M25 Slab Cubes Fail Both Acceptance Criteria - Accept, Core-Test, or Reject?
A residential project in Nashik uses M25 (fck = 25 N/mm²) concrete for a 150 mm thick one-way slab. The RMC supplier's design mix has a water–cement ratio of 0.48 and 340 kg/m³ of cementitious content. Twelve 150 mm cubes were cast from three separate transit mixers and cured under standard conditions. The 28-day results in N/mm², in casting order, are: 22.1, 23.5, 24.0, 26.2, 27.0, 21.8, 25.5, 24.8, 23.0, 22.6, 26.5, 25.2. Slump records for the same three loads were 165 mm, 170 mm and 160 mm against a specified 75–125 mm. The slab was deshuttered at 7 days and shows no visible cracking. The contractor argues that the average of all twelve cubes, 24.35 N/mm², is close enough to 25 and wants to proceed with the floor above. Assume σ = 4.0 N/mm² for this grade.
1. Assess the cube results against the acceptance criteria and state whether the slab concrete can be accepted as it stands. Show the arithmetic for every criterion you apply.
Read the worked answer
Two criteria apply for M25 with an assumed σ = 4.0 N/mm²:
- Individual result: no cube may fall below fck − 3 = 22.0 N/mm². The sixth cube at 21.8 N/mm² fails, marginally.
- Group of four consecutive non-overlapping samples: mean must be ≥ fck + 0.825σ = 25 + 3.30 = 28.3 N/mm². Groups: (22.1, 23.5, 24.0, 26.2) = 23.95; (27.0, 21.8, 25.5, 24.8) = 24.78; (23.0, 22.6, 26.5, 25.2) = 24.33. All three are roughly 3.5–4.4 N/mm² below the requirement.
So the concrete fails both criteria. The grand mean of 24.35 is not an acceptance criterion - averaging twelve results hides the batch-to-batch scatter that the group-of-four rule is designed to catch. Note also that a mix designed for σ = 4.0 should have a target mean of 25 + 1.65 × 4 = 31.6 N/mm²; delivering 24.35 means the mix as supplied or as placed was under-strength by a wide margin, not by a rounding error.
Decision: do not accept and do not cast the floor above on this slab until strength is verified. A 3–4 N/mm² shortfall reduces flexural capacity by roughly 6–12% depending on the limit state being checked, on a slab only 150 mm deep where cover and effective depth leave little reserve. Issue a non-conformance, keep the props in place, and move to cores and NDT.
2. Set out the investigation and disposition path while the strength is in question. Name the tests, the acceptance basis for each, and what you do on site in the meantime.
Read the worked answer
Immediate actions: hold the next pour, tag the slab, and keep the props under it with a note of the deshuttering date.
Step 1 - Verify the test trail before testing concrete. Check sampling frequency (at least one set per 5 m³ or per batch, minimum three samples per day), cube identification, whether cubes were compacted the same way as site concrete, curing tank temperature, and the laboratory's calibration certificate. A cube made from a wheelbarrow sample left in the sun is not evidence about the slab.
Step 2 - NDT screening. Rebound hammer on a grid over the slab to map weak zones, backed by ultrasonic pulse velocity. Both are indicative only; they locate where to core and cannot by themselves be used for acceptance.
Step 3 - Cores. Extract three or four cores of 50 mm diameter or more, with length/diameter ≥ 1, from locations clear of reinforcement, and test in compression. The usual acceptance basis is that the average core strength is not less than about 85% of fck (≈ 21.3 N/mm²) and no individual core falls below about 75% (≈ 18.8 N/mm²); quote the clause in your contract or specification rather than assuming it.
Step 4 - Load test. If cores also fail, run a load test: apply the factored load plus a margin, record deflection, confirm recovery of at least 75% within 24 hours and no new cracking.
Step 5 - Disposition. Cores or load test pass: accept with the record attached to the pour card. They fail: obtain a structural review and choose between strengthening (FRP or jacketing) and demolishing and recasting the affected bays. Do not sign off on the contractor's average-based argument.
3. Using the slump data, identify the most probable causes of the shortfall and specify the changes you would make to the supply and to site quality control before the next slab.
Read the worked answer
The slump is the strongest clue. A mix at w/c 0.48 with 340 kg/m³ should not sit 40–55 mm above the specified upper limit three loads in a row. Each 10 mm of extra slump corresponds roughly to 1–1.5% more water per cubic metre; 45 mm of excess slump implies about 12–18 L/m³, which pushes the effective w/c from 0.48 to about 0.52 and costs roughly 8–12% of strength - consistent with the 3–4 N/mm² shortfall observed.
Probable causes, in order of likelihood: (1) water added at site or in the truck to make placing easier; (2) admixture dosing not adjusted for a change in sand moisture or grading between loads; (3) cement under-weighing at the batching plant, or a fly-ash replacement not reflected in the declared mix; (4) cube-making practice that differed from the slab (compaction, temperature, time to tank).
Changes to impose: revise the mix to a target mean of 31.6 N/mm² - either raise cementitious content by about 20–25 kg/m³ or reduce w/c to about 0.44 with a proper superplasticiser; ban site water addition and require a signed batching/delivery sheet per load; measure sand moisture daily and adjust the batch water; record slump, concrete temperature and cube identity at the point of discharge; take three cubes per 5 m³, cure them beside the slab for the first 24 hours, then in a standard tank; cure the slab by ponding or wet hessian for 7 days minimum. Check 7-day results (about 65–70% of 28-day) before committing the next slab.