- Email: [email protected]
martensitic steel T91
Journal of Nuclear Materials 490 (2017) 305e316
Contents lists available at ScienceDirect
Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat
Ion-irradiation-induced microstructural modifications in ferritic/ martensitic steel T91 Xiang Liu a, *, Yinbin Miao a, b, Meimei Li b, Marquis A. Kirk b, Stuart A. Maloy c, James F. Stubbins a, d a
Department of Nuclear, Plasma, and Radiological Engineering, University of Illinois at Urbana-Champaign, Urbana, IL 61801, USA Nuclear Engineering Division, Argonne National Laboratory, Argonne, IL 60439, USA Materials Science and Technology Division, Los Alamos National Laboratory, Los Alamos, NM 87545, USA d International Institute for Carbon Neutral Energy Research (WPI-I2CNER), Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan b c
g r a p h i c a l a b s t r a c t
a r t i c l e i n f o
a b s t r a c t
Article history: Received 22 July 2016 Received in revised form 18 February 2017 Accepted 25 April 2017 Available online 27 April 2017
In this paper, in situ transmission electron microscopy investigations were carried out to study the microstructural evolution of ferritic/martensitic steel T91 under 1 MeV Krypton ion irradiation up to 4.2 1015 ions/cm2 at 573 K, 673 K, and 773 K. At 573 K, grown-in defects are strongly modified by blackdot loops, and dislocation networks together with black-dot loops were observed after irradiation. At 673 K and 773 K, grown-in defects are only partially modified by dislocation loops; isolated loops and dislocation segments were commonly found after irradiation. Post irradiation examination indicates that at 4.2 1015 ions/cm2, about 51% of the loops were a0 =2〈111〉 type for the 673 K irradiation, and the dominant loop type was a0 〈100〉 for the 773 K irradiation. Finally, a dispersed barrier hardening model was employed to estimate the change in yield strength, and the calculated ion data were found to follow the similar trend as the existing neutron data with an offset of 100e150 MPa. © 2017 Elsevier B.V. All rights reserved.
Keywords: Dislocation loops In situ ion irradiation Irradiation hardening Ferritic/martensitic steel Transmission electron microscopy
1. Introduction In metals and alloys, energetic particle bombardment creates displacement cascades and results in the formation of freely migrating defects and defect clusters. The cascade evolution can be described by four stages, including collisional, thermal spike, quenching, and annealing [1]. The annealing stage is a diffusion
* Corresponding author. 216 Talbot Laboratory, 104 South Wright Street, Urbana, IL 61801, USA. E-mail address: [email protected] (X. Liu). http://dx.doi.org/10.1016/j.jnucmat.2017.04.047 0022-3115/© 2017 Elsevier B.V. All rights reserved.
controlled process that is most influenced by the irradiation temperature. As a result, the size, density, and even spatial distribution of dislocation loops are temperature dependent. In a-iron and Febase body centered cubic (BCC) alloys, a particular effect of the irradiation temperature is that it affects the stability of a0 =2〈111〉 loops with respect to a0 〈100〉 loops [2]. By contrast, almost no radiation-induced a0 〈100〉 loops have been observed in other BCC alloys [3,4]. Modeling efforts have been carried out to explain the formation of a0 〈100〉 loops [2,5e7]. Eyre and Bullough [6] proposed that a0 〈100〉 loops could be formed by reaction 12 ½110 þ 12 ½110/½010
306
and
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
a0 =2〈111〉
1 ½110 2
loops
could
be
formed
by
reaction
1 ½001/1 ½111. 2 2
þ The proposed mechanism involves the formation of a0 =2〈111〉 type faulted loops, which was regarded unfavorable due to the high stacking fault energy. Recently, atomistic simulations have observed the formation of a0 〈100〉 loops involving the interactions of a0 =2〈111〉 loops [5,7]. Dudarev et al. [2] explained the relative stability of a0 〈100〉 and a0 =2〈111〉 loops by the reduction of the shear stiffness near the a g phase transition temperature, and pointed out that a0 =2〈111〉 loops are more likely to transform to a0 〈100〉 loops at high temperatures. Experimentally, both a0 =2〈111〉 and a0 〈100〉 loops have been reported in irradiated a-iron and Fe-Cr model alloys in previous studies [8e18]. Both loop types were reported by Gelles et al. in a series of Fe-Cr-C-Mo ferritic alloys neutron-irradiated at 673 K723 K [9]. Katoh et al. examined multiple Fe-Cr binary ferritic alloys following neutron irradiation at 698 K and found the loops are predominantly a0 〈100〉 type, only few a0 =2〈111〉 loops were confirmed in Fe-12Cr [11]. 86% of the dislocation loops in pure iron irradiated by neutrons at 573 K were identified as a0 〈100〉 loops by ndez-Mayoral and Go mez-Bricen ~ o [17]. Yao et al. irradiated Herna a-iron with Feþ ions and found that the dominant loop type changes from a0 =2〈111〉 to a0 〈100〉 as the irradiation temperature increases from 573 K to 773 K [18]. The results on simple systems such as pure iron and Fe-Cr model alloys provides a good starting point for the irradiation studies on more complex alloy systems such as ferritic/martensitic (F/M) steels. The additional alloying elements atoms in F/M steels introduces extra strain fields, which can decrease the mobility of defects and may also affect the formation, growth, and interactions of dislocation loops. For instance, chains of extended loops observed in pure iron [19] have not been observed in F/M steels. T91 is a representative 9Cr-1MoVNb F/M steel that has promising applications in advanced reactor systems. Ion and neutron irradiations on complex F/M steels such as T91 have been carried out by several groups [20e23]. However, the irradiation-induced dislocation loops in F/M steels have not been fully characterized and it is still not clear how the complex microstructure evolves differently under ion irradiation at different temperatures and what role dislocation loops play in the microstructural evolution process. In this study, transmission electron microscopy (TEM) was used to investigate the real-time microstructural evolution of T91 F/M steel during in situ Krypton ion irradiation at 573 K, 673 K, and 773 K, up to 4.2 1015 ions/cm2. The temperatures were chosen because it was expected that the dominant Burgers vector of dislocation loops might change in the selected temperature range [18]. In situ irradiation damage processes including dislocation loop formation, interactions of loops with grown-in defects (line dislocations and grain boundaries), and dislocation segment/network formation at three different temperatures were obtained. Distinct dislocation structures after irradiation were discussed. The average dislocation loop sizes and densities at different doses and temperatures were compared. The change in yield strength DsY was calculated for each dose and temperature condition using a dispersed barrier hardening model. The calculated data were compared with existing proton data and neutron data. Finally, to a limited extent, attempts were made to correlate the ion data with neutron data.
Table 1 Chemical composition of the as-received T91 in wt.%. Fe Bal. S 0.006
C 0.089 Si 0.28
Cr 9.24 P 0.021
Cu 0.08 V 0.21
Mn 0.47 Nb 0.054
Mo 0.96 Co 0.019
Ni 0.16 N 0.035
into 3 mm discs by electrical discharge machining (EDM). The discs were mechanically thinned to 100 mm and then eletropolished to electron transparency with a Struer Tenupol-5 twin-jet polisher using electrolyte containing 5% perchloric acid and 95% methanol at 30 C and 23 V. The in situ ion irradiation experiments were performed at the IVEM-Tandem facility, Argonne National Laboratory (ANL). The experimental setup is similar to the one described by Li et al. [24]. A Hitachi 9000 NAR electron microscope operated at 200 kV was used for TEM imaging. T91 thin foil specimens were irradiated by 1 MeV Krypton ions at 573 K, 673 K, and 773 K up to a fluence of 4.2 1015 ions/cm2. The incident ion beam was 30 to the electron beam and on average ~15 to the foil normal. At ~15 tilt, grains satisfying desired diffraction conditions were selected for TEM observations. The ion irradiation was paused for 0.5e1 h on average at 6.0 1014, 1.8 1015, and 3.0 1015 ions/cm2 in order to carry out TEM examinations. The samples tilts were adjusted accordingly (typically ± 1e3 ) to ensure that the TEM micrographs were taken under similar diffraction conditions, so that the defect density and size were comparable. The irradiation temperature was measured by a thermocouple attached to the heating cup of a double tilt heating sample holder and was kept within ±3 K during the irradiation. The TEM specimen was located in the heating cup with direct contact, thus the irradiation temperature represented the specimen temperature accurately. The ion flux was kept at 6.25 1011 ions/(cm2$s) for all irradiations. The doses were calculated using:
dpa ¼ 108 n
F N
;
(1)
where N is the number density, F is the fluence in ions/cm2, and n is the damage production in vacancies/(ion$Å) [24]. The damage production for 1 MeV Krypton in T91 is 1.41 vacancies/(ion$Å), obtained from SRIM-2008 Kinchin-Pease quick calculation using a displacement energy of 40 eV [25]. Therefore, the dose rate is 1.04 103 dpa/s for all irradiations and the highest fluence 4.2 1015 ions/cm2 corresponds to 7.0 dpa. 3. Results 3.1. As-received microstructure The as-received microstructure consists of martensite laths with carbide precipitates distributed along the lath boundaries and grain boundaries. Dislocation density varies significantly from grain to grain. Tangled dislocation areas with densities around 6.0e8.0 1013 m2 are commonly found, and some dislocationfree areas were also observed. 3.2. Dynamic microstructural evolution
2. Experimental The T91 steel was normalized at 1311 K for 1 h, air cooled, and then tempered at 1033 K for 1 h and air cooled. The nominal composition is listed in Table 1. The as-received T91 plate was cut
The real-time microstructural evolution of T91 irradiated up to 4.2 1015 ions/cm2 (~ 7 dpa) at three different temperatures (573 K, 673 K, and 773 K) is shown in Figs. 1e4. From these TEM micrographs, it is found that irradiation-induced dislocation loops began to show up at approximately 6.0 1014 ions/cm2 (~ 1 dpa) and loop
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
307
Fig. 1. TEM bright field images showing the dislocation-loop formation and development in T91 thin foil irradiated by 1 MeV Krypton ions at 573 K at different doses: (a) asreceived, (b) 6.0 1014 ions/cm2, (c) 1.8 1015 ions/cm2, (d) 3.0 1015 ions/cm2, and (e) 4.2 1015 ions/cm2. All the images were taken near the [001] zone.
Fig. 2. TEM bright field images showing the dislocation loop formation and development in a T91 thin foil irradiated by 1 MeV Krypton ions at 673 K at different doses: (a) asreceived, (b) 6.0 1014 ions/cm2, (c) 1.8 1015 ions/cm2, (d) 3.0 1015 ions/cm2, and (e) 4.2 1015 ions/cm2. The images were taken near the ½110 zone.
density increased by 1e2 orders of magnitude from 6.0 1014 ions/ cm2 to 1.8 1015 ions/cm2 (~ 3 dpa). After the density of dislocation loops reached about 2 1021 m3 to 4 1021 m3, the interactions of these loops with grown-in defects (line dislocations or grain boundaries) became apparent, as manifested by the changes in grown-in defects shown in Figs. 1e4. For the specimen irradiated at 573 K, the microstructure evolution is shown in Fig. 1. All the TEM micrographs in Fig. 1 were taken under nearly identical kinematical diffraction conditions, in which the sample was tilted near the [001] zone and only g110 was strongly excited. As can be seen, several black-dot dislocation loops formed at 6.0 1014 ions/cm2, examples are the ones marked by red arrows in Fig. 1 (b). From 6.0 1014 ions/cm2 to 1.8 1015 ions/ cm2, the loop density increased significantly from 6.3 1020 m3 to 3.8 1021 m3. Although the loop distribution was relatively uniform, i.e. no apparent dislocation decoration occurred, strong
interactions between grown-in line dislocations and irradiationinduced dislocation loops occurred. Some line dislocations were completely altered and they linked together with surrounding defects. An example is shown in the blue circle in Fig. 1, where gradual changes in the morphology of a dislocation can be readily seen. Observable changes also occurred in the lath boundary shown in Fig. 1. As the dose increased, increasing numbers of dislocation segments attached to the lath boundary. Fig. 2 shows the evolution of grown-in line dislocations and irradiation-induced dislocation loops in the specimen irradiated at 673 K. Similarly, only a few loops were observed at 6.0 1014 ions/ cm2, and the loop density is estimated to be around 4 1020 m3. Some of the loops were indicated by red arrows in Fig. 2 (b). Similar to the 573 K case, noticeable increase (by a factor of ~10) in the loop density was observed between 6.0 1014 ions/cm2 and 1.8 1015 ions/cm2. In Fig. 2 (c)-(e), both black dots and extended loops were
308
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
Fig. 3. Enlarged TEM bright field images showing the evolution of dislocations in T91 irradiated by 1 MeV Krypton ions at 673 K at different doses: (a) 6.0 1014 ions/cm2, (b) 1.8 1015 ions/cm2, (c) 3.0 1015 ions/cm2, and (d) 4.2 1015 ions/cm2.
