Question: A testing firm employee says concrete should be rejected if the slump exceeds the maximum allowable value by as little as 1/4 inch. Is there any published information to help us argue our case?
Section 16.6 of ASTM C 94-98, "Standard Specification for Ready-Mixed Concrete," calls for immediately testing another portion of the same sample if slump falls outside the specified limits. If it still falls outside the limits, the concrete fails that specification's requirements. We don't think this is a reasonable approach in most cases.
ASTM C 143-98, "Standard Test Method for Slump of Hydraulic-Cement Concrete," requires reporting slump to the nearest 1/4 inch. It gives no multilaboratory precision statement, but indicates a multi-operator precision based on field samples with slumps ranging from 1 1/2 to 2 3/4 inches. The 95% repeatability limit based on this data was 0.83 inch. To apply this, a different operator might test the concrete first measured at a 5 1/4-inch slump and obtain a slump of somewhere between 4 1/2 inches and 6 inches. Due to the limited slump range and relatively low slump on which the repeatability limit is based, the actual precision might be lower and result in a repeatability limit even greater than 0.83 inch. With a test of such low precision, rejecting a load based on a minor deviation makes little sense to us.
We like an alternative requirement (see reference). A working limit is the maximum slump for estimating the quantity of mixing water to be used in the concrete, say 5 inches. An inadvertency margin is the allowable deviation from the working limit for batches of concrete that may inadvertently exceed the working limit, say 2 inches.
Batches of concrete in the 5- to 7-inch range are rejected if the contractor or producer fails to act promptly in bringing the slump back to 5 inches. The objective is an average slump not exceeding the working limit. This specification requirement must, of course, be incorporated in the contract documents before the job begins, and not after there's a dispute about slump.
The inadvertency margin approach works well for members not requiring troweled surfaces. It's not appropriate for tight-tolerance floor pours where slump may be limited to a more narrow range to achieve specified flatness.
Lewis H. Tuthill, "Quality Attainment and Common Sense in Nuclear Concrete Construction," Concrete International, March 1979, p. 36.