Thierry Couvant, David Delafosse, Francois Vaillant, Jean-Marie Boursier
Keywords: 304L, strain path, prior strain, strain rate, strain corrosion cracking
Abstract:
Austenitic stainless steels are widespread in primary and auxiliary circuits of pressurized water reactors (PWRs). Moreover, some non-irradiated components suffer stress corrosion cracking (SCC). This degradation could be the result of the synergy between oxidation and strain hardening. Constant elongation rate tests (CERTs) were conducted on austenitic stainless steel AISI 304L in primary PWR environment at 360°C. Particular attention was directed towards pre-straining effects on crack growth rate (CGR) and crack growth path (CGP). Results have demonstrated that the susceptibility of AISI 304L to SCC in high-temperature hydrogenated water was enhanced by pre-straining. It seemed that IGSCC was enhanced by complex strain paths.
Because you are not logged-in to the journal, it is now our policy to display a 'text-only' version of the preprint. This version is obtained by extracting the text from the PDF or HTML file, and it is not guaranteed that the text will be a true image of the text of the paper. The text-only version is intended to act as a reference for search engines when they index the site, and it is not designed to be read by humans!
If you wish to view the human-readable version of the preprint, then please Register (if you have not already done so) and Login. Registration is completely free.
ISSN 1466-8858 Volume 11, Preprint 1 submitted 20 May 2008 © 2008 University of Manchester and the authors. This is a preprint of a paper that has been submitted for publication in the Journal of Corrosion Science and Engineering. It will be reviewed and, subject to the reviewers" comments, be published online at http://www.jcse.org in due course. Until such time as it has been fully published it should not normally be referenced in published work. 1 Effect of Effect of Effect of Effect of SSSStraintraintraintrain----PPPPath on ath on ath on ath on SSSStress tress tress tress CCCCorrosion orrosion orrosion orrosion CCCCracking of AISI 304L racking of AISI 304L racking of AISI 304L racking of AISI 304L S SSStainless tainless tainless tainless SSSSteel in PWR teel in PWR teel in PWR teel in PWR PPPPrimary rimary rimary rimary EEEEnvironment at 360°Cnvironment at 360°Cnvironment at 360°Cnvironment at 360°C Thierry Couvant (1), David Delafosse (2), François Vaillant (1), Jean-Marie Boursier (1). (1) EDF R&D, Avenue des Renardières, 77818 Moret-sur-Loing, France (2) Ecole des Mines de St-Etienne, 157 Cours Fauriel, 42023 St-Etienne cedex 2, France thierry.couvant@edf.fr AbstractAbstractAbstractAbstract Austenitic stainless steels are widespread in primary and auxiliary circuits of pressurized water reactors (PWRs). Moreover, some non-irradiated components suffer stress corrosion cracking (SCC). This degradation could be the result of the synergy between oxidation and strain hardening. Constant elongation rate tests (CERTs) were conducted on austenitic stainless steel AISI 304L in primary PWR environment at 360°C. Particular attention was directed towards pre-straining effects on crack growth rate (CGR) and crack growth path (CGP). Results have demonstrated that the susceptibility of AISI 304L to SCC in high- temperature hydrogenated water was enhanced by pre-straining. It seemed that IGSCC was enhanced by complex strain paths. Keywords: Keywords: Keywords: Keywords: 304L, strain path, prior deformation, strain rate, stress corrosion cracking. IntroductionIntroductionIntroductionIntroduction Austenitic stainless steels are characterised by a good resistance to general corrosion at elevated temperature, permitting their widespread use in primary and auxiliary circuits of pressurised water reactors (PWRs). However, some non-irradiated components suffer stress corrosion cracking (SCC). This degradation could be the result of the strain hardening. Numerous studies have demonstrated the susceptibility to stress corrosion cracking (SCC) of austenitic stainless steels in boiling MgCl 2 solution [1] and in Boiling Water Reactor (BWR) environment [2]. Most of SCC tests have been made in order to find the threshold values for SCC occurrence and to clarify the effect of prior-deformation. Particularly, transgranular SCC (TGSCC) in boiling MgCl 2 solution initiation appears when the amount of elastic energy reach a threshold value, making the strain hardening essential for the initiation and the propagation of SCC [3]. Furthermore, it