Study On Homogeneity And Optimum Asphalt Content Of Asphalt Mixtures Based On CT Technique

Asphalt mixture is a quite complex heterogeneous composite. For a long time, the traditional design system on the macroscale behavior of asphalt mixtures mainly depends on empirical methods, but it requires quantities of time-consuming and costly laboratory tests, and the tests results present considerable difference and poor reproducibility, which makes it unable to meet the requirements of high quality in the modern engineering construction. The macroscale behavior of asphalt mixtures has a close correlation with internal structural characteristics and inclusions behaviors. Thus, it’s quite essential to study on macroscale behavior of asphalt mixtures from the microscale on base of digital image processing.Firstly, the digital image processing of the X-ray computed tomography (CT) image of asphalt mixtures were exposed in this study. Furthermore, a partition method of nine concentric annulus was put forward to confirm the thresholds of coarse aggregates and air voids in the CT image. The results showed that the characterization precision of coarse aggregate and air voids was improved, the characterization loss of coarse aggregates in the edge of CT image was eliminated, and the phenomena of pinhole inside large size mineral aggregates was reduced using the partition method.Secondly, the asphalt mixture specimens produced two compaction methods, three gradations of aggregate, and multi-compaction efforts were identified on base of CT technique. A three-segment partition method along the height of the specimens was put forward to analyze the statistical volume of air voids of internal structure of asphalt mixtures. A prediction model of overall calculated volume of air voids of asphalt mixtures was established by using three characteristic indices of volume of air voids in the middle segment of specimens, and moreover the prediction accuracy of the model were examined. Meanwhile, the discrepancy between the volume of air voids obtained from the prediction model and three-dimension restruction model of asphalt mixture and the measured volume of air voids by laboratory test was analyzed. The results showed that the aggregates gradation had a considerable influence on the validity of overall calculated volume of air voids, whereas the compaction efforts and compaction methods had less influence on it. The prediction accuracy of the model reached 98.98% to the overall calculated volume of air voids. Compared with the measured volume of air voids of asphalt mixture, the volume of air voids of AC and SMA mixtures obtained from the prediction model was lower than that of OGFC mixtures, and the volume of air voids between obtained from the prediction model and that from the three-dimension restruction model exhibited a slight discrepancy.In addition, to test and evaluate homogeneity of internal structure of asphalt mixtures effectively, a homogeneity identifying method based on specific gravity of inclusions (HIMDI) was proposed in consideration of the intraction of coarse aggregate, mastics, and air voids. Asphalt mixtures specimens including three groups of asphalt-aggregate ratios were utilized to analyze the homogeneity of internal structure of the specimens. The effects of the asphalt content, compaction efforts and gradation of aggregate on the homogeneity of internal structure of asphalt mixtures were analyzed, and the actual application of the HIMDI was also discussed. The results showed that the homogeneity of internal structures of single or a set of asphalt mixtures specimens could be characterized through analyzing the homogeneity evaluation parameters proposed by the HIMDI, and the differences between the optimum asphalt-aggregate ratio of asphalt mixtures calculated by the HIMDI and the measured results were 0.07%,0.1%, and 0.55%. The research revealed that the asphalt content and gradation of aggregates had a significant effect on the homogeneity of internal structure of asphalt mixtures, and compaction efforts had a distinction effect on it. The asphalt content, the unsubtantial area of homogeneity, and the parallelity of specimens prepared using same gradation of aggregate of asphalt mixtures were identified by using the evaluation parameter of homogeneity.Besides, based on the prediction model of the volume of air voids and the HIMDI of asphalt mixtures proposed in this study, and taking homogeneity of internal structural of asphalt mixture into consideration, two methods of optimum asphalt content of AC and SMA mixtures were established in terms of the relationship between the volume parameter from Marshall method and the characteristic value of inclusions of internal structural. The results indicated that characteristic value of inclusion of internal structural of mixtures determined by CT technique and DIP can accurately represent volume parameters of AC and SMA mixtures in Marshall method, such as VV, VMA and VFA. For AC and SMA mixtures, the difference between the optimum asphalt content of asphalt mixture obtained by the method proposed in this study and that from Marshall method, were 0.01% and 0.179%, repectively.Finally, using Microsoft Visual Studio and Matlab software by VC++ language, a software programed on volume of air voids prediction, homogeneity evaluation and of calculation on optimum asphalt content was developed based on the HIMDI. Some examples were tested and analyzed using this software. The operation and calculation results e showed this software can meet the design requirements.