Abstract
The term microvibrations generally refers to accelerations in the order of micro-gs and which manifest in a bandwidth from a few Hz up to say 500-1000 Hz. The need to accurately characterise this small disturbances acting on-board modern satellites, thus allowing the design of dedicated minimisation and control systems, is nowadays a major concern for the success of some space missions. The main issues related to microvibrations are the feasibility to analytically describe the microvibration sources using a series of analysis tools and test experiments and the prediction of how the dynamics of the microvibration sources couple with those of the satellite structure. In this thesis, a methodology to facilitate the modelling of these phenomena is described. Two aspects are investigated: the characterisation of the microvibration sources with a semi-empirical procedure which allows derivation of the dynamic mass properties of the source, also including the gyroscopic effect, with a significantly simpler test configuration and lower computational effort compared to traditional approaches; and the modelling of the coupled dynamics when the source is mounted on a representative supporting structure of a spacecraft, including the passive and active effects of the source, which allows prediction of the structure response at any location. The methodology has been defined conducting an extensive study, both experimental and numerical, on a reaction wheel assembly, as this is usually identified as the main contributory factor among all microvibration sources. The contributions to the state-of-the-art made during this work include: i) the development of a cantilever configured reaction wheel analytical model able to reproduce all the configurations in which the mechanism may operate and inclusive of the gyroscopic effect; ii) the reformulation of the coupling theory which allows retrieving the dynamic mass of a microvibration source over a wide range of frequencies and speeds, by means of the experimental data obtained from measurements of the forces generated when the source is rigidly secured on a dynamometric platform and measurements of the accelerations at the source mounting interface in a freefree suspended boundary condition; iii) a practical example of coupling between a reaction wheel and a honeycomb structural panel, where the coupled loads and the panel response have been estimated using the mathematical model and compared with test results, obtained during the physical microvibration testing of the structural panel, showing a good level of agreement when the gyroscopic effect is also taken into account.