Astm stp 1529 2012

ix Kroll Award Papers Explosion Cladding: An Enabling Technology for Zirconium in the Chemical Process Industry J.. 3 Performance of Zirconium Alloys in Light Water Reactors with a Revie

Journal of ASTM International Selected Technical Papers STP1529

Zirconium in the Nuclear Industry: 16th International Symposium

JAI Guest Editors:

Magnus Limbäck Pierre Barbéris

ASTM International

100 Barr Harbor Drive

PO Box C700West Conshohocken, PA 19428-2959

Printed in the U.S.A

ASTM Stock #: STP1529

Library of Congress Cataloging-in-Publication Data ISBN: 978-0-8031-7515-0

ISSN: 1050-7558

Copyright © 2011 ASTM INTERNATIONAL, West Conshohocken, PA All rights reserved This material may not be reproduced or copied, in whole or in part, in any printed, mechanical, electronic, fi lm, or other distribution and storage media, without the

written consent of the publisher

Journal of ASTM International (JAI) Scope

The JAI is a multi-disciplinary forum to serve the international scientifi c and engineering community through the timely publication of the results of original research and critical review articles in the physical and life sciences and engineering technologies These peer-reviewed papers cover diverse topics relevant to the science and research that establish the foundation for standards development within ASTM International

Photocopy Rights

Authorization to photocopy items for internal, personal, or educational classroom use, or the internal, personal, or educational classroom use of specifi c clients, is granted by ASTM International provided that the appropriate fee is paid to ASTM International, 100 Barr Harbor Drive, P.O Box C700, West Conshohocken, PA 19428-2959, Tel:

610-832-9634; online: http://www.astm.org/copyright

The Society is not responsible, as a body, for the statements and opinions expressed in this publication ASTM International does not endorse any products represented in this publication

Peer Review Policy

Each paper published in this volume was evaluated by two peer reviewers and at least one editor The authors addressed all of the reviewers’ comments to the satisfaction of both the technical editor(s) and the ASTM International Committee on Publications

The quality of the papers in this publication refl ects not only the obvious efforts of the thors and the technical editor(s), but also the work of the peer reviewers In keeping with long-standing publication practices, ASTM International maintains the anonymity of the peer reviewers The ASTM International Committee on Publications acknowledges with appreciation their dedication and contribution of time and effort on behalf of ASTM International

au-Citation of Papers

When citing papers from this publication, the appropriate citation includes the paper authors, “paper title”, J ASTM Intl., volume and number, Paper doi, ASTM International, West Conshohocken, PA, Paper, year listed in the footnote of the paper A citation is provided as a footnote on page one of each paper

Second Printing, April 2012Baltimore, MD

This publication, Zirconium in the Nuclear Industry: 16th International

Symposium, contains papers presented at the symposium with the same

name held in Chengdu, Sichuan Province, China, May 9-13, 2010 The sponsor of the symposium was ASTM International Committee B10 on Reactive and Refractory Metals and Alloys

The Symposium Chairman was Magnus Limbäck, Westinghouse Electric Sweden and Co-Chairman Zhao Wenjin, Nuclear Power Institute of China (NPIC), Chengdu, Sichuan Province, China Serving as Guest Editors of this publication are Magnus Limbäck and Pierre Barbéris, Areva/Cezus Research Centre, Ugine, France Arthur Motta, Pennsylvania State University, acted as Associate Editor for the publication of these papers

in Journal of ASTM International (JAI)

Overview ix

Kroll Award Papers

Explosion Cladding: An Enabling Technology for Zirconium in the Chemical Process Industry

J G Banker 3 Performance of Zirconium Alloys in Light Water Reactors with a Review

of Nodular Corrosion

D G Franklin 17 The Evolution of Microstructure and Deformation Stability in Zr–Nb–(Sn,Fe) Alloys

Under Neutron Irradiation

V N Shishov 37 The Development of Zr-2.5Nb Pressure Tubes for CANDU Reactors

B A Cheadle 67

Schemel Award Paper

Photoelectrochemical Investigation of Radiation-Enhanced Shadow Corrosion Phenomenon

Y.-J Kim, R Rebak, Y.-P Lin, D Lutz, D Crawford, A Kucuk, and B Cheng 91

Basic Metallurgy

Dynamic Recrystallization in Zirconium Alloys

J K Chakravartty, R Kapoor, A Sarkar, and S Banerjee 121 Measurement and Modeling of Second Phase Precipitation Kinetics in Zirconium

P Mosbrucker, M R Daymond, and R A Holt 195

Fabrication and Mechanical Properties

Segregation in Vacuum Arc Remelted Zirconium Alloy Ingots

A Jardy, F Leclerc, M Revil-Baudard, P Guerin, H Combeau, and V Rebeyrolle 219 Damage Build-Up in Zirconium Alloys During Mechanical Processing: Application

to Cold Pilgering

A Gaillac, C Lemaignan, and P Barberis 244

Multiscale Analysis of Viscoplastic Behavior of Recrystallized Zircaloy-4 at 400°C

M Priser, M Rautenberg, J.-M Cloué, P Pilvin, X Feaugas, and D Poquillon 269 Polycrystalline Modeling of the Effect of Texture and Dislocation Microstructure

on Anisotropic Thermal Creep of Pressurized Zr-2.5Nb Tubes

W Li, R A Holt, and S Tracy 298 Improved Zr-2.5Nb Pressure Tubes for Reduced Diametral Strain in Advanced

CANDU Reactors

G A Bickel, M Griffi ths, A Douchant, S Douglas, O T Woo, and A Buyers 327 Microstructural Studies of Heat Treated Zr-2.5Nb Alloy for Pressure Tube Applications

N Saibaba, S K Jha, S Tonpe, K Vaibhaw, V Deshmukh, S V Ramana Rao,

K V Mani Krishna, S Neogy, D Srivastava, G K Dey, R V Kulkarni, B B Rath,

E Ramadasan, and S A Anantharaman 349

Hydriding – Hydrogen Effect

High Temperature Aqueous Corrosion and Deuterium Uptake of Coupons Prepared from the Front and Back Ends of Zr-2.5Nb Pressure Tubes

H M Nordin, A J Elliot, and S G Bergin 373 Hydrogen Absorption Mechanism of Zirconium Alloys Based on Characterization

of Oxide Layer

K Une, K Sakamoto, M Aomi, J Matsunaga, Y Etoh, I Takagi, S Miyamura,

T Kobayashi, and K Ito 401

In Situ Scanning Electron Microscope Observation and Finite Element Method Analysis of Delayed Hydride Cracking Propagation in Zircaloy-2 Fuel Cladding Tubes

T Kubo, H Muta, S Yamanaka, M Uno, and K Ogata 433 Study on the Role of Second Phase Particles in Hydrogen Uptake Behavior

C Coleman, V Grigoriev, V Inozemtsev, V Markelov, M Roth, V Makarevicius,

Y S Kim, K L Ali, J K Chakravarrty, R Mizrahi, and R Lalgudi 544 Neutron Radiography: A Powerful Tool for Fast, Quantitative and Non-Destructive

Determination of the Hydrogen Concentration and Distribution in Zirconium Alloys

M Grosse 575

Corrosion – Oxide Layer Characterization

Detailed Analysis of the Microstructure of the Metal/Oxide Interface Region in caloy-2 after Autoclave Corrosion Testing

Zir-P Tejland, M Thuvander, H.-O Andrén, S Ciurea, T Andersson, M Dahlbäck, and L Hallstadius 595

Study of the Initial Stage and Anisotropic Growth of Oxide Layers Formed on Zircaloy-4

B X Zhou, J C Peng, M Y Yao, Q Li, S Xia, C X Du, and G Xu 620 Studies Regarding Corrosion Mechanisms in Zirconium Alloys

M Preuss, P Frankel, S Lozano-Perez, D Hudson, E Polatidis, N Ni , J Wei,

C English, S Storer, K B Chong, M Fitzpatrick, P Wang, J Smith, C Grovenor,

G Smith, J Sykes, B Cottis, S Lyon, L Hallstadius, B Comstock, A Ambard, and M Blat-Yrieix 649 Understanding Crack Formation at the Metal/Oxide Interface During Corrosion of

Zircaloy-4 Using a Simple Mechanical Model

A Ly, A Ambard, M Blat-Yrieix, L Legras, P Frankel, M Preuss, C Curfs, G Parry, and Y Bréchet 682

In Pile Behaviour

Optimization of Zry-2 for High Burnups

F Garzarolli, B Cox, and P Rudling 711 Effects of Secondary Phase Particle Dissolution on the In-Reactor Performance

of BWR Cladding

S Valizadeh, G Ledergerber, S Abolhassan, D Jädernäs, M Dahlbäck, E V Mader,

G Zhou, J Wright, and L Hallstadius 729 Hydrogen Solubility and Microstructural Changes in Zircaloy-4 Due to Neutron

