Date of Award

2022-08-01

Degree Name

Master of Science

Department

Metallurgical and Materials Engineering

Advisor(s)

Ryan B. Wicker

Abstract

Sound localization is the capacity to identify the general location from which a sound is originating. The primary anatomical feature responsible for the ability to localize sound is the external portion of the ear, also known as the auricle or pinna. This faculty is significantly hindered when the ear is obstructed, as is the case with hearing protection devices (HPD), which act as a barrier between sound waves and the ear canal. In the case of electronic HPDs, certain frequencies are entirely filtered out by a digital sound processor (DSP) while the remaining frequencies are delivered directly into the ear from a single direction, further complicating sound localization. As a solution, a device is being developed at the W.M. Keck Center for 3D Innovation (Keck Center) that integrates artificial pinnae into an electronic HPD to imitate the role of its biological counterpart. The effectiveness of this approach, however, is limited unless the device accounts for the custom shape of each individual ear. Because every pinna shape is different and has unique interactions with sound signals, each person perceives sound differently. Currently, the artificial ears attached to the Keck Centerâ??s HPD are modeled after the Knowles Electronics Manikin for Acoustic Research (KEMAR) and do not reflect the actual frequency response of a person with a different ear geometry. The purpose of this research is to explore the feasibility of accurately capturing the morphology of the pinna, producing replicas for a custom fit, and ultimately, developing a model or set of models with common or generic features that successfully reproduce the frequency response of a variety of ear shapes. To ensure a controlled testing environment, research was conducted on the design of anechoic and hemi-anechoic chambers for the installation of an on-campus acoustic testing facility. Anechoic chambers are specialized rooms that absorb and attenuate sound signals to eliminate reverberation. This effect can be achieved by implementing acoustic foam panels, which are specifically shaped to dispel sound waves. For this research, various acoustic foam panels were tested to evaluate their performance based on reflection and attenuation, using American Society for Testing and Materials (ASTM) guidelines as a reference. Ultimately, standard-size melamine foam panels were selected as the main component for the installation as they provided the least reflection and had an attenuation of nearly 20 dB. A CAD layout of the finished test room was designed in Fusion 360, allowing for clear visualization of the room, as well as providing an accurate approximation of the material needed for the installation. 3D Scanning technologies were implemented to model the morphology of the ear, using an artificial left pinna as a reference. Given the possibility of future studies involving the modeling of the pinnae of actual individuals, consideration was made to utilize technologies that would not pose a potential threat to the health, as is the case, for example, with computed tomography (CT) scanners. As such, TrueDepth and photogrammetric approaches were evaluated and compared to a benchmark model obtained with an X-Ray and CT (XCT) system. TrueDepth reconstruction outperformed photogrammetry, with most areas of the scanned model having a max deviation within ±1 mm from nominal data. Therefore, it was selected as the most suitable approach for the modeling of the ear. GN ReSound, the company funding this investigation, provided ear models of different morphologies to conduct acoustic testing and analyze the relation between the shape of the sub-structures of the ear and frequency response. The models were not compatible with the KEMAR mannequin at the Keck Center and had to be edited and corrected. The correction process involved the extraction of ear geometries from the main body and the development of an interface template using the XCT scans of the original KEMAR ears as a reference. Additional alignment and merging procedures had to be defined to ensure a streamlined, repeatable process when combining both models. The final assembly was fabricated with FormLabsâ?? Form3 VP 3D printer, using an elastic engineering resin to allow the parts to adapt to KEMAR's head when placed into position. The parts produced with this material, Elastic 50A, had a layer thickness of 100 microns and, although less flexible than the original GRAS ear models, were highly robust and not easily damaged. Several printing configurations were explored to minimize the number of support structures required while preserving the quality of the overall print. Ultimately, it was concluded that orienting the models in a vertical position with a reduced number of supports yielded a more detailed part. A second version of the final ear model was produced using 3D Systems' Viper SLA 3D printer with a layer thickness of 102 microns, the same technology being employed for the fabrication of the Keck Center's HPD. Due to the lack of flexibility of the material used by the system, the models were separated into two sections to create a silicone-SLA hybrid. The portion with the pinna and its surrounding surface was printed directly with SLA, while the portion of the interface was used to fabricate a mold, which was then casted with silicone to produce a flexible part. A visit to GN's Beltone offices was scheduled to perform acoustic testing on both versions of the ear models. The tests consisted of attaching matching ear models on both sides of KEMAR's head and gradually rotating the mannequin 360 degrees while an audio sweep and a pink noise generator produced frequencies ranging from 100 Hz to 20,000 Hz, which were then captured using directional microphones located inside the ear canals. After analyzing the results, it was concluded that Elastic 50A resin was suitable for the elaboration of ear geometries and that the hybrid models required modifications to their design to improve performance. Additionally, variations in sound perception could be observed when comparing the results of different ear morphologies under the same conditions. Nonetheless, further testing is required to determine if said variations in sound perception are significant enough to be perceived by an individual to justify the fabrication of a generalized pinna shape. The contents discussed in this work serve as a reference for the development of a robust method to elaborate ear models of different morphologies compatible with KEMAR for acoustic testing, extending from the modeling of the pinnae to the final part fabrication. Additionally, future studies will also include data on the tests performed on the hemi-anechoic room designed for this investigation, as the final installation proved to be more challenging and time-consuming than initially anticipated.

Language

en

Provenance

Received from ProQuest

File Size

156 p.

File Format

application/pdf

Rights Holder

Hector Hugo Estrada Medinilla

Download

Included in

Biomedical Commons

Share

COinS