Newton's Second Law for Rotation (τ = Iα)

You now hold all three characters of rotational mechanics — torque (the cause), angular acceleration (the effect), and moment of inertia (the resistance) — and this module snaps them into a single sentence: $\tau_{\text{net}} = I\alpha$ . It reads exactly like $F = ma$ with every symbol swapped through the rotational dictionary, and that is no accident. But here is the part worth savouring: it is not a new law of nature. A rigid body is a swarm of ordinary particles, each obeying plain old $F = ma$ . Multiply each particle's equation by its distance from the axis, add up the whole swarm, and $\tau = I\alpha$ falls out — with the millions of unknowable internal forces cancelling in pairs and the moment of inertia manufacturing itself before your eyes. That derivation is the intellectual heart of the chapter, and it converts everything you already own into rotational currency. Spin-up problems become two-step routines: torque inventory divided by $I$ gives $\alpha$ ; then the rotational SUVAT from three modules back takes over. Flywheels, ceiling fans, grindstones, and the massive pulleys that break the old "same tension both sides" rule — all run on this one line. Every board in this module's set tests it, each in its dialect: CBSE/ICSE with the flywheel-and-cord calculation (an NCERT set piece), NEET with one-step $\alpha = \tau/I$ MCQs, JEE with multi-torque inventories and coupled systems, AP and IB-HL with "reason about the swarm" conceptual items. The full pulley-with-hanging-mass machine gets its own archetype module two steps ahead; here we build the law it runs on — and preview exactly why a hanging weight is not the same as a steady pull.

Key takeaways

Newton's second law for rotation about a fixed axis: $\tau_{\text{net}} = I\alpha$ — net external torque equals moment of inertia times angular acceleration, all about the same axis.
It is derived, not decreed: apply $F = ma$ to every particle, convert to torques, sum — internal forces cancel in Newton's-third-law pairs, leaving only external torques.
Solve recipe: torque inventory (signs! friction included) → $I$ about the actual rotation axis (Steiner if off-centre) → $\alpha = \tau_{\text{net}}/I$ → rotational SUVAT.
Torque pairs with angular acceleration, never with angular velocity: a coasting flywheel needs zero torque — asking "what torque keeps it spinning?" is rotation's version of Aristotle's error.
A cord unwinding from a rim links the linear and angular worlds by $a = R\alpha$ (no slipping) — the bridge that couples $\tau = I\alpha$ to $F = ma$ in mixed systems.
A steady pull $F$ on the cord is not equivalent to hanging a weight of $mg = F$ : the hanging mass accelerates, so the string tension drops below $mg$ — the massive-pulley preview.
Key idea: $\tau_{\text{net}} = I\alpha$ is $F = ma$ summed around the axis — inventory the external torques, divide by the moment of inertia, and rotation becomes the mechanics you already know.

The core insight

About a fixed axis, the net external torque on a rigid body equals its moment of inertia times its angular acceleration, $\tau_{\text{net}} = I\alpha$ — a theorem earned by summing $F = ma$ over the body's particles (internal torques cancelling pairwise), and the engine of every spin-up, brake-down, and massive-pulley problem in the chapter.

Where this lives today

$\tau = I\alpha$ is the sizing equation of the electromechanical world. An electric vehicle's headline trick — instant launch — is a motor that delivers near-maximum torque from zero rpm into a deliberately small drivetrain $I$ ; the spec sheet's torque curve is the left side of equation (1). Robot joints are bought by this arithmetic: a shoulder actuator is specified by the torque needed to give the arm's moment of inertia (Steiner-shifted to the joint) the required angular acceleration, hundreds of times a second. Hard-drive engineers fought spin-up time — how long from rest to 7200 rpm — as a direct $\tau/I$ battle: lighter platters, stronger spindle motors. Washing machines solve the inverse problem, ramping a drum of wet-laundry $I$ (which changes load to load — the controller estimates it from the motor's own $\tau$ and measured $\alpha$ ) up to a 1400 rpm spin. Even the bicycle you ride runs on the massive-pulley preview: your legs' torque at the crank, geared through wheels whose rim-heavy $I$ you feel every time you accelerate from the lights.

