# Why does a jet engine acquire a metallic whine when in full power? [duplicate]

I had a flight yesterday on an AA Embraer ERJ-190. As the plane idled at the end of the runway there was the familiar high-pitched jet whistle, which increased in pitch as the engine started spinning up. But, when it reached full power, the whistle was joined by a far-lower-pitched metallic whine, which persisted until the pilot throttled down after the initial steep climb.

I've heard this before on some (but not all?) planes. By ear, the whistle was about four an a half octaves above the whine, which is about a factor of 24 (I may have been off by an octave, or another factor of two).

On a past flight I calculated that if the main fan spun so that its outer edge was moving at the speed of sound, the whine would be at about the rotational frequency of the fan. If the fan had 24 blades (or 48 if I have the wrong octave) then the whistle could be the fan blades, and the whine the fan itself.

One explanation could be that there's something asymmetrical about the fan, e.g. one blade is a bit too long and hits the housing when at full power. (Of course, this would make no sense from an engineering and reliability standpoint.)

What causes that far-lower-pitched metallic whine?

Edit: I just returned on a JetBlue flight, but also using an ERJ-190. Where before I was seated in row three, this time I was in row 24 of 25. On takeoff, there was only a loud roar; no "whine" or "whistle". So, it's pretty clear that the noises I'm referring to above are emitted from the engine intake.

And, on this page I found a photo of the ERJ-190's GE CF34-10E Efficient Engine's inlet:

Sure enough, the intake fan has 24 blades. So, my above guess on the frequency ratio between my "whine" and "whistle" has some physical basis.

And, I'd like to respond to a comment by @kevin:

Could it be that the "whine" always exists in the frequency spectrum, but it's just below our human ear detection threshold (both pitch and amplitude) during low rpm operation?

When the "whine" starts and ends, my impression is that yes, the pitch rises as it initially strengthens, and falls as it finally fails. However, its volume is strongly pitch-dependent, so that it is far quieter when running at 80% of the maximum pitch/RPM, and inaudible below that. It sounds as if that blade which purportedly hits the housing only does so above a certain RPM (again: this is what it sounds like, it almost certainly is not what is actually happening). More generally, my interpretation is that the cause of the "whine" only happens above a certain RPM.

Edit 2: @fooot was kind enough to point out another question that is basically a duplicate. My text and suppositions are (I think) a lot clearer, but the other question links to two videos that demonstrate exactly what I've been talking about:

Neither of the answers to the other question are correct; one focuses on the fan blade noise (too high-pitched) and the other on pressure equalization vents (not engine-speed related).

## marked as duplicate by fooot♦, kepler22b, Federico♦, Dan Hulme, SMS von der TannMar 7 '16 at 17:46

• Could it be that the "whine" always exists in the frequency spectrum, but it's just below our human ear detection threshold (both pitch and amplitude) during low rpm operation? – kevin Nov 27 '15 at 20:25
• It's most likely the tips of the blades locally exceeding the speed of sound. – Sanchises Nov 27 '15 at 20:44
• "24 or 48" could be a matter of harmonics. The first harmonic is at double the ground frequency, i.e. exactly one octave higher. (And yes, a factor of 24 is 4.58 octaves) – MSalters Nov 27 '15 at 22:59
• The number of blades matches the ratio between the two tones I hear, which strongly implies that one tone is at the per-blade frequency, and the other tone is at the per-rotation frequency. – Daniel Griscom Nov 28 '15 at 19:28
• Interference between static guide vanes ("turbine blades" is a misnomer, these are blades that that don't turn, they are in the compressor section) and the compressor vanes would be be my guess. Like the interaction between a helicopter main rotor and tail rotor, making the "chopper" sound. I'm just a layman, though... – Steve Nov 29 '15 at 2:14

I found the following RPM limits for the GE CF34-10E: 6325 RPM and 18018 RPM for N1 and N2 stages. At 100% power, the RPM for the N2 stage is 17160 RPM.

So basically we're looking at a rotational frequency of about 300 Hz. Well above the 20 Hz threshold for hearing, even if not at 100% power.

Where things get interesting is the fact that the human ear becomes markedly less sensitive to sounds below 300 Hz. So when the engine spins up from 200 Hz to 300 Hz, not only will the increased power probably produce more raw noise, it's also at a frequency that's more audible. But exactly how strong that effect is, is hard to describe. There's even a standard for this, specifically aimed at airplane noises!

• My big question is why a turbine, which is undoubtedly precisely balanced and symmetrical, makes a strong sound at its fundamental rotation frequency. – Daniel Griscom Nov 28 '15 at 15:12
• The turbine has to interact with the air, that's the whole point of a turbine. The rotational motion is periodic. Balancing will reduce asymmetries, but you cannot realistically eliminate them. But even if you did, the input airflow isn't perfectly symmetric either. A small disturbance will likely cause a vibration which dampens only slowly - it's hard to dampen vibrations in fast-moving machinery. – MSalters Nov 29 '15 at 22:35

Not sure what do you mean by "metallic", but the sound of the Embraer ERJ-190 engines is quite different from the sound of larger engines, like those of A-320s or B737s... I guess the fan does produce a much lower frequency sound due to much lower rotational speed. And at full throttle, the 190's engines really do sound "different" compared to lower power settings. I've even noticed some sort of buzzing cyclic variations, both in sound and some slight vibration too. Let's not forget the Embraers are lighter in overall weight and construction, and their feel is indeed different. Anyway, any prolonged contact between rotor and stator is not desirable or common. There is a "seal" of the labyrinth type, but it is non-contacting.

It's possible that you are hearing the blade tips 'brushing' on the honeycomb seals (it's what the blades seal into at the tip). Look up "Turbine shroud seals...."

• Can you provide more detail? Although there can certainly be contact there, it seems like this would cause wear issues if the blades are regularly rubbing on the seals. – fooot Mar 7 '16 at 16:01