Comprehensive nuclear materials 5 09 material performance in lead and lead bismuth alloy Comprehensive nuclear materials 5 09 material performance in lead and lead bismuth alloy Comprehensive nuclear materials 5 09 material performance in lead and lead bismuth alloy Comprehensive nuclear materials 5 09 material performance in lead and lead bismuth alloy Comprehensive nuclear materials 5 09 material performance in lead and lead bismuth alloy Comprehensive nuclear materials 5 09 material performance in lead and lead bismuth alloy
Trang 1K Kikuchi
Ibaraki University, Ibaraki, Japan
ß 2012 Elsevier Ltd All rights reserved.
Abbreviations
ADS Accelerator-driven nuclear
transmutation system
AFM Atomic force microscopy
BEM Backscattered electron microscope
DBTT Ductile-to-brittle transition
temperature
EB Electron beam
EDX Energy-dispersed X-ray analyzer
F/M steel Ferritic–martensitic steel
GESA Gepulste Elektronenstrahlanlage
GIF Generation IV International Forum
ICP Inductive-coupled plasma atomic
emission spectrometer
LA Lead alloy
LBE Lead–bismuth eutectics
LFR Liquid-metal-cooled fast reactor
LINAC Linear accelerator
MA Minor actinides
MEGAPIE MEGA-watt Pilot Experiment
MFM Magnetic force microscopy
MYRRHA Multipurpose hybrid research
reactor for high-tech applications
ODS Oxide dispersion-strengthened steel
OECD/NEA The Organisation for Economic
Co-operation and Development/
The Nuclear Energy Agency
SEM Scanning electron microscopy
WDX Wave-dispersed X-ray analyzer
5.09.1 Recent Lead-Alloy Activity
A brief justification for the utilization of lead or lead bismuth for use as a coolant in nuclear energy sys-tems was given in 2001 by Sekimoto.1 When the possibility of the utilization of nuclear energy was discovered, it was expected to be a primary energy source in the future Fast reactors can utilize the entire energy content of natural uranium The selec-tion of a coolant was an important item for designing fast reactors The neutron slowing-down caused by the coolant should be minimized This is first made possible by decreasing the average atomic density of the coolant in the reactor core, and second by employing a nuclide with a large mass number as the coolant, whose neutron moderating power is low A liquid metal is considered the best coolant for using the second method Initially, liquid mercury was employed but it was not successful in either the United States or Russia Since then, several liquid metals were considered, including lead alloys (LA), and finally, sodium was selected However, public concern about the safety of sodium has increased following sodium leakage incidents, so the develop-ment and deploydevelop-ment of fast reactors on more than a prototype scale has not occurred
In the last 10 years, the study of the utilization of
LA including lead–bismuth eutectics (LBE) has been ongoing for application to nuclear waste transmutation systems and lead–bismuth cooled nuclear reactors
207
Trang 2LBE is a candidate material for a spallation target and a
reactor coolant In the accelerator-driven nuclear
transmutation system (ADS), LBE is a candidate for
both the subcritical-reactor coolant and the spallation
neutron source target In addition, the lead or lead–
bismuth-cooled fast reactor (LFR) is one of the four
reactor types investigated in Generation IV systems
proposed by the Generation IV International Forum
(GIF) A LBE-cooled Long-Life Safety Simple Small
Portable Proliferation-Resistant Reactor has also been
proposed.2
As a result of the investigations on LA,
compre-hensive literature has been published The Working
Group on LBE of the OECD/NEA Nuclear Science
Committee3 published a handbook and review
reports on LA technology The material properties
of lead and lead–bismuth are discussed in detail in
Chapter2.14, Properties of Liquid Metal Coolants
As part of the development of advanced nuclear
systems, including ADS proposed for high-level
radioactive waste transmutation and Generation IV
reactors, heavy liquid metals such as lead or LBE
were investigated as reactor core coolant and
spall-ation targets Heavy liquid metals were also being
envisaged as target materials for high-power neutron
spallation sources The objective of the handbook is
to collate and publish properties and experimental
results on lead and LBE in a consistent format in
order to provide designers with a single source of
qualified properties and data and to guide subsequent
development efforts The handbook covers liquid
lead and LBE properties, material compatibility and
testing issues, key aspects of the thermal-hydraulic
and system technologies, existing test facilities, and
open issues and perspectives
Zhang and Li4reviewed the studies on
fundamen-tal issues in LBE corrosion They included phase
diagrams, thermodynamics, physical properties,
cor-rosion mechanisms, oxygen control, experimental
results, and corrosion results Some
recommenda-tions were proposed for future studies: precipitation
and deposition of corrosion products; oxygen
trans-port; oxide formation and kinetics in LA; coolant
hydrodynamic effects; steel composition,
microstruc-ture, and surface effects; and corrosion models These
