Microstructural examination of neutron-irradiated simple ferritic alloys

Journal of Nuclear Materials 108 & 109 (1982) 515-526 North-Holland Publishing Company

51s

MICROSTRUCITJRAL EXAMINATION OF NEUTRON-IRR4DIA’lTD ALLOYS

SIMPLE FJWRITIC

D.S. GELLES Hanford Engineering Deoelopment Laboratory Richlanrl Washingtrm,99352, USA Received 8 January 1982; accepted

1 February 1982

A series of simple (Fe-Cr-C-MO) ferritic alloy specimens based on the composition of martensitic stainless steels has been examined by transmission electron microscopy following neutron vacation to a maximum fhtenoe of 4.3X 10” n/cm2 (E ~0.1 MeV) over the temperature range 400 to 45O’C. The series includes binary iron-chromium alloys covering the range Fe-3Cr to Fe- 18Cr, ternary alloys of Fe- 12Cr-0.002 to 0.2C and quatematy alloys of Fe- 12Cr with various levels of carbon and molybdenum additions. Swelling and dislocation densities were highest in the Fe-9Cr alloy, where the swelling rate at 425°C was estimated at 0.15W/1022 n/cm2. The dislocation structures in the alloy series included dislocations with a(lO0) and a/2 (11 l} Burgers vectors; the former predominating in the lower chromium alloys. Additions of carbon were found to promote formation of radiation resistant martensite and additions of molybdenum enhanced precipitation. However, in both cases swelling was not completely suppressed. The swelling resistance of this altoy class is attributed to the presence of the irradiation-induced a{ 100) dislocations.

1. Introduction

Martensitic stainless steels appear to provide attractive alternatives to austenitic stainless steels for in-core fusion and fission reactor applications [I]. This is in part a consequence of generally greater resistance to neutron irradiation-induced swelling [2-51 and acceptable high temperature strength [6-81. Furthermore, it appears that ferrite itself is inherently low in swelling 191.Several attempts have been made to explain the low swelling behavior of ferritic steels. Mechanisms based on trapping [ lo,1 l] and the presence of dislocations of two different Burgers vector types [ 12,131 have been proposed. However, there is limited data available to differentiate between these models, and the martensitic alloy class, due to its extreme ~~rost~~tur~ and mocrochemical complexity, provides a host of other possible explanations. It is therefore the purpose of the present effort to provide a data base on this alloy class which may be used for further insight into the swelling resistance of martensitic stainless steels. A series of thirteen ferritic alloys has been prepared based on the martensitic stainless steel base composition (wt%) of Fe- 12Cr- 1MO-O. 1C. The series included Fe-Cr binary alloys covering the range 3 to 18Cr at intervals of 3Cr, Fe-Cr-C alloys based on Fe-12Cr with three levels of carbon from 0.002 to 0.2C and Fe-Cr-C-MO alloys with variations in molybdenum 0022-3115/82,‘~--0000/$02.75

wntent up to 3Mo. A heat treatment selected to produce a tempered martensite structure was used for each alloy condition in order to emphasize behavior anticipated in martensitic stainless steels. This alloy series therefore provides an ideal opportunity to investigate the effects of fast neutron irradiation on the martensitic stainless alloy class.

2. Experimental pmcedure One hundred pound heats of the thirteen alloys listed in table 1 were melted at the Paul D. Merica Research Laboratory, International Nickel Company, Sterling Forest, New York. Compositions provided by the vendor are included in table 1. Chemical overchecks performed by Lukens Steel Company, Coatesville, Pennsylvania were found to be in reasonable agreement with the vendor specification and also indicated impurity levels in each alloy on the order of 0.1 Ni, 0.02 Mn, 0.01 0,, 0.05 Ns and 0.005 P. The alloys were provided in the form of 1.0 cm diameter extruded bar. The material was rolled to 0.03 cm thick sheet and punched to provide 0.3 cm diameter disk specimens. Specimens were engraved with a four digit identification code and then heat treated as follows: 1040°C,‘l h/AC + 760°C/2 h/AC.

