Page 50 - MODELING OF ASPHALT CONCRETE

28 Cha pte r T w o

failure occurs in a brittle manner with a plateau zone showing a strain at failure that is
relatively small (limiting value of approximately 1.0 percent strain). As temperature
increases, a transition from brittle to ductile failure can be observed which, at high
temperatures, converts into a flow zone. The most critical part of this behavior for
pavement applications is the temperature and loading rate at which the transition from
the brittle to the ductile behavior occurs. For many unmodified asphalts, there is some
correlation between stuffiness measured at small strains (rheological prefailure
properties) and this transition. The correlation, however, does not hold for modified
asphalts or specially produced asphalts (Bahia 1995).
Failure properties can be measured using the direct tension test (DTT) to measure the
strength and strain tolerance of asphalts. In the original version of the SHRP binder
specification it was determined that most unmodified asphalts show very similar strength
value and thus strength is not needed as a specification parameter. Strain tolerance,
however, was found to vary significantly depending on asphalt source (chemistry) and
aging. It was also observed that asphalts vary significantly in the temperature zone at
which their failure behavior transitions from brittle to ductile behavior. To ensure that
asphalts at minimum pavement temperatures are within the ductile region, a minimum
value for the strain at failure was included in the specifications. A minimum value of
1 percent strain at failure measured at a specified strain rate was selected as a suitable
criterion to ensure ductility at minimum pavement temperature. Figure 2-5 shows a
schematic explaining the concept of brittle to ductile transition of asphalt binders.
From the earlier discussion of asphalt properties, it is expected that without
measuring the rheological and failure properties at the temperature and loading
frequency ranges that correspond to pavement climatic and loading conditions, selection
of asphalt binders for better-performing pavements and selection of modifiers that can
improve the properties of these binders is very difficult.
In the NCHRP 9-10 project the use of rheological and failure properties to differentiate
between modified asphalts played a different role (Bahia et al. 2001). It was observed that
although binders can show similar linear viscoelastic behavior, their nonlinear behavior
and resistance to damage can vary significantly.
It was clearly observed that the assumption that energy dissipated during
viscoelastic testing (particularly within the linear range) could be dissipated in more

FIGURE 2-5 Schematic showing the concept of brittle—ductile transition for asphalt binders.