- Email: [email protected]
Effects of sliding amplitude and normal load on the fretting wear behavior of alloy 690 tube exposed to high temperature water
Accepted Manuscript Effects of sliding amplitude and normal load on the fretting wear behavior of alloy 690 tube exposed to high temperature water Xianglong Guo, Ping Lai, Lichen Tang, Jiamei Wang, Lefu Zhang PII:
S0301-679X(17)30335-3
DOI:
10.1016/j.triboint.2017.07.001
Reference:
JTRI 4806
To appear in:
Tribology International
Received Date: 3 May 2017 Revised Date:
30 June 2017
Accepted Date: 2 July 2017
Please cite this article as: Guo X, Lai P, Tang L, Wang J, Zhang L, Effects of sliding amplitude and normal load on the fretting wear behavior of alloy 690 tube exposed to high temperature water, Tribology International (2017), doi: 10.1016/j.triboint.2017.07.001. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Effects of sliding amplitude and normal load on the fretting wear behavior of alloy 690 tube exposed to high temperature water Xianglong Guo1, Ping Lai1, Lichen Tang2, Jiamei Wang1, Lefu Zhang1* 1
School of Nuclear Science and Engineering, Shanghai Jiao Tong University, NO. 800
2
RI PT
Dongchuan Road, Shanghai 200240, China
Shanghai Nuclear Engineering Research & Design Institute, No. 29 Hongcao Road, Shanghai 200233, China
SC
Contacting email: [email protected]
M AN U
Abstract： ：The effects of sliding amplitude and normal load on the fretting wear behavior of alloy 690 tube exposed to an environment simulating the secondary side water chemistry of nuclear power plant are studied and the fretting wear process is discussed. The results indicate that with the
TE D
increase of sliding amplitude and normal load, the wear volume of the alloy 690 tube is increasing, however, the wear coefficient firstly
EP
decreases then becomes relatively constant. This is because that the wear mode of the materials is changed from abrasive wear to delamination
AC C
wear with the increase of sliding amplitude and normal load. The synergic effects of corrosion and mechanical wear on fretting wear behavior of the alloy 690 tube are also discussed and a model is proposed to qualitatively reveal the fretting wear process of the alloy 690 tube. Key words: Alloy 690; Fretting wear; Sliding amplitude; Synergic effect 1 Introduction The nuclear power plant (NPP) steam generator (SG) tubes, mostly 1
ACCEPTED MANUSCRIPT
made of Alloy 690 [1-2], experience flow induced vibration in the service life, and this results in the fretting wear between the tubes and the supporting parts due to the contact [3-4]. The damage brought about by
RI PT
the fretting wear can lead to the reduction of the service life of the thin SG tubes [5], so the safety and reliability of the SG in the NPP is highly dependent on the fretting wear characteristics.
SC
Lots of researches have been carried out to study the fretting wear
M AN U
behavior of the Alloy 690 mated with different kinds of materials. Guérout and Fisher [6] summarized the results of former researches (carried out before 1999) on the fretting wear of SG tubes and found that temperature had a significant effect on tube wear damage. Lim and Lee [7]
TE D
studied the fretting wear behavior of Alloy 690 in room temperature and concluded that the normal load and the vibrating amplitude were key parameters that could influence the friction in fretting. Mi and Zhu [8]
EP
investigated the influence of temperature on the fretting wear of Alloy
AC C
690 mated with 405 SS in the air and found that the wear volume increased from room temperature to 90℃ and decreased from 90℃ to 285 ℃ . Xin and Lu [9] revealed the microstructure evolution of subsurface on the Alloy 690TT alloy subjected to dry fretting wear at 285℃ in air and the experimental results indicated that five layers existed in the subsurface. Besides experimental studies, some simulation works have also been carried out to study the fretting wear behavior of Alloy 690, for 2
ACCEPTED MANUSCRIPT
example, Lee and Chai [10] employed the influence function method to investigate the fretting wear problems on the secondary side of Alloy 690 SG tubes and the fretting wear of tube-to-plate contact was successfully
RI PT
simulated. Those researches have contributed a lot to reveal the mechanisms in the fretting wear of Alloy 690 and help the NPP designers to increase the safety and reliability.
SC
However, up until now, the above researches and lots of other
M AN U
studies [11-13] were carried out either in the air or in the water at a temperature much lower than 285℃. Although those researches can provide useful information to speculate the fretting wear behavior of the materials
under
service
conditions,
experiments
under
service
TE D
environments is still of great necessity to evaluate the actual performance of the materials. According to the researches carried out by Fouvry [14-16], various parameters, such as normal load, sliding amplitude and
EP
frequency, can influence the fretting wear behavior of the materials. As
AC C
for SG tube, the effects of those parameters on its fretting wear behavior also need to be studied to evaluate its servicing life. The present study aims at revealing the effects of sliding amplitude
and normal load on the fretting wear behavior of Alloy 690 mated with 405 stainless steel (SS) in the service environment. The damage mechanisms in this condition are discussed based on the chemical and microstructure analysis. A model considering the mechanical and 3
ACCEPTED MANUSCRIPT
corrosion factors is proposed to qualitatively reveal the fretting wear process of alloy 690 tube exposed to high temperature water. 2 Experimental procedure
RI PT
In the present research, the Alloy 690 tubes with an inner diameter of 15.5mm, external diameter of 17.5mm and length of 16mm were used, and the supporting part was made of 405 SS with length of 12mm, width
SC
of 7.6mm and height of 3mm. The roughness (Ra) of the Alloy 690 and
M AN U
405 SS were examined as 0.254µm and 0.376µm, respectively. The composition of Alloy 690 and 405 SS was shown in Table 1. Table 1. Composition of alloy 690 and 405 SS in wt.% Ni
Cr
Alloy 690
Bal.
30.3
405 SS
0.50
12.5
Fe
Al
C
Si
Mn
S
P
9.6
0.25
0.023
0.30
0.23
0.002
0.008
Bal.
0.15
0.056
0.60
0.58
0.013
0.025
TE D
Chemical composition
Table 2 Testing conditions for the fretting wear behavior of alloy 690 tube mated with 405 SS Value (unit)
Temperature Pressure pH Dissolved oxygen (DO) Frequency Cycles
285℃ 8.6MPa 9.75 ＜5 µg/L 5Hz 500000 20µm 60µm 100µm 10N 20N 40N
AC C
EP
Parameter
Sliding amplitude (The normal load is kept constant, and the value is 40N)
Normal force (The sliding distance is kept constant, and the value is 100µm)
A fretting wear test setup with a tube-on-plate configuration was developed to examine the performance of the materials, as schematically 4
ACCEPTED MANUSCRIPT
shown in Figure 1(a). The oscillatory motion, with amplitude from 5µm to 200µm, was controlled by the electric motor. The normal load between the SG tubes and the plates was imposed by springs. A heating and water
RI PT
supply loop system was applied to the testing setup to control the water chemistry in the autoclave, as shown in Figure 1(b). The testing parameters used in present research were listed in Table 2. The solution in
SC
the autoclave was distilled water containing certain amount of ammonia
M AN U
to control the pH as 9.75. The temperature (285℃) was similar to the secondary side of NPP, and the pressure is controlled as 8.6 MPa (higher than the saturated vapor pressure at 285℃) to make sure the water environment in the autoclave. Argon was bubbled into the autoclave to
TE D
control the dissolved oxygen (DO) in the solution. The load and amplitude acting on the specimen was set according to the data estimated
AC C
EP
by the NPP.
5
ACCEPTED MANUSCRIPT Figure 1 Schematic picture of (a) configuration between alloy 690 and 405 SS and (b) the fretting wear test setup
The phase and microstructure of the alloys were examined with X-ray diffraction (XRD), optical microscope (OM) and electron
RI PT
backscattered diffraction (EBSD) respectively. Conventional mechanical polishing followed by vibration polishing were used to prepare the
SC
samples for scanning electron microscopy (SEM) and EBSD analysis. For metallographic observation, alloy 690 was etched in a solution of 4%
etched in aqua regia for 15s.
