Engineering Analysis with Boundary Elements 77 (2017) 36–48

Contents lists available at ScienceDirect

Engineering Analysis with Boundary Elements

journal homepage: www.elsevier.com/locate/enganabound

Optimizing the sacrificial anode cathodic protection of the rail canal MARK
structure in seawater using the boundary element method
⁎
Yong-Sang Kima, Jeongguk Kimb, Dooho Choic, Jae-Yong Limd, Jung-Gu Kima,
a
School of Advanced Materials Science and Engineering, Sungkyunkwan University, 300 Chunchun-Dong, Jangan-Gu, Suwon 16419, Republic of Korea
b
New Transportation Systems Research Center, Korea Railroad Research Institute, 176 Chelodo Bangmulgwan-ro, Uiwang, Gyeonggi 16105, Republic of
Korea
c
Department of Advanced Materials Engineering, Dong Eui University, 176 Eomgwangno, Busan 47340, Republic of Korea
d
School of Mechanical Engineering, Daegu University, 201 Daegudaero, Gyeongsan, Gyeongbuk 38453, Republic of Korea

A R T I C L E I N F O A BS T RAC T

Keywords: This paper deals with the cathodic protection design for axle/wheel and rail of ‘rail-canal system’ in the ocean.
Cathodic protection The cathodic protection design was carried out using the boundary element method and was verified by the
Corrosion physical miniature tests. The optimum cathodic protection designs were determined based on the cathodic
Rail canal system potential distribution and anode lifetime provided by the simulation. The unprotected physical miniature
Boundary element method
experienced widespread corrosion, whereas the protected miniature was covered with calcareous deposit,
Physical miniature test
indicating that the surface was fully protected from corrosion. This study demonstrated that the boundary
element can be applicable to the cathodic protection design of rail-canal structure.

1. Introduction
A ship canal is that provide a shortcut for transport ships to avoid
lengthy detours. With the recent increase in demand for ship canals
desire to reduce the cost of cargo transport, there are plans to extend
both the existing ship canal in Panama, and build additional ship
canals in Nicaragua, Colombia [1,2]. However, conventional canal
waterways are enormously expensive to construct and have long lock
gate operation times. Also, overland routes are interrupted by canal
construction, resulting in environmental problems and restrictions in
national land use. To address these drawbacks in waterway canals, the
Korea Railroad Research Institute (KRRI) has proposed a ‘rail-canal
system’, in which a large vessel is transported not along waterways but
on land, so as to replace conventional waterway canals [3]. The rail-
canal system process is illustrated in Fig. 1. However, the corrosion
problems caused by a marine environment should be addressed to
provide safe and long-term operations for rail-canal systems.
Sacrificial anode cathodic protection (SACP) is a widely used anti-
corrosion method for offshore structures. The mechanism is that
galvanic current flows from a sacrificial anode such as magnesium
(Mg), zinc (Zn) and aluminum (Al) to the protected structure when the
anode and protected structure are electrically connected in a conduct-
ing environment [4]. In general, the design of SACP follows the
international cathodic protection criteria, including NACE RP0176
and DNV RPB041. However, important SACP design factors such as
the selection, location and number of anodes are determined by the
corrosion engineer due to complex conditions and uncertainties [5].
Thus, studies on SACP design for various offshore structures are
required to ensure the adequate design of SACP.
The boundary element method (BEM) is an element method that
only requires meshing of the boundary on the modeling structure, so it
can more effectively mesh and calculate the boundary elements
compared to other element methods [5]. Due to this advantage, BEM
has been widely utilized for the cathodic protection design of enormous
structures such as ships, offshore structures and underground pipeline
systems [5–9]. Various studies [10–12] have showed that the use of
BEM results in excellent cathodic protection system designs.
Simulations of galvanic corrosion related to the SACP using the BEM
method were in good agreement with the experiment results [13,14].
In this study, the SACP of rail canal structure parts (wheel/axle and rail)
in a seawater environment was simulated using BEM. It was difficult to
apply an organic coating to these parts due to the continuous friction and
stress. Thus, a huge current is needed for impressed current cathodic
protection (ICCP) without the organic coating, and this may cause an
electrical accident. Therefore, SACP is applied for the cathodic protection
method of rail canal structures. The optimum selection, location and size of
the sacrificial anodes were determined, and the optimized cathodic
protection design was investigated using physical miniature tests.

