Welcome, Guest
Username: Password: Remember me
Forgot your password? Forgot your username? Create an account
Forum
System Modules
GAIA / SISYPHE
Mass balance info when GAIA coupled with TELEMAC-3D
Flat Threaded

TOPIC: Mass balance info when GAIA coupled with TELEMAC-3D

Mass balance info when GAIA coupled with TELEMAC-3D 4 years 10 months ago #35243

  • qilong
  • qilong's Avatar
  • OFFLINE
  • Expert Boarder
  • Posts: 340
  • Thank you received: 33
Dear Telemac fellows,

I have a question regarding the mass balance info printed out by Telemac-3D coupled with Gaia.

In the following printouts, the ADVECTIVE FLUX THROUGH BOUNDARIES OR SOURCES is given. I'm wondering what does the positive (or negative) sign mean here?

I might be wrong, but my calculation seems to suggest that positive ADVECTIVE FLUX THROUGH BOUNDARIES OR SOURCES is going out of the domain from the boundary. Is it true?
================================================================================
 ITERATION      102    TIME: 17 MN   0.0000 S     (   1020.0000 S)

--------------------------------------------------------------------------------
       ADVECTION-DIFFUSION OF K-EPSILON OR OMEGA STEP
--------------------------------------------------------------------------------
       CALCULATION OF VERTICAL VELOCITY STEP
 DIFFUSION OF AKN
 GRACJG (BIEF) :        4 ITERATIONS, ABSOLUTE PRECISION:   0.2073210E-06
 DIFFUSION OF EPN
 GRACJG (BIEF) :        2 ITERATIONS, ABSOLUTE PRECISION:   0.2738162E-06
--------------------------------------------------------------------------------
       ADVECTION-DIFFUSION OF TRACERS
 DIFFUSION OF TRN1
 EQUNOR (BIEF) :        5 ITERATIONS, RELATIVE PRECISION:   0.4734383E-08

                    GAIA MASS-BALANCE OF SEDIMENTS PER CLASS:
     SEDIMENT CLASS NUMBER                     =        1
     TOTAL BED EVOLUTIONS                      =     614757.8      ( KG )
     BOUNDARIES BEDLOAD FLUX                   =     0.000000      ( KG/S  >0 = ENTERING )
     BEDLOAD FLUX BOUNDARY    1                =     0.000000      ( KG/S  >0 = ENTERING )
     BEDLOAD FLUX BOUNDARY    2                =     0.000000      ( KG/S  >0 = ENTERING )
     EROSION FLUX                              =     3.048749      ( KG/S )
     DEPOSITION FLUX                           =     61478.83      ( KG/S )
     TOTAL MASS                                =    0.6289907E+08  ( KG )
     LOST MASS                                 =    0.2328306E-09  ( KG )
     RELATIVE ERROR 0N MASS                    =    0.3787356E-15

 MAXIMAL EVOLUTION (M)    :    0.7210632E-05 NODE  :  2559
 MINIMAL EVOLUTION (M)    :   -0.1473098E-06 NODE  :  3149
 TOTAL MAXIMAL EVOLUTION  :    0.7429914E-03 NODE  :  2558
 TOTAL MINIMAL EVOLUTION  :     0.000000     NODE  :     3
--------------------------------------------------------------------------------
                MASS BALANCE

  WATER
MASS AT THE PREVIOUS TIME STEP                :    0.6857326E+10
MASS AT THE PRESENT TIME STEP                 :    0.6857291E+10
MASS LEAVING THE DOMAIN DURING THIS TIME STEP :     35588.25
ERROR ON THE MASS DURING THIS TIME STEP       :   -0.8558246E-03
FLUX BOUNDARY    1:     80.00000     M3/S  ( >0 : ENTERING  <0 : EXITING )
FLUX BOUNDARY    2:    -3638.825     M3/S  ( >0 : ENTERING  <0 : EXITING )

  TRACER  1: COH SEDIMENT1   , UNIT : g/l             * M3)

ADVECTIVE FLUX THROUGH BOUNDARIES OR SOURCES  :    0.2668239
DIFFUSIVE FLUX THROUGH THE BOUNDARIES         :     61475.78
MASS AT THE PREVIOUS TIME STEP                :    0.2206871E+09
MASS AT THE PRESENT TIME STEP                 :    0.2200725E+09
MASS EXITING (BOUNDARIES OR SOURCE)           :     614760.5
ERROR ON THE MASS DURING THIS TIME STEP       :    -197.0554

Kind regards,
Qilong
The administrator has disabled public write access.

Mass balance info when GAIA coupled with TELEMAC-3D 4 years 10 months ago #35247

  • pavans
  • pavans's Avatar
  • OFFLINE
  • Moderator
  • Posts: 46
  • Thank you received: 17
Hello Qilong,

ADVECTIVE FLUX THROUGH BOUNDARIES OR SOURCES has positive sign when tracers/sediments are going out of the domain (through boundaries or sources) and negative when they are entering. It represents the sum of the advective fluxes through all the liquid boundaries.

As you say, in your case, the balance through the boundaries says that sediment is going out of the domain.

Best regards,

Sara
The administrator has disabled public write access.
The following user(s) said Thank You: qilong

Mass balance info when GAIA coupled with TELEMAC-3D 4 years 10 months ago #35249

  • qilong
  • qilong's Avatar
  • OFFLINE
  • Expert Boarder
  • Posts: 340
  • Thank you received: 33
Hello Sara,

Thanks for your confirmation. Then I have another question about the mass balance in my model, which is a funnel shaped domain with upstream stream discharge and downstream water level as boundary conditions. Besides, I imposed sediment concentration at downstream boundary, which is dependent on the flow velocity during both inflow and outflow.

The model result seems suggesting that, the sediment mass is created inside the domain. The mass of the tracer is not conserved. On the other hand, GAIA result seems suggesting that the total mass of bed is well conserved in the model (equals to the accumulated deposition mass - accumulated erosion mass).

In the end of the simulation, it says (I think this info is for the sediment in water column only):
                FINAL MASS BALANCE
T =    31536000.0000

--- WATER ---
INITIAL MASS                        :    0.6858485E+10
FINAL MASS                          :    0.5728677E+10
MASS LEAVING THE DOMAIN (OR SOURCE) :    0.1129673E+10
MASS LOSS                           :     134493.6

--- TRACER 1: COH SEDIMENT1   , UNIT : g/l             * M3)
INITIAL MASS                        :    0.3026454E+09
FINAL MASS                          :    0.1331185E+09
MASS EXITING (BOUNDARIES OR SOURCE) :    0.1518944E+10
MASS LOSS                           :   -0.1349417E+10

 END OF TIME LOOP

The GAIA says (I think this info is for the bed layer):
                    FINAL MASS-BALANCE OF SEDIMENTS:

                    GAIA MASS-BALANCE OF SEDIMENTS PER CLASS:
     SEDIMENT CLASS NUMBER                     =        1
     CUMULATED BED EVOLUTIONS                  =    0.7987193E+08  ( KG )
     CUMULATED BOUNDARIES BEDLOAD MASS         =     0.000000      ( KG )
     CUMULATED BEDLOAD BOUNDARY    1           =     0.000000      ( KG )
     CUMULATED BEDLOAD BOUNDARY    2           =     0.000000      ( KG )
     CUMULATED EROSION                         =    0.1495109E+13  ( KG )
     CUMULATED DEPOSITION                      =    0.1495189E+13  ( KG )
     INITIAL MASS                              =     0.000000      ( KG )
     INITIAL MASS ACTIVE LAYER                 =     0.000000      ( KG )
     TOTAL MASS                                =    0.7987193E+08  ( KG )
     CUMULATED LOST MASS                       =   -0.6152344E-01  ( KG )
     CUMULATED RELATIVE ERROR ON MASS          =   -0.3263777E-05
     CUMULATED LOST MASS (INI-FINAL+FLUXES)    =   -0.6152344E-01  ( KG )

So here I don't really understand the result anymore. What could be the reason for the model to create the mass inside the water column? Do you have similar experience with the mass balance of the model?

Kind regards,
Qilong
The administrator has disabled public write access.

Mass balance info when GAIA coupled with TELEMAC-3D 4 years 10 months ago #35250

  • tbe
  • tbe's Avatar
  • OFFLINE
  • Fresh Boarder
  • Posts: 16
  • Thank you received: 5
Hi Qilong!

I'm just wondering, which solvers are you using for the sediment? (i.e. horizontal and vertical).

Thanks
Tom
The administrator has disabled public write access.

