ML12233A725
| ML12233A725 | |
| Person / Time | |
|---|---|
| Site: | Indian Point  |
| Issue date: | 08/20/2012 |
| From: | Andrieu E, Delabrouille F, Legras L, Scott P, Vaillant F, Viguier B Electricite de France, Framatome ANP |
| To: | Atomic Safety and Licensing Board Panel |
| SECY RAS | |
| References | |
| RAS 23331, 50-247-LR, 50-286-LR, ASLBP 07-858-03-LR-BD01 | |
| Download: ML12233A725 (7) | |
Text
ENT000556 Submitted: August 20, 2012 Proceedings of the 12th International Conference on Environmental Degradation of Materials in Nuclear Power System - Water Reactors -
Edited by T.R. Allen, P.1. King, and L. Nelson TMS (The Minerals, Metals & Materials Society), 2005 EFFECT OF THE CHROMIUM CONTENT AND STRAIN ON THE CORROSION OF NICKEL BASED ALLOYS IN PRIMARY WATER OF PRESSURIZED WATER REACTORS F. Delabrouille'*3, L. Legras', F. Vaillant', P. Scott2, B.Viguier3, E. Andrieu3
, EDF (Electricite De France) R&D, MMC, Site des Renardieres, Moret sur Loing cedex 77818, France 2 Framatome-ANP, Tour Areva, 92084 Paris La Defense cedex, France 3 CIRIMAT - UMR CNRS 5085, ENSIACET, 118 route de Narbonne, 31077 Toulouse cedex 4, France Keywords: nickel alloys, SCC, corrosion, chromium.
Abstract level [11]. This study is undertaken in order to determine the effect of chromium and plastic deformation on the oxide films Chromium is known to enhance the corrosion and stress corrosion formed on Ni-xCr-9Fe (5 < x <30 wt%) alloys corroded in cracking behaviour of nickel-based alloys in high temperature primary water.
water. A study has been launched to characterize these effects for alloys containing chromium concentrations ranging from 5% to Alloys and experimental procedures 30%. This present paper compares the growth of oxide scale in these alloys in PWR primary water, and investigates the effects of Seven different alloys are prepared by the steel maker TECPRY applied stress. A detailed characterization of the oxide scale is with compositions ranging from low (5 wt%) to high (30 wt%)
performed by SEM, TEM and SIMS. Increasing chromium chromium contents (Table 1). This range spans the composition content results in an increase in chromium content of the of the industrial alloys (alloy 600 (15%) and alloy 690 (30%). The protective oxide scale and decreases the oxide thickness. Applied content of alloying elements such as Fe, Ti, Al are maintained stress, however acts to increase oxide thickness and decrease the within the specification of alloys 600 and 690 in order to be chromium content of the oxide scale. comparable with the industrial products. The composition balance is done on nickel.
Introduction Reference C Ni Cr Fe Ti Al Si S The nickel-based alloy 600 (Ni-16Cr-9Fe) has been used for B283 0.021 Bal. 5.11 8.25 <0.005 O.oJ5 0.126 0.006 steam generator (SG) tubing in pressurized water reactors (PWR). B216 0.027 Bal. 10.02 10.55 <0.005 0.012 0.107 <0.001 This alloy is susceptible to intergranular stress corrosion cracking B379 0.019 Bal. 14.96 10.07 0.24 0.07 <0.01 <0.001 (IGSCC) in primary water. The higher chromium alloy 690 (Ni-30Cr-9Fe), shows better resistance to IGSCC[l] and is currently B381 0.02 Bal. 19.98 10.11 0.24 0.07 <0.01 <0.001 being used in replacement steam generators and in the B384 0.017 Bal. 24.99 10.12 0.24 0.09 <0.01 <0.001 manufacture of new SG's. To date, the improved resistance of B385 0.017 Bal. 27.13 10.19 0.24 0.08 <0.01 <0.001 alloy 690 has not yet been clearly explained. B388 0.019 Bal. 30.01 10.13 0.24 0.08 <0.01 <0.001 Table 1. ChemIcal composItIon of the alloys studIed (weIght %) .
