To accurately quantify the variation in concrete carbonation depth, selecting an appropriate mathematical curve model is crucial. Currently prevalent models, such as the Fick model and exponential models, confront limitations in prediction accuracy and range of application. Given that a single curve model struggles to precisely describe the pattern of concrete carbonation, this work introduces a mixed-curve-based prediction model for carbonation depth, effectively integrating the Fick model with a hyperbolic model. Compared to the Fick model, the additional term in the mixed-curve model can be viewed as a reasonable correction to better adapt to the complex and varied conditions of concrete carbonation. This hybrid model transcends the limitations of individual models, enhancing fitting precision and broadening the scope of applicability. The new model boasts a concise structure with only two fitting parameters, facilitating ease of application. To validate its superiority, rigorous comparisons were conducted between the proposed model and existing ones, leveraging experimental data from 10 distinct concrete carbonation scenarios. By comparing the average error, standard deviation, and coefficient of determination across these cases, the new model demonstrates a clear advantage over the Fick model and the exponential model. In terms of fitting errors, the average error and standard deviation of the new model are notably lower than those of the other two models. In terms of the coefficient of determination, the values achieved by the new model in all examples are closer to 1 than those of both the Fick model and the exponential model, underscoring the new model’s superior fitting quality and remarkable stability. This research indicates that the combined model presented in this paper holds promising prospects for widespread application in predicting concrete carbonation depth.
