Machine Learning and Optimization Algorithms for Vibration, Bending and Buckling Analyses of Composite/Nanocomposite Structures: A Systematic and Comprehensive Review

Abstract

Composite/nanocomposite structures have been commonly utilized in a variety of applications. The ability to tailor composite materials to attain superior performance in aspects such as weight and strength has made these materials a popular option in numerous applications, such as automotive, marine, aerospace, and civil engineering. With this wide range of use cases, the composite/nanocomposite structures in practice present themselves in different geometries, such as shells, plates, or beams. It is of great importance to make the most out of these materials, as they are often more difficult or costly to manufacture. As such, with such a wide range of applications, it is of the essence to have a good understanding of composite/nanocomposite materials' vibrational, buckling, or bending behaviors to grant us the ability to properly design these composite structures. To improve our design of composite/nanocomposite structures, researchers have used a large selection of optimization methods over the years, and recently, with the advent of machine learning, great focus has been placed on studying and improving composite/nanocomposite structures. This review aims to provide a comprehensive summary of the findings on the studies concerned with the bending, buckling, or vibration behaviors of composite/nanocomposite plate, shell, or beam structures in the context of optimization or machine learning methods from 2014 to 2024. The review is split into two main sections of optimization and machine learning, with subsections for buckling, vibration, and bending, with further subsections for plate, shell, and beam structures. The review is intended to act as a valuable resource for scholars invested in the use of optimization/machine learning methods for the study of vibration/buckling/bending of composite/nanocomposite shell/plate/beam structures.