Corrosion Resistance, Irradiation Growth, and Mechanical Properties

V Chabretou, P B Hoffmann, S Trapp-Pritsching, G Garner, P Barberis,

V Rebeyrolle, and J J Vermoyal 801 Radiation Damage of E635 Alloy Under High Dose Irradiation in the VVER-1000 and

BOR-60 Reactors

G P Kobylyansky, A E Novoselov, A V Obukhov, Z E Ostrovsky, V N Shishov,

M M Peregud, and V A Markelov 827

Creep and Deformation

ZIRLO Irradiation Creep Stress Dependence in Compression and Tension

J P Foster and R Baranwal 853 Experimental Investigation of Irradiation Creep and Growth of Recrystallized

Zircaloy-4 Guide Tubes Pre-Irradiated in PWR

M A McGrath and S Yagnik 875 REFLET Experiment in OSIRIS: Relaxation under Flux as a Method for Determining

Creep Behavior of Zircaloy Assembly Components

S Carassou, C Duguay, P Yvon, F Rozenblum, J M Cloué, V Chabretou,

C Bernaudat, B Levasseur, A Maurice, P Bouffi oux, and K Audic 899

Impact of the Irradiation Damage Recovery During Transportation on the Subsequent Room Temperature Tensile Behavior of Irradiated Zirconium Alloys

B Bourdiliau, F Onimus, C Cappelaere, V Pivetaud, P Bouffi oux, V Chabretou, and A Miquet 929 Shadow Corrosion-Induced Bow of Zircaloy-2 Channels

S T Mahmood, P E Cantonwine, Y.-P Lin, D C Crawford, E V Mader, and K Edsinger 954

Failure Mechanisms and Transients

Characterization of Oxygen Distribution in LOCA Situations

C Duriez, S Guilbert, A Stern, C Grandjean, L Beˇlovský, and J Desquines 993 Effect of Hydrides on Mechanical Properties and Failure Morphology of BWR Fuel

Cladding at Very High Strain Rate

M Nakatsuka and S Yagnik 1021 Simulation of Outside-in Cracking in Boiling Water Reactor Fuel Cladding Tubes

under Power Ramp

K Sakamoto, M Nakatsuka, and T Higuchi 1054 RIA Failure of High Burnup Fuel Rod Irradiated the Leibstadt Reactor: Out-of-Pile

Mechanical Simulation and Comparison with Pulse Reactor Tests

V Grigoriev, R Jakobsson, D Schrire, G Ledergerber, T Sugiyama, F Nagase,

T Fuketa, L Hallstadius, and S Valizadeh 1073 Author Index 1093 Subject Index 1097

Overview

This STP contains the papers presented at the 16th International sium on Zirconium in the Nuclear Industry held in in Chengdu, Sichuan Province, China, May 9-13, 2010 The fi rst symposium was held in Phila-delphia in 1968, and symposia have been held ever since in two to three year intervals The proceedings of each symposium in the series have been documented with an STP

Sympo-This symposium series remains, after forty years, one of the top entation and information source for the research in the area of zirconium alloy performance in a nuclear reactor environment 42 papers and 32 posters were selected for presentation at the 16th Symposium from 130 abstracts submitted The forty-two papers published in these proceedings were peer reviewed and edited, and are also published in the ASTM online journal, JAI In addition, the most signifi cant parts of the discussions that followed the oral presentation of each paper at the symposium are included

pres-in these proceedpres-ings

Four experts in zirconium area received their Kroll Awards at the 16th Symposium: J Banker, D Franklin, V Shishov and B Cheadle Notewor-thy is the fact that the fi rst one deals with zirconium outside the nuclear industry These papers as well as the 26 previous Kroll award papers are now gathered in “The Kroll Medal Papers 1975-2010” published by ASTM, and covering all aspects of zirconium technology

137 attendants from 19 countries attended the 16th Symposium North and South America, Europe, and Asia were represented

The papers were presented during seven sessions, covering the whole spectrum of zirconium metallurgy, from basic metallurgy to accidental con-ditions and transport, through fabrication, creep and growth, and corrosion and hydriding Looking back from the beginning of this symposium series, it appears that these topics remained quite constant over the time

Besides the historical alloys (Zircaloy-2, Zircaloy-4, Zr2.5Nb and Zr-lNb), several studies were devoted to advanced alloys and alloys under develop-ment: ZIRLO™, M5™, X5A, VB, N18, N36, NZ2, E635, ZrNbSnFe with low tin content, Ziron Some optimisation of Zircaloy-2 was proposed

Modelling appears more and more intricate with the experiments, from precipitation to VAR melting, from the effect of texture and dislocations on creep to the oxygen distribution during LOCA situations or the outside-in cracking in BWR fuel cladding

As noted in the past few symposia, advanced techniques are more atically utilised: high temperatures studies of phase transformations using

synchrotron radiation or neutrons diffraction were presented; EBSD allows measuring local texture, studies on corrosion and hydriding benefi t from the precise positioning of thin foils with FIB

Last, it is worth noting that several studies were presented by tic young searchers or PhD students, which demonstrates the dynamism of the research in the zirconium metallurgy area

enthusias-In the fi eld of basic metallurgy, progresses were reported in the following

– texture change and variant selection through the α-β-α phase mation, wich generated some controversial results on the infl uence of

transfor-an externally applied stress

The processing of zirconium was illustrated fi rst by the VAR melting, with modelling of the alloying element segregation allowing effi cient control on the ingot chemistry, then by the investigation of the damage mechanism during pilgering, which is alloy dependent through the number and size of the second phase particles Transmission electron microscopy was used with the aim of investigating the dislocation microstructure of recrystallised Zir-caloy-4 after 400°C creep test to develop a model of the anisotropic visco-plastic behaviour

Three papers were devoted to Zr-2.5Nb tubes The fi rst one presented a self-consistent modelling of their anisotropic behaviour, which takes into ac-count not only the texture but also the dislocation microstructure result-ing from the last cold drawing pass A second paper showed that the tubes for the future ADC will procure reduced creep variability, owing to a better understanding of its dependence on the microstructure and to an improved extrusion process A detailed study of the microstructure evolution during annealing and of its relation with texture and mechanical properties of the tubes was also conductive to improvement in the tube fabrication sequence.The investigation of hydrogen effects and hydriding constituted an im-portant topic of this symposium with eight papers dealing with this topic Besides a paper showing the capability of neutron radiography to investi-gate this area, an investigation of Zr-2.5Nb hydriding during autoclave tests showed that the hydrogen pick-up was linked to the sample position in the tube (back end/front end) and correlated to various microstructural param-eters An investigation of autoclave corroded samples from different alloys

by TEM, RAMAN and SIMS led to the hypothesis according to which the rate controlling process for hydrogen absorption was the diffusion of hydro-gen ions in the oxide barrier layer, the best alloys having a higher protec-tive layer due to compressive stresses and Fe dissolution from the second phase particles Another paper confi rms that the hydrogen pick up during corrosion is closely related to the size, area fraction, and compositions of the second phase particles

The hydride dissolution, re-orientation and stress could be followed in situ by synchrotron radiation diffraction during temperature and stress cy-cles The hydride re-orientation was the subject of a second paper coupling mechanical tests on irradiated cladding and fi nite element modelling, show-ing the importance of the hoop stress, and evidencing some other param-eters Two papers dealt with delayed hydride cracking (DHC) The fi rst one which was derived from an IAEA coordinated research program investigated the effect of the microstructure on DHC, showing its main role is to control the material strength, while the second one presented in situ observation in

an SEM of the crack propagation during a DHC experiment and a FEM of the accepted mechanism

Two papers on the corrosion mechanism and oxide layer tion showed on the one hand that the delayed oxidation of the SPPs in the oxide layer lead to the formation of small cracks, and on the other hand the infl uence of the grain substrate orientation on the oxide layer thickness and epitaxy during steam corrosion tests

characteriza-The in-pile behaviour was the subject of numerous studies In the aim of mitigating the accelerated HPUF tendency at high burn up, it was proposed

to decrease the nickel content and increase the iron and chromium contents

in Zircaloy-2 within the ASTM specifi cation Zircaloy-2 was also the ject of a detailed study of the SPPs’ evolution, showing that Zr(Fe,Cr)2 SPP amorphize while Zr2(Fe,Ni) SPP remain crystalline during irradiation The shadow corrosion due to galvanic coupling between a zirconium alloy and

sub-a more noble metsub-al wsub-as investigsub-ated by photo-electrochemistry; the sub-alloy infl uence was evidenced, and it was postulated that a coating on the fuel as-sembly spacer may mitigate the shadow corrosion

Annealing and DSC experiments on irradiated Zircaloy-4 enabled the study of the infl uence of the neutron damage on the hydrogen solubility, while the morphology of hydrides and SPP was investigated by TEM

The in pile behaviour of advanced alloys were presented: X5A (for PWR) with two different fi nal heat treatments showed improved properties and that the irradiation creep of ZIRLO™ is linear with the (deviatoric hoop) stress The same phenomenon is observed in tension and compression The increase in tin content up to 0.3 % in Zr1NbSnFe does not signifi cantly mod-ify the corrosion resistance nor the hydrogen pick-up compared to Zr1Nb

in the BOR-60 reactor.

