The starting point of this Ph.D. is the industrial issue submitted to the ÉTS by the company Bombardier Recreational Products (BRP) of the noise réduction of the tracked drive mechanism of snowmobiles. The identification of the physical mechanisms responsible for the track noise génération has been realized by measurements made partly before Ihis Ph.D. (measurements on test beds) and partly during this Ph.D. (measurements on track parts submitted to manual impacts). The results of thèse measurements show that the main noise source of the tracked drive mechanism is due to mechanical impacts generated between the bars of the track and other éléments (wheels, rails) of the guiding and driving mechanism.
In this Ph.D., the practical mean investigated to reduce the impact noise generated by the bars is réduction at the source. More precisely, the typical action considered consists in reducing the impact force magnitude by adding, in the contact zone between the colliding bodies, a soft material layer such as elastomeric materials.
The overall goal of this Ph.D. is to develop a method to predict the impact noise réduction obtained by the adding of an elastomeric layer spécimen of small thickness between the impacting body and the impacted structure which is a complex structure (i.e. a structure whose geometry is complex and whose composition involves several materials).
To reach this overall goal, three spécifie goals hâve been fixed: (1) characterize the behavior under impact of différent small thickness elastomeric layers; (2) predict the impact force generated when an elastomeric layer is added on a complex vibrating structure and (3) validate experimentally the whole method by applying it to the impact noise réduction of a bar of the snowmobile track.
To reach the first spécifie goal (characterize the behavior under impact of différent small thickness elastomeric layers), a spécifie expérimental characterization method has been developed. Firstly, an expérimental device has been realized to submit the elastomeric layer spécimens to the reproducible impact conditions of an impact hammer. The measurement of the pénétration depth of the hammer into the elastomeric layer is achieved by recording its motion with a high-speed caméra and by detecting its position by further analysis on the individual images. Secondly, the expérimental curves obtained are analyzed to point out their main characteristics and choose an appropriate impact model. Thirdly, the contact force parameters are estimated from the expérimental results and from the impact model. Using this method, eight impacted elastomeric spécimens hâve been characterized. The results show that a more précise characterization than hardness is obtained.
To reach the second spécifie goal (predict the impact force generated when an elastomeric layer is added on a complex vibrating structure), a simulation model of the impact on a structure whose vibrations are due to bending waves has been used. The physical model developed by the European project « Sounding Object » (Rocchesso et Fontana, 2003) has been chosen. From an analogy between the theory used in this model and the modal formulation used in vibration studies, some first modifications of the original program (MATLAB impactmodal.m script) hâve been made to simulate physically the impact of a mass on a vibrating structure. Some other modifications of the original program hâve been made in order to simulate the rigid body motion of the structure in the case of free boundary conditions (because the structure used for the validation of the method has free boundary conditions).
To reach the third spécifie goal (validate experimentally the whole method by applying it to the impact noise réduction of a bar of the snowmobile track), the first step has been the measurement of the force and the acoustic pressure in two configurations: WITH and WITHOUT the elastomeric layer in the contact zone. The second step has been the simulation of the configuration WITH the elastomeric layer by applying the impact model of a mass on a vibrating structure (presented in Chapter 2). In order to estimate the value of the model parameters describing the track bar, the modal parameters of the six first bending modes of the bar hâve been measured using expérimental modal analysis. Finally, validation of the method has been performed firstly by checking experimentally the hypothesis of linearity by comparisons between the réductions of force spectra obtained thanks to the adding of the elastomeric spécimen and the réductions of noise spectra. Secondly, validations of the method in time and frequency domains hâve been performed by comparisons between simulated and measured force signais. Thèse comparisons show that the discrepancies may be high enough for some spécimens (especially because the rigid motion of the structure is more complicated than a pure translation) but that the order of magnitude of simulated time and frequency signals is satisfactory.
| Date | 11 Jan 2010 |
|---|
| Original language | French |
|---|
| Awarding Institution | - École de technologie supérieure
|
|---|
| Supervisor | Frédéric Laville (Supervisor) |
|---|
Arz, J.-P. (Author),
Laville (Supervisor),
11 Jan 2010Student thesis: Doctoral thesis › Doctorate in Engineering: Engineering