226


closely match measured values for all sensor locations. However, for

most of the other pavements, especially SR 26C, I 10, and SR 715, the

agreement between measured and predicted deflections was good for D ,

D6, and D7 measurements only. The difficulty in matching FWD deflec-

tions with BISAR could be due to one or more of the following:

1. The pavements did not necessarily behave as linear elastic media.

2. It was improper to represent the FWD impulse or dynamic loadings

 with a pseudo-static loading in BISAR.

3. Neglecting the inertia of the pavement in the simulation of NDT

 response using multilayered linear elastic theory.

4. The FWD plate and loading are rigid rather than flexible as assumed

 in the BISAR analysis.

5. The spacing of the geophones and loading configuration are probably

 not suitable to allow the separation of pavement layers during the

 interpretation of FWD deflections.

 It is believed that the above reasons, especially the first three

apply to all NDT devices. It may be argued that the small load (1.0 kip

total) in the Dynaflect system is too small to produce any sensitivity

in the load-deflection response compared to the 9-kip load used in the

FWD. On the other hand, the use of dual loads in the Dynaflect testing

system with the modified sensor configuration enhances the separation of

layer response. Differences between the response of the test pavements

to the FWD and Dynaflect are discussed in detail in Section 6.6.

 The layer moduli which best matched one or more of the FWD deflec-

tions were selected as the pavement tuned layer moduli. These are

listed in Table 6.10. The predicted FWD deflections using the tuned

layer moduli are listed in Table 6.11. The tuned E values were not