Self-Diagnosis of Damage Criticality of Fibrous Composites Based on Multifunctional Characteristics

The Impact Technologies, UDRI and Brown University team have successfully developed an initial concept for autonomously detecting, classifying, and predicting damage in carbon-fiber-reinforced polymer (CFRP) composites using subtle changes in the resistivity signature of the structure. The demonstrated approach fused raw measurement and processed data from a set of optimally placed electrodes, along with an associated electro-mechanical model that was trained to recognize the quantitative relationships between specific CFRP damage types, sizes and location produced by the electrical resistivity signatures. During our proposed Phase II program, we will build upon this Phase I effort and develop an integrated product that utilizes the electro-mechanical models and autonomous damage detection/isolation software to autonomously relate changes in monitored parameters, such as electrical resistance and acoustic emission, to the mechanical and damage state of a CFRP laminate. This will be initially performed on CFRP panels manufactured by UDRI with various states of seeded damage. The Phase II damage detection and prediction system will also be demonstrated on parts provided by our development partners GE Aircraft Engines and the Boeing Phantom Works. GE Aircraft Engines is specifically interested in pursuing this technology for their engine casings and will provide data on such a component. Boeing Phantom Works is interested in using the technology on their composite exhaust duct shroud and/or other sub-components. The Phase II program will employ other robust and low-cost sensing systems such as piezoelectric devices, acoustic emission, accelerometer and strain-based sensors to provide comprehensive insight into the structure’s observed health state. These multiple techniques will serve a two-fold purpose, namely, enable comparison as well as complement each other in case of incomplete damage mapping from one set of sensors, and will enable us to develop approaches that will be applicable in real-time structural health monitoring.