Hardening and microstructural evolution of A533b steels irradiated with Fe ions and electrons

Journal of Nuclear Materials 471 (2016) 243e250

Contents lists available at ScienceDirect

Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat

Hardening and microstructural evolution of A533b steels irradiated with Fe ions and electrons H. Watanabe a, *, S. Arase b, T. Yamamoto c, P. Wells c, T. Onishi b, G.R. Odette c a

Research Institute for Applied Mechanics, Kyushu University, 6-1, Kasuga-kouenn, Kasugashi, Fukuoka, 816-8580, Japan Interdisciplinary Graduate School of Kyushu University, 6-1, Kasuga-kouenn, Kasugashi, Fukuoka, 816-8580, Japan c Dept. Chemical Engineering, UCSB Engineering II, RM3357, Santa Barbara, CA, 93106-5080, USA b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 18 May 2015 Received in revised form 29 December 2015 Accepted 30 December 2015 Available online 4 January 2016

Radiation hardening and embrittlement of A533B steels is heavily dependent on the Cu content. In this study, to investigate the effect of copper on the microstructural evolution of these materials, A533B steels with different Cu levels were irradiated with 2.4 MeV Fe ions and 1.0 MeV electrons. Ion irradiation was performed from room temperature (RT) to 350
C with doses up to 1 dpa. At RT and 290
C, low dose (<0.1 dpa) hardening trend corresponded with DH f (dpa)n, with n initially approximately 0.5 and consistent with a barrier hardening mechanism, but saturating at z0.1 dpa. At higher dose levels, the radiation-induced hardening exhibited a strong Cu content dependence at 290
C, but not at 350
C. Electron irradiation using high-voltage electron microscopy revealed the growth of interstitial-type dislocation loops and enrichment of Ni, Mn, and Si in the vicinities of pre-existing dislocations at doses for which the radiation-induced hardness due to ion irradiation was prominent. © 2016 Elsevier B.V. All rights reserved.

Keywords: Pressure vessel steels Radiation hardening HVEM Ion irradiation Dislocation loop

1. Introduction Radiation-induced embrittlement of reactor pressure vessel (RPV) steels used for the construction of thermal fission reactors is clearly of considerable importance to the safe operation of the reactors and plays a major role in plant life extension considerations. The neutron irradiation of these steels leads to an increase in the ductile-to-brittle transition temperature and a decrease in the upper shelf energy [1,2]. Cu has a strong effect on such embrittlement phenomena, and it has been proposed that the Cu-rich precipitates are responsible. On the other hand, studies on mechanical properties of steels with different Cu levels have shown that socalled matrix defects are dominant during the embrittlement of steels with both low [3] and high Cu fluences [4]. In addition, the Cu-rich precipitation mechanism occurs relatively rapidly compared to that due to matrix defects, which is a slow process with a rate that is typically proportional to the dose. Recently, a series of ion and neutron irradiation studies were conducted for RPV steels and model alloys containing various solutes, and the radiation-induced hardening of these alloys was evaluated [5e10].

* Corresponding author. E-mail address: [email protected] (H. Watanabe). http://dx.doi.org/10.1016/j.jnucmat.2015.12.045 0022-3115/© 2016 Elsevier B.V. All rights reserved.

However, studies of matrix defect formation in RPV steels have been very limited [11]. It is also known that the nucleation and growth processes for dislocation loops during neutron irradiation are strongly controlled by various factors, including the neutron flux, irradiation temperature, chemical composition, and applied stress [12,13]. On the other hand, the effects of dislocation loop formation on hardening are not clearly understood. In this study, therefore, to elucidate the effects of dislocation loop formation on radiation hardening in RPV steels, Fe ion irradiation was performed on steels with three different Cu contents. UCSB's Irradiation VARiable (IVAR) program database [14] and the same A533B steels used in the IVAR program [14] were investigated to compare the effects of the Cu content on radiation-induced hardening due to both the Fe ion and neutron irradiation. Previously, we demonstrated, using conventional transmission electron microscopy (TEM), that ion irradiation of A533B steels at doses of less than 0.5 dpa did not result in the formation of any visible defect clusters [7]. Therefore, in this study, scanning transmission electron microscopy (STEM) was used to analyze the irradiated regions and changes in the radiationinduced hardness resulting from low dose ion irradiation, and the results were compared with the microstructural changes observed during in situ electron irradiation using high-voltage electron microscopy (HVEM), that gives a guidance to the fundamental defect

244

H. Watanabe et al. / Journal of Nuclear Materials 471 (2016) 243e250

clustering processes resulting from irradiation.

