Structural Materials for Fusion Reactors
M. Victoria, N. Baluc and P. Spätig
EPFL-CRPP Fusion Technology Materials, CH-5232 Villigen PSI, Switzerland
e-mail: Philippe.Spatig@psi.ch
Abstract: In order to preserve the condition of an environmentally safe machine, present selection of materials
for structural components of a fusion reactor is made not only on the basis of adequate mechanical properties,
behavior under irradiation and compatibility with other materials and cooling media, but also on their radiological
properties, i.e. activity, decay heat, radiotoxicity. These conditions strongly limit the number of materials
available to a few families of alloys, generically known as low activation materials. We discuss the criteria for
deciding on such materials, the alloys resulting from the application of the concept and the main issues and
problems of their use in a fusion environment.
1. Introduction
The development of adequate structural materials is a major step towards fusion reactors
becoming an efficient source of energy, particularly if the promise of an environmentally safe
machine is to be maintained. The first wall, divertor, limiters and breeding blanket components
are subjected not only to the high energy neutron environment resulting from the fusion
reactions, but also to strong mechanical, heat and electromagnetic loadings. These conditions
create a very severe operation frame for the materials used in these elements and have led to a
long developmental path for the candidate materials.
The main idea in the materials development program to comply with the safety conditions is the
development of low activation materials. The criteria for deciding on such materials, the
resulting material families resulting from the application of the concept and the main issues
and problems of their use in a fusion environment are discussed below.
2. Effects on materials due to the fusion reactor environment.
Eighty percent of the energy released by the D-T fusion reaction are transferred by 14 MeV
neutrons to the first wall and breeding blanket. The remaining 20% are carried by α-particles
issued from the same reaction, that together with other low energy neutral and charged particles
will induce sputtering, erosion and blistering in the plasma facing materials.
About 10% of the energy of the 14 MeV neutrons will be deposited in the first wall, the
remaining energy being transferred mostly to the blanket [1]. Two types of radiation effects
are produced in the materials:
(i) Inelastic (n,x) interactions with nuclei which yield transmutation products and lead to the
production of He, H and other impurities in the bulk of the material.
(ii) The neutrons themselves plus the recoils resulting from the above nuclear reactions transfer
energy to lattice atoms through elastic collisions and displace them from their normal sites.
Through an iterative process, a displacement cascade is formed. About 10% of the
vacancies and interstitial defects originally formed survive the evolution of the cascade and
lead to the formation of a defect microstructure that hardens the material, to the formation of
voids, to the redistribution of elements in the alloy inducing segregation and possibly to
phase transformations.
The neutrons slow down as they penetrate the reactor structure, so the neutron spectrum changes
with penetration depth. This will modify the recoil energy spectra as well as the transmutation
rate, the magnitude of the radiation effects will be typically different in the first wall and blanket.
Radiation damage depends strongly on irradiation temperature Tirr and three regions can be
recognized in relation to the melting temperature of the material TM:
(i) At low temperatures (Tir < 0.3TM) the vacancies do not yet evaporate from their clusters. The
microstructure is dominated by defect clusters and results in radiation hardening and a
degradation of the fracture toughness (embrittlement) of the material, increasing its ductile to
brittle transition temperature.
(ii) In the region of 0.3Tirr < Tirr < 0.5 TM, defects are strongly mobile and this results in
phenomena such as radiation creep and swelling.
(iii) At still higher temperatures, the presence of He leads to embrittlement through the formation
of bubbles at grain boundaries.
All the above described effects limit the life of the components.
3. Low activation materials
While the burning of fission fuel produces long lived actinides, the fusion reaction does not
intrinsically yield other radioactive elements. But fusion neutrons will activate the materials
surrounding the plasma and this fact conditions the waste management and disposal scenarios.
Under accident conditions, the decay heat is an important parameter, since heat enhances
oxidation and possible volatilization and such release to the environment is the main
contamination hazard in a loss of coolant condition [2].