seems that susceptibility to SCC is not a monotonic function of pre-straining in that environment [4]. In high temperature water, intergranular SCC (IGSCC) is also deeply correlated to strain hardening: crack growth tests have shown that the crack growth rate (CGR) increases with the yield strength in several austenitic stainless steels [5]. On the other hand, SCC of austenitic stainless steels in non-polluted primary PWR environment is relatively poorly known, mainly because of the restricted conditions for the occurrence of the phenomena in this environment. For non-irradiated materials, currently, the main cause of initiation of SCC in nominal primary PWR environment is related to materials sufficiently pre-strained [6, 7, 8]. Authors have tried to understand the effect of a pre-straining on SCC in such environment using cold pressed ISSN 1466-8858 Volume 11, Preprint 1 submitted 20 May 2008 © 2008 University of Manchester and the authors. This is a preprint of a paper that has been submitted for publication in the Journal of Corrosion Science and Engineering. It will be reviewed and, subject to the reviewers" comments, be published online at http://www.jcse.org in due course. Until such time as it has been fully published it should not normally be referenced in published work. 2 humped specimens [9, 10]. But the interpretation of constant elongation rate tests (CERTs) with this kind of specimens is not obvious, due to the complex combined local influences of pre-strain, stress and strain rate. In this context, this study has been conducted to highlight the effect of prior deformation and strain path on initiation and propagation of both IG and TGSCC of AISI 304L in PWR primary environment at 360°C. Pre-shearing is not the most common and straight forward method used to pre-harden materials prior to SCC tests. But this method has allowed to reproduce on flat specimens the sequential deformation at the intrados of cold pressed humped specimens where initiation of SCC has been widely observed. ExperimentalExperimentalExperimentalExperimental MaterialMaterialMaterialMaterial The material is a sheet of 30 mm thick. The chemical composition and mechanical properties of the austenitic alloy 304L tested in this work are given in Tables 1 and 2 respectively. Mechanical properties are isotropic in the plane of the sheet as shown by tensile tests in roll and transverse directions. This non-sensitized material was solution annealed at 1150°C and water quenched. The resultant microstructure is characterized by a grain size of about 30 ?m with no evidence of carbide precipitate in the matrix and along grain boundaries. The austenitic grain size was measured through metallographic etching (standard ASTM E 112 [11]). Austenite matrix contains less than 5% of residual d-ferrite and is subject to strain-induced martensite transformation (Ms = -133°C, Md 30 = -3°C) as predicted by Angel [12]. Table 1. Chemical composition (wt%) of studied 304L. C Si Mn S P Cr Ni Co Ti Cu Al Mo N Fe 0.026 0.52 1.49 0.002 0.027 19.23 9.45 0.07 <0.005 0.17 0.033 0.24 0.064 Bal. Table 2. Mechanical properties at 360°C of studied 304L. YS (MPa) UTS (MPa) El. (%) HV0.1 160 450 40 160 PrePrePrePre----SSSShear hear hear hear HHHHardening of the ardening of the ardening of the ardening of the MMMMaterialaterialaterialaterial Pre-shearing tests at room temperature were used to raise the yield strength of the material prior to SCC tests. Similarly to tensile test, we can consider that shear tests lead to an homogeneous strain-hardening with a fair approximation, without any localisation. Two samples (1 ´ 200 ´ 200 mm) have been cut in the middle of the sheet. Samples were silk- screen printed with a 2 mm step grid, in order to control the homogeneity of the deformation at the end of the pre-shear hardening test. Grids were eliminated prior to SCC tests. Pre-shearing were conducted with a frame composed of a fixed body where a mobile tie were sliding at its centre. The shear rate at the ambient air was 2.5 10 -3 s-1 and the shear amplitudes g were respectively 0.2 and 0.4 for the first and the second samples. Pre- shearing led to the formation of a"-martensite (5-7% measured by X-ray diffraction for g = 0.4) and to an increase of hardness (respectively 320 HV 0.1 and 340 HV0.1 for g = 0.2 and 0.4). Each sample had two gage lengths (1 ´ 40 ´ 200 mm) in where secondary specimens ISSN 1466-8858 Volume 11, Preprint 1 submitted 20 May 2008 © 2008 University of Manchester and the authors. This is a preprint of a paper