are key areas for future research
Fazio et al.5 characterized corrosion property for
ferritic–martensitic (F/M) steels and austenitic steels
in stagnant LA on the basis of the results of corrosion
tests This report briefly summarized the current
status on LA activities At a temperature below
450C, adequate oxygen activities in the liquid
metal steels form an oxide layer that behaves as a corrosion barrier In the temperature range above
500C, corrosion protection because of the oxide scales seems to fail A mixed corrosion mechanism has been observed, where both oxide scale formation and dissolution of the steel elements occurred How-ever, in this high-temperature range, it has been demonstrated that the corrosion resistance of struc-tural materials can be enhanced by coating the steel with FeAl alloys Experiments performed in flowing
LA (mostly LBE) confirm that the corrosion mecha-nism of the steels depends on the oxygen content in
LA At relatively low oxygen concentration, the cor-rosion mechanism changes from oxidation to dissolu-tion of the steel elements The experimental activity also extends up to temperatures of 750C for oxide dispersion-strengthened (ODS) alloys and their welded variants in Pb The use of materials at higher temperatures will also require investigation of creep rupture
MEGAPIE was the MEGA-watt Pilot Experiment done at Paul Scherrer Institut (PSI) in 2006 for developing a LBE spallation target The MEGAPIE project was started as an essential step toward demon-strating the feasibility of coupling a high power accel-erator, a spallation target, and a subcritical core assembly The project was expected to furnish impor-tant results regarding safe treatment of components that had come into contact with lead–bismuth.6The design data was obtained and the operational mode was confirmed.7 Corrosion rates were estimated experimentally at 400C for a LBE flow rate of
1 m s1 and 2.2 m s1 where the oxygen content in the LBE was <107wt% No protective oxide layer was produced on the steel surface This oxygen con-tent has been considered representative of the MEGAPIE conditions, as no oxygen control and monitoring system is anticipated to be used in the target The estimated corrosion rates, 40–86mm year1, indicate that in the given testing conditions, the corrosion resistance of the steel does not repre-sent a critical issue, especially since LBE temperature
is expected to be lower (320C) The goals of the experiment were fully accomplished8: 4 months of reliable and essentially uninterrupted operation (beam trips and short beam interruptions permitted)
at a power level as high as the accelerator was able to deliver (about 0.75 MW) excellent performance of the target and the dedicated ancillary systems, the proof of functionality of advanced proton beam safety devices, and, last but not least, a superb neutronic efficiency delivering about 80% more neutrons for
Trang 3the users compared to the previously operated
lead-cannelloni target Verification of performance will be
scheduled in the postirradiation experiment
5.09.2 Utilization of LA
5.09.2.1 The Conceptual Models of ADS
and MYRRHA
Recent activity on materials research and
develop-ment in LA, especially LBE, aims at realizing ADS,
MEGAPIE, LFR, and MYRRHA (multipurpose
hybrid research reactor for high-tech applications).9,10
It is valuable to know each specific environment for
material usage in design studies The material temper-ature at contact with LBE is slightly<500C in the spallation reaction area and<550C in the fuel core area under normal conditions
Figure 1 shows the ADS concept A supercon-ducting linear accelerator (LINAC) is connected with a subcritical fast reactor A high-energy proton beam is injected into the core of the reactor Spall-ation reactions produce a number of neutrons from the lead–bismuth nuclei, which are then used to transmute minor actinides (MA) The interface between the beam duct and lead bismuth is called the beam window For example, a tank type reactor with 800 MW thermal power and LBE-coolant and spallation target was proposed.11–13The proton beam energy was set at 1.5 GeV The beam current varied between 10 and 20 mA according to criticality swings
In the steady-state condition, as the beam window material generates heat by spallation reactions and is cooled by flowing LBE A temperature difference is established between the LBE, the material in contact with the LBE, and the material on the other side of the window, with the temperatures being 400, 450, and 500C, respectively As the MA core cladding material is gamma heated and the fuel adds to the radiation heat, temperatures reach, for example, a maximum of 500, 550, and 600C The maximum average velocity in the particular flow channel of LBE is 1.8 and 2.0 m s1, at the window and in the
MA core region, respectively
Figure 2 shows the conceptual model of MYR-RHA consisting of an inner vessel, guard vessel,
Beam duct
MA
n
P
Superconducting LINAC
ADS DTL
RFQ
Injector
RF
Liq.He
PbBi
Spallation reaction
Beam window
Beam
window
MA (Am,Cm)
Figure 1 The conceptual model of accelerator-driven
nuclear transmutation system with beam window.
