open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I am trying to produce secondary currents in a 135° channel bend with Telemac3d.
The images show the horizontal (V), Vertical (W) and absolute (VW) velocity in a cross section of the channel bend.
(I am still struggling with the use of the Telemac extraction scripts. There is always a triangulation error so I cannot visualize the vectors (RuntimeError: In initialize: Triangulation is invalid) - a bit offtopic but maybe someone knows a solution for that issue too).

The vertical velocity shows somewhat what I'd expect:
Downward motion on the exterior bank (right bank) and updard motion on the interior bank (left bank).
However, the horizontal velocity is only facing from the interior to the exterior bank and therefore no circular motion is produced. Since the horizontal velocity is approximately 10 times bigger than the vertical velocity, the plot of the absolute velocity looks quite similar to the horziontal velocity.

Has someone ever tried modelling secondary currents in a channel bend in detail and has an idea which parameters I should adjust. I attached my .cas file if you want to take a look.

Best regards,
Jakob

Hello Jakobs,

To draw vectors in vertical slices, you can have a look at what has been written for Cooper, bottom_bc, bump e.g. in the vnv.py files.

Looking quickly at your steering file, I have a few remarks/questions:
- which release do you use?
- you can try other advection schemes to see if the results change a lot or not, e.g. LIPS (SCHEME FOR ADVECTION OF... = 5 + SCHEME OPTION FOR ADVECTION OF... = 4) or the method of charcteristics (SCHEME FOR ADVECTION OF... = 1) for every velocity/k-epsilon variables or a combo,
- has the keyword DYNAMIC PRESSURE IN WAVE EQUATION an influence in your computation?
- what about the solving of the different steps in the listing. For each of them, how many iterations do you need to solve every linear solver? Can you upload one listing here please?

Chi-Tuan

Hello Pham,

thank you for the answer!

Turned out, that my script for the visualisation of the cross-profiles had an error...
So the results look qualitativley quite well now. However, I used postel3d for the visualitation, since I could not find the vnv.py files in the examples. I just recently switched to V8P4, but I remember the vnv.py files in the older versions. Can I still find them somewhere?

Best regards,
Jakob

Hello Jakobs,

If you use the automatic installer for release 8.4, the directories were cleaned to keep the necessary files for examples in order to lighten downloads (e.g. geometry file, boundary conditions file, steering file... but no reference file or vnv.py scripts).
Anyway, if you clone the main or v8p4r1 e.g. (or better the latest official release v8p5r0), you will be able to get the vnv.py files.

If you only need one single, you can directly use the gitlab interface, see e.g.:
gitlab.pam-retd.fr/otm/telemac-mascaret/...er.py?ref_type=heads

Hope this helps,

Chi-Tuan