This work is part of the DURACON Ibero-American project, which seeks to characterize concrete durability under environmental conditions, based on reinforced concrete sample exposure in at least two different atmospheres (marine and urban), for each of the 11 countries in the project. Specimens were exposed to the environmental conditions of 13 Mexican sites (8 urban and 5 marine atmospheres). Concrete specimens were 15 x 15 x 30 cm, with 6 rebars each, and three concrete covers (15, 20 and 30 cm). Two concrete mixtures were used with water/cement ratios of 0.45 and 0.65, respectively. Six reinforced and six plain concrete specimens were placed on each exposure site. Environmental data was collected on each exposure site, including rainfall, relative humidity, time of wetness, temperature, wind velocity, and carbon dioxide/chloride concentrations. Corrosion rates and potentials, as well as concrete resistivity were measured in the reinforced samples. Carbonation depths were measured on the plain ones. The present work focused on the measurements of environmental parameters during the first two years of exposure to analyze the potentiality and the probability of carbonation-induced corrosion, and the evaluation of the corrosion initiation period for the reinforcing steel on the 13 Mexican exposure sites.

This research evaluation consisted of a detailed statistical analysis of the recorded data in 72 reinforced concrete specimens, from 12 natural test sites (in nine countries) located in chloride-laden environments (marine airborne-exposure, test sites located between 50-250 m from seashore), during a natural exposure period of 10 years. The parameters evaluated included the concrete physical-mechanical characteristics; meteorochemical information; natural reinforcing steel´s instantaneous corrosion current density (icorr) and cumulative icorr; concrete chloride concentration; surface crack width and rebar cross section loss correlations. This statistical analysis resulted in empirical instantaneous icorr predictions as a function of the exposure microclimates, through linear multiple regressions. These models showed a high linear dependence of the cumulative icorr with the concrete capillary absorption as well as with the meteorochemical parameters. Results obtained in this investigation showed higher corrosion aggressiveness of tropical environments when compared to the non-tropical ones. The cumulative icorr proved to be an effective tool to indicate the corrosive likelihood and to differentiate the stages of Tuutti´s service life model.