TORQUE AND ROTATIONAL DYNAMICS

Pick any door in the building you are in and look at where the handle sits. Never next to the hinge. Always out at the far edge, as far from the pivot as the carpenter could put it. That placement is not aesthetic. It is a physics decision, and it is older than physics itself.

Push on a door near the hinge and it barely budges. Push on the same door at the handle — same force, same direction — and it swings open with ease. The force is identical. The geometry is not. What matters for rotation is not the size of the force but the size of the force times the distance from the pivot, and the handle sits where that product is largest.

That product has a name: torque. It is the rotational analogue of force, and once you have it, almost everything an extended, rotating object does — balancing, spinning up, rolling, toppling — falls out as an application of the same short equation. This topic introduces torque, derives Newton's second law for rotation, and prepares the ground for the three topics that come after, all of which live and die by these formulas.

Torque is a simple idea dressed in vector notation. A force F applied at a point r from a chosen pivot tends to rotate the body about that pivot. How much? It depends not just on the force but on where and in what direction the force is applied:

The cross product ensures that only the component of F perpendicular to r contributes. A force pointing directly along the lever arm does zero torque — it tries to slide the lever through the pivot rather than rotate around it. A force perpendicular to the arm delivers the full F·r. Any intermediate angle gives F·r·sin θ.

The quantity r·sin θ is called the moment arm of the force: the perpendicular distance from the force's line of action to the pivot. Torque is force times moment arm. That is the practical formulation. If you can eyeball the perpendicular distance from a force's line to the pivot, you can read off the torque directly. A door closer spring anchored near the hinge has a short moment arm. The same spring at the handle has a long one, and turns a modest force into a serious tug.

Torque is a vector, pointing along the rotation axis, direction set by the right-hand rule. For planar problems (all forces in one plane) it collapses to a signed scalar: counter-clockwise positive, clockwise negative. Its units are newton-metres — dimensionally the same as joules, but torque and energy are not the same quantity, and the overlap is a bookkeeping coincidence we will earn back in section 7.

Adjust the masses and their distances and watch the rod tip. The condition for balance is m_L·r_L = m_R·r_R — equal torques about the pivot, opposite signs. Double one arm and halve that mass, and the rod still sits level. That is the lever law, stated by Archimedes in the third century BCE and captured in his most famous boast.

Archimedes, sitting in Syracuse in 250 BCE, wrote: Give me a place to stand, and I shall move the Earth. He was not bluffing. His On the Equilibrium of Planes gives the first rigorous proof of the law of the lever, and his argument — purely geometric — is still essentially the proof we teach today. Two unequal weights balance across a fulcrum when their distances from the fulcrum are inversely proportional to their masses. It is the oldest statement of a mechanical law in the form we still use.

The power of a lever is that it trades distance for force. To lift a 500 kg stone against gravity you need 5000 N upward. If you cannot muster 5000 N, the lever rearranges the geometry for you: pivot the stone over a fulcrum, stand at the far end of a long pole, and apply your 500 N at ten times the distance. Torques balance — 500 N × 10 m = 5000 N × 1 m — and the stone lifts. You move your end ten metres for every metre the stone rises. The total work is the same. What the lever bought you is access to the job at all.

That logic runs through almost all pre-industrial technology: the wheelbarrow (load between pivot and effort), the claw hammer (pivot between effort and load), the human forearm (effort between pivot and load), scissors, pliers, bolt cutters, tweezers, fishing rods, crowbars. Each trades force for distance or distance for force, according to the ratio of moment arms. Mechanical advantage is a pure consequence of the geometry of torques — no muscle required, only leverage.

For a rigid body rotating about a fixed axis, the rotational analogue of F = m·a is

where τ is the net external torque about the axis, I is the moment of inertia about that axis, and α = d²θ/dt² is the angular acceleration. The units check: N·m on the left, (kg·m²)·(rad/s²) on the right. Same thing.

The proof is almost a one-liner. A point mass at radius r from the axis obeys F_⊥ = m·a_⊥ = m·r·α. Multiply both sides by r: r·F_⊥ = (m·r²)·α, which is τ = I·α for a single particle. Sum over every particle in a rigid body and both sides are additive. The formula above falls out.

Three everyday problems run on this one line:

A flywheel at rest is hit with a steady torque. Angular velocity climbs linearly — that is α constant, exactly what τ = I·α predicts for a fixed torque. The torque clicks off and the wheel coasts at whatever ω it reached. The structure is identical to a block on a frictionless floor being pushed: the graph of v versus t is a straight ramp, then flat. Rename the axes and it is the same physics.

A body in static equilibrium is not accelerating and not angularly accelerating. For an extended body that means two conditions, both of which must hold at the same time:

Both are non-trivial. A body with Σ F = 0 but Σ τ ≠ 0 will not translate, but it will start to spin. A body with Σ τ = 0 but Σ F ≠ 0 will slide without spinning. Only when both sums vanish does the body truly sit still.

