Jump to content


  • Content count

  • Donations

    0.00 CAD 
  • Joined

  • Last visited

Everything posted by pau22

  1. Hi! I wonder if it's possible to use the ripple solver, either before or after meshing a flip sim, to get extra details on the surface. I've tried to transfer the fluid surface mesh positions to a grid, so the geometry stays constant, without much success. I've looked some other threads but all the examples uses a static geometry. Any idea? Thanks!
  2. Ripple solver on a deforming mesh

    Yes, it works! Thank you Juraj! However, I will try to use displacement at render time because transferring the position offset directly to the flip mesh requires quite a lot of resolution. Cheers!
  3. Hi! I'm trying to replace some non packed pieces from a bgeo sequence to a hi-res version. I cannot simulate again so I need to find a workaround to "transform input geometry" in the dop import. I've tried the 'transform pieces' sop without success. Any suggestion? Thanks! transform_pieces.hip
  4. Finally, I've managed to solve the rotation problem using the attribute reorient as it is explained here: https://www.sidefx.com/forum/topic/27105/?page=1#post-125034 The transform pieces sop now works OK with the orient values. Even though the center of each piece replaced, differs a little compared to the low res geometry: Cheers!
  5. Well the names are already the same because I've packed the geo before making any transformation. I need a way to calculate the orient value from normals and position, not sure if it's possible.
  6. Well after trying different things i could match the position from each piece using packed primitives and the transform pieces sop. Also, transfering the normals I get some rotation but it does not match the original bgeo sequence. Maybe it is not the right approach...
  7. Hi! I think it's an easy problem to solve but I couldn' t figure it out : I need the tank to be opened above the water level, so the splashes get simulated outside the tank boundary. I've tried to make the tank opened and add some collisions to keep it in place but the fluid gets quite unstable. Any suggestions? Thanks!
  8. simulate outside flip tank boundaries

    Yes, but I need more space on the sides and if I stretch it in each directions I can't simulate on the same resolution because I haven't enough memory.
  9. Hi! I have a problem in my sim when I try to increase the resolution. I get close to the results I want when the simulation is about 2 million particles but when I increase it (15 M) the results get worse, the mesh looks chaotic and the particles are really sparse. Also some particle has to much velocity. Should I try to delete the particles based on velocity and don't mesh it? Iv'e tried to change particle collision method(move outside), not reseeding particles and no detect under resolved particles but with not much success. Low res sim: High res sim:
  10. Hi everyone! I have a moving camera projecting an image sequence on a fractured mesh. What I'd like to do is keep projecting each frame to each fractured part until they get activated. Finally, the active pieces should have projected the image from the last frame before activating. Do you know some way to implement this? Since uv's have to be projected before running the sim it hasn't occurred me how to do that. [Edit] I've managed to solve the problem. If someone face the same problem here is the solution I've found: 1- Project UV from camera before RBD 2- Run the sim 3- Import it with "transform input geometry" and plug in the original fractured geometry with the material (image sequence) applied. The uv's will remain the same but the material will keep updating seamlessly with the background. Cheers!
  11. Hello everyone! I have attached a scene with a simple setup. This is a result from a rbd sim. I would like to run a flip simulation from this geometry but first I need to find a way to make the original geometry blend with the mesh of the fluid simulation. The idea is to get a smooth transition between the geometry itself (with shaders and textures) and the polygons from the particles. Any ideas how I could achieve that? Thank you blend_test.hip
  12. blend effect between different geometry

    Hello again. I will follow the thread. I've seen previously effects like this before but actually done in composing. Thank you for the feedback!
  13. blend effect between different geometry

    Yes, in that example everything are points (Pyro and FLIP) but I haven't found anything about converting geo to mesh from points like morph target. I will take a look at the video. Thanks.
  14. glue network and deforming mesh

    Hello! I have a prefractured animated mesh and I'm using bullet solver. When I simulate as a static mesh everything goes fine but when I switch to "create deforming active objects" in the RBD packed object I get odd results and the glue network seems not to be working. I've realized that the glue network only is created once in the beginning of the simulation. Should it be updated every frame if the mesh deforms? Thanks in advance!
  15. glue network and deforming mesh

    Ok. Thanks for the explanation. I've set the "overwrite with SOP" flag to 1 and disabled the time shift called "fetch_from_initial_frame", so that every frame it generates a new network. However, it seems that the glue network is delayed 1 frame, so it does not match the mesh because it's created after simulating each frame. Should I use a different approach?
  16. Hello! After transferring a color attribute to some points I would like to apply them a force, as long as they have the blue color transferred. I've tried the following code in the pop force node without success: vector color= point(0,"Cd",@id); if(color.z !=0 ) force*=@id; What am I doing wrong? Thanks
  17. Thank you Ryan! It's exactly what I wanted to achieve.That helped me understand a little more how the particles and attributes work in Houdini!
  18. Hello, I've tried the some more expressions but the points with transferred attributes still move with the other points, so multiplying the force by the index is not a good option indeed. Could it be a good choice to use a point cloud expression ? I have attached a test file so you can see what i'm trying to do Thanks point_color_force_v1.hip
  19. Hello, I would like to know if that approach is possible using PDB and the new grain solver: 1. Create a Granular solid from a VDB animated mesh and make all the grains move acording to its deformation and position. 2. Break some of the constrains so that some particles will no longer follow the VDB mesh. Thank you!
  20. Thanks Ryan, that's awesome! it works really well. You've made my day!
  21. Thanks! The points are sticking pretty well to the mesh! Now I will have to figure out how make the sim stable and the pop source update the grain position each frame