Hot to use maps/phb with two different crystal structures

Alloy Theoretic Automated Toolkit (ATAT)
1

Dear all,
according to the manual, I can use phb to calculate the phase boundary between two phases with different crystal structures. However, I am not sure how to do this correctly.

Here is my (slightly simplified) situation:
I am interested in the phase boundary between fcc Ni and the hcp Ni3Nb phase (delta phase).
I have done two maps calculations one with an fcc lat.in, one with hcp lat.in, both restricted with
maps -c0=0 -c1=0.25

In the fcc calculation, the energy of pure Ni is then lower than in the hcp calculation (because pure Ni is fcc). However, maps normalizes both calculations so that the pure Ni gs has energy zero on both cases.
Do I have to use the optional ref_energy.in to account for this? If so, how?
Or will phb in some way do this automatically (i don’t see how, but ATAT sometimes does clever things…)?
Any help is appreciated,
Martin.

2

You should

  1. run maps on the fcc lattice (say)
  2. copy the ref_energy.out from fcc to a ref_energy.in file in the hcp directory.
  3. run maps on the hcp lattice
    When I say "run maps" I don’t mean you have to restart everything - just run it with the existing structural energies, so the eci.out is updated to account for the change in reference.
    (Of course, you could set the ref_energy.in to any reference you want, as long as you pick the same reference for both fcc and hcp.)
    Then run phb. You will need to specify the directory where each cluster expansion reside - see phb -h for how to do that.