Checking thrust provided by fan

[gaggio]

Hello,
pretty new Code_Saturne user here, now needing some advice on the usage of fans. This feature seems powerful although not fully documented, however the source code provides good hints.
In my application (study of the wake behind marine propellers) I defined a pair of fans with appropriate blade and hub radius, and the velocity distribution seems to make sense. However, I would like to check that the thrust (force) is consistent with the delta pressure defined in input. The listing only provides the flow for each fan (which I assume to be the volume flow, based on the value) but not the actually applied force. Note that the fan curve has been provided "flat" meaning that only the constant coefficient is non-zero (same delta pressure independent of flow).

Can you suggest a clever way to check this value, perhaps through postprocessing of pressure results in Paraview?

Many thanks,
F

[Yvan Fournier]

Hello,

I need to check this. If I remember correctly, this uses a source term. I also have a piece of an old documentation from one of code_saturne's predecessors regarding the fans model, which might explain the curve coefficients (but does not fo into much detail).

I'll take a look at the code in a day or two and see whant I can recommend.

Best regards,

Yvan

[Yvan Fournier]

Hello,

Checking the fans source code, I see the listing/run_solver.log also contains the pressure jump due to the fan. An approximation of the mean force on the fan would be the pressure jump per surface unit of the fan.

That is probably simplistic, but getting something more precise would probably require also running a computation with a detailed fan (using the turbo-machinery module), which requires a larger mesh and is much more computattionaly costly (but could be used a a reference/comparison).

Best regards,

Yvan

[gaggio]

Thanks for your reply. Actually the pressure jump was given in input, calculated exactly as you propose: thrust divided by area. So verifying that is just checking that the pressure jump put in input is actually accounted for in the simulation, but the obtained force is not checked this way.
I was more thinking about integrating pressure over the cells being assigned to the fan.

I forgot to mention that I'm using version 5.0 (latest available for windows), if relevant.

A further question on the fans: can you provide any reference for the relationships used to assign the pressure jump to each cell as a function of the position of the cell with respect to the fan axis? I read the code and I see numerical coefficients in the cs_fan_compute_force function which I'm not able to recognize or relate to theory (ref https://github.com/code-saturne/code_sa ... e/cs_fan.c).

Thanks,
Regards,
Federico

[Yvan Fournier]

Hello,

For the coefficients I need to find the reference fo this model (I'll try to find it and post it later).

Also, be careful with v5.0: there was a recent fix in the fans model for which some terms were not summed correctly in parallel, so you are safer running in serial mode.

I do not have access to the machine used for Windows builds anymore, so in any case, there will be no new windows builds on our side. But if you are running under Windows 10, you can use the Windows subsystem for Linux to build any current version.

Best regards,

Yvan