Fig. 4. TEM bright field images showing the dislocation-loop formation and development in a T91 thin foil irradiated by 1 MeV Krypton ions at 773 K at different doses: (a) asreceived, (b) 6.0 1014 ions/cm2, (c) 1.8 1015 ions/cm2, (d) 3.0 1015 ions/cm2, and (e) 4.2 1015 ions/cm2. The images were taken near the [001] zone.
observed, and some of them were marked by red arrows. In Fig. 2 (d) and (e), dislocation segments (some were marked by red arrows) formed by coalescence of adjacent dislocation loops and/or line dislocations. The dislocation structure at 4.2 1015 ions/cm2 did not show any apparent alignment. After 1.8 1015 ions/cm2, strong interaction between preexisting line dislocations and dislocation loops occurred, as shown in the changes in the dislocations D1 and D2 in Fig. 2 (d) and (e). Several jogs/kinks (marked by blue triangles) in dislocation D1 were observed, and apparent changes in the curvature of dislocation D2 can be readily seen. These complex changes indicate dislocation climb by absorbing point defects and maybe further interactions with dislocation loops as well. Similar kinks and changes in curvature have also been reported in lower dose ion irradiated Fe-Cr model alloys [26].
Fig. 3 is a set of enlarged TEM images showing more clearly the localized interactions at 673 K. Some dislocation loops (e.g. the one marked in yellow circle) away from grown-in defects grew in size without interference, indicating the interactions were localized. Gradual changes in the lath boundary marked by red circle and the dislocation marked by blue circle were also observed. Compared to Fig. 1, the interaction shown in Figs. 2 and 3 is much more localized: the overall shape of the grown-in line dislocation was only slightly different, although some parts became curved after interacting with surrounding dislocation loops. For the 773 K case, Fig. 4 shows the microstructure evolution of a dislocation-rich region. As can be seen, little dislocation loops were observed at 6.0 1014 ions/cm2. At 1.8 1015 ions/cm2, most dislocation loops were black-dots. At 3.0 1015 ions/cm2, some
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
black-dots grew into much larger dislocation loops. The loops in the lower part of Fig. 4 (d) had the same Burgers vector and were wellaligned. Localized interactions between irradiation-induced dislocation loops and grown-in dislocations were again observed in Fig. 4 (c) to (e). After irradiation, both isolated loops and dislocation segments were found and no obvious dislocation network was observed. It should be noted that the physical processes of in situ irradiation experiments could differ from bulk ex situ irradiation experiments, largely due to the two nearby surfaces of TEM thin foils. The free surfaces can act as sinks and remove a portion of the mobile defects such as a0 =2〈111〉 loops. In the in situ irradiation study of Fe-Cr-Ni austenitic steels at 450 C [27], 20 nm was subtracted considering the depletion of defects near surfaces. F/M systems have lower defect mobility and high density internal dislocation sinks, this effect is expected to be much less. The surface effects were not studied in the present work. 3.3. 〈111〉 〈100〉 loop transition Previous studies reported that the dislocation loops induced by both ion and neutron irradiations in a-iron, Fe-Cr model alloys, and F/M steels are a0 〈100〉 or a0 =2〈111〉 type [8e18]. In this paper, it is also assumed that all the loops are a0 〈100〉 or a0 =2〈111〉 type. The Burgers vectors of dislocations loops were be determined using bright-field TEM images taken under selected diffraction conditions. a0 〈100〉 and a0 =2〈111〉 loops can be distinguished by (1) standard g,b analysis or (2) comparing the orientation of the loops with respect to the diffraction vectors. Fig. 5 shows the dislocation loops in a T91 specimen that was irradiated to 4.2 1015 ions/cm2 at 773 K. Fig. 5 (a) to (c) were obtained by tilting the specimen to different diffraction conditions near the [001] zone. In Fig. 5 (a), only g020 was strongly excited, therefore only a0[001] loops are invisible (g,b ¼ 0) and all other types of loops (a0[010] and all a0 =2〈111〉 loops) should be visible. However, only a0[010] loops marked as “a” showed up in Fig. 5 (a), indicating that there was no a0 =2〈111〉 loops in this area. Both a0[100] and a0[010] loops are visible in Fig. 5 (b) and (c) since g,bs0. The loop marked as “b” showed up in Fig. 5 (b) and (c) but disappeared in Fig. 5 (a), indicating that it is a a0[100] loop. Comparing Fig. 5 (a) to (c), it is found that after irradiation to 4.2 1015 ions/cm2 at 773 K, the dislocation loops are dominantly a0[010] type in this area. Since these loops have the same Burgers vector, they were well-aligned as shown in Fig. 5. Similar TEM
309
examinations were carried out in other grains, and the loops were also found to be dominantly a0 〈100〉 type. Similar grain-level localized loop alignment was commonly observed and should be characteristic for the irradiation at 773 K. For the 673 K case, however, distinct dislocation microstructure was observed. Fig. 6 shows the typical microstructure found in the specimen irradiated to 4.2 1015 at 673 K. It was taken near the [001] zone and only g110 was strongly excited, as shown in the insertion in Fig. 6 (a). Under this diffraction condition, visible loops are a0[010], a0[100], a0/2[111], a0/2½111, a0/2½111, and a0/2½111. The a0[010] and a0[100] loops can be easily distinguished from a0 =2〈111〉 loops by their orientations with respect to the diffraction vectors, i.e. a0[010] loops are perpendicular to g020 and 45 to g110 , a0[100] loops are parallel to g020 and 45 to g110 . Moreover, a0 〈100〉 loops are close to edge-on since the images were taken near the [001] zone axis. The loops in Fig. 6 (b)-(d) were distinguished by this method. The ones in red circles are a0 〈100〉 loops, and the ones in blue circles are a0 =2〈111〉 loops. The overall dislocation structure is shown in Fig. 6 (a). Fig. 6 (b)-(d) are enlarged TEM images showing the a0 =2〈111〉 and a0 〈100〉 loops. Counting only the wellresolved loops in the three areas near Fig. 6 (b), (c), and (d), statistical data of the Burgers vector are shown in Table 2. It is estimated that the 〈111〉=〈110〉 ratio is around 1.02 ± 0.38 for the 673 K case. 3.4. Loop size and density In order to quantitatively characterize the irradiation-induced dislocation loops, the average loop sizes and densities were obtained. For the loop size measurement, TEM kinematic bright field images were used. Therefore, the results might be not accurate for loops less than 10 nm, especially for the small black-dot loops around 2 nm. On average, 40e60 loops were measured for each dose and temperature condition to reduce the statistical errors. Figs. 7e9 show the histograms of the loop size distribution for the irradiation at 573 K, 673 K, and 773 K, respectively. At 573 K, the loops did not show significant increase in size, as shown in Fig. 7. Most loops are small black-dots and only a few extended loops around 6e10 nm were found at 3.0 1015 (~5 dpa) and 4.2 1015 ions/cm2 (~7 dpa). At 673 K, the loops also started as black-dots around 2 nm at 6.0 1014 ions/cm2 (~1 dpa), but the loop size increased significantly as the dose increased, and loops larger than 10 nm were commonly found at 4.2 1015 ions/cm2. At 773 K, few
Fig. 5. Different diffraction vectors: (a) g020 , (b) g110 , and (c) g110 were used to image the dislocation loops in a T91 thin foil that has been irradiated by 1 MeV Krypton ions to 4.2 1015 ions/cm2 at 773 K. The zone axis is [001].
310
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
Fig. 6. Dislocation loops in a T91 thin foil that has been irradiated by 1 MeV Krypton ions to 4.2 1015 ions/cm2 at 673 K: (a) low-mag image showing the overall microstructure, (b) an area with both loops, (c) an area with dominantly a0 =2〈111〉 loops, (d) an area with more a0 〈100〉 loops. All the images were taken with only g110 strongly excited near the [001] zone.
Table 2 Statistical analysis of a0 =2〈111〉 and a0 〈100〉 loops in T91 irradiated to 4.2 1015 ions/cm2 at 673 K. No. of a0 =2〈111〉 loops
No. of a0 〈100〉 loops
〈111〉=〈110〉 ratio
Area 1 Area 2 Area 3
20 14 13
17 10 19
1.18 1.40 0.68
Total
47
46
1.02 ± 0.38
black-dots were observed at 6.0 1014 ions/cm2 and therefore no size histogram was obtained. Large isolated loops around 20 nm were commonly found at 4.2 1015 ions/cm2. Based on these histograms, the average loop sizes were plotted in Fig. 10 (a), in which the error bars were calculated using the student t-distribution and represent the 95% confidence interval. In addition to the loop size, the loop density as well as line dislocation density were also obtained, as shown in Fig. 10 (b). It should be noted that during the in situ irradiations, the thinner areas near the hole created by jet polishing were bended significantly and therefore the thickness fringes method could not be applied for sample thickness measurement. All the TEM examinations were focused on thick enough regions (not too thin to avoid bending, and not too thick to get good contrast). For the loop density calculation, all the sample thickness was assumed to be 100 nm, which is apparently a rough estimate. This uncertainty only affects the loop density calculation, since the line dislocation density is only related to the area of the chosen regions. In the line dislocation density calculation, both grown-in dislocations and dislocation segments formed from irradiation were taken into account. As shown in Fig. 10 (b), for the 573 K irradiation, the loop density began to drop between 1.8 1015 ions/cm2 (~3 dpa) and 3.0 1015 ions/cm2 (~5 dpa), accompanied by a steady increase in the line dislocation density. At 673 K, the drop in loop density occurred between 3.0 1015 ions/cms2 and 4.2 1015 ions/cm2 (~7 dpa), whereas the line dislocation density showed appreciable increase between 1.8 1015 ions/cm2 and 3.0 1015 ions/cm2, due to the formation of dislocation segments. For the irradiation at 773 K, the loop density increased significantly between 6.0 1014 ions/cm2
(~1 dpa) and 1.8 1015 ions/cm2 and then dropped, and the line dislocation density slightly decreased between 1.8 1015 ions/cm2 and 4.2 1015 ions/cm2. Another finding is that at 6.0 1014 ions/ cm2, the loop density at 773 K is significantly lower (by as much as an order of magnitude) than that at 573 K and 673 K. 4. Discussions 4.1. Relationship between loop transition and the dislocation structure In situ TEM irradiation allows the dynamic observation of the microstructural evolution of the same area at different dose levels. For complex systems like F/M steels, differences in the initial microstructure could lead to variations in the irradiated microstructure. In order to compare the representative microstructure of T91 irradiated at different temperatures, post-irradiation examination (PIE) of multiple areas of the irradiated T91 TEM specimens was carried out. Fig. 11 shows the representative dislocation structure of T91 steel irradiated to 4.2 1015 ions/cm2at 573 K, 673 K, and 773 K, respectively. Dislocation networks were commonly found in the sample irradiated at 573 K, as shown in Fig. 11 (a). Some black-dot dislocation loops are also observed in addition to the network dislocations. In contrast to the dislocation network in Fig. 11 (a), isolated dislocation loops were commonly found in the samples irradiated at 673 K and 773 K, as shown in Fig. 11 (b) and (c). However, the difference is that the dislocation loops in Fig. 11 (b) are randomly oriented, whereas the dislocation loops in Fig. 11 (c) are wellaligned.
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
311
Fig. 7. Loop size distribution histograms of T91 irradiated at 573 K at different doses: (a) 6.0 1014, (b) 1.8 1015, (c) 3.0 1015, and (d) 4.2 1015 ions/cm2.
The temperature dependence of the observed dislocation structures should be dependent on the a0 =2〈111〉 a0 〈100〉 dislocation loop transition that happens at elevated temperatures in a-iron, Fe-Cr ferritic model alloys, and F/M steels. The distinction is important since, at elevated temperatures, the a0 =2〈111〉 loops are highly mobile and a0 〈100〉 loops are sessile. Therefore, a0 =2〈111〉 loops can migrate over long distances and interact with distant grown-in defects, leading to formation of dislocation segments and network dislocations at lower dose, whereas a0 〈100〉 can only act as stationary defect sinks once formed. If the a0 〈100〉 loops happen to form near grown-in defects, localized interaction between loops and nearby grown-in defects is possible. As an example, Figs. 2 and 3 shows clearly the process of localized interactions. The stability of a0 〈100〉 loops with respect to a0 =2〈111〉 loops is highly temperature dependent. Yao et al. performed Feþ self-ion irradiation on pure Fe and reported that about 80% of the loops were of the a0 〈100〉 type at 673 K [18]. In the present work, both a0 〈100〉 and a0 =2〈111〉 loops were found at 573 K, the ratio was not determined. However, based on the fact that only at 573 K, loops were observed to escape to the surface and disappeared, mobile a0 =2〈111〉 should be the dominant type. At 673 K, the ratio of a0 =2〈111〉 to a0 〈100〉 loops is around 1.02 (from Table 2). At 773 K,
the loops were found to be predominantly a0 〈100〉 type. Based on our results at three different temperatures, the a0 =2〈111〉 a0 〈100〉 loop transition is confirmed. Although at 673 K, only ~49% of the loops are a0 〈100〉 type, different from the 80% in pure iron reported by Yao et al., the general trend is consistent [18]. The difference could be due to the effect of alloying elements such as chromium, as previous studies suggest that the addition of chromium tends to equalize the proportion of a0 =2〈111〉 and a0 〈100〉 loops [28], or to increase the fraction of a0 =2〈111〉 loops [29]. 4.2. Modeling of irradiation hardening In this study, heavy ion irradiation was employed to emulate neutron irradiation due to the degree of similarity in their weighted average recoil spectra. The disadvantage of heavy ion irradiation is that it produces a non-uniform dose profile over a shallow region (less than 600 nm for 1 Me Kr ions) beneath the sample surface. Our in situ TEM observation avoided this problem since the thickness of thin area is about 100 nm, where the dose profile can be approximated to be constant. The difficulty lies in that it is impossible to directly measure the mechanical properties by small scale tests such as nanoindentation, due to various factors contributing to the
312
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
Fig. 8. Loop size distribution histograms of T91 irradiated at 673 K at different doses: (a) 6.0 1014, (b) 1.8 1015, (c) 3.0 1015, and (d) 4.2 1015 ions/cm2.
