Newton's cradle

The cradle in motion.
Newton's cradle five-ball system in 3D two-ball swing

Newton's cradle, named after Sir Isaac Newton, is a device that demonstrates conservation of momentum and energy using a series of swinging spheres. When one on the end is lifted and released, it strikes the stationary spheres; a force is transmitted through the stationary spheres and pushes the last one upward. The device is also known as Newton's balls or Executive Ball Clicker.[1][2][3][4]


A typical Newton's cradle consists of a series of identically sized metal balls suspended in a metal frame so that they are just touching each other at rest. Each ball is attached to the frame by two wires of equal length angled away from each other. This restricts the pendulums' movements to the same plane.


Newton's cradle with two balls of equal weight and perfectly efficient elasticity. The left ball is pulled away and let go. Neglecting the energy losses, the left ball strikes the right ball, transferring all the velocity to the right ball. Because they are the same weight, the same velocity indicates all the momentum and energy are also transferred. The kinetic energy, as determined by the velocity, is converted to potential energy as it reaches the same height as the initial ball and the cycle repeats.

If one ball is pulled away and is let to fall, it strikes the first ball in the series and comes to nearly a dead stop. The ball on the opposite side acquires most of the velocity and almost instantly swings in an arc almost as high as the release height of the last ball. This shows that the final ball receives most of the energy and momentum that was in the first ball. The impact produces a compression wave that propagates through the intermediate balls. Any efficiently elastic material such as steel will do this as long as the kinetic energy is temporarily stored as potential energy in the compression of the material rather than being lost as heat.

An idealized Newton's cradle with five balls when there are no energy losses and there is always a small separation between the balls, except for when a pair is colliding.

With two balls dropped, exactly two balls on the opposite side swing out and back. With three balls dropped, three balls will swing back and forth, with the central ball appearing to swing without interruption.

Newton's cradle three-ball swing in a five-ball system. The central ball swings without any apparent interruption.


Christiaan Huygens used pendulums to study collisions. His work, De Motu Corporum ex Percussione (On the Motion of Bodies by Collision) published posthumously in 1703, contains a version of Newton's first law and discusses the collision of suspended bodies including two bodies of equal mass with the motion of the moving body being transferred to the one at rest.

The principle demonstrated by the device, the law of impacts between bodies, was first demonstrated by the French physicist Abbé Mariotte in the 17th century.[5][6] Newton acknowledged Mariotte's work, among that of others, in his Principia.

Physics explanation

Newton's cradle can be modeled with simple physics and minor errors if it is incorrectly assumed the balls always collide in pairs. If one ball strikes 4 stationary balls that are already touching, the simplification is unable to explain the resulting movements in all 5 balls, which are not due to friction losses. For example, in a real Newton's cradle the 4th has some movement and the first ball has a slight reverse movement. All the animations in this article show idealized action (simple solution) that only occurs if the balls are not touching initially and only collide in pairs.

Simple solution

The conservation of momentum (mass × velocity) and kinetic energy (0.5 × mass × velocity^2) can be used to find the resulting velocities for two colliding perfectly elastic objects. These two equations are used to determine the resulting velocities of the two objects. For the case of two balls constrained to a straight path by the strings in the cradle, the velocities are a single number instead of a 3D vector for 3D space, so the math requires only two equations to solve for two unknowns. When the two objects weigh the same, the solution is very simple: the moving object stops relative to the stationary one and the stationary one picks up all the other's initial velocity. This assumes perfectly elastic objects, so we do not need to account for heat and sound energy losses. Steel does not compress much, but its elasticity is very efficient which means it does not cause much waste heat. The simple effect from two same-weight efficiently elastic colliding objects constrained to a straight path is the basis of the interesting effect seen in the cradle and gives an approximate solution to all its action.

For a sequence of same-weight elastic objects constrained to a straight path, the effect continues to each successive object. For example, when two balls are dropped to strike three stationary balls in a cradle, there is an unnoticed but crucial small distance between the two dropped balls and the action is as follows: The first moving ball that strikes the first stationary ball (the 2nd ball striking the 3rd ball) transfers all its velocity to the 3rd ball and stops. The 3rd ball then transfers the velocity to the 4th ball and stops, and then the 4th to the 5th ball. Right behind this sequence is the 1st ball transferring its velocity to the 2nd ball that had just been stopped, and the sequence repeats immediately and imperceptibly behind the first sequence, ejecting the 4th ball right behind the 5th ball with the same small separation that was between the two initial striking balls. If they are simply touching when they strike the 3rd ball, the more complete solution described below is necessary in order to be precise.

Other examples of this effect

The interesting effect of the last ball ejecting with a velocity nearly equal to the first ball can be seen in sliding a coin on a table into a line of identical coins, as long as the striking coin and its twin targets are in a straight line. The effect can also be seen when a sharp and strong pressure wave strikes a dense material immersed in a less-dense medium. The atoms, molecules, or larger-scale sub-volumes of the denser material may act like a sequence of elastically-connected balls. It is not a single straight line, but the surrounding atoms, molecules, or sub-volumes experiencing the same pressure wave act to constrain each other (instead of hanging from strings) as if they are in a straight line. For example, lithotripsy shock waves can be sent through the skin and tissue without harm in order to burst kidney stones. The side of the stones opposite to the incoming pressure wave bursts, not the side receiving the initial strike.

