Ngan I., Santiago-Prowald J., Henriksen T.

In early design phases, spacecraft equipment are subjected to random vibration requirements in order to cover the acoustic environment defined by the launcher’s specification. The derivation of this requirement is often based on empirical formulations. Meanwhile, the use of detailed Boundary and Finite Elements models to predict a more realistic environment, is usually carried out at a later stage. Even when the BEM-FEM approach is employed, the instruments and equipment are generally modeled as rigid masses or with spring-mass systems representing the main modes of the payload structure only. The fidelity of the dynamic interaction is thus very limited. On the other hand, if the payloads were replaced by detailed mathematical models, the computational effort would run into the limitations of software and hardware. In this paper, the vibroacoustic environment prediction of the European Space Agency’s HERSCHEL spacecraft, using the BEM-FEM coupled vibroacoustic approach will be outlined. The HERSCHEL spacecraft weighs about 3250 kg at launch and it is about 7 metres hight. The spacecraft carries a 3.5 m aperture telescope and three scientific payloads operating at near absolute zero K. The complete mathematical model of the spacecraft structure has over 1.8 million degrees of freedom. The dimension of such mathematical model poses huge difficulties to the BEM-FEM computation. To tackle this rather unusual large model, different reduction methods such as the classical Guyan and the Craig-Bampton condensation have been implemented in the coupled vibroacoustic analysis. A new technique coined “redundant eigenvector dof reduction” will also be presented. These techniques were initially tested on the cryogenic vessel of the HERSCHEL spacecraft before being applied to the full spacecraft model. Comparison of the results from the analyses using the reduction methods with those from the full models will be summarized and discussed.