Hi all,

I am just wondering how DES can be used for aero-acoustic calculations; especially if one is interested in pressure fluctuations at certain points e.g in a channel. The possible RANS region at the walls should not be able to produce the high frequencies!?

Flo

You are correct when you say that the RANS will not pick up the high frequencies. DES is only useful for those components of the sound spectrum that are due to the flow fluctuations caused by features that are time-accurately resolved in the simulation. You need to know ahead of time if your simulation will provide the sound components you require.

As with most things in life, there's no single right answer. DES is good for aeroacoustics, but it does depend on the case, and what you're interested in. If, your dominant noise source in caused by a turbulent boundary layer, then your comment is correct: DES behaves like a RANS model near the wall, so will not pick up the small scale turbulent fluctuations that generate the noise, in which case you need LES.

If however, most of the noise is generated by larger, detached eddies - that typically contain more energy - then DES is OK. In my experience, in industrial applications, the second scenario is is much the more common.

Hi to both,

for a air conditioning channel with a fin the noise should be produced mainly by detached eddies, so DES could be usefull, but how good are comparison to measurements, if in the measurements one tracks the pressure fluctuations at the wall of the channel with some microphones and the calculation at these points is done with RANS!?

Flo

The lower frequency components of the wall pressure spectrum are due to the larger scale detached eddies that are resolved by the simulation (away from the RANS region). The pressure from these structures propagate through the wall layer and should not be damped out by the RANS turbulence model in the process. This approach has been used successfully to predict airfoil trailing edge noise, which I believe is the same scenario you describe.

Hi,

thanks for your help! I think I got it...

Flo

People!

DES has to be the most misused CFD technology in a long time. It is not appropriate for acoustics of channel flows. DES was designed for flows where there was a clear deliniation of scales. Flows such as bluff bodies and cavities: RANS in the boundary layers (the whole boundary layer, and LES in the wakes). It was not intended to smoothly transisiton from RANS to LES in a boundary layer. That is why it is calle DETATCHED Eddy Simulation. Be careful!

This is interesting. Even though I have little experience on DES, I've come across works on DES-SST and integration with Hybrid Wall. "RANS in the whole boundary layer" statement could be supported by Hybrid wall and the SST improvement might resolve the transition from RANS to LES. Measurement of pressure at the wall could be misleading and irrelevant because what the customer hears is pressure fluctuation progating away from the wall, not at the wall. Wall turbulence-induced structural flutter is not same as air-borne noise at a certain distance away. But, combination of both (FSI) might yield a more realistic noise prediction results.