nCloth balloon puzzle pieces?

#1

Hi.

Just wondering what anybody’s thoughts on using nCloth dynamics as a modeling tool?
For example, if I wanted to have balloon-like puzzle pieces that if one piece were removed the adjacent pieces would fill in the voided space but also know to collide with each other?
(Also, the entire puzzle is constrained to sphere).

Any thoughts would be appreciated.

Thanks

#2

This could be done using a membrane node… see attached scene. It basically executes a cloth solve for a few steps rather than solving over time. To create this scene I created some puzzle pieces then did poly:Combine ( with history) and typed “createMembrane”. On the membrane node I used pressure along with restLengthScale to make the pieces inflate and selfCollision so they would intercollide. One needs to do mesh combine and use self collisions because there is no ui for membranes to collide with other membranes. You can use “collideMembrane” to collide with passive meshes, however. The position of the center piece was animated to show the behavior.

If you want the pieces to expand in a particular way you could model the mesh more densely in regions you want more expansion when using pressure.

#3

Hah cute!

Now why is this hiding outside the Dynamics forum? :wink:

#4

Thanks Duncan.

Do you think it doable that each puzzle piece could also be a metaball blob (particle system)?
Or inconjunction with the membrane node?

Like in this here: http://vimeo.com/58267375 ?

#5

You can’t make a blobby output mesh into cloth because the topology of a blobby mesh typically changes over time, and the cloth requires it to be fixed. However if it is just a collision mesh, then changing topology can work( although collisions are not as reliable as fixed topology ).

The output of the cloth mesh could be turned into a blobby mesh by connecting the cloth mesh to a blobby particle system( each vertex becomes a blob). If the then converting the particle system to poly.
connectAttr -f polySurfaceShape1.outMesh nParticleShape1.inputMesh;
setAttr “nParticleShape1.ignoreSolverGravity” 1;

expression -s “nucleus1.startFrame=frame”;

#6

http://forums.cgsociety.org/showthread.php?p=7652686#post7652686

Just to post it in the right place.

#7

Duncan,

Do you think that membrane handles multiple mesh collisions (and self collisions) in a better way than nCloth layers?

Thx

#8

The membrane node uses nCloth, although a limited subset of nCloth. However it evaluatates time instantly, so it is more like a deformer rather than a dynamic simulation. There is no need to cache a simulation to scrub the time slider, however the resulting solution can jump around more. If you really need something that dynamically responds to collisions then using cloth may be better.

#9

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.