Modelling of rutting behavior of modified polymer asphalt mixture using artificial neural network

rutting is one of the most important distresses in asphalt pavements, which in addition to reducing the service life of the pavement, poses serious risks to road users. Therefore, modeling the rutting behavior of asphalt mixtures is of great importance and if a suitable model is determined, the production of asphalt with a high probability of rutting can be prevented. The aim of this study was to determine the effect of bitumen Performance Grade, different percentages of SBS polymer, temperature, number of loading cycles, and percentage of air void on the rutting strength of asphalt mixtures by examining the creep behavior and using the neural network model. The results of this study showed that the use of polymer and PG64-22 bitumen reduces the cumulative strain and improves the paving resistance of the pavement and the best resistance of the sample contains 6% of polymer with PG64-22 bitumen. Also, by increasing the loading cycle the permanent deformations created in the pavement are significantly increased and the resistance of the mixture to ruttingis reduced. However, the addition of polymer has reduced this rate and improves the process of reducing the resistance and increasing the creep of asphalt mixtures. The results of the modeling showed that the neural network model has a very good performance in terms of predicting creep behavior and rutting strength of modified polymer asphalt mixtures.