You are on page 1of 14

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/299388552

ASTM C1202 rapid chloride penetrability test: a new look

Article  in  Journal of ASTM International · January 2006

CITATIONS READS
0 5,871

1 author:

Moncef L. Nehdi
The University of Western Ontario
342 PUBLICATIONS   4,068 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Development of Design Guidelines for Grouted in-conduit connections View project

Innovations in Pre-placed Aggregate Concrete, also Known as Two-STage Concrete View project

All content following this page was uploaded by Moncef L. Nehdi on 25 March 2016.

The user has requested enhancement of the downloaded file.


Journal of ASTM International, March 2006, Vol. 3, No. 3
Paper ID JAI13403
Available online at www.astm.org

M. T. Bassuoni,1 M. L. Nehdi,1 and T. R. Greenough1

Enhancing the Reliability of Evaluating Chloride Ingress in


Concrete Using the ASTM C 1202 Rapid Chloride
Penetrability Test

ABSTRACT: The rapid chloride penetrability test 共RCPT-ASTM C 1202兲 is commonly used to evaluate the
resistance of concrete to chloride ions ingress owing to its simplicity and rapidity. However, it has been
criticized for various shortcomings such as giving favorable results to supplementary cementitious materials
共e.g., silica fume兲, and bias against calcium nitrite corrosion inhibitors 共CNI兲. Based on the ASTM C 1202
induced voltage concept, this study aims at enhancing the reliability of rapidly evaluating concrete resis-
tance to chloride ions ingress. It is proposed that upon test termination, not only the passing charges are
recorded, but the depth of chloride front migrating into concrete is also measured. Concrete mixtures herein
were prepared incorporating selected materials 共silica fume and CNI兲 that have been known to cause
misleading results for the RCPT test. The effects of silica fume and CNI dosages on RCPT results were
investigated and correlated to porosity trends evaluated by mercury intrusion porosimetry 共MIP兲. The study
reveals that measuring the migrating chloride front in concrete subsequent to the ASTM C 1202 test can
eliminate the bias induced by electrolysis conductivity resulting from silica fume and/or CNI. This improves
the reliability of assessing concrete resistance to chloride ions penetration using the RCPT procedure.
KEYWORDS: rapid chloride penetrability test 共RCPT兲, chloride ions, silica fume, calcium nitrite cor-
rosion inhibitor 共CNI兲, migration coefficient, resistivity, porosity

Nomenclature

A ⫽ Cross-sectional area of specimen 共m2兲


CNI ⫽ Calcium nitrite corrosion inhibitors
CR ⫽ Corrosion rate 共␮m Fe/yr兲
D ⫽ Non-steady-state migration coefficient 共⫻10−12 m2 / s兲
dE / dx ⫽ Electric field strength 共V/cm兲
I ⫽ Current 共A兲
L ⫽ Thickness of specimen 共mm, m兲
MIP ⫽ Mercury intrusion porosimetry
P ⫽ Porosity 共%兲
T ⫽ Average value of initial and final temperatures in anolyte solution 共°C兲
RCPT ⫽ Rapid chloride penetrability test
RMT ⫽ Rapid migration test
t ⫽ Time 共s, h兲
V ⫽ Applied voltage 共V兲
x ⫽ Distance 共cm兲
xd ⫽ Average value of penetration depth 共mm兲
␮ ⫽ Ionic mobility 共cm2 / V · s兲
␳ ⫽ Resistivity 共⍀ . m兲

Manuscript received March 2, 2005; accepted for publication September 14, 2005; published January 2006.
1
Department of Civil and Environmental Engineering, The University of Western Ontario, 1151 Richmond Street, London, ON,
Canada, N6A 5B9.

Copyright © 2006 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959.
2 JOURNAL OF ASTM INTERNATIONAL

Introduction

Background
Corrosion of embedded reinforcement in concrete structures is a key durability issue. Generally, corrosion
related costs are estimated to be in the range of 3 to 5 % of the Gross National Product 共GNP兲 in the U.S.,
Western Europe, and Japan, 25 % of which can be ascribed to corrosion of steel reinforcement in concrete
structures 关1兴. It is well established that chloride ions can destroy the passive layer surrounding steel
reinforcement provided by the high alkalinity of cement paste, thus activating electrochemical corrosion
关e.g., 关2兴兴.
Several approaches have been employed in concrete mixtures to minimize the risk of corrosion of
embedded reinforcing steel, such as incorporating appropriate dosages of supplementary cementing mate-
rials 共silica fume, fly ash, slag, etc.兲 and corrosion inhibiting admixtures. Silica fume possesses an ultra-
fine particle size 共around 0.1 ␮m兲 and produces a pozzolanic reaction with portlandite resulting in a
refined interfacial transition zone 共ITZ兲 between cement paste and aggregates. It helps achieving denser
microstructure, reduced permeability and increased electrical resistance of concrete 关3兴. Consequently, one
of the most prominent applications of silica fume is in the protection of embedded steel reinforcement
against chloride ions ingress 关4兴.
Corrosion inhibiting admixtures delay the onset of steel reinforcement corrosion and reduce its rate.
There are different types of these admixtures with various chemical and physical inhibition mechanisms.
Calcium nitrite-based inhibitors are a common example of inorganic corrosion inhibitors and are often
referred to as CNI. They are usually added in dosages within the range of 5 to 30 L / m3 of concrete
depending on the severity of chloride ions exposure 关5兴. The nitrite ions function through a healing process
to preserve the passive layer surrounding steel reinforcement. When chloride ions attack weak spots of
ferrous oxide on the passivating layer, corrosion is initiated and corrosion products are generated. This
exposes pure iron atoms, which quickly turn into the unstable ferrous oxide state. The healing action
occurs when nitrite ions move rapidly to surround the newly exposed ferrous oxide sites, shielding them
from chloride ions attack. Also, nitrite ions help in converting ferrous oxide into ferric oxide that forms the
protective passive layer 关6兴. Because of this competition between chloride and nitrite ions over vulnerable
spots on steel reinforcement surfaces, the mechanism of most CNIs produced in the market is based on a
critical chloride-to-nitrite ratio of 0.92 关5兴.