that has been submitted for publication in the Journal of Corrosion Science and Engineering. It will be reviewed and, subject to the reviewers" comments, be published online at http://www.jcse.org in due course. Until such time as it has been fully published it should not normally be referenced in published work. 3 have been cut for SCC tests after pre-shearing, for 3 directions defined by f = 45°, 90° and 135° (see Figure 1). Consequently, three strain path changes were followed during subsequent SCC tensile tests. Extensive studies concerning sequential deformations have already been published [13, 14]. Schmitt [15] have proposed a scalar parameter b to characterize a two-stage strain path. In our study b was defined as the double contracted tensor product between the plastic shear mode 1~eduring the pre-strain at room temperature and the subsequent plastic tensile mode 2~e during the CERT in primary PWR environment at 360°C: 2121~.~ ~:~ ee eeb= For a b value close to 1 (f = 45°) the strain tensors were almost identical leading to a pseudo-monotonic test (no important changes were seen in the stress-strain curves). For a value close to -1 ( f = 135°), a reverse test, or pseudo-Baushinger test, was obtained. For a value close to 0 ( f = 90°), the strain tensors of the two sequential deformation paths were perpendicular, i.e. the double-dot product of these tensors is zero. Then, the sequential deformation path led to a cross effect. The micro-mechanical aspects of strain path changes are not dealt in this paper. Figure 1 - Machining of secondary specimens in pre-sheared materials at room temperature. 200 f = f = 45° 40 mm Rolling direction f = 135° ISSN 1466-8858 Volume 11, Preprint 1 submitted 20 May 2008 © 2008 University of Manchester and the authors. This is a preprint of a paper that has been submitted for publication in the Journal of Corrosion Science and Engineering. It will be reviewed and, subject to the reviewers" comments, be published online at http://www.jcse.org in due course. Until such time as it has been fully published it should not normally be referenced in published work. 4 Specimen Specimen Specimen Specimen PPPPreparreparreparreparationationationation Two types of specimens were used in this study. Type A specimens with 79 mm gauge length were used for CERTs on non-pre-sheared material. These 2 mm thick specimens were cut in the middle of the sheet, in its transverse plane. Type B specimens with 25 mm gauge length and 1 mm thick were cut in pre-sheared samples for SCC tests. In the middle of the gauge length of type B specimens, a double notch was machined by electron discharge machining. The radius and depth of the notches were 150 ?m. 3D calculations by finite elements allowed to determine the strain, stress and triaxiality in the notches against the elongation of the specimen at 360°C. CERTCERTCERTCERT PPPProcedurerocedurerocedurerocedure Specimens were ultrasonically rinsed in ethanol and then in distilled water. Tests were carried out in Hastelloy (C-276) autoclaves. Specimens were isolated from the autoclave by oxidized zircalloy to avoid galvanic coupling. Experiments were conducted under open circuit conditions. The environment was primary water at 360°C, containing 1000 ppm B as boric acid and 2 ppm Li as lithium hydroxide. Solution was previously de-aerated by evaporating 20% of the initial volume at 125°C, then a hydrogen overpressure was introduced (30 cc/kg) and controlled using a Pd-Ag thimble. CERTs, were conducted with an apparent applied strain rate ( apε&) of 5 10-8 s-1. For notched specimens, apε& was defined as [Ln(1+ dl/l 0)]/t, where dl was the elongation of the specimen, l0 the initial length and t the duration of the straining. During CERTs the load was measured against elongation. At the end of the tests, specimens were rinsed in distilled water and then microscopically examined in order to find any SCC. The depth of the deepest crack was measured on the fracture surface of failed specimens, using a scanning electron microscope (SEM), or on cross-sections in the case of interrupted tests. ResultsResultsResultsResults Initiation Initiation Initiation Initiation aaaand nd nd nd PPPPropagation of TGSCC ropagation of TGSCC ropagation of TGSCC ropagation of TGSCC without prior deformationwithout prior deformationwithout prior deformationwithout prior deformation Five tests have been carried out. Four tests were stopped before the fracture of the specimen, for different elongations, allowing to establish a relationship between the