Akademik Veri Yönetim Sistemi
THIN-WALLED STRUCTURES, cilt.226, 2026 (SCI-Expanded, Scopus)
This paper introduces a hybrid computational framework for investigating the nonlinear dynamic response of polygonal functionally graded (FG) structural plates reinforced with triply periodic minimal surface (TPMS) architectures. The plates are composed of Polyethylene Terephthalate (PET) with spatially graded properties, providing an optimized stiffness-to-weight ratio for lightweight structural applications. Three representative TPMS configurations-Gyroid (G), Primitive (P), and I-graph/Wrapped Package-graph (IWP)-are considered to assess the influence of cellular geometry on mechanical performance. The Eshelby-Mori-Tanaka (EMT) micromechanical model is employed to describe the heterogeneous material behavior, while the nonlinear governing equations are formulated through Hamilton's principle within a quasi-three-dimensional refined plate theory (Q3D-RPT). The spatial domain is discretized using the generalized differential quadrature (GDQ) method, and temporal integration is performed via the Newmark implicit algorithm. Model predictions are validated through Finite Element Method (FEM) simulations and a deep neural network (DNN) surrogate trained on transient dynamic responses. The excellent agreement between analytical, numerical, and machine learning results confirms the reliability, accuracy, and computational efficiency of the proposed approach, providing a robust framework for the analysis and optimization of advanced polygonal FG-TPMS structural systems.