Capturing the physics of bouncing in a simulation

[mindbreaker]

hi,
srr if this isn't the wright place to post this...

I'm working on engine (i'm a programer , c++) and I'm stuck at bouncing...

I have an object (repressented by a sphere) at an unknow height, with an unknow mass and I drop it, so gravity does it job and it starts to fall down

now , when there is a collision with the floor, I want the object to bounce but dependent of it's texture ( ex.: tennisball - bowl ball ) and physical correct ...
how is this done ? I remember I saw this (almost 8 years ago ) at school ...

note 1: friction (etc) can be ignored (for now , so if this isn't must harder i request this already )
note 2 : this is something that should take place at Earth and at a normal height (so not 150 km above into the sky ) ...

thx anyway

 

[Ben Niehoff]

You can give the object a property called "elasticity". "Elasticity" is a measure of the amount of kinetic energy that can be recovered when the object has deformed due to collision; springier substances that can return to their original shape are more elastic (bowling balls, tennis balls, steel balls and rubber balls are all very elastic; balls of Silly-Putty are very inelastic).

Practically speaking, elasticity is a number between 0 and 1 that represents the fraction of energy retained in collisions. A ball of elasticity 1 will bounce up to its original height every time; a ball of elasticity 0 will fall and then stick to the ground, without bouncing at all. Most objects are somewhere in between; a ball of elasticity 0.8 will always bounce up to 80% of its previous height (thus, the bounces will decrease in height exponentially).

In the case of nonzero friction, you have an additional complication: a ball striking a surface at a glancing angle will, due to friction, experience a torque causing it to spin. Thus, some of the previous translational kinetic energy gets converted into rotational kinetic energy.