0 1982 North-Holland

D.S. Gelles / Neutron-irradiated simple ferritic alloys

516 Table I Compositions Alloy code

ES9 MO E61 E62 El7 E64 E95 E96 E71 E72 E99 E73 E74

(in weight precent) of the simple ferritic alloy series as provided by the vendor Alloy composition

Fe-3Cr Fe-6Cr Fe-9Cr Fe- 12Cr Fe- 15Cr Fe- 18Cr Fe-12Cr-O.lC Fe-12Cr-0.2C Fe- 12Cr-0. IOC-OSOMo Fe- 1ZCr- 1.OMo Fe-12Cr-O.IOC-l.OMo Fe- IZCr-0. IOC-2.OMo Fe- 12Cr-0. IOC-3.0Mo

Cr

C

3.3 6.0 9.6 11.6 15.1 18.3 12.2 12.1 12.1 12.0 12.2 12.4 12.0

0.004 0.007 0.002 0.002 0.001 0.002 0.089 0.18 0.077 0.002 0.091 0.073 0.095

The Fe- 12Cr-O.lC alloy was also prepared in the untempered martensitic condition ( 1040°C/1 h/AC). A very limited number of specimens of the series of simple ferritic alloys was irradiated in the Experimental Breeder Reactor II (EBR-II). Each subcapsule of the experiment was designed to operate at the required irradiation temperature by gas gap regulation of the gamma heat loss and in each subcapsule all specimens were immersed in static sodium and therefore were at the same temperature. Specimens were irradiated in Row 2 of EBR-II for 7 cycles to fluences on the order of 4 X lo** n/cm* (all fluences reported are E > 0.1 MeV) or 20 dpa. Specimens of the binary Fe-3, 6, 9 and 12Cr alloys were obtained following irradiation at 400°C to 3.4 X lo** n/cm*, at 425°C to 4.3 X lo** n/cm* and at 450°C to 2.8 X lo** n/cm*. The rest of the alloys were obtained only following irradiation at 45O’C to 2.8 X lo** n/cm*. The decision to examine only these three low irradiation temperatures was based primarily on the work of Little and Stow [9] who found that for ironchromium binary alloys in the range 0 to 15% Cr peak void swelling occurred at 425°C. Previous program results which showed that irradiation effects were only apparent below 500°C [14] added further support for this selection. Specimens suitable for transmission electron microscopy were prepared using a Metalthin twin jet ele-ctropolishing device with a solution of 5% perchloric acid in 95% butyl alcohol and operating at 9OV, 250 to 300 mA, and a moderate pump speed. In most cases only

Vendor analysis MO -

02

NZ

0.021 0.022 0.026

0.0014 0.0017 0.0017

0.026

0.0022

0.50 1.06 1.03 2.14 3.16

one disk of each alloy/irradiation history condition was available. All but three specimens (Fe-3Cr at 4OO”C, Fe-9Cr at 400°C and Fe-12Cr-1Mo at 45O’C) were successfully prepared, a measure of the high reliability of the procedure. Of further note tias the finding that the Fe-3Cr alloy was particularly sensitive to surface film contamination. This was removed by subsequent flash electropolishing. Microscopy was performed on a JEM-200A electron microscope equipped with a side-entry double-tilting goniometer stage and operating at 200 keV. Beam deflection effects due to magnetic interactions were overcome by use of the dark-field deflection coils and careful objective stigmator adjustment.