1 Inner vessel
11
10 4
6 12
11 9 8 10 7
11 12
5 8 9 7
2 13
3
2 Guard vessel
3 Cooling tubes
4 Cover
5 Diaphragm
6 Spallation loop
7 Subcritical core
8 Primary pumps
9 Primary heat exchangers
10 Emergency heat exchangers
11 In-vessel fuel transfer machine
12 In-vessel fuel storage
13 Coolant conditioning system Figure 2 The conceptual model of subcritical reactor in multipurpose hybrid research reactor for high-tech applications Courtesy of J Bosch ADS Candidate Materials Compatibility with Liquid Metal in a Neutron Irradiation Environment, Doctoral Thesis, ISBN 978-90-8578-241-4, 2008; 7.
Trang 4cooling tubes, spallation loops, primary heat
exchan-gers, and so on, but without a beam window.15In this
system, a high-energy proton beam with an energy of
600 MeV is injected directly into the free surface
of the lead–bismuth in the subcritical reactor core
The MYRRHA project aims to serve as a basis for
the European experimental ADS In the first stage,
the project focuses mainly on demonstrating the ADS
concept, safety research of subcritical systems, and on
nuclear waste transmutation studies Subsequently,
MYRRHA will be used as a fast spectrum irradiation
facility dedicated to research on structural materials,
nuclear fuel, liquid metal technology, and associated
aspects on the one hand and as a radioisotope
produc-tion facility on the other The system consists of a
proton accelerator that supplies a 600 MeV 3–4 mA
proton beam to a LBE spallation target, delivering the
primary neutrons, which in turn couples to a
LBE-cooled subcritical fast core The structural materials for
MYRRHA need to withstand temperatures ranging
between 200 and 550C (normal operating
tempera-ture between 300 and 450C) under high spallation
neutron flux and contact with liquid LBE It is clear
that the candidate materials need to fulfill challenging
requirements such as high thermal conductivity, high
heat resistance, low thermal expansion, low
ductile-to-brittle transition temperature (DBTT) shift,
suffi-cient strength at elevated temperatures with limited
loss of ductility and toughness, low swelling rate, high
creep resistance, and good corrosion resistance.14
Studies of LA for developing ADS are also reported
from the points of view of conceptual ideas16,17 and
related facility.18
5.09.3 Ferritic–Martensitic Steels
One method of using materials such as F/M and
austenitic stainless steels in LA is to keep an oxide
layer on the surface of the base metal in contact with
LA by controlling the oxygen concentration in the
LA.19–21Too little oxygen in LA will lead to
distion of the protective iron oxide Excess oxygen
solu-tion in the LA will lead to the producsolu-tion of lead
oxide that could plug the cooling tubes Theory
pre-dicts that an adequate oxygen concentration in LA
exists between, for example, 106and 104wt% in
the temperature region of 400 and 700C An
alter-native method is to add anticorrosion elements such
as Al to the surface, which leads to a protective oxide
that guards base metals, as mentioned in the section
on surface treatment
The oxide scale is not a simple structure but consists of duplex layers: magnetite Fe3O4 near the
LA side and spinel (FeCr)3O4 near the base metal The original surface exits at the interface between the magnetite and spinel but not at the front surface
of the magnetite near the LA An early question was how the oxide layers on the surface of the base metal grew
Martinelli et al.22–24reported a global study on the oxidation process of Fe–9Cr–1Mo martensitic steel (T91) in static LBE The isotope tracer oxygen-18 was employed in the corrosion test Also, the mass balance
of Fe and Cr was investigated theoretically They explained the Fe–Cr spinel growth rate mechanism as follows: The oxidation reaction can occur because of the presence of nano-channels Nano-channel forma-tion is achieved by the dissociative/perforative growth
in the magnetite The nano-channel allows a fast diffu-sion of oxygen to the T91/spinel interface Oxygen cannot diffuse in the oxide lattice because its rate is insufficient for Fe–Cr spinel formation, but is instead transported via short cut diffusion paths Even if oxygen diffusion in grain boundaries could be possible, oxygen would likely diffuse inside channels The nano-channels are, in some cases, called lead nano-nano-channels because of the results of the LBE oxidation tests Liquid metal does not penetrate evenly in the oxide scales; only lead penetrations are observed Nevertheless, in pure bismuth oxidation tests, bismuth penetrations are also observed in the scales On the other hand, the iron diffusion from T91 to the magnetite/Pb–Bi interface leads to vacancy formation at the T91/Fe–Cr spinel interface Because of the presence of chromium atoms, these vacancies can accumulate to form nano-cavities
at the T91/Fe–Cr spinel interface This accumulation
is quasi complete; very few cavities are annihilated
