What Happens In An Elastic Collision

You’ve seen it a hundred times. Practically speaking, two billiard balls crack together, and the cue ball stops dead while the object ball rockets away with the exact same speed. It looks like magic. It looks like the first ball just handed its soul over to the second one.

That’s an elastic collision. Or at least, it’s the textbook version of one.

Real life is messier. But the physics? The physics is surprisingly clean — if you know where to look.

What Is an Elastic Collision

An elastic collision is any interaction between two objects where the total kinetic energy of the system is conserved. That’s the definition. Full stop.

But “kinetic energy conserved” is a slippery phrase if you don’t unpack it. Consider this: it doesn’t mean each object keeps its own energy. Consider this: it means the sum of their kinetic energies before the crash equals the sum after. Energy gets swapped around, shuffled between the participants, but none of it vanishes into heat, sound, or permanent deformation.

Contrast that with an inelastic collision — a car crumpling against a guardrail, a lump of clay hitting the floor. On the flip side, gone. There, kinetic energy bleeds out. The total energy of the universe is still conserved, obviously, but the mechanical* kinetic energy? Turned into heat, noise, broken molecular bonds.

In a perfectly elastic collision, two things happen simultaneously:

So total momentum is conserved (true for all closed-system collisions, elastic or not). 2. Total kinetic energy is conserved (the special part).

That second constraint is the one that makes the math work out in that satisfying, deterministic way.

The coefficient of restitution

Physicists like to quantify “bounciness” with a number called the coefficient of restitution, usually written as e. It’s the ratio of relative speed after the collision to relative speed before.

For a perfectly elastic collision, e = 1.
Consider this: maybe 0. A superball on concrete? Also, you can hit 0. Consider this: a golf ball off a driver face? On the flip side, 9. 8. Closer to 0.For perfectly inelastic (they stick together), e = 0.
Which means steel ball bearings? Real objects live somewhere in between. 99 if the surfaces are clean and hard.

But e = 1 is an idealization. A limit. Like a frictionless surface or a massless pulley — it doesn’t exist in nature, but it’s the baseline that lets us understand everything else.

Why It Matters / Why People Care

You might wonder: if perfectly elastic collisions don’t really exist, why do we spend weeks on them in physics class?

Because they’re the simplest* model that still captures the core mechanics of momentum transfer. And they show up in surprising places.

Gas molecules and pressure

The kinetic theory of gases treats molecules as tiny, hard spheres undergoing perfectly elastic collisions with each other and the container walls. Plus, no energy loss. It’s why pressure exists. Every time a nitrogen molecule bounces off the wall of your tire, it delivers a tiny impulse. Billions of those impulses per second? That assumption — e = 1 — is what lets us derive the ideal gas law from first principles. That’s 35 PSI.

If those collisions weren’t essentially elastic, gases would cool down spontaneously just by sitting in a container. The atmosphere would collapse. They’d lose kinetic energy to internal vibrations every time they bumped. So yeah, it matters.

Particle physics

At the other end of the scale, high-energy particle collisions in accelerators are often treated as elastic — or at least, the elastic scattering* component is what physicists measure to probe the structure of protons and neutrons. When two protons glance off each other at near light speed without breaking apart, that’s elastic scattering. The math gets relativistic, but the conservation laws are the same.

Engineering and safety

Car bumpers used to be rigid. Plus, a rigid bumper creates a near-elastic collision with whatever you hit — the other car, a pole, a pedestrian. Still, the vehicle stops fast. In practice, the idea was: protect the bodywork. The occupants don’t*. Physics had other ideas. Their internal organs keep moving until the seatbelt (or the dashboard) stops them.

Modern crumple zones are deliberately* inelastic. So naturally, they absorb kinetic energy by deforming plastically. Which means they sacrifice the car to save the people. Understanding elastic collisions is step one in designing that tradeoff.

How It Works

Let’s get into the mechanics. I’ll start with the one-dimensional case because it’s the only one you can solve on a napkin. Then I’ll show why two dimensions makes life harder — and more interesting.

One dimension: the algebra

Two objects. Masses m₁ and m₂. That's why initial velocities u₁ and u₂. Final velocities v₁ and v₂.

Conservation of momentum:

m₁u₁ + m₂u₂ = m₁v₁ + m₂v₂

Conservation of kinetic energy:

½m₁u₁² + ½m₂u₂² = ½m₁v₁² + ½m₂v₂²

Two equations, two unknowns (v₁ and v₂). Solvable. The standard result (derived by subtracting the momentum equation from the energy equation after some factoring) gives the relative velocity reversal:

v₂ - v₁ = u₁ - u₂

The relative speed of approach equals the relative speed of separation. That’s the e = 1 condition, stated in velocities.

