Structural aspects of repair - Concrete Repair - page 86

It is tempting, but unhelpful, to produce a repair material which is markedly
superior to the original material in terms of strength, carbonation resistance,
permeability and diffusion resistance. Achieving better performance in the
repair than the surrounding un-repaired areas will not extend the overall life
of the structure. The more 'durable' material will have differential heating
and wetting and drying characteristics, which can produce differential strains
which will tend to delaminate the repair.
The dominant cause of delamination and failures of repairs to carry
their share of loads on the structure is differential shrinkage as the repair
dries into equilibrium with the adjacent concrete. Shrinkage depends on
water cement ratio (w/c) and aggregate cement ratio (a/c) as shown in Table
3.1. So a repair mix needs to minimise water cement ratio and maximise
aggregate cement ratio. Except for the largest repairs, there is advantage in
reducing the maximum aggregate size to 10 mm to aid compaction. This,
with an aggregate cement ratio of 6, should be the starting point for mix
development. To keep the water cement ratio low and to aid compaction
and adhesion, limited amounts of admixtures (super plasticiser or SBR) and
adjustments to cementitious materials can be used subject to maintaining
compatibility, with a cement content similar or a little less than the original.
Test data on long-term shrinkage is needed, as strain v. time plots for
proprietary mixes are seldom supplied with product details. While water
cement and aggregate cement ratios achieved in site repair trials are a guide
to this, test data should always be obtained for the site mix. The CIRIA
(1993) test is simple and robust and provides results which can be directly
related to long-term site performance. When trial repairs are carried out
sample prisms should be monitored for shrinkage on site where the drying
conditions match those of the structure.
A perennial delusion is that expansive cements will compensate for long-
term shrinkage. The expansion from the development of ettringite over the
period of hydration, typically the first 2 days, needs to occur before the paste
has hardened to the stage where the ettringite damages it. In a contained

Table 3.1 Shrinkage (microstrain) of prisms to 50% relative humidity (Neville, 1995,
Table 9.3) . These shrinkages need to be compared to the tensile strain to cracking for
concrete and mortars of typically 150 to 200 microstrain.

$WZF
DF

800

1200 -

550 850 1050

400 600 750

300 400 550