⁎
Corresponding author.
E-mail address: kimjg@skku.edu (J.-G. Kim).

http://dx.doi.org/10.1016/j.enganabound.2017.01.003
Received 30 June 2016; Received in revised form 5 December 2016; Accepted 4 January 2017
Available online 21 January 2017
0955-7997/ © 2017 Elsevier Ltd. All rights reserved.
Y.-S. Kim et al.
Fig. 1. Schematic of rail-canal system.
Fig. 2. Boundary conditions for the cathodic protection computation.
2. BEM simulation
Adey and Niku [15] have provided the mathematical formulation
for a uniform, isotropic electrolyte domain Ω as shown in Fig. 2. Some
simplifying assumptions were made to assist in the computation
including: (i) the solution must be uniform and electro-neutral, (ii)
there is no concentration gradient in the electrolyte solution, and (iii)
the boundary of the container was electrically insulating, so there was
no flow current leakage from the container. Following these assump-
tions, the potential obeys the Laplace equation:
∇2 Φ = 0
The Laplace equation is solved using the following boundary
conditions:
Φ = Φ0, on Γ1,
I = I0, on Γ2,
Ia = fa (Φa ), on Γ3a,
Ic = fc (Φc ), on Γ3c,
Γ is the entire surface of the electrolyte domain which includes Γ1, Γ2 ,
Γ3a and Γ3c . Φ is the potential, and I is the current density across the
boundary. Φ0 and I0 are given constant values of potential and current
density, respectively. fa (Φa ) and fc (Φc ) are functions that indicate the
relationship between the potential and current density of the anode and
cathode, respectively.
Wheel/axle and rail models were created in the Rhinoceros 3D
drawing software based on the real shape and dimensions of the
structures. The dimension data include the length, height, width,
diameters and so on. The inner parts of structure were not modeled
Engineering Analysis with Boundary Elements 77 (2017) 36–48
to avoid computational errors. Three-dimensional models of wheel/
axle and rail structures with detailed dimensions are shown in Fig. 3(a)
and (b), respectively. Also, the cathodic protection system information
was used to provide the size, location and form for simulating the
cathodic protection process. The size of the anode and the anode
location were changed during the process of cathodic protection design,
and the form of the anode was rectangular to simplify the computation
process. After modeling the structure, the designed structure model
was imported into the BEASY software.
Setting the boundary conditions is an essential step in modeling
corrosion and cathodic protection systems with BEASY. The model
structure and surrounding environment (seawater) are enclosed within
a rectangular box that was approximately 30 times bigger than the
modeling structures. In order to avoid current flow from the object, the
vector sum of the current was zero. The conductivity of the environ-
ment was fixed to 3.75 S/m. The relationship between the current and
potential of the cathode and anode gives the polarization state; there-
fore, the cathode and anode electrochemical properties were deter-
mined according to the input polarization data.
3. Experiments
3.1. Materials
The axle/wheel and rail materials were based on carbon steel. The
axle had a nominal composition (wt%) of 0.45% C, 0.30% Si, 0.75%
Mn, 0.010% P, and 0.001% S, while the wheel had a nominal
composition of 0.64% C, 0.25% Si, 0.73% Mn, 0.015% P, 0.003% S,
and 0.025% Cu. The rail had a nominal composition of 0.85% C, 0.5%
Si, 0.64% Mn, 0.003% P, and 0.003% S. Zn and Al sacrificial anodes
(1) were used in this study. The Zn anode had a nominal composition of
0.050% Al, 0.04% Cu, 0.02% Pd, and 0.001% Mg. The Al anode had a
nominal composition of 0.05% Fe, 0.05% Si, 0.01% Cu, 3.5% Zn, 0.03%
In, and 1.0% Mg. Synthetic seawater from ASTM D-1141 was used, and
(2) its chemical composition is listed in Table 1.
(3)
3.2. Potentiodynamic tests
(4)
The polarization data for the cathode and sacrificial anode materials
(5)
were needed to carry out the simulation. The cell used for carrying out
potentiodynamic tests was a conventional three-electrode cell. A
purified carbon rod was used as the counter electrode, and a saturated
calomel electrode (SCE) was used as the reference electrode. The
working electrodes for axle/wheel, rail and sacrificial anodes (Al, Zn)
were rectangular cubes of 1×1×0.3 cm, and these were mounted in
epoxy resin with an exposed area of 1 cm2. Before the tests, the working
electrode was abraded with wet SiC paper (initially 220, 400, 600 and
800 grades), washed with distilled water, degreased with acetone,
washed with ethanol, and finally dried in nitrogen gas.
Potentiodynamic tests were carried out using a Bio-Logic VSP-300
37
Y.-S. Kim et al. Engineering Analysis with Boundary Elements 77 (2017) 36–48

