Prediction of the Natural Frequency of the Multiwalled Carbon Nanotube-Filled Woven Glass Fibre and Metal Laminated Structure Using Artificial Neural Network

Abstract

The current research explored the predictive model for the natural frequencies (NF) of multiwalled carbon nanotube (MWCNT)-incorporated woven glass fibre and metal laminated structures through an Artificial Neural Networks (ANNs). The MWCNTs incorporated woven glass fibre laminated structures are prepared by hand layup technique. The engineering constants of the aluminium and MWCNTs filled woven glass fibre are determined through the uniaxial tensile test as per ISO 527-5 standards. A set of experiments is performed using the modal impact hammer method to capture the natural frequencies of the proposed MWCNTs-FML plates. The numerical simulation is performed through the finite element (FE) software ABAQUS by adopting shell elements (S58R) having five degrees of freedom (DOFs) per node by utilizing the obtained engineering constants from the tensile test. The NFs obtained from the experimental technique and numerical simulation are consistence. The ANN model is trained on a comprehensive dataset derived from experimental results and FE analysis simulations. Later, the laminate sequences, aspect ratio, and side-to-thickness ratio on the NFs are investigated. The ANNs demonstrates high accuracy in predicting the NF, showcasing its potential as a reliable tool for the design and optimization of advanced composite materials. This approach significantly reduces the computational effort and time required for frequencies analysis, enabling efficient exploration of material configurations and performance predictions in engineering applications.