Jump to content

stevegh

Members
  • Content count

    123
  • Donations

    0.00 CAD 
  • Joined

  • Last visited

  • Days Won

    2

Everything posted by stevegh

  1. oil and water mix

    Different density's is the way to go. What didn't work for you?
  2. Here is an example of how I would do it. ground_break_3_sh.hip
  3. Skin loses shop_materialpath?

    Use the Material SOP after the skin or copy the attribute over?
  4. My bad you don't need to do the empty volume. Notice also how the operation is set to Minimum and the combination is perfect! vdbCombine.hip edit: Union is also perfect.
  5. You need to create an empty volume that is the size/res that you want and then vdb combine sop to add them in (make sure they are vdbs). I'll try to whip up an example...
  6. Table import and special letters

    Well if the table import in H isn't working I'd say brush up on that Python!
  7. Table import and special letters

    You could also use Python (inside or outside of H) to clean up the unwanted letters for you. Each one of those characters has a corresponding integer representation. Poke around stack overflow and I'm sure you will find help with cleaning it up.
  8. @Atom you are always dealing with volumes in FLIP. When you go to create a mesh from particles it will be converted to a volume/sdf first. This is where I would blend in other volumes but if you need a really large extension it might make sense to use volumes to blend into your lake mesh.
  9. FLIP morph custom forces

    You can do it with POPs but using the volume gradient is the way to go. flipAttract.hip
  10. OFFSET SCATTER POINTS

    You can drop an Attribute Wrangle before the scatter sop and enter: @P += @N * chf("shift"); Then press the 'Create spare parameters...' button to the right of the VEX code box which will create a slider for you to control the offset. Negative values will push it inward. tubeShift.hip
  11. Like Carlo said, $F# or $SF# (#=3, 4, 5...) will pad numbers with # number of 0's. When reading file you an also use filename.`padzero(#, $F/$SF)`.ext if that helps at all.
  12. Modification of the velocity of the deforming geometry

    You can create a custom vector attribute on the broom geo , let's call it disturb, throw some curl noise in there or whatever you want using VOPs, and attribute transfer it to the particles and throw that into the v attribute. Just one way out of an infinite number...
  13. Table import and special letters

    Are you using python to parse the CSV? If not I would do that. https://docs.python.org/2/library/csv.html
  14. Imperfect Voronoi fracture

    Have you tried adjusting the 3D tolerance in the fracture sop?
  15. Ultimately it's going to be a volume, I would just make some other volumes to mix in once you are at that stage?
  16. FLIP / FEM - Mutual Interaction

    Here is a quickie. You will probably want to introduce some sort of drag or multiplication to the incoming FLIP points so it doesn't feel like it weighs nothing. flip_push_fem.hip
  17. FLIP / FEM - Mutual Interaction

    Why not grab the velocity grid from FLIP and transfer over the velocity to the FEM points?
  18. Hot Asphalt - RnD

    No particles do not disappear. Try setting up a simple emit FLIP from object on a sphere with a ground plain, turn on viscosity on the FLIP solver and increase the amount in the FLIP object until it start to look like molasses. Particles may appear if you have resampling turned on in the FLIP solver so if that's a problem turn that off. What you will have at the end is a particle sim you can mesh for the 'tar' part of the asphalt and a velocity field that you can advect particles with using a POP advect particles by volume. I don't have time to set this up right now but it should be fairly straight forward. Godspeed!
  19. Hot Asphalt - RnD

    What about a high viscosity FLIP sim and then advect points for the larger chunks?
  20. "Make circle" SOP?

    I know this is kind of a lame way to do it, but you could always do it with some math. makeCircle.hip
  21. Water drops on surface

    You don't need FLIP at all. All you need is a simple setup to move a point along the surface flow which can be calculated with the SDF gradient and some cross products to get your up and right directions. Then you can add some randomness by rotating around the gradient direction and add a trail sop to get the streak. Assign pscale based off of ptnum/numpt and VDB from particles the result with an SDF smooth and voila. I was able to get this result in a couple hours and while it needs tweaking it should get you started. drips.001.hipnc
  22. [SOLVED]A Pulse With Decay?

    Why not just write your own? Give it a couple of parameters like start frame, decay amount/time, maybe add a type like linear/quadratic etc. Or you could use SciPy.
  23. Redshift plugin video's please?

    You can download the demo, which is watermarked.
  24. In case anyone was wondering, just setting the glue strength to 0 and not switching constraints at all seems to work just fine. It should be noted that the bias factor on the cone/twist needs to be at 0 or some weird shit starts to happen.
  25. Hey guys, I am also trying to switch constraint networks based on impact data. In this file I am using a sop solver to detect when the impact happens, and then I switch from glue (which holds everything together as the object falls) to a cone/twist (so that it can bend and break). The main issues I am having right now are: 1) I am losing a buttload of energy when the switch happens. If there were a way to switch before the actual impact I'm sure that would work... 2) The cone constraints are acting funky as hell, whipping random objects all over the place! 3) I have tried many different methods but I cannot seem to get things to come to a complete rest. I have tried artificial dampening, setting objects @active to 0 when below a certain threshold (this works the best) and playing with the collision padding and penetration threshold / substeps to no real avail. Any help or ideas on how to switch constraint networks, or even just have the rdb fractured object stay together while falling, would be greatly appreciated! (Apologies if this file isn't notated properly but it is fairly straightforward. There are, however, 2 different networks. The glue connects all pieces but the cone lets some small pieces fly). clusterConstraints.005.hip
×