- Email: [email protected]
Characterizing the effect of creep on stress corrosion cracking of cold worked Alloy 690 in supercritical water environment
Journal of Nuclear Materials 492 (2017) 32e40
Contents lists available at ScienceDirect
Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat
Characterizing the effect of creep on stress corrosion cracking of cold worked Alloy 690 in supercritical water environment Lefu Zhang a, Kai Chen a, Donghai Du a, Wenhua Gao a, Peter L. Andresen a, b, Xianglong Guo a, * a b
School of Nuclear Science and Engineering, Shanghai Jiao Tong University, Shanghai 200240, PR China Andresen Consulting, Bakersfield, CA 93311, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 March 2017 Received in revised form 15 May 2017 Accepted 15 May 2017 Available online 17 May 2017
The effect of creep on stress corrosion cracking (SCC) was studied by measuring crack growth rates (CGRs) of 30% cold worked (CW) Alloy 690 in supercritical water (SCW) and inert gas environments at temperatures ranging from 450 C to 550 C. The SCC crack growth rate under SCW environments can be regarded as the cracking induced by the combined effect of corrosion and creep, while the CGR in inert gas environment can be taken as the portion of creep induced cracking. Results showed that the CW Alloy 690 sustained high susceptibility to intergranular (IG) cracking, and creep played a dominant role in the SCC crack growth behavior, contributing more than 80% of the total crack growth rate at each testing temperature. The temperature dependence of creep induced CGRs follows an Arrhenius dependency, with an apparent activation energy (QE) of about 225 kJ/mol. © 2017 Published by Elsevier B.V.
Keywords: Alloy 690 Supercritical water Stress corrosion cracking Creep Crack growth rates
1. Introduction Due to its system simplification, reduced plant layout and high thermal efficiency, the supercritical water cooled reactor (SCWR) is recognized as an advanced reactor concept by the Generation IV international forum since 2002 [1]. However, it is still difficult to build a demonstration SCWR, mainly because of the material problems caused by the high temperature, irradiation and the corrosive environment [2]. The operating temperature of primary component materials of a typical reactor vessel type SCWR has been increased to over 510 C, and the water pressure to 25 MPa [3]. High temperature strength and corrosion resistance in the SCW environment are the basic requirements of a candidate material for SCWR primary components. Owing to its excellent SCC performances, Alloy 690 has been selected as one of the major nuclear grade materials for pressurized water reactors (PWR), such as pressure boundary components, reactor vessel head penetrations and steam generator heat exchange tubes, although a steam generator is no longer required in
* Corresponding author. Room A332, School of Mechanical Engineering, No. 800 Dongchuan Road, Shanghai 200240, PR China. E-mail address: [email protected] (X. Guo). http://dx.doi.org/10.1016/j.jnucmat.2017.05.018 0022-3115/© 2017 Published by Elsevier B.V.
some SCWR designs. So far, the existing SCC data for candidate materials in SCW environment are mostly obtained from constant extension rate tensile (CERT) tests or slow strain rate tests (SSRT) [4]. SCC susceptibility is usually measured by the reduction in ductility in the test environment [5], the percentage of intergranular (IG) cracking on the fracture surfaces [6,7], or by the crack depth and density on the gage surface [4,8]. These methods are mostly for quick qualitative comparison of the SCC susceptibility between materials or testing conditions, and often result in under-estimating or overestimating the SCC behavior under plant operating conditions, and thus is not considered a reliable indication of SCC susceptibility except perhaps in extremely aggressive environments. SCC behavior can be quantitatively characterized by CGR experiments at a constant stress intensity factor (K), and Peng et al. obtained some SCC CGR data for 20% CW 316L stainless steel in SCW [9]. SCC CGRs data are of great importance both for the understanding of the SCC behavior and for engineering design of a SCWR, and are far more relevant and trustworthy than SSRT data. The SCC growth of Alloy 690 in SCWR environments remains unclear. Arioka [10] studied the effect of creep on cracking of cold worked Alloy 690 in high temperature environments in the range of 360 Ce460 C, and found that the intergranular creep cracking
L. Zhang et al. / Journal of Nuclear Materials 492 (2017) 32e40
(IG creep cracking) in inert and air environments has a similar morphology and temperature dependence as IGSCC in high temperature subcritical water. This suggests that creep is important in the growth of IGSCC for CW Alloy 690, consistent with the prevailing view that dynamic strain is fundamental to SCC in SCW. SCWRs operate at temperatures higher than the temperatures studied by Arioka et al., so creep will play an even larger role in the cracking of Alloy 690. Cold work is inevitable during the construction of a large system such as a SCWR. For example, welding will introduce residual shrinkage strains, peaking near the fusion line and approximately equivalent to cold work levels up to 20% plastic deformation [11]. In this work, the SCC and creep induced crack growth rate of 30% cold worked Alloy 690 were measured in SCW environment by the reversing direct current potential drop (DCPD) method. The objective of this study was to evaluate the temperature dependence, as well as the creep effect on the cracking of cold worked Alloy 690 in SCW. 2. Experiments 2.1. Testing material The test material is Alloy 690 bar with final mill anneal at 996 C for 20 min followed by air cooling. The chemical composition is listed in Table 1. The material was compressed by one step hydraulic forging up to 30% thickness reduction at room temperature. High-resolution scanning electron microscopy (SEM) backscatter electron (BSE) imaging under low kV conditions proved to be the most effective method for identifying fine intergranular cavities and grain boundary carbides [12,13]. Fig. 1 shows the micrographs of grain boundaries. It can be observed that primary carbides precipitate at grain boundaries semi-continuously along with small isolated TiN particles. These primary carbides are formed during original melting/solidification, usually MC, as well as TiN, and will precipitate along the grain boundaries, even during normal cooling from the annealing temperature, but more so during the 705 C/12 h typical thermal treatment. No micro-cracks or voids can be observed at the interfaces between matrix and carbides although the material was cold worked to 30% reduction in thickness. Fig. 2 shows the electron backscatter diffraction (EBSD) images of asreceived and 30% CW Alloy 690. The EBSD inverse pole figure maps reveal that grains are highly compressed in the CW material as compared with the as-received. Moreover, high misorientation is observed near grain boundaries of CW material, indicating high residual strain along the grain boundaries, which may accelerate IGSCC growth rate. Compact tension (CT) specimens with 12.7 mm thickness and 5% thickness side grooves were prepared in the S-L orientation to obtain the highest SCC CGR under testing corrosive environments [11]. The dimensions of the CT specimen are shown in Fig. 3. Vickers Hardness (HV) of as-received and CW specimens in three different faces were measured. The 1000 g-force was applied on the sample surface and held for 10 s before removal. A HV can then be calculated in units of kilogram-force per square millimeter (kg/mm2). Each value was averaged by at least three individual measurements on the same surface, and the results are given in Table 2. Hardness increases in all three orientations after cold deformation, especially
Table 1 Chemical composition (wt. %) of as-received Alloy 690.
