AECL-4664
ATOMIC ENERGY &£g~k L'ENERGIE ATOMIQUEOF CANADA LIMITED ^ £ 9 DU CANADA UMITEE
THE HYDRIDING OF ZR 3A1-BASED ALLOYS
by
D.J. CAMERON and D. FAULKNER
Whiteshell Nuclear Research Establishment
Pinawa, Manitoba
August 1974
ATOMIC ENERGY OF CANADA LIMITED
THE HYDRIDING OF ZraA1- BASED ALLOYS
by
D.J. Cameron* and D. Faulknert
* Materials Development Branch
t Materials Science Branch
Whiteshell Nuclear Research Establishment
Pinawa, Manitoba, ROE 1LO
Manuscript Prepared November 1973
Printed August 1974
AECL-4664
Hydruration d'alliages It base de Zr3Al
par
D.J. Cameron* et D. Faulknert
*Subdivlsion de développement des matériauxtSubdivision de science des matériaux
Manuscrit rédigé en novembre 1973
Imprimé en août 1974
Résumé
Des échantillons d'alliage contenant 8.3% Al ontfait l'objet d'une hydruration jusqu'à 1300 microgrammesd'hydrogène par gramme d'alliage et des échantillonsd'alliage contenant 9.6% Al ont fait l'objet d'une hydru-ration jusqu'à 1040 microgrammes par gramme à 923K. L'effetdes traitements thermiques postérieurs, à 973,773 et 623 K,sur les relations de phase dans les alliages ayant subi1'hydruration a été étudié par microscopie optique etélectronique de transmission et par diffraction auxrayons X. Aucun hydrure n'a été détecté dans les alliages,bien qu'une infime quantité de précipitations non identifiéesait été observée dans la phase Z,A1 de quelques alliagescontenant 9.6% Al.
L'addition d'hydrogène, même en quantitésrelativement faibles, provoque la décomposition de laphase Z_A1 en ct-Zr et Zr.Al dans l'alliage contenant8.3% Al. Cet effet n'est constaté qu'aux teneurs enhydrogène bien plus importantes dans l'alliage contenant9.6% Al. On estime que dans cet alliage la phase pré-existante Zr2Al agit comme puits d'hydrogène.
Un diagramme schématique de phase ternaire partielledu système Zr-Al-H explique ces observations.
L'Energie Atomique du Canada, LimitéeEtablissement de Recherches nucléaires de Whiteshell
Pinawa, ManitobaROE 1L0
Août 1974
AECL-4664
THE HYDRIDING OF Zr3A1 BASED ALLOYS
by
D.J. CAMERON* AND D. FAULKNERt
ABSTRACT
Samples of 8.3% Al alloy were hydrided to 1300 micrograms
of hydrogen per gram of alloy and samples of 9.6% Al alloy were hydrided
to 1040 micrograms per gram at 923 K. The effect of subsequent
heat treatment at 973, 773 and 623 K on the phase relationships in the
hydrided alloys was studied by optical and transmission electron microscopy,
and X-ray diffraction. No hydrides were found in any of the alloys,
although some minute unidentified precipitation was observed in the
phase of some 9.6% Al alloys.
The addition of hydrogen even in relatively small amounts
causes the decomposition of the Zr3Al phase to Ot-Zr and Zr2Al in the 8.3%
Al alloy. This effect is observed only at much higher hydrogen contents
in the 9.6% Al alloy, in which it is believed that the pre-existing
Zr2Al phase acts as a hydrogen sink.
A schematic partial ternary phase diagram of the Zr-Al-H
system is suggested to explain these observations.
* Materials Development Brancht Material Science Branch
Atomic Energy of Canada Limited
Whiteshell Nuclear Research Establishment
Pinawa, Manitoba, ROE 1L0
Manuscript Prepared November 1973
Printed August 1974
AECL-4664
1.
2.
3.
INTRODUCTION
EXPERIMENTAL
RESULTS
CONTENTS
Page
1
3.1 OPTICAL MICROSCOPY 4
3.2 ELECTRON MICROSCOPY 6
4.
5.
6.
7.
DISCUSSION
CONCLUSIONS
ACKNOWLEDGEMENTS
REFERENCES
" FIGURES
15
15
16
17
- 1 -
1. INTRODUCTION
Hydriding behaviour is an important feature of any zirconium
alloy which is to be used in a nuclear reactor with a hydrogenous coolant.
Hydriding of conventional zirconium alloys results in the formation of
zirconium hydride platelets, which if present in sufficient quantity can
cause embrittlenient of the alloys.