In order to evaluate these parameters, extended cross section data libraries together with decay
and activation data have been developed. In Europe, the present version of the
FISPACT97/EAF97 [3,4] inventory code together with the decay and activation libraries EASY
[5], which includes five reference fusion reactor neutron spectra (first wall, blanket, shield and
two magnetic coils), have been used to show that only a few primary elements can be considered.
They are C, Si, Ti, Fe, Cr and V. Other elements, such as W can be used in limited quantities.
Moreover, all of the evaluations made up to present show the importance of the presence of
typical tramp impurities present in these metals, such as Al, Ni, Ag, Co, Nb and others, that are
detrimental because of their poor radiological properties.
4.Main issues in development of Low Activation Materials
The most intensive development in the past fifteen years has been that of the ferritic-martensitic
steels (Low activation steels or LAS), based on the 7-10CrWVTa composition and which has led
to the production of large casts in Japan (F82H steel) and in Europe (the EUROFER97 steel).
These steels have reasonable thermophysical properties and, from irradiation experience in fast
reactors to well over 100 dpa [6], a substantial resistance to swelling and high temperature
embrittlement. Moreover, they have good compatibility with either water or He coolants and Li-
breeders. They are being developed for a temperature window up to 773 K.
A critical issue in these steels is the existence of a ductile to brittle transition temperature
(DBTT), which increases with irradiation to temperatures well above room temperature, making
the steel unusable as a structural component. It has been shown that reducing the Cr content to
less than 9 % strongly decreases the sensitivity to this phenomena [7] and in Europe a number
of LAS families (OPTIFER, OPTIMAX) have been developed on this basis. It has also been
shown that a positive consequence of using pure materials and clean processing is that low
DBTT temperatures are obtained in the unirradiated steel, which increase only moderately after
irradiation. A typical data set is that obtained for the difference in DBTT, measured from Charpy
tests, between the virgin and irradiated condition (∆DBTT) in the F82H and OPTIMAX A LAS,
shown in Fig. 1. In both cases, the DBTT is below room temperature after irradiation to 2.5 dpa.
0
9 She lf Ene rg y [J]
There are a number of issues still open
8
regarding the behavior of these LAS:
7 (i) Although the normal tensile properties
6
5 ∆
DBT T ∆
DBT T
are practically not affected, there is an
≅ 78 K ≅ 141 K
4
O PT IM AX A
F82H
increasing body of evidence showing that,
1 90 K 2 68 K
3
2 1 85 K 3 26 K
contrary to what was expected in this
1
T [K]
temperature region, the presence of He
0
150 200 2 50 300 350
induces an additional shift in the DBTT,
4 00
already at low He contents [8]. There is a
Figure 1:Shift of the DBTT for the F82H and yet no clear unerstanding of what mecha-
OPTIMAX A steels after 2 dpa irradiation nism is operative
(ii) Although the actual reactor is expected to operate in a quasi-continuous mode, the design of
intermediate or demonstration machines operates still in a pulsed fashion (1000 s pulses in
ITER). Such mode introduces mechanical damage through fatigue. As it is important to
understand the actual dynamical situation under irradiation, which is nearer to the behavior of the
material in the reactor, in-beam fatigue tests have been performed in the F82H steel under a 590
MeV proton beam irradiation [9].
(iii) The effects of ferromagnetism on plasma stability have to be investigated.
