The aim of this Faculty Early Career Development (CAREER) research program is to develop vibration based damage identification strategies which are rigorously suited for time-varying dynamical systems. Structural health monitoring is a key technology for enhancing safety and reliability in many critical engineering systems. Strategies based on monitoring changes in a structure's modal characteristics are attractive due to their non-destructive nature and ability to assess global structural health from limited sensor information. The current evolution of vibration based damage identification strategies largely depend on classical Eigenvector decomposition, where it is presumed that the measured response spectrum has a time invariant correspondence with the true natural frequencies of the structure. One important class of systems where this relationship does not hold is for periodically time varying dynamical systems. Prime examples are rotordynamic structures, such as turbines, shafts, propellers, gear-trains and pumps. This more complex, time varying, spectral behavior poses a significant challenge for damage estimation where a suitable inverse relation must be found to relate measured spectral changes to changes in structural system parameters. This project will develop new methods based on a generalized time-varying modal decomposition framework coupled with a novel Floquet multiplier damage sensitivity paradigm. These new formulations will enable estimation of the severity and location of multiple simultaneous structural damages in systems with inherently time varying dynamical behavior directly from measured time domain or frequency response information. For rotordynamic systems equipped force actuators, new, actively enhanced, damage estimation strategies based on Floquet Eigenvector assignment and periodic feedback control will be explored. Performing damage identification under multiple control tunings, will effectively enrich the data set and thereby compensate for limited sensor information and allow for increased spatial damage resolution.
This research will have an important impact on enhancing the safety and reliability of many infrastructures, transportation and aerospace systems which rely on critical rotating machinery. By providing a new analytical foundation for damage estimation of a large class of time varying dynamical systems, this research will advance the development of real-time, in-situ, methods for prediction of incipit system failures and allow for increased operating safety margins. Additionally, the research results will have significant implications for a wide array of control and signal processing applications where time varying dynamical systems are involved. This project also involves a comprehensive educational plan which utilizes the research and laboratory demonstrations as an entree point to motivate and inspire up-and-coming students to pursue a scientific career. High school outreach activities will be conduced through the High School Introduction to Engineering Systems (HITES) program at the University of Tennessee and will have a valuable influence on education in the greater East Tennessee region.
