MFPT 2018 has ended
View analytic
Thursday, May 17 • 11:30am - 12:00pm
Particle Filtering Techniques for Prognosis of Fatigue Delamination Growth Using Multi-Sensory NDE Methods

Log in to save this to your schedule and see who's attending!

Feedback form is now closed.
Key words: Nondestructive Evaluation, Condition Based Maintenance, Diagnostics and Prognostics
With increasing use of fiber reinforced polymer (FRP) composites in several industries such as aviation, automotive and construction, effective diagnosis and prognosis of composite structures have become an extremely critical task in recent years. The primary goal of prognosis is to accurately predict failure threshold using inspection data from early stages of degradation. Despite outstanding qualities such as light-weight, high specific stiffness and strength, components made of composite materials are often vulnerable to damages caused due to fatigue or external impacts which compromise their performance and hence propel the need of their periodic inspection by robust non-destructive evaluation (NDE) techniques. Overall accurate health prognosis is critical for condition-based-maintenance (CBM) and for reducing life-cycle costs by taking full advantage of the remaining useful life (RUL) of a component.

In this paper, growth of delamination caused by fatigue loading in GFRP Mode I composite samples was periodically inspected by optical transmission scanning (OTS) after regular intervals of cyclic loading. In addition to OTS, measurements were obtained using guided waves (GW) which were excited and sensed by surface-mounted piezoelectric sensors (PZT). An integrated prognostics framework for estimation of damage propagation was proposed which utilizes physical model based on Paris law and CBM data obtained from both NDE techniques. A Bayesian method based on particle filtering was implemented to estimate model parameters using damage-sensitive features extracted from the two sets of NDE measurements. Material and model uncertainties were taken into account during update of model parameters and RUL computation. Prediction from both methods were compared against ground truth observed directly using a digital camera image of cross-section of delaminated sample. Results demonstrating the prognostic capabilities of the two NDE methods on inspection of GFRP composites will be presented.


Yiming Deng

Nondestructive Evaluation Laboratory Dept. of Electrical and Computer Engineering Michigan State University

Thursday May 17, 2018 11:30am - 12:00pm
Cape Lookout