open TELEMAC-MASCARET The mathematically superior suite of solvers

I am running a 3D model that models a water intake structure. For the mesh geometry, the intake structure is within a larger model domain. The grid is 2x2 m for the intake structure area and gradually increases to 100 m towards US and DS boundaries. My vertical layers are user defined with 2 m equal spacing. The model is to be run in steady state (i.e. constant inflows and constant level).

The model runs to complete. But when i review the 3D velocity surface in BlueKenue, i see the coloring is changing even to the final time steps, which suggests the model is not stable.

I also observed slight difference between results from scalar model and parallel model. Is this normal or something is not correct?

I have attached my case file.

Any hints are very much appreciated!

Hailiang

Forget to mention i am using the latest version v7p2r1

Hello,

You may be in a case where a steady state is never reached. It is the case for example with bridge piers in a river, where von Karman eddies appear in the wake of the pier. If it is your case it will also explain differences between scalar and parallel. As a matter of fact the eddies are bifurcations that may be triggered by truncations errors, so the tiniest difference will make them appear. The eddies and vortex shedding found by Telemac are correct, e.g. they have the right frequency, at the condition that you do not use a turbulence model (in this case it gives you a result which is like an average in time and von Karman eddies may disappear).

With best regards,

Jean-Michel Hervouet