3. Results This section will be subdivided into four parts as follows: (1) the microstructures of the alloys prior to irradiation, (2) the microstructures of irradiated ironchromium binary alloys, (3) the microstructures of irradiated iron-chromium-carbon alloys, and (4) the microstructures of irradiated alloys containing iron, chromium, carbon and molybdenum. 3.1. Preirradiation microstructures The microstructures of the fourteen conditions under investigation were extremely varied. Annealed ferrite, recovered ferrite, martensite, tempered martensite and

D. S. Gelles / Neutron-irradiaredsimplpferriric duplex structures were represented

in one specimen or another. Annealed and recovered ferrite may be differentiated by the fact that the latter underwent an austenite to ferrite transformation followed by a temper which resulted in polygonization of the dislocation structure. Annealed ferrite was observed in Fe-ISCr, Fe- 18Cr and Fe- 12Cr- 1Mo. Recovered ferrite was found in Fe-3, 6, 9 and 120. Mar&site or tempered martensite was present in Fe- 12Cr-O. lC and Fe12Cr-0.2C while a duplex structure of tempered martensite and ferrite occurred in the remaining alloys, Fe-12Cr-O.lC-0.5, 1, 2 and 3Mo. Examples of these microstructures at iow magnification are provided in fig. 1. It may be noted that ferrite structures contain extremely low dislocation densities except as subgrain boundaries prior to irradiation. Examples of the’s&grain boundaries in fig 1 were especially selected’ to emphasize that an austenitato-ferrite transformation had taken place. 3.2. Irradiated Fe-Cr alloys

Effects of neutron irradiation were observed in all specimens of the iron-chromium binary alloys examined. Except for two specimens irradiated at 45OT (Fe- 3Cr and Fe- 18Cr), void swelling was observed, and in all cases the dislocation structure had been altered and evidence of precipitate development was found. For the most part void swelling consisted of slightly truncated octahedral voids .tith (110) edges and { 111) faces. The exceptions occurred when voids were coupled to large precipitate particles where unique geometries were invariably produced. As a function of increasing temperature, void density decreased and void size increased as expected. Because void size distributions, in most cases, extended to small sizes, steady states void populations had at best only just been achieved. Grain boundaries generally developed void-free zones. In many cases. voids were in linear arrays, indicating that heterogeneous nucleation on dislocations played a role. Examples are provided in fig. 2. It should be noted that figs. 2(a) and 2(b) are at a magnification approximately twice that of the other micrographs of fig. 2. Examples of the truncated octahedral morphology and linear arrays of voids can be clearly identified. The grain boundary shown in fig. 2(g) provides a typical example of void denuding near grain boundaries. Void swelling measurements are tabulated in table 2. The highest void swelling occurred in specimens of Fe-6Cr. Fe-90 and Fe-12Cr following irradiation at 425°C to 4.3 X 10s2 n/cm2 (E~0.1 MeV) wherein 0.6 to 0.7% swelling was found. If one assumes no incuba-

alloys

517

tion, this corresponds to a swelling rate of approximately O.lS%/lO” n/cm2. In colnparison, other conditions of chromium content and/or irradiation temperature produced lower swelling and lower swelling rates. Void density comparisons indicate that saturation densities are on the order of 8 X 10” cmS3 at 4CWC, 3 X 10” cm-’ at 425*C, and 1 X 10” cmm3 at 450°C. In cases where significant void swelling was measured, mean void sizes followed the expected trend, and higher irradiation temperatures produced larger voids. The evolution of the dislocation structure in Fe-0 alloys as a result of neutron irradiation varied considerably as a function of composition and teat temperature. Some specimens contained primarily dislocation loop arrays, some contained only dislocation networks, and others contained mixtures of the two. The dislocation Burgers vectors qresentcd were found to be of two types: a/2 (111) and u(loO), with both types often present in the same specimen. Dislocation structure often varied from region to region in a given specimen as a function of distance from grain boundaries, subgrain boundarits, or large precipitate particles. The dislocation Ltructnre for each specimen examined is de scribed briefly in table2. Examples of the various types of dislocation networks are provided in fig. 3. Figs. 3(a) and 3(d) allow comparison of the dislocation structure of Fe-3Cr and Fe-6Cr after irradiation at 425°C to a fhrencc of 4.3 x 10” n/ems (E > 0.1 MeV). The dislocation density increases with increasing chromium content. However, at still higher chromium contknts the dislocation density decreases. Figs. 3(c), 3.(f) and 3(h) allow comparison of Fe- 12Cr, Fe- 15Cr and Fe- 180 irradiated at the higher temperature of 4SOT. The structure of the Fe-12Cr consists of a dislocation network, whereas for Fe-15Cr and Fe-18Cr it consists predominately of loops and the dislocation density varies inversely with distance from a grain boundary. Many regions of the Fe-15Cr and Fe-18Cr specimens were effectively dislocation free. Figs. 3(b) and 3(c) allow comparison under different imaging conditions of the same area in a specimen of Fe-6Cr irradiated at 4OOT to 3.4 X lo= n/cm2 (E> 0.1 MeV). For [OlT] contrast, as in fig 3(b), loops are aligned in [Oli] directions whereas for [200] contrast, as in fig. 3(c), a different set of loops is in contrast. From this comparison, it can be deduced that the Burgers vector for the loops must be of type a( IOO} with the loops lying on (100) planes and there are three sets of loops present. However, a similar analysis of loops for the Fe-15Cr alloy after irradiation at 450°C shown in figs. 3(f) and 3(g) reveals that both a(100) and a/2 ( 111) loops have developed. Because loops are present

