Skip to main content

Novel hybrid robust method for uncertain reliability analysis using finite conjugate map

Authors: 

Shun-Peng Zhu, Behrooz Keshtegar, Mansour Bagheri, Peng Hao, Nguyen-Thoi Trung

Source title: 
Computer Methods in Applied Mechanics and Engineering, 371: 113309, 2020 (ISI)
Academic year of acceptance: 
2020-2021
Abstract: 

Due to the limitation of measured data, the random and epistemic uncertainties must be simultaneously handled for the reliability analysis of curvilinear aircraft stiffened panels with multiple cutouts. A hybrid iterative conjugate first-order reliability method (CFORM) and adaptive dynamical harmony search (ADHS) optimization are developed for fuzzy reliability analysis (FRA) of stiffened panels. The proposed global level performs fuzzy analysis using ADHS, while the local level is utilized for reliability evaluation using efficient FORM. In ADHS, two adjusting procedures are implemented to update the new positions of reliability index bounds with dynamical parameters Due to the computational burden of fuzzy reliability analysis with robust iterative formula, a conjugate FORM is applied using finite-step size, which is formulated by applying the Armijo rule. A limited conjugate scalar factor is utilized to determine the conjugate nonlinear describe map. The proposed conjugate formula is examined by various FORM formulas. The fuzzy reliability index is obtained using alpha level set-based ADHS in terms of extracted results by conjugate FORM for stiffened panels. Furthermore, by the sensitivity measure of the fuzzy reliability index, boundaries of uncertainty interval for stiffener heights and thicknesses of the panel were proposed. Moreover, the influences of different stiffener height, design load, stiffener thickness, and skin thickness on buckling mode were discussed. Survey results demonstrated that the proposed FRA is more accurate to handle epistemic uncertainties of curvilinear aircraft stiffened panels than the classic reliability method-based FORM. The conjugate reliability method provides robust and efficient results compared to the modified iterative reliability methods.