Test Cylinder Sampling and Handling

Purpose

A low 28-day break is more often a sampling, molding or curing defect than a concrete defect, and when the number comes back short the shop has to prove which. The finished slab tells you nothing. What decides it is a record showing the sample came from the middle of the load, the cylinders were rodded in the right number of layers, and they spent their first night between 60 and 80 F rather than on a tailgate at freezing. Without it a general contractor reads a low number as bad concrete. With it, an out-of-range curing log turns a strength dispute into a testing question in one call.

Scope

Covers field sampling of ready-mixed concrete, the fresh tests taken with it, molding and initial curing of cylinders, transport to the lab, and reading the result against the ACI 318 acceptance criteria.

It does not cover mix design or plant quality control, the supplier's obligation under ASTM C94; hot or cold weather protection of the structure, owned by the hot weather and cold weather pour standards; or the decision to strip forms, released by field-cured cylinders under the formwork stripping SOP rather than by the standard-cured acceptance set.

Roles and responsibilities

Procedure

1. Fix the number of sets before the first truck. Count what the job needs against ACI 318 frequency: at least one strength test per day a class of concrete is placed, one per 150 cu yd, and one per 5,000 sq ft of slab or wall surface, whichever is greater, plus any field-cured sets the stripping plan calls for. Acceptance: set count and cylinders per set on the work order, remembering a strength test is the average of at least two 6 by 12 in cylinders or three 4 by 8 in. Wrong looks like one set on a 180 cu yd day; stop rule, call the PM before the truck arrives. Hazard: do the arithmetic in the truck, not at a discharging chute.

2. Sample from the middle of the load. Take a composite per ASTM C172 from two or more regularly spaced portions during discharge of the middle portion, never the first or last of it, and remix it in the wheelbarrow. Acceptance: elapsed time from first portion to last no greater than 15 minutes, recorded. Wrong looks like one scoop off the tail of the load, the wettest and most segregated concrete on the truck; stop rule, discard and re-sample from the next truck. Hazard: fresh concrete is strongly alkaline, near pH 12 to 13, and causes deep burns that appear hours later with little pain warning, so wear waterproof gloves and sleeves, ANSI Z87.1 eye protection and rubber boots, and flush any contact with clean water immediately and at length rather than wiping it.

3. Run the fresh tests, then mold inside the clock. Take temperature per ASTM C1064, slump per ASTM C143, air per ASTM C231 by pressure for normalweight or C173 by volumetric for lightweight, and begin molding within 15 minutes of completing the composite. Acceptance: slump and air inside the work order ranges, temperature inside the job's triggers, all four recorded with clock times. Wrong looks like air run twenty minutes late on concrete that has been sitting, which reads low; stop rule, reject a load outside the specified air or slump at the chute and record the ticket number with the reading. Hazard: the pressure-type air meter is a small pressure vessel, so bleed it through the petcocks before releasing the clamps.

4. Mold and rod exactly as ASTM C31 states. Fill 4 by 8 in molds in two equal layers and 6 by 12 in molds in three, rodding each layer 25 times with the rod C31 assigns to that size, penetrating about 1 in into the layer below, then tapping the sides 10 to 15 times with a mallet to close rod holes, and strike level. Acceptance: layer and rod counts as above, molds struck flush, identification on the mold body rather than the lid. Wrong looks like a 6 by 12 filled in two layers, or lids swapped so set identity is lost; stop rule, remake from the same sample inside the 15 minute window, otherwise sample the next truck. Hazard: filled molds are a repeated lift, so stage them at waist height.

5. Control the first 48 hours, where low breaks are made. Put standard-cured cylinders straight into a curing box holding 60 to 80 F for concrete specified below 6,000 psi, or 68 to 78 F at 6,000 psi and above, for up to 48 hours, sealed against moisture loss and protected from vibration. Acceptance: a min and max thermometer or a logger left in the box with the set, both readings inside the range for the whole period. Wrong looks like a set on a tailgate in the sun, or left by the forms on a night that drops below freezing; stop rule, a set whose log shows an excursion is flagged, is not used for acceptance, and the office tells the PM the same day. Hazard: a heated curing box is an appliance in a wet location, so it runs on a GFCI-protected circuit with an undamaged cord out of standing water.

6. Transport and hand over with signatures and times. Move the cylinders in their molds, upright, cushioned and protected from moisture loss, inside the ASTM C31 transport limit of 4 hours. Acceptance: a chain of custody carrying mold identification, sample time, curing start and end, logged minimum and maximum, transport start and arrival, and a signature at each transfer. Wrong looks like cylinders rolling loose in a truck bed overnight, which cracks a specimen before it is loaded; stop rule, flag a set that exceeds the window or arrives damaged rather than quietly testing it. Hazard: a case of cylinders is an awkward lift over a tailgate, so use a cart or a second person.

7. Confirm the lab did its half. Check that the report shows final curing at 73.5 plus or minus 3.5 F in saturated lime water or a moist room, testing within the ASTM C39 age tolerance of plus or minus 20 hours at 28 days, plus measured diameter, age at test, fracture type and any deviation. Acceptance: all present. Wrong looks like a report carrying a strength number, a date and nothing else, which can neither support nor defend anything; stop rule, request the missing data in writing before the report goes out. Hazard: none, an office review.

