open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello, everyone!
I am working in a research for my master's degree that includes the evaluation of shear stress in a channel (half a circle cross section). First of all, I found it very difficult to develop an uniform flow in the channel.

I tried the velocity profile =3 and had this:


and velocity profile =4and had this:


I also tried the velocity profile=1 has similar results.

The solution I found for this issue was modeling an approach channel upstream of the originial channel, in order to "organize" the velocity vectors:





At this point, I observed 2 issues:
1) My model is not reproducing the logarithmic velocity profile, despite I had set the VELOCITY VERTICAL PROFILES = 2;2. Velocities are numericaly different for each layer, but almost the same.

2) The more I increase the number of layers of my 3D model, the more my total shear stress (calculated from v* provided by Telemac, integrating in a cross section along the wet perimeter) increases. I couldn't simulate more than 13 layers, because my computer had no memory enough, so I couldn't have a conclusion about the most realistic shear stress. I am using Nikuradse roughness.




I am wondering if someone here could help me in my research in some of these points I presented here. Anything can be helpful.
Thank you so much.

The FRICTION COEFFICIENT FOR THE BOTTOM of the cas file I sent is not the coefficient I have been using in my simulations.
Acctualy, ks=0.001 (1mm) for my case.