Understanding Pressure Vessel Steels: An Atom Probe ...arc.nucapt.northwestern.edu/refbase/files/MatChar44p235PVSteel... · Understanding Pressure Vessel Steels 237 a vanadium-containing

MATERIALS CHARACTERIZATION 44:235–254 (2000)Published Elsevier Science Inc. 1044-5803/00/$–see front matter655 Avenue of the Americas, New York, NY 10010 PII S1044-5803(99)00056-X

235

Understanding Pressure Vessel Steels:An Atom Probe PerspectiveM. K. Miller,* P. Pareige,*† and M. G. Burke‡

*Microscopy and Microanalytical Sciences Group, Metals and Ceramics Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6376; †Groupe de Metallurgie Physique, UMR CNRS 6634, Faculté des sciences de l’université de Rouen, 76128 Mont Saint Aignan, France; and ‡Bechtel Bettis, Inc., West Mifflin, PA 15122

A review of the contributions of the atom probe field-ion microscopy (APFIM) technique tothe microstructural characterization of pressure vessel steels and to the understanding ofthe embrittlement of these materials during neutron irradiation is presented. Atom probestudies have revealed that the microstructure contains a variety of ultrafine clusters andprecipitates some of which are only formed during neutron irradiation. Furthermore, thereis a complex pattern of segregation of various solutes (including phosphorus, nickel, man-ganese, or molybdenum) to grain boundaries in some pressure vessel materials, and theremay also be additional intergranular precipitation in these materials. Published byElsevier Science Inc.

INTRODUCTION

Since the introduction of the atom probefield-ion microscope (APFIM) [1, 2] in 1967,it has been used extensively for the charac-terization of steels that are used in the pres-sure vessels of nuclear reactors. The pri-mary aim of these studies has been tocharacterize the microstructural featuresthat are responsible for the embrittlementof these materials that occurs during ser-vice, and to evaluate postirradiation an-nealing treatments designed to restore themechanical properties. The embrittlementof these materials is clearly evident fromthe mechanical properties in an increase inthe ductile-to-brittle transition tempera-ture, a reduction in fracture toughness, anda loss in the upper shelf energy [3–8]. Theembrittlement has been correlated with thelevels of copper, phosphorus, nickel, andmanganese in the steel. In particular, thesteel welds in some of the first commercialreactors are a special concern because ofthe elevated copper content of the welds,which was primarily due to the use of cop-

per-coated welding rods in the weldingprocess. Because the service temperature ofthese materials is relatively low, typically288

�C, the thermal diffusion of solute is ex-tremely slow. Although radiation-enhanceddiffusion may significantly increase solutetransport, the scale of the microstructuralfeatures that develop is likely to be ex-tremely small, and are generally under theresolution of conventional techniques.

The atom probe is a particularly power-ful tool for the characterization of these ma-terials due to its high spatial resolution,light element sensitivity, and ability toquantify the composition of ultrafine(

�5nm) precipitates [2, 9]. The atom probetechnique has provided valuable informa-tion about ultrafine precipitates, soluteclustering, and segregation to lath andgrain boundaries. The size, shape, andnumber density of these ultrafine featuresare obtained from high magnification im-ages in the field ion microscope. The com-position of these features and the matrixhave been obtained from both the classicatom probe (i.e., the straight time-of-flight

236 M. K. Miller et al.

atom probe and the energy-compensatedinstrument) and from atomic level recon-struction of the positions of solute atomsobtained in the three-dimensional atomprobe (3DAP). The relatively low doses ofneutron radiation and the extremely smallvolume and activity of a typical field ionspecimen of these materials do not imposesignificant difficulties in the preparation ofspecimens if appropriate safety precautionsare followed [9].

Several other techniques have been usedin attempts to characterize the microstruc-tures of these materials with varying de-grees of success. The main alternatives toatom probe field ion microscopy are smallangle neutron scattering (SANS) and trans-mission electron microscopy. However,both of these techniques suffer from seriouslimitations in the characterization of theseferromagnetic materials. Although theSANS technique is able to gather informa-tion from a large volume of material, the in-terpretation of that data requires a validmodel, and the data are subject to multipleinterpretations as to the solute and vacancycontents of the features present. In addi-tion, SANS cannot reliably deconvolute thecontributions from the multiple types ofprecipitates present in these materials ordistinguish between intergranular or intra-granular precipitates. However, the resultsobtained in the atom probe may be used toprovide a reliable framework for the inter-pretation of SANS data. Transmission elec-tron microscopy of these types of steels isextremely sensitive to specimen prepara-tion artifacts, such as the redeposition ofcopper on the surface of the thin-foil speci-men and surface oxidation. Furthermore,these irradiation-induced clusters do notgenerate sufficient contrast for imaging inthe transmission electron microscope. Inaddition, quantification of the compositionof small features (

�5nm) in the analyticaltransmission electron microscope is ex-tremely difficult or impossible due to theirsize, the presence of significant composi-tion gradients around the features, and theinteraction volume of the electron beamand the thickness of the specimens.

In this paper, a review of the contribu-tions of the atom probe technique to the un-derstanding of the embrittlement of pres-sure vessel steels is presented. Atom probestudies have been performed on plate,forged components and welds from sur-veillance materials, materials obtainedfrom decommissioned reactors, and test re-actor-irradiated materials [9–32]. In addi-tion, unirradiated control material, long-term thermally aged materials, and modelalloys have also been examined [33–43].These characterizations have revealed thatthe microstructure of the matrix of com-mercial pressure vessel steels is extremelycomplex and involves many different typesof intragranular features including copperand phosphorus clusters and precipitates,carbides, nitrides, and oxides. In addition,atom probe studies have revealed a com-plex pattern of segregation of various sol-utes including phosphorus, nickel, manga-nese, molybdenum, and nitrogen to grainboundaries, and have also documented theextent of intergranular precipitation. Thesephenomena are a potential source of em-brittlement because solute segregation andintragranular and intergranular precipita-tion can significantly alter the mechanicalproperties of the matrix and grain bound-aries and even promote mechanisms suchas temper embrittlement. In addition, thehardening of the matrix from the formationof intragranular clusters and precipitateswill make the mechanical properties ofthe grain boundaries more important be-cause the steel will generally fail at theweakest link.

MICROSTRUCTURE AFTERNEUTRON IRRADIATION

Most of the early atom probe studies fo-cussed on Western pressure vessel steels.These steels have included an A302B plateand several A533B type steels and welds.Some typical compositions of these steelsare given in Table 1. In addition to thesesteels, a ferritic–bainitic 16MND5-type steelfrom the French surveillance program and

Understanding Pressure Vessel Steels 237

a vanadium-containing plate steel from theGundremmingen KRB-A reactor have beenexamined. Recently, atom probe character-izations have been performed on the signif-icantly different 15Kh2MFA Cr–Mo–V and15Kh2NMFA Ni–Cr–Mo–V pressure ves-sel steels used in Russian VVER 440 andVVER 1000-type nuclear reactors, respec-tively.

SETTING THE STAGE

The first atom probe study of neutron-irra-diated pressure vessel steels was performedby Miller and Brenner in 1981 on an A302Bplate steel from a surveillance capsule [10,11]. This steel had a composition (atom per-cent—at. %) of Fe-0.17% Cu-0.17% Ni-1.35% Mn-0.45% Si-0.29% Mo-0.12% Cr-0.08% Al-1.1%, C-0.02% P-0.03% N-0.04% S,and was irradiated for 8 years at 290

�C to afluence of 2.2

� 10 23n m

�2. After this expo-sure, the copper and phosphorus levels inthe matrix were found to be significantlyreduced to 0.01 at. % Cu and 0.005 at. % P,respectively. However, it was establishedlater that some preferential evaporationof copper had occurred during the analysis,and so these figures are probably a slightunderestimate of the actual levels in thematrix. A high density of ultrafine (

�

�10nm)darkly imaging copper-enriched featureswas observed in the matrix of the neutron-irradiated steel, as shown in Fig. 1. Thesecopper-enriched regions were enriched inother solutes including phosphorus andaluminum. In addition to the copper-enriched regions, some ultrafine phospho-

rus-enriched regions and some 1nm-diam-eter spherical and needle-shaped Mo2Cprecipitates were also observed, as shownin Fig. 2. Large Mo2C precipitates were alsoanalyzed in the unirradiated and irradiatedmaterials. The composition of one of theseprecipitates was reported as 56 at. % Mo-8% Fe-1% Cr-35% C. No solute segregationat the carbide matrix interface was ob-served. In contrast, segregation was ob-served at both cementite–ferrite interfacesand at grain boundaries in this A302B steel,as shown in Fig. 3 by the brightly imagingdecoration. Atom probe analysis revealedthat Mo, C, V, Cr, and Co were enriched atthe cementite–ferrite interface and thegrain boundaries.