Fig. 3. 3D models of the (a) axle/wheel and (b) rail structures with detailed dimension.

potentiostat. The potential range was from −300 mV vs. open-circuit

Table 1 potential to 0 mVSCE at a scan rate 0.166 mV/s. All polarization tests
Chemical composition of synthetic seawater. were repeated a minimum of three times to confirm the accuracy of the
obtained data.
Component Concentration (g/L).

NaCl 24.53
MgCl2 5.20 3.3. Physical miniature test
Na2SO4 4.09
CaCl2 1.16 Physical miniature tests were carried out to verify the cathodic
KCl 0.695
protection design. The miniature size was 1/10 of that used in the
NaHCO3 0.201
KBr 0.101 numerical model of axle/wheel and rail structures. The axle/wheel was
H3BO3 0.027 connected by silver paste to ensure a good electrical connection. Fig. 4
SrCl2 0.025 shows the physical miniature model used in this study. The sacrificial
NaF 0.003 anode was rectangular, and the size was 1/10 of the numerical model.

38
Y.-S. Kim et al.
Fig. 4. Physical miniature of axle/wheel and rail structures.
In order to evaluate the cathodic protection, one miniature had an
installed sacrificial anode, while the other miniature did not. The
physical miniatures with sacrificial anodes were placed in a tank
containing 20 L synthetic seawater for 45 days. During the immersion
period, images were recorded using a digital camera. After 45 days of
immersion, the weight loss of each miniature model was measured to
confirm the cathodic protection design. Before the weight loss mea-
surements, the corrosion products on the unprotected physical minia-
ture were removed and cleaned for 10 min in a cleaning solution
containing 500 ml HCl, 3.5 g hexamethylenetetramine (C6H12N4) and
enough distilled water to make 1000 ml. The specimens were then
rinsed in distilled water and weighed to 10−4 g for the weight-loss
measurements. Also, the calcareous deposit on the protected physical
miniature was removed using a 3% acetic acid solution and was cleaned
using the same process mentioned above.
4. Results and discussion
A flow chart for the design process for a systematic cathodic
protection system is shown in Fig. 5. The polarization data of each
structure and the sacrificial anode materials are shown in Fig. 6.
Although the axle/wheel structure has a wet-dry cycle during the
operation, the structure is washed off after operation to minimize
corrosion by the concentrated salts during wet-dry cycles. Thus, only
the immersion condition was considered in this study. Both sacrificial
anodes have a lower corrosion potential than the rail structure (axle,
wheel and rail) materials. Therefore, these anodes seem to be appro-
priate for use as sacrificial anodes. For the purpose of the simulation,
the linear region (Tafel region) in the anodic and cathodic parts of the
polarization data was selected and inputted into the material database
in the software. The cathodic protection potential was determined to be
39
Engineering Analysis with Boundary Elements 77 (2017) 36–48
−770 mVSCE according to the cathodic protection criteria. In addition,
the sacrificial anode life was 25 years, in accordance with the general
service life of the rail structure. The axle/wheels and rails structures are
not combined during the design of the cathodic protection system
because the axle/wheels and rails are not fixed to a specific location
during operation. This means that the effect of the combined axle/
wheels and rails continuously changes, and the design of such a
cathodic protection would be complex and have certain limitations.
Thus, the cathodic protection design for this system was separately
carried out, and the combined structures were simulated. Then, the
cathodic protection design was verified by miniature physical tests.
4.1. Cathodic protection design of axle/wheel
Because the wheel is more noble than the axle, a simulation must be
conducted without a sacrificial anode to determine where the sacrificial
anode should be installed. Fig. 7 shows (a) the potential distribution
and (b) the current distribution of the axle/wheel without a sacrificial
anode. An anodic reaction (corrosion) at the connection point between
the axle and wheel would be expected because of the galvanic effect
caused by the potential difference between the wheel (cathode) and
axle (anode) [16–18]. Thus, the sacrificial anode should be installed at
the axle and at the connection regions.
Before the simulation with a sacrificial anode, an adequate sacri-
ficial anode was identified. The axel/wheel structures are subjected to
high stress during operation. The Zn anode was not appropriate as a
sacrificial anode for the axle/wheel because the heavy Zn anode cannot
properly balance the rotating part. Therefore, the light Al anode was
selected in SACP for the axle/wheel structure. Also, to minimize the
influence of the anode weight on the axle/wheel structure, the Al anode
size was fixed to dimensions of 5×3×1 cm, and SACP was determined
Y.-S. Kim et al.
Fig. 5. Flow chart of cathodic protection design in this study.
as a function of the number of anodes.
Fig. 8 shows the potential distribution of the axle/wheel as a
function of Al anode number. A potential below −770 mVSCE was
considered to be an appropriate cathodic protection potential accord-
ing to the international standard. When 1 and 4 Al anodes were
installed, the potential did not satisfy the cathodic protection criterion,
i.e., above −770 mVSCE on the axle/wheel. In contrast, the protection
potential was satisfied in cases where 6, 8 and 12 anodes were installed.
However, both cathodic potential and the lifetime of the sacrificial
anode should be considered in the design of SACP. The required
sacrificial anode life is 25 years, which is the service life of the rail
structure. Actually, the average immersion time of the axle/wheel in
40
Engineering Analysis with Boundary Elements 77 (2017) 36–48
seawater might be half of the service life. Therefore, the adjusted
sacrificial anode lifetime for the axle/wheel was determined to be 12.5
years. Table 2 lists the calculated sacrificial anode life in each case. To
calculate the lifetime of sacrificial anodes (t, day), the following
equation (Faraday’s law) was used [4]:
mnF
t=
Ia (6)
where F is Faraday’s constant (96500 C/mol), n is the number of
equivalents exchanged, a is the atomic weight (a / n=9 at Al), I is the
current measured from the simulation results associated with a 90%
effectiveness of Al [19], and m is the weight (mg) of the sacrificial
Y.-S. Kim et al.
Fig. 6. Polarization curves of axle/wheel, rail structures and sacrificial anodes.
anode (36 g of Al). According to these results, the anode life for service
was satisfied in cases where 6, 8 and 12 anodes were installed.
In summary, the average potential (cathodic potential criteria) and
service life of the anode are important considerations in cathodic
protection design. As shown in Fig. 9, the average potential, Fig. 9(a),
and lifetime of the anode, Fig. 9(b), were satisfied for systems with 6
anodes, which also minimized the number of anodes from an econom-
ics perspective. Thus, the optimized SACP design of the axle/wheel is
the 6 anodes installed in pairs at symmetric locations along the axle.
4.2. Cathodic protection design of rail
The rail has no galvanic effect; therefore, there is no need to
simulate the cathodic protection step. To avoid contact between the
wheel and anode installed on the rail, the anodes were located at the
Fig. 7. Simulation results of axle/wheel without sacrificial anode: (a) potential and (b) current distribution.
41
Engineering Analysis with Boundary Elements 77 (2017) 36–48
side of the rail.
In order to select an appropriate sacrificial anode, the potential
distribution of a 3 m rail section with a Al or Zn anode was simulated as
shown in Fig. 10. Fig. 10(a) shows the potential distribution of the rail
with an Al anode, and Fig. 10(b) indicates the unprotected region with
an Al anode, which was above the −770 mVSCE region. The Al anode
could not provide sufficient protection; therefore, the Al anode was not
an appropriate sacrificial anode for the rail. On the other hand, the Zn
anode decreased the potential of all 3 m rail regions in the range of
−784 and −1051 mVSCE. Thus, the Zn anode could be applied to the rail
structure.
After selecting the sacrificial anode, cathodic protection simulations
were carried out as a function of anode size. Also, the distance between
the anodes was fixed to 6 m to minimize the external influence of other
structures.
Fig. 11 indicates the potential distribution of a 12 m rail as a
function of Zn anode size: (a) 5×3×1 cm, (b) 6×4×2 cm, and (c)
12×5×2 cm. As shown in Fig. 11(a), the small anode does not satisfy
the cathodic protection criterion, whereas the cathodic potential was
satisfied with the other two cases. The designed anode life (25 years)
was also satisfied, and the anode life was calculated using Faraday’s
law. In the anode life calculation, a / n=32.7, the weights were 106.5,
342.2 and 865.2 g, and a 90% effectiveness of Zn [19] was used in the
calculation. Table 3 lists the calculated anode life. As shown in Fig. 12,
the average potential, Fig. 12(a), and life of the anode, Fig. 12(b), were
satisfied only with the 12×5×2 cm (865.2 g) anode. Thus, considering
the potential and anode life, only the 12×5×2 cm (865.2 g) anode
provided the appropriate cathodic protection design for the rail.
4.3. Simulation of combined structures (axle/wheel and rail)
After the cathodic protection design of each structure (axle/wheel
and rail), the potential distributions of combined structures, which are
applied during the cathodic protection design, were investigated as
shown in Fig. 13. The sacrificial anode installed in the middle part of
the rail did not satisfy the cathodic protection potential (−770 mVSCE).
Y.-S. Kim et al. Engineering Analysis with Boundary Elements 77 (2017) 36–48