2. Experimental procedures Three A533B steels with different Cu levels were used in this study. The A533B steels are referred to as A533B(LG) without Cu, A533B(LH) with a low Cu content (0.11 wt%), and A533B(LI) with a high Cu content (0.20 wt%). The results of the chemical analyses [14] of these steels are shown in Table 1. The specimens for ion irradiation and in situ observation via HVEM were annealed (austenitized) at 900
C for 1 h, air cooled, tempered at 664
C for 4 h, air cooled, stress relieved at 600
C for 40 h, and finally air cooled. Ion irradiation with 2.4 MeV Fe2þ was conducted from room temperature (RT) to 350
C using the tandem accelerator at Kyushu University. Fig. 1 (a) shows the SRIM calculation of damage distribution of Fe irradiated by 2.4 MeV Fe2þ ions. Nanoindentation tests were conducted at RT before and after ion irradiation using an Elionix ENT-1100 with a load of 1 gf. A triangular pyramidal diamond indenter (Berkovich type) with a semi-apex angle of 65
was used. The indenter load (L) and displacement (d) were continuously monitored using a computer system. L and d are given by the following equation [15]: L/d ¼ Ad þ B,

(1)

where A and B are dependent on the material but independent of the load and indenter displacement. A is proportional to the Vickers hardness (Hv) and is given as follows [15]: A(GPa) ¼ 0.287 Hv.

(2)

Fig. 1 (b) shows the typical L/ded curves of unirradiated and 1.0 dpa irradiated at 290
C of A533B(LI). In this study, A in the region of about 50e150 nm were compared to obtain the hardness of ion irradiated materials. For the LI (0.2 wt% Cu) sample, atom probe tomography (APT) was also conducted after ion irradiation to 1 dpa at 290
C and 350
C. The atom probe samples were run in a LEAP 3000x HR at the University of California, Santa Barbara in voltage mode using a pulse fraction of 20% at a specimen temperature of 50 K. Previous experiments by Hyde et al. revealed that these conditions prevent the preferential evaporation of Cu in typical RPV Steels [16]. Reconstructions and data analyses were performed using the CAMECA Integrated Visualization and Analysis Software (IVAS). The cluster analysis was performed using the density-based clustering algorithm (DBSCAN) outlined here [17]. DBSCAN is a modified version of the more well-known maximum separation method that measures the distance between solute atoms and their nearest solute neighbors and classifies them as clustered if this distance is less than some threshold distance, dmax. The DBSCAN method measures the distance between a solute and its Nth nearest neighbor, where N was set at five in this case. Furthermore, any clusters with less than Nmin atoms were excluded from the analysis. The precipitate number density was calculated by dividing the total number of clusters present in the sample by the total volume of the sample, which was given by V ¼ NatomsU/h, where U is the atomic volume of Fe and h is the efficiency of the instrument, 0.37% in this Table 1 Chemical compositions of the materials used in the present study (wt%). ID

