Codes referenced
Sampling, capping, loading rate, failure mode observation.
Non-destructive test method (with Nov 2019 amendment).
Slump, temperature, density of fresh concrete.
UPV velocity measurement; rebound hammer test method.
Core extraction, testing, interpretation.
Witnessing vs testing
The witness verifies the test was done right; the lab technician performs the test. Witnessing is a 3-phase activity: pre-checks (equipment calibration valid, sample identity correct, preparation compliant); during-checks (procedure is followed, no shortcuts); post-checks (calculation correct, result signed, witness signs observation form). Missing any phase invalidates the witnessing.
Pre-checks for cube test (IS 516)
- Cube dimensions 150 ± 0.5 mm (or 100 mm for special cases)
- Cube surface planeness < 0.02 mm
- Capping material applied correctly (sulphur or neat cement)
- Compression machine calibration valid (within 12 months)
- Load rate per IS 516 Cl 5.2: 14 MPa/min for M30+, 5-15 MPa/min general
- Cube identity matches the sample reference (delivery ticket → cube ID)
- Curing tank temperature 27 ± 2°C verified
- Cube age matches test schedule (7d, 28d, 56d, 90d)
During-checks for cube test
- Load applied at constant rate (verify dial gauge or machine display)
- No premature load release or pause
- Failure mode observed and recorded (satisfactory cone per IS 516)
- Peak load captured accurately (no manual override of machine)
Post-checks for cube test
- Strength calculation: Peak load (N) / contact area (mm²) → MPa, reported to 0.5 MPa
- Lab technician signed the test report
- Witness signed the witnessing form (not the test report itself)
- Test report filed in QA document register with cube ID cross-reference
Pre-checks for AIV (IS 2386 Pt 4)
- Sample mass 400-450 g of 10-12.5 mm aggregate
- Sample dried to constant mass (oven at 100-110°C)
- 15 blows of hammer (each 380 mm drop, 13.5-14.0 kg mass)
- Sieving on 2.36 mm IS sieve after impact
- AIV = (mass passing 2.36 mm / original mass) × 100; reported to nearest 1%
- AIV acceptance: ≤ 30% for concrete wearing surfaces, ≤ 45% for other concrete (IS 383)
Pre-checks for UPV (IS 13311 Pt 1)
- Surface preparation: smooth, clean, free of dust and loose material
- Couplant applied (gel, paste, or petroleum jelly)
- Transducer spacing measured accurately (typically 150 mm)
- Path length measured perpendicular to surface
- Minimum 3 readings at each location, average used
- UPV grading per IS 516 Amd 1 (Nov 2019): grade-dependent bands (not the legacy IS 13311 fixed bands)
Pre-checks for rebound hammer (IS 13311 Pt 2)
- Surface preparation: smooth, clean, no rough areas
- Calibration check on standard anvil before use
- Minimum 9 readings per test area (IS 13311 Pt 2)
- Readings within ± 20% of median are valid; outliers rejected
- Average rebound number converted to estimated strength via manufacturer calibration curve
Worked example — 15-check cube test witness checklist
| Phase | # | Check | Witness OK |
|---|---|---|---|
| PRE | 1 | Cube dimensions 150 ± 0.5 mm | ✓ 150.2 mm |
| PRE | 2 | Surface planeness < 0.02 mm | ✓ 0.01 mm |
| PRE | 3 | Capping material OK | ✓ sulphur |
| PRE | 4 | Machine cal. valid (≤12 months) | ✓ valid till 2026-09-30 |
| PRE | 5 | Load rate set to 14 MPa/min | ✓ verified on display |
| PRE | 6 | Cube ID matches sample | ✓ MD-2026-M30-F5-C7 |
| PRE | 7 | Tank temp 27 ± 2°C | ✓ 26.5°C |
| PRE | 8 | Cube age 27d (target 28d) | ✓ 27d, acceptable |
| DUR | 9 | Constant load rate applied | ✓ verified |
| DUR | 10 | No premature release | ✓ |
| DUR | 11 | Failure mode observed | ✓ satisfactory cone |
| DUR | 12 | Peak load captured | ✓ 568 kN |
| POST | 13 | Strength calculated: 568/22500 = 25.2 MPa | ✓ 25.2 MPa |
| POST | 14 | Lab tech signed report | ✓ S. Patil |
| POST | 15 | Witness signed observation form | ✓ A. Haridas |
Common deviations seen on Indian sites
A machine with expired calibration produces invalid results. Witness MUST verify calibration certificate before test begins. Without it, the result is unsalvageable.
The witness must be present for the entire test, especially the failure moment. Leaving early means missing the failure mode observation, which is critical for result validity.
A witness who signs without verifying the cube ID has not actually witnessed. The cube tested must be the cube reported. Sample-identity verification is non-negotiable.
A cube that fails in shear rather than cone is suspect — it indicates capping issue, surface defect, or non-conforming material. The failure mode is diagnostic and must be observed.
Field-checklist blocks
A short questionnaire, an inspector's checklist, and three worked examples that show how the pre-, during- and post-checks resolve into a signature, a rejected aggregate or a non-conformance raised in the presence of the witness.