Fig. 9. Loop size distribution histograms of T91 irradiated at 773 K at different doses: (a) 1.8 1015, (b) 3.0 1015, and (c) 4.2 1015 ions/cm2.
measurements. The results of shallow (less than 100 nm) indents will be largely influenced by the size effect and deeper indents will inevitably sample the whole non-uniform dose profile [30]. Quantitative comparison of the results of small scale tests on heavy ion irradiated samples with the data from neutron irradiated samples seems to be almost impossible, at least very challenging. Alternatively, based on the observed defect microstructure it is
possible to calculate the mechanical properties such as change in yield strength DsY using well-accepted models. The dispersed barrier hardening model has been used in previous studies [12,15,21,28,31e34] to establish the microstructure-property relationship and to evaluate the hardening induced by neutron and ion irradiations. In this model, the change in yield strength DsY is related to the defect microstructure by:
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
313
Fig. 10. Statistical information of defects in T91 steel irradiated at different temperatures and doses (a) Average loop sizes; (b) average line dislocation densities and loop densities.
Fig. 11. TEM bright field images showing different dislocation structures in T91 steel irradiated by Krypton ions to 4.2 1015 ions/cm2 at different temperatures: (a) Dislocation network decorated by black-dot dislocation loops at 573 K; (b) random distributed a0 =2〈111〉 and a0 〈100〉 dislocation loops at 673 K; and (c) well-aligned a0 〈100〉 dislocation loops at 773 K.
pffiffiffiffiffiffi
DsY ¼ Mamb rd;
(2)
where M is the Taylor factor, a is the defect barrier strength, m is the shear modulus, b represents the magnitude of the Burgers vector of dislocations, r is the defect density and d is the defect diameter. In our calculation M ¼ 3:06, m ¼ 82 GPa, and b ¼ 2:49 nm, which are consistent with the values used in Fe-Cr alloys [12,15,28,32,34]. In general, the irradiation hardening results from multiple microstructure modifications under energetic ion/neutron bombardment, including voids, precipitates, dislocation loops, line dislocations, etc. The overall effect on the yield strength can be evaluated by the linear superposition of the contributions from various defects: DsY ¼ DsYV þ DsYP þ DsYL þ DsYD þ DsYR , where DsYV , DsYP , DsYL , DsYD and DsYR represent the contribution from voids, precipitates, dislocation loops, line dislocations, and others (such as dislocation decoration, segregation of solute atoms), respectively. In the case of T91, no obvious voids or irradiation-induced a0 precipitates were observed during the in situ irradiation. This is also consistent with the observations in previous studies [21,23,35]. The favorable explanation is that 9e12%Cr F/M steels are well known for their excellent swelling resistance [36] and the final dose 4.2 1015 ions/cm2 (~7 dpa) is still in the transient regime [37]. Thus, only negligible amount of voids is expected at the final dose. Also, the Cr content is 9.24 wt%, which is close to the threshold value for a a0 phase separation in Fe-Cr alloys [38], indicating very few, if any, a0 precipitates are expected. Indeed, no a0 precipitates were reported in F/M steel T91 after proton irradiation to 10 dpa at 400e500 C or neutron irradiation up to 184 dpa at 413 C [35]. As a first order approximation, the contribution from voids and a0 precipitates is neglected due to their marginal amount and
small barrier strength [33,34]. DsYR is also ignored for simplicity. Then we have:
qffiffiffiffiffiffiffiffiffi
DsY ¼ DsYL þ DsYD ¼ MaL mb rL dL þ MaD mb
qffiffiffiffiffiffiffiffi
rirr: D
qffiffiffiffiffiffiffiffiffiffiffiffi
; runirr: D (3)
where aL , rL , and dL are the barrier strength, average density, and average size of dislocation loops, respectively; aD is the barrier unirr: is the line dislocation strength of line dislocations, rirr: D and rD density after and before irradiation, respectively. Here aD ¼ 0:64 and aL is temporarily taken to be 0.40, which are consistent with the values used in literature [15,28,31,32,34]. Using the loop sizes and loop/line dislocation densities from Section 3.4, the increase in yield strength at different dose and irradiation temperature conditions was obtained and is shown in Fig. 12. As can be seen, DsY increased the fastest with increasing dose at 573 K, compared to the 673 K and 773 K irradiations. At 773 K, DsY reached its maximum around 3e5 dpa and then slightly decreased at 7 dpa. The behavior of DsY falls in between at 673 K. For both 573 K and 673 K irradiations no saturation in DsY was observed. The general trend is consistent with the conventional irradiation hardening characteristics of F/M steels, i.e. more pronounced hardening at lower irradiation temperatures and earlier saturation at higher irradiation temperatures [39]. Apparently, the calculated DsY strongly depends on the choice of the barrier strength a. Table 3 lists the calculated DsY at different temperature and dose conditions, for four different aL values. It is easy to see that the 573 K data are relatively insensitive to aL , since the major contribution to the hardening comes from the dislocation segments and network dislocations. At 7 dpa, the 573 K data only
314
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
Fig. 12. Calculated change in yield strength DsY at different irradiation temperatures and doses, estimated using the dispersed barrier hardening model.
changed by 76 MPa when aL changed from 0.1 to 0.4. The 773 K data are most sensitive to aL , because the dominant contribution comes from individual loops. At 7 dpa, the 773 K data changed by 218 MPa when aL changed from 0.1 to 0.4. For simplicity, to the first order approximation, the calculated values in all the figures are based on aL ¼ 0:4, which was commonly used in Fe-Cr model alloys [15,28,31,32,34]. However, we should keep in mind that aL should be lower for lower temperatures, as the fraction of a0 =2〈111〉 loops is higher at lower temperatures. The barrier strength of mobile a0 =2〈111〉 loops should be smaller, compared to that of sessile a0 〈100〉 loops. Efforts were also made to compare the calculated change in yield strength DsY with existing experimental data [21,39e41]. The dose dependence and temperature dependence of irradiation hardening in T91 are shown in Figs. 13 and 14, respectively. For simplicity, only data at similar temperatures (~573 K, 673 K, and 773 K) were shown in Fig. 13 and only data in the similar dose range (7e12 dpa) were shown in Fig. 13. As can be seen in Fig. 13, there is noticeable difference in the data at different irradiation temperatures. Three dotted lines were plotted to show the trends. It is easy to find that the calculated results are reasonably consistent with the existing data. At 573 K, DsY increases with increasing dpa and saturates at around 20 dpa. At 673 K, DsY increases with increasing dpa at a lower rate and saturates earlier (~10 dpa). At 773 K, DsY shows slightly increase and saturates around 3 dpa. In Fig. 14, the dose of all the proton data is the same (7 dpa) as that of the present work. For the neutron data, only these in the range of 9e12 dpa were plotted in order to make the data more comparable. For the neutron data below 723 K, DsY decreases with
Fig. 13. Change in yield strength DsY at different doses. In order to compare the results with existing data at similar temperatures, two sets of proton data [21] and three sets of neutron data [39e41] are shown. The dotted lines are plotted to show the trends.
increasing temperature, and above 773 K DsY 0. The trend is shown in the dotted royal blue line. The ion irradiation data (both proton and krypton) are connected by another dotted violet line, which turns out to be of the similar shape. This manifests that the ion data has very similar temperature dependence as the neutron data. It is also noticed that the ion data was shifted upwards from the neutron data by approximately 100e150 MPa, indicating that more pronounced hardening is expected for ion irradiations. Considering that the dose for the neutron data (9e12 dpa) is higher than that of the ion data (7 dpa), the upward shift in DsY of the ion data becomes even more compelling. One possible reason for the upward shift is related to the thermal annealing effect accompanying low dose rate neutron irradiations. Since the dose rate of ion irradiation is 103 to 104 higher than that of neutron irradiation, the material under ion irradiation experiences much less thermal annealing and therefore the defects (both grow-in dislocations and irradiation-induced defects) have much smaller chance to be annealed and the surviving defects are responsible for more noticeable hardening. On the other hand, the offset also indicates that the microstructure is generally similar, except for differences in the defect density and size. For cases in which ion irradiations cannot produce the comparable microstructure as that of neutron irradiations, the ion-to-neutron is much more challenging. An example is that for Fe-Cr alloys above the a a0 phase separation threshold value, a0 precipitates are only frequently observed in neutronirradiated, rather than ion-irradiated specimens.
Table 3 Calcuated yield strength changes at different irradiation conditions for aL ¼ 0.1, 0.2, 0.3, and 0.4.
DsY
aL
0.1
0.2
0.3
0.4
573 K
1 3 5 7
dpa dpa dpa dpa
14.1 118.3 260.1 521.9
28.2 155.2 294.7 547.2
42.3 192.2 329.2 572.5
56.3 229.2 363.7 597.7
673 K
1 3 5 7
dpa dpa dpa dpa
10.6 41.1 116.6 190.9
21.2 82.1 162.5 243.6
31.9 123.2 208.5 296.3
42.5 164.2 254.5 349.0
773 K
1 3 5 7
dpa dpa dpa dpa
2.1 32.1 73.8 6.7
4.2 64.3 128.1 65.8
6.2 96.4 182.3 138.3
8.3 128.6 236.6 210.9
(MPa)
Fig. 14. Change in yield strength DsY at different irradiation temperatures. The dose for all the proton data [21] and present work is 7 dpa, and all the neutron data [39e41] are in the range of 9e12 dpa. The data are connected by dotted lines to show the trends (the mean value of the proton data and present work at 673 K was used to draw the dotted trend line).
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
Another important conclusion from Fig. 14 is related to the temperature shift. It is commonly believed that in order to correlate ion irradiations with neutron irradiations, the irradiation temperature of ion irradiations should be higher than neutron irradiations. The temperature difference is known as the temperature shift. In contrast to the temperature shift predicted by Mansur [42] and observed in the swelling data of ion irradiations [43], our preliminary results indicate that no apparent temperature shift was observed in the irradiation hardening of F/M steel T91. This is consistent with the swelling data of neutron irradiation in copper [44,45], where no temperature shift was observed. The temperature shift of ion irradiation is still a controversial topic, and it is recently found that the temperature shift could be due to the effect of injected interstitials alone [46]. The reason that no temperature shift is found in our ion irradiations is that all the Krypton ion irradiations were carried out in situ on TEM thin foil specimens that were around 100 nm thick and therefore the problems related to conventional ion irradiations such as injected interstitials and compositional modifications were avoided from the very beginning. 5. Summary In summary, strong dislocation loop-preexisting defects interaction was observed in F/M steel T91 at 573 K, 673 K, and 773 K. The interaction took different forms at different temperatures. At 573 K, both line dislocations and grain boundaries were modified by dislocation loops. The shape of some line dislocations was entirely changed, with more loops attached to the grain boundaries as the dose increased. Dense dislocation network was found after irradiation. For the irradiation at 673 K, grown-in dislocations were locally modified. Jogs/kinks and changes in curvature in grown-in dislocations were observed. At 3.0 1015 ions/cm2, adjacent dislocation loops began to coalesce and dislocation segments started to form. At 4.2 1015 ions/cm2, dislocation segments were commonly observed and some were linked together with preexisting line dislocations. At 773 K, isolated dislocation loops were again observed but in well-aligned fashion rather than random-oriented as the ones found at 673 K. This indicates that these loops have exactly the same Burgers vector in one region. This grain-level localized alignment should be a special feature of a0 〈100〉 loops. These phenomena can be well explained by the 〈111〉 〈110〉 loop transition, which was supported by the post-irradiation examination. Based on the loop size and defect densities, a dispersed barrier hardening model was used to establish the microstructureproperty relationship. The calculated changes in yield strength at different irradiation temperatures and doses were compared with existing data. It was found that the ion data follow the similar trend of the neutron data, expect for an offset of 100e150 MPa. No apparent temperature shift was observed. Acknowledgements The authors would like to thank Pete Baldo and Edward Ryan at Argonne National Laboratory for their assistance with the irradiation experiments. This work was supported by the funding from the DOE Office of Nuclear Energy's Nuclear Energy University Programs (NEUP) under Contract No. DE-NE0008291. The TEM experiments were carried out in part in the Frederick Seitz Materials Research Laboratory Central Facilities, University of Illinois. The electron microscopy with in situ ion irradiation was accomplished at Argonne National Laboratory at the IVEM-Tandem Facility, a U.S.
