# [ODE] gravitational force / unstable joints

Jon Watte (ODE) hplus-ode at mindcontrol.org
Tue Jul 31 09:06:30 MST 2007

```It all depends on what units you put into ODE. If you put in SI units,
then you want gravity to be -9.82 or so (depending on where on Earth you
are). The unit is "force per mass unit," and this means about 10 newtons
per kilogram of mass. It gets turned into an impulse when you step it (i
e, multiplied by the time step size, each time).

Cheers,

/ h+

Tobias Zimmer wrote:
> Thank you again for the fast answer, but I am a bit confused in between!
>
> I found this thread in the petrucci-forum:
> http://ode.petrucci.ch/viewtopic.php?t=306&highlight=frames
> <http://ode.petrucci.ch/viewtopic.php?t=306&highlight=frames>
>
> There, akiross is calculating the following:
>
> "forgot 0.5!
>
> the right formulae is:
> s(t)=0.5at² -> 10 = 0.5*10*1.4² ~ 9.8
>
> So this is correct, of course ode isn't wrong"
>
> This seems to mean that the values in the gravity-function don't directly
> correspond to the values used in physics but have to be calculated
> from them.
> If I understand the formula above correctly, which I doubt, the
> gravity would be
> correct by the use of (0,0,-0.5)...
>
> And the Ray-Car demo is new to me. Is it defined in some of the demos
> submitted with ODE or do I have to download it explicitly?
>
> Thanks again!
> Tobias
>
>
>
> Am 31.07.2007 um 16:02 schrieb Jon Watte (ODE):
>
>> The Ray-Car demo uses fairly realistic values. You might want to look
>> at that. There's also an older demo called "carworld," although
>> that's based on some old release of ODE and GLSL, so it probably
>> doesn't compile right.
>>
>> Cheers,
>>
>> / h+
>>
>>
>> Tobias Zimmer wrote:
>>> I might add, that the buggy-example uses -0.5 as gravity, not -9.81.
>>> since I startet with ODE... why don't the examples start with
>>> "realistic" values?
>>> Again,
>>> Tobias
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> ODE mailing list
> ODE at ode.org
> http://ode.org/mailman/listinfo/ode
```