1. 1 INTRODUCTION
Reinforcement corrosion caused by chloride
attack is the most common cause for
deterioration of marine structures. Predicting the
time of corrosion initiation (T.I) and propagation
(T.P) is useful to tackle the problem. It is
beneficial for developing effective inspection
strategies, to propose accurate planning for
maintenance, strengthening or simply to
successfully conduct the replacement of
damaged elements. This allows prolonging the
service life and reducing the maintenance costs.
Information from chloride ingress
observations can be defined in mathematical
models. The mechanisms of chloride transport
simplified as diffusion problem through chloride
diffusion model. The diffusion as main way of
chloride ion transport requires concrete pores to
be saturated with liquid. Oxygen ingress, that is
present in corrosion process, is possible only in
low relative humidity, below 70%. However,
higher diffusion rate occur in RC structures
subjected to pore water saturation.
In the past 3 decades, there has been an
increase in the amount of research in predicting
the service life of RC structures both
experimental and mathematical models.
Unfortunately, the high number of parameters
involved in the corrosion process make difficult
to come up with accurate predictions for T.I and
T.P. Reinforcement corrosion must be treated
into specific study conditions.
Chloride Diffusion Model reported by
Vedalakshmi, (2012) and potential corrosion
testing such as: half-cell potential (HCP) method
or electrochemical impedance technique (EIS)
have been generally accepted to predict T.I.
El Maaddawy and Soudki, (2007) developed a
mathematical model to determine T.P based on
Chloride corrosion effect on different diameter rebar in marine
structures
A.Manzano-Cabrera
School of Engineering and the Built Environment, Edinburgh Napier University, Edinburgh, UK
ABSTRACT: High chloride content in marine environment is the major cause of reinforcement
corrosion in coastal structures. The exposure of structural elements to seawater level changes triggers
the chloride-induced corrosion. This research aimed to study the mass-loss of different diameter rebar
in RC specimens subjected to immersion and drying cycles in simulated marine environment. An
experimental testing program was conducted to assess 12 and 8 mm diameter rebar in RC samples
throughout corrosion initiation and propagation. Half-Cell Potential testing and Chloride Diffusion
Model were applied to predict the “Time to corrosion initiation” and Maaddawy model, for “Time to
propagation”. This study showed earlier time to initiation compared with researches in spraying
exposure. 12mm rebar experienced greater mass-loss, corrosion rate 61% higher than 8mm rebar
within 182 days of exposure and 22 days faster time of propagation. However, Maaddawy model
does not account the high diffusion in immersion cycles and leaves uncertainties.

2. time from T.I to cracking of the concrete cover.
Mechanical characteristics of concrete and
diameter rebar were taken in account. However,
“the Maaddawy model” referred to T.I from
experimental works based on exposure to
spraying cycles rather than immersion.
Destructive testing such as: mass-loss and
cross section reduction measurements are
general applied to assess the structural capacity
reduction of rebar in corrosion propagation.
This study has proposed the aforementioned
methods, focusing on samples with different
diameter rebar in exposed to 231 days of cyclic
immersion and drying. The results were
compared with works at spraying conditions
reported by Aguiar, (2008) and Vedalakshmi,
(2012). Finally, propose a preventative action
for structural elements subjected to changes in
seawater level in early propagation phase that
may prolong their service life.
1.1 Aim & Objectives
The aim of this research is to study mass-loss
and cross-section reduction in different diameter
rebar in chloride contaminated RC specimens,
subjected to the same cycle of immersion and
drying in a simulated marine environment.
The specific objectives of the study are to:
• Predict T.I of RC specimens through
the use of HCP and chloride diffusion model.
• Predict T.P for 8 and 12mm diameter
rebar in RC samples by using Maaddawy
mathematical model.
• Assess percentage of mass-loss and
corrosion rate of different diameter rebar at 8, 26
and 33 weeks of exposure.
• Analyze apparent diffusion coefficient
(D) from comparison with T.I results obtained in
Aguiar, (2008) and Vedalakshmi, (2012) by
introducing values of governing parameters
related to their work conditions.
