A multilayered plate theory with transverse shear and normal warping functions

TL;DR

This paper introduces a multilayered plate theory incorporating transverse shear and normal warping functions, derived from exact 3D solutions, to improve accuracy in bending analysis of various plates, especially with low length-to-thickness ratios.

Contribution

The paper develops a new multilayered plate theory with five warping functions based on exact 3D solutions, enhancing modeling of transverse shear and normal stretching effects.

Findings

Models are more accurate when their kinematic fields depend on all material properties.

Including length-to-thickness ratio improves model accuracy.

Coupling between x and y directions enhances the model's predictive capabilities.

Abstract

A multilayered plate theory which takes into account transverse shear and normal stretching is presented. The theory is based on a seven-unknowns kinematic field with five warping functions. Four warping functions are related to the transverse shear behaviour, the fifth is related to the normal stretching. The warping functions are issued from exact three-dimensional solutions. They are related to the variations of transverse shear and normal stresses computed at specific points for a simply supported bending problem. Reddy, Cho-Parmerter and (a modified version of) Beakou-Touratier theories have been retained for comparisons. Extended versions of these theories, able to manage the normal stretching, are also considered. All these theories can be emulated by the kinematic field of the present model thanks to the adaptation of the five warping functions. Results of all these theories are confronted and compared to analytical solutions, for the bending of simply supported plates. Various plates are considered, with special focus on very low length-to-thickness ratios: an isotropic plate, two homogeneous orthotropic plates with ply orientation of $0$ and $5$ degrees, a $[0/c/0]$ sandwich panel and a $[-45/0/45/90]_s$ composite plate. Results show that models are more accurate if their kinematic fields (i) depend on all material properties (not only the transverse shear stiffnesses) (ii) depend on the length-to-thickness ratio (iii) present a coupling between the $x$ and $y$ directions.