- Email: [email protected]
Effect of short-range ordering on stress corrosion cracking susceptibility of Alloy 600 studied by electron and neutron diffraction
Available online at www.sciencedirect.com
ScienceDirect Acta Materialia 83 (2015) 507–515 www.elsevier.com/locate/actamat
Effect of short-range ordering on stress corrosion cracking susceptibility of Alloy 600 studied by electron and neutron diffraction ⇑
Young Suk Kim, Wan Young Maeng and Sung Soo Kim Korea Atomic Energy Research Institute, Daeduk-daero 1045, Yuseong, Daejeon 305-353, Republic of Korea Received 26 July 2014; revised 6 October 2014; accepted 7 October 2014 Available online 15 November 2014
Abstract—Slow strain rate tests (SSRTs) were conducted on low-temperature mill-annealed Alloy 600 tubes at 250 and 360 °C in water with either 8 ppm or below 10 ppb of dissolved oxygen (DO). A special tensile specimen design with a hump was employed for these tests. During SSRT in 360 °C water, stress corrosion cracking (SCC) or intergranular cracking of Alloy 600 is enhanced at DO contents below 10 ppb but suppressed at 8 ppm DO. The SCC susceptibility of Alloy 600 is observed to be related to the degree of lattice contraction by short-range ordering, which is enhanced in the presence of hydrogen. By analyzing electron and neutron diffraction patterns before and after SSRT in 360 °C water, for the first ˚ being formed in Alloy 600 during SSRT in 360 °C time, definitive evidence is presented for the short-range ordered phase with a d-spacing of 2.1 A water, which is manifested by the forbidden reflections at the 1/3{4 2 2} positions in h1 1 1i selected area diffraction patterns. Ó 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. Keywords: Stress corrosion cracking (SCC); Intergranular (IG) cracking; Lattice contraction; Short-range ordering (SRO); Alloy 600
1. Introduction Stress corrosion cracking (SCC) of structural materials made of Alloy 600, such as steam generator tubes, heater sleeves and various nozzles, is a hot issue in the nuclear industry. Given the main features of SCC where cracks grow primarily by intergranular (IG) cracking [1,2], several hypotheses have been proposed for several decades assuming that IG cracking is an extrinsic phenomenon that is governed by degradation of grain boundaries by either corrosion [3,4] or oxidation [5,6]. In contrast, IG cracking of austenitic Ni–Cr–Fe alloys was observed to occur even under an inert atmosphere such as argon, where no corrosion or oxidation of the grain boundary is involved [7,8]. These observations suggest that IG cracking may be an intrinsic phenomenon which is against the current hypotheses related to SCC of metals. Another misunderstanding of SCC is the residual tensile stress, which has been regarded as the primary factor leading to earlier SCC failures of the welds and nozzles of Alloy 600. However, if this residual stress is the cause of SCC, the IG cracking of the welds should occur more frequently at the beginning of life, and not after a long incubation of 100,000 h [9]; the reason for this remains unresolved so far. In fact, considering that the residual stress is relieved by 80–100% within 24 h at a temperature of 300 °C and above [10], the residual tensile
⇑ Corresponding
author. Tel.: +82 42 868 2359; fax: +82 42 868 8549; e-mail: [email protected]
stress cannot account for the occurrence of SCC after such long incubation times. Nonetheless, this observation that a long incubation time is required prior to SCC initiation in the structural components made of Alloy 600 suggests that stresses are internally built up in Alloy 600 by an internal factor during a sufficiently long incubation time that are high enough to exceed the critical stress for IG cracking. In other words, IG cracking of Alloy 600 is dictated by an internal mechanism, which is our unique idea. As a matter of fact, Marucco [11] observed that the lattice contractions of up to 0.12% occurred in several nickel alloys due to short-range ordering (SRO) upon aging at 475 °C for 32,000 h. Nevertheless, she did not appreciate the effect of the lattice contractions by SRO on SCC or IG cracking. When the grains are contracted due to SRO, however, the grain boundary would be subjected to tensile stresses that lead to mechanistic cracking of grain boundaries or IG cracking or build-up of higher strains or dislocations in the regions adjacent to the grain boundaries in the case where IG cracking is absent. Support of this suggestion is provided by Kohara and Kuzynski’s observation [12] that a tricrystal of CuAu broke into three pieces by IG cracking 35 seconds after an ordering reaction at 350 °C, and Hou et al.’s observations [13,14] of higher strains or a higher number density of dislocations in the regions adjacent to the grain boundaries. Thus, the more enhanced IG cracking of austenitic Ni–Cr–Fe alloys in water than in argon [7] and the hydrogen-enhanced IG cracking in Alloy 600 [15–17] appear to be related to the enhanced ordering by hydrogen, as already revealed by Flanagan et al. [18].
http://dx.doi.org/10.1016/j.actamat.2014.10.009 1359-6462/Ó 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
508
Y.S. Kim et al. / Acta Materialia 83 (2015) 507–515
Considering that atoms in materials keep changing their positions by either plastic deformation or thermal diffusion, SRO leading to lattice contractions cannot be avoided in alloys with several solute atoms during their use, especially at high temperatures, which is the cause of SCC of structural materials, according to our proposal. The aim of this work is to demonstrate that SCC of Alloy 600 is related to SRO. To this end, slow strain rate tests (SSRTs) were conducted in recirculating water at either 250 or 360 °C on an Alloy 600 tube which had been subjected to low temperature mill annealing. To show if the environmental effect on IG cracking of an Alloy 600 tube is related to a hydrogen-enhanced SRO, the content of dissolved oxygen in water was varied from below 10 ppb to 8 ppm (i.e. by about three orders of magnitude). The lattice spacing was determined before and after SSRT using neutron diffraction to correlate IG cracking susceptibility of Alloy 600 with SRO. Direct evidence for the short-range ordered phase being formed in Alloy 600 during SSRT in 360 °C water is presented by comparing electron and neutron diffraction patterns before and after SSRT.
2. Experimental procedures An Alloy 600 tube mill annealed at 960 °C for 10 min followed by water quenching was used in this study. This low-temperature mill-annealed (LTMA) tube has a typical chemical composition of Alloy 600 with a low carbon
content of 0.01 wt.%, as shown in Table 1, which is in accordance with ASTM B-167. SSRT was conducted on tensile specimens with a hump using a constant extension rate test (CERT) machine and a water loop at a strain rate of 2.5 107 s1. The hump specimens, as shown in Fig. 1, were used to enhance IG cracking in the LTMA Alloy 600, which is herein called Alloy 600. During SSRT, the specimens were exposed to either 250 or 360 °C in pressurized water to 200 atmospheres and containing either 8 ppm or less than 10 ppb of dissolved oxygen (DO). More detailed procedures are given elsewhere [19]. After SSRT, the fracture surfaces of the specimens were examined by scanning electron microscopy (SEM) to observe the cracking pattern and determine the fraction of IG cracking. Transmission electron microscopy (TEM) thin foils were obtained by a focused ion beam from a fractured surface of the tensile specimen and were examined in a JEM 2010 microscope (JEOL, Japan) at 200 kV to obtain images and electron diffraction patterns at the leading edge of a crack that was grown either along the grain boundary or into the grains. The diffraction patterns were obtained from the [1 1 1], [1 1 0] and [1 1 2] zone axes to confirm if diffuse peaks generated by short-range order is a fact or an artifact. A neutron diffraction analysis was conducted using a neutron beam ˚ on the gauge sections wavelength of 1.837225 ± 0.000034 A of the tensile specimens before and after SSRT. The instrumental resolution of the diffractometer was Dd/dffi0.04%. The change in peak position (peak shift) was used to determine a change in lattice parameter and its sensitivity was less than 1/10th of the instrumental resolution.
Table 1. Chemical composition of an LTMA Alloy 600 tube used. Alloy 600
Ni
Cr
Fe
Mn
C
Cu
Si
S
wt.%
75.1
15.4
8.0
0.3
0.01
0.2
1.0
0.001
Fig. 1. Dimension of tensile specimens with a hump used for slow strain rate tests.
Y.S. Kim et al. / Acta Materialia 83 (2015) 507–515
3. Results and discussions Fig. 2 shows the load vs. displacement curves for Alloy 600 in 360 °C water with dissolved oxygen content. In water with 8 ppm DO, Alloy 600 showed strain hardening after yielding and thus a higher ductility of 12.3%. In contrast, in water with DO below 10 ppb, no strain hardening was observed after yielding, indicating the formation of cracks just after yielding, which led to a lower ductility of 5.5%. For the sake of reproducibility, another SSRT was conducted not only at 360 °C but also at 250 °C in water with different DO contents. As shown in Fig. 3, the same effect of the DO content on the cracking susceptibility of Alloy 600 in 360 °C water, as shown in Fig. 2, was observed, confirming that the results of Fig. 2 are reproducible. Note that even in water with DO below 10 ppb, Alloy 600 showed a higher ductility at 250 °C, which was in contrast with the lower ductility at 360 °C. To confirm if the cracks formed after yielding in 360 °C water with DO below 10 ppb are intergranular, the fracture surfaces were examined using an SEM after the tests. As shown in Fig. 4, 100% IG cracking was observed on the fracture surfaces of the Alloy 600 tube in water with DO below 10 ppb, confirming the preferential formation of IG cracks after yielding. In contrast, in water with 8 ppm DO where the Alloy 600 tube showed a higher ductility (Fig. 2), no IG cracking (or transgranular
o 8 ppm DO 360 C <10 ppb DO
Load (N)
2000
1500
1000 -7
2.5x10 /s
500
0 0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Displacement (mm) Fig. 2. Load vs. displacement of an LTMA Alloy 600 tube with dissolved oxygen content in water at 360 °C upon tensile testing at a strain rate of 2.5 107 s1.