Titanium alloys have a number of properties that make them attractive structural material
candidates for fusion reactors. High strength-to-weight ratio, intermediate strength values, good
fatigue and creep rupture properties, small modulus of elasticity, high electrical resistivity, heat
capacity, low coefficient of thermal expansion, low long-term (< 10 years after shutdown)
residual radioactivity (after V and Cr, Ti has the fastest decay rate), a high corrosion resistance
together with good compatibility with coolants such as lithium, helium and water, high
workability and good weldability and commercial availability with established mine and mill
capacity are some of the favorable properties [10]. Because of the numerous current applications
of titanium alloys in the aerospace and medical domains, there exists an extended properties
database and industrial experience on these materials. Ti-alloys can be divided into three major
classes determined by their phase constituency, they are referred to as the alpha (α), beta (β) and
alpha/beta (α/β) alloys, where the α phase is hcp and the β phase is bcc. The alloying elements
used in the titanium system can be divided into two classes according to which phase the element
stabilizes. The substitutional α stabilizers are Al, Zr and Sn, while the β stabilizers are V, Cr, Mn,
Fe, Co, Ni and Mo. Furthermore, the intermetallic TiAl has also been proposed in the Japanese
program. Leguey et al. [11] have recently reported results after irradiation with 590 MeV protons
(mean dose rate: 3x10-6 dpa.s-1; mean He production rate: 50 appm/dpa) of pure Ti, for
irradiation doses ranging between 10-3 and 10-1 dpa. Proton-irradiation at 330 K induces a
significant increase in the critical stress measured at 0.2% plastic strain [11] and a decrease in
the work hardening rate, as shown in Fig. 2. Hardening is of the order of about 40% per dpa for
a dose of 0.03 dpa. It clearly relates to the irradiation-induced formation of a high density of
defect clusters observed with transmission electron microscopy (TEM). About 20% of the
observed defects were identified as dislocation loops with a mean apparent size of 5 nm. No
clear correlation could be established between ductility and irradiation dose, due to large
scattering in the measurements. However, the total strain at fracture of all deformed specimens
(irradiated or not) lies between 18 and 30%. The TEM observation of specimens irradiated to
0.03 dpa and deformed up to fracture shows the simultaneous annihilation of defect clusters as a
result from interactions with mobile dislocations together with the formation of twins and
dislocation cells. No void formation was observed. A comparison of the mechanical properties
of the α -Ti-5Al-2.4Sn and α+ β Ti-6Al-4V alloys has been performed by Marmy et al. [12]
also after irradiation with 590 MeV protons. In the unirradiated alloys it was found that the
resistance to tensile deformation of both alloys is very similar. The critical stress measured at
0.2% plastic strain, σ0.2, has a mean value of about 800 MPa at ambient temperature (that is
about 40% higher than that exhibited by a 9CrWV low activation ferritic/martensitic steel) and
450 MPa at 500oC (the same value as that of the ferritic/martensitic steel). Both alloys exhibit
good ductility. The total elongation of the α+β alloy, between 15 and 20%, is slightly higher
than that of the α alloy, whatever the test temperature between 20 and 500oC. After proton
irradiation, see Fig. 3, hardening and loss of ductility are observed in both alloys. However, these
phenomena are much stronger for the α+β alloy than for the α alloy. In fact, irradiation at 623
K affects the α+β alloy strongly, with an increase of about 40 % in the yield strength and a 65
% decrease in tensile elongation, while the α alloy irradiated under the same conditions shows
only very little hardening and about 3% reduction in elongation. The embrittlement of the α+β
alloy is probably associated with the presence of a radiation induced phase precipitation
observed in TEM. Similar phase instability under irradiation in the Ti6Al4V has already been
observed by a number of investigators. Probably the biggest disadvantage of titanium and its
alloys in a fusion environment is their high chemical affinity with hydrogen which leads to
hydrogen embrittlement and a expected large tritium inventory. Early work established that for
impact embrittlement, a critical level of hydrogen concentration was required and that the onset of
embrittlement was probably associated with precipitation of the hydride phase [13].
Subsequently, it has been shown that additions of aluminum raise the critical level of hydrogen
for embrittlement and this was tentatively attributed to an increase in the solubility limit [14]. The
hydrogen solubility in titanium was first estimated to be less than 100 wppm below 523 K [15].
In general, the alloys that consist primarily of α phase have a lower solubility for hydrogen than
alloys that are primarily β phase. In the case of α+β alloys, the solubility increases with
increasing amounts of β phase. The three major sources of hydrogen isotopes in a fusion reactor
environment are [10]: hydrogen produced in the metal by neutron transmutation reactions,
interactions with the deuterium and tritium (D-T) fuel in the plasma chamber and interactions
with tritium in the breeding material.