on the T91/oxide interface The Fe–Cr spinel grows inside the nanocavity until it is completely filled
At that moment, the oxygen can no longer reach the T91 alloy and the oxidation reaction interrupts itself The formed Fe–Cr spinel thickness then becomes equal to the consumed T91 thickness because of this self-regulation process, as shown in Figure 3
In this process, the limiting step of the Fe–Cr spinel growth rate is thus the ‘iron diffusion’ across the oxide scale
A key issue in maintaining structural integrity is to maintain high performance of the welded materials The corrosion properties between the base metal and the weldment were investigated.25The materials tested were F/M steel F82H26and the electron beam (EB) welding of F82H The chemical composition
Trang 5of F82H is 8Cr–2W–0.2VTa–bal/Fe (wt%) Oxygen
concentration was controlled to (2–4) 105mass %
Welded materials were prepared with a
bead-on-plate weldment with a 15 mm depth of melting
F82H steel was welded after preheating at 300C,
heat-treated at 300C for 2 h, and then annealed at
750C for 2 h for stress relief.Figure 4shows optical
microscope observation of cross-section for F82H
specimens and an impinging-flow simulation around
the specimen It was observed that the welded metal
of F82H revealed a coarse martensitic structure in
comparison with the fine microstructure in the
non-welded region because of melting and resolidification
in the welding process The corrosion depth in F82H
was limited near the surface of the material A failure
of the outside layer in the duplex corrosion layers was
observed The heat-affected zone showed that the
martensitic structure became fine because of the
rapid heating and cooling during the welding process
Regardless of the difference in microstructures, the
corrosion layer showed no apparent difference The
growth of the corrosion depth, defined by the layers
of magnetite and spinel, followed a parabolic law,
where diffusion controls the process The result of
the flow simulation of LBE impingement indicated
that the velocity varied from 0 to 1 m s1 near the
specimen surface At higher temperatures, for
exam-ple, above 500C, the internal oxide layer or
diffu-sion zone was clearly identified Furukawa et al.27
observed three layers, consisting of the duplex layers
mentioned earlier and a diffusion zone in the base
metal beneath the spinel layer in the static LBE
test at 500 and 550C under the oxygen control
to 106wt% for high Cr steel (10.54Cr–1.75W–
MnMoV) with heat treatment: 1070C, 100 min
air-cooled; 770C, 440 min air-cooled
Tan and Allen tested high Cr steel material in the
DELTA loop, at Los Alamos National Laboratory
(LANL) The material tested was HCM12A,
pro-cured from Sumitomo Metal Industries, Ltd., with
composition provided by the supplier: 10.83Cr– 1.89W–1.02Cu–0.64Mn–0.39Ni–0.30Mo–0.27Si–0.19V– 0.11C–0.063N–0.054Nb–0.016P–0.002S–0.001Al–3.1
105B, and balance/Fe (wt%).28The chemical com-position and heat treatment of this material are slightly different from those used in the experiment by Furukawa and Muller HCM12A is one of the third-generation 12Cr ferritic steels with tempered martens-ite,29which was originally developed for heavy section components such as headers and steam pipes for use
at temperatures up to 620C and pressures up to
34 MPa30 with good resistance to thermal shocks.31 The HCM12A was received after being annealed at
1050C and tempered at 770C.28They compared the oxide layer to the porous magnetite layer on the super-critical water exposed sample at 600C, 667 h Tem-peratures at both conditions were different It was found that detachment of most of the magnetite non-protective layer occurred on the LBE-exposed sample
at 530C–600 h earlier in time than models developed
by Zhang and Li From a technical experimental point
of view, it is the issue how to detect the original surface
of base metal in order to evaluate the oxide thickness
A thin yttrium coating layer will help to detect it in the LBE corrosion test
At temperatures above 600C, the oxide layer grew thinner with increasing temperature, which suggests that around this temperature, a change occurred in the mechanism of oxidation At 570C, FeO-wustite
is formed Compared with magnetite, wustite has a lower standard free energy of formation, which ensures its stable existence at low oxygen potential In fact, the layer was formed in the region between magnetite and base metal Also in this temperature range that is beyond the point of oxidation mechanism change, dis-solution attack was observed at several points, and the number of such points increased with prolongation of run duration The observations would suggest lowering the maximum processing temperature in LBE applica-tions from the point of view of the static LBE test.27
Newly formed oxide
Nano-channel Oxygen
Original metal
surface
Iron
Nano-channel
Nano cavity
Figure 3 Self-regulation of the Fe–Cr spinel growth.