The physical picture

Picture the rigid body as what it is: a swarm of particles welded together. Push the swarm into rotation and every particle must ride its own circle — but rigidity means they all share one angular acceleration $\alpha$ . Now bill each particle. A particle of mass $m_i$ at radius $r_i$ needs tangential force $m_i r_i \alpha$ (farther out means a faster speed-up, by the bridge $a_t = r\alpha$ ); and delivering that force costs torque $r_i \times m_i r_i \alpha = m_i r_i^2 \alpha$ (the lever arm bills the radius a second time). Add the swarm: total torque bill $= \left(\sum m_i r_i^2\right)\alpha = I\alpha$ . The moment of inertia you built two modules ago was precisely this bill's coefficient, waiting for its law.

Figure

The swarm view — one shared alpha, torque bills summed particle by particle

The miracle in the sum is what disappears. Every particle also feels ferocious internal forces — the interatomic pulls holding the body together. But Newton's third law delivers them in equal-and-opposite pairs acting along the line joining each pair of particles: two opposite forces on one line share a single lever arm about any axis, so their torques cancel exactly, pair by pair. However violent the internal tug-of-war, it contributes nothing to the rotation — which is why a body cannot spin itself up from inside, and why the law is usable: only the external torques, the ones you can actually list, appear in the inventory. Once you have the law, rotation problems inherit the entire linear playbook. Constant torque means constant $\alpha$ , and the rotational SUVAT of three modules back — $\omega = \omega_0 + \alpha t$ , $\theta = \omega_0 t + \tfrac12\alpha t^2$ , $\omega^2 = \omega_0^2 + 2\alpha\theta$ — takes the handoff. And keep the first-law limit in view: $\tau_{\text{net}} = 0$ means $\alpha = 0$ , not $\omega = 0$ . A flywheel in good bearings coasts indefinitely; torque correlates with the change of spin, never with spin itself. Exam options asserting "a torque is needed to maintain rotation" are the rotational rebirth of Aristotle's oldest mistake.

The math: the law and the working routine

The law, in its fixed-axis working form:

\tau_{\text{net}} = I\,\alpha \qquad (1)

where: $\tau_{\text{net}}$ — the signed sum of all external torques about the rotation axis (anticlockwise positive, each torque computed as $rF\sin\phi$ per the torque module); $I$ — the moment of inertia about that same axis (table entry, plus a parallel-axis shift if the pivot is off-centre); $\alpha$ — the angular acceleration in $\mathrm{rad/s^2}$ . Feel a number: a grindstone of $I = 1.6\ \mathrm{kg\,m^2}$ under a net $8.0\ \mathrm{N\,m}$ spins up at $\alpha = 5.0\ \mathrm{rad/s^2}$ — from rest to $300\ \mathrm{rpm}$ ( $31.4\ \mathrm{rad/s}$ ) in $6.3$ seconds. Everything in the equation refers to one axis: mixing a torque about the hinge with a moment of inertia about the centre is the silent unit-mixing error of this chapter. For systems where a cord, belt, or rim connects rotation to straight-line motion, the two worlds are welded by the no-slip bridge:

a = R\,\alpha \qquad (2)

where: $a$ — the linear acceleration of the cord (or of a point on the rim); $R$ — the radius at which the cord leaves the wheel. This is the differentiated version of $v = R\omega$ from the kinematics module, and it is the third equation that closes every mixed system: one $F = ma$ per hanging body, one $\tau = I\alpha$ for the wheel, one bridge. The pulley-with-mass archetype runs entirely on this trio.

you want	use	notes
spin-up / brake-down rate	$\alpha = \tau_{\text{net}}/I$ , eq. (1)	inventory ALL external torques, with signs
time, angle, final speed	rotational SUVAT with that $\alpha$	radians throughout; convert rpm first
cord/rim systems	eq. (1) + $F = ma$ + bridge (2)	one law per body, then couple
pivot off the CM	$I = I_{\text{cm}} + Md^2$ first	previous module's shift
body coasting at constant ω	$\tau_{\text{net}} = 0$	no torque needed to maintain spin

Derivation AEarning the law — F = ma summed around the axis

We derive $\tau_{\text{net}} = I\alpha$ from nothing but Newton's second law applied to each particle of a rigid body — thereby earning the dictionary rows $F \to \tau$ and $m \to I$ rather than assuming them. The strategy is the swarm sum of the physical picture: write the tangential second law for one particle, trade its private acceleration for the shared $\alpha$ , convert force to torque, and add over the body. Two things to watch: where the $r^2$ gets minted (Steps 2–3), and why the terrifying sum of internal forces exits the problem entirely (Step 4) — without that cancellation the law would be useless, since no one can catalogue interatomic forces.

