Low Cylinder Strengths in Post-Tensioned Slabs: What the Engineer of Record Directs
A five-storey reinforced concrete office building uses post-tensioned slabs on an 8.4 m grid, and the specified characteristic cylinder strength is 35 MPa at 28 days. Of 40 standard-cured cylinders from the last 12 pours, seven fell below 35 MPa; the lowest, from pour P-14 at level 2 grid C-F, is 26.8 MPa. The batch plant changed sand sources three weeks earlier, ambient temperature during those pours averaged 34 C, and the placing crew added water on site in two undocumented loads. The contractor argues the slabs are adequate because the failures are 'only 10% of samples', and proposes to accept them without further testing, citing low service-load utilisation. The post-tensioning specialist has already stressed the tendons, and formwork for level 3 is erected. Your firm is the structural engineer of record; the client wants a recommendation within 48 hours so the floor cycle is not interrupted.
1. Assess the contractor's 'only 10% of samples' argument and the low-utilisation argument, and set out the technical path you would follow to accept or reject the affected pours.
Read the worked answer
The 10% argument confuses a proportion of failures with the acceptance rule. Concrete acceptance is a statistical criterion applied to the results, not a majority vote. Typically the average of any three consecutive strength tests must equal or exceed f'c, and no single test may fall more than about 3.5 MPa (500 psi) below f'c. Seven of 40 results below f'c is already a signal; 26.8 MPa is 8.2 MPa below, well outside that tolerance. Verify the exact criterion against the code edition and the specification in force, but the direction of the answer does not change.
Low utilisation is not automatically a defence. A post-tensioned slab may be governed by transfer strength, anchorage zones, punching shear at columns or deflection rather than by the global bending ratio. Low average utilisation does not prove the local demand at grid C-F is low - but it does open a legitimate route to acceptance by analysis if a proper demand/capacity check with the measured strength shows adequate reserve under the correct load combinations and resistance factors.
Path:
- Audit the cheap explanations first: sampling, cylinder handling, curing, laboratory records and calibration, and whether 26.8 MPa is the average of two cylinders.
- Instruct core testing by an independent accredited laboratory, three cores per questionable area, with typical acceptance at an average of at least 0.85 f'c and no core below 0.75 f'c (confirm against the code).
- Run the demand/capacity check in parallel so cores are not the only line of evidence.
- Decide before level 3 concrete is placed - the pour sequence, not the laboratory turnaround, is the real constraint.
Causes: the sand-source change, 34 C placing and undocumented site water are jointly sufficient to explain the loss. Site water alone can add 0.05-0.10 to the water/cement ratio and several MPa of lost strength. That points to a method and workmanship failure, not a design deficiency.
2. What instructions would you issue on site within the 48 hours, and how would you communicate them to the contractor, the client and the authority?
Read the worked answer
Do not stop the whole site. A blanket stop is disproportionate, will be treated as a delay event, and destroys the cooperation you need to get cores drilled. Issue written instructions scoped to the actual risk:
- Hold striking of formwork and removal of props beneath the affected bays at level 2, grid C-F until results are available.
- Permit level 3 works only where the load path does not rely on the suspect pour and where props are not being removed; identify those zones on a marked-up plan.
- No further stressing, grouting or record changes in the affected area without review.
Technical instructions: a core-testing programme with locations, witnessed sampling and an independent laboratory - not the contractor's laboratory. Require the contractor's method statement for drilling, and the make-good procedure.
Records: a written statement on the site water additions - who authorised them, how much, on which pours - and an immediate project-wide prohibition. Obtain the batch plant's mix records and the approval trail for the sand-source change.
Communication: the instruction goes to the contractor, copied to the client, in one document. Notify the authority having jurisdiction if the code requires reporting of non-conforming work or if the structural permit is affected. Give the client a one-page brief with the decision tree and the dates.
On the 48 hours: give a decision on process, not on outcome. The strength question cannot be answered in 48 hours, and you must say so explicitly - silence is otherwise read as acceptance. State the date on which you will give the technical verdict and what evidence it will rest on.
3. The contractor's commercial team argues that if the engineer will not accept the pours, the engineer should share the remediation cost because the mix design was approved. How do you handle causation, records and liability?
Read the worked answer
Keep the technical recommendation independent of the commercial argument. The report should say what is safe and on what evidence; where the cost falls is a contractual question for others. Accepting an unsafe conclusion to protect the programme - or to avoid a claim - is the one outcome that is not available.
Map causation:
- Undocumented site water is normally prohibited by specification and is a contractor non-conformance.
- An unapproved sand-source change is a supplier and contractor issue.
- Hot-weather concreting at 34 C should have been covered by an approved method statement with placing temperature limits.
- The engineer of record's own exposure sits in the specification, the approval of the mix design, and review comments - so retrieve the submittals and the review record now, while memories and files are fresh.
The 'approved mix design' argument only reaches the engineer if the approved mix, placed as specified, could not have produced 26.8 MPa. The site water and placing temperature break that chain, which is why the records matter more than the argument.
Actions: notify your professional indemnity broker early, without admissions; keep the causation investigation separate from the remediation-cost negotiation; agree a single technical basis (the independent laboratory report) rather than duelling experts. Document the contractor's request to accept the pours without testing and your refusal - that request is evidence of knowledge and will matter if the bay underperforms later. Finally, give the client a clear statement that acceptance by analysis, if achieved, is a documented engineering decision with defined limits on future use, not an informal waiver.