Coriou et al. are the first to examine the effects of chemical composition on stress corrosion cracking in high temperature For general corrosion tests, coupons are machined from the alloys water [2]. Recent studies have shown that chromium has a in the form of rectangular bars 50mm long, 10mm width and 1mm beneficial effect on the resistance to IGSCC [3,4]. Determining thick. Slow strain rate tensile tests (SSRT) in pressurized water the specific cause behind the beneficial role of chromium is are realized on both cylindrical (diameter 4mm, gage length difficult because of the large number of factors influencing 80mm) and plate specimens (3mm width, 1mm thick and 20mm IGSCC behaviour. Chromium influences the formation of a of gage length).
surface film. The oxide film is more protective and has a better mechanical resistance for chromium contents higher than 17% [5].
General corrosion tests are run for 1000 hours0.0116 days <br />0.278 hours <br />0.00165 weeks <br />3.805e-4 months <br />. Corrosion tests are It has also been shown that the passivity and the repassivation performed in an isothermal loop. The primary PWR water is rates in environments increase with Cr content [5,6]. For simulated by a solution of 1000 ppm boric acid and 2 ppm LiOR chromium contents above 12% the alloy is given the designation in demineralised water. The partial pressure of hydrogen is set to "stainless" For equivalent corrosion times, the oxide layer that 0.17 bars. Pressure and temperature are fixed respectively at 18.5 develops in alloy 600 is thicker than that developed in the 690 MPa and 360°C, which represents an accelerated test compared to alloy [7,8]. The creep and deformation rates are decreased by the the conditions experienced in service. SSRT tests are performed in hardening effect of chromium [4,9,10], for example, Vaillant et al.
a static autoclave at 360°C and 15.0 Mpa in a solution observed a decrease in creep rate by a factor of 2 for an increase representative of PWR primary water. The specimens are in chromium content from 15% to 30% for similar grain boundary deformed to rupture under an imposed strain rate of 5 10.8 s",
precipitation and thermal treatment conditions. The effect ofCr on leading to test durations ranging from 350 to 3000 hours0.0347 days <br />0.833 hours <br />0.00496 weeks <br />0.00114 months <br />. The IGSCC has been also attributed to the grain boundary chromium 903
surfaces of all specimens are polished down to grade 1200 grade SiC paper.
The oxide films are analyzed using scanning electron microscopy (SEM), analytical transmission electron microscopy (ATEM) and secondary ion mass spectrometry (SIMS). The corrosion products are characterized using a LEO SUPRA 35 SEM equipped with a Field Emission Gun (FEG) and In-Lens detector operated at 5 kY.
A TECNAI F20 FEG/TEM operating at 200 kY is used for all TEM imaging. Energy filtered electron transmission microscopy (EFTEM) images are acquired at 198 kY using a GATAN Image Filter (GIF), K ratio images are shown obtained with the two windows techniques. Energy dispersive X ray (EDX) analyses are performed in the scanning transmission mode, with a spot size random from I to 1.5 nm, and the specimen tilted to 20° towards the detector using automatic line scanning analysis. There are difficulties in quantifying oxygen concentrations using EDF methods as the pairs of Cr L lines (0.45-0.57keY) and 0 K lines (0.52 keY) can't be resolved due to the relatively poor energy resolution of the EDX detectors. As such the oxygen signal is not shown and is not used for determining the oxide composition, instead the percentage of the different oxides corresponds to the ratio of metallic compounds. The oxygen repartition is studied using energy loss filtered imaging. No surface preparation is used for SEM observations, TEM cross sections are prepared using conventional techniques. The first step is to cut a section of the surface and to embedded in a brass insert. Then the insert is sliced to obtain thin discs, these discs are polished mechanically to a final thickness of 80flm. A bowl-shaped depression is created with dimple grinding. The final ion-milling with GAT AN Precise Ion Polishing System (PIPS) is used to produce electron transparent regions for TEM characterization. SIMS analyses are performed using a CAMECA IMS 4F, the abraded rate in the alloy is calibrated in order to set the depth under the metal-oxide interface at which the elemental images are taken.
