sioio

S. W

tmen

bor, M

faced fuel performance limitations due to Zircaloy dation of the rst wall and plasma by sputtering, ero-sion, blistering, etc., induced activity in structuralmaterials and the challenges in breeding tritiumand in extracting it for future use. While corematerials in ssion reactors faced an environment

erved.

* Tel.: +1 734 763 4675; fax: +1 734 763 4540.E-mail address: gsw@umich.edu

Journal of Nuclear Materials 3670022-3115/$ - see front matter 2007 Elsevier B.V. All rights res1. Introduction

Thirty years ago, the requirements of ssion andfusion systems were largely non-overlapping interms of their challenges and their environments.Table 1 summarizes the principal issues facing amaturing commercial ssion reactor industry andan evolving realization of requirements for materialsin proposed fusion devices. Light water reactors

creep, growth, corrosion and pelletclad interaction,a newly arisen problem of intergranular stress corro-sion cracking of stainless steel BWR pipes, the emer-gence of IGSCC in alloy 600 steam generator tubes,and embrittlement of copper containing welds inreactor pressure vessel steels. Conversely, the fusioncommunity had identied its own set of challenges;creep and fatigue of rst wall materials, dimensionalstability at high temperature and high doses, degra-Abstract

The management of materials in power reactor systems has become a critically important activity in assuring the safe,reliable and economical operation of these facilities. Over the years, the commercial nuclear power reactor industry hasfaced numerous surprises and unexpected occurrences in materials. Mitigation strategies have sometimes solved oneproblem at the expense of creating another. Other problems have been solved successfully and have motivated thedevelopment of techniques to foresee problems before they occur. This paper focuses on three aspects of ssion reactorexperience that may benet future fusion systems. The rst is identication of parameters and processes that have hada large impact on the behavior of materials in ssion systems such as temperature, dose rate, surface condition, gradients,metallurgical variability and eects of the environment. The second is the development of materials performance and fail-ure models to provide a basis for assuring component integrity. Last is the development of proactive materials manage-ment programs that identify and pre-empt degradation processes before they can become problems. These aspects ofLWR experience along with the growing experience with materials in the more demanding advanced ssion reactor systemsform the basis for a set of lessons learned to aid in the successful management of materials in fusion reactor systems. 2007 Elsevier B.V. All rights reserved.Materials degradation in sof relevance to fus

Gary

Nuclear Engineering and Radiological Sciences Depar

1921 Cooley Building, Ann Ardoi:10.1016/j.jnucmat.2007.03.008n reactors: Lessons learnedn reactor systems

as *

t, University of Michigan, 2355 Bonisteel Boulevard,

I 48109-2104, United States

370 (2007) 1120

www.elsevier.com/locate/jnucmat

Today, uprating and life extension of commercial

Table 1Fission/fusion materials challenges, circa 1975

Fission Fusion

Zircaloy corrosion First wall fatigueCladding creep and growth Creep strengthPellet-clad interaction (PCI) Dimensional stability at

high temperature and dosesBWR pipe cracking Induced activitySteam generator tube cracking First wall sputtering,

blistering, erosionRPV embrittlement Breeding materialsVery low capacity factors Tritium retention and

12 G.S. Was / Journal of Nuclear Materials 367370 (2007) 1120LWRs, demands by advanced ssion reactor sys-tems in the Generation IV initiative, [1] and theemergence of accelerator driven transmutation con-cepts [2] have pushed the environments of ssionsystems to higher temperatures, higher neutronuxes, greater uences and signicantly morehelium loading. In fact, as shown by Bloom et al.[3], the environmental conditions of advancedssion systems and magnetic connement fusionsystems are converging in most all aspects. Inessence, both communities are facing very similarchallenges in materials performance.

