Engineering statistical models for carbonation depth

Based on generalized linear models and gamma degradation processes, two models predicting the natural carbonation of reinforced concrete (R.C) structures are proposed. They include the main factors influencing the carbonation process (material, environmental and curing factors) and their effects are evaluated from a large sample of natural carbonation depths measured experimentally worldwide. Then, predicted carbonation values are obtained using both the proposed models and the fib model for sheltered R.C structures exposed in the field. It is revealed that the proposed models have a rather high level of accuracy compared to the fib model and may be proposed as valuable alternatives.