open TELEMAC-MASCARET The mathematically superior suite of solvers

By the way,in the diffusion-propagation step, it shows the calculation is exceeding maximum iterations.

Hello,

After your second post I would rather look to hydrodynamics. Differences between version 6.1 and 6.3 are almost always due to the keywords IMPLICITATION FOR DEPTH and IMPLICITATION FOR VELOCITY which were before set to 1. in some options, whereas now we take the user value or the default value which is 0.55. To be sure add :

IMPLICITATION FOR DEPTH : 1.
IMPLICITATION FOR VELOCITY : 1.

in your Telemac-2D steering file and see what happens.

Another question is what is your parameter :

TREATMENT OF THE LINEAR SYSTEM

If it is 1 (primitive equations) you can try 2 (wave equation) and it will be much faster.

If it is 2, then you can use SOLVER : 1 instead of 3 (the default) and it will be faster also.

With best regards,

Jean-Michel Hervouet

Dear Cath，

Many thanks to you!

Now I can get a sediment result from telemac-6.3 by setting "MUD CONCENTRATION PER LAYER = 90" in my case file.The result is not good.But I will try the new keywords according to your attachment file.

By the way,my first post is missing.

With best regards,

He Jiawei

Dear Jean-Michel Hervouet,

Thanks for your suggestion.

I run the case again after setting "IMPLICITATION FOR DEPTH =1" and "IMPLICITATION FOR VELOCITY =1",but the same problem( the calculation of the diffusion-propagation step exceeds maximum iterations) occurs.

If I add "TREATMENT OF THE LINEAR SYSTEM=2" and "SOLVER=1",an error will happen.

With best regards,

He Jiawei

Hello,

So I conclude that hydrodynamics now works and that the problem is in the sediment part. The message comes from suspension_evol.f in library Sisyphe, it would mean that there is no more sediment available on the bottom but I do not know why we should stop for this. Waiting for other's advice you can try to add this subroutine in your Fortran file and remove the sequence:

CALL PLANTE(1)
STOP

after the error message


On hydrodynamics you can also try faster solutions :

SOLVER : 1 /this will work for sure

and possibly (but check free surface possible oscillations...)

SUPG OPTION = 1;0
DISCRETIZATIONS IN SPACE = 11;11
FREE SURFACE GRADIENT COMPATIBILITY : 0.9

With best regards,

Jean-Michel Hervouet

Dear Jean-Michel Hervouet,

Thanks for your suggestion.

I run the case without coupling with sisyphe and get a good hydrodynamic result(telemac 6.3).It run much faster.
But if I run the case with sisyphe,the value of total minimal evolution would increase to 1.69E2 from 0.9E-1 during only one time step(13h15min-13h30min).
The problem would not happen in telemac v6p1.

With best regards,

He Jiawei

Hi,

There were a number of changes in the keywords from 6.1 to 6.2
which may be the reason (see file enclosed)

For example I don't think that you are specifying the bed concentration in your case file:
MUD CONCENTRATION PER LAYER = 50 per default which is a very soft layer
So you should try to increase this value.

You can also increase the input concentration.

Good luck, Cath

Dear Cath，

Many thanks to you!

Now I can get a sediment result from telemac-6.3 by setting "MUD CONCENTRATION PER LAYER = 90" in my case file.The result is not good.But I will try the new keywords according to your attachment file.

By the way,my first post is missing.

With best regards,

He Jiawei