Big data in nanocomposites: ONN approach and mesh-free method for functionally graded carbon nanotube-reinforced composites

Abstract

Abstract In this paper, the concept of big data in composite materials for design purpose with focus on functionally graded carbon nanotube reinforced composites (FG-CNTRC) has been addressed through mesh-free method and an optimized neural network (ONN) approach. With this regard, mesh-free method as a robust technique was used to analyze the FG-CNTRC for vibrational frequency. The applied nanocomposite is made of aggregated single-walled carbon nanotubes (CNTs) that are embedded in an isotropic polymer as matrix. The material properties are estimated based on the Eshelby–Mori–Tanaka approach. Then a new multi-step approach was used to find optimized neural network for accurate modeling of the nanocomposite which can be used for later goals of optimization and design. Computational time and accuracy of various algorithms were investigated and compared for big data modeling of nanocomposite to come up with the optimal model. Comparative study of the results was carried out to examine and compare the accuracy of the developed ONN model relative to mesh-free method. Furthermore, a comprehensive parametric study was also performed to investigate the effect of geometrical dimensions, CNT distribution and volume fraction on vibrational frequency of the nanocomposite. Highlights Big data in nanocomposites. Analysis of nanocomposite through mesh-free method. Optimized neural network (ONN) for big data mining. Multi-step approach for finding an optimized model. Efficiency of ONN to handle big data.