These and some additional features havesubsequently been observed in many otherneutron-irradiated pressure vessel steelsand can be divided into four generalcategories: matrix composition, ultrafine(

�

�10nm) precipitates, coarse (

�

�10nm)precipitates, and segregation to bound-aries. These categories will be discussed indetail in the following sections.

MATRIX

The characterization of the composition ofthe matrix is important, as the solute con-centrations are required to provide a massbalance and also to establish the baselinefor the determination of the solute enrich-ment factors. The composition of the matrixis also required to establish the concentra-tions of the alloying elements as a functionof different steps in the heat treatment. It is

Table 1 Typical Nominal Compositions of Pressure Vessel Steels

Alloy Cu Ni Mn Si Mo Cr V Al C P S

A302B 0.17 0.17 1.35 0.45 0.29 0.12 — 0.08 1.1 0.02 0.04A533B 0.2–0.35 0.57 1.37 1.01 0.25 0.05 — — 0.55 0.011 0.02216MND5 0.078 0.53 1.26 0.63 0.22 0.17 0.021 0.05 0.74 0.021 0.009KRB-A 0.14 0.71 0.71 0.43 0.36 0.41 0.04 0.08 1.01 0.023 0.02115Kh2MFA 0.30 0.07 0.36 0.36 0.39 2.77 0.33 — 0.67 0.032 0.05215Kh2NMFA 0.05 1.15 0.41 0.49 0.34 2.49 0.11 — 0.79 0.016 0.009

The balance is iron and these data are presented in atomic percent.


also important that contributions from ul-trafine features are not included in the ma-trix composition so that these parameterscan be determined accurately. Several sta-tistical tools were developed specifically toassist in this process [2, 9].

Atom probe studies have revealed theimportance of the stress-relief treatment inhigh copper (

�

�0.2at.% Cu) weld materi-als. It has been established that the copperlevel that remains in the matrix in high cop-per alloys after the stress-relief treatmentand before neutron irradiation is stronglydependent on the annealing temperatureand the cooling rate to room temperature.The measured copper levels have been

found to be in good agreement with ther-modynamic predictions. These studieshave concluded that a lower annealingtemperature and a longer time is most ef-fective in minimizing the susceptibility ofthese materials to embrittlement due tocopper [19].

The copper level in the matrix after neu-tron irradiation has been measured in sev-eral pressure vessel steels, and the resultsare summarized in Table 2. It is evidentfrom these results that a significant deple-tion in the copper content occurs after neu-tron irradiation. There is a general trend atlow fluences that the copper level decreaseswith increasing fluence. The minimumlevel of copper in the matrix is approxi-mately 0.04 to 0.05at.% Cu. No significantdifference has been found between the dif-ferent types of steels. The copper levelsmeasured do not agree well with extrapola-tion of high-temperature thermal annealingdata or equilibrium thermodynamic pre-dictions in model iron–copper alloys. Thisdiscrepancy is not surprising, because theneutron irradiation increases the number ofpoint defects present in the matrix, andhence, changes the free energy of the sys-tem. This change will alter the position ofthe phase fields in the phase diagram. This

FIG. 1. Field ion micrograph of an ultrafine darkly im-aging copper-enriched feature in the matrix of theneutron-irradiated A302B plate steel.

FIG. 2. Field ion micrograph of 1nm-diameter spheri-cal and needle-shaped Mo2C precipitates in the matrixof the neutron-irradiated A302B plate steel.

FIG. 3. Field ion micrograph of a cementite precipitateat a grain boundary in a neutron-irradiated A302Bsteel plate. Both cementite–ferrite interfaces and fer-rite–ferrite interfaces are decorated by a thin film ofmolybdenum carbonitride precipitates.


effect has been observed in neutron-irradi-ated Fe–Cr alloys [44].

Composition variations of the other ele-ments in the alloy, particularly nickel, man-ganese, phosphorus, and silicon, arepresent, but are not as distinct as those ofcopper. In addition, the matrix compositionof the steels is always different from the al-loy composition due to the formation ofcoarse precipitates that depletes the matrixof the elements that are consumed in theseprecipitates (typically carbon, molybde-num, chromium, and vanadium). This ef-fect varies with the different types of steels.

ULTRAFINE CLUSTERS AND PRECIPITATES

The most common characteristic of the neu-tron-irradiated pressure vessel steels is thepresence of darkly imaging copper-enrichedfeatures, which are not observed in unirra-diated materials. Two different morpholo-gies have been observed. Most of these fea-tures were found to be roughly spherical,but a few examples of elongated ribbonshave also been observed. It has been specu-lated that the ribbon morphology is a resultof copper segregation to a moving disloca-tion. The size of these spherical featureswas found to be between

�1 and 4nm. Sev-

eral other solutes have been detected in thecopper-enriched features. It has alreadybeen mentioned that phosphorus and alu-minum were observed in A302B steel.Phosphorus has also been detected in alarge proportion of these features in othersteels.

The microstructural development in neu-tron-irradiated weldments was evaluatedfor materials from H. B. Robinson II andfrom a Rolls Royce test program by Burkeand Brenner using APFIM [13, 14]. In addi-tion to the identification of copper-enrichedsolute clusters in the irradiated microstruc-ture, Burke and Brenner also provided thefirst evidence of Mn and Ni enrichmentsassociated with these irradiation-inducedclusters. These irradiation-induced featureswere termed “solute-rich clusters” becausethey contained significant amounts of cop-per, manganese, nickel, silicon, and occa-sionally phosphorus, as well as high levelsof iron. The H. B. Robinson II data also in-cluded examples that suggested that the ir-radiation-induced clusters had a copper-enriched core with manganese and nickelsurrounding the core. Burke and Brennerobserved that manganese and nickel werealso enriched in these features in a highnickel and high manganese neutron-irradi-

Table 2 Comparison of the Copper Contents in the Matrix of a Variety of Neutron-Irradiated Pressure-Vessel Steels

Reference SteelFluencen m

�2Nominalat. % Cu

Matrixat. % Cu

[17, 18] A533-B 2

� 1021 0.13 0.10

� 0.03[17, 18] KRB-A 8.4 � 1021 0.14 0.13[17, 18] KRB-A 2

� 1022 0.14 0.12[17, 18] KRB-A 2.7

� 1022 0.14 0.11

� 0.01[23] Chooz 4.7 � 1022 0.08 0.05

� 0.01[34] Weld 6.6

� 1022 0.14 0.13

� 0.04[17, 18] KRB-A 8.5

� 1022 0.14 0.09

� 0.01[33] Weld 1.1 � 1023 0.18 to 0.28 0.06

� 0.01[13, 14] Rolls Royce 2

� 1023 0.2 0.1Weld 2.0

� 1023 0.27 0.06

� 0.01[23] Chooz 2.5 � 1023 0.08 0.04

� 0.01[34] Weld 3.5

� 1023 0.14 0.05

� 0.01[23] Chooz 6.6

� 1023 0.08 0.03

� 0.04[23] Chooz 1.2 � 1024 0.08 0.04

� 0.01


ated weld from a Rolls Royce test program(Fe-0.2 at. % Cu-1.5% Ni-1.6% Mn-0.9% Si-0.2% Mo) [14]. Manganese, nickel, and sili-con enrichments have also been observedin A533B and other steels. Atom probecomposition profiles revealed that the spa-tial extent of the manganese, nickel, and sil-icon enrichments was generally slightlylarger than that of the copper, as shown inFig. 4. A set of atom maps of some of thesefeatures in a weld (Fe-0.27 at. % Cu-1.58%Mn-0.57% Ni-0.34% Mo-0.27% Cr-0.58% Si-0.003% V-0.45% C-0.009% P-0.009% S) afterneutron irradiation to a fluence of 2

� 1021nm

�2 (E

� 1MeV) is shown in Fig. 5. In the16MND5 steel from the Chooz A reactor,Pareige et al. [30] found that the numberdensity of 3 to 4nm diameter copper-enriched clusters increased from 3.3

� 1023

to 11

� 1023m

�3 with increasing fluence 2.5to 16

� 1023n m

�2 (E

� 1MeV). These resultswere found to be in good agreement withSANS data. The composition of the clusterswas found to be Fe-

�1 at % Cu-

�4% Ni-

�5% Mn-

�3% Si did not appear to changesignificantly with increasing fluence.

The copper levels measured in these fea-tures (

�1 to �30% Cu) in neutron-irradi-ated materials are always substantiallybelow the equilibrium copper level of the

� copper precipitate. Copper invariablyhas the highest enrichment factor over thematrix level, although the absolute copperlevel measured may be less than the abso-lute levels of manganese and nickel in theclusters in some cases. This effect is due tothe significantly lower initial level of cop-per in the matrix of the steel compared toeither nickel or manganese after the stress-relief treatment. It was also evident thatthese features contained significant levels(�70%) of iron.

These features have been referred to ascopper-enriched clusters, zones, precipi-tates, and copper-stabilized microvoids.One of the main difficulties in establishingthe true identity of these small features isthe lack of proper definitions of what theseterms mean at ultrahigh resolution. A clus-ter is simply an aggregate of solute atomswithin the matrix with no distinct interface,whereas a precipitate is generally definedas a region of solute with a well-definedcrystal structure and interface with the ma-trix. Copper is a substitutional atom in thebody centered cubic (or body centered tet-ragonal) iron lattice, and copper precipi-tates (� Cu) in iron are face-centered cubic(fcc). It is thought that the copper forms anintermediate body-centered cubic precipi-tate as a precursor to the fcc structure [45].A high-resolution transmission electronmicroscopy analysis of precipitates in agedFe-1.3at.% Cu and Fe-1.3at.% Cu-1.1% Nialloys by Othen et al. [46] indicated that thestructure of fine precipitates formed duringthermal aging is 9R. Similar structures havenot yet been observed in pressure vesselsteels. The cluster vs. precipitate identifica-tion is further complicated when other sol-utes and composition gradients are in-volved. It should be noted that a diffusesolute-enriched cluster or well-defined pre-cipitates containing larger numbers of cop-per atoms would both impede the motionof dislocations through the microstructure,and therefore, have an effect on the me-chanical properties. Although vacancieshave been imaged in the field ion micro-scope in pure elements and ordered com-pounds [2], it is not possible to establish the

FIG. 4. Atom-probe composition profiles across a cop-per-enriched feature in a neutron-irradiated A533Bsubmerged arc weld showing that the spatial extent ofthe manganese, nickel, and silicon enrichments wasgenerally slightly larger than that of the copper.


presence or concentration of vacancies inthese features in these complex steels fromeither their appearance in the images in thefield ion microscope or from compositionaldeterminations in the atom probe.

Atom-probe analyses have also revealedthe presence of phosphorus clusters. Theycan sometimes be distinguished from thedarkly imaging copper features by thepresence of some bright spots on the darklyimaging background, as shown in Fig. 6.These �1nm diameter clusters are gener-ally observed only in high phosphorussteels or when the copper level is extremelylow. These phosphorus clusters are oftenassociated with nickel enrichments.

Another common microstructural featureis the ultrafine, brightly imaging sphericaland needle-shaped molybdenum carbideand disc-shaped molybdenum nitride pre-cipitates. Some of these precipitates werefound to be only �1nm in diameter. Theseprecipitates have been observed in A302Bplate [11] and A533B plate and weld mate-rial [9, 15, 17, 18, 24] and in material from

the Gundremmingen KRB-A reactor [17,24]. These precipitates have been observedin both unirradiated and neutron-irradiatedmaterials. The distribution of these featuresin the steel was generally found to be rela-tively inhomogeneous. Their number den-sity was typically slightly lower than that of

FIG. 5. Atom maps of the copper, manganese, nickel, and silicon distribution in a submerged arc weld neutron-irradiated to a fluence of 2 � 1023 n m�2 (E � 1MeV) obtained in a three-dimensional atom probe. Note the solute-enriched regions that result from neutron irradiation.

FIG. 6. Field ion micrograph of a phosphorus clusterin a neutron-irradiated VVER 440 pressure vesselsteel.


the copper-enriched features. Insufficientdata exist to conclusively establish whetherthere is a change in the number density ofthese features after neutron irradiation andwhether these precipitates are responsiblefor some of the embrittlement observedduring neutron irradiation. Isolated exam-ples of small darkly imaging oxide precipi-tates have also been observed [19].

Pareige et al. [32] also detected decorationof molybdenum and phosphorus atoms atdislocations in an unirradiated forging, asshown by the bright spots decorating thedislocation in Fig. 7. No attempts have beenmade to study the possible segregation todislocation or interstitial loops in the irradi-ated materials.

COARSE PRECIPITATES

Coarse precipitates are present in low num-ber densities in the interior of the grains, atgrain boundaries, and on dislocations.These coarse precipitates are present inboth the unirradiated and neutron-irradi-ated materials. The compositions of theseprecipitates are important parameters indetermining the location and levels of allthe solutes in the steel. It should be notedthat the size of these coarse precipitates are

normally larger than the field of view of thefield ion microscope (typically �100nm), sosize measurements are generally estimatedfrom the transmission electron microscope.

A 200 to 600nm fcc copper–manganeseprecipitate was observed on a grain bound-ary in a neutron-irradiated A533B sub-merged arc weld [15, 18], as shown in Fig.8. The composition of the central portion ofthis coarse precipitate was determined tobe 83.6 � 1.2 at. % Cu, 14.5 � 1.1% Mn, 1.5 �0.4% Ni, and 0.3 � 0.2% Fe. The precipi-tate–matrix interface was also found to besubstantially enriched in manganese andnickel, as shown in Fig. 9. These coarsecopper–manganese precipitates are oftenformed at grain boundaries and disloca-tions during the initial treatment of thealloy or the stress-relief treatment and sub-sequent cooling. They account for the dif-ference between the nominal copper con-tent of the alloy and the measured matrixcopper level after the stress relief treatment.

Coarse (�0.1 to 0.5m) darkly imagingcementite precipitates have been observedin A302B, as noted above, and in A533Band the Gundremmingen KRB-A and theRussian VVER 440 and 1000 steels. A sum-mary of the compositions of these cement-ite precipitates is given in Table 3. The com-

FIG. 7. Field ion micrograph of molybdenum andphosphorus atoms decorating a dislocation in an unir-radiated forging.

FIG. 8. Field ion micrograph of a darkly imaging cop-per–manganese precipitate at a grain boundary (ar-rowed) in a neutron-irradiated A533B submerged arcweld.


position of the cementite was found todepend on the composition of the steel. Theenrichment of solutes in these precipitatesaccounts for some of the depletion of theseelements in the matrix analyses.

Coarse brightly imaging Mo2C precipi-tates have been observed in A533B sub-merged arc weld, as shown in Fig. 10(a). Asummary of the compositions of these pre-cipitates is given in Table 4. Phosphorus en-richment at the carbide–matrix interfacehas been observed. The enrichment of sol-utes in these precipitates accounts for someof the depletion of these elements in thematrix analyses.

Smaller brightly imaging roughly spheri-cal V(C,N) precipitates have been charac-terized in both vanadium-containing Gun-dremmingen KRB-A and the Russian VVER440 and 1000 steels. A �10nm diameterV(C,N) precipitate in the ferrite matrix ofthe Gundremmingen KRB-A reactor steel isshown in Fig. 10(b). This precipitate con-tained virtually no molybdenum, manga-nese, or chromium. Phosphorus was de-tected in the precipitate and in the ferritematrix adjacent to the precipitate. Similarprecipitates were observed in the matrixand at boundaries in the Russian VVER 440and 1000 steels [25–29], as shown in Figs.10(c) and (d). The average composition ofthese precipitates in a neutron-irradiatedhigh-phosphorus weld (Fe-1.7 at. % Cr-0.37% Mo-0.2% V-0.95% Mn-0.14% Ni-0.69%Si-0.23% C-0.12% Cu-0.058% P) was deter-mined to be 51.3 � 0.9 at. % V-18.8 � 0.7%C-22.1 � 0.7% N-4.9 � 0.4% Cr-2.3 � 0.3%Mo-0.36 � 0.05% Fe-0.07 � 0.05% B-0.03 �0.03% P.

Blocky M7C3 carbides have also been ob-served in the Russian VVER 440 and 1000steels [26, 27]. An atom probe analysis ofone of these darkly imaging precipitates inthe VVER 440 steel yielded a compositionof 37.6 � 1.7 at. % Cr-22.7 � 1.4% Fe-3.5 �0.6% Mo-2.7 � 0.6% V-0.6 � 0.3% Mn-33.0 �1.6% C. No nitrogen, nickel, silicon, or cop-per was detected in this precipitate. Trans-mission electron microscopy of these pre-

FIG. 9. Atom-probe composition profile from the cen-ter of the copper–manganese precipitates shown inFig. 8 into the ferrite matrix. The interface was alsofound to be substantially enriched in manganese andnickel.

Table 3 Composition of Cementite Precipitates as Measured by the Atom Probe in Different Pressure Vessel Steels

Solute[Ref.]

A302B2.2 � 1023

n m�2 [10, 11]

16MND51.4 � 1024

n m�2 [20]

16MND5�28 h at

600°C [20]

SA-533 Plate29 h at

593–621°C [32]

SA-533 Plate93,000 h at280°C [32]

Fe balance 60.0 � 1.1 61.5 � 0.7 64.0 � 0.8 61.1 � 0.1Mn 9.3 9.4 � 0.7 10.4 � 0.5 8.7 � 0.5 11.9 � 0.7Mo 2.6 to 16 1.65 � 0.3 1.3 � 0.2 1.2 � 0.2 0.9 � 0.2Cr 1.3 2.6 � 0.4 1.7 � 0.2 0.5 � 0.1 —Ni — 0.7 � 0.2 0.7 � 0.1 — 0.2 � 0.1Si — 0.11 � 0.07 0.07 � 0.04 — —C 25a 25.5 � 1.0 22.3 � 0.7 25.6 � 0.7 25.4 � 0.9

Data are in atomic percent.a Normalized to 25% C.


cipitates revealed a characteristic faultedstructure and a size range of 60 to 300nmand 60 to 500nm in the VVER 440 and 1000steels, respectively.

SEGREGATION TO BOUNDARIES

Recently, the susceptibility of Russian pres-sure vessel steels to temper embrittlementhas become a serious concern for continuedsafe operation. This concern is primarilydue to the high phosphorus levels that

have been measured in the bulk materialsand the observation of intergranular frac-ture in some materials. Although the maxi-mum permissible levels for phosphorusand sulphur are 0.020 wt. % in the15Kh2NMFA steel and 0.025 wt. % in the15Kh2MFA steel, a number of cases havebeen found where these values were ex-ceeded.

A general characteristic of the steels isthat the grain and lath boundaries are deco-rated with solutes, and this is manifested assome brightly imaging spots in the field ion

FIG. 10. Field ion micrographs of (a) a brightly imaging Mo2C precipitate in a neutron-irradiated A533B sub-merged arc weld, (b) a �10nm diameter V(C,N) precipitate in the ferrite matrix of the Gundremmingen KRB-A re-actor steel, (c) V(C,N) precipitates in the matrix, and (d) at a boundary in Russian VVER 440 and 1,000 steels.


micrographs, as shown in Fig. 3 for anA302B steel and Figs. 11(a) and (b) for anA533B and a Russian VVER 440 steel. Atomprobe analysis has established that thesebrightly imaging spots are a thin semicon-tinuous film of molybdenum carbonitrideprecipitates. In addition, the boundariesgenerally exhibit phosphorus segregation.

The amount of solute segregation at in-terfaces may be estimated with the use of amethod based on the Gibbsian interfacialexcess [27, 47]. The Gibbsian interfacial ex-cess of element i, i, may be determined di-rectly from an atom probe analysis for allelements with the use of the following rela-tionships:

(1)

where Ni(excess) is the excess number of soluteatoms associated with the interface, Ni is thetotal number of solute atoms in the volumeanalyzed, Ni(�) and Ni(�), are the number ofsolute atoms in the two adjoining regions �and � either side of the dividing surface, andA is the interfacial area over which the inter-facial excess is determined. This methodprovides a fundamental estimate of the levelof segregant at an interface.

The phosphorus coverage measured atsome boundaries in both an A533B andRussian steels is summarized in Table 5[27–29]. It is evident that the phosphorus

Γi Ni excess( ) A Ni Ni α( ) Ni β( ) A⁄––=⁄=

coverage at the boundary increases withneutron irradiation and with the initialphosphorus level in the steel. With the ex-ception of the neutron-irradiated Weld 37material, the fracture surface of all thesematerials was found to be transgranular. Ithas been suggested that because the levelof phosphorus at thee boundaries is signifi-cantly higher than that required to produceintergranular failure in binary Fe–P alloys,the carbon or the carbonitride film at theboundaries is playing an important role insuppressing a change in failure mode.

Because the parameters for phosphorussegregation are reasonably well establishedin steel [48, 49], the atom-probe results fromthe unirradiated materials have been com-pared with predictions from the McLeanmodel of equilibrium segregation [50]. Afree energy change of Gp � �56,700 �12.4 T was used in these calculations [49]. Acomparison of the measured and predictedvalues for some boundaries in unirradiatedpressure vessel steels is shown in Table 5[27–29]. Excellent agreement between themeasured and predicted values was found.It was also noted that these model predic-tions indicate that the phosphorus coverageat the boundary will increase significantlywith postirradiation annealing, and so itwas recommended that this method shouldbe avoided in high phosphorus steels.

Table 4 Composition of Some Typical Coarse Mo2C Precipitates as Measured by the Atom Probe in Different Pressure Vessel Steels

Element

A302B2.2 � 1023

n m�2

A533B1 � 1023

n m�2

16MND51.4 � 1024

n m�2

16MND51.4 � 1024

n m�2

Mo 56 63.6 � 1.4 61.0 � 2.1 63.0 � 6.3Fe 8 2.4 � 0.4 — 1.9 � 1.7Cr 1 — 3.2 � 0.6 —Mn — 2.1 � 0.4 1.8 � 0.5 —C 35 31.4 � 1.4 30.8 � 1.9 32.7 � 6.1O — 0.26 � 0.15 — —N — 0.17 � 0.12 — —P — 0.09 � 0.09 — 1.9 � 1.7

Data are in atomic percent.


POSTIRRADIATION ANNEALING STUDIES

If embrittlement can be reduced or elimi-nated, it should be possible to extend theservice life of a nuclear reactor with consid-erable returns in economics and environ-mental impact. Several studies of mechani-cal properties have demonstrated that theembrittlement may be dramatically re-duced or even eliminated by annealing thepressure vessel at a low temperature (343 to454�C) [51–55]. The upper temperature ofthe annealing treatment is limited by the

magnitude of the stresses that are gener-ated in the vessel. These low temperaturepostirradiation annealing treatments aredesigned to reduce the number of intra-granular precipitates or hardening centersin the material.

Pareige et al. [32] characterized two typesof materials from the B & W Master Inte-grated Reactor Vessel Surveillance Pro-gram. The first material was a SA-508,class 2 forging steel, with a nominal coppercontent of 0.02 wt. %. The second materialwas a typical Mn–Mo–Ni weld wire�Linde

Table 5 Summary of the Average Phosphorus Coverage at Lath and Grain Boundaries as Measured in the Atom Probe and Predicted by the McLean Model of Equilibrium Segregation

Material

Fluence(n m�2)

E � 1MeVP Content

(at. %)i

atoms cm�2Coverage,

monolayersPrediction,monolayers

A533B Unirradiated 0.011 5.7 � 1013 3% 5%1 � 1023 0.011 1.8 � 1014 10%

VVER 440 Unirradiated 0.032 8.2 � 1013 5% 7%VVER 1000 Unirradiated 0.016 1.3 � 1014 8% 5%Weld 28 Unirradiated 0.045 2.0 � 1014 11% 10%

1 � 1023 0.045 4.0 � 1014 24%Weld 37 Unirradiated 0.058 2.8 � 1014 13% 12%

1 � 1024 0.058 9.0 � 1014 54%

FIG. 11. Field ion micrographs of decorated boundaries in (a) an A533B submerged arc weld, and (b) a RussianVVER 440 steel.


