Hi,

I am trying to replicate my usual fluid workflow from Naiad in Houdini and I have a trouble with “surface” distance field.

I reduced it to the minimal example (attached below).

Scene is simple. Sphere in (0,0,0) with radius 0.5 as an emiter. No force is applied, no gravity etc.
I am only using Gas Field Wrangle (is it the best suitable node for it? gas in the name is little bit confusing) to modify velocity field directly.

code is simple:
if( @surface > -0.25)
{
v@vel = set(0,0,0);
}
else
{
v@vel = set(1,0,0);
}

So i am expecting move initiated from the inside in the half of the radius.
And it is true with particle separation == 0.04.
But when i lowered the separation to 0.02, i got totally different behavior (no movement at all).
For me it looks that the SDF value is affected by the separation( and voxel size) but in some strange way.

So the questions:
1) do I understand it correctly?
2) is SDF value resolution indenpendent and in world units?
3) is it the good workflow to use Gas Field Wrangle on the second solver input?

thanks for Your help

Jirka

sorry the topic was shortened, it was about particle separation

another issue in the attached example is, that if you disable visualisation of the particles, only velocity vectors enabled, there are some strange irregularities when replayed. I hope that it is only in the visualisation, and not the artifact of the dynamically resized fields…

edit: it is also in the sim (frame 179)