Time-dependent model for in-situ concrete carbonation depth under combined effects of temperature and relative humidity

Fluctuations in temperatures and relative humidity in the external environment significantly influence the carbonation depth in concrete structures, making it an essential topic when considering concrete durability. This study aims to predict and compare the carbonation depth of concrete under various temperature and relative humidity conditions through both experiment and modeling. The experimental results demonstrate an exponential relationship between carbonation depth and temperature at all ages of the concrete, showing a consistent increase, with rising temperature. Additionally, the carbonation depth of concrete exhibits a parabolic relationship with relative humidity, characterized by an initial increase followed by a subsequent decrease, forming a downward-opening curve. This relationship is well-corrected with the data. Moreover, sensitivity analysis reveals a noticeable variation in sensitivity factors related to temperature and relative humidity after a 28-day carbonation period, with the sensitivity factor for temperature being higher than that for relative humidity. Finally, by employing the Sensitivity Analysis and Least-Square Fitting (SA-LSF) method, a novel model is developed that incorporates both temperature and relative humidity as influential factors in predicting carbonation depth under diverse environmental conditions. The experimental results align well with the model’s predictions, confirming its ability to accurately forecast carbonation depth of concrete. These findings provide significant insights for practical applications, as the model can give valuable predictions for carbonation in concrete. • A model predicts carbonation depth under varying temperature and humidity. • Carbonation depth shows exponential growth with temperature, peaking at 40 °C. • Relative humidity influences carbonation depth parabolically, peaking at 55–70 %. • Sensitivity analysis identifies temperature as the dominant factor over time. • The model aligns well with data, offering accurate predictions for engineering use.