Variability evaluation of gradation for asphalt mixture in asphalt pavement construction

Abstract Variability of gradation for asphalt mixture is always the difficulty faced by the control of pavement construction, involving the pavement performance. The main purpose of this study is to improve the efficiency of detection for gradation variation and clarify the relationship between gradation variation and construction process. The variability of gradation for superpave asphalt mixtures was investigated with nominal maximum aggregate size (NMAS) of 13 mm, 20 mm, and 25 mm in construction. The correlation between the passing percentages at each sieve size and performance of asphalt mixture was first established by coupling grey relational analysis and entropy weight method. A digital image processing technology with multi-thresholds segmentation (DIP-MTS) was developed and verified through the experiment comparison. An approach to reversely calculate the gradation of cold aggregate was established. The mixture gradations during various construction stages were compared. Results illustrate that the passing percentages of sieves with NMAS, 4.75 mm, and 0.075 mm have a significant influence on the pavement performance and should be paid more attention during construction. The detection results of DIP-MTS are closer to the experimental value, and this technology is capable of detecting gradation for moving aggregate. The reverse calculation approach is proved feasible. Gradation variation in transportation is more severe than other construction processes.