Model developments of long-term aged asphalt binders

Artificial neural networks (ANNs) are useful in place of conventional physical models for analyzing complex relationship involving multiple variables and have been successfully used in civil engineering applications. The objective of this study was to develop a series of ANN models to simulate the long-term aging of three asphalt binders (PG 64-22, crumb rubberized asphalt modifier, PG 76-22) regarding seven aging variables such as aging temperature and duration, m-value, mass loss of pressurized aging vessel (PAV) samples, percentages of large and small molecular sizes of high pressure-gel permeation chromatographic (GPC) testing, and binder stiffness. The results indicated that ANN-based models are more effective than the regression models and can easily be implemented in a spreadsheet, thus making it easy to apply. The results also show that the aging temperature, aging duration, percentage of large and small molecular sizes, and binder stiffness are the most important factors in the developed ANN models for prediction of penetration index after a long-term aging process.