Page 204 - MODELING OF ASPHALT CONCRETE

182 Cha pte r Se v e n

FIGURE 7-9 Dynamic modulus mastercurves for control, CR-TB, SBS, and Terpolymer mixtures
in (a) semi-log space and (b) log-log space. (Underwood et al. 2006b, with permission from
Association of Asphalt Paving Technologists.)

calculated, the S values are determined using the iterative refinement technique
presented in Eq. (7-33). The C values are also calculated from Eq. (7-28) using the
measured stresses, pseudostrains, and the initial secant pseudostiffness (I) determined
from the early linear portion (normally pseudostrains less than 500) of the stress-
pseudostrain data. The C values and S values are then cross-plotted for multiple
crosshead rates to check the collapse of the curves, which indicates the negligible
viscoplasticity in the stress-strain data. The resulting C versus S relationship is the
damage characteristic relationship and forms the basis of the VECD model.
The procedure to ensure that negligible viscoplastic strain develops during testing
requires a trial and error approach. Tests are performed with increasing strain rates
until failure occurs in a brittle fashion during the unloading portion of the stress-strain
curve. It is known from previous research (Kim and Chehab 2004) that under such
conditions viscoelastic damage mechanisms dominate the material behavior. Once this