Comprehensive nuclear materials 2 07 zirconium alloys properties and characteristics

Comprehensive nuclear materials 2 07 zirconium alloys properties and characteristics Comprehensive nuclear materials 2 07 zirconium alloys properties and characteristics Comprehensive nuclear materials 2 07 zirconium alloys properties and characteristics Comprehensive nuclear materials 2 07 zirconium alloys properties and characteristics Comprehensive nuclear materials 2 07 zirconium alloys properties and characteristics

C Lemaignan

Commissariat a` l’E´nergie Atomique, Grenoble, France

ß 2012 Elsevier Ltd All rights reserved.

Abbreviations

ASTM American Society for Testing Materials

BWR Boiling water reactor

CANDU Canada Deuterium Uranium (heavy water

pressurized reactor)

DHC Delayed hydride cracking

DSA Dynamic Strain Aging

hcp Hexagonal closed packed

HPUF Hydrogen pick-up fraction

LOCA Loss of coolant accident

MIBK Methyl-isobutyl-ketone

PWR Pressurized water reactor

RBMK Reaktor Bolshoy Moshchnosti Kanalniy

(pressure tube power reactor)

RIA Reactivity induced accident

SPP Second phase particle

TEM Transmission Electron Microscope

TM Transition metal

VVER Voda-Voda Energy Reactor (PWR of

Russian design)

Zirconium (Zr) exhibits a physical property of uppermost importance with respect to the design

of in-core components of thermal neutron power reactors: it has a very low thermal neutron capture cross-section, and its alloys exhibit good engineering properties For an improvement in neutron efficiency

of the water-cooled reactors, the development of industrial-type Zr-based alloys started as early as the beginning of the nuclear reactor design, and is still continuing The engineering properties of Zr and Zr alloys are therefore widely studied Infor-mation exchanges and reviews are available in various sources; for example, the International Atomic Energy Agency issued reviews on Zr alloys for nuclear applications For more detailed, up-to-date information, the reader is referred to a recent one,1or

to the proceedings of the symposia on ‘Zr in the nuclear industry,’ organized at 2–3 year intervals

by ASTM.2

217

It was found early that Zr is naturally mixed in its

ore with its lower companion of the periodic table,

hafnium, the latter being a strong neutron absorber

Purification of Zr from Hf contamination is therefore

mandatory for nuclear applications The

develop-ment of the industrial alloys has been performed

following the classical route: searching for elements

of significant solubility that would improve the

engi-neering properties, without too much impact on

the nuclear ones Tin, niobium, and oxygen are the

main alloying elements, with minor additions of

transition metals (TMs) (Fe, Cr, and Ni) Heat

treat-ments aiming at homogeneous solid solutions, phase

transformations, and precipitation control allow

opti-mizing the structure of the alloys In addition, the

thermomechanical history of the components strongly

impacts their behavior, via the formation of a

crystallo-graphic texture, because of the anisotropy linked to the

hexagonal crystallography of Zr at low temperature

A few Zr alloys are commonly used for structural

components and fuel cladding in thermal neutron

reactors Zircaloy (Zry)-4 is used in pressurized

water reactors (PWRs) and Zircaloy-2 in boiling

water reactors (BWRs) The heavy water-moderated

CANDU reactors, as well as the Russian VVER or

RBMK reactors, use Zr–Nb alloys New alloys are

designed based on variants of the Zr–1% Nb, with

small additions of Fe and sharp control of minor

additions (M5®), or variants of the quaternary alloys,

such as Zirlo® and E635 More complex alloys with

other types of alloying elements are also being tested

in power plants, but the actual experience

accumu-lated on these alloys is too low to consider them as

commonly accepted, from an industrial point of view

Fuel claddings are made out of Zry-2 or Zry-4

Those tubes have different geometries, depending on

reactor design In PWR’s, the fuel cladding rods are

4–5 m long and have a diameter of 9–12 mm for a

thickness of 0.6–0.8 mm BWR fuel rods are usually

slightly larger The design is similar for the Russian

VVER, with Zr–1%Nb In CANDUs, the fuel bundles

are shorter (0.5 m) and the cladding is thinner (0.4 mm)