518

D.S. Gelles / Neutron-irradiated

simple ferritic alloys

Fig. 1. Microstructures of the alloys prior to irradiation. Compositions are as indicated.

in [%O] contrast and do not appear in [ilO] contrast, such loops are of type u/2 (11 I}. Evidence for precipitate development as a result of

irradiation was observed over the whole range of Fe I+-Cr binary alloys examined. In most cases, th e precipit .ates appear as small equiaxed particles distribu [ted randc ,mlY

Fig. 2. Void microstructures for selected ,iron-chromium binary alloys specimens. Compositions and irradiation conditions are as indicated.

through the matrix. They measure appro~mately 20 mn in diametef after irradiation at 450% The phase appears similar to that deduced to be chromium-rich cy’by

Little and Stow [9] and indeed the particle n umber density is greatest in specimens of Fe -1SCr an d Fe18Cr. The precipitate is most easily revealed when

(cmW2)

(cm‘.‘)

Mm)

NM.

8.3 x IO9

4.0x IO’2

4.9x IO”

3.8X IO”

2.0x IO”

58

33.4

26

40

0.10

0.06

0.07

2.8

2.8

2.8

2.8

450

450

450

450

Fe- IZCr-O.lC-0.5Mo

Fe-l2Cr-O.IC-2Mo

Fe- I ZCr.-0. IC-3Mo

NM. = Not Measured.

NM.

5.8 x I09

I09 IO9 10’0 IO9 109

Fe- I2Cr- I MO

0

7.0x 3.2X 1.4x 6.8x 2.5 x

2.8

IO” 10’3 IO” IO” IO” 10”

450

-0

2.2x 4.4x 6.0x 4.2x 1.5x <

Fe- l8Cr

Fe- l5Cr

Fe- l2Cr

38.9 48.5 19.9 30.3 32.1 N.M.

2.8 x IO” 7.8x 10”

33.8 25.4

4.3 2.8 3.4 4.3 2.8 3.4 4.3 2.8 2.8

425 450 400 425 450 400 425 450 450

Fe-9Cr

2.1 x IO9 2.2x IO9

7.7 x IO”

18.4

0.28

3.4

400

Fe-60 0.63 0.10 NM. 0.68 0.24 0.25 0.63 0.29

1.3x 10’0

8.0x IO”

26.1

_

Pd

P”

NM. 1.0x 109 NM.

density

density

diam. d

Dislocation

NM. 0.09 0

AV/ v, @)

Swelling

3.4 4.3 2.8

MeV)

(“C) EB0.l

Fluencc

(n/Cm2

Temp.

for irradiated simple ferritic alloy specimens

400 425 450

measurements

Fe-3Cr

NOY

Table 2 Microstructural

and loops

Network and loops

Network and loops

Network

dub

LOOQSllCSUboun-

Network NUWO7k Netwosk and loops NCtWOIk Network LWpancarboundUiC!Spnd QQtS

NUWOfk Network

same.network

(110) plieacd boQs.

