The M20 Slab, the Hosepipe and the Failing Cubes
A two-storey house is being built with M20 reinforced concrete for columns and a 4.2 m span roof slab. The approved mix has 320 kg of OPC 43 grade cement per cubic metre and a water/cement ratio of 0.50, with a specified slump of 40 mm. On site the mason says the mix is 'too dry to place', so the labourer adds water from a hosepipe until the slump looks like about 100 mm, roughly 18 kg of extra water per cubic metre. Cubes were cast but were left in the open beside the mixer for two days before being put in a tank, then taken out after 7 days; the tank water was warm. Curing of the slab consisted of the mason splashing water twice a day for four days. The slab props were struck after 7 days; beam side shuttering was removed after 24 hours. The 28-day cube results are 16.4, 17.1 and 15.9 N/mm². The contractor wants to plaster over the slab and move on.
1. The cubes are below M20. Before blaming the concrete, what would you check about how the cubes were made, cured and tested, and why does that matter legally and technically?
Read the worked answer
Cube results are only as good as the cubes. Check, with evidence:
- Sampling. Were samples taken from the middle of the discharge, at the correct frequency (one set per 5 m³ or per batch as specified), and were they compacted in three layers with 25 blows per layer on a standard 150 mm cube?
- Curing. These cubes sat in the open for two days, so they dried and were probably overheated; the standard requires 24 ± 0.5 h in the mould in moist conditions, then curing at 27 ± 2 °C until the test. Warm tank water and 7-day removal will both depress strength. This alone can explain several N/mm².
- Storage and transport. Cubes should be stored away from vibration, and brought to the testing laboratory promptly. Cubes rattling in a trolley on a village road lose strength.
- Testing. Confirm the laboratory's calibration certificate for the compression machine, the rate of loading, and the individual cube dimensions (a cube 2 mm out of square reads low).
Why it matters. If the cubes fail because of bad sampling or curing, the structure may still be sound; if the cubes are good, the concrete in the slab is suspect. The cube is evidence about the concrete, not the concrete itself. Legally, acceptance under the code is decided on groups of results and on the specified criteria, not on one bad cube, so keep the test certificates, the sampling record and the laboratory report together. Also check the cement: 320 kg/m³ of OPC 43 with no admixture, stored how? Bags stacked 12 high against a damp wall go lumpy and lose 20-30% of their strength.
2. Quantify the effect of the extra water, and explain what the mason should have been given instead. What limits would you enforce on site?
Read the worked answer
The arithmetic. With 320 kg/m³ of cement at w/c 0.50 the mix already contains 160 kg of water per m³. Adding about 18 kg raises the water to 178 kg, so the effective w/c becomes 178/320 = 0.556 - an increase of 0.056, more than 11%.
The consequence. Strength falls roughly in proportion as w/c rises; a rule of thumb for normal concrete is a drop of about 10-15% in compressive strength for every 0.05 increase in w/c, until the mix becomes too wet to be reliable. Here expect roughly a 10-12% loss, plus bleeding, segregation, more shrinkage cracking and a lower modulus. On an M20 mix that takes the likely strength from about 20-27 N/mm² down towards 18-20 N/mm² before any other factor, and then the bad cube curing is added on top. That is exactly the direction the test results show, which is why the water addition is the first suspect.
What the mason needed. Workability is legitimate, but water is the wrong tool: the correct fixes are a slightly higher cement content (within the mix design), proper grading and moisture control of the sand and aggregate, accurate batching by weight rather than by shovel and bucket, and, if permitted by the specification, a water-reducing admixture or plasticiser dosed to the manufacturer's instructions.
Limits to enforce:
- w/c strictly at or below the approved figure, with a maximum permitted by the specification for the exposure class (check the specified clause; do not exceed it).
- Slump tested at each pour and recorded against the specified 40 mm, with a tolerance agreed with the engineer.
- A locked water measure, an ice-cream or measuring can, and water added only by the supervisor.
- No hosepipe on the platform.
3. The contractor wants to plaster over the slab. What is your technically sound and defensible course of action with the slab and with the concrete already cast?
Read the worked answer
Plaster is not a repair; it hides a problem and adds load. The correct sequence:
- Do not cover. Refuse to allow plastering, and put that instruction in writing with reasons.
- Verify the concrete in place, not just the cubes. Arrange non-destructive testing: rebound hammer for uniformity and ultrasonic pulse velocity for homogeneity. Then take cores from a specified location, avoiding the reinforcement, and test them. A core result is direct evidence about the slab.
- Check the structure. Ask the structural engineer to review the results against the design. If the strength is marginal, a load test or a redesign, such as extra reinforcement or an increase in screed thickness, may be needed; if it is seriously low, partial demolition is the honest answer.
- Check the workmanship that goes with the water. Look at the slab for honeycombing, bleeding channels under the reinforcement, and cracks. Check that the props were struck at 7 days, when the code allows props to a slab spanning up to 4.5 m to be removed after 7 days, so 4.2 m was just acceptable; side forms of beams after 24 h are also acceptable, because they carry no load. The real curing failure is four days of twice-daily splashing against the 7-10 days the specification probably requires.
- Fix the system, not just the batch. Re-brief the mason and labourer, put the water measure and slump testing in place, and record everything: cube results, NDT reports, core results, engineer's decision, and your instructions.
The professional point is that you neither accept the slab because 'it looks fine' nor condemn it because one cube is low. You collect evidence, let the designer decide, and never let the next stage bury the problem.