When the simple solution applies

In order for the simple solution to precisely predict the action, no pair in the midst of colliding may touch a third ball because the presence of the 3rd ball effectively makes the ball being struck appear heavier. Applying the two conservation equations to solve the final velocities of three or more balls in a single collision results in many possible solutions, so these two principles are not enough to determine resulting action.

Even when there is a small initial separation, a third ball may become involved in the collision if the initial separation is not large enough. When this occurs, the complete solution method described below must be used.

Small steel balls work well because they remain efficiently elastic with little heat loss under strong strikes and do not compress much (up to about 30 µm in a small Newton's cradle). The small, stiff compressions mean they occur rapidly, less than 200 microseconds, so steel balls are more likely to complete a collision before touching a nearby third ball. Softer elastic balls require a larger separation in order to maximize the effect from pair-wise collisions.

More complete solution

An ideal cradle is designed to have a small initial separation between the balls, but most do not. This section is needed to accurately describe the action when there is not an initial separation and in subsequent collisions that can involve more than 2 balls. This method simplifies to the simple solution when separations are included.

For two balls colliding, only the two equations for conservation of momentum and energy are needed to solve the two unknown resulting velocities. For three or more simultaneously colliding elastic balls, the relative compressibilities of the colliding surfaces are the additional variables that determine the outcome. For example, five balls have four colliding points and scaling (dividing) three of them by the fourth will give the three extra variables needed (in addition to the two conservation equations) to solve for all five post-collision velocities.

Newtonian, Lagrangian, Hamiltonian, and stationary action are the different ways of mathematically expressing classical mechanics. They describe the same physics but must be solved by different methods. All enforce the conservation of energy and momentum and require knowledge of the compressibilities of the surfaces. Newton's law has been a method used in research papers. It is applied to each ball and the sum of forces is made equal to zero which enforces the conservation of energy and momentum. Frictional force losses could be included this sum, but in this discussion they are not.

Determining the velocities for the case of one ball striking four initially-touching balls is found by modeling the balls as weights with non-traditional springs on their colliding surfaces. Steel is elastic and follows Hooke's force law for springs, , but because the area of contact for a sphere increases as the force increases, colliding elastic balls will follow Hertz's adjustment to Hooke's law, . This and Newton's law for motion () are applied to each ball, giving five simple but interdependent differential equations that are solved numerically.[7] When the fifth ball begins accelerating, it is receiving momentum and energy from the third and fourth balls through the spring action of their compressed surfaces. For identical elastic balls of any type with initially touching balls, all the action for all cradles is the same for at least the first strike. Only the time required to complete a collision will change, which can affect subsequent collisions. 40% to 50% of the kinetic energy of the initial ball (from a single-ball strike) is stored in the ball surfaces as potential energy for most of the collision process. 13% of the initial velocity is imparted to the fourth ball (which can be seen as a 3.3 degree movement if the fifth ball moves out 25 degrees) and there is a slight reverse velocity in the first three balls, with the largest being −7% of the initial velocity in the first ball (it has a slight reverse movement after striking). This separates the balls, but they will come back together just before the fifth ball returns making a determination of "touching" during subsequent collisions complex, but less likely than in the initial strike. Stationary steel balls weighing 100 grams (with a strike speed of 1 m/s) need to be separated by at least 10 µm if they are to be modeled as simple independent collisions. The differential equations with the initial separations are needed if there is less than 10 µm separation, a higher strike speed, or heavier balls.[8]

The Hertzian differential equations predict that if two balls strike three, the fifth and fourth balls will leave with velocities of 1.14 and 0.80 times the initial velocity.[9] This is 2.03 times more kinetic energy in the fifth ball than the fourth ball, which means the fifth ball should swing twice as high as the fourth ball. But in a real Newton's cradle the fourth ball swings out as far as the fifth ball. In order to explain the difference between theory and experiment, the two striking balls must have at least 20 µm separation (given steel, 100 g, and 1 m/s). This shows that in the common case of steel balls, unnoticed separations can be important and must be included in the Hertzian differential equations, or the simple solution may give a more accurate result.

Gravity and the pendulum action influence the middle balls to return near the center positions at nearly the same time in subsequent collisions. This and heat and friction losses are influences that can be included in the Hertzian equations to make them more general and for subsequent collisions.[10]

Adjustment for pressure waves

Pressure waves in steel (like sound) travel about 5 cm in 10 microseconds which is barely fast enough for the Hertzian solution method to not require a substantial modification to adjust for the delay in force propagation across the balls. The effect works to partially obscure the inertia of the balls further away from each other, causing it to act a little closer to the simple solution.