The creep and deformation during or after irradiation, already evoked in the previous section, was illustrated by four more papers No effect of com-mercial reactor irradiation temperature or hydrogen content was found on the Zircaloy-4 guide tube creep Zircaloy-4 irradiation creep law was deduced from relaxation experiments on bent beam specimens in OSIRIS Creep de-formation of Zr-1%Nb and Zircaloy-4 during transportation was associated

to a signifi cant recovery of irradiation damage, preventing the dislocation channelling Shadow corrosion-induced bow of Zircaloy-2 channels was as-sociated to differential hydrogen concentration on channels sides adjacent to and away from the control blades

Four papers dealt with accidental conditions In LOCA conditions, a sion model was developed to compute the oxygen distribution, closely related

diffu-to the mechanical properties, and compared with experiments An tion on the effect of hydrides on mechanical properties during RIA concluded that high strain rates did not seem to impact the stress strain behaviour when the hydrogen content is higher than 400ppm, threshold over which the failure elongation at room temperature decreases drastically When radially oriented hydrides are present in BWR cladding, an outside-in cracking can occur, which was investigated, in isothermal conditions or with a radial ther-mal gradient: the outside-in cracking during the power ramp seems strongly dependent on the distribution of dissolved hydrogen as a result of thermal diffusion Finally, a high burn-up BWR fuel rod, subjected to RIA tests in a research reactor, resulted in cladding failure at room temperature, but not

investiga-at elevinvestiga-ated temperinvestiga-ature A mechanical test was developed to reproduce the ramp test and was shown to be predictive

The John Schemel Award is awarded following each symposium for the best paper presented at the symposium The selection is based upon the technical content of the paper, the usefulness of the work reported to the worldwide reactor components community, and the technical diffi culty in do-ing the work This year, a committee of technical experts in several aspects

of the zirconium industry selected the paper entitled “Photoelectrochemical Investigation of Radiation Enhanced Shadow Corrosion Phenomenon” by Y.-J Kim, R Rebak, Y-P Lin, D Lutz, D Crawford, A Kucuk and B Cheng to receive the John Schemel Award

Pierre BarbérisAreva/Cezus Research Centre

Guest Editor

KROLL AWARD PAPERS

cor-of an explosive detonation to produce a metallurgical bond between metalplates Since its discovery and industrialization in the 1960s, explosion clad-ding has proven to be a highly reliable, robust manufacturing technology Zir-conium clad has become a broadly accepted material of construction for asignificant range of chemical process equipment Zirconium clad equipment isused in manufacture of an extensive range of chemical products or intermedi-ates including nitric acid, acetic acid, urea, and various organic acids Zirco-nium clad and related design, manufacture, and fabrication processes havebeen shown to be highly reliable and to reduce both capital and operating cost.KEYWORDS: zirconium, clad, explosion clad, corrosion control, pressurevessels, heat exchangers

IntroductionZirconium is commonly used as a structural material in the chemical processindustry (CPI) The primary driver is zirconium’s superior corrosion resistanceover a broader range of chemical environments and temperatures than virtually

Manuscript received February 18, 2010; accepted for publication June 23, 2010; published online July 2010.

Cite as: Banker, J G., “Explosion Cladding: An Enabling Technology for Zirconium in the Chemical Process Industry,” J ASTM Intl., Vol 7, No 8 doi:10.1520/JAI103050.

Conshohocken, PA 19428-2959.

3

Reprinted from JAI, Vol 7, No 8 doi:10.1520/JAI103050 Available online at www.astm.org/JAI

any other metal [1–3] It exhibits excellent resistance to corrosive attack in mostorganic and inorganic acids, salt solutions, and strong alkalis Zirconium’s abil-ity to withstand alternating acidic and basic environments is unusual Unlikemany of the other corrosion resistant alloys (CRAs), zirconium is considered to

be non-toxic From the technical perspective, zirconium is an exceptional rial for construction of chemical process pressure vessels, columns, heatexchangers, piping, and valves Zirconium alloy R60702 (ASTM B551 Alloy 702)[4], which is essentially chemically pure zirconium, is the most commonly usedgrade (Table 1) Although R60702 is more expensive than most other CRAs, itcontinues to be a far better value for some severe corrosion environments Mak-ing direct cost comparisons between R60702 and other CRAs is not simple due

mate-to significant variations in corrosion allowances, density, thermal conductivity,mechanical properties, maintenance costs, and failure related costs Today,R60702 is broadly used for thinner walled construction including tubes, pipes,columns, and similar equipments

Solid R60702 has traditionally been used much less in construction of ier equipment (typically driven by environments, which are high temperatureand/or high pressure.) This is primarily driven by two factors, cost and designstrength R60702 is relatively expensive compared to many of the other CRAs.Table 2 presents a comparison of the relative cost of several common CRAs aftermaking adjustments for design strength (at ambient temperature) and density.Second, the design strength allowed for R60702 in the ASME Boiler & Pressure

heav-TABLE 1—ASTM B551, zirconium grades commonly used in CPI equipment for corrosion resistance.

TABLE 2—Comparative costs of some CRAs.

Vessel Code [5] is lower than many of the alternative alloys Further, allowabledesign strength for R60702 drops off more quickly with increasing temperaturethan that of stainless steels and nickel alloys (Fig 1) Consequently, the costproblem is compounded by the need for heavier wall equipment when solidR60702 construction is considered, and even more so when elevated tempera-ture performance is required.

In most chemical process applications, the corrosive media is in contactwith only one face of the structural containment material For reactors and col-umns, typically the interior surface is exposed to the corrosive contents Theexterior surface sees a much less severe environment, frequently ambient condi-tions Costs can be reduced significantly if only a thin layer of zirconium is used

in contact with the process media, and a thicker and lower cost metal is used asthe structural component For some applications, loose linings of zirconiuminside of a steel containment vessel are suitable However, in process applica-tions involving elevated temperatures and/or pressures and/or cyclic conditions,loose linings often fail quickly Loose linings are also inefficient when through-wall heat transfer is important

Construction using clad plates overcomes the shortcomings of loose-linedequipment Clad equipment has decades of demonstrated performance in CPIinstallations The cost reductions achieved with zirconium clad enable the use

of zirconium in many applications where R60702 would otherwise be ered cost prohibitive Figure 2 shows the cost difference between zirconiumclad and solid R60702 over a broad range of thicknesses The cost benefits areachieved by reducing the amount of zirconium used and further by reducingthe overall equipment thickness as a result of the higher allowable design stress

consid-of steel The graph shows two clad thickness combinations Clad consid-of 3 mm thick

FIG 1—Comparative strength R60702 and other alloys

BANKER, doi:10.1520/JAI103050 5

Zr bonded to steel (ASTM A516-70) [6] is commonly used for pressure vessels Athicker cladding layer, 8 mm, is commonly used for heat exchanger tubesheets.

As shown in the graph, clad typically offers cost benefits over solid R60702 struction when the required wall thickness of R60702 exceeds 15 mm Many oftoday’s pressure vessel and heat exchanger applications that use zirconium cladwould not be commercially viable if solid R60702 construction was the onlyoption

con-Clad ManufactureSeveral industrial processes can be used for making clad steel plates when theCRA alloy is metallurgically compatible with steel at elevated temperatures.These include hot roll bonding, weld overlay, and explosion cladding CRAalloys that are compatible with steel include the stainless steels and many of thenickel alloys The clad bond produced during hot roll bonding is very similar tothat produced in the coextrusion cladding of Zr and Zircaloy-2 for manufacture

of barrier and triclad tubing [7] which is also discussed in the earlier KrollMedal paper [8]

When the CRA alloy is not metallurgically compatible with steel, explosioncladding is generally the only viable clad manufacturing process Zirconium isone of these alloys Efforts to produce zirconium clad using hot rollbonding orweld overlay typically result in a brittle and highly cracked product The Zr–Fephase diagram explains this incompatibility [9] The major problems are thehighly stable ZrFe2 intermetallic phase plus the low melting ð934CÞ Zr richeutectic (84 % Zr) These two conditions make hot rollbonding and weld overlay

FIG 2—Comparative cost of zirconium clad versus solid R60702, including ments for design strength Base metal is SA-516 Gr 70 steel

virtually impossible unless interlayer metals, compatible with both Fe and Zr,are inserted On the other hand, explosion cladding works well for making zir-conium clad.

