Dixon isobar issues and Psat varies with H2O

[PennyWieser]

Just as I thought we had got there with Dixon, there is an offset between isobars calculated in VESIcal and VolatileCalc. I think this is partially because the isobar code depends whether the "bulk sample" has H2O and CO2, which makes me think this related to a normalization problem. The 1st graph is for "bulk sample" with H2O=2.24 wt% H2O, the second (which is a better match is for the same sample with 0 wt% H2O)

.

I think H2O and CO2 should be ignored for isobars as the user will probably just pick a composition which has variable volatiles from their file, and then obviously the isobars is calculating volatiles, so the H2O and CO2 entered are arbitrary.

I've done some further testing to explain the discrepencies at both ends of the isobars - although overal Psats are a good match between volatileCalc and VESIcal, they are much worse at H2O=0 or CO2=0 than at slightly higher ones. Any idea why?

[simonwmatthews]

Part of the problem was the molecular weight including volatiles, which I've now changed. However, there's still a discrepancy, and I'm at a total loss for where its coming from. I'll have a fresh look tomorrow, or later this week, but I'm not optimistic.