2. Key factors and processes governing materials

performance

There are numerous important processes, envi-ronments and interactions that impact the perfor-characterized by 320 C temperatures, neutron dosesup to10 dpa and lowHe production, fusion systemanalogues were required to withstand temperaturesabove 500 C, doses over 100 dpa and high heliumloading of plasma facing materials.

extractionmance of materials in a reactor system. Thefactors presented in Table 2 and discussed in the

Table 2Key factors and processes governing materials performance

TemperatureDose rateCompositionmicrostructureCoupling of factors/processesWeldsSurface conditionGradientsIncubationMetallurgical variabilityEnvironment (design)following are distinguished by either their sensitiv-ity, their frequency of occurrence or their unantici-pated consequences in the overall performance ofthe materials or material system.

2.1. Temperature

While temperature is an important factor in mostmaterials degradation processes, it is the sensitivityof degradation to temperature that is sometimessurprising in ssion reactors. An example is shownin Fig. 1, which is a dose-temperature plot of voidformation in an austenitic alloy irradiated in theBN-350 reactor [4]. Because measurements weretaken along the length of a component in the core,the data covers a wide temperature range but areconned to a narrow dose range for that component(identied by strings of data). Also, since the stringsof data come from the same component, uncertain-ties caused by metallurgical variability are elimi-nated. The lled data represent cases whereswelling occurred and the open data represent caseswhere swelling was not observed. There is a strikingdependence on temperature in that all samplesabove 307 C exhibited swelling whereas none ofthe samples below 302 C swelled. This is a verynarrow temperature threshold delineating the swell-ing from the non-swelling regimes, extending over adose range of 357 dpa, and demonstrates theextreme sensitivity of swelling to temperature. Sim-ilar, but not as dramatic, examples exist for swellingin bae bolts in pressurized water reactors (PWR),[5] and in stress corrosion cracking of steam gener-ator tubes [6].

2.2. Dose rate

Equally striking are dose rate dependenciesobserved in ssion reactor systems. For the samedose, a reduction in dose rate by a factor of 10can result in an increase in irradiation hardeningin model pressure vessel steels, by as much as30%, Fig. 2 [7]. Similarly, for 304 stainless steel irra-diated in EBR-II, a reduction in ux by a factor of20 has been observed to produce a decrease in theincubation dose for swelling from 20 dpa to5 dpa [8].

2.3. Composition and microstructure

Composition and microstructure can have pro-

found eects on material behavior, and in some

-760 mm

+590 mm

G.S. Was / Journal of Nuclear Materials 367370 (2007) 1120 1330

40

50

60

Dos

e (d

pa)cases, the results are very surprising. The eect ofcopper on embrittlement of RPV steel welds is wellknown. But a level as low as 0.3% can increase thetransition temperature by over 100 C and dropthe upper shelf energy almost in half followingirradiation to 1 1023 n/m2 at 288 C, Fig. 3 [9].Cu-rich precipitate zones are responsible for about80% of the transition temperature shift. Changesin the Si content of austenitic stainless steel from0.5% to 1.0% is linked to severe IGSCC andenhanced crack growth rates [10]. The addition ofless than 1 wt% Hf to 316 SS suppresses void forma-

0

10

20

280 290 300 31Irradiation tem

Fig. 1. Dose-temperature plot of occurrence of void formation

0.001 0.01 0.1 1 10

t (1023 n/m2)

300

250

200

150

100

50

0

y (M

Pa)

Fig. 2. Eect of dose rate (diamonds = high dose rate, cir-cles = intermediate dose rate, squares = low dose rate) on hard-ening in a model RPV steel containing 0.4% Cu and 1.25% Ni [7].+560 mm

-450 mm

+7500 mm

0 mm

-150 mm

-150 mmtion in a high purity Fe18Cr12Ni alloy to beyond50 dpa compared to less than 2 dpa in the undopedalloy, Fig. 4 [11].

2.4. Coupling of processes

Dose rate and temperature can combine to createsevere eects. In EBR-II thimbles, the ux along thelength of the core varies symmetrically from top tobottom, but the temperature peaks near the top,resulting in greater swelling at the bottom of the

0 320 330 340perature (C)

of an austenitic alloy irradiated in the BN-350 reactor [4].

