SUMMARY OF THE SESSIONS ON AUSTENITIC ALLOYS: STAINLESS STEELS – Chaired by Anders Jenssen, Katsuhiko Kumagai, Taku Arai and Ian de Curières
Six papers were presented in the stainless steel live session on Wednesday. Four further talks were available as recorded download and four summary/poster presentations were given.
Tetsuo Shoji (Tohoku University, Japan) started the session by giving a keynote presentation entitled “Possible mechanisms of SCC of cold worked stainless steels in LWR environments – role of synergy of hydrogen and vacancy”. The work presented aimed at understanding the mechanism of SCC in cold-worked stainless steels in light water reactor environments. Particular emphasis was paid to the role of hydrogen in the cracking mechanism. Thermal desorption analysis (TDA) was employed to determine the types and quantity of hydrogen trapping defects present in the materials. Types 316LN and 316L in solution annealed (SA), SA + surface machined (SM), and SA + SM + low temperature annealed (LTA) conditions were exposed to a hydrogen gas atmosphere or hydrogenated high-temperature water at 288 °C. TDA indicated that hydrogen preferred vacancy sites in the surface machined condition, whereas hydrogen was evenly distributed between vacancy and lattice sites in the SA condition. It was also found that hydrogenated high-temperature water was more efficient in trapping hydrogen in solution annealed as well as deformed material than exposure to a hydrogen gas atmosphere. The results of the study indicate hydrogen can play an important role in SCC by its interaction with lattice defects such as dislocations and vacancies, which can be introduced at the surface of the material by machining, in the bulk by cold-work and welding, and at the crack-tip by loading and growth.
Dave Morton (Naval Nuclear Laboratory, USA) presented results of a study concerning the effect of specimen orientation on the crack growth rate in hydrogen deaerated water (“Stainless steel SCC growth rate orientation and temperature functionality in hydrogen deaerated water”). Testing was performed on 0.6T C(T) specimens (T-L, L-T and S-L orientations) of four heats of type 304 that had been cold-worked to 5, 7.5, 9, 12, 20 and 28 %. One heat was also tested in a sensitized condition, for which the 19 % cold-roll in one pass was carried out either before or after the sensitization heat treatment. The fastest CGR was observed in the S-L orientation, an orientation that is not relevant for a component. It was found that the CGRs of specimens in the T-L orientation were about 2x faster than the L-T orientation, and the effect was observed at all cold-worked levels investigated. In the heavily cold-worked condition (28 %), all three orientations displayed similar Arrhenius temperature dependencies with an activation energy of ~75 kJ/mol. At lower cold-work levels, departure from the Arrhenius temperature dependency was observed at higher test temperatures. The effect, denoted high temperature CGR retardation (HTR), increased with decreasing degree of cold-work. In addition, the extent of HTR was more pronounced in the sensitized and then cold-worked material. This observation suggested that Cr depletion is the explanation for the greater extent of HTR in sensitized and then cold-worked material.
Caitlin Huotilainen (VTT, Finland) gave a presentation entitled “Environmentally-assisted fatigue in stainless steel 304L – effect of surface finish and martensite formation ahead of the crack-tip”. Specimens of type 304L, in polished or ground surface conditions, were tested in 300 °C air or simulated PWR water at nominal strain amplitudes of ±0.3 and ±0.6 %. Post-test examination showed that the striation spacing increased with strain amplitude and crack length, whereas there was no significant difference in striation spacing between polished and ground specimens. The polished samples had a thin (1 µm) ultrafine grained deformation layer at the surface that increased oxidation and passivation and resulted in increased cracking resistance. On the ground specimens, a 10-20 µm thick deformation layer at the surface was present that tended to localize plastic deformation, while grinding marks acted as initiation sites for EAF. Martensite transformation (γ → ε→ α′) was observed at the crack-tip (far from the surface layer) of specimens tested in simulated PWR water, but not in specimens tested in high-temperature air. It was hypothesized that hydrogen lowers the stacking fault energy at the crack-tip, which in turn can promote the formation of deformation induced martensite.