in order to collapse very fast on the UO2pellets

Structural components of zirconium alloys are the

guide tubes, the grids, and the end plates that

main-tain the components of the fuel assemblies They

have to maintain the structural integrity at the stress

levels corresponding to normal or accidental

opera-tions In addition, they should have very low

corro-sion rates in the hot, oxidizing coolant water In

BWRs, each assembly is surrounded by a Zircaloy-2

channel box that avoids cross-flow instabilities of

the two-phase coolant Their geometrical stability is

a mandatory requirement for the neutron physics design of the core

In the case of CANDUs and RBMKs, the moder-ator is separated from the coolant water The coolant water in contact with the fuel rods is contained in pressure tubes, usually made of Zr–Nb alloys They are large components (L
10 m, [
30 cm, and

e
5 mm), with a design life expected to match the reactor life, that is, tens of years, with only minor corrosion and creep deformation

Natural zirconium has an atomic mass of 91.22 amu, with five stable isotopes (90Zr : 51.46%,91Zr: 11.23%,

92

Zr: 17.11%, 94Zr: 17.4%, and 96Zr: 2.8%) The depletion of the most absorbing isotope (91Zr, with

sa
1.25  1028m2) would increase further the inter-est of using Zr alloys in reactors, but would clearly be economically inefficient The cross-section for elastic interaction with neutrons is normal, with respect to its atomic number (sdiff
6.5 barn) Despite its high atomic mass, the large interatomic distance in the hcp crystals lead to a limited specific mass of 6.5 kg dm3 The thermophysical properties correspond to stan-dard metals: thermal conductivity
22 W m1K1 and heat capacity
280 J kg1K1, that is, close to 3R per mole

Below 865C, pure Zr has an hcp structure, with

a c/a ratio of 1.593 (slightly lower than the ideal 1.633) The lattice parameters are a ¼ 0.323 nm and

c ¼ 0.515 nm.3

The thermal expansion coefficients show a strong anisotropy, with almost a twofold differ-ence between the aaandaccoefficients (respectively 5.2 and 10.4 106K1).4This anisotropic behavior of the thermal expansion induces internal stresses due to strain incompatibilities: After a standard heat treat-ment of 500C, where the residual stresses will relax, cooling down to room temperature will result in inter-nal stresses in the range of 100 MPa, depending on grain-to-grain orientations The modulus of elasticity

is also anisotropic, but with lower differences than for thermal expansion (Ea¼ 99 GPa, and Ec¼ 125 GPa).5

For industrial parts, the values recommended are close toa
6.5  106K1andE
96 GPa The tem-perature evolution of the elasticity constants is unusual: the elasticity is strongly reduced as the tem-perature increases (
5% per 100 K).6,7

This abnormal behavior is specific to the hcp metals of the IV-B row

of the periodic table.8

At 865C, Zr undergoes an allotropic

transforma-tion from the low temperature hcp a-phase to the bcc

b-phase On cooling, the transformation is usually

bai-nitic, but martensitic transformation is obtained for

very high cooling rates (above 500 K s1) The bainitic

transformation occurs according to the epitaxy of

the a-platelets on the old b-grains, as proposed by

Burgers9,10: (0001)a // {110}b andh11
20ia // h111ib

Among the 12 different possible variant orientations

of the new a-grains, only a few are nucleated out of a

given former b-grain during this transformation to

minimize the internal elastic strain energy This

pro-cess leads to a typical Ổbasket-weaveỖ microstructure

(Figure 1) As a result, a b-quenching does not

completely clear out the initial crystallographic texture

that had been induced by the former

thermomechani-cal processing.11,12 Although the alloying elements

present in the Zr alloys change the transformation

temperatures, with a 150C temperature domain in

which the a- and b-phases coexist, the crystallographic

nature of the a-b transformation is equivalent to that of

pure Zr Specific chemical considerations (segregations

and precipitations) will be described later

The melting of pure Zr occurs at 1860C,

sig-nificantly above the melting temperature of other

structural alloys, such as the structural or stainless

steels At high pressures, (P > 2.2 GPa) a low-density

hexagonal structure is observed, known as the o-phase

2.07.3.1 Nuclear Grade Zr Base Metal

The most frequently used ore is zircon (ZrSiO4), with

a worldwide production of about 1 million metric

tons per year, out of which only 5% is processed into zirconium metal and alloys

The processing of Zr alloy industrial components

is rather difficult because of the high reactivity of the Zr metal with oxygen It consists of several steps

to obtain the Hf-free Zr base metal for alloy prepa-ration: decomposition of the ore to separate Zr and Si, Hf purification, and Zr chloride or fluoride reduction

