A place to discuss everything related to Newton Dynamics.
Moderators: Sascha Willems, walaber
by zak » Tue Feb 10, 2015 1:18 pm
Why there is no plane in collision primitives?
-
zak
-
- Posts: 87
- Joined: Mon Dec 06, 2004 9:30 am
by Julio Jerez » Tue Feb 10, 2015 3:56 pm
because a plane is not a shape that can be expended as a convex shape.
There is however a plane collisions implemented using the collision tree
../applications\demosSandbox\sdkDemos\toolBox\UserPlaneCollision.cpp
-
Julio Jerez
- Moderator

-
- Posts: 12426
- Joined: Sun Sep 14, 2003 2:18 pm
- Location: Los Angeles
-
by pHySiQuE » Wed Feb 11, 2015 2:47 am
It has no volume.
-
pHySiQuE
-
- Posts: 608
- Joined: Fri Sep 02, 2011 9:54 pm
by Sweenie » Thu Feb 12, 2015 3:01 am
Uhm, maybe I don't get it but how can you except at plane to have a volume?
It only extends in two dimensions so how can it have a volume?
In my world a "plane with a volume" is a box.[EDIT]
Ignore the above...
Sorry physique, thought you were the topic starter and that you questioned why planes doesn't have volume...

Last edited by
Sweenie on Fri Feb 13, 2015 3:26 am, edited 1 time in total.
-
Sweenie
-
- Posts: 503
- Joined: Mon Jan 24, 2005 7:59 am
- Location: Sweden
by Julio Jerez » Thu Feb 12, 2015 4:01 pm
the user end side implementation projects the AABB of the colliding shape over the
infinite plane and make a cut out. From the cut out it make a finite plane parallel to the infinite plane, and the call the regular plane convex collision.
To clarify, there is nothing on the convex-convex collision that say you nee a shape with volume.
what is say is that the combined convex shape must be convex with no zero volume.
as along an one of the shape has volume the minkosky difference will also have a volume
this is why collision of a flat pogoing an a real convey can be calculated.
what you can not do is having tow flay face colliding normal on, because ether combine convey will be flat and lose one dimension.
you could if you want do triangle vs triangle collision by singe out of the two input triangle are flat and use a second function for caution contacts. but the you still have to resolve the problem of witch side of the collision is the correct one to resole interpenetration why is the part when the volume is use to decide the direction to resolve the collision.
Two flat triangles will give to intersection but no collision result ion, and I have not seem any heuristic robust enough that justify investing time of that.
-
Julio Jerez
- Moderator

-
- Posts: 12426
- Joined: Sun Sep 14, 2003 2:18 pm
- Location: Los Angeles
-
Return to General Discussion
Who is online
Users browsing this forum: No registered users and 1 guest