Cu

Ni

Mn

Mo

P

C

Si

S

Fe

LG LH LI

0 0.11 0.20

0.74 0.74 0.74

1.37 1.39 1.37

0.55 0.55 0.55

0.005 0.005 0.005

0.16 0.16 0.16

0.22 0.24 0.24

<0.015 <0.015 <0.015

Balance Balance Balance

case. The total number of solute atoms (Cu, Ni, Mn, and Si) in each cluster was divided by the total number of atoms in the sample to determine the precipitate mole fraction. For samples with large clusters and significant solute depletion from the matrix, there typically was a range of dmax values that yielded a relatively constant number density [18]. In these samples with an extremely high density of very small clusters, a slight change in dmax resulted in a correspondingly large change in the precipitate number density and volume fraction. As a result, it was difficult to determine the “correct” selection parameters to be used for determining the actual volume fraction, precipitate size, and number density. Selection of a very large dmax resulted in the detection of clusters that would be present in any random solid solution. Styman et al. [18] reported that for a typical RPV steel, these “random” clusters can be avoided by using dmax  0.50 nm and Nmin  24 atoms, although the steel analyzed in their paper had a considerably higher solute (Cu, Ni, Mn, and Si) content (z4.25%) compared to the steel presented here (z2.5%). Lower alloy solute concentrations enable the selection of a smaller Nmin or larger dmax without including these random clusters. Consequently, values for dmax in the range of 0.50e0.60 nm and for Nmin in the range of 15e25 atoms were used. Note that a dmax value less than 0.50 nm left visible clusters present in the matrix. Thus, using the above ranges for dmax and Nmin, it was possible to determine the mean and uncertainty values for the precipitate mole fraction and number density. Electron irradiation with in situ observations was also performed using 1.0 MeV electrons and a high-voltage electron microscope (JEM-1000) in the HVEM Laboratory at Kyushu University. Electron irradiation was conducted at 290
C and 350
C. To reduce the temperature rise due to electron beam heating during irradiation, a relatively low electron dose rate of 2.5  104 dpa/s (the same as that for ion radiation) was selected. The TEM samples were prepared by electropolishing using 50 mL of perchloric acid and 950 mL of acetic acid as the electrolyte at RT and 15e30 V. After electron irradiation, solute enrichment around pre-existing dislocations due to irradiation was analyzed via STEM (Hitachi HD2700). 3. Results 3.1. Temperature and dose dependence of radiation-induced hardness due to ion irradiation The dose dependence of the radiation-induced hardness of alloys with different Cu levels after irradiation up to 1.0 dpa at RT, 290
C, 320
C, and at 350
C is shown in Fig. 2 (a)e(d), respectively. Here, radiation-induced hardness is defined by the difference in the hardness before and after irradiation, DH ¼ Hirrad  Hunirrad. In this study, indentation measurements were performed before and after irradiation using the same samples. At all irradiation temperatures, radiation-induced hardening occurred at the beginning of ion irradiation, became saturated at approximately 0.1 dpa, and then gradually increased to 1.0 dpa with the dose. The irradiation temperature dependence of hardening for doses from 0.1 to 1.0 dpa is shown in Fig. 3 (a) through (d), respectively. The dashed line indicates the temperature dependence proposed by Bolton et al. [19] for matrix defects normalized to the observed RT hardening. In this previous study, a qualitatively similar general trend was observed at 0.1 dpa, with the hardness decreasing as the irradiation temperature increased, and the level of radiation-induced hardness was nearly the same with or without added Cu at all temperatures. However, the temperature dependence of the hardness observed in the present study was slightly less than that reported by Bolton et al. In addition, at 1.0 dpa in the present study, greater irradiation

H. Watanabe et al. / Journal of Nuclear Materials 471 (2016) 243e250

245

Fig. 1. (a) SRIM calculation of damage distribution of Fe irradiated by 2.4 MeV Fe ions. (b) Typical L/ded curves of unirradiated and 1.0 dpa irradiated at 290
C of A533B(LI).

Fig. 2. Radiation-induced hardness due to Fe2þ irradiation at (a) RT, (b) 290
C, (c) 320
C, and (d) 350
C.

hardening was observed for LI (0.2 wt% Cu) at 290
C, but the hardening decreased as the irradiation temperature increased. In addition, above 320
C, the dependence of radiation hardening on the Cu content was not prominent, and at 350
C, the hardening level was lower and similar to that of the matrix hardening temperature dependence. The trends in intermediate doses fall between thee two cases. Hardening due to neutron irradiation is known to be recovered at approximately 450
C [20,21]. Therefore, these results indicate that the nucleation of dislocation loops or solute clusters that contributed to radiation-induced hardening at 290
C were not prominent during irradiation above 320
C.