M AN U
hydrofluoric acid, 4% nitric acid and 92% water for 45s. 405 SS was
The morphology of the worn scar was examined with a 3D optical microscope (Bruker Contour GT-I). The volume of the worn scar was
TE D
calculated by the software equipped with the microscope, and the principle of calculating was depth integration across the area. SEM and
EP
energy dispersive spectroscopy (EDS) analysis were also carried out to study the worn scar morphology and composition after fretting wear test.
AC C
3 Results and discussions
3.1 Microstructure of the materials Figure 2 shows the XRD diffraction pattern and microstructure of
alloy 690 and 405 SS. It can be found from the XRD analysis that both alloy 690 and 405 SS have single phase microstructures. In 405 SS, only ferrite is detected, however in alloy 690, only γ-Ni is detected. In order to systematically reveal the status of the materials, the microstructure of the 6
ACCEPTED MANUSCRIPT
alloy 690 and 405 SS is studied by OM and EBSD, as shown in Figure 2(b) to (d). The grain size of Alloy 690 alloy is about 63µm, and the EBSD analysis indicates that a typical austenitic microstructure exists in
RI PT
the tested alloy 690, and twin crystals are also widely distributed in the
TE D
M AN U
SC
materials. For 405 SS, the average grain size is about 52µm.
EP
Figure 2 (a) Phase content of alloy 690 and 405 SS, (b) and (c) microstructure of alloy 690, (d) microstructure of 405 SS
3.2 Wear volume and morphology analysis
AC C
The 3D-profile micrographs of the fretting tube scars tested with
different sliding amplitude and normal load are shown in Figure 3. It can be found that no matter what the sliding amplitude or normal load is, the signs of worn scars are obvious. With the increase of sliding amplitude and normal load, the size and depth of the worn scars is increasing, which indicates that the wear volume of the materials would be increased.
7
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
Figure 3 The 3D-profile micrographs of the fretting tube scars tested at different sliding amplitude/normal load (a) 20µm/40Ｎ Ｎ; (b) 60µm/40Ｎ Ｎ; (c) 100µm/40Ｎ Ｎ; (d) 100µm/20Ｎ Ｎ; (e) 100µm/10Ｎ Ｎ.
The wear volumes of the samples tested with different sliding
amplitude and normal loads are calculated according to the 3D profile micrograph, and the results are shown in Figure 4. The wear volume of the alloy 690 is increased from 3.1x106 µm3 to 7.2x106 µm3 with the sliding amplitude increased from 20µm to 100µm. The changing trend of the wear volume with the increase of normal load is similar to that of 8
ACCEPTED MANUSCRIPT
RI PT
sliding amplitude.
SC
Figure 4 Dependence of wear volume on (a) sliding amplitude and (b) normal load
M AN U
In the present research, the testing cycles and frequency for every specimen are the same, so the sliding distance increases linearly with the increase of sliding amplitude. With the increase of sliding distance, the specimen is subjected to more fretting wear energy (because of the longer
TE D
sliding distance and higher moving speed) and of course, the wear volume is increased with the increase of sliding amplitude. According to the Hertz’s theory [17-18], the fretting contact area is increasing with the
AC C
volume.
EP
increase of the normal load, which contributes to the increase of the wear
The Archard equation [19-21] has been used to study the fretting
wear behavior of SG tubes, and can be expressed by the following equation:
V = K F S ·······································································(1)
where V is the wear volume, K is the wear coefficient, F is the contact force, S is the sliding distance. Although the sliding distance is different 9
ACCEPTED MANUSCRIPT
in the present research, prior research [22] reported that wear coefficient of alloy 690 did not change with increasing distance, so the wear coefficient as a function of normal load and sliding amplitude is
M AN U
SC
RI PT
calculated and shown in Figure 5.
Figure 5 Dependence of wear rate on (a) sliding amplitude and (b) normal load
It is interesting to be observed that when the sliding amplitude is increased from 20µm to 60µm, the wear coefficient is reduced by about
TE D
50%, however, further increase of the sliding amplitude to 100µm has little effects on the wear coefficient, as shown in Figure 5(a). Similar
EP
phenomenon has also been observed with the normal load increased from 10N to 40N. The variation of the wear coefficient indicates that the wear
AC C
mode of the materials is changed. Prior researches [23-25] reported that amplitude was a main variable that determined the wear coefficient and mode, which is in accordance with the results of the present research. As for normal load, which determines the contact area of the specimens, and with the increase of contact area, the wear rate of the materials decrease [26]. To further analyze the wearing behavior of the materials, the 10
ACCEPTED MANUSCRIPT
morphology of the worn scars under different testing conditions is
AC C
EP
TE D
M AN U
SC
RI PT
examined, as shown in figure 6.
Figure 6 Worn surface of the alloy 690 tube tested at different sliding Ｎ; (b) 60µm/40Ｎ Ｎ; (c) 100µm/40Ｎ Ｎ; (d) amplitude/normal load (a) 20µm/40Ｎ 100µm/20Ｎ Ｎ; (e) 100µm/10Ｎ Ｎ.
It can be found that no matter what the normal force or sliding amplitude is, grooves that parallel to the slip direction are observed on the worn surface. In Figure 6(b), (c) and (d), obvious detachments are found on the worn surface, indicating delamination wear. This observation 11
ACCEPTED MANUSCRIPT
reveals that with the normal load≧20N (the sliding amplitude is 100µm) or sliding amplitude≧60µm (the normal load is 40N), the morphology of the worn surface is similar and indicates the same wear mode of
RI PT
delamination. The analysis based on Archard equation indicates that the value of wear coefficient is similar under the three testing conditions, which is in accordance with the morphology analysis that the wearing
SC
mode is delamination. In Figure 6(a) and (e), the morphology of the worn
M AN U
scars is different, and few obvious detachments are observed, however, loose debris is distributing on the grooves, which indicates that when the normal load is 10N (the sliding amplitude is 100µm) or the sliding distance is 20µm (the normal load is 40N), the wear mode becomes
TE D
abrasive wear. The wear coefficient under the two testing conditions is higher (as shown by Fig.5), and this is because that abrasive wear takes the center stage. The above discussion indicates that under different
EP
testing conditions, the wear mode of the materials is changed, and this
AC C
contributes to the evolution of the wear coefficient. Figure 7 shows the EDS analysis results of the specimen after
fretting wear tests. It can be observed that no matter what the sliding amplitude or normal load is, the oxygen content of worn area is higher than that of the unworn area. This is because that in the fretting process, the mechanical energy can be transferred to heat energy, so the temperature of the contact area is increased [27], and the oxidation of the 12
ACCEPTED MANUSCRIPT
materials becomes more severe. The oxide film formed on the contact surface of the materials could be broken under the action of sliding, and the chemical activity of the fresh surface is so high that it is easily to be
RI PT
oxidized, so the corrosion of the worn surface is accelerated and becomes more severe than that of the unworn surface. Besides oxygen content, the content of iron is also increased on the worn surface. The increment of
SC
iron can be mainly attributed to “materials transfer” effect [28-29]. Iron
M AN U
element is transferred from 405 SS to the alloy 690 tube in the fretting
AC C
EP
TE D
wear process, so the content of iron is increased on the worn surface.
13
RI PT
ACCEPTED MANUSCRIPT
SC
Figure 7 EDS analysis of the alloy 690 tube tested at different sliding Ｎ; (b) 60µm/40Ｎ Ｎ; (c) 100µm/40Ｎ Ｎ; (d) amplitude/normal load (a) 20µm/40Ｎ 100µm/20Ｎ Ｎ; (e) 100µm/10Ｎ Ｎ.
