simulation with large number of wall-bounded regions?

Right, I mean still keep intramolecular interactions but discard
intermolecular interactions. This way you could still put multiple
molecules in the same box and then slowly change the dimensions of the
box to pass through your solution composition range.

Stan