3.2. Microstructure due to ion irradiation Fig. 4 shows the dose dependence of TEM microstructures in the LG (without Cu) and LI (0.2 wt% Cu) samples irradiated at 290
C to a dose of 1.0 dpa. Previously, we showed using conventional TEM that ion irradiation of A533B steels at doses less than 0.5 dpa did not result in the formation of any visible defect clusters [7]. However, as observed from Fig. 4, while no small dislocation loops were visible at 0.2 dpa, they were present at 0.5 dpa and higher dose levels (indicated by the arrows in the figure). Notably, the Cu content in the steel did not significantly influence the number density or the size of dislocation loops; the number density and the average

246

H. Watanabe et al. / Journal of Nuclear Materials 471 (2016) 243e250

Fig. 3. Irradiation temperature dependence of the radiation-induced hardness for different doses at 290
C: (a) 0.1 dpa (b) 0.3 dpa (c) 0.5 dpa (d) 1.0 dpa.

selection procedure. Because the two steels contain the same solute concentration, it is inappropriate to compare results from the two irradiations using different combinations of input parameters, and thus the uncertainties should be seen as systematic within the measurement. While the lower bound for the mole fraction at 290
C was less than the upper bound at 350
C, these results should not be interpreted to mean that there is not a statistically significant difference between the two samples, because at any given dmax, the mole fraction was larger at the lower temperature irradiation. Fig. 6 shows the temperature dependence of the chemical composition along with the mole fraction (fp) for A533B steel clusters. At higher temperature, reduction of the Ni, Mn, and Si concentration was more significant than that of Cu. 3.3. Characterization of the defect clusters using HVEM

Fig. 4. Dislocation loop formation following ion irradiation to a dose of 1 dpa at 290
C. Upper and lower photos are the LG (without Cu) and LI (0.2 wt% Cu) samples, respectively.

size of the dislocation loops in the LG (without Cu) and LI (0.2 wt% Cu) samples at 1.0 dpa were 3.3  1022 m3 and 3.2  1022 m3 and 2.1 nm and 2.0 nm, respectively. Fig. 5 (a) and (b) show the APT analyses of the LI (0.2 wt% Cu) sample irradiated at 290
C and 350
C, respectively, with an irradiation dose of 1.0 dpa. Table 2 lists the estimated N, , and f ranges for the cluster analysis parameters dmax (0.50e0.60 nm) and Nmin (15e25 atoms). The precipitate number density in the LI sample irradiated at 290
C was approximately three times greater than the sample irradiated at 350
C. On an average, a given cluster contained the same number of solute atoms, but the higher number density at lower temperatures resulted in a larger mole fraction of 0.36% vs. 0.14% at higher temperature irradiation. Note that the uncertainty in the measurements was determined by varying dmax and Nmin during the cluster

Fig. 7 shows the microstructural evolution of the LI (0.2 wt% Cu) sample during 1.0 MeV electron irradiation at 290
C (upper images) and 350
C (lower images). The irradiation time is shown in the left-hand corner of each photo. Fig. 8 shows the irradiation time dependence of the dislocation loop density at 290
C and 350
C. During irradiation at 290
C, dislocation loops were observed in situ to nucleate in the early stage (455 s, z0.1 dpa) of electron irradiation, and the number of dislocation loops was saturated after 1350 s (z0.4 dpa). These dislocation loops, formed due to irradiation, were identified as interstitial-type dislocation loops using additional electron irradiation at RT [9]. In addition, by increasing the irradiation temperature, saturation of the number of loops occurred at an even earlier stage of electron irradiation (750 s), and growth of the loops became prominent. Furthermore, while interstitial dislocation loops were homogenously formed in the matrix, the enhanced growth of the dislocation loops near pre-existing dislocations was prominent. The total number density of the loop at 290
C was 1.5  1020 (m3) for a low dose of z0.1 dpa, but began to increase at z0.2 dpa and reached a saturation level of 2.0  1022 (m3) at z0.4 dpa. Fig. 9 shows the microstructure and solute atom maps for the LI

H. Watanabe et al. / Journal of Nuclear Materials 471 (2016) 243e250

247

Fig. 5. Atom maps for the LI (0.2 wt% Cu) sample irradiated with ions to 1.0 dpa at (a) 290
C and (b) 350
C.

Table 2 Estimated precipitate number density (N), volume fraction (f), and size () determined using APT.




290 C 350
C

N(1023 m3)

F (%)

(nm)

24.0 ± 13.6 8.4 ± 4.1

0.36 ± 0.19 0.14 ± 0.07

0.71 ± 0.04 0.73 ± 0.74

Fig. 8. Irradiation time dependence of the dislocation loop density at 290
C and 350
C in the LI (0.2 wt% Cu) sample during 1.0 MeV electron irradiation.

Fig. 6. Irradiation temperature dependence of the volume fraction (fp) of precipitates for A533B steel clusters.

