Neural Network Based Model to Estimate Dynamic Modulus E* for Mixtures in Costa Rica

Document Type : Regular Article


National Center for Asphalt Technology, Auburn University, Auburn, United States


Several dynamic modulus (E*) predictive models of asphalt mixtures have been developed as an alternative to laboratory testing. The 1999 I-37A Witczak equation is one of the most commonly used alternatives. This equation is based on mixtures laboratory results in the U.S. In Latin American countries there are significant differences in material properties, traffic information, and environmental conditions compared to the U.S.; therefore, there is a limitation is the use of this equation using local conditions. The National Laboratory of Materials and Structural Models at the University of Costa Rica (Lanamme UCR) has previously performed a local calibration of this equation based on results from different types of Costa Rican mixtures. However, there was still room for improvement using advanced regression techniques such as neural networks (NN). The objective of this study was to develop an improved and more effective dynamic modulus regression model for mixtures in Costa Rica using Neural Networks. Results indicated that the new and improved model based on neural networks (E* NN-model) not only met the model adequacy checking criteria but also exhibited the best goodness of fit parameters and the lowest overall bias.


Google Scholar


Main Subjects

[1]     ARA I. ERES Consultants Division. Guide for Mechanistic–Empirical Design of New and Rehabilitated Pavement Structures. Final report, NCHRP Project 1-37A. Transportation Research Board of the National Academies, Washington, D.C. 2004.
[2]     Clyne TR, Li X, Marastenau MO, Skok EL. Dynamic and Resilient Modulus of Mn/DOT Asphalt Mixtures.“ Final Report MN/RC – 2003-09. University of Minnesota, Minneapolis, MN. 2003.
[3]     Birgisson B, Sholar G, Roque R. Evaluation of a Predicted Dynamic Modulus for Florida Mixtures. Transp Res Rec J Transp Res Board 2005;1929:200–7. doi:10.1177/0361198105192900124.
[4]     Kim YR, King M, Momen M. Typical dynamic moduli values of hot mix asphalt in North Carolina and their prediction. Transp. Res. Board 84th Annu. Meet. Compend. Pap., 2005, p. 5–2568.
[5]     Schwartz CW. Evaluation of the Witczak dynamic modulus prediction model. Proc. 84th Annu. Meet. Transp. Res. Board, Washington, DC, 2005.
[6]     Tran N, Hall K. Evaluating the predictive equation in determining dynamic moduli of typical asphalt mixtures used in Arkansas 2005.
[7]     Robbins MM, Timm DH. Evaluation of Dynamic Modulus Predictive Equations for Southeastern United States Asphalt Mixtures. Transportation Research Record. J Transp Res Board 2011;72:122–9.
[8]     Christensen Jr DW, Pellinen T, Bonaquist RF. Hirsch model for estimating the modulus of asphalt concrete. J Assoc Asph Paving Technol 2003;72.
[9]     Singh D, Zaman M, Commuri S. Evaluation of Predictive Models for Estimating Dynamic Modulus of HMA Mixtures Used in Oklahoma 2010.
[10]    El-Badawy S, Bayomy F, Awed A. Performance of MEPDG dynamic modulus predictive models for asphalt concrete mixtures: local calibration for Idaho. J Mater Civ Eng 2012;24:1412–21.
[11]    Loria LG, Badilla G, Jimenez Acuna M, Elizondo F, Aguiar-Moya JP. Experiences in the Characterization of Materials Used in the Calibration of the AASHTO’Mechanistic-Empirical Pavement Design Guide (MEPDG) for Flexible Pavement for Costa Rica. 2011.
[12]    Far MSS, Underwood BS, Ranjithan SR, Kim YR, Jackson N. Application of Artificial Neural Networks for Estimating Dynamic Modulus of Asphalt Concrete. Transp Res Rec J Transp Res Board 2009;2127:173–86. doi:10.3141/2127-20.
[13]    Ceylan H, Schwartz CW, Kim S, Gopalakrishnan K. Accuracy of Predictive Models for Dynamic Modulus of Hot-Mix Asphalt. J Mater Civ Eng 2009;21:286–93. doi:10.1061/(ASCE)0899-1561(2009)21:6(286).
[14]    Priddy KL, Keller PE. Artificial neural networks: an introduction. vol. 68. SPIE press; 2005.