Stress Corrosion Cracking of Ferritic-Martensitic Steels in Simulated Boiling Water Reactor Environment

Ferritic-martensitic (F-M) steels are being considered for application as structural materials and fuel cladding in high-temperature (>400°C) Generation IV reactor concepts.¹  These steels have many favorable properties, such as higher thermal conductivity, lower susceptibility to stress corrosion cracking (SCC), and reduced swelling under irradiation.²  At these temperatures, F-M steels possess higher strength than both zirconium alloys and austenitic stainless steels.²  F-M steels are also being considered as candidate materials for accident tolerant fuel to replace existing zirconium alloy cladding because the reaction with water and steam at high temperatures is less exothermic for F-M steels. Investigations are aimed at evaluating the performance of F-M steels under both normal operating conditions and under accident scenarios, e.g., loss of coolant accident (LOCA). It is expected that these steels will provide an improvement over current design by reducing the oxidation rate at high temperatures, thus reducing hydrogen gas production and delaying the release of fission products into the environment during LOCA conditions.

F-M steel T91 (UNS K90901)⁽¹⁾ has been extensively used for header and steam pipes in ultra-supercritical fossil plants operating up to 593°C.³  A few studies⁴  were performed for evaluation of oxidation and SCC behavior of T91 under simulated supercritical water (SCW) conditions. It was reported⁴  that oxidation of T91 in SCW at 450°C and 25 MPa (for exposure up to 750 h) followed growth kinetics between parabolic and cubic laws. A dual-layered oxide scale mainly composed of an outer magnetite (Fe₃O₄) layer and an inner Fe/Cr spinel layer formed on T91 steel. Similar results were obtained for exposure of T91 to SCW at temperatures between 400°C and 600°C with dissolved oxygen less than 10 ppb.⁵  Oxidation studies on T91 under SCW condition at 500°C and 25 ppb of dissolved O₂ revealed the typical duplex oxide film in which the scale is composed of a dense outer iron oxide layer and an inner Fe-Cr oxide layer.⁶ 

The effects of proton-irradiation on radiation-induced segregation (RIS)^7-9  and phase-stability in T91¹  were extensively investigated. It was demonstrated⁷  that the inverse Kirkendall mechanism can explain RIS in T91 over the temperature range of 300°C to 500°C.⁷  Irradiation of T91 using protons, as well as heavy ions, in the temperature range of 400°C to 500°C to doses in the range of 10 dpa to 100 dpa was found to alter the average size and Cr/Fe ratio of grain boundary carbides, predominantly M₂₃C₆. Further, three types of precipitates were identified, viz. Ni/Si/Mn-rich, Cu-rich, and Cr-rich precipitates, using atom probe tomography examination.⁷  The irradiation temperature was found to influence the precipitate size, density, and composition for both Ni/Si/Mn-rich and Cu-rich precipitates in T91. The effect of proton-irradiation on SCC of T91 was investigated¹⁰  in deaerated SCW at 400°C, and no SCC was observed during constant extension rate tensile (CERT) testing at a strain rate of 3×10⁻⁷ s⁻¹.

This paper focuses on tensile and SCC behavior of F-M steels T91 and APMT in simulated boiling water reactor (BWR) normal water chemistry (NWC) environment under normal operating conditions. Little or no evaluation of T91 and APMT has been conducted on the corrosion and SCC performance under a simulated BWR-NWC environment. In this investigation, SCC of T91 and APMT were evaluated using CERT tests at 288°C using a slow strain rate of 3×10⁻⁷ s⁻¹ in three different alloy conditions, viz. as-received, cold-worked, and proton-irradiated. The stress-strain behaviors of T91 and APMT at 288°C and at a strain rate of 3×10⁻⁷ s⁻¹ were also evaluated in a high purity argon environment.