Network and loops Loq~ and straight

Description of arrays

structure

loops

loops

(Ill)

(110)

aWf0

a@W LOOPG hmo: a/2 (111). SIMU: a( 100). Network; abubouaduy:

aW) Loops; lm$c: a/2 (111). mull: a( 100). NW, IlCWbauduy:

+W Loopr; hrlge: u/2 (111). small: a( 100). Netwoh; IlWbouoduy:

Lariv a/2 (ill), SUt8llWYbC

Mofeu/2 (Ill) bps but u( K-IO) 8omavhat larger

LaW;(;d;lu

hfo&y4/2

mpnemt

mpnrnt

Qte#at

presalt Both

Both

QIEWlt

Both

Mouly a( loo)

+w

4w

Character o(lO0) vso/2

521

D.S. Gelles / Neutron-irradiated simple ferritic alloys

Fig. 3. Dislocation

microstructures

for selected

iron-chromium

binary alloy specimens. Composition,

irradiation condition

IS

and

imaging conditions are as indicated.

located near foil surfaces and therefore it is easily confused with oxide or surface contaminants. Nevertheless, it cduld be observed within foils under strain

contrast conditions. Diffraction spots or iginating from this precipitate were not observed but the piu titles could be revealed clearly under weak bczam dark :-field

Fii. 4.

Microstructures of iron-chromium-carbon alloys irradiated at 4SOT to a fluence of 2.8X IO” n/cm’. Arrows identify

observable voids.

dislocation contrast conditions. In those cases, however, it appeared that only 25% of the particles could be thus imaged at one time. Thus, the precipitate may be u’ but

further experiments involving extraction techniques are required for verification. A second precipitate phase could be identified after

AS.

Grifes / patron-i~a~ated

simpre ferritic

Fig. 5. Voids and dislocation microstructures in iron-chromium-carbon-molybdenum 2.8 X 10” n/cm*. Compositions and imaging conditions are as indicated.

irradiation in the Fe-3Cr alloy. It consisted of (11 I} rod shaped precipitates and randomly oriented elongated plate4ike precipitates. Examples can be found in

alloys

523

alloys irradiated at 450°C to a fluence of

fig. 3(a). Based on the work of Andrews et al. [IS] and Woodhead and Quarrel [ 161,this precipitr ste is identified as M,C and.therefore its volume fractionn is low due to the low 0.004% level of carbon in the allr3Y*

D.S. Gelles / Neutron-irradiated simple fem.tic alloys

524 3.3. Irradiated Fe-IZCr-C

alloys

larger loops are generally the a/2 (111) type but small Superimposed and moat apparent adjacent to gram boundaries are dislocation tangles of predominantly a( 100) character. Examples of void and dislocation structures found in Fe-12Cr-C-MO alloys after irradiation at 450°C to a fluence of 2.8 X 1O22 n/cm2 are provided in fig. 5. Wherever possible, regions adjacent to grain boundaries are depicted, and it can be noted that figs. 5(c) and S(d), and S(e) and 5(f) show the same areas under different contrast conditions. From fig. 5 it can be noted that variations in molybdenum have less effect on dialocation structure than variati& of carbon, Carbon additions apparently promote dido@ttion evohttion and decrease loop number densities. By comparing fig. 5 with figs. 2(e) through 2(h) and figs. 3(e) through 3(g), it can be demonstrated that additions of molybdenum are comparable to additions of chromium whereas carbon additions can be interpreted as having the opposite effect. The major effect of molybdenum additions to the Fe- 12Cr base alloy is to increase the tendency for precipitate development aa a result of irradiation. As in the previous casea diacua& all spacimenr containing molybdenum develop a fine diqeraion of @axed putitles on the o&r of 30 nm. Diffraction effecti fof this phase arc aa dercribcd previously in section 3.2. Also in proportion to tbeamount of molybdemunprcaanl, large blockypmcipitateatendto&vekapatgraia~ and in association with larga voida. Moat probab& they are M,C or M,C, [15,16J. Exarnph of this bdwior canbefoundinzig.S.TBeaqukxcdprscipitpteismost easily identified in the nriaogaphr under void contrast conditions and tbe blocky precipitate can be seen defining grain boundary regions. a( 100) loops could also be identified.

