Welcome, Guest
Username: Password: Remember me
Forgot your password? Forgot your username? Create an account
Forum
System Modules
TELEMAC-3D
NaN happening when coupled with Waqtel
Flat Threaded
  • Page:
  • 1
  • 2

TOPIC: NaN happening when coupled with Waqtel

NaN happening when coupled with Waqtel 5 years 5 months ago #33758

  • sanaz
  • sanaz's Avatar
  • OFFLINE
  • Senior Boarder
  • Posts: 69
  • Thank you received: 1
Hello,
I am trying to couple TElEMAC 3D with WAQTEL to see the effect of opening in the causeway in the tidal range, water quality and sediment transport around the causeway using TPXO database for the boundary condition.There is only one open boundary with prescribed u and v and prescribed tracer. I've been running a 3D model with salinity and temperature is taken into account . At first when Waqtel wasn't coupled with T3D the results looked reasonable.

in order to be more precise I tried to take the water-atmosphere temperature exchange into consideration. First it was giving me errors and I was not able to run the model, After using VERTICAL TURBULENCE MODEL= k-e, I was able to run the model. However, No matter what time step was used the NaN error occurred.(the model seems is running but it is kind of producing junk).
After further refinement the mesh size gradually decreases from 300m to 20m, I am still having that NaN occurring at the first few iteration.




DIFFUSION OF TRN1
GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS: 200 RELATIVE PRECISION: NaN
MURD3D_POS SCHEME: 14 1 ITERATIONS
DIFFUSION OF TRN2
GRACJG (BIEF) : 0 ITERATIONS, RELATIVE PRECISION: 0.2494701E-15
MURD3D_POS SCHEME: 14 1 ITERATIONS
DIFFUSION OF TRN3
GRACJG (BIEF) : SOLUTION X=0 BECAUSE L2-NORM OF B VERY SMALL: 0.000000
MASS BALANCE
MASS BALANCE

WATER
MASS AT THE PREVIOUS TIME STEP : 0.1432433E+12
MASS AT THE PRESENT TIME STEP : 0.1447052E+12
MASS LEAVING THE DOMAIN DURING THIS TIME STEP : NaN
ERROR ON THE MASS DURING THIS TIME STEP : NaN
FLUX BOUNDARY 1: NaN M3/S ( >0 : ENTERING <0 : EXITING )

TRACER 1: TEMPERATURE , UNIT : C * M3)

ADVECTIVE FLUX THROUGH BOUNDARIES OR SOURCES : NaN
DIFFUSIVE FLUX THROUGH THE BOUNDARIES : NaN
MASS AT THE PREVIOUS TIME STEP : NaN
MASS AT THE PRESENT TIME STEP : NaN
MASS EXITING (BOUNDARIES OR SOURCE) : NaN
ERROR ON THE MASS DURING THIS TIME STEP : NaN

TRACER 2: SALINITY , UNIT : G/L * M3)

ADVECTIVE FLUX THROUGH BOUNDARIES OR SOURCES : 0.000000
DIFFUSIVE FLUX THROUGH THE BOUNDARIES : -0.000000
MASS AT THE PREVIOUS TIME STEP : 0.4010811E+13
MASS AT THE PRESENT TIME STEP : 0.4051745E+13
MASS EXITING (BOUNDARIES OR SOURCE) : 0.000000
ERROR ON THE MASS DURING THIS TIME STEP : -0.4093317E+11

TRACER 3: CONCENTRATION , UNIT : % * M3)

ADVECTIVE FLUX THROUGH BOUNDARIES OR SOURCES : 0.000000




Please see the attached cas files and snapshot of errors.
Many thanks in advance, any advice would be very appreciated!

Sanaz
Attachments:
The administrator has disabled public write access.

NaN happening when coupled with Waqtel 5 years 5 months ago #33761

  • Yunhao Song
  • Yunhao Song's Avatar
  • OFFLINE
  • Senior Boarder
  • Posts: 118
  • Thank you received: 9
Hi Sanaz,

To my limited experience, one way to bypass this tricky NaN error is cutting off dry area in your domain and set 'tidal flats = NO' to see if the results still look reasonable. Hope it helps.