80 flux submerged arc weld, representativeof the materials used to fabricate the belt-line shell course regions of the OconeeUnit-3 and Arkansas Unit-1 vessels. Thecopper content in the matrix of the weldmaterial was depleted from the bulk levelof 0.24at.% Cu to 0.14 � 0.03at.% Cu afterthe stress-relief heat treatment of 29 h at593–621�C followed by a furnace cool to310�C at a rate of �8� C per h. The copperlevel in the matrix was reduced to 0.13 �0.04 and 0.05 � 0.01at.% Cu after neutronirradiation to fluences of 0.66 and 3.5 �1023n m�2 (E � 1MeV), respectively. An-nealing the higher fluence material for 168h at 454�C resulted in a further reductionof the matrix copper content to 0.04 �0.02at.% Cu. Annealing the higher fluencematerial for 29 h at 610�C was found to in-crease the copper level in the matrix to0.17 � 0.04 at.% Cu.

An atom probe characterization of a 200-mm-thick section of the beltline submergedarc weld (designated a Babcock and WilcoxWF-70 weld) of the Midland Unit I pressur-ized water reactor was performed by Millerand Russell [33]. This weld was fabricatedwith the use of copper-coated weldingwires and Linde 80 flux, and is known to bea low upper shelf (LUS) energy, high cop-per weld. This atom probe study has dem-onstrated that the copper in solution in thematrix of an unirradiated pressure vesselsteel weld after a standard postweld heattreatment of 22.5 h at 607 � 14�C was 0.119 �0.007at.% Cu. This value was found to de-crease to 0.088 � 0.012at.% Cu after anneal-ing for 168 h at 454�C and to 0.058 � 0.008at.% Cu after neutron irradiation in a testreactor to a fluence of 1.1 � 1023n m�2 (E �1MeV) at a temperature of 288�C. An addi-tional decrease in copper level to 0.050 �0.010at.% Cu was measured by annealingthe neutron irradiated material for 168 h at454�C. These values and trends are in goodagreement to previous results from a Bab-cock and Wilcox weld [34].

Both studies concluded that annealingneutron-irradiated pressure vessel steelwelds appeared to reduce embrittlementby coarsening the copper-enriched precipi-

tates and decreasing the matrix copper con-tent. Because the measured copper levelswere significantly lower than those in thematrix after the stress-relief treatment,these results predict that the amount ofreirradiation embrittlement due to copperprecipitation should be significantly lowerthan during the initial neutron irradiation.These atom probe studies have also shownthat the susceptibility of pressure vesselsteels to embrittlement could be reduced byannealing the steel at an intermediate tem-perature of 450 to 500�C after the stress-relief treatment and prior to service.

The 16MND5 steel from the Chooz A re-actor was also examined by Pareige et al.[31] after annealing for 2 and 100 h at450�C. The size of the copper clusters wasfound to decrease to 1–2nm diameter afterannealing for 2 h and then increase to 3–4nm diameter after annealing for 100 h. Thecopper level in the copper-enriched clusterswas also found to increase significantlyfrom 1% Cu after the irradiation to 30% Cuafter 2 h and to 60% Cu after 100 h at 450�C.In conjunction with hardness and SANS re-sults, it was concluded that the neutron-induced copper clusters partially dissolvein that the Si, Ni, and Mn return to the solidsolution on annealing and then the remain-ing copper-enriched cluster acts as a nucleifor the pure copper precipitates.

LONG TERM THERMAL AGING

To study the potential effects of long termthermal aging, atom probe characteriza-tions of the microstructure of as-fabricatedand long-term thermally aged (�100,000 hat 280�C) surveillance materials from com-mercial reactor pressure vessel steels havebeen performed by Pareige et al. [34]. Thismicrostructural study focused on the quan-tification of the compositions of the matrixand carbides. The atom probe results indi-cate that there was no significant micro-structural evolution after a long-term ther-mal exposure in typical Mn–Mo–Ni weldwire, Linde 80 flux submerged arc weld,


SA-533 Grade B plate, and SA-508 Class 2forging materials.

The average copper solute concentrationdetermined in the ferritic matrix after ther-mal aging was found to be consistent withthe nominal level for the two plates and theforging materials. Copper depletion from0.24at.% Cu to 0.17at.% Cu was observed inthe matrices of welds of both unaged andthermally aged materials. This depletionoccurred during the stress-relief heat treat-ment that was performed at a temperaturebetween 593 and 621�C and subsequentfurnace cool rather than the thermal agingtreatment.

Molybdenum-containing carbides werefrequently observed in both unaged andthermally aged specimens. The size of theobserved carbides was determined to bebetween 5 and 20nm. The small carbideswere found to have a disc, needle, or spher-ical shape, whereas the larger precipitatesare generally spherical. The Mo:C ratio, inboth the unaged and thermally aged mate-rials, was close to that of Mo2C. Some evi-dence of molybdenum atoms was found inthe vicinity of dislocations and it was some-times associated with carbon and phospho-rus. The grain boundaries were decoratedwith a 12nm-thick layer of molybdenumcarbide precipitates.

The compositions of the cementite pre-cipitates were found to be in good agree-ment with ThermocalcTM predictions formaterials aged at 593–620�C. This agree-ment indicates that long-term thermal ag-ing has no significant impact on the evolu-tion of the microchemistry of cementitecarbides. The cementite compositions weresimilar to those determined from cementiteprecipitates in the Chooz A pressure vesselsteel and weld metal.

It was concluded that the atom probecomparisons of materials under these con-ditions was consistent with the measuredmechanical properties, such as the ductile-to-brittle transition temperature measuredin Charpy V-notch impact tests, and that nosignificant changes in either the micro-structure or the mechanical properties hasbeen observed.

MODEL ALLOYS

Model alloys have been used in parallelwith pressure vessel steels to evaluate theeffect of radiation. These studies can be di-vided in two different categories, namely,studies of model alloys after neutron irradi-ation treatments representative of that in anuclear reactor, and studies of model alloysafter higher temperature isothermal agingtreatments.

The first study on model alloys that is re-lated to the embrittlement of the pressurevessel was performed by Goodman et al. in1973 [35]. In this study, the size, numberdensity, and compositions of the precipi-tates that formed in a binary Fe-1.4at.% Cumodel alloy during thermal aging at 500�Cwere characterized by field ion microscopyand atom probe analysis. In addition, somesupporting transmission electron micros-copy was performed on the material afterthe longer aging times. The mean diameterof the particles at peak strength was deter-mined to be 2.4nm. The mean particle sizewas found to follow a t0.5 relationship withaging time, t, and a copper diffusivity of�2 � 1021m2 s�1 was estimated. The num-ber density of the particles when they werefirst visible was of the order of 1024 m�3,and this value rapidly decreased after thepeak strength of the alloy was reached. Theaverage copper content of the initial precip-itates (�47% Cu after 1 h aging) was foundto be significantly lower than that of theequilibrium � phase, whereas, after agingfor 120 h, they were consistent (93% Cu)with that of the equilibrium phase. The re-sults of Worrall et al. [36] on similar alloyswere in reasonable agreement. A composi-tion of 95% Cu was measured in a later 3Datom probe study of one precipitate in anFe-1.2% Cu-1.4% Ni alloy aged close to thepeak hardness (2 h at 550�C) [37]. However,there is significant confusion about the truesize of this precipitate (2 or 4 to 5nm),which makes comparison with the previ-ous studies difficult [19].

In 1978, Brenner et al. [38] detected veryfine darkly imaging features in a field ionmicroscopy study of an Fe-0.34% Cu model


alloy after neutron irradiation to a fluenceof 3 � 1023n m�2 at a temperature of 290�C.The size of these generally spherical fea-tures ranged from 0.4 to 1.6nm, with a meandiameter of 0.6nm, and were present at anumber density of 8 � 1017cm�3. Althoughno atom probe microchemical analyses wereperformed in this study to differentiate be-tween microvoids and copper-enriched pre-cipitates, Brenner et al. tentatively identi-fied these features as copper-stabilizedmicrovoids, and concluded that they wereresponsible for radiation-induced embrit-tlement of copper-containing ferritic steels.