400 2 3 4
Stress (MPa)
300
o
250 C, < 10 ppb DO o 360 C, 8 ppm DO o 360 C, <10 ppb DO
o
360 C o
250 C
-7
2.5x10 /s 200
100 o
360 C
0
2 3
4
0
2
4
6
8
10
12
14
16
Strain(%) Fig. 3. Tensile behavior of an LTMA Alloy 600 tube with tensile testing temperature and dissolved oxygen content in 360 °C water at a strain rate of 2.5 107 s1.
509
cracking) was observed at the surface of the specimen, as shown in Fig. 5a and b. However, the inner regions were full of IG cracks, as shown in Fig. 5c. This observation shows that a higher DO content in water suppresses IG cracking, especially at the surface. This fact agrees with earlier observations [20,21] that Alloy 600 is immune to SCC at 288 °C in oxygen containing water unless there is a creviced geometry. Nevertheless, IG cracking in the inner regions of the Alloy 600 specimen reveals that IG cracking is an inherent phenomenon governed by an intrinsic factor such as SRO, which occurs independent of the environment. The inner regions being full of IG cracks, as shown in Fig. 5c, agree with Angeliu et al.’s observation [7] of a higher number of IG cracks in austenitic Ni–Cr–Fe alloys preferentially in the inner regions of the fracture surface than on its perimeter. The observations shown indicate that IG cracking of Alloy 600 was enhanced in water with DO below 10 ppb, leading to a lower ductility, and suppressed in water with 8 ppm DO, leading to a higher ductility. Furthermore, IG cracks were observed in the inner regions of the Alloy 600 tube after SSRT at 360 °C, independent of the DO content in the water, revealing that IG cracking is an intrinsic phenomenon independent of the environment. This observation agrees with IG cracking of austenitic Ni–16Cr–9Fe even in argon [7,22]. In other words, SCC of Alloy 600 was enhanced in water with DO below 10 ppb and suppressed in water with 8 ppm DO, which is in contrast with the internal oxidation mechanism suggested by Scott and co-workers [5,6]. Given our previous observation [23] that SCC of Alloy 600 is governed by the degree of lattice contraction due to SRO accompanied under reactor operating conditions, it is expected that the SCC susceptibility of Alloy 600 with DO content in 360 °C water, as shown in Figs. 2 and 3, is related to the degree of lattice contraction accompanied during SSRT. For experimental evidence, the lattice spacing was determined using neutron diffraction before and after SSRT in 360 °C water. As shown in Table 2, lattice contraction occurred in Alloy 600 during SSRT in 360 °C water independent of the DO content. Furthermore, the lattice contraction of the (2 0 0) plane was approximately two times higher in water with DO below 10 ppb than in water with 8 ppm DO, which is in accordance with our expectation. Here, the lattice contraction is termed a fractional change of the d-spacing of either the (1 1 1) or (2 0 0) planes after SSRT in 360 °C water when compared to those before SSRT. The results of Table 2 indicates indirectly that the ordering transformation occurring during SSRT in 360 °C water is independent of the DO content and that lattice contraction due to SRO is the cause of IG cracks formed in the inner regions of Alloy 600 during SSRT in water, even with 8 ppm DO. The enhanced lattice contraction in water with DO below 10 ppb, as shown in Table 2, generates the lower ductility of Alloy 600 and 100% IG cracking, as shown in Figs. 2–4. Considering that Alloy 600 absorbs hydrogen released as a result of its oxidation in 360 °C water with DO below 10 ppb [15–17], the enhanced lattice contraction in water with DO below 10 ppb seems to be related to the hydrogen effect, which enhances SRO. The hydrogen-enhanced lattice contraction of Alloy 600 agrees with the observations of Fukai and co-workers [24,25], showing the remarkable lattice contraction of Ni in hydrogen gas of 3 GPa at high temperatures below 1000 °C [24], and a small lattice contraction of Fe (0.042%) in
510
Y.S. Kim et al. / Acta Materialia 83 (2015) 507–515
a
1mm c
b
100 µm
50µm
Fig. 4. Fracture surfaces of an LTMA Alloy 600 tube after SSRT in 360 °C water with dissolved oxygen content below 10 ppb showing 100% IG cracking: (a) the whole fracture surface; (b, c) the enlarged pictures of the inner regions of the fracture surface.
Fig. 5. Fracture surfaces of an LTMA Alloy 600 tube after SSRT in 360 °C water with 8 ppm dissolved oxygen content showing transgranular cracking at the surface and IG cracking in the inner regions of the specimen: (a) the whole fracture surface, (b) the enlarged picture of the region A shown in (a) and (c) the enlarged picture of the region B shown in (a).
Y.S. Kim et al. / Acta Materialia 83 (2015) 507–515
diffraction using TEM and by neutron diffraction. To examine electron diffraction patterns at the leading edge of the crack tip, TEM thin foils were obtained using a focused ion beam directly from the fracture surfaces of Alloy 600 after SSRT in 360 °C water with different DO contents. Fig. 6 shows a TEM image of an IG crack formed at the outer region of the fracture surface, and electron diffraction patterns along the [1 1 1] zone which were obtained from Alloy 600 after SSRT in 360 °C water with DO below 10 ppb both near the leading edge of an IG crack and a few micrometers away from it, respectively. Note that the leading edge of an IG crack was directly exposed to the environment during SSRT. The IG crack was observed to grow along the grain boundary, in accordance with the SEM results shown in Fig. 4. No superlattice reflections were observed, as shown in Fig. 6, indicating no formation of long-range order (LRO) during SSRT in 360 °C water. Instead, the 1/3{4 2 2} reflections that are normally forbidden by a face centered cubic (fcc) lattice were observed, which seems to be indicative of short-range order formed
Table 2. The lattice contraction percent of Alloy 600 after SSRT in 360 °C water with DO content varying from 8 ppm to less than 10 ppb. Alloy 600
DO content in primary water
The amount of lattice contraction (%)
8 ppm <10 ppb
d(111)
d(200)
0.037 ± 0.004 0.034 ± 0.004
0.020 ± 0.004 0.036 ± 0.004
511
hydrogen gas of 3.3 GPa at a low temperature of 302 °C [25]. Consequently, the observations in Figs. 2–5 and Table 2 indicate that SCC in Alloy 600 occurs by hydrogenenhanced SRO, which agrees excellently with Beachem’s observations [26] that SCC in quenched and tempered steels occurred by hydrogen-assisted cracking mechanisms although he could not describe what deformation mechanism leads to IG cracking in steels. Direct evidence for SRO in Alloy 600 during SSRT in 360 °C water was provided for the first time by electron
a
600 nm b
c
Fig. 6. (a) TEM image of an intergranular crack formed at the fracture surface of Alloy 600 after SSRT in 360 °C water with DO below 10 ppb, (b) the diffraction pattern in the [1 1 1] zone just below the crack tip of the IG crack and (c) the diffraction pattern in the [1 1 1] zone a few micrometers away from the crack tip.
512
Y.S. Kim et al. / Acta Materialia 83 (2015) 507–515
in Alloy 600 as shown in Fig. 6b and c. Note that the forbidden reflections obtained just beneath the leading edge of the crack tip were bright (Fig. 6b), and look like a kind of superlattice reflection; a few micrometers away from it, however, these 1/3{4 2 2} reflections became diffuse (Fig. 6c). Fig. 7 shows a TEM image of a crack formed at the outer region of the fracture surface and the diffraction pattern in the [1 1 1] zone after SSRT in 360 °C water with 8 ppm DO. As shown in Fig. 7a, a transgranular crack was observed to form at the perimeter of the fracture surface, as evidenced by the SEM results shown in Fig. 5a and b. The diffuse 1/3{4 2 2} reflections were observed just beneath the tip of a transgranular crack after SSRT in 360 °C water with 8 ppm DO, as shown in Fig. 7b. Note that the 1/3{4 2 2} reflections observed after SSRT in 360 °C water with different DO contents were not observed before SSRT, indicating the formation of the 1/3{4 2 2} reflections during SSRT in 360 °C water. To show if the
a
20 00nm m
b
Fig. 7. (a) TEM image of a transgranular crack formed at the fracture surface and (b) the diffraction pattern in the [1 1 1] zone of Alloy 600 after SSRT in 360 °C water with 8 ppm DO.
