Modelling crack propagation in structures: Comparison of numerical methods

Crack propagation in concrete structures is a very complicated process, and the distribution of cracks may significantly affect the behaviour of the structures under time-dependent loading. If enough damage or extensive cracks exist in a structure, strengthening or repair using external carbon fibre-reinforced polymer (CFRP) reinforcement may be needed. Therefore, an understanding of the influence of the CFRP strengthening system on behaviour is crucial for the proper design of a structural reinforcement strategy. In this paper, we compare three major methods: the discrete crack method, the smeared crack method and the element-free method. By using these methods to study the fracture pattern of a beam with dapped ends, we can compare the capabilities and efficiencies of the three methods. When modelling the crack formation in reinforced concrete structures, the smeared crack approach gives better results than the discrete model in terms of crack spacing and regularity, however, it does not follow the crack growth well. The element-free Galerkin method seems to be superior to the other methods in the sense that it deals with the irregular cracks well and is very efficient in saving computing time. The preferred choice of method depends on the type of problem and the solution accuracy required, and this also depends on a certain balance between the accurate tracing of individual cracks and the computing efficiency. In addition, new algorithms are required for the efficient simulations of dynamic crack propagation, especially in the case of pre-cracked structures strengthened with prestressed CFRPs. Copyright © 2007 John Wiley & Sons, Ltd.