The chemical compositions of T91 and APMT used in the present investigation are given in Table 1. The T91 alloy used in this investigation was in the form of a forged bar of size 65 mm × 40 mm × 40 mm. The material was received in the normalized (1,050°C for 4.7 h) and tempered (770°C for 4.75 h) condition. The microstructure consisted of tempered martensite laths forming subgrains in a ferrite matrix, described more completely elsewhere.¹¹  The APMT alloy used was also in the form of a forged bar (Heat #241975).

Tensile specimens with a total length of 41 mm and a gage length of 21 mm and square cross section of 2 mm × 2 mm were prepared using electrical discharge machining. Tensile specimens were mechanically polished using SiC paper P#4000 (U.S. grit 1200) followed by electropolishing in a methanol (90%) and perchloric acid (10%) solution at a minimum temperature of −30°C with an applied voltage of 30 V for ~30 s. Electropolishing was done to ensure removal of plastic deformation on the surface introduced during mechanical polishing.

Table 2 describes the different alloy conditions and experimental parameters for CERT tests. Specimens prepared from Alloy T91 are designated as T91-ZZ-YYY and the specimens and specimens prepared from Alloy APMT are designated as AP-ZZ-YYY, where ZZ is the material condition (AR = as-received, CW = cold-worked, IR = irradiated) and YYY is the environment (BWR = boiling water reactor normal water chemistry, ARG = argon). Besides the as-received condition, tensile samples were tested in two additional conditions: a pre-strained condition achieved via tensile loading (in air) to 4% plastic strain at a strain rate of 3×10⁻⁵ s⁻¹, and irradiated to 5 dpa using 2 MeV protons. Details of the irradiation are described next.

One specimen each of Alloy T91 and Alloy APMT in the as-received conditions were irradiated to a dose of 5 dpa using 2 MeV protons in the Michigan Ion Beam Laboratory using a specially designed stage attached to a 1.7 MV Tandetron accelerator. Irradiation was conducted at a dose rate of ~2×10⁻⁵ dpa/s at 360°C. The dose was calculated using SRIM2013^†.¹²  The irradiation parameters produced a nearly uniform damage rate through the first 15 μm of the total proton range of 20 μm; the dpa was calculated using SRIM with displacement energy of 40 eV and in full cascade mode. It may be noted that only one side of the tensile specimen was irradiated and the irradiated area occupied the center 10 mm of the 21 mm gage length across the entire width of the specimen.

CERT tests were conducted in a multi-specimen recirculating autoclave system in the High Temperature Corrosion Laboratory at the University of Michigan. All CERT tests were performed at a temperature of 288°C and a strain rate of 3×10⁻⁷ s⁻¹. For tests in a simulated BWR-NWC environment, the temperature within the autoclave was maintained at 288±1°C and the pressure within the autoclave was maintained at 9.65±0.35 MPa. The conductivity of the water at the inlet was <0.06 μS/cm and at the outlet was <0.1 μS/cm throughout the experiment. The dissolved oxygen level was maintained at 2 ppm by purging a mixture of argon and oxygen throughout the experiment and during the heating of the vessel. To isolate the effect of environment on SCC susceptibility, an additional CERT test was conducted in a high purity argon environment. The pressure of argon gas within the autoclave was maintained slightly above atmospheric pressure to prevent the ingress of air into the autoclave system.

Straining of the T91-IR-BWR specimen was conducted in two steps, with the first strain increment of 6.5%, followed by straining to fracture. Similarly, the straining of AP-IR-BWR was conducted in two steps, each of 5%, for a total of 10%. For tests in BWR-NWC, the stress value reported is the sum of the applied stress and stress resulting from the pressure in the autoclave.

After fracture during CERT tests, the gage surface and fracture surface were examined using a scanning electron microscope (SEM; JEOL-JSM-6480^†). The fracture surface examination was conducted to determine the nature of fracture (ductile, brittle, or both), and gage surface examination of specimens was conducted to determine the presence of cracks and general microstructural features near the necked region. After SEM examination, oxide films were removed from the specimens to better resolve small surface cracks, which might be masked because of the presence of the oxide film formed during the exposure to a simulated BWR environment.