2.07.3.1.1 Ore decomposition Three different processes are currently used for the ZrỜSi separation:

 In alkali fusion, where the zircon is molten in a NaOH bath at 600C, the following reaction takes place:

ZrSiO4ợ4NaOH ! Na2ZrO3ợNa2SiO3ợ2H2O Water or acid leaching allows the precipitation of ZrO2

 The fluo-silicate fusion:

ZrSiO4ợ K2SiF6! K2ZrF6ợ 2SiO2

It produces a potassium hexafluorozirconate which, reacting with ammonia, leads to Zr hydroxide

 The carbo-chlorination process is performed in a fluidized bed furnace at 1200C The reaction scheme is the following:

ZrO2đợSiO2ợ HfO2ỡ ợ 2C ợ 2Cl2

! ZrCl4đợSiCl4ợ HfCl4ỡ ợ 2C The controlled condensation of the gaseous tetra-chloride allows the separation of Zr and Si, but not

of Hf from Zr

2.07.3.1.2 Hf purification and removal The processes described above separate Si from

Zr, but the Zr compounds remain contaminated with the initial Hf concentration The high neutron capture cross-section of Hf (sa
105 barn, compared

to 0.185 barn for Zr) requires its suppression in

Zr alloys for nuclear application Two major pro-cesses are used for this step: the MIBK-thiocyanate solvent extraction and the extractive distillation of tetrachlorides

 In the first case, after reaction of zirconyl chloride (ZrOCl2), obtained by hydrolysis of ZrCl4, with ammonium thiocyanate (SCN-NH4), a solution

of hafnyl-zirconyl-thiocyanate (Zr/Hf )O(SCN)2

is obtained A liquidỜliquid extraction is per-formed with methyl-isobutyl-ketone (MIBK,

Figure 1 Microstructure of a b-quenched Zr alloys, with

a-platelets of four different crystallographic orientations

issued from the same former b-grain.

name of the process) Hf is extracted into the

organic phase, while Zr remains in the aqueous

one Hf-free ZrO2is obtained after several other

chemical steps: hydrochlorination, sulphation,

neu-tralization with NH3, and calcination

 In the dry route, after the transformation of

zircon into its chloride ZrCl4, through the

carbo-chlorination process, Zr and Hf are separated using

a vapor phase distillation, at 350C, within a

mixture of KCl-AlCl3, where the liquid phase is

enriched in Zr, and the vapor in Hf

2.07.3.1.3 Reduction to the metal

The final step to obtain metallic Zr of nuclear grade

is to reduce the Hf-free Zr compounds that have been

obtained by the previous steps Two processes are to

be considered at an industrial scale: the Kroll process

and the electrolysis

 In the Kroll process, the Zr metal is obtained by

the reduction of ZrCl4in gaseous form by liquid

magnesium, at about 850C in an oxygen-free

environment The following reaction occurs:

ZrCl4ðgÞ þ 2MgðlÞ ! MgCl2ðlÞ þ ZrðsÞ

After distillation of the remaining Mg and MgCl2,

under vacuum at 950C, sintering of the Zr

agglom-erate at 1150C gives the metallic sponge cake

 After wet chemical chemistry, the reduction of the

ZrO2 obtained by the MIBK process is often

per-formed by electrolysis It is realized with the mixed

salt K2ZrF6dissolved in NaCl or KCl at 850C under

inert gas, with stainless steel cathode on which Zr is

deposited, and chlorine evolution at the graphite

anode This route is mainly used in the Russian

Federation, the names of the Russian alloys starting

with an ‘E,’ referring to electrolytic processing

High purity Zr can be obtained by the Van Arkel

process It consists of reaction of Zr with iodine at

moderate temperature, gaseous phase transport as

ZrI4, and decomposition of the iodide at high

temper-ature on an electrically heated filament The iodine

released at the high temperature side is used for the

low temperature reaction in a closed loop transport

process, according to the following scheme:

