A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Using Artificial Neural Networks (ANNs) for Hot Mix Asphalt E* Predictions
The dynamic modulus (E*) is the main input to express the influence of climate and traffic speed and loading on hot mix asphalt (HMA) in the Mechanistic-Empirical Pavement Design Guide (MEPDG). However, E* samples’ preparation and testing require extensive time and costly equipment that is not usually available in regular laboratories. Thus, over years researchers developed models for E* predictions. The main purpose of this research is to compare the prediction accuracy of candidate E* predictive models and artificial neural networks (ANNs) model based on laboratory measured E* data. Laboratory E* tests were conducted on 25 Superpave mixes which were recruited from different ongoing construction projects in the Kingdom of Saudi Arabia (KSA). Dynamic shear rheometer tests (DSR) were conducted on the binders contained in the investigated mixes. ANNs trained using the laboratory measured E* data with the same set of parameters used in other popular predictive equations: the modified Witczak and Hirsch models were then used to predict E* values. The ANNs E* predictions using the modified Witczack model set of parameters show higher prediction accuracy compared to the regression models.
Using Artificial Neural Networks (ANNs) for Hot Mix Asphalt E* Predictions
The dynamic modulus (E*) is the main input to express the influence of climate and traffic speed and loading on hot mix asphalt (HMA) in the Mechanistic-Empirical Pavement Design Guide (MEPDG). However, E* samples’ preparation and testing require extensive time and costly equipment that is not usually available in regular laboratories. Thus, over years researchers developed models for E* predictions. The main purpose of this research is to compare the prediction accuracy of candidate E* predictive models and artificial neural networks (ANNs) model based on laboratory measured E* data. Laboratory E* tests were conducted on 25 Superpave mixes which were recruited from different ongoing construction projects in the Kingdom of Saudi Arabia (KSA). Dynamic shear rheometer tests (DSR) were conducted on the binders contained in the investigated mixes. ANNs trained using the laboratory measured E* data with the same set of parameters used in other popular predictive equations: the modified Witczak and Hirsch models were then used to predict E* values. The ANNs E* predictions using the modified Witczack model set of parameters show higher prediction accuracy compared to the regression models.
Using Artificial Neural Networks (ANNs) for Hot Mix Asphalt E* Predictions
El-Badawy, Sherif M. (author) / Khattab, Ahmed M. (author) / Al Hazmi, Al Abbas (author)
Fourth Geo-China International Conference ; 2016 ; Shandong, China
Geo-China 2016 ; 83-91
2016-07-21
Conference paper
Electronic Resource
English
Using Artificial Neural Networks (ANNs) for Hot Mix Asphalt E* Predictions
| British Library Conference Proceedings | 2016
Prediction of fracture toughness using artificial neural networks (ANNs)
| British Library Online Contents | 1997
Recycled Aggregates Concrete Compressive Strength Prediction Using Artificial Neural Networks (ANNs)
| DOAJ | 2021