Temperature (C)

Cha

rpy

Ener

gy (J

)

Submerged-Arc WeldsIrradiation: 1 x 1023 n/m2, 288CNormalized To Unirradiated Curve

Unirradiated

TransitionRegion

Shift

0.06% CuIrradiated

Loss

0.30% CuIrradiated

Upper Shelf

200

150

100

50

0-200 -150 -100 -50 0 50 100 150 200 250 300

Fig. 3. Eect of 0.3% Cu on the DBTT and upper shelf energyfor an RPV steel submerged-arc weld irradiated to 1 1023 n/m2

at 288 C [9].

proles of Ni, Cr and Si were obtained by conduct-

challenge to maintaining component integrity. Reac-tor pressure vessel steel embrittlement, pipe crackingin BWRs, SCC in stainless steel core shrouds andcracking of alloy 182-alloy 600 welds in reactorvessel head penetrations in PWRs are some of thenotable occurrences of problems with welds thathave caused considerable problems in light waterreactors over the years.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 10 20 30 40 50 60

316L SS 316L + Pt 316L + Hf 316L + Hf, optimized

Dose (dpa)

Fig. 4. Suppression of swelling to at least 50 dpa in 316SS dopedwith 1 wt% Hf after irradiation with Ni ions at 500 C [11].

14 G.S. Was / Journal of Nuclear Materials 367370 (2007) 1120ing proton irradiation at a dose rate 100 that forneutron irradiation and at a temperature 85 Chigher, Fig. 6 [12]. The advantage of coupling canbe further realized by composition and microstruc-ture modication. Addition of oversize solutes,grain boundary engineering and oxide dispersantscan be combined to trap vacancies and reduceRIS, thus reducing IASCC, increasing creep resis-tance and enhancing strength [13].

2.5. Surface condition and welds

The importance of surface condition and welds inthe performance of materials, especially in aggres-thimble compared to the top, but in locations thathave experienced the exact same uence, Fig. 5 [8].The coupling of dose rate and temperature wasexploited in the use of proton irradiation to emulateneutron irradiation eects. Nearly identical RIS3A1 flat 31%/dpa3A1 flat 1

Highertemperature

Lowertemperature

Same dpa rate

Fig. 5. Coupling of ux and temperature to produce a variationin swelling at locations where the uence is the same [8].sive environments, is often underestimated. The Jap-anese BWR eet was sidelined for several monthsdue to SCC in core shrouds that was tracked tosevere machining and grinding. Welds have contin-ued to plague the industry and present a substantial

8

12

16

20

24

0

1

2

3

4

5

6

7

8

-12 -8 -4 0 4 8 12

CP-316

Distance from GB (nm)

Cr

Mea

sure

d C

r or N

i (w

t%)

Protons at 360 to 1.0 dpa

Mea

sure

d Si

(wt%

)

Neutrons at 275 to ~1.1x1021n/cm2

Cr Ni Si

Ni

Si

Fig. 6. Exploitation of the coupling between dose rate andtemperature to achieve similar RIS proles in proton-irradiated316 SS irradiated at a dose rate 100 and a temperature 85 Chigher than neutron irradiation [12].2.6. Gradients

While maximum values of key variables are usedas the design point for systems, they are seldom uni-form across the thickness or length of a component.Weld stresses in austenitic stainless steel piping inBWRs result in a residual stress distribution thatvaries through the thickness of the pipe. This varia-tion is multiplied in the stress intensity factor varia-tion at a crack tip, and must be accounted for indetermining component lifetime. The reactor pres-sure vessel contains multiple variable gradients thatcan all impact its integrity [14]. As shown in Fig. 7,neutron uence decreases from the inside to the out-side of the vessel wall, the temperature increases

Fig. 7. Flux, temperature, stress and fracture toughness gradientsin a reactor pressure vessel wall [14].

