Welcome, Guest
Username: Password: Remember me
Forgot your password? Forgot your username? Create an account
Forum
System Modules
GAIA / SISYPHE
bed load flux at outflow boundary
Flat Threaded

TOPIC: bed load flux at outflow boundary

bed load flux at outflow boundary 13 years 2 months ago #2326

  • konsonaut
  • konsonaut's Avatar
  • OFFLINE
  • openTELEMAC Guru
  • Posts: 413
  • Thank you received: 144
Hello,

we are dealing with problems regarding the bed load flux at the outflow. It seems, that Sisyphe calculates NO bed load transport trough the outflow boundary. At the boundary the flux is zero and the sediment coming from upstream is deposited at the nodes of the outflow boundary.
We already checked the boundary conditions file and the LIEBOR integer is set to KSORT (4). In the sense of the description in the manual, we have a free evolution but only deposition which is unphysical in this case.

Has anyone some suggestions how to enable the sediment flux through the boundary or do we make some mistake in the steering file, boundary conditions file,..?

Thanks,

Clemens & Gabi
The administrator has disabled public write access.

Re: bed load flux at outflow boundary 13 years 2 months ago #2343

  • konsonaut
  • konsonaut's Avatar
  • OFFLINE
  • openTELEMAC Guru
  • Posts: 413
  • Thank you received: 144
Hello again,

we made an error in the conlit.f subroutine. We had bedload inflow everywhere along the model boundary, so this explains in part the behaviour at the outflow boundary.
However, without bedload inflow into the model at the outflow boundary Sisyphe calculates in some nodes a lot of deposition till the prescribed water level is reached and alternately in adjacent nodes a lot of erosion. This happens also in the inner model, where Sisyphe calculates a lot of streamwise furrows with alternating deposition and erosion in adjacent nodes.

Has someone some suggestions how to avoid these furrows and the accumulation of sediment at the outflow boundary?

Clemens & Gabi
The administrator has disabled public write access.

Re: bed load flux at outflow boundary 13 years 2 months ago #2346

  • jmhervouet
  • jmhervouet's Avatar
Hello,

Are you doing internal coupling with Telemac-2D or do you just run Sisyphe as a stand-alone program. What you describe looks like what we sometimes get in the latter case.

Jean-Michel Hervouet
The administrator has disabled public write access.

Re: bed load flux at outflow boundary 13 years 2 months ago #2348

  • konsonaut
  • konsonaut's Avatar
  • OFFLINE
  • openTELEMAC Guru
  • Posts: 413
  • Thank you received: 144
Hello,

in the heat of the battle sorry for not mentioning some details.
We use WinXP + Gfortran + scalar Telemac-2D v6.1 internally coupled with Sisyphe.
In following the Telemac-2D and the Sisyphe steering file:

Telemac-2D:
/---------------------------------------------------------------------
/ EQUATIONS
/---------------------------------------------------------------------

FRICTION COEFFICIENT                  =30

LAW OF BOTTOM FRICTION                =3

TURBULENCE MODEL                      =1

/ NON-DIMENSIONAL DISPERSION COEFFICIENTS = 10.0;0.6   für Elder

VELOCITY DIFFUSIVITY                  =0.5

/ TURBULENCE MODEL FOR SOLID BOUNDARIES =2


/---------------------------------------------------------------------
/ EQUATIONS, BOUNDARY CONDITIONS
/---------------------------------------------------------------------

VELOCITY PROFILES            =4;1

PRESCRIBED FLOWRATES         =150;0

PRESCRIBED ELEVATIONS        =0;533.60

OPTION FOR LIQUID BOUNDARIES =1;1


/---------------------------------------------------------------------
/ EQUATIONS, INITIAL CONDITIONS
/---------------------------------------------------------------------

INITIAL ELEVATION  =533.60

INITIAL CONDITIONS ='CONSTANT ELEVATION'


/---------------------------------------------------------------------
/ INPUT-OUTPUT, FILES
/---------------------------------------------------------------------

FORTRAN FILE             ='princi_sis.f'

LIQUID BOUNDARIES FILE   ='Hydrograph_unsteady_to_HQ100_52h.txt'