250 1400
1200 T =RT, T =350°C, Dose=0.3 dpa
Unirradiated
test irr
200
True Stress (MPa)
1000
Stress (MPa)
150 800
as annealed
as annealed 600
100
3x10-3 dpa
400 Ti5Al2.4Sn
50 10-3 dpa
200 Ti6Al4V
3x10-2 dpa
0 0
0 5 10 15 20 25 30 35 0 5 10 15 20
True Strain (%) Strain (%)
Figure 2: Tensile curve on pure Ti before and after Figure 3: Tensile curve of Ti alloys, unirradiated
proton irradiation condition and after 0.3 proton irradiation
Therefore, in order to make practical use of titanium alloys for fusion applications, hydrogen
barrier coatings are necessary to prevent hydrogen intake and eventual embrittlement of the
material. Preliminary attempts of depositing a graded coating with a Cr2O 3 +(Si or Ti)O2 top
layer on the Ti-5Al-2.4V α-alloy have been recently performed at the Materials and Surface
Department of SULZER Innotec AG. These coatings have been tested for structural integrity
and hydrogen permeation and appear quite promising [16].
vanadium alloys are also considered [17] because of their low thermal expansion which coupled
to a low elastic modulus leads to low thermal stresses and a high heat flux capability. Their
compatibility with pure Li makes them a good choice for a liquid lithium coolant breeder blanket
concept. The typical operating window of this design, from 623 to 1023 K and 1 MPa pressure,
requires high temperature strength and good creep properties, conditions that are fulfilled by the
V-(4-5)Cr-(4-5)Ti [18]. The feasibility of fabricating components from the alloy is being
demonstrated by its use in the DIII-D RDP, for which a 1200 kg cast has been produced and
worked into rod, plate and sheet. Postirradiation results [18] after irradiation at temperatures
below 600 K show strong radiation hardening and a reduction of the uniform strain to values
below 1% and an upward shift of the in the DBTT of the order of 560 K. The uniform
irradiation increases at higher irradiation temperatures: is 6% after irradiation at 873 K. A
correlation can be made indicating that brittle fracture behavior occurs whenever the yield
strength exceeds 700 M Pa.
The development of electrically insulating walls for coolant channels is a critical issue for self-
cooled liquid metal systems. Of the candidate coatings AlN deposited by vapor deposition have
been shown unstable, while CaO coatings exhibit good electrical resistance during exposure to
Li at 700 K.
Fibre reinforced SiC-SiC ceramic composites have gained strong interest in the fusion materials
community due to their good low activation and decay heat properties at short and intermediate
decay times, coupled to high mechanical strength for temperatures up to 1273 K. Their
microstructure consists of SiC embedded in a SiC matrix through a fiber-matrix interphase and
are typically synthesized by the chemical vapor infiltration (CVI) process. They have a good
compatibility with He, which makes them primary candidates for a high temperature, He cooled
blanket. Early irradiation results [19] indicated a rapid degradation of the mechanical properties
already at doses of ≈ 1 dpa, due to dimensional instability of the fiber and carbon interphase that
leads to delamination. The present generation of SiC composites, with low oxygen, quasi-
stoichiometric SiC fibers of enhanced crystalline perfection seem to have much improved
properties after irradiation [20]. A number of critical issues are at present of concern:
(i) The relatively low thermal conductivity of SiC composites and the fact that it is further
reduced by irradiation. This has already much improved in the materials produced with the
new Hi Nicalon-S fibers in which a thermal conductivity value of the order of 50 W/mK has
been measured.
(ii) Void swelling in the temperature window of application.
(iii) The (n,x) cross-sections in Si are about one order of magnitude higher than for LAS or V-
alloys (about 150 appm He per dpa). The effects of this large amount of He at high
temperatures is as yet unknown.
(iv) The CVI process produces a microstructure that has about 10% porosity and is therefore
permeable to gases.