33
in the front and back faces (as shown in Fig. 3). Yield strength (YS) increased remarkably after the 30% cold working, which could significantly affect its SCC behavior in SCW environment. 2.2. Experimental procedure The crack growth rate measurements were performed with one specimen using “on the fly” changes to study the effects of environment to obtain consistent data. The SCC tests were conducted at constant stress intensity factor (K) in SCW environment at temperatures between 400 and 550 C and at a pressure of 25 MPa. Creep crack growth rate tests were carried out in high purity argon at the same temperature and at the same K as SCC tests. The crack length was measured by the reversing DCPD system [14]. The feed water to the autoclave was pumped by a double diaphragm high pressure metering pump, and the autoclave pressure was maintained by a precision back pressure regulator, which maintains pressure at ±0.1 MPa fluctuation at 25 MPa. The water chemistry was controlled by the recirculating a water loop connected to the autoclave with a flow rate of about 1 L/h, corresponding to an autoclave refresh rate of 1 vol/h. The feed water to the autoclave was purified by nuclear grade polishing mixed bed resin (Rohm and Hass DS160). High purity argon was continuously bubbled into the water tank to maintain a deaerated water environment. The dissolved oxygen (DO) in the feed water to the autoclave was monitored by using an inline DO sensor, with the value well below 10 ppb. Water purity was continuously monitored by two conductivity meters installed at the inlet and outlet of autoclave. The inlet conductivity was controlled below 0.06 mS/cm and the outlet water conductivity was about 0.2 mS/cm. A pressure balanced seal eliminated the force on the pull rod that would have been induced by the pressure differential between inside and outside of the autoclave. Before SCC testing in SCW, the CT specimen was fatigue precracked by 1 mm in room temperature air at Kmax ¼ 25 MPa√m, load ratio R ¼ 0.3 and frequency f ¼ 1 Hz. Then the pre-cracked specimen was loaded in autoclave for the subsequent transitioning stages in SCW to change the cyclic plastic zone and TG morphology at fatigue crack tip to the monotonic plastic zone and IG morphology of an SCC crack tip. Transitioning was performed in SCW at 400 C and 25 MPa, with R increased from 0.3 to 0.7, and f decreased from 1 Hz to 0.001 Hz, until cracking linear response (which usually was close to the SCC crack growth rate). Then a change was made to constant K conditions to evaluate the SCC crack growth response. The transitioning stages are among the most important elements of a successful SCC crack growth rate test. A hold time of 3000e9000 s at the maximum fatigue load (Kmax) is ideally introduced if the transitioning behavior looks abnormal. The corrosion crack was propagated a minimum length of one grain size (~50 mm) before finishing the transitioning step. To evaluate the contribution of creep to SCC growth rate quantitatively, the creep test was conducted on the same specimen after the SCC crack growth rate was measured in SCW. Before the creep induced crack growth rate measurement, the load on the specimen was released, the high pressure pump was stopped, and the autoclave was depressurization then evacuated using high purity argon for a minimum of 10 h to establish a non-corrosive inert environment in the autoclave for the following creep test. Then the specimen was reloaded up to K ¼ 25 MPa√m to measure the creep induce crack growth rate at each given temperature. 3. SCC and creep CGRs
Elements
Cr
Ni
Mo
P
Ti
C
Fe
Mn
Si
wt %
29.30
60.42
0.01
0.006
0.37
0.034
9.21
0.22
0.06
After testing, each specimen was sliced into two pieces along the center plane of the thickness. One piece was fatigued apart at room
34
L. Zhang et al. / Journal of Nuclear Materials 492 (2017) 32e40
Fig. 1. Microstructure of the 30% CW Alloy 690 with carbides along grain boundaries.
Fig. 2. EBSD inverse pole figures (a, c) and local misorientation maps (b,d) for as-received (a,b) and 30% CW (c,d) Alloy 690 specimens.
Fig. 3. Geometry and dimensions of the CT specimen (in mm).
Table 2 The Vickers hardness in three different orientations of as-received and 30% CW materials (kg/mm2). Faces
Top
Back
Front
As-received 30% CW
241.5 295.5
204.7 301.6
200.3 300.4
present results show that the error in the DCPD measurement was small enough so that no correction is needed. The crack growth curves for each environment are shown in Fig. 5. Fig. 5(a) and (b) show the pre-crack and transition stages, respectively. Fig. 5(b) and (c) show the SCC growth in SCW at 450 C, 500 C and 550 C, with CGRs of 5.3 107 mm/s, 6.0 106 mm/s and 5.9 105 mm/s, respectively. Fig. 5(c) and (d) show the creep growth rates in inert gas environment at 450 C, 500 C and 550 C, with the CGRs of 3.9 107 mm/s, 5.1 106 mm/s and 4.7 105 mm/s, respectively. Using this approach, the creep induced CGRs were quantitatively separated from the SCC growth rates, and it is easy to evaluate the effect of creep on the total CGRs in SCW. Table 3 summarizes the SCC and creep induced CGRs for 30% CW Alloy 690 specimen, showing that both SCC and creep CGRs increased rapidly with increasing temperature. Moreover, creep CGR at each temperature is only slightly lower than the SCC CGRs in SCW at the same temperature. Thus, creep induce cracking dominates the crack growth for highly cold worked Alloy 690 in SCW environments.
4. Discussion temperature to examine the fracture surface, the other piece was polished for crack path observation, as shown in Fig. 4(a). The crack length detected by the DCPD system may not be exactly correct, and the crack length-time curve needs to be corrected by the observed crack length under each condition, as shown in Fig. 4(b). The
4.1. Effect of creep on crack growth rate The data in Table 3 indicate that CW Alloy 690 has high creep cracking susceptibility. Nikbin [15] has predicted the creep crack growth rates for a broad spread of materials by:
L. Zhang et al. / Journal of Nuclear Materials 492 (2017) 32e40
35
Fig. 4. (a) The sliced specimen after test, and (b) fracture surface of the cracked specimen, showing the crack length at different testing conditions.
Fig. 5. The crack growth curve of 30% CW specimen, showing (a) pre-crack, (b) transition, (c) SCC growth and (d) creep crack growth.
. a_ ¼ 8:3 104 C *0:85 ε*f
Table 3 The SCC and creep CGRs at 450 Ce550 C. Specimen
Temperature
SCC
Creep
30%CW 30%CW 30%CW
450 C 500 C 550 C
5.3 107 mm/s 6.0 106 mm/s 5.9 105 mm/s
3.9 107 mm/s 5.1 106 mm/s 4.7 105 mm/s
(1)
While ε*f is the creep failure strain, and C* is the creep fracture mechanics parameter in MJm2h1. The ε*f is considered as 0.3 for Alloy 800H [16], while 0.03 is used in this study considering the cold work effect. C* characterizes the stress state at the creep crack
36
L. Zhang et al. / Journal of Nuclear Materials 492 (2017) 32e40
tip, and is calculated according to ASTM E1457. Nikbin [15] showed that relatively low dependence of creep crack growth rates on temperature was observed when characterized in terms of C*. As a result, the estimated creep crack growth rates by Equation (1) comparing with experimental data at different temperature are shown in Fig. 6, and demonstrates a good agreement (within a factor of 2). The CGRs obtained during the SCC test in SCW consists of two major parts, one part is the actual SCC crack growth rate which is induced by the synergistic effect of intergranular corrosion and load, and the other part is the contribution of high temperature creep. The measured SCC crack growth rate can be considered the total crack growth rate in SCW. From Table 3, creep induced cracking is a major contribution to the total crack growth rate of 30% cold worked Alloy 690 in SCW, and the contribution from creep cracking is about 80% of the total at each test temperatures, as shown in Fig. 7. The morphology of cracking has been examined by SEM, and micrographs are shown in Figs. 8 and 9. Small secondary cracks can be observed away from the main crack (Fig. 8(a)). The final creep crack tip is relatively sharp with a radius of less than 0.1 mm (Fig. 8(b)). Enlarging points C, D and E about 20e50 mm ahead of the crack tip in Fig. 8(b), creep crack (Fig. 8(c)) and discontinuous cavities along grain boundaries can be identified (Fig. 8(d) and e). Cavities preferentially form near the grain boundary carbides. The fracture surfaces in Fig. 9 show discontinuous cavities along the secondary cracks. Although covered with oxides, fracture surfaces and secondary cracks can be distinguished as intergranular, and the morphology is in consistent with Fig. 8(c), d and 8e. The IG cavities are evidence of creep attack during the propagation of main crack. Dynamic crack tip strain rate is one of the major ratedetermining factors that control the crack propagation in high temperature water environments [17]. A crack tip plastic stain equation for a growing crack was proposed by Gao and Hwang [18] to determine the crack tip strain rate in a work hardened material. According to their equation, the plastic strain could reach as high as
Fig. 7. Comparison of crack growth rate induced by Creep to the total in SCW.