GEOMETRY FILE            ='project_Mesh_VarFric20_30.slf'

COUPLING WITH            ='''INTER-SISYPHE'''

STEERING FILE            ='cas_tel'

SISYPHE STEERING FILE    ='cas_sis'

BOUNDARY CONDITIONS FILE ='cas.conlim'

PREVIOUS COMPUTATION FILE = 'res_steady'

RESULTS FILE             ='res_tel_HQ100_STZ'


/---------------------------------------------------------------------
/ INPUT-OUTPUT, GRAPHICS AND LISTING
/---------------------------------------------------------------------

LISTING PRINTOUT PERIOD         =300

VARIABLES FOR GRAPHIC PRINTOUTS ='U,V,B,H,S,F,US'

MASS-BALANCE                    =true

GRAPHIC PRINTOUT PERIOD         =1800

COMPUTATION CONTINUED           =YES

INITIAL TIME SET TO ZERO        =YES


/---------------------------------------------------------------------
/ INPUT-OUTPUT, INFORMATION
/---------------------------------------------------------------------

TITLE ='HQ100_STZ'


/---------------------------------------------------------------------
/ NUMERICAL PARAMETERS
/---------------------------------------------------------------------

CONTINUITY CORRECTION               =true

NUMBER OF TIME STEPS                =374400

FREE SURFACE GRADIENT COMPATIBILITY =0.9

TREATMENT OF THE LINEAR SYSTEM      =2

SUPG OPTION                         =1;1;2;2

TIME STEP                           =0.5


/---------------------------------------------------------------------
/ NUMERICAL PARAMETERS, SOLVER
/---------------------------------------------------------------------

SOLVER =1

/---------------------------------------------------------------------
/ PHYSICAL CONSTANTS
/---------------------------------------------------------------------

WATER DENSITY =1000


Sisyphe:

/---------------------------------------------------------------------
/ INPUT-OUTPUT, FILES
/---------------------------------------------------------------------

GEOMETRY FILE            ='project_Mesh_VarFric20_30.slf'

STEERING FILE            ='cas_sis'

BOUNDARY CONDITIONS FILE ='cas.conlim'

RESULTS FILE             ='res_SIS_HQ100_STZ'


/---------------------------------------------------------------------
/ INPUT-OUTPUT, GRAPHICS AND LISTING
/---------------------------------------------------------------------

LISTING PRINTOUT PERIOD         =300

GRAPHIC PRINTOUT PERIOD         =1800

VARIABLES FOR GRAPHIC PRINTOUTS =
'M,E,H,B,U,V,QSBL,1A1,1A2,1A3,1A4,1A5,1A6,1A7,QS1,QS2,QS3,QS5,QS7'

MASS-BALANCE                    =YES


/---------------------------------------------------------------------
/ INPUT-OUTPUT, INFORMATION
/---------------------------------------------------------------------

/----------------------------------------------
/  GENERAL AND OUTPUT
/----------------------------------------------
TITLE ='HQ100_STZ'


/---------------------------------------------------------------------
/ NONUNIFORM BED MATERIAL
/---------------------------------------------------------------------

NUMBER OF SIZE-CLASSES OF BED MATERIAL     =7

/---------------------------------------------------------------------
/ NUMERICAL PARAMETERS
/---------------------------------------------------------------------

SOLVER ACCURACY                               =1e-6

/ SKIN FRICTION CORRECTION                      =1

/ RATIO BETWEEN SKIN FRICTION AND MEAN DIAMETER =0

TIME STEP                                     =0.5

/ TETA                                          =0.5

ZERO                                          =1e-6


/---------------------------------------------------------------------
/ NUMERICAL PARAMETERS, SOLVER
/---------------------------------------------------------------------

/ FINITE VOLUMES = YES

OPTION FOR THE TREATMENT OF NON ERODABLE BEDS = 3

/---------------------------------------------------------------------
/ SEDIMENT TRANSPORT
/---------------------------------------------------------------------
BED LOAD = YES

SUSPENSION = NO

BED-LOAD TRANSPORT FORMULA                 = 1

HIDING FACTOR FORMULA                      = 1

CONSTANT ACTIVE LAYER THICKNESS = YES

/ CONSTANT ACTIVE LAYER THICKNESS = NO / doesn't work

ACTIVE LAYER THICKNESS = 0.10

FRICTION ANGLE OF THE SEDIMENT =35

SEDIMENT DENSITY               =2650.

