An analysis of existing models for carbonation coefficient applied to tests under natural conditions
Accepted version
Peer-reviewed
Repository URI
Repository DOI
Change log
Authors
Abstract
The carbonation exposure classes adopted by EN 206 enforce minimum cement contents for concrete mix design, therefore placing limits on achievable embodied carbon. Understanding the mechanisms behind carbonation of concrete structures would allow for more accurate specification of appropriately durable concretes, without overspecification of cement content. Various models for concrete carbonation exist in the literature, and these are grouped into: strength based, stoichiometric, microstructural, and empirical models. Examples of each class of model for concrete carbonation are compared against published data from natural carbonation tests to assess their accuracy and driving factors of carbonation resistance. The selected models are able to capture the general trends of the data with some reliability. However, the precision and accuracy of the estimates are low. It is possible that this loss of accuracy is a result of differences in the carbonation mechanism in accelerated carbonation tests, upon which the predicted models are typically calibrated, and natural carbonation processes, for which estimates are required to predict the response of real-world structures.