Eleanor Grieveson (Jacobs, UK) presented newly developed multi-axial fatigue initiation and crack growth specimens (“Design, development and testing of multi-axial fatigue initiation and crack growth specimens”). The motivation of the development is that discounting multi-axial stress state in in-service components can be one of the sources of excessive conservatism in the current safety justifications that is based on uniaxial laboratory experiments. Both crack initiation and growth specimens consist of central “cross” part and “dog-legged” side arms on both sides. Those side arms induce tensile stress upon the central part, generating almost equi-biaxial stress field when a uniaxial load is applied to the entire specimen. So far, fatigue crack initiation tests of type 304 stainless steel have been conducted in air at room temperature and 300 ºC. The stress field evaluated by FEM was experimentally validated using digital image correlation technique, and 2-dimensional “turtle like” cracks have been observed. A fatigue crack growth testing was also carried out on 304 stainless steel at room temperature. It was shown that the crack size can effectively be monitored by DC potential drop technique. Both crack initiation and growth testing in high-temperature water are at the top of the list of future works.
Johan Stjärnsäter (Studsvik Nuclear AB, Sweden) gave a presentation entitled “The effect of specimen size on the stress corrosion crack growth rate in BWR environment”. SCC crack growth rate tests of type 316L stainless steel were conducted using C(T) specimens with three different sizes, 1.5T, 1T and 0.5T, to quantify the effect of specimen size. The test material was a solution annealed, 40 mm-thick plate and was cold-worked uniaxially to 10 % elongation prior to machining the C(T) specimens in T-L orientation. Tests were performed under NWC condition (2 ppm DO at 288 ºC with no impurity injection). Essentially no effect of specimen size on the CGR was observed although the CGRs of the largest specimens (1.5T C(T)) were somewhat lower (by factor of ~2) possibly due to limited intergranular engagement along the crack front. While the larger specimens showed K1.7~2.1 dependencies, the smallest specimens (0.5T C(T)) showed a lower dependency of K0.8. No major effect of crack length was observed within the range of a/W = 0.6~0.7. Based on the observations above, a discussion was made on the K-validity criterion in CGR tests. Although Morton et al. reported a significant increase in CGR of Alloy 600 in PWR environments when [K/YS]2/Beff exceeded 2.0 (where, YS: yield stress, Beff: effective thickness of specimen), this study showed no accelerated CGR even at [K/YS]2/Beff >2.4. This suggests that ASTM E399 criterion is overly conservative and can be exceeded by more than 300 % in case of cold-worked 316L in BWR environments.
The final presentation of this live session given by Sara van Biesen (Laborelec, Belgium) was entitled “Intergranular stress corrosion cracking in SS316 globe valves”. The presentation began with a brief introduction of crack incidents in four grove valves connected to a borated water tank in an actual plant. The incidents were detected by leakage right after the overhaul work that involved removal of seal weld between valve body and bonnet and re-welding afterwards. Additional 6 valves of the same type taken from different unit/circuit were also examined for detailed root cause analyses, which revealed that; (1) IG cracks were found in six valves out of ten, (2) all cracked valves were sensitized while no crack was found in non-sensitized valves, (3) sulfur species were present in the crack of all valves with cracks and (4) additional welding residual stress was supposed to have been imposed by re-welding. From the observations above, it was concluded that the crack was due to IGSCC. As the follow-up, non-destructive examinations have been conducted on 1207 valves of the same type, using metallographic replication technique. Total 77 valves were identified as potentially sensitized and are being replaced (72/77 are completed) and thus, no new leaks from this type of valves due to IGSCC are expected.
Closing remarks were delivered from Anders Jensen, expressing appreciation to all the authors and attendees.
