open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi OpenTelemac-Mascaret users,

I would like to model the presence (and the failure from one time step) of a vertical parapet at the crown of a levee.
I discretized my levee as recommended (5 points at least and chineese hat shape of my crest) to prevent numerical pumping. Annyway, I choose to not represent a 50cm parapet.

Now I want to consider the effect of this parapet (zero overflowing) until one time step and after that time step to consider the structural (and functional) failure of the parapet.

Has this kind of programming been done by one of you?

My idea is to use maskob:

1- To declare a polygon (including the elements I want to remove)

2- with a condition on the time step or time

IF (AT.LT."failure time") THEN


ELSEIF (AT.GE."failure time") THEN
MASKEL = 1.D0 FOR'NORMAL' ELEMENTS
END

3 - Condition inpoly -> MASKEL = 0.D0

DO I=1,BIEF_NBPTS(11,MESH)
IF(INPOLY(X(I),Y(I),XSOM,YSOM,NSOM)) THEN
MASKEL(I)= 0.D0
ELSE
MASKEL(I)= 1.D0
ENDIF
ENDDO

Can it work?

Cheers,
Alexis

Hello Alexis,

If I understand well, your idea is exactly the same as a breach of dyke. Your Parapet can be simply considered as a rise in the bathymetry/topography of the local area of the dyke (which should be refined with more than with 5 points).
At time=time_failure, the breaching process can be triggered.

Am I missing something?

kind regards

Riadh

Hi Riadh,

I see what you mean. I already use the breach module for my levee to simulate the breaching of the earthfill body. The point is that I didn't include the vertical wall in my mesh allowing overflow . If i change the altimetry of the top point I will have steep bathymetry gradient.

My idea was to use maskob to mimic insubmersible vertical wall. Conceptually, the idea is to isolate a part of the mesh (the leveed zone) to keep it dry.

Alexis

Hello Alexis

I will prefer instead to:
1- create the real island and and then hot-restart is you need to include it (or vice versa). the hot start is simply created by projection of the result file on a new serafin file created with BK
2- refine the mesh and raise the bathymetry, this should work.

The masking technique is not mass conservative and could not work with some advection schemes.

If you want to try, yes it can work like you have written. Just some precautions:
- do not forget to add the keyword ELEMENTS MASKED BY USER =yes
- the do loop should be done over NPOIN : DO I=1,NPOIN
- in this way, elements crossing the limit of your area will be masked. If you want to include only elements with all the 3 nodes in the masked area, you have to use :
do IEL=1,NELEM
I1=IKLE(IEL,1)
I2=IKLE(IEL,2)
I3=IKLE(IEL,3)
if((inpoly...of I1).and.(inpoly...of I2).and.(inpoly...of I3))THEN

I hope this helps

with my best regards

Riadh