Influence of Turbulence Intensity and Length Scale on UAV Propeller Noise

Small unmanned aerial vehicles (UAVs) rely on rotating propellers to generate thrust, but the noise they produce still remains a concern, especially in urban environments. Propeller noise depends on blade geometry, rotational speed, and the surrounding airflow and when a propeller encounters turbulent air, irregular changes in air velocity cause its aerodynamic loading to fluctuate and generate additional sound.

Propeller noise consists of tonal and broadband components. Tonal noise occurs at distinct frequencies associated with blade rotation, including the blade-passing frequency and its harmonics, whereas broadband noise extends over a wider frequency range and arises from unsteady aerodynamic processes. In relatively uniform airflow, periodic blade motion and pressure variations on the blade surfaces contribute to the sound field. Under turbulent conditions, fluctuations in aerodynamic loading become an important source of broadband noise, especially as incoming turbulent structures encounter the blade leading edges.

The noise generated by turbulent airflow depends partly on its intensity and the size of its turbulent structures. Turbulence intensity describes the strength of velocity fluctuations relative to the mean airflow, and turbulence length scale represents the characteristic size of these structures. Smaller structures tend to cause more rapid changes in blade loading, whereas larger structures can affect wider portions of a blade and produce more spatially correlated pressure fluctuations. The strength of these fluctuations and how they interact across the blades both affect the sound produced by the propeller.

Studying these effects requires control over the incoming airflow. In experiments and numerical simulations, turbulence is often generated using upstream grids, airfoils, or cylinders. Changing the configuration of these objects can affect both turbulence intensity and length scale, making it difficult to separate their individual effects. Simulating the turbulent wakes produced by these objects also requires substantial computational resources. Generating turbulence with independently controlled properties provides a way to investigate how each characteristic affects propeller noise.

In a recent research paper published in Aerospace Science and Technology, Mr. Mingyu Shao, Professor Shiyi Chen and Professor Hanbo Jiang from Eastern Institute of Technology, Ningbo, developed a numerical approach combining synthetic turbulence generation, computational fluid dynamics, and aeroacoustic analysis to study noise from a small UAV propeller. By independently controlling turbulence intensity and length scale, they investigated how incoming turbulent structures affect blade loading, pressure fluctuations, and the strength, frequency distribution, and direction of the emitted sound.

Experiments and Findings

The researchers used a momentum source method to generate controlled turbulence upstream of the propeller without producing unwanted acoustic disturbances. They then simulated how the turbulent airflow interacted with the rotating blades using an acoustic-wave-preserved artificial compressibility method. The sound produced by the propeller was calculated from changes in surface pressure and blade motion.

The team compared numerical predictions with published experimental measurements for a small propeller before studying turbulent inflow and found that the predicted aerodynamic forces and acoustic characteristics agreed well with the reference measurements. Separate comparisons with analytical turbulence spectra confirmed that the generated airflow reproduced the intended turbulence conditions.

The simulations involved a propeller rotating at 90 revolutions per second in airflow moving at 10 m/s. Four turbulence length scales, from 0.025 to 0.15 m, were tested at an intensity of 20%. Additional calculations at intensities of 10% and 5% provided a direct comparison of the influence of turbulence strength.

The airflow behind the propeller changed substantially when turbulence was introduced. Under uniform inflow conditions, the rotating blades generated an organized wake containing helical tip vortices. Incoming turbulence disturbed these structures and caused pronounced variations in thrust. Smaller turbulent structures produced a more fragmented wake, whereas larger structures left recognizable tip vortices farther downstream. The average thrust, however, remained relatively similar across the investigated conditions.

The differences were much greater in the acoustic response and with uniform airflow, sound radiation was strongest near the plane of propeller rotation, and the spectrum contained distinct tonal peaks. Turbulent inflow increased broadband noise and shifted the strongest radiation toward the propeller’s rotational axis. The dominant acoustic contribution came from unsteady aerodynamic loading rather than the periodic motion of the blades.

Another feature was the appearance of broad spectral humps around the blade-passing frequency and its harmonics, commonly called the haystack phenomenon. These features develop when successive blades encounter turbulent disturbances that produce related, but not identical, pressure fluctuations. The resulting sound contains energy distributed around the blade-passing frequencies instead of being concentrated entirely at narrow tonal peaks.

Professor Hanbo Jiang and colleagues found turbulence intensity had a substantial effect on sound levels and reducing the intensity from 20% to 10% lowered the overall sound pressure level by approximately 4.5 dB. A further reduction to 5% produced an additional decrease of about 4 dB. Noise at the blade-passing frequency also weakened as turbulence intensity decreased.

The influence of turbulence length scale on propeller noise was non-monotonic. Among the investigated conditions, the highest overall sound pressure level occurred at a turbulence length scale of 0.05 m. Increasing the length scale to 0.15 m reduced the sound pressure level by approximately 4 dB. This reduction was associated with a redistribution of turbulent kinetic energy toward lower frequencies, resulting in weaker high-frequency fluctuations in aerodynamic loading.

Interestingly, the smallest turbulent structures also generated lower far-field noise than those of intermediate scale, despite producing comparable local acoustic source strengths near the blade leading edges. This behavior highlights the importance of spatial coherence in acoustic radiation. Smaller turbulent structures induced less correlated pressure fluctuations across the blade surfaces and between the blades, reducing the efficiency of coherent sound radiation. Intermediate-sized turbulent structures, in contrast, produced more strongly correlated loading fluctuations, enhancing the acoustic response around the blade-passing frequency.  These findings suggest that turbulence-induced propeller noise is governed not only by the magnitude of local pressure fluctuations but also by their spatial and temporal organization.

Analysis of the blade surfaces identified the leading edges as the main regions of turbulence-induced acoustic activity. Pressure fluctuations were strongest near the outer blade sections, where rotational velocities were higher, and became weaker farther along the blade surface. This spatial distribution helped explain why the interaction between incoming turbulence and the leading edges contributed so strongly to the additional noise.

Engineering Significance

The relationship between turbulence length scale and propeller noise is important for acoustic prediction. Turbulence intensity alone cannot explain how much sound a propeller will produce. The study by Professor Hanbo Jiang and colleagues shows that the distribution of turbulent energy across frequencies and the correlation of pressure fluctuations on the blades also affect acoustic radiation. Rapid pressure changes contribute to high-frequency noise, and the way pressure fluctuations combine across the blades influences the lower-frequency response. These relationships help explain why turbulent flows of similar intensity can produce different noise levels.

The numerical approach reported in the study provides a practical way to investigate these effects under controlled conditions and by changing turbulence intensity and length scale independently, the method enables researchers to assess their individual effects without having to simulate the turbulent wake of an upstream object.

For UAV noise prediction and control, the study findings show why the characteristics of incoming airflow matter alongside blade geometry and rotational speed. Understanding these relationships can help engineers evaluate noise under different turbulent conditions and investigate ways to reduce acoustic emissions.