Modelling in Mechanics

Modelling an object as a $\text{particle}$ means treating all its mass as concentrated at a single point. Its size and shape are ignored, so it cannot rotate and air resistance can be neglected as there is no surface area.

This is reasonable when the object is small compared with the distances involved, or when rotation is irrelevant. It fails when the question concerns turning — you cannot take moments about a particle, which is why the Moments chapter models objects as $\text{rigid bodies}$ instead.

A $\text{rod}$ is a rigid body modelled as one-dimensional: it has length but no thickness. A $\text{uniform}$ rod has its mass evenly distributed, so its weight acts at the midpoint — the assumption that makes moments problems solvable without integration.

A $\text{smooth}$ surface exerts no friction, so the only contact force is the normal reaction perpendicular to the surface. A $\text{rough}$ surface does have friction, and its magnitude is governed by the coefficient of friction.

A $\text{light}$ string has negligible mass, which means the tension is the same throughout it. An $\text{inextensible}$ string does not stretch, which means connected objects share the same acceleration — the assumption that makes connected-particle problems work.

A $\text{smooth pulley}$ has no friction at its axle, so the tension is unchanged as the string passes over it. Without that assumption the tensions on the two sides would differ and the problem would need the pulley’s moment of inertia.

$\text{Air resistance is negligible}$ removes any drag force, so a projectile has constant horizontal velocity and constant vertical acceleration $g$. It is reasonable for dense compact objects at modest speeds, and unreasonable for parachutes, feathers or anything moving fast.

$\text{Gravity is uniform}$ takes $g$ as constant, which holds well near the Earth’s surface but fails over large altitude changes.

$\text{A beam is rigid}$ means it does not bend, so forces act where they are applied. $\text{A surface is horizontal}$, or a $\text{plane is inclined at a fixed angle}$, remove complications from uneven ground.

Questions often ask how a model could be improved. A good refinement removes the assumption that matters most in that context and says what would have to be added.

For a projectile over a long distance, include air resistance. For a heavy chain rather than a light string, account for the string’s mass so the tension varies. For an object that visibly rotates, model it as a rigid body rather than a particle.

Refinements have a cost, and saying so shows understanding: a more realistic model usually needs more data — a drag coefficient, a mass distribution — and more difficult mathematics.