Further developments of the System Equivalent Reduction Expansion Process for Finite Element Models validation of spacecrafts

Reduction of large size Finite Element models is a common technique in dynamic analysis. In particular validation of Finite Element models requires comparison between the FE model and the experimental data usually carried out using the Modal Assurance Criteria or Cross Orthogonality Checks. This is particularly true in space industry, where the FE model of a satellite has several hundreds of thousand of Degrees of Freedoms (DOFs), while experimental data, which can be acquired at a limited number of locations, typically correspond to a few hundred of DOFs. The System Equivalent Reduction Process (SEREP) technique has been identified as a particularly suitable method for spacecraft FE models. However, the reduced system matrices obtained using SEREP can be rank deficient, which limits the use of the reduced model for further analysis (e.g. forced response modelling). The present paper shows an alternate form of the SEREP technique in order to overcome this issue. The method is described and a numerical example, using data from the FE model of the Aeolus satellite, is provided in order to show the robustness and suitability of the proposed methodology.