Explosion CladdingThe Detaclad explosion cladding technology was patented and industrialized bythe DuPont Chemical Co in the 1960s [10–13] Explosion cladding utilizes theenergy of an explosive detonation to achieve welding between two metal compo-nents During the explosion cladding event, sheets of the CRA alloy are welded

to plates of steel over the full contacting face area Clad plate sizes can rangefrom a few square centimetres up to around 50 m2 The explosion clad interface

is strong, tough, and highly reliable Typical bond shear strength of zirconiumclad manufactured Dynamic Materials Corporation (DMC) is around 200 MPa.Figure 3 depicts the Detaclad explosion clad manufacturing process A broaderdiscussion of the technology is presented in the references cited above

Figure 4 presents a photomicrograph of an explosion clad interface Thewavy bond morphology is typical of the process Optical microscopy shows noevidence of heating

 No significant melt layers can be observed

 There is significant cold work in the waves and no recrystallization isevident

 Scanning electron microscopy studies show no evidence of diffusion.Consequently explosion cladding is considered to be a cold welding process.Early researchers referred to it as a “solid state electron sharing bond.” TheFIG 3—DMC’s Detaclad explosive cladding process for manufacture of zirconium clad

BANKER, doi:10.1520/JAI103050 7

bond interface was initially assumed to be similar in nature to a high energygrain boundary.

In more recent years, transmission electron microscopy (TEM) studies haveprovided better insight into the nature of the interface and the bonding mecha-nism TEM studies indicate that very significant heating is occurring at theinterface but on a scale barely observable with optical microscopy [14–16].These TEM studies indicate the presence of a thin metallurgical weld,0:1–0:2 lm thick Further, they report indications of an amorphous structure ofthe resolidified weld metal This suggests that the time at elevated temperature

is extremely short and the cooling rates are extremely fast Other research dataindicate that the metals are heated to the welding conditions and returned towell below the melting point in less than 20 microseconds The result is a meta-stable metallurgical condition at the interface The stable and deleteriousphases shown in the Zr–Fe phase diagram do not have time to form Today, theexplosion clad interface is frequently referred to as a “very thin, hyper cooled,fusion weld.” One needs to be cautious when drawing broad conclusions fromthese studies The number of studies of this type has been severely limited bythe high cost of TEM work and the relatively small budgets of the explosioncladding companies and research groups It is estimated that less than 20 speci-mens were extensively analyzed by TEM prior to 2008 Considering the area an-alyzed in a single TEM specimen, this is an extremely small sampling ratio.Modern nano-scale analytical capabilities are beginning to provide furtherknowledge of the interface Figure 5 presents TEM data from work performed

by Song and others at Max-Planck Institute in 2008 and 2009 Although thework is on titanium-steel samples, we would expect similar results with zirco-nium clad Ongoing studies at Max-Planck Institute, Oak Ridge National

FIG 4—Typical explosion clad interface (50)

Laboratory, and Colorado School of Mines are beginning to provide cantly more insight and understanding of the product and process.

signifi-Zirconium Clad Development and ManufactureZirconium-steel clad is generally considered to be the most difficult metal com-bination to produce by explosion cladding [17,18] Contributing factors includethe limited number of slip planes of the Zr crystal lattice, the high energy of for-mation of the ZrFe2intermetallic and the low melting temperature of the Zr–Fe(84–66 wt %) eutectic

DuPont researchers developed zirconium clad manufacturing parameters

in the 1960s In the initial work, they observed the following

 It was much easier to clad R60702 to titanium than to clad R60702direct to steel

 It was relatively easy to clad titanium to steel

 Consequently, the use of a titanium interlayer between R60702 and steelwas deemed to greatly simplify the process

In the 1980s, researchers at the Explosive Fabricators, Inc (now DMC)developed procedures for reliable industrial manufacture of high qualityzirconium-steel clad without an interlayer This effort involved the development

of a number of process variants, which are unique to zirconium Further, theyobserved that using zirconium with higher purity and lower strength than thatmandated in the ASTM B551 for R60702 was highly beneficial In support ofthis effort, ATI Wah Chang agreed to supply a variant of R60702 with reducedoxygen content (1000 ppm maximum), reduced strength, and enhanced ductil-ity This alloy, which was commonly referred to as “Zr 702 Low Ox” for manyyears, became the alloy of choice for zirconium clad In 2003, that higher purityvariant was established as R60700 and added to ASTM B551

Today, clads of both zirconium bonded direct to steel and zirconiumbonded to steel using a titanium interlayer can be found in the marketplace.Factors affecting the manufacturing route are cost, material availability, cus-tomer preference, and clad manufacturer’s know-how Considerable productiondata over the past 25 years demonstrate that zirconium clad produced by direct

FIG 5—TEM analysis of titanium-steel interface (scale of picture at left:

1 cm=250 nm)

BANKER, doi:10.1520/JAI103050 9

bonding can be a highly reliable material for construction of process equipment[19].

Zirconium Clad for the Chemical Process IndustryClad production for the first large CPI zirconium clad project was performed byExplosive Fabricators, Inc., in 1981 The clad plates were used to construct a ro-tary kiln for a zirconium producer During the 1980s, zirconium clad pressurevessels and heat exchangers were selected for a number of highly corrosivechemical processes These included pressure vessels for manufacture of aceticacid and various industrial and specialty chemical products They also includedheat exchangers for manufacture of nitric and numerous other acids Detailsregarding the size and operating conditions of several of these units are pre-sented in “Commercial Applications of Explosive Clad” published by ASTM in

1996 [20,21] Figures 6–11 show several of these units and components ofconstruction

The pressure vessel for acetic acid manufacture provides a good case study

of vessel design options and considerations This unit was used in a licensedmethanol carbonylation process The process had been developed in the late1960s and early 1970s and had originally used nickel alloy equipment The spe-cific nickel alloy being used provided adequate corrosion protection; however,the alloy exhibited a tendency to fail by stress corrosion cracking unless thewelded structure was precisely heat treated and water quenched after complete

FIG 6—Nitric acid cooler condenser, zirconium tubes, zirconium-stainless steel cladtubesheet

fabrication The reliable manufacturing window was exceptionally narrow.Ongoing fabrication challenges encouraged the process management team toseek a more predictable material The background of the replacement of nickelalloy with zirconium in this process is presented in Sanders’ Kroll Award Paper[22] Much of the process equipment was relatively light wall and could easilyjustify a change from nickel alloy to zirconium on cost reduction On the otherhand, the main pressure vessel presented a significant cost challenge It was

 3:5 m diameter  5:0 m tall and operated under conditions that required wellover 100 mm wall thickness of R60702 The R60702 design was considerablymore expensive than nickel alloy, so sufficiently that the increased performance

of zirconium could not alone justify the higher capital cost In the early 1980s,one of the process licensees chose to proceed with installation of a zirconiumclad main pressure vessel This would be the first heavy-walled zirconium cladvessel ever constructed and required numerous design and fabrication innova-tions The zirconium clad pressure vessel had lower cost than the traditionalnickel alloy unit Further, the superior corrosion performance of zirconiumallowed the unit to be operated under more aggressive conditions, improvingprocess yields Subsequently zirconium clad has become the standard for thisand similar acetic acid pressure vessels

FIG 7—Interior view of zirconium clad column during fabrication, 2.5 m diameter

BANKER, doi:10.1520/JAI103050 11

Figure 6 shows a typical nitric acid heat exchanger This unit is a good casestudy of the benefits of clad in shell-and-tube heat exchanger design The tube-sheet is explosion clad Zr/stainless steel The clad tubesheet enables the designer

to select zirconium for the tube-side components and stainless steel for the shellside components but to still be able to fully weld the tubes to the tubesheet andFIG 8—Zirconium clad heads hot formed after cladding Largest are 3.7 m diameter

FIG 9—Zirconium clad reactor with external steel heating jacket 3.0 m diameterwith wall thickness of 25 mm steel +3 mm Zr

the tubesheet to the shell, eliminating potential leakage problems Combiningthe benefits of Zr tube-to-tubesheet welds, the cost reductions of a steel shell,and the fully-welded shell to tubesheet joint is not possible if solid zirconiumtubesheets are used.