G.S. Was / Journal of Nuclear Materials 367370 (2007) 1120 152.7. Incubation

Some of the biggest surprises in material perfor-mance occur because of an incubation period thatdelays the emergence of a process. A classic exampleis void swelling in which composition, cold workfrom the inside to the outside and the stress in thevessel peaks at the inside surface and decreasesthrough wall. As a result of these gradients, the frac-ture toughness varies through wall in a complicatedmanner that must be understood in order to ensuresafe operation of the vessel.and dose rate can push the onset of measurable

0

20

40

60

80

100

1023 1024 1025 1026 1027

%IG

SCC

Neutron Fluence (n/m2, E>1 MeV)

Kodama et al. 1993Clark and Jacobs 1983Jacobs et al. 1993Kodama et al. 1992

Jacobs et al. 1993Kodama et al. 1992

304 SS

316 SS

304SS

316SS

Fig. 8. Incubation dose for the emergence of IASCC in 304 and316 SS in BWRs [16].swelling from a few dpa to greater than 50 dpa[15]. IASCC in austenitic stainless steels has beenobserved to occur only after an incubation periodof 13 dpa (721 1020 n/cm2), Fig. 8 [16]. Moreextreme examples are cracking of alloy 182 weldsor failure of steam generator tubes, Fig. 9, in whichhot leg cracking required 610 years to emerge as aproblem [17]! Unfortunately, once failures begin tooccur, the rates can be extremely rapid. Althoughthe emergence of hot leg cracking at the tubesheetand tube support plates in the Ringhals steam gen-erators required almost a decade to appear as aproblem, it took less than half that time for theseprocesses to dominate all other failure modescombined.

2.8. Metallurgical variability

Material properties can vary substantially with

Service time (EFPY)

Cum

ulat

ive

fract

ion

of tu

bes

faile

d

20 30101

aggregate of all mechanisms

freespanIGSCC

lower bundle IG Aimpingement erosion/corrosion

wear/frettingat TSPs

miscellaneous causes

corrosionfatigue

0.9

0.1

0.01

0.001

0.0001

Fig. 9. Incubation in the degradation mechanisms in a PWRsteam generator [17].metallurgical condition. Dierences among alloyheats or batches can contribute to this variabilityin ways that are not fully understood. Yet heat-to-heat variability translates into variability in compo-nent performance that adds an uncertainty to theability of plant operators to predict material perfor-mance. Fig. 10 shows the signicance of heat-to-heat variabilities in the stress corrosion cracking ofsteam generator tubing [18]. Thirty six heats of tub-ing were used in a single steam generator in a singleplant. Following 75000 h of service, the perfor-mance in the tubes varied widely and the strongestvariable explaining the variation of tube crackingthroughout the SG was the heat. Note that whilesome heats had as much as one third of tubesaected, others had none. Yet all heats met the samespecications on composition and properties.

rator

16 G.S. Was / Journal of Nuclear Materials 367370 (2007) 11202.9. Environment (design)

One of the most important factors aecting com-ponent integrity, that is too often overlooked ordiscounted in the design process, is the environment.All power generation systems, fusion included, uti-lize a working uid in converting the heat generatedfrom the fuel to mechanical and then electricalenergy. At the temperatures and pressures requiredfor operation of nuclear sources, these uids presentaggressive conditions to the components they con-tact. Not surprisingly, corrosion continues toaccount for a majority of the capacity factor lossdue to materials degradation in light water reactorsystems. Steam generator tube cracking, vessel pen-etration weld cracking, bae bolt cracking in PWRsand steam line cracking and core shroud cracking in

Fig. 10. Heat-to-heat variability in steam geneBWRs, along with turbine disc cracking and clad-ding corrosion and SCC are only a few examplesof environmentally-induced failures in light waterreactors.

