TELEMAC-MASCARET Forum: SUPG + wetting/drying errors (2/2)

Welcome, Guest

TOPIC: SUPG + wetting/drying errors

SUPG + wetting/drying errors 7 years 2 months ago #27720

SDAC

Apologies for bringing this up again, but I want to double check I've got this right as my model keeps collapsing into "exceeding maximum iterations." My time step is already very low, as is my threshold for viscosity correction on tidal flats. I think it may be from how I'm implementing the advection scheme.

Looking at the quote below:
riadh wrote:

Hello

SUPG does not work with tidal flats. You should rather use scheme 13 or 14 (LEO Postma or MURD)
Or you can try the newest option with the implicit option of the PSI (or N scheme). To activate that:
1- use advection scheme =3 (N) or 4 (psi , recommanded)
2- SCHEME OPTION FOR ADVECTION OF VELOCITIES = 4
3- NUMBER OF CORRECTIONS OF DISTRIBUTIVE SCHEMES = 1 (defautl, you can increase it if necessary)
4- NUMBER OF SUB-STEPS OF DISTRIBUTIVE SCHEMES = 1 (default, you can increase it if necessary)

I hope that this helps

Riadh ATA

There is no keyword for ADVECTION SCHEME in my dictionary for Telemac 3D, so I assumed it was SCHEME FOR ADVECTION OF K-EPSILON (for example) and set that to 5 (PSI). Trying without the above immediately makes the model fail. Using the above keyword it seemed fine but fails with the error eventually. Is this method correct? I couldn't find anymore keywords for advection to experiment with.

I've attached my steering file if anyone is able to give it a quick eye over, as I'm trying to whittle down possible causes.

Much appreciation.

The administrator has disabled public write access.

SUPG + wetting/drying errors 7 years 2 months ago #27724

riadh	Hello What do you mean by "Using the above keyword it seemed fine but fails with the error eventually. Is this method correct". Which error did you get? Please attach your full model in order to check it up. kind regards Riadh
	The administrator has disabled public write access.

SUPG + wetting/drying errors 7 years 2 months ago #27746

SDAC

Hello,

Apologies. I mean, using SCHEME FOR ADVECTION OF VELOCITIES/K-EPSILON with SCHEME OPTION FOR ADVECTION OF VELOCITIES/K-EPSILON, I don't get the "exceeding maximum iterations error" until much later on in the run. This only happens at a few steps however - for example, see the sortie file for the final few steps below.

I've also noticed that if I change the advection schemes to MoC the run immediately crashes. It only works when one of the advection scheme options is MoC and the other PSI.It doesn't matter which way round it is, be it for the advection of K-EPSILON or velocities or vice versa.

Many thanks!

MASS BALANCE

WATER
MASS AT THE PREVIOUS TIME STEP : 235.9275
MASS AT THE PRESENT TIME STEP : 235.9292
MASS LEAVING THE DOMAIN DURING THIS TIME STEP : -0.1660764E-02
ERROR ON THE MASS DURING THIS TIME STEP : 0.1394906E-12
FLUX BOUNDARY 1: -0.6089236 M3/S ( >0 : ENTERING <0 : EXITING )
FLUX BOUNDARY 2: 0.7750000 M3/S ( >0 : ENTERING <0 : EXITING )
GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS: 60 RELATIVE PRECISION: 0.6815615E-03

================================================================================
ITERATION 22990 TIME 0 D 0 H 3 MN 49.9000 S ( 229.9000 S)
================================================================================
ELEVATION AT THE BOUNDARY READ IN THE ASCII FILE

ADVECTION STEP
JACOBI: 1 ITERATIONS, ABSOLUTE PRECISION = 1.4570756982115318E-006
MURD3D_LIPS: 1 ITERATIONS
JACOBI: 1 ITERATIONS, ABSOLUTE PRECISION = 6.5586666136687913E-008
MURD3D_LIPS: 1 ITERATIONS
JACOBI: 2 ITERATIONS, ABSOLUTE PRECISION = 5.4761220149245184E-009
MURD3D_LIPS: 1 ITERATIONS
DIFFUSION OF WN
GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS: 60 RELATIVE PRECISION: 0.3752594E-02

PROPAGATION AND DIFFUSION WITH WAVE EQUATION
GRACJG (BIEF) : 2 ITERATIONS, ABSOLUTE PRECISION: 0.1693246E-06
POSITIVE DEPTHS OBTAINED IN 3 ITERATIONS

DYNAMIC PRESSURE STAGE
GRACJG (BIEF) : 29 ITERATIONS, ABSOLUTE PRECISION: 0.9279437E-04

VELOCITY PROJECTION STEP

ADVECTION-DIFFUSION OF K-EPSILON OR OMEGA STEP
JACOBI: 1 ITERATIONS, RELATIVE PRECISION = 1.0730035273803215E-007
MURD3D_LIPS: 1 ITERATIONS
DIFFUSION OF AKN
GRACJG (BIEF) : 107 ITERATIONS, RELATIVE PRECISION: 0.8126581E-06
JACOBI: 1 ITERATIONS, RELATIVE PRECISION = 3.0072953468832222E-008
MURD3D_LIPS: 1 ITERATIONS
DIFFUSION OF EPN
GRACJG (BIEF) : 151 ITERATIONS, RELATIVE PRECISION: 0.9679319E-06