Zr + 2 I2=> ZrI4 => Zr + 2 I2

250–300 ⬚C 1300–1400 ⬚C

This source of metallic Zr (called ‘iodide Zr’) is used

in Russia in addition to Zr obtained by the electrolytic

process for the melting of the alloys (typically 30%

‘iodide Zr’ in the first electrode to be melted) 2.07.3.2 Alloy Melting

Whatever the processing route followed for the pro-duction of Zr metal, the sponge or the chips obtained

by scrapping out the electrodes are the base products for alloy ingot preparation The melting of the alloys

is performed using the vacuum arc remelting (VAR) process This process is specific to highly reactive metals such as Zr, Ti, or advanced superalloys For industrial alloy preparation, an electrode is prepared by compaction of pieces of base metal frag-ments (sponge or scraps) with inclusion of the alloy-ing elements Typically, the elements to be added are the following: O (in the form of ZrO2 powder), Sn,

Nb, Fe, Cr, and Ni to the desired composition In addition, a strict control of minor elements, such as C,

N, S, and Si, is ensured by the producers, at concen-trations in the range of 30–300 ppm, according to their requirements to fulfill the engineering properties

A few specific impurities are strictly controlled for neutron physics reasons: Cd and Hf due to their impact on neutron capture cross-section, U for the contamination of the coolant by recoil fission fragments escaping from the free surface of the clad-ding, and Co for in-core activation, dissolution trans-port, contamination, and g-irradiation

The compact stack is melted in a consumable electrode electric vacuum furnace with water chilled

Cu crucible Electromagnetic fields are often used for efficient stirring of the liquid pool and reduced seg-regations After three to four melts, the typical dimensions of the final ingots are 0.6–0.8 m diameter and 2–3 m length, that is, a mass of 4–8 tons

2.07.3.3 Forging Industrial use of Zr alloys requires either tube- or plate-shaped material The first step in mechanical processing is forging or hot rolling in the b-phase, at a temperature near 1050C, or at lower temperatures

in the aþ b range or even in the upper a range The high oxidation kinetics of Zr alloys in air at high temperatures restricts the high temperature forging process to thick components, that is, with minimum dimensions larger than 10 cm, at least Final dimen-sions after forging correspond to 10–25 cm diameter for billets and 10 cm for slabs

A b-quenching is usually performed at the end of the forging step This heat treatment allows complete

dissolution of the alloying elements in the b-phase

and their homogenization above 1000C, followed by

a water quench During the corresponding bainitic

b to a transformation, the alloying elements are

redistributed, leading to local segregations: O and

Sn preferring the middle of the a-platelets, while

the TMs (Fe, Cr, and Ni) and Nb are being rejected

to the interface between the platelets.13These

segre-gations lead to plastic deformation strains highly

localized at the interplatelet zones for materials having

a b-quenched structure (heat-affected zones, welds, or

b-quenched without further thermomechanical

pro-cessing) As described later, this b-quench controls the

initial size distribution of the precipitates in Zircaloy,

and further recovery heat treatments should be

per-formed below the b–a transus only

2.07.3.4 Tube Processing

For seamless tube production, first a hot extrusion is

performed in the temperature range of 600–700C

For pressure tube fabrication, this step is followed by

a single cold drawing step and a final stress relieving

heat treatment For cladding tubes, the extrusion

produces a large extruded tube (‘Trex’ or ‘shell’), of

50–80 mm in diameter and 15–20 mm in thickness,

which is further reduced in size by cold rolling on

pilger-rolling mills

After each cold working step of plate or tube

material, an annealing treatment is mandatory to

restore ductility It is usually performed in the range

of 530–600C to obtain the fully recrystallized

material (RX) The resultant microstructure is an

equiaxed geometry of the Zr grains with the

precipi-tates located at the a-grain boundaries or within the

grains The location of the precipitates at the grain

boundaries is not due to intergranular precipitation

but because they pin the grain boundaries during

grain growth (Figure 9) These different heat

treat-ments contribute to the control of the cumulative

annealing parameter to be described below For better

mechanical properties of the final product, the

tem-perature of the last annealing treatment can be

reduced to avoid complete recrystallization This is

the stress-relieved (SR) state, obtained with final heat

treatment temperature of 475C, which is

character-ized by elongated grains and a high density of

dis-locations, and by relief of the internal stresses,

leading to a greater ductility than cold-worked

mate-rials It is mostly applied to the PWR claddings, while

for BWRs, a complete recrystallization is performed

at 550–570C

2.07.3.4.1 Crystallographic texture development

Two plastic deformation mechanisms are operating during low temperature deformation of the Zr alloys: dislocation slip and twinning As reviewed by Tenckhoff,14the most active deformation mechanism depends on the relative orientation of the grain in the stress field

