was not available, and the HVS PG 76-22 mix, for which binder viscosity measurements
were not available. The three conditions considered are:
* Mix/Laydown Condition from Witzcak and Fonseca (1996),
* RTFO aged binder results from Brookfield Rotational Viscometer Test, and
* RTFO aged binder results from the DSR test.
Tables 4-9 through 4-20 list the predicted dynamic modulus values for all test
temperatures and testing frequencies (Proposed by Witzcak and Fonseca-1996).
Similarly, Figures 4-24 through 4-26 show the resulting comparisons between predicted
and measured dynamic moduli for the three conditions studied. In order to evaluate the
relative quality of the predictions, linear regressions with zero intercept were performed
for the three cases. The results of the regression analysis are shown on Figures 4-23
through 4-26. The coefficient describing the slope of the regression line is a measure of
the quality of fit, the closer the slope coefficient is to unity, the less of a bias is built into
the prediction. A slope that is less than one indicates an unconservative prediction, in
which the predicted dynamic modulus is higher than the measured dynamic modulus.
Similarly, a slope that is greater than unity indicates a conservative prediction, in which
the predicted dynamic modulus is lower than the measured dynamic modulus. Similarly,
the R2 value is a measure of the goodness of fit of the regression line. A high R2 value
indicates a good fit, whereas a low R2 indicates an inadequate fit. The results from the
regression analysis show that the RTFO aged binder results from Brookfield Rotational
Viscometer test provide a slope that is closest to unity (0.6857), and the highest R2 value
(0.845). The Mix/Laydown binder viscosity conditions proposed by Witzcak and
Fonseca (1996) provide very similar results. However, the RTFO aged binder results
from the DSR test have a slope, which is higher