Alloys based on the intermetallic compound Zr3Al are being
developed as a potential pressure tube material'1"3' for the various
CANDU* reactor systems. The theoretical single-phase Zr3Al alloy contains
8.97% Al, but the optimum alloy composition will probably lie in the
range 8.0 to 9.5% Al. Alloys containing less than ̂ 8.5% Al can have a
Zr3Al matrix with a dispersion of a-Zr solid solution, while those
containing more aluminum have dispersions of both a-Zr solid solution
and Zr2Al in the Zr3Al matrix^1). It has not been possible to eliminate
a-Zr from any of the alloys, and the conjunction of the a-Zr and 2r2Al phases
in the alloys of higher aluminum content suggests that the phase relationships
might best be described in terms of a pseudoternary system, Zr-Al-ImpuritiesC ),
These observations are supported by Schulson^ '.
This investigation was conceived as a study of the general
features of the hydriding of these rather complex alloys, particularly
in terms of the limits of hydrogen solubility, and the type, morphology,
and distribution of any hydrides which might be formed.
Previous studies of hydrides formed in alloys from the
zirconium-aluminum system have been concerned with the hydrides formed
after gross hydriding. It has been reported that heavy hydriding of
Zr3Al results in the formation of zirconium hydride and a hydride based
on Zr2Ali and that Zr2Al forms a hydride with a hydrogen to metal ratio
of 0.80(6> .
Canada Deuterium Uranium
- 2 -
In zirconium alloy reactor components the effect of hydrogen
in the hundreds of yg/g*range of concentration can be important, and the
previous work gives no indication of the effect these levels of hydrogen
will have on ZraAl-based alloys.
In this study nominal hydrogen concentrations of 200, 500 and
1000 yg/g were introduced into samples of a two-phase alloy containing
8.3% Al, and a three-phase alloy with 9.6% Al. The samples were studied
by optical and transmission electron microscopy, and electron and X-ray
diffraction, following hydriding and homogenization at 923 and 973 K
respectively, and after heat treatments at lover temperatures.
2. EXPERIMENTAL
The apparatus used to hydride the samples has been described
previously*6'. The weighed samples were exposed at 923 K to a known
volume of hydrogen at a predetermined pressure. The pressure was monitored
during hydriding, and when it had dropped to a value corresponding to the
required uptake of hydrogen, hydriding was stopped. After hydriding, the
specimens, in pairs with the same aluminum and nominal hydrogen content,
were sealed in Vycor capsules and homogenized at 973 K for 0.60 Ms (7 days).
Subsequent lower temperature heat treatments were also performed in
sealed capsules.
Metallograp1 xc preparation was conventional for these alloys/7^
except that an anodizing voltage of 30 V was used to reveal the phases
present. This voltage was used because of information that the oxide
film above any zirconium hydride would appear gold under these conditions,
but: would break down if a higher potential was used^8).
Specimens for electron microscopy were prepared as thin 3 mm
diameter discs. An electrolytic jetting technique was used to produce
*Pg/g - micrograms of hydrogen per gram of alloy
- 3 -
a dished prof i le , and final perforation was achieved using a conventional
electropolishing c e l l . A solution of 10% perchloric acid in ethanol was
used in both cases, the final thinning being performed at 223 K. The
specimens were examined in a Hitachi HU-200E electron microscope operating
at 200 kV.
3 . RESULTS
A piece of one sample from each pair with different aluminum
and hydrogen contents was analysed for hydrogen. The results are shown
in Table 1.
TABLE 1
THE HYDROGEN CONTENTS OF THE SAMPLES
NOMINAL HYDROGENCONTENT, yg/g
200
500
1000
ACTUAL HYDROGEN CONTENT, yg/g8.3% Al 9.6% Al
190
650
1300
230
450
1040
A wedge-shaped segment of the disc samples was used for the
analysis, and although there was some inhomogeneity in some of the
samples, this was generally radial, so the results should represent a
reasonably accurate average hydrogen content. I t was assumed that the
other sample of each pair, having been hydrided and homogenized under
identical conditions, would have a similar hydrogen content.
- 4 -
3.1 • OPTICAL MICROSCOPY
Before hydriding, the samples had two distinct microstructures.
The 8.3% Al specimens had a Zr3Al matrix containing a dispersion of
a-Zr solid solution particles with an average diameter of ̂ 3 to 5 ym
occupying an estimated volume fraction of ̂ 15%. The Zr3Al matrix of the
9.6% Al alloy contained ̂ 25 to 30% by volume of Zr2Al with a large
elongated morphology. About 10% of the volume of this alloy was occupied
by a finer dispersion of a-Zr solid solution. The original structures
can be deduced from those of the alloys with the lowest hydrogen contents.
Figures 1 to 3 show micrographs of samples of 8.3% Al containing
190, 650, and 1300 yg/g hydrogen. In addition to the a-Zr solid solution,
irregular shaped darker areas can be seen dispersed throughout the ZrsAl
matrix. Some of these areas are marked for easier identification.
