Properties of concrete - Concrete Mix Design, Quality Control and Specification - page 93

any excess of either coarse aggregate or mortar would be left on top. If the
top half were discarded, then the proportions of the bottom half would be
a reasonable guide to the desirable sand percentage to use. This is a useful
exercise for students since it illustrates the concept of filling the voids in the
coarse aggregate with mortar and demonstrates that an ideal mix cannot
be overvibrated once it is fully compacted in place (in that the remaining
concrete will not further segregate however long it is vibrated).
Very high strength depends on a number of other things besides w/c
ratio. These include the strength of the coarse aggregate, and the bond
between the matrix and the coarse aggregate. It used to be very difficult
to achieve a strength much in excess of 90 MPa (13,000 psi). Strengths
of double this amount can be obtained given a strong coarse aggregate,
silica fume, and a superplasticising admixture. Day recalls carrying out
trial mixes for 60 MPa concrete in the late 1970s before either silica fume
or superplasticiser were available. Of the two coarse aggregates tried, the
stronger one gave unsatisfactory results. This was because it was such a
hard impermeable material that the matrix did not bond to it sufficiently.
With silica fume and superplasticising admixtures now available, excellent
bond was developed and the stronger coarse aggregate gives better results
than the other and both can easily exceed 100 MPa.
There are two words of caution about using very high concrete
strengths. One is that concrete in a structure cannot be saturated with
water as can test cylinders or cubes in a water bath. It will have a w/c
insufficient to provide full hydration and will therefore self-desiccate and
not develop the full strength of the test specimens. At best it may be pos-
sible to prevent the loss of any of the mixing water by polythene wrap-
ping immediately on demolding or placing the concrete in permanent
formwork such as a steel pipe column. So perhaps high strength test
specimens should be polythene wrapped rather than water-bath cured,
although this should probably be restricted to a few comparison tests,
since it may be undesirable for quality control from the viewpoint of
introducing variability into the results. The opposite problem occurs
when the high strength test specimens dry out due to poor sampling and
early protection. Because of the low penetrability, the specimens do not
absorb water on immersion. The test specimens may give satisfactory
early strength but significantly reduced strength compared to the in situ
concrete at later stages. The provision of saturated lightweight particles
in a mix to provide internally the water for curing (Bentz et al., 2005)
helps maximise the performance of very high strength concrete and also
helps address the problem of autogenous shrinkage. Another suggestion
has been to use a proportion of reactive magnesia to perform a similar
function (see Chapter 13).
The other problem with very high strength concrete (actually very
low penetrability concrete) is that of explosive failure in a fire situation.