Fig. 8. Potential distribution of axle/wheel SACP as a function of Al anode number: (a) 1, (b) 4, (c) 6, (d) 8, (e) 12 anodes.

The protection current is not sufficiently supplied to the far region of

Table 2 the rail structure from the sacrificial anode because of the axle/wheel
Lifetime of sacrificial anode (Al) for the each cathodic protection design of axle/wheel
structure.
structures near the sacrificial anode on the rail. However, the dis-
satisfaction value of the cathodic potential is small (about 10 mVSCE),
Number of Al anode Total current from anode ( μ A) Calculated anode life and the axle/wheels and rails are not always combined in specific
(year) regions as mentioned above. Therefore, the influence of the combined
1 132728 7.9
condition does not have a significant effect on the cathodic protection.
4 96876 10.8 In addition, the combined effect would be less than that observed in
6 79488 13.3 actual conditions because the contact resistance between the axle/
8 64584 16.3 wheels and rails increases due to the formation of scale on the surface
12 56304 18.7
of the structure used for cathodic protection.

42
Y.-S. Kim et al.
Fig. 9. Determination of cathodic protection design of axle/wheel as a function of the (a) average potential and (b) anode life.
4.4. Physical miniature test
Physical miniature tests were carried out to verify the cathodic
protection design of the axle/wheel and rail. In previous studies [5,8],
the differences between computational and experimental results were
attributed to geometric simplifications and a lack of material interac-
tions during polarization. The relationships between corrosion and
cathodic protection observed in computational and experimental
results are compared by means of the miniature physical tests.
Although the combined effect is small, the combined situation is the
worst case in the simulation result. Thus, miniature physical tests were
conducted in combined structures. Fig. 14 shows the physical minia-
ture results for axle/wheel and rail after 2 days of immersion in
synthetic seawater. The results of physical miniature test for the rail
43
Engineering Analysis with Boundary Elements 77 (2017) 36–48
showed uniform corrosion because there is no galvanic corrosion. The
connection region between the axle and wheel was preferentially
corroded in the case of the unprotected axle/wheel structure after a
2-d immersion test. This result is similar to the simulation result in
Fig. 7, which shows a higher current density in the connection region.
This means that the computational simulation result was highly
correlated with the miniature physical test at the galvanic corrosion
part. Meanwhile, white scale precipitated on the whole surface in the
protected physical miniature test. Generally, the cathodic current
during cathodic protection increases the dissolved oxygen reduction
reaction (7), which generates hydroxyl ions on the polarized surface.
This increases the surface pH and increases the carbonate ion
concentration (8), resulting in the precipitation of inorganic layers
such as calcium carbonate and magnesium hydroxide. These mixed
Y.-S. Kim et al. Engineering Analysis with Boundary Elements 77 (2017) 36–48

Fig. 10. Potential distribution of 3 m rail installed (a) one Al anode and (b) unprotected region, above −770 mVSCE, (c) one Zn anode.

Fig. 11. Potential distribution of 12 m rail as a function of Zn anode size; (a) 5×3×1 cm, (b) 6×4×2 cm and (c) 12×5×2 cm.

44
Y.-S. Kim et al. Engineering Analysis with Boundary Elements 77 (2017) 36–48

Table 3 Mg2++2OH−→Mg(OH)2 (s) (9)

Lifetime of sacrificial anode (Zn) for the each cathodic protection design of rail structure.

