The Estimation of the Dynamic Modulus of Asphalt Mixtures Using Artificial Neural Networks

The dynamic modulus is the main input material property of asphalt mixtures for the modern mechanistic-empirical asphalt pavement design methods. The dynamic modulus is determined in laboratory by different procedures but in all cases, they require sophisticated equipment and well-trained personnel. When these experimental results are not available, they could be estimated using different predictive models based on the aggregate gradation, volumetric properties of the mixture and binder characteristics. This paper presents the application of the Artificial Neural Network (ANN) technique in order to develop a robust prediction model of the dynamic modulus of asphalt mixtures. The experimental data used for the training and validation processes were collected from different construction projects in Argentina. The measured and estimated dynamic modulus results using the ANN model were compared and discussed showing that the ANN model developed in this study is promising to estimate the dynamic modulus of bituminous mixes for practical applications.