Questionnaire — ask before you sign the observation form
- Is this a test that has to be witnessed because the result drives acceptance, signoff or payment, or is it a random QC test that only needs calibrated equipment and an accredited lab?
- Is the calibration certificate on the machine current? A machine with an expired calibration produces results that cannot be salvaged, whatever they read.
- Does the cube on the bench match the sample reference on the delivery ticket, and is its age the one the test schedule calls for?
- Is the load rate set to the requirement: 14 MPa/min for M30 and above, or 5–15 MPa/min generally, per IS 516 Cl 5.2?
- Will someone be present for the failure moment, because the failure mode is diagnostic and cannot be reconstructed afterwards?
- Which form am I signing? The witness signs the observation form, not the test report itself.
Inspector's checklist — what to verify at the test
- Pre, equipment: cube dimensions 150 ± 0.5 mm, surface planeness below 0.02 mm, capping material correctly applied (sulphur or neat cement), machine calibration valid within 12 months.
- Pre, load rate: verified on the machine display before the test starts, at 14 MPa/min for M30 and above or 5–15 MPa/min generally, per IS 516 Cl 5.2.
- Pre, sample identity and age: cube ID matched to the sample reference, and the age matching the test schedule of 7, 28, 56 or 90 days.
- Pre, curing: tank temperature verified at 27 ± 2 °C.
- During: constant load rate held, no premature release or pause, failure mode observed and recorded, peak load captured with no manual override of the machine.
- Post, calculation: peak load in newtons divided by contact area in mm², reported to 0.5 MPa; technician has signed the test report; the witness has signed the observation form.
- Post, filing: the test report is in the QA document register with the cube ID cross-reference against it.
- For the other tests: AIV pre-checks (400–450 g of 10–12.5 mm aggregate, dried at 100–110 °C, 15 blows from a 380 mm drop with a 13.5–14.0 kg mass, sieved on the 2.36 mm sieve), UPV pre-checks (surface clean, couplant applied, spacing measured, path measured perpendicular, minimum 3 readings averaged) and rebound pre-checks (anvil calibration, minimum 9 readings per test area, only readings within ± 20% of the median valid).
What happens if… three worked examples
What happens. The contact area is 150 × 150 = 22,500 mm², so the strength is 568,000 ÷ 22,500 = 25.24 N/mm², recorded as 25.2 MPa. The tank was at 26.5 °C, inside 27 ± 2 °C; the cube was at 27 days against a 28-day target; the calibration was valid; the load rate was verified on the display; the failure mode was a satisfactory cone and the peak was captured without a manual override.
Likely outcome. The technician signs the test report and the witness signs the observation form, and the report goes into the QA document register against the cube ID. The witness's signature goes on the observation form only. That separation is the point of the exercise: the technician performed the test, the witness verified it, and the record shows both roles kept honestly.
Preventive correction. Record arrival and departure times in the observation form. They are the evidence that you were there for the whole test, including the failure.
What happens. The aggregate impact value is (135 ÷ 420) × 100 = 32.1%, reported to the nearest 1% as 32%. The acceptance limit depends on where the aggregate is going: ≤ 30% for concrete wearing surfaces, ≤ 45% for other concrete. So 32% fails for the wearing-surface pour and passes for everything else.
Likely outcome. The aggregate is rejected for the wearing-surface element and accepted for the structural pours, with both decisions written on the same test record. Treating it as a single pass or fail across the project would be wrong in one direction or the other, and it is the kind of judgement a witnessing form should capture rather than hide.
Preventive correction. State the intended use of the aggregate on the test request. The sample itself tells you the number; the use tells you which limit applies.
What happens. The witness is still in the room, which is the only reason this is diagnosable. A cube that fails in shear rather than cone is suspect, and the causes are specific: a capping issue, a surface defect, or non-conforming material. The test is stopped, the failure mode is documented in the observation form, an NCR is raised per ISO 9001 Cl 10.2.1, and a re-test is required.
Likely outcome. The witness's signature does not validate a failed test. If the failure is structural, the consultant advises the client of the non-conformance in writing immediately. If the witness had left before the failure moment, there would be no failure mode to document, the re-test could not be justified, and the whole result would be open to challenge.
Preventive correction. Stay for the whole test. Signing a form for a test you did not personally observe is not witnessing, and it is professional misconduct rather than a paperwork slip.
Frequently Asked Questions
When must a test be physically witnessed by the consultant?
What if a test fails in the witness's presence?
Can a witness approve a test result they did not personally observe?
Related articles
Amit Haridas
Founder & Proprietor, ConcreteInfo. 25+ years experience in construction QA/QC, concrete technology, RMC plant operations, consulting and technical training across India. For engagement-letter templates or consulting engagement setup, contact amit@concreteinfo.in.
Technical updates, not sales emails
One email when a new article goes out on concrete technology, QA/QC and construction quality. No promotions, and your address is never sold or passed to advertising networks. Subscribing is separate from making an enquiry.