Tomas Slancik

try unchecking Advanced/Grain Collisions/Assume Uniform Radius

your setup looks good

HypoSection
... I've tried:

v@avoid_force = {1, 0, 0};

But it's still 0, 0, 0 in the geometry spreadsheet.

this would be {0,0,0} only for particles with 0 velocity as there is an early return for those , otherwise you should see {1, 0, 0}

as for

apart from 0 velocity this also depends on f@pscale and whether the particles are actually close enough to impart the avoidforce

It does work, maybe bump up the values a bit to see

your setup applies divergence only in source area, but you can also test on different setups, like a simple setup where you feed the simulated temperature (scaled as needed) to divergence to get it expanding in hot areas etc...