Kater tab, one knife edge slides; the mirror match falls off the bar, so only the true match remains.
Matching the periods exactly is slow work. Bessel’s formula uses a pair that nearly matches, with $h_1$ and $h_2$ the edges’ distances from the center of mass:
$8π^2/g = \frac{T_1^2 + T_2^2}{h_1 + h_2} + \frac{T_1^2 − T_2^2}{h_1 − h_2}$
The first term needs only $h_1 + h_2 = ℓ$, known precisely. The second needs $h_1 − h_2$, which depends on the center of mass, but it is multiplied by $T_1^2 − T_2^2$, small when the periods nearly match, so a center of mass found roughly by balancing the bar is good enough.
Wide swings and energy
Dividing $Iθ̈ = −mgd\sinθ$ by $I$ gives $θ̈ = −(g/L_{eq})\sinθ$: exactly the simple pendulum’s equation with $L_{eq}$ for $L$, at every amplitude. So a physical pendulum’s wide swings follow the same exact law as a simple pendulum’s, with the complete elliptic integral of the first kind, computed quickly by the arithmetic–geometric mean (the AGM: replace two numbers by their average and their geometric mean, again and again, until they agree):
$T = T_0/\mathrm{AGM}(1, \cos(θ_0/2))$
where $T_0$ is the small-swing period. The ratio is 1.0019 at 10°, 1.0732 at 60°, 1.1803 at 90° and grows without limit toward 180°. A wire has no such factor: its period ratio is 1 at every twist.
Energy gives the motion without solving the equation. Without friction,
$\tfrac12Iω^2 + mgd(1 − \cosθ) = mgd(1 − \cosθ_0)$
so the turning points are at ±θ₀, where $ω = 0$ and the torque is largest, and the swing is fastest at the bottom. For a torsion pendulum the potential energy is $\tfrac12κθ^2$ instead.
A wire’s stiffness, and weighing an inertia
A round wire of radius $r$, length $ℓ$ and shear modulus $G$ has
$κ = πGr^4/2ℓ$
The fourth power of the radius makes thickness the strongest knob: twice the diameter makes a wire sixteen times as stiff and its period a quarter as long. That is why sensitive torsion balances, like Henry Cavendish’s of 1798, which weighed the Earth, hang from very fine wires or fibers.
A torsion pendulum is the standard way to measure an unknown rotational inertia. Time the disk alone ($T_0$) and with the object centered on it ($T_1$). On the same wire $T^2 = 4π^2I/κ$, so the ratio of the squared periods is the ratio of the inertias, and
$I_x = I_{disk}(T_1^2/T_0^2 − 1)$
The wire’s $κ$ cancels. If the disk’s own inertia is not known either, a calibration object of known inertia fixes it: that is how the Torsion Balance tab works.
The center of percussion
Strike a pivoted body with an impulse $J$ at a distance $s$ below the pivot. The blow sets it spinning about the pivot at $ω = Js/I$, which carries the center of mass along at $ωd = Jsd/I$. Left to itself, with nothing at the pivot, the blow would give the center of mass the speed $J/m$. If these two speeds differ, the pivot must push to make up the difference, and a hand holding it feels a jolt. They agree when
$s = I/md = L_{eq}$
The center of percussion is the center of oscillation. For a stick held at one end it is two thirds of the way down; a ball hit there sends no sting into the hands. Real bats, rackets and hammers are not uniform, and the hands are not a fixed pivot, so this is one of several “sweet spots” people define.
Reference
Common mistakes
Part 1
- Using $2π\sqrt{L/g}$ with the body’s length: a spread-out body is not a bob on a string.
- Putting gravity’s torque at the lowest point, or using $d\cosθ$: the weight acts at the center of mass, with lever arm $d\sinθ$.
- Using $I_{cm}$ for a body that turns about a pivot, or $md^2$ alone: $I = I_{cm} + md^2$.
- Expecting a shorter period the closer the pivot is to the center: the period has a minimum, then grows without limit.
- Thinking a heavier body swings differently: the mass cancels from every period here.
- Treating a torsion pendulum like a gravity pendulum: its period has no $g$ and does not change with the twist.
Part 2
- Taking the mirror pivot (the same $d$ on the other side) as the conjugate pivot: the other one is at $k^2/d$.
- Averaging two unequal periods and using $g = 4π^2ℓ/T^2$: use Bessel’s formula instead.
- Using the diameter for $r$ in $κ = πGr^4/2ℓ$: the error is a factor of sixteen.
- Forgetting to subtract the disk, or to square the ratio, in $I_x = I_{disk}(T_1^2/T_0^2 − 1)$.
The equation sheet
| Part 1 | |
|---|---|
| gravity’s torque | $τ = −mgd\sinθ$ |
| parallel axes | $I = I_{cm} + md^2$ |
| about the center | $I_{cm}$: rod $mL^2/12$, disk $\tfrac12mR^2$, hoop $mR^2$ |
| the period | $T = 2π\sqrt{I/mgd}$ |
| equivalent length | $L_{eq} = I/md$ |
| a torsion pendulum | $T = 2π\sqrt{I/κ}$ |
| Part 2 | |
|---|---|
| with $k$ | $T = 2π\sqrt{(k^2 + d^2)/gd}$ |
| the fastest pivot | $T_{min} = 2π\sqrt{2k/g}$ at $d = k$ |
| conjugate pivot | $d′ = k^2/d$ |
| matched periods | $g = 4π^2ℓ/T^2$ |
| wide swings | $T = T_0/\mathrm{AGM}(1, \cos(θ_0/2))$ |
| a round wire | $κ = πGr^4/2ℓ$ |
| an unknown inertia | $I_x = I_{disk}(T_1^2/T_0^2 − 1)$ |
| energy | $\tfrac12Iω^2 + mgd(1 − \cosθ) = mgd(1 − \cosθ_0)$ |
Symbols: $m$ the mass; $d$ the distance from the pivot to the center of mass; θ the angle from straight down (θ₀ at release); $I$ the rotational inertia about the pivot and $I_{cm}$ about the center of mass; $k$ the radius of gyration; $L$ a rod’s length and $R$ a radius; $L_{eq}$ the equivalent length; $κ$ the torsion constant; $G$ the shear modulus, $r$ a wire’s radius and $ℓ$ its length (or the knife edges’ separation); $h_1$, $h_2$ the knife edges’ distances from the center of mass; $T$ the period and $T_0$ its small-swing value; $ω$ the angular speed; $g$ = 9.81 m/s².
Further reading
- OpenStax, University Physics Volume 1, by Samuel J. Ling, Jeff Sanny and William Moebs, §10.5 “Calculating Moments of Inertia” and §15.4 “Pendulums,” an open textbook (CC BY-NC-SA 4.0). It writes $L$ for the distance from the pivot to the center of mass, where this Guide writes $d$.
- This lab’s Field Manual lists the model, its typical values and what it leaves out. Simple Harmonic Motion & Pendulums treats the simple pendulum, damping and resonance.
Physical Pendulums & Torsion v1.0 · © 2026 Dr. David Meier · CC BY-NC-SA 4.0 · glidepathscience.org