An approach to reliability based design of composite structures under dynamic response

Abstract

Uncertainties in deviations of physical properties lead to a probabilistic failure analysis of the composite materials. In laminate composite structures with random physical properties its dynamic response is also a random variable. Reliability in dynamic conditions is defined as the probability of the structures not to fail within a specified time interval. Thus it is necessary to define failure criteria and the associated limit state functions. In this work, it is assumed that the structure fails if the maximum displacement, strain or stress exceeds some specified values. An exact solution to this problem is not known and we will use an approximate solution, available for Gaussian processes, where high threshold is assumed, so that its crossing is a rare event. It is considered that displacements, strains or stresses are normal processes with zero mean. The failure probability is obtained using a procedure based on Level 2 reliability analysis and Lind-Hasofer method adapted for dynamic loading conditions and considering the physical properties of the ply as random variables. In the present work it is intended to develop a simple model of probabilistic analysis of composite structures under dynamic loading conditions aiming to further developments in reliability based optimal design. In general the difficulties encountered when linking reliability and optimal design are associated to discontinuities in the derivatives of limit state functions. Furthermore it is possible to find multiple most probable failure points (MPPs). To avoid these problems it is proposed a new methodology based on evolutionary strategy aiming the global most probable failure point (MPP) search for composite structures under dynamic response.