2 EXPERIMENTAL PROGRAMME
2.1 Specimen preparation
For initial corrosion testing, it was used 3 OPC
RC cubes (100x100x100) mm with 8 and 12 mm
of steel rebar (B-500 S (B)) 110mm long
embedded at 20mm of concrete cover from 4
faces. 9 OPC RC beams (400x100x100) mm
were used along propagation phase. 8 and 12
mm rebar steel (B-500 S (B)) 500mm long
embedded at 20mm of cover. After 28 days of
curing, the average compressive strength for all
the samples was found to be 26.1 MPa according
to EN 12390-4. All rebar were weighed before
being embedded to determine the mass-loss and
cross-section reduction after exposure. 3 OPC
concrete beams (400x100x100) mm were used
throughout 231 days of programme to approach
the required moisture content.
2.2 Exposure conditions
The specimens were subjected to 19 hours of
immersion in water tank at indoor with 3.5% of
salinity (35grams of salt by liter of water
according with a specific gravity of 1025) and 5
hours of drying in laboratory, which constituted
1 exposure cycle. This procedure was repeated
daily to simulate the wetting and drying cycles
of structural elements subjected to changes in the
water level in marine environment. A total of
231 cycles were conducted with an average
temperature between 22.5C to 27C and RH
ranged between 40 to 45%.
2.3 Testing procedure
The potential corrosion of rebar in 3 RC cubes
was measured by HCP testing, which is done
using voltmeter and a silver/silver reference
electrode connected to the exposed rebar ends in
each specimen, providing corrosion value in –
mV. According to the American Standard Test
Method (ASTM), T.I is determined at values
more negative than -350mV. 9 RC beams were

3. assessed weekly to draw the corrosion trends in
propagation phase. The destructive test of mass-
loss was conducted at the end of 2, 56, 182 and
231 days and was done by calculating weight
differences from the initial state.
3 MATHEMATICAL MODELS
The T.I (days) was mathematically predicted by
using the chloride diffusion model reported by
Vedalakshmi, in Equation 1. The apparent
diffusion coefficient, D (m2
/s) chloride content,
Cx (%) at depth of concrete cover, x (mm) and
chloride surface, Cs (%), were introduced from
assumptions in literature and reported research’s
works. The T.I value reported by Aguiar and
Vedalakshmi were simulated using this model to
compare and analyze the governing parameters
in immersion and spraying conditions.
T. I =
X2
4D�erf−�1−�
Cx
Cs
���
2 (1)
The corrosion rate (µm/year) of 8 and 12mm
diameter rebar were calculated at the end of 56,
182 and 231 days, by using an empirical
Equation 2 reported by Vedalakshmi from their
weight loss, w (mg) related to each period, T (s),
density of iron, D (gr/cm3
) and rebar cross-
section, A (cm2
).
Cr(µm/yr) =
87600 x w
DAT
(2)
The T.P (days) value was calculated through
Equation 3 - Maaddawy model which takes into
account the mechanical characteristics of
concrete cover, C (mm) and diameter of rebar, D
(mm). T.P for the 12 and 8mm diameter rebar
was predicted by introducing 3 different
corrosion current densities, icorr, (µA/cm2
)
which were referred to corrosion conditions:
very high (111 µA/cm2
), high (10 µA/cm2
) and
low/moderate (1 µA/cm2
) corrosion.
T. P = �
7117.5(D+2δ0)(1+γ+ψ)
icorrEcf
� … (3)
. . . �
2Cfct
D
+
2δ0Ecf
(1+γ+ψ)(D+2δ0)
�
Where, effective modulus of elasticity of
concrete, Ecf, (N/mm2
), Poisson’s ratio of
concrete, γ is approved as 0.18 for CSA standard,
porous layer, δ0, assumed 10µm thick, tensile
strength of concrete, ƒct=0.7√ƒck given by BIS
456 (N/mm2
) and the compressive strength of
concrete at 365 days of age, ƒck, (N/mm2
).
4 RESULTS & ANALYSIS
Time to corrosion initiation (T.I) achieved at the
second cycle of exposure (2days). HCP showed
values more negative than -350mV which
implies 90% probability of corrosion according
to ASTM. This T.I agreed with the experimental
work conducted by Aguiar. The percentage of
mass-loss in all rebar at T.I was lower than
0.04% and therefore mass-loss and cross-
sectional area reduction were neglected.