The road to broad CPI industry acceptance of zirconium clad involvedmany efforts in addition to simply learning how to manufacture the clad Theseincluded the development of fabrication techniques [23,24], the development ofASTM B898 for reactive metal clad [25,26], and the development of a zirconiumgrade specifically designed for explosion cladding, ASTM B551 Alloy 700

FIG 10—Zirconium clad tubesheet blanks for a Urea process heat exchanger sheets are 2750 mm diameter, 500 mm steel thickness, and 10 mm R60700 thickness

Tube-BANKER, doi:10.1520/JAI103050 13

Today, zirconium explosion clad is broadly accepted as a reliable material

of construction for CPI equipment The cost reductions of clad constructionhave enabled the use of zirconium in a number of CPI processes where solid zir-conium construction had previously been rejected due to it very high cost Areview of manufacturing records indicates that three zirconium clad pressurevessels were put into service in the 1980s, 12 in the 1990s, and over 30 in the2000–2009 time period Over this period, hundreds of zirconium heat exchang-ers with clad tubesheets were put into service Virtually all of this equipmenthas provided highly reliable service and continues in operation When therehave been problems, they have typically been traced to untrained equipmentfabricators or engineers duplicating equipment previously designed in stainlesssteel without making modifications for the differences in basic physical and me-chanical properties between stainless steel and zirconium [27]

ConclusionZirconium is an exceptional material for containment of corrosive media in CPIprocesses The corrosion resistance of zirconium is exceptional in a broad range

of aggressively corrosive environments For heavier wall equipment, it is cult for the superior benefits of zirconium to justify the increased cost of solidzirconium when compared to alternative but inferior CRAs Explosion claddinghas enabled CPI companies to benefit from the superior performance features

diffi-of zirconium in heavy-wall pressure vessels and heat exchangers in a costeffective way Zirconium clad has become a broadly accepted material of

FIG 11—Typical large shell-and-tube heat exchanger with reactive metal cladtubesheets

construction for a broad range of chemical process equipment Zirconium claddesign, manufacture, and fabrication process, which have been developed overthe past 30 years, are highly reliable, robust, and reduce both capital and oper-ating cost.

References

Technology, M G Fontana and R W Staehle, Eds., Plenum Press, New York, 1976, Vol 5, p 173.

ASM Handbook, Vol 13, 9th ed., ASM International, Metals Park, OH, 1987, p 707.

Sheet, Strip and Plate,” Annual Book of ASTM Standards, Vol 02.04, ASTM tional, West Conshohocken, PA, pp 510–520.

American Society of Mechanical Engineers, New York, 1998.

Car-bon Steel, for Moderate- and Lower-Temperature Service,” Annual Book of ASTM Standards, Vol 01.04, ASTM International, West Conshohocken, PA,

pp 281–284.

Armijo, J S., “Zircaloy-2 Lined Zirconium Barrier Fuel Cladding,” Zirconium in the Nuclear Industry: 11th International Symposium, ASTM STP 1295, Garmisch- Partenkerchen, Germany, 1996, E R Bradely and G P Sabol, Eds., ASTM Interna- tional, West Conshohocken, PA, pp 676–694.

Zirconium-Barrier Fuel Cladding,” Zirconium in the Nuclear Industry: Tenth International posium, ASTM STP 1245, Baltimore, MD, 1994, A M Garde and E R Bradeley, Eds., ASTM International, West Conshohocken, PA, pp 3–18.

1958, p 742.

Research Council Bulletin 104, American Welding Society, New York, April 1965.

15, 3rd ed., John Wiley & Sons, New York, 1981, pp 275–296.

Pat-ent 3,137,937 (1964).

Soldering, ASM Handbook, Vol 6, ASM International, Metals Park, OH, pp 303–305.

Affected by the Bonding Process During Oblique Collision of Metallic Surfaces,” Fifth International Symposium Explosive Working of Metals, Gottwaldov, Oct 12–14, 1982, Czechoslovak Scientific and Technical Society, Gottwaldov, Czechoslovakia.

Explosively-Welded Clads and Bonding Mechanism,” Mater Sci Forum, Vol 465–475, 2004, pp 465–474.

BANKER, doi:10.1520/JAI103050 15

[16] Lafont, M C., Masri, T., and Nobili, A., “Recent Developments in Characterization

of a Titanium-Steel Explosion Bond Interface,” Reactive Metals in Corrosive cations Conference Proceedings, Sun River, OR, 1999, J Haygarth and J Tosdale, Eds., Wah Chang, Albany, OR, pp 89–98.

Equipment: Applications, Design, Fabrication,” Proceedings of 1997 Zirconium/ Organics Conference, Gleneden Beach, OR, September 1997, ATI Wah Chang, Albany, OR, pp 71–78.

Clad,” Proceedings 1999 Reactive Metals in Corrosive Applications Conference, Sun River, OR, September 1999, ATI Wah Chang, Albany, OR, pp 83–88.

Explo-sion Clad Manufacturing,” Proceedings CorroExplo-sion Solutions Conference, Sun River,

OR, September 2001, ATI Wah Chang, Albany, OR.

Vol 24, No 2, 1996, pp 91–95.

ASTM Symposium on Industrial Applications of Titanium and Zirconium, Phoenix,

AZ, November 1994, ASTM International, West Conshohocken, PA.

Carbonylation Acetic Acid Process,” The Kroll Medal Papers (1975–2010): Unique and Lasting Contributions to the Technology of Zirconium, ASTM International, West Conshohocken, PA, 2010.

Highly Corrosive HCl Service,” Proceedings Corrosion Solutions Conference, Sun River, OR, September 2001, ATI Wah Chang, Albany, OR.

Consider-ations,” Proceedings Corrosion Solutions Conference, Park City, OR, September

2009, ATI Wah Chang, Albany, OR.

Plate,” Annual Book of ASTM Standards, Vol 02.04, ASTM International, West shohocken, PA, pp 1112–1118.

Metal Plate Specification,” Proceedings Corrosion Solutions Conference, Sun River,

OR, September 2001, ATI Wah Chang, Albany, OR.

Exchanger Failures,” Proceedings Corrosion Solutions Conference, Park City, UT, September 2009, ATI Wah Chang, Albany, OR.

David G Franklin1

Performance of Zirconium Alloys in Light Water Reactors with a Review of Nodular Corrosion

ABSTRACT: This paper provides historical context and future direction onselected contributions to understanding zirconium-alloy deformation and cor-rosion in light water reactors More detailed discussion of the relative impor-tance of electrochemical potential and hydrogen pickup on nodular corrosion

is provided Since nodular corrosion was observed about 50 years ago, eral explanations have been proposed None have explained all the observa-tions, especially the sharp transition between nodular-free and nodular-covered regions of some fuel rods, with the transition being reversed betweenhigh- and low-power regions However, the effect of electrochemical potentialdominates any effects of hydrogen, other than the effect of hydrogen on elec-trochemical potential

sev-KEYWORDS: Zircaloy, corrosion, deformationIntroduction

Concerning the Kroll Medal award, all of the writer’s supporting colleaguescannot be mentioned for there are too many, but some must be At the BettisAtomic Power Laboratory, the birthplace of the zirconium alloys, the writerbenefited greatly from the contributions of other Bettis staff, past and present.Those privileged to work within the Naval Nuclear Propulsion Program bene-fited from the culture built and left as a legacy by Admiral Rickover, the firstKroll Medal recipient Pieter Kreyns and Bruce Kammenzind deserve specialthanks While at the Electric Power Research Institute, Albert Machiels, Ste-phen Gehl, Joseph Santucci, and Rosa Yang provided much support; Professor

Manuscript received February 9, 2010; accepted for publication May 7, 2010; published online July 2010.

1

Retired from Bettis Atomic Power Laboratory now at 2212 Grier Woods Ct., Las Vegas,

NV 89134.

Cite as: Franklin, D G., “Performance of Zirconium Alloys in Light Water Reactors with

a Review of Nodular Corrosion,” J ASTM Intl., Vol 7, No 6 doi:10.1520/JAI103032.

Conshohocken, PA 19428-2959.

17

Reprinted from JAI, Vol 7, No 6 doi:10.1520/JAI103032 Available online at www.astm.org/JAI

Eugene Lucas at the University of California in Santa Barbara was instrumental

in preparation of the monograph on creep of zirconium alloys [1]; andRonald Adamson contributed greatly to my understanding of deformationphenomena

Three ongoing or recent developments result in the need for additionalunderstanding about the performance of zirconium alloys in nuclear applica-tions First, new reactor designs that could result in different operating condi-tions continue to be developed Second, the drive to reduce fuel-cycle costcontinues and results in additional demand on zirconium-alloy structural mate-rials, especially fuel-element cladding Third, a change in the U.S plans to closethe fuel cycle may result in much longer spent-fuel storage times prior to dis-posal, which could increase demand for the zirconium-alloy core structuralmaterials These developments are considered in reviewing the contributions.One topic, nodular corrosion, is selected for review in detail

General Aspects of Performance of Zirconium Alloys in NuclearApplications

Corrosion and deformation continue to be important phenomena for future clear applications Other properties not addressed herein have varying degrees

nu-of importance, for example, hydrogen effects remain important, especially atextended exposures, while physical properties like conductivity are relativelywell known Although off-normal conditions, for example, loss-of-coolant condi-tions, will not be addressed directly, one historical perspective is worth notingbecause of the reminder it provides In 1984, several of us concluded that work

on high-temperature properties was sufficient and that this area of study should

be reviewed and summarized for the next ASTM symposium on Zirconium inthe Nuclear Industry with an eye toward reducing resources applied to thisarea We asked F.J Erbacher of the Institut fu†r Reaktorbauelemente in Ger-many to prepare a review paper for the 1985 Strasbourg, France symposium

