Field Calibration of Shrinkage Crack Width Model for CRCP in Tropical Climate

Abstract

Crack width (CW) is one of the most important parameters to be considered in performance evaluation and prediction for continuously reinforced concrete pavements (CRCP) because it affects the load transfer efficiency across the cracks themselves. Therefore, accurate CW prediction during CRCP design is very important to anticipate pavement performance or even for changes regarding variables such as the longitudinal steel content. This study brings together CW data collected on the first conventional CRCP constructed in Brazil, using different concrete mixtures as well as natural and recycled aggregates and two cement types. A modified mechanistic-empirical pavement design guide (MEPDG) model for cracking measurement on any depth was used to predict CW. The results were compared with those obtained from the field in a pavement with two types of steel reinforcement (black and galvanized), allowing a local calibration. This is required for introducing CRCP technology in tropical countries like Brazil considering that MEPDG can be adjusted, calibrated, and implemented for local conditions. As the models require the shrinkage strains for CW prediction, up-to-date general models were tested to predict the shrinkage strains seeking the better fit for field data. Such shrinkage models include ACI 209R-92, Eurocode 2 (EC2), fib Model Code, Model B3 (used by MEPDG), and the Brazilian one specified in the NBR 6118 standard. The better CW fit for the raw data was obtained through Eurocode 2 model for the galvanized continuously reinforcement case. For black steel reinforcement, the crack opening was overestimated when calculated by any model except B3 and ACI 209R-92. It was also observed that the cement type thoughtlessness in some models inflicted significant differences from field data when not calibrated previously.