DWPM-Mixture-Model

Phase seperation calculation using the DWPM mixture rule.

Main function call examples.

The functions can be used by calling the main.py script with optionals passed from the command line, for example initializing the state and parameter classes s p for acetone water phenol with two valid root phases and then enter the Python shell:

$ python -i main.py -c acetone benzene water -p x y

>>> s
<nComp.state instance at 0x7f28190cf368>
>>> p
<data_handling.MixParameters instance at 0x7f28190d01b8>

[To be implemented soon:] To optmise new paramters and save the results add the -optim -save options

$ python main.py -c acetone benzene water -p x y -optim -save

To calculate all phase seperations and equilibrium points specifify the pressure and temperature only

$ python main.py -c acetone water -p x y -P 101e3 -T 298.15

To calculate equilibrium at a specific feed point specifify the pressure, temperature and a feed composition

$ python main.py -c acetone water -p x y -P 101e3 -T 298.15 -z 0.5 0.4

Several plots for binary and ternary systems are available. The plot optional -pltg can be added to plot the Gibbs surface and any equilibrium tangent planes at a P, T point.

$ python main.py -c acetone water -p x y -P 101e3 -T 298.15 -pltg

The -pltiso will plot either isotherms or isobars with the current optimised parameters at the data $P$, $T$, $\bar{x}$ data points to compare the fit

$ python main.py -c acetone water -p x y -pltiso

Specifying -pltiso with a temperature or pressure point will plot an isotherm or isotherm at that point

$ python main.py -c acetone water -p x y -T 298.15 -pltiso

Basic description of files.

1. main.py main script to call, the specifications in config.cfg is used to determine the scripts to be run. See Config files for detail.
2. nComp.py contains the core functions used to simulate multicomponent equilibrium problems and optimise parameter models.
3. pure.py is used to fit single component Van der Waals EOS temperature dependent activity coefficient parameters to the Adachi-Lu and Soave models. These parameters are used in the multicomponent functions. The single component data is contained in \Data\pure and every component simulated in multicomponent system requires vapour pressure data or parameters stores in the .csv files in \Data\pure.
4. Van_der_Waals.py contains the functions the volume root, saturation pressure and state dependent functions which are called in both pure.py and nComp.py (note that for example, the function a_T(s, p) is used in calculating the temperature dependent activity coefficient parameters for both the pure and the mixed phases states (for example to calculated for component 1: s.c[1] = a_T(s.c[1], p.c[1]) while for the mixed parameter phase: s.m = a_T(s.m, p.m)).
5. DWPM Hessian.ipynb [temporary file] contains some of the analytical derivatives which will be written into nComp.py functions used to calculate the Hessian soon.
6. csvDict.py contains some functions used in the current data handling scheme using the .csv files stored in \Data. To be replaced soon.
7. tgo.py contains the topographical global optimisation function used to find the global minima of many sub-problems in nComp.py.

Config files

The simulation inputs are read from a file called config.cfg in the same format as config example.cfg and needs to placed in the directory before running the main.py file. The [paths] section needs to contain a directory to a Data folder containing the .csv files used in the project. Any .csv file with the required headings as read in data_handling.py can be read.

Running unittests

The *_tests.py files contain unittests. To run unittests covering all the code run test_all.py.

Specific test suites in each file can be easily run from the command line as example:

$ python2 -m unittest -v tgo_tests.TestTgoFuncs

Specific function tests inside a unittest can be run for example:

$ python2 -m unittest -v tgo_tests.TestTgoSubFuncs.test_t1