5%, and the dynamic strain rate is also relatively high even at a distance of 50 mm ahead of the crack tip. During the SCC test, the load applied on the CT specimen produced a plastic zone along with a plastic strain field at the crack tip. Thus, the main IG crack propagation is promoted by the combined effect of strain rate at crack tip and corrosion of grain boundary. At the same time, dislocation movement is enhanced ahead of the main crack tip under the effect of plastic strain field. The enhanced dislocation movement tends to cause pile up and accumulation at the grain boundaries, especially at GB carbides. As a result, grain boundaries are weakened by cavities, and total IG CGR is accelerated. Arioka studied the SCC and creep behavior of cold-worked Alloy 690 in high temperature water environment [10,19,20]. His study indicated that creep plays an important role in IGSCC. The high density of cavities is formed because of the diffusion of CW induced
Fig. 6. Comparison of SCC and creep CGRs.
L. Zhang et al. / Journal of Nuclear Materials 492 (2017) 32e40
37
Fig. 8. The cross sectional view of the crack path, (a) main crack, (b) creep crack tip, (c,d,e) grain boundary ahead of the crack tip.
Fig. 9. The SEM images of cavity at the secondary cracks.
vacancies. According to Arioka, CW induced vacancies and hydrogen induced vacancies are the main sources of vacancy in the material. In this study, the material was highly cold worked (30%) and has large amounts of GB carbides, as shown in Fig. 1. Moreover, similar cavities around IG carbides were also observed in Figs. 8 and 9. Thus, the vacancies were mainly formed due to the CW effect and nucleation at GB carbides. When vacancies accumulate ahead of the crack tip, driven by stress gradient, cavities tend to form and affect the bond strength of grain boundaries. Consequently, the bond strength is assumed to be weakened and hence the SCC and creep crack growth rates increase. 4.2. Effect of temperature The SCC and creep growth rates of 30% CW Alloy 690 at different temperatures are shown in Fig. 10. Both SCC and creep growth rates increase with temperature, and follow an Arrhenius dependency. In the present research, the activation energy (QE) is calculated and marked in Fig. 10. The QE of SCC and creep is 228 kJ/mol and 222 kJ/ mol, respectively, which is comparable to the QE of oxide growth of nickel-base alloy (~200 kJ/mol) [21]. In previous SCC growth rate experiments in SCW by Peng et al. [9], the CGR of 20% CW 316L stainless steel decreased with the temperature increasing from 400 C to 450 C. The authors thought that the decrease of the CGR was due to the crack tip blunting induced by the rapid increase of the material oxidation rate and softening at higher temperature. In this study, the Alloy 690 has much lower corrosion/oxidation rate
Fig. 10. Temperature dependence on SCC and creep CGRs for 30%CW Alloy 690.
compared to 316L [4], thus the crack tip is less blunted, which could explain the opposite effect of temperature on SCC for Alloy 690 vs. 316L. Was has studied the SCC behavior of nickel-base alloys in pure, deaerated (<10 ppb O2) SCW with CERT experiments, and found that the QE is between 84 and 87 kJ/mol [4], which is much
38
L. Zhang et al. / Journal of Nuclear Materials 492 (2017) 32e40
lower than that in this study. As mentioned in the Introduction Section, we believe that there must be some differences between CERT and CT tests. Firstly, the forced strain rate applied by CERT is larger than that of constant K test for CT specimen. Secondly, the strain rate is very important in CERT test, which might give completely opposite results. Only enough low strain rate (107~108 s1) can obtained reliable data, especially for materials with high SCC resistance. Thirdly, the results obtained by CERT are usually given by more than one specimen and have relatively large scatter, while CT test is conducted with only one specimen with “on the fly” change, which has good consistency. Fourthly, the CERT is mainly for crack initiation and SCC susceptibility tests, focusing on the surface cracks, while CT test focuses on the long crack growth rate with a different crack tip chemistry from the bulk water. Above all, the different Q could be attributed to the above differences, and it is reasonable to believe that the results obtained by CERT test is more suitable for SCC initiation, while more suitable for SCC growth by CT test. The CGRs obtained in this study are plotted to compare with results of works by Arioka et al. [19] and Gao et al. [22], as shown in Fig. 11 (data in this study are plotted with red solid points). The SCC and creep CGRs of 30% CW Alloy 690 in this work are below that of 20% CW Alloy TT690 obtained in PWR primary water and gas environments in Arioka's work [19] in Fig. 11(a), and the creep crack growth of Alloy 718 in Gao's work [22] in Fig. 11(b), but most of the studies show a similar slope vs. temperature, except the dotted line for alloy 718 in oxygenas shown in Fig. 11(b). The major reason can be attributed to the difference of load in the experiments: literature data were obtained at a higher K value of 30 MPa√m in PWR water vs. 40 MPa√m in gas. The difference in the degree of cold work has less effect if comparing the increase of yield strength in Alloy 690. The similarity in slope for most CGR responses in Fig. 11 represent a similar temperature dependency and similar activation energy (QE), consistent with the results of this study. The different slope for Alloy 718 in oxygen may have been caused by the enhancement of oxygen on crack growth [22]. While for the lower crack growth rates compared with Arioka's data, Bruemmer [12,13] and Arioka [10] reported that slower SCC growth rates were observed in annealed Alloy 690 than TT690 in PWR primary water, mainly due to the grain boundary carbides. Thus, the difference could be attributed to the material conditions, especially the difference in IG carbide coverage. The concept of an environmental correction factor Fen [23] can be borrowed to quantify the corrosion effect of a high temperature
environment on the stress corrosion crack growth rate of a material,
Fen ¼ CGRSCC CGRcreep:
(2)
The temperature dependence of Fen is shown in Fig. 12. All Fen at 450e550 C lie well below 1.5, indicating a relatively mild corrosion effect of SCW environment on crack growth of cold worked Alloy 690. Fig. 11(a) shows remarkably higher SCC crack growth rates in PWR environment compared with gas environment, suggesting an additional environmental enhancement effect on SCC growth in addition to the cavity-related creep mechanism. However, the SCW environmental enhancement effect on the SCC behavior of 30% CW Alloy 690 in this study is negligible, and can be taken as in high temperature inert environment (although there was undoubtedly some trace levels of oxygen and water vapor). SCC growth is the result of the combined effect of mechanical driving damage and chemical or/and electrochemical reactions at the crack tip [24]. In this study, the 30% CW material has a relatively high creep CGRs (107 - 105 mm/s) in SCW, indicating a high mechanical driving damage during the crack growth. This high creep growth rate gives
Fig. 12. The temperature dependence of Fen for 30%CW Alloy 690.
Fig. 11. Comparison of SCC and creep CGRs for 30% CW Alloy 690 in this study with (a) 20% CW TT690 by Arioka et al. [19] and (b) Inconel 718 Alloy by Gao et al. [22].