MEAN DIAMETER OF THE SEDIMENT  =0.0004;0.008;0.01;0.017;0.028;0.05;0.07


/---------------------------------------------------------------------
/ SEDIMENTOLOGY
/---------------------------------------------------------------------

SECONDARY CURRENTS =true

We are quite new in modelling sediment transport with Sisyphe, so maybe we are doing some errors in setting up the whole model?

Clemens
The administrator has disabled public write access.

Re: bed load flux at outflow boundary 13 years 2 months ago #2351

  • jmhervouet
  • jmhervouet's Avatar
Hello,

Unfortunately all our sediment specialists here are on the leave for the RCEM congress in China.

If you are beginners with Sisyphe I would suggest that you start with only one class of sediments, and do not add the secondary currents. Then if you only have bedload, you are not obliged to call Sisyphe at every time step, you can use the COUPLING PERIOD FOR SISYPHE (in your case at least 10 because your time step is very small).

minor details :

* in coupling the keyword for the time-step in Sisyphe is not used, the time step is imposed by telemac-2D
* for the VARIABLES FOR GRAPHIC PRINTOUTS you can use a jocker, e.g.
1A* instead of 1A1,1A2,1A3,1A4,1A5,1A6,1A7

then sediment specialists will probably have other advice to give...

with best regards,

Jean-Michel Hervouet
The administrator has disabled public write access.

Re: bed load flux at outflow boundary 13 years 1 month ago #2650

  • clebouteiller
  • clebouteiller's Avatar
Hello,

I don't know if this helps, but I came across some problems myself while defining Sisyphe boundary conditions for bedload transport, and I think that they are actually two types of conditions:
- LIEBOR which refers to the bed elevation (free or imposed), and which can be set in the .conlim file,
- LIQBOR which refers to the bedload discharge (free or imposed), which can only be set in conlit.f subroutine.
I found that the manual was a bit vague about it but there is the following definition in subroutine bedload_main.f:
LIEBOR = "type of boundary condition for bed evolution"
LIQBOR = "type of boundary condition for qs"

So perhaps when LIEBOR=4 with free bed evolution, LIQBOR could be set by default to a fixed zero outflow discharge?

I am not sure of this interpretation and you might have already found a solution to your problem, but let me know if it makes sense and if it corresponds to your experience.

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

Re: bed load flux at outflow boundary 13 years 1 month ago #2653

  • konsonaut
  • konsonaut's Avatar
  • OFFLINE
  • openTELEMAC Guru
  • Posts: 413
  • Thank you received: 144
Hello Caroline,

thank you for your comments!
We already solved the problems.

1. We had problems with an imposed stage discharge curve where we defined a falling water level by an increasing discharge at the outlet, which doesn't work. It caused strong flow oscillations (inflow/outflow) at the outlet. Solution was to prescribe the water level in the liquid boundary file which works well and which is acceptable if the modelled reach is not too long.

2. As described in the previous posts, Sisyphe calculated a lot of streamwise furrows with alternating deposition and erosion in adjacent noded in the inner model and at the outlet.
We did some parameter tests and the simple solution was to turn off the slope effect correction, which by default is turned on! Now we have very good results. A problem with slope correction is, where do you apply the correction: on the acting side or on the reacting side and then you must be careful with setting the limiters. However there exist a lot of slope correction formulas which maybe are more suitable for river modelling.

If you want a free bed evolution e.g. at the outlet, then you set LIEBOR = KSORT(4) and for an imposed bed evolution LIEBOR = KENT(5) in the boundary file. For an imposed bed load discharge at a boundary LIEBOR = KSORT(4) in the boundary file or conlit.f and LIQBOR = KENT in the conlit.f for that boundary. With this configuration we don't have a zero outflow discharge.

Best regards,
Clemens
The administrator has disabled public write access.
Flat Threaded
Moderators: Pablo, pavans
Forum
System Modules
GAIA / SISYPHE
bed load flux at outflow boundary

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