depth of the main stress corrosion cracks and the strain hardening of the material resulting from the CERT in primary PWR environment (Figure 2). The fracture surface analysis of the failed specimen allowed to identify a purely transgranular stress corrosion cracking. Cracks were initiated and propagated on every faces of the specimen. Furthermore, a large amount of ISSN 1466-8858 Volume 11, Preprint 1 submitted 20 May 2008 © 2008 University of Manchester and the authors. This is a preprint of a paper that has been submitted for publication in the Journal of Corrosion Science and Engineering. It will be reviewed and, subject to the reviewers" comments, be published online at http://www.jcse.org in due course. Until such time as it has been fully published it should not normally be referenced in published work. 5 cracks was equally observed on the gauge surface (»130 cracks.mm-2). More generally, cracks were uniformly distributed for elongations higher than 17%. For short elongations (< 17%), the identification of the crack path may have been difficult, because depths were systematically lower than the grain size. Very short cracks (2 ?m) were detected for 0.10 strain, while the transition from initiation to propagation was observed for cracks deeper than 50 ?m after 0.27 strain. The elongation (0.34 strain) of the failed specimen was rather close to the elongation observed in inert gas. Figure 2 allowed to propose, for an apparent strain rate of 5 10 -8 s-1, a slow crack growth rate (CGR) in the initiation stage, for a strain in the range 0.10 - 0.27. The CGR, 0.05 ?m.h -1, was calculated from the slope of the curve represented on Figure 2. Further tests would be necessary to precise the CGR. Crack depth versus strain hardening resulting from CERTs is shown on Figure 3. Strain hardening was quantified by Vickers micro-hardness measured near the initiation areas. Thus, a micro-hardness threshold for true initiation was found close to 250 HV 0,1 and close to 310 HV 0,1 for propagation. However, it was assumed that micro-hardness did not increase in the vicinity of the edges of the cracks during the propagation stage, except at the crack-tip. So, this hypothesis led us to consider that the micro-hardness measured at the end of the test was representative of the strain hardening for initiation. Similarly, the evolution of crack depth versus the equivalent Von Mises stress have been represented on Figure 4, on which initiation corresponded to a stress of 430 MPa (0.10 strain) and the transition to propagation to a stress of 700 MPa (0.34 strain). Additional tests associated to finite element calculations should precise the value of the transition stress during this kind of test. This stress threshold was above the stress measured by X-ray diffraction at the room temperature at the apex of RUBs where no initiation has been observed after 21000 h. Consequently, the overshoot of these thresholds (stress and strain hardening) is not a sufficient condition for initiation and propagation of SCC, at least for short duration tests in laboratory. ISSN 1466-8858 Volume 11, Preprint 1 submitted 20 May 2008 © 2008 University of Manchester and the authors. This is a preprint of a paper that has been submitted for publication in the Journal of Corrosion Science and Engineering. It will be reviewed and, subject to the reviewers" comments, be published online at http://www.jcse.org in due course. Until such time as it has been fully published it should not normally be referenced in published work. 6 y = 3736,3x - 940,09 y = 311,19x - 32,746 0 50100150200250300350400450 0,00 0,05 0,10 0,15 0,20 0,25 0,30 0,35 0,40 StrainDepth of the deepest crack (μm) Initiation of TGSCC Propagation of TGSCC Figure 2. Initiation and propagation stages during CERTs with type A specimens exposed to PWR environment (360°C). Depth of the main crack vs. strain at the end of the test. 050100150200250300350 180 220 260 300 340 Vickers microhardness at the crack-tip Crack depth (μm) 050100150200250300350 160 260 360 460 560 660 760 Equivalent V.M. Stress (MPa) Crack depth (μm) Figure 3 - Micro-hardness at crack tips after CERTs in primary PWR environment (360°C, e&= 5 10-8 s-1). Type A specimens. Figure 4 - Final stress for CERTs in primary PWR environment (360°C, e&= 5 10-8 s-1). Type A specimens. Initiation Initiation Initiation Initiation aaaand nd nd nd PPPPropagation of SCC ropagation of SCC ropagation of SCC ropagation of SCC DDDDuring uring uring uring CECECECERTs RTs RTs RTs wwwwith ith ith ith PPPPrererere----sheared