Oxide films formed on tensile specimens after CERT tests in simulated BWR-NWC environment were characterized using Raman analysis. Raman analysis was conducted on specimens after exposure to a simulated BWR-NWC, as well as to high purity argon, to determine the nature of the oxide film formed during exposure. Raman analysis was done using a Renishaw inVia^† Raman microscope equipped with a RenCam CCD^† detector with resolution of 1 cm⁻¹, 1,800 lines/nm grating, and 50 μm slit. An objective lens of 20× was used to give a laser beam spot size of ~1 μm. The spectra were excited using 633 nm red laser with power of 9 mW. Specimens were not tilted during spectra acquisition and the incident angle for the laser beam on specimen was 90°. The acquisition time was 15 s and each spectrum was collected twice. Each spectrum was scanned between wave numbers 1,450 cm⁻¹ to 150 cm⁻¹. All experimental parameters were kept identical between different specimens.

Figure 1 shows SEM micrographs of the as-received alloy after etching with Villella’s reagent. The Villella’s reagent consisting of 95 mL of ethyl alcohol, 5 mL of hydrochloric acid, and 1 g of picric acid. Etching was done by swabbing specimens for less than 1 min. The as-received microstructures of both alloys consisted of fine, discrete precipitates at grain boundaries and within the matrix. It is reported that the microstructure of this heat of T91 consists of chromium-rich carbides M₂₃C₆ at prior austenite grain boundaries (PAGB) and on subgrain boundaries,¹¹  and (V, Nb) carbonitrides precipitated mainly on dislocations within subgrains.¹¹  Secondary carbonitrides containing V and Nb usually form during tempering. A higher magnification view (Figure 1[b]) indicates that the microstructure was full of precipitates distributed uniformly throughout the matrix. The SEM analysis indicated no retained austenite in the microstructure. Similar microstructure was obtained for the as-received specimen of the APMT alloy, as shown in Figure 1(c). Precipitates were observed at PAGB, as well as within the matrix, similar to that for the T91 steel.

Stress-strain behavior of different alloy conditions in CERT tests are shown in Figure 2 and the results are summarized in Table 3 for T91 specimens. The yield strength for T91 specimens in different conditions ranged between 427 MPa and 490 MPa and the total elongation was in the range of 4.7% to 11.5%. The uniform elongation was in the range of 2.8% to 7.7%, with the lowest uniform elongation observed for the T91-CW-BWR specimen. As noted in Table 3, the total elongation and uniform elongation for the T91-AR-ARG and T91-AR-BWR were similar.

The total elongation for T91-CW-BWR and T91-IR-BWR was 6% and 8%, respectively. It should be noted that T91-CW-BWR was pre-strained to 4% before the CERT test. Also, straining of sample T91-IR-BWR was interrupted at 6.5% strain (total elongation) for examination of stress corrosion cracks on the surface. This step is referred to as first strain increment and straining from this point onward until the fracture is referred to as second strain increment. It was expected that the total elongation in the CERT test would be lower for these two alloy conditions.

As shown in Figure 2(a), work-softening was observed for the T91-CW-BWR and T91-IR-BWR samples. The reduction in area was the highest for T91-AR-BWR specimen (64%) and the reduction in area for all specimens was in the range of 39% to 64%. The lower reduction in area for T91-CW-BWR and second strain increment for the T91-IR-BWR condition was expected because of prior deformation in these two alloy conditions. The different slope of the elastic region of the stress-strain curve for the T91-AR-ARG was a result of the use of a different loading system with different compliance.

