WIT Press

Composite Plate Stiffness Identification Using Neural Networks

Price

Free (open access)

Paper DOI

10.2495/CP980391

Volume

21

Pages

4

Published

1998

Size

381 kb

Author(s)

Chen Xiaolin, Hua Hongxing & Hugo Sol

Abstract

A new methodology is presented for composite plate stiffness identification. The orthotropic plate stiffness coefficients are identified from its dynamic properties (eigenvalues and eigenvectors) through a backward propagation neural network, which is constructed by three multilayer subnets. An example is used to demonstrate the effectiveness and limitations of the method. 1 Introduction The higher performance composite material are more and more applied in the aeronautical, transport, and sport industry. A strategical element for the design of structures of composite materials is the capability of simulating the response of the structure (deformations, stresses

Keywords