can be used side by side in research and routine pavement evaluation

 studies. This should include an on-board computer (PC) which can

 compute moduli for a four-layer pavement system using the regression

 equations presented in Appendix F. The PC should be capable of

 printing out deflection response and layer moduli profiles which

 could be superimposed in graphic format for visual interpretation of

 lineal segments of highway pavements.

2. The field tests reported herein were carried out on fifteen pavement

 sections at fixed levels of temperature and moisture conditions. It

 is recommended that additional work be conducted on other flexible

 pavements at various seasons of the year to establish the effects

 (if any) of moisture and temperature on the prediction equations.

3. The FWD layer moduli prediction equations were found to be reliable,

 provided the field measured deflections simulated the conditions

 used to develop the equations. It was suggested, among other fac-

 tors, that the rigidity of the plate, and the load-geophone configu-

 ration in the FWD testing system influenced field measured deflec-

 tions. This influence makes FWD deflections deviate from those

 simulated from the multilayered linear elastic theory. It is there-

 fore recommended that further study be conducted to establish proce-

 dures to adjust field measured FWD deflections which would enhance

 the use of the developed layer moduli prediction equations.

 It is also believed that the FWD load-configuration does not

 provide separation of the pavement layers' response as was obtained

 with the Dynaflect. A modified FWD system utilizing a dual loading

 system appears worth pursuing in future NDT research.