Regards,
Yunhao
The administrator has disabled public write access.
The following user(s) said Thank You: sanaz

NaN happening when coupled with Waqtel 5 years 5 months ago #33762

  • sanaz
  • sanaz's Avatar
  • OFFLINE
  • Senior Boarder
  • Posts: 69
  • Thank you received: 1
Hello Yunhao,
Thank you so much for ur help.
I set the Tidal flat to NO but it is giving the following error:
PROPAGATION AND DIFFUSION WITH WAVE EQUATION
GRACJG (BIEF) : 3 ITERATIONS, RELATIVE PRECISION: 0.4760077E-06
application called MPI_Abort(MPI_COMM_WORLD, 2) - process 0 CALCOT: PLANES 1 AND 2

INTERCROSS AT POINT 22
LOWER POINT : 0.0000000000000000
HIGHER POINT: -1.0578974443413009E-010
DIFFERENCE : -1.0578974443413009E-010
DEPTH : -5.2894872217065045E-010

PLANTE: PROGRAM STOPPED AFTER AN ERROR
RETURNING EXIT CODE: 2

job aborted:
rank: node: exit code[: error message]
0: L0003127CA.corp.pbwan.net: 2: process 0 exited without calling finalize
_____________
runcode::main:
:
|runCode: Fail to run
|C:\opentelemac-mascaret\mpich2\bin\mpiexec.exe -wdir C:\opentelemac-mascaret\v8p0r2\Project\newtd\telemac3d.cas_2019-06-17-09h45min48s -n 1 C:\opentelemac-mascaret\v8p0r2\Project\newtd\telemac3d.cas_2019-06-17-09h45min48s\out_mesh_transf.exe
|~~~~~~~~~~~~~~~~~~
|application called MPI_Abort(MPI_COMM_WORLD, 2) - process 0
|~~~~~~~~~~~~~~~~~~




What I understood from the manual First, if the user has ascertained that his/her model has no tidal area throughout the simulation,
the processing of such areas can be deactivated by setting the keywords TIDAL FLATS to NO.
my main goal is to study the tidal range in the domain. So if I set the Tidal flat to NO is not it going to effect the accuracy of the tidal range ??


Thanks in advance for your help
The administrator has disabled public write access.

NaN happening when coupled with Waqtel 5 years 5 months ago #33784

Dear Sanaz,

ATMOSPHERE-WATER EXCHANGE MODEL = 2

combined with tidal flat is not a stable option (in versions <= v8p0). You can try

ATMOSPHERE-WATER EXCHANGE MODEL = 1

In addition tips form aleroy may help you:
www.opentelemac.org/index.php/kunena/21-...se-with-some-figures

Best regards,
Jacques
The administrator has disabled public write access.
The following user(s) said Thank You: sanaz

NaN happening when coupled with Waqtel 5 years 5 months ago #33801

  • sanaz
  • sanaz's Avatar
  • OFFLINE
  • Senior Boarder
  • Posts: 69
  • Thank you received: 1
Thanks a lot Jacques for your reply. I will try on that and will let you know the result.

Sanaz
The administrator has disabled public write access.

NaN happening when coupled with Waqtel 5 years 5 months ago #33826

  • sanaz
  • sanaz's Avatar
  • OFFLINE
  • Senior Boarder
  • Posts: 69
  • Thank you received: 1
Hello dear Jacques,
when I used ATMOSPHERE-WATER EXCHANGE MODEL=2 which uses the complete balance. At the first few seconds it started giving Nan error. When I changed that to ATMOSPHERE-WATER EXCHANGE MODEL=1 which use the linearized formula at the free surface it gave me NAN error after 20 mins of running the model.
I tried to refine the mesh and also use a small value for time step but still it is giving me the NAN ERROR. It is just The NAN error occurred later.
I am not sure if I further refine the mesh it would be helpful.
I would be thankful to have your comments on my issue.

Thanks in advance,
Sanaz
The administrator has disabled public write access.

NaN happening when coupled with Waqtel 5 years 5 months ago #33827

Hello,

Did you try the tips from aleroy?

mainly:
SCHEME FOR ADVECTION OF VELOCITIES = 5
 SCHEME FOR ADVECTION OF K-EPSILON = 5
 SCHEME FOR ADVECTION OF TRACERS = 5
 / predictor-corrector with local implicitation implicite
 SCHEME OPTION FOR ADVECTION OF VELOCITIES = 4
 SCHEME OPTION FOR ADVECTION OF K-EPSILON = 4
 SCHEME OPTION FOR ADVECTION OF TRACERS = 4

Best regards,
J
The administrator has disabled public write access.