MASS BALANCE

WATER
MASS AT THE PREVIOUS TIME STEP : 235.9442
MASS AT THE PRESENT TIME STEP : 235.9458
MASS LEAVING THE DOMAIN DURING THIS TIME STEP : -0.1664393E-02
ERROR ON THE MASS DURING THIS TIME STEP : -0.1476355E-11
FLUX BOUNDARY 1: -0.6085607 M3/S ( >0 : ENTERING <0 : EXITING )
FLUX BOUNDARY 2: 0.7750000 M3/S ( >0 : ENTERING <0 : EXITING )
GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS: 60 RELATIVE PRECISION: 0.1270915E-03
GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS: 60 RELATIVE PRECISION: 0.1457160E-02
GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS: 60 RELATIVE PRECISION: 0.3849069E-02
GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS: 60 RELATIVE PRECISION: 0.6766516E-03
GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS: 60 RELATIVE PRECISION: 0.1641764E-03
GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS: 60 RELATIVE PRECISION: 0.7750892E-04
GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS: 60 RELATIVE PRECISION: 0.7907776E-04
GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS: 60 RELATIVE PRECISION: 0.6615756E-03
GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS: 60 RELATIVE PRECISION: 0.1051835E-03

================================================================================
ITERATION 23000 TIME 0 D 0 H 3 MN 50.0000 S ( 230.0000 S)
================================================================================
ELEVATION AT THE BOUNDARY READ IN THE ASCII FILE

ADVECTION STEP
JACOBI: 1 ITERATIONS, ABSOLUTE PRECISION = 1.4932488775125832E-006
MURD3D_LIPS: 1 ITERATIONS
JACOBI: 1 ITERATIONS, ABSOLUTE PRECISION = 4.4083124923655096E-008
MURD3D_LIPS: 1 ITERATIONS
JACOBI: 2 ITERATIONS, ABSOLUTE PRECISION = 6.6388320066141194E-010
MURD3D_LIPS: 1 ITERATIONS
DIFFUSION OF WN
GRACJG (BIEF) : EXCEEDING MAXIMUM ITERATIONS: 60 RELATIVE PRECISION: 0.5385157E-04

PROPAGATION AND DIFFUSION WITH WAVE EQUATION
GRACJG (BIEF) : 2 ITERATIONS, ABSOLUTE PRECISION: 0.9426263E-07
POSITIVE DEPTHS OBTAINED IN 3 ITERATIONS

DYNAMIC PRESSURE STAGE
GRACJG (BIEF) : 30 ITERATIONS, ABSOLUTE PRECISION: 0.8920692E-04

VELOCITY PROJECTION STEP

ADVECTION-DIFFUSION OF K-EPSILON OR OMEGA STEP
JACOBI: 1 ITERATIONS, RELATIVE PRECISION = 1.3820272284006365E-007
MURD3D_LIPS: 1 ITERATIONS
DIFFUSION OF AKN
GRACJG (BIEF) : 66 ITERATIONS, RELATIVE PRECISION: 0.7198646E-06
JACOBI: 1 ITERATIONS, RELATIVE PRECISION = 1.2326446086327471E-007
MURD3D_LIPS: 1 ITERATIONS
DIFFUSION OF EPN
GRACJG (BIEF) : 53 ITERATIONS, RELATIVE PRECISION: 0.9498905E-06

MASS BALANCE

WATER
MASS AT THE PREVIOUS TIME STEP : 235.9608
MASS AT THE PRESENT TIME STEP : 235.9625
MASS LEAVING THE DOMAIN DURING THIS TIME STEP : -0.1667420E-02
ERROR ON THE MASS DURING THIS TIME STEP : -0.1459260E-11
FLUX BOUNDARY 1: -0.6082580 M3/S ( >0 : ENTERING <0 : EXITING )
FLUX BOUNDARY 2: 0.7750000 M3/S ( >0 : ENTERING <0 : EXITING )

FINAL MASS BALANCE
T = 230.0000

--- WATER ---
INITIAL MASS : 278.4407
FINAL MASS : 235.9625
MASS LEAVING THE DOMAIN (OR SOURCE) : 42.47818
MASS LOSS : -0.1499245E-11
CALL OF P_EXIT IN ITS VOID VERSION

CORRECT END OF RUN

The administrator has disabled public write access.

SUPG + wetting/drying errors 7 years 2 months ago #27755

SDAC	Please see model files attached. Kind regards. Attachments: SDAC.cas (3KB) Drag_model.f (14KB) Mesh01.cli (115KB) Mesh01.slf (1,452KB)
	The administrator has disabled public write access.