open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello all,

I have for you a few questions about my simulations's results.

1. The velocity profiles given by 2D and 3D results (2D and 3D results of the same simulation launched with TELEMAC3D) are slightly different. I know that in 3D results, I have as many velocity profiles as number of horizontal planes used. But I want to compare 2D and 3D results, and I don't know if the 2D velocity profile was computed for the free surface or the average velocities of the chosen section?

2. I have my free surface below than the highest point in my bottom; so, the horizontal planes used converge to the same points (like illustrated below):



PICTURE -01-

In this point (encircled area, picture -01- ), my velocity reaches very high values (picture below):



PICTURE -02-
*Velocity distribution (telemac3D, the highest plane = free surface)*


By cons, when the free surface exceeds the highlight of my bottom (i.e.: I impose a constant elevation higher than the highest point in my bottom), the problem does not arise, velocity's pikes disappear; in border, of course.



PICTURE -03-
Distribution of my horizontal planes





PICTURE -04-
*Velocity distribution when my free surface exceeds the highest point in my bottom*

Is the convergence of planes to same points witch cause the high velocity values? Why?
did the program sum the velocity values of different planes at the same point?

Thank you in advance.

Hi

I thought your problem comes from the type of profile you impose on your boundary.
It look likes you have velocity profile = 1
In such case, it's better to have 4 or 5

Regards

I'll run simulation now and I'll let you know...

regards.

Hello,

Answer to the first question : the 2D velocity given in the 2D results file in Telemac-3D is the depth-averaged velocity.

Regards,

JMH

Re: Answer to the first question : the 2D velocity given in the 2D results file in Telemac-3D is the depth-averaged velocity.

Question: Does the depth-averaged velocity in the 2D results file in Telemac-3D include the vertical component of the velocity in the averaging, e.g. depth-averaging over U,V, and W. Or, is the depth-averaged velocity calculated in the 2D results file in Telemac-3D calculated only from the horizontal components of velocity, e.g. depth-averaging over U and V.

Hello,

It is the depth-averaged U and V, W not considered (and it is not very clear to me how taking W into account would modify the depth-averaged horizontal velocity...).

JMH

Hi,

Someone can explain me what TELEMAC do exactly when I put velocity profile = 4 or 5??? and why it's better than "velocity profile = 1" for my case???

"I think that I was trying to pass a linear inflow by different height along my front inlet section while I have a variable geometry at the entrance. Am I right???"

PS: Indeed, I got better results. I lost the Inlet effect described in pictures below:



*velocity profile=1*




*velocity profile=4*


Thank you in advance.

Hello,

With a profile 4 the velocities at the entrance are proportional to the square root of the depth (something like what you get in a canal with a Chézy law, so closer to reality than constant velocity). Profile 5 is like 4 but depth is computed with the lowest free surface of the entrance, not with the local free surface (this is to overcome a problem that happens sometimes when there both a main channel and a flood plain at the entrance, and the floodplain does not dry out at the entrance when the free surface level decreases).

Regards,

Jean-Michel Hervouet

Hi,

When I impose profile 4 (inlet and outlet velocity) I get this velocity distribution:



***Distribution of the velocity for velocity profile=4 in the inlet and outlet (2Dresults for TELEMAC3D)***

But I get also the disturbance in the middle of my field. I checked the velocity of the bottom and I found non-zero velocity from the input to the disturbance.



***Distribution of the velocity in the bottom for velocity profile=4(inlet and outlet)***


I thought it was because I impose a constant vertical profile, so I've changed to a logarithmic profile for imposing a zero velocity in the bottom and the disturbance just went backward.





**Velocity distribution for horizontal profile=4 and vertical profile=2 (inlet and outlet)**

The same thing happened to the bottom velocity:






So, how can I impose a zero velocity in the bottom???
Is that the logarithmic profile (vertical profile=2)of TELEMAC3D has the following form?






Thank you in advance.
Regards.

Hello,

Yes, if not specified, the velocity profile on the vertical will be constant, only after some distance the logarithmic profile will appear. If you prescribe a logarithmic profile at the entrance, you will not have 0 on the bottom but a value of velocity theoretically taken at some distance of the real bottom (e.g. at a distance equal to 1/10 of the first layer). Then the theoretical logarithmic profile prescribed at the entrance may not be equal to the numerical logarithmic profile obtained by the program (with a given number of planes on the vertical, so there may be some adaptation between the entrance and the profile that is Inside the domain. So basically your results seem "normal". See subroutine tfond.f for further détails.

With best regarda,

Jean-Michel Hervouet