be due to nonuniqueness in the backcalculation of layer moduli. Table

6.13 also shows that slight changes in E3 and E (SR 15A and SR 26A)

could also lead to the same deflection solutions.

 The problem of nonuniqueness in layer moduli determination is pre-

vented with the use of the prediction equations. The subgrade modulus

has already been shown to be uniquely related to the deflection(s) at

the farthest sensor in the Dynaflect and FWD testing systems. Also the

high degree of reliability of predicting E1 from asphalt rheology rela-

tionship generally fixes the E1 value to be used in the tuning of

deflection basin. The use of the prediction equations in addition to E

predictions from rheology tests eliminate guesswork in selecting initial

moduli. Therefore the methodology presented in this dissertation

ensures unique solutions and is not user-dependent with regard to

selecting input moduli values.

6.5.5 Effect of Stress Dependency

 As previously mentioned, laboratory studies generally suggest that

the moduli of subgrade materials and granular bases are stress depen-

dent. One of the advantages of the FWD testing system is its ability to

apply variable and heavier loads to assess the stress dependency of

pavement materials. The load-deflection response shown in Figures 6.1

through 6.14 indicated that FWD deflections were within reason linearly

related to the applied loads for most of the test sections except for SR

26B, SR 15A and SR 15B. For these pavement sections, the tendency was

toward a nonlinear response. However, in all cases the load-deflection

response did not pass through the origin. This could be due to the

inertia of the pavement system to loading, and perhaps the influence of

the static loading due to the plate.