M icrostructure and tensile properties of heavily irradiated 5052 o aluminum
Journal of Nuclear Materials 97 (1981) 33-43North-Holland Publishing Company MICROSTRUCTURE AND TENSILE PROPERTIES OF HEAVILY IRRADIATED 5052-O ALUMINUM ALLOY * K. FARRELL Oak Ridge National Laboratory, Oak Ridge, Tennessee 37830, USA Received 14 July 1980; in revised form 3 November 1980 During neutron irradiation of an AI-2.2% Mg solid solution alloy in the High Flux Isotope Reactor to fast and thermal fluences >10z7 neutrons(n)/m2 at 328 K (0.35 Tm) about seven percent insoluble, transmutant silicon was produced. Some of this silicon reacted with the dissolved magnesium to form a fine precipitate of MgzSi. A tight dislocation structure was also created. The alloy showed good resistance to cavity formation. These microstructural features are responsible for pronounced strengthening and an associated marked loss in ductility as revealed by tensile tests at 323, 373, and 423 K (0.35, 0.40 and 0.45 T,,.,). These changes were greater than in magnesium-free aluminum and in alloys containing preexist- ing, thermallyaged Mg2Si precipitate. Increasing the thermal-to-fast flux ratio from 1.7 to 2.1 caused further strengthening beyond that expected from a simple increase in silicon level.1. Introduction 1O24 fast n/m2, 1.26 X 102’ total n/m2) at -340 K (-0.36 T,,,) showed an increase in strength and a Many aluminium alloys depend upon thermally- small loss in ductility for the 0 temper; the H temperinduced precipitates to provide satisfactory mechani- material showed a smaller increase in strength and ancal strength for service applications. The SOOO-Series increase in ductility, suggesting some recovery of coldalloys are exceptions in that they are solid solution work during irradiation. Irradiations of the 5052-Ostrengthened by magnesium, one of the few elements alloy to much higher fluences [up to 3.6 X 1O26 n/m2with significant solid solubility in alumlnium. The (fast) and 5.8 X 1O26 n/m2 (thermal)] at -330 K5052 alloy consists nominally of 2.5 weight percent caused substantial increases in creep-rupture strengthmagnesium and 0.25 chromium. It is normally used at 323 K and considerable loss in ductility with ain “0’‘-fully annealed, or “H”-partially strained con- propensity for intergranular failure . Examinationditions. Such alloys have good formability and resis- of the microstructures [2,3] revealed the presence oftance to corrosion in flowing water. They have found a fine precipitate of Mg2Si which was held largelyapplications in some water-cooled reactors for situa- responsible for the irradiation-induced increases intions where the strength requirements lie between strength. This Mg2Si precipitate is produced by thethose of the softer 1000 Series alloys and the stronger interaction of transmutation-produced silicon withprecipitation-hardened 6000 Series alloys. the dissolved magnesium. The silicon is created primarily by thermal neutrons through the sequential reactions 27Al(n, 7) + 28Al, “Al + 28Si t 0, for2. Previous irradiation studies which the cross-section is high, 2.3 X 10s2’ m2 for Tensile tests [l] on 5052 and 5154 (3.5 Mg) alloys 2200 m/s neutrons (i.e., E 5 0.025 eV). Sufficientlyin 0 and H tempers irradiated to low fluences (1 X high temperatures and/or the presence of atomic dis- placements from fast neutrons allow transport of the silicon and development of Mg2Si precipitates. In* Research sponsored by the Division of Material Sciences, US Department of Energy, under Contract W-7405eng essence, irradiation of originally solid solution 5000 26 with Union Carbide Corporation. Series aluminum alloys in a reactor of mixed neutron 33
