114


log (E ) = 0.587 0.037 t 10.19 D + 8.01 log(D ) + 2.226 log(D )
 3 1 1 3 6

 5.119 log (D D ) + 17.255 log (D D )
 1 3 1 5
 6.101 log (D D ) 7.051 log (D D )
 2 5 4 7

 Eqn. 4.23

 (N = 192, and R2 = 0.958)



 The database used to develop Equation 4.23 was then expanded to

include E3 = 15 ksi, E4 = 5 ksi and t, = 8.0 in. combinations.

Subsequent regression analysis resulted in a 5-variable prediction

equation listed under Case 2.


Case 2. For 150.0 < E < 300.0 ksi, 1.5 < t < 8.0 in.,
 1 1

 42.5 < E < 85.0 ksi, 15.0 < E < 75.0 ksi,
 2 3
 and 5.0 < E < 40.0 ksi,
 4


log (E) =
 0


3.8646 0.27061 log (t ) + 1.1212 log (D )
 1 6
- 1.849 log (D D ) + 12.009 log (D D )
 2 3 2 4

- 12.3637 log (D D )
 2 5


 (N = 400, and R2 = 0.935)



 4.3.3.4 Prediction Equations for E4. The subgrade modulus, E4,

was found from the sensitivity analysis to contribute significantly to

the NDT deflections compared to the moduli of the upper layers. Changes

in E4 affected the deflections to the greatest degree. Using the 9-kip


Eqn. 4.24