<abstract xmlns="http://www.w3.org/1999/xhtml"><p>Rutting occurs due to accumulation of incrementally small permanent deformations from each load application and it can cause irreparable problems in pavements. On the other hand, the Marshall Mix design which is known as the main method in Iran, the lack of a simple test to determine specimen resistance to permanent deformation as the major reason for asphalt rutting is noticeable. Although today many devices are used for rutting measurement, none of them have the ability to use in wider field. In addition, prevalent methods of evaluating rutting potential are usually costly and time consuming. Mentioned parameters illustrate the necessity of developing a simple method, not only having fine precision, but also are able to predict rutting performance with low cost in the short term in laboratory. In this research, after performing the main tests on specimens, IDT test results and Marshall Parameters were used to develop a mathematical model to estimate specimen rut depth. The model is validated by using ANN and makes it possible to evaluate mixtures rutting potential while OBC is being determined in laboratory. So not only is there no need to use expensive instruments of rutting test, but also a remarkable time saving in mix design procedure is achievable.</p></abstract>

Rutting occurs due to accumulation of incrementally small permanent deformations from each load application and it can cause irreparable problems in pavements. On the other hand, the Marshall Mix design which is known as the main method in Iran, the lack of a simple test to determine specimen resistance to permanent deformation as the major reason for asphalt rutting is noticeable. Although today many devices are used for rutting measurement, none of them have the ability to use in wider field. In addition, prevalent methods of evaluating rutting potential are usually costly and time consuming. Mentioned parameters illustrate the necessity of developing a simple method, not only having fine precision, but also are able to predict rutting performance with low cost in the short term in laboratory. In this research, after performing the main tests on specimens, IDT test results and Marshall Parameters were used to develop a mathematical model to estimate specimen rut depth. The model is validated by using ANN and makes it possible to evaluate mixtures rutting potential while OBC is being determined in laboratory. So not only is there no need to use expensive instruments of rutting test, but also a remarkable time saving in mix design procedure is achievable.

Predict of Asphalt Rutting Potential Based on IDT and Validation with ANN

Department of Civil Engineering, Nowshahr Branch

Journal of Applied Engineering Sciences

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

{"article-title":"Predict of Asphalt Rutting Potential Based on IDT and Validation with ANN"}

Rutting occurs due to accumulation of incrementally small permanent deformations from each load application and it can cause irreparable problems in pavements....

Predict of Asphalt Rutting Potential Based on IDT and Validation with ANN

Published Online: Dec 21, 2019

Page range: 131 - 138

Received: Aug 06, 2019

Accepted: Sep 30, 2019

DOI: https://doi.org/10.2478/jaes-2019-0018

Keywords
IDT, Rut Depth, Vertical Deformation, Artificial Neural Network

© 2019 Divandari Hassan, published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Predict of Asphalt Rutting Potential Based on IDT and Validation with ANN

Published Online: Dec 21, 2019

Page range: 131 - 138

Received: Aug 06, 2019

Accepted: Sep 30, 2019

DOI: https://doi.org/10.2478/jaes-2019-0018

KeywordsIDT, Rut Depth, Vertical Deformation, Artificial Neural Network

© 2019 Divandari Hassan, published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Keywords
IDT, Rut Depth, Vertical Deformation, Artificial Neural Network