Step 1. One particle, tangentially. A mass element $m_i$ at radius $r_i$ feels some net tangential force $F_{t,i}$ — external pushes plus internal pulls from its neighbours — and obeys $F_{t,i} = m_i a_{t,i}$ . Radial forces point through the axis, have zero lever arm, and cannot influence the turning; they are the axle's business, not the rotation's.

Step 2. Trade private for shared. Rigidity gives every particle the same $\alpha$ , and the kinematic bridge converts it: $a_{t,i} = r_i \alpha$ . So $F_{t,i} = m_i r_i \alpha$ — the first factor of $r$ , billed by the demand.

Step 3. Convert to the honest currency. Multiply by the lever arm: $\tau_i = r_i F_{t,i} = m_i r_i^2\,\alpha$ — the second factor of $r$ , billed by the means. Each particle's turning bill is its mass taxed by distance squared, exactly the ingredient of the moment of inertia.

Step 4. Sum the swarm and watch the internal forces die. Summing $\tau_i$ over the body counts internal torques too — but internal forces come in Newton's-third-law pairs: equal, opposite, and acting along the line joining the two particles. Two opposite forces sharing one line of action share one lever arm about the axis, so each pair's torques cancel exactly. The sum keeps only the external contributions:

\tau_{\text{net}}^{\text{ext}} = \left(\sum_i m_i r_i^2\right)\alpha = I\,\alpha

Step 5. Read what was proved. The law is a theorem of $F = ma$ plus rigidity — no new postulate entered. The same cancellation, run on forces instead of torques, is what gave $F_{\text{net}}^{\text{ext}} = Ma_{\text{cm}}$ in the centre-of-mass module: a body can neither shove nor spin itself from inside. And the $I$ that appears is manufactured about the actual rotation axis — which is why an off-centre pivot demands the parallel-axis shift before anything else. Verified: dimensions $\mathrm{kg\,m^2 \times s^{-2}} = \mathrm{N\,m}$ ✓; the massless limit $I \to 0$ gives $\alpha \to \infty$ for any torque — why ideal pulleys transmit tension unchanged ✓; matches the standard construction (NCERT Ch. 7, HC Verma Ch. 10, Halliday–Resnick–Walker Ch. 10). Worked example 1 closes the loop numerically.

Figure

The pull versus the weight — same 25 newtons, different physics

Go deeper · Derivation B · 2 minThe pull versus the weight — why a hanging mass is a different problem

Derivation BThe pull versus the weight — why a hanging mass is a different problem

We prove a distinction that examiners mine every year, and that previews the pulley-with-mass archetype: a cord wound on a flywheel pulled with a steady force $F$ produces a different angular acceleration than the same cord carrying a hanging mass of weight $mg = F$ . The strategy: solve both setups with the same three-equation machinery — the wheel's $\tau = I\alpha$ , the hanging body's $F = ma$ , and the bridge $a = R\alpha$ — and watch the tension refuse to equal the weight. Keep an eye on the physical reason: the hanging mass accelerates downward, and a downward-accelerating body must weigh more than its supporting tension.

Step 1. Setup one — the steady pull. Force $F$ applied tangentially at radius $R$ of a wheel with moment of inertia $I$ . Torque inventory: $\tau = FR$ (the axle's forces pass through the axis: zero lever arm). The law:

\alpha_{\text{pull}} = \frac{FR}{I}

No second body, no bridge needed — the force is guaranteed by the hand that applies it.

