Actuator-sensor pair excitation tuning and self diagnostics for damage identification of a sandwich plate

An active structural health monitoring system has been developed to monitor metallic honeycomb thermal protection systems (TPS) for future aerospace applications. Damage was diagnosed on the outer panel surface with dual actuator-sensor pairs attached to the inner Surface by implementing the Method of Virtual Forces. This paper expands on previous work by investigating the control of the piezo based actuators and condition monitoring of the sensing system to enable self-diagnostics. The virtual force method produces a matrix of forcing function vectors, which are equivalent to the effective forces the damage exerts on the structure. Previous work has shown that the forcing vectors are correlated with the damage and actuator forces used to interrogate the system. In this work the dual-actuator sensor pair is tuned to a specific frequency bandwidth relative to an input sensor or response sensor. Experimental results of the effects of tuning on mass damage on a TPS panel are provided. The tuning procedure was also shown to be sensitive to damage in the sensor array. A simple four degree of freedom (DOF) lumped parameter model of a TPS panel is presented with a simulated sensor mass and bond stiffness on each DOF to study the effects of sensor attachment damage. The model is verified with experimental results on a TPS panel where the threaded sensor attachment is loosened to simulate a stiffness change. The virtual forces that are estimated for attachment damage in both the model and experimental results were shown to span the full frequency bandwidth. The ability and limits of the sensor array to self-heal from attachment damage was demonstrated.