33 replies to this topic

[tjeeds]

Here ya go, just had to do a little work to get the pieces to the origin (in the foreach contained in voronoi shatter) and back (in the dops position node), otherwise the center of mass isn't properly calculated. Pretty simple setup otherwise.

nanocell, on 18 September 2009 - 04:34 AM, said:

Dude, that's great! Would you mind sharing your .HIP file? I'd love to look at your setup.

CheerZ,
Van Aarde.

[tjeeds]

.

Attached Files

•  bullet_voronoi.hip   498.81K   102 downloads

[horizon1231]

by checking tjeeds' example, I follow with a setup.add each piece with $CEX,$CEY,$CEZ att, but my model interpenetrating.

another question,how to define a grid to Ground?

Attached Files

•  bullet_od.hip   606.93K   46 downloads

[Stephen]

horizon1231, on 12 November 2009 - 11:13 PM, said:

by checking tjeeds' example, I follow with a setup.add each piece with $CEX,$CEY,$CEZ att, but my model interpenetrating.

another question,how to define a grid to Ground?

Set your RBD fractured objects "Group Mask" to read in just your pieces (in your case large_*). For your ground go to your RBD object under the Physical tab uncheck "Compute Mass" and set your "Mass" to 0

hope that helps

-S

Edited by Stephen, 13 November 2009 - 10:09 AM.

[horizon1231]

Stephen, on 13 November 2009 - 08:37 AM, said:

Set your group mask to read in just your pieces and not everything.

-S

ok,it works.

but that only works for a while,after 190 frames,pieces blast away.

Edited by horizon1231, 13 November 2009 - 11:49 PM.

[horizon1231]

after frame 190, pieces blast away.very weird

Attached Files

•  Untitled.jpg   107.11K   60 downloads

[almatea]

this may be solwed by apply position dop only in first simulation frame

[rony]

that's because /obj/dopnet1/position1/activation is set to 1, change that to $SF==1 so that it doesn't override the bullet transforms at rest.

[rony]

so I'm in the process of implementing the RBD Glue Object into the bullet solver for our studio.  I'm going to use btCompoundShape for that.  Now I don't want to drastically change the way things are currently set up so I'm going to be using a map of a vector of map to store the compoundShapes separately.  
Something like this: std::map<std::string, std::vector<std::map<int, bulletBody> > >, where the string is the name of the rbdglueobject. So in other words I'll end up with this type of description

|-RBD_GLUE_OBJECT_1
|    |
|    |-RBD_GLUE_CHILD_OBJECT_1
|    |-RBD_GLUE_CHILD_OBJECT_2
|    |-RBD_GLUE_CHILD_OBJECT_3
|-RBD_OBJECT_2
|- ...


This is a bit too long to type but it should be convenient to use the already implemented functions that use map<int, bulletBody>.

Any thoughts/suggestions?


Cheers,
Ron

Edited by rony, 08 February 2010 - 09:57 AM.

[neonshaun]

Anyone else make any progress with btCompoundShape or glue?
Care to share some tips? haha...

-S-

Edited by neonshaun, 25 April 2010 - 07:12 PM.