(0.2 wt% Cu) sample irradiated for 1000 s (0.25 dpa) at 290
C. The STEM-bright field (BF) image (Fig. 8 (a)) shows the pre-existing dislocations and dislocation loops formed as the result of electron irradiation that were not obvious in the STEM-dark field (DF) image. In addition, strong enrichment of the Mn, Si, and Ni atoms was detected at the dislocations. On the other hand, at the lower dose, prominent Cu and MneNieSi clusters were not observed following electron irradiation.

Fig. 7. Microstructural evolution of the LI (0.2 wt% Cu) sample during 1.0 MeV electron irradiation at 290
C (upper images) and 350
C (lower images). The number density of interstitial-type dislocation loops at 290
C was greater than that at 350
C.

248

H. Watanabe et al. / Journal of Nuclear Materials 471 (2016) 243e250

Fig. 9. Microstructure (STEM-BF and -DF images) and solute atom maps for the LI (0.2 wt% Cu) sample irradiated with electrons for 1000 s at 290
C.

4. Discussion 4.1. Radiation-induced hardening due to ion irradiation 4.1.1. Low-dose trend common to all three alloys To understand the irradiation dose dependence of the irradiation-induced hardening observed in the present study, logelog plots of the irradiation-induced hardening at RT and 290
C as a function of the displacement dose level (dpa) were prepared and are shown in Fig. 10 (a) and (b), respectively. The dose dependence of the samples was compared using the simple power law expression developed by Byun and Farrell [22], DsYS f (dpa)n, where DsYS and dpa are the radiation-induced increase in the yield stress and the irradiation dose, respectively. They also showed that the mean values for n obtained from many metals, including RPV steels, are approximately 0.5 for the low-dose regime. using the relationship DH z (1/3) DsYS [23], the radiation-induced hardness obtained in this study (DH) was controlled by the irradiation dose (dpa), and DH f (dpa)n was obtained. As shown in Fig. 10 (c), DH had no systematic or significant dependence on the Cu content at doses less than z0.1 dpa, while a slight dependence was noticeable above 0.3 dpa. Hence, for a better statistical analysis, the exponent was evaluated for each temperature using all of the alloy data at doses up to 0.05 or up to 0.1 dpa. Hardness increase due to ion

irradiation was small, especially by low dose irradiation up to 0.1 dpa, but as can be seen in Fig. 10, the values for the exponent n, which represent the slopes of the logelog plots, were greater than 0.5 for the data up to 0.05 dpa, but z0.5 at RT and z0.4 at 290
C for the data up to 0.1 dpa. All of these fittings cover all of the data in the corresponding ranges within the scatter, indicating that the overall n value of z0.5 is consistent with the results reported by Byun and Farrell [22], i.e., the mean value for n was approximately 0.5, which is the same as the theoretical value obtained for barrier hardening models, which assumes that production of stable defect clusters with the same size and strength factors is proportional to the dose. The hardening trends at 290
C were then compared to the corresponding trends observed for the same steels after neutron irradiation at 290
C, but at much lower dose rates from 1010 to 109 dpa/s (contained in the large IVAR tensile database [14]). Fig. 11 shows the logelog plot of the increase in yield stress as a function of dpa taken from the IVAR database along with the plot for the data obtained at 290
C in the present study. As can be seen in the figure, the values for the exponent n at the start of irradiation were z0.5 at all Cu levels, but lower n values were observed for the steels with higher Cu concentrations, i.e., the n value decreased as the Cu content increased. It is well-known that radiation-induced hardening of Cu-containing A533B steels during neutron

Fig. 10. Dose dependence of ion irradiation-induced hardness at (a) RT and (b) 290
C, (c) temperature and Cu cross plots of the hardening at each dose level.

H. Watanabe et al. / Journal of Nuclear Materials 471 (2016) 243e250

249

hardening model. The contribution of each individual loop or precipitate to the yield stress (sj), apart from all other contributions, is given by the relationship:

sj ¼ ajMmb(Njdj)1/2/3 (j ¼ l:Loop or p:CRP).