He and S Leistikow provided an excellent paper [2], and no sessions on thistopic were included for the next five symposia However, our optimism was notjustified, due in part to the industry drive to increase exposures This topic reap-peared as a session at the 13th symposium in 2001 We were reminded that allaspects of performance must be monitored and considered as the industry pro-gresses, conditions change, and knowledge expands

Concerning deformation, the impacts of mispredicting deformation can beunacceptable Several of the early and dramatic issues with the use of zirconiumalloys as fuel cladding were due to misunderstanding deformation In 1988,Ron Adamson and the writer provided a review of the effects of creep andgrowth of Zircaloy-clad fuel rods [3], showing some of the observed effects ofnot adequately taking into account the effects of creep and growth Two dra-matic examples are (1) cladding creep and collapse into fueled regions vacated

by densified fuel and (2) fuel-rod interactions due to growth For new staff ing on in-reactor fuel performance, a review of those dramatic examples isworthwhile, both to understand the past and to realize the importance of theirwork Since that time, much work has been done, especially on engineering

aspects of creep and growth, with much of the work reported in the ASTM posium series on Zirconium in the nuclear industry These studies extend thebasis for predicting behavior to new alloys and increased exposures Finally, incontrast to the performance of stainless steels in liquid-metal fast breeder reac-tors, zirconium alloys did not swell significantly from the formation ofirradiation-induced voids for light water reactor (LWR) conditions, so an engi-neering need to model this phenomenon did not develop.

sym-Cladding creep was a concern in the 1960s and 1970s due to its impact onthe fuel-to-cladding gap, which, in addition to the example above of claddingcollapse, influenced fuel-pellet stored energy, mechanical interactions, includ-ing stress corrosion cracking, and fuel-rod growth In pressure-tube reactors,for example, the Canadian reactors, creep of the pressure tube also was an issuedue to the long exposure times The effects of irradiation, which enhanced creepduring irradiation but inhibited creep postirradiation, were surprising andproved to be a challenge to measure and model During this period, especiallybefore fuel densification was resolved, in-reactor deformation and mechanicalproperties were considered by many to be more challenging than corrosion Ini-tially, modeling was fairly simple with some phenomenological aspects in corre-lations In the middle 1970s, we attempted to include grain orientation andtexture into creep modeling [4], but the general tools were not advanced enough

to result in direct applications As described by Nichols in his Kroll acceptancepaper [5], this situation was being remedied by advances both in material-modeling techniques and computing power Tenckhoff in his Kroll acceptancepaper [6] summarized texture aspects

In the 1970s, the Materials Research Council asked A.L Bement, G.E.Lucas, and the writer to prepare a monograph on creep of zirconium alloys innuclear applications, which ASTM published [1] That effort covered engineer-ing, modeling, and mechanistic studies Recent models generally are phenome-nological incorporations of detailed deformation behavior, for example, byGelebart et al [7] and Allen et al [8] In parallel, the major Nuclear Steam Sup-ply System providers and some research organizations, for example, the Depart-ment of Energy and the Electric Power Research Institute in the United States,developed larger databases Still, additional benefit can be obtained from the de-velopment of data that support more-mechanistic models

Concerning corrosion, the selection of zirconium by the U.S Naval clear Propulsion Program as the base metal was, in part, based on the needfor corrosion resistance Therefore, it has been studied and reviewed manytimes, starting with those who preceded the writer at Bettis, for exampleKass [9] and Hillner [10], and continuing through recent reviews by careerexperts such as Brian Cox Cox’s 2005 review 11 provides a good perspective

Nu-on nodular corrosiNu-on, although the cNu-onclusiNu-ons below are not the same in allrespects Despite the extensive studies on corrosion, the fundamentals are notfully revealed

The need for additional research on deformation and corrosion is driven bythe three changes discussed in the introduction With new alloys and new condi-tions associated with either new reactor designs or increased exposures, addi-tional studies will be needed to confirm or revise existing models An important

FRANKLIN, doi:10.1520/JAI103032 19

recent change is due to the decision in the United States to reconsider the bestway to deal with high-level waste Specifically, the Department of Energy is con-sidering options other than disposal at a Yucca Mountain repository Someoptions would result in very-long-term storage of spent-fuel assemblies Theresult could be that there is a need to ensure fuel cladding and assembly integ-rity during very-long-term storage, in part due to the possibility of subsequenthandling Cladding creep, rupture, and handling damage need to be understoodfor the conditions selected, although those conditions are not known fully.Hydrogen effects may be important due to the effect on mechanical perform-ance Work is being performed in this area, for example, as reported by Tsaiand Billone [12] at the 14th ASTM symposium and by a session with threepapers at the 15th symposium [13–15].

For very-long-term storage, inert atmospheres would prevent corrosion andhydrogen pickup If water or air environments need to be considered in eitherstorage or disposal, Bettis addressed such corrosion for Zircaloy based on 30-year autoclave data [16] We concluded that the postirradiation corrosion ratereturns to or returns close to the nonirradiated corrosion rate, allowing the use

of these nonirradiated data This conclusion was based on in-reactor flux andtemperature-shift tests and on two previous postirradiation studies, one by Gar-zarolli et al [17] and one by Cheng et al [18] Although the latter paper had con-fusing results for 400C steam, the more-relevant results for 316C water werenot ambiguous, nor were the results of Garzarolli et al More recently, Iltis et al.[19] studied postirradiation corrosion effects and the associated causes, espe-cially the effects of irradiation-induced iron dissolution and in-reactor environ-mental parameters Their results suggested that iron in solution and itsreprecipitation could be controlling the postirradiation results reported previ-ously [17,18] However, their 400C postirradiation tests were confusingbecause there appeared to be irradiation enhancement between 60 and 210days, but not for the next test period of over 100 days Since the Cheng et al.[18] results were also mixed at 400C, model databases may need to be limited

to postirradiation testing performed at lower more-prototypical temperatures toobtain unambiguous results Considering that inert environments may be usedfor storage and that the corrosion rates can be predicted if water or air is used,the corrosion and hydrogen pickup can be predicted for long storage times forcladding with normal exposures Some fundamental results that were intended

to apply to in-reactor corrosion may help to understand postirradiation formance, especially the memory effects for changing from in-reactor to postir-radiation corrosion For example, studies on the influence of solutes in theoxide on corrosion, as discussed by Cheng and Adamson [20], and the results ofIltis et al [19] on iron effects, may be a part of the memory effect For high-exposure cladding, there may be a need for additional postirradiation testing ifair environments must be presumed Some relevant fundamental work is beingdone on postirradiation corrosion properties, for example, using impedancespectroscopy, which began in the 1950s and became fairly sophisticated by1980s [21] Although valuable, these techniques are not as direct as the postirra-diation corrosion measurements and the results need to be considered subservi-ent to the direct measurements

Nodular CorrosionNodular corrosion was reported first in the 1960s [22–25] There was a concernthat this accelerated localized corrosion would lead to cladding failures, andmuch work was initiated to understand nodular corrosion It was not initiallyrealized that there was a self-limiting aspect to nodular corrosion for most LWRconditions This phenomenon does not appear only as isolated surface spots Itcan appear as accelerated corrosion that covers much or all of a surface It isdistinguished better by the conditions that cause the accelerated corrosion than

by its appearance At reactor temperatures, nodular corrosion requires ing conditions and irradiation, as summarized in many reviews [26,27].Although dissolved coolant O2 concentration can be simulated out of reactor,sufficient irradiation is difficult to simulate [28] Nodule formation in LWRsmay be promoted by lower temperatures, which would reduce the rate at whichthermal processes remove the effects of irradiation, but such an effect has notbeen demonstrated conclusively

oxidiz-At the 1985 ASTM symposium, several papers addressed important aspects

of nodular corrosion Important work at General Electric (GE) was presentedand the paper by Cheng and Adamson [20] provided a particularly importantcontribution, which is discussed later The most perplexing of the observations

of nodular corrosion, and as yet not well explained, is the sharp change in lar coverage in a boiling-water reactor (BWR) at the blanket-to-fuel transition,which we discussed and correlated with heat flux and other variables [29] Inlow-power fuel rods, there was a sharp transition from no nodules to nodules atapproximately the axial height of the transition from the lower blanket to thefueled region High-flux rods had the opposite observation That paper dis-cussed the possible impact of heat flux and other variables on the observed nod-ule formation, but no fundamental or mechanistic understanding of theseobservations was available Effects of heat flux had been reported earlier, forexample, by Urquhart and Vermilyea [30], who reported that more nodulesformed on channels with no heat flux than on fuel-rod cladding Marlowe et al.[31] reported two similar observations on the effects of heat flux: (1) Gd-doped

nodu-UO2rods, which had low-heat-fluxes early in life, had more nodules than UO2rods and (2) nonfueled regions of test rods had more nodules than fueledregions Mishima and Aoki 32 reported grayish-white corrosion at the bottom offuel rods, which could be full nodular coverage, although they did not describe

it as nodular corrosion Etoh 33 also investigated the effects of heat flux andfound a correlation between nodule thickness and heat flux

The previous evaluation of the nodule coverage change at approximatelythe enrichment transition [29] was not able to correlate this change with more-fundamental variables than heat flux Previous studies of nodular corrosionassociated nodule formation with electrochemical potential and hydrogenpickup However, the relative importance of these and other variables, and howthey affect nodule formation are not understood fully Since the roles of hydro-gen and electrochemical potential are the primary candidates in the literaturefor causing nodule formation, they will be the focus of this discussion The dis-cussion is limited to exposures typical for the occurrence of nodules, which

FRANKLIN, doi:10.1520/JAI103032 21

usually appear in the first reactor cycle [32], well before corrosion in BWRsaccelerates after four or five cycles [34] This limitation on exposure is impor-tant because the effects of hydrogen pickup may be different at high exposuresand large hydrogen pickup Etoh et al [35] demonstrated this limited time pe-riod for nodule formation in postirradiation tests on nodular-susceptible Zirca-loy-2.