sheared sheared sheared SSSSpecimens pecimens pecimens pecimens Type B specimens were used in this part of the study. Except one specimen, all specimens were double-notched. Above all, pre-shear hardening associated with complex strain paths allowed to reproduce with flat specimens and a perfect homogeneous pre-strain, the mechanical state followed at the intrados of cold pressed humped specimens. The use of flat specimens also permitted a quantitative approach of the effect of strain hardening on SCC, by contrast to the use of cold pressed humped specimens implying contact problems, friction, high heterogeneity and weak reproducibility of the strain hardening. Nevertheless, the section of the specimen was limited to 1 mm thick because of technological ISSN 1466-8858 Volume 11, Preprint 1 submitted 20 May 2008 © 2008 University of Manchester and the authors. This is a preprint of a paper that has been submitted for publication in the Journal of Corrosion Science and Engineering. It will be reviewed and, subject to the reviewers" comments, be published online at http://www.jcse.org in due course. Until such time as it has been fully published it should not normally be referenced in published work. 7 considerations. Two pre-shearing levels were employed (g = 0.2 and g = 0.4) and three strain paths were followed: pseudo-monotonic ( b = +1), reverse or pseudo-Baushinger (b = -1) and cross test ( b = 0). Stress strain curves of the CERTs are plotted in Figure 5, as a function of pre-shear level and strain path change. The equivalent cumulated strain was the sum of the equivalent pre- shear at room temperature (g / 3) and the equivalent strain at the end of the CERT at 360°C. Thus, a limited strain to failure was observed for {g = 0.4 ; b = -1}, associated to a maximum crack depth of 140 ?m, while a strain to failure of 0.22 was noted for the monotonic test, associated to highest crack depth (300 ?m). Finally, the effects of the strain path changes on yield stress were minor for g = 0.4, but significant for g = 0.2. More precisely, stress-strain curves of reverse tests have confirmed that the material presented a Baushinger behaviour. 0100200300400500600 0 0,04 0,08 0,12 0,16 0,2 0,24 0,28 0,32 Cumulated strain Equivalente VM stress (MPa) g = 0, b = (+1) g = 0,2, b = -1 g = 0,2, b = 0 g = 0,4, b = -1 g = 0,4, b = 0 g = 0,4, b = +1 Figure 5 - Stress-strain curves of CERTs in primary PWR environment for several strain path changes (360°C, e&= 5 10-8 s-1). The equivalent stress and the equivalent cumulated tensile strain are considered. g is the pre-shearing and b is the strain path. Figure 6 describes the maximum stress corrosion crack depth initiated at the notch of specimens as a function of their elongation and strain path. Results corresponding to {g = 0.2 ; b = 0} were in agreement with those corresponding to {g = 0.2 ; b = -1}: the shortest elongation to rupture was observed for the cross test ( b = 0), which was the most severe strain path change. CGR was calculated in the initiation stage, considering tests relating to reverse and cross test with 0.2 pre-shear and considering an elongation rate of 4.7 ?m.h -1. The CGR, equal to 0.22 ?m.h -1, was clearly higher than the one measured in non-pre- ISSN 1466-8858 Volume 11, Preprint 1 submitted 20 May 2008 © 2008 University of Manchester and the authors. This is a preprint of a paper that has been submitted for publication in the Journal of Corrosion Science and Engineering. It will be reviewed and, subject to the reviewers" comments, be published online at http://www.jcse.org in due course. Until such time as it has been fully published it should not normally be referenced in published work. 8 strained specimens, with a factor 4. Therefore, strain hardening clearly led to an increase of CGR. 01002003004005006007008009001000 0 1 2 3 4 5 6 7 Elongation (mm) Max crack depth (μm) g = 0,2 / b = -1g = 0,2 / b = -1g = 0,2 / b = -1g = 0,2 / b = -1 g = 0,2 / b = 0g = 0,2 / b = 0g = 0,2 / b = 0g = 0,2 / b = 0 g = 0 / b = (+1)g = 0 / b = (+1)g = 0 / b = (+1)g = 0 / b = (+1) g = 0,4 / b = -1g = 0,4 / b = -1g = 0,4 / b = -1g = 0,4 / b = -1 g = 0,4 / b = 0g = 0,4 / b = 0g = 0,4 / b = 0g = 0,4 / b = 0 g = 0,4 / b = +1g = 0,4 / b = +1g = 0,4 / b = +1g = 0,4 / b = +1 Figure 6. Maximum crack depth versus elongation and strain paths. CERTs (360°C, e&= 5 10-8 s-1) on notched type B specimens. Effect of Effect of Effect of Effect of SSSStrain train train train LLLLocalisation on SCocalisation on SCocalisation on SCocalisation on SCCCCC In