Fig. 11. Dose dependence of irradiation-induced hardening in UCSB's IVAR experiments for 3  1011 n/cm2/s neutron flux conditions compared to that due to ion irradiation observed in the present study.

irradiation is mainly due to the formation of Cu clusters. This behavior indeed led to a large and systematic increase in hardening with Cu content, as shown in Fig. 11. However, as can be seen in Figs. 10 and 11, in the present study, the hardening of the ionirradiated samples did not depend on the Cu content in the lowdose region (<0.05 dpa) and over a wide range of irradiation temperatures. In addition, as mentioned above, no loops were detected over the dose range either. Then, what causes the significant hardening during the initial stages of ion irradiation? Given that this initial hardening is produced within a very short time scale (z100 s), the hardening features formed directly in the displacement cascades are the most promising suspects. It is known that socalled unstable matrix damage (UMD) contributes to significant hardening during high dose rate test reactor irradiation [4,20]. UMD occurs in aged displacement cascades as solute-vacancy cluster complex remnants that thermally anneal during irradiation [4,20,26]. Inecascade formation and thermal annealing of vacancy clusters and their complex with solute are also found in MD simulations [27,28], while a rate-theory model shows that they can build-up and grow to one of the major hardening features at low temperatures such as 100
C that prohibit the thermal annealing [29]. Even at 290
C. UMD builds up to a higher steady-state concentration at higher dose rates [4,20,26]. In the case of ion irradiation, there is no time for such displacement cascade remnants to thermally anneal; they saturate instead due to volume exclusion effects, as first proposed by Makin and Minter [30], resulting in a high steady-state concentration of defects. Such behavior can explain the initial hardening trend observed during ion irradiation. A more detailed model will be found in our future publications. Cu effects on this enhanced hardening are only seen at very high doses of z0.1 dpa or higher, where the higher Cu concentration systematically causes greater hardening. Fujii et al. also showed that the addition of Cu barely influenced the radiation-induced hardness in FeeMneNi model alloys during 3.0 MeV Fe2þ ion irradiation up to 0.5 dpa at 290
C [10]. 4.1.2. High irradiation dose trend Under high dose conditions, dislocation loops and Cu clusters (CRPs) were observed via TEM and atom probe tomography. Contributions by these features are thought to lead to the additional hardening at the top of the fitted model in Fig. 11, which presents the base hardening for all of the cases discussed above. This additional hardening was then compared to the estimated hardening of the observed dislocation loops and CRPs using the barrier

(3)

Here, aj is the hardening efficiency, which depends on the nature of the defect and is often assigned a value of 0.4 for small dislocation loops [24], while for CRPs, the size dependent strength factor ap(rp) ranging from 0.1 to 0.2 was taken from a RussellBrown (RB)-type model [25] empirically modified by fitting the IVAR Dsy values and data from CRP databases [31]. The parameters Nj and dj are the number densities and diameters of the loops or CRPs, respectively. Dislocation loops in bcc structures undergo contrast change depending on the Burgers vector (b) and the diffraction vector (g). Dislocation loops are invisible when the value of g is 200 for gb ¼ 0. For steels, N is the measured value multiplied by 1.5. The Taylor factor (M ¼ 3.06), the shear modulus (m ¼ 80700 MPa), and the Burgers vector (b ¼ 0.25  109 m) were used. The obtained isolated hardening contributions were then combined with the unirradiated dispersed barrier strengthening contributions (au), such as those from Mo2C precipitates, using the ap(d) dependent strength superposition model [32]. Here au was set at 200 MPa. Further details are described elsewhere [32]. The estimated radiation-induced hardness (DHv) due to ion irradiation at 290
C and a dose of 1.0 dpa for the LG (without Cu) and LI (0.2 wt% Cu) samples were 110 and 152, respectively. The corresponding measured DHv values were 95 ± 19 and 176 ± 33, respectively, and were in reasonable agreement with the estimated values. These results suggest that radiation-induced hardness at the highest dose level (z1.0 dpa) for both the LG and LI (without Cu) samples can be explained by a combination of cascade induced defects, dislocation loops, and CRP formation (for LI only). Nevertheless, further fundamental studies of defect formation during the early stages of irradiation (up to 0.1 dpa) are essential to confirm these hardening models. 4.2. Effects of dislocation loops and dislocations on radiationinduced hardness As can be seen in Fig. 9, the Mn, Ni, and Si atoms were enriched at dislocations (1000 s, 0.25 dpa), and the level of enrichment corresponded to the dose prior to dislocation loop saturation. These results suggest that the Mn, Ni, and Si atoms were combined with interstitials, and that pre-existing dislocations acted as strong defect sinks for these interstitialesolute complexes. It is known that Mn-, Ni-, and Si-rich phases form in a number of austenitic stainless steels and RPV steels during irradiation. In austenitic steels, a number of G-phases are known to exist that are enriched in face-centered cubic Ni and Si [33,34]. Such G phases and an Ni3Si phase have been shown to nucleate only at defect sinks for interstitials during irradiation. The sub-sized Ni and Si atoms combine with the interstitials to form mixed-dumbbells that migrate to defect sinks, including grain boundaries and dislocations. In RPV steels, on the other hand, G phases or MneNieSi lateblooming phases have also been observed to form as a result of neutron irradiation [35]. In Fe-based model alloys, the author demonstrated that the addition of Mn is effective for increasing the interstitial-type dislocation loop density at 290
C, and a binding energy of 0.22 eV was obtained for the interaction of Mn atoms and interstitials [9]. The formation of Mn clusters detected via APT and interstitial-type loops at RT clearly indicated that the oversized Mn atoms migrated via an interstitial mechanism [9]. In the present study, additional diffraction spots attributed to phasetransformations or new-phase formation were not detected.