Rapid Chloride Penetrability Test (RCPT)


Several techniques have been employed, though some of which remain questionable, for the evaluation
and prediction of corrosion damage in reinforced concrete. These techniques include chemical analysis of
chlorides, rapid chloride penetrability test 共RCPT兲, half-cell potential, ponding test, linear polarization,
resistivity, and radiography 关7兴. The RCPT is commonly used to evaluate the resistance of concrete to
chloride ions ingress through electrical conductivity measurements. In the mid-1970s, because of the
extensive use of deicing salts, the Federal Highway Administration 共FHWA兲 was interested in evaluating
the ingress of chloride ions in the concrete cover towards steel rebars. Construction Technology Labora-
tories of Skokie, Illinois, was assigned the task of developing an instrument that gives a rapid indication
of the resistance of concrete to the infiltration of chlorides. At that time, there were rapid methods for
assessing the permeability of concrete to air and water; however, these results did not correlate well with
that of diffusion of chloride ions in concrete 关8兴.
The original concept of the RCPT depended on ionic mobility. Chloride ions are forced into concrete
through the induction of an external direct voltage on the concrete surface, and ionic mobility was
calculated by:
x dE
␮= 共1兲
t dx

where
␮ ⫽ Ionic mobility 共cm2 / V · s兲
x ⫽ Distance 共cm兲
BASSUONI ET AL. ON CHLORIDE INGRESS 3

t ⫽ Time 共s兲
dE / dx ⫽ Electric field strength 共V/cm兲

The results of this applied voltage method agreed well with results of long-term chloride immersion
tests, such as the 90-day ponding test 共AASHTO T 259兲 关9兴. In a preliminary phase, it was concluded that
applying 80 V DC for eight hours was effective in differentiating between low and high chloride ions
penetrability concrete. High currents and high temperatures accompanied extremely permeable samples,
and vice versa for low permeability samples 关10兴. In 1981, a final laboratory version of the test was
developed to determine the chloride permeability class of concrete. In this test, a cylindrical plain concrete
specimen of approximately 100 mm in diameter and 50 mm in thickness is placed across 3 % sodium
chloride 共NaCl兲 and 0.3 N sodium hydroxide 共NaOH兲 solutions for 6 h, while applying 60 V DC. The
samples are preconditioned under vacuum for 3 h and submerged in water for 18 h for full saturation
before testing. The output of this apparatus is the total current integrated over time, or charges passed in
Coulombs 关10兴. The test method was issued under AASHTO T 277-83 including a table for determining
the chloride permeability class 共negligible, very low, low, moderate and high兲. This table was prepared
based on cores from slabs provided by the FHWA without considering the effect of cement type, aggregate
type, and density of concrete. Thus, it was recommended that users may construct their own RCPT
classification tables considering the effect of local materials tested by the RCPT 关9兴. In 1991, the RCPT
was issued under the ASTM C 1202 designation, without major deviations from the AASHTO version
关10兴.
The RCPT spread widely since 1986 accompanying the extensive use of silica fume, particularly in
bridge overlays 关10兴. Also, many concrete standard specifications 共e.g., CSA A23.1 issued by National
Standards of Canada兲 have used the RCPT passing charge levels as a durability indicator for concrete
quality. However, the RCPT has since been criticized for various shortcomings including its bias to
supplementary cementing materials such as silica fume and against CNI 关e.g., 关11兴兴. Several researchers
关e.g., 关12,13兴兴 questioned the reliability and suitability of this test method to assess chloride ions
penetrability into concrete. It was also suggested to adopt the rapid migration test 共RMT兲 developed by
Tang and Nilsson 关14兴 and issued under NT BUILD 492 共Nordic Standards兲 as an alternative.
Basically, the RMT is similar to the RCPT in the sense that chloride ions are forced to migrate into
concrete by applying a strong electrical field 关14兴. Under a specific direct voltage, a concrete disk of
100 mm in diameter and 50 mm in thickness is placed across 10 % sodium chloride and 0.3 N sodium
hydroxide solutions for certain time duration depending on the initial current. At the end of testing, the
specimen is axially split and sprayed with a silver nitrate solution to determine an average chloride
penetration depth from which a diffusion coefficient is calculated 关15兴. Contrary to RCPT passing charge
results, RMT chloride penetration measurements are insensitive to pore solution conductivity, and are less
vulnerable to testing variations 关12兴. It should be emphasized that both the RCPT and RMT are accelerated
tests based on ionic mobility under an induced direct voltage, but the measured parameters are different in
each test.
Using the ASTM C 1202 concept and procedure, this study aims at enhancing the reliability of rapidly
evaluating concrete resistance to chloride ions ingress by obtaining multiple measurements at the end of
testing. Julio-Betancourt and Hooton 关16兴 have recently proposed modifying the ASTM C 1202 to a
1-min conductivity test to eliminate heat effects on ions mobility and final Coulomb values. Nonetheless,
based on the conductivity criterion alone, both the conventional and the newly proposed procedure can
yield unreliable assessment for chloride ions penetrability due to the incorporation of pozzolanic materials
and/or CNI in concrete mixtures.
The current authors believe that there is a need to obtain an additional measurement 共chloride
penetration depth兲 from the RCPT, which better correlates to physical characteristics of the pore structure
and in turn chloride ions penetrability. Hence, the reliability of assessing the resistance of concrete to the
penetration of chloride ions, and in turn the risk of reinforcement corrosion, using this accelerated
procedure 共ASTM C 1202兲 can be improved by relying on multiple measurements that indicate both
chloride ions penetrability and concrete resistivity. These are chief factors in the service life prediction of
structures exposed to chlorides. The total service life is the sum of time-to-corrosion initiation of steel
4 JOURNAL OF ASTM INTERNATIONAL

TABLE 1—Chemical composition of OPC and silica fume.

