i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 7 0 3 6 e7 0 4 6

Available online at www.sciencedirect.com

ScienceDirect

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

CFX study of flow accelerated corrosion via mass

transfer coefficient calculation in a double elbow

Korosh Keshtkar, Mohammadreza Nematollahi*, Ali Erfaninia

Shiraz University, School of Mechanical Engineering, Mollasadra St, 71348-51154, Shiraz, Iran

article info abstract

Article history: Using CFX numerical simulation, this study aims at calculating Mass transfer coefficient
Received 27 October 2015 (MTC) as a parameter that indicate the FAC rate on the wall of the double elbow. A double
Received in revised form elbow in the feeder water piping system of a typical combined cycle power plant that had
6 February 2016 been damaged due to the FAC was considered and was studied. In this study the influence
Accepted 8 February 2016 of several parameters such as the mean flow velocity and the dimensionless distance
Available online 21 March 2016 between the two elbows (L/D) on the MTC at the elbows was investigated. The geometrical
factor and the close proximity of two elbows with respect to the mass transfer coefficient
Keywords: were studied.
Flow accelerated corrosion It is found that mass transfer coefficient increases by Reynolds number. The maximum
Mass transfer coefficient MTC was occurred in the intrados of second elbow. It is found that there is a very good
Double elbow coincidence between the locus in which the maximum MTC was calculated and the locus
CPE in which minimum wall thickness was measured.
It is observed that as the remoteness between the elbows (L/D) increases, the Close
Proximity Effect (CPE) is decreasing for L/D < 7. As L/D gets amplified (L/D > 7) the CPE is
increasing very slowly and became constant. This CPE value is lower in comparison with
that of L/D < 7. This is due to the fact that, when L/D > 7, the second bend experiences less
impact due to the first bend and, also, there is a fully developed flow between both the
bends.
Copyright © 2016, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights
reserved.

Factors such as the geometrical configuration and the orien-

Introduction
Flow accelerated corrosion (FAC) is a slow piping degradation
process that is caused by the fluid flow damaging or thinning
the protective layers of piping components. It is essentially a
three step process: (a) a series of electrochemical reactions at
the metaleoxide interface, (b) dissolution of iron production
(Fe2þ) at the oxide/water interface (c) transfer of the corrosion
products to the bulk flow across the diffusion boundary layer.
tation of the piping component, fluid temperature and piping
material can significantly affect FAC [1,2].
Although the FAC is characterized by a general reduction in
the pipe wall thickness for a given piping component, it
frequently occurs over a limited area within this component
due to the local high area of turbulence.
Several catastrophic failures have been reported in several
power plants around the world since 1981 due to FAC [3]. At
Surry Unit 2 power plant in 1986 severe elbow rupture

