VIBRO-ACOUSTIC MODELING OF SANDWICH AND UNRIBBED OR RIBBED PANELS WITH THICK LAMINATE COMPOSITE SKIN USING STATISTICAL ENERGY ANALYSIS
Abderrazak Mejdi  1@  
1 : esi-group  -  Website
12555 high bluss drive 174 , san diego , ca, 92130 -  United States

This paper investigates the vibro-acoustic response of stiffened and unstiffened laminate composite structures and sandwich structures based on a Statistical Energy Analysis (SEA) approach. SEA is a modeling procedure which uses energy flow relationships for the theoretical estimation of the dynamic response as well as the sound transmission through structures in resonant motion. The accuracy of SEA is related to the accurate estimates of its parameters (modal density, Damping Loss Factor (DLF) and the Coupling Loss Factor (CLF)). Wave and modal based approaches are developed to predict the SEA parameters for both stiffened and unstiffened composite panels and sandwich panels. For composite structures each layer is assumed to be a thick laminate with orthotropic orientation. Moreover, rotational inertia and transversal shearing, membrane and bending deformations are accounted for. First order shear deformation theory is used. The developed approach handles symmetrical and asymmetrical constructions of an unlimited number of transversal incompressible layers. Moreover, for the case of a ribbed panel with thick composite skin, the effect of variable spacing of the ribs is accounted for. The sandwich model uses a discrete displacement field for each layer and allows for out-of-plane displacements and shearing rotations. The accuracy of this modeling approach is confirmed through comparison to measured test data and alternate validated theoretical results. The advantages of the new developed models compared to the classical models are also investigated. Representative examples of aircraft interior noise predictions for typical load cases are shown and the use of SEA models as a tool for guiding construction of multi-layer lightweight structures to meet acoustic performance and weight targets and optimize designs are presented. Conclusions about the overall applications and improvements offered by these approaches, current limitations, and future work to extend and improve these approaches are given.


e
Online user: 1