L. Zhang et al. / Journal of Nuclear Materials 492 (2017) 32e40
less time for the SCW to corrode the crack tip, and creep dominates the crack growth. This can be the major reason to explain the low Fen (<1.5) of Alloy 690 in SCW environment. 4.3. Mechanistic analysis The slip-oxidation model, also called Ford-Andresen (F-A) model [24] is among the most widely accepted model, and has been widely used for predicting CGRs of various materials, including stainless steels, nickel alloys and low alloy steels in LWR environment. In this model, the failure due to SCC exhibits IG cracking characteristics, and oxide film on the crack front is ruptured by dislocation movement at crack tip, which exposes bare metal at the crack tip to the corrosive environment, with subsequent oxidation reforming the oxide film again (repassivaion), and the slipoxidation process repeats. The results in this study confirmed that formation of cavity ahead of crack tip by creep effect plays an important role in the SCC growth rates for CW Alloy 690 in higher temperature environments. The testing temperatures were much higher and the SCW environments were more corrosive than the operating conditions in current LWRs. The higher temperature leads to a greater effect of creep on IG cracking, and the mechanism of SCC in SCW becomes more complex, and includes the combined effect of creep and corrosion. Inspired by the film-induced cleavage model [25,26], a creepinduced extension model (as shown in Fig. 13) is proposed to explain the SCC behavior of CW Alloy 690 in SCW. First, the crack grows by the bare surface dissolution and oxide growth, and this process is governed by the same rate-determining steps as described in the F-A model. At the same time, for a long crack as illustrated in Fig. 13(a), there exists a plastic zone at the crack tip if a load is applied, and the plastic zone creates a strain field and strain gradient ahead of the crack tip. For highly cold worked materials, the yield strength is dramatically higher, and there is a high strain gradient and high density of vacancies at the crack tip. Thus, the CW induced vacancies in the grain interior are activated and tend to diffuse into the grain boundaries ahead of the crack tip under the effect of strain gradient. The accumulation of vacancies tends to
39
form cavities at the grain boundary. As a result, the grain boundary is weakened. In the creep-induced extension model described in Fig. 13(b), the crack tip will readily penetrate a distance a* into the weakened grain boundary ahead of the crack tip when the oxide ruptures, and this extension increases the average crack growth rate. The extent of this additional creep-induced crack advance is governed by the creep susceptibility of the material at the environment temperature. For 30% CW Alloy 690 in this study, the creep CGRs are in the range of 107 e105 mm/s at 450e550 C. Thus, the creep-induced extension a* is much higher than the bare surface dissolution and oxide growth in slip-oxidation model. However, in this creep-induced extension mechanism there are several factors that require clarification, such as oxygen diffusion through the cavities into the matrix and its effect on creep and oxidation, and vacancy diffusion coefficient ahead of the crack tip. Thus, further work is necessary to complete this model.
5. Conclusion The crack growth rates of 30% CW Alloy 690 in SCW and inert gas environments at temperatures between 450 and 550 C were evaluated to study the relative effect of creep and SCC on the total crack growth rate. The following conclusions were drawn: 1) CW Alloy 690 exhibits a relatively high susceptibility to IG cracking in SCW, and creep contribute the largest portion to the total crack growth rate. 2) Extensive cavity formation was observed ahead of the crack tip, and cavities tend to nucleate near GB carbides. Cavities weaken the bond of grain boundary and accelerate the IG crack growth. 3) Temperature strongly affects the CGRs in SCW. The temperature dependence follows an Arrhenius dependency with an activation energy QE of about 225 kJ/mol for both SCC and creep crack growth. 4) A creep-induced extension mechanism based on F-A model can be used to explain the SCC behavior of CW Alloy 690 in SCW.
Fig. 13. A revised F-A model illustrating the SCC mechanism of Alloy 690 in SCW.
40
L. Zhang et al. / Journal of Nuclear Materials 492 (2017) 32e40
Acknowledgment Thanks to the microstructural characterization of Instrumental Analysis Center of SJTU. References [1] Generation IV, F., Technology roadmap update for generation IV nuclear energy systems, OECD Nuclear Energy Agency for the Generation IV International Forum, 2014. [2] L. Zhang, Y. Bao, R. Tang, Selection and corrosion evaluation tests of candidate SCWR fuel cladding materials, Nucl. Eng. Des. 249 (8) (2012) 180e187. [3] X.J. Liu, X. Cheng, Coupled thermal-hydraulics and neutron-physics analysis of SCWR with mixed spectrum core, Prog. Nucl. Energy 52 (7) (2010) 640e647. [4] G.S. Was, et al., Corrosion and stress corrosion cracking in supercritical water, J. Nucl. Mater. 371 (1e3) (2007) 176e201. [5] R. Novotny, et al., Stress corrosion cracking susceptibility of austenitic stainless steels in supercritical water conditions, J. Nucl. Mater. 409 (2) (2011) 117e123. [6] Z. Shen, et al., The effect of temperature on the SSRT behavior of austenitic stainless steels in SCW, J. Nucl. Mater. 454 (1e3) (2014) 274e282. [7] H. Je, A. Kimura, Stress corrosion cracking susceptibility of oxide dispersion strengthened ferritic steel in supercritical pressurized water dissolved with different hydrogen and oxygen contents, Corros. Sci. 78 (2014) 193e199. [8] S. Teysseyre, et al., Effect of irradiation on stress corrosion cracking in supercritical water, J. Nucl. Mater. 371 (1e3) (2007) 107e117. [9] Q. Peng, et al., Stress corrosion crack growth in type 316 stainless steel in supercritical water, Corrosion 63 (11) (2007) 1033e1041. [10] K. Arioka, et al., Dependence of stress corrosion cracking of alloy 690 on temperature, cold work, and carbide precipitationdrole of diffusion of vacancies at crack tips, CORROSION 67 (3) (2011), 035006e35011-035006-18. [11] P.L. Andresen, M.M. Morra, Stress corrosion cracking of stainless steels and nickel alloys in high-temperature water, CORROSION 64 (1) (2008) 15e29. [12] Bruemmer, S.M., et al., Cold-work effects on stress corrosion crack growth in Alloy 690 tubing and plate materials, in 17th International Conference on Environmental Degradation of Materials in Nuclear Power Systems-water
Reactors: Ottawa, Canada. [13] S.M. Bruemmer, et al., High-resolution characterizations of grain boundary damage and stress corrosion cracks in cold-rolled alloy 690, in: 15th International Conference on Environmental Degradation of Materials in Nuclear Power Systems-water Reactors, 2011 (Hoboken, US). [14] D. Du, et al., SCC crack growth rate of cold worked 316L stainless steel in PWR environment, J. Nucl. Mater. 456 (2015) 228e234. [15] K.M. Nikbin, An engineering approach to the prediction of creep crack growth, J. Eng. Mater. Technol. 108 (2) (1986) 186e191. [16] G.A. Webster, Creep crack growth, in: I. Milne, R.O. Ritchie, B. Karihaloo, I. Milne, R.O. Ritchie, B. Karihaloo (Eds.), Comprehensive Structural Integrity, Pergamon Oxford, 2003, pp. 241e271. [17] P.L. Andresen, M.M. Morra, Stress corrosion cracking of stainless steels and nickel alloys in high-temperature water, CORROSION 64 (1) (2008) 15e29. [18] Y.C. Gao, K.C. Huang, Elastic-plastic Field in Steady Crack Growth in a Strainhardening Material, 1981 (Cannes, France). [19] K. Arioka, et al., Degradation of alloy 690 after relatively short times, CORROSION 72 (10) (2016) 1252e1268. [20] K. Arioka, W.R. Whitney award lecture: change in bonding strength at grain boundaries before long-term SCC initiation, CORROSION 71 (4) (2014) 403e419, 2014. [21] G.S. Was, S. Teysseyre, Z. Jiao, Corrosion of austenitic alloys in supercritical water, CORROSION 62 (11) (2006) 989e1005. [22] M. Gao, D.J. Dwyer, R.P. Wei, Chemical and microstructural aspect of creep crack growth in INCONEL 718 Alloy, in: Superalloys 718, 625, 706 and Various Derivatives, The minerals, Metals and Materials Society, 1994. [23] M. Higuchi, et al., Revised and New Proposal of Environmental Fatigue Life Correction Factor (Fen) for Carbon and Low-Alloy Steels and Nickel Base Alloys in LWR Water Environments, 2006, pp. 93e102. [24] P.L. Andresen, Emerging issues and fundamental processes in environmental cracking in hot water (Reprinted from Proceedings of the CORROSION/2007 research topical symposium “Advances in Environmentally Assisted Cracking”, 2007), CORROSION 64 (5) (2008) 439e464. [25] A. Paskin, et al., Environmentally induced crack nucleation and brittle fracture, Acta Metall. 30 (9) (1982) 1781e1788. [26] K. Sieradzki, R.C. Newman, Brittle behavior of ductile metals during stresscorrosion cracking, Philos. Mag. A 51 (1) (1985) 95e132.
