Cyclic Symmetry Analysis: Capturing Features

[Saladsamurai]

I am wondering what kind of approaches people take to a cyclic symmetry analysis in FEM when you have multiple repeating features that don't divide to the same integer. Take the example below for example. I have a N_HOLES = 136 and I have N_SLOTS = 52.

I am not sure what to do here. The lowest common divisor is 2 which would mean doing a 180° sector, which is hardly better than running full 3D. I know that in Aerospace, they use small sector cuts based on Turbine geometry all the time. Obviously not all of the surrounding features divide the same as the turbine geometry.

But what should be done about those features? Do we just leave them as they are and let the the FEM repeat those features even if it leads to "overlapping geometry"? Or do we change the number of those features such that they are all divisible by the number of sectors.

If my question is unclear, please let me know.

 

[AlephZero]

You have to use your skill and judgment. First decide what questions the model is supposed to answer. ("Tell me everything about the structure" isn't a useful question to ask the model!)

If those slots and holes are minor features of the whole model, you could probably approximate them the same way as you would in an axisymmetric model. On the other hand if they are the important features you might need to make a full model.