Newton's Cradle Lab (2D)

Front view of a cradle with velocity arrows, contact flashes and live graphs of energy, momentum and the contact force pulse.

Switch to 3D

Presets

Playback

Chain

Number of balls, how many are lifted, the lift angle and the mass of the lifted balls.

Contact and environment

Restitution, Hertz contact time, damping and gravity.

Time

Slow motion makes the collision pulse visible.

Layers

Model checks

The science

What a Newton's cradle shows

Lift one ball and release it: one ball leaves the far end at the same speed. Lift two: two leave. Momentum p = sum m v and kinetic energy E = sum 1/2 m v^2 are both conserved in the collision, and for equal masses the only outcome that keeps both is "k balls in, k balls out". With unequal masses the outcome follows from the same two equations (a heavy ball hitting a light chain pushes out several balls).

Pendulums

Each ball of mass m_i hangs on strings of length L = 0.20 m, so the angle obeys theta'' = -(g/L) sin(theta) + F_t / (m L) where F_t is the contact force component along the swing direction. A ball lifted by theta0 arrives at the bottom with v = sqrt(2 g L (1 - cos theta0)). The period is not 2 pi sqrt(L/g) at large angles but 4 sqrt(L/g) K(sin(theta0/2)) with the complete elliptic integral K.

Hertz contact between the balls

Two touching spheres squeeze each other with a force F = k d^(3/2) (d = overlap). A collision lasts t_c = 2.943 d_max / v0 with d_max = (5 m* v0^2 / (4 k))^(2/5), so softer contact means a longer pulse. The pulse travels through the chain ball by ball (graph 4). To lose energy, a damping term is added (Hunt and Crossley), F = k d^(3/2) (1 + 3/2 alpha d'), and the page tunes it so the restitution at 1 m/s equals the slider value e.

Numerical checks on this page

(1) The simulated swing period agrees with the exact elliptic-integral period. (2) A lossless two-ball impact lasts as long as the Hertz formula predicts and the balls exchange velocities. (3) Two of five balls lifted, no loss: about all the kinetic energy ends up in the two balls that leave.

Honest simplifications

The motion is planar and the strings are rigid and massless; there is no friction between the balls and no rotation. Contact is made softer than real steel (a few tenths of a millisecond in reality, 1.5 ms by default) so it can be integrated in real time with a leapfrog step of at most 20 microseconds; the behaviour in the cradle is the same. The momentum shown is the horizontal momentum of the balls: string tension also acts, so it is only conserved inside a collision, not over a whole swing.

What to try
  • Classic: one ball, then Two balls swing in and Three balls: the same number of balls leaves as arrived.
  • Slow the time speed to 0.02x and watch graph 4: the contact force pulse moves along the chain.
  • Heavy ball hits light chain and Light ball hits heavy chain: unequal masses break the "k in, k out" rule.
  • Soft putty balls: low restitution, the energy graph drops after each collision and the balls stick together.
  • Increase the contact time slider: a softer ball makes the pulse longer and the transfer less clean.
Ball angles versus time
Energy versus time
Horizontal momentum versus time
Contact forces in the last collision