Accuracy Improvement for Two-Dimensional Finite-Element Modeling while Considering Asphalt Mixture Meso-Structure Characteristics in Indirect Tensile Test Simulation

Abstract

A two-dimensional finite-element simulation consumes less computational resources and has higher efficiency than a three-dimensional simulation. However, the existence of great differences between different slice images makes it difficult to represent the meso-structure of the whole specimen by a single or small number of images, while aggregate content and particle distribution have a significant impact on the results of two-dimensional finite-element simulations. To obtain accurate calculation results, it is necessary to solve a large number of two-dimensional models to suppress the errors by calculating the average value of the results. In this paper, based on a large number of slice images obtained by X-ray computed tomography (CT) scanning technology, two-dimensional finite-element models for indirect tensile test simulations were established while considering the meso-structure of asphalt mixtures. The relationship between the simulation results and the aggregate content was analyzed via an image processing method, and an aggregate content indicator was proposed and applied in the image selection process. According to the results, the use of the aggregate content indicator for the image selection process can greatly reduce the number of finite-element models and improve the calculation efficiency on the premise of ensuring the accuracy of the simulation results.