Cube non-conformity in a raft foundation pour
A 120 m³ raft foundation is being cast with C30/37 ready-mixed concrete. The raft is 600 mm thick and lightly reinforced. Ten sets of 150 mm cubes were taken and tested at 28 days. The results, in MPa, were 33.5, 32.0, 29.0, 25.5, 31.0, 34.0, 30.5, 28.0, 35.0 and 31.5. The mean is 31.0 MPa and the sample standard deviation is about 2.9 MPa. The specified characteristic strength is 30 MPa. The client's engineer wants to reject the whole raft because the lowest cube is below 30 MPa. The contractor argues that the average is above 30 MPa and that the raft is lightly stressed, so the concrete should be accepted. The next pour is scheduled for tomorrow, using the same mix and the same testing laboratory.
1. Assess conformity of the cube results against a recognized acceptance rule, showing your calculations. Is rejection of the whole raft justified?
Read the worked answer
Mean = 31.0 MPa; sample standard deviation s = 2.9 MPa; lowest = 25.5 MPa. Under EN 206/BS 8500 site-cube conformity criteria for the applicable production case, the group mean must normally be at least f_ck + 4 = 34 MPa for the initial assessment, and no individual result should fall below f_ck − 4 = 26 MPa. The mean of 31.0 fails the 34 MPa requirement, and 25.5 fails the 26 MPa individual limit. Under ACI 318, the average of three consecutive results must equal or exceed f_c' and no individual result may fall below f_c' − 3.5 MPa = 26.5 MPa; again 25.5 fails. So the results indicate non-conformity and must be investigated. Rejection of the entire 120 m³ raft is not automatically justified, however: ten cubes sample a small fraction of the pour, and a single low result can come from sampling, curing or testing error. The correct decision is to withhold acceptance, review the records and carry out an in-place assessment rather than condemn the whole raft on the cube average alone.
2. Describe a defensible core-testing and assessment program, including how you would decide between local repair and acceptance.
Read the worked answer
First audit the batching plant records, delivery times, cube storage, curing and testing-machine calibration, and identify which truck or section produced the low cubes. Take cores from locations representing the critical sections and the suspect concrete, not only from convenient spots. Test them to EN 12504-1 or ASTM C42 and correct for drilling damage, moisture and age. Estimate in-place strength using EN 13791 or ACI 214.4, and compare it with both the specified characteristic strength and the actual stress demand in the raft. If the corrected in-place strength at the critical sections is adequate and the low cubes are explained by testing error, accept the raft with documentation and monitoring. If isolated zones are low, repair or strengthen locally, for example by local thickening or FRP, after checking the raft's load path. If the assessment shows widespread deficiency, reject or redesign. The decision should be made by the designer, not by the contractor or the testing laboratory alone.
3. What changes should be made before tomorrow's pour if the same mix and laboratory are used?
Read the worked answer
Tomorrow's pour should not proceed as a repeat of the same uncontrolled process. Before casting, check the mix design and batching accuracy, aggregate moisture, water/cement ratio, cement content, admixtures, delivery time and curing regime. Review the cube results by truck and by test age; if the low result is a testing artefact, the laboratory's procedures and calibration must be corrected. Increase sampling and require cubes from the first and last trucks. Set a target mean strength with an adequate margin, typically f_ck + 2s or more, rather than aiming just above 30 MPa. Consider a trial mix or a mix adjustment if the data show a low mean. The client's engineer should approve the revised quality plan, and the next pour should be held until the cause of the low cube is understood.