Although it is not resolvable in this micrograph these areas had a very
fine duplex structure. The anodizing characteristics of these phases and
comparison with other samples showed that the duplex structure was
composed of a-Zr and Zr2Al phases. The sample containing 650 yg/g
hydrogen was not homogeneous. The perimeter was composed of the original
a-Zr solid solution dispersion surrounded by the fine duplex structure.
The centre still had a Zr3Al matrix, but there was substantial decomposition
to the duplex structure. Figure 2 is a micrograph of the edge of this
region, and shows light patches of Zr3Al surrounded by the fine two-phase
a-Zr + Zr2Al structure. The original a-Zr solid solution can also be
seen, but in this region it is rather finer than in Figure 1. The
addition of 1300 yg/g hydrogen causes the complete elimination of the
Zr3Al phase, Figure 3, and the sample has a duplex matrix containing the
original a-Zr solid solution dispersion.
The hydriding of the 9.6% Al alloy is shown in Figures 4 to
6. At a hydrogen content of 230 yg/g, Figure 4, the Zr2Al phase displays
two anodizing effects. In some areas it was brown, while in others it
was a very pale yellow. Figure 4 shows the edge of these regions, where
- 5 -
both effects are observed. The original Ot-Zr solid solution is the
darkest phase in this micrograph. No decomposition of the matrix was
observed. When 450 yg/g hydrogen had been added the Zr2Al phase was a
uniform dark brown and is the dark phase in Figure 5. The a-Zr now
appears lighter than the matrix. Again no decomposition of the Zr3Al
occurred. At 1040 yg/g hydrogen, decomposition was apparent. This can
be seen in Figure 6 as the ragged-edged dark areas and spots. The duplex
structure of the decomposition product is shown at higher magnification
in Figure 7. The anodizing characteristics of the two phases were
identical to those of the ct-Zr and Zr2Al phases.
Samples of each alloy of each hydrogen content were heat-
treated at 773 K and 623 K for 0.60 Ms (7 days) and at 598 K for 0.26 Ms
(3 days). Optical microscopy revealed no detectable difference in the
structures.
To determine whether any hydrides not visible optically had
been formed, the sample of 8.3% Al containing 1300 yg/g hydrogen, and that
of 9.6% Al with 1040 yg/g hydrogen were subjected to X-ray diffraction
after the 623 K heat treatment. The results are shown in Table 2. The
8.3% Al alloy consists mainly of Zr2Al with the 12U crystal modification,
and a-Zr solid solution. Two very weak diffractions corresponding to the
strongest lines from the Zr3Al crystal structure were also present,
although this phase was not observed optically. The 9.6% Al alloy
contains substantial quantities of all three phases. The d-values of the
a-Zr solid solution are reduced somewhat by the dissolved Al, but those
of the other two phases are very close to the published values. No
suggestion of zirconium hydride or any unknown phase was found. That the
Zr2Al phase is tetragonal in both alloys is perhaps significant, as it
is usually, but not always, found that in unhydrided alloys this phase
is hexagonal. It is possible that interstitial impurities, such as
hydrogen, stabilize the tetragonal structure relative to the hexagonal
form, in a manner analogous to that observed in the ZrsAl3 phase( '.
If this is the case, the first stage in hydriding of alloys containing
ZT2A1 could be the transformation from the hexagonal to the tetragonal
- 6 -
form. This would account for the two distinct etching characteristics
of the Zr2Al phase in the 9.6% Al alloy containing 230 yg/g hydrogen.
3.2 ELECTRON MICROSCOPY
Samples with each aluminum and hydrogen content were examined
in the electron microscope both prior and subsequent to the 598 and 62 3 K
heat treatments. No detectable differences were observed after heat
treatment, and the subsequent observations refer to either material.
All major phases present were positively identified by electron diffraction
using the d-values in Table 2. A number of general observations were
made:
1. In all specimens a-Zr was present, and this phase was thinned
preferentially during preparations.
2. In specimens containing Zr3Al and Zr2Al as well as a-Zr,
the Zr2Al is thinned more rapidly than the Zr3Al.
3. The Zr3Al phase could usually be identified by tilting the
specimen in the microscope, since this phase is characterized by the
presence of twins, Figure 8, and stacking faults, Figure 9, under suitable
diffraction conditions.
4. The ZraAl grain size is significantly smaller than that of
the Zr3Al.
Figures8 and 9 show the structure of the 8.3% Al alloy
containing 190 yg/g hydrogen. This material consists essentially of a
Zr3Al matrix with some small (̂ 3 ym) a-Zr grains. No Zr2Al was detected
in these specimens, but statistically this is not surprising in view of
the small amount of decomposition revealed optically, and the relatively
small areas studied in the electron microscope. The average Zr3Al grain
size was approximately 7 ym.
