An Artificial Neural Network Approach to Elastic-Creep Life Assessment

Abstract

This paper describes a new multiaxial model for predicting creep lives of structures. The model hypothesises that the creep damage is proportional to the creep internal energy density absorbed by the material. The authors argue that the model is the most appropriate for characterizing gross creep damage from a macroscopic point of view because it takes into accounts both the multiaxial internal deformation and loading. The verification and application of the model are demonstrated by applying it to various notched bars for which the experimental data are available. The results show that the proposed model is capable of predicting creep lives of the structures conservatively.