Effect of HydrogenDetrimental effects of hydrogen pickup were considered early in the develop-ment of zirconium alloys [36] Several investigators suggested that increasinghydrogen concentration near the metal-to-oxide interface contributed to noduleformation [20,37–39] However, hydrogen concentration, both in the coolantand in Zircaloys that formed nodules in reactor, did not correlate, and probablycorrelated inversely with nodule formation The primary observations were:

 Increasing H2 concentration in the coolant suppressed nodule tion in LWRs [40,41]

forma- Hydrogen pickup fraction in Zircaloy in BWRs was low at the exposureswhere nodules formed [25,32]

 Specialized test-reactor experiments showed no correlation, as shown

by Aomi et al [41] and Takagawa et al [42] with experiments in theHalden test reactor Increasing dissolved H2 in the water to 400 partsper billion (ppb) eliminated nodules, despite an increase in hydrogenconcentration in the samples

 Hydrogen pickup correlation with second-phase particle (SPP) size wasinconsistent with the effect of solutes on nodule formation Forinstance, Huang et al [43] and Tagstrom et al [44] found that thehydrogen pickup increased with increasing solute concentration andwith decreasing initial SPP sizes This observation, when combinedwith Cheng and Adamson’s [20] demonstration that nodule resistanceincreased with decreasing initial SPP size, suggested that, if anything,nodule formation correlated with low hydrogen pickup

 The sharp transition in corrosion coverage at the enrichment transitiondid not support hydrogen being a significant factor because there was

no basis for an associated variation in hydrogen concentration at theenrichment transition

Hydrogen transport through the oxide has been related to the potential acrossthe oxide [45] However, an increase in the coolant dissolved H2content decreasesthe oxide potential drop, as demonstrated in the section below titled Electrochemi-cal Potential Therefore, increased water H2content decreases the driving force ofthe potential gradient on Hþdiffusion through the oxide In the end, the sources

of each of the drivers for hydrogen diffusion, and oxygen migration for that ter, are the changes in Gibbs free energy from oxidation and reduction reactions,

mat-so the individual processes need to not be considered in imat-solation

There are suggestions that a greater hydrogen concentration at the base ofnodules than away from nodules supports hydrogen as the cause [38] However,

if there is increased hydrogen at the base of nodules, it is likely the result of the

breakdown of the oxide film rather than the cause of the breakdown, given theobservations discussed above In supporting hydrogen as the cause, Rudling andWikmark [38] referred to pressurized water reactors (PWRs) not forming nod-ules because PWR temperatures are high enough for hydrogen to diffuse fromthe base of an SPP before forming a hydride, while this diffusion is slower inBWRs due to the lower BWR temperatures However, when a PWR operates withlow dissolved H2in the coolant, nodules do form, as observed in Obrigheim andother reactors, and discussed above Therefore, the correlating difference in BWRand PWR nodule susceptibility is water chemistry, specifically, water hydrogenconcentration and electrochemical potential, and not temperature, with increas-ing dissolved H2suppressing rather than promoting nodule formation That is not

to say that temperature does not affect nodule susceptibility The rates of nation of radiolytically produced species and the transport of anion vacancies andelectrons through the oxide increase with increasing temperature The impacts ofsuch changes are discussed below in the section on polarization

recombi-In conclusion, hydrogen concentrations, at least at concentrations observed

in BWRs through several cycles, do not contribute to nodule formation This clusion does not apply to zirconium alloys at high exposures with high hydrogenconcentrations, but this limitation is not important since nodules form early

con-Effects of Second-Phase Particles (SPPs)The nodule-formation requirement for both irradiation flux and low coolant H2

concentration is related to the concentration of oxidizing species at the oxidebarrier-layer interface with the water This concentration affects the electro-chemical potential2 at that surface Irradiation produces oxidizing species,which increase the potential at the external surface Coolant H2concentration,recombination, transfer into the steam phase, and transport through the oxideremove oxidizing species, which decreases the external-surface potential

By 1985, the effects of SPPs on nodular corrosion, especially in autoclavetests, were extensively studied, as evidenced by a number of papers at the sev-enth ASTM symposium [20,46–51] It was clear that nodule formation corre-lated with large SPPs and resistance to nodule formation correlated with smallSPPs The Cheng and Adamson paper [20] suggested a mechanism based on sol-ute concentration in solution, of which most parts have been supported by sub-sequent studies Ogata [52] suggested that nodule frequency correlated withregions of low-solute concentration, supporting Cheng and Adamson In a varia-tion on the Cheng and Adamson model, Etoh [33] suggested that irradiationflux homogenizes solute concentrations, especially Fe and Ni solutes, through-out the matrix Subsequent reprecipitation creates neighboring regions of low-solute concentration Etoh still considered nodules to form in regions of low-solute concentrations, but the proposed way these regions formed was different.Although many supported the conclusion that nodules form in matrix devoid ofsolutes, it is not accepted universally, the above-discussed model by Rudling

2

When potential is used herein, it refers to the electrochemical potential.

FRANKLIN, doi:10.1520/JAI103032 23

and Wikmark [38] being an example Based on the preponderance of the mation, it is concluded that the Cheng and Adamson model for nodule forma-tion at regions low in solute concentration is likely correct.

infor-Effects of Variables of Lesser ImportanceCox included in his comprehensive review of corrosion [53], which he updated

in 2005 [11], a discussion of variables that can impact corrosion In the presentdiscussion, only a few of the variables presented by Cox will be discussed Me-chanical and structural aspects were suggested early on to contribute to orcause oxide breakdown or nodule formation The large volume expansion onZrO2formation creates large compressive stresses in the oxide, which promotesdeformation and fracture However, Zhou [54] could not find oxide damage tosupport it as the cause of nodule formation He concluded nodules form partially

as a result of deformation in the metal, which produces dislocation short circuitsfor oxygen diffusion into the metal, which increases the anion-vacancy concen-tration in the oxide and oxygen diffusion through the oxide Nodule formationmay also be influenced by texture [55], grain size, and cold work Charquet et al.[56] considered the effects of grain size and texture and concluded that they havelittle effect for nodule-resistant fine-SPP and nodule-susceptible large-SPP mate-rials However, for intermediate materials, if several favorably oriented grainsare located together, he suggested that nodule formation could be promoted.Oxide dissolution and reprecipitation was shown by Nishino et al [28] tooccur under certain conditions and to increase with gamma irradiation Thisphenomenon has also been associated with the transformation from tetragonal

to monoclinic zirconia They suggested that this process produced microcracks

in the oxide, which contributed to the oxide breakdown and nodule formation.This process would be promoted by high pH, particularly from LiOH additions.Although this process cannot be excluded from contributing to oxide break-down, the demonstrated effect of coolant dissolved H2 concentration to sup-press nodule formation and the relatively low pH in BWRs makes it unlikelythat this process is a significant contributor to nodule formation

These and other factors may have some effect, but the results are mixed Aclue to their importance is provided by the observations that we are trying tounderstand; the sharp transitions in nodular corrosion along the axis of fuelrods [29] at blanket-to-fuel transitions Material properties such as grain size,texture, and deformation do not vary systematically with such sharp transitionsalong the axis of the fuel cladding, but the variation in nodular formation does

To be important, these variables would need to be shown to work in concertwith a variable that varies axially, which has not been shown, and for whichthere is no theoretical basis Therefore, these properties are at most minor con-tributors to nodule susceptibility

