Compaction Results That Do Not Add Up on a Four-Foot Embankment

T 99 (standard Proctor) uses lower compactive effort (about 12,400 ft-lb per cubic foot, 5.5-lb rammer, 12-inch drop, three lifts) than T 180 (modified, about 56,000 ft-lb per cubic foot, 10-lb rammer, 18-inch drop, five lifts). Maximum dry density from T 99 is therefore lower than from T 180 for the same soil.

Consequence: dividing field density by the lower T 99 maximum inflates the reported percentage. A reported 96.2 percent may be several points lower when compared against the correct T 180 maximum - the lot may be failing outright.

Required actions before more fill:

1. Suspend acceptance of the affected lots.
2. Verify the curve - lab report, project number, sample source, date, and soil classification.
3. Have a T 180 curve run on a representative sample of the actual borrow (or verify by a one-point Proctor per T 272).
4. Confirm the nuclear gauge has a current correlation to sand cone (T 191) on this soil and that calibration is current.
5. Recompute the affected lots and identify how far back the wrong curve was used; those lifts must be re-evaluated.

Document the nonconformance and the retest plan under the engineer's direction.

Do not accept, and do not release the area for base aggregate.

• 92.8 percent at STA 215+00 fails the 95 percent requirement. Unless the specification expressly allows lot averaging - and even then a failing sublot usually still triggers rework - a failing test is a failing test. Proofrolling is a supplementary check for unstable material, never a substitute for density acceptance.
• The 11-inch lift explains the low density. Compactive effort from the surface cannot reach the bottom of an over-thick lift, so the lower portion stays loose. The lift itself is a nonconformance independent of the density result.

Corrective action:

1. Scarify or excavate the over-thick lift to within 8 inches loose; the failing area may extend beyond the single test point, so define the limits with additional tests rather than one point.
2. Moisture is 3.5 percent above optimum - aerate/disc and allow drying to within the -1 to +2 band.
3. Recompact with the approved equipment and pass count.
4. Retest at the failed location after rework plus enough additional points to bracket the repaired area. A retest relocated to a convenient spot is not a valid retest.

Record thickness measurements, moisture, equipment, passes, and all test locations.

The material is outside the -1 to +2 percent band, so it cannot be accepted as placed.

Options, in order of practicality:

1. Aerate by disking, blading, and turning until the moisture falls into the band, then recompact.
2. Blend in drier borrow material.
3. Delay the lift and let it dry (weather-dependent, and note the schedule impact).
4. Do not accept 'roll it more' as a fix - overrolling wet fine-grained soil causes pumping, shearing, and a crust over soft material.

Document: moisture at each test location and the test method; whether the wetting came from rain, haul-road watering, or the borrow source itself (a borrow pit above optimum will keep producing wet fill); the corrective action taken; and drying time and weather during it.

If the contractor disputes the moisture, verify with an oven-dry sample (T 265). Nuclear moisture readings are an estimate and require correlation; the oven-dry result is the referee value. A borrow source that is consistently wet of optimum is a source issue worth raising to the engineer before more fill is scheduled.