Question: On a recent project we had some low-strength cylinders. The engineer ordered the concrete removed. We want to take cores, but are not being allowed. The testing lab says cores are a problem, as the concrete is older than 28 days. How can we handle this situation without just removing the concrete?

Answer: Look at your specification. Usually, commercial construction in the U.S. will reference ACI 301, the specification for structural concrete. In Canada, CSAA 23.1, the equivalent of ACI 318, has similar requirements for testing cores, including an assumption that the concrete will have a certain rate of failure with respect to compressive strength. Specifically, that only one result in 10 will fall below the specified strength, and that no average of a set of three consecutive tests will fall below the specified strength with no individual test below the compressive strength by more than 500 psi or 10%, whichever is greater.

To achieve this, statistical requirements are set on the concrete, such as requiring a design compressive strength higher than that specified. Still some failures can be expected even if these requirements are met.

ACI 301 also gives direction as to what is to be done in the event compressive strength results on cores have fallen below the specified strength. Concrete will generally be considered structurally adequate if the average of any three cores meets 85% of the specified strength, and no individual core is less than 75% of the compressive strength.

Obtaining and testing cores must be done in accordance with ASTM C 42. Unless otherwise specified, this testing requires conditioning of the cores for five days after they are obtained. Often this is too long to wait, and cores will be broken in the “as received” condition. If these results show acceptable performance, then there is not a major concern with the testing program, since the strengths will only be higher with conditioning. Similarly, if they are well below the acceptance criteria, the moisture condition testing is not important. What is important is if the result shows a failure, but only marginally, it’s likely that proper moisture conditioning will result in samples that pass.

Frequent objections are raised to using these reduced strength requirements for cores, when the “28-day strength wasn’t made, and now you’re measuring strength at later age.”

Inducing fracture

The purpose of taking a core is to judge whether the material in situ is structurally adequate. Drilling cores cuts the coarse aggregate particles and may induce fracture. This is why there is a lower standard for the cores versus the test cylinders. Similarly, the structure will be in service beyond 28 days, and an argument that the 28th test date is sacrosanct reveals the ignorance of the speaker.

Of course the preparation, handling, and load application have to be in strict accordance with ASTM C 42. Until recently, this required either sulfur caps or end-grinding of the samples. The latest edition of ASTM C 42, released in October 2015, allows the use of unbonded capping systems, conforming to ASTM C 1231.

Remember that a 4-inch nominal core drill takes a core of about 3.7 inch diameter. If a conventional ASTM C 1231 unbonded capping system is used, the gap between the edge of the sample and the restraining ring will not be in compliance, resulting in a bias towards lower strength. Four-inch diameter cores, taken with a 4 1/4-inch diameter barrel, will be required.

Even if ASTM C 42 is not specifically mentioned in the specification, ACI 301 contains provisions for coring. This is not a “get out of jail free” card, but is a prudent step, specified by the engineer of record, and represents the next step to be taken if the cylinder tests appear to indicate variability within the concrete.

Kevin MacDonald is president of Beton Consulting Engineers. E-mail or visit