54 Pearmeability Prediction Model for Concrete Mixes Used in Kansas PCC Pavements

To properly characterize the permeability response of PCC pavement structures, Kansas Department of Transportation (KDOT) generally runs the Rapid Chloride Permeability test (RCPT) to determine the resistance of concrete to penetration of chloride ions. RCPT typically measures the number of coulombs passing through concrete samples over a period of six hours at a concrete age of 7, 28, and 56 days. In this study, back-propagation Artificial Neural Network (ANN)- and Regression-based permeability response prediction models for Rapid Chloride are developed by using the database provided by KDOT in order to reduce the duration of the testing period. Backprop ANN learning technique proved to be an efficient methodology to produce relatively accurate permeability prediction models. Comparison of the prediction accuracy of the developed models proved that ANN models have outperformed their counterpart regression models. Developed ANN-Based permeability prediction models are effective and applicable in characterizing the permeability response of concrete mixes.