How to sample concrete, when to test, and how to interpret cube test results per IS 456. Both individual cubes AND running averages must satisfy criteria. Failure of either triggers investigation, core testing, or load tests. The criteria are for mean of 3 cubes per sample (1 sample = 3 cubes from 1 batch).
Acceptance Criteria — Cl 16.1 (Compressive Strength)
BOTH individual AND average criteria must be met
| Rule | Value | Frequency | Note |
|---|
| Individual cube (mean of 3 from one sample) | ≥ fck − 3 N/mm² (M15 grade) | — | OR ≥ 0.85 × fck for M20+ |
| Mean of any 4 consecutive samples (M15) | ≥ fck + 0.825 × σ | Or ≥ fck + 3 | Whichever larger; σ = SD |
| Mean of any 4 consecutive samples (M20+) | ≥ fck + 0.825 × σ | Or ≥ fck + 4 | Whichever larger |
| Standard deviation σ (n ≥ 30) | Computed from results | — | Update monthly |
| Assumed σ (n < 30, M20) | 4.0 N/mm² | — | Cl 9.2.4.2 Table 8 |
| Assumed σ (n < 30, M25) | 4.0 N/mm² | — | — |
| Assumed σ (n < 30, M30) | 5.0 N/mm² | — | — |
| Assumed σ (n < 30, M40+) | 5.0 N/mm² | — | Or 6.0 for M55+ |
Notes
• Cube size = 150×150×150 mm (standard) — 100 mm cubes also allowed with correction factor
• ALWAYS cast a duplicate set of cubes — gives backup if primary set fails
• Cubes must be tested at 28 days (±24 hours) — not earlier, not later
• Cure cubes UNDER WATER, not by spraying — ensures uniform hydration
• 1 sample (3 cubes) test result = mean of the 3 cubes (reject if any one differs by ±15% from mean)
• Always log slump, batch time, weather, and ambient temperature with cube records
• For PPC/PSC concrete, IS 456 allows acceptance based on 56-day strength — get this in your specification
• If a sample fails, do not condemn the structure immediately — investigate per Cl 17
• Core test is the most reliable verification — costs are justified for borderline failures