PSI - Issue 77

Francisco Afonso et al. / Procedia Structural Integrity 77 (2026) 584–592

589

6

F. Afonso et al. / Structural Integrity Procedia 00 (2026) 000–000

(b) Fan during operation, as seen from Prophesee’s Metavision SDK 4.

(a) Bands on fan’s top.

Fig. 5: Neuromorphic camera and standard lens setup, measuring the fan’s top.

3.2. Accelerometer with unbalanced fan

In parallel with the previous measurements, the unbalanced fan’s vibrations were also measured using a tri-axial accelerometer. This test was conducted with the same earlier version of the mechanical fan holders. Figure 6 shows the accelerometer axis orientations and the fast Fourier Transform (FFT) vibration spectrum. The dominant frequency occurs around 31.0 Hz (X axis peak), in close agreement with the previous setup results. The Y and Z axes show significantly smaller amplitudes, indicating a predominant motion along the X axis. It should be noted that using an accelerometer adds weight to the fan, which may slightly a ff ect its vibration frequency.

(b) Accelerometer FFT spectrum.

(a) Accelerometer axes orientation.

Fig. 6: Accelerometer mounted on fan and measured results.

3.3. Neuromorphic camera with macro lens and balanced fan

For this setup, the aforementioned grid was used and attached to the fan’s side. As the specimen vibrates, the vertical and horizontal grid lines generate vibrations in both directions, combining the e ff ects of the first setup’s horizontal and vertical bands. This configuration was tested without an accelerometer mounted on the specimen. Figure 7 shows the setup and the corresponding measurements obtained using Prophesee’s Metavision SDK 4.