Rapid field testing techniques for determining soil density and water content.

Proper control of soil water content and soil compaction during earthwork construction operations is critical in achieving adequate performance of structural fills. Conventional methods (e.g., sand cone, rubber balloon, nuclear density gauge) are not always reliable and may require more time than is available to expedite construction. To improve upon the current approach, new rapid and reliable field testing technologies are needed. As part of this effort, this paper describes an evaluation of relatively inexpensive equipment that uses time domain reflectometry (TDR) and capacitance measurement techniques to determine soil density and water content in the field. Comparisons were made to the conventional nuclear and drive core methods for four different soil types and were used to estimate the accuracy and precision of each device and grounds for their successful use. The devices used were an IMKO TRIME TDR, Campbell Scientific DMM600 Duff Moisture Meter, and Humboldt nuclear moisture-density gauge. Gravimetric sampling via drive cores was used as the reference test method. Results show that TDR measurements of volumetric soil moisture are accurate without soil-specific calibration and, given a priori knowledge of gravimetric moisture content, could be used to estimate soil dry unit weight. The Duff Moisture Meter was used in conjunction with a modified drive core developed by the author to measure both volumetric moisture content and the total unit weight of a soil sample. For moisture content determination relative to oven-dried moisture content, the nuclear and TDR methods were less erratic overall than the DMM600. No method compared favorably to drive core dry unit weight; the nuclear method had the lowest RMSE and SEP values, but also the lowest R2 values overall compared to the TDR and DMM600 methods. The average time to perform each test was two to three minutes for the DMM600, less than one minute for the TDR (two replications), and two to five minutes for the nuclear moisture-density gauge (two replications).