open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear all,
I'm using telemac3D version v7p0 with a tracer in my simulation. I already have finish my first simulation and I would like to continue this simulation. My problem is that the initial distribution of tracer (next simulation) are the same in first simulation (at t=0)! However, the distribution of speed, depth seem correct. Does we can get out the value of tracer at the end of first simulation in my fortran? (the name of this variable). Thank you very much for any your suggestion.
Best regards
Thong NGUYEN
HCMUT, VN

Hello,

Yes it should work. There is a known bug with 2D variables, in telemac3d.f at line about 199, the line DO I=37,36+NTRAC should be changed into 38,37+NTRAC because an orbital velocity has been added in between. This would make a problem only if you do a computation continued with a 2D file to start with.

Another point to check is the names of your tracers. If you give the keyword NAMES OF TRACERS they must be 32 characters, 16 for the name and 16 for the unit. If you do not give all the characters maybe the names are different in two different runs.

If it is none of these cases I'll have to do a test here on another case.

With best regards,

Jean-Michel Hervouet

Dear Mr. J-M Hervouet,
Thank you very much for your quick response. I have checked the simulation with your two suggestions. The result isn't changed. That's mean the initial value of the next run are zero. Would you like to check this problem with other example?
P/S: I have checked this problem with all two version: v6p3 (after DO I=37,36+NTRAC be changed into 38,37+NTRAC as your suggestion) et v7p0 (already modified this bug).
With my best regards
Thong NGUYEN
HCMUT, VN

Hello Thong,

I tried both version 7.0 and 7.1, with a computation continued with 3 tracers called TOTO1, 2 and 3, with 32 characters in the name, and both versions work. You can see in the history of telemac3d.f that a bug was corrected in version 6.3, that had the same behaviour as you report, but now I do not see what it can be. What you can do:

Take telemac3d.f in your Fortran file. Find the sequence IF(.NOT.DEBU) THEN...
this is where the previous file is read.

At the end of this sequence add:

PRINT*,'NAME 1=',TEXTP3(ADR_TRAC),' TROUVE=',TROUVE(ADR_TRAC)
PRINT*,'NAME 2=',TEXTP3(ADR_TRAC+1),' TROUVE=',TROUVE(ADR_TRAC+1)

You will then see the names of the first 2 tracers taken for looking in the file, and array TROUVE will tell you if they have been found (TROUVE=1 means: found).

If you see nothing add:

PRINT*,'DEBU=',DEBU

before the sequence.

All this will tell you:

* if you are really in a computation continued
* if the variables are found

I suppose that you have checked that your tracers are really in the 3D file, not only in the 2D file.

Then if you do not find I would need your case...

With best regards,

Jean-Michel Hervouet

Dear Mr. J-M. Hervouet,
Thank you very much for your help. I don't know why the value of variable DEBU is T in telemac3d.f? while I have already defined SUITE DE CALCUL =OUI and SUITE 2D =OUI in steering file! I add an instruction IF(SUIT2) DEBU=.FALSE. at the head of SUBROUTINE CONDIM, the problem is resolved.
With my best regards
Thong NGUYEN
HCMUT, VN

I'm having this exact problem at the moment. Unfortunately I can't seem to get around the error. I'm using Telemac3d v7p0 and can't seem to get it to read in the tracer (salinity) from the a hotstart file. I'm making my hotstart file with Bluekenue, it does seem to read in depth and UV velocity.

Like Thong, the variable DEBU is T, so its not entering the IF(.NOT.DEBU) sequence.

I have also tried to write out a restart file but the file is corrupted each time.

Any help would be greatly appreciated,

Thanks

Glen

NIWA, NZ

Hello,

I can have a closer look at your files but first I want to be sure that there is no misunderstanding:

You have 3 possibilities to continue a computation:

1) with a previous 2D file;
2) with a previous 3D file;
3) with the hotstart mode.

The hotstart mode needs a specific file that is built by Telemac-3D itself in the previous run, and it is useful only to avoid any difference of digits, it is meant for automatic derivation or data assimilation studies that are sensitive to this, so let's discard it for the moment.

Then you cannot have 2 of these 3 possibilities at a time. In your case I guess that what you want is possibility 2, so you should not be in hotstart mode and you should not ask for a continuation with a 2D file. The previous computation result file is the 3D one.

If this does not clarify the matter just tell be and I'l have a closer look,

With best regards,

Jean-Michel Hervouet

Dear Mr. J-M. Hervouet

I apologise I wasn't very descriptive about my case.

What I trying to do is make a 2D hotstart file and initialise my 3D model using 2D continuation.

I'm starting the model with completely freshwater I only have 4 vertical layers to speed up computation. I introduce a salt tracer at the boundary and run it for 90+ days until I get a salt wedge in the river system. I want to then increase the number of vertical layers from 4 to 15 and run the model for 20 days where I introduce a flood event.

So I'm using Bluekenue to take the last time step from the 2D model results to and making a hotstart file similar to the Baxter river flood-inundation tutorial.

Out of interest, I have tried starting the model with the domain completely saline (basically the reverse of above. But this seems to introduce an instability which causes blow-ups? Please correct me if there is a better way to approach this...

Hello,

You could also continue a computation with 15 layers starting from a 3D file with 4 layers, there is some interpolation programmed in that case, though perhaps merely based on layer numbers and not coordinates.

For starting the computation, the experience we have, and I have seen other users doing that also, is to start with initial conditions that are not too far from reality, but it is easier to do in lakes and if it is not possible in your case the freshwater start is probably the best in your case.

With best regards,

Jean-Michel Hervouet

Hello Thong,

Sorry I had not seen this post and I discover it today. The problem in your case is indeed that you should not have SUITE 2D = OUI if you start from a 3D file. I'll add a checking of this in the next release and I'll stop if both types of continuations are at YES.

With best regards,

Jean-Michel Hervouet