Contents lists available at ScienceDirect
Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat
Characterizing the effect of creep on stress corrosion cracking of cold worked Alloy 690 in supercritical water environment Lefu Zhang a, Kai Chen a, Donghai Du a, Wenhua Gao a, Peter L. Andresen a, b, Xianglong Guo a, * a b
School of Nuclear Science and Engineering, Shanghai Jiao Tong University, Shanghai 200240, PR China Andresen Consulting, Bakersfield, CA 93311, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 March 2017 Received in revised form 15 May 2017 Accepted 15 May 2017 Available online 17 May 2017
The effect of creep on stress corrosion cracking (SCC) was studied by measuring crack growth rates (CGRs) of 30% cold worked (CW) Alloy 690 in supercritical water (SCW) and inert gas environments at temperatures ranging from 450 C to 550 C. The SCC crack growth rate under SCW environments can be regarded as the cracking induced by the combined effect of corrosion and creep, while the CGR in inert gas environment can be taken as the portion of creep induced cracking. Results showed that the CW Alloy 690 sustained high susceptibility to intergranular (IG) cracking, and creep played a dominant role in the SCC crack growth behavior, contributing more than 80% of the total crack growth rate at each testing temperature. The temperature dependence of creep induced CGRs follows an Arrhenius dependency, with an apparent activation energy (QE) of about 225 kJ/mol. © 2017 Published by Elsevier B.V.
Keywords: Alloy 690 Supercritical water Stress corrosion cracking Creep Crack growth rates
1. Introduction Due to its system simplification, reduced plant layout and high thermal efficiency, the supercritical water cooled reactor (SCWR) is recognized as an advanced reactor concept by the Generation IV international forum since 2002 [1]. However, it is still difficult to build a demonstration SCWR, mainly because of the material problems caused by the high temperature, irradiation and the corrosive environment [2]. The operating temperature of primary component materials of a typical reactor vessel type SCWR has been increased to over 510 C, and the water pressure to 25 MPa [3]. High temperature strength and corrosion resistance in the SCW environment are the basic requirements of a candidate material for SCWR primary components. Owing to its excellent SCC performances, Alloy 690 has been selected as one of the major nuclear grade materials for pressurized water reactors (PWR), such as pressure boundary components, reactor vessel head penetrations and steam generator heat exchange tubes, although a steam generator is no longer required in
* Corresponding author. Room A332, School of Mechanical Engineering, No. 800 Dongchuan Road, Shanghai 200240, PR China. E-mail address: [email protected] (X. Guo). http://dx.doi.org/10.1016/j.jnucmat.2017.05.018 0022-3115/© 2017 Published by Elsevier B.V.
some SCWR designs. So far, the existing SCC data for candidate materials in SCW environment are mostly obtained from constant extension rate tensile (CERT) tests or slow strain rate tests (SSRT) [4]. SCC susceptibility is usually measured by the reduction in ductility in the test environment [5], the percentage of intergranular (IG) cracking on the fracture surfaces [6,7], or by the crack depth and density on the gage surface [4,8]. These methods are mostly for quick qualitative comparison of the SCC susceptibility between materials or testing conditions, and often result in under-estimating or overestimating the SCC behavior under plant operating conditions, and thus is not considered a reliable indication of SCC susceptibility except perhaps in extremely aggressive environments. SCC behavior can be quantitatively characterized by CGR experiments at a constant stress intensity factor (K), and Peng et al. obtained some SCC CGR data for 20% CW 316L stainless steel in SCW [9]. SCC CGRs data are of great importance both for the understanding of the SCC behavior and for engineering design of a SCWR, and are far more relevant and trustworthy than SSRT data. The SCC growth of Alloy 690 in SCWR environments remains unclear. Arioka [10] studied the effect of creep on cracking of cold worked Alloy 690 in high temperature environments in the range of 360 Ce460 C, and found that the intergranular creep cracking
L. Zhang et al. / Journal of Nuclear Materials 492 (2017) 32e40
(IG creep cracking) in inert and air environments has a similar morphology and temperature dependence as IGSCC in high temperature subcritical water. This suggests that creep is important in the growth of IGSCC for CW Alloy 690, consistent with the prevailing view that dynamic strain is fundamental to SCC in SCW. SCWRs operate at temperatures higher than the temperatures studied by Arioka et al., so creep will play an even larger role in the cracking of Alloy 690. Cold work is inevitable during the construction of a large system such as a SCWR. For example, welding will introduce residual shrinkage strains, peaking near the fusion line and approximately equivalent to cold work levels up to 20% plastic deformation [11]. In this work, the SCC and creep induced crack growth rate of 30% cold worked Alloy 690 were measured in SCW environment by the reversing direct current potential drop (DCPD) method. The objective of this study was to evaluate the temperature dependence, as well as the creep effect on the cracking of cold worked Alloy 690 in SCW. 2. Experiments 2.1. Testing material The test material is Alloy 690 bar with final mill anneal at 996 C for 20 min followed by air cooling. The chemical composition is listed in Table 1. The material was compressed by one step hydraulic forging up to 30% thickness reduction at room temperature. High-resolution scanning electron microscopy (SEM) backscatter electron (BSE) imaging under low kV conditions proved to be the most effective method for identifying fine intergranular cavities and grain boundary carbides [12,13]. Fig. 1 shows the micrographs of grain boundaries. It can be observed that primary carbides precipitate at grain boundaries semi-continuously along with small isolated TiN particles. These primary carbides are formed during original melting/solidification, usually MC, as well as TiN, and will precipitate along the grain boundaries, even during normal cooling from the annealing temperature, but more so during the 705 C/12 h typical thermal treatment. No micro-cracks or voids can be observed at the interfaces between matrix and carbides although the material was cold worked to 30% reduction in thickness. Fig. 2 shows the electron backscatter diffraction (EBSD) images of asreceived and 30% CW Alloy 690. The EBSD inverse pole figure maps reveal that grains are highly compressed in the CW material as compared with the as-received. Moreover, high misorientation is observed near grain boundaries of CW material, indicating high residual strain along the grain boundaries, which may accelerate IGSCC growth rate. Compact tension (CT) specimens with 12.7 mm thickness and 5% thickness side grooves were prepared in the S-L orientation to obtain the highest SCC CGR under testing corrosive environments [11]. The dimensions of the CT specimen are shown in Fig. 3. Vickers Hardness (HV) of as-received and CW specimens in three different faces were measured. The 1000 g-force was applied on the sample surface and held for 10 s before removal. A HV can then be calculated in units of kilogram-force per square millimeter (kg/mm2). Each value was averaged by at least three individual measurements on the same surface, and the results are given in Table 2. Hardness increases in all three orientations after cold deformation, especially
Table 1 Chemical composition (wt. %) of as-received Alloy 690.
