Identification of material compositions of functionally graded beams using a Levenberg-Marquardt backpropagation neural network

Abstract

This paper uses Layerwise (LW) Higher-order Shear Deformation Theory (HSDT) to analyze the vibration frequency and critical load of a functionally graded simply supported beam under axial load It is assumed that the material properties of the FG beam change with the thickness coordinates and the effective material properties of the FG beam can be calculated by the rule of mixtures of two phases The numerical example results show that they are in agreement with the exact solutions provided in the literature This paper also uses neural network to predict the vibration frequency of the FG simply supported beam under axial load and identify the material properties In the process of training neural networks a training group and a test group are generated in advance according to the mixed layered high-order shear deformation theory The former is used to train the neural network and the latter is used to test the accuracy of the neural network After proper training the vibration frequency predicted by the neural network has quite good accuracy and the calculation time is greatly saved compared with the mixed layered high-order shear deformation theoretical solution which also achieves the material identification that can not be achieved by the theoretical solution

Date of Award	2020
Original language	English
Supervisor	Chih-Ping Wu (Supervisor)