Vibration analysis of sandwich functionally graded material plate with cut-outs using artificial neural network technique

Abstract

In the present article, the effect of geometric discontinuities on the vibrational response of porous sandwich functionally graded material (SFGM) plates with double FGM facesheets has been investigated using the artificial neural network (ANN) technique. Generalized governing equations for the SFGM plate have been derived based on nonpolynomial based higher-order shear deformation theory (HSDT). Geometric discontinuities have been incorporated in terms of cut-outs in the SFGM plates. The FGM layers integrate a porosity model, while the core layer is considered a ceramic layer within the SFGM plate structure. Further, a C° continuous isoparametric finite element formulation with a four-noded, isoparametric quadrilateral element with seven degrees of freedom (DOFs) per node has been employed to accomplish the results. The accuracy of the present results has been demonstrated through convergence and validation studies. A comprehensive study has been carried out to investigate the influence of volume fraction index, even and uneven porosity distribution and cut-outs on the frequency parameter of SFGM plates. However, the finite element method (FEM) is computationally challenging. Therefore, the motivation behind adopting ANN technique to develop a predictive model for reasonable accuracy with less computational time. The ANN technique is proposed to predict the NDFP of the SFGM plate using cut-outs under various conditions using numerical simulation datasets. An optimised ANN model has been developed with an architecture of (6–5–10–5–1), exhibits best performance based on mean error value, which accurately predicts the Non-Dimensional Frequency Parameter (NDFP) of the SFGM plate.