Weight of Zn anode Total current from anode Calculated anode life (year) Ca2++CO32−→CaCO3 (s) (10)
(g) ( μ A)
The calcareous deposit decreases the active surface reaction, which
106.5 211544 6.3 decreases the anode dissolution and the rate of reaction (7). Thus, the
342.2 243936 11.3 calcareous deposit has a beneficial effect on cathodic protection,
865.2 272313 25.8 indicating that cathodic protection occurred on the structure [21].
Fig. 15 shows the physical miniature test results for (a) unprotected
and (b) protected system after 45 days immersion in synthetic sea-
water. A thick corrosion product was formed on the unprotected
deposits are generally called calcareous deposits [20]. physical miniature specimen, whereas calcareous deposits precipitated
on the surface of the protected miniature physical specimen. As
O2+2H2O+4e−→4OH− (7)
mentioned above, this indicates that the protected structure had
sufficient protection potential (negative potential) regardless of the
HCO3−+OH−→H2O+CO32− (8) connection between the axle/wheel and rail structures, as suggested in

Fig. 12. Determination of cathodic protection design of rail as a function of the (a) average potential and (b) anode life.

45
Y.-S. Kim et al.
Fig. 13. Potential distribution of combined structures (axle/wheel and rails), which are equipped with cathodic protection.
Fig. 14. Photograph of unprotected and protected physical miniatures after immersion in synthetic seawater for 2 days.
the computational results. After removing the corrosion product and
calcareous deposit, corrosion rates were determined by the weight loss
method. The corrosion rates of the physical miniature specimens with
and without cathodic protection are listed in Table 4. The corrosion
rate was calculated based on the following equation [4]:
87600W
μ m /yr =
DAT
where W is the weight loss in grams, D is the density of the low carbon
steel (7.85 g/cm3), A is the exposed area in cm2 (axle/wheel: 239 cm2,
rail: 76 cm2), and T is the time in hours (1080 h). Weight loss occurred
on the unprotected structure, while weight loss was not measured in
the protected structure. The corrosion rates for unprotected axle/wheel
and rail structures were 148 and 177 μ m/yr, respectively. This means
that the corrosion only occurred in the unprotected structure. This
indicates that the cathodic protection design was adequately applied,
and the simulation results well reflected the actual state.
5. Conclusion
In the present work, the cathodic protection design of axles, wheels
and rails used in rail canal systems in seawater was carried out using
Engineering Analysis with Boundary Elements 77 (2017) 36–48
the boundary element method. The design was verified using physical
miniature tests. In the axle/wheel cases, the Al anode and the number
of anodes were design factors for minimum load influence. Considering
the cathodic potential criterion and anode life, the case where 6 anodes
were installed was the optimum design for the axle/wheel structure. In
the rail case, the optimum cathodic protection design for the rail was a
12×5×2 cm (865.2 g) Zn anode installed at 6 m intervals based on
(11)
potential and anode life. The influence which degrades the corrosion
protection and the life of anodes at the certain location of the structure,
due to the increase of the protection area, was indicated in the
simulation of the combined structure (axle/wheel and rail). Although
the axle/wheel and rail were not fixed to a specific location, satisfactory
the cathodic protection could be achieved. The cathodic protection
design from the simulation was verified by physical miniature tests.
The unprotected physical miniature specimen was corroded and
showed corrosion products on the surface, whereas the protected
miniature physical specimen formed calcareous deposits on the surface
without weight loss. Thus, the simulated cathodic protection design
results showed trends similar to those of the experiments, and these
results can be applied to the cathodic protection design of an actual rail
canal system.
46
Y.-S. Kim et al. Engineering Analysis with Boundary Elements 77 (2017) 36–48

Fig. 15. Physical miniature models of axle/wheel and rail, (a) unprotected and (b) protected, after 45 days immersion in synthetic seawater.