315
Department of Energy Facility funded by the DOE Office of Nuclear Energy, operated under Contract No. DE-AC02-06CH11357 by UChicago Argonne, LLC and the DOE Office of Nuclear Energy under DOE Idaho Operations Office Contract DE-AC07-051D14517 as part of a Nuclear Science User Facilities experiment. References [1] G.S. Was, Fundamentals of Radiation Damage Materials Science: Metals and Alloys, Springer Science & Business Media, 2007. [2] S.L. Dudarev, R. Bullough, P.M. Derlet, Effect of the ag phase transition on the stability of dislocation loops in bcc iron, Phys. Rev. Lett. 100 (2008) 135503, http://dx.doi.org/10.1103/PhysRevLett.100.135503. [3] R.C. Rau, F.S. D’aragona, R.L. Ladd, Neutron damage in molybdenum irradiated at high temperatures, Philos. Mag. 21 (1970) 441e452, http://dx.doi.org/ 10.1080/14786437008238430. [4] T. Matsui, S. Muto, T. Tanabe, TEM study on deuterium-irradiation-induced defects in tungsten and molybdenum, J. Nucl. Mater 283e287 (2000) 1139e1143, http://dx.doi.org/10.1016/S0022-3115(00)00169-0. [5] H. Xu, R.E. Stoller, Y.N. Osetsky, D. Terentyev, Solving the Puzzle of <100> interstitial loop formation in bcc iron, Phys. Rev. Lett. 110 (2013) 265503, http://dx.doi.org/10.1103/PhysRevLett.110.265503. [6] B.L. Eyre, R. Bullough, On the formation of interstitial loops in b.c.c. metals, Philos. Mag. 12 (1965) 31e39, http://dx.doi.org/10.1080/ 14786436508224943. [7] J. Marian, B.D. Wirth, J.M. Perlado, Mechanism of formation and growth of <100> interstitial loops in ferritic materials, Phys. Rev. Lett. 88 (2002) 255507, http://dx.doi.org/10.1103/PhysRevLett.88.255507. [8] B.C. Masters, Dislocation loops in irradiated iron, Philos. Mag. 11 (1965) 881e893, http://dx.doi.org/10.1080/14786436508223952. [9] D.S. Gelles, Microstructural examination of neutron-irradiated simple ferritic alloys, J. Nucl. Mater 108e109 (1982) 515e526, http://dx.doi.org/10.1016/ 0022-3115(82)90523-2. [10] A. Okada, T. Yasujima, T. Yoshiie, I. Ishida, M. Kiritani, Mechanical property changes and microstructures of iron irradiated with fission and fusion neutrons, J. Nucl. Mater 179e181 (1991) 1083e1087, http://dx.doi.org/10.1016/ 0022-3115(91)90280-K. [11] Y. Katoh, A. Kohyama, D.S. Gelles, Swelling and dislocation evolution in simple ferritic alloys irradiated to high fluence in FFTF/MOTA, J. Nucl. Mater 225 (1995) 154e162, http://dx.doi.org/10.1016/0022-3115(94)00669-5. [12] S. Porollo, A. Dvoriashin, A. Vorobyev, Y. Konobeev, The microstructure and tensile properties of FeeCr alloys after neutron irradiation at 400 C to 5.5e7.1 dpa, J. Nucl. Mater 256 (1998) 247e253, http://dx.doi.org/10.1016/S00223115(98)00043-9. [13] Y.V. Konobeev, A.M. Dvoriashin, S.I. Porollo, F.A. Garner, Swelling and microstructure of pure Fe and FeeCr alloys after neutron irradiation to ~26dpa at 400 C, J. Nucl. Mater 355 (2006) 124e130, http://dx.doi.org/10.1016/ j.jnucmat.2006.04.011. [14] M. Matijasevic, E. Lucon, A. Almazouzi, Behavior of ferritic/martensitic steels after n-irradiation at 200 and 300 C, J. Nucl. Mater 377 (2008) 101e108, http://dx.doi.org/10.1016/j.jnucmat.2008.02.063. [15] M. Matijasevic, A. Almazouzi, Effect of Cr on the mechanical properties and microstructure of FeeCr model alloys after n-irradiation, J. Nucl. Mater 377 (2008) 147e154, http://dx.doi.org/10.1016/j.jnucmat.2008.02.061. [16] M. Matijasevic, W. Van Renterghem, A. Almazouzi, Characterization of irradiated single crystals of Fe and Fee15Cr, Acta Mater 57 (2009) 1577e1585, http://dx.doi.org/10.1016/j.actamat.2008.11.042. ndez-Mayoral, D. Go mez-Bricen ~ o, Transmission electron microscopy [17] M. Herna study on neutron irradiated pure iron and RPV model alloys, J. Nucl. Mater 399 (2010) 146e153, http://dx.doi.org/10.1016/j.jnucmat.2009.11.013. [18] Z. Yao, M.L. Jenkins, M. Hern andez-Mayoral, M.a. Kirk, The temperature dependence of heavy-ion damage in iron: a microstructural transition at elevated temperatures, Philos. Mag. 90 (2010) 4623e4634, http://dx.doi.org/ 10.1080/14786430903430981. ndez-Mayoral, Z. Yao, M.L. Jenkins, M.A. Kirk, Heavy-ion irradiations [19] M. Herna of Fe and FeeCr model alloys Part 2: damage evolution in thin-foils at higher doses, Philos. Mag. 88 (2008) 2881e2897, http://dx.doi.org/10.1080/ 14786430802380477. [20] M. Grosse, Y. Dai, S. Van Petegem, Irradiation-induced structural changes in martensitic steel T91, J. Nucl. Mater 356 (2006) 112e117, http://dx.doi.org/ 10.1016/j.jnucmat.2006.05.026. [21] G. Gupta, Z. Jiao, A.N. Ham, J.T. Busby, G.S. Was, Microstructural evolution of proton irradiated T91, J. Nucl. Mater 351 (2006) 162e173, http://dx.doi.org/ 10.1016/j.jnucmat.2006.02.028. [22] X. Jia, Y. Dai, Microstructure in martensitic steels T91 and F82H after irradiation in SINQ Target-3, J. Nucl. Mater 318 (2003) 207e214, http://dx.doi.org/ 10.1016/S0022-3115(03)00101-6. [23] Z. Jiao, N. Ham, G.S. Was, Microstructure of helium-implanted and protonirradiated T91 ferritic/martensitic steel, J. Nucl. Mater 367e370 (2007) 440e445, http://dx.doi.org/10.1016/j.jnucmat.2007.03.118. [24] M. Li, M.a. Kirk, P.M. Baldo, D. Xu, B.D. Wirth, Study of defect evolution by TEM with in situ ion irradiation and coordinated modeling, Philos. Mag. 92 (2012)
316
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
2048e2078, http://dx.doi.org/10.1080/14786435.2012.662601. [25] R.E. Stoller, M.B. Toloczko, G.S. Was, A.G. Certain, S. Dwaraknath, F.A. Garner, On the use of SRIM for computing radiation damage exposure, Nucl. Instrum. Methods Phys. Res. Sect. B Beam Interact. Mater. Atoms 310 (2013) 75e80, http://dx.doi.org/10.1016/j.nimb.2013.05.008. [26] D. Kaoumi, J. Adamson, M. Kirk, Microstructure evolution of two model ferritic/martensitic steels under in situ ion irradiation at low doses (0e2dpa), J. Nucl. Mater 445 (2014) 12e19, http://dx.doi.org/10.1016/ j.jnucmat.2013.10.047. [27] S. Jublot-Leclerc, X. Li, L. Legras, M.-L. Lescoat, F. Fortuna, A. Gentils, Microstructure of Au-ion irradiated 316L and FeNiCr austenitic stainless steels, J. Nucl. Mater 480 (2016) 436e446, http://dx.doi.org/10.1016/ j.jnucmat.2016.08.006. [28] W.-Y. Chen, Irradiation Damage in Neutron-irradiated Fe-cr Model Alloys, University of Illinois at Urbana-Champaign, 2014. ndez-Mayoral, C. Heintze, E. On ~ orbe, Transmission electron micro[29] M. Herna scopy investigation of the microstructure of FeeCr alloys induced by neutron and ion irradiation at 300 C, J. Nucl. Mater 474 (2016) 88e98, http:// dx.doi.org/10.1016/j.jnucmat.2016.03.002. [30] P. Hosemann, D. Kiener, Y. Wang, S.A. Maloy, Issues to consider using nano indentation on shallow ion beam irradiated materials, J. Nucl. Mater 425 (2012) 136e139, http://dx.doi.org/10.1016/j.jnucmat.2011.11.070. , R. Lebensohn, S. Maloy, Modeling and simulation of irradi[31] C. Deo, C. Tome ation hardening in structural ferritic steels for advanced nuclear reactors, J. Nucl. Mater 377 (2008) 136e140, http://dx.doi.org/10.1016/ j.jnucmat.2008.02.064. [32] W.-Y. Chen, Y. Miao, J. Gan, M.A. Okuniewski, S.A. Maloy, J.F. Stubbins, Neutron irradiation effects in Fe and Fe-Cr at 300 C, Acta Mater 111 (2016) 407e416, http://dx.doi.org/10.1016/j.actamat.2016.03.060. [33] K.G. Field, X. Hu, K.C. Littrell, Y. Yamamoto, L.L. Snead, Radiation tolerance of neutron-irradiated model FeeCreAl alloys, J. Nucl. Mater 465 (2015) 746e755, http://dx.doi.org/10.1016/j.jnucmat.2015.06.023. [34] F. Bergner, C. Pareige, M. Hern andez-Mayoral, L. Malerba, C. Heintze, Application of a three-feature dispersed-barrier hardening model to neutronirradiated FeeCr model alloys, J. Nucl. Mater 448 (2014) 96e102, http:// dx.doi.org/10.1016/j.jnucmat.2014.01.024. [35] J. Van den Bosch, O. Anderoglu, R. Dickerson, M. Hartl, P. Dickerson,
[36]
[37]
[38]
[39]
[40]
[41]
[42] [43]
[44] [45]
[46]
J.A. Aguiar, P. Hosemann, M.B. Toloczko, S.A. Maloy, SANS and TEM of ferriticemartensitic steel T91 irradiated in FFTF up to 184dpa at 413 C, J. Nucl. Mater 440 (2013) 91e97, http://dx.doi.org/10.1016/j.jnucmat.2013.04.025. G.R. Odette, M.J. Alinger, B.D. Wirth, Recent developments in irradiationresistant steels, Annu. Rev. Mater. Res. 38 (2008) 471e503, http:// dx.doi.org/10.1146/annurev.matsci.38.060407.130315. F. Garner, M. Toloczko, B. Sencer, Comparison of swelling and irradiation creep behavior of fcc-austenitic and bcc-ferritic/martensitic alloys at high neutron exposure, J. Nucl. Mater 276 (2000) 123e142, http://dx.doi.org/ 10.1016/S0022-3115(99)00225-1. G. Bonny, D. Terentyev, L. Malerba, On the aea0 miscibility gap of FeeCr alloys, Scr. Mater 59 (2008) 1193e1196, http://dx.doi.org/10.1016/ j.scriptamat.2008.08.008. O. Anderoglu, T.S. Byun, M. Toloczko, S.A. Maloy, Mechanical performance of ferritic martensitic steels for high dose applications in advanced nuclear reactors, Metall. Mater. Trans. A 44 (2013) 70e83, http://dx.doi.org/10.1007/ s11661-012-1565-y. R.L. Klueh, J.M. Vitek, Elevated-temperature tensile properties of irradiated 9 Cr-1 MoVNb steel, J. Nucl. Mater 132 (1985) 27e31, http://dx.doi.org/10.1016/ 0022-3115(85)90389-7. A. Alamo, J.L. Bertin, V.K. Shamardin, P. Wident, Mechanical properties of 9Cr martensitic steels and ODS-FeCr alloys after neutron irradiation at 325 C up to 42dpa, J. Nucl. Mater 367e370 (2007) 54e59, http://dx.doi.org/10.1016/ j.jnucmat.2007.03.166. L.K. Mansur, Correlation of neutron and heavy-ion damage, J. Nucl. Mater 78 (1978) 156e160, http://dx.doi.org/10.1016/0022-3115(78)90514-7. N.H. Packan, K. Farrell, J.O. Stiegler, Correlation of neutron and heavy-ion damage, J. Nucl. Mater 78 (1978) 143e155, http://dx.doi.org/10.1016/00223115(78)90513-5. M. Labbe, J.P. Poirier, Aux neutrons rapides, J. Nucl. Mater 46 (1973) 86e98. M. Labbe, J.P. Poirier, Letters to the editors - lettres aux redacteurs effet du flux instantane sur le gonflement du cuivre pur irradie aux neutrons rapides, J. Nucl. Mater 49 (1973) 232e234. F.A. Garner, L. Shao, M.B. Toloczko, S.A. Maloy, V.N. Voyevodin, Use of self-ion bombardment to study void swelling in advanced radiation- resistant alloys, in: 17th Int. Conf. Environ. Degrad. Mater. Nucl. Power Syst., Ontario, 2015.