Results and discussion (c)
- Oxides on undeformed specimens Figure I. SEM images of the oxide scale developed in simulated primary water on alloys with various Cr contents (a) 5% Cr, (b)
The microstructures of the external corrosion layer produced on 15% Cr and (c) 30% Cr.
coupons are first examined by SEM for the different alloys. The alloys presenting the lowest chromium contents (5 and 10%) Two categories of filaments are detected; nickel-rich filaments, exhibit oxides in filaments shaped on their surface. The filament and filaments containing both iron and nickel (probably NiFe204 density increases with decreasing chromium content until the spinels). Beneath the filament layer small discrete crystallites are surface is completely covered for the 5 wt% Cr alloy (Figure la). observed, as indicated by the white arrows in Figure la. The filaments disappear with increasing chromium content and are replaced by a distribution of crystallites as illustrated in Figures lb (15% Cr) and Ic (30% Cr). The crystallites present a bimodal size distribution of small crystallites in the order of 10nm, and large ones ranging from 100 nm to few microns (randomly distributed) (Figure lc). The average size of the small crystallites decreases with chromium content The oxide scales are also observed in the cross-section using TEM in order to examine in more detail the different layers of the scale, both in structure and composition.
904
I Chromium I Chromium Nickel and iron rich crystallite I rich scale I de leted zone 80 70 --*-Cr
-.to-Fe
- - 60
.<:: ---y---- Ni Ol
.~ 50 ***--*--Ti
~ 40 (a) -El 2 30 c
100 Alloy chromium rich scale Crystallite 8 20 90 80
- c Ol 70 10 20 30 40 50 60 70 80 Distance (nm) 90 100 110 120 130 140
'iii
- 60 Figure 3. TEM cross-section of the oxide scale formed on the 15%
~ 50 Grain boundary Cr alloy in simulated primary water.
.l!l c
2 40 c
0 30
()
20 10 200 400 600 800 distance (nm)
Figure 2. TEM cross-section of the oxide scale developed in simulated primary water in the 5% chromium alloy.
Figure 2 illustrates the entire thickness of the oxide formed on the 5% Cr alloy. The filaments observed in Figure la, can be seen on the coupon surface although they are somewhat damaged due to Crystallite IChromiumlChromium depletel:i rich scale I zone I Alloy the TEM sample preparation. The crystallites seen in SEM are also visible beneath the filaments and do not form a continuous layer. An inner oxide layer can be seen below the crystallites. This consists of chromium rich oxide grains that have penetrated into :E 60 Ol the metal as opposed to oxide grains that have grown outwards .~ Grain bound~
from the metal surface. ;oR
~40 (J)
The outer oxide layer formed on the 15% Cr alloy, is composed of "E the crystallites observed previously in SEM (Figure 3). This layer 2c o
is not compact and an EDX scan indicates enrichment of the () 20 crystallites by Ni and Fe (see the EDX line scan in Figure 3). A 10nm thick inner oxide layer is observed enriched with chromium up to 50%. In this layer the impurities like titanium could segregate (see Figure 3). Chromium is depleted over a distance of 20 40 60 80 100 120 approximately one hundred nanometres beneath the metal-oxide Distance (nm) interface. Figure 4. TEM cross-section of the oxide scale formed on the 30%
Cr alloy in simulated primary water.
The layered structure of the oxide scale developed on the 30 %
alloy is shown in Figure 4. The outer surface displays crystallites as observed by SEM (Figure l.a). This layer is not continuous, consisting of discrete separated grains, meaning it is unlikely that this layer acts as a protective layer. A second layer, 10nm thick, consisting of very small oxide grains is observed below the outer surface. The structure of the oxide layers is more clearly defined in the X ray profile analysis (Figure 5). The outer crystallites are enriched in chromium and iron (40% of Cr and 40% of Fe). The second oxide layer is highly enriched in chromium, reaching levels of up to 70%. The parent metal is impoverished of chromium over a thickness of 30nm below the second oxide layer.
