mrhumbility Posted March 9, 2014 Share Posted March 9, 2014 (edited) Hey, So I'm having an issue where when I have fluids inside an animated container, they wont hold their volume. As soon as the geometry moves they just begin to collapse in on themselves. I have reseeding turned off so I can see that all the points still exist, they just exist on top of each other. If the geometry is not animated it will hold its volume. I have been running into this issue for a while now whenever I try to fill a container with fluid, anybody know what I may be doing wrong? Ill attach a simplified sample file as well. Thanks! note - Changing the velocity type in the collisions tab of the fluid solver helps but the issue still happens. Just slower. collapsing_fluids.hipnc Edited March 9, 2014 by mrhumbility Quote Link to comment Share on other sites More sharing options...
Doum Posted March 10, 2014 Share Posted March 10, 2014 Hey james, I opened your scene and checked what was going wrong. If you want to have a proper velocity collision, since your static object is transferred to volume, you have to change in the flip solver/volume motion/Collisions the velocity type to volume. It will work for not so fast moving collider, but if your object is going to move faster, I know that the flip will start to compress the same way because of the pressure solve and I don<t know how to solve that. I think the better way for fast moving collider is to slow it down for sim collision and retime your sim after. Good luck Doum Quote Link to comment Share on other sites More sharing options...
kartik Posted March 20, 2014 Share Posted March 20, 2014 hi i am new to flip plz let me know how this 3 flip fluids can intreact...??? attached file and image thanks.. flips_interaction.hipnc Quote Link to comment Share on other sites More sharing options...
eetu Posted March 21, 2014 Share Posted March 21, 2014 Just to add, if you do your animation on OBJ level and not within your SOPs (which is often a good idea for non-deforming animation), you don't need to change the collision type. collapsing_fluids_ee.hipnc Quote Link to comment Share on other sites More sharing options...
mrhumbility Posted March 28, 2014 Author Share Posted March 28, 2014 So it seems that animating the geometry on the OBJ level fixes the issue. What I have not been able to figure out it how to make Houdini think that animation is happening on the OBJ level for geometry that is being brought in pre animated. In my situation I have a cup that was animated in Maya and exported as an alembic file. I read this in and then use an object merge to bring the animated geometry into a new clean geo container. Then I read it into the DOP network, but the animation is not being read as OBJ level transformations. Any hints on how to do this? Quote Link to comment Share on other sites More sharing options...
LPriest Posted March 28, 2014 Share Posted March 28, 2014 It's much simpler than all of this, Mainly you're using the wrong kind of collider, delete your static rbd object, in the SOP level of your collision object compute the velocity on your points with a trail SOP then in the populate containers shelf use the source collision tool to create a collider for you're simulation. This will generate a collision and collisionvel.xyz volumes required to correctly calculate your collisions. This way you can also calculate colliders for cached geometry which is the most frequent use case anyway. If you need to do things at high speeed then smiply make sure you cache your collisions at substeps and increase dopsubsteps. The RBD solvers > FLIP solver merging should only really be used for active RBDs to fluid two way interaction. Please don't do the slow down sim and running overlength method, thats very expensive, physically innacurate and a bit of a hack. Quote Link to comment Share on other sites More sharing options...
mrhumbility Posted May 2, 2014 Author Share Posted May 2, 2014 (edited) Hey all, So I spent a long time trying to use the collision volume to act as a container but I wasn't able to have the fluid source node create an accurate enough volume representation to behave the way I needed it to. A friend of mine helped me get this working though and as usual in my experience, it turned out to be a simple drop down. In the fluid solver in the volume motion > collisions tab the velocity type needed to be point instead of rigid. This allowed me to use rbd's for the collision type which was faster than the velocity volume method, and still have the fluids hold their separation. Here is a sample file of how I have it working now. Fixed_Collapsing_Fluids.hipnc Edited May 2, 2014 by mrhumbility 1 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.