33
in the front and back faces (as shown in Fig. 3). Yield strength (YS) increased remarkably after the 30% cold working, which could significantly affect its SCC behavior in SCW environment. 2.2. Experimental procedure The crack growth rate measurements were performed with one specimen using “on the fly” changes to study the effects of environment to obtain consistent data. The SCC tests were conducted at constant stress intensity factor (K) in SCW environment at temperatures between 400 and 550 C and at a pressure of 25 MPa. Creep crack growth rate tests were carried out in high purity argon at the same temperature and at the same K as SCC tests. The crack length was measured by the reversing DCPD system [14]. The feed water to the autoclave was pumped by a double diaphragm high pressure metering pump, and the autoclave pressure was maintained by a precision back pressure regulator, which maintains pressure at ±0.1 MPa fluctuation at 25 MPa. The water chemistry was controlled by the recirculating a water loop connected to the autoclave with a flow rate of about 1 L/h, corresponding to an autoclave refresh rate of 1 vol/h. The feed water to the autoclave was purified by nuclear grade polishing mixed bed resin (Rohm and Hass DS160). High purity argon was continuously bubbled into the water tank to maintain a deaerated water environment. The dissolved oxygen (DO) in the feed water to the autoclave was monitored by using an inline DO sensor, with the value well below 10 ppb. Water purity was continuously monitored by two conductivity meters installed at the inlet and outlet of autoclave. The inlet conductivity was controlled below 0.06 mS/cm and the outlet water conductivity was about 0.2 mS/cm. A pressure balanced seal eliminated the force on the pull rod that would have been induced by the pressure differential between inside and outside of the autoclave. Before SCC testing in SCW, the CT specimen was fatigue precracked by 1 mm in room temperature air at Kmax ¼ 25 MPa√m, load ratio R ¼ 0.3 and frequency f ¼ 1 Hz. Then the pre-cracked specimen was loaded in autoclave for the subsequent transitioning stages in SCW to change the cyclic plastic zone and TG morphology at fatigue crack tip to the monotonic plastic zone and IG morphology of an SCC crack tip. Transitioning was performed in SCW at 400 C and 25 MPa, with R increased from 0.3 to 0.7, and f decreased from 1 Hz to 0.001 Hz, until cracking linear response (which usually was close to the SCC crack growth rate). Then a change was made to constant K conditions to evaluate the SCC crack growth response. The transitioning stages are among the most important elements of a successful SCC crack growth rate test. A hold time of 3000e9000 s at the maximum fatigue load (Kmax) is ideally introduced if the transitioning behavior looks abnormal. The corrosion crack was propagated a minimum length of one grain size (~50 mm) before finishing the transitioning step. To evaluate the contribution of creep to SCC growth rate quantitatively, the creep test was conducted on the same specimen after the SCC crack growth rate was measured in SCW. Before the creep induced crack growth rate measurement, the load on the specimen was released, the high pressure pump was stopped, and the autoclave was depressurization then evacuated using high purity argon for a minimum of 10 h to establish a non-corrosive inert environment in the autoclave for the following creep test. Then the specimen was reloaded up to K ¼ 25 MPa√m to measure the creep induce crack growth rate at each given temperature. 3. SCC and creep CGRs
Elements
Cr
Ni
Mo
P
Ti
C
Fe
Mn
Si
wt %
29.30
60.42
0.01
0.006
0.37
0.034
9.21
0.22
0.06
After testing, each specimen was sliced into two pieces along the center plane of the thickness. One piece was fatigued apart at room
34
L. Zhang et al. / Journal of Nuclear Materials 492 (2017) 32e40
Fig. 1. Microstructure of the 30% CW Alloy 690 with carbides along grain boundaries.
Fig. 2. EBSD inverse pole figures (a, c) and local misorientation maps (b,d) for as-received (a,b) and 30% CW (c,d) Alloy 690 specimens.
Fig. 3. Geometry and dimensions of the CT specimen (in mm).
Table 2 The Vickers hardness in three different orientations of as-received and 30% CW materials (kg/mm2). Faces
Top
Back
Front
As-received 30% CW
241.5 295.5
204.7 301.6
200.3 300.4
present results show that the error in the DCPD measurement was small enough so that no correction is needed. The crack growth curves for each environment are shown in Fig. 5. Fig. 5(a) and (b) show the pre-crack and transition stages, respectively. Fig. 5(b) and (c) show the SCC growth in SCW at 450 C, 500 C and 550 C, with CGRs of 5.3 107 mm/s, 6.0 106 mm/s and 5.9 105 mm/s, respectively. Fig. 5(c) and (d) show the creep growth rates in inert gas environment at 450 C, 500 C and 550 C, with the CGRs of 3.9 107 mm/s, 5.1 106 mm/s and 4.7 105 mm/s, respectively. Using this approach, the creep induced CGRs were quantitatively separated from the SCC growth rates, and it is easy to evaluate the effect of creep on the total CGRs in SCW. Table 3 summarizes the SCC and creep induced CGRs for 30% CW Alloy 690 specimen, showing that both SCC and creep CGRs increased rapidly with increasing temperature. Moreover, creep CGR at each temperature is only slightly lower than the SCC CGRs in SCW at the same temperature. Thus, creep induce cracking dominates the crack growth for highly cold worked Alloy 690 in SCW environments.
4. Discussion temperature to examine the fracture surface, the other piece was polished for crack path observation, as shown in Fig. 4(a). The crack length detected by the DCPD system may not be exactly correct, and the crack length-time curve needs to be corrected by the observed crack length under each condition, as shown in Fig. 4(b). The
4.1. Effect of creep on crack growth rate The data in Table 3 indicate that CW Alloy 690 has high creep cracking susceptibility. Nikbin [15] has predicted the creep crack growth rates for a broad spread of materials by:
L. Zhang et al. / Journal of Nuclear Materials 492 (2017) 32e40
35
Fig. 4. (a) The sliced specimen after test, and (b) fracture surface of the cracked specimen, showing the crack length at different testing conditions.
Fig. 5. The crack growth curve of 30% CW specimen, showing (a) pre-crack, (b) transition, (c) SCC growth and (d) creep crack growth.
. a_ ¼ 8:3 104 C *0:85 ε*f
Table 3 The SCC and creep CGRs at 450 Ce550 C. Specimen
Temperature
SCC
Creep
30%CW 30%CW 30%CW
450 C 500 C 550 C
5.3 107 mm/s 6.0 106 mm/s 5.9 105 mm/s
3.9 107 mm/s 5.1 106 mm/s 4.7 105 mm/s
(1)
While ε*f is the creep failure strain, and C* is the creep fracture mechanics parameter in MJm2h1. The ε*f is considered as 0.3 for Alloy 800H [16], while 0.03 is used in this study considering the cold work effect. C* characterizes the stress state at the creep crack
36
L. Zhang et al. / Journal of Nuclear Materials 492 (2017) 32e40
tip, and is calculated according to ASTM E1457. Nikbin [15] showed that relatively low dependence of creep crack growth rates on temperature was observed when characterized in terms of C*. As a result, the estimated creep crack growth rates by Equation (1) comparing with experimental data at different temperature are shown in Fig. 6, and demonstrates a good agreement (within a factor of 2). The CGRs obtained during the SCC test in SCW consists of two major parts, one part is the actual SCC crack growth rate which is induced by the synergistic effect of intergranular corrosion and load, and the other part is the contribution of high temperature creep. The measured SCC crack growth rate can be considered the total crack growth rate in SCW. From Table 3, creep induced cracking is a major contribution to the total crack growth rate of 30% cold worked Alloy 690 in SCW, and the contribution from creep cracking is about 80% of the total at each test temperatures, as shown in Fig. 7. The morphology of cracking has been examined by SEM, and micrographs are shown in Figs. 8 and 9. Small secondary cracks can be observed away from the main crack (Fig. 8(a)). The final creep crack tip is relatively sharp with a radius of less than 0.1 mm (Fig. 8(b)). Enlarging points C, D and E about 20e50 mm ahead of the crack tip in Fig. 8(b), creep crack (Fig. 8(c)) and discontinuous cavities along grain boundaries can be identified (Fig. 8(d) and e). Cavities preferentially form near the grain boundary carbides. The fracture surfaces in Fig. 9 show discontinuous cavities along the secondary cracks. Although covered with oxides, fracture surfaces and secondary cracks can be distinguished as intergranular, and the morphology is in consistent with Fig. 8(c), d and 8e. The IG cavities are evidence of creep attack during the propagation of main crack. Dynamic crack tip strain rate is one of the major ratedetermining factors that control the crack propagation in high temperature water environments [17]. A crack tip plastic stain equation for a growing crack was proposed by Gao and Hwang [18] to determine the crack tip strain rate in a work hardened material. According to their equation, the plastic strain could reach as high as
Fig. 7. Comparison of crack growth rate induced by Creep to the total in SCW.
