Hi everyone
I was just running some basic fluid tests, when I run into this problem.
Basically I just created a fluid container, a sphere and then made the sphere a liquid object.
So the sphere is emitting a single volume of liquid, as expected and hits the bottom of the liquid container.
After that I get some splashes - also as expected - and some “interesting” behavior. It seems like the amount of liquid is increasing.
After the fluid settles down I have much more liquid volume than the initial emitter (sphere).
Another issue is that when I increase the amount of sub steps, I get even more liquid volume.
Attached are two videos and the hip file to illustrate the problem. Maybe someone can tell me what I am doing wrong here.
Thanks,
Daniel
weird fluid volume behavior
3783 8 3- schyzomaniac
- Member
- 69 posts
- Joined: 7月 2005
- Offline
- Skybar
- Member
- 166 posts
- Joined: 3月 2013
- Offline
- hcmistryvfx
- Member
- 29 posts
- Joined: 1月 2011
- Offline
- schyzomaniac
- Member
- 69 posts
- Joined: 7月 2005
- Offline
I downloaded two new builds (latest production and daily), but I still get this behavior.
hcmistryvfx, if you don't mind, could you re-save my file with your Houdini version and send it back?
Or just add a screenshot of your last simulated frame? I am curious how it looks on your end.
I tried using shelf tools as well as manually building the node network. But I always end up with the same issue that the fluid is “growing”.
Triple checked all parameters, but running out of ideas.
cheers,
Daniel
hcmistryvfx, if you don't mind, could you re-save my file with your Houdini version and send it back?
Or just add a screenshot of your last simulated frame? I am curious how it looks on your end.
I tried using shelf tools as well as manually building the node network. But I always end up with the same issue that the fluid is “growing”.
Triple checked all parameters, but running out of ideas.
cheers,
Daniel
- schyzomaniac
- Member
- 69 posts
- Joined: 7月 2005
- Offline
Skybar, I can only find the reseeding option on the FLIP solver, but not the fluid solver - which I am using here.
Is there a similar parameter that does the same but is named differently?
This is starting to drive me crazy. Can't even do the most simple simulation.
Is no body else experiencing those issues?
Is there a similar parameter that does the same but is named differently?
This is starting to drive me crazy. Can't even do the most simple simulation.
Is no body else experiencing those issues?
- schyzomaniac
- Member
- 69 posts
- Joined: 7月 2005
- Offline
hcmistryvfx, could you check if your point count is increasing or changing at all during simulation?
Mine is increasing steadily.
Mine is increasing steadily.
hcmistryvfx
I opened up file and tested and everything seems to be right.
I am not getting more fluid in my container.
Something might be wrong with Houdini Version so just try with different version.
I am using 12.5.375
- Alejandro Echeverry
- Member
- 691 posts
- Joined: 6月 2006
- Offline
For this type of sims I recommend you to use FLIP solver instead.
But if you want to continue with the fluid solver, decrease the voxel size, and in the markers section activate “build outside markers”, and crank up the number of markers; This ensure a better surface tracking, and the outside markes prevent the gain of volume; Also if you experiment flicker in the surface go to the surface tab and play with the reinitialization rate. This solver is quite complicated for thin fluids, you need to decrease the voxel size to obtain nice results but at a high computational cost.
But if you want to continue with the fluid solver, decrease the voxel size, and in the markers section activate “build outside markers”, and crank up the number of markers; This ensure a better surface tracking, and the outside markes prevent the gain of volume; Also if you experiment flicker in the surface go to the surface tab and play with the reinitialization rate. This solver is quite complicated for thin fluids, you need to decrease the voxel size to obtain nice results but at a high computational cost.
Feel The Knowledge, Kiss The Goat!!!
http://www.linkedin.com/in/alejandroecheverry [linkedin.com]
http://vimeo.com/lordpazuzu/videos [vimeo.com]
http://www.linkedin.com/in/alejandroecheverry [linkedin.com]
http://vimeo.com/lordpazuzu/videos [vimeo.com]
- schyzomaniac
- Member
- 69 posts
- Joined: 7月 2005
- Offline
alejandro, thanks a lot for your suggestions and tips.
Sorry for replying so late, but I just had time to try the things you mentioned out.
While they definitely help to improve the simulation, I somehow remember this was easier a couple of versions ago. And with easier I mean the result would have been more predictable without tweaking so many parameters. Only decreasing the voxel size would already to help to get a better result.
But maybe I remember things wrong - who knows?
cheers,
Daniel
Sorry for replying so late, but I just had time to try the things you mentioned out.
While they definitely help to improve the simulation, I somehow remember this was easier a couple of versions ago. And with easier I mean the result would have been more predictable without tweaking so many parameters. Only decreasing the voxel size would already to help to get a better result.
But maybe I remember things wrong - who knows?
cheers,
Daniel
- Skybar
- Member
- 166 posts
- Joined: 3月 2013
- Offline
schyzomaniacOh my bad, I just assumed you used FLIP. I haven't used that fluid solver much so I wouldn't know
Skybar, I can only find the reseeding option on the FLIP solver, but not the fluid solver - which I am using here.
Is there a similar parameter that does the same but is named differently?
This is starting to drive me crazy. Can't even do the most simple simulation.
Is no body else experiencing those issues?
-
- Quick Links