A high degree of safety and reliability is required in variety of automated systems, yet to date, there is no one perfect method or system that is capable of ensuring reliability by identifying equipment faults as they occur and predicting imminent failures of components. David Chelidze, assistant professor of mechanical engineering and applied mechanics at URI, has been working to solve the problem of damage diagnosis and prognosis.

Chelidze was responsible for establishing the nonlinear dynamics laboratory at URI, which focuses on topics in nonlinear dynamics and vibrations, including damage diagnosis and prognosis in engineered, geophysical and biological systems, failure mechanics, system and parameter identification, modal testing and analysis, dynamics and stability of engineered systems.

In September, 2003, Chelidze received a Phase I grant from STTR (Small Business Technology Transfer) to pursue “Data Driven Damage Diagnosis and Prognosis.”

The Missile Defense Agency (MDA), a part of the Department of Defense (DoD), is funding the project, which is under the supervision of Air Force Research Office and is administered by the Naval Surface Warfare Center. The project is collaborated with Migma Systems, Inc., which is following up in the development of Damage Diagnosis and Prognosis technology and their related products for process predictive maintenance applications.

Chelidze says, “During Phase I of our research we developed methods capable of tracking simultaneously occurring multiple damage processes and failure prognosis given appropriate damage models for air-borne laser system (ABL). We submitted a proposal for Phase II. The main objective during this phase is to apply these methods in collaboration with Boeing to develop an empirical damage model for their overwrapped composite pressure vessel (OCPV) systems used in ABL. Essentially, we want to develop the damage diagnosis system for OCPV, which will have a prognostic ability also.”

The focus of the next stage is on developing data-driven, integrated damage diagnosis and prognosis technology to enable on-line, accurate and useful failure predictions for ABL systems. The developed software system will utilize advance nonlinear dynamics based signal analysis and dynamics reconstruction methods to identify active damage modes, track their evolution, develop appropriate empirical damage models and predict remaining useful life. According to Chelidze, “One of the unique aspects of our approach is that the methods developed in the lab not only can diagnose the damage but they can be used to pinpoint the active damage states, verify the theoretical damage laws or develop empirical models.”

Fault diagnosis of system components, in particular, can lead to greater plant availability, extended plant life, higher quality products, and smoother system operations.  Empirical modeling techniques present good estimates of reliability for similar or modified products.