Les mousses adaptatives pour l'amelioration de l'absorption acoustique: Modelisation, mise en oeuvre, mecanismes de controle

The objective of this thesis is to conduct a thorough numerical and experimental analysis of the smart foam concept, in order to highlight the physical mechanisms and the technological limitations for the control of acoustic absorption. A smart foam is made of an absorbing material with an embedded actuator able to complete the lack of effectiveness of this material in the low frequencies (<500Hz). In this study, the absorbing material is a melamine foam and the actuator is a piezoelectric film of PVDF.;A 3D finite element model coupling poroelastic, acoustic, elastic and piezoelectric fields is proposed. The model uses volume and surface quadratic elements. The improved formulation (u,p) is used. An orthotropic porous element is proposed. The power balance in the porous media is established. This model is a powerful and general tool allowing the modeling of all hybrid configurations using poroelastic and piezoelectric fields.;Three smart foams prototypes have been built with the aim of validating the numerical model and setting up experimental active control. The comparison of numerical calculations and experimental measurements shows the validity of the model for passive aspects, transducer behaviors and also for control configuration.;The active control of acoustic absorption is carried out in normal incidence with the assumption of plane wave in the frequency range [0-1500Hz]. The criterion of minimization is the reflected pressure measured by an unidirectional microphone. Three control cases were tested: off line control with a sum of pure tones, adaptive control with the nFX-LMS algorithm for a pure tone and for a random broad band noise. The results reveal the possibility of absorbing a pressure of 1.Pa at 1.00Hz with 100V and a broad band noise of 94dB with a hundred Vrms starting from 250Hz. These results have been obtained with a mean foam thickness of 4cm. The control ability of the prototypes is directly connected to the acoustic flow. An important limitation for the broad band control comes from the high distortion level through the system in the low and high frequency range (<500Hz, > 1500Hz).;The use of the numerical model, supplemented by an analytical study made it possible to clarify the action mode and the dissipation mechanisms in smart foams. The PVDF moves with the same phase and amplitude of the residual incidental pressure which is not dissipated in the foam. Viscous effect dissipation is then very weak in the low frequencies and becomes more important in the high frequencies. The wave which was not been dissipated in the porous material is transmitted by the PVDF in the back cavity.;The outlooks of this study are on the one hand, the improvement of the model and the prototypes and on the other hand, the widening of the field of research to the control of the acoustic transmission and the acoustic radiation of surfaces. The model could be improved by integrating viscoelastic elements able to account for the behavior of the adhesive layer between the PVDF and foam. A modelisation of electro-elastomers materials would also have to be implemented in the code. This new type of actuator could make it possible to exceed the PVDF displacement limitations. Finally it would be interesting for the industrial integration prospects to seek configurations able to maximize acoustic absorption and to limit the transmission and the radiation of surfaces at the same time.