Methods for determining representative density–depth profiles using nuclear density gauges

Abstract

Most fills when compacted in layers will, to some degree, experience variations in density with depth. When measured using a nuclear density gauge in direct transmission mode, such variations in density through the layer will lead to unrepresentative density–depth profiles. This is because each measurement used to generate the profile is the mean density between the surface and the depth of measurement, and no correction for this is currently made. Uniform sands and silts, and pulverised fuel ash (PFA), often go through a process of over-stressing during compaction, leading to significantly lower densities in the upper part of a compacted layer. Under these circumstances the errors recorded in nuclear density gauge determinations of the density–depth profile may become significant. In extreme cases the usual guidance is to remove the upper part of the compacted layer prior to density testing. This paper describes numerical techniques to determine the mean density between successive measurement depths for the cases of both equal and unequal depth increments, thus giving a density–depth profile that is much closer to reality.