open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear JMH,

I am sorry to interrupt you again.

I have tried several ways as the suggestions you provided latst time. they are separately as following,
OK, this is a common problem in long rivers, and a constant elevation is not a solution. here are 4 possibilities:

1) you put a constant depth everywhere and the water will flow down in its correct position during the first time steps (ugly, lots of supercritical flows, but it generally works).

2) BlueKenue and Fudaa allow you to give a not too unrealistic free surface elevation to start the computation.

3) Start from a dry river, except at the entrance, and fill it till reaching a steady state, use this as restart conditions for further computations (it may take time).

4) program yourself a tentative initial free surface in condin.f (see the section allowed for INITIAL CONDITION = 'SPECIAL')

Firstly, I tried the last one, though I can run it successfully, but because I don't know how to define the function for the H%R(I) at the time, then I tried to use other suggestions.

Secondly, I tried the first one, I set a constant depth, the details I did this are: I used the keyword in the cas file,

INITIAL CONDITIONS = 'CONSTANT DEPTH'
INITIAL DEPTH = 0.4

then I set the HUOB and HVOB as 700m3/s (flow in the upstream) in the open boundary file with prescribed Q in the conlim file, is this correct? but the water depth of result file is 0.4 in all the places. am I wrong to do so? then how can I use the 'constant depth'? Attached files is my cas file and cli file.

thirdly, I want to try the third suggestion, but I don't know how to start from a dry river, can you give me more details, I can use the keyword 'STOP IF A STEADY STATE IS REACHED' and set it as YES, then should I use other keywords to start a dry river?


I am really sorry to ask you so much questions, but I really think for several days and don't have ideas for them, any help will be appreciated!

Best regards!
yajuan

Hi

1 - What do you mean by "I don't know how to define the function for the H%R(I) at the time"? You want to introduce intial conditions, so the time is the intial time and there is no need to use it in subroutine condin.

2- For your second solution, the implementation is completely wrong! the prescribed flowrates must be introduced in the steering file with keyword PRESCRIBED FLOWRATES = 0;700

Please see the user manual, it is very clear and explicit. You can also use the validation cases which contain similar cases.

With my best regards

Riadh

Dear Riadh,

Thank you for your reply!

I tried to run telemac2d as you told me just now, I set the PRESCRIBED FLOWRATES = 0;700 and use the following in the steering file.

INITIAL CONDITIONS = 'CONSTANT DEPTH'
INITIAL DEPTH = 0.4

it can run indeed. but I checked with the conlim file(*.cli), the first boundary is 455(open boundary with prescribed Q), the second boundary is 544 (open boundary with prescribed H), so I think if I use the keyword PRESCRIBED FLOWRATES, it should be PRESCRIBED FLOWRATES = 700; 0, why should it be 0;700?

Best regards!
yajuan

Hello
To be sure about boundaries' numbering, try to run telemac only with 1 time-step. Then, you will read in the listing output (something like ):
THERE ARE 2 LIQUID BOUNDARIES:
LIQUID BOUNDARY 1: STARTS WITH NODE NUMBER ...
ENDS WITH NODE NUMBER ...

and the same thing for liquid boundary number 2. Therefore you can verify the numering of these boundaries

Eventually, you can verify directly in the result file if the boundary conditions are well prescribed.

With my best regards

Riadh ATA