diffuse reflections shown in Figs. 6 and 7 are a fact, the diffraction patterns were obtained not only from the [1 1 1] zone axis but also from the [1 1 0] and [1 1 2] zone axes. As shown in Fig. 8, the diffuse reflections corresponding to 1/3{4 2 2} and 1/2{1 3 1} in an fcc lattice were observed in the [1 1 1] and [1 1 2] zone axes, respectively, but none of them was observed in the [1 1 0] zone axis except twinning and its streaks at the {1 1 1} twin plane. The same 1/3{4 2 2} forbidden reflections as shown in Fig. 8 had also been observed in an aged Ni–25 at.% Cr [27], a hydrogenated 310 stainless steel deformed at 295 K [28], an Al–5Zn–1Mg aged at room temperature (RT) for 7 days [29], an austenitic Fe–Mn–Al alloy after tensile deformation at 77 K to 0.02% [30], Hg3In2Te6 crystals grown by the Bridgman method [31] and a 316 stainless steel after tensile deformation at RT in 70 MPa H2 gas [32]. Note that the bright 1/3{4 2 2} reflections as shown in Fig. 6b appear in the last three cases (i.e. Fe–Mn–Al deformed at 77 K, Hg3In2Te6 crystals and 316 stainless steel deformed in 70 MPa H2 gas at RT). Besides, when the [1 1 1] diffraction pattern was tilted by 19° around the [1 1 0] axis until the [1 1 2] diffraction pattern is obtained, the 1/2{1 3 1} forbidden reflections were observed, as shown in Fig. 9b, which had also been observed in an aged Ni– 28.4Cr–5.1Fe [33] and aged Al–5Zn–1Mg [29], Hg3In2Te6 crystals [31], and a 316 stainless steel deformed in 70 MPa H2 gas [32]. Thus, it is evident that the 1/3{4 2 2} forbidden reflections are a real fact but not caused by double diffraction. For concrete evidence that the diffuse or bright 1/3{4 2 2} reflections are related to the short-range ordered phase, neutron diffraction was conducted on the gauge sections of Alloy 600 before and after SSRT in 360 °C water with different DO contents. As shown in Fig. 9, despite the absence of sharp superlattice reflections corresponding to the LRO phase, the new diffraction peaks, which should not be normally observed in fcc metals, appeared next to the diffraction peak of the (1 1 1) plane in Alloy 600 after SSRT in 360 °C water with different DO contents whose ˚ . Note that the intensity of d-spacing corresponds to 2.1 A the new diffraction peak was very low before SSRT but increased to some extent after SSRT, showing that the new diffraction peak was formed additionally during SSRT in 360 °C water. The d-spacing of the short-range ordered phase formed on the 1/3{4 2 2} plane shown in Figs. 6 ˚ , using the meaand 7 was also determined to be 2.1 A sured values of the d-spacing of the various lattice planes of Alloy 600 that were determined by neutron diffraction. This excellent agreement between the d-spacing of the SRO phase determined by TEM and by neutron diffraction shows that the diffuse or bright 1/3{4 2 2} reflections are direct evidence for the formation of the short-range ordered phase. Consequently, considering that hydrogen absorbed in Alloy 600 is localized primarily at the crack tip [34] or in the locally deformed regions [15–17], the bright 1/3{4 2 2} reflections just below the leading edge of an IG crack (Fig. 6b) seem to be related to hydrogen-enhanced SRO, which is evidenced by a higher lattice contraction of Alloy 600 in water with DO below 10 ppb, as shown in Table 2. Support is lent to the hydrogen-enhanced ordering transformation by Flanagan et al.’s observation [18] that the long-range ordered phase was formed in Pd3Mn upon annealing only under 5 MPa of H2 but the short-range ordered phase was instead formed in vacuum upon annealing, even at a higher temperature. However, it remains yet
Y.S. Kim et al. / Acta Materialia 83 (2015) 507–515
a
513
b
1/3(422)
½(131)
[111]
[112] c
[110] Fig. 8. Diffraction patterns taken along the [1 1 1], [1 1 2] and [1 1 0] zone axes of Alloy 600 after SSRT in 360 °C water with DO below 10 ppb.
Alloy 600 tube SSRT Before test 8 ppm DO <10ppb DO
10000
Intensity (cps)
5000
1/3(422)
400
200 0.25
0.20
0.15
0.10
d-spacing (nm) Fig. 9. Neutron diffraction patterns of the gauge sections of the Alloy 600 tube before and after SSRT in 360 °C water with different DO contents.
to be resolved how hydrogen accelerates the ordering transformation in Alloy 600. In short, the lattice contraction occurs because of the accompanying SRO during SSRT in 360 °C water, which was enhanced in water with DO below 10 ppb due to the hydrogen effect. This enhanced lattice contraction due to the hydrogen-enhanced SRO is the cause of 100% IG cracking during SSRT in 360 °C water containing DO below 10 ppb, as shown in Fig. 4. In water with 8 ppm DO with a lesser lattice contraction, IG cracks were formed only in the inner regions and not at the surface due to little
hydrogen-enhanced SRO, as shown in Fig. 5. With anisotropic contractions of the lattice planes in grains due to SRO [35], the grain boundary is subjected to tensile stresses. The larger the lattice contraction becomes, the larger the tensile stress developed at the grain boundary. This tensile stress induced at the grain boundary due to the lattice contraction leads to higher strains in the regions adjacent to the grain boundary when compared to those in the interior regions of grains [13,14]. When the tensile stress at the grain boundary arising from the lattice contractions is large enough to exceed the grain boundary strength, grain boundary cracking or IG cracking occurs spontaneously without applied stress: evidence for this was provided by Kohara and Kuzynski’s observation [12]. IG cracking of CuAu by the ordering transformation even without applied stress [12] reveals that the lattice contraction due to the ordering transformation generates large tensile stresses at the grain boundary. When the tensile stress at the grain boundary induced by SRO is rather low due to a small lattice contraction of below 0.1% as shown in Table 2, IG cracking occurs only with a tensile stress applied. To provide supportive evidence, tensile tests were conducted at RT on a model alloy of Ni3Fe (76.18 wt.% Ni–23.66 wt.% Fe) with 90 ppm carbon and 60 ppm Cr, which was subjected to either water quenching or furnace cooling after solution annealing at 1050 °C for 1 h. The furnace-cooled (FC)-Ni3Fe exhibited a lattice contraction of 0.001% to 0.019%, depending on orientation, when compared to the water-quenched (WQ)-Ni3Fe, as shown in Table 3. Given this observation shown in Table 3, it is evident that the former has SRO to some extent and the latter has disorder or a low degree
514
Y.S. Kim et al. / Acta Materialia 83 (2015) 507–515
Table 3. Lattice contractions of the Ni3Fe model alloy with cooling rate. Cooling rate
WQ-Ni3Fe FC-Ni3Fe
The amount of lattice contraction (%) d(200)
d(220)
d(222)
0 0.001 ± 0.004
0 0.018 ± 0.004
0 0.019 ± 0.004
of SRO. Upon tensile tests at RT with a strain rate of 6 105 s1, as expected, the former indeed showed IG fracture or cracking but the latter exhibited ductile fracture with fine dimples, as shown in Fig. 10. Fig. 10 evidently indicates that IG cracking is related to the lattice contraction by SRO. In other words, albeit the small amount, less than 0.1%, the lattice contraction due to the formation of the short-range ordered phase is the cause of IG cracking in Alloy 600 and the FC-Ni3Fe upon tensile testing at elevated temperatures and at room temperature, respectively. Note that these results agree with our previous observations [23,35] that the SCC susceptibility of Alloy 600 is dictated by the degree of accompanying lattice contraction under the reactor operating conditions. Consequently, we propose that SCC of Alloy 600 is related to lattice contractions due to SRO which is enhanced by hydrogen, i.e. hydrogen-enhanced SRO. Given that SRO occurs by the lattice diffusion of atoms, the rate of SRO becomes slower at a lower temperature of 250 °C when compared to that at 360 °C, resulting in a smaller lattice contraction at 250 °C. In other words, the lower degree of lattice contraction due to the slow rate of SRO at 250 °C suppresses IG cracking after yielding despite the hydrogen effect, leading to a higher ductility in 250 °C
a
b
10 µ m Fig. 11. Ductile fracture of Alloy 600 after SSRT in 250 °C water with DO below 10 ppb.
water with DO below 10 ppb, as shown in Fig. 3. Ductile fracture with dimples was observed on the fracture surfaces of Alloy 600 deformed in 250 °C water with DO below 10 ppb, as shown in Fig. 11. This observation indicates that IG cracking or SCC of Alloy 600 is governed primarily by the degree of SRO. In other words, the cracking mechanism is SRO leading to lattice contraction. 4. Implication of the proposed mechanism SRO always occurs at high temperatures such as reactor operating conditions in all austenitic Fe–Cr–Ni alloys, including nickel alloys of Alloys 600/690/182/82/152/52 and Fe-base alloys of 304 or 316 stainless steels. Given our observations that SCC of Alloy 600 is related to hydrogen-enhanced SRO, SCC of Alloy 690 or 304/316 stainless steels would also be related to SRO and their SCC susceptibility will depend on the rate of SRO or the degree of lattice contraction arising from SRO. Given this hypothesis, the SCC characteristics of Alloy 600 would be the same as those of others. In fact, there are many similarities between SCC of nickel alloys including Alloy 600 and that of austenitic stainless steels: both nickel alloys and austenitic stainless steels have shown (1) IG cracking, even in an argon environment with no corrosion or oxidation of grain boundaries [7,8,22]; (2) the formation of planar dislocations at the tip of a leading crack after SCC tests in simulated reactor operating conditions [36–38], which are characteristic of short-range order [39,40]; and (3) the enhanced IG cracking with decreasing strain rate [41–43]. Considering that the enhanced IG cracking with decreasing strain rate had been observed in body centered cubic (bcc) steels during SCC and hydrogen-assisted cracking [26], IG cracking in fcc and bcc metals during SCC and hydrogen-assisted cracking (or hydrogen embrittlement) may be governed by a single phenomenon such as hydrogen-enhanced SRO. 5. Conclusions
Fig. 10. Fracture surfaces of the WQ- and FC-Ni3Fe after tensile tests at 6 105 s1 and room temperature.
SCC or IG cracking of an LTMA Alloy 600 tube was enhanced at DO below 10 ppb during SSRT in 360 °C water but not in 250 °C water. However, it was suppressed during SSRT in 360 °C water with 8 ppm DO, which is in contrast with the internal oxidation mechanism. Enhanced
Y.S. Kim et al. / Acta Materialia 83 (2015) 507–515
IG cracking and the lower ductility of Alloy 600 at DO concentrations below 10 ppb during SSRT in 360 °C water are related to two times higher lattice contractions of the (2 0 0) plane at DO below 10 ppb than that at 8 ppm DO. This observation is perfectly consistent with our previous results that SCC of Alloy 600 is governed by lattice contraction due to SRO. A supplementary experiment confirms that a small lattice contraction of below 0.02% due to SRO caused IG cracking under applied tensile stress. The higher lattice contraction of Alloy 600 during SSRT in 360 °C water with DO below 10 ppb is related to hydrogenenhanced SRO, leading to the formation of a higher degree of short-range order. Just beneath the leading edge of an IG crack in Alloy 600 in 360 °C water with DO below 10 ppb, the bright 1/3{4 2 2} reflections were observed but the diffuse reflections appeared a few micrometers away from the leading edge of the IG crack or ahead the transgranular crack grown in 360 °C water with 8 ppm DO. Using electron and neutron diffraction patterns obtained from the fracture surfaces or the gauge sections of Alloy 600 after SSRT in 360 °C water, it is demonstrated for the first time that the 1/3{4 2 2} reflections correspond to the short-range ˚ . A higher ductility ordered phase with a d-spacing of 2.1 A and suppressed SCC of Alloy 600 in 250 °C water with DO below 10 ppb seem to be related to the slow rate of SRO at 250 °C. In conclusion, we propose that SCC in Alloy 600 is related to hydrogen-enhanced SRO and the dominant cracking mechanism is SRO leading to lattice contraction. Acknowledgments This work has been carried out as a part of the Nuclear R&D Program sponsored by the Korean Ministry of Science, ICT and Future Planning. Special thanks are due to S.S. Lee, who conducted neutron diffraction experiments at KAERI, to Y.W. Kim, who obtained electron diffraction patterns using TEM and to E.J. Kim, who made TEM samples using FIB. YSK is particularly grateful for the comments by Michael Kaufman (Professor, Department Head in Colorado School of Mines).
References [1] [2] [3] [4] [5]
[6]
[7] [8] [9] [10]
D. Van Rooyen, Corrosion 31 (1975) 327. S.M. Bruemmer, G.S. Was, J. Nucl. Mater. 216 (1994) 348. R.C. Newman, Corrosion 50 (1994) 682. P.L. Andresen, in: Corrosion 96, NACE: 1996. Paper #96258. P.M. Scott, M. Le Calvar, in: Proceedings of 6th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, San Diego, CA, American Society for Metals, Materials Park, OH, 1993, p. 657. P.M. Scott, in: Proceedings of 9th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Newport Beach, CA, TMS, Warrendale, PA, 1999, p. 3. T.M. Angeliu, D.J. Paraventi, G.S. Was, Corrosion 51 (1995) 837. T. Onchi, K. Dohi, N. Soneda, M. Navas, M.L. Castano, J. Nucl. Mater. 340 (2005) 219. P.M. Scott, PWSCC of Nickel Base Alloys and Mitigation in PWRs, presented at the INL Seminar on SCC in LWRs, Idaho Falls, Idaho, 19–20 March 2013, p. 32. K.F. Amouzouvi, R.C. Styles, L. Clegg, D.P. McCooeye, L. Zhao, A.K. Koul, in: Proceedings of 6th International Conference on Shot Peening, San Francisco, CA, 1996. p. 166.
515
[11] A.J. Marucco, Phys. IV France 03 (1993) C7-331. [12] S. Kohara, G.C. Kuzynski, Acta Met. 2 (1956) 221. [13] J. Hou, Q.J. Peng, Z.P. Lu, T. Shoji, J.Q. Wang, E.-H. Han, et al., Corros. Sci. 53 (2011) 1137. [14] J. Hou, Q.J. Peng, J.Q. Wang, E.-H. Han, W. Ke, Corros. Sci. 53 (2011) 2956. [15] N. Totsuka, E. Lumarska, G. Gragnolino, Z. SzklarskaSmialowska, Corrosion 43 (1987) 505. [16] F.H. Hua, P.B. Rebak, in: Proceedings of the Second International Conference on Environment-Induced Cracking of Metals (EICM-2), Banff, AB, 2004. [17] N. Totsuka, Z. Szklarska-Smialowska, Corrosion 43 (1987) 734. [18] T.B. Flanagan, A.P. Craft, T. Kuji, K. Baba, Y. Sakamoto, Scr. Metall. 20 (1986) 1745. [19] W.Y. Maeng, M.S. Choi, U.C. Kim, J. Mater. Sci. 39 (2004) 655. [20] R.A. Page, Corrosion 39 (1983) 409. [21] R.A. Page, A. McMinn, Metall. Trans. A 17A (1986) 877. [22] J.K. Sung, G.S. Was, Corrosion 47 (1991) 824. [23] Y.S. Kim, S.S. Kim, D.W. Kim, in: Proceedings of 16th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Asheville, NC, TMS, Warrendale, PA, 2013. [24] Y. Fukai, Y. Shizuku, Y. Kurokawa, J. Alloys Compd. 329 (2001) 195. [25] T. Hiroi, Y. Fukai, K. Mori, J. Alloys Compd. 404–406 (2005) 252. [26] C.D. Beachem, Metall. Trans. 3 (1972) 437. [27] A. Marucco, B. Nath, J. Mater. Sci. 23 (1988) 2107. [28] D.P. Abraham, C.J. Altstetter, Metall. Mater. Trans. A 26A (1995) 2859. [29] X.J. Jiang, B. Noble, G. Waterloo, J. Tafto, Metall. Mater. Trans. A 31A (2000) 339. [30] K. Sato, M. Ichinose, Y. Hirotsu, Y. Inoue, ISIJ Int. 29 (1989) 868. [31] L. Luo, W. Jie, Y. Xu, Y. He, L. Xu, L. Fu, R. Soc. Chem. 16 (2014) 5073. [32] B.H. Choe, Private communication, 2012. [33] A. Marucco, Mater. Sci. Eng., A 189 (1994) 267. [34] F. Foct, O.D. Bourvier, T. Magnin, Metall. Mater. Trans. A 31A (2000) 2025. [35] S.S. Kim, D.W. Kim, Y.S. Kim, Met. Mater. Int. 19 (2013) 969. [36] T. Couvant, L. Legras, A. Herbelin, A. Musienko, G. Ilevbare, D. Delafosse, G. Cailletaud, J. Hickling, in: Proceedings of 14th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Virginia Beach, VA, TMS, Warrendale, PA, 2009, p. 182. [37] T. Nakamura, M. Koshiishi, T. Torimaru, Y. Kitsunai, K. Takakura, K. Nakata, M. Ando, Y. Ishiyama, A. Jenssen, in: Proceedings of 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems, Whistler, BC, TMS, Warrendale, PA, 2007. [38] F. Vaillant, J.M. Boursier, L. Legras, B. Yrieix, E. Lemaire, J. Champredonde, C. Amzallag, in: Proceedings of 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems, Whistler, BC, TMS, Warrendale, PA, 2007. [39] V. Gerold, H.P. Karnthaler, Acta Met. 37 (1989) 2177. [40] G. Thomas, Acta Met. 11 (1963) 1369. [41] M.P. Manahan, R. Kohli, J. Santucci, P. Sipush, Nucl. Eng. Des. 113 (1989) 297. [42] U. Ehrnsten, T. Saukkonen, W. Karlsen, H. Hanninen, in: Proceedings of 14th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Virginia Beach, VA, TMS, Warrendale, PA, 2009, p. 910. [43] G.S. Was, K. Lian, Corrosion 54 (1998) 675.