34 K. Farrell /Microstructure and tensile properties of aluminum alloyspectrum gradually converts them to precipitation- ment with the usual practice for other metals, and wehardened 6000-type alloys. remain consistent with previously published data on The present work takes these observations to aluminum. However, as a constant reminder of thehigher neutron fluences and explores the effects of input from the thermal fluence, a scale showing thetest temperature on the tensile properties of irra- levels of transmitted silicon for q&J& = 1.7 isdiated 5052-O alloy. included on the abscissa of graphs of mechanical properties.3. Experimental conditions 4. Microstructural changes The chemical composition of the commercially-produced 5052 alloy was 2.2 wt% Mg-0.2 Cr-<0.3 The microstructure of the alloy before irradiationNi-0.18 Fe-<O.l Si. Button-headed tensile speci- consisted of a solid-solution matrix containing a fewmens of gauge length 28.6 mm and guage diameter inclusion particles. At the lowest fast fluences of3.2 mm were machined from cold-swaged rod and l-l .5 X 102’ n/m2 (“2 dpa), radiation damage struc-were annealed for 0.5 h at 617 K (0.66 T,,,) in a salt ture was observed in the form of small, unfaultedbath to give a fully-recrystallized “0” temper. loops very close to grain boundaries and around The specimens were irradiated in the peripheral residual inclusions (fig. 1). At some of the inclusions,target positions in the Oak Ridge High Flux Isotope the loops had grown into dislocation tangles. TheReactor (HFIR) in contact with the cooling water at transmutation-produced silicon level was about 5.8 X328 K (0.35 T,) for periods up to three and one-half 10e2 at% at this fluence, and no Mg,Si precipitatesyears. HFIR is predominantly a thermal reactor with were discernible. With increasing fluence, a generally-a strong fast neutron component. The fast neutron dispersed, closely-knit, tangled dislocation structure(E > 0.1 MeV) fluences in these experiments ranged was developed and a finely-divided precipitate ofup to 1.8 X102’ n/m2, corresponding to 260 dis- Mg,Si appeared (fig. 2a). The precipitate was con-placements per atom (dpa). Associated gas generation tinuous to the grain boundaries, i.e., no obvious near-was estimated to be up to 8.5 X lo-’ atomic fraction grain boundary denuded regions. The identity of thehelium and 5 X lo4 atomic fraction hydrogen. The precipitate is evident from the (001) diffraction pat-thermal neutron fluences (E < 0.025 eV) were higher tern showing streaking in ( 100) and (010) passingthan the fast fluences by a factor of about 1.7, and through the 100 and 110 type reciprocal latticewere responsible for the creation of up to 6.9 at% points. Such features are characteristic of thermally-(7.15 wt%) silicon. produced Mg,Si phase in aluminum alloys [3,4]. The In this work, both the fast fluence and the thermal rings of spots in the pattern are not characteristic offluence make significant contributions to the micro- thermally-produced Mg,Si and further comment onstructural changes and hence to the mechanical prop- them is withheld for the discussion.erties response. It becomes a moot point whether the A few cavities were present after irradiation to fastfast fluence or the thermal fluence should be con- fluences above about 5 X 1O26 n/m2, correspondingsidered the primary independent variable. The two to about 70 dpa. These cavities were most oftenare not independent and both are important. The sili- located in loose clusters within the grains andcon from the thermal neutrons is largely insoluble attached to residual inclusions; some were also notedand forms precipitates of Mg,Si with dissolved mag- on grain boundaries. The cavity concentration wasnesium. Atomic displacements by the fast neutrons low, about 101’/m3, and their mean size was 25 nmassist and control the formation of the precipitates by for fast fluences in the vicinity of 5 X 10z6 n/m’. Atenhancing diffusion rates and providing nucleation the maximum fast fluence of 1.8 X 1O27 n/m2 (260sites. The fast neutrons also contribute to damage dpa) there was no significant change in the concentra-microstructure by forming dislocations and cavities. tion of cavities but their size had increased to aboutFor the sake of simplicity we discuss our data in 50 nm. Several cavities are shown in fig. 2b; note thatterms of the fast fluence. This way we retain agree- the size of the precipitate has increased, too. Reflec-
K. Farrell f Microstructure and tensile properties of aluminum alloyFig. 1. Dislocation loops developing near grain boundaries and around inclusions at 1.5 X 1O2s n/m* 00.1 MeV) and 2.5 X lo*’n/m* (<0.025 eV).Fig. 2. (a) Precipitates and dislocations at 5.7 x 10z6 n/m* , with (001) diffraction pattern. (5) Cavities and larger matrix precipi-tates at 1.8 X 1O26 n/m*. (c) Heavy discontinuous precipitation at grain boundaries. Thermal fluences are -1.7 times the quotedfast fluences.
