Intelligent Damage Identification and Prognosis for Composite Structures

Impact Technologies, in collaboration with the University of Dayton Research Institute (UDRI) and GE Aircraft Engines, propose to develop a practical yet innovative damage identification and prognosis system for composite structures that uses an optimized suite of reliable COTS sensors/actuators coupled with advanced damage detection and modeling algorithms. Robust and low-cost sensing systems such as piezoelectric devices, acoustic emission sensors, inherent resistivity measurements, and strain-based sensors will be assessed and down-selected to provide comprehensive insight into the structure's observed health state. Next, the measurements will be complemented with both micro- and macro-scale structural models to both provide virtual measurements for critical locations not directly measured and to represent the baseline health state of the structure to be used in the fault detection and isolation process. The proposed modeling methodology will also be augmented using adaptive model updating techniques that will use the available (or virtually sensed) measurements on the structure to continually assess the current damage state and predict future damage based on historical loading profiles. Specifically, model-based data fusion and an extended Kalman filter are proposed to continually update the FE and composite models so that the currently identified model parameters (potentially damaged) can be assessed against the healthy baseline model. Advanced statistical and neural network classifiers will be used to determine damage type, location and severity. A Phase I demonstration is planned that utilizes an integrated sensor suite and model-based structural health monitoring (SHM) process to detect, isolate and predict damage in a realistic carbon fiber reinforced polymer (CFRP) composite laminate.