Scratch Posted March 22, 2015 Share Posted March 22, 2015 Hey Folks, I am currently working on a water-drop simulation, small scale, 8cm x8cm tank, 1cm water drop. The sim turned out fine, but I was asked to emitt some additional droplets from the splash crown. My first idea was to use the white-water solver to emit some particles from the crown, since it gives quite some control on emitting particles from fluid-sims, but I can't get it to work on my scene-scale. It's allways offsetted from my sim. I have attached a scene file to illustrate the problem. Any ideas what goes wrong there? In addition, I'd be curious if you guys have any idea how to emit some extra droplets from the splash crown? I'm a bit stuck here... My ideas so far (not sure if any of them works, they are all rough and untested) - whitewater solver: emit a few particles -> merge them with the original sim -> give the whitewater particles a much much bigger pscale -> mesh it alltogehter -> white water particles appear as droplets - custom particle sim: filtering the fluid-sim particles by expressions (somehow the same as the whitewater source does it), and do a custom particle sim -> merge it with the original sim and mesh it, or mesh it seperately...whatever works best. - droplet function on the flip-solver: you (may) get some droplet-particles incl. droplet attribute -> pscale bigger on droplet particles -> mesh them together with the sim, so the droplet particles get meshed as visible, bigger water drops ->? Thank you in advance for your time and help guys! Looking forward to your answers! white_water_test_01.hipnc Quote Link to comment Share on other sites More sharing options...
sergio Posted March 22, 2015 Share Posted March 22, 2015 Hello Philipp, I don't think you need white water at all. I made a couple changes in your file: -added an initial velocity to sphere -scaled that velocity in source volume in autodopnetwork -added surface tension -decreased smoothing i think if you decrease particle seperation you may get some more droplets. just my thoughts. Didn't go through droplets settings but it may work. white_water_test_02.hipnc Quote Link to comment Share on other sites More sharing options...
Scratch Posted March 22, 2015 Author Share Posted March 22, 2015 (edited) Nice! Thanks alot for that quick reply! Thats certainly an option which i will consider/show my Sup tomorrow! Not sure if that is the result the supervisor/client is after...they showed me pictures like this: http://th01.deviantart.net/fs13/200H/i/2007/180/c/8/splash_crown_WIP_by_chocozilla.jpg Don't know if that is even possible. I was able to get droplets/tendrils which shoot off the main crown structure, but they never form little-droplet-spheres, its more like small cylindric shapes, or misshaped random forms. Seems to be very hard to achive... Surface Tension 150? Wow, thats a huge number...in my sim at work, I choose values around 1-5 (forgot to put it in my example file, so nice thinking man!). Is it common to use such a high surface tension? I was wondering what number to put in there, and I seem to be way off Edited March 22, 2015 by Scratch Quote Link to comment Share on other sites More sharing options...
sergio Posted March 22, 2015 Share Posted March 22, 2015 I know Alejandro echeverry has a custom microtension solver for these stuff but if you are in a hurry i remember realflow had a plugin for that. Surface Tension 150? Wow, thats a huge number i remember i saw yancy lindquist was using high numbers like that in a sim. thats why i tried gong high like that . Any way hope you figure out i will be following this thread. Quote Link to comment Share on other sites More sharing options...
Scratch Posted March 22, 2015 Author Share Posted March 22, 2015 (edited) Very very interesting, i didn't know Alejandro's work/paper. The tests he uploaded on vimeo look amazing! I think something like that is what I am missing. He writes in the vimeo-comments that this is far more complex than the surface tension microsolver (which is clear when you watch the videos). Surface Tension Values: Well, if a guy like Yancy Lindquist is using these values, it like this, it can't be too wrong can it? EDIT: In one of his videos ( ) he mentions he uses values of 3 "which is pretty high"...not sure now...for me personally...values of around 1-5 would make sense, because the initial value-range for this gas-microsolver is 0-1 by default...150 may be a crazy high value, maybe for certain special cases, but i guess...not for the standard sim. But I am just guessing here, so I may be completely wrong... Anyway, thanks sergio, I'll stay on it and let you know when I have some progress to show. Edited March 22, 2015 by Scratch Quote Link to comment Share on other sites More sharing options...
Diego A Grimaldi Posted March 23, 2015 Share Posted March 23, 2015 Hey Philipp, Without opening your file but just reading the conversation, this thread may be useful http://forums.odforce.net/topic/18111-flip-smorganicsheeter-effect/page-4 Quote Link to comment Share on other sites More sharing options...
