Deformable Geometry

Just curious.  Has anyone implemented a game or animation with deformable geometry?

Jeff E.

No, but I've been thinking about it for a while.

Have you done any research on it?  If you like I can post here what research I find.

There is a python script for Blender that do deformations… the code in it could be ported to do some tests with jME.

Did you see TestCloth?

This is some kind of a Deformable Geometry i guess. :slight_smile:

Ahh. Here tis:

http://en.wikipedia.org/wiki/.kkrieger

impressive, huh?

Core-Dump said:

Did you see TestCloth?
This is some kind of a Deformable Geometry i guess. :)


I'm looking at TestCloth now.  Thanks!  And thanks to everyone else who posted!
Jeff

I think it's done tx to finite element methods. In statics it's pretty easy but in dynamics I think it's a bit more complicated  ://

Real time deformation of solids, Part 1

Real time deformation of solids, Part 2

@nymon: I planned to read these articles along with reviewing the TestCloth demo.  Thanks for the links.  My wish is to have a racing animation with vehicles showing damage from collisions.  Will reply soon.



Jeff

nymon said:

Real time deformation of solids, Part 1
Real time deformation of solids, Part 2


Their method looks strange  ! It's like they just consider the system in statics (before/after deformation) and in the end they just explain something about interpolating the deformation through time (not sure it would work).
Normaly we should solve for node's velocities and update their position.

There are some other methods using spring-networks.

I hadn't read through the whole article before I posted links to it. Have you found papers describing others ways to do it?

Sorry, didnt check for replies  ://.



This article looks much better:



http://www.engin.brown.edu/courses/En222/Notes/FEMdynamic/FEMDynamic.htm

Looking at ClothPatch.java, how did you guys come up with the interpolation value?  I'm trying it out with an .obj model.  I was thinking about a set value multiplied by the length of the edges between the vertices.  What do you think?



Jeff