Evaluation of Backcalculation Methods for Nondestructive Determination of Concrete Pavement Properties

Abstract

Different backcalculation algorithms often give different answers in backcalculated pavement properties. This is because of the differences in the type of pavement models, solution search procedure, and deflection matching criteria used in the backcalculation analysis. Regardless of the theory applied and the backcalculation algorithm adopted, a logical basis of selection of the backcalculation procedure for practical applications would be to assess whether the backcalculated pavement properties could provide good estimates of the actual pavement properties. Today, the ease and the convenience of access to the Long-Term Pavement Performance (LTPP) database of actual measured data enable a highway agency to adopt this approach to select a backcalculation algorithm that meets its needs. With the LTPP-measured data, this approach was applied to evaluate the relative merits of four backcalculation algorithms (two versions of ILLIBACK, NUSBACK, and LTPP best-fit method) by a comparison of the computed elastic modulus of concrete pavement slab and the modulus of subgrade reaction of concrete pavements against the LTPP measured values. The performance of the four algorithms was greatly affected by the constraints imposed by the deflection theory adopted and was significantly dependent on their respective criteria used to match the computed and measured deflections. The number of sensors used in the backcalculation, as well as the choice of sensor configuration, can significantly affect the performance of the backcalculation algorithms.