It is necessary to point out that with the increase of the sliding
M AN U
amplitude and normal force, the content of the oxygen on the worn surface is increasing, which indicates that the oxidation becomes more severe. This is mainly because that the higher the sliding amplitude and normal force, the more mechanical energy being transferred to thermal
TE D
energy, and the oxidation of the worn surface becomes more severe. 3.3 Fretting wear process and mechanism analysis
EP
The above discussion reveals that when the fretting couples are exposed to high temperature water, corrosion and mechanical wear
AC C
simultaneously affect the behavior of the materials, so in the following paragraphs, the fretting wear process of alloy 690 tube mated with 405 SS is analyzed to distinguish the factors that influence it. Before fretting wear tests, the tube and plate are put into the autoclave and heated up to 285℃, and in this process (takes 2.5 hours), corrosion of the materials happens. Corrosion on the contact surface is not as severe as other parts of the specimens because the tight coupling 14
ACCEPTED MANUSCRIPT
between the tube and plate can hinder the diffusion of water, and this could help to retard corrosion of the materials. The corrosion resistance of alloy 690 is much higher than that of 405 SS due to its higher Cr content
RI PT
[30], so the thickness of the oxide scale on the alloy 690 tube is much lower than that on 405 SS, as schematically shown in Figure 8 (a).
For the oxide scale grown on a Fe-Cr-Ni based alloy exposed to high
SC
temperature water, the oxide scale thickness (h) as a function of testing time (t) can be expressed by the following equation [31-32]:
M AN U
h = kt n ·············································································(2)
where the exponential index n represents the degree of oxidation rate, k(m/sn) is the chemical reaction constant that follows Arrhenius equation
k = A exp( −
TE D
as:
Q ) ·································································(3) RT
EP
where A is a constant, Q is the activation energy, R is gas constant, T is the temperature (although in the heating up process, the temperature is
AC C
changing, in the present research, it is treated as constant for the simplification of analysis). For the contact surface between the tube and plate, the corrosion is
retarded, and a parameter, α, is introduced to reflect this phenomenon. So the thickness of the oxide film on the contact area can be expressed by the following equation:
15
ACCEPTED MANUSCRIPT hc = A exp( −
Q n )t α ·····························································(4) RT
where the value of α is lower than 1. If no fretting is applied, an indentation can be formed on the contact
RI PT
area due to the different corrosion rate between the contact area and the other part of the tube surface, as schematically shown in Figure 8 (b).
SC
Suppose the width of the rectangular contact area is b and the length of
the following equation: Vc = hlb − hc lb = A exp( −
M AN U
the tube is l, then the wear volume of the indentation can be calculated by
Q n )t (1 − α )lb ···········································(5) RT
When fretting begins, corrosion and mechanical wear happen simultaneously in this process. Suppose each fretting cycle takes time of
TE D
t0, then in this process, the thickness of the oxide film on the contact area increases as much as
Q n )t0 α ····························································(6) RT
EP
h0 = A exp( −
AC C
The oxide scale is moved under repeated sliding, and the surface depassivation rate of the materials depends on the fretting wear parameters, such as the normal load and sliding amplitude. A parameter, η (values between 0 and 1), is introduced to reflect the surface depassivation rate for one cycle, and then the mechanical wear depth, h0f, can be written as following: h0 f = A exp(−
Q n )t0 αη ··························································· (7) RT 16
ACCEPTED MANUSCRIPT
After N cycles of fretting wear test, the materials volume loss due to mechanical wear can be calculated by the following equation: V f = h0 f Nlb = A exp(−
Q n )t0 αη Nlb ··············································· (8) RT
RI PT
In the present research, the measured wear volume is a combination of Vc and Vf, so the total wear volume of the materials can be expressed by the following equation:
Q n Q n )t (1 − α )lb + A exp( − )t0 αη Nlb ················(9) RT RT
SC
V = Vc + V f = A exp( −
M AN U
If the oxide scale on the contact surface is not broken in the fretting wear process, as shown in Figure 8(c) and (e), equation (9) can be used to calculate the wear volume of the alloy 690 tube. However, the worn scar surface analysis indicates that delamination are observed on the fretting
TE D
worn surface, so the passive film is broken in the fretting wear process. Former researches [33-35] also reported that the passive films can be
EP
broken due to the mechanical stresses brought about by sliding. If the oxide scale is broken when the sliding cycle is N1, the value of η changes
AC C
to 1 and the metal matrix is worn directly, so an additional part, Vm, can be added to equation (9) to reflect the wear volume of the matrix alloy, as shown in Figure 8 (d), (f) and (g). Then equation (9) can be changed as: V = Vc + V f + Vm = A exp( − hm ( N − N1 )lb
Q n Q n )t (1 − α )lb + A exp( − )t0 α Nlb + RT RT ········ (10)
where hm is the mechanical wear depth of the matrix alloy for one slide. The fresh surface of the metal alloys exposed by such mechanical 17
ACCEPTED MANUSCRIPT
factors owns high chemical activity and can be easily oxidized. This can reduce the value of Vc because the corrosion on the contact surface is accelerated and the oxide film grows quicker than other parts of the
RI PT
specimen. The corrosion of the materials can also influence the wear of the materials in the following ways: (1) the oxide film is harder than the matrix, and the friction coefficient of the oxide film is smaller [34], so
SC
formation of oxide scale reduces wear; (2) the oxide scale can be flaked
M AN U
because of fatigue of the materials under fretting, and this accelerates wear; (3) the EDS analysis (Figure 7 ) reveals that the oxygen content on the worn scar is much higher than that of other parts on the surface of the alloy 690 tube. This phenomenon could be attributed to the fact that the
TE D
“glaze layer” is formed on the contact surface, as schematically shown in Figure 8(h). At high testing temperature (above 200℃), the broken oxide scale can be mixed, grinded and sintered into a compact layer, which is
EP
fully oxidized in this process and the oxygen content is high [36-37]. The
AC C
“glaze layer” is very protective, so the wear rate of the materials can be reduced.
However, it is necessary to point out that the above discussion is
only a qualitative analysis to help to distinguish the mechanical and corrosive effects on the fretting wear behavior of the materials. Further researches are necessary to quantitate the parameters by experiment or modelling to reveal the synergic effects of corrosion and mechanical 18
ACCEPTED MANUSCRIPT
M AN U
SC
RI PT
wear.
4 Conclusion
TE D
Figure 8 Schematic graph of the fretting wear process for alloy 690 tube exposed to high temperature water
In the present research, the effects of sliding amplitude and normal
EP
load on the fretting wear behavior of alloy 690 tube exposed to high
AC C
temperature water is studied, and the results can be summarized as following:
(1) With the increase of sliding amplitude and normal load, the wear volume of the alloy 690 tube is increasing, however, the wear coefficient of the materials firstly decreases then becomes relatively constant. (2) Sliding amplitude and normal load determines the wear mode of the materials. When the sliding amplitude or normal load is low, the wear 19
ACCEPTED MANUSCRIPT
mode is abrasive wear, however, delamination wear dominates when the sliding amplitude or normal load increases. (3) The fretting wear process is analyzed and the synergic effect of
RI PT
corrosion and mechanical wear on fretting wear behavior of the alloy 690 tube is illustrated. Mechanical wear results in the broken of the oxide scale formed on the contact surface and this promotes further corrosion of
5 Acknowledgement
M AN U
SC
the materials.
This work is financially supported by “National Basic Research Program of China” (No. 2007CB209800), and we also would like to
TE D
acknowledge the help from Instrument Analysis Center of SJTU for the electron microscope analysis. References
EP
[1] Marcos A. Bergant, Alejandro A. Yawny, Juan E. Perez Ipiña.
AC C
J-resistance curves for Inconel 690 and Incoloy 800 nuclear steam generators tubes at room temperature and at 300℃, J. Nucl. Mater. 2017; 486; 298-307.
[2] Hee-Sang Shim, Myung Sik Choi, Deok Hyun Lee, Do Haeng Hur. A prediction method for the general corrosion behavior of Alloy 690 steam generator tube using eddy current testing, Nucl. Eng. Des.2016; 297; 26-31. 20
ACCEPTED MANUSCRIPT
[3] Jae Yong Yun, Gyeong Su Shin, Dae Il Kim, Ho Sik Lee, Woong Soon Kang, Seon Jin Kim. Effect of carbide size and spacing on the fretting wear behavior of Inconel 690 SG tube mated with SUS 409, Wear 2015;
RI PT
338-339; 252-257. [4] Z.H. Wang, Y.H. Lu, J. Li, T. Shoji. Effect of pH value on the fretting wear behavior of Inconel 690 alloy, Tribol. Int. 2016; 95; 162-169.
