ANISOTROPIC FAILURE CRITERION FOR BRITLE MATERIALS USING ARTIFICIAL NEURAL NETWORKS

Abstract eng:
Brittle materials are complex composite materials and a model is necessitated to realistically predict their load carrying capacity under any stress-state. During the last decades, a lot of failure criteria have been proposed for these materials in the literature. However, the majority of these models may substantially over-estimate the strength of the failure surface envelope. To overcome this problem a new approach is proposed which is based on Artificial Neural Networks (ANNs) techniques. The proposed approach is applied to a characteristic brittle anisotropic material such as the masonry material. The Neural Network managed to produce closed failure curves extending in all four quadrants of principal stresses, thus fully covering the compression and the tension stress areas. The comparison of the derived results with experimental findings demonstrates the promising potential of using ANNs for the reliable and robust approximation of the failure surface under biaxial stress state.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket