Which rust deposition model should be used in predicting concrete cover cracking due to reinforcement corrosion?

The critical amount of corroded steel that causes concrete cover cracking can be readily calculated based on thick-walled cylinder theory. However, the results may vary significantly depending on how the rust deposition is considered. There are several rust deposition hypothesis proposed in the literature for modelling concrete cover cracking of RC structures due to reinforcement corrosion. Among them, three are considered representative ones and have been widely cited in the literature. They are: (i) assumes a certain amount of rust product carried away from the rust layer and deposited within the open cracks proposed by Pantazopoulou and Papoulia; (ii) assumes all of the rust products build up around the bar and all of them are responsible for the expansive pressure proposed by Bazant; (iii) assumes certain amount of rust products deposited into a porous zone around the bar/concrete interface proposed by Liu and Weyers. In this paper, all three rust deposition hypotheses were examined for the critical amount of corrosion to induce cover cracking. When compared to the test data available from the literature, it showed that the porous zone model proposed by Liu and Weyers gives the best predictions. Thus it may be concluded that assuming a porous zone around the steel/concrete interface would be reasonable and may be adopted in developing concrete cover cracking predictive model.