Trang 6Oxide layer
(a)
Heat-affected zone Welded zone
(e)
Oxide layer
(f) Oxide layer
(b)
Tip
(d)
Oxide layer
(c)
(g)
1 mm
20µm 20µm
1 mm
Specimen
Y LBE flow
x
1.089
Velocity magnitude
1.012 0.9338 0.7783 0.6228 0.4673 0.3118 0.1563 0.7851E−01 0.7603E−03
z
M/S Local MX = 1.089 Local MN = 0.7602E-03
*Presentation grid*
Pb–Bi
STAR
PROSTAR 3.10
Figure 4 Optical microscope observation of cross-section for F82H specimens and an impinging-flow simulation around the specimen (a) Macro structure, including welded zone and heat-affected zone, (b) macro structure at the specimen end where lead–bismuth eutectics impinges from the right hand side indicated with an arrow, (c) micro structure of welded zone tested at 450C for 1000 h, (d) micro structure of tip region tested at 450C for 1000 h, (e) cross-section of tip region tested at 450C for 3000 h, (f) cross-section of tip region tested at 500C for 1000 h, and (g) simulated flow profile
of lead–bismuth eutectics around the specimen.
Trang 7Hosemann et al attempted nano-scale
characteri-zation of HT-9 (11.95Cr–1Mo–0.6Mn–0.57Ni–
0.5W–0.4Si–0.33V–bal/Fe (wt%)) by using atomic
force microscopy (AFM), using a function of
mag-netic force microscopy (MFM) and C-AFM C-AFM
is a contact mode electrical characterization
tech-nique that involves applying a voltage typically
between the conductive AFM tip and the sample
while monitoring variations in the local electrical
properties in a range of picoamperes to
microam-peres The HT-9 tube was tested at 550C in flowing
LBE under 106wt% oxygen for 3000 h.32 It was
found that the oxide consists of at least four different
layers with different grain structures and therefore
conductivity/magnetic properties The outer layers
seem to be Fe3O4and have good conductivity, while
the inner layer is Cr enriched and has lower
conduc-tivity or is insulating This is in agreement with the
literature where Cr additions lower the conductivity
of Fe3O4 The outer layer can be divided into two
distinct areas based on a change in grain structure
The inner oxide layers adopt the grain structure from
the bulk steel High pore density within these layers
suggests that these are fast diffusion paths allowing Fe
diffusion outward and O diffusion inward The LBE
corrosion experiment in the DELTA Loop on T91,
HT-9, and EP823 conducted for 600 h at 535C
showed multilayer oxides on the tested materials
The wave-dispersed X-ray analyzer (WDX)
measure-ments on the cross-sections revealed two Cr and Fe
containing oxide layers and no Fe3O4layer It appears
that the main difference between observed oxide
layers is the Fe content and the microstructure
Nano-indentation tests across the oxide layers
were performed.33 The results showed lower values
of E-modulus in these oxide layers than that of the
bulk steel layers and higher hardness values for the
oxides than that of the bulk steel The inner oxide
layer is softer than the outer oxide layer This might
be due to the fact that the inner oxide layer has higher
porosity than the outer layer
Yamaki and Kikuchi34 conducted a mechanical
test of oxide scales The beam window at the
bound-ary of the high-energy proton beam and reactor core,
as shown in Figure 1, is loaded by thermal stress
and buckling load in the deep LBE of the reactor.35
The specimen was a ring made from the F/M steel
pipe, HCM12A The inner surface of the pipe had
been exposed to flowing LBE during the loop
opera-tion at 400–500C for 5500 h under an oxygen
con-centration in the range from 1 105to 5 105wt%
Apparently, the oxide layer had a duplex structure
Possibly they were outside the magnetite and inner
side spinel Figure 5 shows the results of the ring compression test The HCM12A ring was com-pressed by 50% and unloaded Near position A, cracking occurred because of excess strain to the spinel layer rather than the Fe3O4 layer This was caused by the fact that the Young’s modulus of Fe–Cr spinel layer was lower than that of Fe3O4 layer by 10% with the same hardness in both layers Near position E, the oxide layer was spalled off from the boundary between the base metal and the spinel