Contents lists available at ScienceDirect
Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat
Ion-irradiation-induced microstructural modifications in ferritic/ martensitic steel T91 Xiang Liu a, *, Yinbin Miao a, b, Meimei Li b, Marquis A. Kirk b, Stuart A. Maloy c, James F. Stubbins a, d a
Department of Nuclear, Plasma, and Radiological Engineering, University of Illinois at Urbana-Champaign, Urbana, IL 61801, USA Nuclear Engineering Division, Argonne National Laboratory, Argonne, IL 60439, USA Materials Science and Technology Division, Los Alamos National Laboratory, Los Alamos, NM 87545, USA d International Institute for Carbon Neutral Energy Research (WPI-I2CNER), Kyushu University, 744 Motooka, Nishi-ku, Fukuoka 819-0395, Japan b c
g r a p h i c a l a b s t r a c t
a r t i c l e i n f o
a b s t r a c t
Article history: Received 22 July 2016 Received in revised form 18 February 2017 Accepted 25 April 2017 Available online 27 April 2017
In this paper, in situ transmission electron microscopy investigations were carried out to study the microstructural evolution of ferritic/martensitic steel T91 under 1 MeV Krypton ion irradiation up to 4.2 1015 ions/cm2 at 573 K, 673 K, and 773 K. At 573 K, grown-in defects are strongly modified by blackdot loops, and dislocation networks together with black-dot loops were observed after irradiation. At 673 K and 773 K, grown-in defects are only partially modified by dislocation loops; isolated loops and dislocation segments were commonly found after irradiation. Post irradiation examination indicates that at 4.2 1015 ions/cm2, about 51% of the loops were a0 =2〈111〉 type for the 673 K irradiation, and the dominant loop type was a0 〈100〉 for the 773 K irradiation. Finally, a dispersed barrier hardening model was employed to estimate the change in yield strength, and the calculated ion data were found to follow the similar trend as the existing neutron data with an offset of 100e150 MPa. © 2017 Elsevier B.V. All rights reserved.
Keywords: Dislocation loops In situ ion irradiation Irradiation hardening Ferritic/martensitic steel Transmission electron microscopy
1. Introduction In metals and alloys, energetic particle bombardment creates displacement cascades and results in the formation of freely migrating defects and defect clusters. The cascade evolution can be described by four stages, including collisional, thermal spike, quenching, and annealing [1]. The annealing stage is a diffusion
* Corresponding author. 216 Talbot Laboratory, 104 South Wright Street, Urbana, IL 61801, USA. E-mail address: [email protected] (X. Liu). http://dx.doi.org/10.1016/j.jnucmat.2017.04.047 0022-3115/© 2017 Elsevier B.V. All rights reserved.
controlled process that is most influenced by the irradiation temperature. As a result, the size, density, and even spatial distribution of dislocation loops are temperature dependent. In a-iron and Febase body centered cubic (BCC) alloys, a particular effect of the irradiation temperature is that it affects the stability of a0 =2〈111〉 loops with respect to a0 〈100〉 loops [2]. By contrast, almost no radiation-induced a0 〈100〉 loops have been observed in other BCC alloys [3,4]. Modeling efforts have been carried out to explain the formation of a0 〈100〉 loops [2,5e7]. Eyre and Bullough [6] proposed that a0 〈100〉 loops could be formed by reaction 12 ½110 þ 12 ½110/½010
306
and
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
a0 =2〈111〉
1 ½110 2
loops
could
be
formed
by
reaction
1 ½001/1 ½111. 2 2
þ The proposed mechanism involves the formation of a0 =2〈111〉 type faulted loops, which was regarded unfavorable due to the high stacking fault energy. Recently, atomistic simulations have observed the formation of a0 〈100〉 loops involving the interactions of a0 =2〈111〉 loops [5,7]. Dudarev et al. [2] explained the relative stability of a0 〈100〉 and a0 =2〈111〉 loops by the reduction of the shear stiffness near the a g phase transition temperature, and pointed out that a0 =2〈111〉 loops are more likely to transform to a0 〈100〉 loops at high temperatures. Experimentally, both a0 =2〈111〉 and a0 〈100〉 loops have been reported in irradiated a-iron and Fe-Cr model alloys in previous studies [8e18]. Both loop types were reported by Gelles et al. in a series of Fe-Cr-C-Mo ferritic alloys neutron-irradiated at 673 K723 K [9]. Katoh et al. examined multiple Fe-Cr binary ferritic alloys following neutron irradiation at 698 K and found the loops are predominantly a0 〈100〉 type, only few a0 =2〈111〉 loops were confirmed in Fe-12Cr [11]. 86% of the dislocation loops in pure iron irradiated by neutrons at 573 K were identified as a0 〈100〉 loops by ndez-Mayoral and Go mez-Bricen ~ o [17]. Yao et al. irradiated Herna a-iron with Feþ ions and found that the dominant loop type changes from a0 =2〈111〉 to a0 〈100〉 as the irradiation temperature increases from 573 K to 773 K [18]. The results on simple systems such as pure iron and Fe-Cr model alloys provides a good starting point for the irradiation studies on more complex alloy systems such as ferritic/martensitic (F/M) steels. The additional alloying elements atoms in F/M steels introduces extra strain fields, which can decrease the mobility of defects and may also affect the formation, growth, and interactions of dislocation loops. For instance, chains of extended loops observed in pure iron [19] have not been observed in F/M steels. T91 is a representative 9Cr-1MoVNb F/M steel that has promising applications in advanced reactor systems. Ion and neutron irradiations on complex F/M steels such as T91 have been carried out by several groups [20e23]. However, the irradiation-induced dislocation loops in F/M steels have not been fully characterized and it is still not clear how the complex microstructure evolves differently under ion irradiation at different temperatures and what role dislocation loops play in the microstructural evolution process. In this study, transmission electron microscopy (TEM) was used to investigate the real-time microstructural evolution of T91 F/M steel during in situ Krypton ion irradiation at 573 K, 673 K, and 773 K, up to 4.2 1015 ions/cm2. The temperatures were chosen because it was expected that the dominant Burgers vector of dislocation loops might change in the selected temperature range [18]. In situ irradiation damage processes including dislocation loop formation, interactions of loops with grown-in defects (line dislocations and grain boundaries), and dislocation segment/network formation at three different temperatures were obtained. Distinct dislocation structures after irradiation were discussed. The average dislocation loop sizes and densities at different doses and temperatures were compared. The change in yield strength DsY was calculated for each dose and temperature condition using a dispersed barrier hardening model. The calculated data were compared with existing proton data and neutron data. Finally, to a limited extent, attempts were made to correlate the ion data with neutron data.
Table 1 Chemical composition of the as-received T91 in wt.%. Fe Bal. S 0.006
C 0.089 Si 0.28
Cr 9.24 P 0.021
Cu 0.08 V 0.21
Mn 0.47 Nb 0.054
Mo 0.96 Co 0.019
Ni 0.16 N 0.035
into 3 mm discs by electrical discharge machining (EDM). The discs were mechanically thinned to 100 mm and then eletropolished to electron transparency with a Struer Tenupol-5 twin-jet polisher using electrolyte containing 5% perchloric acid and 95% methanol at 30 C and 23 V. The in situ ion irradiation experiments were performed at the IVEM-Tandem facility, Argonne National Laboratory (ANL). The experimental setup is similar to the one described by Li et al. [24]. A Hitachi 9000 NAR electron microscope operated at 200 kV was used for TEM imaging. T91 thin foil specimens were irradiated by 1 MeV Krypton ions at 573 K, 673 K, and 773 K up to a fluence of 4.2 1015 ions/cm2. The incident ion beam was 30 to the electron beam and on average ~15 to the foil normal. At ~15 tilt, grains satisfying desired diffraction conditions were selected for TEM observations. The ion irradiation was paused for 0.5e1 h on average at 6.0 1014, 1.8 1015, and 3.0 1015 ions/cm2 in order to carry out TEM examinations. The samples tilts were adjusted accordingly (typically ± 1e3 ) to ensure that the TEM micrographs were taken under similar diffraction conditions, so that the defect density and size were comparable. The irradiation temperature was measured by a thermocouple attached to the heating cup of a double tilt heating sample holder and was kept within ±3 K during the irradiation. The TEM specimen was located in the heating cup with direct contact, thus the irradiation temperature represented the specimen temperature accurately. The ion flux was kept at 6.25 1011 ions/(cm2$s) for all irradiations. The doses were calculated using:
dpa ¼ 108 n
F N
;
(1)
where N is the number density, F is the fluence in ions/cm2, and n is the damage production in vacancies/(ion$Å) [24]. The damage production for 1 MeV Krypton in T91 is 1.41 vacancies/(ion$Å), obtained from SRIM-2008 Kinchin-Pease quick calculation using a displacement energy of 40 eV [25]. Therefore, the dose rate is 1.04 103 dpa/s for all irradiations and the highest fluence 4.2 1015 ions/cm2 corresponds to 7.0 dpa. 3. Results 3.1. As-received microstructure The as-received microstructure consists of martensite laths with carbide precipitates distributed along the lath boundaries and grain boundaries. Dislocation density varies significantly from grain to grain. Tangled dislocation areas with densities around 6.0e8.0 1013 m2 are commonly found, and some dislocationfree areas were also observed. 3.2. Dynamic microstructural evolution
2. Experimental The T91 steel was normalized at 1311 K for 1 h, air cooled, and then tempered at 1033 K for 1 h and air cooled. The nominal composition is listed in Table 1. The as-received T91 plate was cut
The real-time microstructural evolution of T91 irradiated up to 4.2 1015 ions/cm2 (~ 7 dpa) at three different temperatures (573 K, 673 K, and 773 K) is shown in Figs. 1e4. From these TEM micrographs, it is found that irradiation-induced dislocation loops began to show up at approximately 6.0 1014 ions/cm2 (~ 1 dpa) and loop
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
307
Fig. 1. TEM bright field images showing the dislocation-loop formation and development in T91 thin foil irradiated by 1 MeV Krypton ions at 573 K at different doses: (a) asreceived, (b) 6.0 1014 ions/cm2, (c) 1.8 1015 ions/cm2, (d) 3.0 1015 ions/cm2, and (e) 4.2 1015 ions/cm2. All the images were taken near the [001] zone.
Fig. 2. TEM bright field images showing the dislocation loop formation and development in a T91 thin foil irradiated by 1 MeV Krypton ions at 673 K at different doses: (a) asreceived, (b) 6.0 1014 ions/cm2, (c) 1.8 1015 ions/cm2, (d) 3.0 1015 ions/cm2, and (e) 4.2 1015 ions/cm2. The images were taken near the ½110 zone.
density increased by 1e2 orders of magnitude from 6.0 1014 ions/ cm2 to 1.8 1015 ions/cm2 (~ 3 dpa). After the density of dislocation loops reached about 2 1021 m3 to 4 1021 m3, the interactions of these loops with grown-in defects (line dislocations or grain boundaries) became apparent, as manifested by the changes in grown-in defects shown in Figs. 1e4. For the specimen irradiated at 573 K, the microstructure evolution is shown in Fig. 1. All the TEM micrographs in Fig. 1 were taken under nearly identical kinematical diffraction conditions, in which the sample was tilted near the [001] zone and only g110 was strongly excited. As can be seen, several black-dot dislocation loops formed at 6.0 1014 ions/cm2, examples are the ones marked by red arrows in Fig. 1 (b). From 6.0 1014 ions/cm2 to 1.8 1015 ions/ cm2, the loop density increased significantly from 6.3 1020 m3 to 3.8 1021 m3. Although the loop distribution was relatively uniform, i.e. no apparent dislocation decoration occurred, strong
interactions between grown-in line dislocations and irradiationinduced dislocation loops occurred. Some line dislocations were completely altered and they linked together with surrounding defects. An example is shown in the blue circle in Fig. 1, where gradual changes in the morphology of a dislocation can be readily seen. Observable changes also occurred in the lath boundary shown in Fig. 1. As the dose increased, increasing numbers of dislocation segments attached to the lath boundary. Fig. 2 shows the evolution of grown-in line dislocations and irradiation-induced dislocation loops in the specimen irradiated at 673 K. Similarly, only a few loops were observed at 6.0 1014 ions/ cm2, and the loop density is estimated to be around 4 1020 m3. Some of the loops were indicated by red arrows in Fig. 2 (b). Similar to the 573 K case, noticeable increase (by a factor of ~10) in the loop density was observed between 6.0 1014 ions/cm2 and 1.8 1015 ions/cm2. In Fig. 2 (c)-(e), both black dots and extended loops were
308
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
Fig. 3. Enlarged TEM bright field images showing the evolution of dislocations in T91 irradiated by 1 MeV Krypton ions at 673 K at different doses: (a) 6.0 1014 ions/cm2, (b) 1.8 1015 ions/cm2, (c) 3.0 1015 ions/cm2, and (d) 4.2 1015 ions/cm2.