8. Read the result against the acceptance criteria, then act. Compare the strength test, the average of the specimens from one sample, against both ACI 318 criteria: every average of three consecutive strength tests at or above f'c, and no individual strength test below f'c by more than 500 psi where f'c is 5,000 psi or less, or by more than 0.10 f'c above that. Acceptance: both met. Wrong looks like one low cylinder inside a set being reported as a failing test, which it is not; stop rule, on a genuine failing test do not begin repair or removal, check curing and protection of the structure first, then follow the specification's investigation path, which normally leads to cores per ASTM C42 evaluated against the ACI 318 core criteria. Hazard: coring is drilling into concrete and carries respirable silica exposure, so wet-cut with continuous water under the silica dust control SOP.

The record this produces

One chain of custody sheet per sample, plus one fresh-test line per load sampled. Fields: job, pour identifier, date, technician name and certification, batch ticket number, truck and load number, cubic yards discharged at sampling, ambient and concrete temperature, slump, air and the method used, sample start and finish times, molding start, mold identifications and specimen size, curing container, logged minimum and maximum with times, transport out and in, receiving signature, and any flag with the value that triggered it.

It is filed beside the batch tickets so the two read together. The PM reads it the day a number comes back short, and the curing log is the first line they look at. The engineer decides from it whether a low test means investigate the concrete or discount the test. The estimator reads it across a season, because a shop whose flagged sets cluster in November has a curing box problem, not a supplier problem.

One pass, with a step that failed

A 180 cu yd warehouse floor, f'c 4,000 psi, air 5 to 7 percent, slump 4 to 6 in, slab area 9,600 sq ft, placed in one day. Step 1 arithmetic: one test per day gives 1, 180/150 rounds up to 2, 9,600/5,000 rounds up to 2, so the greatest of the three is 2 sets. Each set was five 4 by 8 cylinders, two broken at 7 days and three at 28, since a strength test on 4 by 8 specimens is the average of at least three.

Sample A came at roughly 42 cu yd discharged, sample B at roughly 130 cu yd, both from the middle of their loads over 9 and 11 minutes. Fresh tests read 68 F, 5.0 in and 6.2 percent air on A, and 71 F, 5.5 in and 5.8 percent on B, both inside the specified ranges.

Step 5 failed on set B. Set A sat in the heated box inside the site trailer and logged a minimum of 66 F and a maximum of 74 F. Set B went into a second box left on the north side of the building because the trailer was full, and its logger recorded a minimum of 47 F against the 60 F floor. The technician flagged set B with the 47 F written beside the 60 to 80 F criterion, and the office called the PM that afternoon.

The 28-day report landed as expected. Set A read 4,410, 4,520 and 4,380 psi, averaging 4,437 psi, above f'c and well above the individual-test floor of 4,000 less 500, which is 3,500 psi. Set B read 3,510, 3,640 and 3,580 psi, averaging 3,577 psi. That is above the 3,500 psi floor, so at face value it is not a failing test, but it sits below f'c and would have driven the conversation if it had arrived unexplained.

Because the excursion was on paper, the argument took one call. The engineer accepted set B as informational only and asked for cores from that half of the floor. Three cores tested per ASTM C42 read 3,640, 3,720 and 3,880 psi, averaging 3,747 psi. Against f'c of 4,000 psi that average is 94 percent and the lowest single core 91 percent, both clear of the ACI 318 core criteria of 85 percent on the average of three and 75 percent on any single core. The floor stayed.

Unrecorded, the same day leaves the shop owning a 3,577 psi test with no explanation on a 4,000 psi floor, and the cheapest route out is still coring, now paid for as a concrete problem rather than settled as a testing one.

When the load or the site will not cooperate

The truck is half discharged before anyone is ready: sample the next load rather than the tail of this one, and note which loads the record represents. Air comes back out of range mid-placement: stop discharge, reject the balance and record the ticket number and reading, because a rejection backed by a written value is a supplier conversation and one backed by opinion is a fight. No curing box on site: the set does not get molded at all, because a cylinder with no controlled first night is not an acceptance specimen and molding it only creates a number someone will use later. Where the contractor's own agency samples, the shop still records its fresh-test readings and the curing arrangement, because the shop gets asked to explain a low break.

References

• ASTM C172/C172M, sampling freshly mixed concrete, for the middle-portion and 15 minute composite requirements.
• ASTM C31/C31M, making and curing test specimens in the field, for layers, rodding, initial curing ranges and transport time.
• ASTM C39/C39M for compressive strength and age tolerances, with C1064, C143 and C231 or C173 for temperature, slump and air.
• ACI 318, sampling frequency, the definition of a strength test, acceptance criteria and the core evaluation path, with ASTM C42 for cores, in the edition your building department has adopted.
• See related: the cold weather and hot weather pour standards, and the formwork stripping SOP for field-cured sets.