Nuclear Gauge Versus Sand Cone Discrepancy
During compaction control on a 12-in. lift of silty sand embankment fill, a nuclear gauge in direct-transmission mode with a 6-in. probe reports wet density 128.4 pcf, moisture 9.8%, and 96.8% of the laboratory maximum dry density (standard Proctor, 132.6 pcf). The same spot is immediately tested by sand cone: dry density 120.1 pcf and moisture 8.9% by oven drying, or 90.6% compaction. The gauge's daily standard count is 1.5% below its 30-day average. The lift contains gravel up to 1.5 in., and the specification requires 95% standard Proctor and permits nuclear testing only if the gauge is calibrated for the material and lift thickness. The contractor points to the gauge result and asks for acceptance so the next lift can be placed before rain.
1. Which test result should control the acceptance decision, and why?
Read the worked answer
The sand cone result should control unless it is shown to be invalid. Sand cone is a direct physical measurement of excavated volume and mass, while the nuclear gauge is an indirect correlation-based method. A difference of about 8 pcf and more than 6 percentage points of compaction is far beyond normal test variability and requires investigation. The gauge is especially suspect here because the lift contains gravel up to 1.5 in., the 6-in. probe does not cover the full 12-in. lift, and the standard count has drifted. That does not automatically make the sand cone perfect in gravelly soil; the hole must be large enough and the excavated material must include all particles. But on the present evidence, the acceptance decision should be based on 90.6%, which fails the 95% requirement.
2. Evaluate the nuclear gauge standard count and probe depth as sources of error in this lift.
Read the worked answer
A daily standard count 1.5% below the 30-day average must be checked against the agency or manufacturer tolerance. Many programs require the density standard count to be within about 1% of the established average; if that is the project criterion, 1.5% is outside and the gauge should not be used until the cause is found and the gauge is recalibrated or repaired. Even if 1.5% is within a looser tolerance, it indicates drift and reduces confidence. The 6-in. probe in a 12-in. lift measures an average density over only the upper portion of the lift; if the lower half is looser, the gauge can overstate the average. Gravel larger than the probe spacing also causes scatter. These factors support rejecting the gauge result for acceptance and using sand cone or another verified method.
3. State the acceptance decision and the corrective actions you would require before the next lift.
Read the worked answer
The lift fails at 90.6% standard Proctor. I would not accept it and would not allow the next lift to cover it. The contractor should scarify or rework the lift, adjust moisture toward optimum, and recompact. Because the sand cone moisture is 8.9% and the material is below the required density, additional water may be needed, but the optimum moisture should be confirmed from the Proctor curve rather than assumed. After recompaction, I would retest at randomly located spots with sand cone, or with a nuclear gauge only after its standard count and material calibration are verified. I would also check lift thickness and gravel segregation. Acceptance should be based on passing tests before rain arrives, not on the earlier gauge reading.