Step 2. Setup two — the hanging mass. Replace the hand by a mass $m$ with $mg = F$ . Now the cord's tension $T$ is an unknown: it is whatever the coupled system decides. Wheel: $TR = I\alpha$ . Hanging mass, taking down as positive: $mg - T = ma$ . Bridge: $a = R\alpha$ .

Step 3. Solve the trio. From the wheel equation, $T = I\alpha/R = Ia/R^2$ . Substitute into the mass equation:

mg - \frac{I}{R^2}a = ma \qquad \Longrightarrow \qquad a = \frac{g}{1 + \dfrac{I}{mR^2}}, \qquad \alpha_{\text{hang}} = \frac{a}{R} = \frac{mg\,R}{I + mR^2}

Step 4. Compare. $\alpha_{\text{pull}} = FR/I = mgR/I$ versus $\alpha_{\text{hang}} = mgR/(I + mR^2)$ : the hanging setup is always slower, because the denominator gained the hanging mass's own inertia $mR^2$ . Equivalently, the tension came out as $T = mg\,\dfrac{I}{I + mR^2} < mg$ — the falling mass accelerates, so the cord cannot carry its full weight. Announcing " $T = mg$ " freezes the mass in mid-air and contradicts the motion it is supposed to have.

Step 5. Read the limits, which bracket the physics. $I \to \infty$ (an immovable wheel): $a \to 0$ , $T \to mg$ — the mass just hangs, and the pull-vs-hang distinction disappears because nothing accelerates. $I \to 0$ (a massless wheel): $a \to g$ , $T \to 0$ — free fall, the cord slack in all but name. Every pulley-with-mass problem in the archetype module lives between these brackets. Verified: numbers — wheel $I = 0.40\ \mathrm{kg\,m^2}$ , $R = 0.20\ \mathrm{m}$ , and $F = 25\ \mathrm{N}$ vs hanging $m = 2.5\ \mathrm{kg}$ ( $mg = 24.5 \approx 25\ \mathrm{N}$ at $g = 9.8$ ): pull gives $\alpha = 25 \times 0.2/0.4 = 12.5\ \mathrm{rad/s^2}$ ; hang gives $a = 9.8/(1 + 0.4/(2.5 \times 0.04)) = 9.8/5 = 1.96\ \mathrm{m/s^2}$ , $\alpha = 9.8\ \mathrm{rad/s^2}$ , $T = 2.5(9.8 - 1.96) = 19.6\ \mathrm{N} < 24.5\ \mathrm{N}$ ✓. Matches the treatment of coupled systems in HC Verma Ch. 10; the NCERT flywheel example is Setup one.

Limiting cases

case	result	meaning
$\tau_{\text{net}} = 0$	$\alpha = 0$ : $\omega$ constant	rotational first law — flywheels coast
$I \to 0$ at fixed $\tau$	$\alpha \to \infty$	why ideal (massless) pulleys pass tension unchanged
$I \to \infty$	$\alpha \to 0$	immovable wheel: hanging masses just hang
force through the axis	contributes zero torque	axle reactions never enter the inventory
constant $\tau$	constant $\alpha$ — SUVAT applies	the kinematics module takes the handoff
same $\tau$ on hoop vs disc (equal $M, R$ )	disc spins up twice as fast	$\alpha \propto 1/I$
pull $F$ vs hanging $mg = F$	$\alpha_{\text{hang}} < \alpha_{\text{pull}}$ , $T < mg$	the accelerating mass keeps part of its weight

Common trap"The wheel spins at constant speed, so a steady torque must be driving it — and the cord's tension equals the hanging weight"

Two mark-killers, one root: importing intuitions that rotation does not honour. First: torque pairs with angular acceleration, never with angular velocity. Constant spin means zero net torque — a coasting flywheel, a fan after the motor cuts out (until friction's torque brakes it), the Earth itself. If a real machine needs steady motor torque to hold constant speed, that torque is balancing friction's counter-torque, netting zero: the inventory, not the intuition, decides. Second: in any system where the hanging body actually accelerates, the string tension is strictly less than the weight ( $T = mg - ma$ ) — writing $T = mg$ pins the mass motionless and quietly contradicts the acceleration you are solving for. The same slip wears a rotational costume: assuming the two sides of a massive pulley carry equal tensions. They cannot — a pulley with moment of inertia needs a net torque $(T_1 - T_2)R$ to spin up, so the tensions must differ. "Light pulley" in a question is your licence for one tension; a stated pulley mass revokes it.