Dislocation slip occurs mostly on prism plane with

an a Burgers vector It is referred to as the {10
10} h1
210i, or prismatic, system The total strain imposed during mechanical processing of the Zr alloys cannot, however, be accounted for only with this single type

of slip, as the different orientations of the crystal would only give two independent shear systems

At high deformations, and as the temperature is increased, (c þ a) type slip is activated on {11
21} or {10
11} planes These are the pyramidal slip systems, having higher resolved shear stresses (Figure 2) Different twinning systems may be activated depending on the stress state: for tensile stress in the c-direction, {10
12} h
1011i twins are the most frequent, while the {11
22}h
1
123i system is observed when compression is applied in the c-direction The resolved shear stresses of the twin systems have been shown to be higher than the one necessary for slip, but due to the dependence of the Schmid factor

on orientation, twinning is activated before slip, for some well-oriented grains Therefore, there are five independent deformation mechanisms operating

in each grain, and thus the von Mises criterion for grain-to-grain strain compatibility is fulfilled

At the large strains obtained during mechanical processing, steady-state interactions occur between the twin and slip systems that tend to align the basal planes parallel to the direction of the main deformation.15,16For cold-rolled materials (sheets or tubes), the textures are such that the majority of the

Figure 2 The two Burgers vectors (a and c þ a) for strain dislocations in Zr alloys, and the two slip planes (prismatic and pyramidal) in hcp a-Zr.

grains have theirc-axis tilted 30–40 away from the

normal of the foil or of the tube surface toward the

tangential direction, as can be seen in theh0001i pole

figure of a cladding tube (Figure 3)

During tube rolling, the spread of the texture can

be reduced by action on the ratio of the thickness to

diameter reductions (Q factor): a reduction in

thick-ness higher than the reduction in diameter gives a

more radial texture, that is, a texture with thec poles

closer to the radial direction.16

After cold processing, theh10
10i direction is

paral-lel to the rolling direction, and during a

recrystalliza-tion heat treatment a 30 rotation occurs around the

c-direction and the rolling direction is then aligned

with theh11
20i direction for most of the grains

2.07.4.1 Alloying Elements and

Phase Diagrams

Like any metal, pure Zr exhibits rather poor

engi-neering properties To improve the properties of a

given metal, the metallurgical engineering

proce-dures are always the same: It consists in finding

addi-tions, any species of the periodic table could be

considered, with significant solubility, or heat

treat-ments producing new phases that could improve the

properties The relative solubility of the various

alloy-ing elements in the a- and b-phases is therefore one

basis for the choice of additions, as well as for

devel-oping the heat treatments, for microstructure control

For the nuclear applications, neutron physics

requirements restrict the possibilities, by rejection

of the isotopes having high interaction cross-sections,

or isotopes that would transmute to isotopes of high capture cross-section or having high irradiation impact (Co) Elements such as Hf, Cd, W, and Co have therefore not been considered for alloy devel-opments With low nuclear impact, O, Sn, and Nb have been selected (Al and Si having also low nuclear impacts were not retained because of degradation in corrosion resistance), while other TMs (Fe, Cr, Ni, etc.) can be accepted up to limited concentrations (below 0.5% total)

The additions have to improve the engineering properties The main properties to be improved are the corrosion behavior in hot water and the mechan-ical strength (yield stress, ductility, and creep) As described below, Sn and Nb are added for corrosion resistance, and elements forming secondary phases (Nb and Fe, Cr, and Ni) or solid solutions are also used for increasing the mechanical properties Last, the microstructure obtained after the ther-momechanical processing should not change without control under irradiation Therefore, hardening obtained by precipitation or strain hardening can be considered only if the irradiation-induced evolution

of the initial microstructure will be compensated by the development of irradiation-induced microstruc-tural defects In this respect, the evolution of preci-pitates in Zircaloys is of high importance for corrosion behavior and geometrical integrity These points are discussed inChapter5.03, Corrosion of Zirconium Alloys and Chapter 4.01, Radiation Effects in Zirconium Alloys

