Corrosion induced crack patterns are influenced by many things. Rebar diameter, spacing, cover depth, and boundary conditions all play a role. In the videos, two different scenarios are compared: both specimens have 2 rebars of 10mm diameter, but one has a cover depth of 10 mm, while the other one has a cover depth of 20 mm. Crack growth in both specimens is shown. The difference is obvious.

However, in reality, corrosion rarely causes uniform expansion of the reinforcing steel. Therefore, it is possible that real pressure on the surrounding concrete acts in a different way (Ohtsu and Uddin, "Mechanisms of corrosion-induced cracks in concrete at meso- and macro-scales", Journal of Advanced Concrete Technology Vol.6 , No. 3, 419-429, October 2008):

The effect of different scenarios on the crack patterns in our model is visible in the following snapshots (at 5000 steps- yellow for visibility):

(top to bottom- radial, horizontal and vertical loading scheme, respectively)

It is clear that the loading condition makes a difference in the cracking behavior of the cover. Hope everyone is convinced! Enjoy the holidays!

When the reinforcing steel bar inside concrete corrodes, it starts expanding, since the volume of the oxides that form during the process is higher than the volume of the original steel. It, therefore, exerts pressure on the surrounding concrete, which eventually cracks. To study the mechanism, the lattice model can be used. Here is the result of the first simulation of corrosion induced cracking. Enjoy!

RCM test

Rapid chloride migration testing is increasingly used in practice to assess the chloride resistance characteristics of a concrete mix. It is fast and simple, and can be performed with no fancy equipment whatsoever. Standardized as NT BUILD 492 in Scandinavia, it is also increasingly popular in the Netherlands. But, is there a faster way of testing? Well, of course there is: a numerical simulation! If we could simulate the test, different concrete mixes could be tried and the best one chosen. The first simulation of the test is here. Enjoy!

Here is the comparison of our simulations with some experimental data from the literature: