Quality control - Concrete Mix Design, Quality Control and Specification - page 241

such a failure. Both these possibilities should be avoided by the provision of
a circular shield during testing.
10.6 DATA RETRIEVAL AND ANALYSIS/
CONAD SYSTEM
10.6.1 Coping with data
A basic challenge in the quality control of concrete is to cope with the avail-
ability of possibly excessive amounts of data. There is no doubt that facts
can be harder rather than easier to deduce if included in more data than
a person can cope with. It should not be forgotten that quality control is
an exercise in cost reduction and that cost includes the cost of the quality
control. A better quality concrete can be purchased at a higher price, but
the task of quality control is to deliver concrete of a chosen quality at the
minimum cost.
So the value of given data should be considered alongside the cost of
acquiring, storing, analysing, and employing those data. In particular no
substantial cost should be incurred in acquiring and storing data that will
definitely not be used. On the other hand, storage of huge amounts of data
is no longer a problem, providing it can be acquired at negligible cost and
effort and the precise data needed can be automatically retrieved with little
effort.
An example of inadequate cost-benefit occurred in the 1980s in New
York where inspectors were employed to manually write down batch
quantities at substantial cost, but no analysis of the acquired data was
carried out. In contrast batch quantities (intended and actual) are auto-
matically acquired electronically by the ConAd system, are automatically
matched with test data on tested loads, and errors can be automatically
displayed either numerically or graphically. In the latest development, the
system can automatically e-mail or telephone selected personnel to advise
of errors, and can predict the strength of a miss-batched load. Long-term
trends in inaccuracy can be precisely displayed graphically. Of course
these facilities require both suitable batching equipment and a suitable
analysis program.
Other data that can be automatically acquired include details of the orig-
inal order, so that a field testing officer only needs to record a batch num-
ber and his actual measurements. Also many laboratory testing machines
are able to output test results direct to a laboratory computer. This not
only saves time but also avoids the possibility of error in transference and
the necessity to check for such errors. Not only crushing loads but also
weights and dimensions of compression specimens are often automatically
recorded.