P. Sas, K. De Langhe

The following paper deals with SEA, first described in the 70s by Lyon 1 and mainly developed for evaluating structural high frequency vibrations and structural-acoustic coupling. Over the last ten years, a profound change has taken place in computer industry and data acquisition possibilities. As a result from this evolution, and because the growing demand from aerospace and automotive industry for substantial improved high frequency modelling methods, SEA has become a technique to solve these specific kinds of problems adequately. Few publications however address the accuracy and the limits of the SEA method, especially when applied to mechanical systems such as aircraft or car bodies. It is the aim of this paper to contribute to this field by determining through systematic experiments the conditions under which SEA can provide acceptable accurate estimates of vibration levels in coupled systems, together with quantitative descriptions of the sensitivity of confidence levels to variation in system parameters such as coupling strength, perturbations of geometry etc. In frequency bands with low modal density, MA and FE are especially useful to study the dynamic behaviour of structures. Yet problems arise if vibrations are considered at high frequencies : (- Many modes per unit frequency. Because of the high modal density, an individual mode study is practically impossible. (- The CPU time (FE solution) increases with increasing frequency. The numerical accuracy however mostly decreases. Higher cost for a lower accuracy should not be justified. (- Output from large FE solutions is unmanageable and in most cases difficult to interprete. (- Because of the large number of parameters involved, sensitivity analysis and consequently design optimisation becomes difficult. (- Traditional structural dynamics concentrates on low frequencies, because the corresponding modes are the dominant ones. Still, lightweight structures, eventually containing an acoustical cavity, give rise to high frequency problems. In the present paper, emphasis is given to the experimental identification of an SEA model these experiments yields usefull guidelines for the application of SEA.

DGLR/AIAA 14th Aeroacoustics Conference, 1992

Conference Paper

englisch

21,0 x 29,7 cm, 6 Seiten

DGLR-Bericht, 1992, 1992-03, DGLR/AIAA 14th Aeroacoustics Conference - Proceedings; S.988-993; 1992; Deutsche Gesellschaft für Luft- und Raumfahrt - Lilienthal-Oberth e.V., Bonn

NA

aeroacoustics

Bibliothek

1992