Influence of PWHT on the sulfide stress cracking susceptibility of 9%Ni low carbon steel

Influence of PWHT on the sulfide stress cracking susceptibility of 9%Ni low carbon steel

Engineering Failure Analysis 104 (2019) 331–340 Contents lists available at ScienceDirect Engineering Failure Analysis journal homepage: www.elsevie...

5MB Sizes 0 Downloads 5 Views

Engineering Failure Analysis 104 (2019) 331–340

Contents lists available at ScienceDirect

Engineering Failure Analysis journal homepage: www.elsevier.com/locate/engfailanal

Influence of PWHT on the sulfide stress cracking susceptibility of 9%Ni low carbon steel

T



S.S.M. Tavaresa,d, , M.L. Lauryaa, H.N. Farnezea, R.V. Landimb, J.A.C. Velascob, J.L.M. Andiac a

Centro Federal de Educação Celso Suckow da Fonseca, Programa de Pós-Graduação em Engenharia Mecânica e Tecnologia de Materiais, Rio de Janeiro, Brazil b Instituto Nacional de Tecnologia, Laboratório de H2S (LAH2S), Rio de Janeiro, Brazil c Instituto SENAI de Tecnologia de Solda, Rio de Janeiro, Brazil d Universidade Federal Fluminense, Departamento de Engenharia Mecânica, Rua Passo da Pátria, 156, CEP 24210-240 Niterói, Brazil

A R T IC LE I N F O

ABS TRA CT

Keywords: 9Ni steel Sulfide stress corrosion cracking Post weld heat treatment

The 9Ni steel low carbon steel is usually applied to pipes, pressure vessels and forged parts working at cryogenic temperatures, due to its high toughness at −196 °C and mechanical strength. Recently, this material was selected to facilities for storage and re-injection of CO2 gas in modern off shore oil and gas platforms. In this new application the steel may be subjected to high pressures in as environment with condensed water, CO2 and traces of H2S. The effect of such a harsh environment on the simulated heat affected zone (HAZ) was investigated by slow strain rate tensile (SSRT) tests. Two post weld heat treatments (PWHT) were reproduced, single tempering (ST) at 575 °C and double tempering (DT) at 670 °C (2 h) and 600 °C (2 h). Specimens without PWHT were more susceptible to sulfide stress cracking (SSC), while the single tempering at 575 °C decreased both the hardness and the SSC susceptibility. Double tempering also increased the resistance to SSC, but was not more effective than the single tempering. As general conclusion, PWHT is strongly recommended to 9Ni low carbon steels subjected to H2S containing environments.

1. Introduction Low carbon Ni alloyed steels are used in low temperature and cryogenic services. The increase of Ni content decreases the ductilebrittle transition temperature (DBTT) and, as a consequence, the minimum service temperature of the steel. For instance, the 9%Nilow carbon steel (“9Ni”) has a DBTT lower than -196 °C, if a correct heat treatment is applied. According to the specification standards ASTM A333 [1], A553 [2] and A353 [3], the recommended heat treatments for 9Ni steel are the quenching and tempering or double normalizing and tempering. Also according to the specifications, the tempering temperature must be within the 565–605 °C range [1–3]. The resulting microstructure consists of tempered martensite and about 5 to 10% of reversed austenite. Austenite is believed to increase the toughness due to an effect of scavenging intertitials, which purifies the matrix (tempered martensite) [4–6]. Tempering in a lower temperature can cause temper embrittlement. For instance, specimens quenched and tempered at 400 °C

⁎ Corresponding author at: Centro Federal de Educação Celso Suckow da Fonseca, Programa de Pós-Graduação em Engenharia e Tecnologia de Materiais, Rio de Janeiro, Brazil E-mail address: [email protected] (S.S.M. Tavares).

https://doi.org/10.1016/j.engfailanal.2019.05.017 Received 13 February 2019; Received in revised form 27 May 2019; Accepted 28 May 2019 Available online 29 May 2019 1350-6307/ © 2019 Elsevier Ltd. All rights reserved.

Engineering Failure Analysis 104 (2019) 331–340

S.S.M. Tavares, et al.

showed intergranular fracture and very low impact toughness at −196 °C [6]. On the other hand, tempering above 605 °C results in increasing amounts of fresh martensite, which also decreases the toughness [6]. Quenching and double tempering (DT) can also be applied to 8–9%Ni steels, with the first tempering carried out at 670 °C, between A1 and A3 temperatures, and the second one at lower temperature (600 °C or 620 °C). This treatment is also known as “Quenching, lamellarizing and tempering (QLT)” [7,8]. One of the effects of this treatment is to increase the reverse austenite content to more than 10%, and this is believed to be a key factor to increase the toughness of QLT processed steels. In addition to the scavenging effect of austenite, some authors also attribute the positive effect of this phase on toughness to its transformation into martensite during deformation (TRIP-transformation induced plasticity effect) [5]. In a previous work [6], the average impact toughness at -196 °C of QLT (or DT) specimens of 9%Ni steel containing 15.6% of austenite was 220 J, while the specimen single tempered at 575 °C had 4.9% of reverse austenite and fractured with 184 J. Both values are much superior to the minimum required (34 J) in the specification standards [2,3]. Welding of 9Ni steels is typically done with dissimilar filler metal. Ni alloy 625 is the most used due to its excellent weldability [9,10]. Due to the low carbon content, 9Ni steel has good weldability, and pre-heating is recommended only in special cases. However, a peak of hardness may be measured in the heat affected zone (HAZ) in the as welded condition. Post weld heat treatment (PWHT) may be necessary, but is only mandatory for thickness higher than 51 mm [11]. Consequently, in practice, PWHT is not performed in thinner components, because it costs time and money. Recently, the 9Ni steel low carbon steel has been used in systems of CO2 reinjection for Enhanced Oil Recovery (EOR) of Floating Production Storage and Offloading (FPSO). In this new application the steel may be subjected to condensed water, CO2 and traces of Hydrogen Sulfide (H2S), which may cause Sulfide Stress Cracking (SSC) and hydrogen Embrittlement (HE). In this work, tensile specimens of 9%Ni low carbon steel were subjected to localized thermal cycle in a Gleeble machine to reproduce the coarse grain heat affected zone (CGHAZ). Two post weld heat treatments were performed: tempering at 575 °C and double tempering (or QLT). Then the specimens were tested by slow strain rate tensile (SSRT) tests in H2S containing media. The goal was to investigate the benefits and disadvantages of post weld heat treatments (PWHT) for 9%Ni low carbon steels. 2. Experimental The chemical composition of the 9Ni studied in this work is shown in Table 1. Cylindrical specimens with 6.5 mm of diameter were machined and processed in a Gleeble® thermal simulator machine model 3500. In order to produce a thermal cycle typical of the coarse grained heat affected zone (CG-HAZ) the peak temperature was set to 1350 °C and the heat input was 1.5 kJ/mm. A Rykalin2D model was used to simulate the thermal cycle (Fig. 1). A thermocouple was carefully welded in the center of the specimen to measure the real temperature and compare to the model. The extension of the region submitted to the thermal cycle was around 9.0 mm, as shown in Fig. 2. Some samples processed in the Gleeble® machine were heat treated, in order to reproduce typical post weld heat treatments (PWHT). Samples without treatment were named AW (“as welded”). Two PWHT were performed, a single tempering at 575 °C (condition T575), and a double tempering at 670o C (2 h) and 600 °C (2 h) (condition DT-670-600). In both PWHT the specimens were water cooled to room temperature. Vickers hardness (HV10) was measured in the region were the thermocouple was welded and in two positions at 3.0 mm left and right. Microstructures were analyzed by optical and scanning electron microscopy (SEM). Two etching procedures were used to prepare specimens for microscopy. Previous austenite grain boundaries (PAGB) were revealed by immersion in a solution of 15 g picric acid, 5 ml of potassium chromate, 625 ml of distilled water and 275 ml of neutral detergent solution. The previous austenite grain size (PAGS) was determined with the aid of Image J software. For general microstructure characterization the specimens were etched with nital 2%. Austenite phase quantification was assessed by magnetization saturation (mS) measurements in a Vibrating Sample magnetometer, using the same methodology described in [6]. The saturation magnetization intrinsic of martensite (mS(i)) was the value measured in a sample quenched from 1000 °C and cold rolled to transform all retained austenite into martensite (231.0 emu/g). The mS values were determined by plotting M × 1/H and extrapolating for 1/H = 0. The austenite volume fraction (AVF) was obtained using the linear Relation (1):

AVF = 1 − mS/mS(i)

(1)

