Numerical Simulation And Design Optimization For Interior Acoustic-Structural Coupled Systems

Nowadays, the techniques of noise control are paid more and more attention in engineering design. For interior noise problems, more research papers can be found in many fields, such as the cabins of automobiles, ships and airplanes. Because of the complex shape, it can not be solved by traditional theory methods. Experiment techniques and numerical methods are more suitable for these problems. In this dissertation, the numerical simulations are used to study the interior noise for analysis and optimizing the acoustic problems of the cabins at design stages.This dissertation presents the finite element equation and proposes reasonable and high efficient solution algorithms for interior acoustic-structural coupled systems based on the research history and actuality of numerical methods. Then the optimization models are presented for size, shape and topology design variables. Furthermore, the sensitivity analysis equations and optimization argorithms are derived. Specially, the random response analysis, its sensitiviy and design optimization methods are studied in this dissertation.New functional models of dynamic analysis and optimization design for interior acoustic-structural coupled systems are implemented in JIFEX, which is a large-scale engineering structural analysis and optimization software. The main content of this dissertation includes the following chapters:In chapter 1, the research motivation and the main problems of noise control techniques are investigated. Several numerical methods used in acoustic fields, such as FEM, BEM, SEA (Statistical Energy Analysis), are discussed. Then the software-JIFEX is introduced. Finally, the brief description of the research work in the dissertation is presented.In the second chapter, based on the wave equation for linear acoustics and various boundary conditions, the finite element equations for lightly damped acoustic cavity are given. The equations are solved based on the assumption that the acoustic pressure vector, which indicates the magnitude and relative phases at all locations in the system, is entirely real. The semi-analytical method is used to calculate the sensitivity of stiffness, mass and damping matrices comparing with finite difference method. In the last section, the eigenfrequency and sound pressure level response of a 3D-car model are optimized to show the validity and efficiency of the sensitivity analysis and design optimization methods. In chapter 3, the finite element equations of acoustic-structural coupled systems are given. The static condensation method is used to avoid the singular structural mass matrix. In order to calculate the eigenpairs of non-symmetrical matrix, the adjoint subspace iteration method is given. Also, the relation of the left and right eigenvectors and the normalized orthogonality conditions are confirmed. The responses of coupled systems are calculated using the modal method, direct method and iteration method. Numerical examples are given to illustrate the validity of these methodsIn chapter 4, the sensitivity of the eigenfrequency and sound pressure with respect to size and shape design parameters for those coupled systems are derived. Then the design optimization models are also presented. The topology optimization method for the coupled acoustic-structural systems is studied with the topology optimization theory. The sensitivity of acoustic pressure with respect to topology design variables and topology optimization method of stiffener distribution for the acoustic-structural coupled systems are proposed to study the stiffener material distribution of structure for interior sound pressure level (SPL) reduction of the coupled system. Numerical examples are given to show the validity and the efficiency of the sensitivity analysis and design optimization methods.In the fifth chapter, the interior acoustic-structural coupled systems subjected to random excitations are studied. An iteration method is proposed to improve the efficient and accuracy of the computation. The pseudo-excitation method is used to calculate the random response in iterations. It was found that the pseudo excitation method is more efficiency than traditional random analysis methods for iteration calculation of interior acoustic-structural coupled systems beside its high efficiency.In chapter 6, the sensitivity analysis formulas for the random responses of coupled acoustic-structural systems are derived and compared with finite difference method. The node sound pressure level (SPL) or average SPL of acoustic pressure power spectral density (PSD) is selected as the objective function with the structural weight constraint. The thickness of the structural plate is chosen as design variables. The design optimization model of the coupled systems subjected to random excitations is proposed. Furthermore, the topology optimization models of the coupled systems subjected to random excitations are presented through improving the SIMP and RAMP topology optimization models. Numerical examples are given to show the validity and efficiency of the sensitivity analysis and design optimization methods.In the conclusion, the main contributions of the dissertation are summarized and the further work is suggested.