250

H. Watanabe et al. / Journal of Nuclear Materials 471 (2016) 243e250

During electron irradiation using HVEM, the specimen surface also acted as a strong defect sink for interstitials, likely due to the failure of G-phase formation at dislocations. 5. Conclusions Ion irradiation with 2.4 MeV Fe2þ and electron irradiation using HVEM were performed on A533B steels with different Cu contents. The main results are summarized as follows: 1) The effect of the Cu content on the irradiation-induced hardening of the steels was prominent only at higher dose levels (1.0 dpa) at 290
C. At lower dose levels (z0.1 dpa) and temperatures above 320
C, nearly the same extent of change in the hardness was observed for all of the A533B steels, regardless of the Cu content. 2) Logelog plots of irradiation hardening at 290
C, fitted to the dose dependence of the irradiation hardening using DH f (dpa)n, revealed an overall value for n of z0.5, which was consistent with a dislocation barrier hardening mechanism, for doses of 0e0.05 dpa, followed by saturation behavior. The initial hardening trend, which exhibited no or minimal dependence on the Cu concentration or temperature, is consistent with hardening due to the defects that form in aged cascades but quickly become spatially saturated at very high dose rates. 3) Analysis of the microstructure resulting from electron irradiation revealed that the growth of the saturated number density of dislocation loops was prominent at high dose levels for which ion irradiation led to loop and CRP formation. At low doses, on the other hand, when hardening induced by ion irradiation rapidly increased and reached the first saturation, dislocation loops were still nucleating and growing, mostly near dislocations. Acknowledgments

[4]

[5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]

[20]

[21]

[22] [23] [24] [25] [26] [27] [28]

This study was supported by Japan Society for the Promotion of Science (JSPS) KAKENHI Grant Numbers 23360418, 26630487, and 26289362. This study was also supported partly by the Kansai Electric Power Company and the Collaborative Research Program of the Research Institute for Applied Mechanics, Kyushu University.

[30] [31]

References

[32]

[1] G.R. Odette, Scr. Met. 11 (1983) 1183. [2] G.E. Lucas, G.R. Odette, P.M. Lombrozo, J.W. Sheckherd, in: Effects of Radiation on Materials: 12th International Symposium, ASTM STP 870, American Society for Testing Materials, Philadelphia, 1985, p. p900. [3] M. Suzuki, K. Onizawa, M. Kizaki, in: Effects of Radiation on Materials: 17th

[29]

[33] [34] [35]