SC
[5] J.D. Kwon, H.K. Jeung, I.S.Chung, D.H.Yoon, D.K.Park, A study on
Int. 2011; 44; 1483-1487.
M AN U
fretting fatigue characteristics of Inconel 690 at high temperature, Tribol.
[6] F. M. Guérout and N. J. Fisher, Steam Generator Fretting-Wear Damage: A Summary of Recent Findings, J. Press. Vess. Tech. 1999, 121;
TE D
304-310.
[7] Min-Kyu Lim, Se-Doo Oh, Young-Ze Lee. Friction and wear of Inconel 690 and Inconel 600 for steam generator tube in room
EP
temperature water, Nucl. Eng. Des. 2003; 226; 97-105.
AC C
[8] X. Mi, Z.B. Cai, X.M. Xiong, H. Qian, L.C. Tang, Y.C. Xie, J.F. Peng, Min-hao Zhu. Investigation on fretting wear behavior of 690 alloy in water under various temperatures, Tribol. Int. 100 (2016) 400-409. [9] L. Xin, B.B. Yang, Z.H. Wang, J.Li, Y.H. Lu, T. Shoji. Effect of normal force on fretting wear behavior and mechanism of Alloy 690TT in high temperature water, Wear 2016; 368-369; 210-218. [10] Choon Yeol Lee, Li Si Tian, Joon Woo Bae, Young Suck Chai. 21
ACCEPTED MANUSCRIPT
Application of influence function method on the fretting wear of tube-to-plate contact, Tribol. Int. 2009; 42; 951-957. [11] Jae-do Kwon, Dae-kyu Park, Seung-wan Woo, Dong-hwan Yoon,
RI PT
Ilsup Chung. A study on fretting fatigue life for the Inconel alloy 600 at high temperature, Nucl. Eng. Des. 2010; 240; 2521-2527.
[12] Jae Yong Yun, Ho Sik Lee, Do Haeng Hur, Woong Soon Kang,
SC
Cheol Hyun Bae, Seon Jin Kim. Effect of oxidation film on the fretting
Wear 2016; 368-369; 344-349.
M AN U
wear behavior of Alloy 690 steam generator tube mated with SUS 409,
[13] J. Li, B.B. Yang, Y.H. Lu, L. Xin, Z.H. Wang, T. Shoji. The effects of electrochemical polarization condition and applied potential on
TE D
tribocorrosion behaviors of Inconel 690 alloys in water environment, Mater. Des. 2017; 119; 93-103.
[14] T. Dick, C. Paulin, G. Cailletaud, S. Fouvry, Experimental and
EP
numerical analysis of local and global plastic behaviour in fretting wear.
AC C
Tribol. Int. 2006; 39; 1036-1044. [15] C. Paulin, S. Fouvry, C. Meunier, Finite element modelling of fretting wear surface evolution: Application to a Ti-6A1-4V contact. Wear 2008; 264; 26-36.
[16] B. van Peteghem, S. Fouvry, J. Petit, Effect of variable normal force and frequency on fretting wear response of Ti-6Al-4V contact. Wear 2011; 271; 1535-1542. 22
ACCEPTED MANUSCRIPT
[17] R. D. Mindlin. Compliance of elastic bodies in contact. ASME j. appl. Mech. 1949; 16; 259-268. [18] Zhongrong Zhou, Léo Vincent. Fretting wear, Science Press, Beijing,
RI PT
2002. [19] J. Jiang, F.H. Stott, M.M. Stack. Some frictional features associated
with the sliding wear of the nickel-base alloy N80A at temperatures to
SC
250℃, Wear 1994; 176; 185-194.
M AN U
[20] Xiangzhen Xue, Sanmin Wang, Bo Li. Modification methodology of fretting wear in involute spline, Wear 2016; 368-369; 435-444. [21] Arnab Ghosh, Ben Leonard, Farshid Sadeghi. A stress based damage mechanics model to simulate fretting wear of Hertzian line contact in
TE D
partial slip, Wear 2013; 307; 87-99.
[22] Jae Yong Yun, Myung Chul Park, Gyeong Su Shin, Ji Haeng Heo, Dae Il Kim, Seon Jin Kim. Effects of amplitude and frequency on the
EP
wear mode change of Inconel 690 SG tube mated with SUS 409, Wear
AC C
2014; 313; 83-88.
[23] I. Chung, M. Lee. An experimental study on fretting wear behavior of crosscontacting Inconel 690 tubes, Nucl. Eng. Des.2011; 241 (10); 4103-4110.
[24] T.-H. Kim, S.-S. Kim, Fretting wear mechanisms of Zircaloy-4 and Inconel 600 contact in air, KSME Int. J. 2011; 15 (9); 1274-1280. [25] Y.-H. Lee, I.-S. Kim, S.-S. Kang, H.-D. Chung, A study on wear 23
ACCEPTED MANUSCRIPT
coefficients and mechanisms of steam generator tube materials, Wear 2011; 250 (1–12); 718-725. [26] G.S. Park, G.G. Kim, S.J. Kim, Sliding wear behaviors of steam
RI PT
generator tube materials in high temperature water environment, J. Nucl. Mater. 2006; 352 (1–3); 80-84.
[27] X. Jin, P.H. Shipway, W. Sun. The role of frictional power
SC
dissipation (as a function of frequency) and test temperature on contact
M AN U
temperature and the subsequent wear behavior in a stainless steel contact in fretting, Wear 2015; 330-331; 103-111.
[28] Xiaoling Zhang, R.G.Vitchev, W.Lauwerens, Jiawen He, J.-P.Celis. Transfer of molybdenum sulphide coating material onto corundum balls
TE D
in fretting wear tests, Thin Solid Films 2004; 446; 78-84. [29] Volker Franco Steier, Remy Kalombo Badibanga, Cosme Roberto Moreira da Silva, Murilo Magalhães Nogueira, José Alexander Araújo.
EP
Effect of chromium nitride coatings and cryogenic treatments on wear
AC C
and fretting fatigue resistance of aluminum, Electr. Pow. Syst. Res. 2014; 116; 322-329.
[30] S. Fyfitch. Corrosion and Stress Corrosion Cracking of Ni-Base Alloys, Comp. Nucl. Mater. 2012; 5; 69-92. [31] R.W. Revie, H.H. Uhlig, Corrosion and Corrosion Control, fourth ed., A John Wiley Sons, New Jersey 2008; 220-223.
24
ACCEPTED MANUSCRIPT
[32] Branko N. Popov, Chapter 11-High-Temperature Corrosion, Corros. Eng. 2015; 481-523 [33] A. Iwabuchi, T. Tsukamoto, Y. Tatsuyanagi, Electrochemical
RI PT
approach to corrosive wear of SKD61 die steel in Na2SO4solution, Wear 1992; 156; 301-313.
[34] Y. Sato, A. Iwabuchi, M. Uchidate, H. Yashiro. Dynamic corrosion
SC
properties of impact–fretting wear in high-temperature pure water, Wear
M AN U
2015; 330-331; 182-192.
[35] J. Pellier, J. Geringer, B. Forest. Fretting-corrosion between 316L SS and PMMA: Influence of ionic strength, protein and electrochemical conditions on material wear. Application to orthopaedic implants, Wear
TE D
2011; 271; 1563-1571.
[36] J. Hardell, S. Hernandez, et al., Effect of oxide layers and near
EP
surface transformations on friction and wear during tool steel and boron steel interaction at high temperatures, Wear 2015; 330-331; 223-229.
AC C
[37] F.H. Scott. High temperature sliding wear of metals. Tribol. Int. 2002; 35; 489-495.
25
ACCEPTED MANUSCRIPT
Highlights 1.A fretting wear test setup with a tube-on-plate configuration was developed to examine the performance of the materials in 285℃ water.
RI PT
2.With the increase of sliding amplitude and normal load, the wear volume of the alloy 690 tube is increasing.
