LatticePhase partial molar property calculations are incorrect

[speth]

Problem description

Although the calculation is often done separately for efficiency, the molar enthalpy, internal energy, entropy, Gibbs free energy, and volume should be equal to the mole fraction weighted sum of the corresponding partial molar properties. For LatticePhase, most of these do not agree. The partial_molar_enthalpies seem to be particularly bad.

Additionally, the implementations of getStandardChemPotentials and getGibbs_RT (which is supposed to calculated the nondimensional standard chemical potentials) are inconsistent.

Steps to reproduce

>>> p = ct.Solution('test/data/thermo-models.yaml', 'Li7Si3-interstitial')
>>> p.gibbs_mole, sum(p.X * p.chemical_potentials)
-349994.3375144891, -349042.0221149627)
>>> p.enthalpy_mole, sum(p.X * p.partial_molar_enthalpies)
(261425.4600475561, 174272.0)
>>> p.volume_mole, sum(p.X * p.partial_molar_volumes)
(0.09557086388415284, 0.09661515524531131)
>>> p.entropy_mole, sum(p.X * p.partial_molar_entropies)
(843.3376518097174, 843.3376518097147) # this one is OK
# partial_molar_int_energies is not implemented

System information

• Cantera version: 2.6.0 or main at 1ab81ac

Additional context

This was discovered as part of #1299, which implements Cantera/enhancements#114.