Date of Award
2022
Degree Type
Dissertation
Degree Name
Doctor of Philosophy in Mechanical Engineering and Applied Mechanics
Department
Mechanical, Industrial and Systems Engineering
First Advisor
David Chelidze
Abstract
The hidden nature and detrimental effects on engineering structures make fatigue damage detection and monitoring essential to structural health monitoring (SHM). The advances in novel engineering materials and additive manufacturing meet the contemporary structural design requirements of finding an optimal combination of high strength, lightweight, and multiple functionalities. Such advances also allow the adaptation and exploitation of structural nonlinearity. The highly irregular dynamic response of the nonlinear structures aggravates the adverse effect of fatigue damage. It challenges the application of conventional fatigue monitoring and prognosis methods for the load interaction effects. A novel fatigue damage monitoring and prognosis framework are proposed in this dissertation for damage modeling and damage prognosis. It consists of two parts, one extracts the time history for analysis, the other based on statistics of VAL to predict fatigue life. First, the geometry-informed phase space warping (GIPSW) theory is developed that resolves the underlying damage dynamics from the system’s states. Its corresponding algorithm, named GIPSW for reliable monitoring (GIPSWARM), is proposed to reliably and accurately estimate fatigue damage-time histories. The effectiveness of the GIPSWARM is validated through a series of quasi-Monte-Carlo numerical experiments. And it is applied to identify various fatigue damage mechanisms caused by raster-angles during fused deposition modeling. Additionally, a fatigue life prediction method is proposed which incorporates the overload effects into the cumulative damage model. The accuracy of the method is tested using an experiment-simulation hybrid data set which outperforms the conventional methods. It can be applied when only a limited amount of system response or its statistical description is available. This new framework can serve as a data-driven SHM toolbox for damage evaluation, modeling, and prognosis.
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.
Recommended Citation
Li, He-Wen-Xuan, "ON DATA-DRIVEN FATIGUE DAMAGE MONITORING AND PREDICTION IN NONLINEAR MECHANICAL SYSTEMS" (2022). Open Access Dissertations. Paper 1453.
https://digitalcommons.uri.edu/oa_diss/1453