Behavioral prediction of materials degradation in load-bearing structures

Abstract

A new methodology is developed for the analysis and prevention of failure in load-bearing structures and machine components subject to several competing or complementary deterioration modes. Top-down modeling is used to integrate several specialist disciplines into a multidisciplinary model. Both mechanistic and probabilistic aspects of the system under study are analyzed. Constitutive equations governing static equilibrium behavior of the structure are decoupled from the deterioration (irreversible) phenomena and separate formulations developed for the equilibrium and irreversible phenomena. The overall mechanistic formulation for the physical system under study is then obtained by using the formulations for the deterioration phenomena as feedbacks into transfer functions describing the equilibrium phenomena. This mechanistic formulation is amenable to solution by numerical-analog computer codes. An overall dynamic probabilistic formulation is then obtained directly from the mechanistic equations rather than independently by the reverse, bottom-up approach. The resulting set of dynamic probabilistic formulations is also amenable to solution by numerical-analog techniques. The model and the methodology are implementable for a great variety of analyses including design, materials selection, inspection optimization, information management schemes and cost-benefit analyses. Failure modes treated in this study include corrosion, corrosion fatigue, brittle fracture, and deflection using an example from the papermaking industry.