Oversize Correction Turns a Passing Density Into a Failure
Roadway embankment fill is a gravelly sand with about 30% of the total dry mass retained on the 3/4-in. sieve. The laboratory ran ASTM D698 on the minus-3/4-in. fraction only and reported a maximum dry density of 124.5 pcf at 9.8% optimum moisture; the coarse fraction has a bulk specific gravity of 2.62 by ASTM C127. The field technician runs an ASTM D1556 sand cone with a 12-in.-diameter hole on the full material and computes a dry density of 125.8 pcf at 8.9% moisture. The specification requires 95% relative compaction of the corrected maximum dry density. A junior technician divides 125.8 by 124.5, gets 101%, and submits the location as passing. Acceptance of the next 8-in. lift depends on this number.
1. Compute the corrected maximum dry density for the full material.
Read the worked answer
Apply the ASTM D4718 oversize correction, which combines fine and coarse fractions by volume per unit mass:
unit weight corrected = 100 / [ Pc / (Gsb x 62.4) + (100 - Pc) / unit weight fine ]
Pc = 30%, Gsb x 62.4 = 2.62 x 62.4 = 163.5 pcf, fine-fraction maximum = 124.5 pcf.
30 / 163.5 = 0.1835 70 / 124.5 = 0.5623 sum = 0.7458 corrected maximum = 100 / 0.7458 = 134.1 pcf
So the corrected maximum is about 134 pcf, roughly 10 pcf above the fine-fraction curve. That shift is the point of the correction: the dense gravel occupies less volume per unit mass than the fines it replaces, so the full mixture must be denser to reach equivalent relative compaction. D4718 also corrects optimum moisture downward, but the density correction governs acceptance here.
2. Compute relative compaction and state the disposition.
Read the worked answer
Relative compaction = 125.8 / 134.1 = 0.938, or 93.8%, against a 95% requirement. Required dry density is 0.95 x 134.1 = 127.4 pcf, so the location is about 1.6 pcf short. It fails.
The 101% is a false pass produced by comparing the full material's field density to the fine fraction's laboratory maximum: the numerator contains 30% gravel at Gs 2.62 while the denominator contains none. Any mixture with dense gravel will beat the fine-fraction curve almost automatically, so the error always leans toward passing and will never surface in the contractor's own checks.
Disposition: reject the location, notify the engineer of record, correct the acceptance spreadsheet so every later test on this fill uses the corrected maximum, and check field moisture against the corrected optimum rather than the 9.8% fine-fraction optimum.
3. What assumptions and limits must be documented when using this correction?
Read the worked answer
Document that the 30% oversize was measured on a representative sample of the same material on the same day, not carried over from an earlier submittal. Confirm Gsb is a measured value by ASTM C127 rather than an assumed 2.65. The correction assumes the oversize is uniformly distributed in the fill and that particles are not so large they bridge or sit in the test hole.
ASTM D4718 practice covers materials with a moderate oversize fraction, about 40% or less, and a top size of roughly 3 in.; outside that, use a larger mold with D1557 or a field method suited to rockfill, and get the engineer's direction. If oversize content varies from lift to lift, one corrected maximum is not enough: recalculate for each sample's Pc or develop a family of curves. Note also that a 12-in. sand cone hole is only marginally adequate for 3/4-in. gravel; a larger hole reduces scatter.