After the heat treatment, the cylindrical specimens were machined to final dimensions of tensile specimens used in SSRT tests. These were small-size specimens of ASTM E-8, with gauge length 11.4 mm and diameter 2.9 mm. The strain rate adopted was 10−6 s−1. Test solution was composed by distilled water with 4000 ppm of acetic acid (CH3COOH), pH =2.7 ± 0.3, which was a simulation of condensate water in the CO2 reinjection system. The solution was de-aerated with N2, and then saturated with a mixture of CO2 (99.2%) and H2S (0.8%). During the SSRT test the flux of CO2/H2S gas was maintained, which means that the vapor phase of Table 1 Chemical composition of 9Ni low carbon steel studied (%wt.) C

Ni

Mn

P

S

Fe

0.063

9.19

1.12

0.004

0.0055

Bal.

332

Engineering Failure Analysis 104 (2019) 331–340

S.S.M. Tavares, et al.

Fig. 1. Thermal cycle produced in the Gleeble thermal simulator to create a coarse grained heat affected zone (CG-HAZ).

Fig. 2. Macrostructure of the longitudinal section of the specimen submitted to Gleeble thermal simulator.

the system had a concentration of H2S equal 8.000 ppm and 8 mbar of H2S partial pressure. The tests were conducted at 22 ± 1 °C. After the fracture, the two parts were cleaned with alcohol and preserved for observation in the scanning electron microscope (SEM) within 24 h. SSRT results in aggressive environment were compared to tests conducted in inert environment (air). The tests were performed in triplicate. Three parameters of susceptibility to hydrogen embrittlement were determined from the SSRT tests, as defined below: Plastic elongation (El) ratio [12]: (2)

KEl = ElH/Elair Reduction in area (RA) ratio [12]:

(3)

KRA = RAH/RA air Energy (Area of σ x ε curve) ratio:

(4)

KENERGY = Area (σxε)H /Area(σxε)air

Fig. 3. (a) Photomicrograph of specimen as welded; (b) image after processing with image J software. 333

Engineering Failure Analysis 104 (2019) 331–340

S.S.M. Tavares, et al.

Table 2 Vickers hardness results (HV10). AW (as welded)

T575

DT-670-600

350 ± 9

257 ± 7

246 ± 7

3. Results and discussion Fig. 3(a) shows the image of specimen AW (as welded) etched to reveal the previous austenite grain boundaries (PAGB). The photomicrograph was improved and cleaned with Image J software, resulting in the image of Fig. 3(b), which was used for PAGS determination by quantitative metallography. From the analyses of 10 fields the PAGS was 33.4 ± 5.7 μm. Table 2 shows the average Vickers hardness of specimens AW (without heat treatment), T575 and DT670-600. The hardness of the base metal was similar to that of T575 condition. The as welded (AW) specimen has a hardness value compatible to the microstructure of fresh low carbon martensite and bainite, as will be shown. The results confirm that a peak of hardness is observed in the CGHAZ of 9Ni steels. One of the concepts of NACE 0175/ISO 15156 standard [13] for materials handling with H2S in oil and gas production is that the hardness must be limited to avoid failure due to sulfide stress corrosion cracking. The limit hardness depends on the material and service conditions. Although 9Ni steels are not listed in this standard, till now, there is a tendency to adopt the same hardness limit of carbon and low alloy steels to 9Ni steels, which is 250 HV. In this case, the T575 treatment results in a hardness slightly superior, while the DT670-600 treatment results in a hardness slightly inferior to 250 HV. The general microstructure of specimen as welded is shown in Fig. 4(a). The detailed image in Fig. 4(a) shows the predominance of martensite (M) and granular bainite (GB). The microstructure of specimen T575 consists of tempered martensite and bainite, as shown in Fig. 5. The microstructure of specimen DT670-600 has an austenite/martensite lamellar morphology (Fig. 6) typically found in steels tempered at high temperatures [14] or with QLT (or DT) treatment [6–8,15]. The quantification of austenite phase with SEM images was not reliable. Thus, the amount of austenite was precisely determined by the magnetic method. Fig. 7 shows the comparison of the magnetization curves of specimens AW, T575 and DT-670-600. The higher the magnetization the lower the austenite volume fraction (AVF). Table 3 shows the mS values and the austenite volume fraction (AVF) using Eq. (1). Fig. 8(a), (b) and (c) show comparisons between SSRT curves in the test solution and in air, of specimens AW, T575 and DT-670600, respectively. Table 4 shows the average values and statistic error of the parameters KEL, KRA and KENERGY. The as welded specimens show the highest embrittlement factors in the SSRT tests. The fracture surface of the AW specimen tested in air show a ductile behavior with large dimples (Fig. 9(a–b)). When the material as welded is tested in the test solution with H2S the fracture surface was also characterized by microdimples, but with smaller size than in the specimen tested in air. Secondary cracks were also observed (Fig. 10(a–b)). Tempering at 575 °C resulted in a pronounced decrease of all embrittlement factors in relation to the as welded condition. From the point of view of hydrogen embrittlement, this PWHT should be applied regardless of the thickness, not only for joints higher than 51 mm. The fracture surface of a specimen T575 tested in the test solution was also characterized by very small dimples and some secondary cracks (Fig. 11(a–d)). The results of specimens DT-670-600 are better than the as welded specimens, but can be considered equivalent to those of T575. Despite of the lower hardness obtained (246 HV), double tempering was not better than the single treatment T575 to reduce hydrogen