SC
3.Sliding amplitude and normal load determines the wear mode of the materials。
M AN U
4.The synergic effect model of corrosion and mechanical wear on fretting
AC C
EP
TE D
wear behavior of the alloy 690 tube is proposed.
S0301-679X(17)30335-3
DOI:
10.1016/j.triboint.2017.07.001
Reference:
JTRI 4806
To appear in:
Tribology International
Received Date: 3 May 2017 Revised Date:
30 June 2017
Accepted Date: 2 July 2017
Please cite this article as: Guo X, Lai P, Tang L, Wang J, Zhang L, Effects of sliding amplitude and normal load on the fretting wear behavior of alloy 690 tube exposed to high temperature water, Tribology International (2017), doi: 10.1016/j.triboint.2017.07.001. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Effects of sliding amplitude and normal load on the fretting wear behavior of alloy 690 tube exposed to high temperature water Xianglong Guo1, Ping Lai1, Lichen Tang2, Jiamei Wang1, Lefu Zhang1* 1
School of Nuclear Science and Engineering, Shanghai Jiao Tong University, NO. 800
2
RI PT
Dongchuan Road, Shanghai 200240, China
Shanghai Nuclear Engineering Research & Design Institute, No. 29 Hongcao Road, Shanghai 200233, China
SC
Contacting email: [email protected]
M AN U
Abstract： ：The effects of sliding amplitude and normal load on the fretting wear behavior of alloy 690 tube exposed to an environment simulating the secondary side water chemistry of nuclear power plant are studied and the fretting wear process is discussed. The results indicate that with the
TE D
increase of sliding amplitude and normal load, the wear volume of the alloy 690 tube is increasing, however, the wear coefficient firstly
EP
decreases then becomes relatively constant. This is because that the wear mode of the materials is changed from abrasive wear to delamination
AC C
wear with the increase of sliding amplitude and normal load. The synergic effects of corrosion and mechanical wear on fretting wear behavior of the alloy 690 tube are also discussed and a model is proposed to qualitatively reveal the fretting wear process of the alloy 690 tube. Key words: Alloy 690; Fretting wear; Sliding amplitude; Synergic effect 1 Introduction The nuclear power plant (NPP) steam generator (SG) tubes, mostly 1
ACCEPTED MANUSCRIPT
made of Alloy 690 [1-2], experience flow induced vibration in the service life, and this results in the fretting wear between the tubes and the supporting parts due to the contact [3-4]. The damage brought about by
RI PT
the fretting wear can lead to the reduction of the service life of the thin SG tubes [5], so the safety and reliability of the SG in the NPP is highly dependent on the fretting wear characteristics.
SC
Lots of researches have been carried out to study the fretting wear
M AN U
behavior of the Alloy 690 mated with different kinds of materials. Guérout and Fisher [6] summarized the results of former researches (carried out before 1999) on the fretting wear of SG tubes and found that temperature had a significant effect on tube wear damage. Lim and Lee [7]
TE D
studied the fretting wear behavior of Alloy 690 in room temperature and concluded that the normal load and the vibrating amplitude were key parameters that could influence the friction in fretting. Mi and Zhu [8]
EP
investigated the influence of temperature on the fretting wear of Alloy
AC C
690 mated with 405 SS in the air and found that the wear volume increased from room temperature to 90℃ and decreased from 90℃ to 285 ℃ . Xin and Lu [9] revealed the microstructure evolution of subsurface on the Alloy 690TT alloy subjected to dry fretting wear at 285℃ in air and the experimental results indicated that five layers existed in the subsurface. Besides experimental studies, some simulation works have also been carried out to study the fretting wear behavior of Alloy 690, for 2
ACCEPTED MANUSCRIPT
example, Lee and Chai [10] employed the influence function method to investigate the fretting wear problems on the secondary side of Alloy 690 SG tubes and the fretting wear of tube-to-plate contact was successfully
RI PT
simulated. Those researches have contributed a lot to reveal the mechanisms in the fretting wear of Alloy 690 and help the NPP designers to increase the safety and reliability.
SC
However, up until now, the above researches and lots of other
M AN U
studies [11-13] were carried out either in the air or in the water at a temperature much lower than 285℃. Although those researches can provide useful information to speculate the fretting wear behavior of the materials
under
service
conditions,
experiments
under
service
TE D
environments is still of great necessity to evaluate the actual performance of the materials. According to the researches carried out by Fouvry [14-16], various parameters, such as normal load, sliding amplitude and
EP
frequency, can influence the fretting wear behavior of the materials. As
AC C
for SG tube, the effects of those parameters on its fretting wear behavior also need to be studied to evaluate its servicing life. The present study aims at revealing the effects of sliding amplitude
and normal load on the fretting wear behavior of Alloy 690 mated with 405 stainless steel (SS) in the service environment. The damage mechanisms in this condition are discussed based on the chemical and microstructure analysis. A model considering the mechanical and 3
ACCEPTED MANUSCRIPT
corrosion factors is proposed to qualitatively reveal the fretting wear process of alloy 690 tube exposed to high temperature water. 2 Experimental procedure
RI PT
In the present research, the Alloy 690 tubes with an inner diameter of 15.5mm, external diameter of 17.5mm and length of 16mm were used, and the supporting part was made of 405 SS with length of 12mm, width
SC
of 7.6mm and height of 3mm. The roughness (Ra) of the Alloy 690 and
M AN U
405 SS were examined as 0.254µm and 0.376µm, respectively. The composition of Alloy 690 and 405 SS was shown in Table 1. Table 1. Composition of alloy 690 and 405 SS in wt.% Ni
Cr
Alloy 690
Bal.
30.3
405 SS
0.50
12.5
Fe
Al
C
Si
Mn
S
P
9.6
0.25
0.023
0.30
0.23
0.002
0.008
Bal.
0.15
0.056
0.60
0.58
0.013
0.025
TE D
Chemical composition
Table 2 Testing conditions for the fretting wear behavior of alloy 690 tube mated with 405 SS Value (unit)
Temperature Pressure pH Dissolved oxygen (DO) Frequency Cycles
285℃ 8.6MPa 9.75 ＜5 µg/L 5Hz 500000 20µm 60µm 100µm 10N 20N 40N
AC C
EP
Parameter
Sliding amplitude (The normal load is kept constant, and the value is 40N)
Normal force (The sliding distance is kept constant, and the value is 100µm)
A fretting wear test setup with a tube-on-plate configuration was developed to examine the performance of the materials, as schematically 4
ACCEPTED MANUSCRIPT
shown in Figure 1(a). The oscillatory motion, with amplitude from 5µm to 200µm, was controlled by the electric motor. The normal load between the SG tubes and the plates was imposed by springs. A heating and water
RI PT
supply loop system was applied to the testing setup to control the water chemistry in the autoclave, as shown in Figure 1(b). The testing parameters used in present research were listed in Table 2. The solution in
SC
the autoclave was distilled water containing certain amount of ammonia
M AN U
to control the pH as 9.75. The temperature (285℃) was similar to the secondary side of NPP, and the pressure is controlled as 8.6 MPa (higher than the saturated vapor pressure at 285℃) to make sure the water environment in the autoclave. Argon was bubbled into the autoclave to
TE D
control the dissolved oxygen (DO) in the solution. The load and amplitude acting on the specimen was set according to the data estimated
AC C
EP
by the NPP.
5
ACCEPTED MANUSCRIPT Figure 1 Schematic picture of (a) configuration between alloy 690 and 405 SS and (b) the fretting wear test setup
The phase and microstructure of the alloys were examined with X-ray diffraction (XRD), optical microscope (OM) and electron
RI PT
backscattered diffraction (EBSD) respectively. Conventional mechanical polishing followed by vibration polishing were used to prepare the
SC
samples for scanning electron microscopy (SEM) and EBSD analysis. For metallographic observation, alloy 690 was etched in a solution of 4%
etched in aqua regia for 15s.
