FST41-872

Fusion Science and Technology
ISSN: 1536-1055 (Print) 1943-7641 (Online) Journal homepage: http://www.tandfonline.com/loi/ufst20
Hydrogen Isotopes Permeability in Eurofer 97
Martensitic Steel
A. Aiello, I. Ricapito, G. Benamati & R. Valentini
To cite this article: A. Aiello, I. Ricapito, G. Benamati & R. Valentini (2002) Hydrogen Isotopes
Permeability in Eurofer 97 Martensitic Steel, Fusion Science and Technology, 41:3P2, 872-876,
DOI: 10.13182/FST41-872
To link to this article: http://dx.doi.org/10.13182/FST41-872
Published online: 10 Aug 2017.
Submit your article to this journal
Article views: 4
View related articles
Citing articles: 3 View citing articles
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=ufst20
Download by: [University of Florida]
Date: 28 October 2017, At: 07:47
HYDROGEN ISOTOPES PERMEABILITY
IN EUROFER 97 MARTENSITIC STEEL
R. Valentini
University of Pisa
Dept. of mechanical engineering
Pisa, Italy
Downloaded by [University of Florida] at 07:47 28 October 2017
A. Aiello, I. Ricapito, G. Benamati
ENEA Fusion Division - CR Brasimone
40032 Camugnano Italy
tel.+39.0534.8011
ABSTRACT
In considering structural materials for fusion
reactors a detailed understanding of the transport
parameters and solubility of hydrogen and its isotopes is
an important issue which deal with safety and blanket
performance aspects.
The experimental activities were focused on the
determination
of
hydrogen/deuterium
transport
parameters through Eurofer 97 in the temperature range
423+723K using a time dependant permeation technique
The hydrogen permeation and diffusivity at room
temperature and density of trapping sites were also
evaluated using Devanathan 's technique.
Hydrogen / deuterium permeation experiments on
Eurofer 97 showed a non-negligible decrease in
permeability with respect to other fusion oriented
martensitic steels, even if it remains about one order of
magnitude higher compared with that of austenitic AISI
316L steel.
I
INTRODUCTION
The martensitic steel Eurofer 97 represents the latest
generation of martensitic steels for fusion applications
and is presently the candidate material for the first wall
and structural components of the demonstration fusion
reactor DEMO. It belongs to the Reduced Activation
Martensitic (RAM) steels of the 7-10% Cr family.
A precise determination of the permeability,
diffusivity and solubility of hydrogen isotopes through
such materials is of crucial importance to evaluate tritium
behaviour and inventory in the blanket modules. These
martensitic steels have approximately one order of
magnitude higher permeability than austenitic steels and a
872
complete characterisation of hydrogen isotope transport
parameters is necessary to evaluate the tritium behaviour
in the DEMO blanket.
Moreover, the hydrogen isotope transport in martensitic
steels at temperatures lower than 200-250 °C is generally
strongly affected by trapping phenomena. An
understanding of trapping mechanisms is necessary to
understand the real operative conditions at which
hydrogen embrittlement (HE) can take place.
In order to gather data on all of these phenomena,
permeation experiments have been performed on Eurofer
97 in the temperature range 413+123 K with the device
"PERI" (gas phase technique) and at room temperature
with Devanathan's electrochemical cell. The obtained
results have been compared with those of other candidate
materials proposed for the first wall and blanket of a
fusion reactor, such as F82H steel, which has been
analysed in the past'1'2]. The composition of Eurofer 97
together with that of F82H is reported in tab. 1.
Eurofer
97
F82H
C
Si
Mn
P
0.11
0.04
0.48
<0.005
0.003
0.09
0.11
0.16
0.002
0.002
Cr"
Eurofer
97
F82H
~
Eurofer
97
F82H
Mo"
NT
S_
V
W
835
<0.001
0U021
CL20"
L08"
7.66
<0.01
0.02
0.16
2.00
Al
Nb
Cu
Co"
Ti ~
0.002
0.006
0.006
0.009
0.002
<0.01
<0.01
0.01
0.001
<0.01
Tab.l: Composition of Eurofer 97 compared with that of
F82H
FUSION SCIENCE AND TECHNOLOGY
VOL. 41
MAY 2002 872
Aiello et al.
II
EXPERIMENTAL SET UP
II.I
Hydrogen
isotopes
HYDROGEN ISOTOPES PERMEABILITY IN EUROFER 97 MARTENSITIC STEEL
permeation
at
high
Downloaded by [University of Florida] at 07:47 28 October 2017
temperature
Eurofer 97, used for the hydrogen/deuterium
permeation experiment underwent the following heat
treatment by the manufacturer in order to produce a fully
ferrite free martensitic phase: normalisation at 975°C for
15 minutes, cooling to RT in air, tempering at 740°C for
45 minutes, cooling to RT in air. The samples used in the
permeation apparatus consisted of discs of 48 mm in
diameter and about 1 mm in thickness. They were
machined from a sheet supplied by Boehler Edelstahl
GmbH (Germany).
The
permeation device "PERI" is described
elsewhere'11 and a schematic view is reported in fig. 1.
Fig. 1:
Schematics of the permeation apparatus PERI for the
gas-phase measurements.
High pressure deuterium gas, with a nominal purity
of 99.7 %, is taken from a cylinder, although hydrogen
gas has also been used. The hydrogen had an impurity
content less than 1 vppm of H 2 0 and 0 2 , and was
produced by a hydrogen generator connected to a
hydrogen purifier. The hydrogen/deuterium gas is
admitted to the sample via a control valve which, together
with a pressure transducer and a Baratron, enables the
pressure to be set at any value between 100-1.5-105 Pa.
The sample is heated by a resistance furnace. Gas
permeates through the sample causing a rise in pressure in
the outlet volume. For each run, an analysis of the inlet
and permeated gas is carried out by a quadrupole mass
spectrometer in order to check for possible contaminants.
The National Instruments system (Labview) is used for
data acquisition.
II.II
Hydrogen permeation at room temperature
Hydrogen
permeation
experiments
FUSION SCIENCE AND TECHNOLOGY
VOL. 41
at
temperature have been performed using the Devanathan's
technique. A 1 mm thick, square shaped specimen of
Eurofer 97 as received was inserted in a double sided
electrolytic cell, with the specimen separating the two
sides. The anodic cell was filled with a solution of NaOH
0.1 M, and the cathodic one with a solution of H 2 S0 4 0.1
M and thiourea as a hydrogen recombination inhibitor.
The two cells were aerated using a flux of N2 to reduce
the background noise. The cathodic cell contains the
working electrode, represented by one side of the
membrane, a platinum counter electrode and a calomel
standard potential electrode. The exit side of the
membrane was polarised at a potential of 200 mV, high
enough to reduce protons to hydrogen atoms before their
recombination. Hydrogen was produced using a cathodic
density of current of 10 mA/cm2, low enough to prevent
hydrogen embrittlement phenomena.
Ill
RESULTS AND DISCUSSION
III.I
High temperature H2/D2 permeation tests
The experiments were performed in a temperature
range of 473+723 K using a hydrogen or deuterium
upstream pressure of about 75000 Pa. The results of
permeability <P and effective diffusivity D for Eurofer 97
in comparison with F82H steel'2' are presented as
Arrhenius plots in figures 2 and 3, respectively.
looon- (1/K)
Fig. 2: Permeability of hydrogen/deuterium through Eurofer 97
compared to F82H
low
MAY 2002
873
Aiello et al.
HYDROGEN ISOTOPES PERMEABILITY IN EUROFER 97 MARTENSITIC STEEL
The correlation factors obtained by the two fitting
procedures and experimental curves were very high, and
the parameters obtained with the two routines were in
good agreement.
Permeability, lattice diffusivity and Sieverts constant
Ks i for deuterium in Eurofer 97 were calculated as:
IHt/U
e~~*^(m2
D, = 1.5-10'7
Ks
l
= 1.02-KT'e
Downloaded by [University of Florida] at 07:47 28 October 2017
O = 1.53 -10~ 8 e"
Fig. 3 Diffusivity of hydrogen/deuterium in Eurofer 97
compared to F82H
The ratio <P(m/<&(D2) is about 1.6 in this temperature
range, a little higher than that expected by the classical
theoretical value of diffusion coefficient ratio
-Jl = (mD / mH)V2.
Variation of the steady state flux
with deuterium pressure was determined at different
temperatures. In the pressure range 103-105 Pa a near
half-power pressure dependence of the permeation fluxes
is observed. As a consequence, the whole transport
mechanism is controlled by the diffusion through the
bulk, at least under these operative conditions.
The experimental data of permeated flux were fitted
using linear and general fitting techniques in order to give
precise values for diffusivity and permeability. In the
case of linear fitting, permeability is calculated by
linearisation of the pressure-time curve referred to the
low pressure side in steady state flux regime, and
diffusivity is determined by the time lag.
In the case of general fitting the parameters obtained
from the linear fitting are adjusted performing a least
square best fit of the solution of the general diffusion
equation (eq.l) in the whole experimental data set. In eq.l
D, p and d represent respectively permeability,
diffusivity, upstream pressure and specimen thickness.
Q(t)=
f
„
7
71
874
D
d
6D
Zr-T-exp
t l n2
-S1)
23810 f
RT
38280
RT
m'1
mol
mol
m
•s
Pa
(2)
2
Pa " 2
Significant hydrogen trapping was observed below
573K. In this region, the diffusion coefficient drops
sharply below the values obtained by a direct
extrapolation at lower temperatures. It is worth noting
that in F82H the trapping phenomena were evident only
at temperature below 523 K.
In modelling diffusivity data to extract the density
and mean energy of the trapping sites, it has been
assumed that the hydrogen atoms in the material can
move through ordinary lattice sites, characterised by a
lattice diffusion energy Ed, or through trapping sites, in
which a potential well deep enough produces stable
bound of hydrogen atoms to specific trapping centres at a
certain temperature. The mean interaction energy Et can
be evaluated. Using Ni and N t to represent the density of
lattice and trapping sites, the effective diffusivity D eff may
be related to the lattice one Di as131:
D,r
=
,
AN
\ + —
>
Ml.
e RT
(3)
It is evident from eq.3 that the trapping effects
become negligible at high temperatures where Defr
approaches Di.
Using the non linear fitting of diffusivity data (fig.
4) to extract the values of E, and N t it is assumed that the
concentration of lattice sites is 5.2-1029 m"3 assuming six
tetrahedral lattice sites per host atom (bcc iron) [1] .
(D
-D—T-t
d
)
FUSION SCIENCE AND TECHNOLOGY
VOL. 41
MAY 2002
Aiello et al.
HYDROGEN ISOTOPES PERMEABILITY IN EUROFER 97 MARTENSITIC STEEL
Taking into account these data and considering the
grain dimension of Eurofer 97 (ASTM 10) and its
composition, the main trapping sites are probably the
interfaces between martensitic laths together with carbide
precipitates and dislocations.
III.II
Diffusivity and density of trapping sites were
measured by experiments of hydrogen permeation at
room temperature using Devanathan's technique. The
relationship between the hydrogen permeated flux J and
the anodic current in the electrochemical cell is:
Downloaded by [University of Florida] at 07:47 28 October 2017
Fig. 4: Arrhenius plot of deuterium diffusivity in
Eurofer 97
The obtained values, compared with those of F82H
and Manet II[2], are presented in tab. 2.
3
N, (m" )
E, (J/mol)
Ed (J/mol)
MANETII
1.6 1025
48500
13210
F82H
1.6-1023
55938
Eurofer 97
1.04-1024
57898
13950
14474
Tab.2: trapping parameters in MANET II, F82H and
Eurofer 97 martensitic steels
The most interesting result is that the density of
trapping sites in Eurofer 97 is about one order of
magnitude higher than in F82H. This circumstance was
confirmed by experiments of permeation at low
temperature, as will be seen in the next section.
Some published data for the trapping energy in
various trapping sites in ferritic steels taken from
literature'3"101 are reported in tab. 3.
Type of trap
Single vacancy
Atomic traps
Substitutional Cr atoms
Substitutional Mo atoms
Substitutional V atoms
Substitutional Mn atoms
Interstitial C atoms
Substitutional Ni atoms
Interstitial N atoms
Grain boundaries
Second phase particles
A1N, ecarbide
MnS
Dislocation
Trapping
energy (kJ/mol)
50, 46.6, 78.3
Ref.
