Jump to content


Popular Content

Showing most liked content on 05/08/2015 in all areas

  1. 1 point
    Sort of . I'll try squeeze in some time this weekend. Thanks M
  2. 1 point
    I took a look at your file. A few things I notice: there is no temperature being sourced into the simulation, therefore bouyancy wont work. In H13 and before, the source volume would by default, use density to source into temperature if a temperature source field wasnt present. Now in H14, you must explicity tell the source volume to use the source density field to source into temperature as well. Source volume node > SOP to DOP bindings tab. You also need to account for the gravity in the sim. Currently it's force is greater than the bouyancy force, so the smoke will still go downwards. byuancy1.hip
  3. 1 point
    There is the Sapphire R290 Vapor X version which holds 8Gigs. I was thinking this could help with doing RnD in rather Low Resolutions. Not rendering but Simulating using OpenCL. The OpenCL support is afaik better on the AMD side than on the Nvidia Side of things.
  4. 1 point
    Yup. As of now it works fine. I played only a little bit with the whole GPU acceleration for fluids. So I cant really say anything about that. Most of the Stuff I had to do needed mor ram than the 8Gigs that come with the card… If you have any specific questions I am more than happy to reply.
  5. 1 point
    BTW. I'm still waiting for this "small" wrapper, @magneto
  6. 1 point
  7. 1 point
    The hips don't lie!
  8. 1 point
    Hi all! New version of the setup for H14. The scene is much better organised and optimised. There also some new features which makes this setup actually very useful. Have Fun! DOP_DynamicFracture_H14_v09.hiplc
  9. 1 point
    just force attribute (or fexternal in H12.5) attribute applied in normal direction within SOP Solver should take you most of the way just make sure you update normals (Facet SOP) prior to computing force if you are using CVEX based VOPs/wrangle nodes
  10. 1 point
    here is the full list Stiffness parameters: Primitive, float attributes: shellstretchstiffness, shellshearstiffness, shellweakbendstiffness, shellstrongbendstiffness, solidstretchstiffness, solidshearstiffness, solidvolumestiffness damping Parameters: primitive, float attributes: shellstretchdamping, shellsheardamping, shellweakbenddamping, shellstrongbenddamping, solidstretchdamping, solidsheardamping, solidvolumedamping Surface Mass Density: Primitive Float: surfacemassdensity Volume Mass Density Primitive Float volumeemassdensity Seam Angle: primitive float seamangle fracturepart: Primitive Integer Partitions the object into unbreakable parts. Fracture Threshold: Vertex Float fracturethreshold Rest position Point or Vertex Vector rest initial rest positions Point or Vertex Vector restinitial Target position Point or Vertex Vector targetP Target velocity Point or Vertex Vector targetv Target Stiffness Point Float targetstiffness Target Geometry Point Float targetdamping hard constraint to 'targetP' and 'targetv' Point Float pintoanimation External force applied to each point Point Vector force or fexternal Point Float Animates each point that has value > 0.5 Point Float pintoanimation
  11. 1 point
    I think you have a little too much time on your hands...
  12. 1 point
    If I'm not misunderstanding; you can add a birth group in the source pop, then add an attribute to the birth group $T. Only particles at birth will recieve this attribute, and this will prevent uneccessary per-frame/point calcluations.
  13. 1 point
    you can use an object merge as well. specify your camera in the object transform and it will orient the scene along the -z world axis (as if your camera was pointing down the axis). works like a charm