Five further papers were presented as recorded talks (the abstracts are shown here).
Hyeon Bae Lee (KAIST, Korea), “Effect of dissolved hydrogen and zinc concentration on oxide layers formed on type 316 stainless steel exposed to the simulated PWR primary water environment”: The effect of varying dissolved hydrogen and zinc concentrations on the oxide layers formed in the simulated PWR environment was evaluated for type 316 stainless steel. The oxide layers formed in each environment were investigated by using electron microscope and electrochemical analyses. Overall, zinc addition of 30 ppb into PWR primary water substantially increased corrosion resistance of type 316 stainless steel, while increasing DH showed an opposite effect. The measured electrochemical properties of oxide layers showed significant improvement with zinc addition, but slightly deteriorated as DH concentration increased. When zinc is added, the ZnFe2O4 and ZnCr2O4 were formed at inner and outer oxide layers, resulting in lower defect density and higher reaction resistance of oxide film. Meanwhile, both reaction resistance and bode impedance slightly decreased with increasing DH concentration. Overall, the best corrosion resistance was observed in the simulated PWR environment with normal DH concentration and zinc addition.
Tai-Cheng Chen (INER, Taiwan), “Stress corrosion cracking behavior of CF8A stainless steels subjected to different heat treatments in a simulated BWR environment”: In this study, four heat treatments were applied on the as-casted CF8A stainless steel in order to investigate their effect on the SCC susceptibility of the CF8A in a simulated BWR environment. These heat treatments include low temperature solution annealing (1050 °C/2 h, C sample), thermal aging (385 °C/20000 h, A sample), recovery annealing (500 °C/1 h, R sample) and high temperature solution annealing (1200 °C/2 h, S sample). The results indicated that the SCC susceptibility of the tested samples in simulated BWR environment increased in the sequence, S < C < R < A. Thermal aging caused the hardening of δ-ferrite, resulting in the crack initiation along the γ/δ interface. Recovery annealing could mitigate the detrimental effect of thermal aging by partially recovering the embrittled δ-ferrite. High temperature solution annealing resulted in a decrease in δ-ferrite content and a significant change in δ-ferrite morphology, which accounted for the low SCC susceptibility of the S sample relative to that of the other ones in a simulated BWR environment.
Yida Xiong (Tohoku University, Japan), “Effects of dissolved oxygen concentration on low-cycle fatigue life of 316LN austenitic stainless steels in borated and lithiated high-temperature water“: The effects of dissolved oxygen concentration on the low-cycle fatigue life of 316LN in borated and lithiated high-temperature water were discussed. The results showed that the low-cycle fatigue life of 316LN at dissolved oxygen concentration of 50 or 100 ppb was significantly longer than that at dissolved oxygen concentrations of < 5 ppb or 2 ppm. Based on the analysis of the specimens, we proposed that the hydrogen induced cracking caused the shorter low-cycle fatigue life at the dissolved oxygen concentration of < 5 ppb and the higher corrosion potential and the higher potential gradient between the crack-mouth and crack-tip caused the shorter low-cycle fatigue life at the dissolved oxygen concentration of 2 ppm.