M AN U
hydrofluoric acid, 4% nitric acid and 92% water for 45s. 405 SS was
The morphology of the worn scar was examined with a 3D optical microscope (Bruker Contour GT-I). The volume of the worn scar was
TE D
calculated by the software equipped with the microscope, and the principle of calculating was depth integration across the area. SEM and
EP
energy dispersive spectroscopy (EDS) analysis were also carried out to study the worn scar morphology and composition after fretting wear test.
AC C
3 Results and discussions
3.1 Microstructure of the materials Figure 2 shows the XRD diffraction pattern and microstructure of
alloy 690 and 405 SS. It can be found from the XRD analysis that both alloy 690 and 405 SS have single phase microstructures. In 405 SS, only ferrite is detected, however in alloy 690, only γ-Ni is detected. In order to systematically reveal the status of the materials, the microstructure of the 6
ACCEPTED MANUSCRIPT
alloy 690 and 405 SS is studied by OM and EBSD, as shown in Figure 2(b) to (d). The grain size of Alloy 690 alloy is about 63µm, and the EBSD analysis indicates that a typical austenitic microstructure exists in
RI PT
the tested alloy 690, and twin crystals are also widely distributed in the
TE D
M AN U
SC
materials. For 405 SS, the average grain size is about 52µm.
EP
Figure 2 (a) Phase content of alloy 690 and 405 SS, (b) and (c) microstructure of alloy 690, (d) microstructure of 405 SS
3.2 Wear volume and morphology analysis
AC C
The 3D-profile micrographs of the fretting tube scars tested with
different sliding amplitude and normal load are shown in Figure 3. It can be found that no matter what the sliding amplitude or normal load is, the signs of worn scars are obvious. With the increase of sliding amplitude and normal load, the size and depth of the worn scars is increasing, which indicates that the wear volume of the materials would be increased.
7
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
Figure 3 The 3D-profile micrographs of the fretting tube scars tested at different sliding amplitude/normal load (a) 20µm/40Ｎ Ｎ; (b) 60µm/40Ｎ Ｎ; (c) 100µm/40Ｎ Ｎ; (d) 100µm/20Ｎ Ｎ; (e) 100µm/10Ｎ Ｎ.
The wear volumes of the samples tested with different sliding
amplitude and normal loads are calculated according to the 3D profile micrograph, and the results are shown in Figure 4. The wear volume of the alloy 690 is increased from 3.1x106 µm3 to 7.2x106 µm3 with the sliding amplitude increased from 20µm to 100µm. The changing trend of the wear volume with the increase of normal load is similar to that of 8
ACCEPTED MANUSCRIPT
RI PT
sliding amplitude.
SC
Figure 4 Dependence of wear volume on (a) sliding amplitude and (b) normal load
M AN U
In the present research, the testing cycles and frequency for every specimen are the same, so the sliding distance increases linearly with the increase of sliding amplitude. With the increase of sliding distance, the specimen is subjected to more fretting wear energy (because of the longer
TE D
sliding distance and higher moving speed) and of course, the wear volume is increased with the increase of sliding amplitude. According to the Hertz’s theory [17-18], the fretting contact area is increasing with the
AC C
volume.
EP
increase of the normal load, which contributes to the increase of the wear
The Archard equation [19-21] has been used to study the fretting
wear behavior of SG tubes, and can be expressed by the following equation:
V = K F S ·······································································(1)
where V is the wear volume, K is the wear coefficient, F is the contact force, S is the sliding distance. Although the sliding distance is different 9
ACCEPTED MANUSCRIPT
in the present research, prior research [22] reported that wear coefficient of alloy 690 did not change with increasing distance, so the wear coefficient as a function of normal load and sliding amplitude is
M AN U
SC
RI PT
calculated and shown in Figure 5.
Figure 5 Dependence of wear rate on (a) sliding amplitude and (b) normal load
It is interesting to be observed that when the sliding amplitude is increased from 20µm to 60µm, the wear coefficient is reduced by about
TE D
50%, however, further increase of the sliding amplitude to 100µm has little effects on the wear coefficient, as shown in Figure 5(a). Similar
EP
phenomenon has also been observed with the normal load increased from 10N to 40N. The variation of the wear coefficient indicates that the wear
AC C
mode of the materials is changed. Prior researches [23-25] reported that amplitude was a main variable that determined the wear coefficient and mode, which is in accordance with the results of the present research. As for normal load, which determines the contact area of the specimens, and with the increase of contact area, the wear rate of the materials decrease [26]. To further analyze the wearing behavior of the materials, the 10
ACCEPTED MANUSCRIPT
morphology of the worn scars under different testing conditions is
AC C
EP
TE D
M AN U
SC
RI PT
examined, as shown in figure 6.
Figure 6 Worn surface of the alloy 690 tube tested at different sliding Ｎ; (b) 60µm/40Ｎ Ｎ; (c) 100µm/40Ｎ Ｎ; (d) amplitude/normal load (a) 20µm/40Ｎ 100µm/20Ｎ Ｎ; (e) 100µm/10Ｎ Ｎ.
It can be found that no matter what the normal force or sliding amplitude is, grooves that parallel to the slip direction are observed on the worn surface. In Figure 6(b), (c) and (d), obvious detachments are found on the worn surface, indicating delamination wear. This observation 11
ACCEPTED MANUSCRIPT
reveals that with the normal load≧20N (the sliding amplitude is 100µm) or sliding amplitude≧60µm (the normal load is 40N), the morphology of the worn surface is similar and indicates the same wear mode of
RI PT
delamination. The analysis based on Archard equation indicates that the value of wear coefficient is similar under the three testing conditions, which is in accordance with the morphology analysis that the wearing
SC
mode is delamination. In Figure 6(a) and (e), the morphology of the worn
M AN U
scars is different, and few obvious detachments are observed, however, loose debris is distributing on the grooves, which indicates that when the normal load is 10N (the sliding amplitude is 100µm) or the sliding distance is 20µm (the normal load is 40N), the wear mode becomes
TE D
abrasive wear. The wear coefficient under the two testing conditions is higher (as shown by Fig.5), and this is because that abrasive wear takes the center stage. The above discussion indicates that under different
EP
testing conditions, the wear mode of the materials is changed, and this
AC C
contributes to the evolution of the wear coefficient. Figure 7 shows the EDS analysis results of the specimen after
fretting wear tests. It can be observed that no matter what the sliding amplitude or normal load is, the oxygen content of worn area is higher than that of the unworn area. This is because that in the fretting process, the mechanical energy can be transferred to heat energy, so the temperature of the contact area is increased [27], and the oxidation of the 12
ACCEPTED MANUSCRIPT
materials becomes more severe. The oxide film formed on the contact surface of the materials could be broken under the action of sliding, and the chemical activity of the fresh surface is so high that it is easily to be
RI PT
oxidized, so the corrosion of the worn surface is accelerated and becomes more severe than that of the unworn surface. Besides oxygen content, the content of iron is also increased on the worn surface. The increment of
SC
iron can be mainly attributed to “materials transfer” effect [28-29]. Iron
M AN U
element is transferred from 405 SS to the alloy 690 tube in the fretting
AC C
EP
TE D
wear process, so the content of iron is increased on the worn surface.
13
RI PT
ACCEPTED MANUSCRIPT
SC
Figure 7 EDS analysis of the alloy 690 tube tested at different sliding Ｎ; (b) 60µm/40Ｎ Ｎ; (c) 100µm/40Ｎ Ｎ; (d) amplitude/normal load (a) 20µm/40Ｎ 100µm/20Ｎ Ｎ; (e) 100µm/10Ｎ Ｎ.
It is necessary to point out that with the increase of the sliding
M AN U
amplitude and normal force, the content of the oxygen on the worn surface is increasing, which indicates that the oxidation becomes more severe. This is mainly because that the higher the sliding amplitude and normal force, the more mechanical energy being transferred to thermal
TE D
energy, and the oxidation of the worn surface becomes more severe. 3.3 Fretting wear process and mechanism analysis
EP
The above discussion reveals that when the fretting couples are exposed to high temperature water, corrosion and mechanical wear
AC C
simultaneously affect the behavior of the materials, so in the following paragraphs, the fretting wear process of alloy 690 tube mated with 405 SS is analyzed to distinguish the factors that influence it. Before fretting wear tests, the tube and plate are put into the autoclave and heated up to 285℃, and in this process (takes 2.5 hours), corrosion of the materials happens. Corrosion on the contact surface is not as severe as other parts of the specimens because the tight coupling 14
ACCEPTED MANUSCRIPT
between the tube and plate can hinder the diffusion of water, and this could help to retard corrosion of the materials. The corrosion resistance of alloy 690 is much higher than that of 405 SS due to its higher Cr content
RI PT
[30], so the thickness of the oxide scale on the alloy 690 tube is much lower than that on 405 SS, as schematically shown in Figure 8 (a).
