Troubleshooting - Concrete Mix Design, Quality Control and Specification - page 273

12.2 CAUSES OF CRACKING IN CONCRETE SLABS
The causes of cracking in concrete are sufficiently well known to permit
their automatic diagnosis in most cases. Day wrote an expert computer
system for this purpose. An expert system is a computer program that asks
questions of a user in order to be able to diagnose the cause of the user's
problem; the better ones are also able to explain why the particular ques-
tion is being asked, on request by the user.
The first question to be asked is the age of the concrete at cracking. If
the age was less than 10 hours, the crack would be classified as a plas-
tic or presetting crack caused by either excessive evaporation from the
surface or by restrained or differential bleeding settlement. If the age
was more than 10 hours but less than 48 hours (and especially if the
crack occurred in the early morning following pouring) the crack would
probably be a thermal contraction crack. If the age exceeded 2 days (and
was after termination of covered or moist curing if any) it may be due to
drying shrinkage.
To determine whether plastic or presetting cracks are caused by evapora-
tion or settlement, questions are asked about the shape, size, and orienta-
tion of the crack, and about whether the concrete bled substantially or
was subjected to drying winds and low humidity. Plastic shrinkage cracks
may be quite wide on occasions, but they are usually short and randomly
orientated. However, they can sometimes be concentrated in an area of
the slab that is more exposed to wind and can form parallel lines. In the
latter case they may be more difficult to distinguish from settlement cracks
occurring over a steel mesh, except that it would not be likely that evapora-
tion cracks would be parallel to the direction of the mesh, or at the same
spacing. As already noted the settlement cracks can occur over reinforcing
bars, installed plumbing, or the like. They can also occur at lines where the
section deepens, such as dropped capitals for columns, haunched beams, or
the edge of thickened areas of a slab (waffle slabs, slab/beam).
A classic situation for thermal cracking exists when a concrete wall is
poured between restraints. The restraints may be a heavy foundation beam
with starter bars or substantial columns with projecting reinforcement.
When a wall in such a situation is poured on a warm afternoon using a mix
rich in a Portland cement the width of the crack to be anticipated on strip-
ping next morning can be calculated if a maximum reading thermometer
is inserted. Such cracks are often widest at the base, near to the restraining
foundation beam, and taper away to nothing 2 or 3 metres up the wall.
A commonly encountered situation is where a crack runs parallel to, and
often close to, a sawn control joint. It is easy to see that either the joint
was not deep enough to be effective or, more likely, it was actually cut
after the slab had already cracked, although perhaps before it had opened
sufficiently to be noticeable.