An approach for predicting longitudinal free vibration of axially functionally graded bar by artificial neural network

Abstract

In this study, an Artificial Neural Network (ANN) model is designed for the analysis of the longitudinal free vibration of an axially Functional Graded (FG) bar. The material of the bar is varied from one end to the other end according to power law function. Aluminum and Alumina and the mixture of these two materials are used in the analysis. The radius of the bar is also varied along to the longitudinal direction according to power law function. In addition, the effect of the variation in the slenderness ratio is also investigated. An ANN code is written in Matlab in order to obtain a generalized approach for vibration analysis of the FG bar. In the ANN training, some results of free vibration analysis obtained from the SolidWorks program based on Finite Element Method are used. The predicted values that are not used in the training are also obtained in SolidWorks for comparison. A Matlab program code, which based ANN method is used to find new values to be predicted. The process is also made for comparison with the ANN based Matlab nntool program. When the results obtained from SolidWorks and prepared Matlab code and Matlab nntool are compared, it is seen that the results are in harmony. AMS subject classifications: 74H45, 68T07 © IMechE 2022.