Flexible pavements, made up of asphalt concrete, are commonly used for highways all around the world. These pavements suffer from distresses like reflective cracking, rutting and fatigue cracking. Performance-based design methods have been adopted to enhance the service life of flexible pavements. Other measures to enhance the life cycle of these pavements include reinforcement with materials such as geosynthetics. However, there is a gap in the literature on development of performance-based design models for reinforced pavements. In this study, artificial neural network (ANN) models are developed for predicting flexural stiffness and rutting depth of reinforced asphalt pavements using design parameters from the simple laboratory procedures for Marshall and rut depth tests. A multilayer feedforward neural network (MLFNN) was found suitable in this study when a larger dataset was available with a flexural stiffness model. On the other hand, radial basis neural network (RBNN) was found to give higher accuracy with the smaller dataset of rut depth available in this study. In both cases, ANNs were found to predict the parameters with sufficient accuracy. These models show that reinforced asphalt designs with central gradation have the best design. The models developed in this study will be helpful to design long-lasting pavements with geosynthetic reinforcement without the requirement for high-tech testing facilities.


    Zugriff

    Download

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Artificial Neural Network Models for Performance Design of Asphalt Pavements Reinforced with Geosynthetics


    Weitere Titelangaben:

    Transportation Research Record: Journal of the Transportation Research Board


    Beteiligte:


    Erscheinungsdatum :

    06.06.2020




    Medientyp :

    Aufsatz (Zeitschrift)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch