[ODE] QuickStep tests and units....

[Alen Ladavac]

If we are not mistaken, the boxes in testcrash are 1x1x1m (what would make
1m^3 volume) and weight only 1kg. What makes them a little lighter than thin
air (air has characteristic density of about 1.16kg/m^3 at room
temperature). I know that, theoretically, if you scale all masses with same
factor, simulation should appear same, but sometimes it is better to have
realistic values, just to be on the safe side, and to make sure you are
testing something that makes sense. I'd recommend to use some density value
like 800kg/m^3 (wood), it might help solver stability, as increased inertia
tensors will prevent numerical errors from spinning the boxes.

Btw, in our tests we run with gravity of 30m/s^2, as it is more appealing
for FPS games; 9.81 is just too low.

Just my 2e-2$,
Alen


----- Original Message -----
From: "Megan Fox" <shalinor at circustent.us>
To: "Nguyen Binh" <ngbinh at glassegg.com>; "Ed Jones" <ed.jones at oracle.com>
Cc: "Ode" <ode at q12.org>
Sent: Thursday, May 20, 2004 05:04
Subject: RE: Re[2]: [ODE] QuickStep comparison to Tokamak -
large-scalestacking


> That test is using the standard 1 unit = 1 meter scale, with boxes that
are
> 1x1x1.
>
> Heck, it's right here -
> http://shalinor.circustent.us/elium/builds/PhysDemo.zip - if anyone wants
to
> fiddle with it.  Powered by Nebula 2, etc etc, blah - if it crashes,
you're
> one of the small minority that has that problem with my builds (no idea
why,
> I'm overdue for a WinXP reinstall anyways).
>
>
> Boxes that are 1/2 meter cubed shouldn't be any more or less stable in a
> properly-behaving sim, though - and they certainly aren't in Tokamak, not
> sure in ODE as I've only ever used 1x1x1's.
>
> -Megan Fox
>
>
> > Hi Ed,
> >
> > EJ> Where I am -9.80665 is normal!
> > EJ> ;)
> >
> >     ;)
> >     But if you apply -9.8 gravity in test_boxstack.cpp you'll make ODE
> >     much more unstable because such value is too large for,say, a
> >     0.5*0.5*0.5 box, So i you drop a box from 5 (m) into a box in the
> >     ground,you'll notice very deep penetration.
> >     So I just afraid that people use geoms like common ODE sample but
> >     with very high gravity(you could see in test_boxstack.cpp, gravity
> >     is -0.5).
> >
> >     Actually, we should take special care to units in physics to avoid
> >     numerical problem.
> >
> > --
> > Best regards,
> >
> > ---------------------------------------------------------------------
> >    Nguyen Binh
> >    Software Engineer
> >    Glass Egg Digital Media
> >
> >    E.Town Building
> >    7th Floor, 364 CongHoa Street
> >    Tan Binh District,
> >    HoChiMinh City,
> >    VietNam,
> >
> >    Phone : +84 8 8109018
> >    Fax   : +84 8 8109013
> >
> >      www.glassegg.com
> > ---------------------------------------------------------------------
> >
> > _______________________________________________
> > ODE mailing list
> > ODE at q12.org
> > http://q12.org/mailman/listinfo/ode
> >
> >
>
> _______________________________________________
> ODE mailing list
> ODE at q12.org
> http://q12.org/mailman/listinfo/ode
>