Exploring Vibrations of UAV Propellers: A Preliminary Study for Characterization and Fault Detection

The propulsion systems of unmanned aircraft such as multicopters are very vulnerable to any kind of damage. Thus, the presented study introduces the possibility of using vibration sensing to determine the condition of the rotating propeller to detect possible damage or contamination. For the purpose of this study, a test stand was designed and a total of 3 different propeller damages were tested. These included minor and major damage to the tip of the propeller blade as well as disruption of the leading edge profile geometry. Propeller vibration was captured within the operating range of propeller (from 3 500 to 6000 RPM) in the axial X-axis and radial Y-axis. The measured data were processed by spectral analysis, i.e., by calculating the amplitude spectrum using a Fourier Transform. The outputs in the form of spectograms for each damage type and individual RPM were further processed by Principal Component Analysis (PCA). From the PCA, it was possible to identify the vibration frequency spectrum of the tested propeller and the individual damaged states.