NaN happening when coupled with Waqtel 5 years 5 months ago #33830

  • sanaz
  • sanaz's Avatar
  • OFFLINE
  • Senior Boarder
  • Posts: 69
  • Thank you received: 1
Hello Jacques,
Yes I tried both number and it gave me the following error:
ADVECTION STEP
MURD3D: ITERATION NO. REACHED 100 , STOP.application called MPI_Abort(MPI_COMM_WORLD, 2) - process 0

ALFA = 1.3346022521864294E-002 GUILTY POINT = 289128

PLANTE: PROGRAM STOPPED AFTER AN ERROR
RETURNING EXIT CODE: 2

job aborted:
rank: node: exit code[: error message]
0: L0003127CA.corp.pbwan.net: 2: process 0 exited without calling finalize


However, I just found that The keyword SCHEME OPTION
FOR ADVECTION OF VELOCITIES must be used coupled with the keyword SCHEME FOR ADVECTION
OF VELOCITIES. As I just used only the keyword SCHEME FOR ADVECTION
OF VELOCITIES. Now I put a new run seems to be working. will let you know the result.


Thanks alot for your help. will get back to you soon.

Sanaz
The administrator has disabled public write access.

NaN happening when coupled with Waqtel 5 years 5 months ago #33844

  • sanaz
  • sanaz's Avatar
  • OFFLINE
  • Senior Boarder
  • Posts: 69
  • Thank you received: 1
Hello Jacques,
I run the model with the time step of 10 sec with the original mesh(from 300 m to 10 m in some area of interest) and using the scheme value recommended. It started the NAN error after 5 hours. I would really appreciate if you could have a look in the error. Also I attached the two cas files.

Thanks alot in advance,
Sanaz



=======================================================================
ITERATION 2222 TIME 0 D 6 H 10 MN 20.0000 S ( 22220.0000 S)
================================================================================
ADVECTION STEP
ADVECTION STEP
JACOBI: 1 ITERATIONS, RELATIVE PRECISION = 6.5583550893430515E-006
MURD3D_LIPS: 1 ITERATIONS
JACOBI: 1 ITERATIONS, RELATIVE PRECISION = 2.9945573260633078E-006
MURD3D_LIPS: 1 ITERATIONS
JACOBI: 2 ITERATIONS, RELATIVE PRECISION = 2.9085444806878468E-006
MURD3D_LIPS: 1 ITERATIONS
DIFFUSION OF WN
GRACJG (BIEF) : 6 ITERATIONS, RELATIVE PRECISION: 0.9470397E-05
PROPAGATION AND DIFFUSION WITH WAVE EQUATION
PROPAGATION AND DIFFUSION WITH WAVE EQUATION
GRACJG (BIEF) : 40 ITERATIONS, RELATIVE PRECISION: 0.9926640E-06
POSITIVE DEPTHS OBTAINED IN 8 ITERATIONS
DYNAMIC PRESSURE STAGE
DYNAMIC PRESSURE STAGE
GRACJG (BIEF) : 23 ITERATIONS, RELATIVE PRECISION: 0.9780348E-04
VELOCITY PROJECTION STEP
VELOCITY PROJECTION STEP
ADVECTION-DIFFUSION OF K-EPSILON OR OMEGA STEP
ADVECTION-DIFFUSION OF K-EPSILON OR OMEGA STEP
JACOBI: 11 ITERATIONS, RELATIVE PRECISION = 3.1793835883390772E-007
MURD3D_LIPS: 1 ITERATIONS
DIFFUSION OF AKN
GRACJG (BIEF) : 1 ITERATIONS, RELATIVE PRECISION: 0.4058804E-08
JACOBI: 10 ITERATIONS, RELATIVE PRECISION = 3.2045367797734355E-007
MURD3D_LIPS: 1 ITERATIONS
DIFFUSION OF EPN
GRACJG (BIEF) : 1 ITERATIONS, RELATIVE PRECISION: 0.2327863E-07
ADVECTION-DIFFUSION OF TRACERS
ADVECTION-DIFFUSION OF TRACERS
JACOBI: 6 ITERATIONS, RELATIVE PRECISION = 6.3600864200781847E-007
MURD3D_LIPS: 1 ITERATIONS
DIFFUSION OF TRN1
GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS: 200 RELATIVE PRECISION: NaN
JACOBI: 1 ITERATIONS, RELATIVE PRECISION = 1.5278477491101602E-007
MURD3D_LIPS: 1 ITERATIONS
DIFFUSION OF TRN2
GRACJG (BIEF) : 1 ITERATIONS, RELATIVE PRECISION: 0.2588168E-06
JACOBI: 1 ITERATIONS, ABSOLUTE PRECISION = 0.0000000000000000
MURD3D_LIPS: 1 ITERATIONS
DIFFUSION OF TRN3
GRACJG (BIEF) : SOLUTION X=0 BECAUSE L2-NORM OF B VERY SMALL: 0.000000
MASS BALANCE
MASS BALANCE