36 K. Farrell ~~~crastru~tllre and tensile properties ~~all(rn~nurn a&ytions from MgsSi precipitates persisted in the diffrac- temperatures up to 523 K, but cavities disappeared attion patterns at these high fluences. Also, in the grain 523 K. At 573 K there was obvious coarsening of theboundaries there was pronounced accumulation of precipitate particles and a loss of dislocations. Atprecipitate in large, spongy-like discontinuous par- 623 K there was further precipitate coarsening in theticles (fig. 2c). matrix and the formation of large particles on grain The microstructures in the strained gauge sections boundaries.of the more heavily irradiated specimens were indis-tinguishable from those in the unstrained heads of thespecimens. This is not surprising since it would be 5. Tensile propertiesvery difficult, if not impossible, to recognize slip dis-locations in a structure conta~~ng preexisting, dense, Tensile tests were conducted at a strain rate ofradiation-induced dislocation tangles. Only if there 7.4 X 10-‘/s in air at temperatures of 323, 373, andwas an unusual plastic deformation mechanism, such 423 K (0.35, 0.40, and 0.45 T,), where the radia-as dislocation channeling, would the slip dislocations tion-produced microstructures were stabIe. In tests atbe discernible. The conclusion is that the plastic 323 K, the unirradiated alloy displayed a well-de~neddeformation mechanism was normal for a precipita- yield point and continuous serrated, or jerky, flowtion-hardened material. throughout the test, as seen in the upper half of The thermal stability of the radiation damage fig. 3. The frequency of serrations was particularlystructure was tested by cutting disks from specimens high after the maximum load was exceeded and neck-irradiated to fluences of about 6 X 1O26n/m2 (fast) ing had commenced. Progressive irradiation elimi-and 1 X 102’ n/m2 (thermal) and annealing them for nated both the yield point and the jerky flow, and0.5 h at temperatures of 423, 473, 523, 573, and caused pronounced streng~en~g and loss in ductil-623 K, followed by air cooling. TEM examination ity. Strengthening occurred primarily throughrevealed no change in the precipitate structure for increased yield or initial flow stress. Because the flow 600 80 k? 0 0 2 4 6 8 i0 12 $4 16 18 20 2 22 24 26 rv” 2; I I 1.6 x lO27 n/m* -- 500 T,_, = 423 K 60 -j L 400 I I ELONGATION Pa) Fig. 3. Effects of radiation on tensile curves at 323 and 423 K.
K. Farrell /Microstructure and tensile properties of aluminum alloy 31 dpo 2 xtoo (0’ (02 (03 I I I 1 too A I I I llrrl I I111111 I I llllll I lllllll1 600 500 60 400 = 8 ;; /6 fk.2 7. FLOW ST+ % I 3 : 300 z 200 too 0 FLUENCE. > 0.t McV (n/m’) I lllll I I I111111 1 l 1 tt1111 0.1 t (0 wt X *%i Fig. 4. Fluence dependence of tensile properties at 323 K (0.35 T,).stress increased more rapidly than the UTS, the work- reached at a fluence of about 2 X 102” n/m2. At thehardening rate tended to decrease at the higher highest fluence the 0.2% flow stress was increased byfluences. The fast fluence dependence of the tensile a factor of 6.5 and the UTS by a factor of 2.7. Theproperties at 323 K is shown in fig. 4. The appro- ductility (expressed here as plastic elongation, thepriate abscissa of thermal fluence is obtained by elastic strain being ignored) was progressively reducedtranslating the fast fluence scale to the right by a fac- from 26% elongation for the unirradiated material totor of 1.7. At the lower fluences the increases in flow about 8% at a fast fluence of about 2 X 102’j n/m2,stress and UTS were only fractional, and the flow above which the ductility remained seemingly inde-stress remained considerable lower than the UTS. pendent of fluence. The loss in ductility was causedSubstantial hardening occurred with increasing entirely by loss of uniform strain. The fracture pathfluence, the 0.2% flow stress increasing more rapidly was transgranular by locally ductile tearing and wasthan the UTS with the closest convergence being invariant with fluence.