A comparison between a thermally agedmodel Fe-1.1at.% Cu-1.4% Ni alloy and aneutron-irradiated Fe-0.22at.% Cu-0.70%Ni model steel was performed by Pareige etal. [39]. This study revealed that the com-position of the copper clusters was differ-ent in the model alloy than in the pressurevessel steel. The low temperature coppersolubilities have been determined in thesame Fe-1.1at.% Cu-1.4% Ni alloy by Milleret al. [40] after long term isothermal agingat 300, 400, 500, 550, and 600�C. The resultsare summarized in Table 6. A small deple-tion in copper was found in the matrix afteraging for 4,000 h at 300�C, and significantlylarger depletions were observed at the highertemperatures. The kinetics of copper pre-cipitation was found to have an activationenergy of 250kJ mol�1. An atom map of thecopper distribution in this material agedfor 10,000 h at 300�C is shown in Fig. 12.Some small copper-enriched regions are ev-ident, indicating that phase separation canoccur in highly supersaturated Fe–Cu–Nialloys at these low temperatures.

Pareige [41] has also characterized an Fe-1.4% Cu model alloy after neutron irradia-tion to a fluence of 5.5 � 1023n m�2 at a tem-perature of 290�C, and a flux of 2.8 � 1017nm�2s�1. A copper atom map showing ahigh density (2 � 1024m�3) of 2 to 4nm di-ameter copper precipitates obtained fromthis material in the tomographic atomprobe [41] is shown in Fig. 13(a). In thistype of representation, each dot indicatesthe location of an individual copper atom,and the copper precipitates are evident

from the local high density of spots. A com-position profile through one of the copperprecipitates and the distribution of themaximum copper concentrations measuredin these precipitates are shown in Fig. 13(b)and (c), respectively. It is evident fromthese data that there is a significant distri-bution in the copper level from one precipi-tate to another in this neutron-irradiatedmaterial, and that the copper level is signif-icantly lower (0.04% Cu) than that pre-dicted from materials thermally aged athigher temperatures. Pareige also studiedthe same material after electron irradiationat the same temperature to a fluence of 9 �1023e m�2 with a flux of 4 � 1017e m�2 s�1,and found 2nm-diameter precipitates con-taining 95% Cu at a number density of 1024

m�3. This indicates that, even if the NRTdisplacements per atom are not the same(5.4 � 10�5 and 7.5 � 10�2 for electron andneutron, respectively), the phase transfor-mation under these irradiation conditionsdoes not follow the same path.

Some model pressure vessel steels werealso characterized by Miller et al. [42, 43] in1987. In this systematic study, five modelalloys were produced from a split meltwith a base composition representative of a

Table 6 Change in Copper Content of the Matrix of an Fe-1.1% Cu, 1.4% Ni Model Alloy as a Function ofAging Time and Temperature

Time(h)

Copper concentration, at. %

300°C 400°C 500°C

0.5 — — 0.47 � 0.111 — 0.92 � 0.08 0.31 � 0.104 — — 0.13 � 0.03

10 — 0.91 � 0.09 0.13 � 0.03100 — 0.54 � 0.08 0.073 � 0.02168 0.93 � 0.06 — —

1,000 0.82 � 0.06 0.12 � 0.03 0.082 � 0.034,000 0.80 � 0.05 0.085 � 0.01 —

550�C 600�C 850�C AQ)

5 — — 0.91 � 0.041,000 0.12 � 0.03 0.18 � 0.03 —


low manganese (0.01at.% Mn) pressurevessel steel and with low and high levels ofcopper (�0.008 and 0.22at.% Cu), nickel(0.011 and �0.67% Ni), and phosphorus(�0.007 and 0.041at.% P). These materialswere examined after neutron irradiation for1600 h to a fluence of 4.6 � 1023n m�2 (E �1MeV) at a temperature of 288�C. The ma-trix copper level was found to be signifi-cantly depleted in the 0.22at.% Cu steelsafter neutron irradiation (0.03at.% Cu inthe FeCu, 0.07% Cu in the Fe–Cu–Ni and0.08% Cu in the Fe–Cu–Ni–P alloy) com-pared to the 0.19 at. % Cu measured in aFe–Cu alloy annealed for 1600 h at 288�C.No clusters or precipitates were observedin the thermally aged alloy. A high numberdensity of ultrafine (�2nm) darkly imagingcopper-enriched clusters were observed inall neutron-irradiated copper-containingsteels. In the Fe–Cu–Ni alloy, the clusterscontained approximately 50% Cu, andwere also enriched in nickel. Similar resultswere observed in an irradiated Fe-0.31at.%Cu-0.51% Ni-0.46% C alloy. The maximumphosphorus-to-copper ratio in the phos-phorus-free Fe–Cu and Fe–Cu–Ni alloyswas �2%. However, in the Fe–Cu–Ni–P al-loy, phosphorus-to-copper ratio of the clus-ters was found to be between 25 and 50% in�30% of the cases, indicating a copper–phosphide cluster or precipitate. The nickellevel was also found to be higher in the vi-cinity of these clusters. In the copper-freeFe–Ni–P and Fe–Ni alloys, several ultrafinespherical phosphorus-rich clusters weredetected. The number density of the clus-ters in the phosphorus-containing alloywas significantly larger due to the higher

phosphorus content of the alloy. Bothnickel and carbon were found to be associ-ated with these phosphorus clusters.

FIG. 12. Copper atom map in an Fe-1.1%Cu-1.4% Ni alloy aged for 10,000 h at 300�C. Small copper-enriched re-gions are evident. Box is 13 � 13 � 101nm. Individual random copper atoms have been omitted for clarity.

FIG. 13. (a) Copper atom map (15 � 15 � 30nm), (b)composition profile through a 2.5nm-diameter copperprecipitate, and (c) the distribution of the copper con-tent of the copper precipitates from a neutron-irradi-ated Fe-1.4% Cu alloy (T � 290�C, dose � 5.5 1023

n m�2, flux � 2.8 1017 n m�2 s�1).


SUMMARY

Atom probe studies have provided the firstdirect evidence of copper-enriched clustersand phosphorus clusters in the matrix ofneutron-irradiated pressure vessel steels.The copper-enriched clusters were alsofound to be enriched in other solutes partic-ularly nickel, manganese, phosphorus, andsilicon. In addition, several types of intra-granular and intergranular precipitateshave been characterized in these materials.Atom probe studies have also revealed acomplex pattern of segregation of varioussolutes including phosphorus, nickel, andmanganese to grain and lath boundaries insome pressure vessel steels. Furthermore,atom probe analysis has provided the onlytrue quantitative measurement of the ma-trix solute content so that the true coppercontent of the matrix prior to irradiationand after irradiation can be determined.These data have demonstrated the impor-tance of the stress-relief temperature forweldments in reducing the amount of cop-per available for cluster�precipitate forma-tion in high-copper steels. Also, the atomprobe has been able to provide unique mi-crochemical data that have been very im-portant in the interpretation of results ob-tained from other analytical techniquessuch as small angle neutron scattering(SANS).

The information obtained by the atomprobe technique has provided experimen-tal evidence that has led to a basic under-standing of the causes of embrittlement ofthese pressure vessels and the mechanismsthat operate during postirradiation anneal-ing treatments.

The authors wish to thank S. P. Grant for hiscontinued interest and encouragement of atomprobe studies of RPV materials. This researchwas sponsored by the Division of MaterialsSciences, U.S. Department of Energy, undercontract DE-AC05-96OR22464 with LockheedMartin Energy Research Corp., by the Office ofNuclear Regulatory Research, U.S. NuclearRegulatory Commission under inter-agencyagreement DOE 1886-N695-3W with the U.S.

Department of Energy, and through the SHaREProgram under contract DE-AC05-76OR00033with Oak Ridge Associated Universities. Thisresearch was conducted utilizing the SharedResearch Equipment User Program facilities atORNL. Studies of the French pressure vesselsteels and some model alloys were performedat the “Equipe de Sonde Atomique et Micro-structures” of the GMP-UMR CNRS 6634 ofthe Rouen University and were financiallysupported by Electricité de France DépartementEtude des Matériaux” under EDF�CNRS con-tracts.