The stress-strain behaviors of T91-AR-ARG and T91-AR-BWR were identical, and T91-CW-BWR exhibited work-softening immediately upon reaching the yield-point. The T91-CW-BWR specimen was pre-strained prior to CERT test. The irradiated sample, T91-IR-BWR, underwent two strain increments, the first to 6.5% strain and the other an additional 1.6% to failure at a total of 8.1% strain. The dislocation density for the T91-CW-BWR was higher than for T91-AR-ARG and T91-AR-BWR. Work-softening was reported earlier²  for T91 during CERT tests at a temperature of 500°C in high purity argon, as well as SCW environments. It is reported that for ferritic-martensitic steel, the movement of lath and block boundaries can absorb dislocations.¹³  This corresponds to a dynamic recovery process that leads to work-softening.¹³  The depth of the irradiated zone of sample T91-IR-BWR was approximately 20 μm for 2 MeV protons. Because the volume of the irradiated zone is negligible compared to the un-irradiated region, the influence of the irradiated microstructure on bulk stress-strain behavior was expected be minimal, if not negligible.

Dynamic strain aging (DSA) was observed for all specimens; for T91-AR-ARG, the onset of DSA occurred at a strain value of 5.7%. The strain at onset of DSA (critical strain for DSA) at a given temperature and strain rate was similar for T91-AR-ARG, T91-IR-BWR, and T91-AR-BWR specimens, and was in the range of 5.7% to 6.0%. The critical strain for the onset of DSA for T91-CW-BWR was significantly lower, at 2.7%. The onset of DSA occurred immediately upon reaching the yield point for T91-CW-BWR. The stress-amplitude in the DSA region, for different alloy conditions, is also tabulated in Table 3 and was in the range of 26.5 MPa to 33 MPa for all conditions. The stress amplitude for the T91-CW-BWR was the lowest among all conditions, with the average value of 26.5 MPa.

Dynamic strain aging was reported earlier for T91 steel in different regimes for temperature and strain rate.^14-16  It was reported that T91 steel exhibited DSA in the temperature range of 100°C to 450°C and the strain rate in the range of 8×10⁻³ s⁻¹ to 4×10⁻⁵ s⁻¹. The serrations observed for T91 steel in this experiment (at 288°C and at a strain rate of 3×10⁻⁷ s⁻¹) are Type C serration. Type C serrations for T91 steel have not been reported thus far. It may be noted that the critical strain associated with the onset of Type C serration is higher than¹⁷  for Type A and Type B serrations, and the physical processes responsible for Type C serration are different than for Type A, B, or D serrations.

Figure 2(b) shows stress-strain curves for the APMT specimens and results are summarized in Table 3. The total elongation for AP-AR-ARG was 23.3% compared to 11.5% for T91-AR-ARG, and the uniform elongation for AP-AR-ARG was ~18% vs. 7.7% for T91-AR-ARG. The reduction in area for AP-AR-ARG was ~67%. The yield strength for all specimens, except AP-CW-BWR, was in the range of 375 MPa to 430 MPa. The yield strength for AP-CW-BWR was higher (570 MPa) because of prior straining (4%). Dynamic strain aging was observed for AP-AR-ARG and AP-IR-BWR, with the average stress amplitude of 30 MPa and 18 MPa, respectively. As observed for T91 specimens, DSA in APMT specimens were of Type C; however, the strain at which DSA began was higher for the APMT alloy (16%) compared to T91 steel (6%). The average stress amplitude in the DSA region was more or less same for both alloys.

