open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,
maybe my question is quite stupid.
I'm implementing a 3D model nested to a sort of 3D regional model.
I need to impose (after interpolation) 2d (u,v) velocities on each of the vertical layers.
I implemented mesh transformation = 3 with 17 horizontal planes and set the ZPLANE (i.e. from -2 m to -50 m ) in accordance to the regional model.
How is the 3D mesh developed when the bottom is lower than -50 m? I mean the number of planes is always 17?? In this case, how does the code manage the velocity profiles I gave it?
for instance, for u- component and depth of the node equal to -8 m (i.e numb of layers = 4), I gave the vertical profile = 0.1,0.0.2,0.01,0.01, 0, 0, 0, 0,0, 0... (etc up to 17th layer).
The result is a vertical profile 0.1,0.0.2,0.02 ,0.02,0.02,.... 0.01,0.01,0.01.
Maybe I wasn't clear... Need a feedback, thanks.
Gabriella

Hello,

I am not sure to understand all the question but here are some answers:

Even when you prescribe the elevations of intermediate planes, the first one will be the bottom and the last one the free surface. For example if you prescribe -50 and the bottom is at -49, the first plane will be at -49 (this may even cause void elements, which are dealt with).

For setting a profile, you can look in bord3d.f the use of the function vel_prof_z.f: if you are setting a user profile you must choose VELOCITY VERTICAL PROFILES : ...;0;... (0 is for the liquid boundary with a profile, others will have a 1), then you implement vel_prof_z for the case 0. Then the velocities UBORL and VBORL (prescribed values of lateral velocities, initialised as normal to the boundary and with a norm equal to 1) will be multiplied by this profile, and if you have a prescribed discharge, it will be multiplied again by a coefficient. So you do not find in the results the profile you gave because it has been multiplied by a constant, this is the explanation of your problem if I understand it correctly, but I am not sure...

With best regards,

Jean-Michel Hervouet

Thanks Jean-Michel.
I'm actually modifying the subroutine BOR3D directly and not imposing a vertical profile in the subroutine you mentioned, since the model is quite complex and I got the velocities from the regional model.
I mean BORD3d read a file with u-v components, varying in time and 3D space.
Since the mother model has fixed z-levels, I imposed mesh transformation =3 and set layers levels as in the mother model.
But I dont have clear how TEL3d treats the fixed horizontal layers when the bottom depth is lower than the layer level (lower than N).
G

Hello,

Basically the prescribed elevations are obeyed, except for the bottom and free surface, and except when a plane would go under the bottom or higher than the free surface, and there is for this a hardcoded tolerance (an internal plane will not approach the bottom or the free surface less than a given distance which is hardcoded in the beginning of subroutine calcot.f). You can refer to this subroutine to see all the details of implementation.

With best regards,

Jean-Michel Hervouet

Thanks for your advice.
So, how calcot.f treats the pink line when it intercepts the bottom (i.e. A - B )??
Could I impose for that layer (for instance), zero velocities at the nodes along A - B??

Hello,

Between A and B the pink line will follow (more or less depending on hardcoded parameter) the bottom, so the velocities must be as on the bottom, zero for example but not necessarily. An important point is that you should give the same values for points that will be superimposed in the mesh, so that you do not trigger infinite vertical gradients.

Regards,

JMH

Thanks a lot.
It's like that layers below the bottom converge to the bottom values, equally for the free surface.
I didn't understand the second point you mentioned about, but i'll go back later on this.
Regards

G