CERTs, specimen often fails early after the initiation of the cracking and little information on crack propagation is obtained, especially in pre-strained materials. That is why SCC was focused at particular locations by the use of double notches in the gauge section of the specimen. The effect of strain-localization during CERT appeared comparing two specimens, one with notches and on the other without notch. Pre-strain hardening and subsequent strain path, characterised by {g = 0.4 ; b = -1}, were identical. Results showed that notches favoured initiation of SCC and especially TGSCC. Indeed, observations of fracture surfaces revealed a multitude of intergranular short cracks (< 40 ?m) initiating at the surface of the smooth specimen and a main transgranular stress corrosion crack (80 ?m) for the notched specimen. In the notched specimen, a lot of intergranular short cracks were observed outside the notches. In brief, 0.4 pre-shear hardening associated with a subsequent reverse SCC test was favourable to intergranular initiation, even if the presence of the notch mostly enhanced TGSCC. Effect of the Effect of the Effect of the Effect of the SSSStrain train train train PPPPath on the CGPath on the CGPath on the CGPath on the CGP Pre-shearing had different implications for trans- and intergranular crack depths. The maximum intergranular crack depth was an increasing function of pre-shear hardening for complex strain paths (see Figure 7). Moreover, for reverse and cross SCC tests, the IG crack depth was independent on the strain path for 0.2 pre-shear. In return, for 0.4 pre-shear, IG ISSN 1466-8858 Volume 11, Preprint 1 submitted 20 May 2008 © 2008 University of Manchester and the authors. This is a preprint of a paper that has been submitted for publication in the Journal of Corrosion Science and Engineering. It will be reviewed and, subject to the reviewers" comments, be published online at http://www.jcse.org in due course. Until such time as it has been fully published it should not normally be referenced in published work. 9 crack was clearly deeper for the most mechanically severe strain path (cross test or b = 0). Consequently, it could be concluded that a pseudo-monotonic strain path do not significantly favour initiation of IGSCC (Figure 8) whatever the pre-shear hardening level. Secondly, it could be assumed that IGSCC was enhanced by strain path changes. 020406080100120140160 0 0,1 0,2 0,3 0,4 gggg Max. IGSCC depth (μm) b = -1b = -1b = -1b = -1 b = 0b = 0b = 0b = 0 b = +1b = +1b = +1b = +1 Figure 7. Intergranular crack depth versus pre-shearing for several strain paths. CERTs (360°C, e&= 5 10 -8 s-1) on notched type B specimens. Figure 8. Intergranular crack propagation. CERTs (360°C, e&= 5 10-8 s-1) on notched pre-sheared type B specimen. ISSN 1466-8858 Volume 11, Preprint 1 submitted 20 May 2008 © 2008 University of Manchester and the authors. This is a preprint of a paper that has been submitted for publication in the Journal of Corrosion Science and Engineering. It will be reviewed and, subject to the reviewers" comments, be published online at http://www.jcse.org in due course. Until such time as it has been fully published it should not normally be referenced in published work. 10 Figure 9 show the effect of strain path on maximum transgranular stress corrosion crack depth. First, it could be noticed that TGSCC (Figure 10) initiated whatever the strain path. For a pseudo-monotonic strain path ( b = +1) the depth of transgranular cracking was independent of the level of the pre-shear hardening (» 400 ?m). This suggest that CGR in a pre-strained material (g = 0.4) would be more rapid than in the non-pre-strained one (g = 0), since the durations of the tests were radically different (respectively 537 h and 1368 h for respectively g = 0.4 and g = 0). Then, the case of complex strain path changes was observed. A significant reduction of the maximum crack depth for {g = 0.4 ; b ¹ 1} was noticed, compared to the pseudo-monotonic strain path (g = 0.4 ; b = +1). Thus, the depth of transgranular cracks were 5 times less important for a complex strain path than for a pseudo-monotonic strain path. For g = 0.2, the two complex strain paths have been distinguished. TGSCC was favoured by the pre-strain for b = -1: TGSCC reached 1 mm depth. The moderate 0.2 pre-shear significantly increased the CGR. For the most severe strain path ( b = 0), the depth of crack did not exceeded 400 ?m. This result suggests that mechanically severe strain path ( b = 0) could have limited the propagation of transgranular cracking. 