Investigating the ANN model for cracking of HMA in terms of temperature, RAP and fibre content

Using high contents of the RAP material in newly produced asphalt mixtures leads to a reduction in the crack resistance of the mixtures. This shortcoming can be compensated using some methods such as polymer modification, rejuvenation, fibre-reinforcement, etc. In this paper, the effect of Cyclogen as a rejuvenator and glass fibre was investigated on the cracking behaviour of recycled asphalt mixtures. For this aim, the fracture energy and critical value of J integral (Jc)of semi-circular bending (SCB) asphalt specimens in terms of RAP content, Fibre content and testing temperature were investigated in this study and the results were used to develop an experimental model to predict the crack resistance of asphalt mixtures in terms of temperature and fibre and RAP content using the artificial neural network (ANN) and the multiple regression methods. The results indicated that the fracture energy and Jc valueof asphalt mixtures declined by increasing the RAP contentat intermediate temperatures. However, the reduction in fracture energy and Jc valuecould be compensated using the fibres. Both investigated models were appropriate for predicting the cracking behaviour of fibre-reinforced recycled asphalt mixtures. Nevertheless, the level of correlation was much higher for the ANN approach than the multiple regression model.

Automated summary

This paper investigates the cracking behavior of recycled asphalt mixtures and finds that using high contents of RAP material reduces the crack resistance of the mixtures. This shortcoming can be compensated by using Cyclogen as a rejuvenator and glass fiber. The crack resistance of asphalt mixtures can be predicted using artificial neural network (ANN) and multiple regression methods, with the ANN approach showing a higher level of correlation.