I’m modelling a realtime physics simulation of a Mousetrap Board game.
Using the 2.40 test builds and Bullit physics how would I get a ball bearing to drop faster without it screwing up the collision detection. In my current scene the ball is painfully slow (the scene still has a high FPS). I can get the ball to go faster if I angle the surfaces upon which the ball rolls. But I’m not too keen on that approach as it is distorting the accuracy of the models. Here’s a link to the scene if anyone fancies a look at it?
Arr Matey!! What you can do is turn off the collision for all the faces, and add invisible platforms for the ball to roll on just the way you want it.
I plan to do this eventually and I have turned off collision for all the faces that should not come in contact with the ball. Currently my computer is powerful enough to run the simulation at > 60 fps. I don’t think the scene is cpu/physics limited (but I am willing to be corrected)
xarton As the bearing falls down the rickety steps? Try increasing the restitution of the steps, as they’re actually quite elastic.
I have played around with the material properties and I cannot get a serious speedup.
I think it might be time to ask Erwin for some suggestions
I like the ‘invisible path’ idea, at least in certain places. If the movement is slow or jerky at particular places on the course, then the machine may indeed be CPU-limited in those areas.
It is not a performance problem at all. In the end, the slope really doesn’t go down in the corner, it actually goes slightly up…
I fixed a couple of things in Bullet-Blender integration. Hooked up the ‘formfactor’ (an inertia tweak), friction, restitution, and some scaling related things.
However, the rounded ramp really doesn’t have a proper slope in the corner. Please try to fix the artwork, it should roll much better. Try to play with the ‘formfactor’ Also, you really don’t need to switch off collision detection on all the other faces. It is not a performance problem.
Using your new build the ball moves much faster now. I made all the rolling surfaces point down instead of up and I reduced the form factor of the ball.
I have also added a constraint to the see saw which gets hit by the new larger ball. The code for the constraint is taken from your p2p demo. I’m ignorant of the correct terminology but this appears to be a point constraint (I think Maya calls them nails?). Obviously for a see-saw I need some kind of hinge constraint. Has this been implemented in Bullet yet?
If anyone is interested here is the latest blend file. Use the left and right cursor keys to rotate the scene.
Yes I could use an IPO but I’m trying to model the scene as a realtime simulation using the new Bullet engine. Faking it with Ipos will be a last resort!
