open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I have tested your case and all seems to work, except that in condim you initialize the tracer 2 which does not exist (here the compiler sees it and stops, maybe it erases some memory in yoour case).

Anyway the solution to the mystery could be that you have :

TREATMENT OF FLUXES AT THE BOUNDARIES =2;2

2 instead of 2;2 would be sufficient here but the "2" means that the program will ensure that the correct flux of tracer is entered at your boundary. 1 would mean that you want to see the prescribed value, and this cannot be done concurrently with a correct flux.

So to be sure you could change the 2 into 1 and see what happens at your boundary, but as soon as it is OK the solution 2 is better. With solution 1 the flux entered is wrong and eventually it is the correct flux that you want in your domain.

With best regards,

Jean-Michel Hervouet

Dear Jean-Michel,

Thank you very much for looking into it. I made the changes and it is running. Fingers crossed I will get good results. Sorry for confusion with two tracers in Condim. I had it ready for case with two tracers, as this is my goal, but since I had problems with Tra-prof-z subroutine, I reduced to one tracer only, but, silly me, did not adjust the Condim. I have changed Treatment of fluxes at the boundaries to 1, and if it works file I will change it to 2 as per your suggestion. Thank you so much, your help is much appreciated. I would not never be able to find the error myself. I will let you know of results. If everything goes well, then I will try to run it for two tracers. Thanks again.

Kind Regards!

Violeta

Dear Jean-Michel,

Thank you again for sorting my vertical tracers profile problem. I amended condime as you suggested and it works fine. Just letting you know if you would like to amend the BORD3D.f subroutine if anyone wishes to use Tracers values as I did. Thanks a mil.
Sorry for coming back with similar question. I have tried to increase the number of tracers to 2 (now Temperature and Salinity) in TRA_PROF_Z but it is not working. I have been using below code. I have amended the Condim subroutine for two tracers as well this time. I think my problem is how to tell the system to differentiate between Tracer 1 and Tracer 2. Please advise. Thanks a mil.

IF(IOPT.EQ.0) THEN
IF(ITRAC.NE.0.OR..NOT.SEDI) THEN
IF(IPLAN.GT.8) THEN
TRA_PROF_Z(1)=1.D0
TRA_PROF_Z(2)=25.D0
ELSE
TRA_PROF_Z(1)=4.34D0
TRA_PROF_Z(2)=35.D0
ENDIF
ENDIF

Kind Regards!

Violeta

Hello,

TRA_PROF_Z is not an array, it is called for every tracer, so you need to write :

IF(IOPT.EQ.0) THEN
IF(ITRAC.NE.0.OR..NOT.SEDI) THEN
IF(IPLAN.GT.8) THEN
IF(ITRAC.EQ.1) THEN
TRA_PROF_Z=1.D0
ELSEIF(ITRAC.EQ.2) THEN
TRA_PROF_Z(2)=25.D0
ENDIF
ELSE
IF(ITRAC.EQ.1) THEN
TRA_PROF_Z(1)=4.34D0
ELSEIF(ITRAC.EQ.2) THEN
TRA_PROF_Z(2)=35.D0
ENDIF
ENDIF
ENDIF

With best regards,

Jean-Michel Hervouet

Dear Jean-Michel,

I have made some progress on stratification case with the open boundary. The results are better now, but still not there yet. If I run it for longer the stratification is not stable.

I wonder what else I could do to improve the case. Looks like boundary tracers vertical profile works fine now, but stratification in domain is not perfect yet.

So far I have 10 layers (30 layers result is more stable but still stratification is smoothing after days, but with 10 layers smoothing starts after just 20 days). I have used Hydrostatic Inconsistency Filter with Trisou subroutine, FORMUL=3. Also for Vertical Tracers profile I updated BORD3D and TRA_PROF_Z. I also run the case with 30 layers and gradual change in tracers gradient, say for salinity change 35-33-31-27-25 assigned to specific layers but with same results.

Could it be that I do not have enough number of layers in interface area? Then later, if I add tides to the case and mixing occurs, then how would I know what level are affected exactly to refine them? The max depth of investigated area is 900m. Or maybe I have wrong Friction Law and Friction Coefficient values (Strickler with friction coefficient for the bottom-30 and for lateral solid boundaries -30)? I have attached some graphical results for salinity at 0.0 time and after two months. Also velocity at boundary is increasing (0.4m/s) but I do not have any prescribed velocities there? Please advise. Thanks in advance.

Kind Regards!

Violeta

Dear Violeta,

Thank you very much ,I have learmed a lot form you topics.I am doing the similar simulation like you have done and have the similar problems like you .I am really curious about whether you solve the problems, I really hope you cam give me some advises!

Best regards !
Andy

Dear Andy,

It has been a while since I run those cases. As I remember it worked fine for me. My case was complicated, as I had to set up specific tracers (temp and salinity) values in domain and different values for these tracers at the open boundary (30 layers in total). Also, I think the choice of mesh transformation fix the problem. See my fortran and boundary files attached. In fortran file watch for TRA_PROF_Z and CONDIM sub-routines. Also, check which mesh transformation I used. It is all in fortran file. I saved it as .cas extension. I am afraid, I can't remember details as I did it more than 3 years ago. In this case I have strong stratification that is applicable to the Greenland fjords and tidal model. Good luck.

Kind Regards,

Violeta