Trr No. 1260 Pavement Design, Management And Performance., pp. 216-265
The following conclusions can be drawn : 1. Setting a depth to bedrock of 20 ft produced a better fit between measured and calculated depth deflections. 2. The average errors on the thicker sections (9 and 12) were acceptable (less than 9 percent per sensor), indicating that linear elasticity for backcalculating E values is reasonable for thick pavements. 3. The average errors on the thin pavement (1 in. asphalt over 16 in. granular base) were high. The errors were greater than 20 percent, indicating that linear elasticity does a relatively poor job at predicting deflections within these thin pavements. The Theoritical deflection consistently underpredicted measured deflections. Difficulty of backcalculating layer moduli for thin pavements may result in a poor match. Work is now under way to determine if, by making different assumptions in the modeling process, the precent error could be reduced. The nonlinearity of the thin pavement is probably due to stiffening of the underlying granular layers. This makes it a candidate for analysis by other methods. Finite element techniques, for example, account for material stress sensitivity and use the Mohr-Coulomb failure criteria in analyzing pavement response. Work is under way to determine how the nonlinearity of the thin pavement may be accounted for. The MDD appears to be an excellent tool for validating backcalculation procedures and its use is recommended over the traditional laboratory testing approach.