Page 9 of 13
Re: turbine modelling
Posted: Mon Jul 09, 2012 8:18 am
by stage75
Hi James,
Do you have an exemple of two subroutines coupling "ussatc.f90" ? this could help me, I think
I've never done an imaginary boundary in salome !
Re: turbine modelling
Posted: Mon Jul 09, 2012 9:40 am
by JamesMcNaughton
You don't call things inlet outlet etc in Salome, as per your previous posts you have things called Face_02 etc, you define what they are in saturne.
It's not imaginary (maybe this is the wrong word), it's the interface between your coupling it just isn't a wall or inlet or anything, its just an interface, call it interface_inner / interface_outer.
Hence your ussatc will be the same but where you had put Face_59 / Face_31 (which is the name of the boundaries you want to couple) you put interface_outer / interface_inner.
Re: turbine modelling
Posted: Mon Jul 09, 2012 12:27 pm
by stage75
I tried the method what you told me but unfortunately I didn't have the recirculation area (vortex around the blades and baffles!) !
Re: turbine modelling
Posted: Mon Jul 09, 2012 2:02 pm
by JamesMcNaughton
If you show some pictures of the flow-field I might be able to suggest something depending on what the problem is.
Re: turbine modelling
Posted: Mon Jul 09, 2012 2:14 pm
by stage75
The is an example of the velocity, the result of the velocity distribution is not consistentof reality ! I think
Re: turbine modelling
Posted: Mon Jul 09, 2012 2:31 pm
by JamesMcNaughton
You should refine your mesh quite a bit
Re: turbine modelling
Posted: Mon Jul 09, 2012 2:38 pm
by stage75
Yes I know but I test first if I get the desired results first, I'll refine yes, I'm just waiting for the right result
Re: turbine modelling
Posted: Mon Jul 09, 2012 3:25 pm
by JamesMcNaughton
The top picture you post has one cell between the outer wall and the interface. This will not be enough to get anywhere near realistic solutions.
Try refining your mesh, it needn't be too far but your current mesh is not suitable. I'm not familiar with Salome's capabilities but look at putting prism layers at the interface and on the walls which will help.
cheers
Re: turbine modelling
Posted: Tue Jul 10, 2012 12:00 pm
by stage75
Hi James,
I try to raffine the mesh but I obtiened this error ,
Code: Select all
Code :
/home/itouche/Téléchargements/tmp/installer/ncs-2.0.4/src/base/cs_sles.c:1746: Erreur fatale.
Jacobi: error (divergence) solving for Turb dis
You have an idea what is it?
THanks
Re: turbine modelling
Posted: Wed Jul 11, 2012 8:01 am
by JamesMcNaughton
The solver is diverging for turbulent dissipation.
It's hard to give advice when that is all the info about the error you give.
Given that you've just refined the mesh it could be that your new mesh is not of good enough quality (have you checked for this?). Or the cells are much smaller and you've kept the same time-step (what is your CFL range?).
Other people may be able to help you further but you should give more details such as the listing file.
Thanks