turbine modelling

[stage75]

Hi,

I want to model a turbine rotating around a tree (using code_saturne), but I do not know what boundary conditions to be imposed on the turbine so that it rotates. Because I have the choice between (wall, inlet ad outlet) !!!

Thanks all,

-------------------------------------------------------------------------------------------------
Bonjour tout le monde,

Je veux modéliser sur le code saturne une turbine en rotation autour d'un arbre, mais je ne sais pas quelles sont les conditions aux limites à imposer à la turbine (les pales) pour qu'elle tourne en rotation. Parce que j'ai le choix entre (wall, inlet et outlet)

merci

[mercierg]

Hi,

One available option for turbine modelisation is the code/code coupling. You will simulate two instances of code_saturne (one for the whole domain, the other one for the blades aera) with an exchange interface, defined as boundary conditions.

In order to simulate the rotation, you choose the mobile mesh option, with a rotation vector OMEGA.

Be careful to use 2.1.6 version at least, and be aware of the limitation inherent to the sliding interface. You'll certainly have to limitate your time step, has well has the cell size on the interface. Moreover, the two instances process force you to create two independant meshes.

-----------------------
A few steps to create a coupling simulation :

-create a saturne study:

Code: Select all

code_saturne create -s BASSIN -c PISCINE -c TURBINE

which gives you the appropriate environment, with the runcase_coupling file.

-Prepare the to cases separetly, as normal ones, but let coupling boundary without defined condition.

-Add cs_user_coupling.c (depending on the release) file in the SRC repertory and modify it accordingly

-Define the rotation in usini1/cs_user_parameters file (look for omegax) with

Code: Select all

 icorio = 0
 omegax = 0.d0
 omegay = 0.d0
 omegaz = 0.d0

if you have no rotation, add the keyword "ifaccp=1" nearby, and "use cplsat" in the module definition of the fonction.

-Launch your simulation from runcase_coupling file.


I hope i'm clear enough. Probably you'll have problem with this coupling stuff, but feel free to ask. Once again be aware of the limitations the slinding interface creates.

and remember to have fun with it,
Guillaume

[stage75]

Thank's, I'll try to apply the steps you described to me and I'll come back to you.

Thank's

Hi,

One available option for turbine modelisation is the code/code coupling. You will simulate two instances of code_saturne (one for the whole domain, the other one for the blades aera) with an exchange interface, defined as boundary conditions.

In order to simulate the rotation, you choose the mobile mesh option, with a rotation vector OMEGA.

Be careful to use 2.1.6 version at least, and be aware of the limitation inherent to the sliding interface. You'll certainly have to limitate your time step, has well has the cell size on the interface. Moreover, the two instances process force you to create two independant meshes.

-----------------------
A few steps to create a coupling simulation :

-create a saturne study:

Code: Select all

code_saturne create -s BASSIN -c PISCINE -c TURBINE

which gives you the appropriate environment, with the runcase_coupling file.

-Prepare the to cases separetly, as normal ones, but let coupling boundary without defined condition.

-Add cs_user_coupling.c (depending on the release) file in the SRC repertory and modify it accordingly

-Define the rotation in usini1/cs_user_parameters file (look for omegax) with

Code: Select all

 icorio = 0
 omegax = 0.d0
 omegay = 0.d0
 omegaz = 0.d0

if you have no rotation, add the keyword "ifaccp=1" nearby, and "use cplsat" in the module definition of the fonction.

-Launch your simulation from runcase_coupling file.


I hope i'm clear enough. Probably you'll have problem with this coupling stuff, but feel free to ask. Once again be aware of the limitations the slinding interface creates.

and remember to have fun with it,
Guillaume

[stage75]

Hi,

This type of coupling between the rotation turbine and the fluid medium is not feasible on a version 2.0?

This is the version we have on the server. is it necessary to change this version 2.0 to this type of simulation.

[Yvan Fournier]

Hello,

See [url=http://code-saturne.org/forum/viewtopic.php?f=10&t=959[this thread[/url] in the old forum posts for a setup for version 2.0 (the setup may actually be slightly older than version 2.0, so you may need to check and adapt it). In any case, use the --new-runcase option when running "code_saturne create" with version 2.0

Still, rotor/stator interaction with version 2.0 may contain bugs (not features) corrected since, so I second Guillaume's recommendation to use at least version 2.1.6...

Regards,

Yvan

[stage75]

Thank's Yvan,

Regarding the version 2.1.6 is not available at EDF. The latest version is 2.1.4!

[Yvan Fournier]

Hello,

2.1.6 (and 2.1.7) are available in the downloads section of this site.

Regards,

Yvan

[stage75]

Thanks,

I still can't be rotated the turbine in saturne!

Is it possible to make this coupling using the GUI or not? because the link that you passed to me Yvan does'nt use GUI.

for used this rotation of turbine can i use the window of bundary condition (moved mesh in saturne to define the rotationel movment?)

there is anyone can tell me the differents steps using the GUI?

I have three meshes:

mesh 1; turbine
mesh 2 ;piscine
mesh 3: the fluid volume

can i imposed the rotationel in the file of "moved mesh" at the mesh 1 for imposed the rotation of the turbine??

thank's

[Pisolino]

Hi,
you should create a small amount of fluid around your turbine (cylinder geometry is preferred) called 'rotor' inside your total fluid domain (stator), then you subtract the turbine's volume from rotor and set the icorio rotation for the rotor domain, the inner surfaces of rotor (that are the surfaces fluid-turbine) must be set to wall. the surfaces between rotor and stator are the coupled boundaries.

Andrea

[Pisolino]

maybe take also a look here
http://code-saturne.org/forum/viewtopic ... 7&start=10