38 K. Earrell /Microstructure and tensile properties of aluminum alloy dw 2x10° 40 to2 103 I I I 1 so I I 1 Id,, I I 11rlll I I I lllll I I lllll’ 600 5052 - 0 Al 80 - 5t-r 5 328~ Ttest = 373 K a ; = 7.4 x (0-5 s-4 500 70 60 400 ; 50 2 g 300 FLOW STRESS k 40 c v) 30 g 200 z 0 F s 20 5 = 100 i0 C I I IIJJ 0 0 iO26 4027 4028 FLUENCE,> 0.8 MeV (n/m*) I 1 I llill I I I I Illil I I 1111111 0.1 4 IO wt % ‘*Si Fig. 5. Fluence dependence of tensile properties at 373 K (0.4 Tm). At the two higher test temperatures (figs. 5 and 6), higher fhrences, and necking instab~~ty occurred im-similar patterns of fluence dependency of tensile mediately after yield (lower part of fig. 3). Corre-properties prevailed, but the strengths and ductilities spondingly, the totai elongation fell to about 3% withwere less than those at 323 K. Minimum uniform no measureable uniform elongation. Under theseelongation was reached at progressively lower flu- conditions, fracture continued to follow a locally-ences with increasing test temperature. At 423 K, the ductile transgranular path. At the highest fluence a0.2% flow stress and UTS values were equal at the facetted surface was evident. It was not clear, though,
39 500 70 T,,,#= 423 K c: = 7.4 x10-5 <’ 400 G g 300 s E 2 200 100 0 0 0 10= FLUENCE. aO.1 MeV (n/m*) I I III1 I I I IllIll I I I I11111 0.1 1 10 WI % *%i Fig. 6. Fluencedependenceof tensile propertiesat 423 K (0.45 Z’,).whether the facets were intergranular. A single test particularly apparent at 373 K (fig. S), where enoughperformed on one of the highest fluence specimens at specimens were tested to demonstrate a pronounced423 K and a strain rate 40 times higher than that in upward shift in the flow stress and UTS curves.fig. 6 showed small increases in strength and increasesin total and uniform elongations to 4.6 and 1.9%,respectively, with non-facetted fracture. 6. Discussion At all test temperatures it was found that a fewspecimens, which had been irradiated in sites wherethe thermal-to-fast flux ratio was -2.1, appeared to This work confirms and extends the earlier reportsconsistently display higher strengths than the more [l-3] of radiation-induced strengthen~g and micro-common specimens with a thermal-to-fast flux ratio structurat changes in 5052-O aluminum alloy, andof 1.7. The data for these aberrant specimens are reveals some new features. It also provides the mostindicated by the dotted lines in figs. 4-6. They are comprehensive description of the effects of neutron
40 K. Farrell /Microstructure and tensile properties of aluminum alloyfluence on the tensile properties of the originally clusters but are favored for formation of gas bubbles,5052 alloy. suggests that the cavities are developed on gas bub- Before proceeding to a discussion of the more bles. Their disappearance during heating at 523 K,obvious findings of this work, attention is drawn to which agrees with the temperatures required to sintertwo microstructural aspects that might otherwise pass voids in purer aluminum [8,14], indicates that theyunnoticed, but which are considered to be valuable are not helium filled. They could, possibly, be filledobservations. The first is that the 5052-O alloy has with the more mobile hydrogen, since hydrogenvery good resistance to cavity formation and swelling. forms bubbles readily in aluminum [ 151 and theyThis becomes more readily apparent when the void become unstable above 473 K [ 15,161 as the hydro-parameters are compared with those of high purity gen solubility in aluminum increases.aluminum [ 71, commercial-purity 1 loo-grade alumi- The second unobtrusive aspect of the microstruc-num [6,8,9] and the Mg,Si precipitation-hardened ture is that the cavity distribution in the 5052 alloy is6061 alloy [ IO,1 11, irradiated under similar condi- markedly different in one major respect from those intions. In the high-purity aluminum, cavities are the other aluminum materials. In pure aluminum,created at a fast fluence as low as 1.5 X 1O23 n/m2, 1100 Al and 6061 Al, the first cavities are developedand at 1O26 n/m2 there is >5% swelling. To achieve in sheets lying close to and on each side of grainthe same level of swelling in 1100 Al, a fast fluence of boundaries. These grow to be the largest cavities andabout 102’ n/m2 is required. At this high fluence provide planes of mechanical weakness and potentialthere is less than 1% swelling in 6061 alloy and less fracture sites. No such cavities were found in