3D finite-element modelling of splitting crack propagation

Although many experimental results on bond strength are available and splitting is commonly recognised as an important aspect of bond strength, only a few studies on splitting crack development have been published so far. Although, splitting is a 3D phenomenon, the only measurement easily taken in a lest is the splitting crack width on the specimen surface. As a result, the crack development inside the specimen is unmonitored. Such being the limits of testing, finite-element (FE) analysis can be a valuable alternative for the study of splitting crack 3D nature and for a better understanding of bond behaviour. However, FE studies are limited by the intrinsic difficulty in correctly modelling the bond. In fact, the models proposed so far are either too complex or too limited for describing actual anchorages and for studying splitting development. The main purpose of this research work is to develop a 3D FE model in which bond is modelled by a simple kinematic interface. The model developed is useful for establishing rational design criteria for anchorages and splices. Such a model is validated by simulating a series of pull-out tests concerning the interaction between a bar and a stirrup, when the concrete cover is split and the stirrup crosses the splitting plane. The comparison between experimental and numerical results shows that the proposed model describes quite well the behaviour of an anchored bar in terms of both bond-slip response and splitting crack width. In spite of its simplicity, the proposed model can be easily expanded to cover more complex structural cases and to develop more rational criteria for the design of anchorages and splices in RC members.