frozen rotor e Coriolis Source Terms

[attene]

Dear all,

I would like to know from you upon the setting up of the Coriolis Source Terms through the GUI (version 5.0.8).

I am modeling a model tidal turbine:

My domain consists of two parts (cell zones): rotating domain (with a turbine in it ), stationary domain (rest of the tank).
No interface exists between the two aforementioned parts or in other words the mesh in conformal.

In my case, for the time being, I am interested in using the frozen rotor approach for steady calculations. The reason why I am stick on the frozen rotor is that I am doing a code to code validation against past simulations performed with ansys fluent.

As I suppose the NS equations (in the rotating domain) are solved in the relative frame of motion: Do I need to take into account the coriolis source terms in the Volume conditions? What about the centrifugal terms?

Can I then write all the quantities in the abs frame of motion? (for post processing reason)?

ps: I am also look into how to calculate out-of-plane and in-plane bending moments as well as power and thrust coefficient. I will post a specific topic on that.

Regards,

FA

[Yvan Fournier]

Hello,

The frozen rotor model should include all required terms (it has been improved in version 5.2 and somewhat reformulated, but I am not familiar enough with it to explain the details).

For postprocessing, I attach some user routines from one of our test cases, which might be useful in your case (though I'm not sure if this is exactly what you are looking for, it is pretty "classical" postprocessing for turbomachinery cases).

Regards,

Yvan

[attene]

Hello Yvan,

Thank you!!
I will try to do it one on my own taking inspiration from the one you posted.

For the same purpose, Can I use as well cs_user_extra_operations.f90 (maybe the name is a bit different in the last versions)?

Regards,

FA

[Yvan Fournier]

Hello,

Yes, both the C and Fortran versions of cs_user_extra_operations can be used. The Fortran version might be removed in a future version, but not before at least a year (probably more; conversion to C is not a high priority so it is done mainly when rewriting parts of the code for other reasons).

Best regards,

Yvan

[attene]

Hello,

I think the following part is what suits my needs, but I do need to change something to calculate the Thrust and both the in-plane and out-of plane bending moments acting on the blades. I was wondering whether there is any other specific function like cs_post_moment_of_force to do that or I need to manipulate this function by defining new axis and invariant points. Any suggestions?

Regards

FA

Code: Select all

 void
cs_user_extra_operations(void)
{

  cs_lnum_t n_elts;
  cs_lnum_t *elt_list;

  BFT_MALLOC(elt_list, n_b_faces, cs_lnum_t);
  cs_selector_get_b_face_list("wall", &n_elts, elt_list);

  cs_real_t c = cs_post_moment_of_force(n_elts, elt_list, axis);

  BFT_FREE(elt_list)
  cs_real_t power = c*omega
}

END_C_DECLS

[Yvan Fournier]

Hello,

I do not know of any other "standard" turbomachinery functions, so modifying sc_post_moment_of_force or similar functions is probably required.

If your application is generic enough, do not hesitate to submit feature requests (being specific enough in the bugtracker and/or the forum), or propose patches.

Best regards,

Yvan

[attene]

Hello,

thank you for your answer!
I will try first to calculate the power using cs_post_moment_of_force and start looking on how to implement the other
physical quantities I need to calculate.

One question regarding the surface boundaries to call through cs_post_moment_of_force corresponding to the selection criteria ( "wall" in the example above, three blades in my case). Do I need consider all of them in one call of cs_post_moment_of_force or I need to call the function more than one time (three in my case), then summing the contributes?

Regards,

FA

[Yvan Fournier]

Hello,

You can consider all the blades together with an appropriate selection criteria. The only reason for computing them separately is if you need to postprocess individual contributions (which might be useful for some transient approches, probably not for an averaged approach).

Best regards,

Yvan

[attene]

Hello,

Thanks!
I guess that a good post-processing tool like Paraview can also do this job...
I am going to explore this possibility a well! On this regard I would have a question regarding the "stress" that code saturne can compute: Does it take into account for both pressure and viscous forces?

Regards,

FA

[Yvan Fournier]

Hello,

Yes, the "stress" takes both of these into account (so it is better than simply integrating the pressure over the surface).

Best regards,

Yvan