Hooman Gholamzadeh (Queens University, Canada), “Dealloying of Ni-Fe-Cr alloys in boiling caustic solutions and its importance in stress corrosion cracking”: Some Ni- and Fe-based alloys are susceptible to dealloying and SCC in hot caustic and lead caustic environments, with a film-induced cleavage mechanism often proposed. In this mechanism, the brittle, nanoporous film that forms on the surface as a result of dealloying fractures under stress and may act as a precursor to SCC. The geometry of this film (e.g., thickness and pore size) and its bonding at the film-metal interface play an important role in crack formation and injection into the substrate material by stress concentration and energy transmission. The geometry of this film could be altered by pH, temperature, time of exposure, and alloy composition. In this study, electrochemical measurements and SCC testing of several Ni- and Fe-based alloys were performed in boiling caustic solutions to understand the effects of alloy composition and time of exposure on the geometry of dealloyed layers. The focus was largely on Alloy 800, the tubing material in CANDU steam generators, and the caustic environment was relevant to the extreme end of conditions proposed for secondary side heat transfer crevices. U-bend sample exposure followed by FIB crack extraction and transmission electron microscopy (TEM) analysis was conducted to evaluate nanoscale chemistry and material changes due to SCC. Also, several post-exposure slow strain rate tensile tests and in-SEM tensile tests followed by fractography were used to study the effects of the presence of a dealloyed layer on the response of Alloy 800 to a tensile load. Results suggest that a brittle dealloyed layer could act as a precursor to cracking of an inherently ductile material even if the corrosive environment is removed after its formation. Also, TEM characterization from this study provides support for dealloying of Alloy 800 and confirms the potential role nanoporosity plays in initiating SCC in boiling caustic environments, provided that sufficient stress is applied.
Xiangyu Zhong (Tohoku University, Japan), “Investigations on the environmental assisted cracking behavior of cold-worked stainless steel in BWR environment“: SCC and environmentally-assisted fatigue (EAF) are the main degradation mechanisms in pressure boundary components of the NPPs due to a variety of transients during operation such as thermal stratification and striping, flow-induced turbulence, and start-up and shut-down. In the present study, the SCC and EAF behavior of 20 % cold-worked 316L stainless steels with and without Sc and Zr addition in simulated BWR environment with NWC (DO = 0.2 ppm) and HWC (DH = 0.5 ppm) were investigated by using SSRT and fatigue tests with hollowed cylindrical specimens. The crack morphology and crack depth distribution were analyzed. Multiple sites of crack growth and coalescence were observed on the inner surface. A lot of longer cracks and short cracks were observed in the necking area. Only a few short cracks were observed in the area far away from the necking area. It indicates that the crack initiated at the early stage before the maximum load. The continued nucleation of new cracks may result in the reactivation of dormant cracks and coalescence between the dormant crack and a subsequently nucleated crack. The number of cracks for all the materials in HWC is larger than that of NWC. The depth of cracks for the materials in HWC is larger than that of NWC. The crack open displacement for the materials in HWC is larger than that of NWC. These evidences suggest that hydrogen plays a role in crack initiation and propagation. SCC susceptibility in HWC is higher than that in NWC, where dissolved hydrogen promoted SCC susceptibility on these cold-worked austenitic stainless steels. The results also show that the combined addition of Sc and Zr can reduce the crack initiation and propagation in both NWC and HWC and therefore improve the SCC and EAF resistance significantly.
Finally, four stainless steel-related summary/poster presentations were also given.
Liberato Volpe (PSI, Switzerland), “Microstructural characterization of machined 316L stainless steel and its environmentally-assisted cracking behavior under BWR conditions”: EAC is one of the major causes of failure events occurring in light water reactors (LWRs). Despite the important role of surface material condition on the EAC susceptibility, there are no comprehensive and rich datasets correlating the surface conditions and microstructure with EAC initiation. Therefore, the effect of standard and novel surface machining processes on the EAC behavior of a cold-worked (CW) type 316L austenitic stainless steel (SS) has been studied within the European Horizon 2020 MEACTOS (Mitigating Environmentally-Assisted Cracking Through the Optimization of Surface Condition) project. In this work, the EAC behavior of a CW 316L SS was investigated as function of three industrial surface machining processes and compared with a reference surface (RS) obtained with a P2000 SiC grinding paper. The surface and near surface microstructure induced during specimen preparation were investigated with a variety of complementary characterization techniques, such as scanning electron microscopy (SEM), transmission electron microscopy (TEM) and nano-hardness analyses. Moreover, CW 316L flat tapered tensile specimens with both RS and machined surfaces, were exposed to a simulated boiling water reactor (BWR) environment at 288 °C under slow strain rate testing conditions to quantify the stress threshold for EAC initiation. The main advantage of using a tapered specimen geometry is that in one single experiment a stress gradient is present along the gauge section, and therefore a critical stress for EAC initiation can be determined in a reasonable time frame. The microstructural material characterization prior to BWR exposure showed that the different surface processes induced the formation of a plastically deformed ultra-fine grained layer and affected the surface and near surface hardness. The microstructural and mechanical investigations prior and after exposure were then correlated with the specific surface machining processes and with the EAC stress threshold values in order to develop a better mechanistic understanding between near surface microstructure, industrial surface condition, EAC and performance of the austenitic CW 316L SS under BWR conditions.