Percent by
Mass OPC Silica Fume
SiO2 19.5 92.0
CaO 61.3 0.3
Al2O3 5.2 1.0
Fe2O3 2.4 1.0
MgO 2.5 0.6
K 2O 1.2 0.8
SO3 4.1 0.2
Na2O 0.1 0.3

rebars and corrosion propagation time until the occurrence of concrete cracking and spalling. Concrete
cover and chloride diffusivity govern the corrosion initiation stage, while concrete resistivity may
determine the corrosion propagation rate 关e.g., 关17兴兴.

Experimental Program

Materials
Seventeen concrete mixtures with water-to-cementitious materials ratios 共w/cm兲 of 0.30, 0.45, and 0.60
were prepared incorporating selected materials that have been known to cause misleading RCPT results.
The mixtures were divided into two groups: the first group incorporated CSA Type 10 共ASTM Type I兲
ordinary portland cement 共OPC兲 and silica fume with and without CNI. The second group involved only
OPC with and without CNI.
OPC with a specific gravity of 3.17 and a specific surface area 共Blaine Fineness兲 of 412 m2 / kg was
used, and densified silica fume with an average particle size of 0.1 ␮m, a specific surface area of
19530 m2 / kg, and a specific gravity of 2.12 was used. The total cementitious materials content was
400 kg/ m3. The oxide analysis of the cement and silica fume is presented in Table 1. The fine aggregate
was natural siliceous sand with a fineness modulus of 2.80, a saturated surface dry specific gravity of 2.65
and absorption of 1.5 %. Crushed stone with a maximum nominal size of 19 mm, a saturated surface dry
specific gravity of 2.68 and absorption of 0.8 % was also used. An ASTM C 494 Type F high-range water
reducing admixture was employed, and its dosage was adjusted to maintain a slump range of
160 to 190 mm. An inorganic corrosion inhibitor based on 30 % solution of calcium nitrite 共CNI兲 with a
density of 1.26 kg/ L was used at three dosages of 10, 20 and 30 L / m3 of concrete.

Procedures
Constituent materials were mixed in a mechanical mixer in accordance with the ASTM C 192 standard
procedure. Table 2 shows the proportions of the tested mixtures. All specimens were demolded at 24 h and
moist cured at 23° C and 95 % RH for 27 days. Concrete disks of 100 mm in diameter and 50 mm in
thickness were sliced from the top portion of 100 mm⫻ 200 mm cylinders for the RCPT 关18兴, simulating
concrete above the rebar level. Three replicates were prepared for each mixture. Samples were precondi-
tioned according to the ASTM C 1202 standard procedure. While the anolyte solution 共NaOH兲 concen-
tration was kept at 0.3 N as per the ASTM C 1202 guidelines, the concentration of the catholyte solution
共NaCl兲 was raised to 10 % complying with the RMT standard procedure 关19兴 to increase the likelihood of
a visible chloride front.
According to the ASTM C 1202, a 60 V DC is applied for 6 h or until high temperature of solutions
is observed. Comparatively, 6 h is the minimum test duration in the NT BUILD 492 Appendix 2 method,
but at a low voltage of 10 V DC. The NT BUILD 492 varies the test duration 共6 – 96 h兲 with the induced
voltage 共10– 60 V DC兲 based on the measured initial current 关19兴. In this study, it was decided that
applying 60 V DC for 6 h, according to the current ASTM C 1202 procedure, would limit the test
variables and keep the testing duration short. This is also a simpler procedure and is more common in
North American testing laboratories. At the end of testing, the specimens were axially split and sprayed
with 0.1 M silver nitrate solution forming a white precipitate of silver chloride after about 15 min 共colo-
BASSUONI ET AL. ON CHLORIDE INGRESS 5

TABLE 2—Proportions of concrete mixtures (per m3 of concrete).

Cement Silica Coarse


Content Fume Fine Aggregate Aggregate CNI 28-Day fc⬘
Mix ID 共kg兲 共kg兲 w/cm 共kg兲 共kg兲 共L兲 共MPa兲
OPC-1 400 ¯ 0.30 800 1150 ¯ 59.4
OPC-2 400 ¯ 0.45 740 1050 ¯ 44.2
OPC-3 400 ¯ 0.60 685 950 ¯ 26.7
OPC-4 400 ¯ 0.30 750 1150 20 62.3
OPC-5 400 ¯ 0.45 695 1050 20 41.3
OPC-5A 400 ¯ 0.45 720 1050 10 42.6
OPC-5B 400 ¯ 0.45 670 1050 30 45.0
OPC-6 400 ¯ 0.60 635 950 20 30.0
SF-1 368 32 0.30 790 1150 ¯ 60.8
SF-2 368 32 0.45 730 1050 ¯ 47.2
SF-2A 384 16 0.45 740 1050 ¯ 46.2
SF-3 368 32 0.60 675 950 ¯ 29.6
SF-4 368 32 0.30 740 1150 20 65.5
SF-5 368 32 0.45 680 1050 20 51.1
SF-5A 368 32 0.45 710 1050 10 45.6
SF-5B 368 32 0.45 650 1050 30 48.3
SF-6 368 32 0.60 620 950 20 33.0

rimetric method兲. The chloride penetration depth for each specimen was determined by averaging the
depth of the precipitate at a minimum of five different locations across the width of the sample.
For the mercury intrusion porosimetry 共MIP兲 testing, small cores were carefully drilled out, to mini-
mize coarse aggregates inclusion, from the top portion of 100 mm⫻ 200 mm cylinders in order to deter-
mine the pore size distribution and average porosity for the tested mixtures at 28 days. For each mixture,
at least five samples were tested to obtain statistically representative average results with a standard
deviation of less than ±15 % 关20兴. Preconditioning of samples was done by oven drying at 60± 5 ° C until
a constant mass to reduce the potential of drying shrinkage cracks associated with higher temperatures.
The apparatus had a pressure range from sub-ambient to 207 MPa. By assuming a cylindrical geometry of
pores, a mercury contact angle of 140° and surface tension of 0.484 N / m, the high pressure limit
共207 MPa兲 yields a pore radius of 3.6 nm, implying intrusion of all capillary pores since the smallest size
of capillary pores reported in literature is about 5 nm 关20兴. Most of the CSH gel pores can not be intruded
under such a pressure; however, such pores have little, if any, contribution to mass transport mechanisms
in the concrete matrix.

Results and Discussion

Rapid Chloride Penetrability Test (RCPT)

Passing Charges—The RCPT results at 28 days, shown in Fig. 1, were adjusted according to the
standard sample diameter 共95 mm兲 as per the ASTM C 1202 guidelines. The test continued for 6 h or until
solutions reached a high temperature of 80° C, which was defined as an upper limit to avoid damage of test
cell compartments. Figure 1 shows that the ultimate recorded passing charges ranged from 1088 to 8662
Coulombs, which correspond to low and high chloride penetrability, respectively, according to the
ASTM C 1202. However, the catholyte 共NaCl solution兲 used herein had 10 % concentration compared to
a standard concentration of 3 %, which makes the ASTM C 1202 penetrability classes inapplicable for the
categorization of the tested specimens. Figure 1 also shows extrapolated passing charges recorded at
30 min, which ranged from 964 to 7023 Coulombs. Linear extrapolation to 6 h was done by multiplying
the 30-min values by 12 to discount heat effects associated with some of the tested mixtures; in such cases,
extrapolated data may provide better estimation than the ultimate recorded charges 关11,21兴.
It has been shown that the Joule effect characterizes the RCPT so that a linear correlation between
passing charges and temperature rise exists, thereby influencing results significantly 关16兴. The relationship
between the recorded and extrapolated charges is illustrated in Fig. 2. With reference to the equity line,
when exceeding about 4000 Coulombs on the recorded charges axis, there is a noticeable disassociation
6 JOURNAL OF ASTM INTERNATIONAL

FIG. 1—RCPT recorded and extrapolated passing charges.

between the two datasets in most cases. For example, for mixture OPC-5 prepared with 100 % OPC, w/cm
of 0.45 and CNI 共20 L / m3兲, the ratio of extrapolated to recorded passing charges is 0.55. This implies that
the heat effect on the recorded passing charges of such a mixture was profound. Moreover, it seems that
the particular combination of cementitious material, w/cm, and CNI dosage influences the degree of

FIG. 2—Recorded vs. extrapolated passing charges.


BASSUONI ET AL. ON CHLORIDE INGRESS 7

TABLE 3—Resistivity, chloride penetrability and porosity of mixtures.

Test 30-min Penetration Migration Apparent


Duration Current Resistivity Depth Coefficient Porosity
Mix ID 共h兲 共mA兲 共⍀ . m兲 共mm兲 共⫻10−12 m2 / s兲 共%兲
OPC-1 6.0 135 69.78 8 7.24 9.6
OPC-2 6.0 191 49.45 24 23.80 17.8
OPC-3 5.5 306 30.83 45 52.23 20.0
OPC-4 6.0 121 78.17 5 4.42 8.6
OPC-5 6.0 267 35.28 45 47.87 19.9
OPC-5A 6.0 261 36.09 39 40.96 19.2
OPC-5B 4.0 344 27.42 35 54.33 19.5
OPC-6 3.7 395 23.88 42 73.56 21.7
SF-1 6.0 49 192.24 5 4.46 8.6
SF-2 6.0 101 93.73 14 13.52 13.5
SF-2A 6.0 151 62.38 21 20.70 16.0
SF-3 6.0 116 81.56 26 25.86 18.4
SF-4 6.0 68 138.53 5 4.31 8.5
SF-5 6.0 221 42.62 17 17.29 12.3
SF-5A 6.0 143 66.11 21 20.47 16.8
SF-5B 6.0 186 50.75 28 28.80 15.6
SF-6 6.0 211 44.64 34 35.15 19.9

discrepancy between extrapolated and recorded data 共Fig. 1兲. Although relying on extrapolated passing
charges may better indicate chloride penetrability in concrete mixtures experiencing high temperature rise
关11,21兴, results should be taken cautiously considering conductivity biases resulting from different material
types and proportioning.
Table 3 lists the electrical resistivity for all mixtures corresponding to electrical current measurements
at 30 min from the start of the RCPT. The values were calculated by:
VA
␳= 共2兲
IL

where
␳ ⫽ Resistivity 共⍀ · m兲
V ⫽ Applied voltage 共V兲
A ⫽ Cross-sectional area of the specimen 共m2兲
I ⫽ Current 共A兲
L ⫽ Thickness of the specimen 共m兲

Previous work 关16兴 recommended the calculation of 1-min resistivity to avoid temperature effects 共up to
40 ° C兲 that could happen even at 30 min from the start of the test. Nevertheless, no such temperature
rise was observed in the present study for all tested mixtures, and the maximum temperature gradient at
30 min did not exceed 6 ° C. Also, there was a marginal difference between the initial and 30-min
currents resulting in comparable resistivity values at 1 and 30 min. Thus, only the 30-min resistivity
was considered herein conforming to passing charge extrapolations.
As shown in Fig. 1, for mixtures prepared without CNI 共OPC-1, OPC-2, OPC-3, SF-1, SF-2, and
SF-3兲, extrapolated and recorded passing charges decreased with reduction in the w/cm, which is well
documented in the literature. From Table 3, the electrical resistivity for these mixtures increased as the
w/cm decreased, resulting in lower passing charges. Figure 1 also shows that passing charges 共recorded or
extrapolated兲 for silica fume mixtures without CNI 共SF-1, SF-2 and SF-3兲 had about 60 % lower values
than that of corresponding OPC mixtures without CNI 共OPC-1, OPC-2, OPC-3兲 at similar w/cm. This is
ascribed to the increased resistivity as implied by Table 3. Silica fume affects the ionic concentration of the
pore solution since it reduces the free alkalis content 关13兴. The influence of silica fume on conductivity can
be appreciated by comparing data for mixtures SF-2A and SF-2, where doubling the silica fume dosage
from 4 to 8 % led to 32 % reduction and 50 % increase in the extrapolated passing charges and resistivity
values, respectively. Moreover, by comparing results for mixtures SF-3 prepared with 8 % silica fume at
w/cm of 0.60 and those for mixture SF-2A prepared with 4 % silica at w/cm of 0.45, it is shown that
8 JOURNAL OF ASTM INTERNATIONAL

increasing the silica fume dosage reduced passing charges despite the increase in w/cm. Since this is
believed to be due to a reduction of pore solution conductance induced by silica fume, the RCPT passing
charge indication has been criticized for exhibiting significant preference to pozzolanic additives 关e.g.,
关22兴兴.
The impact of CNI addition on RCPT results is presented in Fig. 1 and Table 3. It was observed that
mixtures prepared with OPC at w/cm of 0.45 and 0.60 and incorporating CNI experienced high temperature
rise, which affected the recorded results. Thus, extrapolated results are considered preferable for the CNI
mixtures since at this early stage of the test 共30 min兲, temperature effects were minimal. Mixture OPC-4
prepared with w/cm of 0.30 and 20 L / m3 CNI had slightly lower extrapolated passing charges and higher
30-min resistivity than that of the corresponding mixture OPC-1 without CNI. Conversely, at higher w/cm
共0.45 and 0.60兲 all OPC mixtures prepared with CNI had higher extrapolated charges and lower resistivity
than that of the corresponding OPC mixtures without CNI, irrespective of the CNI dosage employed.
Increasing the CNI dosage from 10 to 20 L / m3 did not significantly affect extrapolated passing charges
and resistivity for the OPC mixtures, but a 30 L / m3 dosage contributed to a noticeable decrease in
resistivity and in turn an increase in extrapolated passing charges.
Similar to the OPC mixtures with CNI, the silica fume mixtures with CNI had higher extrapolated
charges and lower resistivity than that of the corresponding silica fume mixtures without CNI, irrespective
of the CNI dosage used. However, unlike for OPC mixtures with CNI, the variation of the CNI dosage for
silica fume mixtures had a pessimum level, whereby a combination of 8 % silica fume, w/cm of 0.45 and
20 L / m3 共SF-5兲 showed higher extrapolated passing charges and lower resistivity than that of the other
silica fume mixtures, irrespective of the w/cm or the CNI dosage. CNI adversely affects the electrical
conductance of concrete 关23兴, but when added with silica fume, it seems that there is a pessimum
CNI-silica fume combination depending on the w/cm. The distortion of the RCPT results in the presence
of CNI is already recognized in the ASTM procedure 关18兴, which implies the unsuitability of the passing
charges indication for evaluating chloride ions ingress in concrete mixtures incorporating CNI.
Concrete resistivity calculated from the RCPT can be used to estimate the length of corrosion
propagation stage until spalling of concrete. Generally, a linear correlation between the inverse of concrete
resistivity and the steel corrosion rate has been established 关e.g., 关17,24兴兴. For instance, when concrete
resistivity exceeds 70 ⍀ · m, the corrosion rate of steel rebars is expressed by:

CR = 1000/␳concrete 共3兲

where
CR ⫽ Corrosion rate 共␮m Fe/yr兲
␳concrete ⫽ Concrete resistivity 共⍀ · m兲

Then, the duration of the corrosion propagation stage can be calculated by dividing the amount of
corrosion that causes cracking by the corrosion rate 关17兴. In addition to resistivity, by considering oxygen
availability and the concentration of hydroxide ions 共OH−兲, a more accurate estimation of the corrosion
rate can be obtained.
Although taking measurements from the RCPT at short time intervals 共e.g., 1-min 关16兴 or 30-min兲 can
eliminate the temperature effect on the RCPT passing charge and resistivity results, such a procedure may
not correctly assess the chloride ions penetrability in concrete. Concrete with low resistivity 共i.e., high
passing charges兲 does not necessarily have a high chloride ions diffusivity as will be shown below. Thus,
it is inappropriate to use passing charges alone to evaluate chloride ions penetrability in concretes
incorporating materials affecting the ionic concentration of the pore solution. There is need to take an
additional physical measurement that is less dependent on the electrical conductance of concrete.

Penetration Depth of Chlorides—At the end of the RCPT, the average penetration depth of chloride
ions was determined, from which chloride diffusion coefficients 共migration coefficients兲 due to non-
steady-state migration were calculated 共Table 3兲. The average penetration depth ranged from
4.95 to 42.23 mm, corresponding to migration coefficients 共D兲 ranging from 4.31 to 73.56⫻ 10−12 m2 / s,
which were calculated by 关19兴:
BASSUONI ET AL. ON CHLORIDE INGRESS 9

FIG. 3—Whitish chloride front migrating into concrete prepared with: (a) 8 %, and (b) 4 % silica fume at
w/cm of 0.45 without CNI.

D=
0.0239共273 + T兲L
共V − 2兲t

xd − 0.0238 冑 共273 + T兲Lxd
V−2
冊 共4兲

where
D ⫽ Non-steady-state migration coefficient 共⫻10−12 m2 / s兲
V ⫽ Applied voltage 共V兲
T ⫽ Average value of initial and final temperatures in the anolyte solution 共°C兲
L ⫽ Thickness of the specimen 共mm兲
xd ⫽ Average value of penetration depth 共mm兲
t ⫽ Time 共h兲

From Table 3, the effect of the w/cm on D followed the common trend observed in the literature
without irregularities. For all OPC and silica fume mixtures prepared with and without CNI, D decreased
with a reduction of the w/cm as expected. Mixture SF-2A prepared with 4 % silica at w/cm of 0.45
had lower D than that of mixture SF-3 prepared with 8 % silica fume at w/cm of 0.60. This supports the
argument that higher passing charges 共lower resistivity兲 do not necessarily mean higher chloride
diffusivity. Also, unlike the RCPT results, the effect of increasing the silica fume dosage on decreasing
concrete conductance did not affect chloride diffusivity, since the w/cm was the governing factor.
The addition of silica fume had a pronounced effect on reducing D 共average reduction of 43 %兲 likely
due to the pozzolanic and micro-filler effects that produce denser microstructure, lower porosity, and
refined ITZ 关3兴. Figure 3 illustrates the chloride front for mixtures SF-2A and SF-2 tested under similar
conditions. At a w/cm of 0.45, the increase of the silica fume dosage from 4 to 8 % reduced the average
chloride penetration depth from 20.88 to 14.20 mm 共32 % reduction兲. It is worth noting that by visually
inspecting chloride fronts on split samples, a reasonable qualitative assessment of chloride penetrability in
concrete can be obtained without biases induced by pore solution conductivity.
The addition of CNI to OPC mixtures did not only increase RCPT passing charge results, but also
increased D by a factor of up to 2.28 compared to that of the corresponding OPC mixtures without CNI
共except mixture OPC-4兲. A similar trend for the influence of CNI on increasing chloride penetration was
reported in the literature 关25–27兴. At a w/cm of 0.45, increasing the CNI dosage from 10 to 30 L / m3 for
the OPC mixtures led to a moderate linear increase in D. Similarly, CNI addition in silica fume mixtures
led to a higher D by a factor of up to 2.17 compared to that of the corresponding silica fume mixtures
without CNI 共except for mixture SF-4兲.
At a w/cm of 0.45, increasing the CNI dosage from 10 to 20 L / m3 for the silica fume mixtures led to
about 15 % reduction in D. Li et al. 关27兴 reported a similar decrease in chloride diffusivity for cement
pastes prepared with fly ash and silica fume when CNI dosage increased from 15 to 20 L / m3. Perhaps this
is due to the interaction between CNI and pozzolanic materials. Generally, the increase in chloride
diffusivity due CNI addition may be attributed to the set accelerating effect induced by CNI on cement
hydration, which is more pronounced at high CNI dosage 共30 L / m3兲, resulting in a relatively less tortuous
pore microstructure, but not necessarily higher apparent porosity as will be shown later by the MIP results.
Unlike the case of the RCPT passing charges and resistivity results, there was no pessimum CNI-silica
fume combination for D, which indicates that such a pessimum behavior is an artifact induced by pore
solution conductivity. In agreement with previous research 关12兴, this implies that D is marginally affected
10 JOURNAL OF ASTM INTERNATIONAL

FIG. 4—Pore size distribution of tested mixtures at 28 days: (a) OPC mixtures without CNI, (b) OPC
mixtures with CNI, (c) Silica fume mixtures without CNI, (d) Silica fume mixtures with CNI.

by conductive ions in the pore solution.


Unlike the RCPT passing charges, measured penetration depths can give better representation of
chloride ions penetrability in concrete through physical evaluation. Some investigations attempted to
correlate chloride ions diffusivity in concrete to electrical conductivity 关e.g., 关28兴兴. Diffusion of chloride
ions into concrete is dependent on its pore structure 关28兴, while conductivity of concrete is mainly governed
by ionic concentration in the pore solution. Thereby, some investigators argue that it is inappropriate to use
conductivity as an indicator for chloride ions penetrability in concrete made with different constituents and
proportions 关13兴. More research in this area is still needed for reaching definitive conclusions.

Mercury Intrusion Porosimetry (MIP)


Apparent porosity for all tested mixtures was determined after 28 days using MIP, and the results are listed
in Table 3. Depending on the mixture composition, the apparent porosity ranged from 8.5 to 21.7 %.
Figures 4共a兲–4共d兲 show the measured pore size distribution for all tested mixtures. MIP has been criticized
for its inability to intrude CSH gel and closed pores, as well as measuring pore entrances rather than actual
pore sizes, which is termed the ink-bottle effect 关29兴. Hence, MIP rather evaluates the apparent porosity,
which is a useful index correlated to the pore structure, mechanical strength and overall quality of con-
crete.
From Table 3 and Figs. 4共a兲–4共d兲, and in agreement with chloride ions penetration depth measure-
ments, the reduction of the w/cm led to a decrease in cumulative mercury intrusion, and in turn a reduced
apparent porosity for the OPC and silica fume mixtures. In most cases, the pore size distribution of the
silica fume mixtures was shifted to smaller pore sizes below 0.1 ␮m compared to that of the corresponding
BASSUONI ET AL. ON CHLORIDE INGRESS 11

FIG. 5—Relationship between porosity and chloride migration coefficient.