ScienceDirect Acta Materialia 83 (2015) 507–515 www.elsevier.com/locate/actamat
Effect of short-range ordering on stress corrosion cracking susceptibility of Alloy 600 studied by electron and neutron diffraction ⇑
Young Suk Kim, Wan Young Maeng and Sung Soo Kim Korea Atomic Energy Research Institute, Daeduk-daero 1045, Yuseong, Daejeon 305-353, Republic of Korea Received 26 July 2014; revised 6 October 2014; accepted 7 October 2014 Available online 15 November 2014
Abstract—Slow strain rate tests (SSRTs) were conducted on low-temperature mill-annealed Alloy 600 tubes at 250 and 360 °C in water with either 8 ppm or below 10 ppb of dissolved oxygen (DO). A special tensile specimen design with a hump was employed for these tests. During SSRT in 360 °C water, stress corrosion cracking (SCC) or intergranular cracking of Alloy 600 is enhanced at DO contents below 10 ppb but suppressed at 8 ppm DO. The SCC susceptibility of Alloy 600 is observed to be related to the degree of lattice contraction by short-range ordering, which is enhanced in the presence of hydrogen. By analyzing electron and neutron diffraction patterns before and after SSRT in 360 °C water, for the first ˚ being formed in Alloy 600 during SSRT in 360 °C time, definitive evidence is presented for the short-range ordered phase with a d-spacing of 2.1 A water, which is manifested by the forbidden reflections at the 1/3{4 2 2} positions in h1 1 1i selected area diffraction patterns. Ó 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. Keywords: Stress corrosion cracking (SCC); Intergranular (IG) cracking; Lattice contraction; Short-range ordering (SRO); Alloy 600
1. Introduction Stress corrosion cracking (SCC) of structural materials made of Alloy 600, such as steam generator tubes, heater sleeves and various nozzles, is a hot issue in the nuclear industry. Given the main features of SCC where cracks grow primarily by intergranular (IG) cracking [1,2], several hypotheses have been proposed for several decades assuming that IG cracking is an extrinsic phenomenon that is governed by degradation of grain boundaries by either corrosion [3,4] or oxidation [5,6]. In contrast, IG cracking of austenitic Ni–Cr–Fe alloys was observed to occur even under an inert atmosphere such as argon, where no corrosion or oxidation of the grain boundary is involved [7,8]. These observations suggest that IG cracking may be an intrinsic phenomenon which is against the current hypotheses related to SCC of metals. Another misunderstanding of SCC is the residual tensile stress, which has been regarded as the primary factor leading to earlier SCC failures of the welds and nozzles of Alloy 600. However, if this residual stress is the cause of SCC, the IG cracking of the welds should occur more frequently at the beginning of life, and not after a long incubation of 100,000 h [9]; the reason for this remains unresolved so far. In fact, considering that the residual stress is relieved by 80–100% within 24 h at a temperature of 300 °C and above [10], the residual tensile
⇑ Corresponding
author. Tel.: +82 42 868 2359; fax: +82 42 868 8549; e-mail: [email protected]
stress cannot account for the occurrence of SCC after such long incubation times. Nonetheless, this observation that a long incubation time is required prior to SCC initiation in the structural components made of Alloy 600 suggests that stresses are internally built up in Alloy 600 by an internal factor during a sufficiently long incubation time that are high enough to exceed the critical stress for IG cracking. In other words, IG cracking of Alloy 600 is dictated by an internal mechanism, which is our unique idea. As a matter of fact, Marucco [11] observed that the lattice contractions of up to 0.12% occurred in several nickel alloys due to short-range ordering (SRO) upon aging at 475 °C for 32,000 h. Nevertheless, she did not appreciate the effect of the lattice contractions by SRO on SCC or IG cracking. When the grains are contracted due to SRO, however, the grain boundary would be subjected to tensile stresses that lead to mechanistic cracking of grain boundaries or IG cracking or build-up of higher strains or dislocations in the regions adjacent to the grain boundaries in the case where IG cracking is absent. Support of this suggestion is provided by Kohara and Kuzynski’s observation [12] that a tricrystal of CuAu broke into three pieces by IG cracking 35 seconds after an ordering reaction at 350 °C, and Hou et al.’s observations [13,14] of higher strains or a higher number density of dislocations in the regions adjacent to the grain boundaries. Thus, the more enhanced IG cracking of austenitic Ni–Cr–Fe alloys in water than in argon [7] and the hydrogen-enhanced IG cracking in Alloy 600 [15–17] appear to be related to the enhanced ordering by hydrogen, as already revealed by Flanagan et al. [18].
http://dx.doi.org/10.1016/j.actamat.2014.10.009 1359-6462/Ó 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
508
Y.S. Kim et al. / Acta Materialia 83 (2015) 507–515
Considering that atoms in materials keep changing their positions by either plastic deformation or thermal diffusion, SRO leading to lattice contractions cannot be avoided in alloys with several solute atoms during their use, especially at high temperatures, which is the cause of SCC of structural materials, according to our proposal. The aim of this work is to demonstrate that SCC of Alloy 600 is related to SRO. To this end, slow strain rate tests (SSRTs) were conducted in recirculating water at either 250 or 360 °C on an Alloy 600 tube which had been subjected to low temperature mill annealing. To show if the environmental effect on IG cracking of an Alloy 600 tube is related to a hydrogen-enhanced SRO, the content of dissolved oxygen in water was varied from below 10 ppb to 8 ppm (i.e. by about three orders of magnitude). The lattice spacing was determined before and after SSRT using neutron diffraction to correlate IG cracking susceptibility of Alloy 600 with SRO. Direct evidence for the short-range ordered phase being formed in Alloy 600 during SSRT in 360 °C water is presented by comparing electron and neutron diffraction patterns before and after SSRT.
2. Experimental procedures An Alloy 600 tube mill annealed at 960 °C for 10 min followed by water quenching was used in this study. This low-temperature mill-annealed (LTMA) tube has a typical chemical composition of Alloy 600 with a low carbon
content of 0.01 wt.%, as shown in Table 1, which is in accordance with ASTM B-167. SSRT was conducted on tensile specimens with a hump using a constant extension rate test (CERT) machine and a water loop at a strain rate of 2.5 107 s1. The hump specimens, as shown in Fig. 1, were used to enhance IG cracking in the LTMA Alloy 600, which is herein called Alloy 600. During SSRT, the specimens were exposed to either 250 or 360 °C in pressurized water to 200 atmospheres and containing either 8 ppm or less than 10 ppb of dissolved oxygen (DO). More detailed procedures are given elsewhere [19]. After SSRT, the fracture surfaces of the specimens were examined by scanning electron microscopy (SEM) to observe the cracking pattern and determine the fraction of IG cracking. Transmission electron microscopy (TEM) thin foils were obtained by a focused ion beam from a fractured surface of the tensile specimen and were examined in a JEM 2010 microscope (JEOL, Japan) at 200 kV to obtain images and electron diffraction patterns at the leading edge of a crack that was grown either along the grain boundary or into the grains. The diffraction patterns were obtained from the [1 1 1], [1 1 0] and [1 1 2] zone axes to confirm if diffuse peaks generated by short-range order is a fact or an artifact. A neutron diffraction analysis was conducted using a neutron beam ˚ on the gauge sections wavelength of 1.837225 ± 0.000034 A of the tensile specimens before and after SSRT. The instrumental resolution of the diffractometer was Dd/dffi0.04%. The change in peak position (peak shift) was used to determine a change in lattice parameter and its sensitivity was less than 1/10th of the instrumental resolution.
Table 1. Chemical composition of an LTMA Alloy 600 tube used. Alloy 600
Ni
Cr
Fe
Mn
C
Cu
Si
S
wt.%
75.1
15.4
8.0
0.3
0.01
0.2
1.0
0.001
Fig. 1. Dimension of tensile specimens with a hump used for slow strain rate tests.
Y.S. Kim et al. / Acta Materialia 83 (2015) 507–515
3. Results and discussions Fig. 2 shows the load vs. displacement curves for Alloy 600 in 360 °C water with dissolved oxygen content. In water with 8 ppm DO, Alloy 600 showed strain hardening after yielding and thus a higher ductility of 12.3%. In contrast, in water with DO below 10 ppb, no strain hardening was observed after yielding, indicating the formation of cracks just after yielding, which led to a lower ductility of 5.5%. For the sake of reproducibility, another SSRT was conducted not only at 360 °C but also at 250 °C in water with different DO contents. As shown in Fig. 3, the same effect of the DO content on the cracking susceptibility of Alloy 600 in 360 °C water, as shown in Fig. 2, was observed, confirming that the results of Fig. 2 are reproducible. Note that even in water with DO below 10 ppb, Alloy 600 showed a higher ductility at 250 °C, which was in contrast with the lower ductility at 360 °C. To confirm if the cracks formed after yielding in 360 °C water with DO below 10 ppb are intergranular, the fracture surfaces were examined using an SEM after the tests. As shown in Fig. 4, 100% IG cracking was observed on the fracture surfaces of the Alloy 600 tube in water with DO below 10 ppb, confirming the preferential formation of IG cracks after yielding. In contrast, in water with 8 ppm DO where the Alloy 600 tube showed a higher ductility (Fig. 2), no IG cracking (or transgranular
o 8 ppm DO 360 C <10 ppb DO
Load (N)
2000
1500
1000 -7
2.5x10 /s
500
0 0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Displacement (mm) Fig. 2. Load vs. displacement of an LTMA Alloy 600 tube with dissolved oxygen content in water at 360 °C upon tensile testing at a strain rate of 2.5 107 s1.
