Towards a 3D-printed acoustic sensor inspired by hair-like structures of arachnids and insects

Martinelli, Samuele and Reid, Andrew and Windmill, James FC (2024) Towards a 3D-printed acoustic sensor inspired by hair-like structures of arachnids and insects. In: Anglo-French Physical Acoustics Conference 2024, 2024-01-17 - 2024-01-19, Lodge on Loch Lomond Hotel, Loch Lomond, Scotland.

[thumbnail of Martinelli-etal-AFPAC-2024-Towards-a-3D-printed-acoustic-sensor-inspired] Text. Filename: Martinelli-etal-AFPAC-2024-Towards-a-3D-printed-acoustic-sensor-inspired.pdf
Accepted Author Manuscript
Restricted to Repository staff only until 19 January 2025.
License: Strathprints license 1.0
Download (1MB) | Request a copy

Abstract

Biological studies of the different sensing methodologies in insects in the past century gave an understanding of the plethora of different sensory methodologies present in insects that react to acoustic stimuli [1]. Among these, the trichoid sensilla is of particular interest because by a change of the hair structure, or stiffness of the basal area, the sensilla can be tuned to a specific low frequency, near field sound, and are present in different insects and arachnids [2]. Additionally, it is believed that from this sensilla other sensing capabilities are derived, such as airflow, acceleration [2, 3] and, maybe, even odour [3, 4] and infrared sensing [5]. This research aims to replicate a similar approach to develop 3D-printed sensors that would allow filtering of different frequencies at the acquisition phase, i.e., cancelling the need for computational processes like the Fast Fourier Transform (FFT) to determine the frequency content of a sound after acquisition. This has been proved possible in previous research [6], where, by changes in the structure (shown in Figure 1), the hair could react to different narrow frequency-bands when stimulated by a periodic chirp. Study with an X-ray microCT scanner found that the 3D-printer hair sensors have very fine dimensional tolerances. Following this, improvements to the manufactured sensors have been made. This work presents results (Figure 2) obtained by the same testing conducted in [6], with new, more accurate, 3D-printed sensor structures. Moreover, COMSOL simulations showed an absolute mean error in the peak frequency of 5.1%, with a maximum of 14.2% and a minimum of 0.3%.”

CORE (COnnecting REpositories)