Physics-based machine learning for modeling of laminated composite plates based on refined zigzag theory

Archive of Applied Mechanics, Год журнала: 2025, Номер 95(5)

Опубликована: Апрель 25, 2025

Abstract Physics-based machine learning techniques have recently gained prominence for their ability to model complex material and structural behavior, particularly in laminated composite structures. This study introduces an innovative approach, being the first employ physics-informed neural networks (PINNs) conjunction with refined zigzag theory (RZT) stress analysis of plates. A multi-objective loss function integrates governing partial differential equations (PDEs) boundary conditions, embedding physical principles into analysis. Using multiple fully connected artificial networks, called feedforward deep tailored handle PDEs, PINNs are trained using automatic differentiation. training process minimizes a that incorporates PDEs underlying laws. RZT, suitable thick moderately plates, simplifies formulation by only seven kinematic variables, eliminating need shear correction factors. The capability proposed method is validated through several benchmark cases analysis, including 3D elasticity solutions, analytical experimental results from three-point bending test based on displacement measurements reported literature. These show consistent agreement referenced confirming accuracy reliability method. Comprehensive evaluations conducted examine effects softcore presence, elastic foundation, various lamination schemes, differing loading conditions distribution

Язык: Английский

A new iterative method for free vibration analysis of a FG Timoshenko beam

International Journal Of Engineering & Applied Sciences, Год журнала: 2025, Номер 17(1), С. 17 - 43

Опубликована: Май 1, 2025

In this study, the free vibration behavior of functionally graded Timoshenko beams is analyzed. The equations motion are derived using Hamilton’s principle, resulting in fourth-order differential equations. By solving these equations, displacement and rotation functions obtained. Applying appropriate boundary conditions yields a system four linear which constitute coefficient matrix for various support scenarios. fundamental frequencies determined by identifying points where determinant equals zero. To efficiently locate points, novel iterative method proposed. results validated through comparisons with existing studies literature illustrated comprehensive tables figures.

Язык: Английский

Free vibration analysis of annular and flat plates with sinusoidal corrugations using isogeometric analysis and machine learning-based stiffness modeling

Thin-Walled Structures, Год журнала: 2025, Номер unknown, С. 113492 - 113492

Опубликована: Май 1, 2025

Язык: Английский