open TELEMAC-MASCARET The mathematically superior suite of solvers

Hello,

I am working on the comparison of two models (initial state and project state) in order to assess the impact of new buildings on water heights and velocity. The initial state model is stable and validated. I am using the same mesh/initial conditions/boundary conditions for both models, the only difference is the bathymetry at the location of the projects. The boundary condition for both models is a tidal wave.
Unfortunately, these modifications lead to significant variations within non-modified areas located very far from the projects.(attached files)
I modified some parameters and run calculations but it seems that none of that improved the results.

Here is the steering file:
CONSTANTES PHYSIQUES
/

---

SEUIL POUR LES PROFONDEURS NEGATIVES =0.01


/

---

/ ENTREES-SORTIES, FICHIERS
/

---

FICHIER DU CALCUL PRECEDENT ='condInit.ser'

FICHIER DES RESULTATS ='res_projets.res'

FICHIER DES FRONTIERES LIQUIDES ='cas.liq'

FICHIER DES PARAMETRES ='cas'

FICHIER DE GEOMETRIE ='maillage_projets_2.slf'

FICHIER DES CONDITIONS AUX LIMITES ='cas.conlim'


/

---

/ ENTREES-SORTIES, GENERALITES
/

---

SUITE DE CALCUL =true


/

---

/ ENTREES-SORTIES, GRAPHIQUES ET LISTING
/

---

PERIODE POUR LES SORTIES LISTING =900

PERIODE DE SORTIE LISTING =900

VARIABLES POUR LES SORTIES GRAPHIQUES=
B,U,V,H,C,S,F,L,MAXZ,MAXV,TMXZ,TMXV

PERIODE POUR LES SORTIES GRAPHIQUES =900

BILAN DE MASSE =true


/

---

/ EQUATIONS
/

---

COEFFICIENT DE DIFFUSION DES VITESSES =1

COEFFICIENT DE FROTTEMENT =30

LOI DE FROTTEMENT SUR LE FOND =3


/

---

/ EQUATIONS, CONDITIONS LIMITES
/

---

OPTION POUR LES FRONTIERES LIQUIDES =2;2;2;2;2;2;2


/

---

/ PARAMETRES NUMERIQUES
/

---

PAS DE TEMPS =1 (I also tried 0.5s)

CORRECTION DE CONTINUITE =true

TRAITEMENT DU SYSTEME LINEAIRE =2 (I also tried 1)

COMPATIBILITE DU GRADIENT DE SURFACE LIBRE =0.

PROCESSEURS PARALLELES =24

CALCUL COMPATIBLE DES FLUX =true

NOMBRE DE PAS DE TEMPS =126000


/

---

/ PARAMETRES NUMERIQUES, SOLVEUR
/

---

MAXIMUM D'ITERATIONS POUR LE SOLVEUR =300


/

---

/ PARAMETRES NUMERIQUES, VITESSE-CELERITE-HAUTEUR
/

---

MASS-LUMPING SUR LA VITESSE =0.9

MASS-LUMPING SUR H =0.9

Thank you in advance!

Hello,

Maybe you need to redo your mesh with your new buildings (how do you take them into account?).

With only one figure, we may not see everything you have noticed. What kind of boundary is it at the bottom of the figure?

Anyway, looking only at the difference of water depth at one specific time may be hazardous if you have a little shift in time for the wave propagation e.g.
Maybe the differences between the two models are not so big (rather look at the maximum of peaks e.g.).

Hope this helps,

Chi-Tuan

Hello,

Thank you for the quick reply!

The mesh takes into account the new buildings. I used hardlines to delineate the buildings and a very low Strickler coefficient (because I don't want to create holes in the mesh). Actually, this is a model that was built many years ago and I am using it for this project.

The boundary at the bottom of the figure is an imposed water level (a tidal wave).

I also checked other time steps and the differences range from 0.02 to more than 1.5 m (the peak is around 74700). The flooded area is bigger when running the project state, knowing that the flow is not supposed to be influenced by the projects in these areas...

Thank you

Hello,

To my opinion, even light modifications in your TELEMAC-2D model can have small effects in the results, e.g. a small influence for the wave propagation in particular it can propagate ligthly slower or faster and that can explain your differences for instantaneous difference of water depth. That is why I suggested you to compare the maximum of peaks but separately for your 2 computations, not using the pure difference of results at the same time. In that case, comparing instantaneous water depth or velocity is not relevant for me.

You seem to also have changed friction coefficient for an area and that also may change the results. I do not know what magnitude of water depth you have.

Do you compare the results using the same release of TELEMAC-2D? Which one do you use?

Hope this helps,

Chi-Tuan

Hello,

Thank you for your reply.

I changed the bathymetry and the friction coefficient in the areas where the projects are located.

I compared the maximum of peaks, for example the maximum water heights and the differences can go to up 20cm. Also, the extent of the flooded areas is not the same in some zones located far from the projects.

This is problematic because I have to make the flooding maps (maximum water heights and velocity) for both cases and also the map of the differences to show the impact of the projects. These maps should cover the full extent of the model and not only the projects and their surroundings.

I am using the v7p1r0 version for both calculations and Fudaa-Prepro 1.4.2 for pre/post-processing.

Thank you