I have been trying a few various tutorials on the web and each time I am having trouble with collisions.

I am creating a simple sphere, which I then click on the geometry node and click 'FLIP fluid from object' from the shelf, this creates the autoDOPNetwork.

In another geometry node I create two Spheres (one large, and one small) and use a boolean to subtract the small from the large. I then click on the geometry node for this and create a static object. This updates the autoDOPNetwork accordingly.

On every tutorial this collision appears to work first time. Mine is not, and I am not sure why.
I would expect the particles to fall from the fluid object, and when they hit the inside surface of the collision object, they would create a splash and eventually settle.

I have tried looking at the static object in the autoDOPNetwork and if I use the collision guide with ray intersect I can choose a low division size to give me a solid looking collision guide. This seems to work with regards to particles not penetrating my container sphere, but my next problem is the particles appear to not settle. As the 'splash' settles down the particles slowly disappear. I would expect them to settle down so I see a visible volume of liquid at the bottom of the sphere. Instead I only ever see a thin layer about the collision surface. Why does this volume not increase?

Finally, what are the best particle settings for realistic water? I have played with this a bit, and whilst computationally heavy, the particle separation of 0.02 is good, what size should the particle radius be? Are there any other useful settings for realistic water?