How Does Concrete Strength Testing Actually Work on Site?

Concrete strength testing casts sample cylinders during placement, cures them under standard conditions, and crushes them at 7 and 28 days to verify the concrete meets its specified compressive strength. Acceptance is based on the 28-day average of each set; drilled cores investigate when cylinders fail or are missing. The whole system exists because concrete is the one structural material made on site, under field conditions, from ingredients that vary. Steel arrives with a mill certificate; concrete arrives as a promise in a mixer truck. The cylinder program is how that promise gets verified — and it is the only verification that happens before the structure is loaded.

It starts before the first truck arrives: the lab is engaged, the mix designs are submitted and reviewed, and the cylinder frequency is written into the special inspection plan. During each placement the technician samples from the middle of the load — not the first or last concrete out of the truck — casts the cylinders in standard molds, and protects them from disturbance, freezing, and drying for the first 24 hours. The 7-day break is the early warning: it typically reaches about two-thirds of the 28-day strength, so a low 7-day result flags a problem while there is still time to investigate. The 28-day break is the acceptance event. The code's acceptance criteria look at the average of the set and the lowest individual result — a single low cylinder does not automatically fail the concrete, but a pattern of low results triggers the investigation sequence: review, cores, structural evaluation, and only then remediation.

Most concrete disputes I have seen were not about bad concrete — they were about bad samples. Cylinders left in the sun, dropped during transport, or cured in a job trailer instead of the lab produce low breaks that condemn perfectly good concrete. The sampling and curing discipline matters as much as the mix design. The non-negotiables I set on every project.

• Write the cylinder frequency into the special inspection plan before the first pour — no retroactive sampling exists
• Sample from the middle of the load and protect cylinders from heat, freezing, and disturbance in the first 24 hours
• Treat the 7-day break as an early warning system, not a formality — investigate low results immediately
• Keep batch tickets with every set so a failure can be traced to a specific load and plant
• Have the core-drilling contingency planned in advance so a failed set does not stall the schedule