Figures 3(a) and (b) show SEM micrographs of the fracture surface of the T91-AR-ARG specimen, exhibiting ductile fracture (Figure 3[a]) with typical dimple-like structure. A few oxide particles can be seen on the edges of the dimpled region. A few locations on the fracture surface showed brittle fracture. On the surface of the T91-AR-ARG specimen, severe deformation could be observed near the necked region (Figure 3[b]) and deformation bands were visible close to the fracture surface. On the gage surface and away from the fracture region, wavy-slip bands could also be observed, as shown in Figure 3(b). However, the number of slip bands was fewer and most of the slip bands in the region away from the neck were oriented perpendicular to the loading direction. Figures 3(c) and (d) show SEM micrographs of the fracture surface of AP-AR-ARG, showing ductile-brittle fracture (Figure 3[c]). A higher magnification image of a dimpled region of the same specimen is shown Figure 3(d). In contrast to the fracture surface of T91-AR-ARG, oxidation of the edges of dimpled regions was not observed for AP-AR-ARG. It may be noted that T91-AR-ARG and AP-AR-ARG were strained simultaneously during a single straining experiment. Lack of oxidation on dimpled regions of the AP-AR-ARG fracture surface could possibly be a result of better oxidation resistance of APMT alloy because of its higher Cr content (22 wt% versus 9 wt% for T91) and the presence of Al as an alloying element.

Similar to the T91-AR-ARG specimen, the fracture surface of T91-AR-BWR specimen showed mixed mode fracture. In addition, as shown in Figure 4(a), oxidation was observed in the ductile region of the fracture surface. A few locations on the fracture surface also showed brittle-fracture (Figure 4[b]). A total of three cracks, perpendicular to the loading direction, were observed in the region away (~1.5 mm) from the neck, as shown in Figure 4(c). Further away in the necked region, slip bands were observed oriented perpendicular to the loading direction.

Sample AP-AR-BWR was not strained to failure, but to a total strain of 11% (Table 3). A typical surface appearance of this specimen after CERT test in simulated BWR environment is shown in Figure 4(d). A few needle-shaped oxide particles (covering <5% area) were observed, possibly because of oxidation of martensitic plates/laths with oxide particles retaining the shape of the martensitic plates/laths. No stress corrosion cracks were observed on the surface of AP-AR-BWR.

Microstructure features on the fracture surface and of the T91-CW-BWR specimen were similar to those found on the T91-AR-BWR specimen. The mode of fracture for T91-CW-BWR was also mixed ductile and brittle fracture, as observed for T91-AR-ARG and T91-AR-BWR specimens. Oxidation of the dimple region was observed and edges were more oxidized than the interior on the fracture surface. Severe deformation was observed in the neck region, while slip bands away from the fracture region were preferentially oriented perpendicular to the loading direction. No stress corrosion cracks were observed on the surface of the T91-CW-BWR specimen. Microstructure features on AP-CW-BWR after the CERT test (total strain: 11%) in simulated BWR environment were identical to that observed for AP-AR-BWR. Also, no stress corrosion cracks were observed on the surface of AP-CW-BWR.

After the first strain increment (strained to 6.5%) of irradiated sample T91-IR-BWR, a large crack was observed in the center of the gage section, as shown in Figure 5(a). Severe oxidation was observed within the newly created crack surface, as well as in the vicinity of this crack. The surface appearance of the region surrounding the crack shown in Figure 5(a) indicates that the region had undergone severe plastic deformation compared to other regions on the surface. Two additional cracks were observed, as shown in Figure 5(b), near the vicinity of the crack shown in Figure 5(a). The morphology of the cracks shown in Figures 5(a) and (b) indicates that the cracking was not intergranular.

As observed for other specimens, e.g., T91-AR-ARG in Figure 3(b), wavy deformation bands were also observed ~2 mm away from the center of the gage section. Deformation bands further away from this region were oriented perpendicular to the loading direction. That is, in regions of severe deformation (e.g., in the vicinity of the center of the gage section), plastic-deformation bands appeared randomly oriented. Further away in regions of comparatively less plastic-deformation, deformation bands were perpendicular to the loading direction. Deformation bands observed in T91 steel were wavy in nature, likely a result of the large number of slip systems in the body-centered cubic (bcc) structure.¹⁸  Dislocations can readily move from one slip system to another by cross-slip, giving rise to wavy slip bands. In contrast to this, slip bands in face-centered cubic (fcc) materials are linear because of difficult cross-slip resulting from low stacking-fault energies of such materials.¹⁸ 

No cracks were observed on any of the un-irradiated sections of sample T91-IR-BWR specimen. Further observation revealed that the morphology the oxide particles was the same in the irradiated and un-irradiated regions.

