open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello!

I am modelling an archipelago system with wind driven currents with k-epsilon.
I made a couple of tests that showed that surface velocities near shorelines are pretty much impacted on the type of vertical discretization.

All my sounds end up to a minimal depth of 0,5 m (no tidal flats), which means that the vertical discretization is very thin compared to the channel if using a classical sigma transform.

A test performed with 2 planes at fixed elevations near the surface (-0,25 and -0,5 m, water surface at 0) and sigma underneath showed lower velocities near shores, but the higher velocities are still observed at these locations.

I don't know in detail the physics of wave driven currents, so I am wondering if this result is physical or not (do we expect higher velocities in shallow areas?). I use THRESHOLD DEPTH FOR WIND = 2.0, but it is not reasonable to go much deeper since my area of interest has depths up to 6 m.

I need to have a rather good discretization near the water surface in order to model floating piers that are ca 0,2 m deep. So I can't skip to have several planes in the first meter under the surface...

I found some intresting info in THIS POST about the Berre lake model which also used k-epsilon.
How was the vertical mesh discretization defined near the surface in this model?

Thank you very much in advance for your help!

Regards,
PL

Hello,

I think that what you find is correct, there is more influence of the wind on shallow waters (this is why we introduced a threshold, without it and near tidal flats, thin layers of water are very influenced).
The Berre lake mesh is done with planes with fixed elevation (which gives a number of crushed elements) to follow stratifications. The refinement at the free surface is not so important, because the stress due to wind is introduced directly in the k-epsilon equation (so we do not need a refinement that would give accurate velocity gradients on the vertical).

regards,

JMH