020040060080010001200 0 0,1 0,2 0,3 0,4 gggg Max. TGSCC depth (μm) b=-1b=-1b=-1b=-1 b=0b=0b=0b=0 b=+1b=+1b=+1b=+1 Figure 9. Transgranular crack depth versus pre-shearing for several strain paths. CERTs (360°C, e&= 5 10 -8 s-1) on notched type B specimens. ISSN 1466-8858 Volume 11, Preprint 1 submitted 20 May 2008 © 2008 University of Manchester and the authors. This is a preprint of a paper that has been submitted for publication in the Journal of Corrosion Science and Engineering. It will be reviewed and, subject to the reviewers" comments, be published online at http://www.jcse.org in due course. Until such time as it has been fully published it should not normally be referenced in published work. 11 Figure 10. Transgranular crack propagation. CERTs (360°C, e&= 5 10-8 s-1) on notched pre-sheared type B specimen. DiscussionDiscussionDiscussionDiscussion First, CERTs have supported the primary idea that strain hardening was a prerequisite condition for SCC initiation and propagation in primary PWR environment. Several curves have led to propose thresholds for both initiation and propagation. No SCC has been observed for micro-hardness below 240 HV 0.1, and no propagation under 310 HV0.1. Likewise, an equivalent stress close to 700 MPa seemed necessary for propagation. Nevertheless, the overstepping of these thresholds was not a guarantee for SCC initiation or propagation as demonstrated by sequential tests. Second, CERTs have demonstrated the important effect of the strain path on SCC mechanisms and more precisely on the crack growth path. Briefly, the monotonic strain paths led to pure TGSCC while complex strain paths (reverse and cross SCC tests) favoured IGSCC. Furthermore, IGSCC was an increasing function of strain hardening while TGSCC was first favoured by strain hardening, then decreased when the strain hardening became too important. ISSN 1466-8858 Volume 11, Preprint 1 submitted 20 May 2008 © 2008 University of Manchester and the authors. This is a preprint of a paper that has been submitted for publication in the Journal of Corrosion Science and Engineering. It will be reviewed and, subject to the reviewers" comments, be published online at http://www.jcse.org in due course. Until such time as it has been fully published it should not normally be referenced in published work. 12 Therefore, some mechanical considerations about the strain path changes could put in light some SCC aspects. During plastic strain, the most highly stressed slip systems were activated, leading to the dislocation motion in these planes. After a sufficient amount of monotonic deformation ( b = +1) the dislocation structures evolved toward steady-state configurations as cell block boundaries (CBBs), where dislocations were stored. Generally CBBs were formed, in fcc structure, along the most active {111} cristallographic slip planes. In a reverse test ( b = -1), most of the slip systems that were active during the pre-strain were also active during the second loading, but they were operating in the opposite sense. According to Hu [14], the beginning of the reverse loading lead to the rapid disappearing of unstable dislocation pile-ups, giving rise to an asymmetry of slip resistance. In a cross test ( b = 0), the active slip systems from the first deformation path remained latent while new slip systems were activated. A high resistance to dislocation motion was obtained, because the CBBs formed during the first stage operated as obstacles for the new slip systems. Consequently, changes of strain paths could have two major effects on SCC: first, CBBs formed during the pre-deformation could lead to strong obstacles to dislocations motion and increase SCC in agreement with corrosion enhanced plasticity models [1]. Second, short transient behaviours as Baushinger effect (decreasing of yield strength on reloading in a reverse sense) or cross effect, resulting from micro-plasticity, could have major implications on the enhancement of SCC mechanisms in 304L. One of the noticeable features was that the effects resulting from reverse and cross tests appeared to vanish after an equivalent tensile strain of 0.15-0.20. Afterwards, the initial plastic anisotropy was totally replaced by the anisotropy induced by the new deformation mode. In accordance to SCC observations, it could be assumed that TGSCC was dramatically reduced in the cross SCC test because the motion of dislocations into the grains, during the second strain, was slowed down by the dislocation forest induced