Page 95 - vol5_n2

225 IBRACON Structures and Materials Journal • 2012 • vol. 5 • nº 2
J. TANESI | M. G. da SILVA | V. GOMES
[19], Cascudo [20]). Furthermore, it is an easy test to run.
Although, CEB 192 (CEB [21]) proposes guidelines for the cor-
rosion risk assessment using electrical resistivity, there is no
mathematical model that uses electrical resistivity as an input
to predict service life. As a consequence, electrical resistivity
cannot be understood as a performance-based specification
requirement but as a requirement for hybrid specifications.
Since the current is carried by the ions in the pore solution,
several factors can influence the electrical resistivity, such as
the size and quantity of pores (for example, with increased
water-binder ratio), moisture, degree of hydration, concrete
carbonation and salt ingress (Millard; Gowers [22]). Therefore,
the electrical resistivity should be analyzed in conjunction with
other techniques (Andrade; Alonso [23]).
3.2.2.2.2 Discussion on possible requirements related
to carbonation induced corrosion
n Carbonation coefficient – Carbonation is not the direct cause
of concrete deterioration, but its effects can significantly impact
the durability of the structure. For example, shrinkage due to
carbonation and the reinforcement despassivation can deci-
sively contribute to reinforcement corrosion.
The carbonation depth is usually measured by spraying phe-
nolphthalein over the concrete surface. A very simple math-
ematical model (model Ushida-Hamada, 1928) may be used to
obtain the carbonation coefficient k. In this case, a maximum
k can be specified in order to prevent the carbonation to reach
the reinforcement.
The natural carbonation is a very slow process that may take years.
In addition, several factors can influence the natural carbonation
test, such as internal humidity of concrete, the environmental condi-
tions (temperature and humidity) and curing condition. As a conse-
quence, it is practically impossible to reproduce in the laboratory the
conditions to which the structure is expected to be subjected to, so
the performance presented in laboratory studies may not necessar-
ily represent the concrete behavior in service.
The accelerated carbonation can reduce testing time, but does
not represent the exposure conditions in service and does not
correlate with natural carbonation exposure.
Sanjuán et al [24]; Bier et al. [25] and Tanesi [2] observed that
different mixtures respond differently to accelerated testing. In
other words, there seems to be different degrees of microstruc-
ture change for different concentrations of CO2. Although coeffi-
cient of carbonation is in theory a good requirement, the fact that
the accelerated carbonation is not necessarily comparable with
the natural carbonation, makes its use basically impractical.
3.2.2.2.3 Discussion on possible general durability
requirements
n Absorption – mass transport plays an important role in various
deterioration mechanisms, such as corrosion, carbonation, sul-
fate attack, chloride penetration and alkali-aggregate reaction.
Nevertheless, this property is not recommended as a require-
ment since there are no clear criteria for this property neither
a mathematical model for service life prediction that uses this
property was found. Nevertheless, this property can be used
only as a tool to compare mixtures.
n Drying Shrinkage – free drying shrinkage can be measured
by NBR 8 490/1984 (ABNT [26]) and ASTM C 157-08 (ASTM
[27]). These standards are similar. Since they measure the free
shrinkage, they do not represent the behavior of the structure,
which is normally restrained. NBR 8 490/1984 (ABNT [26]) only
provides a limited indication of volume stability.
No service life prediction model was found that includes shrink-
age as one of its inputs. However this can be a useful tool for
mixture comparison.
n Cracking – cracking is usually due to restrained shrinkage.
There are two standardized test methods to assess cracking
susceptibility ASTM C 1581-04 (ASTM [28]) and AASHTO PP
34-99 (AASHTO [29]), both ring tests, which was originally de-
veloped by Coutinho in 1954 (Tanesi [30]). These test methods
present high variability and do not measure a fundamental ma-
terial property. They do not simulate the conditions that may
be encountered in the field. This test method typically presents
the age for the onset of the first crack. No recommendation for
cracking was found in the current Brazilian specifications.
3.2.2.2.4 Special requirements
n Sulfate resistance – NBR 13 583/1996 (ABNT [31]) assesses
sulfate resistance of mixtures. There is no recommendation in
the Brazilian specifications with respect to a maximum allowed
expansion that would guarantee the mixture to be resistant to
sulfate attack.
n Abrasion – in special cases, as in floors and pavement, it may
be necessary to specify abrasion resistance. Abrasion resis-
tance is usually indicated by means of the mass loss or the
depth of abrasion. Several test methods have been used for
the evaluation of this property, as for example, ASTM C 779-05
(ASTM [32]) and ASTM C 944-05 (ASTM [33]).
n Alkali-silica reaction – when the aggregates to be used in the
job are prone to alkali-silica reaction, recommendations should
be provided in the specification in order to avoid the reaction,
either by changing the aggregate source or by adopting pre-
ventive measures, such as the addition of blast-furnace slag.
There are several test methods for the detection of the po-
tential for the alkali-silica reaction, such as ASTM C 1260-07
(ASTM [34]), ASTM C1293-08 (ASTM [35]), ASTM C289-07
(ASTM [36]) and ASTM C 1567-08 (ASTM [37]).
n Step 5: the applicability and reliability of the available service
life predictive models should be reviewed. Table 3 presents a
summary of some requirements relating to durability that are
used as inputs for the prediction of the service life.
If no models are available for some of the selected require-
ments, the criteria should be chosen according to the existing
technical recommendations.
n Step 6: The service life model should be selected according to
appropriate technical principles and should be consistent with
the deterioration mechanisms expected to affect the structure.
Then, the models should be applied, incorporating the mini-
mum service life established in step 1.
n Step 7: The performance criteria is determined by applying the
desired minimum service life to the models. In addition, the cri-
teria for the special requirements should be determined.
n Step 8: Requirements for quality control have to be specified.
They should correlate with the performance requirements se-