Dear Giada

I do not know the ThermoCalc idiosyncracies  but Gibbs free Energy  always consists

of an enthalpic term (average energy of interactions) an  entropic term (e.g mixing entropy

term expressible  in terms of site occupation fractions in your case) and a pressure times volume term alternatively if you know either empirically or via  theory the chemical potentials and the fractions it is the sum of all products of chempotentials times numnerical fractions present.Very probably either ThermoCalc or you miss some of the  cterms.necessary for  a consistent value.I suspect the enthalpic term is making trouble

which  depends more on lattice symmetry (i.e BCC vs FCC)

 Dear Giada,

In your database, the sublattice model should be defined by compound energy formalism and the excess parameters by the Redlish-Kister model.

Nevertheless, many problems can also come from your macro files (TCM) even if your parameters are properly defined in your database.

Please check :

Have you defined a reference state may different to SER defined by default, is it the good one ?

For example, the command in Fe-C should be :

set-reference Fe BCC * * (1st star is for T, 2nd for pressure)

set-reference C GRAPHITE * *

In BCC, the site fractions are represented by y(BCC_A2,*)

The energy functions, for example HM or HMR, depends on the way you define the condition of your calculation. One good way to avoid errors is to fix n=1 as a condition, your results are then referenced to 1 mol in J/mol/atoms. Then, HM and HMR are equivalent.

I hope that this advices could be useful.

Regards

Dear Dikeos Mario and Gossè Stèphane

First of all thanks for the help.

I have found the mistake, in case of interstitial elements it is necessary to multiply by the molar fraction of the substitutional otherwise the Gibbs energy is normalized. For this reason the 2 results were different.