Jump to content


  • Content count

  • Donations

    0.00 CAD 
  • Joined

  • Last visited

Community Reputation

0 Neutral

About AiyaPK

  • Rank

Personal Information

  • Name
    Wilson L
  1. Point/SOP level parent-constraint

    thx julian for the tip! I'm going to use your example as a basis and work from there. Your example gave me some pretty good insight into how to use matrices so hopefully I'll be able to find my answer. One question though, in one of your attrib wrangles, you had @m.ay=0.1 I was curious and randomly tried out m.az and m.bx... are these how you access the individual attributes of a matrix in vex? so 'a' is actually the 4th row for translation, and 'b' is for the 3rd row? I tried 'c' and 'd' but they don't work.. Is there a help card detailing this?
  2. As the title suggests. So suppose I have a single point somewhere in space, this would be my parent point. Now suppose I add a child point at (1,1,1) from the parent point. How could I ensure that if I rotate this parent point in my animation, the child point would maintain its distance and 'orientation' with respect to the parent point? I realize parent/child constraining is a feature of chops? but I think I would like to maintain control and flexibility back at the point level. I think this somehow involves matrices and vector math but not entirely sure what's the proper method. My thinking is to give the parent point an arbitrary orientation (let's call it N). There would also be a vector from the parent to the child point. And then maybe there would be a length calculated so it is known how many units away this child is. This would be the rest state. My parent point would get animated, at which point, the parent N would be pointing somewhere else. I would query this new N to compare to the template N, and from there work out where my child point is suppose to be... Is this the proper way of approaching this problem or is there some incredibly simple tool I'm missing for this exact purpose?
  3. Setting up for FEM destruction

    So for the ground plane collision issue (#1), my solution so far seems to be manually setting the 'pin to animation' value to 1 for the 'base' points so they just don't move... Using this as my equivalent for "active" attribute for RBD packed prims. Seems to work so far... Originally I was trying to manually crank up "targetstrength" point attribute but it doesn't seem to do anything, which is when I decided to just use the pin to animation. Will look into targetstrength some more later to see why it's not working...
  4. Setting up for FEM destruction

    Kinda figured out the answer to #2 by changing the fracture threshold and also varying up shapestiffness attribute. However, in doing so, I noticed that adding shapestiffness as a primitive attribute doesn't work. But adding it as a point attribute does?.... From the docs I was under the impression that shapestiffness should be a prim attribute.... Still haven't worked out the answer to #1 yet... though still messing around w/ my test scene. I initially tried wrangling the points below the ground plane to just have a @P.y of 0, but that ends up screwing up the sim entirely.... not sure what's going on there yet...
  5. Setting up for FEM destruction

    Hi, Quite new to using solids for destruction. I have a flying torus smashing into a stationary nuclear cooling tower (a modified tube) that's seated on the Y plane (no points below 0). Used the shelf tools 'fracture solid object' on both, and set up initial velocity on the torus. Things smash and collide, all cool.... My question is, what is a typical workflow for setting up objects for FEM destruction. More specifically: 1) The solid embed sop remeshes the tube, but some of the points are now below the ground plane collider. So at the start of the shot, the ground plane bumps the bottom row of points (or tets I suppose) upwards. How could I have it setup so the points at the bottom just start at y=0? 2) The tower just collapses as the time goes. The stiffness doesn't really do much to stop it either. Is this a matter of setting up target constraints and turning them off right at the point when the collision happens?.... but wouldn't that just cause the tower to start collapsing on itself from that point onwards? 3) Unrelated to this specific setup, but say I wanted to sim a car collision w/ curmpling handled by fem. Not every piece of the car is suppose to crumple, so lets say the wheels and the axles, how would I keep them w/ the deforming car? Is this also a case for constraints? W/ RBD constraints/destruction, we could manually set up constraints and such (thus setup takes a lot of time), I was under the impression from other people that setup time for FEM destruction should be a lot quicker? I know that sounds like a harsh over generalization, but I kinda want to see exactly where in the spectrum FEM falls into since I haven't had much experience setting it up Thanks!
  6. Hi, I wouldn't be surprised if this question has been answered here before (apologies if it has), but i'm not even sure what's the best search terms to use for this problem: Suppose I have a sphere object and I split it to northern hemisphere and southern hemisphere. Suppose I delete the northern hemisphere and add some new point attribute to the "equator" points of the southern hemisphere (which I have labelled as @equator. I would now like to transfer those attributes back to the orignal sphere before I did any deleting/manipulation. What's the best way to achieve this? I recall watching this video tutorial once that used the old Point SOP and plugged the original sphere into input1, and the southern hemisphere into input2. By using the older houdini (i'm assuming the video was using H12.5?...) it's possible to reference the equator points from input2 by using $EQUATOR2 where $EQUATOR is that from the first input. My ultimate question is how would I perform this task using the current Point SOP, and with attributes instead of variables. Is this something that is maybe better served via the point expression in a wrangle sop?... Thanks!