1
$\begingroup$

I'd like to simulate the following problem:

  1. Define some square, horizontal and hard area onto which cylinders (aspect ratio R) sediment.
  2. Preferably, the source of cylinders sequentially generates N of these cylinders from a large distance above the plane. There is gravity.
  3. The cylinders are hard, if possible deformable (bending), non-interpenetrable, and friction / damping forces are large to stabilize the system (avoid infinite bouncing).

Essentially I need a physics simulation software which handles collisions of many-body systems. The number of cylinders would be around 100'000.

I am in fine interested in the stacking dynamics of the cylinders. A measure of this would be the z-position of the center of mass of the sedimented mesh of cylinders in function of N/A (aerial density) and the cylinder aspect ratio.

I have knowledge of Python, and can work with scripting Blender, and should be able to find my around making scripts (C,Python,...) work on GPUs. Blender apparently incorporates Bulletphysics. The Molecular Add-on for Blender that I tried however doesn't have S.I values of physical units (Gravity is mapped between 0 and 1, I don't know what this is), and I couldn't get the simulation to converge.

Any suggestion would be appreciated !

Improve this question
$\endgroup$
2
  • $\begingroup$ For what context? Does the software have to run real-time-ish (i.e. for a game)? Or is correct physics primary? Do you need hydrodynamics, which are very relevant for real sedimentation (i.e. in fluids)? $\endgroup$
    – AJK
    Jan 2 '17 at 2:51
  • $\begingroup$ Real-time is not needed. Correct physics is what I need. Including hydrodynamics could be interesting, but everything is static, so I think the surrounding liquid comes into play only through viscous damping. Cheers $\endgroup$
    – Mathusalem
    Jan 3 '17 at 15:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.