thethan 0.1% in the present 5052-O alloy. This reduced 5052 alloy, even though the initial damage struc-swelling in the 5052 alloy is achieved through signifi- ture - dislocation loops - was developed in the vicin-cant decreases in both cavity concentrations and ity of grain boundaries. It was noted, too, that therecavity growth. In particular, the incubation dose for was no denuding of the Mg,Si precipitate in the vicin-cavity formation is rather high, about 5 X 1O26 n/m2, ity of grain boundaries.which is more than three thousand-fold that for pure The formation of Mg,Si precipitate [2,3] is con-aluminum . Such strong resistance to cavitation firmed. Moreover, it is found that this precipitatemust be imparted, initially at least, by the magnesium persists to the highest fluences examined despite theand other minor impurities. While these remain in presence of considerable excess silicon and the possi-solution, they are assumed to act as trapping and ble precipitate restructuring effects of the very highrecombination sites for vacancies and interstitial atomic displacement levels. About 1.2 at% silicon isatoms, thus depressing the vacancy supersaturation, required to fix the total magnesium content as sto-as discussed in an earlier paper [ 121. As the magne- chiometric Mg2Si precipitate. This quantity is reachedsium is drawn from solution to form the Mg,Si pre- at a fluence of about 3 X 1O26 fast n/m’ (“5 X 1O26cipitates, trapping and recombination are presumably thermal n/m’). Since silicon is essentially insoluble inshifted to the precipitates whose high spatial density aluminum, any excess beyond this 1.2% must bemight provide overlapping point-defect capture zones. accomodated as precipitate. About 7% silicon is gen- Gases are known to enhance cavity formation dur- erated at the highest fluences. In magnesium-freeing irradiation : yet in this 5052 alloy cavity aluminum [5,6] the silicon is manifest as elementalnucleation is restrained despite high levels of trans- silicon. However, in the present irradiated alloys, themuted gas. This could, of course, be due entirely to a electron diffraction patterns are too complicated tovery low superaturation of vacancies. Another factor permit easy recognition of spots from elemental sili-could be trapping of gases by the high concentrations con. Furthermore, in the presence of a co-generatedof dislocations and precipitates, making such gases precipitate of Mg,Si, there is no guarantee that theunavailable for cavity nucleation. Nevertheless, the excess silicon will assume elemental form. It may befact that many of the cavities which do form after taken up by the Mg,Si or perhaps form a new precipi-prolonged irradiation, are associated with grain boun- tate. These considerations are currently under studydaries and stable particle-matrix interfaces, sites to account for the many extra precipitate spots thatwhich are unsuitable for nucleation of pure vacancy give rise to the discontinuous rings in the diffraction
K. Farrell /Microstructure and tensile properties of aluminum alloy 41patterns. Some of the complications arise from super- 6061 Al [ 1 l] over the full fluence range, and is seenimposed spots from particles freed from the matrix in the present 5052 alloy at fluences greater than 3 Xduring electropolishing then trapped in random orien- 1O26 n/m’. Below this fluence, however, in the rangetations on the foil surfaces. At present, no firm con- where Mg is slowly drawn from solid solution in theclusions regarding the extra spots are available. When matrix, the flow stress increases with fluence at asuch conclusions are reached they will be published. much faster rate than does the UTS. This is probably Whatever the final outcome of the search for the because the matrix is unstable. As the Mg is depletedsource of the extra diffraction spots, there is no from solution the flow stress and work hardening ratedoubt that the dominant radiation-induced micro- of the matrix are reduced and thus partially counter-structural defects in this irradiated 5052-O alloy are act the hardening by the Mg2Si precipitate: The resul-the dislocations and the finely-dispersed precipitates. tant rates of increase in the flow stress and UTS ofA quantitative accounting of the tensile strengths in the alloy will be disparate until a stable matrix isterms of hardening by these microstructural defects is attained, presumably when all dissolved Mg isnot attempted here