Simultaneous Time and Profile Transformation for multistage axial fan modeling in CFX
 

Hi,

I am using transient blade row modeling for 2 Stage axial Fan (2 rotors - 2 stators with no of blades as R1-28, S1-46, R2-60 and S2-59) and have a total of 03 interfaces. Pitch ratio limit (0.75-1.4) is already accounted for by modeling an appropriate number of passages and also the fluid is compressible.

Under 'Transient Blade Row Model' setting (as per guidance from Ansys CFX solver modeling guide 2021-topic 6.1.8.5.3) I want to model the setup with both TT and PT disturbances across rotor-stator interfaces as the stage count is greater than 1.5, Hence, I want to use only one TT disturbance at my S1-R2 interface as this is the R2 inlet (probable shock location) and other R1-S1, R2-S2 interfaces as PT disturbance.

Now I want to ask that do I need to create the TT and PT disturbances separately, or I just create one TT disturbance at S1-R2 interface and the CFX will automatically take the other 2 as PT disturbances (considering I have used TRS- interface for all 03 interfaces)? If not then how to create the separate TT and PT disturbance at the same time as I only have the provision of opting either Time Transformation or Profile transformation at one time (Transient Blade Row Modles>Options- dropdown menu).

In the Ansys modeling guide for this type of case setup only the TT- disturbance is created but not much clarity is there regarding PT disturbance creation.

Any help is really appreciated.
Best Regards

Profile Transformation is the default behavior for a transient rotor-stator domain interface.

You only activate the transformation that accounts for the phase lag across a domain interface, i.e. Time Transformation or Fourier Transformation

Hey,
i'm also working with time transformation. For the stationary solution i have predefined some functions that are referenced by expressions at certain locations like inlet/outlet or at an interface. However ansys tells me that these locators have a disallowed dimensionality and only point locators are supported. i am not quite sure why this is a problem for the TT. I did not find anything in the guidlines. Does anyone know more about it?

Best regards

Each point at the mesh is at a different time; therefore, any area/volume integral will require reconstructing the solution for the same time at all geometric points and then computing the requested function.

The true physical time is easy to reconstruct for a single point, and it is still independent of the time for the neighbor nodes.

Thank you for the explanation

Right, Thank you.