400 2 3 4
Stress (MPa)
300
o
250 C, < 10 ppb DO o 360 C, 8 ppm DO o 360 C, <10 ppb DO
o
360 C o
250 C
-7
2.5x10 /s 200
100 o
360 C
0
2 3
4
0
2
4
6
8
10
12
14
16
Strain(%) Fig. 3. Tensile behavior of an LTMA Alloy 600 tube with tensile testing temperature and dissolved oxygen content in 360 °C water at a strain rate of 2.5 107 s1.
509
cracking) was observed at the surface of the specimen, as shown in Fig. 5a and b. However, the inner regions were full of IG cracks, as shown in Fig. 5c. This observation shows that a higher DO content in water suppresses IG cracking, especially at the surface. This fact agrees with earlier observations [20,21] that Alloy 600 is immune to SCC at 288 °C in oxygen containing water unless there is a creviced geometry. Nevertheless, IG cracking in the inner regions of the Alloy 600 specimen reveals that IG cracking is an inherent phenomenon governed by an intrinsic factor such as SRO, which occurs independent of the environment. The inner regions being full of IG cracks, as shown in Fig. 5c, agree with Angeliu et al.’s observation [7] of a higher number of IG cracks in austenitic Ni–Cr–Fe alloys preferentially in the inner regions of the fracture surface than on its perimeter. The observations shown indicate that IG cracking of Alloy 600 was enhanced in water with DO below 10 ppb, leading to a lower ductility, and suppressed in water with 8 ppm DO, leading to a higher ductility. Furthermore, IG cracks were observed in the inner regions of the Alloy 600 tube after SSRT at 360 °C, independent of the DO content in the water, revealing that IG cracking is an intrinsic phenomenon independent of the environment. This observation agrees with IG cracking of austenitic Ni–16Cr–9Fe even in argon [7,22]. In other words, SCC of Alloy 600 was enhanced in water with DO below 10 ppb and suppressed in water with 8 ppm DO, which is in contrast with the internal oxidation mechanism suggested by Scott and co-workers [5,6]. Given our previous observation [23] that SCC of Alloy 600 is governed by the degree of lattice contraction due to SRO accompanied under reactor operating conditions, it is expected that the SCC susceptibility of Alloy 600 with DO content in 360 °C water, as shown in Figs. 2 and 3, is related to the degree of lattice contraction accompanied during SSRT. For experimental evidence, the lattice spacing was determined using neutron diffraction before and after SSRT in 360 °C water. As shown in Table 2, lattice contraction occurred in Alloy 600 during SSRT in 360 °C water independent of the DO content. Furthermore, the lattice contraction of the (2 0 0) plane was approximately two times higher in water with DO below 10 ppb than in water with 8 ppm DO, which is in accordance with our expectation. Here, the lattice contraction is termed a fractional change of the d-spacing of either the (1 1 1) or (2 0 0) planes after SSRT in 360 °C water when compared to those before SSRT. The results of Table 2 indicates indirectly that the ordering transformation occurring during SSRT in 360 °C water is independent of the DO content and that lattice contraction due to SRO is the cause of IG cracks formed in the inner regions of Alloy 600 during SSRT in water, even with 8 ppm DO. The enhanced lattice contraction in water with DO below 10 ppb, as shown in Table 2, generates the lower ductility of Alloy 600 and 100% IG cracking, as shown in Figs. 2–4. Considering that Alloy 600 absorbs hydrogen released as a result of its oxidation in 360 °C water with DO below 10 ppb [15–17], the enhanced lattice contraction in water with DO below 10 ppb seems to be related to the hydrogen effect, which enhances SRO. The hydrogen-enhanced lattice contraction of Alloy 600 agrees with the observations of Fukai and co-workers [24,25], showing the remarkable lattice contraction of Ni in hydrogen gas of 3 GPa at high temperatures below 1000 °C [24], and a small lattice contraction of Fe (0.042%) in
510
Y.S. Kim et al. / Acta Materialia 83 (2015) 507–515
a
1mm c
b
100 µm
50µm
Fig. 4. Fracture surfaces of an LTMA Alloy 600 tube after SSRT in 360 °C water with dissolved oxygen content below 10 ppb showing 100% IG cracking: (a) the whole fracture surface; (b, c) the enlarged pictures of the inner regions of the fracture surface.
Fig. 5. Fracture surfaces of an LTMA Alloy 600 tube after SSRT in 360 °C water with 8 ppm dissolved oxygen content showing transgranular cracking at the surface and IG cracking in the inner regions of the specimen: (a) the whole fracture surface, (b) the enlarged picture of the region A shown in (a) and (c) the enlarged picture of the region B shown in (a).
Y.S. Kim et al. / Acta Materialia 83 (2015) 507–515
diffraction using TEM and by neutron diffraction. To examine electron diffraction patterns at the leading edge of the crack tip, TEM thin foils were obtained using a focused ion beam directly from the fracture surfaces of Alloy 600 after SSRT in 360 °C water with different DO contents. Fig. 6 shows a TEM image of an IG crack formed at the outer region of the fracture surface, and electron diffraction patterns along the [1 1 1] zone which were obtained from Alloy 600 after SSRT in 360 °C water with DO below 10 ppb both near the leading edge of an IG crack and a few micrometers away from it, respectively. Note that the leading edge of an IG crack was directly exposed to the environment during SSRT. The IG crack was observed to grow along the grain boundary, in accordance with the SEM results shown in Fig. 4. No superlattice reflections were observed, as shown in Fig. 6, indicating no formation of long-range order (LRO) during SSRT in 360 °C water. Instead, the 1/3{4 2 2} reflections that are normally forbidden by a face centered cubic (fcc) lattice were observed, which seems to be indicative of short-range order formed
Table 2. The lattice contraction percent of Alloy 600 after SSRT in 360 °C water with DO content varying from 8 ppm to less than 10 ppb. Alloy 600
DO content in primary water
The amount of lattice contraction (%)
8 ppm <10 ppb
d(111)
d(200)
0.037 ± 0.004 0.034 ± 0.004
0.020 ± 0.004 0.036 ± 0.004
511
hydrogen gas of 3.3 GPa at a low temperature of 302 °C [25]. Consequently, the observations in Figs. 2–5 and Table 2 indicate that SCC in Alloy 600 occurs by hydrogenenhanced SRO, which agrees excellently with Beachem’s observations [26] that SCC in quenched and tempered steels occurred by hydrogen-assisted cracking mechanisms although he could not describe what deformation mechanism leads to IG cracking in steels. Direct evidence for SRO in Alloy 600 during SSRT in 360 °C water was provided for the first time by electron
a
600 nm b
c
Fig. 6. (a) TEM image of an intergranular crack formed at the fracture surface of Alloy 600 after SSRT in 360 °C water with DO below 10 ppb, (b) the diffraction pattern in the [1 1 1] zone just below the crack tip of the IG crack and (c) the diffraction pattern in the [1 1 1] zone a few micrometers away from the crack tip.
512
Y.S. Kim et al. / Acta Materialia 83 (2015) 507–515
in Alloy 600 as shown in Fig. 6b and c. Note that the forbidden reflections obtained just beneath the leading edge of the crack tip were bright (Fig. 6b), and look like a kind of superlattice reflection; a few micrometers away from it, however, these 1/3{4 2 2} reflections became diffuse (Fig. 6c). Fig. 7 shows a TEM image of a crack formed at the outer region of the fracture surface and the diffraction pattern in the [1 1 1] zone after SSRT in 360 °C water with 8 ppm DO. As shown in Fig. 7a, a transgranular crack was observed to form at the perimeter of the fracture surface, as evidenced by the SEM results shown in Fig. 5a and b. The diffuse 1/3{4 2 2} reflections were observed just beneath the tip of a transgranular crack after SSRT in 360 °C water with 8 ppm DO, as shown in Fig. 7b. Note that the 1/3{4 2 2} reflections observed after SSRT in 360 °C water with different DO contents were not observed before SSRT, indicating the formation of the 1/3{4 2 2} reflections during SSRT in 360 °C water. To show if the
a
20 00nm m
b
Fig. 7. (a) TEM image of a transgranular crack formed at the fracture surface and (b) the diffraction pattern in the [1 1 1] zone of Alloy 600 after SSRT in 360 °C water with 8 ppm DO.
