open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi,

I'm setting up a simple rectangular channel with flow direction from left to right. The turbulence model that I would like to use is k-epsilon. And I've prescribed the K and E profiles at the inlet as well as for the initial conditions. I've also prescribed flow rate at the inlet and constant depth at the outlet, such that U is about 2 m/s.

The simulation did not crash with the settings in the steering file (see attached channel.cas). However, I noticed that there are instabilities at the outlet corners of the channel (see screenshots of U and V). I've tried changing several settings in the steering file but the simulations either crashed or the issue was still there.

Can you please kindly suggest the changes that I need to implement to prevent this happening? Thank you!

Hello,

Are you sure you have correctly defined your liquid boundary in the corners? The convention is "A liquid segment is one between two points of liquid type"

Chi-Tuan

Hi Chi-Tuan, yes I think I have correctly defined the boundary in the corners, which looks like this (see screenshot and .cli file).

Thanks for the picture.

There was another question I would have asked but I have the answer with this picture. You have 2 overconstrained triangles at the 2 corners which is really bad (the 3 nodes belong to the boundaries). Try to proceed differently for your mesh (e.g. cutting them into 2) so that at least there is a node inside the domain.

Chi-Tuan

Hi Chi-Tuan. After adding 2 hardpoints near each corner to split the overconstrained triangles, the problem is now solved. Thanks again!

Good news.

Next time, when you first run a computation with a new geometry file, you should add the option CHECKING THE MESH = YES for 1 time step. It will write some lines and comments about your mesh.

Hope this helps,

Chi-Tuan