Dynamic Paint with Mesh Cache

I simulated a bunch of rigid bodies falling onto the ground.
I then turned them into a mesh cache (.mdd file) and loaded them back into the scene with a mesh cache modifier.
I am now trying to use dynamic paint to generate particles at the location where the individual rigid bodies hit the ground.
This works partially. The problem is getting the paint to disappear after the initial impact.
With normal objects you can use the velocity ( multiply alpha ) to reduce the wetness with decreasing velocities.
However, with mesh caches it appears that the velocity is not taken from the geometry but from the object which of course is allways 0.

Is there a workaround for this or is there a way generate these types of wetmaps differently?

I also tried linking individual objects to the mesh caches vertices and using these objects, I tried using hair particles on the mesh cache, I tried particles with physicas set to none, but blender appears to be unable to detect the velocity of anything that is somehow attached to the mesh cache.

Here are two images. The only thing I changed is the “multiply Alpha” checkbox in the Velocity settings:

Have you tried to play with Dissolve subpanel of Dynamic Paint canvas ?

Yes. The problem is that the chunks come to a rest which means that they keep lying at the same spot.

This means that they keep producing new paint and therefore the dissolve is negated.

That is why I wanted to do it with velocity. If they stop moving they stop producing new paint. This works fine with normal objects but for mesh caches the speed is allways zero which means that they never produce paint. Even when they are moving.

I am not sure that what you want to achieve is possible with current particles.
I don’t think that particles emission will take into account an image sequence, anyways.
DP+Hair particles should work. But DP+emitter particles that should not.

I would probably try to make several small fracture simulations and use a mesh sequence cache modifier supporting an alembic file, instead.

I am not creating an image sequence but using the vertex groups function of the wetmap. The ground plane has a very high vertex count.
I would then use the vertex weight in the density slot of the particles and in the vertex group slot of the smoke sim.
Regarding the several fractures: This is one of 40 or so fracture sims. I can not divide it up any further and alembic is out of the question because it is absurdly slow.