Moleculars physics

Excellent. It works great with 2.79.7. Thank you!

Support Blender 2.80

6 Likes

Hi Pavel,

many thanks for the Molecular-Script.
I wanted to test the Self Collision, but nothing happens.
Can you give me a little hint?
http://pasteall.org/blend/index.php?id=51339

Blender from today
Win 7 Pro 64Bit

It is necessary to increase the lifetime of the particles to 150.
And you need to make the size of a Cube object equal to 2x2x2 meters.
And change the particle size values.
https://drive.google.com/open?id=1kFJAoQzARBw6lJUza-qCgs9qS7SChYla

This is so Great!
Thank you very much Pavel.

Pavel!! Thank you so much!!! I was hoping someone would take over the script!

Pavel,

I am unable to get the script to run, this is the error in the console:

I’m running 2.80.44 on Windows 10

Read prefs: C:\Users\semil\AppData\Roaming\Blender Foundation\Blender\2.80\config\userpref.blend
found bundled python: K:\v44-blender-2.80.0-git.69b2f5268114-windows64\2.80\python
cmolcore imported with success! v1.01
Molecular Sim start--------------------------------------------------
Traceback (most recent call last):
File “C:\Users\semil\AppData\Roaming\Blender Foundation\Blender\2.80\scripts\addons\molecular\operators.py”, line 60, in execute
simulate.pack_data(True)
File “C:\Users\semil\AppData\Roaming\Blender Foundation\Blender\2.80\scripts\addons\molecular\simulate.py”, line 18, in pack_data
for psys in obj.particle_systems:
AttributeError: ‘NoneType’ object has no attribute ‘particle_systems’

location: :-1

location: :-1

Can you give a link to the blend scene? So I can reproduce the error.

Hello,
first of all thanks to everyone contributing to this amazing stuff.
I ran into some issues trying to have my particles behave like I want and got some other questions. I would be very glad if someone can help.

  1. I got a scene with ~23 000 particles dropping into a cylinder with a cone attached. Whenever I drop my particles they don’t stop to move. They keep moving, sometimes jittering.

Animation: https://gfycat.com/understatedadorableboubou

I used high Substep settings, compared to some scenes with much more particles I found online. Maybe it is hard to solve because I use small particles?(Blender size:0.0025 = real world:4mm) Need them in this size in order to have the scene real world size. For their behavior I aim for something like the first 2 examples of this video: https://www.youtube.com/watch?v=_5QkPnPDcfc

I uploaded the .blend with the settings from the video. I ran the same simulation with higher friction(0.66) and higher Substep(up to 512). Doesn`t make too much difference and the particles still keep moving till the end.

.blend: http://pasteall.org/blend/index.php?id=51424
  1. Is it possible to do the Substep calculation while keeping the same particle size in the system? Otherwise I can`t test with less particles and then calculate for higher resolution.

  2. CPU usage: Molecular doesn`t use much of my CPU, even though it should support Multithreading. What I am doing wrong? I attach a comparison of the usage while Molecular-Simulation vs. Cycles-Render.

Images: https://imgur.com/a/KjBFnva
  1. Is there an easy workflow for 30fps to 60fps?

Thank you and keep up the great work!

Sorry for the edits and strange formatting. There are some limitations as a new user.

Hi Pavel,

I have the same error as steve.

Traceback (most recent call last):
File "E:\Blender-28\scripts\addons\molecular\operators.py", line 60, in execute simulate.pack_data(True)
File "E:\Blender-28\scripts\addons\molecular\simulate.py", line 18, in pack_data for psys in obj.particle_systems:
AttributeError: 'NoneType' object has no attribute 'particle_systems'

Blender from today: blender-2.80-a813e259d630-win64
Win 7 Pro

particle_systems-no_attribute.blend (1.1 MB)

Very happy to see this one ported. Great job

I don’t know how I missed this. 2.80? Fantastic. Thanks, @Pavel_Blend

Thank you for the great addon!
Just my scene made in Blender through Molecular Script and my addon Bravais Lattices Creator
Atomic visualization of HSS in contact with synthetic diamond crystal


Bravais Lattices Creator:
https://blenderartists.org/t/bravais-lattices-creator
http://alexiy.nl/category/blc/
1 Like

a few fixes for the molecular addon, should work in 2.80 now again :slight_smile:

6 Likes

Thank you scorpion81,
is working again.

Unfortunately, I can’t even handle a simple collision.
The objects (cubes) go into each other.
From picture 147 the cubes move very fast away from each other?

Is it a bug or a mistake on my part?

Are there any basic tutorials for Moleculars physics?

Blend:
http://pasteall.org/blend/index.php?id=51935

Blender from today (741641f4c3b8-win64)
Win 7

I have made a fix which correctly resets the cache if a molecular re-simulation is being invoked. I tested your blend with that version and it showed different collision behavior compared to the bake. This is in 1.0.8 whose source is on github already. I also made a version bump inside the cython binary (but no further change). I will post the “proper” release zips later. In the meantime you could download the latest python files from github and use the existing .pyd file in order to test.

Update: 1.0.8 release zips are up ! :slight_smile:

Some older video tutorials…

4 Likes

That’s great!

Works perfectly, only very few overlaps in the objects.
I’ll have to do a little work on the parameters.

Thanks a lot for the script and the video links, scorpion81.

Took me a long time to find a working version of this script, all the links to it online only lead to the old devs’ page where it runs out at 1.0.7 (which totally doesn’t work).

Your version seems to work well though, I really hope you keep it updated in the future as Blender would be completely crippled if they ever lost this script. Honestly don’t know why it’s not in the basic toolbox like it is with C4D.

1 Like

Does it work on the newly released Blender 2.8? I tried but am getting an error when baking the simulation

It worked for me, but you have to make sure you’re using the actual new version. It’s a pain to find, because pretty much all of the links for Molecular take you to the original github which only has out of date files. You need to be on 1.0.8