Fig. 4. Microstructure of AW specimen. 334

Engineering Failure Analysis 104 (2019) 331–340

S.S.M. Tavares, et al.

Fig. 5. Microstructure of specimen T575.

Fig. 6. Microstructure of specimen DT-670-600.

Fig. 7. Magnetization curves of specimens AW, T575 and QT-670-600.

335

Engineering Failure Analysis 104 (2019) 331–340

S.S.M. Tavares, et al.

Table 3 mS and AVF of specimens AW, T575 and QT-670-600. Specimen

mS (emu/g)

AVF

AW T575 QT-670-600

230.0 224.0 199.6

0.004 0.030 0.136

Fig. 8. Comparisons between curves of SSRT tests of specimens (a) AW; (b) T575 and (c) DT670-600.

Table 4 SSRT results. Parameters KEL, KRA and KENERGY (average from three tests). Condition

KEL

KRA

KENERGY

As welded T575 DT-670-600

0.126 ± 0.030 0.364 ± 0.200 0.257 ± 0.120

0.161 ± 0.011 0.239 ± 0.060 0.216 ± 0.050

0.194 ± 0.037 0.417 ± 0.216 0.310 ± 0.100

embrittlement susceptibility of the HAZ. This is an interesting result because T575 is simplest and cheapest than DT-670-600. Fig. 12 (a–b) show the surface of fracture of the specimen DT-670-600 tested in the solution with H2S. The fracture surface differs from T575, because contain a mixture of quasi-cleavage (~70%) and small dimples (~30%). Semi-quantitative analyses of fracture surfaces were performed in all samples tested, in order to estimate the areas with dimples and with quasi-cleavage modes, with the results summarized in Table 5. 336

Engineering Failure Analysis 104 (2019) 331–340

S.S.M. Tavares, et al.

Fig. 9. Images of the fracture surfaces of specimen as welded (AW) tested in air: (a) general view; (b) dimples in the center part of the specimen.

Fig. 10. Images of fracture surfaces of specimens as welded (AW) tested by SSRT in the test solution: (a) general view; (b) detailed image from region A.

Some previous works investigated the effect of reversed austenite on hydrogen embrittlement (HE) in martensitic [15,16] and supermartensitic stainless steels [17]. Fan et al. [15] studied a low carbon (0.042%) martensitic UNS S41500 steel and found that the increase of reverse austenite improved the HE resistance. They compared the mode of fracture of single tempered (ST) with double tempered (DT) specimens subjected to SSRT tests with hydrogen cathodic charging in 1 mol/l NaOH aqueous solution with thiourea. The ST specimens presented intergranular cracks, while DT had quasi-cleavage fracture. In the present work the mode of fracture of double tempered 9%Ni steel was also characterized by quasi-cleavage, but the single tempered (T575) specimen showed microdimples with some transgranular secondary cracks. Li et al. [16] found that the best HE resistance of a 13-8Mo PH steel was obtained with tempering at 650 °C, and attributed this to the maximum amount of reverse austenite (16%) resulted from this treatment. On the other hand, Solheim et al. [17] concluded that the ductility of SMSSs submitted to hydrogen charging was strongly reduced by the presence of retained austenite. Austenite act as trap for hydrogen and increases its solubility in the steel. During deformation the austenite transforms into a martensite with high hydrogen content, which contribute decisively to the cracking mechanism. Kim et al. [18] found similar results with a 5.5%Ni low carbon steel. They pointed that the reverse austenite transformation into martensite ahead of the crack tip generates a local strain which promotes the fracture. The results obtained in this work seem to corroborate those of Solheim et al. [17] and Kim et al. [18], since specimens QT-670-600 had much higher AVF than specimen T575, but its resistance to hydrogen embrittlement was not superior to that of specimen T575. It must be observed, however, that all these