Jaromir Janousek (CVR, Czech Republic), “Initiation of environmentally assisted cracking of AISI 316L steel by exposure to superheated hydrogenated steam”: Superheated H2-steam system has been shown to offer the possibility to accelerate oxidation kinetics while keeping the electrochemical conditions similar to those in a primary water environment. This system was used to initiate EAC on AISI 316L stainless steel SENT-8 specimens to evaluate the effect of applied stress conditions. The specimens were loaded under constant stress with 565 MPa for temperature 360 °C and 545 MPa for 440 °C. Applied superheated H2-steam environment corresponded to 12 cc/kg of hydrogen content in pure water at 320 °C. A crack detection was performed by direct current potential drop (DCPD) system. This work was assessed in the framework of a European collaborative research project MEACTOS (Mitigating Environmentally-Assisted Cracking Through Optimisation of Surface Condition).
Mychailo Toloczko (PNNL, USA), “Chloride-induced SCC growth of 304/316 SS in concentrated NaCl solutions”: In the USA and some other countries, LWR spent fuel is being stored in above ground dry cask storage (DCS) systems much longer than anticipated. These systems utilize a 304 or 316 SS container that is placed within a concrete bunker. Convective air flow cools the stainless steel container but also exposes it to airborne salts. In the USA, these DCS systems are currently stored at the reactor sites that include seaside and inland locations. A variety of salts have been observed on the stainless steel containers, including NaCl. There is concern that a deliquescent salt solution will eventually form on the container surface, thus allowing chloride induced stress corrosion cracking (CISCC) to occur. Early results are presented here on CISCC growth rate measurements of 304 and 316 SS in aqueous solutions. Currently, the effect of stress intensity, cold-work level, and corrosion inhibitor additions is being evaluated.
Keietsu Kondo (JAEA, Japan), “Effect of cold-work and long-term aging on SCC initiation susceptibility of austenitic stainless steel”: The effect of the long-term thermal aging (LTA) on the SCC susceptibility to initiation in cold-worked L-grade austenitic stainless steels has been investigated in JAEA. In this study, the thermal aging treatments at 288 °C for more than 14000 h were conducted on Type 316L SS samples after cold-rolling with 20 % thickness reduction, and creviced bent beam (CBB) testing was performed to both aged and unaged samples. The results showed that an obvious increase of IGSCC susceptibility was induced after the LTA treatment for 14000 h. In order to reveal the enhancement factor of the SCC susceptibility in LTA treated specimens, mechanical and microstructural properties of those samples have been examined. The Vickers micro-hardness testing showed only a slight increase of hardness after the LTA treatment for 14000 h. Examination on microstructural changes induced by the LTA treatment were conducted by means of SEM/EBSD and TEM, and those results suggested that movement and rearrangement of dislocations toward the lower energy configuration such as dislocation cell structure occurred during LTA treatment. Additionally, SEM observations on the sample surface before and after CBB testing indicated that the deformation behavior in LTA treated sample was different from that in non-LTA. We are currently researching to reveal the correlation between the materials change by LTA treatment and the SCC susceptibility, with focusing on not only the microstructural properties but also corrosion behavior during CBB testing.