System Identification Technique and Neural Networks for Material Lifetime Assessment Application

Abstract

Modeling of a material lifetime to assess the material useful lifetime during its service in design has been always challenging task. In the present study, a framework of system identification technique based upon nonlinear autoregressive exogenous inputs (NARX) was introduced and presented for material lifetime assessment using neural networks (NN). Using the framework, the task of material lifetime assessment was accomplished in a fashion of one-step ahead prediction with respect to stress level. In addition, by sliding over one-step to one-step of the stress level, the task of prediction dynamically covered all loading spectrum. As a result, material lifetime assessment can be fashioned for a wide spectrum of loading in an efficient manner based upon limited material lifetime data as the basis of the NARX regressor. The multilayer perceptron (MLP)-NARX and radial basis functions NN (RBFNN)-NARX models were developed to predict fatigue lives of composite materials under multiaxial and multivariable loadings. Several multidirectional laminates of polymeric based composites were examined in this study.