337

Engineering Failure Analysis 104 (2019) 331–340

S.S.M. Tavares, et al.

Fig. 11. Images of the fracture surface of specimen T575 tested in the test solution with pH2S = 8 mbar: (a) general view; (b) detailed image from region A; (c) region B and (d) region C.

results were obtained with SSRT, which is surely more severe than other tests where constant and lower loads are applied to the specimen. The stability of the austenite in specimens T-575 and DT-670-600 were compared after the SSRT tests by cutting samples of the fracture region 0.5 mm away from the edge, with electro-erosion, and performing the magnetic quantification of austenite. The results are shown in Table 6 among with AVFs before the SSRT tests. When SSRT was conducted in air, almost all austenite phase of T575 and DT-670-600 has transformed into fresh martensite due to the high plastic deformation in inert environment. The specimens T-575 and DT-670-600 tested in air had reduction of area (RA) 79.6% and 83.%, respectively. The specimens T-575 and DT-670-600 fractured in the test solution were much less ductile, with RA = 16.0% and 14.6%, respectively. As a consequence, the austenite transformation was reduced in the tests with test solution. Even so, a higher portion of austenite was converted into martensite in the specimen DT-670-600 than in T-575. The stability of austenite is strongly dependent on its chemical composition, mainly C and Ni contents. Specimen T-575 was tempered close to the A1 temperature, where a small amount of austenite is formed, but this austenite is very stable because contains high C and Ni. Specimen QT-670-600 has higher austenite content, with lower C and Ni contents, which makes it less stable under deformation. The lower stability of the austenite of QT-670-600 aids to explain why this specimen is not more resistant to hydrogen embrittlement and sulfide stress corrosion cracking than specimen T-575. 4. Conclusions The main conclusions of the present study are: 338

Engineering Failure Analysis 104 (2019) 331–340

S.S.M. Tavares, et al.

Fig. 12. Images of the fracture surface in the center of specimen DT670–600 tested in the test solution: (a) Regions with dimples and quasi-cleavage; (b) detail of the region with quasi-cleavage. Table 5 Percentages of dimples and quasi-cleavage areas in fracture surfaces. Condition

SSRT

Fracture appearance

As welded

Air Test solution Air Test solution Air Test solution

Dimples (100%) Small dimples (100%) and secondary cracks Dimples (100%) Small dimples (100%) and secondary cracks Dimples (100%) Quasi-cleavage (70%) and small dimples (30%)

QT575 DT670-600

Table 6 Austenite volume fraction of T-575 and DT-670-600 before and after SSRT tests (specimens cut 0.5 mm away from the fractures surface). Specimen

Before SSRT

SSRT in air

SSRT in test solution

T-575 DT-670-600

0.030 0.136

0.010 0.020

0.025 0.045

The simulated coarse grain heat affected zone (CGHAZ) of 9%Ni steel is susceptible to sulfide stress corrosion cracking in slow strain rate tensile tests in a aqueous solution with 4000 ppm of acetic acid (CH3COOH), pH between 2.7 and 3.3, and 8 ppm of H2S. Single tempering at 575 °C reduced the hardness from 350 ± 9 HV10 (as welded) to 257 ± 7 HV, and was found to reduce the embrittlement susceptibility in SSRT test. This heat treatment should be specified as post weld heat treatment (PWHT) for 9%Ni steels independently of the thickness. Double tempering treatment (670 °C/2 h + 600 °C/2 h) decreased the hardness to 246 ± 7 HV10. Despite of the lower hardness and the higher austenite volume fraction, this PWHT was not more effective to reduce the embrittlement susceptibility of the CGHAZ of 9%Ni steel than the single tempering at 575 °C.

Acknowledgements Authors acknowledge to Brazilian research agencies CAPES, CNPq and FAPERJ for financial support.

