Modeling and Prediction of Corrosion-Fatigue Failures in AF1410 Steel Test Specimens
2009
Rusk, David T. | Hoppe, Wally | Braisted, William | Powar, Nilesh
Prepared in cooperation with Univ. of Dayton Research Institute, Structural Integrity Div., Dayton, OH.
Показать больше [+] Меньше [-]The global maritime operating environment of U.S. Naval Aviation assets necessitates their prolonged exposure to severe corrosive environments. The resulting corrosion damage on flight critical structural components has a significant adverse impact on fleet readiness and total ownership costs. To address these issues, NAVAIR has initiated a multiyear research program to investigate and quantify the fatigue life reduction due to corrosion on high-strength steels, and to develop models and metrics to implement actionable maintenance criteria for corrosion damage. An Equivalent Stress Riser (ESR) model has been proposed for AF1410 steel that extends traditional notched fatigue analysis methods to applications with multiple corrosion notches present. An approximate elasticity solution featuring a Fast Fourier Transform of the corroded surface data was developed, and is used to identify local stress concentration areas that may initiate fatigue cracks under loading. Groups of image pixels with stress concentration values above a certain threshold are treated as idealized semi-elliptical notches, with corrosion-fatigue (C-F) notch factor values assigned based on notch geometry and a calibration to an experimental database of failures on bare, unnotched AF1410 C-F test plates. Median life prediction results for abrasive blasted and polished sets of test specimens show a level of conservatism between 1.0 and 3.0 for most predictions, with more severe levels of corrosion on the polished set having a higher level of conservatism. Validation tests were performed using cadmium plated C-F test plates, with the ESR life predictions showing a level of conservatism similar to the predictions on the bare, abrasive blasted C-F test plates. The ESR model was adapted to account for stress gradient effects, and demonstrated by performing a life prediction on a partial WLI scan of the corroded surface of an F/A-18C/D arresting shank.
Показать больше [+] Меньше [-]Ключевые слова АГРОВОК
Библиографическая информация
Эту запись предоставил AVANO