References

1. E. W. Muller, J. A. Panitz, and S. B. McLane: Theatom-probe field ion microscope. Rev. Sci. Instrum.39:83–86 (1968).

2. M. K. Miller, A. Cerezo, M. G. Hetherington, andG. D. W. Smith: Atom Probe Field Ion Microscopy,Oxford University Press, Oxford, UK (1996).

3. G. R. Odette and G. E. Lucas: Irradiation embrit-tlement of reactor pressure vessel steels: Mecha-nisms, models and data correlations. In RadiationEmbrittlement of Nuclear Reactor Pressure VesselSteels; An International Review (second volume)ASTM STP 909, L. E. Steele, ed., American Societyfor Testing and Materials, Philadelphia, PA pp.206–241 (1986).

4. Radiation Embrittlement of Nuclear Reactor PressureVessel Steels, An International Review, ASTM STP1170, L. E. Steele, ed., American Society for Testingand Materials, Philadelphia, PA (1993).

5. N. N. Alekseenko, A. Amaev, I. Gorynin, and V. A.Nikolaev: Radiation Damage of Nuclear Power PlantPressure Vessel Steels, American Nuclear Society,La Grange Park, IL (1997).

6. J. R. Hawthorne: Exploratory Studies of Element In-teractions and Composition Dependencies in RadiationSensitivity Development, USNRC Report NUREG�

CR-4437 (MEA-2113) (November 1985).

7. G R. Odette, C. L. Liu, and B. D. Wirth: On thecomposition and structure of nanoprecipitates inirradiated pressure vessel steels. In MicrostructureEvolution During Irradiation, Symposium Proceed-ings, Vol. 439, I. M. Robertson, G. S. Was, L. W.Hobbs, and T. D. de la Rubia, eds., Materials Re-search Society, Pittsburgh, PA, pp. 457–469 (1997).

8. G. R. Odette: Radiation-induced microstructuralevolution in reactor pressure vessel steels. In Mi-crostructure of Irradiated Materials, Symposium Pro-ceedings, Vol. 373, I. M. Robertson, L. E. Rehn, S. J.


Zinkle, and W. J. Phythian, eds., Materials Re-search Society, Pittsburgh, PA, pp. 137–148 (1995).

9. M. K. Miller, M. G. Hetherington, and M. G.Burke: Atom probe field-ion microscopy: A tech-nique for microstructural characterization of irra-diated materials on the atomic scale. Metall. Trans.20A: 2651–2661 (1989).

10. M. K. Miller and S. S. Brenner: An Atom ProbeStudy of Irradiated Pressure Vessel Steel. Proc. 28thInt. Field Emission Symposium, 27–31 July 1981,L. W. Swanson and A. Bell, eds., The OregonGraduate Center, Beaverton, OR, pp. 242–244(1981).

11. M. K. Miller and S. S. Brenner: FIM�atom probestudy of irradiated pressure vessel steels. Res.Mechan. 10:161–168 (1984).

12. M. K. Miller, J. A. Spitznagel, S. S. Brenner, andM. G. Burke: Microanalytical Investigation of LightWater Reactor Materials Using the Atom Probe. Proc.2nd Int. Symp. on Environmental Degradation ofMaterials in Nuclear Power Systems—Water reac-tors, J. T. A. Roberts, J. R. Weeks, and G. J. Theus,eds., American Nuclear Society, La Grange Park,IL, pp. 523–528 (1986).

13. S. P. Grant, S. L. Earp, S. S. Brenner, and M. G.Burke: Phenomenological Modeling of Radiation Em-brittlement in Light Water Reactor Vessels with AtomProbe and Statistical Analysis. Proc. 2nd Int. Symp.on Environmental Degradation of Materials inNuclear Power Systems—Water Reactors, J. T. A.Roberts, J. R. Weeks, and G. J. Theus, eds., Ameri-can Nuclear Society, La Grange Park, IL, pp. 385–392 (1986).

14. M. G. Burke and S. S. Brenner: A microstructuralinvestigation of irradiated pressure vessel steelweld metal. J. Phys. 47-C2:239–244 (1986).

15. M. K. Miller and M. G. Burke: Characterization ofirradiated A533B pressure vessel steel weld. J.Phys. 48-C6:429–434 (1987).

16. M. K. Miller and M. G. Burke: Microstructural char-acterization of PWR Steel Using the Atom Probe Field-Ion Microscope. Proc. 3rd Int. Symp. on Environ-mental Degradation of Materials in NuclearPower Systems—Water Reactors, G. J. Theus andJ. R. Weeks, eds., The Metallurgical Society, Pitts-burgh, PA, pp. 141–149 (1988).

17. M. G. Burke and M. K. Miller: Solute clusteringand precipitation in pressure vessel steels underlow fluence irradiation conditions. J. Phys. 49-C6:283–288 (1988).

18. M. K. Miller and M. G. Burke: Fine Scale Micro-structural Characterization of Pressure Vessel Steelsand Related Materials Using APFIM Proc. 14th Int.Symp. On Effects of Radiation on MaterialsASTM-STP 1046, vol. 2, N. H. Packan, R. E. Stoller,and A. S. Kumar, eds., American Society for Test-ing and Materials, Philadelphia, PA, pp. 107–126(1990).

19. M. K. Miller and M. G. Burke: An atom probe fieldion microscopy study of neutron-irradiated pres-sure vessel steels. J. Nucl. Mater. 195:68–82 (1992).

20. P. Pareige, J. C. Van Duysen, and P. Auger: AnAPFIM study of the microstructure of a ferrite al-loy after high fluence neutron irradiation. Appl.Surf. Sci. 67:342–347 (1993).

21. P. Auger, P. Pareige, M. Akamatsu, and J.-C. VanDuysen: Microstructural characterization of atomclusters in irradiated pressure vessel steels andmodel alloys. J. Nucl. Mater. 211:194–201 (1994).

22. M. Akamatsu, J. C. Van Duysen, P. Pareige, and P.Auger: Experimental evidence of several contribu-tions to the radiation damage in ferritic alloys. J.Nucl. Mater. 225:192–195 (1995).

23. P. Auger, P. Pareige, M. Akamatsu, and D.Blavette: APFIM investigation of clustering inneutron-irradiated Fe-Cu alloys and pressure ves-sel steels. J. Nucl. Mater. 225:225–230 (1995).

24. M. K. Miller and M. G. Burke: An APFIM Survey ofGrain Boundary Segregation and Precipitation in Irra-diated Pressure Vessel Steels, Proc. 16th Int. symp.on Effects of Radiation in Materials-ASTM STP1175, A. S. Kumar, D. S. Gelles, R. K. Nanstad, andE. A. Little, eds., American Society for Testing andMaterials, Philadelphia, PA, pp. 492–502 (1993).

25. M. K. Miller, R. Jayaram, P. J. Othen, and G.Brauer: Atom Probe Field Ion Microscopy Character-izations of VVER Steels. Proc. 6th Int. Symp. on En-vironmental Degradation of Materials in NuclearPower Systems—Water Reactors, R. E. Gold andE. P. Simonen, eds., The Minerals, Metals and Ma-terials Society, Warrendale, PA, pp.161–169 (1993).

26. M. K. Miller, R. Jayaram, P. J. Othen, and G.Brauer: APFIM characterization of 15Kh2MFA Cr-Mo-V and 15kh2NMFA Ni-Cr-Mo-V type steels.Appl. Surface Sci. 76�77:242–247 (1994).

27. M. K. Miller, R. Jayaram, and K. F. Russell: Char-acterization of phosphorus segregation in neu-tron-irradiated Russian pressure vessel steel weld.J. Nucl. Mater. 225:215–224 (1995).

28. M. K. Miller, R. Jayaram, and K. F. Russell: Charac-terization of Phosphorus Segregation in Neutron-Irra-diated Pressure Vessel Steels by Atom Probe Field-IonMicroscopy. Proc. Microstructure of Irradiated Ma-terials, Vol. 373, I. M. Robertson, L. E. Rehn, S. J.Zinkle, and W. J. Phythian, eds., Materials Re-search Society, Pittsburgh, PA, pp. 113–118 (1995).

29. M. K. Miller and K. F. Russell: APFIM character-ization of a high phosphorus Russian RPV weld.Appl. Surface Sci. 94�95:378–383 (1996).

