Integrated Evaluation of the Aeroacoustics and Psychoacoustics of a Single Propeller

Abstract

Aeroacoustic noise in multiple rotor drones has been increasingly recognized as a crucial issue, while noise reduction is normally associated with a trade-off between aerodynamic performance and sound suppression as well as sound quality improvement. Here, we propose an integrated methodology to evaluate both aeroacoustics and psychoacoustics of a single propeller. For a loop-type propeller, an experimental investigation was conducted in association with its aerodynamic and acoustic characteristics via a hover stand test in an anechoic chamber; the psychoacoustic performance was then examined with psychoacoustic annoyance models to evaluate five psychoacoustic metrics comprising loudness, fluctuation strength, roughness, sharpness, and tonality. A comparison of the figure of merit (FM), the overall sound pressure level (OASPL) and psychoacoustic metrics was undertaken among a two-blade propeller, a four-blade propeller, the loop-type propeller, a wide chord loop-type propeller, and a DJI Phantom III propeller, indicating that the loop-type propeller enables a remarkable reduction in OASPL and a noticeable improvement in sound quality while achieving comparable aerodynamic performance. Furthermore, the psychoacoustic analysis demonstrates that the loop-type propeller can improve the psychological response to various noises in terms of the higher-level broadband and lower-level tonal noise components. It is thus verified that the integrated evaluation methodology of aeroacoustics and psychoacoustics can be a useful tool in the design of low-noise propellers in association with multirotor drones.

Keywords: aerodynamic noise; broadband noise; loop-type propeller; psychoacoustic; tonal noise.

Figure 1

The hover stand setup in an anechoic chamber for aerodynamic and acoustic measurements.

Figure 2

Five propellers: (a) two-blade propeller (#1), (b) four-blade propeller (#2), (c) loop-type propeller (#3), (d) wide chord loop-type propeller (#4), and (e) DJI Phantom propeller (#5).

Figure 3

Radial distributions of chord length (a) and pitch angle (b) of five propellers.

Figure 4

Comparison of (a) the thrust coefficient ($C_t$), (b) torque coefficient ($C_Q$), and (c) figure of merit (FM) among the two-blade propeller (#1), four-blade propeller (#2), loop-type propeller (#3), wide chord loop-type propeller (#4), and DJI Phantom III propeller (#5).

Figure 5

Comparison of mechanical power vs. thrust at four stable thrusts of five propellers.

Figure 6

Comparison of the power spectral density of background noise, unloaded motor noise, and DJI Phantom III propeller (#5) ($5400 RPM, \theta =25 \mathrm{deg}$).

Figure 7

Power spectral density of five propellers at a rotational speed of 5400 RPM at Mic 2. (a) 100–20 kHz, (b) 100–4 kHz, and (c) 4.5–10 kHz.

Figure 7

Power spectral density of five propellers at a rotational speed of 5400 RPM at Mic 2. (a) 100–20 kHz, (b) 100–4 kHz, and (c) 4.5–10 kHz.

Figure 8

Comparison of OASPL among the two-blade propeller (#1), four-blade propeller (#2), loop-type propeller (#3), wide chord loop-type propeller (#4), and DJI Phantom III propeller (#5) at a rotational speed of 5400 rpm (a) and at a stable thrust of 3 N (b).

Figure 9

Contour maps of the SPL in dB ($Mic 2, \theta =25 \mathrm{deg}$) over a frequency range of 50–2000 Hz associated with propeller noises: (a) two-blade propeller (#1), (b) four-blade propeller (#2), (c) loop-type propeller (#3), (d) wide chord loop-type propeller (#4), and (e) DJI Phantom propeller (#5) at various rotational speeds.

Figure 9

Contour maps of the SPL in dB ($Mic 2, \theta =25 \mathrm{deg}$) over a frequency range of 50–2000 Hz associated with propeller noises: (a) two-blade propeller (#1), (b) four-blade propeller (#2), (c) loop-type propeller (#3), (d) wide chord loop-type propeller (#4), and (e) DJI Phantom propeller (#5) at various rotational speeds.

Figure 10

Contour maps of the SPL in dB ($Mic 2, \theta =25 \mathrm{deg}$) over a frequency range of 6000–10,000 Hz associated with propeller noises: (a) two-blade propeller (#1), (b) four-blade propeller (#2), (c) loop-type propeller (#3), (d) wide chord loop-type propeller (#4), and (e) DJI Phantom propeller (#5) at various rotational speeds.

Figure 10

Contour maps of the SPL in dB ($Mic 2, \theta =25 \mathrm{deg}$) over a frequency range of 6000–10,000 Hz associated with propeller noises: (a) two-blade propeller (#1), (b) four-blade propeller (#2), (c) loop-type propeller (#3), (d) wide chord loop-type propeller (#4), and (e) DJI Phantom propeller (#5) at various rotational speeds.

Figure 11

Comparison of the tonal noise, $OASP{L}_i \left(i=2, 4,\dots , 36\right)$, below 5400 rpm at four locations of Mic 1 (a), Mic 2 (b), Mic 3 (c), and Mic 4 (d).

Figure 12

Comparison of the average broadband noise ($OASP{L}_{band}$) beyond 6000 Hz vs. rotation speed at four locations of Mic 1 (a), Mic 2 (b), Mic 3 (c), and Mic 4 (d).

Figure 13

Comparison of noise loudness among the two-blade propeller (#1), four-blade propeller (#2), loop-type propeller (#3), wide chord loop-type propeller (#4), and DJI Phantom III propeller (#5) at a rotational speed of 5400 rpm (a) and at a stable thrust of 3 N (b).

Figure 14

Comparison of noise sharpness among the two-blade propeller (#1), four-blade propeller (#2), loop-type propeller (#3), wide chord loop-type propeller (#4), and DJI Phantom III propeller (#5) at a rotational speed of 5400 rpm (a) and at a stable thrust of 3 N (b).

Figure 15

Comparison of noise roughness among the two-blade propeller (#1), four-blade propeller (#2), loop-type propeller (#3), wide chord loop-type propeller (#4), and DJI Phantom III propeller (#5) at a rotational speed of 5400 rpm (a) and at a stable thrust of 3 N (b).

Figure 16

Comparison of fluctuation strength among the two-blade propeller (#1), four-blade propeller (#2), loop-type propeller (#3), wide chord loop-type propeller (#4), and DJI Phantom III propeller (#5) at a rotational speed of 5400 rpm (a) and at a stable thrust of 3 N (b).

Figure 17

Comparison of tonality among the two-blade propeller (#1), four-blade propeller (#2), loop-type propeller (#3), wide chord loop-type propeller (#4), and DJI Phantom III propeller (#5) at a rotational speed of 5400 rpm (a) and at a stable thrust of 3 N (b).

Figure 18

Comparison of psychoacoustic annoyance (Di’s model) among the two-blade propeller (#1), four-blade propeller (#2), loop-type propeller (#3), wide chord loop-type propeller (#4), and DJI Phantom III propeller (#5) at a rotational speed of 5400 rpm (a) and at a stable thrust of 3 N (b).

Figure 19

Mean psychoacoustic annoyance (Di’s model) vs. thrust and FM among the two-blade propeller (#1), four-blade propeller (#2), loop-type propeller (#3), wide chord loop-type propeller (#4), and DJI Phantom III propeller (#5) at a rotational speed of 3000–5700 rpm.