5%, and the dynamic strain rate is also relatively high even at a distance of 50 mm ahead of the crack tip. During the SCC test, the load applied on the CT specimen produced a plastic zone along with a plastic strain field at the crack tip. Thus, the main IG crack propagation is promoted by the combined effect of strain rate at crack tip and corrosion of grain boundary. At the same time, dislocation movement is enhanced ahead of the main crack tip under the effect of plastic strain field. The enhanced dislocation movement tends to cause pile up and accumulation at the grain boundaries, especially at GB carbides. As a result, grain boundaries are weakened by cavities, and total IG CGR is accelerated. Arioka studied the SCC and creep behavior of cold-worked Alloy 690 in high temperature water environment [10,19,20]. His study indicated that creep plays an important role in IGSCC. The high density of cavities is formed because of the diffusion of CW induced
Fig. 6. Comparison of SCC and creep CGRs.
L. Zhang et al. / Journal of Nuclear Materials 492 (2017) 32e40
37
Fig. 8. The cross sectional view of the crack path, (a) main crack, (b) creep crack tip, (c,d,e) grain boundary ahead of the crack tip.
Fig. 9. The SEM images of cavity at the secondary cracks.
vacancies. According to Arioka, CW induced vacancies and hydrogen induced vacancies are the main sources of vacancy in the material. In this study, the material was highly cold worked (30%) and has large amounts of GB carbides, as shown in Fig. 1. Moreover, similar cavities around IG carbides were also observed in Figs. 8 and 9. Thus, the vacancies were mainly formed due to the CW effect and nucleation at GB carbides. When vacancies accumulate ahead of the crack tip, driven by stress gradient, cavities tend to form and affect the bond strength of grain boundaries. Consequently, the bond strength is assumed to be weakened and hence the SCC and creep crack growth rates increase. 4.2. Effect of temperature The SCC and creep growth rates of 30% CW Alloy 690 at different temperatures are shown in Fig. 10. Both SCC and creep growth rates increase with temperature, and follow an Arrhenius dependency. In the present research, the activation energy (QE) is calculated and marked in Fig. 10. The QE of SCC and creep is 228 kJ/mol and 222 kJ/ mol, respectively, which is comparable to the QE of oxide growth of nickel-base alloy (~200 kJ/mol) [21]. In previous SCC growth rate experiments in SCW by Peng et al. [9], the CGR of 20% CW 316L stainless steel decreased with the temperature increasing from 400 C to 450 C. The authors thought that the decrease of the CGR was due to the crack tip blunting induced by the rapid increase of the material oxidation rate and softening at higher temperature. In this study, the Alloy 690 has much lower corrosion/oxidation rate
Fig. 10. Temperature dependence on SCC and creep CGRs for 30%CW Alloy 690.
compared to 316L [4], thus the crack tip is less blunted, which could explain the opposite effect of temperature on SCC for Alloy 690 vs. 316L. Was has studied the SCC behavior of nickel-base alloys in pure, deaerated (<10 ppb O2) SCW with CERT experiments, and found that the QE is between 84 and 87 kJ/mol [4], which is much
38
L. Zhang et al. / Journal of Nuclear Materials 492 (2017) 32e40
lower than that in this study. As mentioned in the Introduction Section, we believe that there must be some differences between CERT and CT tests. Firstly, the forced strain rate applied by CERT is larger than that of constant K test for CT specimen. Secondly, the strain rate is very important in CERT test, which might give completely opposite results. Only enough low strain rate (107~108 s1) can obtained reliable data, especially for materials with high SCC resistance. Thirdly, the results obtained by CERT are usually given by more than one specimen and have relatively large scatter, while CT test is conducted with only one specimen with “on the fly” change, which has good consistency. Fourthly, the CERT is mainly for crack initiation and SCC susceptibility tests, focusing on the surface cracks, while CT test focuses on the long crack growth rate with a different crack tip chemistry from the bulk water. Above all, the different Q could be attributed to the above differences, and it is reasonable to believe that the results obtained by CERT test is more suitable for SCC initiation, while more suitable for SCC growth by CT test. The CGRs obtained in this study are plotted to compare with results of works by Arioka et al. [19] and Gao et al. [22], as shown in Fig. 11 (data in this study are plotted with red solid points). The SCC and creep CGRs of 30% CW Alloy 690 in this work are below that of 20% CW Alloy TT690 obtained in PWR primary water and gas environments in Arioka's work [19] in Fig. 11(a), and the creep crack growth of Alloy 718 in Gao's work [22] in Fig. 11(b), but most of the studies show a similar slope vs. temperature, except the dotted line for alloy 718 in oxygenas shown in Fig. 11(b). The major reason can be attributed to the difference of load in the experiments: literature data were obtained at a higher K value of 30 MPa√m in PWR water vs. 40 MPa√m in gas. The difference in the degree of cold work has less effect if comparing the increase of yield strength in Alloy 690. The similarity in slope for most CGR responses in Fig. 11 represent a similar temperature dependency and similar activation energy (QE), consistent with the results of this study. The different slope for Alloy 718 in oxygen may have been caused by the enhancement of oxygen on crack growth [22]. While for the lower crack growth rates compared with Arioka's data, Bruemmer [12,13] and Arioka [10] reported that slower SCC growth rates were observed in annealed Alloy 690 than TT690 in PWR primary water, mainly due to the grain boundary carbides. Thus, the difference could be attributed to the material conditions, especially the difference in IG carbide coverage. The concept of an environmental correction factor Fen [23] can be borrowed to quantify the corrosion effect of a high temperature
environment on the stress corrosion crack growth rate of a material,
Fen ¼ CGRSCC CGRcreep:
(2)
The temperature dependence of Fen is shown in Fig. 12. All Fen at 450e550 C lie well below 1.5, indicating a relatively mild corrosion effect of SCW environment on crack growth of cold worked Alloy 690. Fig. 11(a) shows remarkably higher SCC crack growth rates in PWR environment compared with gas environment, suggesting an additional environmental enhancement effect on SCC growth in addition to the cavity-related creep mechanism. However, the SCW environmental enhancement effect on the SCC behavior of 30% CW Alloy 690 in this study is negligible, and can be taken as in high temperature inert environment (although there was undoubtedly some trace levels of oxygen and water vapor). SCC growth is the result of the combined effect of mechanical driving damage and chemical or/and electrochemical reactions at the crack tip [24]. In this study, the 30% CW material has a relatively high creep CGRs (107 - 105 mm/s) in SCW, indicating a high mechanical driving damage during the crack growth. This high creep growth rate gives
Fig. 12. The temperature dependence of Fen for 30%CW Alloy 690.
Fig. 11. Comparison of SCC and creep CGRs for 30% CW Alloy 690 in this study with (a) 20% CW TT690 by Arioka et al. [19] and (b) Inconel 718 Alloy by Gao et al. [22].