Worked examples

1 · Board level — the NCERT flywheel (CBSE/ICSE pattern, 3 marks)

A cord of negligible mass is wound round the rim of a flywheel — a uniform disc of mass $20\ \mathrm{kg}$ and radius $0.20\ \mathrm{m}$ mounted on a frictionless horizontal axle. A steady pull of $25\ \mathrm{N}$ is applied to the cord. Find (a) the angular acceleration, (b) the angular speed after the wheel turns through $10\ \mathrm{rad}$ , (c) the length of cord unwound by then.

Step 1. Moment of inertia (disc, central axis): $I = \tfrac12 MR^2 = \tfrac12 \times 20 \times 0.04 = 0.40\ \mathrm{kg\,m^2}$ .

Step 2. Torque inventory: the pull acts tangentially at the rim, $\tau = FR = 25 \times 0.20 = 5.0\ \mathrm{N\,m}$ ; the axle's forces pass through the axis (zero lever arm) and the axle is frictionless. So (a) $\alpha = \tau/I = 5.0/0.40 = \boxed{12.5\ \mathrm{rad/s^2}}$ .

Step 3. (b) Rotational SUVAT from rest: $\omega^2 = 2\alpha\theta = 2 \times 12.5 \times 10 = 250$ , so $\omega = \boxed{15.8\ \mathrm{rad/s}}$ (about 151 rpm).

Step 4. (c) Cord unwound $= R\theta = 0.20 \times 10 = \boxed{2.0\ \mathrm{m}}$ — the integrated bridge, $s = R\theta$ . Marking-scheme sentences: "torque = force × perpendicular distance from the axis" and "the axle reaction exerts no torque about the axis."

2 · Board level — braking a ceiling fan (ICSE/NEET pattern)

A ceiling fan has moment of inertia $1.2\ \mathrm{kg\,m^2}$ about its axle and spins at $240\ \mathrm{rpm}$ . When switched off, friction at the bearings applies a steady retarding torque of $3.0\ \mathrm{N\,m}$ . Find (a) how long the fan takes to stop, (b) how many revolutions it makes while stopping.

Step 1. Convert first: $\omega_0 = 240 \times \dfrac{2\pi}{60} = 8\pi = 25.1\ \mathrm{rad/s}$ .

Step 2. Deceleration: $\alpha = -\tau/I = -3.0/1.2 = -2.5\ \mathrm{rad/s^2}$ .

Step 3. (a) Time: $t = \omega_0/|\alpha| = 25.1/2.5 = \boxed{10.1\ \mathrm{s}}$ .

Step 4. (b) Angle: $\theta = \dfrac{\omega_0^2}{2|\alpha|} = \dfrac{(25.1)^2}{5.0} = 126\ \mathrm{rad}$ , so revolutions $= 126/2\pi = \boxed{20}$ . Sanity: average spin rate $120$ rpm for $10$ s $= 20$ revolutions ✓.

3 · Competitive level — one torque, three bodies (NEET/JEE-Main pattern, ~2 minutes)

Equal torques act on a hoop, a solid disc, and a solid sphere, each of the same mass $M$ and radius $R$ , each spinning about its own central axis from rest. Rank their angular accelerations, and find the ratio of times each takes to reach the same angular speed $\omega$ .

Step 1. $\alpha = \tau/I$ with the catalogue: hoop $I = MR^2$ , disc $\tfrac12 MR^2$ , sphere $\tfrac25 MR^2$ . So $\alpha_{\text{hoop}} : \alpha_{\text{disc}} : \alpha_{\text{sphere}} = 1 : 2 : \tfrac52$ — the sphere snaps up fastest, the hoop slowest: $\boxed{\alpha_{\text{sphere}} > \alpha_{\text{disc}} > \alpha_{\text{hoop}}}$ .

