The introduction of new devices to measure pavement density requires an evaluation method that is both accurate and fair. Evaluations are commonly done by taking a density measurement with the gauge and comparing this density to the density obtained from a core taken at the same location. This approach must meet two requirements; first, enough cores must be taken at each location to make the comparisons meaningful, and second, evaluation of these devices must be done with as many projects in as many locations as possible to account for all materials used in pavement construction. Unfortunately, given the difficulties in obtaining field cores, large numbers of cores that can be used for evaluations are not available from all projects. Therefore, any evaluation of density gauges must balance rigor with practicality. The correlation coefficient was selected to evaluate the new nonnuclear pavement density gauge. However, because only a limited number of cores are available at each site for development of the correlation coefficient, this parameter by itself cannot be used to evaluate the new gauge. Instead, results from side-by-side comparisons with the accepted nuclear density gauge were used to aid in the evaluation of the new density-measuring device. It was concluded, on the basis of data from 76 projects in six different states, that the proposed nonnuclear density gauge must be further developed before it can be used to determine pavement density. In its current form, the densities obtained with the device do not correlate with the measured densities to the same degree that the densities measured with existing devices do. Use of this nonnuclear device will introduce more uncertainty in pavement density measurements than that which already exists.
Abstract