L. Zhang et al. / Journal of Nuclear Materials 492 (2017) 32e40
less time for the SCW to corrode the crack tip, and creep dominates the crack growth. This can be the major reason to explain the low Fen (<1.5) of Alloy 690 in SCW environment. 4.3. Mechanistic analysis The slip-oxidation model, also called Ford-Andresen (F-A) model [24] is among the most widely accepted model, and has been widely used for predicting CGRs of various materials, including stainless steels, nickel alloys and low alloy steels in LWR environment. In this model, the failure due to SCC exhibits IG cracking characteristics, and oxide film on the crack front is ruptured by dislocation movement at crack tip, which exposes bare metal at the crack tip to the corrosive environment, with subsequent oxidation reforming the oxide film again (repassivaion), and the slipoxidation process repeats. The results in this study confirmed that formation of cavity ahead of crack tip by creep effect plays an important role in the SCC growth rates for CW Alloy 690 in higher temperature environments. The testing temperatures were much higher and the SCW environments were more corrosive than the operating conditions in current LWRs. The higher temperature leads to a greater effect of creep on IG cracking, and the mechanism of SCC in SCW becomes more complex, and includes the combined effect of creep and corrosion. Inspired by the film-induced cleavage model [25,26], a creepinduced extension model (as shown in Fig. 13) is proposed to explain the SCC behavior of CW Alloy 690 in SCW. First, the crack grows by the bare surface dissolution and oxide growth, and this process is governed by the same rate-determining steps as described in the F-A model. At the same time, for a long crack as illustrated in Fig. 13(a), there exists a plastic zone at the crack tip if a load is applied, and the plastic zone creates a strain field and strain gradient ahead of the crack tip. For highly cold worked materials, the yield strength is dramatically higher, and there is a high strain gradient and high density of vacancies at the crack tip. Thus, the CW induced vacancies in the grain interior are activated and tend to diffuse into the grain boundaries ahead of the crack tip under the effect of strain gradient. The accumulation of vacancies tends to
39
form cavities at the grain boundary. As a result, the grain boundary is weakened. In the creep-induced extension model described in Fig. 13(b), the crack tip will readily penetrate a distance a* into the weakened grain boundary ahead of the crack tip when the oxide ruptures, and this extension increases the average crack growth rate. The extent of this additional creep-induced crack advance is governed by the creep susceptibility of the material at the environment temperature. For 30% CW Alloy 690 in this study, the creep CGRs are in the range of 107 e105 mm/s at 450e550 C. Thus, the creep-induced extension a* is much higher than the bare surface dissolution and oxide growth in slip-oxidation model. However, in this creep-induced extension mechanism there are several factors that require clarification, such as oxygen diffusion through the cavities into the matrix and its effect on creep and oxidation, and vacancy diffusion coefficient ahead of the crack tip. Thus, further work is necessary to complete this model.
5. Conclusion The crack growth rates of 30% CW Alloy 690 in SCW and inert gas environments at temperatures between 450 and 550 C were evaluated to study the relative effect of creep and SCC on the total crack growth rate. The following conclusions were drawn: 1) CW Alloy 690 exhibits a relatively high susceptibility to IG cracking in SCW, and creep contribute the largest portion to the total crack growth rate. 2) Extensive cavity formation was observed ahead of the crack tip, and cavities tend to nucleate near GB carbides. Cavities weaken the bond of grain boundary and accelerate the IG crack growth. 3) Temperature strongly affects the CGRs in SCW. The temperature dependence follows an Arrhenius dependency with an activation energy QE of about 225 kJ/mol for both SCC and creep crack growth. 4) A creep-induced extension mechanism based on F-A model can be used to explain the SCC behavior of CW Alloy 690 in SCW.
Fig. 13. A revised F-A model illustrating the SCC mechanism of Alloy 690 in SCW.
40
L. Zhang et al. / Journal of Nuclear Materials 492 (2017) 32e40
Acknowledgment Thanks to the microstructural characterization of Instrumental Analysis Center of SJTU. References [1] Generation IV, F., Technology roadmap update for generation IV nuclear energy systems, OECD Nuclear Energy Agency for the Generation IV International Forum, 2014. [2] L. Zhang, Y. Bao, R. Tang, Selection and corrosion evaluation tests of candidate SCWR fuel cladding materials, Nucl. Eng. Des. 249 (8) (2012) 180e187. [3] X.J. Liu, X. Cheng, Coupled thermal-hydraulics and neutron-physics analysis of SCWR with mixed spectrum core, Prog. Nucl. Energy 52 (7) (2010) 640e647. [4] G.S. Was, et al., Corrosion and stress corrosion cracking in supercritical water, J. Nucl. Mater. 371 (1e3) (2007) 176e201. [5] R. Novotny, et al., Stress corrosion cracking susceptibility of austenitic stainless steels in supercritical water conditions, J. Nucl. Mater. 409 (2) (2011) 117e123. [6] Z. Shen, et al., The effect of temperature on the SSRT behavior of austenitic stainless steels in SCW, J. Nucl. Mater. 454 (1e3) (2014) 274e282. [7] H. Je, A. Kimura, Stress corrosion cracking susceptibility of oxide dispersion strengthened ferritic steel in supercritical pressurized water dissolved with different hydrogen and oxygen contents, Corros. Sci. 78 (2014) 193e199. [8] S. Teysseyre, et al., Effect of irradiation on stress corrosion cracking in supercritical water, J. Nucl. Mater. 371 (1e3) (2007) 107e117. [9] Q. Peng, et al., Stress corrosion crack growth in type 316 stainless steel in supercritical water, Corrosion 63 (11) (2007) 1033e1041. [10] K. Arioka, et al., Dependence of stress corrosion cracking of alloy 690 on temperature, cold work, and carbide precipitationdrole of diffusion of vacancies at crack tips, CORROSION 67 (3) (2011), 035006e35011-035006-18. [11] P.L. Andresen, M.M. Morra, Stress corrosion cracking of stainless steels and nickel alloys in high-temperature water, CORROSION 64 (1) (2008) 15e29. [12] Bruemmer, S.M., et al., Cold-work effects on stress corrosion crack growth in Alloy 690 tubing and plate materials, in 17th International Conference on Environmental Degradation of Materials in Nuclear Power Systems-water
Reactors: Ottawa, Canada. [13] S.M. Bruemmer, et al., High-resolution characterizations of grain boundary damage and stress corrosion cracks in cold-rolled alloy 690, in: 15th International Conference on Environmental Degradation of Materials in Nuclear Power Systems-water Reactors, 2011 (Hoboken, US). [14] D. Du, et al., SCC crack growth rate of cold worked 316L stainless steel in PWR environment, J. Nucl. Mater. 456 (2015) 228e234. [15] K.M. Nikbin, An engineering approach to the prediction of creep crack growth, J. Eng. Mater. Technol. 108 (2) (1986) 186e191. [16] G.A. Webster, Creep crack growth, in: I. Milne, R.O. Ritchie, B. Karihaloo, I. Milne, R.O. Ritchie, B. Karihaloo (Eds.), Comprehensive Structural Integrity, Pergamon Oxford, 2003, pp. 241e271. [17] P.L. Andresen, M.M. Morra, Stress corrosion cracking of stainless steels and nickel alloys in high-temperature water, CORROSION 64 (1) (2008) 15e29. [18] Y.C. Gao, K.C. Huang, Elastic-plastic Field in Steady Crack Growth in a Strainhardening Material, 1981 (Cannes, France). [19] K. Arioka, et al., Degradation of alloy 690 after relatively short times, CORROSION 72 (10) (2016) 1252e1268. [20] K. Arioka, W.R. Whitney award lecture: change in bonding strength at grain boundaries before long-term SCC initiation, CORROSION 71 (4) (2014) 403e419, 2014. [21] G.S. Was, S. Teysseyre, Z. Jiao, Corrosion of austenitic alloys in supercritical water, CORROSION 62 (11) (2006) 989e1005. [22] M. Gao, D.J. Dwyer, R.P. Wei, Chemical and microstructural aspect of creep crack growth in INCONEL 718 Alloy, in: Superalloys 718, 625, 706 and Various Derivatives, The minerals, Metals and Materials Society, 1994. [23] M. Higuchi, et al., Revised and New Proposal of Environmental Fatigue Life Correction Factor (Fen) for Carbon and Low-Alloy Steels and Nickel Base Alloys in LWR Water Environments, 2006, pp. 93e102. [24] P.L. Andresen, Emerging issues and fundamental processes in environmental cracking in hot water (Reprinted from Proceedings of the CORROSION/2007 research topical symposium “Advances in Environmentally Assisted Cracking”, 2007), CORROSION 64 (5) (2008) 439e464. [25] A. Paskin, et al., Environmentally induced crack nucleation and brittle fracture, Acta Metall. 30 (9) (1982) 1781e1788. [26] K. Sieradzki, R.C. Newman, Brittle behavior of ductile metals during stresscorrosion cracking, Philos. Mag. A 51 (1) (1985) 95e132.