Correlative Study Of Macroscopic And Mesoscopic Parameters Based On Mesostructural Asphalt Concrete

Asphalt concrete has complicated mechanical and thermal properties, so it is hard to use macroscopic parameter to quantitatively evaluate its properties. Mesostructures nowadays have been used to predict macroscopic material properties of asphalt concrete. In order to study influences of mesostructures and material parameters on macroscopic mechanical and thermal properties of asphalt concrete, this thesis presented a three dimensional aggregate packing algorithm. On the base of considering distribution laws of real aggregate volumes, a four-phase three dimensional mesostructural numerical modeling method containing cube aggregate, asphalt mastic, ITZ and air void was presented. Material parameters obtained by back calculation were brought into models to be analyzed. Thus, macroscopic mechanical and thermal properties of asphalt concrete were predicted.Firstly, according to packing algorithms of circle and square aggregate, two dimensional numerical models under different gradations were built. When back calculated parameters of asphalt mastic and parameters of aggregate were brought into numerical models, effects of aggregate shape, aggregate gradation and spatial position on compressive deformations on top of two dimensional models were studied. Secondly, based on the packing algorithm of two dimensional aggregate, an improved four-phase three dimensional numerical modeling method was presented. Cut planes of real specimens were respectively identified to count the number of different aggregate cross section areas, whose results were compared with that of cut planes in different numerical models. Thus, applicable scope of Walraven function and the reasonability of three dimensional numerical modeling method were verified. In order to predict macroscopic mechanical and thermal properties of asphalt concrete by using mesostructures and material parameters, three dimensional numerical models considering distribution laws of real aggregate volumes were respectively built. Material parameters of each phase were brought into models to implement simulations, whose results were respectively verified by mechanical and thermal experiment data. Finally, homogeneous models were used to predict mechanical and thermal properties of asphalt concrete. Results indicate that the numerical modeling method considering distribution laws of real aggregate volumes is reasonable. Mesostructures and material parameters can be used to effectively predict macroscopic mechanical and thermal properties of asphalt concrete.