905
These features are observed on all alloys with chromium contents higher than 10 %. Table 2 summarises the characteristics (composition and dimensions) of the zone depleted of chromium as a function of chromium content. The EDX analysis shows that the external crystallites are progressively enriched in chromium at the expense of iron and nickel for alloys with increasing nominal chromium content. By contrast, the inner Cr rich layer behaved differently: a minimum chromium content is required to form this layer (- 10 %), this layer being unchanged in thickness and composition for higher nominal chromium contents. These features are summarized in Table 2.
Chromium 5% 10% 15% 20% 25% 30%
content Max. size of big fewJ.!m fewJ.!m 140nm 90nm 50nm 50nm crystallites Composition NiFe,04 NiFe,04 10%Cr 25%Cr 45%Cr 50%Cr pfbig and Ni and Ni 50% Fe 50% Fe 25% Fe 40% Fe clystallites rich rich 40%Ni 25%Ni 20%Ni 10%Ni Irhickness 0 100 to Cr in the 100nm 5-IOnm 5-10 nm 5-IOnm 5-IOnm 200nm inner layer Max. % 0 chromium in 25' 80 80 75 80 90 the inne layer' Max. depth 0
-100 -100 Cr in the - 500nm 100nm 60nm nm nm depleted zone Table 2. Charactenstlcs of the oXIde layers appeanng on coupons as a function of the chromium content of the alloy.
'This percentage corresponds the percentage chromium is given by the ratio of chromium to other alloying elements, it's not the real composition of the oxide, because the oxygen is not quantified.
, For the 5 % Cr alloy the inner oxide does not form a layer but instead is constituted by the penetration oflarge particles from place to place.
In order to identifY the consequences of the corrosion process inside the alloys, chemical analysis of the coupons have been performed by SIMS in the "reversed mode". That is, the analysis is performed in image mode from the metal side of the oxide layer, and successive removals of the metal enabled a through thickness investigation of the distribution up to the oxide [12].
This procedure presents two main advantages as compared to classical SIMS analysis where the abrasion starts from the oxide to the metal: i) A flat clean surface can be obtained from the metal side in contrast with the rough outer oxide layer and ii) the sensitivity to oxygen can be adjusted to detect very small amounts of 0 atoms (The detector rapidly saturates in its detection of oxygen when starting from the outer oxide surface. Contamination of the metal by oxides is also avoided). The three images of Figure S are taken chronologically during such a "reversed" abrasion sequence, Figure Sa is 4.S/lm under the metal oxide interface, to Figure Sc which is taken on the oxide scale itself.
This analysis demonstrates that oxygen penetrates over large distances into the metal. It must be noted that the oxygen penetration is not uniform but it's localized at certain points in the microstructure. Further abrasion shows the correspondence of these points with the triple lines in the grain structure. These results complement the observations by SIMS performed by Newman et al [13], who detected chromium oxide at the grain Figure S. SIMS images of the oxygen signal during "reversed boundary under the surface of alloy 600. The present results abrasion" showing the deep penetration of oxygen at the triple indicate the role of triple lines and the large penetrations of lines and grain boundaries of the 20% Cr alloy. The images oxygen that can be observed (penetrations up to 9 /lm are seen). correspond respectively to the depth under the metal surface: (a) 4.S /lm, (b) 1.9 /lm, (c) 0.3 /lm.
906
The overall consequences of corrosion on nickel-based alloys is 100 schematized in Figure 6. An outer oxide layer consisting of large sparse crystallites and a more uniform distribution of small '\.!'ro/"""\ -"'"
80 crystallites is seen. Below the outer layer is a thin compact layer of chromium rich oxide (chromium can be enriched up to 90 % ). :E 0>
Beneath this layer, the alloy is depleted in chromium over a .~ 60 distance of a few tens of nanometres. An "affected zone" is ~
observed in the metal, which takes on the appearance of a heavily J!J 40 c
deformed and fine grained microstructure as reported in the 2c literature [14]. The deep penetration of oxygen at the triple 0
() 20 junctions of grain boundaries is also indicated. This schema is only valid for alloys with a chromium content higher than 10 %. .-
Below this composition numerous filaments are present on top of the oxide scale and the chromium rich oxide layer is not 50 100 150 200 250 300 350 continuous. Distance (nm) 10mn Chrom!um deplded Zone Figure 8. TEM cross-section of the oxide scale developed on the 5% Cr tensile specimen tested in simulated primary water.
