[ODE] Constraint theorizing

Wed Feb 26 03:24:01 2003

>2.  Nothing is ever new.  I saw this exact thing being done with point
masses
>"simulating" rigid bodies
there are already out there engines that use relaxation methods based
methods:

see the free physics engines supplied by oxford dynamics:
http://www.oxforddynamics.co.uk

relaxation is at positional level. This is very similar to what described
in a paper
presented at eurograph a couple of years ago("a procedural approach to
solving constraints...", j. Won Lee, eurographics 2000 )

The first time i came across the idea of this kind of solution was while
reading the Havok documentation a long time ago.

Also Xavier Provot used relaxation for cloth simulation a few years ago.

however when we work by altering directly positions  forces are not
directly computed and this may cause
problems, especially in the friction computation, but it may still be
acceptable.

in the the method implemented by Sergio,  error correction is used instead.
this implies that recovering forces
are effectively computed. So it's physically more consistent(realistic) and
i think fits better in the ODE framework.
For example it would work in conjunction with the Featherstone's method
where multibodies are
parametrized in state-space and we cannot direcly displace objects
around(imagine a chain colliding against the floor).
As usual the best solution is application dependent as i found that a
combination of different methods worked best for me.

cheers,
Antonio

"Gary R. Van Sickle" <g.r.vansickle@worldnet.att.net>@q12.org on 26/02/2003
03:34:17

Sent by:    ode-admin@q12.org

To:    <ode@q12.org>
cc:
Subject:    RE: [ODE] Constraint theorizing

1.  Ok, you're BOTH Gods!
2.  Nothing is ever new.  I saw this exact thing being done with point
masses
"simulating" rigid bodies
(http://www.ioi.dk/Homepages/thomasj/publications/gdc2001.htm) and it
floated
around in my head for weeks without me seeing a direct connection to "real"
rigid body simulation.  And I'm sure this guy got it from somebody else
too.  We
all add our parts to the sum of the whole... or words to that effect ;-).
3.  Nothing is ever easy.  But you guys got it done.  Again, GODS!  GODS I
SAY!
;-)

--
Gary R. Van Sickle
Brewer.  Patriot.

> -----Original Message-----
> From: Gary R. Van Sickle [mailto:g.r.vansickle@worldnet.att.net]
> Sent: martes, 25 de febrero de 2003 7:29
> To: ode@q12.org
> Subject: RE: [ODE] Constraint theorizing
>
>
> > There were some discussion about such things in this list
> > (iterative schemes and the like) I have tested (succesfully)
> > a straighforward iterative  technique which solves *only*
> > 1 constraint at each step (thanks Antonio). That is, if you
> > have a system  of say C= 10 constraints, then you solve
> > every constraint separately without taking into account
> > the other constraints. The idea is that the time step is
> > subdivided into a sort of "microsimulations". Something
> > like that:
> >
> > parameter: N substeps, C constraints, Dt (ODE timestep)
> >
> > for i=0..N-1 do
> >   for c = 0..C-1 do
> >         Solve constraint c-th
> >         Apply forces to constraint bodies
> >   next
> >   Integrate bodies by (Dt/N)
> > next
> >
> > The idea is that as N->inf the movement of the bodies is
> > very small so the effect of every constraint is very
> > localized and this approach becomes exact just in the limit.
> > In my tests a small number is iterations is enough to
> > get decent results.
> >
> > Please note that the core of this scheme is to solve a
> > very small (up to 6x6) LCP constraint system.  I think
> > this approach gives enough room to a lot of performance
> > improvements. The next step should be to find a fast
> > code able to solve the small system.  Also, note that
> > is no longer necessary to store a big matrix for the
> > constraint coefficients.
>
> This is like excellent to the third power!  And believe it or not, I
> actually
> thought of just this scheme during my run this evening, not that I'd have
> had
> the ability to pull it off.  So let me see if I can summarize.  ODE as-is
is
> O(C^3).  The new scheme is O(CN), with N a small integer.  That
improvement
> in
> scalability alone is fabulous.  PLUS, we lose the stack problems that
have
> been
> plagueing people.
>
> You are a GOD!
>
> --
> Gary R. Van Sickle
> Brewer.  Patriot.
>
>
> _______________________________________________
> ODE mailing list
> ODE@q12.org
> http://q12.org/mailman/listinfo/ode
> _______________________________________________
> ODE mailing list
> ODE@q12.org
> http://q12.org/mailman/listinfo/ode

_______________________________________________
ODE mailing list
ODE@q12.org
http://q12.org/mailman/listinfo/ode

**********************************************************************
This email and any files transmitted with it are confidential and
intended solely for the use of the individual or entity to whom they
are addressed. If you have received this email in error please notify
postmaster@scee.net

This footnote also confirms that this email message has been checked
for all known viruses.

**********************************************************************
 SCEE 2002