(v) Technologically, methods for fabricating large components and of joining the composite to
itself and to other metals need to be developed.
6. Conclusions
• In order to realize the potential safety and environmental advantages of fusion, low activation
materials are being developed within a large international collaboration. The materials choice
in this case is based not only on adequate mechanical properties, behavior under irradiation,
and compatibility with other materials and cooling media, but also on their radiological
properties.
• The alternative alloy classes being studied are ferritic-martensitic steels, V and Ti alloys and
SiC-SiC ceramic composites. At present, the ferritic-martensitic steels have achieved the
greatest technological maturity.
• Common to all the alternatives is the lack of irradiation in a fusion relevant neutron
environment, which limits our knowledge of the behavior of these materials in the proper
reactor conditions (by example, on synergistic He effects on embrittlement, creep and
swelling). Therefore, the availability of an adequate neutron source, such as the proposed
International Fusion Materials Irradiation Facility (IFMIF) is of primary importance for the
future development of the program.
•
Acknowledgements
The present work has been supported by the European Fusion Technology Programme.
References
[1] D.R. Harries, “Ferritic Martensitic Steels for Use in Near Term and Commercial Fusion
Reactors”, Proceedings of the Topical Conference on Ferritic Alloys for Use in Nuclear
Energy Technologies, Snowbird, Utah, June 19-23, 1983.
[2] K. Ehrlich, Phil. Trans. R. Soc. Lond. A 357 (1999) 595.
[3] D.R. Harries, G.J. Butterworth, A. Hishinuma, F.W. Wiffen, J. Nucl. Mater. 191-194 (1992)
92.
[4] R.A. Forrest, J.C. Sublet, FISPACT-97 User Manual, UKAEA FUS 358 Report, 1997.
[5] R.A. Forrest, J. Kopercky, “The European Activation System (EASY)”, IAEA Advisory
Group Meeting on FENDL-2, Vienna, November 1991.
[6] D.S. Gelles, J. Nucl. Mater. 233-237 (1996) 293.
[7] R.L. Klueh and G.J. Alexander, J. Nucl. Mater. 233-237 (1996) 336.
[8] R. Lindau, A. Möslang, D. Preininger, M. Rieth and H.D. Röhrig, “Influence of He on
impact properties of reduced-activation ferritic-martensitic steels”, Proc. Intern. Conf. On
Fusion Reactor Mater. (ICFRM8) Sendai, Japan, October 1997.
[9] P. Marmy, “In beam fatigue properties of the F82H steel”, Proc. Intern. Conf. On Fusion
Reactor Mater. (ICFRM9), Colorado, USA, September 1999.
[10] J.W. Davis and D.L. Smith, J. Nucl. Mater. 85 & 86 (1979) 71.
[11] T. Leguey, C. Bailat, N. Baluc and M. Victoria, to be submitted at the MRS Symposium on
‘Microstructural Processes in Irradiated materials’, Fall Meeting 2000.
[12] P. Marmy, Report on ITER TASK BL 14.2: Titanium Alloys Irradiation Testing (2000).
[13] C.M. Craighead, G.A. Lenning and R.I. Jaffee, Trans. AIME 194 (1952) 1317.
[14] L.W. Berger, D.N. Williams and R.I. Jaffee, Trans. AIME 212 (1958) 509.
[15] G.A. Lenning, C.M. Craighead and R.I. Jaffee, Trans. AIME 200 (1954) 367.
[16] T. Leguey, Private comunication.
[17] S.J. Zinkle, H. Matsui, D.L. Smith, A.F. Rowcliffe, E. van Osch, K. Abe and V.A. Kazakov,
J. Nucl. Mater. 258-263 (1998) 205.
[18] B.A. Loomis and D.L. Smith, J. Nucl. Mater. 212-215 (1994) 799.
[19] P. Fenice, A.J. Frias Rebelo, R.H. Jones, A. Kohyama and L.L. Snead, J. Nucl. Mater. 258-
263 (1998) 215.
[20] A. Kohyama, this Conference.
