Thank you. Got it. But, I still can`t understand phase missing in 500K while no phase missing in 450K and 550K. In fact, I found phase suddenly miss sometimes in Piecewise temperature limits. do you have some suggestions?

Can you share the database file? I can try to reproduce the issue and find a solution.

al-cu.txt

A small update here. I am able to reproduce the issue. At this composition, ALCU_ETA becomes the only stable phase and there's a discontinuity in the chemical potentials at this point, which causes pycalphad to get stuck in a loop of trying to add ALCU_THETA, ALCU_ZETA, ALCU_DELTA phase, depending on the chemical potentials/driving forces.

This is a bug in the minimizer that may have been introduced when we changed some solver internals. Depending on your application, a possible workaround is to nudge the composition to be slightly Cu-rich or Cu-poor, e.g.

eq = equilibrium(dbf, comps, phases, {v.P: 101325, v.T: 450, v.X("CU"): 0.50 + 1e-6, v.N: 1}, output='HM', verbose=False)

This will give you two phases, but the phase fraction of the second phase will be small.

Hello, recently, I have been performing equilibrium calculations for aluminum alloys with multiple components. I have encountered the same issue of missing phases that we discussed before. However, this time, I have more elements, and the values are constrained to 5-6 decimal places, yet the problem persists. Additionally, since the TDB file for this calculation belongs to our research group, I cannot share it. Is there a solution to this issue?