The T.I at 38 days obtained through chloride
diffusion model by assuming 3.5% of Cs, 0.39%
of Cx and D equal to 30x10-12
m2
/s reported by
Bertolini, L. (2013) as value found in OPC
structures with a w/c 0.86 at 28days of age.
However, Aguiar reported 50 days by using a
transformation factor developed by Li (2000).
Applying the mathematical model at conditions
used by Aguiar with a w/c ratio equal to 0.6, the
D was 23.3 x10-12
m2
/s and the T.I was 50 days
as the T.I reported. The D obtained by
Vedalakshmi, however, was 5.64 x10-12
m2
/s;
this is due to longer drying cycle stage (2 days)
and lower Cs (3%), resulting in considerably
longer T.I (747 days).
Two conclusions were drawn, T.I through
chloride diffusion model in both cases,
overestimated the T.I from HCP testing, T.I was
significantly affected by the D which is related
to the w/c ratio of the structure and drying time.

4. At the end of 56 and 182 days, from 6 specimens
assessed, the percentage of mass-loss of 12mm
rebar showed higher value, an average of: 0.14
to 0.20% against 0.09 to 0.11% registered in the
8mm rebar, represented in Figure 1 below. At
231 days, smaller diameter showed a slight
increase reaching 0.15%. However, 12mm rebar
did not show an increase at 231 days but it still
had the highest value of mass-loss.
Figure 1. Average percentage of mass-loss in 182 days
After 56 days, 12mm rebar reached 29%
higher corrosion rate (Cr) than 8mm rebar. At
the end of 182 days, 12mm rebar achieved the
highest Cr differences between each rebar, a
61% higher Cr than 8mm. The difference
reduced slightly at 231 days with a Cr value for
12mm rebar 35% higher than 8mm. The Cr
agreed well with the percentages of mass-loss.
They represented a high grade of corrosion
within 182 days according with studies made by
Andrade and Alonso, (1996).
Time to propagation (T.P) through Maaddawy
model in very high corrosion scenario, was
1289 days, just 2 days faster for 12mm rebar and
therefore interpreted as the same time. Under
moderate corrosion condition, T.P differences
between each diameter rebar was 217days,
referred to a total T.P of 143,261 and 143,044
days. The T.P was denied because it did not have
a significant impact. However, the T.P in a high
corrosion condition showed that bigger diameter
rebar achieved 22 days earlier the T.P than 8mm
rebar within a range of 39 years. The T.P for both
diameter in that case, agreed with the Cr results
and mass-loss within early propagation phase of
exposure (182 days). In addition to this, T.P in
this case represents a value that differs
considerably with the design working life for
civil engineering structures (100 years)
according with BS EN 1990:2002.
5 CONCLUSIONS
According with the difference of Cr between the
12 and 8mm diameter rebar and the percent of
mass-loss in early propagation phase. This study
proposes to use smaller diameter rebar in
secondary reinforcements (stirrups) in areas of
structural elements affected by changes of
seawater level and long period of immersion. In
addition to this, it is suggested reducing the
space between stirrups to not compromise the
structural capacity. The service life of the
structure would increase when this measure is
undertaken. Applying the Maaddawy model, the
T.P of structural elements with that measure
could prolong at least 22 days. Although, the
impact of that delay within 39 years may be
insignificant, this study challenges the
applicability of Maaddawy model for RC
structures subjected to immersion cycles. The
model does not account for the high diffusion
rate occurrence in immersion compared with
spraying cycles and its consequent effect on the
Cr of reinforcements.
6 REFERENCES
- Aguiar, S. (2008). An investigation of corrosion-
induced structural deterioration in reinforced concrete
coastal elements. Thesis PHD Civil Engineering,
University of Dundee.
- El Maaddawy, T. and Soudki, K. (2007). A model
for prediction of time from corrosion initiation to
corrosion cracking. Cement and Concrete
Composites, 29(3), pp.168-175.
- Vedalakshmi, R. (2012). Prediction of service life of
concrete structures using corrosion rate model.
Proceedings of the ICE - Structures and Buildings,
165(2), pp.95-108.