Fig. 4. TEM bright field images showing the dislocation-loop formation and development in a T91 thin foil irradiated by 1 MeV Krypton ions at 773 K at different doses: (a) asreceived, (b) 6.0 1014 ions/cm2, (c) 1.8 1015 ions/cm2, (d) 3.0 1015 ions/cm2, and (e) 4.2 1015 ions/cm2. The images were taken near the [001] zone.
observed, and some of them were marked by red arrows. In Fig. 2 (d) and (e), dislocation segments (some were marked by red arrows) formed by coalescence of adjacent dislocation loops and/or line dislocations. The dislocation structure at 4.2 1015 ions/cm2 did not show any apparent alignment. After 1.8 1015 ions/cm2, strong interaction between preexisting line dislocations and dislocation loops occurred, as shown in the changes in the dislocations D1 and D2 in Fig. 2 (d) and (e). Several jogs/kinks (marked by blue triangles) in dislocation D1 were observed, and apparent changes in the curvature of dislocation D2 can be readily seen. These complex changes indicate dislocation climb by absorbing point defects and maybe further interactions with dislocation loops as well. Similar kinks and changes in curvature have also been reported in lower dose ion irradiated Fe-Cr model alloys [26].
Fig. 3 is a set of enlarged TEM images showing more clearly the localized interactions at 673 K. Some dislocation loops (e.g. the one marked in yellow circle) away from grown-in defects grew in size without interference, indicating the interactions were localized. Gradual changes in the lath boundary marked by red circle and the dislocation marked by blue circle were also observed. Compared to Fig. 1, the interaction shown in Figs. 2 and 3 is much more localized: the overall shape of the grown-in line dislocation was only slightly different, although some parts became curved after interacting with surrounding dislocation loops. For the 773 K case, Fig. 4 shows the microstructure evolution of a dislocation-rich region. As can be seen, little dislocation loops were observed at 6.0 1014 ions/cm2. At 1.8 1015 ions/cm2, most dislocation loops were black-dots. At 3.0 1015 ions/cm2, some
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
black-dots grew into much larger dislocation loops. The loops in the lower part of Fig. 4 (d) had the same Burgers vector and were wellaligned. Localized interactions between irradiation-induced dislocation loops and grown-in dislocations were again observed in Fig. 4 (c) to (e). After irradiation, both isolated loops and dislocation segments were found and no obvious dislocation network was observed. It should be noted that the physical processes of in situ irradiation experiments could differ from bulk ex situ irradiation experiments, largely due to the two nearby surfaces of TEM thin foils. The free surfaces can act as sinks and remove a portion of the mobile defects such as a0 =2〈111〉 loops. In the in situ irradiation study of Fe-Cr-Ni austenitic steels at 450 C [27], 20 nm was subtracted considering the depletion of defects near surfaces. F/M systems have lower defect mobility and high density internal dislocation sinks, this effect is expected to be much less. The surface effects were not studied in the present work. 3.3. 〈111〉 〈100〉 loop transition Previous studies reported that the dislocation loops induced by both ion and neutron irradiations in a-iron, Fe-Cr model alloys, and F/M steels are a0 〈100〉 or a0 =2〈111〉 type [8e18]. In this paper, it is also assumed that all the loops are a0 〈100〉 or a0 =2〈111〉 type. The Burgers vectors of dislocations loops were be determined using bright-field TEM images taken under selected diffraction conditions. a0 〈100〉 and a0 =2〈111〉 loops can be distinguished by (1) standard g,b analysis or (2) comparing the orientation of the loops with respect to the diffraction vectors. Fig. 5 shows the dislocation loops in a T91 specimen that was irradiated to 4.2 1015 ions/cm2 at 773 K. Fig. 5 (a) to (c) were obtained by tilting the specimen to different diffraction conditions near the [001] zone. In Fig. 5 (a), only g020 was strongly excited, therefore only a0[001] loops are invisible (g,b ¼ 0) and all other types of loops (a0[010] and all a0 =2〈111〉 loops) should be visible. However, only a0[010] loops marked as “a” showed up in Fig. 5 (a), indicating that there was no a0 =2〈111〉 loops in this area. Both a0[100] and a0[010] loops are visible in Fig. 5 (b) and (c) since g,bs0. The loop marked as “b” showed up in Fig. 5 (b) and (c) but disappeared in Fig. 5 (a), indicating that it is a a0[100] loop. Comparing Fig. 5 (a) to (c), it is found that after irradiation to 4.2 1015 ions/cm2 at 773 K, the dislocation loops are dominantly a0[010] type in this area. Since these loops have the same Burgers vector, they were well-aligned as shown in Fig. 5. Similar TEM
309
examinations were carried out in other grains, and the loops were also found to be dominantly a0 〈100〉 type. Similar grain-level localized loop alignment was commonly observed and should be characteristic for the irradiation at 773 K. For the 673 K case, however, distinct dislocation microstructure was observed. Fig. 6 shows the typical microstructure found in the specimen irradiated to 4.2 1015 at 673 K. It was taken near the [001] zone and only g110 was strongly excited, as shown in the insertion in Fig. 6 (a). Under this diffraction condition, visible loops are a0[010], a0[100], a0/2[111], a0/2½111, a0/2½111, and a0/2½111. The a0[010] and a0[100] loops can be easily distinguished from a0 =2〈111〉 loops by their orientations with respect to the diffraction vectors, i.e. a0[010] loops are perpendicular to g020 and 45 to g110 , a0[100] loops are parallel to g020 and 45 to g110 . Moreover, a0 〈100〉 loops are close to edge-on since the images were taken near the [001] zone axis. The loops in Fig. 6 (b)-(d) were distinguished by this method. The ones in red circles are a0 〈100〉 loops, and the ones in blue circles are a0 =2〈111〉 loops. The overall dislocation structure is shown in Fig. 6 (a). Fig. 6 (b)-(d) are enlarged TEM images showing the a0 =2〈111〉 and a0 〈100〉 loops. Counting only the wellresolved loops in the three areas near Fig. 6 (b), (c), and (d), statistical data of the Burgers vector are shown in Table 2. It is estimated that the 〈111〉=〈110〉 ratio is around 1.02 ± 0.38 for the 673 K case. 3.4. Loop size and density In order to quantitatively characterize the irradiation-induced dislocation loops, the average loop sizes and densities were obtained. For the loop size measurement, TEM kinematic bright field images were used. Therefore, the results might be not accurate for loops less than 10 nm, especially for the small black-dot loops around 2 nm. On average, 40e60 loops were measured for each dose and temperature condition to reduce the statistical errors. Figs. 7e9 show the histograms of the loop size distribution for the irradiation at 573 K, 673 K, and 773 K, respectively. At 573 K, the loops did not show significant increase in size, as shown in Fig. 7. Most loops are small black-dots and only a few extended loops around 6e10 nm were found at 3.0 1015 (~5 dpa) and 4.2 1015 ions/cm2 (~7 dpa). At 673 K, the loops also started as black-dots around 2 nm at 6.0 1014 ions/cm2 (~1 dpa), but the loop size increased significantly as the dose increased, and loops larger than 10 nm were commonly found at 4.2 1015 ions/cm2. At 773 K, few
Fig. 5. Different diffraction vectors: (a) g020 , (b) g110 , and (c) g110 were used to image the dislocation loops in a T91 thin foil that has been irradiated by 1 MeV Krypton ions to 4.2 1015 ions/cm2 at 773 K. The zone axis is [001].
310
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
Fig. 6. Dislocation loops in a T91 thin foil that has been irradiated by 1 MeV Krypton ions to 4.2 1015 ions/cm2 at 673 K: (a) low-mag image showing the overall microstructure, (b) an area with both loops, (c) an area with dominantly a0 =2〈111〉 loops, (d) an area with more a0 〈100〉 loops. All the images were taken with only g110 strongly excited near the [001] zone.
Table 2 Statistical analysis of a0 =2〈111〉 and a0 〈100〉 loops in T91 irradiated to 4.2 1015 ions/cm2 at 673 K. No. of a0 =2〈111〉 loops
No. of a0 〈100〉 loops
〈111〉=〈110〉 ratio
Area 1 Area 2 Area 3
20 14 13
17 10 19
1.18 1.40 0.68
Total
47
46
1.02 ± 0.38
black-dots were observed at 6.0 1014 ions/cm2 and therefore no size histogram was obtained. Large isolated loops around 20 nm were commonly found at 4.2 1015 ions/cm2. Based on these histograms, the average loop sizes were plotted in Fig. 10 (a), in which the error bars were calculated using the student t-distribution and represent the 95% confidence interval. In addition to the loop size, the loop density as well as line dislocation density were also obtained, as shown in Fig. 10 (b). It should be noted that during the in situ irradiations, the thinner areas near the hole created by jet polishing were bended significantly and therefore the thickness fringes method could not be applied for sample thickness measurement. All the TEM examinations were focused on thick enough regions (not too thin to avoid bending, and not too thick to get good contrast). For the loop density calculation, all the sample thickness was assumed to be 100 nm, which is apparently a rough estimate. This uncertainty only affects the loop density calculation, since the line dislocation density is only related to the area of the chosen regions. In the line dislocation density calculation, both grown-in dislocations and dislocation segments formed from irradiation were taken into account. As shown in Fig. 10 (b), for the 573 K irradiation, the loop density began to drop between 1.8 1015 ions/cm2 (~3 dpa) and 3.0 1015 ions/cm2 (~5 dpa), accompanied by a steady increase in the line dislocation density. At 673 K, the drop in loop density occurred between 3.0 1015 ions/cms2 and 4.2 1015 ions/cm2 (~7 dpa), whereas the line dislocation density showed appreciable increase between 1.8 1015 ions/cm2 and 3.0 1015 ions/cm2, due to the formation of dislocation segments. For the irradiation at 773 K, the loop density increased significantly between 6.0 1014 ions/cm2
(~1 dpa) and 1.8 1015 ions/cm2 and then dropped, and the line dislocation density slightly decreased between 1.8 1015 ions/cm2 and 4.2 1015 ions/cm2. Another finding is that at 6.0 1014 ions/ cm2, the loop density at 773 K is significantly lower (by as much as an order of magnitude) than that at 573 K and 673 K. 4. Discussions 4.1. Relationship between loop transition and the dislocation structure In situ TEM irradiation allows the dynamic observation of the microstructural evolution of the same area at different dose levels. For complex systems like F/M steels, differences in the initial microstructure could lead to variations in the irradiated microstructure. In order to compare the representative microstructure of T91 irradiated at different temperatures, post-irradiation examination (PIE) of multiple areas of the irradiated T91 TEM specimens was carried out. Fig. 11 shows the representative dislocation structure of T91 steel irradiated to 4.2 1015 ions/cm2at 573 K, 673 K, and 773 K, respectively. Dislocation networks were commonly found in the sample irradiated at 573 K, as shown in Fig. 11 (a). Some black-dot dislocation loops are also observed in addition to the network dislocations. In contrast to the dislocation network in Fig. 11 (a), isolated dislocation loops were commonly found in the samples irradiated at 673 K and 773 K, as shown in Fig. 11 (b) and (c). However, the difference is that the dislocation loops in Fig. 11 (b) are randomly oriented, whereas the dislocation loops in Fig. 11 (c) are wellaligned.
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
311
Fig. 7. Loop size distribution histograms of T91 irradiated at 573 K at different doses: (a) 6.0 1014, (b) 1.8 1015, (c) 3.0 1015, and (d) 4.2 1015 ions/cm2.
The temperature dependence of the observed dislocation structures should be dependent on the a0 =2〈111〉 a0 〈100〉 dislocation loop transition that happens at elevated temperatures in a-iron, Fe-Cr ferritic model alloys, and F/M steels. The distinction is important since, at elevated temperatures, the a0 =2〈111〉 loops are highly mobile and a0 〈100〉 loops are sessile. Therefore, a0 =2〈111〉 loops can migrate over long distances and interact with distant grown-in defects, leading to formation of dislocation segments and network dislocations at lower dose, whereas a0 〈100〉 can only act as stationary defect sinks once formed. If the a0 〈100〉 loops happen to form near grown-in defects, localized interaction between loops and nearby grown-in defects is possible. As an example, Figs. 2 and 3 shows clearly the process of localized interactions. The stability of a0 〈100〉 loops with respect to a0 =2〈111〉 loops is highly temperature dependent. Yao et al. performed Feþ self-ion irradiation on pure Fe and reported that about 80% of the loops were of the a0 〈100〉 type at 673 K [18]. In the present work, both a0 〈100〉 and a0 =2〈111〉 loops were found at 573 K, the ratio was not determined. However, based on the fact that only at 573 K, loops were observed to escape to the surface and disappeared, mobile a0 =2〈111〉 should be the dominant type. At 673 K, the ratio of a0 =2〈111〉 to a0 〈100〉 loops is around 1.02 (from Table 2). At 773 K,
the loops were found to be predominantly a0 〈100〉 type. Based on our results at three different temperatures, the a0 =2〈111〉 a0 〈100〉 loop transition is confirmed. Although at 673 K, only ~49% of the loops are a0 〈100〉 type, different from the 80% in pure iron reported by Yao et al., the general trend is consistent [18]. The difference could be due to the effect of alloying elements such as chromium, as previous studies suggest that the addition of chromium tends to equalize the proportion of a0 =2〈111〉 and a0 〈100〉 loops [28], or to increase the fraction of a0 =2〈111〉 loops [29]. 4.2. Modeling of irradiation hardening In this study, heavy ion irradiation was employed to emulate neutron irradiation due to the degree of similarity in their weighted average recoil spectra. The disadvantage of heavy ion irradiation is that it produces a non-uniform dose profile over a shallow region (less than 600 nm for 1 Me Kr ions) beneath the sample surface. Our in situ TEM observation avoided this problem since the thickness of thin area is about 100 nm, where the dose profile can be approximated to be constant. The difficulty lies in that it is impossible to directly measure the mechanical properties by small scale tests such as nanoindentation, due to various factors contributing to the
312
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
Fig. 8. Loop size distribution histograms of T91 irradiated at 673 K at different doses: (a) 6.0 1014, (b) 1.8 1015, (c) 3.0 1015, and (d) 4.2 1015 ions/cm2.
