 |
Langley Research CenterTurbulence Modeling Resource |
Notes on How to Use this Website
If your purpose is : Finding a Turbulence Model Description
1. From the
home page near the top, select a turbulence model from the section titled "Turbulence Models."
2. The model and
published variants (if any) are described on the selected page.
If you do not see a particular turbulence model listed that you are interested in,
please consider volunteering to help us add a description of the model to the website (see
Guidelines for submitting a new turbulence model description).
It should be published (in an accessible reference).
If your purpose is : Verification 1. From the
home page, look at the section titled "Turbulence Model Verification Cases and Grids."
The links are listed several different ways. There is a descriptive title (e.g., "2D Coflowing Jet") as well as a
short-hand notation (e.g., "VERIF:2DCJ"). There is also a table
titled "Models Currently Applied to Verification Cases".
If you are trying to verify your implementation of one of the listed models (left column), then continue on to step 2.
If the model you are interested in is not listed in this table, then you cannot perform verification using this site
at this time;
please consider volunteering to help us to add a description of your model to the website if it is not already in place (see
Guidelines for submitting a new turbulence model description),
and to get verification cases in place.
2. Select one of the
verification cases and go to that page.
3. Find the link for
obtaining the case's sequence of grids (to be used for the grid convergence study) on the Intro Page for the case
of interest.
4. Download the grids,
and run your code on the full sequence of grids (in other words, if there are 5 grids, then run your code 5 times, once on each
grid). Be sure to use boundary conditions that are consistent
with the ones used herein. Be sure to fully converge each of your verification solutions iteratively.
5. Compare your results
with those posted on the verification page. For grid convergence (results as a function of h), your solution should approach the
same answer as h approaches zero. Also, your results on the finest grids should very closely match the posted
results. Note that if you are running an incompressible code, your results may not match, as almost all results on this
site are compressible code results.
6. If your results do not match the website's,
then you may have an implementation difference (check the turbulence model equations), a bug, or you may not have
sufficiently converged your solutions iteratively. Read the entire Verification page(s) for your case thoroughly,
to make sure that your boundary conditions are appropriate, and that you have handled other important issues appropriately.
Note also that your code's mean-flow equations should probably be nominally
second-order accurate (or higher), as first-order mean-flow terms are typically inadequate.
7. Note the Model Readiness Rating
(MRR) of the turbulence model you are interested in. The higher the rating, the more
confidence there is in the posted results. For example, MRR Level 3 or higher indicates very high confidence, because multiple
CFD codes from different organizations (coded independently) have been demonstrated to produce consistent results for
Verification cases.
8. For verification of the SA model (and in one case
SA-QCR2000) there is another section of verification cases that have been developed for use in AIAA workshops. See the section on
the main page titled: "Additional Turbulence Model Verification Cases and Grids."
If your purpose is : Validation 1. From the
home page, look at the section titled "Turbulence Model Validation Cases and Grids."
The links are listed several different ways. There is a descriptive title (e.g., "2D Mixing Layer") as well as a
short-hand notation (e.g., "2DML"). There are also two tables, one listing the validation cases along with the models
currently tested for them, and another listing the type of flow physics covered by each validation case.
Choose the case you are interested in.
2. If you want to run the case
yourself, then find the link for
obtaining the case's sequence of grids on the Intro Page for the case
of interest and continue on to step 3. If you have not yet verified your implementation of the model,
then you should go through that step first (see "If your purpose is : Verification" above). If you are only interested in seeing how different models behave
for the case of interest, then simply view the plots on the validation pages.
3. Download the grids,
and run your code on the appropriate grid level (typically the finest or second-finest grid level).
You can run on all grid levels to check the grid convergence if desired.
Be sure to use boundary conditions that are consistent
with the ones used herein. Be sure that each of your solutions is sufficiently converged iteratively.
Corresponding data for each validation case is
provided on the Validation Intro Page.
4. Compare your results
with those posted on the validation page. Your solution should closely match the posted results for the model.
If it does not, then be sure that you have verified the implementation of your model (see "If your purpose is : Verification" above).
Note that if you are running an incompressible code, your results may be different, as almost all results on this
site are compressible code results.
5. Note that for the Validation results
posted on this site, some effort was made to ensure reasonable grid resolutions
for the finest grid levels provided, but thorough grid convergence studies were not completed for these cases,
and in some cases (specifically noted) iterative convergence was problematic.
Therefore, posted Validation results should be considered representative, but not "truth." In other words, the intent
of the Validation cases is to indicate the general behavior of different turbulence models for a wide variety
of problems of interest. It is expected that differences in results between
the models will typically be larger than differences caused by insufficient grid resolution.
If your purpose is : Finding Data 1. From the
home page, go to one of the sections under "Turbulent Flow Validation Databases."
There are many experimental, DNS, and LES databases contained there.
2. Note that data is also
provided on the Intro Pages for each Validation case.
If your purpose is : Other 1. Search around the site!
There is a section on the method of manufactured solutions, and several sections on Numerical Analysis,
which focus on numerical details for specific problems of interest.
Return to: Turbulence Modeling Resource Home Page
Recent Significant Site Updates Responsible NASA Official:
Ethan Vogel
02/15/2018 - Added further clarification to Note 5 under Validation
02/27/2017 - Added Note 5 under Validation
07/08/2016 - Added Note 7 under Verification
Page Curator:
Clark Pederson
Last Updated: 06/15/2024