Figures 5(c) and (d) show micrographs of the surface of T91-IR-BWR, taken from the necked region near the fracture location. Secondary cracks were observed at these locations that possibly initiated from the central crack in the neck region. These cracks were not perpendicular to the loading direction. Also, severe plastic deformation was observed near the fracture surface in the necked region. The cracks that appeared in the center of the necked region after the first strain increment were perpendicular to the loading direction. And, as was the case for all other specimens, severe plastic deformation could be observed near the fracture surface. Like all other specimens, mixed mode fracture was observed with oxidation of the dimpled region at a few locations. Also, the edges were more oxidized compared to the interior of the fracture surface. Only three cracks were observed in the irradiated region of T91-IR-BWR and no cracks were observed on un-irradiated sections on the same specimen.

The morphology of cracks observed in the irradiated zone of T91-IR-BWR specimen was similar to that of cracks typically observed in the necked region of a ductile material. These cracks were shorter in length and the width/length ratio was higher than typical stress corrosion cracks. This indicates that the observed cracks grew more because of applied stress and the role of environment for the growth of such cracks was limited. Also, intergranular fracture was not observed on the fracture surface of T91-IR-BWR. Proton-irradiated alloys are not expected to undergo cracking in the absence of corrosive medium, e.g., simulated BWR environment.

It is known that irradiation leads to localized deformation in the form of dislocation channels formed during tensile loading of irradiated alloys.^19-20  It was also demonstrated that localized deformation plays the primary role in increasing susceptibility to irradiation assisted SCC of austenitic alloys in simulated BWR environment.²⁰  In these experiments, the number of cracks observed was the same in the un-irradiated and irradiated cases. However, in the irradiated sample, secondary cracks were found that possibly resulted from the primary crack. The presence of these secondary cracks in the irradiated region on T91-IR-BWR indicate that irradiation may enhance susceptibility of T91 steel to SCC in simulated BWR environment, but the degree of enhancement is small.

The irradiated specimen of APMT alloy (AP-IR-BWR) was strained to a total strain of 10%. The straining was interrupted at 5% strain to observe the presence of cracks (if any) on the surface. A second strain increment of an additional 5% strain was then applied under the same conditions. Surface appearance of the irradiated region of AP-IR-BWR is shown in Figure 6. Dislocation channels could be observed, indicating localized deformation because of prior irradiation. Figure 6(a) shows dislocation channels after the first strain increment, showing preferential oxidation on dislocation channels. Some of dislocation channels created a step on the surface, as indicated by the arrow in Figure 6(b). Further, the density of oxide particles increased, as expected after the second strain increment, as shown in Figure 6(c). Also, dislocation channels at this particular location showed little or no oxidation, while the surrounding region had undergone oxidation.

No stress corrosion cracks were observed (either in the irradiated or the un-irradiated region) on the surface of AP-IR-BWR after 10% strain. The absence of cracking in the irradiated region of AP-IR-BWR indicates that APMT alloy was resistant to cracking in simulated BWR-NWC environment, at a strain rate of 3×10⁻⁷ s⁻¹. Deformation bands were observed only for AP-IR-BWR but not for T91-IR-BWR. Unlike austenitic stainless steels,^19-20  dislocation channels in the irradiated region of APMT alloy were wavy in nature. As explained earlier, this could be a result of the martensitic structure or the number of available slip systems.

