open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi everyone,

I simulated an overtopping embankment flow with telemac 2d. It seems to me the boundary inflow shape is not realistic and this happened during overtopping and downstream boundary as well. Please any comment or suggestion. I would like to couple with sisyphe to include the erosion of the embakment with the rest of the floor should remains unchanged but in fact I dont have any idea about the required parameters in sisyphe cas file or how to prepare it, any help or example that could help us.

Thanks
AMANJ

and also the inflow rate is 10 lit/sec is seems to me the wave or rate or coming to overtop is too big.

Hello,

seeing your pictures I think you have more than one problem.
It is a good idea to start only with hydrodynamics and see how Telemac performs and what it needs to simulate such cases. Is picture one the initial condition? I would set the initial water surface level higher than the dam and lower it to a defined outlet water surface value via a ramp function in the liquid boundaries file. Furthermore the time step may be too large, and Telemac provides several stabilizing methods, which you can find in the manual.

Since we did some sediment transport dam-break simulations and for maybe an exponential self-learning curve, you will find enclosed the Telemac files for the simulation of a dam-break case.


Best regards,
Clemens

Thank you dear Clemens,

I will attach my steering files until you have more idea about my situation.
/

---

/ TELEMAC2D Version v6p3 22-Sep-2014
/ nom inconnu
/

---

/

---

/ EQUATIONS
/

---

FRICTION COEFFICIENT =40

LAW OF BOTTOM FRICTION =3

VELOCITY DIFFUSIVITY =0.6


/

---

/ EQUATIONS, BOUNDARY CONDITIONS
/

---

PRESCRIBED FLOWRATES =0.0;0.01

PRESCRIBED ELEVATIONS =0.01;0.0


/

---

/ EQUATIONS, INITIAL CONDITIONS
/

---

INITIAL ELEVATION =0.25

INITIAL CONDITIONS ='CONSTANT ELEVATION'


/

---

/ INPUT-OUTPUT, FILES
/

---

SISYPHE STEERING FILE ='sisyphe.cas'

STEERING FILE ='Test1.cas'

BOUNDARY CONDITIONS FILE ='BC.cli'

RESULTS FILE ='result.slf'

GEOMETRY FILE ='geometry.slf'

COUPLING WITH ='SISYPHE'


/

---

/ INPUT-OUTPUT, GRAPHICS AND LISTING
/

---

LISTING PRINTOUT PERIOD =250

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

MASS-BALANCE =true

GRAPHIC PRINTOUT PERIOD =100


/

---

/ NUMERICAL PARAMETERS
/

---

TIME STEP =0.01

NUMBER OF TIME STEPS =25000

DISCRETIZATIONS IN SPACE =11;11

TREATMENT OF THE LINEAR SYSTEM =2

Dear Clemens,

Can you please guide me to change your cas file from parallel into single processor because my setting is based on single processor and I tried to run your project But I couldnt.

Thanks for your help in advance
AMANJ

Hello,

commenting the lines in both the Telemac-2D and Sisyphe steering files should do the work:
/ PARALLEL PROCESSORS=2

The Telemac-2D Manual is a good resource!

Regarding you TEL2D steering file:
it seems to be a laboratory case? If so, you have maybe small cell sizes and in combination with expected flow velocities, I would lower the eddy viscosity to a values around 1.E-4 - 1.E-6. Perhaps the inflow wiggles disappear. You should adapt the time step to roughly expected Courant numbers Cr, in order that Cr around 1 or lower. Furthermore I would set the Solver=1.
The roughness seems also very high for a laboratory case, but I don't know the configuration.
Generally I would try to capture the configuration of the test case, especially the inflow and outflow conditions.

Hope this helps and
best regards,
Clemens

Hello

I'd like to add that you start with a constant free surface of 0.25 and you prescribe an elevation of 0.01
This will generate highly transient flow in the outlet area which could explain your wiggles.
So try to verify if the instability is related to the outlet or in the inlet? I'm wondering if you are switching both conditions (read the upper part of the listing, before the first time step).
Furthermore, it is recommended to let stabilizing the hydrodynamics in order to skip the highly transient regime and then couple with Sisyphe

with my best regards

Riadh ATA

Hello dear friends,

I coupled Sisyphe with T2D and I cant see the result of sediment transport and strange change happen in my surface profile. Here I posted the results before and after couple and both sis and t2d steering files. Pls any recommendation, my primary objective is to simulate erosion of overtopping an embankment.

Many Thanks,
AMANJ

Hello,

What happens looks like you have an enormous friction. So probably BED ROUGHNESS PREDICTION =true is to be set to false for some reason to investigate (possibly different friction laws in the two codes are not possible). However I find many strange things in your parameter files; The dimensions are very small, the time-step is very small. I would not start with Finite Volumes and I would try a more robust set of parameters, take for example (to be put at the end of your steering file) :

/
EQUATIONS ='SAINT-VENANT EF'
TIDAL FLATS : YES OPTION FOR THE TREATMENT OF TIDAL FLATS : 1
TREATMENT OF NEGATIVE DEPTHS : 2 CONTINUITY CORRECTION : YES
H CLIPPING : NO
TYPE OF ADVECTION : 1;5
SUPG OPTION : 0;0
FREE SURFACE GRADIENT COMPATIBILITY : 0.9
MATRIX STORAGE : 3 MATRIX-VECTOR PRODUCT : 1
DISCRETIZATIONS IN SPACE : 11;11
TREATMENT OF THE LINEAR SYSTEM:2 SOLVER:2 PRECONDITIONING:11
SOLVER ACCURACY = 1.E-5
IMPLICITATION FOR DEPTH : 1.
IMPLICITATION FOR VELOCITY : 1.
MASS-LUMPING ON H : 1.

It should allow larger time steps and a much faster execution time.

I hope it helps,

Jean-Michel Hervouet

Hello dear Jean-Michel Hervouet,

I followed your correction and it helped a lot. Too many thanks. I would like to ask you about three points. First, for me is not clear about prescribed flow rate and elevation, I know the U/S is prescribed flow rate but why we need two values? and where is prescribed elevation? what is difference between constant elevation and varied one?. Second, I wonder to understand and know more about initial value, when I say initial depth is equal to some value is it mean that every where is the same on my domain? how to change for some specific region only, for example, i just want to have initial value on my U/S of my Embankment and downstream value is zero or different value, how to set this?. When I coupled Sisyphe with Telemac and calculate erosion, I wanna know how to visualize the erosion process with surface flow? how to define the rigid bed to separate from erodible bed?

I know I opened many topics, but I need your helps please!

Kind Regards,
AMANJ