Développement d’un système de mesure automatisé pour la caractérisation acoustique de matériaux en laboratoire et en milieu de travail

Abstract

Workers are exposed to loud noises in their workplace, which impacts their health and safety. Hearing loss, tinnitus, lack of sleep or stress can occur. There is a real necessity to characterize the noise propagation and to develop solutions to lessen the impact of noise on a worker’s health and life. There are several noise reduction solutions amongst which the reduction of the propagation of the acoustic waves, by soundproofing the premises for instance. My research project evolves in this context and more specifically in the implementation of a new characterization method of sound absorbing materials. In order to choose efficient sound absorbing materials for given premises, it is mandatory to know its acoustic properties for sound fields narrowing those to which they will be submitted to once installed. However, there are only two normalized methods to this day : the impedance tube method and the reverberant room method. These methods are limited : they are carried out in conditions far removed from reality as they are difficult to implement and require specific sample mounting conditions that are not quite repeatable and thus remain long and costly That is the reason why a new method, more robust and closer to the actual conditions has been developed : the sound field synthesis (Robin et al., 2014, 2018, 2019). The principle is to move an acoustic source along a parallel plane at the surface of a material to be characterized, then to generate a signal at each location and acquire the acoustic pressures received by two microphones centered above the material. This then allows to estimate the absorption coefficient by synthesizing a complex acoustic field. This method meets its limits in low frequencies emanating from the uncertainties of the sources and microphone’s positions or the simplified spherical waves model. It is also difficult to implement as the source is moved manually. Hence the idea to develop an automated test bench appeared and to propose a new method using a more precise model of acoustic field (Allard’s model). This model is developped at the university of Sherbrooke as part of the global project, funded by the IRSST. My project focuses on the automatization of a test bench in order to limit measurement errors and uncertainties and their propagation in the results. Firstly, I have implemented the conception, fabrication and automatization of the test which comprises a user interface allowing the movement of the acoustic source, the data generation and acquisition, the signal treatment and the data post treatment. Then, the validation of the proposed new method was done and it requires reference measurements in a Kundt’s tube, small alpha cabin, and in a reverberant room as well as to characterize the macroscopic parameters of the five studied materials so as to feed the theoretical models as a comparison. The obtained results are encouraging. The unrolling of the measure, encompassing the sample and microphones mounting, is easier than the previous method. The results from the new post treatment method show an improvement in low frequencies in spite of a few oscillations coming from parasite reflexions of the elements of the test bench.

Date	7 Dec 2022
Original language	French
Awarding Institution	École de technologie supérieure
Supervisor	Thomas Dupont (Supervisor) & Olivier Robin (Co-supervisor)