285
log E, = 4.970 + 0.1773(t.) 1.6966 log(t) 0.1069(DJ
O i 1 H
+ 0.2552(D7) 2.6546 1og(D1) 3.9906 log(D3)
+ 1.8241 log(Dc) + 3.5092 log(D0 D_)
b o
Eqn. 6.13
R2 = 0.887 and N = 22
Error analysis indicated that one pavement (SR 15B) had -34.6
percent prediction error. The actual Eg value was 50.0 ksi, while
the predicted value was 32.7 ksi. Others had prediction errors
generally less than +15 percent.
Equation 6.13 for the prediction of Eg applies to a slightly
wider range of variables than Equation 4.24 which was selected from
the theoretical analysis. However, the latter is more simplified
and contains fewer variables than Equation 6.13. Also, since the
R2 value is greater, Equation 4.24 should be used for E3 predic
tions.
6.7.3.2.4 Subgrade modulus, E Regression analyses of E^
against either Dg, or Dy, or both resulted in the following equa
tions :
E = 53.697(0 )"i.04i Eqn. 6>14
4 6
(R2 = 0.997 and N = 22)
E = 39.690(D )1.004 Eqn< 6>15
(R2 = 0.999 and N = 22)