* Corresponding author.
E-mail address: nema@shirazu.ac.ir (M. Nematollahi).
http://dx.doi.org/10.1016/j.ijhydene.2016.02.072
0360-3199/Copyright © 2016, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 7 0 3 6 e7 0 4 6
happened in the downstream of T-bend and caused 4 fatal-
ities. At Millstone 3 in 1990 failure occurred in the downstream
of control valves, caused the failure of two parallel trains but
no injuries. At Louviisa-1 in 1990 failure occurred in the
downstream of feeder water systems, without injuries. At
Prairie Power Plant in 1995, FAC failure occurred at the
downstream of T-bend caused the two fatalities. At Fort Col-
houn in 1997 failure occurred at the bend, but no injuries. At
Mihama 3 in 2004 failure occurred in the downstream of
orifice, caused five fatalities and several injuries. Very
recently, at Iatan fossil power plant in 2007 failure occurred at
the downstream of the control valve, caused the two fatalities
and a huge capital of plant loss [4].
The FAC in piping systems of power plants have been
observed where complex flow occurs.
Piping elbows have been identified as one of the most
common components prone to FAC. Flow in a 90-degree bends
is subject to severe changes in the flow direction, leading to
the development of secondary flows and/or flow separation.
The secondary flows induce a pressure drop along the elbow,
which can significantly increase the wall shear stresses [4], as
well as the flow turbulence generated close to the wall. These
mechanisms are known to be the hydrodynamic governing
factors responsible for FAC [2,5].
Two or three successive bends are widely used in industry.
In particular, applications include large-diameter pipe work or
ducting in chemical processes, desalination plants, water
supply and nuclear power stations, where the manufacture of
single bends may be either impractical or uneconomical.
In this study one such critical geometry, namely, the dou-
ble elbow is considered and the FAC is studied via mass
transfer coefficient calculation in it.
The double elbow belongs to the feeder pipes of a typical
combined cycle power plant. Using Computational Fluid
Dynamic (CFD), this study aims at calculating mass transfer
coefficients on the wall of the double elbow in different flow
regimes as well as in different close proximity of the
elbows.
In this study by calculating the mass transfer coefficient
from the inner wall of the pipe to the bulk flow, the effect of
several hydrodynamic parameters on the FAC is investigated.
On the contrary of surveys in the literatures in which flow
under FAC condition is analyzed, this study focuses on the
mass transfer coefficient as the most important parameter
affecting FAC rate.
The results will be used to promote design and enhance the
safety and reliability of the piping systems in fossil and nu-
clear power plants in order to minimize the degradation of
pipes due to the FAC.
Material and methods
In order to investigate the FAC at the double elbow, by simu-
lating a given double elbow with known operating condition
the CFX numerical simulation is performed to calculate the
Mass transfer coefficient (MTC) as a parameter that indicate
FAC rate. The influences of mean flow velocities as well as the
distance between the two elbows on the MTC are studied
numerically.
7037
In order to validate the numerical results, ultrasonic
thickness measurement of damaged double elbow is
compared with the trend of MTC that was calculated
numerically.
Numerical simulation and modeling
In the present study, typical combined cycle power plant
operating conditions of the feeder piping system were
considered. The uniform velocities are used as the inlet
boundary condition. No-slip, constant temperature and con-
stant concentration of the species are the boundary condi-
tions imposed on the wall. At the outlet, the zero-gradient
properties are considered to be linear for pressure.
Operating conditions and geometrical characteristics are
presented in Table 1. Figs. 1 and 2 show the schematic of the
double elbow and the meshes, respectively.
The numerical simulation of the governing equations
subjected to the boundary conditions was performed by uti-
lizing the commercially available software, namely ANSYS
CFX15. The RNG ke3 models are used for turbulence simula-
tion. The RNG ke3 model was derived using a rigorous statis-
tical technique (called renormalization group theory). It is
similar in form to the standard ke3 model, but includes the
following refinements:
 The RNG model has an additional term in its 3 equation that
significantly improves the accuracy for rapidly strained
flows.
 The effect of swirl on turbulence is included in the RNG
model, enhancing accuracy for swirling flows.
 The RNG theory provides an analytical formula for turbu-
lent Prandtl numbers, while the standard ke3 model uses
user-specified, constant values.
 While the standard ke3 model is a high-Reynolds-number
model, the RNG theory provides an analytically-derived
differential formula for effective viscosity that accounts
for low-Reynolds-number effects.
These features make the RNG ke3 model more accurate
and reliable for a wider class of flows than the standard ke3
model [6].
A mesh sensitivity study was done to check on the in-
fluence of the mesh resolution on the results and to mini-
mize numerical influences introduced by the size of meshes
and their distributions. For mesh sensitivity analysis, four
Table 1 e Operating conditions and geometrical
characteristics of the double elbow.
Parameter Quantity
Outside diameter of main pipe 115 mm
The thickness of main pipe 5 mm
Fluid water
Temperature 120  C
Pressure 2 bar
Material of the pipe Carbon steel
Radius of curvatures 170 mm
L/D 2.85
Schmidt number 42
7038 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 7 0 3 6 e7 0 4 6

Fig. 1 e The schematic of the double elbow and meshes.

Fig. 2 e The 3D view of the mesh in the pipe.

meshes of differing size were used. The meshes with 382376
nodes, 609034 nodes, 1206870 nodes and 3055931 nodes
were used. The mesh independence of the simulation is
depicted in Fig. 3 by showing the changes of the mean flow
velocity along the center line of the pipe for the four
meshes. In order to have the best convergence of solutions
as well as reasonable Yþ the mesh with 609034 nodes was
used for simulation.
The mesh types of Hexahedral and tetrahedral were uti-
lized. The mesh quality metrics such as the average skewness
and the average orthogonal quality were 0.211 and 0.914,
respectively. A proper inflation with 15 layers and first layer
thickness of 0.5 mm was used for near wall meshing. Fig. 2
depicts the 3D view of mesh in the pipe.
The dimensionless wall distance value (yþ) for the near-
wall cells is less than 60 with the scalable wall function
which is used by CFX for near wall treatment. In all simula-
tions residual RMS Error values have reduced to an acceptable
value of 106. Fig. 4 shows the residual RMS Error values for
three successive numerical solutions.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 7 0 3 6 e7 0 4 6
Fig. 3 e The mean velocity along the center line of the pipe for different meshes in order to show the mesh independency of
the numerical solutions.
Mass transfer coefficient (MTC)
Generally, the iron reacts with water to form a surface oxide
layer on the inner surface of pipelines. The rate of iron
removal (the FAC rate) is controlled by the rate of diffusion of
dissolved iron species through the boundary layer of water
near the surface in the bulk water. This diffusion (or mass
transport of iron away from the surface) depends directly on
the concentration of soluble iron species at the oxide surface
and inversely on the thickness of the boundary layer. Thus, a
decrease of the boundary layer thickness because of the
increased water flow rate or because of local turbulence cau-
ses an increase of corrosion rate thus increasing the FAC rate
[7e13].
The FAC mechanism involves convective mass transfer of
the ferrous ions in the water. The convective mass transfer for
single phase flow is known to be dependent on the hydrody-
namic parameters near the wall interface, such as flow ve-
locity, local turbulence, geometry, and surface roughness. In
addition, the physical properties of the transported species or
the water do not affect the local transport rate in adiabatic
flow, especially when the temperature changes in piping
system are negligible. Over a limited length of piping
component, FAC rate is considered as a direct function of the
mass flux of ferrous ions.
The convective mass transfer of ferrous ions from the
oxide water interface through the boundary layer of water into
the bulk of water was analyzed and was related to MTC and
ferrous ion concentration as follow:
Mass flux of ferrous ions ¼ FAC rate ¼ MTCðcw  cb Þ
Where MTC is the mass transfer coefficient, cw is the con-
centration of the ferrous ions at the oxide water interface and
cb is the concentration of the ferrous ions in the bulk fluid. For
the long piping the mass transfer analysis is not used directly
to calculate FAC rate till we know the concentration difference
of ferrous ions at the oxide water interface and in the bulk
7039
water [14e23]. But, the concentration difference depends on
the first and the second mechanism of FAC and it is not
possible to calculate in this study.
This analysis can be applied when the piping is short and
the FAC rate is dominated by mass transfer coefficient [16,24].
The mass transfer coefficient was analyzed extensively to
predict the wall thinning locations in the double elbow under
operating conditions of a typical combined cycle power plant.
Under the turbulent flow condition the MTC was calculated
based on the ChiltoneColburn equation [17]. This equation is
written in terms of the wall shear stress (t), mean velocity (U),
density (r) and Schmidt number (Sc) and is as follows:
m  t  2
MTC ¼ Sc 3 (2)
s rU
In this study the influence of several parameters such as
the mean flow velocity, the dimensionless distance between
the two elbows (L/D) on the MTC at the elbows was investi-
gated. The geometrical factor and the close proximity of two
elbows with respect to the mass transfer coefficient were
studied.
Results and discussion
The changes of the maximum MTC (MTCMax) on the pipe wall
of the double elbow versus Reynolds number are shown in
Fig. 5. The maximum mass transfer coefficient is occurring in
the intrados of the second elbow. It is found from Fig. 5 that
(1) the MTCMax increases due to increasing the Reynolds number
of the flow inside the pipe.
For different Reynolds number the change of MTC along
the X ordinate on the wall of pipe between two successive
elbows, which was shown in Fig. 1, is depicted in Fig. 6. The
locus of the second elbow is shown by the dash lines in the
graph. It is found that the MTC increases along the X ordi-
nate along the pipe. The vortices which are formed by the
7040 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 7 0 3 6 e7 0 4 6

Fig. 4 e The residual RMS Error values for three successive numerical solutions in order to show the convergence of the
numerical solutions.

first elbow in the flow field cause an increase of turbulence
at the locus of the second elbow. The turbulence in turn
increases the mass transfer coefficient on the second elbow
wall. It is found from Fig. 6 that MTC increases by Reynolds
number.
The secondary flow induces a pressure drop along the
distance between two elbows, which can significantly in-
crease the wall shear stresses and the turbulence intensity
close to the wall. The higher the velocity the higher the in-
tensity of secondary flow; which in turn enhances the rate of
mass transfer. In addition, as the flow velocity increases the
thickness of the hydrodynamic boundary layer decreases, as
a consequence the thickness of the diffusion layer of the wall
of the second elbow decreases as well. That decrease in
diffusion layer thickness enhances the rate of mass transfer
over the elbow.
Fig. 7 shows the effect of dimensionless distance between
the two elbows (L/D) on the MTCMax at the second elbow for
two Reynolds numbers. By increasing L/D, MTCMax decrease
gradually and it is found that for L/D higher 7 there are no
significant changes in MTCMax and it remains constant.
The size and the geometry of a piping component directly
influence the flow velocity and hence the local mass transfer
rate. In addition, a component with complex geometries tends
to experience higher FAC rate such as elbow, tee, reducers and
valves, etc. The effect of turbulence on the FAC rate is
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 7 0 3 6 e7 0 4 6 7041

Fig. 5 e The changes of maximum mass transfer coefficient versus Reynolds number.

Fig. 6 e The changes of MTC along the X ordinate for different Reynolds number.

represented by the geometry enhancement factor as such components tend to experience more severe FAC as
described by Chexal et al. [25]. Generally, a component which discussed by Kastner et al. [26] and Poulson [27]. In fact, the
is located next to another one is subjected to more turbulence double elbow, where an elbow located downstream the other
that further increases the FAC rate. One would expect that elbow, is a strong evidence of the proximity effect.
7042 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 7 0 3 6 e7 0 4 6

Fig. 7 e The effect of dimensionless distance between two elbows (L/D) on the MTCMax at the second elbow for two Reynolds
numbers.

Fig. 8 e The effect of dimensionless distance between two elbows (L/D) on the close proximity in double elbow for two
Reynolds numbers.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 7 0 3 6 e7 0 4 6 7043

Fig. 9 e Distribution of geometrical factor in double elbow (Re ¼ 2.941eþ5 and Sc ¼ 42).

Fig. 10 e The calculated MTC results on the wall of the second elbow.
7044 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 7 0 3 6 e7 0 4 6
Fig. 11 e The calculated MTC along X ordinate on the intrados line of the second elbow.
In general, for the geometry with the close proximity of
bends, the downstream bend experiences a higher FAC rate. In
the present study, the close proximity effect (CPE) of bends is
investigated for fixed Re ¼ 1.764e þ 5 and Re ¼ 3.528e þ 5 as
well as for different L/D using the following correlation [16].
MTCs  MTCf
CPE% ¼  100
MTCf
where MTCs denotes the maximum value of MTC at the
downstream bend and MTCf is the maximum value of MTC at
the upstream bend (reference bend). These simulated results
are shown in Fig. 8 and it is observed that as the remoteness
between the elbows (L/D) increases, the CPE is decreasing for
L/D < 7, since the first bend effect is less on the second elbow.
As L/D gets amplified (L/D > 7) the CPE is increasing very
slowly and became constant. This CPE value is lower in
comparison with that of L/D < 7. This is due to the fact that,
when L/D > 7, the second bend experiences less impact due to
the first bend and, also, there is a fully developed flow between
both the bends [16].
One of the indices for mass transfer evaluation in double
elbow is the geometrical factor, which is the ratio of the mass
transfer coefficient of the flow through the elbow to that of the
circular pipe at the same Reynolds number [28]. It should be
mentioned that the geometrical factor has been used for the
estimation of the wall-thinning rate in the pipeline of nuclear/
fossil power plants. Fig. 9 shows the distribution of the
geometrical factor calculated for double elbow for
Re ¼ 2.941eþ5 and Schmidt (Sc) number 42. The locus of first
and the second elbow are shown by the dash lines. The results
(3) show a maximum value 1.7 at the locus of second elbow.
Fig. 10 depicts the calculated MTC on the second elbow. It is
clear that the maximum MTC is occurring in the intrados of
the second elbow.
The calculated MTC along X ordinate on the intrados line of
the second elbow is shown in Fig. 11.
Fig. 12 shows the result of Ultrasonic Technique (UT) wall
thickness measurement of double elbow along the X ordinate
and calculated MTC. It is found that there is a very good
coincidence between the locus in which the maximum MTC
was calculated and the locus in which minimum thickness
was measured.
Conclusion
In this study the mass transfer coefficient (MTC) as a
parameter that indicate FAC rate was studied in a double
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 7 0 3 6 e7 0 4 6 7045

Fig. 12 e The result of Ultrasonic Technique (UT) wall thickness measurement of double elbow along the X ordinate and
calculated MTC.

elbow under operational condition of a typical combined Reynolds number must be set in the range in which MTC
cycle power plant. get minimized. Additionally, the L/D for two elbows must
The CFX numerical simulation was performed to calculate be greater than 7.
the MTC by utilizing the ChiltoneColburn analogy. In this
study the influence of several parameters such as the mean
flow velocity, the dimensionless distance between the two
elbows (L/D) on the MTC at the elbows was investigated. The
geometrical factor and the close proximity of two elbows with
Feature works
respect to the mass transfer coefficient were studied. The re-
Since there isn't any experimental study under the same
sults of this study are as follows:
condition of this research published in the literatures, Particle
Image Velocimetry (PIV) Study, as well as chemical investi-
 The mass transfer coefficient increase by Reynolds num-
gation of the mass transfer coefficient can be done to validate
ber. The maximum MTC is occurring in the intrados of
the numerical results as feature works.
second elbow.
 There is a very good coincidence between the locus in
which the maximum MTC was calculated and the locus in
which minimum thickness was measured by UT.
Acknowledgment
 As the remoteness between the elbows (L/D) increases, the
Close Proximity Effect (CPE) is decreasing for L/D < 7, since
There are special thanks to all individuals who helped the
the first bend effect is less on the second elbow.
authors to do this research.
 As L/D gets amplified (L/D > 7) the CPE is increasing very
slowly and became constant. This CPE value is lower in
comparison with that of L/D < 7. This is due to the fact that,
when L/D > 7, the second bend experiences less impact due references
to the first bend and, also, there is a fully developed flow
between both the bends.
 The results will be used to promote design and enhance the [1] Ahmed WH, Bello MM, Al-Sarkhi A, El Nakla M. Flow and
safety and reliability of the piping systems in fossil and mass transfer downstream of an orifice under flow
accelerated corrosion conditions. Nucl Eng Des
nuclear power plants in order to minimize the degradation
2012;252:52e67.
of pipes due to the FAC. [2] Muhammadu MM, Sheriff J, Hamzah E. A review of literature
 In order to minimize the FAC rate in the two elbows, which for flow accelerated corrosion of mitred bends. Int J Emerg
is utilized in industrial areas, it is recommended that the Technol Adv Eng 2013;3:668e77.
7046 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 1 ( 2 0 1 6 ) 7 0 3 6 e7 0 4 6
[3] Kanster W, Erve M, Henzel N, Stellwag B. Calculation code for
erosion corrosion induced wall thinning in piping system.
Nucl Eng Des 1990;119:431e8.
[4] Ahmed WH. Evaluation of the proximity effect on flow
accelerated corrosion. Ann Nucl Energy 2010;37:598e605.
[5] Chen X, McLaury BS, Shirazi SA. A comprehensive
procedure to estimate erosion in elbows for gas/liquid/
sand multiphase flow. J Energy Resour Technol
2006;128:70e8.
[6] ANSYS version 15 Users Guide (2014).
[7] Cubicciotti D. Flow assisted corrosion of steel and the
influence of Cr and Cu additions. J Nucl Mater
1988;152:259e64.
[8] Ho Moon J, Ho Chung H, Sung Ki W, Kim Uh Ch, Pho Jae S.
Dependency of single phase FAC of carbon steel and low
alloy steels for NPP system piping on pH, orifice distance and
material. Nucl Eng Technol 2005;37:375e84.
