Micromechanical Modeling Approach Of Asphalt Mixture Based On The Influence Of Aggregates On Simulation Solution And Error Estimation

The micromechanical method can effectively explore the mechanical behavior of asphaltic mixture components.In the micromechanical simulations of asphalt mixtures,the complexity of coarse aggregates and asphalt mortar in shape needs a lot of finite elements to represent,which makes the simulation very time consuming and brings high risk to fail.To reach a balance between solution accuracy and computational cost,it is necessary to simplify the morphology of asphaltic mixture components.To reach this goal,this paper proposes a methodology to simplify the morphology of coarse aggregates as well as asphalt mortar based on three-dimensional(3-D)reconstruction of aggregates from X-ray computed tomography(CT)imaging.In this procedure,the influence of aggregates on simulation solution is taken into consideration.Furthermore,solution error due to the morphological simplification of asphaltic mixture components is quantified with finite element interpolation theory.Specifically,the main contents of the presented work are as follows:(1)The aggregates in mixtures are 3-D reconstructed using computer graphics and image processing techniques.Specifically,two-dimensional(2-D)profiles of aggregates in X-ray CT images are detected to be mapped to aggregates for further reconstruction of 3-D aggregates.The 3-D modeling was conducted using ACIS,the modeling toolkit developed by Spatial Cooperation.(2)The algorithm is developed to simplify the geometry of aggregate and asphalt components.Surface triangulation of aggregate model was conducted to obtain the clusters of approximately-flat regions on surface by parameters.Based on that,topological relationship between adjacent clusters is identified,surface texture of aggregate models was removed by self-developed algorithms to obtain the polyhedral representations of aggregate models.The model of asphalt mortar is then obtained by Boolean operations.(3)The influence index of an aggregate was presented and quantified based on a trial simulation,which provides the criterion to simplify the aggregate morphology.Aggregates were divided into groups to simplify with different parameters.This facilitates a balance between solution accuracy and computational effort.The final simplified model is obtained.(4)The solution error due to the morphological simplification of aggregates was accurately estimated using finite element interpolation theory.Parameters of simplification will be refined if the estimated error is larger than expected.Therefore,the simplification can be proved effective by ensuring the reliability of simulation.