- 7 -
TABLE 2
X-RAY DIFFRACTION DATA FROM TWO HYDRIDED Zr3Al-BASED ALLOYS
8.1300
I
-
12
-
36
30
204
38
20
133
27
16
37
32
17
-
7
13
20
30
42
-
44
-
26
16
12
3% Al,Mg/g H
d, mxlO"8
-
3.43-
2.786
2.746
2.676
2.566
2.519
2.451
2.390
2.181
2.168
2.144
1.888-
1.716
1.703
1. >16
1.594
1.572-
1.457-
1.365
1.345
1.318
9.1040
I
15
-
25
7
16
104
-
216
14
15
88
-
18
13
12
-
10
8
13
22
72
27
16
-
-
93
6% AlUg/g H
d, mxlO"8
4.365
-
3.085
2.787
2.744
2.677
-
2.521
2.456
2.389
2.183-
2.1401.952
1.782
-
1.700
1.614
1.591
1.569
1.544
1.455
1.393-
-
1.318
ASTH CARD
Zr2Al (12U)CARD #14-436
I/Io
-
40
-
-
40
100
-
-
-
40
-
50
50
-
10
30
30
40
60
-
-
50
-
-
-
d, nuclO"6
-
3.41
-
-
2.746
2.675-
-
-
2.392-
2.167
2.144
-
1.713
1.702
1.615
1.592
1.571-
-
1.393
-
-
-
INDEX DIFFRACTION DATA
CARD
I/Io
—
-
-
33
-
-
32
-
100
-
-
-
17
-
-
-
17
-
-
-
18
-
18
12
-
a-Zr#5-0665
d, mxl0~e
-
-
-
2.798
-
-
2.573
-
2.459
-
-
-
1.894
-
-
-
1.616
-
-
-
1.463
-
1.368
1.350
-
Zr3AlCard #7-115
I/Io
60
-
60
-
-
-
-
100
-
-
80
-
40
40
-
-
-
-
-
80
40
-
-
-
80
d, mxl0~B
4.38
-
3.09
-
-
-
-
2.52
-
-
2.18
-
1.95
1.78
-
-
-
-
-
1.54
1.46
-
-
-
1.32
- 8 -
In the 8.3% Al samples containing 650 yg/g hydrogen, both
Zr3Al and Zr2Al were identified. The Zr2Al grains were interspersed
with small (< 1 ym) a-Zr grains, as shown in Figure 10, and this structure
presumably corresponds to the duplex structure observed optically. A
hydrogen content of 1300 yg/g causes complete transformation to the
duplex structure of Zr2Al + a-Zr shown in Figure 11. The Zr2Al grain
size is ̂ 1 ym. No structural changes were observed in the 1300 yg/g
specimen after a 0.60 Ms (7 days) heat treatment at 623 K.
In contrast to the 8.3% Al alloy, specimens of material con-
taining 9.6% Al did not show significant structural changes with the
addition of hydrogen. The ZraAl, Zr2Al and a-Zr phases were identified
in all cases. Although the Zr2Al appears optically as elongated grains,
the upper right part of Figure 12 shows that these elongated grains in
fact consist of numerous small interconnected grains. The grain size
of Zr2Al is *v 1.5 ym while that of the ZraAl is ̂ 7 ym. No areas of
duplex Zr2Al + a-Zr were detected in the 1040 yg/g hydrogen sample, but
again this may have been due to the relatively small areas examined.
Perhaps the most significant observation in the 9.6% Al alloy
is the presence of strain-field contrast effects in the Zr2Al phase of
alloys containing 450 and 1040 yg/g hydrogen (Figure 13). This type of
contrast is typical of coherent precipitates, and may result from the
rejection of hydrogen to form hydrides as the temperature is lowered.
No significant coarsening of these precipitates was observed following
the heat treatment at 598 K.
4. DISCUSSION
To explain the hydriding effects it is necessary to consider
the possible phase relationships in the ternary system Zr-Al-H. Then
- 9 -
we can more correctly refer to the hydrogen additions causing the stabi-
lization of a three-phase field, rather than decomposing the Zr3Al. The
starting point in constructing a schematic ternary system is to consider
the possible effects of the hydrogen on the peritectoid equilibria:
e-Zr + Zr2Al * Zr3Al
and B-Zr + Zr3Al % a-Zr
in the Zr-Al binary system.
In the binary system, Figure 14, these equilibria occur at
invariant temperatures. However, the addition of a third component gives
another degree of freedom to the system, and the three phases can then
co-exist over a range of temperature and composition in the form of a
triangular tube space in the ternary diagram.