Most of the binary phase diagrams with Zr are already known and many ternary or higher-level

fR= 0.64

fR= 0.55

AD

80⬚

60⬚

30⬚

4⬚

0⬚

60⬚

1 2

35 3 AD

4

TD

2 TD 3

5 3

Figure 3 h0001i Pole figure of two cladding tubes with slightly different mechanical processing routes.

diagrams of industrial interest are now known.17The

need for a better control of the processing of the

current alloys and the aim of finding new alloys and

structures without too much experimental work have

been a driving force for the modern trend in

numeri-cal simulation for material science It is now also

possible to extrapolate the binary data to

multicom-ponent systems In that respect, a thermodynamic

database for Zr alloys, called ZIRCOBASE, has

been developed under the Calphad methodology.18

This database contains 15 elements and is frequently

updated The most complex ternary or quaternary

phase diagrams available are optimized or computed

using this database, and, in the case of missing basic

thermodynamic data, with the contribution of

‘ab-initio’ computations.19

The phase diagrams presented

in this review were obtained according to this

procedure

Oxygen is highly soluble in the a-phase, and

stabi-lizes at high temperature (Figure 4) Oxygen has

to be considered as an alloying element This use of

oxygen for strengthening is rare in metallurgy,

com-pared to the use of nitrogen However, the use of

nitrogen for strengthening would severely deteriorate

the corrosion resistance, and nitrogen is removed as

much as possible The purpose of oxygen additions

is to increase the yield strength by solution

strength-ening, without degradation of the corrosion

resis-tance The O content is not specified in the ASTM

standards, but usually it is added to concentrations in

the range of 600–1200 ppm, and this has to be agreed

between producer and consumers High O

concen-trations (O> 2000 ppm) reduce the ductility of the

alloys; therefore, O additions above 1500 ppm are

not recommended In addition, O atoms interact

with the dislocations at moderate temperatures,

leading to age-strengthening phenomena in temper-ature ranges depending on strain rate.20The oxygen

in solid solution in a-zirconium is an interstitial in the octahedral sites In the Zr–O system, the only available stable oxide is ZrO2 A monoclinic phase

is stable at temperatures up to about 1200C, above which it transforms to a tetragonal structure The impact on corrosion of the different phases of ZrO2, according to temperature and pressure is dis-cussed in Chapter 5.03, Corrosion of Zirconium Alloys

Tin tends to extend the a-domain, and has a maximal solubility in the hcp Zr of 9 wt% at 940C (Figure 5) It was originally added at concentrations

of 1.2–1.7% to increase the corrosion resistance, especially by mitigating the deleterious effects of nitrogen The amount of Sn needed to compensate the effect of 300 ppm of N is about 1% of Sn How-ever, in N-free Zr, Sn has been observed to deterio-rate the corrosion resistance Therefore, the modern trend is to reduce it, but only slightly, in order to maintain good creep properties.21

Iron, chromium, and nickel, at their usual concentra-tions, are fully soluble in the b-phase (Figure 6) However, in the a-phase, their solubilities are very low: in the region of 120 ppm for Fe and 200 ppm for

Cr at the maximum solubility temperature.22 In the pure binary systems, various phases are obtained: ZrFe2 and ZrCr2 are Laves phases with cubic or hexagonal structure, while Zr2Ni is a Zintl phase with a body-centered tetragonal C16 structure These precipitates are called the Second Phase Par-ticles (SPPs)

In the Zircaloys, the Fe substitutes for the corresponding TM and the intermetallic compounds found in Zircaloy are Zr2(Ni,Fe) and Zr(Cr,Fe)2

0

500

1000

1500

2000

2500

10 20

Atomic percent oxygen

L

a-Zr

b-Zr

30 40 50 60 70

a2⬘⬘

a⬘

Figure 4 Zr–O binary phase diagram.

2000 1800 1600 1400 1200 1000 800 600 400 200

0 5 10 15 20 25 30

Atomic percent tin

a-Zr

b-Zr b-Zr + Zr5 Sn3

L

Figure 5 Zr–Sn binary phase diagram.