OPC mixtures, demonstrating the significant effect of silica fume on pore size refinement. All silica fume
mixtures had lower cumulative intrusion/apparent porosity than that of the corresponding OPC mixtures.
By increasing the dosage of silica fume from 4 to 8 %, the apparent porosity was decreased by 16 %,
complying with measurements of chloride ions penetration depth.
Unlike for the RCPT passing charge measurements, it seems that CNI addition affected the apparent
porosity results marginally for both the OPC and silica fume mixtures. Complying with trends for chloride
ions migration results, at low w/cm 共0.30兲, the CNI mixtures had slightly lower porosity than that of the
corresponding mixtures without CNI. However, at higher w/cm ratios 共0.45 and 0.60兲, all of the CNI
mixtures had relatively higher porosities than that of the corresponding mixtures without CNI, except for
mixture SF-5, which exhibited a pessimum combination for passing charges as explained earlier. The
effect of varying the CNI dosage on the apparent porosity for the OPC mixtures at w/cm of 0.45 was
insignificant since mixtures OPC-5, OPC-5A and OPC-5B had comparable porosities and identical pore
size distributions 共Fig. 4共b兲兲. This implies that the moderate linear increase observed in the migration
coefficients, associated with increasing CNI dosage, may be due to increased capillary pore continuity
occurred at similar total porosities of OPC mixtures. At w/cm of 0.45, the effect of interaction between
CNI dosage and silica fume on total porosity was similar to migration coefficient results.
The discussion above indicates that a correlation exists between chloride ions migration coefficients
and porosity of tested mixtures. Figure 5 illustrates this relationship between D and the apparent porosity
for all tested mixtures, which is expressed by:

D = 0.0152P2.662 共5兲

where
D ⫽ Non-steady-state migration coefficient 共⫻10−12 m2 / s兲
P ⫽ Porosity 共%兲

The correlation coefficient 共␳兲 and the coefficient of determination 共R2兲 between D and porosity data
are 0.89 and 0.94, respectively. The standard error of estimate for D is 6.3⫻ 10−12 m2 / s, which is
approximately 22.5 % of the mean value of migration coefficients 共28⫻ 10−12 m2 / s兲. This reflects a
significant association between the two datasets. The above relationship holds as a qualitative indication
that chloride migration coefficients obtained from the RCPT are sensitive to pore structure characteristics.
On the other hand, comparing the RCPT passing charge results to that of migration coefficient showed that
D is insensitive to electrolysis biases induced by CNI and/or silica fume.
12 JOURNAL OF ASTM INTERNATIONAL

Conclusions and Recommendations


Previous research suggested considering RCPT measurements after short time periods from the start of the
test 共1-min resistivity and 30-min passing charges extrapolation兲 to minimize the temperature effect on
passing charge results. Nevertheless, it was shown herein that such extrapolated passing charges should be
taken cautiously considering conductivity biases resulting from combining various proportions of different
materials. Hence, it is inappropriate to use passing charges alone to evaluate chloride ions penetrability in
concretes incorporating materials affecting the ionic concentration of the pore solution.
The study showed that multiple measurements can be successfully obtained from the RCPT: passing
charges and chloride migration coefficients. While passing charges are mainly influenced by pore solution
conductivity, migration coefficients are sensitive to the physical characteristics of the pore structure.
Hence, by using the RCPT-ASTM C 1202 procedure followed by the colorimetric method, both resistivity
and chloride penetrability can be evaluated. These are chief factors in the process of corrosion of embed-
ded steel in reinforced concrete. Also, when materials that affect pore solution chemistry 共e.g., silica fume
and CNI兲 are incorporated, the migration coefficient can eliminate conductivity biases and show a reason-
ably reliable assessment for chloride penetrability that is correlated to the pore structure of concrete. This
improves the overall reliability of rapidly evaluating chloride ions penetrability in concrete using the
RCPT method, if the colorimetric method is added to the current procedure. Yet, further research is
recommended to investigate the effect of reducing the catholyte concentration from the 10 % used in the
present study to its standard value 共3 %兲 on migration coefficients. Also, excessive temperature rise that
can occur with some concretes need to be controlled since high temperature is known to influence the
microstructure of concrete and increase chloride diffusivity.

