I recently found the following image showing the different sources of noise in vehicles. I notice that Aerodynamics is actually quite a big part of vehicle noise, as shown below by the blue line:

enter image description here

Image source

I can imagine this is partly because of the flow going over the car. However, I think a part of this noise also comes from separating flows (for example at the rear window).

I was wondering if it is possible to reduce the aerodynamic noise of vehicles by applying separation control.

As such, I'm wondering:

How much noise does a separating flow generate?

I'm guessing this highly specific to the local geometry and velocity, but I would already be very happy with some examples.

  • 3
    $\begingroup$ We already have questions about aerodynamic noise so I don't believe this is off topic. This answer seems to address the question at least partly. $\endgroup$
    – fooot
    Commented Jan 22, 2016 at 15:46
  • $\begingroup$ I agree this is not off-topic. It is just hard to answer $\endgroup$
    – vasin1987
    Commented Jan 22, 2016 at 18:29
  • 3
    $\begingroup$ The Aviation.SE help center lists "Aerodynamics" as a topic, but with the stipulation "(related to aircraft)". This question is clearly related to automotive aerodynamics, with no clear connection to aviation. $\endgroup$
    – J W
    Commented Jan 31, 2016 at 4:00

2 Answers 2


Flow separation makes a lot of noise. Usually, it is masked by engine noise and/or damped by insulation. Any pilot of a light plane may climb to a safe altitude and then idle the engine, start a glide, and then slowly pull the stick until the wings start to stall. He will hear a lot of noise, and feel a strong vibration, too... The cause is the turbulence associated with the flow separation.


There are several sources of noise and a complete theory explaining how the noise is created. There are several mechanisms working in a car.

One mechanism appears when you have, as you mention, detached flow. When a flow is detached the flow structures after the detachment point is similar to the following one: enter image description here

As you can see the air behaves creating swirls. Those swirls interact between each other and generating noise.

So, if you are able to avoid the creation of the swirls you are able as well to reduce the noise generated.

Actually avoiding the creation of swirls and detachment is not the only mechanism, you can always "create different swirls" that generate smaller noise. For example, the antennas are not perfect cilinders and end like a cilinder with an string around in order to generate a different detachment structure, in this case to avoid vibration due to specific tones, this change of structure will also change the noise generate and I think (I can not say for sure) in this case reduces the noise.

enter image description here

  • 7
    $\begingroup$ In the antenna you show , the "string" is actually the active part of the antenna around a dielectric whip. The helical form is to shorten it. The bottom is a loading coil for resonating the monopole, the tip is a protection. I'm not sure they play an aerodynamic role. Actually this is a common helix-loaded whip or a rubber ducky when the base is not a coil but just the same helix than on the upper part. $\endgroup$
    – mins
    Commented Jan 24, 2016 at 13:12
  • 1
    $\begingroup$ It destroys, aerodynamically, the structure of the von karman swirls, breaking them in smaller (and less energetic) swirls. My backgroun is not electronics and communications, I was not aware that as well they had the function. Thanks! $\endgroup$ Commented Jan 24, 2016 at 13:22
  • 1
    $\begingroup$ It reminds me of a Megastructures episode where a bridge is subject to rain wind induced vibrations on the cables when there is a certain quantity of water droplets, solved by adding them a helical filet (source). I can't remember which bridge it was. $\endgroup$
    – mins
    Commented Jan 24, 2016 at 14:43
  • $\begingroup$ Is exactly the same phenomena. You made me anxious to see that episode!!! $\endgroup$ Commented Jan 24, 2016 at 15:20

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .