Structural Materials for Hybrid Systems: A Challenge in Metallurgy
D. Gorse, J.-L. Boutard
J. Phys. IV France
PrefaceD. Gorse and J.-L. Boutard
The Symposium on "Structural materials for Hybrid systems: A challenge in metallurgy" was held during
the "Journées d'Automne 2001" of the "Société Française de Métallurgie et de Matériaux", Paris
from October 29 through 31, 2001. The editors of this volume, D. Gorse and J.-L. Boutard, would
like to acknowledge GDR GEDEON, CEA - Direction de l'énergie Nucléaire, CNRS - Département des Sciences
Chimiques, et Centre d'études de Chimie Métallurgique for sponsoring this symposium.
This symposium was divided into three
sessions dealing with i) thermodynamics, intergranular
penetration and liquid metal embrittlement, ii) irradiation effects in structural materials and iii)
compatibility of structural materials with lead alloys in relation with R& D studies for MEGAPIE. The
intent of this symposium was to provide a forum for discussing the most recent results obtained in the
frame of the materials research program of the "Groupement De Recherche (GDR) GEDEON". Special emphasis
was given to all factors susceptible to affect the durability of structural materials for spallation
targets, like irradiation effects under proton and neutron mixed spectrum, Liquid Metal Corrosion (LMC)
and Embrittlement (LME).
The material research program of GEDEON is a joint CEA-CNRS venture. In 1997, the GDR GEDEON gave
opportunity to metallurgists and nuclear physicists of both organizations to collaborate for validating
the concept of Pb-Bi spallation targets as a key component for Accelerator-Driven Systems (ADS).
Historically, since 1995, GDR GEDEON has promoted ADS, also called Hybrid System, as an option for
waste management. The starting point of the material program was the 1st GEDEON Workshop on "Materials
For Hybrid Systems" held in Paris in 1997, where reference materials of the 9-12 Cr martensitic steels
series were selected: EM10 (9Cr-1Mo) and the modified 9Cr-1Mo, T91 (9Cr-1Mo-V-Nb). However, besides
their specific concern for ADS, our studies are also of interest for the next generation of LM spallation
targets in EU, U.S.A. and Japan.
These proceedings contain manuscripts from 90% of the presented papers. The organizers would
to thank all their Colleagues who presented papers, contributed with manuscripts and attended the
sessions at the symposium. For sake of clarity, this volume is divided into five sections: 1) general
R& D for spallation targets, 2) irradiation effects in liquid metal spallation targets, 3)
oxygen control: thermodynamics and monitoring, 4) resistance to liquid metal corrosion and embrittlement
of structural materials for spallation targets and 5) basic studies of intergranular penetration and
liquid metal embrittlement.
Section 1 begins with a description of the spallation neutron source facility SINQ and of ongoing
R& D programs at PSI (Switzerland), including MEGAPIE, the joint initiative by six European research
institutions and JAERI (Japan), DOE (USA) and KAERI (Korea) to design, build, operate and assess the
performance of a liquid lead-bismuth spallation target for 1MW of beam power (G. Bauer et al.
materials aspects related to the MEGAPIE target and to the LiSoR (Liquid Solid Reactions under
irradiation) experiment are reviewed by T. Auger et al.
The advantages and drawbacks of solid tungsten
spallation targets, compared to liquid Pb-Bi eutectic spallation targets are examined by R. Enderlé
, presenting the CEA point of view.
Section 2 is dedicated to irradiation effects in Liquid Metal (LM) spallation targets
a crucial problem for the feasibility of ADS. P. Jung is pointing out the specificity of the irradiation
conditions in LM targets by comparison with fast neutron fission and fusion reactors, and the
metallurgical consequences like irradiation and helium-induced embrittlement. The author emphasizes
the importance of spallation residues whose deleterious effects on in-service properties of target
container and window are largely unknown. Until recently, say 1997, only predictions of spallation
residues production were available. This motivated the GEDEON workshop on "Nuclear Data for materials
studies in the frame of hybrid systems" held in 1999 in Vichy (France). The paper presented by C.
Villagrasa et al.
highlights the progress achieved during the past five years, promoted by GDR GEDEON.
The reverse kinematics experiments performed during that period allowed to obtaining reliable values
for production cross sections of residual nuclei produced by spallation reactions. The authors also
mention improvements on the computation codes. The case of the target window, thin target simply made of
iron from the nuclear physics point of view, is thoroughly investigated. The authors also treat the
production of residues in heavy metals, like lead.
The experimental data reported below results from the cooperation of CEA,
CNRS and Forschungszentrum
Jülich within the program SPIRE of the 5th European Framework Program.
From the metallurgists point of view, O. Danylova
reported an interesting attempt to
simulate by doping production of three spallation residues, namely phosphorus, sulfur and titanium,
certainly detrimental regarding the mechanical properties (fracture toughness...) of the 9% Cr martensitic
steels pre-selected for the target window and container. The authors describe the inherent difficulties
encountered for obtaining the microstructure and modified steel composition expected under irradiation.
In all cases, production
of hydrogen and helium is expected very high, ~90 000 appm/year for
hydrogen and ~5000 appm/year for helium, corresponding to a damage level of ~100 dpa/year in the target
window. These estimated values are just mentioned to show the importance of the metallurgical problem
to be faced. These data motivated studies of helium implantation effects by J. Henry et al.
9Cr-1Mo (EM10) and modified 9Cr-1Mo-V-Nb (T91) pre-selected martensitic steels. Homogeneous implantation
of 5000 appm He in T91 and EM10 thin foils followed by TEM examination and tensile testing revealed an
important hardening and embrittlement at 250 °C. At 500 °C both materials exhibit far less hardening and
retain significant ductility. In the latter case, those findings were correlated with the appearance of
helium bubbles detected in Transmission Electron Microscopy (TEM) examination. At 250 °C, no He bubble
is detected by TEM. Conversely, Small Angle Neutron Scattering (SANS) allows to identify a high density
of objects ~1-2 nm, most probably tiny He bubbles. The low temperature He embrittlement is most probably
due to the conjunction of grain hardening and decrease of the cohesive energy of prior austenitic grain
boundaries due to He segregation.
Besides hydrogen and helium, significant quantities of other spallation residues will be
accumulated in the target window during one year of operation. About 1000 appm of calcium form after
one year of irradiation with 1 GeV protons. This exemplary insoluble spallation residue could also
contribute to embrittle the window. G. Amiri et al.
studied the implantation of both calcium and sulfur
(up to 1-2 at% corresponding to a 20 dpa damage level) at 500 °C, assumed to be the highest temperature
of the target window in service conditions. P. Pareige et al.
detected the formation of Ca nanoclusters
using the 3D atomic probe. Neither calcium sulfides nor carbides were detectable whatever the
temperature (300 °C or 500 °C) and damage levels. E. Cadel, P. Pareige and M.O. Ruault showed how
powerful is the 3D atom probe to characterize both chemical and structural evolution of irradiated
steel specimens, at a nanometric scale allowing for interpretation of radiation damages "from first
Section 3 is dedicated to oxygen control, including thermodynamic aspects and monitoring. The GDR
GEDEON material program promoted the development of oxygen sensors in liquid metals. V. Ghetta et al.
studied carefully the possibilities, limitations of existing sensors and proposed new developments
allowing to interpret e.m.f measurements at temperatures well below 450 °C. Intercomparisons between
oxygen sensors, using various static facilities (BIP, JACOMEX and COLIMESTA) were carried out by J.-L.
Courouau et al.
This study is now going on including our EU partners in the closely related TECLA
program granted by the 5th framework program. C. Lesueur et al.
developed original methods for measuring
wettability, permitting to determine the stability of native oxides (on Al, Fe, Ni and also T91 steel)
in contact with molten lead. A. Maitre et al.
presented a preliminary thermodynamic study of the quinary
Bi-Fe-Hg-O-Pb system, beginning with the Bi-Hg-Pb system. No solid ternary phase was found in Pb-Bi
eutectic in the temperature range expected for the spallation target. Then the authors focused on the
possible formation of oxygenated compounds of low melting point either at the T91 steel/Pb or at the
T91 steel/Pb-Bi eutectic interface.
Section 4 concerns the resistance to LMC and LME of structural materials pre-selected
spallation target of ADS, with special emphasis on the liquid Pb-Bi eutectic target concept. The
compatibility of 9% Cr martensitic steels (like EM10 or T91...) with liquid Pb and Pb-Bi eutectic was
examined. Qualification of the T91 steel/LBE system is required to build the liquid Pb-Bi target in
PSI in 2005 and for further applications. A comparative study of the long-term corrosion behaviour of
various martensitic and austenitic steels in forced circulation loop was carried out by F.
Balbaud-Celerier et al.
allowing to vary the temperature and oxygen content in LBE covering the range
10-8wt% -10-6wt.% . This work results from a fruitful collaboration with IPPE (Obninsk, Russia). The
two following papers of Section 4 deal with the tensile behaviour of T91 steel in contact with lead
and its alloys. The most severe metallurgical (J.-B. Vogt et al.
) and environmental (J.-L. Pastol
) conditions are considered in order to estimate the risk of LME for the target window in service.
J.-B. Vogt et al.
observed a transgranular (TG) brittle fracture after hardening heat treatment of T91
steel after tensile testing of notched specimens at 350 °C in oxygen-saturated lead. Note that cases
of TG-LME are rather scarcely observed. On the other hand, J.-L. Pastol et al.
obtained an embrittling
effect on notched specimens by optimizing the environmental parameters (in LBE at 350 °C under He-4% H2
cover gas...) and tensile testing conditions. F. Gamaoun et al.
showed that a few days exposure of T91
steel in lead under either reducing or oxidizing conditions gives rise to a significant porosity in the
steel. This could justify the increased stress relaxation kinetics of specimens submitted to 4-point
bending tests under the same conditions.
Besides the R& D research related to the spallation target for ADS, a non-negligible
part of the
GEDEON material research program was dedicated to intergranular penetration and LME. Some exemplary
LME studies are reported in Section 5 of this volume, illustrating various aspects of this very old,
tricky and exotic problem in metallurgy. One can find in this volume studies involving various Solid
Metal/Liquid Metal (SM/LM) couples, going from high purity metal couples (Ni/Pb, Ni/Bi...) to industrial
metallic alloys put in contact with liquid metals or alloys of varying purity. LME is a multi scale
phenomenon that cannot be explained on the sole basis of either thermodynamics, metallurgical or
kinetics criteria. In this section, the role of the stress-strain state is emphasized. The importance
of the microstructure of the grains boundaries (facetting...) on LM penetration is also pointed out.
This aspect of the problem can now be dealt with thanks to Atomic Force Microscopy (AFM), High
Resolution Transmission Electron Microscopy (HRTEM), Electron Energy Loss Spectroscopy (EELS) and
Auger Electron Spectroscopy (AES) techniques. Crack propagation kinetics can now be followed at ESRF
(Grenoble) using synchrotron radiation imaging techniques. It is now possible to follow the propagation,
structure and chemistry of the embrittling phase in the crack, at microscopic and at nanometric scale.
It is also possible, in principle, to take into account the role of defects in grain-boundaries (GB) on
the penetration of the embrittling species and conversely to investigate the role of the defects produced
by wetting on the GB microstructure. We do hope that these techniques open a new and attractive field
of research to improve the understanding of LME.
The HRTEM study carried out by A. Charai et al.
enlightens the importance of crystallography
on wetting. Two exemplary couples are thoroughly investigated: the Mo/Ni couple giving rise to an
interfacial nanometer thick interface for one type of Mo bicrystal wetted by nickel and the Ni/Pb
couple pointing out the role of the GB plane on wetting.
Y. Brechet et al.
investigated not only the microstructure effects on "LM" penetration into
cracks (cracks kinetics and morphologies), but also the plasticity effects ahead of the crack tip,
studying a variant of the mythic Al/Ga couple, say the 7010 Aluminum alloy in contact with Gallium,
allowing for modifications of the precipitation state at the GBs by applying various heat treatments.
Y. Brechet et al.
conclude in favour of a step by step penetration process, quasi at atomic scale and
give an estimate of crack propagation velocity, of the order of a few micrometers per second, in good
agreement with experimental observations. The author stresses the importance of defects in the GBs
(precipitates...) at different scales which can explain discrepancies between the experimental crack
velocity and the one calculated using a continuum elastic model. The paper presented by W. Ludwig et al.
concentrates on two synchrotron radiation imaging techniques (X-ray micro-radiography and
micro-fluorescence) available at ESRF (Grenoble) in order to characterize the penetration kinetics
of embrittling species in the crack. Beside the Al/Ga couple, it is shown that those techniques are
suitable to follow intergranular propagation of cracks in the Ni/Bi system, even when the embrittling
layer becomes nanometric.
Structuration of liquid Al in contact with Cu(100) and Cu(111) surfaces of a bulk Cu crystal
was studied by P. Geysermans et al.
using Molecular Dynamics simulations. The authors showed an
epitaxial relationship, whatever the density of the solid and contacting liquid phases. This feature
is of importance to understand the structure of the wetting layers in GBs. J. Bernardini et al.
emphasize the role of macroscopic defects like facets in GBs on "LM" penetration and propose a model
allowing to describe the real shape of the cracks they observed with Ni annealed in contact with pure
lead, inconsistent with Mullins grooving theory. Cu/Bi, Cu/LBE (Lead-Bismuth Eutectic) and Ni/LBE
systems were studied by K. Wolski et al.
using AES semi-quantitative analysis of fractured specimens
under vacuum within an Auger Electron spectrometer, enlightening the embrittling role of the
nanometer-thick wetting layer in the crack tip area.
In summary, during the past five years close cooperation between researchers
within the material
research program of GEDEON co-sponsored by the 5th EU Framework program allowed to make significant
progress. Research dedicated to radiation damage in structural materials for ADS and spallation targets
was very active and a number of impressive results were obtained. The low temperature He embrittlement
is certainly a matter of concern. The controlling mechanisms and their dependence with temperature, He
and dpa production rate has to be clarified. Likewise, compatibility studies of structural materials
with lead and lead alloys, particularly LBE, gave rise to a number of remarkable results. One can
mention here specifically the progress made with oxygen sensors. Significant is also the renewal of
interest for LME during the past five years. Promising results on LME and intergranular penetration
are already obtained and should be obtained in this attractive field of research. Of course, a great
deal of work remains to be done in order to validate the concept of Pb-Bi spallation target for ADS.
Especially, the effect of spallation elements production in structural materials and in LBE has to be
assessed as well as LMC and LME under irradiation.
© EDP Sciences 2002