The formation of these precipitates, and more

complex ones in industrial alloys, is analyzed in

detail for the control of the corrosion behavior of

the Zircaloys Indeed, a strong correlation has been

observed between precipitate size distributions and

corrosion kinetics, the behavior being opposite for

BWRs and PWRs A better uniform corrosion

resis-tance is obtained for Zircaloys used in PWRs if they

contain large precipitates, while better resistance to

the localized forms of corrosion is seen in BWRs in

materials that have finely distributed small

precipi-tates.23,24With an increase in the particle diameters

from 0.05 to 0.1 mm or higher, the in-pile corrosion of

Zircaloy cladding diminishes appreciably However,

nodular corrosion may occur in BWR cladding with a

further increase in the particle diameters above about

0.15 mm25(Figure 7)

Due to the low solubility of the transition metals

(Fe, Cr and Ni) in the Zr matrix, coarsening of the

precipitates, after the last b-quench, occurs at very

low rates, during the intermediate annealing heat treatments, following each step of the rolling process Therefore, the precipitate growth integrates the ther-mal activation times of each recovery, and their tem-peratures and durations can be used to control the size of SPPs This integrated coarsening activation time is referred as the ‘A’ or ‘SA’ parameter

The A-parameter calculates the integral of the activation processes for the different anneal durations and temperatures The annealing parameter is defined as A ¼ Si (tiexp (Q/RTi), where ti is the time (in hours) of theith annealing step, at tempera-ture Ti(in K);Q/T is the activation temperature of the process involved The activation energy for the process should have been taken as the one controlling the coarsening, that is, the diffusion However, as the early studies were undertaken with the aim of improving the corrosion resistance, an unfortunate practice has been induced to take 40 000 K as the value of Q/T A more correct value would be

2000 1800 1600 1400 1200 1000 800 600 400 200

Atomic percent iron

b-Zr

a-Zr

¬

0

L

2000

1800

1600

1400

1200

1000

800

600

400

200

Atomic percent chromium

L

b-Zr

a-Zr

¬

¬

¬

¬

(b)

g-Cr2Zr a-Cr2Zr

2000 1800 1600 1400 1200 1000 800 600 400 200 0

(c)

a-Zr

¬

b-Zr

¬

L

Atomic percent nickel

Figure 6 Zr-rich site of the Zr-transition metal binary phase diagram: (a) Zr–Fe, (b) Zr–Cr, (c) Zr–Ni.

32 000 K, which fits very well with the

recrystalliza-tion kinetics The influence of the A-parameter

on the corrosion of Zircaloy is discussed in more

detail in Chapter 5.03, Corrosion of Zirconium

Alloys High resistance to uniform corrosion in

PWR is obtained for the A-parameter close to

(1.5–6.0) 1019h In BWR, theA-parameter value

for the Zircaloy-2 cladding in BWR has to be in

the range (0.5–1.5) 1018h (Figure 7).25 This

corresponds to precipitates larger than 0.18 mm

The SA approach has been developed for the

Zircaloys and is clearly not applicable for other

alloys, such as the Zr–Nb alloys

Niobium (columbium) is a b-stabilizer that can

extend the bcc domain to a complete solid solution

between pure Zr and pure Nb at high temperatures

(Figure 8) A monotectoid transformation occurs at

about 620C and around 18.5 at.% Nb The

solubil-ity of Nb in the a-phase is maximal at the monotectic

temperature, and reaches 0.65%

Water b-quenching of small pieces leads to the precipitation of a0 martensite supersaturated in Nb Tempering at intermediate temperature results

in b-Nb precipitation within the a0 needles and subsequent transformation of a0 into a When quenching is performed from an aþ b region,

a uniform distribution of a- and b-grains is obtained, and the Nb-rich b-phase does not trans-form By aging at temperatures in the range of 500C, the metastable Nb-rich b-phase can be decomposed into an hcp o-phase This gives a sharp increase

in mechanical strength because of the fine micro-structure obtained by the b-o transformation.26