A possible schematic development of the ternary system is
shown by means of isothermal sections of decreasing temperature in
Figure 15 (a) to (e). In the construction of the diagram it is assumed
that Zr3Al and Zr^Al have negligible variations in Zr:Al ratios, but
do dissolve hydrogen. These compound solutions are therefore represented
by lines in isothermal sections, and planes in the three-dimensional
diagram. For convenience, the phases ot-Zr, 3-Zr, Zr3Al, and Zr2Al are
often designated a, 3, 3 and 2 respectively in some of the diagrams and
text. Also, it is implicitly understood that when these phases are
referred to in the ternary system, they actually represent solid solutions
of hydrogen in the appropriate phase.
Figure 15 (a) is an isotherm between the temperatures of the
two peritectoids in the binary system. The broken line IQ is not part
of the isothermal section, but represents the path of the 3 phase
composition of the three-phase field as the temperature drops. The
peritectoid reaction:
0-Zr + Zr2Al t Zr3Al
- 10 -
starts as a l ine on the binary Zr-Al phase, and as i t fa l l s into the
ternary diagram i t opens into a triangular three-phase f ie ld , 3 + 3 + 2 .
As the triangle drops, i t sweeps out behind i t two, two-phase regions,
3 + 3 and 3 + 2. The other side of the triangle i s presumably bounded
by the 3 + 2 phase f ield originating above the binary peritectoid temper-
ature.
At a temperature just below the second peritectoid reaction
in the binary, the ternary isotherm can be schematically represented by
Figure 15 (b). The f irst peritectoid 8 + 3 + 2 , triangle has moved
further away from the Zr-Al axis of the diagram and behind i t the second
peritectoid triangle has formed, representing the a + 8 + 3 phase f i e ld .
This triangle has isolated the 8 + 3 phase field from the Zr-Al axis ,
and i s in turn sweeping out two more two-phase f i e lds , a + 3 and a + b ,
adjacent to the axis . The line FQ again i s not part of the isothermal
section, but traces the path of the 8 phase composition of the second
reaction as the temperature drops.
In Figure 15 (c ) , which i s a section below the a-8 transformation
temperature in pure zirconium, both these phase triangles have moved
further into the diagram, and closer together. Because the solubi l i ty
of Al in Zr decreases rapidly with fal l ing temperature, the corner of
the second triangle which represents the a-Zr composition has moved
closer to the Zr-H axis than has the corner representing the 8 composition,
which i s s t i l l following the l ine FQ. This has caused a rotation of the
side of the triangle joining the a and 8 composition corners, which in
the solid geometry of the three dimensional system causes a twist in
this face of the three-phase f i e ld , and also in the adjacent a + 8 phase
f ie ld. As a result of this rotation, the triangle now represents the
eutectoid equilibrium,
3 Z « + 3
rather than the peritectoid equilibrium,
8 + 3 t a.
- l i -
l t is now assumed that the two triangles merge with a common
side, QP in Figure 15 (d) , pinching off the 3 + 3 phase field. This forms
the quadrilateral OPRQ representing the four-phase quasi-peritectoid:
8 + 3 t a + 2
This reaction, by the dictates of the phase rule, is invariant in temper-
ature, and as the temperature drops further the four-phase field splits
along the other diagonal, OR, to form two more three-phase fields,
a + 3 + 2 and a + g + 2, with an intervening two-phase field, a + 2. This
is represented by Figure 15 (e). In the three-phase field, a + 3 + 2,
reaction is complete, while the second triangle represents the eutectoid
8 ~Z a + 2
and as the temperature drops further the 3 corner follows a path such
as that of the broken line in Figure 15 (e). I t is this isothermal section
which possibly can explain the hydriding observations, and to which
further reference will be made.
As a further aid to understanding the suggested ternary
equilibrium diagram, Figure 16 shows an exploded three-dimensional repre-
sentation of the phase fields. The important feature is the development
of the four-phase, quasi-peritectoid reaction shown in the foreground.
The peritectoid 8 + 2 Z 3 starts as the line IJK in the phase of the
Zr-Al binary. I t falls into the ternary diagram as an expanding triangle
until i t reaches QPR. Similarly, the peritectoid B + 3 Z <* starts at
FGH and falls to OQP. The twist in the face FGQO which denotes the
change to a eutectoid, 8 Z 3 + a, can clearly be seen. The two triangles
join at the four-phase invariant OPRQ, denoting the quasi-peritectoid
B+ 3 J a + 2, As the temperature decreases, the invariant Bplits along the
other diagonal, OR, forming the two new three-phase fields with the wedge
shaped a + 2 phase field between them.