For the oxide scale grown on a Fe-Cr-Ni based alloy exposed to high
SC
temperature water, the oxide scale thickness (h) as a function of testing time (t) can be expressed by the following equation [31-32]:
M AN U
h = kt n ·············································································(2)
where the exponential index n represents the degree of oxidation rate, k(m/sn) is the chemical reaction constant that follows Arrhenius equation
k = A exp( −
TE D
as:
Q ) ·································································(3) RT
EP
where A is a constant, Q is the activation energy, R is gas constant, T is the temperature (although in the heating up process, the temperature is
AC C
changing, in the present research, it is treated as constant for the simplification of analysis). For the contact surface between the tube and plate, the corrosion is
retarded, and a parameter, α, is introduced to reflect this phenomenon. So the thickness of the oxide film on the contact area can be expressed by the following equation:
15
ACCEPTED MANUSCRIPT hc = A exp( −
Q n )t α ·····························································(4) RT
where the value of α is lower than 1. If no fretting is applied, an indentation can be formed on the contact
RI PT
area due to the different corrosion rate between the contact area and the other part of the tube surface, as schematically shown in Figure 8 (b).
SC
Suppose the width of the rectangular contact area is b and the length of
the following equation: Vc = hlb − hc lb = A exp( −
M AN U
the tube is l, then the wear volume of the indentation can be calculated by
Q n )t (1 − α )lb ···········································(5) RT
When fretting begins, corrosion and mechanical wear happen simultaneously in this process. Suppose each fretting cycle takes time of
TE D
t0, then in this process, the thickness of the oxide film on the contact area increases as much as
Q n )t0 α ····························································(6) RT
EP
h0 = A exp( −
AC C
The oxide scale is moved under repeated sliding, and the surface depassivation rate of the materials depends on the fretting wear parameters, such as the normal load and sliding amplitude. A parameter, η (values between 0 and 1), is introduced to reflect the surface depassivation rate for one cycle, and then the mechanical wear depth, h0f, can be written as following: h0 f = A exp(−
Q n )t0 αη ··························································· (7) RT 16
ACCEPTED MANUSCRIPT
After N cycles of fretting wear test, the materials volume loss due to mechanical wear can be calculated by the following equation: V f = h0 f Nlb = A exp(−
Q n )t0 αη Nlb ··············································· (8) RT
RI PT
In the present research, the measured wear volume is a combination of Vc and Vf, so the total wear volume of the materials can be expressed by the following equation:
Q n Q n )t (1 − α )lb + A exp( − )t0 αη Nlb ················(9) RT RT
SC
V = Vc + V f = A exp( −
M AN U
If the oxide scale on the contact surface is not broken in the fretting wear process, as shown in Figure 8(c) and (e), equation (9) can be used to calculate the wear volume of the alloy 690 tube. However, the worn scar surface analysis indicates that delamination are observed on the fretting
TE D
worn surface, so the passive film is broken in the fretting wear process. Former researches [33-35] also reported that the passive films can be
EP
broken due to the mechanical stresses brought about by sliding. If the oxide scale is broken when the sliding cycle is N1, the value of η changes
AC C
to 1 and the metal matrix is worn directly, so an additional part, Vm, can be added to equation (9) to reflect the wear volume of the matrix alloy, as shown in Figure 8 (d), (f) and (g). Then equation (9) can be changed as: V = Vc + V f + Vm = A exp( − hm ( N − N1 )lb
Q n Q n )t (1 − α )lb + A exp( − )t0 α Nlb + RT RT ········ (10)
where hm is the mechanical wear depth of the matrix alloy for one slide. The fresh surface of the metal alloys exposed by such mechanical 17
ACCEPTED MANUSCRIPT
factors owns high chemical activity and can be easily oxidized. This can reduce the value of Vc because the corrosion on the contact surface is accelerated and the oxide film grows quicker than other parts of the
RI PT
specimen. The corrosion of the materials can also influence the wear of the materials in the following ways: (1) the oxide film is harder than the matrix, and the friction coefficient of the oxide film is smaller [34], so
SC
formation of oxide scale reduces wear; (2) the oxide scale can be flaked
M AN U
because of fatigue of the materials under fretting, and this accelerates wear; (3) the EDS analysis (Figure 7 ) reveals that the oxygen content on the worn scar is much higher than that of other parts on the surface of the alloy 690 tube. This phenomenon could be attributed to the fact that the
TE D
“glaze layer” is formed on the contact surface, as schematically shown in Figure 8(h). At high testing temperature (above 200℃), the broken oxide scale can be mixed, grinded and sintered into a compact layer, which is
EP
fully oxidized in this process and the oxygen content is high [36-37]. The
AC C
“glaze layer” is very protective, so the wear rate of the materials can be reduced.
However, it is necessary to point out that the above discussion is
only a qualitative analysis to help to distinguish the mechanical and corrosive effects on the fretting wear behavior of the materials. Further researches are necessary to quantitate the parameters by experiment or modelling to reveal the synergic effects of corrosion and mechanical 18
ACCEPTED MANUSCRIPT
M AN U
SC
RI PT
wear.
4 Conclusion
TE D
Figure 8 Schematic graph of the fretting wear process for alloy 690 tube exposed to high temperature water
In the present research, the effects of sliding amplitude and normal
EP
load on the fretting wear behavior of alloy 690 tube exposed to high
AC C
temperature water is studied, and the results can be summarized as following:
(1) With the increase of sliding amplitude and normal load, the wear volume of the alloy 690 tube is increasing, however, the wear coefficient of the materials firstly decreases then becomes relatively constant. (2) Sliding amplitude and normal load determines the wear mode of the materials. When the sliding amplitude or normal load is low, the wear 19
ACCEPTED MANUSCRIPT
mode is abrasive wear, however, delamination wear dominates when the sliding amplitude or normal load increases. (3) The fretting wear process is analyzed and the synergic effect of
RI PT
corrosion and mechanical wear on fretting wear behavior of the alloy 690 tube is illustrated. Mechanical wear results in the broken of the oxide scale formed on the contact surface and this promotes further corrosion of
5 Acknowledgement
M AN U
SC
the materials.
This work is financially supported by “National Basic Research Program of China” (No. 2007CB209800), and we also would like to
TE D
acknowledge the help from Instrument Analysis Center of SJTU for the electron microscope analysis. References
EP
[1] Marcos A. Bergant, Alejandro A. Yawny, Juan E. Perez Ipiña.
AC C
J-resistance curves for Inconel 690 and Incoloy 800 nuclear steam generators tubes at room temperature and at 300℃, J. Nucl. Mater. 2017; 486; 298-307.
[2] Hee-Sang Shim, Myung Sik Choi, Deok Hyun Lee, Do Haeng Hur. A prediction method for the general corrosion behavior of Alloy 690 steam generator tube using eddy current testing, Nucl. Eng. Des.2016; 297; 26-31. 20
ACCEPTED MANUSCRIPT
[3] Jae Yong Yun, Gyeong Su Shin, Dae Il Kim, Ho Sik Lee, Woong Soon Kang, Seon Jin Kim. Effect of carbide size and spacing on the fretting wear behavior of Inconel 690 SG tube mated with SUS 409, Wear 2015;
RI PT
338-339; 252-257. [4] Z.H. Wang, Y.H. Lu, J. Li, T. Shoji. Effect of pH value on the fretting wear behavior of Inconel 690 alloy, Tribol. Int. 2016; 95; 162-169.