Specimens of Fe- 120 with additions of 0.1 and 0.2% carbon developed similar structures after irradiation to a fluence of 2.8 X 10z2 n/cm2 (E > 0.1 MeV) at 450°C. The preirradiation tempering of the Fe-12CrO.lC ahoy produced no significant microstructural differences and addition of a further 0.1% C resulted in only a somewhat larger subgrain size and larger Ms-,G precipitate particles. Of particular note are the observations that small voids at very low concentrations were identified after irradiation in the 0.1% carbon alloy in both conditions and a fine precipitate was observed within subgrains. Examples of these microstructural features are provided in fig.4. Voids observable in figs. 4(a) and 4(b) are identified with arrows. The huger dark regions delineate the subgrain structure resulting from the martensite tempering process. 3.4. Irradiated Fe-12Cr-C-Mo

alloys

The microstructures of specimens containing iron, chromium, carbon and molybdenum after irradiation were not very different from those of the less complex alloys. However, due to the fact that molybdenum acts as a ferrite former and, as a result, most of these alloys contain a duplex structure of tempered martensite and ferrite, the complexity of analysis was increased twofold. The Fe- IZCr-C-MO ahoy -ens were more swelling resistant than the Fe- 12Cr alloy, but less swelling resistant than the Fe- 12Cr-O.lC alloy. The swelling observed in the ferrite phase of the molybdenumcontaining alloys was significant but less than that of Fe-12Cr. Because no swelling was found in the tempered martensite phase of the molybdenumumtaining alloys, it is concluded that the molybdenum additions in tempered martensite inhibit void nucleation and growth. As a result the overall swelling was intermediate between that of Fe- 12Cr and Fe- 12Cr-0.1 C. Voids tended to be nonuniformly distributed, and often appeared in rows. A marked occurrence of void-precipitate association was also found. An absence of voids was noted adjacent to grain boundaries. Void swelling measurements are included in table2 for the molybdenum-containing alloys. The dislocation structures generated in the ferrite phase. of these specimens as a result of irradiation at 450°C are intermediate between those in Fe-12Cr and Fe- 18Cr. The dislocation structures consist primarly of perfect dislocation loops which had developed almost to the point where a dislocation tangle was created. The

4. Dbcurios Sufficient data have been obtained to more clearly define the effects of compositional variation on void swelling in ferritic alloys. Fig. 6 compares the void awelling results of the present study with those of Little and Stow [9] for iron-chromium binary alloys. Surprisingly good agreement is found if one ignores fluena variations. It can be noted that the two curves almost superimpose following irradiation at 450°C versus 46O’T. A peak in the magnitude of swelling is identified at 12% chromium. The data points obtained in the present study are almost within experimental error in the comparison at 420°C versus 425“C except that a much lower minimum at 3% chromium is indicated by the

D.S. G&es / Neutron-irradiated

0.0

Fig. 6. Cotiparision of void swelling measurements as a function of chromium content for various irradiation conditions. Results from both the present study and Little and Stow [I,21 are shown.