because previous work on 1100 removed. These changes in the matrix will also beAl [6,8,17], 6061 Al [lo] and 5052 Al  has reflected in the ductility. Appropriately, the uniformshown satisfactory correlations, and there is little to elongation does not reach a minimum until a fluencebe gained in repeating the exercise. But it is worth of about 2 X 1O26 n/m?, which is an order of magni-mentioning that in 1100 Al, which contained little or tude higher than the corresponding points in 1100no magnesium and which developed many cavities, and 6061 alloys [6,1 l] irradiated and tested underthe larger part of the increase in strength could be the same conditions.attributed to radiation-induced dislocation tangles, The strength levels attained in this 5052 alloy arewith a significant contribution from silicon precipi- much higher than those reached in 1100-O ,tate and a minor boost from cavities. Similarly, in 6061-O [l l] and 6061-T6 [l l] aluminum alloys irra-6061 Al where Mg,Si precipitates preexisted. How- diated and tested under the same conditions. Theever, in 5052 Al where the Mg,Si precipitate was precipitate was also more finely distributed in thecreated during the irradiation, the concentration of 5052 alloy which would favor greater strength. This isprecipitates was so high as to overwhelm hardening consistent with the conclusions of King and Jostonsfrom other sources . Our fingings qualitatively  who pointed out that the radiation-inducedagree with this conclusion, and the observed precipi- Mg,Si precipitate in 5052 alloy is finer than thattates and dislocations are assumed to be responsible produced in 6061 alloy by thermal aging.for the considerable increases in tensile strength and The fineness of the precipitate may also be at thelosses in ductility. Catities make little contribution; root of the increase in strength with increases inmuch of the hardening occurred before cavities were thermal-to-fast flux ratio. These strength increases areformed. For fast fluences up to about 3 X 10z6 n/m’ greater than would be expected from a simplethe major precipitate is MgzSi. For higher fluences increase in the silicon level. For example, in fig. 5 thethe continuously-generated silicon must sustain the 0.2% flow stress at 1 X 102(j n/m2 on the unbrokencontinued hardening. curve which represents &/$r = 1.7 is 3 10 MPa and Actually, such correlations are not closely applica- the estimated silicon concentration is 0.4 wt%; at ble to the 5052 alloy over the whole of its fluence $th/@r = 2.1 (the dotted curve) the flow stress is 375 range. The correlations are based on the assumptions MPa and the silicon level is -0.5%. Yet the flow stress that the flow stress and work hardening properties of corresponding to -0.5% silicon on the r&t,/& = 1.7 the alloy matrices are not affected by the irradiation. curve is only 325 MPa. The difference may be due to The increments of hardening due to the radiation- the higher silicon production rate at &,/$r = 2.1 induced dislocation obstacles are then added to the causing faster, and hence finer, formation of Mg2Si unirradiated flow stress or UTS to derive the new particles. stresses. In this way, the fluence dependencies of the The decreases in strength with increasing test tkm- flow stress and the UTS should roughly parallel one perature are not caused by recovery or coarsening of another, which is what is observed in 1100 Al  and the microstructure, and they are larger than the 5% or
42 K. Farrell /Microstructure and tensile properties of aluminum alloyso decrease expected for the change in shear modulus 7. Conclusions [ 181. Thermally-assisted surmounting of obstacles bydislocations is a likely cause. (1) Irradiation of the 5052 aluminum-2.2% mag- The loss in ductility from irradiation appears to be nesium, solid solution alloy to high fluences in aa natural consequence of the radiation hardening. The mixed neutron spectrum at 328 K converts the alloybulk of the loss occurred through reduction in uni- to a precipitation-hardened 6000 type alloy.form strain, a reflection of the increased resistance to (2) The precipitates are derived from insolublemovement of slip dislocations provided by the defect transmutation-produced silicon, and include Mg,Si.structure. Microstructural examination revealed noth- (3) The alloy displays a high resistance to cavitying untoward about the deformation mechanism, and formation.the fractures were apparently transgranular and (4) The damage microstructure is stable up tolocally ductile, even those occurring at the lowest 523 K in short-term anneals.bulk strains. In only one case, a high fluence speci- (5) The flow stress and tensile strength aremen tested at 423 K, did the fracture path change, increased by factors up to 6.5 and 2.7, respectively,and this displayed a facetted surface similar to the with accompanying loss of ductility associated withintergranular failures observed by King and Jostsons marked reduction in uniform strain.  