MAPPING CRACKING PATTERN OF MASONRY WALL PANEL BASED ON TWO-STEP CRITERION FOR MATCHING ZONE SIMILARITY

摘要

This paper puts forward an innovative criterion in the cellar automata (CA) technique for both matching zone similarity and mapping failure pattern of masonry wall panel. The criterion proposed in this paper is a two-step matching method. Firstly, calculate the state values of the cells in the base and unseen panels using the CA technique. Then, the first step of the criterion is to take a cell and its eight neighbourhoods as a data block and choose all the best-match blocks within the base panel corresponding to a data block within the unseen panel, according to a proposed minimum risk principle. Finally, map cracking patterns of unseen panels use the criterion for mapping cracking pattern. The cracking patterns of unseen panels are mapped using the tested cracking patterns of several simply-supported base panels and the methods developed above. The mapped results are verified by the corresponding experimental results. The proposed criterion for matching zone similarity can greatly improve the existing CA technique for mapping the cracking pattern of an unseen panel; particularly, the convergence of the improved CA technique obtains a great improvement. Also, this mapping task is realized on the basis of the fine CA cell lattices of the panels, using the proposed method.