Step 2. Times to reach $\omega$ : $t = \omega/\alpha \propto I$ , so $t_{\text{hoop}} : t_{\text{disc}} : t_{\text{sphere}} = 1 : \tfrac12 : \tfrac25 = \boxed{5 : 2.5 : 2}$ (or $10 : 5 : 4$ in integers). Timing note: ~90 seconds — pure catalogue recall plus one inversion. The trap is ranking by $I$ and forgetting to invert for $\alpha$ ; options listing the hoop as fastest are pricing exactly that. The deeper reading: at equal torque, mass distribution alone decides the spin-up race — the concept module's rim test, now dynamical.

4 · Competitive level — two-phase spin: drive, then coast to rest (JEE-Main stretch, ~4 minutes)

A grindstone of moment of inertia $1.6\ \mathrm{kg\,m^2}$ starts from rest. A motor applies a torque of $10\ \mathrm{N\,m}$ for $8.0\ \mathrm{s}$ against a constant bearing-friction torque of $2.0\ \mathrm{N\,m}$ ; the motor is then switched off. Find (a) the maximum angular speed, (b) the time to stop after switch-off, (c) the total revolutions from start to stop.

Step 1. Phase one inventory: $\tau_{\text{net}} = 10 - 2 = 8.0\ \mathrm{N\,m}$ , so $\alpha_1 = 8.0/1.6 = 5.0\ \mathrm{rad/s^2}$ . (a) $\omega_{\max} = \alpha_1 t = 5.0 \times 8.0 = \boxed{40\ \mathrm{rad/s}}$ .

Step 2. Phase one angle: $\theta_1 = \tfrac12 \alpha_1 t^2 = \tfrac12 \times 5.0 \times 64 = 160\ \mathrm{rad}$ .

Step 3. Phase two: only friction remains, $\alpha_2 = -2.0/1.6 = -1.25\ \mathrm{rad/s^2}$ . (b) $t_2 = 40/1.25 = \boxed{32\ \mathrm{s}}$ .

Step 4. Phase two angle: $\theta_2 = \dfrac{\omega_{\max}^2}{2|\alpha_2|} = \dfrac{1600}{2.5} = 640\ \mathrm{rad}$ . (c) Total $= 800\ \mathrm{rad} = 800/2\pi = \boxed{127\ \text{revolutions}}$ . Timing note: ~3.5 minutes. The whole question is the torque inventory changing between phases: driving phase $10 - 2$ , coasting phase $-2$ alone. The planted error is using $10\ \mathrm{N\,m}$ in phase one (ignoring friction) — it inflates $\omega_{\max}$ to 50 and everything downstream. Two-phase spin problems are the rotational twin of the driven-then-braked kinematics questions from U2, and the same bookkeeping discipline wins them.

Exam traps

One axis for everything. $\tau$ , $I$ and $\alpha$ in equation (1) must all refer to the same axis. A pivot at the rod's end demands $I = \tfrac13 ML^2$ , not $\tfrac1{12}ML^2$ — the parallel-axis shift comes before the dynamics.

Inventory means all of them. Friction at the bearings, the brake pad, the second cord — every external torque enters with its sign. The single most common stretch-question error is running phase-one arithmetic with the driving torque alone.

Radians or nothing. Convert rpm to $\mathrm{rad/s}$ before any SUVAT step ( $N$ rpm $= N \times 2\pi/60\ \mathrm{rad/s}$ ). Mixing revolutions into $\omega^2 = 2\alpha\theta$ is a silent factor- $2\pi$ leak.

Torque pairs with $\alpha$ , not $\omega$ . "What torque maintains constant spin?" — zero net. Constant-speed machines balance drive against friction; the exam statement "net torque is required for uniform rotation" is always false.

$T \neq mg$ on an accelerating cord. The tension is an unknown; solve the coupled system. Equivalently: massive pulley ⇒ two different tensions — $(T_1 - T_2)R$ is precisely what spins it.

Axle forces never enter. Reactions at the axis have zero lever arm. If your torque inventory contains "force from the bearing," the pivot choice already killed it.

The bridge needs no slipping. $a = R\alpha$ holds only while the cord grips the rim (or the belt grips the wheel). A slipping belt decouples the two worlds — the question will say "does not slip" when you may weld them.