IASCC is a prime example of the prevalence ofenvironmentally-induced degradation. IASCC is aform of intergranular stress corrosion cracking thatoccurs due to a combination of irradiation and anaggressive environment. It does not occur withoutthe environment and may occur, but much moreslowly, in the environment in the absence of irradi-ation. Yet it is a generic problem, aecting dozens ofcomponents in most austenitic alloys in all waterreactor designs.

Steam generator tube cracking constitutes one ofthe most pervasive corrosion problems occurring inlight water reactors. Its occurrence is a combinationof a susceptible alloy and a design that promotedconditions conducive to corrosion. Staehle andGorman [17] have identied some 25 distinctcombinations of corrosion mode and location fordegradation of steam generator tubes in steam gen-erators with drilled-hole tube support plates.

As described earlier, advanced ssion reactorconcepts incorporate higher temperatures, higherdose rates and higher doses in an eort to boost e-ciency and economy of power generation. The veryhigh temperature gas reactor (VHTR), the lead-cooled fast reactor and the supercritical waterreactor (SCWR) all present signicant corrosionchallenges due to aggressive environments and hightemperatures. Heat exchangers proposed for theVHTR include the printed circuit board design thatbenets from compactness and eciency, but con-sists of very thin walls and numerous welds between

tube degradation after 75000 h of service [18].layers. Both of these distinguishing features willpresent considerable challenges in maintainingmaterial integrity at very high temperatures.

The SCWR will operate with supercritical watertemperatures in the 600 C range. Signicant experi-ence with supercritical fossil plants exists in the USand other countries, but the engineering is notdirectly transferable. For example, a ferriticmar-tensitic re tube in a fossil plant may accumulateseveral hundred micrometers of oxide over 20 yearsof operation near 500 C. For a 10 mm thick tube,this amount of oxidation is acceptable. But due toneutron economy, the SCWR fuel cladding thick-ness is closer to 0.6 mm and the water rod designcalls for 0.4 mm thick walls. At the same corrosionrate as in a fossil plant, the lifetime would be limitedto the order of a year at best, without accounting forthe eect of irradiation on corrosion and SCC.

G.S. Was / Journal of Nuclear Materials 367370 (2007) 1120 17Because the bulk of the power generation plant issimilar for all systems, and because the environmentfor materials in advanced ssion reactor systems ismoving closer to that in fusion systems, fusion willface the same problems as those just discussed. Asconcepts evolve into designs, the fusion communityshould look to the LWR industry and the advancedssion reactor program as a repository of a rich his-tory of material performance data, and for guidancein dealing with generic issues involving materialsperformance in power generation systems. How-ever, material performance is only part of the issuethat the fusion community must confront. The otheris component failure and failure prediction.

3. The importance of failure (or performance)

prediction

The operation of materials in high temperature,aggressive environments will lead to degradation,and ultimately, failure. The capability to predictand track the approach to failure has enormous eco-nomic and performance consequences. The develop-ment of sophisticated performance predictioncapability is required in order to provide defensiblearguments to regulators for the continued safe oper-ation of a system containing aws. The inability todo so will result in the application of unnecessarilyconservative assumptions that will force prematureshutdown and component replacement, all of whichcan be extremely costly.

Over the years, the LWR industry has evolvedfrom deterministic failure prediction to probabilisticprediction methodology. Historically, laboratoryand commercial experience were used to develop aphenomenological model for the failure of a compo-nent. However, such deterministic approaches oftenhave poor accuracy due to the multitude of vari-ables aecting the outcome, a poor knowledge ofthe failure mechanism and uncertainty in many ofthe materials parameters and dependencies. As aresult, probabilistic failure prediction has gainedimportance in predicting the behavior of complexsystems.

An early example of probabilistic failure predic-tion is the failure of fuel rods in light water reactors.Using fuel rod pre-characterization parameters,reactor operating history, thermo-mechanical fuelbehavior models and deterministic failure modelsas input, Christensen [19] used pattern recognitionmethodology to predict fuel failures in an operating

reactor. The prediction called for 6 3 of the 217fuel assemblies in the Maine Yankee reactor cycle4 (September 1978January 1980) to contain at leastone leaking rod. Sipping results in January 1980revealed that there were 9 assemblies with at leastone leaking rod. Interestingly, the contribution ofthe deterministic model to this extremely accurateresult was so small as to be inconsequential. Thisearly example highlighted the potential for statisti-cal modeling of failure processes in nuclear systems.

