Incorrect mass in insert/pack using mass_in_region or volumefraction_region

Hi,

I find that the actual mass inserted is about three times greater than the requested mass when using insert/pack with a lognormal distribution. It is about 10% greater than the requested mass when using a gaussian distribution and slightly greater when using a uniform random distribution. Using volumefraction_region instead of mass_in_region gives about the same results. I can live with a few percent error, but it looks like something is very wrong with the mass calculation for the lognormal distribution. Here is a short script that demonstrates each case.

David