I am working on a similar project and wanted to know the basics groundwork one might think of in LBM to simulate a process like precipitation.

The following are my concerns:

(i) There is always a chance of setting up bounce-back condition anywhere in the system

(ii) The solids formed will migrate to the bottom of the system

(iii) They might also get stick to the walls of the domain

Keeping in mind all these consequences, what is necessarily to be taken into care for such simulations?