2022 article
Fatigue Damage Diagnostics-Prognostics Framework for Remaining Life Estimation in Adhesive Joints
Lyathakula, K. R., & Yuan, F.-G. (2022, May 10). AIAA JOURNAL.
UN Sustainable Development Goal Categories
11. Sustainable Cities and Communities
(Web of Science)
Source: Web Of Science
Added: May 23, 2022
This work presents an integrated damage diagnostics–prognostics framework for remaining useful life (RUL) estimation in the adhesively bonded joints under fatigue loading. A matching pursuit algorithm is proposed as the diagnostics technique for estimating the damage extent followed by the fatigue damage growth (FDG) simulator as the predictive model for simulating fatigue degradation. The framework calibrates the FDG simulator by quantifying uncertainties in fatigue model parameters using the damage extent data. Bayesian inference via the Markov chain Monte Carlo method is used to quantify uncertainties and estimate the probabilistic RUL from the quantified uncertainties. The FDG simulator encompasses a physics-based fatigue damage degradation model with an artificial neural network-based hybrid machine-learning model for tracing the damage progression. In the diagnostic technique, ultrasonic guided waves are excited into the structure using a pair of piezoelectric wafers, and the damage extent is quantified by reconstructing the reflected signal from the bond region. The proposed diagnostic technique is verified using the ultrasonic signal obtained from the finite element simulations. The damage prognostics part of the integrated framework is verified by estimating RUL in a mixed-mode failure joint specimen using the experimental fatigue damage progression data. In addition, the integrated framework is then verified by estimating RUL in two adhesively bonded joints: a single lap joint and a tapered single lap joint using Gaussian noise added synthetic data and diagnostic damage extent data.