diffuse reflections shown in Figs. 6 and 7 are a fact, the diffraction patterns were obtained not only from the [1 1 1] zone axis but also from the [1 1 0] and [1 1 2] zone axes. As shown in Fig. 8, the diffuse reflections corresponding to 1/3{4 2 2} and 1/2{1 3 1} in an fcc lattice were observed in the [1 1 1] and [1 1 2] zone axes, respectively, but none of them was observed in the [1 1 0] zone axis except twinning and its streaks at the {1 1 1} twin plane. The same 1/3{4 2 2} forbidden reflections as shown in Fig. 8 had also been observed in an aged Ni–25 at.% Cr [27], a hydrogenated 310 stainless steel deformed at 295 K [28], an Al–5Zn–1Mg aged at room temperature (RT) for 7 days [29], an austenitic Fe–Mn–Al alloy after tensile deformation at 77 K to 0.02% [30], Hg3In2Te6 crystals grown by the Bridgman method [31] and a 316 stainless steel after tensile deformation at RT in 70 MPa H2 gas [32]. Note that the bright 1/3{4 2 2} reflections as shown in Fig. 6b appear in the last three cases (i.e. Fe–Mn–Al deformed at 77 K, Hg3In2Te6 crystals and 316 stainless steel deformed in 70 MPa H2 gas at RT). Besides, when the [1 1 1] diffraction pattern was tilted by 19° around the [1 1 0] axis until the [1 1 2] diffraction pattern is obtained, the 1/2{1 3 1} forbidden reflections were observed, as shown in Fig. 9b, which had also been observed in an aged Ni– 28.4Cr–5.1Fe [33] and aged Al–5Zn–1Mg [29], Hg3In2Te6 crystals [31], and a 316 stainless steel deformed in 70 MPa H2 gas [32]. Thus, it is evident that the 1/3{4 2 2} forbidden reflections are a real fact but not caused by double diffraction. For concrete evidence that the diffuse or bright 1/3{4 2 2} reflections are related to the short-range ordered phase, neutron diffraction was conducted on the gauge sections of Alloy 600 before and after SSRT in 360 °C water with different DO contents. As shown in Fig. 9, despite the absence of sharp superlattice reflections corresponding to the LRO phase, the new diffraction peaks, which should not be normally observed in fcc metals, appeared next to the diffraction peak of the (1 1 1) plane in Alloy 600 after SSRT in 360 °C water with different DO contents whose ˚ . Note that the intensity of d-spacing corresponds to 2.1 A the new diffraction peak was very low before SSRT but increased to some extent after SSRT, showing that the new diffraction peak was formed additionally during SSRT in 360 °C water. The d-spacing of the short-range ordered phase formed on the 1/3{4 2 2} plane shown in Figs. 6 ˚ , using the meaand 7 was also determined to be 2.1 A sured values of the d-spacing of the various lattice planes of Alloy 600 that were determined by neutron diffraction. This excellent agreement between the d-spacing of the SRO phase determined by TEM and by neutron diffraction shows that the diffuse or bright 1/3{4 2 2} reflections are direct evidence for the formation of the short-range ordered phase. Consequently, considering that hydrogen absorbed in Alloy 600 is localized primarily at the crack tip [34] or in the locally deformed regions [15–17], the bright 1/3{4 2 2} reflections just below the leading edge of an IG crack (Fig. 6b) seem to be related to hydrogen-enhanced SRO, which is evidenced by a higher lattice contraction of Alloy 600 in water with DO below 10 ppb, as shown in Table 2. Support is lent to the hydrogen-enhanced ordering transformation by Flanagan et al.’s observation [18] that the long-range ordered phase was formed in Pd3Mn upon annealing only under 5 MPa of H2 but the short-range ordered phase was instead formed in vacuum upon annealing, even at a higher temperature. However, it remains yet
Y.S. Kim et al. / Acta Materialia 83 (2015) 507–515
a
513
b
1/3(422)
½(131)
[111]
[112] c
[110] Fig. 8. Diffraction patterns taken along the [1 1 1], [1 1 2] and [1 1 0] zone axes of Alloy 600 after SSRT in 360 °C water with DO below 10 ppb.
Alloy 600 tube SSRT Before test 8 ppm DO <10ppb DO
10000
Intensity (cps)
5000
1/3(422)
400
200 0.25
0.20
0.15
0.10
d-spacing (nm) Fig. 9. Neutron diffraction patterns of the gauge sections of the Alloy 600 tube before and after SSRT in 360 °C water with different DO contents.
to be resolved how hydrogen accelerates the ordering transformation in Alloy 600. In short, the lattice contraction occurs because of the accompanying SRO during SSRT in 360 °C water, which was enhanced in water with DO below 10 ppb due to the hydrogen effect. This enhanced lattice contraction due to the hydrogen-enhanced SRO is the cause of 100% IG cracking during SSRT in 360 °C water containing DO below 10 ppb, as shown in Fig. 4. In water with 8 ppm DO with a lesser lattice contraction, IG cracks were formed only in the inner regions and not at the surface due to little
hydrogen-enhanced SRO, as shown in Fig. 5. With anisotropic contractions of the lattice planes in grains due to SRO [35], the grain boundary is subjected to tensile stresses. The larger the lattice contraction becomes, the larger the tensile stress developed at the grain boundary. This tensile stress induced at the grain boundary due to the lattice contraction leads to higher strains in the regions adjacent to the grain boundary when compared to those in the interior regions of grains [13,14]. When the tensile stress at the grain boundary arising from the lattice contractions is large enough to exceed the grain boundary strength, grain boundary cracking or IG cracking occurs spontaneously without applied stress: evidence for this was provided by Kohara and Kuzynski’s observation [12]. IG cracking of CuAu by the ordering transformation even without applied stress [12] reveals that the lattice contraction due to the ordering transformation generates large tensile stresses at the grain boundary. When the tensile stress at the grain boundary induced by SRO is rather low due to a small lattice contraction of below 0.1% as shown in Table 2, IG cracking occurs only with a tensile stress applied. To provide supportive evidence, tensile tests were conducted at RT on a model alloy of Ni3Fe (76.18 wt.% Ni–23.66 wt.% Fe) with 90 ppm carbon and 60 ppm Cr, which was subjected to either water quenching or furnace cooling after solution annealing at 1050 °C for 1 h. The furnace-cooled (FC)-Ni3Fe exhibited a lattice contraction of 0.001% to 0.019%, depending on orientation, when compared to the water-quenched (WQ)-Ni3Fe, as shown in Table 3. Given this observation shown in Table 3, it is evident that the former has SRO to some extent and the latter has disorder or a low degree
514
Y.S. Kim et al. / Acta Materialia 83 (2015) 507–515
Table 3. Lattice contractions of the Ni3Fe model alloy with cooling rate. Cooling rate
WQ-Ni3Fe FC-Ni3Fe
The amount of lattice contraction (%) d(200)
d(220)
d(222)
0 0.001 ± 0.004
0 0.018 ± 0.004
0 0.019 ± 0.004
of SRO. Upon tensile tests at RT with a strain rate of 6 105 s1, as expected, the former indeed showed IG fracture or cracking but the latter exhibited ductile fracture with fine dimples, as shown in Fig. 10. Fig. 10 evidently indicates that IG cracking is related to the lattice contraction by SRO. In other words, albeit the small amount, less than 0.1%, the lattice contraction due to the formation of the short-range ordered phase is the cause of IG cracking in Alloy 600 and the FC-Ni3Fe upon tensile testing at elevated temperatures and at room temperature, respectively. Note that these results agree with our previous observations [23,35] that the SCC susceptibility of Alloy 600 is dictated by the degree of accompanying lattice contraction under the reactor operating conditions. Consequently, we propose that SCC of Alloy 600 is related to lattice contractions due to SRO which is enhanced by hydrogen, i.e. hydrogen-enhanced SRO. Given that SRO occurs by the lattice diffusion of atoms, the rate of SRO becomes slower at a lower temperature of 250 °C when compared to that at 360 °C, resulting in a smaller lattice contraction at 250 °C. In other words, the lower degree of lattice contraction due to the slow rate of SRO at 250 °C suppresses IG cracking after yielding despite the hydrogen effect, leading to a higher ductility in 250 °C
a
b
10 µ m Fig. 11. Ductile fracture of Alloy 600 after SSRT in 250 °C water with DO below 10 ppb.
water with DO below 10 ppb, as shown in Fig. 3. Ductile fracture with dimples was observed on the fracture surfaces of Alloy 600 deformed in 250 °C water with DO below 10 ppb, as shown in Fig. 11. This observation indicates that IG cracking or SCC of Alloy 600 is governed primarily by the degree of SRO. In other words, the cracking mechanism is SRO leading to lattice contraction. 4. Implication of the proposed mechanism SRO always occurs at high temperatures such as reactor operating conditions in all austenitic Fe–Cr–Ni alloys, including nickel alloys of Alloys 600/690/182/82/152/52 and Fe-base alloys of 304 or 316 stainless steels. Given our observations that SCC of Alloy 600 is related to hydrogen-enhanced SRO, SCC of Alloy 690 or 304/316 stainless steels would also be related to SRO and their SCC susceptibility will depend on the rate of SRO or the degree of lattice contraction arising from SRO. Given this hypothesis, the SCC characteristics of Alloy 600 would be the same as those of others. In fact, there are many similarities between SCC of nickel alloys including Alloy 600 and that of austenitic stainless steels: both nickel alloys and austenitic stainless steels have shown (1) IG cracking, even in an argon environment with no corrosion or oxidation of grain boundaries [7,8,22]; (2) the formation of planar dislocations at the tip of a leading crack after SCC tests in simulated reactor operating conditions [36–38], which are characteristic of short-range order [39,40]; and (3) the enhanced IG cracking with decreasing strain rate [41–43]. Considering that the enhanced IG cracking with decreasing strain rate had been observed in body centered cubic (bcc) steels during SCC and hydrogen-assisted cracking [26], IG cracking in fcc and bcc metals during SCC and hydrogen-assisted cracking (or hydrogen embrittlement) may be governed by a single phenomenon such as hydrogen-enhanced SRO. 5. Conclusions
Fig. 10. Fracture surfaces of the WQ- and FC-Ni3Fe after tensile tests at 6 105 s1 and room temperature.