Recent crack growth rate investigations^21-22  on the same alloys have shown that both steels (T91 and APMT) have low crack growth under less aggressive loading conditions in simulated BWR-NWC environments with 2 ppm of dissolved O₂ and 30 ppb of sulfates as H₂SO₄. For example, 23% cold-forged APMT alloy showed a crack growth rate of 1.2×10⁻⁸ mm/s, which is very low compared to austenitic alloys such as Type 304 stainless steel (UNS S30400) and Alloy 600 (UNS N06600).²¹  Further, the crack growth rate stalled in T91 steel when loading conditions were made less aggressive. This confirms the results obtained in the present investigation that F-M steels T91 and APMT have low susceptibility to SCC in BWR-NWC environments. This could possibly be a result of better oxidation resistance of APMT alloy because of the presence of highly oxidization elements Al and Cr. The presence of these elements may result in rapid repassivation of the surface during tensile loading. Also, the Cr content of T91 alloy was 8.2 wt% compared to 22 wt% for APMT alloy, making T91 less resistant to oxidation.

Raman spectra obtained on T91 steel in different conditions are shown in Figures 7(a), (b), and (c) for T91-AR-ARG, T91-AR-BWR, and T91-IR-BWR, respectively. The peaks obtained were compared with peaks of different possible compounds that can form on Fe-Cr alloys in similar environments. Comparison with the available literature data²³  indicated that the oxide film formed on all alloy conditions consisted mainly of α-Fe₂O₃ and Fe₃O₄. Intensities of various peaks for specimens exposed to BWR-NWC were higher than corresponding peak intensity for the specimen tested in high purity argon environment. The Raman peak at 667 cm⁻¹ is a characteristic peak for Fe₃O₄ and has the highest intensity, as shown in Figures 7(a) through (c). The rest of the peaks were of α-Fe₂O₃.

Figure 7(d) shows Raman spectra for AP-IR-BWR, obtained using identical parameters used for T91 specimens. Raman spectra for APMT consisted of the major peak of Fe₃O₄ at 663 cm⁻¹, while all other peaks were of α-Fe₂O₃. Raman spectra for AP-AR-BWR and AP-CW-BWR were identical in nature to that of AP-IR-BWR, both for peak positions and intensities of peaks. This implies that the nature of oxide film formed on different alloy conditions after exposure to simulated BWR-NWC condition was the same, unlike T91-AR-ARG, which showed an oxide film (Figure 7[a]) after CERT test in high purity argon. Raman analysis on the surface AP-AR-ARG did not reveal any oxide peaks, indicating the presence of a very thin oxide film or no oxide film.

The thickness of the oxide film in the irradiated region of sample T91-IR-BWR was estimated by sectioning the sample using the focused ion beam technique. The oxide film formed on the surface was protected by platinum deposited before sputtering. Figure 8 shows a cross-sectional view of oxide film; oxide particles, platinum deposition, and the metal/oxide interface are marked in the micrograph. The thickness of the oxide film was observed to vary between 200 nm and 650 nm, with an average thickness of 405 nm.

The absence of stress corrosion cracks (after CERT tests with a strain rate of 3×10⁻⁷ s⁻¹) in APMT specimens, both in the un-irradiated and proton-irradiated conditions, indicate that APMT is resistant to SCC in a simulated BWR-NWC environment. The absence of significant stress corrosion cracks in T91 specimens in as-received (T91-AR-BWR) and pre-deformed (T91-CW-BWR) conditions in the same environment indicates that T91 has low susceptibility to SCC in a BWR-NWC environment. Irradiation resulted in increased secondary cracking, but the overall enhancement of SCC was extremely small. Overall, results confirm that in comparison to austenitic stainless steels, ferritic-martensitic steels are much more resistant to SCC in BWR-NWC environment.

Support for this work was provided by the United States Department of Energy under the grant DE-NE0000568. The authors gratefully acknowledge the facilities provided by Michigan Ion Beam Laboratory (MIBL) and the High Temperature Corrosion Laboratory at the University of Michigan. The authors would also like to thank Fabian Naab of MIBL for his support in conducting the irradiations.