1102 "'"
At tlte lmaxlm..mj Figure 6. Schematic of the oxide structure observed for nickel based alloys with chromium contents above 10% oxidized in simulated primary water
- Oxides on deformed specimens Complete characterization of the corrosion products on the SSRT specimens has also been performed. Figure 9 show the cross section of the 5% Cr alloy. The corrosion layers are strongly modified in comparison with the layers formed on coupon. The outer scale consists of oxide filaments, however the inner oxide is slightly enriched by Cr and Fe (up to 10% Cr and up to 10% Fe).
Figure 8 shows the cross section of the 15 % Cr alloy as imaged using the jump ratio technique in GIF. The structure of the oxide scale on deformed material is strongly modified with respect to sectIon of the oxide scale developed on the 15%
the unstressed sample. The oxide-metal interface, indicated by the Cr alloys during SSRT testing to rupture (exposure time: 2500 dashed white line in the Fe ratio image, does not present a thin hours) as observed in EFTEM.
layer of Cr rich oxide (Figure 9). Instead oxygen penetrates more deeply into the metal to form Cr enriched oxide inside the parent metal. The structure of the oxide scale is similar to that developed in unstressed samples oflower Cr contents.
907
CONCLUSION inner Chromium 80+-________~C~~~s~ta~lIi~te=s________~~~~d~e~le~led~Z~On~e~
The consequence of corrosion in high temperature water on nickel based alloys as a function of chromium content has been characterized using SEM, TEM and SIMS. All alloys formed a non-continuous external layer of crystallites (non protective). For alloys above 10% chromium a continuous and compact layer of chromium oxide develops that covers the metal surface. The penetration of oxygen under the oxide scale along triple junctions is evidenced by SIMS imaging, this penetration may reach several micrometers. The SSRT tests showed that chromium content has a strong influence on the rate of oxide growth. Plastic deformation is shown to have a significant influence on the structure of the corrosion film. In particular the detrimental influence of strain on 20 40 60 80 100 120 Distance (nm) the protective chromium oxide layer has been demonstrated. The designation "stainless" can be linked up to the existence of oxide rich inner layer. Further paper will deal with the effect of chromium on IGSCC behaviour and oxide growth within cracks.
References
[I] K. SMITH, A. KLEIN, P. SAINT-PAUL and J. BLANCHET, proceedings of second international Symposium In Environmental Degradation of Materials In Nuclear Power Systems-Water reactors, pp12-1 to 10, NACE, 1989.
[2] H. CORIOU, L. GRALL, Y. LEGALL, S. VETTIER, "scc of Inconel in high temperature water", 3rd metallurgy symposium, Saclay, France, p.161, 1960.
[3] T. YONEZAWA, N. SASAGURI, K. ONIMURA, "Effects of metallurgical factors on stress corrosion cracking of Ni-base alloys", Japan Atomic Industrial Forum, vol. 2, p.490, 1988.
Figure 10. Cross section of the oxide scale developed on 30% Cr [4] G.S. WAS, "Micromechanical and microstructural effects on alloys during SSRT (exposure time: 2500 hours0.0289 days <br />0.694 hours <br />0.00413 weeks <br />9.5125e-4 months <br />). SCC of nickel base alloys", PARKINS symposium on fundamental aspects of stress corrosion cracking, 1991.
[5] T.M. ANGELIU, G.S. WAS, "The effect of chromium, carbon The layered structure of the oxide scale developed on the SSRT and yttrium on the oxidation of nickel-base alloys in high specimen of the 30 % alloys is shown in Figure 10. At the outer temperature water", Journal of Electrochemical. Society, vo1.140, surface, crystallites of about a hundred nanometers size are seen. p.1877,1993.
