PSI - Issue 52

Ilias N. Giannakeas et al. / Procedia Structural Integrity 52 (2024) 655–666 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

665 11

delamination damage size estimation. Specifically, the experimental observations from (Giannakeas, Sharif Khodaei, and Aliabadi 2022) were utilized to quantify the uncertainty associated with diagnosis and localization. For diagnosis, various levels of temperature difference were considered and the uncertainty on the HI ( ) was extracted for each one. For the damage location on the other hand, the localization accuracy and precision was used to generate inputs that have errors in the exact position of the impact. Then, using the damage estimation methodology developed in (Giannakeas et al. 2023), a Monte Carlo sampling approach was used to propagate the uncertainties and evaluate their effect on the delamination size estimate. From the results presented it is clear that both the uncertainties regarding the HI and the localization must be accounted for during the design of the SHM system as they can considerably affect its operation. Temperature difference between the baseline and the current signals is expected as it is not feasible to collect baseline measurements over a wide range of temperatures. Assuming = 10℃ for the health indicator , the mean estimated delamination area ̃ is increased by an average of 15.13% across all impact events while at = 15℃ the increase is 30.45%. The temperature difference had a bigger impact on the standard of deviation. When = 10℃ for instance, the standard deviation ̃ was increased by approximately 30% for the first three impacts and by 160% for the last impact event leading to significant increase in the spread of the predictions. In the case of localization uncertainty, ̃ was not affected significantly and even with = 20mm , the increase is approximately 14.9% across all impact events. The localization uncertainty however did have a significant impact on the standard deviation of the predictions. For = 15mm , ̃ increased on average by 35.2% while for = 20mm it increased by 49.62%. It is noted that a conservative scenario was used in the localization where a temperature difference of 10 ℃ was assumed during the extraction of the and values. The results highlight that understanding how uncertainties propagate through the system is important to capture its behavior. In the case of the uncertainty on , when > 10℃ the accuracy of the predictions deteriorates significantly and can lead to erroneous indications. In the case of the localization accuracy, the mean prediction does not seem to be affected however the spread of the predictions is increased significantly. Such results can be leveraged to set up limits or ranges within which the performance of the system is deemed satisfactory. In this study the uncertainty of the HI and the localization were tested separately. In future study, the interaction between the different inputs will also be studied and consider the simultaneous effect of the different sources of uncertainty. Acknowledgements The research leading to these results has gratefully received funding from the European JTICleanSky2 program under the Grant Agreement n◦ 314768 (SHERLOC). References Aliabadi, M. H., and Zahra Sharif Khodaei. 2017. Structural Health Monitoring for Advanced Composite Structures . Vol. 8. Singapore: World Scientific. Banerjee, Portia, Rajendra Prasath Palanisamy, Lalita Udpa, Mahmood Haq, and Yiming Deng. 2019. “Prognosis of Fatigue Induced Stiffness Degradation in Gfrps Using Multi- Modal Nde Data.” Composite Structures 229: 111424. Chao, Manuel Arias, Chetan Kulkarni, Kai Goebel, and Olga Fink. 2022. “Fusing Physics - Based and Deep Learning Models for Prognostics.” Reliability Engineering & System Safety 217: 107961. Dienel, Christoph P, Hendrik Meyer, Malte Werwer, and Christian Willberg. 2019. “Estimation of Airframe Weight Reduction by I ntegration of Piezoelectric and Guided Wave – Based Structural Health Monitoring.” Structural Health Monitoring 18 (5 – 6): 1778 – 88. Giannakeas, Ilias N, Fatemeh Mazaheri, Omar Bacarreza, Zahra Sharif Khodaei, and Ferri MH Aliabadi. 2023. “Probabilistic Resi dual Strength Assessment of Smart Composite Aircraft Panels Using Guided Waves.” Reliability Engineering & System Safety 237: 109338. Giannakeas, Ilias N, Zahra Sharif Khodaei, and MH Aliabadi. 2022. “An Up -Scaling Temperature Compensation Framework for Guided Wave – Based Structural Health Monitoring in Large Composite Structures.” Structural Health Monitoring , 14759217221095416. Giljohann, Sebastian, and Uwe Klingauf. 2014. “Cost -Benefit Analysis and Specification of Component- Level PHM Systems in Aircrafts.” In . Vol. 6. Giurgiutiu, Victor. 2015a. “SHM of Aerospace Composites – Challenges and Opportunities.” Proceedings of the Composites and Advanced Materials Expo, Dallas, CA, USA , 26 – 29. ——— . 2015b. Structural Health Monitoring of Aerospace Composites . Elsevier Academic Press. Goh, Joslin, Derek Bingham, James Paul Holloway, Michael J Grosskopf, Carolyn C Kuranz, and Erica Rutter. 2013. “Prediction a nd Computer Model Calibration Using Outputs from Multifidelity Simulators.” Technometrics 55 (4): 501 – 12. Higdon, Dave, James Gattiker, Brian Williams, and Maria Rightley. 2008. “Computer Model Calibration Using High - Dimensional Output.” Journal of the American Statistical Association 103 (482): 570 – 83.