in their creep specimens. The implication is that (6) The changes in tensile properties are consistentslow strain rates and/or high test temperatures may with hindrance of slip dislocations by dispersedencourage intergranular separation. At pormal tensile phases.test rates, the alloy demonstrated good resistance to (7) Increasing the thermal/fast neutron flux ratiograin boundary failure, more so than 1100 and 6061 from 1.7 to 2.1 gives further strengthening beyondalloys [6,11,19] under similar conditions. The that expected from a simple increase in silicon level.absence of grain boundary denuded zones and near-grain boundary cavities may be contributing factors.Nevertheless, the overall ductility of 5052 alloy was Acknowledgementslower than that of 1100 and 6061 alloys at fastfluences greater than 2 X 1O26 n/m’, and was very This work is part of the Oak Ridge Nationallow at 423 K. Modest necking strains persisted at all Laboratory Aluminum Surveillance Program whichtest temperatures. owes a great deal to the early efforts of R.T. King and One final comment concerns the suppression of A.E.’ Richt. Their contributions are highly appre-serrated yielding by irradiation. Serrated yielding is ciated. E. Boling, J. Woods, W. Henry, E. Gilbert andoften attributed to the sudden release of dislocations J.T. Houston provided technical assistance, and Helenfrom an impeding atmosphere of solute atoms, fol- Sharpe and Linda Pollard typed this manuscript.lowed by re-segregation of the atmosphere and a repeat process. Presumably in the unirradiated 5052 alloy, the dissolved magnesium provides the Suzuki References atmosphere. Interestingly, irradiation eliminates ser- [l] R.V. Steele and W.P. Wallace, The effects of neutron rated flow long before all of the magnesium is esti- flux on the mechanical properties of aluminum alloys, mated to be removed from solution. Serrated flow LRL-145 (Livermore Research Laboratory, Livermore, disappears after irradiations in the mid 102’ n/m2, at CA, 1954).  R.T. King and A. Jostsons, Met. Trans. 6A (1975) 863. which point there is only about one-sixth of the  A. Jostsons and R.T. King, Scripta Met. 6 (1972) 447. transmutation-produced silicon required to draw all  M.H. Jacobs, Phil. Mag. 26 (1972) 1. the magnesium from solution as Mg2Si. It is quite [S] K. Farrell, J.O. Stiegler and R.E. Gehlbach, Metallo- possible, though, that nucleation of Mg2Si embryo is graphy 3 (1970) 275: complete at such fluences, and is on a fine enough  K. Farrell and A.E. Richt, in: Effects of Radiation onscale to inhibit segration of magnesium to disloca- Structural Materials, ASTM STP 683 (1979) p. 427.  N.H. Packan, J. Nucl. Mater. 40 (1971) 1.tions.  A. Jostsons and E.L. Long, Jr., Radiation Effects 16 (1972) 83.
K. Farrell /Microstructure and tensile properties of aluminum alloy 43  K. Farrell, R.T. King, A. Jostsons and E.L. Long, Jr.,  J.T. Houston and K. Farrell, J. Nucl. Mater. 40 (1971) Nucl. Metallurgy 19 (1973) 153. 225.[lo] R.T. King, A. Jostsons and K. Farrell, in: Effects of  C.E. Ells and W. Evans, Met. Trans. AIME 227 (1963) Radiation on Substructure and Mechanical Properties of 437. Metals and Alloys, ASTM STP 529 (1973) p. 165. [la] L.H. Milacek and R.D. Daniels, J. Appl. Phys. 39 (1968)Ill] K. Farrell and R.T. King, in: Effects of Radiation on 2803. Structural Materials, ASTM STP 683 (1979) p. 440.  K. Farrell and R.T. King, Met. Trans. AIME 4 (1973) K. Farrell and J.T. Houston, J. Nucl. Mater. 83 (1979) 1223. 57.  P.M. Sutton, Phys. Rev. 91 (1953) 816. K. Farrell, Experimental observations of effects of inert  K. Farrell and A.E. Richt, in: Properties of Reactor gas on cavity formation during irradiation, in: Proc. of Structural Alloys after Neutron or Particle Irradiation, Consultant Symp. on Rare Gases in Metals and Ionic ASTM SIP570 (1975) p. 311. Solids, Harwell, 1979, AERE-R9733 (1980); also Oak Ridge National Laboratory Report ORNL/TM-7193 (1980).