Origins · 1 minIsaac Newton and how this was discovered

Isaac Newton

1643–1727 · Cambridge

There is no rotation chapter in the Principia — and that is exactly what makes $\tau = I\alpha$ remarkable. Newton's 1687 masterwork states the laws for particles: point masses, forces, accelerations. Every result of this module is those particle laws, summed — the derivation above is a miniature of how the Principia's framework conquered territory its author never explicitly mapped. Newton himself came heartbreakingly close: his third law's action–reaction pairs are precisely what make the internal torques cancel, and his analysis of the Moon's motion is torque-free angular momentum in embryo. The explicit rotational law waited nearly eight decades for Euler, who named the moment of inertia and wrote rigid-body dynamics as differential equations in 1765 — earning the law its occasional textbook name, "Euler's equation for fixed-axis rotation." The honest attribution is the one the derivation shows: the physics is Newton's, the bookkeeping is Euler's, and the exam question is both of theirs. Newton spent his later decades as Master of the Royal Mint and President of the Royal Society — running, among other things, the torque-driven machinery of coin presses whose dynamics his laws, summed around an axis, describe.

History

1673

— Huygens's compound-pendulum analysis solves a rotational dynamics problem before the general law exists.

1687

— Newton's Principia states the particle laws; their swarm sum is this module's derivation.

1765

— Euler's Theoria motus corporum solidorum writes rigid-body dynamics in mature form: τ = Iα and far beyond.

1884

— James Thomson coins "torque"; the machine age adopts the law as the currency of every rotating shaft.

1900

s — dynamometers make τ and ω the measured certificates of every engine and motor built since.

Going further

Texts. NCERT Physics Class 11, Ch. 7 (the flywheel example and τ = Iα) · HC Verma, Concepts of Physics Vol. 1, Ch. 10 · Halliday–Resnick–Walker Ch. 10. For the same law with the full vector machinery: Torque & Fixed-Axis Dynamics (UG deep dive).

Provenance. All results are standard. A verifier should recompute: the flywheel set ( $I = 0.40\ \mathrm{kg\,m^2}$ , $\alpha = 12.5\ \mathrm{rad/s^2}$ , $\omega = \sqrt{250} = 15.8\ \mathrm{rad/s}$ , cord $2.0$ m); the fan ( $\omega_0 = 8\pi$ , $t = 10.1$ s, $20$ rev); Example 3's ratios ( $1 : 2 : 2.5$ and times $10 : 5 : 4$ ); Example 4's two phases ( $40\ \mathrm{rad/s}$ , $32$ s, $160 + 640 = 800$ rad $= 127$ rev); the pull-vs-hang derivation ( $12.5$ vs $9.8\ \mathrm{rad/s^2}$ , $T = 19.6\ \mathrm{N} < 24.5\ \mathrm{N}$ ).

Prerequisites. Torque & the Turning Effect · Moment of Inertia: the Concept · Angular Displacement, Velocity & Acceleration. Leads to. Rotational KE & Work Done by Torque · Pulley With Mass (archetype) · Torque & Fixed-Axis Dynamics (UG deep dive).

Is τ = Iα a new law of nature, or does it follow from what I already know?

It follows — completely. Newton's laws govern particles; a rigid body is particles welded together. Write $F = ma$ tangentially for each particle, convert each equation to torque by multiplying with the particle's radius, and add: rigidity (one shared $\alpha$ ) lets the sum collapse, the internal forces cancel in third-law pairs, and $\tau_{\text{net}} = (\sum mr^2)\alpha$ emerges with the moment of inertia self-assembled. Nothing was postulated. That derivation-not-decree status matters practically: it tells you exactly when the law holds (rigid body, fixed axis, torques and $I$ about that axis) and exactly what to do when conditions change — go back to the particle laws and re-sum, which is how the rolling and archetype modules extend it.

Why do the internal forces really cancel? The atoms in a spinning body pull on each other constantly.