To mentally reform the solid three dimensional diagram, the
a + 2 phase field is placed between the 8 + a + 2 and a + 3 + 2 three-
phase fields, and this construction is pushed back to contact the a,
- 12 -
a + 3, 3, and 3 + 2 phase fields which are collapsed along the Zr - Al
axis. The downward curving wedge of the 3 + 3 phase field is slid into
the slot between the two high temperature three-phase fields a + 3 + 3
and g + 3 + 2 . It is rather more difficult to see how the a + 3 field
fits in. The surface EGOST covers the a phase field; the surface VTOQ
covers one side of the 3 + ot + 2 field, and OQFR covers the twisted face
of the a + 3 + 3 phase field. Figure 17 shows the reconstructed schematic
partial diagram. The line IQV is important. To the left of this line
the exposed surfaces are covered by the 3-solid solution phase field,
while to the right, they are presumably covered by the 3 + 2 field.
Neither of these fields is included as they are not relevant to the
subsequent discussion.
Having developed a possible schematic ternary diagram, it
is now necessary to explain how it may be used to account for the experimental
observations. Figure 18 is an enlarged version of Figure 15 (e) , in
which an attempt has been made to draw the a-Zr + Zr3Al + Zr2Al phase
field in such a manner that the observations made after homogenizing at
973 K may be explained. Also shown in the diagram are broken lines
indicating various hydrogen contents, and the paths of the ternary alloy
compositions when hydrogen is added to 8.3 and 9.6% Al binary alloys.
Consider the 8.3% Al alloy. Initially this is two-phase;
Zr3Al and a-Zr solid solution. Following the path of the composition
of this alloy as hydrogen is added, there is initially some solubility
of hydrogen in the two original phases. However, when ^ 150 yg/g hydrogen
has been added the alloy enters the three-phase a-Zr + Zr3Al + Zr2Al
field: hence the limited decomposition of the ZrsAl in the alloy containing
190 yg/g hydrogen. When 650 ug/g has been added, the alloy is well into
the three-phase field, accounting for the substantial decomposition of
the matrix observed in this sample. At 1300 yg/g hydrogen the alloy is
in the a-Zr + Zr2Al phase field. The suggested isotherm therefore
explains the experimental observations on the 8.3% Al alloy very well.
- 13 -
Turning to the 9.6% Al alloy, the suggested ternary predicts
that the decomposition of Zr3Al should start at 450 yg/g hydrogen, and
should be quite substantial at a 1000 yg/g hydrogen content. At first
sight there appears to be a discrepancy between the diagram and the
experimental results, as only slight decomposition was observed in a
sample containing 1040 yg/g hydrogen. However, it must be remembered
that this alloy originally contained all three phases: a-Zr, Zr3Al and
Zr2Al. This is presumably due to the alloy being in a three-phase field
of the pBeudo-ternary system Zr-Al-Impurities during the heat treatment
used to transform the alloy. That this can easily happen can be seen
by inspection of Figure 15 (a) and (b). The three-phase 3 + 3 + 2 field,
stabilized here by hydrogen, could easily be stabilized by oxygen, or
a combination of impurities in the original transformation heat treatment
of the alloys. Due to the high solubility of oxygen in (3-Zr, it is
likely that the triangle will be distorted somewhat from the shape shown
in these figures* but the principle is the same.
When hydrogen is added at 973 K, the hydrogen content in all
three phases originally present must reach the equilibrium values indicated
by the corners of the three-phase field in Figure 18, before any decom-
position of the Zr3Al will occur. As the 9.6% Al alloy initially
contains three phases, there is in effect already considerable inherent
decomposition, and therefore excess a-Zr and ZrzAl. As it is suggested
that the ZrjAl phase particularly has a substantial capacity for
dissolving hydrogen, it is not surprising that additional decomposition
of the Zr3Al phase is observed only when the hydrogen content reaches
1040 yg/g.
Therefore, while it is not suggested that Figure 18 is a
definitive 973 K isotherm of the Zr-Al-H ternary, the estimated position
of the a-Zr + Zr3Al + Zrzkl phase field is close enough to allow a
reasonable explanation of the experimental results from specimens
homogenized at this temperature.
The results from the lower temperature heat treatments do
not provide any data which would enable the suggested ternary diagram
to be extended to lower temperatures. However, the diffusion rates of
zirconium and aluminum are probably so slow that attainment of equilibrium
would be impossible in reasonable times. It is interesting to note that
signs of incipient precipitation of what may be hydrides are observed
in the Zr2Al phase, which is the one in which it is suggested hydrogen
has the highest solubility at high temperatures.
As it has been shown previously that a continuous matrix of
Zr3Al is required to maintain good corrosion resistance^2), the possibility
of this phase decomposing during reactor service is of concern. Whether
this will be a problem will depend on the movement of the a-Zr + Zr$Al +
Zr2Al triangular, three-phase field as the temperature is lowered. If
this moves closer to the Zr-Al axis and/or becomes thinner, the amount
of hydrogen needed to induce substantial decomposition.will be reduced.