Acknowledgments
Table 4
Corrosion rates of axle/wheel and rail physical miniature by weight loss method. 1. This research was supported by a Grant from R & D Program of the
Physical miniature Initial Weight after Initial weight Corrosion rate
Korea Railroad Research Institute, Republic of Korea.
weight immersion −weight after 2. This research was supported by a Global Ph.D. Fellowship Program
(g) (g) immersion ( μ m/yr) through the National Research Foundation of Korea (NRF) funded
(g) by the Ministry of Education (2015H1A2A1033362).
Axle/ Protected 611.1 611.1 0 –
w- Unprotected 608.2 2.9 148.4 References
he-
el [1] The Panama Canal Authority (ACP). Proposal for the expansion of the panama
canal by the construction of the third set of locks project. Panama: ACP; 2006.
Rail Protected 149.3 149.3 0 – [2] HKND (Hong Kong Nicaragua Canal Development) Group. Nicaragua canal project
Unprotected 148.2 1.1 177.1 description. Hong Kong: Environmental Resources Management (ERM); 2014.
[3] Seo SI, Lim JY, Kim JG, Choi DH, Kang DH. An innovative Rail-Canal system for
effective shipping logistics. Rotterdam, The Netherlands: The international

47
Y.-S. Kim et al.
Conference on Logistics and Maritime systems; 2014.
[4] Jones DA. Principle and prevention of corrosion, 2nd ed.. New Jersey: Prentice
Hall; 1996.
[5] Lan Z, Wang X, Hou B, Wang Z, Song J, Chen S. Simulation of sacrificial anode
protection for steel platform using boundary element method. Eng Anal Bound
Elem 2012;36:903–6.
[6] Zamani NG. Boundary element simulation of the cathodic protection system in a
prototype ship. Appl Math Comput 1988;26:119–34.
[7] Purcar M, Van den Bossche, Bortels L, Deconinck J, Wesselius P. Numerical 3-D
simulation of a cathodic protection system for a buried pipe segment surrounded by
a load relieving U-shaped vault. Corrosion 2003;59:1019–28.
[8] DeGiorgi VG, Thomas ED, Lucas KE. Scale effect and verification of modelling of
ship cathodic protection system. Eng Anal Bound Elem 1998;22:41–9.
[9] Zamani NG, Chuang JM, Porter JF. BEM simulation of cathodic protection systems
employed in infinite electrolytes. Int J Numer Method Eng 1987;24:605–20.
[10] Brebbia CA, Niku SM. Computational application of boundary element methods for
cathodic protection systems. In: Proceedings of the international conference on
boundary element methods in engineering. Italy: Villa Olmo, Lake Como; 1985.
[11] Niku SM, Adey RA, Brebbia CA. BEASY-CP, CAD system for mathematical
modeling of galvanic corrosion and cathodic protection problems. Washington, DC,
USA: CADMO; 1986.
[12] Adey RA, Niku SM. Computational modeling of galvanic corrosion. Pheonix,
48
Engineering Analysis with Boundary Elements 77 (2017) 36–48
Arizaona: In proceedings of the ASTM conference; 1986.
[13] Jia JX, Song GL, Atrens A. Experimental measurement and computer simulation of
galvanic corrosion of magnesium coupled to steel. Adv Eng Mater 2007;9:65–74.
[14] Jia JX, Song GL, Atrens Am John St D, Baynham J, Chandler G. Evaluation of the
BEASY program using linear and piecewise linear approaches for the boundary
conditions. Mater Corros 2004;55:845–52.
[15] Adey RA, Niku SM. Computer modeling of galvanic corrosion. ASTM Spec Tech
Publ 1998;978:96–117.
[16] Santos WJ, Santiago JAF, Telles JCF. Optimal positioning of anodes and virtual
sources in the design of cathodic protection system using the method of funda-
mental solutions. Eng Anal Bound Elem 2014;46:64–74.
[17] Bellezze T, Fratesu R. Assessing the efficiency of galvanic cathodic protection inside
domestic boilers by means of local probes. Corros Sci 2010;32:3023–32.
[18] Zhang GA, Yu N, Yang LY, Guo XP. Galvanic corrosion behaviour of deposit-
covered and uncovered carbon steel. Corros Sci 2014;86:202–12.
[19] Gurrappa I. Cathodic protection of cooling water system and selection of appro-
priate materials. J Mater Process Technol 2015;166:256–67.
[20] Ru Lee, Ambrose JR. Infulence of cathodic protection parameters on calcareous
deposit formation. Corrosion 1988;44:887–91.
[21] Rousseau C, Baraud F, Leleyter L, Jeannin M, Gil O. Calcareous deposit formed
under cathodic protection in the presence of natural marine sediments: A 12 month
experiment. Corros Sci 2010;52:2206–18.