7.5 Variation of Subgrade Stiffness with Depth

 The analysis presented in Section 7.4 indicated the potential for

CPT prediction of the modulus of layered pavement systems as compared to

the DMT. However, the correlation of NDT tuned moduli to CPT qc values

were generally better for the base and subbase layers than the subgrade

layer (see Table 7.5). The poor correlation for the subgrade was attri-

buted, among others, to the technique of determining the effective

subgrade layer. This layer was assumed to extend from the subbase-

subgrade interface to a depth of 1 m from the surface of the pavement

without regard to the stratification of the underlying subsoils. It is

argued that depending on the relative stiffnesses of the upper layers,

the zone of influence of the dynamic loads from NDT or actual wheel

loadings could exceed the EPT value of 1 m used in the analysis.

 For a pavement section with very weak subsoil conditions, the

 effective subgrade layer used in the analysis is in reality an embank-

 ment which was placed on the natural subgrade to facilitate construction

 of the pavement. Therefore, this layer would not be a true represen-

 tation of the subsoil conditions on the site. However, the subgrade

 modulus determined from the NDT deflections and used in the correlation

 represents a homogeneous and isotropic semi-infinite layer, as pre-

 viously explained. In reality, infinitely thick subgrades and homo-

 geneous subgrades with a well-defined depth seldom exist in the field.

 This is substantiated by the stiffness profiles presented in Figures 7.1

 to 7.7 in which there was a general tendency for the stiffness to

 decrease significantly to a depth of 2 m. This would mean that in

 principle the subgrade must be considered as two or more sublayers

 depending on the stratigraphy, and consequently, increasing the total