Conversely, if it moves away from the axis or becomes wider, the problem
will be less serious.
If the solubility of hydrogen in ZrjAl remains high at lower
temperatures, there is a distinct advantage in having~this phase present
in the alloys to act as a hydrogen sink. As the alloys with the best
corrosion'2^ and creep resistance^9) are found in the composition range
8.5 to 9.5% Al it is quite likely that this phase will be present in the
optimum alloy.
The potential problem revealed by this study must be the
subject of further investigation if these alloys continue to show promise
as a future reactor material. The magnitude of the problem would probably
best be determined by studying the effects of hydrogen on the phase
relationships when it is introduced at a slow rate at, or near, the
temperature of interest. In this regard, studies currently being performed
on the rate of hydriding of the alloys in the WR-1 reactor coolant might
prove useful, but not exhaustive.
- 15 -
5. CONCLUSIONS
1. The presence of 190 pg/g hydrogen i s suff icient to in i t iate
decomposition of the ZraAl phase of an 8.3% Al alloy to a-Zr + Zr2Al at
973 K.
2. The decomposition of ZrsAl in this alloy i s complete when
the hydrogen content reaches 1300 Vg/g at this temperature.
3. The onset of decomposition of the Zr3Al phase of a 9.6% Al
alloy occurs at a hydrogen content above 1000 Vg/g. This may be because
the excess Zr2Al phase originally present in i t ia l ly acts as a hydrogen
sink.
4. These observations can be explained by the ex is t an ce of a
three-phase ct-Zr + ZraAl + Zr2Al region in the Zr-Al-H tenary system,
in which the Z^Al phase dissolves a substantial quantity of hydrogen.
5. Heat treatments at lower temperatures for relatively short
times fa i l to change the structure of the alloys or to precipitate
identif iable hydride precipitates.
6 . ACKNOWLEDGEMENTS
The authors are grateful to G.A. Ledoux, A.E. Unger,
F. Havelock, and E.E. Sexton for their assistance with the experimental
program, and to Dr. B.J.S. Wilkins for clarifying their thoughts about
the suggested ternary phase diagram.
-16-
REFERENCES
1. D.J. Cameron and A.E. Unger: The Potential of Alloys Based on ZrJilas Structural Materials in an Organic Cooled Reactor. WNRE-137*, June 1973.
2. D.J. Cameron and A.E. Unger: The Corrosion of Zr£l - Based Alloys byCarbon Dioxide-Water Vapour Mixture.
Part I: The Kinetics and their Practical ImplicationsAtomic Energy of Canada Limited, ReportAECL - 4662, September 1973.
Part 2: The Oxidation Process, Atomic Energy of CanadaLimited, Report AECL-4665, to be published.
3. Ordered Alloys for "CANDU" Reactors.
(a) E.M. Schulson, The Tensile and Corrosion Behaviour ofOrdered ZrJil - Based Alloys, J . Nucl, Materials,50, 127(1974).
(b) E.M. Schulson and D.J. Cameron, Canadian Patent Application.
(c) E.M. Schulson and R.B. Turner, unpublished results.
4. D.J. Cameron: The Micro structures of As-Melted and Heat Treated ZrJLl -Based Alloys. Atomic Energy of Canada Limited, Report AECL - 4759,to be published.
5. E.M. Schulson: A Preliminary Study of Cast and Transformed Micro-structures of ZrAl - Based Alloys, Atomic Energy of Canada Limited,Report AECL - 4652, January 1973.
6. R.L. Beck: Investigation of the Hydriding Characteristics of Inter-metallic Compounds. Denver Research Institute, Report DRI - 2059,October 1962.
7. J.F,R. Ambler, E.M. Schulson and G.P. Kiely: Methods for Revealingthe Microstructures of the Zrjil - Based Alloys. To be publishedin J. Nucl. Materials.
8. M.L. Picklesimer in reply to D.J. Cameron at the Symposium onZirconium in Nuclear Applications; Portland, Oregon. August 1973.
9. L.G. Bell and D.J. Cameron: High Temperature Mechanical Tests onZrAl - Baaed Alloys. Atomic Energy of Canada Limited, AECL Reportto bs published.
* Unpublished internal report of the Whiteshell Nuclear Research Establishmentof Atomic Energy of Canada Limited.
FIGURE 1: MICROGRAPH OF Zr3A1 ALLOY CONTAINING8.3% A1 AND 190 ng/g HYDROGEN. (500X)
THE MATRIX IS Zr3Al. THE REGULARSHAPED DISPERSION IS ot-Zr, AND THEIRREGULAR DARKER AREAS ARE DUPLEXZr2Al +aZf, SOME OF WHICH AREINDICATED BY ARROWS.