Declaration Authors declare that there is no conflict of interests on the content of the article “Influence of PWHT on the sulfide stress cracking susceptibility of 9%Ni low carbon steel” submitted to Engineering Failure Analysis. 339

Engineering Failure Analysis 104 (2019) 331–340

S.S.M. Tavares, et al.

References [1] ASTM A333-16, Standard Specification for Seamless and Welded Steel Pipe for Low-Temperature Service and Other Applications with Required Notch Toughness, ASTM International, West Conshohocken, PA, USA, 2014. [2] ASTM A553-10, Standard Specification for Pressure Vessel Plates, Alloy Steel, Quenched and Tempered 8 and 9% Nickel, ASTM International, West Conshohocken, PA, USA, 2010. [3] ASTM A353-09, Standard Specification for Pressure Vessel Plates, Alloy Steel, Double-Normalized and Tempered 9% Nickel, ASTM International, West Conshohocken, PA, USA, 2009. [4] E.F. Nippes, J.P. Balaguer, A study of the weld heat affected zone toughness of 9%Ni steel, Weld. J. (1986) 237–243. [5] K.J. Kim, L.H. Schwartz, On the effects of intercritical tempering on the impact energy of Fe-9Ni-0.1C, Mater. Sci. Eng. 33 (1978) 5–20. [6] S.S.M. Tavares, C.R. Rodrigues, C.A.S. de Oliveira, C.B. Woyames, J. Dille, Influence of heat treatments on microstructure and toughness of 9%Ni steel, J. Mater. Eng. Perform. 27 (2018) 1530–1536. [7] X.J. Zhang, Microhardness characterisation in developing high strength, high toughness and superior ballistic resistance low carbon Ni steel, Mat. Sci. Tech. 28 (7) (2012) 818–822. [8] J.-B. Ju Jae-il Jang, D. Kwon B-W Lee, Woo-Sik Kim, Effects of microstructural change on fracture characteristics in coarse grained heat-affected zones of QLTprocessed 9% Ni steel, Mater. Sci. Eng. A340 (2003) 68–79. [9] R.E. Avery, D. Parsons, Welding stainless and 9% nickel steel cryogenic vessels, Weld. J. (1995) 40–50. [10] A.E.F.M. Khourshid, M.A. Ghanem, The influence of welding parameters on brittle fracture of liquefied natural gas storage tank welded joint, Mater. Sci. Appl. 4 (2013) 198–204. [11] ASME Code, B31.3-2010, Process Pipping, ASM, USA, 2011. [12] ASTM G129-00, Standard Practice for Slow Strain Rate Testing to Evaluate the Susceptibility of Metallic Materials to Environmentally Assisted Cracking, ASTM International, West Conshohocken, PA, USA, 2010. [13] NACE 0175/ISO 15156 - Part 3, Petroleum and Natural Gas Industries – Materials for Use in H2S Containing Environments in Oil and Gas Production – Cracking Resistant CRAs (Corrosion Resistant Alloys) and Other Alloys, 2nd edition, International Organization for Standardization, 2009. [14] X. Lei, Y. Feng, J. Zhang, A. Fu, C. Yin, D.D. Macdonald, Impact of Reverse Austenite on the pitting corrosion behavior super 13Cr martensitic stainless steel, Electrochim. Acta 191 (2016) 640–650. [15] Y.H. Fan, B. Zhang, H.L. Yi, G.S. Hao, Y.Y. Sun, J.Q. Wang, E.-H. Han, W. Ke, The role of reversed austenite in hydrogen embrittlement fracture of S41500 martensitic stainless steel, Acta Mater. 139 (2017) 188–195. [16] X. Li, J. Zhang, J. Chen, S. Shen, G. Yang, T. Wang, X. Song, Effect of aging treatment on Hydrogen embrittlement of PH 13-8Mo martensite, Mater. Sci. Eng. A 651 (2016) 474–485. [17] K.G. Solheim, J.K. Solberg, J. Walmsley, F. Rosenqvist, T.H. Bjørnå, The role of retained austenite in hydrogen embrittlement of supermartensitic stainless steel, Eng. Fail. Anal. 34 (2013) 140–149. [18] Y.W. Kim, H.J. Kim, W. Morris Jr., The influence of precipitated austenite in hydrogen embrittlement, Metall. Trans. A 17A (1993) 1157–1164.

340