Fig. 9. Loop size distribution histograms of T91 irradiated at 773 K at different doses: (a) 1.8 1015, (b) 3.0 1015, and (c) 4.2 1015 ions/cm2.
measurements. The results of shallow (less than 100 nm) indents will be largely influenced by the size effect and deeper indents will inevitably sample the whole non-uniform dose profile [30]. Quantitative comparison of the results of small scale tests on heavy ion irradiated samples with the data from neutron irradiated samples seems to be almost impossible, at least very challenging. Alternatively, based on the observed defect microstructure it is
possible to calculate the mechanical properties such as change in yield strength DsY using well-accepted models. The dispersed barrier hardening model has been used in previous studies [12,15,21,28,31e34] to establish the microstructure-property relationship and to evaluate the hardening induced by neutron and ion irradiations. In this model, the change in yield strength DsY is related to the defect microstructure by:
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
313
Fig. 10. Statistical information of defects in T91 steel irradiated at different temperatures and doses (a) Average loop sizes; (b) average line dislocation densities and loop densities.
Fig. 11. TEM bright field images showing different dislocation structures in T91 steel irradiated by Krypton ions to 4.2 1015 ions/cm2 at different temperatures: (a) Dislocation network decorated by black-dot dislocation loops at 573 K; (b) random distributed a0 =2〈111〉 and a0 〈100〉 dislocation loops at 673 K; and (c) well-aligned a0 〈100〉 dislocation loops at 773 K.
pffiffiffiffiffiffi
DsY ¼ Mamb rd;
(2)
where M is the Taylor factor, a is the defect barrier strength, m is the shear modulus, b represents the magnitude of the Burgers vector of dislocations, r is the defect density and d is the defect diameter. In our calculation M ¼ 3:06, m ¼ 82 GPa, and b ¼ 2:49 nm, which are consistent with the values used in Fe-Cr alloys [12,15,28,32,34]. In general, the irradiation hardening results from multiple microstructure modifications under energetic ion/neutron bombardment, including voids, precipitates, dislocation loops, line dislocations, etc. The overall effect on the yield strength can be evaluated by the linear superposition of the contributions from various defects: DsY ¼ DsYV þ DsYP þ DsYL þ DsYD þ DsYR , where DsYV , DsYP , DsYL , DsYD and DsYR represent the contribution from voids, precipitates, dislocation loops, line dislocations, and others (such as dislocation decoration, segregation of solute atoms), respectively. In the case of T91, no obvious voids or irradiation-induced a0 precipitates were observed during the in situ irradiation. This is also consistent with the observations in previous studies [21,23,35]. The favorable explanation is that 9e12%Cr F/M steels are well known for their excellent swelling resistance [36] and the final dose 4.2 1015 ions/cm2 (~7 dpa) is still in the transient regime [37]. Thus, only negligible amount of voids is expected at the final dose. Also, the Cr content is 9.24 wt%, which is close to the threshold value for a a0 phase separation in Fe-Cr alloys [38], indicating very few, if any, a0 precipitates are expected. Indeed, no a0 precipitates were reported in F/M steel T91 after proton irradiation to 10 dpa at 400e500 C or neutron irradiation up to 184 dpa at 413 C [35]. As a first order approximation, the contribution from voids and a0 precipitates is neglected due to their marginal amount and
small barrier strength [33,34]. DsYR is also ignored for simplicity. Then we have:
qffiffiffiffiffiffiffiffiffi
DsY ¼ DsYL þ DsYD ¼ MaL mb rL dL þ MaD mb
qffiffiffiffiffiffiffiffi
rirr: D
qffiffiffiffiffiffiffiffiffiffiffiffi
; runirr: D (3)
where aL , rL , and dL are the barrier strength, average density, and average size of dislocation loops, respectively; aD is the barrier unirr: is the line dislocation strength of line dislocations, rirr: D and rD density after and before irradiation, respectively. Here aD ¼ 0:64 and aL is temporarily taken to be 0.40, which are consistent with the values used in literature [15,28,31,32,34]. Using the loop sizes and loop/line dislocation densities from Section 3.4, the increase in yield strength at different dose and irradiation temperature conditions was obtained and is shown in Fig. 12. As can be seen, DsY increased the fastest with increasing dose at 573 K, compared to the 673 K and 773 K irradiations. At 773 K, DsY reached its maximum around 3e5 dpa and then slightly decreased at 7 dpa. The behavior of DsY falls in between at 673 K. For both 573 K and 673 K irradiations no saturation in DsY was observed. The general trend is consistent with the conventional irradiation hardening characteristics of F/M steels, i.e. more pronounced hardening at lower irradiation temperatures and earlier saturation at higher irradiation temperatures [39]. Apparently, the calculated DsY strongly depends on the choice of the barrier strength a. Table 3 lists the calculated DsY at different temperature and dose conditions, for four different aL values. It is easy to see that the 573 K data are relatively insensitive to aL , since the major contribution to the hardening comes from the dislocation segments and network dislocations. At 7 dpa, the 573 K data only
314
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
Fig. 12. Calculated change in yield strength DsY at different irradiation temperatures and doses, estimated using the dispersed barrier hardening model.
changed by 76 MPa when aL changed from 0.1 to 0.4. The 773 K data are most sensitive to aL , because the dominant contribution comes from individual loops. At 7 dpa, the 773 K data changed by 218 MPa when aL changed from 0.1 to 0.4. For simplicity, to the first order approximation, the calculated values in all the figures are based on aL ¼ 0:4, which was commonly used in Fe-Cr model alloys [15,28,31,32,34]. However, we should keep in mind that aL should be lower for lower temperatures, as the fraction of a0 =2〈111〉 loops is higher at lower temperatures. The barrier strength of mobile a0 =2〈111〉 loops should be smaller, compared to that of sessile a0 〈100〉 loops. Efforts were also made to compare the calculated change in yield strength DsY with existing experimental data [21,39e41]. The dose dependence and temperature dependence of irradiation hardening in T91 are shown in Figs. 13 and 14, respectively. For simplicity, only data at similar temperatures (~573 K, 673 K, and 773 K) were shown in Fig. 13 and only data in the similar dose range (7e12 dpa) were shown in Fig. 13. As can be seen in Fig. 13, there is noticeable difference in the data at different irradiation temperatures. Three dotted lines were plotted to show the trends. It is easy to find that the calculated results are reasonably consistent with the existing data. At 573 K, DsY increases with increasing dpa and saturates at around 20 dpa. At 673 K, DsY increases with increasing dpa at a lower rate and saturates earlier (~10 dpa). At 773 K, DsY shows slightly increase and saturates around 3 dpa. In Fig. 14, the dose of all the proton data is the same (7 dpa) as that of the present work. For the neutron data, only these in the range of 9e12 dpa were plotted in order to make the data more comparable. For the neutron data below 723 K, DsY decreases with
Fig. 13. Change in yield strength DsY at different doses. In order to compare the results with existing data at similar temperatures, two sets of proton data [21] and three sets of neutron data [39e41] are shown. The dotted lines are plotted to show the trends.
increasing temperature, and above 773 K DsY 0. The trend is shown in the dotted royal blue line. The ion irradiation data (both proton and krypton) are connected by another dotted violet line, which turns out to be of the similar shape. This manifests that the ion data has very similar temperature dependence as the neutron data. It is also noticed that the ion data was shifted upwards from the neutron data by approximately 100e150 MPa, indicating that more pronounced hardening is expected for ion irradiations. Considering that the dose for the neutron data (9e12 dpa) is higher than that of the ion data (7 dpa), the upward shift in DsY of the ion data becomes even more compelling. One possible reason for the upward shift is related to the thermal annealing effect accompanying low dose rate neutron irradiations. Since the dose rate of ion irradiation is 103 to 104 higher than that of neutron irradiation, the material under ion irradiation experiences much less thermal annealing and therefore the defects (both grow-in dislocations and irradiation-induced defects) have much smaller chance to be annealed and the surviving defects are responsible for more noticeable hardening. On the other hand, the offset also indicates that the microstructure is generally similar, except for differences in the defect density and size. For cases in which ion irradiations cannot produce the comparable microstructure as that of neutron irradiations, the ion-to-neutron is much more challenging. An example is that for Fe-Cr alloys above the a a0 phase separation threshold value, a0 precipitates are only frequently observed in neutronirradiated, rather than ion-irradiated specimens.
Table 3 Calcuated yield strength changes at different irradiation conditions for aL ¼ 0.1, 0.2, 0.3, and 0.4.
DsY
aL
0.1
0.2
0.3
0.4
573 K
1 3 5 7
dpa dpa dpa dpa
14.1 118.3 260.1 521.9
28.2 155.2 294.7 547.2
42.3 192.2 329.2 572.5
56.3 229.2 363.7 597.7
673 K
1 3 5 7
dpa dpa dpa dpa
10.6 41.1 116.6 190.9
21.2 82.1 162.5 243.6
31.9 123.2 208.5 296.3
42.5 164.2 254.5 349.0
773 K
1 3 5 7
dpa dpa dpa dpa
2.1 32.1 73.8 6.7
4.2 64.3 128.1 65.8
6.2 96.4 182.3 138.3
8.3 128.6 236.6 210.9
(MPa)
Fig. 14. Change in yield strength DsY at different irradiation temperatures. The dose for all the proton data [21] and present work is 7 dpa, and all the neutron data [39e41] are in the range of 9e12 dpa. The data are connected by dotted lines to show the trends (the mean value of the proton data and present work at 673 K was used to draw the dotted trend line).
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
Another important conclusion from Fig. 14 is related to the temperature shift. It is commonly believed that in order to correlate ion irradiations with neutron irradiations, the irradiation temperature of ion irradiations should be higher than neutron irradiations. The temperature difference is known as the temperature shift. In contrast to the temperature shift predicted by Mansur [42] and observed in the swelling data of ion irradiations [43], our preliminary results indicate that no apparent temperature shift was observed in the irradiation hardening of F/M steel T91. This is consistent with the swelling data of neutron irradiation in copper [44,45], where no temperature shift was observed. The temperature shift of ion irradiation is still a controversial topic, and it is recently found that the temperature shift could be due to the effect of injected interstitials alone [46]. The reason that no temperature shift is found in our ion irradiations is that all the Krypton ion irradiations were carried out in situ on TEM thin foil specimens that were around 100 nm thick and therefore the problems related to conventional ion irradiations such as injected interstitials and compositional modifications were avoided from the very beginning. 5. Summary In summary, strong dislocation loop-preexisting defects interaction was observed in F/M steel T91 at 573 K, 673 K, and 773 K. The interaction took different forms at different temperatures. At 573 K, both line dislocations and grain boundaries were modified by dislocation loops. The shape of some line dislocations was entirely changed, with more loops attached to the grain boundaries as the dose increased. Dense dislocation network was found after irradiation. For the irradiation at 673 K, grown-in dislocations were locally modified. Jogs/kinks and changes in curvature in grown-in dislocations were observed. At 3.0 1015 ions/cm2, adjacent dislocation loops began to coalesce and dislocation segments started to form. At 4.2 1015 ions/cm2, dislocation segments were commonly observed and some were linked together with preexisting line dislocations. At 773 K, isolated dislocation loops were again observed but in well-aligned fashion rather than random-oriented as the ones found at 673 K. This indicates that these loops have exactly the same Burgers vector in one region. This grain-level localized alignment should be a special feature of a0 〈100〉 loops. These phenomena can be well explained by the 〈111〉 〈110〉 loop transition, which was supported by the post-irradiation examination. Based on the loop size and defect densities, a dispersed barrier hardening model was used to establish the microstructureproperty relationship. The calculated changes in yield strength at different irradiation temperatures and doses were compared with existing data. It was found that the ion data follow the similar trend of the neutron data, expect for an offset of 100e150 MPa. No apparent temperature shift was observed. Acknowledgements The authors would like to thank Pete Baldo and Edward Ryan at Argonne National Laboratory for their assistance with the irradiation experiments. This work was supported by the funding from the DOE Office of Nuclear Energy's Nuclear Energy University Programs (NEUP) under Contract No. DE-NE0008291. The TEM experiments were carried out in part in the Frederick Seitz Materials Research Laboratory Central Facilities, University of Illinois. The electron microscopy with in situ ion irradiation was accomplished at Argonne National Laboratory at the IVEM-Tandem Facility, a U.S.
