Assessment of an aircraft propeller noise model by verification and experimental validation

More Info


The need to contribute to a sustainable future of the aviation industry has spiked an interest in turboprop engines. This is due to the higher propulsive efficiency, the better take-off and landing performance on short runways, and the new developments in propeller-driven hybrid and electric aircraft. The most significant drawback is the noise they generate, which is undesirable. To reduce the propeller noise, new advanced propeller designs are needed. Propeller noise models can be used as a valuable tool during the design phase. The models can only be confidently used when they have been properly validated. The more data is used to validate, the higher the confidence will be.

A master student that graduated from the Aircraft Noise and Climate Effects section at the TU Delft created a propeller noise model called HeliX-tool. This model is validated by another model and wind tunnel measurements from NASA but not yet by full-scale outdoor experimental data. To increase the confidence of the HeliX-tool, this model is compared to full-scale outdoor experimental data and the quality is assessed in this thesis.

The HeliX-tool models the steady loading and thickness noise and is based upon the helicoidal surface theory. This theory needs the aerodynamic and performance characteristics of the propeller blade which are obtained by implementing the two external analysis tools, XFOIL and XRotor.

The experiment is performed with the Pipistrel Velix Electro at Teuge airport to obtain the full-scale outdoor data to validate the model. The Pipistrel Velis Electro is a fully electric-powered two-seater owned by E-flight academy. The microphone array that is used to obtain the full-scale outdoor data has 64 microphones placed in an Underbrink configuration.
The propeller geometry is an important input for the HeliX-tool and is obtained by 3D scanning the propeller blade of the Pipistrel Velis Electro with the FaroBlu HD laser.

The validation metrics are the SPL and OSPL values. The frequency spectra are generated for both the full-scale outdoor data and the HeliX-tool when the aircraft is overhead the microphone array and are compared to each other. In general, the HeliX-tool gives higher SPL values compared to the full-scale outdoor data, which also results in higher OSPL values from the HeliX-tool.
The location of the most dominant noise source can be found, with beamforming, although the resolution is limited by the Rayleigh limit. Functional beamforming is used to reduce the side lobes and it can be seen that the dominant noise source shifts in location. It can be concluded that there might be a noise source present within the Rayleigh limit but this cannot be confirmed.

The values of the validation metrics are within 4 dB and from this can be concluded that the confidence in the HeliX-tool is increased. It can be used as a tool in the design phase of new advanced propellers but is limited to Mach numbers below 0.7 due to the Prandtl­Glauert compressibility correction used in XRotor.