Vibration and acoustics in porous insulating materials - The help of FE numerical simulations for the analysis of experiments in rooms and tubes Pierre Lamary a , José M.C. Dos Santos b , Renato Pavanello c , Olivier Tanneau d , Sylvie Le Moyne e a, Volage Limited, Aeronautic Engineering and Consulting, Ashford, UK a,b,c VibroAcoustic Lab, FEM, UNICAMP, State University of Campinas, Campinas, Brazil d,e LISMMA, SupMeca, ISMEP, Institut of Mechanics of Paris, Saint-Ouen, France a pierre.lamary@volage-ltd.com, b zema@dmc.fem.unicamp.br; c pava@dmc.fem.unicamp.br; d olvier.tanneau@supmeca.fr; e sylvie.le_moyne@supmeca.fr  Abstract To illustrate our purpose we first recall the results of a 4 years study, carried out at the ISMEP, of trimmed fuselage panels of aircraft and multilayered insulating systems. We underline the complementarities of real tests in transmissibility rooms with a set of FE and analytical methods we have developed. Two major difficulties exist i) high performance panel tests rapidly reach the limit of the experimental device, ii) short wave length in poroelastic material do not allow us direct 3D finite element calculations in all the frequency band [0 ; 6,000 Hz] of interest. Therefore, 3D models are completed with 2D models and an analytical method. We show that all this numerical tools are needed to understand the tests and to investigate some particular points as the mounting of the panel in the acoustic room. An other problematic question with poro-elastic materials is that the characterization of the material itself, which is of prior importance for inputting data in numerical models, is highly difficult to carry out. This second point motivates present research at UNICAMP based on the use of FE numerical calculations to determine the absorption and the transmissibility of porous samples in tubes. The construction of modified tubes are also envisaged for a better control of the boundary conditions during tests while the FE model will allow us to recover some fundamental characteristics by inverse calculation. We learned from our first simulations that high precisions calculations are needed to substitute the real device with a numerical one. We list at the end of the paper, new research involving virtual and real tests: curved panels, active control, noise inside a simplified aircraft cabin. From sub-structures to samples of materials, FE calculation proves itself of most practical benefits to exploit tests involving porous materials.