Testing - Concrete Mix Design, Quality Control and Specification - page 182

penetrability). Potentially a contractor could be required to continue or
resume water curing until an acceptable permeability is achieved.
7.6.6 Chloride diffusivity
Ions will naturally diffuse through the water-filled pores and microcracks
present in concrete from areas of high concentration to areas of low
concentration. This process will occur without any hydraulic gradient
being necessary and thus is different from pressure permeability. Chloride
diffusivity plays an important role in long-term performance, particularly
after a high salt concentration has been established in the surface layer as
a result of sorpivity.
Laboratory measurement of diffusivity used to be conducted using a dif-
fusion cell. The time and equipment required to effectively measure diffu-
sivity using this procedure limited its use to research applications.
Bulk diffusion tests, such as the Nordtest NT Build 443 or ASTM
C1556, expose saturated concrete to highly concentrated chloride solution
for a minimum of 35 days. Profile grinding and testing for chloride content
enable the chloride diffusion coefficient to be calculated using Fick's second
law. Therefore a chloride diffusion coefficient can be measured in two to
three months. This may be suitable for verification of trial mix properties
but not ongoing compliance testing. One appeal of chloride diffusion test-
ing is that it can be used directly in service life prediction models based on
Fick's law. An increasing number of specifications require chloride diffu-
sion to be measured, often with the shorter coulomb test or migration test
to facilitate compliance testing during construction.
Chloride diffusivity can reduce by orders of magnitude during the service
life and therefore a single measurement of early age diffusion alone cannot
predict the long-term performance. The improvement over time is primar-
ily due to the composition of the cementitious binder. This is why many
specifications require minimum replacement levels of fly ash or GGBS in an
attempt to ensure certain improvement in diffusivity over time. However,
prescriptive requirement on cementitious materials is contrary to the aim of
performance specification. Thomas and Stanish (2003) measured chloride
diffusion for periods from 90 to 180 days and 90 to 1550 days to establish
the time-dependent effect as shown in Figure 7.16. Clearly projects cannot
wait for 4 years to get a result but the research suggests a series of early-age
tests could help confirm model assumptions.
Another limitation of bulk diffusion tests is the difference between
actual chloride penetration over time and the measured chloride diffusion
coefficient. Figure 7.17 shows the calculated chloride diffusion coefficient
from the in situ chloride profiles after 19 years exposure and the chloride
diffusion from NT Build 443 test on the uncontaminated concrete (Vallini
and Aldred, 2003). The chloride diffusion from profiles is generally one to