[PeterShi]

Hi all,

I am using incompressible LES to study airfoil-self noise. I am wondering, do I need to have sponge layers in this case? I feel like the sponge layer is needed in compressible LES solvers, but not sure about the incompressible solver.

Would like to hear any inputs. Thank you.

Best regards,
Peter

[LuckyTran]

Sound is due to compressibility... Why are you doing incompressible LES? Am I missing something here?

Sponge layers are used for damping out unwanted fluctuations for various reasons. You may or my not need it for your problem. For example a fixed pressure boundary at an outlet reflects pressure fluctuations back into the domain and can lead to numerical issues. A sponge can be used to damp out these fluctuations as a compromise and allow the fixed pressure boundary condition to still be applied without the unwanted reflection.

[PeterShi]

Originally Posted by LuckyTran

Sound is due to compressibility... Why are you doing incompressible LES? Am I missing something here?

Sponge layers are used for damping out unwanted fluctuations for various reasons. You may or my not need it for your problem.

If you use compressible LES, you can calculate the sound "directly". But it will take you very long time to finish the simulation.

Instead, if you use incompressible LES, you calculate the near field pressure fluctuations first, then use FW-H to compute the far field noise. It is a hybrid method, which saves you a lot of time

In my case, my Mach number is very low, the flow is almost incompressible. Therefore, I use incompressible LES.

Hope my explanation makes sense to you.

Best,
Peter

[FMDenaro]

Originally Posted by PeterShi

Hi all,

I am using incompressible LES to study airfoil-self noise. I am wondering, do I need to have sponge layers in this case? I feel like the sponge layer is needed in compressible LES solvers, but not sure about the incompressible solver.

Would like to hear any inputs. Thank you.

Best regards,
Peter

Maybe you are referring as to the generation of inflow conditions for LES? In such a case, you should read about the method of Kaltenbach, illustrated in Sec.10.3 of the Sagaut's book.

[LuckyTran]

Originally Posted by PeterShi

Instead, if you use incompressible LES, you calculate the near field pressure fluctuations first, then use FW-H to compute the far field noise.

Okay now I follow you. So you using LES to obtain the background flow and model the noise separately.


Well then you probably don't need a sponge. You don't really have this reflection problem in incompressible LES. I'm trying to think of what you would need/want to damp out in incompressible LES and nothing comes to mind (right now).