When the Core Comes Back Low

Consider a case documented in The Construction Specifier, from a multi-storey car park in the United States. Their site casts cylinders where an Indian site would cast cubes, but the problem is the same one, and the way it was handled is the lesson. The slabs were specified at 40 MPa. When the 28-day cylinders for one pour came back between 32 and 36 MPa, averaging 34, well under the acceptance line, the job stalled. Later pours were re-sequenced in case the slabs had to come out.

Nobody reached for a jackhammer. The engineer of record went back to the design, worked out what the slabs actually had to carry, and set a structural minimum of 35 MPa. Then the team drilled cores, four separate rounds of them, and tested them against the rule written for exactly this situation.

Concrete in the member represented by a core test shall be considered acceptable if the average equivalent cube strength of the cores is equal to at least 85 percent of the cube strength of the grade of concrete specified for the corresponding age and no individual core has a strength less than 75 percent.

IS 456:2000, Clause 17.4. The American code, ACI 318, uses the same 85 and 75 percent.

The cores held up, and petrography found the culprit. The air content in the concrete had run from nine to more than fifteen percent, against a specification of six and a half plus or minus one and a half. A little entrained air is put in on purpose, to survive freeze and thaw, but every extra percent of it costs strength, and this mix had far too much. The concrete was not weak because it was badly placed. The petrography found it well consolidated. It was weak because it left the plant carrying too much air, which is a batching problem you fix at the plant, not a slab you demolish. About five months after the pour, and four months after the 28-day break first raised the alarm, the engineer and the owner accepted it.

This is the distinction the whole issue turns on. A cube, cured in a tank under ideal conditions, measures the potential of the concrete, the best the mix can do. A core measures what the structure actually received, after real compaction, real curing, and the drilling itself.

Determination of the actual strength of concrete in a structure is not easy because it is dependent on the history of curing and the adequacy of compaction of concrete.

Khoury, Aliabdo and Ghazy, "Reliability of Core Test," Alexandria Engineering Journal, 2014.

A core reads lower than its cube for a list of ordinary reasons. It is more slender than a cube, it may cut through a bar or a void, it dries out differently before the test, and it is drilled across the direction the concrete was cast. A gap between the two is expected. It is not, by itself, a defect.

Which is why IS 456 does not ask the core to match the cube. It asks for at least three cores, prepared and tested to IS 516, and it accepts the concrete when they average at least 85 percent of the specified grade with no single core below 75. A core is allowed to read below the grade, and the concrete still passes. The reflex to condemn a pour on one low, uncorrected core reading, and the opposite reflex to wave through concrete that breaks the 75 percent floor, are the two symmetric ways engineers get this wrong.

The lesson: When a core comes back low, do not reach for the jackhammer, and do not sign it off either. Correct each core to an equivalent cube for its shape and its moisture, put at least three of them against the IS 456 rule of 85 percent on average and 75 percent as a floor, and if it still fails, find out why before you act, because the answer is usually in the mix or the curing, not the structure. The cube tells you what the concrete could be. The core tells you what you have. Reading the gap between the two is the job.

Read more: Concrete Acceptance Criteria for Dam QA/QC →