The inner oxide layer consists of a layer enriched in chromium [6] K. YAMANAKA, J. MURAYAMA, proceedings of fourth (up to 70%), as illustrated by the line scan crosses on Figure 10. international symposium on environmental degradation of This layer is about 10 nm thick. A 20nm layer impoverished of materials in nuclear power systems-water reactors, Houston, p.96 chromium is observed in the parent metal below the metal oxide 1989.
interface. [7] L. GUINARD, O. KERREC, D. NOEL, S. GARDEY and F.
COULET, Water Chemistry of Nuclear Reactor Systems 7, These results confirm the general tendency of plastic deformation BNES, Bournemouth, 62, 1996.
to weaken the role of Cr in the development of the Cr oxide [8] F. CARRETTE, << Relachement des produits de corrosion des protective layer. A few observations have been also undertaken to tubes en alliage 690 de generateur de vapeur du circuit primaire characterize the structure of oxide within cracks developed during des reacteurs a eau pressurisee >>, Ph-D Thesis (in French), IN?
tensile tests. Such observations indicate that the structure of Toulouse, 2002.
oxides and alloy modifications within cracks are very different [9] G.S. WAS, "Deformation and intergranular cracking from what happens at the free surface. Two explanations for this behaviour of Ni-Cr-Fe alloys at high temperature", EPRI Airlie difference can be given : i) the confinement of the corroding Workshop, Warrenton, 1993.
medium in the crack alters the electro-chemical conditions with [10] F. VAILLANT, J.D. MITHIEUX, O. DE BOUVIER, D.
respect to the surface and ii) mechanical stresses are strongly V ANCON, G. ZACHARIE, "Influence of chromium content and concentrated at the crack tip resulting in modifications of the microstructure on creep and PWSCC resistance of nickel base corrosion rate. These finding highlight the essential importance of alloys", Proceedings 9th Symposium on environmental characterizing the SCC directly in the crack, in agreement with degradation of materials in nuclear power system, 1999.
the observations by Thomas et al. [15]. Recent work in this view [II]. J.J. KAI, G.P. YU, C.H. TSAI, M.N. LIA, and S.C. YAO, showed that such observation may be very useful for "The Effects of Heat Treatment on the Chromium Depletion, discriminating between the different mechanisms that may operate Precipitate Evolution, and Corrosion Resistance of Inconel Alloy during laboratory or in service stress corrosion cracking of nickel 690" Metallurgical Transactions A., Vo1.20A, p.2057, 1,989.
based alloys [16]. [12] J. PANTER; B. VIGUIER; J.M. CLOUE, J.; M.
FOUCAULT, P. COMBRADE, and E. ANDRIEU; << Influence of 908
oxide films on primary water stress corrosion cracking initiation of alloy 600 >>, Journal of Nuclear Materials, submitted 2005.
[13] R.C. NEWMAN; T.S. GENDRON and P.M. SCOTT; "Internal oxidation and embritlement of Alloy 600", Ninth International Symposium on Environment Degradation of Materials in NPS-WR, Newport Beach, CA, USA 1-5 Aug 1999, The TMS pp 79 - 93, 2000.
[14] F. CARRETTE, M.e. LAFONT, G. CHATAINIER, L.
GUINARD and B. PIERAGGI, << Analysis and TEM examination of corrosion scales grown on Alloy 690 exposed to pressurized water at 325 c>> Surface and Interfaces analysis., 34:, pp. 135-138,2002.
[15] L. THOMAS and S. BRUEMMER, "High resolution analytical electron microscopy characterization of corrosion and cracking at buried interfaces", Surface and Interfaces analysis, 31, p.571-581,2001.
[16] F. DELABROUILLE, "Caracterisation par MET de fissures de corrosion sous contrainte d'alliages it base de nickel: influence de la teneur en chrome et de la chimie du milieu" Ph -D Thesis (in French), INP Toulouse, 2004.
909