From there, the final velocities are:

v₁ = (m₁ - m₂)/(m₁ + m₂) u₁ + (2m₂)/(m₁ + m₂) u₂
v₂ = (2m₁)/(m₁ + m₂) u₁ + (m₂ - m₁)/(m₁ + m₂) u₂

These formulas are worth memorizing if you solve collision problems regularly. Plus, the center-of-mass frame* is faster and less error-prone. But honestly? More on that in the tips section.

Special cases that build intuition

Equal masses, one initially at rest
m₁ = m₂, u₂ = 0.
Plug it in: *v₁

= 0, v₂ = u₁.
Think about it: the projectile stops dead. This is Newton’s cradle. The target takes off with the exact same velocity. Think about it: this is billiards (ignoring spin). The momentum and energy transfer completely.

Massive projectile, light target at rest
m₁ ≫ m₂, u₂ = 0.
v₁ ≈ u₁, v₂ ≈ 2u₁.
The heavy object barely notices. The light one gets flung forward at twice the projectile’s speed. A golf club hitting a ball. A neutron moderating in a reactor — though there the masses are comparable, so the factor is less than two.

Light projectile, massive target at rest
m₁ ≪ m₂, u₂ = 0.
v₁ ≈ −u₁, v₂ ≈ 0.
The projectile bounces back with nearly the same speed. The wall (or planet) doesn’t budge. A tennis ball off a brick facade. An alpha particle off a gold nucleus — Rutherford’s scattering, essentially elastic at large impact parameters.

Equal masses, head-on with equal and opposite velocities
m₁ = m₂, u₂ = −u₁.
v₁ = −u₁, v₂ = u₁.
They exchange velocities. Each reverses direction. Symmetry demands it.

Two dimensions: the geometry

In 2D, momentum is a vector. Conservation gives two scalar equations (x and y). Kinetic energy gives one scalar equation. That's why three equations. But you have four unknowns (v₁ₓ, v₁ᵧ, v₂ₓ, v₂ᵧ). The system is underdetermined — impact parameter matters.

If you found this helpful, you might also enjoy chord and arc of a circle or number of protons neutrons and electrons in beryllium.

The collision isn’t defined by masses and initial velocities alone. You need the geometry: where the centers are at contact. Velocities don’t change. That said, that defines the line of centers* — the direction along which the impulse acts. Perpendicular to that line, nothing happens. Along that line, it’s a 1D elastic collision.

The algorithm:

Find the unit vector n̂ along the line of centers at impact (from center 1 to center 2).
Decompose initial velocities into components parallel and perpendicular to n̂:
- u₁∥ = u₁ ⋅ n̂, u₁⊥ = u₁ − *u₁∥*n̂
- Same for u₂
Perpendicular components are unchanged: v₁⊥ = u₁⊥, v₂⊥ = u₂⊥
Parallel components follow the 1D formulas with u₁∥, u₂∥ as inputs.
Reconstruct: v₁ = *v₁∥*n̂ + v₁⊥, v₂ = *v₂∥*n̂ + v₂⊥

At its core, why billiards players obsess over cut angles. A head-on shot stops the cue ball. Still, the cue ball’s final direction depends entirely on the impact parameter. A glancing blow sends it forward at a tangent. The object ball always moves along the line of centers.

The center-of-mass frame: the physicist’s shortcut

Transform to the frame where total momentum is zero. In this frame:

p₁ = −p₂ at all times.
Kinetic energy is K = p₁²/(2m₁)* + p₂²/(2m₂)* = p²/2μ*, where μ = m₁m₂/(m₁+m₂)* is the reduced mass.
Elastic collision means |p| is conserved. Only the direction* of p changes.

The scattering angle θ in the CM frame is the free parameter (determined by impact parameter). And once you know θ, you rotate the momentum vector. Then transform back to the lab frame.

This frame makes two things obvious:

Think about it: **Speeds don’t change in the CM frame. Consider this: ** Only directions. The collision is a pure rotation in momentum space.
The lab-frame kinematics are just vector addition. v₁ = V_cm + p/m₁, v₂ = V_cm − p/m₂, where V_cm = (p₁+p₂)/(m₁+m₂) is constant.

