• Before the calculation initial and boundary conditions need to be copied from 0.org to 0 directory.

  # cp -r 0.org 0

   

• Simulation will be running in parallel thus the parallel decomposition has to be done. The method of decomposition and the number of employed processors/cores is in system/decomposeDict.

  # decomposePar > log.simulation-decomposePar 2>&1

   

• It is useful to run renumberMesh utility before the computation. It provides better numerical properties of matrices resulting from the finite volume method discretization of Navier-Stokes during the computation.

  # mpiexec -np 6 renumberMesh -overwrite -parallel > \
  log.simulation-renumberMesh 2>&1

   

• The number of used processors/cores behind the switch -np must correspond with the number of processors/cores in the system/decomposeDict.

   

• The chosen solver for this tutorial case is pimpleFoam. To launch it in parallel use the following command:

  # mpiexec -np 6 pimpleFoam -parallel > log.simulation-pimpleFoam 2>&1

   

• See Allrun-rotating.sh script which contains the same course of steps that is described above.

   

echo "Running Simulation..."
   cp -r 0.org 0
   decomposePar > log.simulation-decomposePar  2>&1 
   mpiexec -np $numProcs renumberMesh -overwrite -parallel > log.simulation-renumberMesh 2>&1
   mpiexec -np $numProcs pimpleFoam -parallel > log.simulation-pimpleFoam 2>&1

Subsections

• Deleting of unnecessary fields during the simulation
• Plotting the torque