Mass balance info when GAIA coupled with TELEMAC-3D 4 years 10 months ago #35251

  • qilong
  • qilong's Avatar
  • OFFLINE
  • Expert Boarder
  • Posts: 340
  • Thank you received: 33
Hi Tom!

I use SOLVER FOR DIFFUSION OF SUSPENSION = 3, which is conjugate gradient on a normal equation, I believe.

I also use the following numerical options for the tracer in TELEMAC-3D:

SCHEME FOR ADVECTION OF TRACERS = 1
SCHEME FOR DIFFUSION OF TRACERS = 1

and no tidal flats in the domain.

Kind regards,
Qilong
The administrator has disabled public write access.

Mass balance info when GAIA coupled with TELEMAC-3D 4 years 10 months ago #35253

  • tbe
  • tbe's Avatar
  • OFFLINE
  • Fresh Boarder
  • Posts: 16
  • Thank you received: 5
Hello again Qilong

OK, the fact that there are no tidal flats in your model rules out any vertical solver instability issues.

In that case I am not sure what the problem could be at present without looking in detail at your setup.

If there is a problem then it is likely to be in the calculation of the mass balance rather than actual mass being created. If you can be a bit patient, this is something that will be investigated at the upcoming GAIA developer event at EDF in a couple of weeks. So if there is an issue I hope it can be fixed then.


Best Regards
Tom
The administrator has disabled public write access.

Mass balance info when GAIA coupled with TELEMAC-3D 4 years 10 months ago #35255

  • qilong
  • qilong's Avatar
  • OFFLINE
  • Expert Boarder
  • Posts: 340
  • Thank you received: 33
Hi Tom!

Thanks for your suggestion!

I'll check the mass information in detail and see if I could find anything there. Meanwhile, I will wait for the news from the next GAIA developers meeting.

Based on what I computed from the model result, the information (as shown in the first post) about the initial mass of suspension, the final mass of suspension, the accumulated advective flux given by TELEMAC-3D are fine. The total mass of bed given by GAIA is also correct.

If my model setup is needed for for further investigating, just let me know and I'm more than happy to share it.

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

Mass balance info when GAIA coupled with TELEMAC-3D 4 years 10 months ago #35256

  • tbe
  • tbe's Avatar
  • OFFLINE
  • Fresh Boarder
  • Posts: 16
  • Thank you received: 5
Hi Qilong

One other thought that just occurred to me.

I am wondering if it has something to do with the treatment of fluxes at the boundaries. (keyword TREATMENT OF FLUXES AT THE BOUNDARIES)

This needs to be set to 2 for each boundary to ensure the Dirichlet values at the boundary are treated correctly in the mass balance calculation. I believe that this is now the default value, but perhaps you changed it?

Also, how are you applying the concentration at the boundary? Have you modified the Fortran? Perhaps it would help if you could upload your steering file?

Tom
The administrator has disabled public write access.

Mass balance info when GAIA coupled with TELEMAC-3D 4 years 10 months ago #35259

  • qilong
  • qilong's Avatar
  • OFFLINE
  • Expert Boarder
  • Posts: 340
  • Thank you received: 33
Hi Tom,

The model steering files are attached.

I didn't notice the treatment of fluxes at the boundaries. But in the steering file, I commented this keyword, and I checked the .dico file, indeed it now uses option 2 as the default value.

I modified the fortran files, one for prescribing the concentration profiles (rouse profiles dependent on the velocity, assuming erosion equals to deposition) at boundaries (for the upstream boundary it's zero concentration), and the other one for initializing the model with a concentration field.

Perhaps prescribing concentration profiles causes trouble here?

Kind regards,
Qilong
The administrator has disabled public write access.

Mass balance info when GAIA coupled with TELEMAC-3D 4 years 10 months ago #35260

  • tbe
  • tbe's Avatar
  • OFFLINE
  • Fresh Boarder
  • Posts: 16
  • Thank you received: 5
Qilong

Thanks for the files. I can't immediately think what is the cause of the mismatch, but your example should help us locate it during the GAIA developments.

Best wishes

Tom
The administrator has disabled public write access.
The following user(s) said Thank You: qilong
Flat Threaded
Moderators: Pablo, pavans
Forum
System Modules
GAIA / SISYPHE
Mass balance info when GAIA coupled with TELEMAC-3D

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