Active Health Monitoring of Complex Satellite Structures

Structural Health Monitoring methods have been previously developed in an effort to truncate the number of lengthy validation tests required for satellite checkout. Among successfully applied methods, active wave-based techniques have been able to locate and quantify structural changes, although additional insight into how the wave modes propagate in a complex structure and interact with damage is desired. In this work, an FE model of an isogrid plate is developed, and variability is introduced for each state observation. Different states are modeled as masses distributed at several discrete locations. Signals are obtained through an explicit time-domain analysis. Matching Pursuit Decomposition is used to extract physically significant parameters from the signals, the features are reduced to a more manageable subset, and Support Vector Machines are utilized to identify important features in the sensor signals. The results of this study show that even in complex rib-reinforced structures, the highest weighted class-differentiating features correspond to the antisymmetric wave mode interaction with damage. I. Introduction atellite Assembly, Integration, and Testing (AI&T) can take significant time and entails several challenges impacting the satellite timeline. Tests to qualify the satellite for the harsh launch and space environment can take months. Of these tests, dynamic testing using a shaker table provides information about the state of the global structure and how it handles mechanical loading over a range of frequencies defined by the launch vehicle payload development guide but does little to identify issues with components and their interfaces. The only existing option to locate issues with components and their interfaces are the trained senses of skilled technicians monitoring unacceptable deformations or hearing uncharacteristic sounds during active testing. Structural Health Monitoring (SHM) methods are seen as an opportunity to truncate the testing timeframe while providing higher fidelity information than current tests