| The main objective of this research project was to study the feasibility and effectiveness of different modelling strategies available for the prediction of the propagation of structure-borne noise. A literature review of the helicopter cabin noise problematic is conducted in Chapter 1 in order to show the relevance of modelling correctly structure borne noise.;The outcome of this project is a modelling methodology that is computationally effective and leads to satisfactory results in the prediction of structure-borne noise.;Two modeling strategies are reviewed in Chapter 2, the deterministic and stochastic approaches, precisely the Finite Elements Method (FEM) and Statistical Energy Analysis (SEA). A hybrid theory that allies both FEM and SEA is introduced as a potential solution to overcome the mid-frequency problem that occurs at the intersection of the domains of validity of FEM and SEA. The hybrid methodology is also used to alleviate the conceptual limitations of SEA (structure-borne excitation, complex junctions...).;The theoretical background required to compare the deterministic, stochastic and hybrid approaches is introduced in Chapter 3, with emphasis on the hybrid approaches which constitute the originality of the present work.;The experimental case study of an in-vacuo structural problem and a coupled fluid-structure problem is presented in Chapter 4. This case study serves as a baseline for the comparison of the different methods in Chapter 5. In this chapter, models developed using the different strategies available are presented and discussed in terms of computational cost as well as capability to lead to adequate results.;Chapter 6 summarizes the main remarks regarding each technique and discussed the validity and relevance of the recently developed hybrid FEM-SEA and SEA-CMT techniques with a specific attention to the case of modeling structure-borne noise in helicopter cabins. |