Frequency and dynamic stability analysis of the sandwich structure with honeycomb core and GOP reinforcement using deep neural network

Abstract

In the presented paper, by using deep neural network (DNN), the vibrational response associated with a three-layered circular sector plate, in which the core is made of honeycomb structures and the face sheets are made of composites reinforced with graphene oxide powders is studied. Two-directional generalized differential quadrature method (2D-GDQM), as the solution method, is employed to discretize and solve ordinary differential equations. By using a momentum-based optimizer in a neural network system, the optimum value associated with different parameters is acquired. First-order shear deformation theory (FSDT) is utilized in order to attain the displacement fields based on which the governing equations are extracted. The frequency of the system is attained by using the adaptive learning method. Additionally, the results as well as the formulation of the current study are validated by using other papers. It was shown that the DNN-based model in studying these three-layered systems has less error than other models. Lastly, the impact of various parameters for the honeycomb core as well as reinforced composite face sheets, was investigated in detail.