2 replies to this topic

[ChristianW]

I've just been quickly working on a blood type sim that hugs the collision geo and have a bit of a problem with inheriting velocity.
At the minute it kind of swims over the sphere as it moves even though it sticks to the surface. I tried to inherit the movement of the sphere before solve so it will move with it before solving gravity and dripping down the sphere, something like ($TX + ($VX / $FPS)) or ($TX _+ ($VX * $TIMESTEP)). I've had it kinda doing the right thing, but I loose the nice dripping shapes.

Anyone have a better idea of getting the spheres (read any geo I attach this too) movement applied to the particles before it solves the forces.
Thanks
Christian.

P.s. About the scene. It should be pretty self explanatory, the SOP solver generates a couple of attributes, like impact, Normals and distance that are used in the post POP solver to force it back onto the collision geo. In the Collision SOP there is a PC that has velocities that I'm looking up to transfer the velocities to the particles and multiply by $STICKYDISTANCE which will only effect the particles near the collision object.

Attached Files

•  ldev_stickyBlood_v08_velocityTransferForce.hip   524.44K   46 downloads

[ChristianW]

Bump. Still having issues with this.
Has anyone a reliable way of manually inheriting movement from collision object? I'm going to try it as a force when I get a bit more time to play with it tomorrow.

[pclaes]

If your collision object does not have changing topology, you could sample the uv's of the hit position (Your object will need to have uv's per primitive). Then put the object at the next frame back at the uv position and apply the "local forces & recalculate a new uv hit position for the next frame. This would probably cause stuff to stick/slide depending on how much you blend between stickyness and forces.

As a force works better for things like fluids or particles. But it does give that forcefield look at times.