ikarus Posted August 3, 2010 Share Posted August 3, 2010 (edited) for rest positions why not just do what the liquid solvers do and have hav a rest variable updated every X frames in the sop solver? Edited August 3, 2010 by ikarus Quote Link to comment Share on other sites More sharing options...
ikarus Posted August 3, 2010 Share Posted August 3, 2010 for rest positions why not just do what the liquid solvers do and have hav a rest variable updated every X frames in the sop solver? changed your example file to reflect this <green nodes> Quote Link to comment Share on other sites More sharing options...
ikarus Posted August 3, 2010 Share Posted August 3, 2010 changed your example file to reflect this <green nodes> without attaching the file...x_x new force scale d.hipnc Quote Link to comment Share on other sites More sharing options...
Macha Posted August 3, 2010 Author Share Posted August 3, 2010 without attaching the file...x_x Yes, I did exactly that previously. It was my first idea. But I had so many update problems that I worked around it. The only difference I can discern between your and my attempt is that you used a 3-float and I used a 3-vector. I'll have to test it a little and see if it is stable. Quote Link to comment Share on other sites More sharing options...
Macha Posted August 4, 2010 Author Share Posted August 4, 2010 Instead of using the Channel Editor to make my own ramp curve I thought I will go the other way around and make a curve from my function encoded in the Vex nodes. I did this by copying the parameters into a spare channel in a null node. This works except that the sampling is way too low. Is there a way to increase sampling in a a specific channel? See attached images. A hires version from an external app, and the Houdini low sampled res. If I can upres the samples it should look OK and enable me to predict the force behavior as I set it. Quote Link to comment Share on other sites More sharing options...
edward Posted August 4, 2010 Share Posted August 4, 2010 Make sure your Channel Editor is set to display at the highest quality. After that, make sure you're driving it with $FF, NOT $F. Quote Link to comment Share on other sites More sharing options...
Macha Posted August 4, 2010 Author Share Posted August 4, 2010 (edited) Here is today's version. I can simulate 30000 particles with this in a reasonable time. The rest force is now proportional to the main force so that it is easier to get a stable setup. I have some ideas about a better and simpler force function that I will try next. It's not a clean file, so if if anybody is looking at it, sorry for the mess. After that, make sure you're driving it with $FF, NOT $F. $FF! That's it! Owner.4684.hipnc Edited August 4, 2010 by Macha Quote Link to comment Share on other sites More sharing options...
Macha Posted August 4, 2010 Author Share Posted August 4, 2010 Hmmm...thinking aloud here...hmmm... At the moment I calculate the magnitude of the force vector with a function. Perhaps it would be better, more stable, precise, customizable and quicker if I was to use a stress tensor instead. Has anybody done this before? I'm shaky with tensors to say the least, but maybe...maybe... Quote Link to comment Share on other sites More sharing options...
Macha Posted August 6, 2010 Author Share Posted August 6, 2010 (edited) I found a much simpler, faster, stabler and more powerful way to melt and crumble particles. I was thinking a bit statistically and figured out I could just base it on density instead of more complicated forces. I think if I can spend some time on this I could make some interesting effects. Perhaps I could reseed points into this based on density, or change their size, etc. If needed I could sum up vectors again and find out if particles are on an edge, or estimate density based on distance, create anisotropic advection (made up term but I believe I could justify it) etc etc. All very very interesting!!! Edited August 6, 2010 by Macha Quote Link to comment Share on other sites More sharing options...
Macha Posted August 6, 2010 Author Share Posted August 6, 2010 Uups, forgot to attach the file... densityb send.hipnc 1 Quote Link to comment Share on other sites More sharing options...
Macha Posted August 6, 2010 Author Share Posted August 6, 2010 and images... Quote Link to comment Share on other sites More sharing options...
ikarus Posted August 7, 2010 Share Posted August 7, 2010 easy density calculation using a pc iterate loop, can base ur forces off of that or w/e countWhile.hip Quote Link to comment Share on other sites More sharing options...
Macha Posted August 8, 2010 Author Share Posted August 8, 2010 (edited) easy density calculation using a pc iterate loop, can base ur forces off of that or w/e Ah, I see, identical to my previously posted setup. You're cheering me up, great minds think alike! In fact, if it is just simple radius based density the proximity pop is probably sufficient. Edited August 8, 2010 by Macha Quote Link to comment Share on other sites More sharing options...
Macha Posted August 9, 2010 Author Share Posted August 9, 2010 (edited) Works a lot better and more stable now, taking advantage of vopsop multithreading pretty nicely. It has a very rudimentary volume preservation ability. As if by magic it creates these interesting mushroomy vortice fluid motions as well, especially at hires. Edited August 9, 2010 by Macha Quote Link to comment Share on other sites More sharing options...
Macha Posted August 12, 2010 Author Share Posted August 12, 2010 Here is the latest. The particles start crumbling from the surface, tumble along the outer edges and fall down, hopefully preserving a notion of volume on the heap below. I'll give this a rest for now. Mawi, Pclaes and Putoparri, thanks for helping me with this. Vimeo is here: Quote Link to comment Share on other sites More sharing options...
ffyanwei Posted August 13, 2010 Share Posted August 13, 2010 wonderful! i like it Quote Link to comment Share on other sites More sharing options...
andreu.lucio Posted August 13, 2010 Share Posted August 13, 2010 Wow! Great results. Quote Link to comment Share on other sites More sharing options...
ikarus Posted August 13, 2010 Share Posted August 13, 2010 lookin pretty good, you should try making a building out of particles and seeing how it behaves. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.