much different from those obtained from the rheology tests (Table 6.5).
This does not only confirm the reliability of the indirect method of
predicting EV, but also suggests that the E1 values obtained from the
prediction equations are suspect. Also the tuned E2 and E values are
much different from the predicted values listed in Table 6.4. However,
the subgrade modulus values compare well which emphasizes the uniqueness
of the relationship between E, and the furthest NDT sensor deflec-
tion(s).
The use of the field measured FWD deflections as input into the
prediction equations did not yield reasonable layer moduli, especially
for El. Predicted deflections were generally different from the field
measured values and likewise the tuned moduli differed considerably from
the predicted, with the possible exception of E It was suggested that
this discrepancy was due to the fact that FWD measurements do not neces-
sarily satisfy the assumptions of layered theory used in the theoretical
analysis. To support this, the predicted deflections obtained from
BISAR using the tuned moduli were then used as input into the developed
equations to compute the layer moduli. These deflections, listed in
Table 6.11, are for all purposes BISAR generated deflections and should
satisfy the assumptions inherent in the theoretical analysis.
Table 6.12 lists the calculated (predicted) moduli compared with
the tuned values. There is favorable agreement between the two for each
layer modulus. For the SR 12 tests site, the considerable difference in
E values is due to the poor prediction accuracy of Equation 4.18 for
very thin asphalt concrete layers. The relatively good agreement
between predicted and tuned moduli in Table 6.12 indicates that the
prediction equations could be used to estimate layer modlui. However,