Published Online: 1 March 2012
Page Count: 13
Meehan, Christopher L.
Assistant Professor, Dept. of Civil and Environmental Engineering, 301 DuPont Hall, Univ. of Delaware, Newark, DE
Tehrani, Faraz S.
Graduate Student, Purdue Univ., School of Civil Engineering, West Lafayette, IN
Senior Project Engineer, Paul C. Rizzo Associates, Inc., Pittsburgh, PA
(Received 12 October 2010; accepted 5 October 2011)
This paper presents and compares the results from a series of in situ density-based and modulus-based compaction control tests that were conducted during construction of a coarse-grained soil embankment. To simulate current construction practices as closely as possible, these in situ tests were performed on an embankment that was constructed and compacted by a vibratory smooth drum roller in a series of lifts. During construction of the test embankment, the compaction process was monitored using the nuclear density gauge device and a number of alternative modulus-based devices, including the lightweight deflectometer, the dynamic cone penetrometer, and the soil stiffness gauge. Comparison of the in situ test results illustrates that point-to-point variability in measured values is quite common for each of these test devices, to varying degrees for the different devices that were examined. Consistent increases in measured soil properties from pass-to-pass of the compactor are considered critical for proper control of the compaction process, with some devices faring better than others in this area of performance. The measured modulus values correlated poorly to the nuclear density gauge dry unit weights, and also correlated poorly with other measured moduli when the results from different devices were compared. This lack of agreement was likely caused by a variety of factors including: variations in the magnitude of strain and rate of strain application between the different modulus-based devices, variations in the tested volume between the different devices, and variations in the local moisture content and matrix suction conditions. Finally, the effect of soil moisture content was shown to be critically important when interpreting the results from modulus-based tests, and the utility of multiple regression analyses was explored for including this effect.
Paper ID: GTJ103479