Re: COLLISION 2

[John Denker <jsd@MONMOUTH.COM>]

At 02:05 PM 11/28/99 -0500, Chuck Britton wrote:

> In the inelestic case we 'just' need to lock the compressed spring in
> place once the relative motion has ceased. I don't see any basic
> difficulty in 'capturing' the collision energy in this way.

That is an approximation, but only a rough approximation, to what is really
going on.  In particular, if you apply that model to repeated couplings and
uncouplings, it will break down.

More generally, whenever you model a dissipative process by a simple
nondissipative process, you violate the fluctuation-dissipation theorem and
various other laws of physics.

> Is it
> similar to imagining two hydrogen atoms colliding and forming the
> diatomic molecule?

Actually your model of railroad cars is *better* than the corresponding
model of hydrogen.  You are free to add "unusual" but imaginable parts to
the railroad car, but you cannot add parts to a hydrogen atom.

In fact, the chance is utterly negligible that a two-body collision between
hydrogen atoms would result in formation of the molecule.

You can see this in the reaction rates:  the gas-phase recombination
reaction rate goes like the cube of the density, indicating that a
three-body collision is necessary in order to balance energy and momentum.

> The molecule would have to be in a excited state to contain the
> kinetic energy that is lost.

Energy can't be lost;  it can only be transferred or stored.  In this case,
the molecule would be so excited that it would immediately
dissociate.  Guaranteed.