Phenomena Driving Nodule FormationThe following factors will be evaluated for their contributions to nodular corro-sion: Electrochemical potential of the environment, modification of the

potential across the oxide by changes in the controlling reaction or changes inconcentration of the oxidizing species at the oxide barrier-layer external sur-face, and the effect that oxygen potential has on the nodule formation Theeffect of each is reviewed and then related to the observed sharp axial changes

in nodule formation near enrichment changes [29]

Electrochemical PotentialEarly work of researchers like Cox at AECL [57], Urquhart et al [37,58] at GE,and Rosecrans [21] at the Knolls Atomic Power Laboratory, demonstrated theimportance of the electrochemical nature of Zr-alloy corrosion Urquhart et al.measured 600 mV for nodule-susceptible Zircaloy in 500C 1500-psi steam The-oretical calculations for the in-reactor oxygen concentrations associated withthe conditions for nodule formation can be used to estimate these potentials.For example, in the International Atomic Energy Agency (IAEA) [27,59] reviews,radiolysis calculations illustrated that the irradiation of nonoxygenated waterresults in an increase in the O2 concentration, fO2g, by over eight orders ofmagnitude and fO

2g by over six orders of magnitude Hydrogen addition,which changes the water from nodule promoting to nodule resistant, to a sys-tem with 200 ppb dissolved fO2g was predicted to decrease fO2g by four orders

of magnitude and fO2g by even more Using the Nernst equation one can mate the effect on potential for such changes in concentration, that is, using

esti-RT lnða1=a2Þ=nF, where R is the gas constant ð8:314 J=deg  molÞ, T is absolutetemperature, a is activity (a1 and a2being the activities for a change betweenwater chemistries 1 and 2), n is the charge transfer, and F is Faraday’s constant(96 500 C/eq) With activity taken to be concentration for this application, theincrease in potential for concentration changes ða1=a2Þ of 105for n ¼ 2 at 300C

is about 280 mV This simple calculation is only for perspective as it ignores thepossibility of changes in the controlling reactions Still, it demonstrates thatthere is not an inconsistency between the expected effect of irradiation onpotential and the observed potentials that are associated with nodule formation.The effect of the other oxidizing species with large concentration changes is notstraightforward, but should promote nodule formation These results suggestthat nodule formation is associated with corrosion potentials of no more than

600 mV Now we consider when such corrosion potentials (electrochemicalpotential (ECP)) occur in reactors

Measurements of ECP in BWRs were made to support programs to reducethe stress corrosion cracking of stainless steels For example, Cowan [60], Het-tiarachchi et al [61,62] reported ECP as a function of coolant H2and O2 con-centrations, with and without the catalyzing effects of noble-metal surfaces Inthe absence of noble-metal surfaces, ECP decreased from above 200 mV toabout 400 mV (standard hydrogen electrode (SHE)) with increasing H2 con-centration With noble-metal surfaces, ECP decreased from about 50 to 500

mV, with much less H2required for the ECP decrease In summary, ECP ured in BWR conditions that experienced nodular corrosion are several hun-dred millivolts without H2 additions H2 additions reduce the ECP by 450–600

meas-mV Since the interpretation of in-reactor BWR measurements for the effects of

FRANKLIN, doi:10.1520/JAI103032 25

coolant H2and O2concentrations on ECP is complicated by the effects of ing, we consider recent measurements without boiling.

boil-Rishel et al [63], using a special loop in the Halden reactor in Norway to licate PWR conditions, measured the differences in potential between oxygenatedand hydrogenated nonboiling coolant conditions The reported potentials are notcorrected to be referenced to the SHE, but the application here needs only the rel-ative potential differences between conditions with H2added (time steps 22 and23) and with O2added (time steps 29 through 37) into the coolant For the high-flux in-core regions, that is, those regions within the booster rods, the potentialfor oxygenated conditions was 500–700 mV different from that for hydrogenatedconditions and somewhat less for the region outside the booster rods Theseexperiments provide two relevant results on the effects of irradiation, fO2g, and

rep-fH2g on corrosion potential at the oxide-to-coolant interface:

(1) The potential differences between (a) the last time periods with H2added to the water prior to adding O2and (b) subsequent time periodswith O2addition and no added H2were 500–700 mV after allowing sev-eral time steps needed to reach equilibrium in the high-flux locations.This change in potential may include both changes in concentrations ofoxidizing species and controlling reactions

(2) In the moderate-flux region above the booster rods, the similar potentialdifference is 100–200 mV, demonstrating the significant effect of irradi-ation flux on electrochemical potential for oxygenated conditions, whilethere is relatively little effect for hydrogenated conditions Apparently,relatively small changes in irradiation flux in high-flux oxygenated-cool-ant conditions change significantly the local corrosion potential

These results must be considered with caution because the reported pretation was that the addition of O2resulted in a measurement of more reduc-ing conditions, which means that there may have been a problem in theinterpretation of the results or in the conduction of the tests Ruiz et al [64]measured the effect of hydrogen concentration in the plenum of a BWR andfound that the potential decreased by about 550 mV as hydrogen concentrationincreased to about 2.5 parts per million, confirming the effect of coolant hydro-gen on potential in BWR oxygenated coolant In general, the strong effect of

inter-fO2g and fH2g on potential and the correlations between potential and nodulesusceptibility support potential being a major factor in nodule susceptibility.This importance of potential, which is ascribed above to be highly influenced bythe concentration of radiolytic species, is consistent with aspects of Cox’s 2005review third conclusion [11] Furthermore, the above experiments and calcula-tions demonstrate that the potential difference between hydrogenated and oxy-genated conditions in the core coolant environment is at least 600 mV

These differences in potential need to be related to the effect of potential onnodule formation Hurst and Tyzack [65] suggested that oxidation potential ofthe environment played an important role Urquhart et al demonstrated that

600 mV more potential drop develops across the oxide of nodule-susceptiblealpha-annealed Zircaloy-4 than for nonsusceptible beta-heat-treated Zircaloy-4

in 500C 1500-psi steam In the absence of irradiation flux, lower temperatures,even 10C lower than the required 500C and certainly by 450C, eliminate

nodule susceptibility [66] This result suggests that the required potential fornodule formation is not far below 600 mV for the oxide thicknesses and condi-tions tested As a nodule-formation potential, 600 mV is not fixed However, for aset of conditions, there is a nodule-formation potential Also, Zircaloy potentio-dynamic polarization scans show a fairly sharp breakdown sensitivity withincreasing potential [67–69] at about 550 mV, depending on the conditions.These effects of potential changes on the protective nature of the oxide are con-sistent with the above observed changes in potential with changes in coolant

fO2g and fH2g concentrations, which themselves are associated with LWR ule susceptibility

nod-The gradient in potential at the transition at the bottom of the enriched tion of a BWR can be estimated to support the evaluation of the effect of poten-tial on the observed transition in nodule coverage at the enrichment transition.The Rishel et al [63] potential measurements provide a lower bound to thepotential gradient by comparing the data from the probes in the booster-rodregion to the reference probe above that region In general, there is a 400 mVdifference over 27 cm between the top of the high-flux region and probe #5 dur-ing the period of high oxygen in the coolant, resulting in a 15 mV/cm gradient.This estimated sensitivity is a lower bound because the upper probe still is in arelatively high-flux region compared to a blanket region in a BWR Consideringthat the flux gradient will be greater compared to those in Rishel’s tests, theeffect of a flux gradient can produce a significant change in potential in transi-tioning to the enriched region of a core This result is significant since the poten-tial can be close to the oxide-breakdown potential

por-Above, in estimating the effects of concentrations of oxidizing species, theimpact of cations at the oxide surface on the controlling reactions was not con-sidered The Zircaloy alloying elements Cr, Fe, and Ni have negative ECPs atstandard states, that is, Eo values, in the range of 740 mV to 250 mV SHE.These potentials are in the range of the 500 mV ECP that Andresen [70] esti-mates can occur for the maximum impact of H2 injection, which is known tosuppress nodule formation Therefore, depending on their concentrations at theoxide surface, their cations may decrease the ECP in a BWR that does not have

H2injection, and may have a suppressing effect on nodule formation The tribution of irradiation is to move these elements from the intermetallic SPPsinto the ZrO2lattice as cations This result is consistent with the suggested benefi-cial effects of solute cations in the model of Cheng and Adamson 20 In contrast,

con-Cu cations have positive Eovalues relative to H2oxidation, and Cuþ=Cu have itive Eovalues relative to O2reduction as well The presence of Cu cations raisesthe ECP and, in sufficient concentration, may counteract the coolant H2lowering

pos-of the ECP, as observed by Cowan [60] This effect pos-of Cu on ECP may be in tion to Marlowe’s [31] suggested effect of Cu on crud thermal conductivity

addi-Surface Potential Shifts and Polarization EffectsCheng and Adamson [20] associated nodule resistance with solute concentra-tion in the oxide matrix, and with the associated effect on oxide anion-vacancyconcentration This finding was adopted by others studying nodule and uniform

FRANKLIN, doi:10.1520/JAI103032 27