This utility converts a Chemkin-format reaction mechanism with reversible reactions to one with only irreversible reactions.
It does this by fitting reverse Arrhenius coefficients using a nonlinear least-squares minimization.
irrev_mech requires Python (2.x or 3.x) and the SciPy stack, but no additional libraries (i.e., Chemkin, Cantera). It has only been tested on Python 2.7, but probably works on earlier versions.
irrev_mech can be used locally via
$ python -m irrev_mech [options]
or installed as a package using pip install --user irrev_mech or via python setup.py install, and called using
$ irrev_mech [options]
Use the option -h or --help to see the full usage instructions. To generate an irreversible mechanism, from the command line, use python -m irrev_mech -c mech_name -t thermname where mech_name and therm_name are the names of the chemical kinetics reaction mechanism file and thermodynamic database, e.g.:
$ python irrev_mech.py -c mech.dat -t therm.dat
You can also run irrev_mech without a thermodynamic database if the information is held in the chemistry mechanism file (after the species are declared), e.g.:
$ python -m irrev_mech -c mech.dat
The new model file has the name mech_irrev.txt.
The default temperature range used for parameter fitting is 300 K to 5000 K. This can be changed by specifying the -r or --range command line option, e.g.:
$ python -m irrev_mech -c mech.dat -t therm.dat -r 1000 3000
irrev_mech is released under the MIT license, see LICENSE for details.
If you use this software as part of a scholarly publication, please cite the software directly using the DOI in the badge at the top of this file (and found here: https://zenodo.org/badge/latestdoi/5316746).
The three reverse Arrhenius coefficients are determined using a nonlinear least-squares minimization, using the SciPy function scipy.optimize.leastsq. The fit is performed using a large number of calculated reverse rate coefficients, densely sampled over the specified temperature range. The initial guess for this minimization comes from an analytical fit to three temperatures (the high and low values of the range, plus a midpoint). See the appendix of Niemeyer and Sung's paper for more details:
- KE Niemeyer and CJ Sung. "Accelerating moderately stiff chemical kinetics in reactive-flow simulations using GPUs." J. Comput. Phys., 256:854-871, 2014. doi:10.1016/j.jcp.2013.09.025
The use of the least-squares minimization to obtain better fits was suggested by Taylor et al.:
- BD Taylor, DA Schwer, and A Corrigan. "Implementation of Thermochemistry and Chemical Kinetics in a GPU-based CFD Code." 53rd AIAA Aerospace Sciences Meeting, Kissimmee, FL, January 2015. doi:10.2514/6.2015-0842
Created by Kyle Niemeyer. Email address: kyle.niemeyer@gmail.com