F/M steels can guard the base metal by forming a spinel oxide layer The formation mechanism is con-trolled by the iron diffusion rating The magnetite oxide layer is not protected against contact with LBE Under the tensile stresses, excess strains will spall off the oxide layer from the interface between the base metal and the spinel layer Over 570C, the oxide formation mechanism is changed by the forma-tion of wustite
and Austenitic Steels
Muller et al demonstrated that the effect of Al-alloying into the surface of the base metal was the F/M steel OPTIFER IVc (10Cr–0.58Mn–0.56C–0.40W– 0.28V–bal/Fe (wt%)) using electron pulse treatment, GESA (Gepulste Elektronenstrahlanlage).36,37There was no corrosion attack visible in any part of the alloyed portion after 1500 h exposure to liquid lead
at 550C with 8 106at.% oxygen The alumina layer that must have formed at the surface during oxidation in lead might be very thin and could not be detected Only the unalloyed part of the surface was covered with thick oxide scales The results also suggested that the Fe–Cr spinel layer ends at the original specimen surface This surface treatment had a similar result when applied to an austenitic steel base metal 1.4970(16.5Cr–13.8Ni–1.91Mn– 0.81Si–MoTi–bal/Fe (wt%))
Weisenburger et al examined T91 tubes with modified FeCrAlY coatings in LBE These coatings are often used for turbine blade protection.38 The coating had an average thickness of 30mm after appli-cation by a plasma spray method and was remelted using the pulsed large area GESA EB to gain a dense coating layer and to improve the bonding between the coating and the bulk material They intended to simulate a cladding material’s environment by using a pressurized tube type specimen The results showed
Trang 8that the tangential wall stress of about 112.5 MPa
induced by an internal tube pressure of 15 MPa
increased the Fe diffusion and led to enhanced
mag-netite scale growth Coated specimens, however, have
no magnetite layer This is another advantage of the
coating Energy-dispersed X-ray analyzer (EDX) line
scans of the cross-section of the coated T91 tube
specimen after 2000 h exposure to LBE at 600C
show that the top oxide scale must consist mainly of
alumina followed by a thin layer enriched in Cr
These layers protect the steel not only from LBE
attack but also from oxygen diffusion into the coating
and bulk material The coating process needs some
improvement to avoid coating regions with
alumi-num concentrations below 4 wt%; otherwise, the
oxide layer will grow in the same manner as the
original material Steels with 8–15 wt% Al alloyed
into the surface suffer no corrosion attack for all
experimental temperatures and exposure times.39
Technical concerns about surface treatment are the
effect of cyclic loading on the low cycle fatigue
endurance in air and LBE Low cycle fatigue tests
were conducted in LBE containing 106wt%
dis-solved oxygen with T91 steel at 550C T91 was
employed in two modifications, one in the as-received state and the other after alloying FeCrAlY into the surface by pulsed EB treatment (GESA process) Tests were carried out with symmetrical cycling (R ¼ 1) with a frequency of 0.5 Hz and a total elonga-tionDet/2 between 0.3% and 2% No fatigue effects from LBE could be detected Results in air and LBE showed similar behavior Additionally, no difference was observed between surface treated and nontreated T91 specimens.14
A melting process of coating materials enhances bonding between the coating and the bulk materials Heat deposition because of a pulsed EB exposure successfully demonstrated that remelted alumina or FeCrAlY coating was effective in protecting the base metal property from LBE attack
Dispersion-Strengthened Steel
Takaya et al.40investigated the corrosion resistance of ODS steels with 0–3.5 wt% Al and 13.7–17.3 wt% Cr,
at 550 and 650C for up to 3000 h in stagnant LBE
(a) Flat plate
Initial
diameter
(24 mm)
D
10 mm
27.5 mm E
G F H I
Load
Load
Specimen
(b)
(c)
(d)
(e)
(f)
Base metal
Oxide scales
Crack Delamination
d width
d interval
10 mm
100 mm
C D
E
Figure 5 The ring compression test (a) HCM12A ring model before compression, (b) the ring model after compression
by 50%, (c) the ring after unloading, (d) simulation of maximum principal strain distribution induced at loading, (e) the cross-section near position A, and (f) the cross-section near position E.
