| Rutting is one of the main distresses of asphalt pavement in China.Its deformation is mainly contributed by the permanent deformation of asphalt mixture layer.The rutting resistance of asphalt mixture mainly depends on the performances of and interactions between components for asphalt mixture,especially the interaction between mineral aggregates.However,for a long time,researchers have paid more attention to the contact or interlocking conditions of mineral aggregates,lacking a unified understanding on the overall spatial distribution of and interaction between mineral aggregates.Asphalt mixture is a kind of multiscale composite material.The particle interaction phenomenon exists in all scales of asphalt mastic,asphalt mortar and asphalt mixture.As an important mechanical performance index of asphalt mixture to resist deformation,dynamic modulus has become a key multi-scale index that connects the micro-structure and macro-performance of asphalt mixture.Therefore,it is necessary to investigate the correlaiton mechanism between the distribution characteristics of micro-structure and the dynamic modulus performance and creep behavior of asphalt mixture.For this purpose,in this study,the dynamic modulus and creep compliance of asphalt mixture are predicted by combining probability statistics and decoupling hierarchical multi-scale analysis methods according to the characteristics of aggregate particles in terms of distribution and interaction in spatial scale.Main works are as follows:First of all,a new understanding of the stiffening effect of filler on asphalt mastic was conducted through the statistics theory according to the cooperative behavior of particles.Based on the nearest surface function,probability statistics theory,micro-mechanics and matrixparticle interaction,a general technical framework for analyzing the stiffening mechanism of asphalt mastic was constructed to accomplish the utilization of the micro-structure information within asphalt mastic.According to the mechanism of asphalt-filler interaction,the concepts of permeable transition zone and interpenetrating filler particles were introduced,and a new method to calculate the filler critical volume fraction was established,which takes into account the distribution characteristics of filler and breaks through the limitations of empirical fitting of existing conceptual models.The statistical distribution of filler in asphalt mastic was obtained through the nearest surface quantities of filler in asphalt mastic,and the reinforcement mechanism of filler particles interaction was revealed from a statistical perspective.Secondly,three viscoelastic micro-mechanical models(DM,MTM,SCM)were improved by combining the micromechanics homogenization equivalence with the fractal statistics principles and considering the distribution characteristics of filler.Based on the improved MTM and SCM,a combined SC-MT model was established by a two-step homogenization method.Based on the rheological test results of asphalt mastics with different filler volume fractions,the applicability of the model was verified,and the meaning of model parameter was analyzed.The distribution characteristics of dynamic shear modulus of asphalt mastic and its reinforcement effect with dimensionless particle size were predicted by using the four models,and the influence of different parameters on the distribution characteristics was analyzed.Results show that the SC-MT combined model can effectively predict the dynamic shear modulus of asphalt mastic.The larger the filler volume fraction,the more significant the influence of filler particle size on the dynamic shear modulus of asphalt mastic,meaning that the more significant the particle effect in the reinforcement mechanism of mastic.Thirdly,according to the spatial distribution characteristics of filler particles in asphalt mastic,a homogenization equivalent scheme of asphalt mastic considering the nearest surface spacing was established.The local fluctuation characteristics of filler volume fraction were analyzed by using Vorono(?) diagram and Stienen model,and the relevant volume fractions of the equivalent model was resolved.Based on this,the model(MH-MRP)to predict the dynamic shear modulus of asphalt mastic was deduced and established.The distribution coefficients of filler particles at each grade were determined by using the nearest surface function.The master curve of dynamic modulus of asphalt mastic was established by using the standard Sigmaid function and Williams-Landel-Ferry equation.Based on the test results of dynamic modulus for asphalt mastics with different filler volume fractions,the applicability of the model was verified,and the predicted results of MH-MRP model and traditional models(GSC model,MT model,DEMT model)were compared and analyzed.Results show that there is a great difference between the distribution coefficients of each graded filler in asphalt mastic,and the distribution coefficient decreases significantly with the increase of filler volume fraction.MH-MRP model has the best prediction effect,and the prediction accuracies of all models decrease with the increases of filler volume fraction and temperature.Finally,based on the MH-MRP model of asphat mastic,the MH-MRP model of asphalt mastic was constructed through the multi-step homogenization method and by taking coarse aggregate and void as inclusions.Based on the nearest surface function and fractal gradation theory,a method to calculate the nearest surfaces distance between coarse aggregates was established.According to the design gradation of mineral aggregates in asphalt mixture,the nearest surface spacing between each graded coarse aggregates was calculated by the nearest surface function,and its variation law was analyzed.The applicability of the model was predicted and verified according to the results of the designed multi-scale dynamic modulus tests.Considering that the interaction between coarse aggregate particles,the influence of the nearest surface spacing between coarse aggregates on the creep behavior of asphalt mixture was investigated.Based on the improved MRP model and the viscoelastic constitutive relationship of asphalt mortar,a micro-creep model of asphalt mixture was established through the elasticity to viscoelasticity correspondence principle,and the predicted results were compared with the tested results.Then the micro-factors affecting the creep behavior of asphalt mixture were analyzed.Results show that the MH-MRP model of asphalt mixture reflects the interaction characteristics of coarse aggregate particles and can well predict the effective dynamic modulus.The creep behavior of asphalt mixture is very sensitive to the distribution coefficient of coarse aggregate.The creep compliance decreases with the decrease of distribution coefficient and increases with the increase of void content.The higher the distribution coefficient of coarse aggregate is,the more sensitive the creep behavior to void content is.The creep prediction model is expected to realize the optimization design of gradation for coarse aggregate particles to improve the rutting resistance of asphalt mixture. |
PDF Full Text Request