30. P. Pareige, P. Auger, S. Miloudi, J. C. Van Duysen,and M. Akamatsu: Microstructural evolution ofthe CHOOZ A PWR surveillance program mate-rial: Small angle neutron scattering and tomo-graphic atom probe studies. Ann. Phys. C2 22:117–124 (1997).8

31. P. Pareige, P. Auger, S. Welzel, J. C. Van Duysen,


and S. Miloudi: Annealing of a Low Copper Steel:Hardness, SANS, Atom Probe and ThermoelectricPower Investigations. Proc. 19th Int. Symp. Effectsof Radiation on Materials, ASTM STP 1366, M. L.Hamilton, A. S. Kumar, S. T. Rosinski, and M. L.Grossbeck, eds., American Society for Testing andMaterials, Philadelphia, PA (in press).

32. P. Pareige, K. F. Russell, R. E. Stoller, and M. K.Miller: Influence of long-term thermal aging onthe microstructural evolution of nuclear reactorpressure vessel materials: An atom probe study. J.Nucl. Mater. 250:176–183 (1997).

33. M. K. Miller and K. F. Russell: Atom probe charac-terization of copper solubility in the Midland weldafter neutron irradiation and thermal annealing. J.Nucl. Mater. 250:223–228 (1997).

34. P. Pareige, R. E. Stoller, K. F. Russell, and M. K.Miller: Atom probe characterization of the micro-structure of nuclear pressure surveillance materi-als after neutron irradiation and after annealingtreatments. J. Nucl. Mater. 249:165–174 (1997).

35. S. R. Goodman, S. S. Brenner, and J. R. Low: AnFIM-atom probe study of the precipitation of cop-per from iron 1.4 At. Pct copper: Part I field ionmicroscopy, and part II, atom probe analysis.Metal. Trans. 4:2363–2378 (1973).

36. G. M. Worrall, J. T. Buswell, C. A. English, M. G.Hetherington, and G. D. W. Smith: A study of theprecipitation of copper particles in a ferrite matrix.J. Nucl. Mater. 148:107–114 (1987).

37. J. T. Buswell, C. A. English, M. G. Hetherington,W. J. Phythian, G. D. W. Smith, and G. M. Worrall:An Analysis of Small Clusters Formed in Thermallyaged and Irradiated FeCu and FeCuNi Model Alloys.Proc. 14th Int. Symp. On Effects of Radiation onMaterials ASTM-STP 1046, vol. 2, N. H. Packan,R. E. Stoller, and A. S. Kumar, eds., American So-ciety for Testing and Materials, Philadelphia, PA,pp. 127–153 (1990).

38. S. S. Brenner, R. Wagner, and J. A. Spitznagel:Field-ion microscope detection of ultra-fine de-fects in neutron-irradiated Fe-0.34 Pct Cu alloy.Metal. Trans. 9A:1761–1764 (1978).

39. P. J. Pareige, K. F. Russell, and M. K. Miller:APFIM studies of the phase transformation inthermally aged ferritic FeCuNi alloys: Compari-son with aging under neutron irradiation. Appl.Surface Sci. 94/95: 363–369 (1996).

40. M. K. Miller, K. F. Russell, P. Pareige, M. J. Starink,and R. C. Thomson: Low temperature copper sol-ubilities in Fe-Cu-Ni. Mater. Sci. Eng. A250: 49–54(1998).

41. P. Pareige, P. Auger, P. Bas, and D. Blavette: Directobservation of copper precipitation in neutron-irradiated FeCu alloys by 3D atomic tomography.Scripta Metall. Mater. 33: 1033–1036 (1995).

42. M. K. Miller, D. T. Hoelzer, F. Ebrahimi, J. R. Haw-thorne, and M. G. Burke: Characterization of irra-

diated model pressure vessel steels. J. Phys. 48-C6:423–428 (1987).

43. M. K. Miller, D. T. Hoelzer, F. Ebrahimi, J. R. Haw-thorne, and M. G. Burke: Microstructural Character-ization of Irradiated Fe-Cu-Ni-P Model Steels. Proc.3rd Int. Symp. on Environmental Degradation ofMaterials in Nuclear Power Systems—Water Re-actors, G. J. Theus and J. R. Weeks, eds., The Met-allurgical Society, Pittsburgh, PA, pp 133–139(1988).

44. M. K. Miller, R. E. Stoller, and K. F. Russell: Effectof neutron-irradiation on the spinodal decomposi-tion of Fe-43% Cr model alloy, J. Nucl. Mater. 230:219–225 (1996).

45. J. T. Buswell, C. A. English, M. G. Hetherington,W. J. Phythian, G. D. W. Smith and G. M. Worral:Analysis of Small Clusters formed in Irradiated andThermally Aged Copper-Containing Model Alloys,Proc. 14th Int. Symp. on Effects of Radiation onMaterials: ASTM STP 1046, N. H. Packan, R. E.Stoller and A. S. Kumar, eds., American Societyfor Testing and Materials, Philadelphia, PA, p.127(1990).

46. P. J. Othen, M. L. Jenkins, G. D. W. Smith, and W. J.Phythian: Transmission electron microscope in-vestigations of the structure of copper precipitatesin thermally-aged Fe-Cu and Fe-Cu-Ni. Philos.Mag. Lett. 64:383–391 (1991).

47. M. K. Miller and G. D. W. Smith: Atom probeanalysis of interfacial segregation. App. Surface Sci.87�88:243–250 (1995).

48. A. Vatter, C. A. Hippsley, and S. G. Druce: Reviewof thermal aging data and its application to oper-ating reactor pressure vessels. Int J. Pressure Ves.Piping, 54:31–48 (1993).

49. S. G. Druce, G. Gage, and G. Jordan: Effect of ag-ing on properties of pressure vessel steels. ActaMetall. 34:641–652 (1986).

50. D. McLean: Grain Boundaries in Metals, OxfordUniversity Press, London, p. 116 (1957).

51. J. R. Hawthorne, H. E. Watson, and F. L. Loss: Ex-ploratory investigations of cyclic irradiation andannealing effects of notch ductility of A533B welddeposits. In Effects of Radiation on Structural Materi-als, ASTM STP 683, J. A. Sprague and D. Kramer,eds., American Society for Testing and Materials,Philadelphia, PA, pp. 278–294 (1978).

52. J. R. Hawthorne, H. E. Watson, and F. L. Loss: Ex-perimental Investigation of Multicycle Irradiation andAnnealing Effects on Notch Ductility of A533-B WeldDeposits. Proc. 10th Conf. on Effects of Radiationin Materials, ASTM STP 725, D. Kramer, H. R.Prager, and J. S. Perrin, eds., American Society forTesting and Materials, Philadelphia, PA, pp. 63–75(1981).

53. R. G. Lott, T. R. Mager, R. P. Shogan, and S. E. Ya-nichko: Annealing and reirradiation response of ir-radiated pressure vessel steels. In Radiation Embrit-tlement of Nuclear Reactor Pressure Vessel Steels, An


International Review, vol. 2, ASTM STP 909, L. E.Steele, ed., American Society for Testing and Ma-terials, Philadelphia, PA, pp. 242–259 (1986).

54. J. R. Hawthorne: Steel Impurity Element Effects onPost irradiation Properties Recovery by Annealing.13th Int. Symp. Influence of Radiation on Materi-als Properties Part II, ASTM STP 956, F. A. Garner,C. H. Hanager, and N. Igata, eds. American Soci-ety for Testing and Materials, Philadelphia, PA,pp. 461–479 (1987).

55. J. R. Hawthorne and A. L. Hiser: Investigation of Ir-radiation-anneal-reirradiation (IAR) Property Trendsof RPV Welds. Proc. 5th Int. Symp. on Environmen-tal Degradation of Materials in Nuclear PowerSystems—Water Reactors, D. Cubicciotti, E. P. Si-monen, and R. E. Gold, eds., American NuclearSociety, La Grange Park, IL, pp. 671–678 (1992).

Received December 1998; accepted February 1999.

Documents

Understanding Pressure Vessel Steels: An Atom Probe ...arc.nucapt.northwestern.edu/refbase/files/MatChar44p235PVSteel... · Understanding Pressure Vessel Steels 237 a vanadium-containing