For equal masses, V_cm = (u₁+u₂)/2. The final velocities are symmetric about V_cm. They always leave at 90° to each other in the lab

The right‑angle result for equal masses

When (m_{1}=m_{2}=m), the centre‑of‑mass velocity is simply the average of the two initial lab velocities, [ \mathbf V_{\rm cm}= \frac{\mathbf u_{1}+\mathbf u_{2}}{2}. Here's the thing — ] In the centre‑of‑mass frame the two particles approach with equal and opposite momenta (\pm\mathbf p). After an elastic collision the magnitude of (\mathbf p) is unchanged, but its direction rotates by some angle (\theta). Day to day, transforming back to the lab, [ \mathbf v_{1}= \mathbf V_{\rm cm}+ \frac{\mathbf p}{m},\qquad \mathbf v_{2}= \mathbf V_{\rm cm}- \frac{\mathbf p}{m}. ] Subtracting the two expressions gives [ \mathbf v_{1}-\mathbf v_{2}= \frac{2\mathbf p}{m}. In practice, ] But (\mathbf p) is orthogonal to (\mathbf V_{\rm cm}) only when the impact parameter is such that the line of centres is perpendicular to the relative velocity. Plus, in the generic case the two final velocities satisfy [ (\mathbf v_{1}-\mathbf V_{\rm cm})\cdot(\mathbf v_{2}-\mathbf V_{\rm cm})=0, ] which is precisely the statement that the vectors from the centre‑of‑mass to each particle are perpendicular after the collision. Consequently the angle between (\mathbf v_{1}) and (\mathbf v_{2}) in the laboratory frame is always (90^{\circ}), independent of the initial speeds or the impact parameter. This elegant result underlies the classic “corner pocket” shot in billiards: after a perfectly elastic, equal‑mass collision, the cue ball and the object ball depart at right angles.

General masses and the scattering angle

When the masses differ, the centre‑of‑mass frame still provides a clean picture, but the final angle is no longer fixed. Day to day, let the reduced mass be (\mu=m_{1}m_{2}/(m_{1}+m_{2})). Still, in the CM frame the two momenta are (\pm\mathbf p) with (|\mathbf p|=\mu,u_{\rm rel}), where (u_{\rm rel}=|\mathbf u_{1}-\mathbf u_{2}|). After collision the momentum rotates by (\theta); the lab‑frame velocities become [ \mathbf v_{1}= \mathbf V_{\rm cm}+ \frac{\mathbf p}{m_{1}},\qquad \mathbf v_{2}= \mathbf V_{\rm cm}- \frac{\mathbf p}{m_{2}}. ] The angle (\phi) between (\mathbf v_{1}) and (\mathbf v_{2}) follows from the dot product: [ \cos\phi= \frac{(\mathbf V_{\rm cm}+ \mathbf p/m_{1})\cdot(\mathbf V_{\rm cm}- \mathbf p/m_{2})} {|\mathbf V_{\rm cm}+ \mathbf p/m_{1}|;|\mathbf V_{\rm cm}- \mathbf p/m_{2}|}. ] Because (\mathbf p) can point in any direction (set by the impact parameter), (\phi) varies continuously from a near‑head‑on configuration (small (\phi)) to a grazing hit (large (\phi)). In the limiting case (m_{1}\gg m_{2}) (a heavy cue ball striking a light target), the target can be ejected at nearly twice the cue‑ball speed while the cue ball’s direction changes only slightly—a classic “mass‑ratio” effect seen in Newton’s cradle.

Practical implications

Billiards and pool. The line‑of‑centres rule explains why a player can control the cue ball’s post‑collision path by adjusting the cut angle. A shot aimed directly at the object ball (zero cut) leaves the cue ball stopped (for equal masses). A shallow cut imparts a large tangential component, sending the cue ball off at a shallow angle while the object ball follows the line of centres.
Particle physics. The same geometry governs scattering of charged particles in a detector. The impact parameter determines the scattering angle (\theta) in the centre‑of‑mass frame, which is then transformed to laboratory angles to reconstruct interaction dynamics.
Engineering collisions. In mechanisms such as ball bearings or robotic manipulators, designers exploit the perpendicular outcome for equal‑mass spheres to achieve predictable redirection of motion without additional actuators.

Concluding remarks

Two‑dimensional elastic collisions are richer than their one‑dimensional counterparts because the geometry of the impact—specifically the line of centres—introduces an extra degree of freedom. By decomposing velocities into components parallel and perpendicular to this line, the problem reduces to

a pair of independent one-dimensional interactions. Still, along the line of centres, the normal components of velocity undergo a standard elastic exchange—swapping entirely for equal masses—while the tangential components remain completely unaffected because the impulsive contact force acts solely along the normal. This elegant decomposition not only simplifies the mathematical analysis but also provides profound physical insight into the simultaneous conservation of momentum and kinetic energy.

Whether analyzing the trajectory of a cue ball, the scattering of subatomic particles, or the impact forces in mechanical systems, the underlying principles remain universally applicable. The transition from one to two dimensions transforms a straightforward linear exchange into a geometric interplay of vectors, revealing how the spatial orientation of the impact dictates the final energy and momentum distribution. In the long run, mastering these fundamental dynamics allows us to predict and manipulate the physical world with remarkable precision, naturally bridging the gap between theoretical mechanics and everyday physical phenomena.