Regression Model for Prediction of Fracture Energy of Reclaimed Warm Asphalt Mixtures

Fracture energy is the resistance of asphalt mixture to cracking due to the effect of traffic loading and environmental conditions. The objective of this research is to predict a regression model for estimating fracture energy of reclaimed warm asphalt (RAP) mixtures. RAP is used with 20%, 40%, 60% and 80% by weight of aggregate mass. Petroleum wax is used to produce an ecofriendly asphalt mixtures. The used wax contents are 2%, 4%, 6%, 8%, 10%, 12% and 14%. The characteristics of the reclaimed warm asphalt mixtures were investigated by using Marshal test while the fracture energy was measured using the indirect Tensile Strength test. A regression model had been analyzed using the SPSS program based on the results of the fracture energy calculation as well as the mixtures characteristics tests. The model related the fracture energy as a dependent variable with the independent variables that included RAP percentage, wax percentage, temperature, Optimum Asphalt Content (OAC) and mix gradation. The predicted regression model is tested and analyzed for normality and collinearity and showed normal distribution for the residuals. The predicted regression model has R2 value = 0.543 and RMSE value = 362.040 J/m2 and MAE value = 261.473 J/m2. Also, the predicted model is validated by using the validation data set and it had succeeded in the test. The model analysis showed that the degree of temperature is significantly influences the fracture energy, so it is recommended to use the reclaimed warm asphalt mixtures in moderate temperature areas.