MembraneEOS.jl

Wrapper for various equation of state (EOS) implementations to be used in the membrane package ecosystem.

This repository will hopefully soon wrap Clapeyron.jl for EOS calculations, while providing convenience functions and data for membrane / polymer focused tasks.

Sections

The Gist

To get the properties of a mixture of chemicals in a given composition and conditions (state), we use an equation of state (EOS). This works by...

  1. Constructing the parameters you will be needing for the EOS. (see EOS Parameters)
  2. (optional) Constructing a KIJ matrix to account for deviations from the EOS predictions. (see KIJ Matrices)
  3. Construction of the EOS model itself. (see Models)
  4. Querying properties from the model. (see Exposed Functionality)
  • the two ways to get parameters
    • by name
    • by values
  • Constructing mixed models
    • KIJ initialization and setting
  • unit limitations

Quick Start

Parameter generation

To construct a model, you first construct the parameters that go into the model. using Peng Robinson as an example, which uses "cubic parameters"...

co2 = CubicParameters("CO2")

They can also be constructed via values directly (see CubicParameters)

ch4 = CubicParameters(190.8, 45.79, 0.012, 16.04) # Tc, Pc, ω, MW

Model construction

These parameters can then be combined to create an EOS and extract a property from it.

components = [co2, ch4]
pr = PR(components)
pressure(pr, 22.414, 273.15, [0.5, 0.5]) # returns in mpa
0.1008297268871727

Unideal interactions

KIJ matrices, which encode deviations from the default EOS interactions, follow the same order that your parameters do in the input to the EOS model constructor. We can use the helper function MembraneEOS.initmatrix to create an ideal KIJ matrix for some parameters. 

kij = MembraneEOS.initmatrix(components)
kij[1, 2] = 0.09
kij[2, 1] = 0.09
pr = PR(components, kij)
pressure(pr, 22.414, 273.15, [0.5, 0.5])
0.10085658760678402

Fast lookups for common components in gas separations.

Some default data is available for automatically looking up and using EOS parameters and kij values.

See "TestChemicalLookup.jl" for a bit more on how custom databases are used.

If looking up from a data base, you can skip the parameters and simply construct using strings (this will also try to initialize a KIJ matrix)

Note

If you mix directly created parameters with tabulated ones, the KIJ search will not match any directly specified parameters. Eventually mixing these two methods of EOS model generation will throw an error to avoid ambiguity and unexpected, hard to debug behavior.

pr = PR(["CO2", "CH4"])
pressure(pr, 22.414, 273.15, [0.5, 0.5])
0.10085658760678402