References

关1兴 Scannell, W., Sohanghpurwala, A., and Islam, M., “Assessment of Physical Condition of Concrete
Bridge Components,” FHWA-SA-97-002, HTA-10/1-96 (3M)P, U.S. Department of Transportation,
Federal Highway Administration, 1996, pp. 2–6.
关2兴 Bentur, A., Diamond, S., and Berke, N., Steel Corrosion in Concrete, E & FN SPON, London, 1998.
关3兴 Gjørv, O. E., “Effect of Condensed Silica Fume on Steel Corrosion in Concrete,” ACI Mater. J., Vol.
92, No. 6, 1995, pp. 591–598.
关4兴 Whiting, D. and Dziedzic, W., “Resistance to Chloride Infiltration of Superplasticised Concrete as
Compared with Currently Used Concrete Overlay Systems,” FHWA/OH-89/009, U.S. Department of
Transportation, Federal Highway Administration, 1989, 86 pp.
关5兴 Nmai, C. K., Farrington, S. A., and Bobrowski, G. S., “Organic-Based Corrosion-Inhibiting Admix-
ture for Reinforced Concrete,” Concr. Int., Vol. 14, No. 4, 1992, pp. 45–51.
关6兴 Berke, N. and Rosenberg, A., “Technical Review of Calcium Nitrite Corrosion Inhibitor in Con-
crete,” Transp. Res. Rec., No. 1211, 1989, pp. 18–27.
关7兴 Broomfield, J., Corrosion of Steel in Concrete, E & FN SPON, London, 1997.
关8兴 Whiting, D., “Rapid Measurement of the Chloride Permeability of Concrete,” Public Roads, Vol. 45,
No. 3, 1981, pp. 101–112.
关9兴 Whiting, D. and Kuhlman, L., “Curing and Chloride Permeability,” Concr. Int., Vol. 9, No. 4, 1987,
pp. 18–21.
关10兴 Whiting, D. and Mitchell, T., “History of Rapid Chloride Permeability Test,” Transp. Res. Rec., No.
1335, 1992, pp. 55–61.
关11兴 Hooton, R. D., Pun, P., Kojuncdic, T., and Fidjestol, P., “Influence of Silica Fume on Chloride
Resistance of Concrete,” Proceedings of PCI/FHWA International Symposium on High Performance
Concrete, New Orleans, Louisiana, 1997, pp. 245–256.
关12兴 Stanish, K., Hooton, R. D., and Thomas, M. D. A., “A Novel Method for Describing Chloride Ion
Transport due to an Electrical Gradient in Concrete: Part 2. Experimental Study,” Cem. Concr. Res.,
Vol. 34, No. 1, 2004, pp. 51–57.
关13兴 Shi, C., “Effect of Mixing Proportions of Concrete on its Electrical Conductivity and the Rapid
Chloride Permeability Test 共ASTM C 1202 or AASHTO T 277兲 Results,” Cem. Concr. Res., Vol. 34,
No. 3, 2004, pp. 537–545.
BASSUONI ET AL. ON CHLORIDE INGRESS 13

关14兴 Tang, L. and Nilsson, L., “Rapid Determination of the Chloride Diffusivity in Concrete by Applying
an Electric Field,” ACI Mater. J., Vol. 89, No. 1, 1992, pp. 49–53.
关15兴 Tang, L., “Electrically Accelerated Methods for Determining Chloride Diffusivity in Concrete-
Current Development,” Mag. Concr. Res., Vol. 48, No. 176, 1996, pp. 173–179.
关16兴 Julio-Betancourt, G. A. and Hooton, R. D., “Study of the Joule Effect on Rapid Chloride Permeabil-
ity Values and Evaluation of Related Electrical Properties of Concrete,” Cem. Concr. Res., Vol. 34,
No. 6, 2004, pp. 1007–1015.
关17兴 Polder, R. B., “The Influence of Blast Furnace Slag, Fly Ash, and Silica Fume on Corrosion of
Reinforced Concrete in Marine Environment,” Heron, Vol. 41, No. 4, 1996, pp. 287–300.
关18兴 ASTM C 1202, “Electrical Indication of Concrete’s Ability to Resist Chloride Ion Penetration,” Vol.
4.02, Annual Book of American Society for Testing Materials, Philadelphia, 2000.
关19兴 NT BUILD 492, “Chloride Migration Coefficient from Nonsteady-State Migration Experiments,”
NORDTEST, Finland, 1999.
关20兴 Kumar, R. and Bhattacharjee, B., “Porosity, Pore Size Distribution and In Situ Strength of Concrete,”
Cem. Concr. Res., Vol. 33, No. 1, 2003, pp. 155–164.
关21兴 Stanish, K., Hooton, R. D., and Thomas, M. D. A., “Testing the Chloride Penetration Resistance of
Concrete: A Literature Review,” FHWA/DTFH61-97-R-00022, U.S. Department of Transportation,
Federal Highway Administration, 1997, 30 pp.
关22兴 Streicher, P. E. and Alexander, M., “A Chloride Conduction Test for Concrete,” Cem. Concr. Res.,
Vol. 25, No. 6, 1995, pp. 1284–1294.
关23兴 Ozyildirim, C., “Rapid Chloride Permeability Testing of Silica-Fume Concrete,” Cem. Concr. Aggr.,
Vol. 16, No. 1, 1994, pp. 53–56.
关24兴 Lopez, W. and Gonzalez, J. A., “Influence of the Degree of Pore Saturation on the Resistivity of
Concrete and the Corrosion Rate of Steel Reinforcement,” Cem. Concr. Res., Vol. 23, No. 2, 1993,
pp. 368–376.
关25兴 Pfeifer, D., Landgren, R., and Zoob, A., “Protective Systems for New Prestressed and Substructure
Concrete,” FHWA/RD-86/193, U.S. Department of Transportation, Federal Highway Administration,
1987, 126 pp.
关26兴 Nmai, C. and McDonald, D., “Long-Term Effectiveness of Corrosion-Inhibiting Admixtures and
Implications on the Design of Durable Reinforced Concrete,” The Role of Admixtures in High-
Performance Concrete, L. S. Johal, Ed., RILEM, Monterrey, Mexico, 1999, pp. 1–17.
关27兴 Li, Z., Peng, J., Ma, B., “Investigation of Chloride Diffusion for High-Performance Concrete Con-
taining Fly Ash, Microsilica and Chemical Admixtures,” ACI Mater. J., Vol. 96, No. 3, 1999, pp.
391–396.
关28兴 Garboczi, E. G., “Permeability, Diffusivity and Microstructural Parameters,” Cem. Concr. Res., Vol.
20, No. 5, 1990, pp. 591–601.
关29兴 Diamond, S., “Mercury Porosimetry; an Inappropriate Method for the Measurement of Pore Size
Distribution in Cement Based Materials,” Cem. Concr. Res., Vol. 30, No. 10, 2000, pp. 1517–1525.

View publication stats