SCC or IG cracking of an LTMA Alloy 600 tube was enhanced at DO below 10 ppb during SSRT in 360 °C water but not in 250 °C water. However, it was suppressed during SSRT in 360 °C water with 8 ppm DO, which is in contrast with the internal oxidation mechanism. Enhanced
Y.S. Kim et al. / Acta Materialia 83 (2015) 507–515
IG cracking and the lower ductility of Alloy 600 at DO concentrations below 10 ppb during SSRT in 360 °C water are related to two times higher lattice contractions of the (2 0 0) plane at DO below 10 ppb than that at 8 ppm DO. This observation is perfectly consistent with our previous results that SCC of Alloy 600 is governed by lattice contraction due to SRO. A supplementary experiment confirms that a small lattice contraction of below 0.02% due to SRO caused IG cracking under applied tensile stress. The higher lattice contraction of Alloy 600 during SSRT in 360 °C water with DO below 10 ppb is related to hydrogenenhanced SRO, leading to the formation of a higher degree of short-range order. Just beneath the leading edge of an IG crack in Alloy 600 in 360 °C water with DO below 10 ppb, the bright 1/3{4 2 2} reflections were observed but the diffuse reflections appeared a few micrometers away from the leading edge of the IG crack or ahead the transgranular crack grown in 360 °C water with 8 ppm DO. Using electron and neutron diffraction patterns obtained from the fracture surfaces or the gauge sections of Alloy 600 after SSRT in 360 °C water, it is demonstrated for the first time that the 1/3{4 2 2} reflections correspond to the short-range ˚ . A higher ductility ordered phase with a d-spacing of 2.1 A and suppressed SCC of Alloy 600 in 250 °C water with DO below 10 ppb seem to be related to the slow rate of SRO at 250 °C. In conclusion, we propose that SCC in Alloy 600 is related to hydrogen-enhanced SRO and the dominant cracking mechanism is SRO leading to lattice contraction. Acknowledgments This work has been carried out as a part of the Nuclear R&D Program sponsored by the Korean Ministry of Science, ICT and Future Planning. Special thanks are due to S.S. Lee, who conducted neutron diffraction experiments at KAERI, to Y.W. Kim, who obtained electron diffraction patterns using TEM and to E.J. Kim, who made TEM samples using FIB. YSK is particularly grateful for the comments by Michael Kaufman (Professor, Department Head in Colorado School of Mines).
References [1] [2] [3] [4] [5]
[6]
[7] [8] [9] [10]
D. Van Rooyen, Corrosion 31 (1975) 327. S.M. Bruemmer, G.S. Was, J. Nucl. Mater. 216 (1994) 348. R.C. Newman, Corrosion 50 (1994) 682. P.L. Andresen, in: Corrosion 96, NACE: 1996. Paper #96258. P.M. Scott, M. Le Calvar, in: Proceedings of 6th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, San Diego, CA, American Society for Metals, Materials Park, OH, 1993, p. 657. P.M. Scott, in: Proceedings of 9th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Newport Beach, CA, TMS, Warrendale, PA, 1999, p. 3. T.M. Angeliu, D.J. Paraventi, G.S. Was, Corrosion 51 (1995) 837. T. Onchi, K. Dohi, N. Soneda, M. Navas, M.L. Castano, J. Nucl. Mater. 340 (2005) 219. P.M. Scott, PWSCC of Nickel Base Alloys and Mitigation in PWRs, presented at the INL Seminar on SCC in LWRs, Idaho Falls, Idaho, 19–20 March 2013, p. 32. K.F. Amouzouvi, R.C. Styles, L. Clegg, D.P. McCooeye, L. Zhao, A.K. Koul, in: Proceedings of 6th International Conference on Shot Peening, San Francisco, CA, 1996. p. 166.
515
[11] A.J. Marucco, Phys. IV France 03 (1993) C7-331. [12] S. Kohara, G.C. Kuzynski, Acta Met. 2 (1956) 221. [13] J. Hou, Q.J. Peng, Z.P. Lu, T. Shoji, J.Q. Wang, E.-H. Han, et al., Corros. Sci. 53 (2011) 1137. [14] J. Hou, Q.J. Peng, J.Q. Wang, E.-H. Han, W. Ke, Corros. Sci. 53 (2011) 2956. [15] N. Totsuka, E. Lumarska, G. Gragnolino, Z. SzklarskaSmialowska, Corrosion 43 (1987) 505. [16] F.H. Hua, P.B. Rebak, in: Proceedings of the Second International Conference on Environment-Induced Cracking of Metals (EICM-2), Banff, AB, 2004. [17] N. Totsuka, Z. Szklarska-Smialowska, Corrosion 43 (1987) 734. [18] T.B. Flanagan, A.P. Craft, T. Kuji, K. Baba, Y. Sakamoto, Scr. Metall. 20 (1986) 1745. [19] W.Y. Maeng, M.S. Choi, U.C. Kim, J. Mater. Sci. 39 (2004) 655. [20] R.A. Page, Corrosion 39 (1983) 409. [21] R.A. Page, A. McMinn, Metall. Trans. A 17A (1986) 877. [22] J.K. Sung, G.S. Was, Corrosion 47 (1991) 824. [23] Y.S. Kim, S.S. Kim, D.W. Kim, in: Proceedings of 16th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Asheville, NC, TMS, Warrendale, PA, 2013. [24] Y. Fukai, Y. Shizuku, Y. Kurokawa, J. Alloys Compd. 329 (2001) 195. [25] T. Hiroi, Y. Fukai, K. Mori, J. Alloys Compd. 404–406 (2005) 252. [26] C.D. Beachem, Metall. Trans. 3 (1972) 437. [27] A. Marucco, B. Nath, J. Mater. Sci. 23 (1988) 2107. [28] D.P. Abraham, C.J. Altstetter, Metall. Mater. Trans. A 26A (1995) 2859. [29] X.J. Jiang, B. Noble, G. Waterloo, J. Tafto, Metall. Mater. Trans. A 31A (2000) 339. [30] K. Sato, M. Ichinose, Y. Hirotsu, Y. Inoue, ISIJ Int. 29 (1989) 868. [31] L. Luo, W. Jie, Y. Xu, Y. He, L. Xu, L. Fu, R. Soc. Chem. 16 (2014) 5073. [32] B.H. Choe, Private communication, 2012. [33] A. Marucco, Mater. Sci. Eng., A 189 (1994) 267. [34] F. Foct, O.D. Bourvier, T. Magnin, Metall. Mater. Trans. A 31A (2000) 2025. [35] S.S. Kim, D.W. Kim, Y.S. Kim, Met. Mater. Int. 19 (2013) 969. [36] T. Couvant, L. Legras, A. Herbelin, A. Musienko, G. Ilevbare, D. Delafosse, G. Cailletaud, J. Hickling, in: Proceedings of 14th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Virginia Beach, VA, TMS, Warrendale, PA, 2009, p. 182. [37] T. Nakamura, M. Koshiishi, T. Torimaru, Y. Kitsunai, K. Takakura, K. Nakata, M. Ando, Y. Ishiyama, A. Jenssen, in: Proceedings of 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems, Whistler, BC, TMS, Warrendale, PA, 2007. [38] F. Vaillant, J.M. Boursier, L. Legras, B. Yrieix, E. Lemaire, J. Champredonde, C. Amzallag, in: Proceedings of 13th International Conference on Environmental Degradation of Materials in Nuclear Power Systems, Whistler, BC, TMS, Warrendale, PA, 2007. [39] V. Gerold, H.P. Karnthaler, Acta Met. 37 (1989) 2177. [40] G. Thomas, Acta Met. 11 (1963) 1369. [41] M.P. Manahan, R. Kohli, J. Santucci, P. Sipush, Nucl. Eng. Des. 113 (1989) 297. [42] U. Ehrnsten, T. Saukkonen, W. Karlsen, H. Hanninen, in: Proceedings of 14th International Symposium on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Virginia Beach, VA, TMS, Warrendale, PA, 2009, p. 910. [43] G.S. Was, K. Lian, Corrosion 54 (1998) 675.