Nuclear Gauge vs. Cores on a Thin Overlay
A 2.0 in (50 mm) mill-and-fill overlay is being placed on a granular base. The specification requires a minimum of 92.0% of Gmm. The QC technician uses a nuclear density gauge in backscatter mode because the lift is too thin for direct transmission. At 10 random locations the gauge averages 91.8%. The daily standard count is within the manufacturer's tolerance, and the gauge was zeroed on the same base. The contractor disputes the result. Six 4 in (100 mm) cores are cut at randomly selected locations, trimmed, and tested by AASHTO T 166; the average core density is 93.6% of Gmm. The gauge and core locations are not the same. The owner's representative must decide whether to accept the lot, retest, or require removal. The contractor argues that the gauge is the specified acceptance method. The mix is a 12.5 mm surface mix with 5.4% AC and Gmm 2.498.
1. Which density result should govern acceptance of this lot, and why?
Read the worked answer
The core results should govern acceptance unless the specification explicitly requires uncorrelated gauge readings. The gauge was used in backscatter mode on a 2.0 in lift. Backscatter measurements have a depth of influence that extends beyond the asphalt lift, so the granular base density and the asphalt-base interface affect the reading. That biases the gauge low or high depending on the base. The cores directly measure the asphalt lift and were tested by AASHTO T 166, so they are the more defensible measure of in-place density. The fact that the gauge standard count passed does not remove the thin-lift bias; a standard count only verifies gauge stability, not correlation to the mix and thickness. With cores averaging 93.6% of Gmm against a 92.0% minimum, the lot should be accepted on core results, provided the cores are representative and properly trimmed. The gauge can remain a process-control tool, not the acceptance basis, until it is correlated to cores for this mix and lift thickness.
2. What could cause the 1.8 percentage-point difference between the gauge and the cores, and what immediate checks would you make?
Read the worked answer
The most likely cause is the backscatter depth of influence. On a 2.0 in lift, the gauge responds partly to the base, so a lower-density base can pull the gauge reading below the asphalt-only density. Other possible causes include surface texture or loose material under the gauge, incorrect gauge calibration or offset, and core damage or poor trimming. Immediate checks: verify the daily standard count and the calibration block reading; confirm the gauge was properly seated and the surface was clean and flat; review the gauge calibration and offset for this mix; inspect the cores for cracking, missing edges, or uncompacted surfaces; confirm the cores were dried and tested consistently under AASHTO T 166; and verify Gmm by AASHTO T 209 on the same mix. If possible, take cores at the exact gauge locations to separate location variability from method bias. A consistent gauge-low bias across paired locations points to the thin-lift backscatter effect rather than random coring error.
3. What actions should be taken before the gauge is used for production control on the remaining paving?
Read the worked answer
Establish a correlation between the gauge and cores on a test strip or on the first production lot. Use at least five paired gauge-core locations covering the expected density range, take the cores at the exact gauge locations, and compute the offset or regression needed to convert gauge readings to core-equivalent density. Do not apply the correlation outside the mix, lift thickness, and base conditions used to develop it. Where lift thickness permits, use direct transmission instead of backscatter because it reduces the base influence. For this lot, retain the cores as the acceptance record and document the gauge bias. For the remaining production, use the gauge for rapid control after correlation, but continue random core verification at the specified frequency. If the specification names the gauge as the acceptance method, obtain the owner's written approval for the correlation and acceptance protocol before proceeding.