Trang 9containing 106 and 108wt% oxygen The ODS
steels were manufactured by hot extrusion of
mechan-ically alloyed powders at 1150C, and consolidated
bars were annealed by 60 min of heat treatment at
1150C, followed by air cooling Chemical
composi-tions of ODS materials are (13.7–17.3) Cr–(1.9–3.5)
Al–(0.34–0.36)Y2O3–TiSiMn–bal/Fe (wt%)
Protec-tive Al oxide scales formed on the surfaces of the
ODS steels with 3.5 wt% Al and 14–17 wt% Cr, and
no dissolution attack was seen in any of the cases
Addition of Al is very effective in improving the
corrosion resistance of ODS steels in LBE On the
other hand, the ODS steel with 16 wt% Cr and no Al
showed no corrosion resistance, except in the case of
exposure to LBE with 106wt% oxygen at 650C
Thus, the corrosion resistance of ODS steels in LBE
may not be improved solely by increasing Cr
concentration
There is additional data reported by Hosemann
et al on ODS alloys in LBE Specimens were exposed
to flowing LBE in the DELTA Loop at LANL
at 535C for 200 and 600 h The oxygen content
in the LBE was about 106wt% The detailed
manufacturing process was not disclosed Conclusively,
PM2000, which has a chemical composition of 20Cr–
5.5Al–0.5Y2O3–0.5Ti–bal/Fe (wt%), showed a very
dense, thin, and protective oxide layer because of its
higher Al content The compositions of the oxide layers
found on the Al alloyed materials change with depth
Elements are oxidized based on the amount of oxygen
available for oxidation and the free energy of the oxide
It appears that at least 5.5 wt% Al in the alloy is
necessary to form a protective Al-enriched oxide.41
The oxide scale has a duplex structure below 500C
Over 500C, a diffusion zone in the base metal is
apparently observed The oxide layer appears to consist
of three layers, that is, the duplex layers plus the
diffu-sion zones The oxide scale does become unstable The
outer magnetite layer is prone to be spalled off in the
flowing LA In such an environment, the Al coating is
found to be effective in enhancing corrosion resistance
Remelting processes, for example, by GESA EB
expo-sure, make a good Al layer The disadvantage of the
coating method is the disintegration or cracking due to
an uncontrolled process This cracking could be
attrib-uted to the local reduction of Al content The ODS
alloy is developed for cladding materials Materials
development is progressing in the direction of
high-Cr Al-ODS alloys The recommended Al composition
in ODS alloys varied from 3.5 to 5.5 An adequate
amount of Al will balance the corrosion resistance and
mechanical strength Excess Al will reduce mechanical
strength The reason why the Al enrichment in Fe-base steel improves corrosion resistance in LA will be determined in future investigations
ODS steel aims at enhancing the strength of material applicable to the cladding materials of a fast reactor The addition of Al to ODS improves corrosion resistance in LBE at the fuel cladding temperatures On the other hand, the excess addi-tion of aluminum reduces the strength of materials
An optimization is needed to balance the two fac-tors at around 5%
5.09.6 Austenitic Stainless Steels
Austenitic stainless steels are candidate materials for the spallation target window in ADS In MEGAPIE, however, F/M steel, T91, was used for the beam window in flowing LA, and this was acceptable for a limited duration (4 months) The lifetime of the beam window of the T91 liquid Pb–Bi container in the MEGAPIE target was summarized based on the pres-ent knowledge of LBE corrosion, embrittlempres-ent, and radiation effects in the relevant condition.42 It was suggested that the lower bound of the lifetime of the T91 beam window was determined when the steel became brittle at the lowest operation temperature,
230C, with a safety margin of 30% Evaluation using the DBTT data and fracture toughness values
of T91 specimens tested in LBE, a dose limit of about
6 dpa, corresponding to 2.4Ah proton charge to be received by the target in about 20 weeks in the nor-mal operation condition, was set
