open TELEMAC-MASCARET The mathematically superior suite of solvers

Dear all,

I would like to achieve the mass-conservative of tracer in my model with the existence of source /sink and tidal flat.

So on the basis of option 1-characteristics method. I have tried SCHEME FOR ADVECTION OF TRACERS 3,4,5 and 13,14.



Yet when I choose the SCHEME FOR ADVECTION OF TRACERS =3/4/5. This error information show up:'CVTRVF: 200 SUN-ITERATIONS REQUIRED FOR THE DISTRIBUTIVE SCHEME. DECREASE THE TIME-STEP ' .

Also, when I choose the SCHEME FOR ADVECTION OF TRACERS =13/14, 'GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS' happens after some time.

I have reduce the time step as much as I can, but still no working. Is anyone know how to fix this? Many thanks!

Best Regards,
Bin

Hello,

Have you tried to simplify your case and see if the problem still occurs (e.g. no culverts and/or use classical values for some keywords such as OPTION FOR THE DIFFUSION OF TRACERS = 1 first which should better work with tidal flats,
DISCRETIZATIONS IN SPACE = 11 just to test, TURBULENCE MODEL = 1 just to see the influence of the turbulence model).

Even if it is not mass-conservative, but just to try if you also have a problem, what about the method of characteristics for advection scheme for tracers?

Last test if possible, include corfon subroutine in your FORTRAN FILE and decrease the bottom elevation for dry areas so that they are always wet during your computation to see if the problem comes from tidal flats or not (it seems to be the problem).


Hope this helps,

Chi-Tuan

Dear Chi_Tuan,

Many thanks for your reply.

I have solved this problem by using coupled linear system(wave equation I used before). I found out that there are some relationship between choose of linear system and the advection velocity.

But I will try several test with your suggestion to see if there will be a better result. Thanks again.

Best Regards,
Bin