Welcome, Guest
Username: Password: Remember me

TOPIC: mesh density instability calculation

mesh density instability calculation 10 years 5 days ago #14887

  • julesleguern
  • julesleguern's Avatar
Is that the fact of using a mesh with several density of nodes may create instabilities in the model with oscillations of the free surface?

I have a mesh with a density of 10 meters and an area of 1 meter density. When I launch the simulation in parallel with 8 procs, I see some instability near the transition zone between one density and an other. This area is also a boundary of two procs. This instability create an oscillation of the free surface that propogate toward the upstream of my model.

Someone have an idea?

Thanks

Jules

my .CAS file :

/ INFORMATIONS DU CALCUL /
/
/
/ /
FICHIER DE GEOMETRIE : bottom.slf
FICHIER DES CONDITIONS AUX LIMITES : condlim.cli
FICHIER DES RESULTATS : res_casier2.slf
/FICHIER DES SECTIONS DE CONTROLE : sections_in.txt
/FICHIER DE SORTIE DES SECTIONS DE CONTROLE : sections_out.txt
/FICHIER FORTRAN : fortran.f
PROCESSEURS PARALLELES : 8
/
/
/ INFORMATIONS GENERALES /
/
/
/
TITRE : 'casier_test2'
/
PAS DE TEMPS : 1
DUREE DU CALCUL : 47000
/DATE DE L'ORIGINE DES TEMPS :
/HEURE DE L'ORIGINE DES TEMPS :
/
VARIABLES POUR LES SORTIES GRAPHIQUES : 'H,U,V,L,S'
PERIODE POUR LES SORTIES GRAPHIQUES : 100
/NUMERO DU PREMIER PAS DE TEMPS POUR LES SORTIES GRAPHIQUES :
/
/VARIABLES A IMPRIMER :
PERIODE POUR LES SORTIES LISTING : 50
/NUMERO DU PREMIER PAS DE TEMPS POUR LES SORTIES LISTING :
BILAN DE MASSE : OUI
INFORMATIONS SUR LE SOLVEUR : OUI
/DEBUGGER : 1
/INFORMATIONS SUR LE MODELE K-EPSILON : OUI
/
/
/ CONDITIONS INITIALES /
/
/
/
CONDITIONS INITIALES :'COTE CONSTANTE'
COTE INITIALE : 0
/SUITE DE CALCUL : OUI
/REMISE A ZERO DU TEMPS : OUI
/FICHIER DU CALCUL PRECEDENT : res_test18.slf
/ENREGISTREMENT POUR SUITE DE CALCUL :
/
/
/ CONDITIONS AUX LIMITES /
/
/
/
COTES IMPOSEES : 0.1 ; -0.1 ; -0.1
DEBITS IMPOSES : 100 ; 0 ; 0
/OPTION FOR LIQUID BOUNDARIES : 1 ; 1 ; 1 ; 1
FICHIER DES FRONTIERES LIQUIDES : hydro.txt
FICHIER DES COURBES DE TARAGE : tarage.txt
COURBES DE TARAGE : 0 ; 1 ; 1
PROFILS DE VITESSE : 4 ; 4 ; 4
CALCUL COMPATIBLE DES FLUX : OUI
/
/
/ PARAMETRES PHYSIQUES /
/
/
/
LOI DE FROTTEMENT SUR LE FOND : 3
COEFFICIENT DE FROTTEMENT : 30
/NOMBRE MAXIMUM DE DOMAINES DE FROTTEMENT : 10
MODELE DE TURBULENCE : 1
COEFFICIENT DE DIFFUSION DES VITESSES : 1.0E-6
/
/
/ PARAMETRES NUMERIQUES /
/
/
/EQUATIONS : SAINT-VENANT EF
DISCRETISATIONS EN ESPACE : 11 ; 11
TRAITEMENT DU SYSTEME LINEAIRE : 2
FORME DE LA CONVECTION : 1 ; 5
SOLVEUR : 1
PRECISION DU SOLVEUR : 1.E-6
MAXIMUM D'ITERATIONS POUR LE SOLVEUR : 100
BANCS DECOUVRANTS : OUI
OPTION DE TRAITEMENT DES BANCS DECOUVRANTS : 1
TRAITEMENT DES HAUTEURS NEGATIVES : 2
CORRECTION DE CONTINUITE : OUI
OPTION DE SUPG : 0 ; 0 ; 0 ; 0
MASS-LUMPING SUR H : 1
COMPATIBILITE DU GRADIENT DE SURFACE LIBRE : 0.9
/
&ETA
The administrator has disabled public write access.

mesh density instability calculation 10 years 5 days ago #14890

  • jmhervouet
  • jmhervouet's Avatar
Hello,

What changes when the mesh size changes is the Courant number, and this is a hint. You could have inf-sup oscillations, in this case the free surface gradient compatibility should have an influence, try 0 instead of 0.9 just to see.
Another thing is that you have no diffusion of velocities, try 1. instead of 1.E-6.
I do not believe in a parallelism problem, you can check that you get the same problem in scalar mode.

If nothing works also try :

TRAITEMENT DES HAUTEURS NEGATIVES : 1
OPTION DE SUPG : 0 ; 2 ; 0 ; 0

if you do not mind having slightly negative depths in dry zones, it does not spoil mass conservation and is faster.

With best regards,

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

mesh density instability calculation 10 years 4 days ago #14897

  • julesleguern
  • julesleguern's Avatar
Hello Jean-Michel,

Thanks for your response. I will try these options and give a back on the post.
The administrator has disabled public write access.

mesh density instability calculation 9 years 11 months ago #14940

  • julesleguern
  • julesleguern's Avatar
Hi Jean-Michel,

I tried all your suggestions but I always see some local ocsillations of the free surface and the Courant number. I have a minimum depth of -1.39m in dry zones with treatment of negative depths=1.

If I put the SUPG option : 0;2;0;0 I had to put the type of advection : 1;2 ?

I will try a simulation with a variable time step and fix the Courant number to 1 (or inf to 1). Do you think it's a good idea ?

Any suggestions?

Best regards.

Jules
Attachments:
The administrator has disabled public write access.

mesh density instability calculation 9 years 11 months ago #14948

  • jmhervouet
  • jmhervouet's Avatar
Hello,

> If I put the SUPG option : 0;2;0;0 I had to put the type of advection : 1;2 ?

No, it is always 1;5 in fact we just imagine that the term div(hu) is split into h div(u) + u.grad(h), the second term is like an advection of h and we add the SUPG term related to this advection term, in some sense we partly upwind the term div(hu). This cannot work with the treatment that gives no negative depths.

Another more expensive possibility is to play with quasi-bubble elements for the velocities :

SPACE DISCRETIZATIONS : 12;11

Regards,

JMH
The administrator has disabled public write access.
Moderators: pham

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