315
Department of Energy Facility funded by the DOE Office of Nuclear Energy, operated under Contract No. DE-AC02-06CH11357 by UChicago Argonne, LLC and the DOE Office of Nuclear Energy under DOE Idaho Operations Office Contract DE-AC07-051D14517 as part of a Nuclear Science User Facilities experiment. References [1] G.S. Was, Fundamentals of Radiation Damage Materials Science: Metals and Alloys, Springer Science & Business Media, 2007. [2] S.L. Dudarev, R. Bullough, P.M. Derlet, Effect of the ag phase transition on the stability of dislocation loops in bcc iron, Phys. Rev. Lett. 100 (2008) 135503, http://dx.doi.org/10.1103/PhysRevLett.100.135503. [3] R.C. Rau, F.S. D’aragona, R.L. Ladd, Neutron damage in molybdenum irradiated at high temperatures, Philos. Mag. 21 (1970) 441e452, http://dx.doi.org/ 10.1080/14786437008238430. [4] T. Matsui, S. Muto, T. Tanabe, TEM study on deuterium-irradiation-induced defects in tungsten and molybdenum, J. Nucl. Mater 283e287 (2000) 1139e1143, http://dx.doi.org/10.1016/S0022-3115(00)00169-0. [5] H. Xu, R.E. Stoller, Y.N. Osetsky, D. Terentyev, Solving the Puzzle of <100> interstitial loop formation in bcc iron, Phys. Rev. Lett. 110 (2013) 265503, http://dx.doi.org/10.1103/PhysRevLett.110.265503. [6] B.L. Eyre, R. Bullough, On the formation of interstitial loops in b.c.c. metals, Philos. Mag. 12 (1965) 31e39, http://dx.doi.org/10.1080/ 14786436508224943. [7] J. Marian, B.D. Wirth, J.M. Perlado, Mechanism of formation and growth of <100> interstitial loops in ferritic materials, Phys. Rev. Lett. 88 (2002) 255507, http://dx.doi.org/10.1103/PhysRevLett.88.255507. [8] B.C. Masters, Dislocation loops in irradiated iron, Philos. Mag. 11 (1965) 881e893, http://dx.doi.org/10.1080/14786436508223952. [9] D.S. Gelles, Microstructural examination of neutron-irradiated simple ferritic alloys, J. Nucl. Mater 108e109 (1982) 515e526, http://dx.doi.org/10.1016/ 0022-3115(82)90523-2. [10] A. Okada, T. Yasujima, T. Yoshiie, I. Ishida, M. Kiritani, Mechanical property changes and microstructures of iron irradiated with fission and fusion neutrons, J. Nucl. Mater 179e181 (1991) 1083e1087, http://dx.doi.org/10.1016/ 0022-3115(91)90280-K. [11] Y. Katoh, A. Kohyama, D.S. Gelles, Swelling and dislocation evolution in simple ferritic alloys irradiated to high fluence in FFTF/MOTA, J. Nucl. Mater 225 (1995) 154e162, http://dx.doi.org/10.1016/0022-3115(94)00669-5. [12] S. Porollo, A. Dvoriashin, A. Vorobyev, Y. Konobeev, The microstructure and tensile properties of FeeCr alloys after neutron irradiation at 400 C to 5.5e7.1 dpa, J. Nucl. Mater 256 (1998) 247e253, http://dx.doi.org/10.1016/S00223115(98)00043-9. [13] Y.V. Konobeev, A.M. Dvoriashin, S.I. Porollo, F.A. Garner, Swelling and microstructure of pure Fe and FeeCr alloys after neutron irradiation to ~26dpa at 400 C, J. Nucl. Mater 355 (2006) 124e130, http://dx.doi.org/10.1016/ j.jnucmat.2006.04.011. [14] M. Matijasevic, E. Lucon, A. Almazouzi, Behavior of ferritic/martensitic steels after n-irradiation at 200 and 300 C, J. Nucl. Mater 377 (2008) 101e108, http://dx.doi.org/10.1016/j.jnucmat.2008.02.063. [15] M. Matijasevic, A. Almazouzi, Effect of Cr on the mechanical properties and microstructure of FeeCr model alloys after n-irradiation, J. Nucl. Mater 377 (2008) 147e154, http://dx.doi.org/10.1016/j.jnucmat.2008.02.061. [16] M. Matijasevic, W. Van Renterghem, A. Almazouzi, Characterization of irradiated single crystals of Fe and Fee15Cr, Acta Mater 57 (2009) 1577e1585, http://dx.doi.org/10.1016/j.actamat.2008.11.042. ndez-Mayoral, D. Go mez-Bricen ~ o, Transmission electron microscopy [17] M. Herna study on neutron irradiated pure iron and RPV model alloys, J. Nucl. Mater 399 (2010) 146e153, http://dx.doi.org/10.1016/j.jnucmat.2009.11.013. [18] Z. Yao, M.L. Jenkins, M. Hern andez-Mayoral, M.a. Kirk, The temperature dependence of heavy-ion damage in iron: a microstructural transition at elevated temperatures, Philos. Mag. 90 (2010) 4623e4634, http://dx.doi.org/ 10.1080/14786430903430981. ndez-Mayoral, Z. Yao, M.L. Jenkins, M.A. Kirk, Heavy-ion irradiations [19] M. Herna of Fe and FeeCr model alloys Part 2: damage evolution in thin-foils at higher doses, Philos. Mag. 88 (2008) 2881e2897, http://dx.doi.org/10.1080/ 14786430802380477. [20] M. Grosse, Y. Dai, S. Van Petegem, Irradiation-induced structural changes in martensitic steel T91, J. Nucl. Mater 356 (2006) 112e117, http://dx.doi.org/ 10.1016/j.jnucmat.2006.05.026. [21] G. Gupta, Z. Jiao, A.N. Ham, J.T. Busby, G.S. Was, Microstructural evolution of proton irradiated T91, J. Nucl. Mater 351 (2006) 162e173, http://dx.doi.org/ 10.1016/j.jnucmat.2006.02.028. [22] X. Jia, Y. Dai, Microstructure in martensitic steels T91 and F82H after irradiation in SINQ Target-3, J. Nucl. Mater 318 (2003) 207e214, http://dx.doi.org/ 10.1016/S0022-3115(03)00101-6. [23] Z. Jiao, N. Ham, G.S. Was, Microstructure of helium-implanted and protonirradiated T91 ferritic/martensitic steel, J. Nucl. Mater 367e370 (2007) 440e445, http://dx.doi.org/10.1016/j.jnucmat.2007.03.118. [24] M. Li, M.a. Kirk, P.M. Baldo, D. Xu, B.D. Wirth, Study of defect evolution by TEM with in situ ion irradiation and coordinated modeling, Philos. Mag. 92 (2012)
316
X. Liu et al. / Journal of Nuclear Materials 490 (2017) 305e316
2048e2078, http://dx.doi.org/10.1080/14786435.2012.662601. [25] R.E. Stoller, M.B. Toloczko, G.S. Was, A.G. Certain, S. Dwaraknath, F.A. Garner, On the use of SRIM for computing radiation damage exposure, Nucl. Instrum. Methods Phys. Res. Sect. B Beam Interact. Mater. Atoms 310 (2013) 75e80, http://dx.doi.org/10.1016/j.nimb.2013.05.008. [26] D. Kaoumi, J. Adamson, M. Kirk, Microstructure evolution of two model ferritic/martensitic steels under in situ ion irradiation at low doses (0e2dpa), J. Nucl. Mater 445 (2014) 12e19, http://dx.doi.org/10.1016/ j.jnucmat.2013.10.047. [27] S. Jublot-Leclerc, X. Li, L. Legras, M.-L. Lescoat, F. Fortuna, A. Gentils, Microstructure of Au-ion irradiated 316L and FeNiCr austenitic stainless steels, J. Nucl. Mater 480 (2016) 436e446, http://dx.doi.org/10.1016/ j.jnucmat.2016.08.006. [28] W.-Y. Chen, Irradiation Damage in Neutron-irradiated Fe-cr Model Alloys, University of Illinois at Urbana-Champaign, 2014. ndez-Mayoral, C. Heintze, E. On ~ orbe, Transmission electron micro[29] M. Herna scopy investigation of the microstructure of FeeCr alloys induced by neutron and ion irradiation at 300 C, J. Nucl. Mater 474 (2016) 88e98, http:// dx.doi.org/10.1016/j.jnucmat.2016.03.002. [30] P. Hosemann, D. Kiener, Y. Wang, S.A. Maloy, Issues to consider using nano indentation on shallow ion beam irradiated materials, J. Nucl. Mater 425 (2012) 136e139, http://dx.doi.org/10.1016/j.jnucmat.2011.11.070. , R. Lebensohn, S. Maloy, Modeling and simulation of irradi[31] C. Deo, C. Tome ation hardening in structural ferritic steels for advanced nuclear reactors, J. Nucl. Mater 377 (2008) 136e140, http://dx.doi.org/10.1016/ j.jnucmat.2008.02.064. [32] W.-Y. Chen, Y. Miao, J. Gan, M.A. Okuniewski, S.A. Maloy, J.F. Stubbins, Neutron irradiation effects in Fe and Fe-Cr at 300 C, Acta Mater 111 (2016) 407e416, http://dx.doi.org/10.1016/j.actamat.2016.03.060. [33] K.G. Field, X. Hu, K.C. Littrell, Y. Yamamoto, L.L. Snead, Radiation tolerance of neutron-irradiated model FeeCreAl alloys, J. Nucl. Mater 465 (2015) 746e755, http://dx.doi.org/10.1016/j.jnucmat.2015.06.023. [34] F. Bergner, C. Pareige, M. Hern andez-Mayoral, L. Malerba, C. Heintze, Application of a three-feature dispersed-barrier hardening model to neutronirradiated FeeCr model alloys, J. Nucl. Mater 448 (2014) 96e102, http:// dx.doi.org/10.1016/j.jnucmat.2014.01.024. [35] J. Van den Bosch, O. Anderoglu, R. Dickerson, M. Hartl, P. Dickerson,
[36]
[37]
[38]
[39]
[40]
[41]
[42] [43]
[44] [45]
[46]
J.A. Aguiar, P. Hosemann, M.B. Toloczko, S.A. Maloy, SANS and TEM of ferriticemartensitic steel T91 irradiated in FFTF up to 184dpa at 413 C, J. Nucl. Mater 440 (2013) 91e97, http://dx.doi.org/10.1016/j.jnucmat.2013.04.025. G.R. Odette, M.J. Alinger, B.D. Wirth, Recent developments in irradiationresistant steels, Annu. Rev. Mater. Res. 38 (2008) 471e503, http:// dx.doi.org/10.1146/annurev.matsci.38.060407.130315. F. Garner, M. Toloczko, B. Sencer, Comparison of swelling and irradiation creep behavior of fcc-austenitic and bcc-ferritic/martensitic alloys at high neutron exposure, J. Nucl. Mater 276 (2000) 123e142, http://dx.doi.org/ 10.1016/S0022-3115(99)00225-1. G. Bonny, D. Terentyev, L. Malerba, On the aea0 miscibility gap of FeeCr alloys, Scr. Mater 59 (2008) 1193e1196, http://dx.doi.org/10.1016/ j.scriptamat.2008.08.008. O. Anderoglu, T.S. Byun, M. Toloczko, S.A. Maloy, Mechanical performance of ferritic martensitic steels for high dose applications in advanced nuclear reactors, Metall. Mater. Trans. A 44 (2013) 70e83, http://dx.doi.org/10.1007/ s11661-012-1565-y. R.L. Klueh, J.M. Vitek, Elevated-temperature tensile properties of irradiated 9 Cr-1 MoVNb steel, J. Nucl. Mater 132 (1985) 27e31, http://dx.doi.org/10.1016/ 0022-3115(85)90389-7. A. Alamo, J.L. Bertin, V.K. Shamardin, P. Wident, Mechanical properties of 9Cr martensitic steels and ODS-FeCr alloys after neutron irradiation at 325 C up to 42dpa, J. Nucl. Mater 367e370 (2007) 54e59, http://dx.doi.org/10.1016/ j.jnucmat.2007.03.166. L.K. Mansur, Correlation of neutron and heavy-ion damage, J. Nucl. Mater 78 (1978) 156e160, http://dx.doi.org/10.1016/0022-3115(78)90514-7. N.H. Packan, K. Farrell, J.O. Stiegler, Correlation of neutron and heavy-ion damage, J. Nucl. Mater 78 (1978) 143e155, http://dx.doi.org/10.1016/00223115(78)90513-5. M. Labbe, J.P. Poirier, Aux neutrons rapides, J. Nucl. Mater 46 (1973) 86e98. M. Labbe, J.P. Poirier, Letters to the editors - lettres aux redacteurs effet du flux instantane sur le gonflement du cuivre pur irradie aux neutrons rapides, J. Nucl. Mater 49 (1973) 232e234. F.A. Garner, L. Shao, M.B. Toloczko, S.A. Maloy, V.N. Voyevodin, Use of self-ion bombardment to study void swelling in advanced radiation- resistant alloys, in: 17th Int. Conf. Environ. Degrad. Mater. Nucl. Power Syst., Ontario, 2015.