FIGURE 2: MICROGRAPH OF Zr3Al ALLOY CONTAINING8.3% Al AND 650 yg/g HYDROGEN. (500X)
THE MATRIX IS DUPLEX a-Zr + Zr2Al.THE GREY AREAS ARE UNDECOMPOSED Zr3Al.
FIGURE 3: MICROGRAPH OF Zr3Al ALLOY CONTAINING8.3% Al AND 1300 yg/g HYDROGEN. (500X)
SHOWING COMPLETE DECOMPOSITION TODUPLEX a-Zr + Zr2Al. THE ORIGINAL a-ZrDISPERSION CAN BE SEEN AS THE LARGERLIGHT AREAS.
FIGURE 4: MICROGRAPH OF Zr3Al ALLOY CONTAINING9.62 Al AND 230 yg/g HYDROGEN. (500X)
THE VERY DARK, FINER DISPERSION IS a-Zr.THE LARGER ELONGATED REGIONS ARE Zr2AlWHICH HAS ETCHED BOTH LIGHT AND DARK.THE MATRIX IS Zr3Al.
FIGURE 5: MICROGRAPH OF Zr3A1 ALLOY CONTAINING9.6% Al AND 450 yg/g HYDROGEN. (500X)
THE Zr2Al IS THE DARK PHASE, AND THEa-Zr IS LIGHT.
FIGURE 6: MICROGRAPH OF Zr3A1 ALLOY CONTAINING9.6% Al AND 1040 yg/g HYDROGEN. (500X)
THE START OF DECOMPOSITION INTO DUPLEXa-Zr + Zr2Al CAN BE SEEN EASILY.
-20 -
FIGURE 7 : MICROGRAPH OF Zr 3 Al ALLOY CONTAINING
9.6% AT, 1040 yg/g HYDROGEN. (1500X)
SHOWING THE DUPLEX NATURE OF THE
DECOMPOSED REGIONS.
FIGURE 8: Zr3Al GRAINS CONTAINING TWINS IN AZr-8.3% Al ALLOY CONTAINING 190 yg/gHYDROGEN. (7.500X)
FIGURE 9: STACKING FAULTS IN THE Zr3Al PHASE OFZr-8.3% Al ALLOY CONTAINING 190 pg/gHYDROGEN. GRAIN BOUNDARY DISLOCATIONSARE VISIBLE AT BOTTOM RIGHT. (31,500X)
FIGURE 10: PARTIAL DECOMPOSITION OF Zr3Al TODUPLEX Zr2Al(2) + Zr(Z) IN A Zr-8.3% AlALLOY CONTAINING 650 yg/g HYDROGEN.(48.000X)
FIGURE 11: DUPLEX Zr2Al(2) + Zr(Z) STRUCTURE IN AZr-8.356 Al ALLOY CONTAINING 1300 ug/gHYDROGEN. (22.500X)
I
FIGURE 12: Zr2Al(2) AND Zr3Al(3) PHASES IN A Zr-9.ALLOY CONTAINING 450 ug/g HYDROGEN.(7.500X)
Al FIGURE 13: STRAIN CONTRAST EFFECTS IN THE Zr2AlPHASE OF A Zr-9.6% Al ALLOY CONTAIN-ING 1030 yg/g HYDROGEN. SUB-BOUNDARYDISLOCATION NETWORKS ARE ALSO VISIBLE.(48.000X)
NJ
- 2 4 -
1300
8000 10 20 Zr9AI 30 Zr2AI
COMPOSITION, AT % Al
FIGURE 1 4 : THE RELEVANT PORTION OF THE BINARY Z r - A l PHASE DIAGRAM
-25-
10 20ATOMIC %
(0)
10 20ATOMIC %
mi
FIGURE 1 5 : A SERIES OF SUGGESTED ISOTHERMAL SECTIONS IN THE
Z r - A l - H TERNARY PHASE DIAGRAM, AT SUCCESSIVELY
LOWER TEMPERATURES
tv)
FIGURE 16: AN EXPLODED VIEW OF THE PHASE FIELDS IN THE SUGGESTED TERNARY DIAGRAM
Ill
M
I
w
FIGURE 17: A VIEW LOOKING INTO THE ZIRCONIUM RICH CORNER OF THE Zr-Al-H SCHEMATIC TERNARY DIAGRAM
- 2 8 -
1500 ppm
fl vP
10 20
HYDROGEN, Atomic
30
FIGURE 18: THE SUGGESTED 973 K ISOTHERMAL SECTION OF THE TERNARY DIAGRAM,SHOWING THE PROGRESSION THOUGH THE VARIOUS PHASE FIELDS ASHYDROGEN IS ADDED TO 8.3 and 9.6% Al ALLOYS