26.1
27
57
10.6
3.3
-11.6
12.5
59,32
5
5
5
5
6
5
6
7
65
72
20 to 30
8
9
10
J{t) =
m - L
(4)
where F is the Faraday constant and L the thickness of the
specimen. When the measured current reaches a steady
state value corresponding to the hydrogen saturation of
the specimen, the diffusivity can be evaluated using the
time lag or breakthrough methods.
In both cases the current is integrated versus time,
determining the total amount of hydrogen permeated
through the sample during the experiment. It can be seen
that the cumulative amount of permeated gas
asymptotically approaches a straight line (fig. 5).
3,4
Tab. 3: Trapping energies from literature
875 FUSION SCIENCE AND TECHNOLOGY
Hydrogen permeation at room temperature
VOL. 41 MAY 2002
Fig. 5: evolution of density current integrated over the time in
Devanathan's experiment
HYDROGEN ISOTOPES PERMEABILITY IN EUROFER 97 MARTENSITIC STEEL
Aiello et al.
This line intercepts the time axis at a value defined
as:
tL =
L2
using the time lag method, or
6-Deff
Downloaded by [University of Florida] at 07:47 28 October 2017
tB =
L2
15.3 -Deff
using the breakthrough method.
In the first case the time is measured starting from
the moment at which electric current is applied to the
inlet side of the sample. In the second case the time is
measured starting from the moment in which the
hydrogen from the outlet side of the specimen appears.
Independently of the method used to fit the experimental
data (breakthrough or time lag), the effective diffusivity
measured is in the range 6-10"12-s-4.9-10"um2/s, as can be
seen in tab. 4.
respect to F82H martensitic steel. For instance, at 350°C
the permeability of deuterium in Eurofer 97 is about 2.5
times less than in F82H. Also the diffusivity is lower in
Eurofer 97 and this difference is particularly marked at
250°C where the trapping phenomena are more evident in
Eurofer 97 than in F82H.
In addition, also if the interaction energy for the
trapping sites is practically the same for the two steels,
the density of traps in Eurofer 97 is about one order of
magnitude higher than in F82H. This circumstance was
also confirmed by permeation tests carried out at room
temperature using Devanathan's technique.
REFERENCES
[1]
[2]
Parameter
Steady state current
(A/cm2)
Dbreakthrough (m'/s)
Dtime tag (m2/s)
First
permeation
14
Second
permeation
10.4
6.00e-12
2.40e-11
3.30e-ll
4.90e-11
[3]
[4]
[5]
[6]
Tab.4: Diffusion coefficients determined using
Devanathan's technique
The second permeation run, carried out just after a
degassing of the specimen at the end of the first
permeation test, gives the highest values of diffusivity
because the irreversible traps had already been filled
during the first run.
A concentration of trapped hydrogen of 0.48 wppm
was measured, which corresponds to a density of
irreversible traps of 2.2-1024 m"3. This value is in quite
good agreement with the density of traps calculated from
fitting the diffusivity data, measured at high temperature,
with eq. 3.
IV
[7]
[8]
[9]
[10]
E. Serra, A. Perujo, G. Benamati, Journal of
Nuclear Materials, 245, 108-114 (1997)
E. Serra, G. Benamati, , Material Science and
Technology, 14, 573-578 (1998)
S.M. Myers, S.T. Picraux, R.E. Stroltz, J. Appl.
Phys, 50, 5710 (1979)
K.B. Kim, S. Pyun, Arch. Eisenh. 53, 397 (1982)
A.I. Shirley, C.K. Hall, Scr. Metall., 17, 1003
(1983)
J.J. Au, H.K. Birnbaum, Acta Metall., 26, 1105
(1978)
J.P. Hirth, Metall. Trans., 11 A, 861 (1980)
H.H. Podgurski, R.A. Oriani, Metall. Trans., 3,
2055 (1972)
J. Chene, J.O. Garcia, C.P. De Oliveira et oth., J.
Microsc. Spectrosc. Electron., 4, 37 (1979)
C.A. Wert, in "Topics in Applied Physics,
Hydrogen in metals", eds. G. Alefeld , 305
(1978)
CONCLUSIONS
Hydrogen / deuterium permeation experiments on
Eurofer 97, carried out in the temperature range 473-723
K, showed a non-negligible decrease in permeability with
876
FUSION SCIENCE AND TECHNOLOGY
VOL. 41
MAY 2002