[9] Shoji T, Lu Z, Takeda Y, Sato Y. Towards proactive materials
degradation management in NPP e today and future. In: CD
proceedings of the 14th Asia Pacific corrosion control
conference (APCCC), October 21e24; 2006.
[10] International Atomic Energy Agency. Material degradation
and related managerial issues at nuclear power plants. In:
Proceedings of technical meeting held during February
15e18, 2005 at Vienna by International atomic energy Agency
(IAEA); 2006.
[11] Remy FN, Bouchacourt M. Flow-assisted corrosion: a method
to avoid damage. Nucl Eng Des 1992;133:23e30.
[12] Kain V, Roychowdhury S, Mathew T, Bhandakkar A. Flow
accelerated corrosion and control strategies in the secondary
circuit pipelines in Indian nuclear power plants. J Nucl Mater
2008;383:86e91.
[13] Watanabe Y, Sue K, Abe H. Effects of Cr content and
environmental factors on FAC rate of carbon steels. In:
Proceedings of 2009 ASME pressure vessel and piping
division conference, July 26e30; 2009.
[14] Pietralik JM, Smith BAW. CFD applications to flow
accelerated corrosion in feeder bends. In: 14th
International conference on nuclear engineering (ICONE-
14); 2006.
[15] Pietralik JM. Mass transfer effects in feeder flow-
accelerated corrosion wall thinning. In: 18th CNS
International conference on CANDUMaintenance, 16e18
November; 2008.
[16] Pietralik JM, Schefski CS. Flow and mass transfer in bends
under flow accelerated corrosion wall thinning conditions. J
Eng Gas Turb Power 2011;133:1e7.
[17] Fingjun L, Lin Y, Li X. Numerical simulation for carbon steel
flow-induced corrosion in high-velocity flow seawater. J
Anti-Corros Method Mate 2008;55(2):66e72.
[18] Zeng L, Zhang GA, Guo XP, Chai CW. Inhibition effect of
thioureidoimidazoline inhibitor for the flow accelerated
corrosion of an elbow. Corros Sci 2015;90:201e15.
[19] El-Gammal M, Ahmed WH, Ching CY. Investigation on wall
mass transfer characteristics downstream of an orifice. Nucl
Eng Des 2012;242:353e60.
[20] Singh JL, Kumar Umesh, Kumawat N, Kumar Sunil, Kain V,
Anantharaman S, et al. Flow accelerated corrosion of carbon
steel feeder pipes, from pressurized heavy water reactors. J
Nucl Mater 2012;429:226e32.
[21] Takano T, Yamagata T, Ito A, Fujisawa N. Mass transfer
measurements behind an orifice in a circular pipe flow for
various combinations of swirlintensity and orifice bias. J
Power Energy Syst 2012;6:402e11.
[22] Fujisawa N, Yamagata T, Kanno S, Ito A, Takano T. The
mechanism of asymmetric pipe-wall thinning behind an
orifice by combined effect of swirling flow and orifice bias.
Nucl Eng Des 2012;252:19e26.
[23] Yamagata T, Ito A, Sato Y, Fujisawa N. Experimental and
numerical studies on mass transfer characteristics behind
an orifice in a circular pipe for application to pipe-wall
thinning. Exp Therm Fluid Sci 2014;52:239e47.
[24] Jianu OA, Rosen MA, Naterer G f, Wang Z. Two-phase bubble
flow and convective mass transfer in water splitting
processes. Int J Hydrogen Energy 2015;40:4047e55.
[25] Chexal B, Horowitz J, Jones R, Dooley B, Wood C,
Bouchacourt M, et al. Flow-accelerated corrosion in power
plants. Electric power research institute report no. TR106611.
1996.
[26] Kastner W, Erve M, Henzel N, Stellwag B. Calculation code for
erosion corrosion induced wall thinning in piping systems.
Nuc Eng Des 1990;119:431e8.
[27] Poulson B. On environmental degradation of materials in
nuclear power systems. In: Proc of 13th Int Conf, Whistler,
B.C., Canada; 2007.
[28] Keller H. Erosionskorrosion an nassdampfturbinen. VGB-
Kraftwerkstechnik 1974;54:292e5.