SC
[5] J.D. Kwon, H.K. Jeung, I.S.Chung, D.H.Yoon, D.K.Park, A study on
Int. 2011; 44; 1483-1487.
M AN U
fretting fatigue characteristics of Inconel 690 at high temperature, Tribol.
[6] F. M. Guérout and N. J. Fisher, Steam Generator Fretting-Wear Damage: A Summary of Recent Findings, J. Press. Vess. Tech. 1999, 121;
TE D
304-310.
[7] Min-Kyu Lim, Se-Doo Oh, Young-Ze Lee. Friction and wear of Inconel 690 and Inconel 600 for steam generator tube in room
EP
temperature water, Nucl. Eng. Des. 2003; 226; 97-105.
AC C
[8] X. Mi, Z.B. Cai, X.M. Xiong, H. Qian, L.C. Tang, Y.C. Xie, J.F. Peng, Min-hao Zhu. Investigation on fretting wear behavior of 690 alloy in water under various temperatures, Tribol. Int. 100 (2016) 400-409. [9] L. Xin, B.B. Yang, Z.H. Wang, J.Li, Y.H. Lu, T. Shoji. Effect of normal force on fretting wear behavior and mechanism of Alloy 690TT in high temperature water, Wear 2016; 368-369; 210-218. [10] Choon Yeol Lee, Li Si Tian, Joon Woo Bae, Young Suck Chai. 21
ACCEPTED MANUSCRIPT
Application of influence function method on the fretting wear of tube-to-plate contact, Tribol. Int. 2009; 42; 951-957. [11] Jae-do Kwon, Dae-kyu Park, Seung-wan Woo, Dong-hwan Yoon,
RI PT
Ilsup Chung. A study on fretting fatigue life for the Inconel alloy 600 at high temperature, Nucl. Eng. Des. 2010; 240; 2521-2527.
[12] Jae Yong Yun, Ho Sik Lee, Do Haeng Hur, Woong Soon Kang,
SC
Cheol Hyun Bae, Seon Jin Kim. Effect of oxidation film on the fretting
Wear 2016; 368-369; 344-349.
M AN U
wear behavior of Alloy 690 steam generator tube mated with SUS 409,
[13] J. Li, B.B. Yang, Y.H. Lu, L. Xin, Z.H. Wang, T. Shoji. The effects of electrochemical polarization condition and applied potential on
TE D
tribocorrosion behaviors of Inconel 690 alloys in water environment, Mater. Des. 2017; 119; 93-103.
[14] T. Dick, C. Paulin, G. Cailletaud, S. Fouvry, Experimental and
EP
numerical analysis of local and global plastic behaviour in fretting wear.
AC C
Tribol. Int. 2006; 39; 1036-1044. [15] C. Paulin, S. Fouvry, C. Meunier, Finite element modelling of fretting wear surface evolution: Application to a Ti-6A1-4V contact. Wear 2008; 264; 26-36.
[16] B. van Peteghem, S. Fouvry, J. Petit, Effect of variable normal force and frequency on fretting wear response of Ti-6Al-4V contact. Wear 2011; 271; 1535-1542. 22
ACCEPTED MANUSCRIPT
[17] R. D. Mindlin. Compliance of elastic bodies in contact. ASME j. appl. Mech. 1949; 16; 259-268. [18] Zhongrong Zhou, Léo Vincent. Fretting wear, Science Press, Beijing,
RI PT
2002. [19] J. Jiang, F.H. Stott, M.M. Stack. Some frictional features associated
with the sliding wear of the nickel-base alloy N80A at temperatures to
SC
250℃, Wear 1994; 176; 185-194.
M AN U
[20] Xiangzhen Xue, Sanmin Wang, Bo Li. Modification methodology of fretting wear in involute spline, Wear 2016; 368-369; 435-444. [21] Arnab Ghosh, Ben Leonard, Farshid Sadeghi. A stress based damage mechanics model to simulate fretting wear of Hertzian line contact in
TE D
partial slip, Wear 2013; 307; 87-99.
[22] Jae Yong Yun, Myung Chul Park, Gyeong Su Shin, Ji Haeng Heo, Dae Il Kim, Seon Jin Kim. Effects of amplitude and frequency on the
EP
wear mode change of Inconel 690 SG tube mated with SUS 409, Wear
AC C
2014; 313; 83-88.
[23] I. Chung, M. Lee. An experimental study on fretting wear behavior of crosscontacting Inconel 690 tubes, Nucl. Eng. Des.2011; 241 (10); 4103-4110.
[24] T.-H. Kim, S.-S. Kim, Fretting wear mechanisms of Zircaloy-4 and Inconel 600 contact in air, KSME Int. J. 2011; 15 (9); 1274-1280. [25] Y.-H. Lee, I.-S. Kim, S.-S. Kang, H.-D. Chung, A study on wear 23
ACCEPTED MANUSCRIPT
coefficients and mechanisms of steam generator tube materials, Wear 2011; 250 (1–12); 718-725. [26] G.S. Park, G.G. Kim, S.J. Kim, Sliding wear behaviors of steam
RI PT
generator tube materials in high temperature water environment, J. Nucl. Mater. 2006; 352 (1–3); 80-84.
[27] X. Jin, P.H. Shipway, W. Sun. The role of frictional power
SC
dissipation (as a function of frequency) and test temperature on contact
M AN U
temperature and the subsequent wear behavior in a stainless steel contact in fretting, Wear 2015; 330-331; 103-111.
[28] Xiaoling Zhang, R.G.Vitchev, W.Lauwerens, Jiawen He, J.-P.Celis. Transfer of molybdenum sulphide coating material onto corundum balls
TE D
in fretting wear tests, Thin Solid Films 2004; 446; 78-84. [29] Volker Franco Steier, Remy Kalombo Badibanga, Cosme Roberto Moreira da Silva, Murilo Magalhães Nogueira, José Alexander Araújo.
EP
Effect of chromium nitride coatings and cryogenic treatments on wear
AC C
and fretting fatigue resistance of aluminum, Electr. Pow. Syst. Res. 2014; 116; 322-329.
[30] S. Fyfitch. Corrosion and Stress Corrosion Cracking of Ni-Base Alloys, Comp. Nucl. Mater. 2012; 5; 69-92. [31] R.W. Revie, H.H. Uhlig, Corrosion and Corrosion Control, fourth ed., A John Wiley Sons, New Jersey 2008; 220-223.
24
ACCEPTED MANUSCRIPT
[32] Branko N. Popov, Chapter 11-High-Temperature Corrosion, Corros. Eng. 2015; 481-523 [33] A. Iwabuchi, T. Tsukamoto, Y. Tatsuyanagi, Electrochemical
RI PT
approach to corrosive wear of SKD61 die steel in Na2SO4solution, Wear 1992; 156; 301-313.
[34] Y. Sato, A. Iwabuchi, M. Uchidate, H. Yashiro. Dynamic corrosion
SC
properties of impact–fretting wear in high-temperature pure water, Wear
M AN U
2015; 330-331; 182-192.
[35] J. Pellier, J. Geringer, B. Forest. Fretting-corrosion between 316L SS and PMMA: Influence of ionic strength, protein and electrochemical conditions on material wear. Application to orthopaedic implants, Wear
TE D
2011; 271; 1563-1571.
[36] J. Hardell, S. Hernandez, et al., Effect of oxide layers and near
EP
surface transformations on friction and wear during tool steel and boron steel interaction at high temperatures, Wear 2015; 330-331; 223-229.
AC C
[37] F.H. Scott. High temperature sliding wear of metals. Tribol. Int. 2002; 35; 489-495.
25
ACCEPTED MANUSCRIPT
Highlights 1.A fretting wear test setup with a tube-on-plate configuration was developed to examine the performance of the materials in 285℃ water.
RI PT
2.With the increase of sliding amplitude and normal load, the wear volume of the alloy 690 tube is increasing.
SC
3.Sliding amplitude and normal load determines the wear mode of the materials。
M AN U
4.The synergic effect model of corrosion and mechanical wear on fretting
AC C
EP
TE D
wear behavior of the alloy 690 tube is proposed.