Page 4 of 13
Re: turbine modelling
Posted: Thu Jun 21, 2012 2:56 pm
by Yvan Fournier
Hello,
Selection criteria are based on mesh "groups", and not face numbers. To check groups available in your mesh, use the mesh checking funtions of the code ("code_saturne check mesh -m <file>" in text mode), and check preprocessor log and visualize preprocessing/mesh chaecking output under ParaVis/ParaView for example.
To build groups in SALOME, there are many options, but the simplest is to build groups in GEOM, then build face groups based on the geometry under SALOME.
The Code_Saturne user guide has a chapter explaining everything about selection criteria, and how it works (colors/groups) depending on the mesh format used. Search the pdf for "selection".
Regards,
Yvan
Re: turbine modelling
Posted: Thu Jun 21, 2012 4:15 pm
by stage75
So if I understand, on Salome each color corresponds to a number of this color ?
Re: turbine modelling
Posted: Thu Jun 21, 2012 4:45 pm
by Yvan Fournier
No, avoid color numbers with SALOME, and use group names. In Code_Saturne's getfbr(), you may use group names instead of color numbers. Please read the corresponding chapter in the user manual.
Re: turbine modelling
Posted: Fri Jun 22, 2012 8:20 am
by stage75
Thank you Yvan,
I started to read the chapter
Re: turbine modelling
Posted: Fri Jun 22, 2012 9:56 am
by stage75
I think that my fatal error comes from my boundary conditions definitions :
So I remember that I should model a rotating iturbine in a basin of water. therefore my overall geometry is constituted by a turbine and a basin and then the water level in the basin (figure attached).
So by boundary conditions are:
THe differents faces of the turbine are the WALLS (name group: Turbine-faces) ,definied in "usclim.f90" like:
Code: Select all
[code]call getfbr('Turbine-faces', nlelt, lstelt)
do ilelt = 1, nlelt
ifac = lstelt(ilelt)
! Wall: zero flow (zero flux for pressure)
! friction for velocities (+ turbulent variables)
! zero flux for scalars
do iphas = 1, nphas
itypfb(ifac,iphas) = iparoi
enddo
enddo
[/code]
THe walls of the basin : WALLS (name group: basin faces)
Code: Select all
call getfbr('basin faces', nlelt, lstelt)
do ilelt = 1, nlelt
ifac = lstelt(ilelt)
! Wall: zero flow (zero flux for pressure)
! friction for velocities (+ turbulent variables)
! zero flux for scalars
do iphas = 1, nphas
itypfb(ifac,iphas) = iparoi
enddo
enddo
THe "upper face of the basin" : symetry (Upper-face)
[
Code: Select all
code]call getfbr('Upper-face', nlelt, lstelt)
do ilelt = 1, nlelt
ifac = lstelt(ilelt)
! Wall: zero flow (zero flux for pressure)
! friction for velocities (+ turbulent variables)
! zero flux for scalars
do iphas = 1, nphas
itypfb(ifac,iphas) = isymet
enddo
enddo[/code]
So, since i want to couple the turbine with all geometry in subroutine "ussatc.f90" of SCR.1and SRC.2 i take:
Code: Select all
do ii = 1, nbcsat
if (ii .eq. 1) then
! numsat = 1
call defsat(numsat, namsat, 'all[]', ' ', ' ', 'all[]', iwarns)
What do you think? because i have the fatat error in this case
Thanks
Re: turbine modelling
Posted: Sun Jun 24, 2012 10:42 am
by stage75
Hi eveone,
please how can I do to define the coupling between my two instances (how can adapt my subroutine "ussatc.f90")!
I did in SRC.1
Code: Select all
do ii = 1, nbcsat
if (ii .eq. 1) then
! numsat = 1
call defsat(numsat, namsat, 'all[]', ' ', ' ', '[b]number of faces of my first intance[/b]', iwarns)
In SRC.2 I did,
do ii = 1, nbcsat
if (ii .eq. 1) then
! numsat = 1
call defsat(numsat, namsat, 'all[]', ' ', ' ', '
number of faces of my second instance', iwarns)
Is this right please!!
Re: turbine modelling
Posted: Sun Jun 24, 2012 8:07 pm
by Yvan Fournier
Hello,
I have tested and debugged internals, but not set up, rotor/stator cases with Code_Saturne, so I am not too familiar with the use of ussatc.f90.
With no details on the error (listing.*), I cannot tell you much, and I do not have details on the rest of your setup.
What I would recommend is to test both domains independently (using a reasonably close boundary condition on the faces where coupling shoud occur), and solve all data setup errors for this case first.
As a second step, couple the 2 domains, using ussatc.f90.
This is easier with version 2.2, as you each domain setup can be run independently, while with 2.0, you need to rebuild a separate case for each unstructured domain (but is tis still feasible).
Regards,
Yvan
Re: turbine modelling
Posted: Mon Jun 25, 2012 8:45 am
by stage75
Hello,
I get this error in "listing.2"
Code: Select all
Unmatched Code_Saturne couplings:
---------------------------------
Couplage Code_Saturne :
id de couplage : 0
nom local : ""
numéro local : -1
/home/itouche/Téléchargements/tmp/installer/ncs-2.0.4/src/base/cs_sat_coupling.c:2076: Erreur fatale.
Au moins 1 couplage Code_Saturne a été défini pour lequel
aucune communication avec une instance de Code_Saturne n'est possible.
Pile d'appels :
1: 0x7ff0eabd0562 <cs_sat_coupling_all_init+0xab2> (libsaturne.so.0)
2: 0x7ff0eaa9b345 <cs_run+0xc5> (libsaturne.so.0)
3: 0x7ff0eaa9bda5 <main+0x1f5> (libsaturne.so.0)
4: 0x7ff0e8709c4d <__libc_start_main+0xfd> (libc.so.6)
5: 0x40be89 <> (cs_solver)
Fin de la pile
Re: turbine modelling
Posted: Mon Jun 25, 2012 12:41 pm
by Yvan Fournier
This type of error occurs very early, before defining boundary conditions or even matching faces between meshes.
I just checked the examples for the definitions in DEFSAT for both ussatc.f90 files. NUMSAT should have value 2 for SRC.1, and value 1 for SRC.2 (or value -1 for automatic matching in both cases).
Re: turbine modelling
Posted: Mon Jun 25, 2012 12:53 pm
by stage75
In this case I get :
Code: Select all
The group or attribute "Face_53" in the
selection criterion:
"Face_53"
does not match any edge face.
However this face I well defined in my boundary conditions!.