RKSymFoam

This code contains slightly adjusted versions of the solvers used in the paper "A symmetry-preserving second-order time-accurate PISO-based method." by E.M.J. Komen, J.A. Hopman, E.M.A. Frederix, F.X. Trias and R.W.C.P. Verstappen. One notable adjustment is made in the pressure gradient interpolation to follow the paper "An energy-preserving unconditionally stable fractional step method on collocated grids." by D. Santos Serrano, F.X. Trias Miquel, G. Colomer Rey and C.D. Pérez Segarra. The pseudo-symplectic Runge-Kutta integrators included in this work are extracted from the paper "Explicit Runge-Kutta schemes for incompressible flow with improved energy-conservation properties." by F. Capuano, G. Coppola, L. Rández, and L. de Luca. For a description of the method, please refer to these papers.

Authors

The main structure of the solver, including the Runge-Kutta schemes was developped by Edo Frederix, of the Nuclear Research and Consultancy Group (NRG), Westerduinweg 3, 1755 LE Petten, The Netherlands. The symmetry-preserving method was applied to this structure by Jannes Hopman, of the Heat and Mass Transfer Technological Center, Technical University of Catalonia, C/Colom 11, 08222 Terrassa, Spain. The pseudo-symplectic schemes were added by Josep Plana-Riu, of the Heat and Mass Transfer Technological Center, Technical University of Catalonia.

License

RKSymFoam is published under the GNU GPL Version 3 license, which can be found in the LICENSE file.

Prerequisites

OpenFOAM v2012. While it may compile against other versions, this is not tested and currently not supported.
Python with numpy and matplotlib

Usage

Make sure that OpenFOAM v2012 is loaded into your environment
Compile all libraries and apps with

./Allwmake

Test cases

All test cases can be found in the "cases" directory, including a Taylor-Green Vortex and a channel flow.
Both test cases contain a "templateCase" directory, which should be copied before it run from inside the new directory with:

./run.sh <solver> <simulation type> <Runge-Kutta scheme>

Solvers permitted by this script: "icoFoam", "pimpleFoam" and "RKSymFoam"
Simulation types permitted by this script: "LES", "laminar" (laminar should be selected to run DNS)
Runge-Kutta schemes permitted by this script: "BackwardEuler", "Kutta" (classical Runge-Kutta 3 scheme)
icoFoam does not read the <simulation type> and <Runge-Kutta scheme> arguments, so they can be omitted
pimpleFoam does not read the <Runge-Kutta scheme> argument, so it can be omitted
The user is encouraged to experiment with different settings after getting familiar with the structure of the code, to do so change the <VAR*> variables inside "system/controldict.m4", "constant/turbulenceProperties.m4" and "system/fvSolution.m4" and rename the files to omit the ".m4" extension. The case can now be run without the "run.sh" script, but simply as any other OpenFoam case.
The available Runge-Kutta schemes can be found in "libraries/RungeKuttaSchemes", the Butcher Tableaus are given in the "<*.C>" file and a reference is given in the "<*.H>" file

Taylor-Green Vortex

Demonstrating the loss of kinetic energy due to numerical dissipation over time
Validation cases: icoFoam and RKSymFoam using Backward Euler scheme

Channel flow

Demonstrating accuracy of the solver to simulate turbulence
Demonstrating the ability to include LES models
Validation cases:
DNS cases: icoFoam (Backward Euler) and RKSymFoam (Backward Euler and Runge-Kutta 3)
LES cases: pimpleFoam (Backward Euler) and RKSymFoam (Backward Euler and Runge-Kutta 3)
Run cases on 8 processors with the same command by first adjusting the following lines in "run.sh" to:

#- Run serial
#runApplication $(getApplication)

#- Run parallel 
runApplication decomposePar
runParallel $(getApplication)
runApplication reconstructPar

Post-processing

To post-process the cases, run "plot.py" from the "cases/<case>/postProcess" directory using

python plot.py

For example: To postprocess "<run_directory>/<case_1_name>" and "<run_directory>/<case_2_name>", edit the <runDir> and <solvers> variables in "plot.py" to:

runDir = '<run_directory>'

solvers = ['<case_1_name>', '<case_2_name>']

Resulting plots will be found in the "postProcess/results" directory

Validation

A set of predetermined cases was run and post-processed by the authors.
To run the same cases and reproduce the results, navigate to the "cases/validation" directory and run all cases with:

./launchCases.sh

After completing the cases, run "plot.py" inside "cases/<case>/postProcessing", the version of "plot.py" in this repository is set up to post-process the validation cases
The resulting plots can be compared with the results readily available in "cases/<case>/postProcessing/validationResults"

Using RKSymFoam in your own OpenFOAM cases

The entries in "system/fvSchemes" are not read by RKSymFoam, except potentially for the turbulence model, all other schemes can be set to:

    default         none;

In "system/fvSolution", the subdictionaries for "p" and "pFinal" are named "pCorr" and "pCorrFinal" respectively.
A subdictionary named "RungeKutta" has to be added to "system/fvSolution", for example:

RungeKutta
{
scheme          BackwardEuler;
nOuter          1;
nInner          2;
pnPredCoef      1;
pRefCell        0;
pRefValue       0;
}

All available schemes are based on the Butcher Tableau and can be found in the "libraries/RungeKuttaSchemes" directory
Cases are run exactly the same way as by any other OpenFOAM solver
A transport model has to be chosen in "constant/tansportProperties", similar to the usage of pimpleFoam.
A turbulence model has to be chosen in "constant/turbulenceProperties" file, similar to the usage of pimpleFoam.
If you want to run a DNS, set the transport model to Newtonian (1) and the simulation type to laminar (2), as demonstrated below.
1. In "constant/transportProperties" add the line:

transportModel  Newtonian;

2. Create the file "constant/turbulenceProperties":

FoamFile
{
    version     2.0;
    format      ascii;
    class       dictionary;
    location    "constant";
    object      turbulenceProperties;
}

simulationType laminar;

Contact & support

For bug reports or support, feel free to contact Jannes Hopman at jannes.hopman@upc.edu. Please note that this code is not maintained nor regularly updated, and is only tested with OpenFOAM v2012. Questions related to other versions will thus not be answered.

Disclaimer

RKSymFoam is provided by the copyright holders and contributors "as-is" and any express or implied warranties, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose are disclaimed. In no event shall the copyright owner or contributors be liable for any direct, indirect, incidental, special, exemplary, or consequential damages (including, but not limited to, procurement of substitute goods or services; loss of use, data, or profits; or business interruption) however caused and on any theory of liability, whether in contract, strict liability, or tort (including negligence or otherwise) arising in any way out of the use of this software, even if advised of the possibility of such damage.

janneshopman/RKSymFoam