Blender 2.59 molecular script v026

Another test with my python molecular script.
For this one I fill suzan with particles.
I modifing my script to get independant stiffness for shearing and
I put density of material instead of guessing a weight approximatly.
With density and approximative volume , the real weight is founded.

1 blender unit = 1 meter
1 blender weight unit = 1kg

Stats:

Molecules: ±212 000
Simulation Time: 5 days
Substep: 150
Render Time: 3h30

Category:

People & Blogs

[anpaw76]

This looks absolutely amazing! And, well - 5 days sounds quite serious... Did you use normal PC for that?

[PyroEvil]

@anpaw76 Processor: Intel Quadcore 3.2Ghz ( i7 server processor) 24gig of ram. The problem here is not really the computer. It's because of : I did it in python , python use one processor, I use blender sph for self-collision and I'm not a coder. I did it just for fun and not sure to find the faster algorythm. Now I rewriting all my code to be "standalone" and Blender is just here to display and render the particles.I done all the physics by myself including self-collision(see my recent videos).

[PyroEvil]

Erratum: Lagoa is really faster then that. Probably did the same simulation in hours when you know how to optimize it!

[d3sertst0rm]

Just what i was looking for in this script o_o And a 5 day sim time O_O

[PyroEvil]

@d3sertst0rm Yes , 5days it's a lot of time :) I probably do some test with module in Cython to check if I can optimize and get better result. It's take 50sec per substep to calculate but from this 50sec , 10sec is from blender to do its SPH (I use it to do self collision between particles only) , collision with obstacles and refresh all this particles. But a collegue did a amical competition with me by doing similar test with XSI Lagoa and it's take a lot of time too(1day I think,not sure)

[MoolahNasreddin]

Cooool! )) Looks like a big cookie is crashing!

[amuddypizza]

COOL! Looks like legos.