study than was anticipated by Little and Stow. Finally the data obtained after irradiation at 400°C in the present study lie between the Little and Stow curves for 38O’C and 420°C. These two data sets indicate that (1) a very strong suppression of void swelling occurs in the composition range of 3% chromium, (2) peak swelling response for alloys with chromium contents in excess of 12% shifts to lower temperatures (apparently thermal behavior controls to lower temperatures in-reactor at higher chromium levels), and (3) the swelling rate of Fe-Cr alloys is at least 0.15%/1022 n/cm2. Compositional variations also clearly affect dislocation structure evolution. However, it should be first noted that the observation of ~(100) dislocation loops in irradiated ferritic alloys is not unanticipated. Masters identified a( 100) loops in pure iron irradiated with Fe+ ions at 550°C [17]. Little and Eyre noted similar response in annealed silicon-killed mild steel irradiated with 1 MeV electrons at 550°C [18]. Also, Little et al. report many ~(100) dislocation segments in pure iron following fast neutron irradiation at 42O’C and suggested they form from growth and coalescence of

present

525

simple ferritic alloys

a( 100)~type interstitial loops [ 121.What is perhaps most surprising is the observation that a/2 (111) dislocations can develop in competition with a( 100) loops. Furthermore, as the chromium content is increased in Fe-Cr binary alloys, the fraction of u/2 (111) dislocations increases. This observation is important because it runs counter to models developed by Little et al. [12] and Bullough et al. [ 131 to explain void suppression in ferritic steels. A further comment regarding the dislocation structure evolution as a function of chromium content may be made based on elastic constant anisotropy. It is well known that iron has highly anisotropic elastic constants. Data for Fe-19Cr elastic constants, however, show that the anisotropy decreases with increasing chromium content. This comparison is shown in table 3. It is therefore possible that the increase in the fraction of u/2 (111) dislocations with increasing chromium content is due to a decrease in elastic anisotropy ratio. The swelling resistance of ferritic alloys can be attributed to the presence of the radiation-induced u (100) dislocations. Effects of chromium and carbon which results in precipitation and segregation of course must play a role in controlling swelling incubation and swelling rate. However, the u( 100) Burgers vector makes the ferritic alloy class unique. Most other materials generate radiation-induced steady state dislocation structures comprised of glissile dislocations which can readily form mobile dislocation networks. Examples are u/2 (110) dislocations in austenitic alloys and u/2 (111) dislocations in body centered cubic alloys such as molybdenum. However, the ~(100) dislocations in ferritics can be considered sessile (unable to move conservatively) based on the planar geometries of large loops following irradiation. Also, intersections of a( 100) dislocations on different planes will tend to create u( 110) stair-rod configurations thereby preventing dislocations from moving passed one another. A network of ~(100) dislocations should therefore be much less mobile. A dislocation network combining u( 100) and u/2 (111) dislocations should alter this situation. Segments

Table 3 Elastic constants of pure iron [ 191 and Fe- 19.43Cr (201 Composition

Cl,

C,,

C44

A=-

2C44 cll-cl2

Pure iron Fe- 19.43Cr

(lO’kg/cm*)

(lo5 kg/cm2)

(10’ kg/cm’)

24.2 22.41

14.65 11.84

11.20 11.36

2.36 2.15

526

D. S. Gelles / Neutron-irradiated

of a/2 (111) between the a (100) segments can provide a means whereby dislocaltion line length may be reduced and network mobility improved. It is anticipated that this is the reason that swelling can be higher in Fe- 12Cr than in Fe-3Cr. Note that based on point defect capture bias, ~(100) dislocations should climb faster than a/2 (111) dislocations due to the larger Burgers vector which produces a larger strain field [21]. A further effect of note is that due to the carbon content in Fe- 12Cr-C-MO alloys. Effects of carbon on irradiation response in ferritic steels are generally difficult to demonstrate due to the dramatic effect of carbon content on microstructures. However, the present study includes carbon variations for an Fe-12Cr1Mo base composition where a ferrite phase is always present and, although the Fe- 12Cr-0. lC- 1Mo specimen was not examined, Fe-12Cr-O.lC-OSMo and -2Mo were effectively identical and therefore provide the necessary information. Figs. S(a) through 5(f) can be used to demonstrate that increased carbon content leads to enhanced microstructural evolution at 45O’C. This is perhaps the reverse of what might be expected based on solute segregation arguments.

