Translating Square ALE (2D)

Translating Square ALE (2D)#

Description of the setup#

Figure 12 shows the geometry of the test case you will run in this tutorial.

This test case aims to validate the ALE (Arbitrary Lagrangian–Eulerian) implementation in TrioCFD against the SPHERIC benchmark Test 6, which represents a 2D incompressible flow around a translating square cylinder in a rectangular tank.

Geometry of the translating square setup

Figure 12 Geometry of the test case#

Tutorial setup#

This tutorial will use a TrioCFD validation form as its base. It is a jupyter Notebook that is used to start a list of TrioCFD simulations and displays various postprocessed results.

First, go to an empty directory and copy the validation form from which we will start: TranslatingSquare_ALE_CoarseMesh. Note that you must have sourced the TrioCFD environment before this step.

cp -r $(find $project_directory -name TranslatingSquare_ALE_CoarseMesh) . && cd TranslatingSquare_ALE_CoarseMesh

Now, use the following commands:

  • Run_fiche, which will start jupyter and open the notebook in the web browser.

    • Do that and execute all cells

    • Note: the run takes around 5 minutes. You can directly jump to the next step to save time.

  • Run_fiche -export_pdf, which will run all the content of the notebook and generate a pdf report from the output of the cells.

    • Resulting pdf is located in ./build/rapport.pdf

  • Now, copy the file ./build/rapport.pdf and the build directory:

    • cp build/rapport.pdf initial.pdf

    • mv build initial_build

Modifying the test cases in the validation form#

Now, edit the file src/TranslatingSquare_ALE_CoarseMesh.data:

  • Replace ALE { amont } convection scheme by ALE { muscl }

  • Change the Solver_moving_mesh_ALE:

    • replace PETSC GCP { precond ssor { omega 1.5 } seuil 1e-7 impr } into petsc cholesky { }

Now, use the command:

Run_fiche -export_pdf

This will rerun the simulations, postprocessings and generate the pdf report.

Now, open both the new report and the initial one:

cp build/rapport.pdf modified.pdf
evince initial.pdf modified.pdf &

What do you think of the results ? Are they better/worse/unchanged ?

Compare the *.TU files between the initial_build and build directories.

What can you conclude ?

Contents