Bouncing Ball Bearings Beautifully.

Some of my favourite demonstrations of scientific principles are ones that you expect to behave one way but, in reality, act completely different. To me this demonstrates the value of experimentation and observation as you can never be sure until you do something for yourself. It also usually means that there’s some kind of interesting physical phenomena at play that I’m not yet familiar with, something which usually means an enjoyable trip down a Wikipedia hole. The following video is one such demonstration, showcasing an interesting property of amorphous metals.

In this demonstration (the whole channel is worth watching by the way) we can see the difference between an amorphous metal surface and a traditional one when a ball bearing is dropped on it. The difference in bounce height is quite staggering, enough to make you think initially that there’s some form of spring hidden in the cylinder. The actual reason for the difference, which is briefly touched on in the video, is far more interesting than it being a simple trick.

The material that the atomic trampoline is made of has some rather unique properties. Regular metals are usually of a crystalline structure meaning that their component atoms are highly ordered. Amorphous metals on the other hand (sometimes referred to as metallic glass) have a highly disorganised structure, owing to the fact that they’re usually alloys (made up of several different metals) and their creation process stops the formation of a crystalline structure.

This disorganisation prevents the formation of defects called dislocations which appear in crystalline metals. When a ball bearing strikes the regular metal surface these dislocations glide through the other parts of the metal’s structure, dissipating a lot of the energy. In the amorphous metal however there are no such dislocations and so much less of the energy is lost with each bounce. Of course the lack of dislocations does not negate other losses due to sound and heat which is why the ball bearing doesn’t bounce infinitely.

What I’d love to see is the same experiment redone in a vacuum chamber with both the ball bearing and the surface made from amorphous metals. I’m sure we could get some really absurd bouncing times with that!