Scratch Posted March 23, 2015 Author Share Posted March 23, 2015 Awesome Link, this is exactly what I am looking for, but it seems that the asset of Alejandro is still in development not ready to be released yet. Do you know any available tools to use at the moment? I found Sheeter Deamon for Realflow (Fusion CI's somehow inhouse solution for fluid sheeting), and I found 2 interesting papers, but both are not implemented in houdini (yet...I hope that this gets more attention, since it is something we all need and noone can actually do it to full extend) http://cg.informatik.uni-freiburg.de/publications/siggraphasia2013/2013_SIGGRAPHASIA_surface_tension_adhesion.pdf http://vacation.aid.design.kyushu-u.ac.jp/~and/sheetflip/index.html Any ideas? In my case, the sheeting (holes in the mesh) is not the problem, i am only missing the spherical droplets which detach from the main splash body, and that seems to be impossible without a more accurate surface tension model. But I had an idea to use the droplet-function of the flip solver to get me some seperated droplet particles with droplet attribute on them. For droplet particles (droplet attribute > 0.8 or so) i will set a very high pscal, and for non-droplet particles, I will leave pscale untouched....doing so, i hope to get some bigger droplets while meshing the sim. I let you guys know if that works! Cheers so far for sharing all your thoughts! Odforce rocks! Quote Link to comment Share on other sites More sharing options...
Diego A Grimaldi Posted March 23, 2015 Share Posted March 23, 2015 Hey Philipp, On the link I posted there's an awesome file posted by Pavel...You should look into that one...I think it may help you solve your problem. Good luck! Quote Link to comment Share on other sites More sharing options...
Scratch Posted March 23, 2015 Author Share Posted March 23, 2015 (edited) Ah ok? I musst have overlooked the file, thanks for the reminder. I'll look into it right away! EDIT: hmm...it seems to be created with Houdini Indie, can't open it with H13 ... but the sheeting is not so much the problem of my sim (that somehow works, i get relatively few holes in my mesh because of a bigger influence scale value on my vdb from particlefluid node. Or is Pavels solution also handling the droplet part?) What I was asked to do is 1) generate some more spherical droplets which break off the main splash and in addition 2) to emphasize the "boarder/rim" which forms at the edge of the splash when fluid particles stack up. Edited March 23, 2015 by Scratch Quote Link to comment Share on other sites More sharing options...
Diego A Grimaldi Posted March 23, 2015 Share Posted March 23, 2015 I think his file helps more with the border than the droplets...weird, I was able to open it in H14.. Quote Link to comment Share on other sites More sharing options...
Scratch Posted March 23, 2015 Author Share Posted March 23, 2015 (edited) I guess our H13 build is older than the Indie Version (Indie was released after H13 as far as I remember), for H14 (e.g. apprentice) it works fine, converting the actual session to a Indie session. But hey, thank you again for your support, I highly appreciate your efforts! I will keep testing and experimenting! Edited March 23, 2015 by Scratch Quote Link to comment Share on other sites More sharing options...
Scratch Posted March 23, 2015 Author Share Posted March 23, 2015 (edited) Hey again, I managed to emit additional droplets from my splash Workflow abuse a whitewater source node to filter the liquid particles so only the splash edge is left (based on speed and curvature, I did not use accel and vorticity). This is the source for a simple particle sim. I emit a few points, gravity, kill below -0,5 y position to clean up and thats it. Then I use the copy sop to copy size-randomized point clouds onto these particles (using the pointreplicate sop and some stamp expressions). Afterwards, these new particles get merged with the liquid sim particles and everything gets meshed together. Clearly, it's a fake, but it works pretty good! Edited March 23, 2015 by Scratch Quote Link to comment Share on other sites More sharing options...
Diego A Grimaldi Posted March 23, 2015 Share Posted March 23, 2015 great, would u mind sharing ur setup in case someone else needs it? Quote Link to comment Share on other sites More sharing options...
Scratch Posted March 23, 2015 Author Share Posted March 23, 2015 (edited) Sure, I will! I can't share the actual production scene, but I will put up a little demo-file which illustrates the workflow real quick. Edited March 23, 2015 by Scratch Quote Link to comment Share on other sites More sharing options...
Scratch Posted March 24, 2015 Author Share Posted March 24, 2015 Here is the hip-file as promised. Thanks again for your support folks! I am still working to find a method to get the thicker border edge of the splash. Still no progress there...but I'll keep you posted. additional_droplets.hip 3 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.