Abstract:

In this study, the rheological properties of the bitumen modified with Magnetic Activated Carbon (MAC) were investigated and the results were predicted with artificial neural networks. Within the scope of the study, modified bitumens were obtained by adding 5%, 10% and 15% MAC to the B160/220 penetration grade bituminous binder. Then, a frequency sweep test at ten different frequencies (0.01-10Hz) and four different temperatures (40 °, 50 °, 60 °, 70 ° C) was carried out on original and modified bitumens with the Dynamic Shear Rheometer (DSR) device. The results showed that the addition of magnetic activated carbon improves the elastic properties of the binder by increasing the complex modulus and decreasing the phase angle values. Then, complex modulus and phase angle values, which vary depending on the frequency, addition content and temperature value, were predicted by artificial neural networks. The results showed that complex modulus and phase angle values of the bitumen can be obtained with very high accuracy and low error with artificial neural networks.

Keywords: