loudsubs Posted July 8, 2011 Share Posted July 8, 2011 Hello All, I'm working on a project that involves a shark that swims in the road. Where ever it swims, the pavement inherits a thick fluid property within a close radius. Once the shark has past through, the pavement goes back to normal. This web image is the concept art that I am using for the look I want: I've been developing a FLIP fluid sim which is pretty close to what I want, but it can still be improved. Here's a flipbook of my progress so far: http://vimeo.com/26154413 Aside from the normal shelf FLIP setup, I've added a sopsolver that multiplies velocity by an attribute transfer. I've also added the gas surface tension microsolver which seems to pull the particles back to their original state. The fluid reforms relatively well, but there is still very subtle movement and jittering. I'm wondering if anyone with a little more experience with fluids could take a look at my scene file and possibly give me some advice as to how to settle the fluid down. The goal is to get it to a standstill and very flat along the top surface if possible. I will be projecting the road from my backplate onto the fluid geometry. I don't know very much about SDFs, but would this be a situation where I could use an SDF of the original shape to gradually have the fluid points attract back to the SDF? And is there a way to have points slowly attract to an SDF, as opposed to just snapping them to it? Thanks for any input! fluidTest02.hipnc Quote Link to comment Share on other sites More sharing options...
bhaveshpandey Posted July 8, 2011 Share Posted July 8, 2011 Hey Arrev, this is how you could get rid of jitters. I have disabled your sop solver for now(its not needed..at least as of now). it uses a rest attribute to seek its position should it be moved by the fin. Hope this helps. cheers fluidTest02.hipnc Quote Link to comment Share on other sites More sharing options...
michael Posted July 9, 2011 Share Posted July 9, 2011 often the best way to do this sort of thing is to simply mesh it out and use SOP tricks to 'harden' the points you don't want to move.... a point sop and a blendshape would do what you want... Quote Link to comment Share on other sites More sharing options...
loudsubs Posted July 9, 2011 Author Share Posted July 9, 2011 Thanks for the ideas guys. @ Bhavesh: cool setup, that's nearly what I want. I'm still decoding the vops, this is a good example to learn from. I wouldn't have known how to to set this up. The rest attribute is a nifty idea. No more jitters. The displaced points try really hard to get back to their original positions. It does create somewhat turbulent movement so I'm messing around with some of the params to try and smooth things. I tried changing the rest attribute value in the tz channel to just $TZ instead of the point expression. I'm gonna keep playing, thanks for having a look. @ Michael: Are you suggesting to try and do the entire effect in SOPs? I orininally was looking into trying to use deformers such as the magnet, but I felt it lacked the organic, fluid movement I was looking for. Or maybe it is possible to calculate the blend based on an attribute (proximity to the shark fin)? So use the fluids sim but blend back to the original point position on frame one (similar to Bhavesh). I'm thinking out loud, lol. I'll be looking into this. Thanks. Quote Link to comment Share on other sites More sharing options...
bhaveshpandey Posted July 9, 2011 Share Posted July 9, 2011 (edited) I dont quite remember the params I promoted on that voppop, but you might want to push down the number of points and threshold distance lookup responsible for transferring the velocity from shark fin points to the particles. That should get you less turbulence(wave like movement) in the areas near the fin. It would be clever to apply a ramp vop over the distance between the queried position and the position returned by the Point cloud filter node (basically remapping the distance so you could explicitly specify the Region affected by the fin), and control it using that, it would give you increased control over the area which should receive turbulence. Edited July 9, 2011 by bhaveshpandey Quote Link to comment Share on other sites More sharing options...
ikarus Posted July 9, 2011 Share Posted July 9, 2011 (edited) since your going to have to patch the fluid surface into the surrounding road geometry you can make a mask attribute on your fluid surface based on velocity that would flatten the points with little velocity to the road surface level, or alternatively just ray sop the those points to the road surface if their velocity is low enough Edited July 9, 2011 by ikarus 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.