Page 370 - Mechanics of Asphalt Microstructure and Micromechanics

T
362 Ch a p t e r e n

In some cases it is even worse as different testing devices or procedures may yield con-
flicting results.
Mix design methods typically include the following steps: 1) selection of binder and
aggregate; 2) characterization of binder and aggregate; 3) selection of a gradation; 4)
determination of the optimum AC content and volumetric characteristics; and 5) evalu-
ation of the performance of the mixtures. Once the materials (asphalt binder and ag-
gregate) are selected, the most important steps are the determination of the gradation,
the optimum asphalt content, and evaluation of the performance through mechanical
testing such as the simple performance test (SPT) or the simulative test (ST). If results of
the mechanical tests do not meet the specifications, another gradation shall be tried.
This process typically requires two weeks. Nevertheless, even if the mechanical test
results meet the specification requirements, the mixes may not perform satisfactorily
when placed in the field. This is a very challenging problem: is there any way by which
a mix can be designed with predicted performances of reasonable accuracy?
Before answering this question, the important performance criteria should be ana-
lyzed. Historically, performance has been measured in terms of PSI (Present Service
Index), roughness index, and individual distresses such as rutting, cracking (thermal
cracking, fatigue cracking), and moisture damage. These distresses have been demon-
strated. However, it was still not clear what had happened inside the mixture even after
the SuperPave mix design was developed. In recent years, several research projects
have focused on the understanding of the fundamental mechanisms of the develop-
ment of these distresses. A common feature of these projects is to investigate what fac-
tors truly control the development of these distresses. As AC is a highly heterogeneous
material composed of aggregates, asphalt binder, and air voids, stresses and strains are
significantly localized in different constituents. Therefore, an understanding of how
each constituent and/or its interaction affect the mixture performance would help de-
velop a mix that performs well.
With recent development in X-ray computed tomography (XCT) and computation-
al simulation (the digital specimen and digital test technique, for example), mix design
can be significantly enhanced and even replaced to some degree with a computation-
based approach. Digital mix design may take the following four steps.

Step 1: Representing Aggregate Particles
Recent efforts at Virginia Tech and the National Institute of Standards and Technol-
ogy are developing databases for 3D aggregate images acquired through XCT and laser
scanning. These images are not simple pictures of the particles. They are computer vi-
sualizations, including all the detailed information such as the surface coordinates of
these particles. The particle shape, angularity, and texture information can be abstract-
ed from these images. In other words, aggregate evaluations can be achieved through
X-ray tomography imaging and image analysis at the same time they are represented
for packing. The details of the methods on how to represent the particles can be found
in Wang et al., (2004).
Step 2: Packing Particles to Form a Gradation and Gradation Optimizing
Once individual particles are represented in 3D, a set (or a gradation) of particles
can be assembled together to form a skeleton structure. In this structure, particle-parti-
cle contacts can be assessed (Figure 10.47). By manipulating the particle locations and
orientations and placing smaller particles in the voids among larger particles, an opti-
mum gradation of the aggregates can be achieved through maximizing the aggregate-