by the pre-strain hardening. Furthermore, TGSCC could be favoured by planar glide at the crack-tip, while IGSCC would rather be enhanced by the strain incompatibilities. Additional tests and observations should be carried out to strengthen this hypothesis. ConclusionsConclusionsConclusionsConclusions Series of test was conducted with two types of specimens to investigate the effect of strain hardening on the initiation of stress corrosion cracking in 304L stainless steels exposed to primary PWR enviroment. Pre-shearing tests were used to clarify the effect of the strain path on SCC and more precisely on the crack growth path as suggested by previous studies conducted with humped specimens. The main conclusions were that pre-straining was ISSN 1466-8858 Volume 11, Preprint 1 submitted 20 May 2008 © 2008 University of Manchester and the authors. This is a preprint of a paper that has been submitted for publication in the Journal of Corrosion Science and Engineering. It will be reviewed and, subject to the reviewers" comments, be published online at http://www.jcse.org in due course. Until such time as it has been fully published it should not normally be referenced in published work. 13 necessary for SCC initiation, and intergranular cracking was enhanced by strain incompatibilities due to changes of strain path. Additional efforts, based on aggregate computations and superficial strain localisation evaluation could be done to clarify the effect of strain path and strain hardening on oxidation and initiation of stress corrosion cracking of austenitic stainless steels. Transmission electronic microscopy should also be necessary to identify implied mechanisms. ReferenceReferenceReferenceReferencessss [1] T. Magnin, Chierragatti, Oltra, Acta Metallurgica et Materialia, 38383838, 7, p1313, 1990. [2] 'Slip dissolution model", P. Ford, Corrosion sous contrainte phénoménologie et mécanismes , Ed. de Physique, 1990. [3] 'Etude du mécanisme de fissuration par CSC des aciers inoxydables austénitiques : influence de l"écrouissage, de la température et du potentiel électrochimique", M. Puiggali, Thesis Bordeaux I University, 1981. [4] T. Magnin, A. Chambreuil, Bayle, Acta Materialia, 44444444, p1457,1996. [5] 'IGSCC of non-sensitized stainless steels in high temperature water", P.L. Andresen, M.M. Morra, Journal of Nuclear Materials, in press, 2007. [6] 'Stress corrosion cracking of AISI 304 and 316 stainless steels in lithiated water at 350°C", S.W. Sharkawy, Z. Xia, Z. Szklarska-Smialowska, Journal of Nuclear Materials, 195195195195, pp184-190, 1992. [7] 'Hydrogen induced IGSCC of Ni-containing fcc alloys in high temperature water", N. Totsuka, Z. Szklarska-Smialowska, 3rd Environmental Degradation of Materials in Nuclear Power Systems-Water Reactor , pp691-696, 1988. [8] 'Comparative studies of SCC in two austenitic stainless steels and alloy 600 on exposure to lithiated water at 350°C", Z. Szklarska-Smialowska, Z. Xia, S.W. Sharkawy, Corrosion, 48484848, 6, pp455-462, 1992. ISSN 1466-8858 Volume 11, Preprint 1 submitted 20 May 2008 © 2008 University of Manchester and the authors. This is a preprint of a paper that has been submitted for publication in the Journal of Corrosion Science and Engineering. It will be reviewed and, subject to the reviewers" comments, be published online at http://www.jcse.org in due course. Until such time as it has been fully published it should not normally be referenced in published work. 14 [9] 'Stress corrosion cracking of AISI 304 and 316 stainless steels in lithiated water at 350°C", S.W. Sharkawy, Z. Xia, Z. Szklarska-Smialowska, Journal of Nuclear Materials, 195195195195, pp184-190, 1992. [10] 'Comparative studies of SCC in two austenitic stainless steels and alloy 600 on exposure to lithiated water at 350°C", Z. Szklarska-Smialowska, Z. Xia, S.W. Sharkawy, Corrosion Science, 48484848, 6, pp455-462, 1992. [11] Annual Book of ASTM Standards, Part 11, ASTM, Philadelphia 1978, 205. [12] 'Formation of martensite in austenitic stainless steels", Angel, Journal of the Iron and Steel Institute , 177177177177, 1, pp165-174, 1954. [13] 'A cristal plasticity based worked-hardening/softening model for B.C.C. metals under changing strain paths", Peeters, Kalidindi, Van Houtte, Aernoudt, Acta Materialia, 48484848, pp2123-2133, 2000. [14] Hu, Rauch, Teodosiu, International Journal of Plasticity, 8888, p839, 1992. [15] 'Yield loci for polycristalline metals without texture", Schmitt, Aernoudt, Baudelet, Material Science and Engeneering, 75757575, pp13-20, 1985.