It is becoming increasingly common to specify a service life for concrete structures of 100 years or more. However, we do not yet have the proper tool to be able to predict the longevity of concrete with any real accuracy, since we do not have much history with modern concretes, or adequate experience with short-term tests for the prediction of long-term behavior. This paper deals with the ambiguities in how we specify and verify service life, and the difficulties in establishing robust predictive models.
Subscribe to magazine