Earplugs remain the most effective short term solution to tackle the problem of occupational hearing loss. However, they are not always effective and adapted as desired and their in situ performances are difficult to assess. Wearing earplugs is associated with several issues such as risk co-factors in work accidents, insufficient in situ protection of the worker compared to the laboratory conditions, difficulty of using standardized measurement techniques in work places. Their design is often based on empirical methods where comfort is not taken into account.
To address some of these issues, a collaborative research project between l’École de technologie supérieure and the Institut de recherche Robert Sauvé en santé et sécurité du travail has been launched in 2010. This thesis is part of this project, its objective is to contribute to the better understanding of the problem of the sound transmission through the earplug-earcanal system to improve the design of the protector, to contribute to a better assessment of its effectiveness and thus ultimately to improve the workers protection. This objective is accomplished via a methodological development of a predicting tool to simulate the attenuation in the framework of a step by step modeling approach which integrates gradually geometrical and physical complexities. This development is carried out in four steps, the first 3 being presented in the form of peer reviewed journal articles and the forth constitutes an additional chapter in this article-based thesis.
In the first step, only the geometrical aspect of the ear canal is studied via the question of the validity of a simplified 2D axisymmetric description of the ear canal geometry in the case of rigid boundary conditions. This choice of boundary conditions allows in a first approximation to set aside the problems associated with the integration of the earcanal constitutive tissues. This question is investigated through comparisons of attenuation predictions with 2D and 3D models, and for individual or averaged over a group cases. The emphasis is put on finding the most reliable 2D geometry reconstruction method (in terms of attenuation prediction) which can be used, in order to define a 2D axisymmetric geometry based on geometrical characteristics of a 3D one.
In the second step, the previous 2D model is extended to take into account a layer of skin on the ear canal walls. This also corresponds to a configuration of the synthetic earcanal included in the acoustical test fixtures that are commonly used to measure the attenuation. The role of the skin and its effect on the earplug attenuation are studied. The contributions of the different acoustic pathways due to an airborne excitation are quantified. More generally, the investigation concerns the energy circulating within the domain. In addition, statistical analyses were performed to quantify the effect on the attenuation of the mechanical parameters of both the skin and the earplug.
In the third step, the previous model is extended to take into account the others tissues surrounding the ear canal (bone and soft tissues). Moreover, the previous geometry (the cylindrical one) is partially modified to obtain an average 2D axisymmetric ear canal geometry with a variable cross section. This choice is motivated by the results obtained in the previous steps. This model is first validated by comparisons with measurements on human subjects and then exploited to quantify the impact of various factors known to vary the attenuation when measured in laboratory conditions. These factors are: the possible presence of leaks, the insertion depth of the earplug, the inter-individual variation of the ear canal geometry and the variation of the mechanical parameters associated with the surrounding tissues. These factors are introduced one by one in the model and their impact on the attenuation is quantified and thereafter compared to standard deviations obtained from attenuation measurements on human subjects. Such comparisons are used to evaluate the predominance (as a function of the frequency) of the aforementioned factor effects on the attenuation.
In the fourth step, the study goes into the possibility of replacing the surrounding tissues of the ear canal by mechanical impedance boundary conditions. This work ultimately aims to simplify the models developed in the third step, by using models of steps 1 and 2 improved by the mechanical impedance boundary conditions rather than the more conventional condition limits (the fixed one) used till now. Two tissue replacement scenarios have been tested in the 2D axisymmetric configuration described in the first approximation. In the first scenario (similar to the model developed in step 1), the effect of all the tissues (skin, soft tissue and bone) were reduced to a mechanical impedance. A second scenario, a little less simplified, where only the bone and soft tissue domains are replaced is then considered. This second scenario allowed to correct the limitations obtained in the first scenario and helped to achieve more realism in terms of attenuation predictions, closer than those obtained with the model developed in step 3. These tests were performed on a unique geometry that does not allow generalizing the method validation which could be extended in future works to 3D models or 2D axisymmetric models with variable cross sections.
From a scientific point of view, each of the steps described above helped to better understand the key mechanisms involved in the attenuation of an earplug, the pathways leading the acoustic transmission in the case of an airborne sound excitation through the occluded ear canal, and to investigate the possible simplification which can be done to predict a realistic attenuation with a numerical model. This work helped to consolidate the idea that simplifying the 3D geometry of the occluded ear canal by a 2D axisymmetric one is possible, with certain limitations related to the geometrical reconstruction method used to define the 2D geometry. The importance of taking into account the skin in a finite element model of the occluded ear canal is highlighted. Using an average 2D axisymmetric model which integrates the ear canal surrounded tissues (or alternatively replaced by mechanical impedance boundary conditions) permitted to simulate realistic average attenuations, measurable on human subjects. The utilization of these different models allowed to a better comprehension of how the energy flows in the occluded ear canal and to quantify the effect of critical factors (leakage, insertion depth of the plug, the mechanical and geometrical parameters) responsible for variations in attenuation.
From a technological point of view, the sensitivity analyses performed on the earplug mechanical parameters with the modeling tools developed in this thesis provide concrete ways for the manufacturers to improve their product effectiveness. Following the same idea, statistical analyses performed on the mechanical parameters of the artificial skin included in artificial tests fixtures can lead to improved standards of requirements on their design or help manufacturers to approach mechanical parameters closer to human subjects.
From an occupational health and safety point of view, the advances described in this thesis and the development of the different modeling tools can help to better design the hearing protection products, and their efficiency will be better evaluated. Ultimately, this will improve working conditions by minimizing the risk of damage to the worker’s hearing.
Further research perspectives mentioned at the end of this thesis consist mainly in extending the models to different aspects not yet considered: an enlarged range of earplugs, with various materials and geometrical shapes, the introduction of the bone conduction, the modeling of dual protection when wearing earplug is coupled to earmuffs, the integration in the model of the head and the torso, to study the effects of this integration, and the consideration of impact noise.
| Date | 17 Dec 2014 |
|---|
| Original language | French |
|---|
| Awarding Institution | - École de technologie supérieure
|
|---|
| Supervisor | Frédéric Laville (Supervisor) & Franck Sgard (Co-supervisor) |
|---|
Viallet, G. (Author),
Laville (Supervisor) & Sgard (Co-supervisor),
17 Dec 2014Student thesis: Doctoral thesis › Doctorate in Engineering: Engineering