Advancement of Performance Test Methods of Bituminous Mixtures for Practical Implementation and Fundamental Understanding of Deformation and Fracture

Abstract

Design and material selection of bituminous mixtures (asphalt concrete) have a decisive role in mixture and pavement performance. Traditionally, the design of the mixtures has focused on the volumetric fractions of components. However, recently there have been an increase in the use of additives and recycled asphalt pavement (RAP) materials that can significantly affect the characteristics of a bituminous mixture with minimal changes to the volumetric fraction. As a result, performance-based mixture design whose key components are performance tests and performance criteria has been proposed for designing mixtures for performance rather than volumetrics. There are numerous performance test methods whose overabundance and complexity lead to results that often contradict each other. To tackle the challenge, this study first examined two test methods, the semicircular bending and gyratory stability tests for the primary distresses of fracture and permanent deformation, respectively. The experimental-statistical approach was used to determine the recommended values for critical testing variables that are repeatable and practical. The two performance tests were then used to successfully perform performance-based mixture design on high-RAP mixtures with different types of rejuvenators and dosages. Secondly, to enhance the understanding of inelastic deformation and fracture in mixtures, this study proposed an inverse method based on optimizing local displacements from digital image correlation (DIC) and finite element method (FEM) simulation. The inverse method was utilized to determine constitutive and fracture properties of example materials that are elastic (polyetheretherketone, PEEK) and viscoelastic (fine aggregate matrix, FAM). A three-point bending configuration was selected for testing coupled with the DIC. MATLAB module was used to calculate the objective function of displacement difference from both experimental and numerical results. The objective function was then minimized using derivative-free non-linear optimization. The DIC-FEM inverse method produced reasonable elastic modulus of PEEK, viscoelastic relaxation modulus of FAM, and cohesive zone fracture characteristics of FAM, while fracture toughness of PEEK requires further examination as PEEK showed a brittle fracture and was not fully characterized by the DIC in this study. Further studies that employ the developed DIC-FEM inverse method for heterogeneous materials such as asphalt concrete mixtures are recommended as the method can capture important local phenomena within the mixture. This enhanced understanding can help develop the mixture performance tests in a more accurate and efficient manner.