Cube results that miss the individual limit on an M25 slab pour
You are the site engineer for a G+7 residential block cast with M25 design-mix concrete (OPC, 20 mm coarse aggregate, RMC supply). The supplier changed its cement source ten days ago. From the fifth pour, three 150 mm cubes were cast from one batch and tested wet at 28 days in a NABL-accredited laboratory after curing at 27 ± 2 °C. The results are 22.4, 23.1 and 21.8 N/mm². The site has no established standard deviation because only 14 test results exist so far. The contractor argues that the average, 22.43 N/mm², is above 22 N/mm², so the concrete is acceptable, and wants to cast the seventh-floor slab tomorrow. Site records show the slump for that pour was 110 mm against a specified 75–100 mm, and the batching plant's water meter was recalibrated two days afterwards. The structural consultant has asked you for a written recommendation before the pour.
1. Apply the IS 456 acceptance criteria for M25 concrete. Is the contractor's 'average is above 22' argument valid?
Read the worked answer
No - the batch does not comply. Two separate requirements must both be satisfied under IS 456 (clause 15.4 and Table 11):
- Individual result: no cube may fall below fck − 3 N/mm². For M25 that floor is 22.0 N/mm². The 21.8 N/mm² cube fails, even though it is short by only 0.2 N/mm²; the code gives no rounding concession.
- Group result: the mean of four consecutive non-overlapping samples must reach the acceptance value. With no established standard deviation the minimum is fck + 3 = 28 N/mm². Three cubes cannot demonstrate this at all, and 22.43 N/mm² is far below 28 N/mm².
The contractor has confused the individual floor (22 N/mm²) with the group target. Also note that the mean is barely above the individual floor, which points to a systematic loss of strength rather than one rogue cube.
2. What should you do before the seventh-floor slab is cast, and on what basis could the already-cast sixth floor be accepted?
Read the worked answer
Hold the pour and issue a non-conformance report; do not let the contractor average the failure away.
- Verify cube identity, curing and testing records, and check the compression machine's calibration certificate.
- Assess the suspect concrete already in place on the sixth floor using rebound hammer and ultrasonic pulse velocity for uniformity, then extract cores from representative locations.
- Accept the in-place concrete only if core results satisfy the usual IS 456 / IS 516 core criteria: average of three cores at least 0.85 fck and no single core below 0.75 fck, with the cores taken away from highly stressed zones.
- If cores fail, refer the member for structural review rather than quietly increasing the cement content.
- Permit the next pour only from a re-validated mix, with slump checked at the point of discharge and cubes cast under strict supervision.
3. Identify the probable root causes using the slump and calibration records, and state the quality-control changes you would impose.
Read the worked answer
The slump of 110 mm against a 75–100 mm band is the strongest clue. Excess water raises the water-cement ratio and directly lowers strength. Because the water meter was recalibrated two days after the pour, the batch was probably over-watered, either at the plant or by the driver adding water on site to improve workability.
The cement source change is a second factor: a different cement's fineness, water demand and strength-gain profile can shift the mix, so the design mix should have been re-validated with trial cubes before use.
Controls to impose: lock and seal the batching plant water calibration with a per-batch flow printout; reject trucks outside the slump band at discharge; obtain the supplier's cement test certificate and a formal change-control record; cast six cubes per batch for the next three pours; retain all 14 existing results in the statistical record for eventual σ estimation.