How can the binary parameters _epsilon_k_ab_ and _kappa_ab_ be explicitly submitted for a mixture?

[moritzhs6]

Dear all,

we are developing a Python program for calculating LLE using FeOs for PC-SAFT calculations of chemical potential. In our case: One component has both a negative and positive association site. The other component only has a negative association site.

Short Python example:
hmf = PureRecord(identifier=Identifier(name="HMF"), molarweight=126.11, model_record=PcSaftRecord(m=2.72266, sigma=3.83791, epsilon_k=433.125, kappa_ab=0.0052651, epsilon_k_ab=2524.61, na=1, nb=1))
mthf = PureRecord(identifier=Identifier(name="MTHF"), molarweight=86.132, model_record=PcSaftRecord(m=2.7191, sigma=3.6567, epsilon_k=265.09, kappa_ab=0, epsilon_k_ab=0, na=0, nb=1))
pures = [hmf, mthf]

In order to take the induced association in the mixture into account, we would like to explicitly specify epsilon_k_ab and kappa_ab as parameters for the mixture between the two components. However, when creating the equation of state we were not able to find a way for submitting epsilon_k_ab and kappa_ab to the PcSaftParameters object.

Short Python example (unfortunately, this formulation didnt work and only adopts k_ij and not the other two parameters):
binary = BinaryRecord(id1=Identifier(name="HMF"), id2=Identifier(name="MTHF"),model_record=PcSaftBinaryRecord(k_ij=0.015, kappa_ab=0.50, epsilon_k_ab=1800.0))
params = PcSaftParameters.from_records(pures, [binary])
eos = EquationOfState.pcsaft(params)

How can we access the modelRecord constructor to submit the binary parameters epsilon_k_ab and kappa_ab for a mixture to the PC-SAFT EoS?

Thanks for your help!

Best regards,
Moritz

[prehner]

Hello Moritz,

the parameter handling was revised a bit from going from feos 0.8 to 0.9. One aspect we added was to increase the flexibility of the association scheme so that any (well most I guess) kinds of hydrogen bonding arrangements can be modeled.

With the current feos version the code would look like this:

import feos

hmf = feos.PureRecord(identifier=feos.Identifier(name="HMF"), molarweight=126.11, m=2.72266, sigma=3.83791, epsilon_k=433.125, association_sites=[{"kappa_ab": 0.0052651, "epsilon_k_ab": 2524.61, "na": 1, "nb": 1}])
mthf = feos.PureRecord(identifier=feos.Identifier(name="MTHF"), molarweight=86.132, m=2.7191, sigma=3.6567, epsilon_k=265.09, association_sites=[{"nb": 1}])
pures = [hmf, mthf]

binary = feos.BinaryRecord(id1=feos.Identifier(name="HMF"),	id2=feos.Identifier(name="MTHF"), k_ij=0.015, association_sites=[{"kappa_ab": 0.50, "epsilon_k_ab": 1800.0}])
params = feos.Parameters.from_records(pures, [binary])
eos = feos.EquationOfState.pcsaft(params)
display(params)
eos.parameters

eos.parameters contains the fully assembled matrix of association parameters, so that you can check whether everything was read in correctly. I move this to a discussion, because it could be interesting for other users as well.