Cellular Automata Model for Predicting the Failure Pattern of Laterally Loaded Masonry Wall Panels

Abstract

This paper introduces a technique that directly predicts the failure patterns of laterally loaded masonry panels based on the results of existing typical panels tested in the laboratory. The technique is based on the use of the cellular automata (CA). In this technique, the CA modeling is established to propagate boundary effects to zones within a panel. The corresponding rules for the state values of zones are derived from the proposed CA model, using appropriate transition functions. These state values are then used by the CA to establish zone similarity between two panels. Finally, the zone similarity is applied to establish locations of cracks on the panel. The technique is used in a novel way that eliminates the use of any numerical tools such as finite-element analysis (FEA). This technique is purely based on comparing the failure pattern of the base panel (a panel whose failure pattern is known from the laboratory tests) and unseen panels (panels not tested in the laboratory by the writers or with unknown failure patterns) subjected to the same type of loading and with similar boundary conditions to predict the failure load of the unseen panels.