Application of genetic algorithms for the identification of elastic constants of composite materials from dynamic tests

Abstract

This work has been placed within the framework of the identification of stiffness properties of composite materials from dynamic tests. More precisely, the used approach has been inserted in the general context of model updating. The genetic algorithms method has been used as a complementary technique allowing a first estimation of the elastic coefficients. In other words, the initial finite element model is estimated. The refinement of solutions can thus be made by a classical updating method, such as the sensitivity method. The procedure allows the simultaneous estimation of several properties from a single test. Properties of extension, bending, twisting and transverse shear effects can be identified. Results obtained by numerical simulation show the efficiency and robustness of the genetic algorithms. Copyright © 1999 John Wiley & Sons, Ltd.