They cancel in pairs, by Newton's third law plus one geometric fact. Atom A pulls atom B; B pulls A back with the exact opposite force; and both forces act along the same line — the line joining the two atoms. Two equal-opposite forces on one line of action have one common lever arm about any axis, so their torques are equal and opposite: each pair nets exactly zero, and the sum over millions of pairs is still zero, however ferocious the individual forces. This is why no body can spin itself up from inside (conservation of angular momentum in embryo — two modules ahead), and why the law is usable at all: the unknowable internal chaos exits, leaving an inventory you can actually write down.

How is solving rotation problems different from the F = ma problems I know?

Structurally, not at all — that is the dictionary's point. The routine transliterates: draw the body, choose the axis (rotation's version of choosing axes), inventory external torques with signs (instead of forces), find $I$ about that axis (instead of $m$ ), divide for $\alpha$ (instead of $a$ ), then run SUVAT in $\theta, \omega, \alpha$ . The genuinely new bookkeeping items are three: $I$ depends on the axis (Steiner-shift before dynamics), forces through the axis vanish from the inventory (a simplification $F = ma$ never offered), and mixed systems need the bridge $a = R\alpha$ as an extra equation. Students who treat rotation as a foreign language struggle; students who treat it as the same language with swapped nouns finish early.

What exactly does "fixed axis" exclude — and what happens with a rolling wheel?

"Fixed axis" means the rotation axis neither moves nor changes direction: axles, hinges, pivots. It excludes tumbling (the axis direction wanders — genuinely harder, a university topic) and, at first sight, rolling (the axis translates with the wheel). Rolling is rescued by a happy theorem: $\tau = I\alpha$ remains valid about the centre of mass even while the CM accelerates — the centre of mass is that privileged. So a rolling body is handled by $F = ma$ for the centre plus $\tau_{\text{cm}} = I_{\text{cm}}\alpha$ about it, coupled by the rolling condition — exactly the machinery of the rolling module two steps ahead. What you may not do is apply the law about an arbitrary accelerating point; the CM (and the fixed axle) are the safe choices at this level.

The cord problems keep saying "the cord does not slip." What breaks if it slips?

The bridge. $a = R\alpha$ encodes "the cord and the rim move together where they touch" — every centimetre of cord that pays out turns the wheel by exactly one rim-centimetre. If the cord (or belt) slips, the two motions disconnect: the wheel can spin without paying out cord, and the system gains an extra degree of freedom that the three-equation method cannot close — you would need the friction physics of the slipping contact instead. Exams signal the regime explicitly: "does not slip" or "sufficiently rough" licenses the bridge; "smooth peg" or "slips freely" revokes it (a cord over a smooth peg just slides — the peg spins not at all and merely redirects the tension). Reading that one phrase correctly decides which problem you are actually solving.

Why does a fan keep turning after the switch is off — and what finally stops it?

Nothing needs to keep it turning — that is the rotational first law. With the motor off, if the bearings were perfect and the air absent, the net torque would be zero and the blades would coast at constant $\omega$ forever. What actually stops the fan is the negative torque inventory that never sleeps: bearing friction and air drag on the blades. Their combined retarding torque gives $\alpha = -\tau_{\text{fric}}/I$ , and the fan winds down over the seconds-to-minutes your bedroom experiment shows (Example 2 is precisely this calculation). The exam point: "no driving torque" and "no torque" are different statements. The everyday point: rotation is never self-sustaining against friction, but it is never self-stopping without it — flywheel energy storage works by making the friction inventory almost empty (magnetic bearings, vacuum chambers) so the coast lasts hours.

When the pulley has mass, where does the "lost" tension difference go?

Into the pulley's own spin-up — the books balance exactly. With a massive pulley, the cord's two sides pull with $T_1$ (heavy side) and $T_2$ (light side), and the difference provides the torque: $(T_1 - T_2)R = I\alpha$ . Nothing is lost; a third body joined the system and demanded its share of the driving force. For a uniform disc pulley the effect has a beautifully simple accounting: substituting $I = \tfrac12 MR^2$ and the bridge turns the torque equation into $T_1 - T_2 = \tfrac12 M a$ — the pulley behaves like an extra half-its-mass of ordinary inertia hidden in the cord. That single line is the key that unlocks the entire Pulley With Mass archetype, where the full variant family (single mass, Atwood, table systems) is drilled to exam speed.