In the usual form of the Zr–2.5% Nb, the cold work condition after aþ b extrusion and air-cooling, the microstructure consists of Zr grains with layers of b-Nb rich phase (close to eutectoid composition) Owing to the affinity of Fe for the b-phase, most of this element is found in the minor b-grains These b-grains are metastable and decompose, upon aging,

to a mixture of a-Zr and pure b-Nb The Nb dis-solved in the a-hcp Zr phase is itself metastable and the irradiation-induced precipitation of the supersat-urated Nb solid solution is believed to be the origin

of the improvement in corrosion resistance under irradiation of these alloys.27

In the case of Zr–1% Nb used for VVER and RBMK, or M5® in PWRs, the concentration of Nb

in the Zr matrix after processing corresponds to the maximum solubility near the monotectoid tempera-ture, which is higher than the solubility at the service temperature Owing to the slow diffusion of Nb, the equilibrium microstructure cannot be obtained ther-mally However, the irradiation-enhanced diffusion allows precipitation of fine b-Nb needles in the grains after a few years in reactors.28

In-reactor

10

5

3

2

1

30

0.02

0.02

Average diameter of precipitates (mm)

0.8

10

3

1

PWR

Out-of-pile

500 ⬚C/16 h

350 ⬚C/ 1a BWR

Figure 7 Effect of precipitate size on the corrosion

kinetics of Zircaloys Reproduced from Garzarolli, F.;

Stehle, H Behavior of structural materials for fuel and

control elements in light water cooled power reactors, IAEA

STI/PUB/721; International Atomic Energy Agency: Vienna,

1987; p 387.

2000

1600 1400 1200 1000 800 600 400 200

a-Zr

(b-Zr, b-Nb)

¬

Atomic percent niobium

Figure 8 Zr–Nb binary phase diagram.

Sulfur has recently been observed to be extremely

efficient in improving the creep resistance, even at

concentration as low as 30–50 ppm This chemical

species, formerly not considered as important, is now

deliberately added during processing to reduce the

scatter in behavior and to improve the high

tempera-ture mechanical properties.29 The efficiency of such

low concentrations on the creep properties has been

explained by the segregation of the S atoms in the core

of the dislocations, changing their core configurations

It does not affect the corrosion properties.30

In the case of complex alloys, other

thermody-namical interactions are expected and intermetallic

compounds including three or four chemical

ele-ments are observed The chemistry and the

crystal-lography of these phases may be rather complex

Two examples will be given of the complex structure and behavior of these intermetallics

 For the Zr–Cr and Zr–Ni binary alloys, the stable forms of the second phase are Zr2Ni or ZrCr2 These phases are effectively the ones observed in the Zircaloys, with Fe substituting for the corresponding TM Therefore, the general formu-lae of the intermetallic compounds in Zircaloys are

Zr2(Ni,Fe) and Zr(Cr,Fe)2 The crystal structure of the Zr(Cr,Fe)2 precipitates is either fcc (C15) or hcp (C14), depending on composition and heat treatment Both structures are Laves phases, with characteristic stacking faults as seen in Figure 9 The equilibrium crystallographic structure is dependent upon the Fe/Cr ratio, cubic below 0.1 and above 0.9, and hexagonal in the middle Under irradiation, these precipitates transform to amor-phous state and release their Fe in the matrix, with strong impact on corrosion behavior under irradiation.31

 In the Zr–Nb–Fe ternary, other intermetallic com-pounds can be observed (Figure 10): the hexago-nal Zr(Nb,Fe)2phase and the cubic (Zr,Nb)4Fe2.32 Although of apparent similar composition, the two phases are indeed different: Nb can substitute Fe in the hexagonal phase, while it will substitute Zr in the cubic phase In these alloys, due to the slow diffusion of Nb, metastable phases are often present and the equilibrium microstructure after industrial heat treatments may be far from the stable one Therefore, the final microstructure is strongly dependent on the exact thermomechanical history

200 nm 3mm

Figure 9 Microstructure of recrystallized Zry-4: Zr(Fe,Cr) 2

precipitates in the Zr(Sn–O) matrix (TEM at two different

scales).

0

0.02 0.04 0.06 0.08 0.1

Nb (wt%)

β-Zr + hex β-Nb + hex + cub Hex + cub β-Zr phase boundary Domain limit Domain limit

α-Zr + cub

α-Zr + hex

α-Zr + cub + hex

α-Zr + β-Zr (metastable) + β-Nb + hex

α-Zr + β-Nb + hex

α-Zr + Zr3Fe + cub

α-Zr + β-Nb α-Zr

Figure 10 Zr-rich corner of the Zr–Nb–Fe ternary phase diagram at 580C.