The calculation of existing pavement structural capacity in terms of 18-kip equivalent single axle load (18-kip ESAL) repetitions by the mechanistic-empirical method is a multistep analysis process. The variability in this calculation process is presented with respect to falling weight deflectometer (FWD) input deflection data in the backcalculation scheme, backcalculated layer moduli, and the number of FWD tests over a section of the road. As can be expected, the sensor readings of FWD on a long section of the road show more variability than on a short section. This variability in sensor readings is magnified when the layer moduli are backcalculated (i.e., small variability in sensor data over a section of a pavement will result in high variability of calculated layer moduli). However, this variability is independent of the length of the section of the roadway over which deflection testing is done. All the layer moduli and their interaction affect the calculated structural capacity. The variation in backcalculated layer moduli is magnified when the number of 18-kip ESALs the pavement can carry before fatigue failure is estimated. Interaction of high asphalt concrete modulus and base modulus tends to produce low asphalt concrete strain and consequently a high number of 18-kip ESAL applications. The opposite is true for interaction of low asphalt concrete modulus and low base modulus. As a result, the computed structural capacity becomes highly variable, especially when the number of tests done on a long section of the pavement is small. The frequency of testing does not affect the estimated 18-kip ESALs over a short section of pavement. However, for long sections it affects the mean estimated 18-kip ESALs. For a mile-long section, five FWD tests were found to be a viable choice for estimation of 18-kip ESALs. However, the coefficient of variation of estimated 18-kip ESALs over a long section may or may not decrease with increasing number of tests. (Author/publisher)
Abstract