Stochastic Frequency Analysis of Variable Stiffness Laminated Composite Beam

Abstract

AbstractIn this paper, stochastic free vibration analysis of composite beam using curvilinear fibers is addressed. The material properties and the fiber orientations are considered as uncertain parameters. First-order shear deformation theory (FSDT) is adopted to construct a finite element model (FEM) in MATLAB environment. Manufacturing constraints are considered while selecting the fiber orientations for the vibration analysis. Frequency analysis is performed for different boundary conditions and with various lamination sequences. Variance-based global sensitivity analysis is performed to describe the sensitive patterns of each input variable for the free vibration. Further, stochastic frequency analysis is performed using a polynomial neural network (PNN). The accuracy and efficiency of PNN model is compared with Monte Carlo simulation (MCS).KeywordsStochastic-free vibrationCurvilinear fiberVariable stiffness laminated compositePolynomial neural network (PNN)Monte Carlo simulation (MCS)