WATER
MASS AT THE PREVIOUS TIME STEP : 0.1454358E+12
MASS AT THE PRESENT TIME STEP : 0.1454358E+12
MASS LEAVING THE DOMAIN DURING THIS TIME STEP : -28885.54
ERROR ON THE MASS DURING THIS TIME STEP : -0.3312621E-04
FLUX BOUNDARY 1: 2888.554 M3/S ( >0 : ENTERING <0 : EXITING )

TRACER 1: TEMPERATURE , UNIT : C * M3)

ADVECTIVE FLUX THROUGH BOUNDARIES OR SOURCES : 0.3182915E+08
DIFFUSIVE FLUX THROUGH THE BOUNDARIES : NaN
MASS AT THE PREVIOUS TIME STEP : 0.1596364E+13
MASS AT THE PRESENT TIME STEP : NaN
MASS EXITING (BOUNDARIES OR SOURCE) : NaN
ERROR ON THE MASS DURING THIS TIME STEP : NaN

TRACER 2: SALINITY , UNIT : G/L * M3)

ADVECTIVE FLUX THROUGH BOUNDARIES OR SOURCES : 0.4516572E+08
DIFFUSIVE FLUX THROUGH THE BOUNDARIES : -0.000000
MASS AT THE PREVIOUS TIME STEP : 0.2070255E+13
MASS AT THE PRESENT TIME STEP : 0.2069804E+13
MASS EXITING (BOUNDARIES OR SOURCE) : 0.4516572E+09
ERROR ON THE MASS DURING THIS TIME STEP : -4674.618

TRACER 3: CONCENTRATION , UNIT : % * M3)

ADVECTIVE FLUX THROUGH BOUNDARIES OR SOURCES : 0.000000
DIFFUSIVE FLUX THROUGH THE BOUNDARIES : -0.000000
MASS AT THE PREVIOUS TIME STEP : 0.000000
MASS AT THE PRESENT TIME STEP : 0.000000
MASS EXITING (BOUNDARIES OR SOURCE) : 0.000000
ERROR ON THE MASS DURING THIS TIME STEP : 0.000000

================================================================================
ITERATION 2223 TIME 0 D 6 H 10 MN 30.0000 S ( 22230.0000 S)
================================================================================
Attachments:
The administrator has disabled public write access.

NaN happening when coupled with Waqtel 5 years 4 months ago #33848

Dear Sanaz,

I propose you to comment or remove all numerical options you set in your t3d.cas file (particularly from line 207 to the end of file). Except the one I mentioned:
SCHEME FOR ADVECTION OF VELOCITIES = 5
SCHEME FOR ADVECTION OF K-EPSILON = 5
SCHEME FOR ADVECTION OF TRACERS = 5
 / predictor-corrector with local implicitation implicite
SCHEME OPTION FOR ADVECTION OF VELOCITIES = 4
SCHEME OPTION FOR ADVECTION OF K-EPSILON = 4
SCHEME OPTION FOR ADVECTION OF TRACERS = 4

LAW OF BOTTOM FRICTION = 5
NON-HYDROSTATIC VERSION = NO / DEFAULT VALUE… REMOVE KEYWORD
IMPLICITATION FOR VELOCITIES = 0.55

Best regards,
The administrator has disabled public write access.
Flat Threaded
  • Page:
  • 1
  • 2
Moderators: pham
Forum
System Modules
TELEMAC-3D
NaN happening when coupled with Waqtel

The open TELEMAC-MASCARET template for Joomla!2.5, the HTML 4 version.