More current examples include recent work byStaehle [20] on the development of statistical modelsfor steam generator tube failure. Staehle developeda relation between the dependent variable, e.g.,depth of SCC penetration, and the independentvariables such as temperature, stress, and pH thendetermined the dependence of the statistical modelparameters on the independent variables. The objec-tive is to model the statistical parameters accordingto the dependencies of the primary independentvariables in order to better model the crack penetra-tion depth (and hence, the failure distribution) intime.

Probabilistic fracture mechanics is also beingapplied to degradation of reactor pressure vessels.A model is constructed for the stress intensity factoras inuenced by design, physical properties of thecomponent, aw data and thermal/hydraulic systemperformance. Another model is developed for theirradiation temperature shift that depends on alloychemistry, uence, temperature, etc. The modelsare combined with a fracture mechanics model toprovide the probability of crack initiation and theprobability of thru-wall cracking, Fig. 11.

Prediction of materials degradation and failureare important for maintaining the integrity of thecomponent and the safe operation of the plant.But the light water reactor industry has gone onestep further in materials management in reactor sys-tems proactive management of materials.

4. Proactive materials management

Two separate but similar initiatives have beenestablished to actively seek out and identify thepotential weak-links in materials performance andaddress the weaknesses before they emerge in plantoperation. This proactive approach is designed topre-empt problems before they occur. One such pro-gram is the Industry Initiative on Management ofMaterials Issues that was established in 2003 bythe chief nuclear ocers of the nations utilities

under the auspices of the Nuclear Energy Institute

tor p

18 G.S. Was / Journal of Nuclear Materials 367370 (2007) 1120and the Electric Power Research Institute [21]. Thepurpose of the initiative is to assure the safe, reliableand ecient operation of US nuclear power plantsby providing for:

consistent materials degradation issue manage-ment processes across the industry,

prioritization of materials degradation issues,and

proactive, integrated and coordinated approachesto issue resolution.

Fig. 11. Probabilistic approach to failure of a reacThe approach utilizes two tools; degradationmatrix tables and issue management tables. Thedegradation matrix tables seek to identify the mate-rials used for major passive components and sys-tems, identify the most probable degradationmodes of each component of each system of a par-ticular reactor type, and develop the tools to man-age the potential threats to system integrity. Theissue management table focuses on identifying spe-cic components, materials of construction, failuremechanisms and locations, consequences of failureand mitigation/repair/replacement options, andgoes so far as to identify knowledge gaps and prior-itize work to resolve the gaps.

The NRC is in the process of developing a Proac-tive Materials Degradation Assessment programwith the objectives of identifying materials and loca-tions where degradation can reasonably be expectedin the future, and developing and implementing aninternational cooperative research program for thecomponents and degradation of interest that willreview, evaluate and develop remedies as needed.

Both programs have in common an exhaustiveand thorough examination of every system, sub-sys-tem and component in the plant. While time con-suming and expensive, only by combing througheach and every component in every sub-system inevery system, can potential problems be identiedand addressed. The fusion community can benetfrom these pioneering eorts by implementing pro-active programs as specic designs begin to take

ressure vessel steel. (Courtesy S. Rosinsky, EPRI).form.

5. Summary

Materials degradation has played a large role inthe loss of capacity factor in the ssion reactorindustry. Degradation is costly and can limit thecompetitiveness of the power system. But it is alsoinevitable as corrosion, for example, is nothingmore than the return of a metal to its thermody-namically stable state, and engineered metal-basedsystems are the antithesis of this process.