International Symposium, ASTM STP 1270, American Society for Testing Materials, Philadelphia, 1996, p. p351. G.R. Odette, E.V. Mader, G.E. Lucas, W.J. Phythian, C.A. English, in: Effects of Radiation on Materials: 16th International Symposium, ASTM STP 1275, American Society for Testing Materials, Philadelphia, 1993, p. p373. W.J. Phythian, C.A. English, J. Nucl. Mater. 205 (1993) 172. K. Fuji, K. Fukuya, J. Nucl. Mater. 336 (2005) 323. H. Watanabe, S. Masaki, A. Masubuchi, N. Yoshida, Y. Kamada, J. Nucl. Mater. 417 (2011) 932. K. Yabuuchi, H. Yano, R. Kasada, H. Kishimoto, A. Kimura, J. Nucl. Mater. 417 (2011) 988. H. Watanabe, S. Masaki, A. Masubuchi, N. Yoshida, K. Dohi, J. Nucl. Mater. 439 (2013) 268. K. Fujii, T. Ohkubo, K. Fukuya, J. Nucl. Mater. 417 (2011) 949. T. Nishi, N. Hashimoto, S. Ohnuki, T. Yamamoto, G.R. Odette, J. Nucl. Mater. 417 (2011) 936. K. Fujii, K. Fukuya, R. Kasada, A. Kimura, T. Ohkubo, J. Nucl. Mater. 407 (2010) 151. H. Watanabe, A. Hiragane, S. Shin, N. Yoshida, Y. Kamada, J. Nucl. Mater. 442 (2013) 776. G.R. Odette, T. Yamamoto, D. Klingensmith, D. Gragg, G.E. Lucas, UCSB MRPG PV1e2009, Santa Barbara, 2009. T. Suzuki, M. Inamura, J. Ohta, A. Kohyama, Seisan Kenkyu 42 (1990) 287 (Japanese). J.M. Hyde, M.G. Burke, B. Gault, D.Wf Saxey, P. Styman, K.B. Wilford, T.J. Williams, Ultramicroscopy 111 (2011) 676e682. L.T. Stephenson, P.M. Moody, P.V. Liddicoat, S.P. Ringer, Microsc. Microanal. 13 (2007) 448e463. P.D. Styman, J.M. Hyde, K. Wilford, G.D. Smith, Ultramicroscopy 132 (2013) 258e264. C.J. Bolton, J.T. Buswell, R.B. Jones, R. Mskovic, R.H. Priest, “The Modeling of Irradiation Embrittlement in Submerged Arc Welds,” Effects of Radiation on Materials, in: 17th International Symposium, ASTM STP 1405, American Society for Testing and Materials, West Conshohocken, PA, 2001, p. 103. D. S. Gelles et al. E. Mader, G.E. Lucas, G.R. Odette, in: Effects of Radiation on Materials: 15th International Symposium, ASTM STP 1125, American Society for Testing Materials, Philadelphia, 1992, p. p151. R.K. Nastand, P.h. Tipping, R.D. Kalkhof, M.A. Sokolov, in: Effects of Radiation on Materials: 21th International Symposium, ASTM STP 1447, American Society for Testing Materials, Philadelphia, 2004, p. p149. T.S. Byun, K. Farrell, J. Nucl. Mater. 326 (2004) 86. J.R. Cahoon, W.H. Broughton, A.R. Kutzak, Metal. Trans. 2 (1971) 1979. G.E. Lucas, J. Nucl. Mater. 206 (1993) 287. K.E. Russell, L.M. Brown, Acta Metall. 20 (1972) 969. G.R. Odette, T. Yamamoto, D. Klingensmith, Philos. Mag. 85 (2005) 779. R.E. Stoller, J. Nucl. Mater. 276 (2000) 22. B.D. Wirth, G.R. Odette, R.E. Stoller, in: V. Bulatov, et al. (Eds.), Multiscale Modeling of Materials, MRS Soc. Symp. Proc, 677, Materials Research Society, Warrendale, PA, 2001. AA5.2. X. Hu, D. Xu, T.S. Byun, B.D. Wirth, Modelling and Simulation in Materials Science and Engineering Modelling Simul, Mater. Sci. Eng. 22 (2014), 065002. M.J. Making, F.J. Minter, Acta Mettallugica 8 (1960) 691. P.B. Wells, T. Yamamoto, B. Miller, T. Milot, J. Cole, Y. Wu, G.R. Odette, Acta Mater. 80 (2014) 205e219. E.D. Eason, G.R. Odette, R.K. Nanstad, T. Yamamoto, J. Nucl. Mater. 433 (2013) 240. H.J. Beattie, F.L. Versnyder, Nature 178 (1956) 208. A. Mateo, L. Llans, M. Anglada, A. Redjaimia, G. Metauer, J. Mater. Sci. 32 (1997) 4533. G.R. Odette, R.K. Nanstad, JOM 61 (7) (2009) 17.