binary mixture

[achremos]

Bubble + dew point routines use the composition as input and estimate the pressure. But this needs to be expanded.

What is missing (I think) is a routine that uses the xi or yi as initial guesses and provides resulting values of the xi and yi at specified conditions of T and P.

[jaclark5]

You'll find that there is already a flash calculation type, although as a warning this type of thermodynamic calculation is notoriously unstable to converge.

[achremos]

Will initial guesses for compositions improve the stability & convergence of the calculation by searching at regions of interest?

In SW bubble point requires an initial guess.

[jaclark5]

Yes it will! Make sure you're using a gradient decent optimization method in that case so that the initial guess is used. You can also apply a tighter pressure range to search.