Machine learning (ML) method for static response estimation of FG plates using the novel 3D functions based on Shear-locking free meshless method

Seyed A. Vakili , Farzad Shahabian , Mohammad H. GhadiriRad

Abstract

This paper analyzes the static response of power-law thick functionally graded plates (P-FGPs) using the refined Element-Free Galerkin (EFG) method. The C1 continuity requirements of the displacement field are accurately and effectively fulfilled. A method is also presented that eliminates the shear-locking phenomenon through the use of specific shape functions. The stretching effect is approximated using higher order shear deformation theory (HSDT), and the shear correction factor is not required. According to Reddy's power law rule of mixture, the Young's modulus and Poisson's ratio of the two-phase metal–ceramic membrane vary continuously through the thickness. Furthermore, a three-dimensional function based on machine learning is employed to estimate the central deflection. This study introduces a novel three-dimensional estimating function for the central deflection of FGPs based on the results of the EFG method and sigmoid-cubic functions, representing the first application of this approach in the literature. Comparison with existing results demonstrates that the proposed estimation function provides an excellent fit to the response curve and is highly efficient for analyzing the static bending behavior of thick FGPs.

Key Words

bending; deformation theory; Element-Free Galerkin; higher order shear estimation function; power law functionally graded plates

Address

Seyed A. Vakili and Farzad Shahabian — Civil Engineering Department, Faculty of Engineering, Ferdowsi University of Mashhad, Iran
Mohammad H. GhadiriRad — Department of Civil Engineering, Faculty of Engineering Science, Quchan University of Technology, Iran

PDF Viewer

Preview uses the same access rules as Full Text PDF (subscription, purchase, or open access).

← Back to Volume 14, No. 6