Intelsat VI is the latest series of communication satellites under design and construction by Hughes Aircraft Company and a group of international subcontractors for Intelsat. BAe is a major subcontractor responsible for the design and manufacture of the cradle and carrier which is used to interface the Intelsat VI satellite to the STS Shuttle and provide the facility for satellite deployment from the STS using the 'Frisbee' technique. BAe were responsible for the design and qualification of the structure of the perigee stage (carrier) and orbiter interface cradle. This included the production and verification of a mathematical model representing the dynamics of the system. SDRC were contracted to perform the testing and assist in the test/analysis correlation. The task, was complicated by the size and mass of the test article (which included a dummy spacecraft and a plaster filled rocket motor case) making it impractical to design a test fixture to reproduce the theoretical boundary conditions. Modal surveys of the test fixture alone, on its foundations, resulted in the definition of a modal model which was used to both correlate an analytical model of the fixture, and provide estimates of foundation stiffnesses. It was also employed to predict the modes of the cradle carrier assembly when mounted in the fixture, multi-point random was the technique selected for this investigation. The modal survey of the assembly, including the fixture, was carried out using SDRC's multi-point sine dwell and multi-phased stepped sine modal analysis system. Linearity and reciprocity checks were undertaken. Orthogonality of the test modes to an analytical mass matrix, along with other correlation procedures, was checked to verify the mathematical model. Stringent test acceptance requirements were set and met for the measured test data base. Use was made of the experimental modal model to eliminate the effect of the test fixture by grounding the fixture to carrier attachment points, making use of residual stiffness terms, to improve the accuracy of the prediction. Comparisons of measured and predicted frequency response functions were also part of the correlation process. Some adjustment of the idealisation of the solid rocket motor was found to be necessary to obtain good agreement between test and theory.
展开▼
