Gravity value

dreamkatana · Post by **dreamkatana** » Mon Jul 11, 2005 8:16 pm

Hi, it's amazing this new Physics System, but I notice that the gravity (9.8 ) is not working like in the real world, the objects are follow too slow, or maybe the old SUMO Physics that were doing they go fast, I do not know.

Nice work.

Erwin Coumans · Post by **Erwin Coumans** » Mon Jul 11, 2005 11:06 pm

Apparently time is 'lost', due to some timing issues in the blender engine. I will ask Kester about this.

calli · Post by **calli** » Tue Jul 12, 2005 4:33 pm

Hi Erwin,

cool work! Very stable results on the Wall demos.

I also noticed the slow-motion falling (it was even an issue with solid when scene scaling was weird).

However: if you enable Game->"Enable all Frames" in the menu, all falls reasonably fast.

Regards,
Carsten.

Erwin Coumans · Post by **Erwin Coumans** » Fri Jul 15, 2005 2:52 pm

can you test the latest 15 july 2005 build ?
I'm waiting for Feedback.

calli · Post by **calli** » Mon Jul 18, 2005 12:20 pm

Hm.

Without "All Frames" it still looks a bit slow. But maybe I this is because there is no real comparison and scale. Nobody of us has seen 2x2x2m boxes with 1kg weight fallin on a stack

I guess you use 1m equals 1 Blenderunit (be)?

What goes into the simulation? I guess:

- mass
- damping (but somehow strange, 0 seems to be most damping, 1.0 no damping)?

- DYN from Material? (could not see a difference yet)

- more?

Small objects seem to have the tendency to fall through a plane/box if the reach a high speed (well 10cm box 1kg falling from 20m)

If you remove a box from the stack the others seem to be quite sticky, I would expect in reality that the ones above the gap would fall nearly with not much friction.

I need to get more feeling for Bullet physics.

Cant wait to have the motion actuator working to shoot something through a stack of boxes...

Guest · Post by **Guest** » Mon Jul 18, 2005 1:15 pm

>Small objects seem to have the tendency to fall through a plane/box if the reach a high speed (well 10cm box 1kg falling from 20m)

The actual time of impact (continuous collision) is not enabled yet, althought the whole infrastructure and code is there. This will prevent missing collisions.

- friction is unreasonably high at the moment (I use a hardcoded constant).

not all things are hooked up actually. I focussed on your statement on open source:

"release early, release often!"

Thanks Calli !

calli · Post by **calli** » Mon Jul 18, 2005 2:53 pm

I guess the Guest is Erwin

Thanks for your response, I will test further.

Carsten

kjknohw · Post by **kjknohw** » Fri Jan 20, 2006 5:22 am

If you want objects to have initial velocities, just use overlap. The object is overlapping an immobile object, and will shoot out.

Real-Time Physics Simulation Forum

Gravity value

Gravity value

new build 15 july 2005 fixes gravity ?