Which pivot? A surprising and useful fact: if Σ F = 0, the net torque is the same about every point. So for a body already in force equilibrium, you can pick any convenient pivot when you check rotational equilibrium. Engineers usually pick a point where an unknown force acts, which kills that unknown from the torque equation and leaves fewer things to solve for.

This pair of conditions is the daily bread of structural analysis. A ladder leaning against a frictionless wall stays up only while the friction at the floor supplies enough horizontal force and while the torques about the base sum to zero; push the top much further out and the torque equation fails before the force equation does, which is why ladders slip at the bottom, not topple from the top. A crane with a load, a bookshelf bracket, a weighing scale, a person standing on one leg — all of them are just Σ F = 0 and Σ τ = 0 written down twice and solved.

A wheel rolling cleanly on a road is doing two things at once: its center of mass translates forward, and the wheel rotates about that center. Rolling without slipping is the constraint that locks the two motions together:

where R is the wheel's radius. The point of the tyre in contact with the road is instantaneously at rest — that is what "no slip" means — and the rest of the wheel is doing a mix of rotation and translation around that instantaneous pivot.

Differentiate both sides: a_CM = α·R. For a wheel of mass m and moment of inertia I pushed by a force F at the axle, combining F = m·a_CM with τ = I·α gives

Less than the F/m you'd get for a sliding block of the same mass. Some of the force is being spent spinning the wheel up rather than shoving its center forward.

For a solid disk, I = ½·m·R², which gives a = (2/3)·F/m. For a hollow hoop, I = m·R², which gives a = ½·F/m. Same mass, same radius, same force — the hoop accelerates slower because all its mass sits at maximum radius and costs more to spin.

This is the money shot of rotational dynamics: race a solid cylinder and a hollow one down the same ramp, starting together from rest. Both have the same weight. Both feel the same gravitational pull. The solid one wins every single time, and it is not close. The hollow cylinder's mass is all at the rim, its moment of inertia is twice as large, and a larger fraction of its gravitational energy is siphoned into rotational kinetic energy instead of forward motion. The same logic explains why flywheels for energy storage are built with mass at the rim (large I, lots of energy stored per unit ω) while rotors meant to spin up fast are built with mass near the axis (small I, easy to accelerate). Wheel design is always a choice between those two.

A force acting through a distance does work: W = F·d. A torque acting through an angle does rotational work, and the formula is the direct translation:

for a constant torque through angle θ (in radians). For a varying torque, W = ∫τ·dθ. The factor-of-2π in a full turn — the mystery reason torque and energy share units of N·m — is exactly what makes this work out. One full revolution converts a torque of τ newton-metres into 2π·τ joules of work.

Differentiate both sides with respect to time. Rotational power is

Torque times angular velocity. This is the equation an automotive engineer lives in. A car engine produces a torque curve — how much twist the crankshaft delivers at each rpm — and the power at any rpm is that torque times the angular velocity. The reason cars have gearboxes is that an engine has a narrow band of rpm where torque is highest, but the wheels need a wide range of speeds. Gears trade rotational speed for torque with the same r·F geometry we started this topic with. Low gear: the engine spins fast, the wheels spin slow, and the wheel torque is large — good for climbing. High gear: the engine spins slow relative to the wheels, and torque at the wheel is small but ω is large — good for cruising. Same power in either case, differently distributed between τ and ω.

The same equation runs every rotating machine in the economy: turbines, drills, fans, mills, lathes, helicopter rotors, washing-machine drums, bicycle cranks, wind turbines. James Watt rated his steam engines in horsepower — defined as 550 ft·lbf/s, which is F·v for a pulling horse but just as well τ·ω for the engine shaft. The numbers have modernised; the equation has not.

We now have a working framework for rigid-body rotation. Torque plays the role of force. Moment of inertia plays the role of mass. Angular acceleration plays the role of linear acceleration. τ = I·α is the rotational F = m·a. Static equilibrium requires Σ F = 0 and Σ τ = 0. Rolling without slipping ties linear and rotational motion with v = ω·R. Work and power get new forms, W = τ·θ and P = τ·ω, with the same physical content as before.

One quantity keeps sliding into every formula without ever being interrogated: the moment of inertia I. Why is a solid disk's moment of inertia ½·m·R² while a hoop's is m·R²? Why does a sphere beat a cylinder down a ramp? Why does a pendulum's period depend on where its pivot sits relative to its center? Those are all questions about how mass distributes itself around an axis, and they are the subject of the next topic. In FIG.10 we'll build the full moment-of-inertia toolkit — integrals for standard shapes, the parallel-axis theorem, and the principal axes of a general rigid body. After that, rotation analysis stops being a set of tricks and becomes something you can calculate from scratch.