Department of Energy Argonne National Laboratory Office of Science NEWTON's Homepage NEWTON's Homepage
NEWTON, Ask A Scientist!
NEWTON Home Page NEWTON Teachers Visit Our Archives Ask A Question How To Ask A Question Question of the Week Our Expert Scientists Volunteer at NEWTON! Frequently Asked Questions Referencing NEWTON About NEWTON About Ask A Scientist Education At Argonne Coefficient of Restitution
Name: Jim
Status: Student
Grade: 6-8
Location: AP
Country: United States
Date: March 2009

We are working on a module on Sports Materials. We have done an activity where we measure the rebound of sports balls. We calculate the coefficient of restitution (COR) using:

COR = (rebound height/drop height)^0.5

A COR of 1 would be a perfectly elastic collision, and a COR of 0 would be a perfectly inelastic collision. We have dropped ping pong, tennis, basketball, soccer ball, baseball, golf ball, and others.

The class decided to do an extension of this activity. We tried different metal balls. We have dropped metal balls (varying mass) on metal surfaces, different balls on hard surfaces, and done both of these on soft surfaces. We got results that seemed to make sense. We then dropped (60 cm height)steel balls (1", 1/2", 1/8" diameters) on a drum membrane. Why is it that the steel on steel trials the impact (and presumably deformation/heat production) the lower the COR, but for steel on a drum membrane the greater the impact (and presumably deformation/heat production) the higher the COR?

I accept, for the steel on steel, the collisions have nonlinear effects, as shown by the square root relationship shown above. The greater the collision, the greater the losses (so balls dropped from greater heights or are thrown at greater speeds will have lower CORs). But for the drum membrane, the greater the impact speed, the higher the COR and the more elastic the collision. Are small impacts being absorbed by the membrane without stretching it?

I have done this with constant volume, but different density balls (lead, steel, and glass), and again I find opposite trends. For the steel plate, the marble, then steel, then lead ball bounce highest, but it is the opposite for the drum. Perhaps the radius of curvature of the impact has something to do with the outcome. Perhaps the membrane is a polymer and reacts significantly different than a non-polymer. I am not sure. Again, Why is the membrane more elastic the more it is deformed? What would you guess would happen if I continued the experiments for balls of brass, rubber, cork, and aluminum?


Polymers have a vexing property called glass transition that is the point at which the polymer moves from a solid glassy state to a solid rubbery state. The vexing thing about it is that it is not set to a particular temperature like melting point. It is not a first order transition.

On top of this, is that most polymers will exhibit viscoelastic properties giving it both viscous and elastic properties at any given temperature. This means that the polymer can both absorb energy and dissipate it as heat, and also return the energy in an elastic way.

The combined situation makes polymer collisions very non-linear.

For example, if you take silly putty and stretch it slowly, or allow its own weight to make it stretch, the putty will flow like a liquid (be viscous). However, if you pull on it really fast, it snaps and breaks like a solid rubber (be elastic with a low energy to failure). This points out that the flow characteristics of silly putty is a function of the frequency and amplitude of the force acting on it. I've even seen videos where silly putty was formed into a hollow ball with thin walls and with a very high frequency, high amplitude impact - as with a very fast hammer strike, make the silly putty break like glass - with a sound very much like breaking glass.

Another example is in the case of dropping steel balls onto solid glassy polystyrene. Above and below the glass transition, the ball will bounce. Below the glass transition the polymer is acting like a solid glass and not much energy dissipation takes place. Above the glass transition the polymer is acting like a rubber which returns the energy to the ball (as the polymer returns to its low energy state) and not much energy dissipation takes place. But, at the glass transition, the ball does not bounce at all. At the glass transition all the energy is dissipated as the polymer flows - and flows to an equally low energy state so no rebound occurs.

I am guessing because the drum skin is made of some polymeric material that is stretched tight, it tends to exhibit elastic, viscous and energy dissipation - depending on impact strength and frequency (like I said- very non-linear). You might even find that depending on drum-head material and manufacturer, that you would get varying results.

Greg (Roberto Gregorius)

Hi Jim,

These are great questions, and as you have noticed there are tons of variables involved. I also want to congratulate you on performing some great experiments with your students. Your challenge now is how to better organize and visualize your data -- that will help you to come up with explanations for the data (and generate new experiments to test them).

Looking at the big picture, there are lots of factors at play here beyond just the material of the ball and the height it is dropped. The biggest factors are the deformation of the ball and the deformation of the surface when they impact, the time frames in which they deform, and how much of the energy of that deformation is returned to the ball. Depending on the materials (ball and surface), the deformation could be large or small, and depending on the velocity/energy involved in the impact, the return could be (comparatively) fast or slow. All of these factors affect the apparent COR.

As you can see, things get complicated very fast. If you want to cut to the chase, I did find a very good article helping explain some of the details from the American Journal of Physics.

This article does not discuss the surface as much as the ball, but it does a good job of breaking down the factors involved. The article is probably at too high a level for the vast majority of your students, but you might find it useful, and perhaps you can 'translate' in terms that are appropriate for them.

With respect to your drum experiments, I would guess that yes, the membrane is able to absorb some amount of energy before it returns it elastically. Polymers have a 'time constant' where they deform when moving slowing but like a solid when moving quickly (think of silly putty stretching slowing, but breaking/snapping when pulled quickly). The more energy in the collision, the more solid the polymer will act, and thus the more elastic the collision. So the heavier/higher the ball, the closer to 'full' COR you will see. Did you plot the CORs against drop height? Do they reach an asymptote?

I would recommend you try to get quantitative and use graphs to examine your data. Start plotting your data of COR versus height and COR versus diameter (for the same ball). Perhaps visualization will help you develop better explanations for your data. I would also be interested in hearing what experiments you plan next to test your hypothesis.

Hope this helps,
Burr Zimmerman

This may not completely answer your question but in some cases increased force increases elastic behavior. An example of this is playing pool. A a light strike of cue ball and object ball will leave both moving after the collision - not a very good elastic collision. But a very firm hit will stop the cue ball dead in its tracks and send the object ball streaking away in a highly elastic collision.

Robert Avakian

Click here to return to the Engineering Archives

NEWTON is an electronic community for Science, Math, and Computer Science K-12 Educators, sponsored and operated by Argonne National Laboratory's Educational Programs, Andrew Skipor, Ph.D., Head of Educational Programs.

For assistance with NEWTON contact a System Operator (, or at Argonne's Educational Programs

Educational Programs
Building 360
9700 S. Cass Ave.
Argonne, Illinois
60439-4845, USA
Update: June 2012
Weclome To Newton

Argonne National Laboratory