equilibrium: Detecting single-phase regions

[broshe]

equilibrium: Detecting single-phase regions

[richardotis]

Right now pycalphad can't tell the difference between a single-phase region and a really small miscibility gap. Notice that the two compositions on the tie line are really close together, near the numerical limit of the solver. In this case the phase fractions returned don't mean anything.

For now, pycalphad will always return a tie region with the maximum number of phases allowed by the Gibbs phase rule, and you tell the single phase regions apart by checking how far apart the compositions / site fractions are.

Eventually pycalphad will automatically convert really small tie lines of the same phase into single phase regions automatically.

[richardotis]

Fixed in 31022d3