Feasibility of Fracture Simulation on Bullet engine

vini
Posts: 8
Joined: Tue Dec 29, 2009 6:46 pm

Feasibility of Fracture Simulation on Bullet engine

Post by vini »

Greetings.

My graduation thesis theme was simulating dynamic fractures on rigid bodies on top of NVDIA'S PhysX engine. I'm going to continue my research in the area on my Master Degree and I'm considering to migrate to Bullet Physics engine, since I had a lot of problems with NVDIA'S PhysX engine.

The fracture model is based on continuum mechanics and a Finite Element Method for domain partitioning. I have based my work on the O'Brien papers (Real-Time Deformation and Fracture in a Game Environment and Graphical Modeling and Animation of Brittle Fracture), and it uses tetraedron shaped elements as the FEM basis.

The initial idea is to extend the sofbody simulation to support fractures. Since I am new to the engine, I would like to know if there is support to following features:
  • - Real-time tetrahedra's node splitting;
    - Softbodies' collision callbacks;
    - Dynamic stress measurement due to softbody deformation.
What is the feasibility of this work upon the Bullet Engine? Is anyone interested/working in some related project?

TYA.
User avatar
Erwin Coumans
Site Admin
Posts: 4221
Joined: Sun Jun 26, 2005 6:43 pm
Location: California, USA

Re: Feasibility of Fracture Simulation on Bullet engine

Post by Erwin Coumans »

It would be nice if you could help improving Bullet soft body simulation. The features you mention are currently not available, but all the code is open source obviously so it can be extended/added.
vini wrote: The initial idea is to extend the sofbody simulation to support fractures. Since I am new to the engine, I would like to know if there is support to following features:
  • - Real-time tetrahedra's node splitting;
The soft body collision detection acceleration structure (a dynamica aabb tree, btDbvt) allows splitting of the data. There is already 'splitting' or 'cutting' of cloth. This could be extended for tetrahedra.
  • - Softbodies' collision callbacks;
This can be added.
  • - Dynamic stress measurement due to softbody deformation.
There is not real 'stress' value, but this could be derived from existing values (applied impulse/force).

Thanks,
Erwin
vini
Posts: 8
Joined: Tue Dec 29, 2009 6:46 pm

Re: Feasibility of Fracture Simulation on Bullet engine

Post by vini »

Thanks for the reply Erwin!

I am willing to contribute to the engine if its possible. The stress (strain) values are measured by deformation. I will improve the stress models that I have implemented in order to be as the last O'Brien paper (Real-Time Deformation and Fracture in a Game Environment), which is faster and uses implicit integration schemes. Perhaps a more integrated solver could be discussed.

By now I am a little busy by now, but when I have time I will start studying the Bullet SDK. If it is possible, I would like some help to implement the features that I've listed.

Regards.
01codebit
Posts: 1
Joined: Fri Feb 03, 2012 1:32 pm

Re: Feasibility of Fracture Simulation on Bullet engine

Post by 01codebit »

Hello vini!

I'm developing my master thesis application using Bullet, CGAL, Qt and Visualization Library (VL) dealing with elastic and plastic deformations, fractures and so on.

I'm studying the way to do it: if I will achieve something interesting, I will write here!

Have fun and good work!
Hi
Xcoder79
Posts: 42
Joined: Sun Aug 07, 2011 5:27 am

Re: Feasibility of Fracture Simulation on Bullet engine

Post by Xcoder79 »

I would be willing to contribute to this project Softbody Fracture would be very helpful I thought about doing this awhile ago