[ODE] mass ratio idea

[Petr Sovis]

In my opinion, it isn't, becouse even statical object can have velocity 
(fake - just for handling contacts). I don't know if it's possible to add 
relative velocities for normal contact direction in ODE  but maybe it's 
possible.

one constraint line for normal ( n ) direction of contact:
  J v' = c  ---> J v' = c + static_object_velocity * n * fps;   c - 
represent bouncing
for tangencial == friction directions (t1, t2)
t1:  J v' = 0  ---> J v' = rv1 = static_object_velocity * t1 * fps;
t2:  J v' = 0  ---> J v' = rv2 = static_object_velocity * t2 * fps;

static_object_velocity should be absolute velocity in P of "static" object.
static_object_velocity = linvelstatic + angvelstatic x (P - COMstatic).

rv1, rv2 are in ODE "contact.surface.motion1" and "contact.surface.motion2" 
parameters for contact joint (Same thing as conveyor belt), so only problem 
remains in n direction line. It could be solvable by cheating bounce contact 
joint parameter.

Best regards, Petr.

P.S. Sorry for mislead if I'm wrong.
P.S.S. I'm realy not sure about signs in equations (I always try and debug 
to write them correctly).

----- Original Message ----- 
From: "Andrzej Szombierski" <qq at kuku.eu.org>
To: <ode at q12.org>
Sent: Monday, September 27, 2004 6:09 PM
Subject: Re: [ODE] mass ratio idea


> On Mon, 27 Sep 2004, Tyler Streeter wrote:
>
>> Here's an idea for dealing with the problem of having
>> two objects collide with a huge mass ratio:
>>
>> Whenever two objects (both with bodies) collide,
>> calculate the ratio of the larger mass to the smaller
>> mass.  If the ratio is above some threshold (e.g. the
>> large mass is a train car and the small mass is a
>> baseball), ignore one side of the collision.  In other
>> words, temporarily treat the train car as a static
>> object (with no body).
>>
>> Any thoughts?
>
> Seems nice, but I think there may be a problem if both bodies are moving.
> If one of them is treated as static, the relative velocity at the contact
> point will be calculated incorrectly, possibly resulting with non-physical
> behavior.
>
> (but that's my guess - haven't tried it)
>
> -- 
> :: Andrzej Szombierski :: qq at kuku.eu.org :: http://kuku.eu.org ::
> :: anszom at bezkitu.com  :: radio bez kitu :: http://bezkitu.com ::
>
> _______________________________________________
> ODE mailing list
> ODE at q12.org
> http://q12.org/mailman/listinfo/ode
>