Domain decomposition in acoustic and structure–acoustic analysis
Research output: Contribution to journal › Article
Finite element analysis in acoustics, structure–acoustics, and structure–hydroacoustic engineering applications leads to large systems of equations, and is still a challenge to current high-performance computer systems. The demands for extremely large models are due to large physical domains or the desire to resolve high-frequency levels, at least as long as the different modes are separated, say to 500 Hz. This obviously calls for small elements. It is therefore of great interest to have procedures for domain splitting, fluid–structure and fluid–fluid, thus splitting the analysis procedure into smaller problems. Use of multiple processing also lies down the road. Furthermore changes in design often affect only part of the geometry, thus, only that particular domain needs to be recalculated, for instance in the case of sound quality engineering. The interest and efforts put into numerical methods related to fluid–structure interaction are still spreading. This applies both to engineering application studies and applied mathematical issues related to the topic. Thorough investigations of issues of basic nature can be found in the literature [Harari et al., Arch. Comput. Meth. Eng. 3 (2–3) (1996) 131–309; H. Morand, R. Ohayon, Fluid Structure Interaction, Wiley, 1995]. A reduction procedure of the structure and multiple fluid domains is exercised. The coupled problem is formulated as a symmetric standard problem. A subsequent analysis in the time domain can be performed on a subset of the eigenmodes. The selection of participating modes can be made on the basis of the information provided by the coupling characteristics.
|Research areas and keywords||
Subject classification (UKÄ) – MANDATORY
|Journal||Computer Methods in Applied Mechanics and Engineering|
|Publication status||Published - 2001|