anyone know the correct physics formula for dropping, say, a hard plastic cube on a concrete floor, maybe on grass? "it''s all about coefficients, u just gotta know those coeffecients" - me hopeful and waiting -Succinct ////////////////////////////// I have no name that you may call me. I am Succinct, so call me as such. \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

I can recommend the technique I am using. It is very effective.

Simply assign a coefficient of elasticity to each material you are simulating (between 0 and 1). Then, when two materials collide, multiply their two coefficients together and use the result to scale your rebound impulses.

So, rubber might have a coefficient of say 0.95, and putty might have a coefficient of say, 0.1. It works very well for me.

will that incorporate "the jerk" into the model?
hrmm, that''s really what i''m trying to model, "the jerk"

displacement
derv: velocity
derv: acceleration
derv: "the jerk"

i think i just like saying "the jerk"
lol



//////////////////////////////

I have no name that you may call me. I am Succinct, so call me as such.

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

I think I speak for everyone when I say, What the hell is "the jerk"?

Anyway, about your original question, I guess Novalis took your quote about coefficients literally, but maybe you wanted to know a little more...

Upon collision, you''ll want to basically add a normal force to to your cube. What I mean is, take the normal vector for your concrete floor (probably (0,1,0) if y is up), and add some scaled version of that to your cube''s velocity. The scale (magnitude) will depend on how hard the cube is hitting the ground, i.e. how fast the cube is moving in the direction of the floor''s normal. It will also depend on how much elasiticity you want, as Novalis mentioned.

If you want your cube to spin realistically upon collision, then, well, that''s a little beyond me. I think you''d have to divide the normal force into a regular "pushing" force and a torque... I dunno what axis you spin the cube on, though, or how you combine different spins, or how you even represent the angular velocity in the first place.

If you want to find out how high the cube will bounce then Im not shure I can help as that depends on some surface property. But in general you can find the upward force that the ground will exert of the cube by multiplying the cubes mass by its acceleration (due to gravity), that is find force. When the cube hits the ground that same force will be reflected back on the cube, the cube will then move by that force. From there to find how far the cube will bounce will depend on some form on elasticity constant that you would have to create.

for other collisions,

conservation of momentum

mass1*velocity1+mass2*velocity2=mass1*velocity_final1+mass2*velocity_final2

if its inelastic then the two objects that collided would be modeled by

mass1*velocity1+mass2*velocity2=(mass1+mass2)velocity_final

there we go, thx guys.

seems like hs physics, but i haven't been in hs for a few years

thx for the reminder.

btw, the jerk is...

first and most basic, when relating motion of an object to time, there is
displacement (s):
s = f(t),
where t is time and f is some function describing the
object's position over time.

secondly, there is velocity (v)
v = f'(t)
where f' is the derivative of f

acceleration is the derivative of velocity, and the derivative of acceleration is "the jerk" or change in acceleration, with respect to time

that's the jerk, i mean i wasn't just trying to be an idiot

//////////////////////////////

I have no name that you may call me. I am Succinct, so call me as such.

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
Edited by - Succinct on November 18, 2000 9:59:42 AM