5. Conelllsl0M As a result of microstructural examinations of a series of irradiated Fe-Cr-C-MO simple ferritic alloys, it can be concluded that changes to the base composition of a 12Cr martensitic stainless steel can be expected to produce the following results. 1. Reduction of chromium contents to the Fe-3Cr range minim&s void swelling and produces predominantly a( 100) dislocation structures. 2. Increases in chromium contents above Fe-12Cr shifts microstructural development to lower temperatures and produces predominantly u/2 (111) dislocation structures. 3. Peak swelling of 0.7% was found in a specimen of Fe-90 irradiated at 425’Y, and this corresponds to a swelling rate on te order of 0.15!%/1022 n/cm2. 4. Fe-12Cr-O.lC with a tempered martensite structure can develop void swelling during irradiation at 45O’C to a fluence of 2.8 X 10” n/cm2 (E>O.l MeV) but the swelling is very low. 5. The primary effect of variations in molybdenum

simple ferritic alloys

content is to change the size and amount of precipitates at grain boundaries.

Ret erences Ill See for example, D.S. Gelles, 12Cr-1Mo Steel for Fission

and Fusion Applications, in: Proc. Intern. Conf. on Ferritic Steels for High Temperature Applications, ASh$ to be published. 121J. Brler, A. Maillard, G. Brun, J. Lehmann and J.M. Dupouy, in: Irradiation Behavior of Metallic Materials for Fast Reactor Core Components (1979) paper AZ. [31 J.J. Huet, A. DeBremaecker, M. Snykers and P. Van Asbroeck, in: Irradiation Behavior of Metallic Materials for Fast Reactor Core Components (1979) paper Al’. [41 E.A. Little and D.A. Stow, J. Nut. Mater. 87 (1979) 25. PI R.W. Powell, D.T. Peterson, M.K. Zimmerschied and J.F. Bates, in: Proc. 2nd Topical Meeting on Fusion Reactor Materials, J. Nyc. Mater. 103 & 104 (1981) 969. WI M.M. Paxton, B.A. Chin, E.R. Gilbert and R.E. Nygren, J.Nuc. Mater. 80 (1979) 144. 171 M.M. Paxton, B.A. Chin and E.R. Gilbert, J. Nut. Mater. 95 (1980) 185. PI W. Vandermeulen, A. DeBremaecker, S. de Burbure, J.J. Huet and P. Van Asbroeck, in: Behavior of Metallic Materials for Fast Reactor Core Components (1979) paper Al. 191 E.A. Little and D.A. Stow, Met. Sci. J. 14 (1980) 89. 1101 M.R. Haynes and TM. Williams, J. Nut. Mater. 74 (1978) 151. 1111 E.A. Little, J. Nut. Mater. 87 (1979) 11. 1121 E.A. Little, R. Bullough and M.H. Wood, Proc. Roy. Sot. (London) A372 (1980) 565. r131 R. Bullough, M.H. Wood and E.A. Little, in: Effects of Radiation on Materials Tenth conference, ASTM-STP-725 (1981) p. 593. 1141 D.S. Gelles, in: Proc. 2nd Topical Meeting on Fusion Reactor Materials, J. Nucl. Mater. 103 & 104 (1981) 975. [I51 K.W. Andrews, H. Hughes and D.J. Dyson, J. Iron Steel Inst. 210 (1972) 337. iI61 J.H. Woodhead and A.G. Quarrell, The Role of Carbides in Low Alloy Creep Resisting Steels (Climax Molybdenum Company, London, UK, 1965). 1171 B.A. Masters, Phil. Mag. 11 (1965) 881. 1181 E.A. Little and B.L. Eyre, Met. Sci. J. 7 (1973) 100. 1191 W.P. Mason, Pieaoelectric Crystals and Their Application to Ultrasonics (Van Nostrand, New York, 1950). WI H. Maxumoto and M. Kikuchi, Trans. JIM 12 (1971) 90. 1211 W.G. Wolfer, Scripta. Met. 9 (1975) 801.