open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi!

I want to run a simulation with 35 sources and at least 15 tracers (with a source file defining variable flow from the sources). This doesn't work - is there any way to go around the problem? I only get cryptic error messages like:

SCARACT
TOO MANY LOST TRACEBACKS IN 2 PROCESSORS

or

MURD3D: ITERATION NO. REACHED 100 , STOP.
ALFA = 1.3813717026794594E-006
GUILTY POINT = 1

Thankful for any help!

Charlotta

Hi

Hard to say with the only informations you post ...

As it seems you're running in parallel, maybe the first test should be to try in scalar mode...

Kind regards

Actually I found the problem and it was a quite easy one this time: In Sweden we use ',' as decimal separator and I had forgot to change to '.' in the sources file.

However, now the problem is that I cannot get results from all the tracers in the results file (I have 15 tracers). Is there a way to get around this?

Best regards,
Charlotta

Hi
What did you ask in your steering file for the output?
If you ask for T in the graphic printout variables, it seems to me that all the tracers are writen in the file.

Regards

I have asked for:
VARIABLES FOR 3D GRAPHIC PRINTOUTS = 'Z,U,V,W,TA*'

Hello,

I think that to have the tracers beyond 9 you need to have:

VARIABLES FOR 3D GRAPHIC PRINTOUTS = 'Z,U,V,W,TA*,TA1*'

A known problem is if you use Thompson boundary conditions with many tracers this will trigger computing characteristics for more than 10 variables and this is not programmed (and is a very heavy change).

With best regards,

Jean-Michel Hervouet

Thanks Jean-Michel,

I am running with Thomson boundary conditions, so I will have a try with Classic boundary and use the TA1* and hopefully it will work.

Best regards,
Charlotta

That sounds right!
You don't have any message in the listing?
The simulation runs well but you don't have the tracers results in the results files. Did you have a part of them or none?

Regards

And my next problem with a similar model is that I cannot get more than 19 sources to work - is there any way to get around this? I have seen other forum posts with more sources so it should work I guess...

Best regards,
Charlotta

Hello,

Yes it should work, someone tried 500 sources, but 2 arrays are allocated by source so it becomes very heavy. In the next version keywords will be available for giving a maximum number of sources and allocate a number of arrays.

In version 7.0 the maximum of sources is 100, so I do not see why 19 is a limit, what kind of error do you get ?

Regards,

JMH