Possibilities to validate design models for corrosion in carbonated concrete using condition assessment data

• A corrosion propagation phase model was compared with data from existing concrete facades.
• Propagation phase until cracking was calculated backwards from statistical data.
• The DuraCrete model overestimated the propagation phase in this carbonation initiated case.
• The time of wetness and concrete resistivity had a high influence to the model.
• Overestimation may also result from the limit state model for corrosion penetration.

This study analyses the actual observed corrosion damage and active corrosion in concrete facades in Nordic outdoor climate and compares this data with the DuraCrete design model for carbonation initiated corrosion propagation phase. Active corrosion was examined in this study by calculating backwards the age of the building and the initiation time by carbonation from a large condition assessment data from concrete facades and balconies using statistical simulation methods. The earliest visible corrosion damage on concrete facades was observed in condition assessments already after 8–15 years from construction. In a large group of buildings the damage occurrence was found to reasonably well follow a normal distribution. The design model was found in the case of concrete facades and balconies in Finnish climate conditions to overestimate considerably the propagation phase predictions in all of the studied structure types compared to the statistical analysis. The overestimation in the model is due to the high influence of concrete resistivity and the definition of corrosion penetration needed for the initiation of a crack.