Predicting the chloride diffusion coefficient and surface electrical resistivity of concrete using statistical regression-based models and its application in chloride-induced corrosion service life prediction of RC structures

Abstract

The chloride diffusion coefficient and surface electrical resistivity (SER) are two most important characteristics of concrete durability. This study aims to develop prediction models of the rapid chloride migration tests (RCMT) and SER results according to the concrete mix properties and their correlation. The proposed models can be employed in the designing phase for service life prediction and in the construction phase to control the quality of the concrete. In this paper, the results of the RCMT and surface electrical resistivity (SER) test (Wenner method) of concrete specimens at the reference age (28 days) were collected, and statistical models were prepared to predict the RCMT and SER based on the concrete mix properties. Furthermore, the correlation of the RCMT results and SER results is shown, which can be practical in using surface electrical tests as an alternative for RCMT, due to its simplicity in the design and quality control phase. In order to validate the model and compare its precision against other prediction models, a dataset of RCMT resulting from other studies was gathered. Then, the prediction performance of different models was compared with each other. It was observed that the model proposed in this study makes more precise predictions in chloride diffusion coefficient than the models proposed by other authors. Finally, the predicted and experimental chloride diffusion coefficients along with the service lives of RC structures with different concrete cover and chloride diffusivity were estimated using a full probabilistic model. The results indicated that the calculated corrosion initiation probability using the predicted and experimental chloride diffusion coefficient was almost similar, demonstrating the applicability of the predicted chloride diffusion coefficient in service life prediction.