While material behavior in reactor systems isinuenced by many parameters and processes, itssensitivity to some parameters or combinations ofparameters is notable. In particular, temperature,dose rate, composition/microstructure, surface con-dition and the coupling of multiple parameters haveplayed key roles in ssion materials performance.Processes such as incubation time for the occur-rence of a degradation mode and gradients in thick

G.S. Was / Journal of Nuclear Materials 367370 (2007) 1120 19sections have also driven materials performance.The variability in metallurgical condition adds aconfounding parameter to materials performanceprediction. Environmental eects have played amajor role in LWR materials performance and willlikely be a key factor in fusion systems as well.

When concepts evolve into designs, the assuranceof safe operation of a reactor system will depend onthe development of failure prediction models. Thesemodels have evolved frompurely deterministic to sta-tistical or probabilistic, in an eort to accommodatethe numerous variables aecting materials behaviorand poorly dened degradation mechanisms.

In an eort to move beyond a purely reactivemode of materials performance, proactive materialsmanagement programs are being developed to iden-tify weak-links in components and systems and todevelop the knowledge and tools in a pre-emptivefashion. The identication of key parameters andprocesses that aect materials performance in reac-tor systems, the evolution of prediction methodol-ogy for materials performance and failure, and theadvent of proactive materials management pro-grams provides a rich toolbox for the ecient devel-opment of future fusion reactor systems. The fusioncommunity has much to gain by drawing on theeorts of the ssion reactor industry to addressmaterials degradation issues, develop models fortheir occurrence and begin to establish proactivematerials management plans to pre-empt problemsthat can be foreseen.

Acknowledgements

The author gratefully acknowledges the followingindividuals for providing gures and references forthis paper: G.R. Odette, UCSB; J. Busby, E. Bloom,R. Stoller and R. Nanstad, ORNL; F. Garner and S.Bruemmer, PNNL, and S. Rosinsky, EPRI.

References

[1] A Technology Roadmap for Generation IV Nuclear EnergySystems, US Department of Energy, Oce of NuclearEnergy, Science and Technology, GIF-002-00, 2002.

[2] Report to congress on advanced fuel cycle initiative: thefuture path for advanced spent fuel treatment and transmu-tation research, US Department of Energy, Oce of NuclearEnergy, Science and Technology, 2003.

[3] E.E. Bloom, S.J. Zinkle, F.W. Wien, J. Nucl. Mater. 329-333 (2004) 12.

[4] F.A. Garner, S.I. Porollo, A.N. Vorobjev, YU.V. Konobe-eva, M. Dvoriashin, Void swelling at low displacement rates

in annealed 12X18H10T stainless steel at 3060 dpa and280332 C, in: F.P. Ford, S.M. Bruemmer, G.S. Was(Eds.), Ninth International Symposium on EnvironmentalDegradation of Materials in Nuclear Power Systems WaterReactors, The Minerals, Metals & Materials Society, War-rendale, PA, 1999, p. 1051.

[5] F.A. Garner, L.R. Greenwood, Survey of recent develop-ments concerning the understanding of radiation eects onstainless steels used in the LWR power industry, in: G.S.Was, L. Nelson, P. King (Eds.), Eleventh InternationalConference of the Environmental Degradation of Materialsin Nuclear Power Systems Water Reactors, AmericanNuclear Society, LaGrange Park, IL, 2003, p. 887.

[6] T.B. Cassagne, P. Combrade, M.A. Foucault, A. Gelpi,Inuence of mechanical and environmental parameters onthe crack growth behavior of alloy 600 in PWR primarywater, in: P.J. Tunturi (Ed.), Proceedings of 12th Scandina-vian Corrosion Congress & Vol. 2, EUROCORR92, Vol. 2,Corrosion Society of Finland, Helsinki, 1992, p. 55.

[7] G.R. Odette, T. Yamamoto, D. Klingensmith, Philos. Mag.85 (47) (2005) 779.

