Testing - Concrete Mix Design, Quality Control and Specification - page 162

The upper limit for which the slump test can be used is very dependent
on the type of concrete. Harsh, gap-graded concrete (MSF of 20 or less; see
Table 8.1) will fall apart on a slump test at slumps not much higher than
50 mm. On the other hand, continuously graded mixes of high sand con-
tent (MSF of 27 or more) will give a measurable and reasonably repeatable
slump up to 200 mm or more.
The technique of carrying out a slump test is also important in obtaining
a true reading and it should be realised that the slump itself is measured
in different ways in the United States (to midpoint), United Kingdom (to
highest point), and Australia (to average see 117).
What is important is not necessarily to stop using the slump test but to
realise and allow for its limitations. For example a limiting slump value
is often included in a job specification. With few exceptions, this is not
the best way to achieve the specifier's objective. First, there should be
an objective for the specification of anything, rather than it having been
included in a previous specification and so mindlessly continued in the cur-
rent document. The objectives may be to avoid high shrinkage, segregation,
and bleeding, or to avoid an excessive water to cement (w/c) ratio leading
to inadequate strength or durability. However, any of these faults can be
encountered at almost any slump, however low, and avoided at any slump,
however high. It is also easy to detect from a theoretical mix submission
which mixes will be subject to one or the other of these problems. The
contractor should therefore be permitted to submit his mix for approval
at whatever slump he chooses, provided it is designed to accommodate his
own slump limit without detriment. It is quite possible to produce fully
flowing (250 mm slump or more) concrete having none of the potential
faults noted and to produce almost all these faults in a 50 mm slump mix.
The rejection of a truckload of concrete on the basis of slump should
also be approached in a reasonable manner. The slump test is both quite
sensitive to small changes of water content and very easy to perform inac-
curately. Certainly the truck driver should always be allowed to insist on
the test being repeated. An extra 10 mm of slump probably involves about
an extra 3 liters of water per cubic metre of concrete and may depress
strength by about 1 MPa. The person charged with concrete acceptance
should be kept continuously aware of the current strength margin of the
mix in question and therefore of whether it is essential to reject slightly
overslump concrete on strength grounds (and similarly for any shrinkage
limit that may have been specified). It is more usual to find that a need to
reject first arises on the grounds of wet properties or surface appearance.
Slump variation will cause colour variation on a fair faced wall and slump
in excess of that designed for can involve segregation, bleeding, delayed
finishing, or floors of poor wear resistance.
Rejecting concrete for some petty reason such as a nominally noncompli-
ant slump, temperature, and delivery time may result in an actual problem