Hi everyone, I am new to code saturne. I would like to carry out a mesh sensitive analysis on my model to save on computational time. However, I am unsure on how to do this. I have created the model in SALOME, successfully exported and run a simulation in Code_Saturne and viewed the results in paravis. I don't now how to picks areas of interest that I can monitor to see the the influence of altering my mesh has on their values.
I am unsure of how to go about this, any suggestions?
Much appreciated, Meb15aa
Mesh Independent study
Forum rules
Please read the forum usage recommendations before posting.
Please read the forum usage recommendations before posting.
-
- Posts: 4085
- Joined: Mon Feb 20, 2012 3:25 pm
Re: Mesh Independent study
Hello,
This is a very broad question indeed.
If meshing with Salome, parametrizing your mesh refinement is useful for this. If you wrote a script for meshing, this should be trivial. Otherwise, if you build the mesh using the Salome graphical interface, modifying a Python dump or use of the notebook are adequate solutions.
Using Code_Saturne, the study manager can help you run the different refinements as a series, and possibly help organize plots, so it is a useful tool.
If the goal is to save time, a senitivity to other aspects, such as linear system precision (reasonably, do not reduce it too much), gradient computation options, and time-stepping options (especially if this is a steady flow) can also be useful.
But in any case, you need to know what areas and variables of interest to look for if for example you want to plot comparative results for the different refinements. And this heavily depends on your geometry and setup (boundary conditions, turbulence model, ...) The best practices described on this website and more general CFD best practices can help here, as well as basic physical understanding of the flow (for which you initial ParaVis visualization may provide some intuition).
Best regards,
Yvan
This is a very broad question indeed.
If meshing with Salome, parametrizing your mesh refinement is useful for this. If you wrote a script for meshing, this should be trivial. Otherwise, if you build the mesh using the Salome graphical interface, modifying a Python dump or use of the notebook are adequate solutions.
Using Code_Saturne, the study manager can help you run the different refinements as a series, and possibly help organize plots, so it is a useful tool.
If the goal is to save time, a senitivity to other aspects, such as linear system precision (reasonably, do not reduce it too much), gradient computation options, and time-stepping options (especially if this is a steady flow) can also be useful.
But in any case, you need to know what areas and variables of interest to look for if for example you want to plot comparative results for the different refinements. And this heavily depends on your geometry and setup (boundary conditions, turbulence model, ...) The best practices described on this website and more general CFD best practices can help here, as well as basic physical understanding of the flow (for which you initial ParaVis visualization may provide some intuition).
Best regards,
Yvan
Re: Mesh Independent study
Thank you Yvan, your post was very helpful