[8] F.A. Garner, D.J. Edwards, S.M. Bruemmer, S.I. Porollo,Yu.V. Konobeev, V.S. Neustroev, V.K. Shamardin, A.V.Kozlov, Recent developments concerning potential voidswelling of PWR internals constructed from austeniticstainless steels, in: Proceedings of the Fontevraud 5, Contri-bution of Materials Investigation to the Resolution ofProblems Encountered in Pressurized Water Reactors, 2327 September, 2002, paper #22.

[9] R.K. Nanstad, R.E. Stoller, M.K. Miller, M.A. Sokolov, In-service degradation and life extension of nuclear reactorvessels: combining experiments and modeling, in: L.G.Mallinson (Ed.), Ageing Studies and Lifetime Extension ofMaterials, Kluwer Academic/Plenum Publishers, New York,2001, p. 565.

[10] G.S. Was, P.L. Andresen, Stress corrosion cracking behaviorof alloys in aggressive nuclear reactor core environments, in:Proceedings of the Topical Research Symposium on Corro-sion in Aggressive Environments, Corrosion/05, NACEInternational, Houston, 2005.

[11] J. Gan, E.P. Simonen, S.M. Bruemmer, L. Fournier, B.H.Sencer, G.S. Was, J. Nucl. Mater. 325 (2004) 94.

[12] G.S. Was, J.T. Busby, T. Allen, E.A. Kenik, A. Jenssen,S.M. Bruemmer, J. Gan, A.D. Edwards, P. Scott, P.L.Andresen, J. Nucl. Mater. 300 (2002) 198.

[13] Design of radiation-tolerant structural alloys for generationIV nuclear energy systems, NERI nal report for project 02-110, US Department of Energy, December, 2005.

[14] R.D. Cheverton et al., J. Press. Vessel Technol.Trans.ASME 105 (May) (1983) 102.

[15] F.A. Garner, Irradiation performance of cladding andstructural steels in liquid metal reactors, in: B.R.T. Frost(Ed.), Materials science and technology: a comprehensivetreatment, Nuclear Materials, vol. 10A, Springer, Berlin,1994, p. 419.

[16] G.S. Was, Recent developments in understanding irradiationassisted stress corrosion cracking, in: Proceedings of the 11thInternational Conference Environmental Degradation ofMaterials in Nuclear Power Systems Water Reactors,American Nuclear Society, La Grange Park, IL, 2003, p.965.

[17] R.W. Staehle, J.A. Gorman, Progress in understanding and

mitigating corrosion on the secondary side in PWR steam

generators, in: Tenth International Conference of the Envi-ronmental Degradation of Materials in Nuclear PowerSystems Water Reactors, Special Topic Paper, NACEInternational, Houston, TX, 2002, p. 1.

[18] P.M. Scott, Corrosion 56 (8) (2000) 771.[19] R. Christensen, SPEAR-BETA fuel performance code sys-

tem, vol. 1: general description, Electric Power ResearchInstitute, Palo Alto, CA, EPRI NP-2291, vol. 1, 1982.

[20] R.W. Staehle, Bases for predicting the earliest failures due tostress corrosion cracking, in: Proceedings of the Chemistry

and Electrochemistry of Corrosion and Stress Corrosioncracking: A Symposium Honoring the Contributions ofR.W. Staehle, The Minerals, Metals and Materials Society,Warrendale, PA, 2001, p. K1.

[21] R.L. Jones, Management of materials degradation/agingissues in light water reactors, Presented in panel session onManagement of Materials Degradation in Current andFuture Power Reactors, American Nuclear Society AnnualMeeting, San Diego, 2005.

20 G.S. Was / Journal of Nuclear Materials 367370 (2007) 1120

Materials degradation in fission reactors: Lessons learned of relevance to fusion reactor systemsIntroductionKey factors and processes governing materials performanceTemperatureDose rateComposition and microstructureCoupling of processesSurface condition and weldsGradientsIncubationMetallurgical variabilityEnvironment (design)

The importance of failure (or performance) predictionProactive materials managementSummaryAcknowledgementsReferences