Testing - Concrete Mix Design, Quality Control and Specification - page 138

Rather we should regard the analysed pattern of test results as an
important part (but only part) of the evidence we require to establish
whether the totality of concrete being delivered to the project (or
leaving the plant) is or is not of the required quality.
4. Before concrete of a particular grade is even ordered, it should be
established that it is almost certain to be satisfactory. This may be
done on the basis of trial deliveries, laboratory trials, analysis of
past data, or even just the reputation of the supplier. This assess-
ment needs to take into account variability as well as mean strength.
For an important project it may be inadvisable to obtain concrete
from a supplier who cannot show either or both substantial analy-
ses of past data showing low variability or a computer batching
plant that records the actual batched weights of every truck load
delivered.
5. A particular individual (perhaps with assistants on a major or
widely spread project) should have the responsibility of visually
inspecting every truck of concrete and rejecting or further testing
any suspect loads.
6. When a truck is sampled and test specimens cast, there should
normally be at least three specimens. This is to permit an early-age
test and a pair of 28-day tests although it is better to have a pair
of results for the early age. The early age (not later than 7 days) is
because any necessary mix adjustments must be carried out long
before 28-day tests are carried out. The 28-day test is necessary to
establish the current significance of the early age results. Two 28-day
specimens are needed partly because the average pair difference is the
best measure of testing quality and partly so that one can be brought
forward to confirm or amend a low early-age test result.
7. The sampling procedure should also include measuring and record-
ing slump and concrete temperature, and also cylindercube density
on receipt at the laboratory. This is because such information is less
expensive to obtain than the compressive strength, yet at least dou-
bles the value we can extract from it. Entrained air tests are also
useful, but this test is a little more expensive so it is not invariably
justified. J.M. Shilstone (1987) has suggested that the fresh density
of concrete may be a better quality indicator than slump. If taken
it should certainly be combined with an air content determination,
but it involves on site weighing equipment and it is not so simple to
attain the required precision. Also it is not such a direct check on the
relative water content of successive loads. It may be that hardened
specimen density is sufficient providing that it is measured on receipt
of the specimens at the laboratory (i.e., within 24 hours) and that it
is immediately followed up by air testing when a significant density
change is experienced. It may be that fresh density measurement is