In the ADS design, for example, the beam window material will produce about 1000 appm (3Heþ4
He)
a year by 1.5 GeV proton beam bombardment in the reactor core for austenitic stainless steel, Japanese Primary Candidate Alloy (JPCA), and F/M steel, F82H.43 The helium production of 1000 appm He suggests that the DBTT will increase by 400–
500C.44 This increase will set the design tempera-ture at the beam window at 450–500C The use of
a F/M steel may lead to a brittle fracture, and those materials should be avoided in operations for extended times Therefore, austenitic steel is the candidate material The production of hydrogen and helium in JPCA was slightly larger, 3–4%, than that of F82H because of the addition of nickel and boron JPCA, in which the chemical composition
is 0.50Si–1.77Mn–0.027P–0.005S–15.60Ni–14.22 Cr–2.28Mo–0.24Ti–0.0031B–0.0039N–bal/Fe (wt%), was developed to reduce the helium embrittlement
Trang 10of austenitic steel for first wall and blanket
struc-tural components in fusion reactors.45The optimized
JPCA material is manufactured by vacuum induction
melting, vacuum arc melting, and solution-annealing
at 1100C for 1 h The TiC precipitates within the
matrix and on the grain boundaries serve as trapping
centers for the helium produced during neutron
irra-diation However, dissolution of the MC precipitates
initiates the onset of helium embrittlement as well as
high swelling during high fluence neutron
irradia-tion The improved stability of the MC precipitates,
which formed in the matrix during irradiation,
pre-vents loss of ductility at 500C and below
The corrosion properties of an austenitic stainless
steel at low temperature demonstrated good
endur-ance for material usage in LBE during a short time,
approximately at 300 and 470C for 3000 h for 1.4970
austenitic stainless steel,46 and at 420C for 2000 h
for 1.4970 austenitic stainless steel and 316L at an
oxygen concentration of 106wt%.47No dissolution
was seen in the aforementioned results A thin oxide
scale may protect the material from attack in LBE
As demonstrated in Figure 4, a corrosion test
under impinging flow was also conducted on JPCA
and its EB welded bar at an oxygen concentration of
2–4 105wt%.48 The EB welded metal of JPCA
exhibited a dendritic structure 1 mm in width, but a
heat-affected zone was not visible Scanning electron
microscopy (SEM) observation showed no corrosion
layer for the specimens tested at 450C and 1000 h
But at 3000 h, a thin corrosion layer could be observed at 1–2mm in depth For the weld joint, the depth of the corrosion layer as well as corrosion morphology showed the same results as with the parent material The results of X-ray diffraction ana-lyses showed how the oxide layer developed at 450 and 500C Figure 6 shows X-ray diffraction ana-lyses of the JPCA specimens under the conditions of
1000 h at 450C (top, JPCA-1), 1000 h at 500C (middle, JPCA-2), and 3000 h at 450C (bottom, JPCA-3) Oxidation of the JPCA at 450C pro-gressed in the same manner as at 500C
5.09.7 Precipitation Formation
The dissolution of Ni, Cr, and Fe from structural materials into LA was studied Saturated solubility
of Ni in LA is calculated to be a couple of wt% at 450–500C.20Corrosion–erosion tests have been con-ducted at the JLBL-1 facility of the Japan Atomic Energy Agency (JAEA) The main circulating loop was made of SS316 austenitic stainless steel, and con-sisted of the specimens at high and low temperatures, filters, a surge tank, a cooler, an electromagnetic flow meter, a surface-level meter, thermocouples, and a drain tank.49 The loop was operated at a maximum temperature of 450C with a temperature difference
of 50–100C and average flow velocity of 1 m s1 The oxygen concentration was estimated to be 107wt%,
0
500
1000
1500
2000
200
400
600
800
1000
1200
200
400
600
800
1000
1200
1400
1600
JPCA-3
JPCA-2
JPCA-1
Cr0.19, Fe0.70, Ni0.11 Bi
M3O4
Figure 6 X-ray diffraction analyses of JPCA specimens under the condition of 1000 h at 450C (top, JPCA-1), 1000 h
at 500C (middle, JPCA-2), and 3000 h at 450C (bottom, JPCA-3).