WIP: Add a mass dependent maximal smoothing length

The used density seems to correspond to the input density I put in the parameter file, the only thing is that I multiply 2 large numbers so I need to place them in a double before I put the final minimum density in a float. I tested some isolated disk galaxies for a couple of 100 of years and that seems to run, let me know if you have more suggestions to test this piece of code/ suggestions for the code.