Novel Model to Predict Critical Strain Energy Release Rate in Semi-Circular Bend Test as Fracture Parameter for Asphalt Mixtures Using an Artificial Neural Network Approach

Growing use of recycled asphalt materials in pavement means the current volumetric-based Superpave mixture design may not address durability concerns arising from replacement a proportion virgin binder with ones. To this limitation, performance-based testing is introduced to supplement conventional volumetric assessing cracking performance mixtures. Louisiana Department Transportation and Development’s Specifications for Roads Bridges specify criterion critical strain energy release rate, J c , obtained semi-circular bend (SCB) test as complement practice evaluate resistance Quality control/assurance practices, however, require SCB samples be long-term aged five days at 85°C, which time-consuming process. Therefore, it beneficial able estimate mixtures based on measured plant-produced Asphalt aging complex, various variables are involved process, including properties chemical/rheological characteristics binder. With capability artificial neural network (ANN) complex relationships between input output variables, study aims predict fracture parameter, using ANN. A total 34 were selected study. tests chemical rheological characterization conducted. Stepwise regression analysis was used determine significant parameters correlation . determined parameters, ANN gradient descent backpropagation approach then applied develop validate predictive model. It shown that developed model more accurately than linear non-linear models.