Effect of different types of balls

Using an efficiently elastic material other than steel such as rubber or glass, larger or smaller balls, or flattening the surface at the contact points will not change the action (final velocities) of the cradle as long as the same change is made in each ball. This is clear from the simple solution for when the balls only collide in pairs. It is also true for the complete solution because the compressibility of the surfaces relative to each other (not the actual amount of compression) determines the percent of kinetic energy transferred to each ball and therefore the final velocities. Steel is better than most materials because it allows the simple solution to apply more often in subsequent collisions. Hollow thin-shell spheres or objects with traditional springs on their surfaces will have a smaller exponent than the Hertzian 1.5 value which will cause their action to be a different, further away from the ideal simple solution if their surfaces are initially touching.

Heat and friction losses

This discussion has neglected energy losses from heat generated in the balls from non-perfect elasticity, friction in the strings, friction from air resistance, and sound generated from the clank of the vibrating balls. The energy losses are the reason the balls eventually come to a stop, but they are not the primary or initial cause of the action to become more disorderly, away from the ideal action of only one ball moving at any instant. The increase in the non-ideal action is caused by collisions that involve more than two balls at a time, effectively making the struck ball appear heavier. The size of the steel balls is limited because the collisions may exceed the elastic limit of the steel, deforming it and causing heat losses.


The most common application is that of a desktop executive toy. Another use is as an educational physics demonstration, as an example of conservation of momentum and conservation of energy.

A similar principle, the propagation of waves in solids, was used in the Constantinesco Synchronization gear system for propeller / gun synchronizers on early fighter aircraft.

Invention and design

Large Newton's cradle at American Science and Surplus

The experimental use of pendulum devices, to demonstrate the law of impacts between bodies, was first described by Mariotte in the 17th century.

There is much confusion over the origins of the modern Newton's cradle. Marius J. Morin has been credited as being the first to name and make this popular executive toy. However, in early 1967, an English actor, Simon Prebble, coined the name "Newton's cradle" (now used generically) for the wooden version manufactured by his company, Scientific Demonstrations Ltd. After some initial resistance from retailers, they were first sold by Harrods of London, thus creating the start of an enduring market for executive toys. Later a very successful chrome design for the Carnaby Street store Gear was created by the sculptor and future film director Richard Loncraine.

The largest cradle device in the world was designed by MythBusters and consisted of five one-ton concrete and steel rebar-filled buoys suspended from a steel truss.[11] The buoys also had a steel plate inserted in between their two halves to act as a "contact point" for transferring the energy; this cradle device did not function well. A smaller scale version constructed by them consists of five 6 inches (15 cm) chrome steel ball bearings, each weighing 33 pounds (15 kg), and is nearly as efficient as a desktop model.

The cradle device with the largest diameter collision balls on public display, was on display for more than a year in Milwaukee, Wisconsin at retail store American Science and Surplus. Each ball was an inflatable exercise ball 26 inches (66 cm) in diameter (enclosed in cage of steel rings), and was supported from the ceiling using extremely strong magnets. It was dismantled in early August 2010 due to maintenance concerns.

See also


  1. "". Retrieved 3 November 2011.
  2. "". Retrieved 3 November 2011.
  3. "". Retrieved 3 November 2011.
  4. "". Retrieved 3 November 2011.
  5. "Harvard website page on Newton's Cradle". Retrieved 2007-10-07.
  6. Wikisource:Catholic Encyclopedia (1913)/Edme Mariotte
  7. Herrmann, F.; Seitz, M. (1982). "How does the ball-chain work?" (PDF). American Journal of Physics. 50. pp. 977–981.
  8. Lovett, D. R.; Moulding, K. M.; Anketell-Jones, S. (1988). "Collisions between elastic bodies: Newton's cradle". European Journal of Physics. 9 (4): 323. Bibcode:1988EJPh....9..323L. doi:10.1088/0143-0807/9/4/015.
  9. Hinch, E.J.; Saint-Jean, S. (1999). "The fragmentation of a line of balls by an impact" (PDF). Proc. R. Soc. Lond. A. 455. pp. 3201–3220.
  10. Hutzler, Stefan; Delaney, Gary; Weaire, Denis; MacLeod, Finn (2004). "Rocking Newton's Cradle" (PDF). American Journal of Physics. 72. pp. 1508–1516.
  11. "Newton's Crane Cradle (5 Oct. 2011)" at the Internet Movie Database


  • Herrmann, F. (1981). "Simple explanation of a well-known collision experiment". American Journal of Physics. 49 (8): 761. Bibcode:1981AmJPh..49..761H. doi:10.1119/1.12407. 
  • B. Brogliato: Nonsmooth Mechanics. Models, Dynamics and Control, Springer, 2nd Edition, 1999.
Wikimedia Commons has media related to Newton's cradle.
This article is issued from Wikipedia - version of the 12/1/2016. The text is available under the Creative Commons Attribution/Share Alike but additional terms may apply for the media files.