Beruflich Dokumente
Kultur Dokumente
w-PHASE
IN ZIRCONIUM
BASE
ALLOYS*
B. A. HATT
and J. A. ROBERTS?
The metastable o-phase in zirconium base alloys has been found to be truly hexagonal with c/cl -= 0.622 h 0.002. The orientation relationships with respect to the parent p-phase (b.c.c.) are
The interpretation of X-ray data from high solute content alloys quenched from the p-phase indicates that the w formed during quenching is heavily faulted. The o structure in this state is designated diffuse satisfactorily accounts for the observed diffraction effects w. A model for the @ + o transformationwhich This is based on the gliding of { 112)~ planes in (111 )b directions in a distinct sequenoe. The is developed. magnitude of the glide component is realistically altered to account for intensity differences in the diffractions from diffuse o and bulk w. LA PHASE w DANS LES ALLIAGES DE ZIRCONIUM avec un
On a montre que la phase metastable w dans des alliages de zirconium Btait hexagonale rapport c/u = 0,622 + 0,002. Les relations dorientation avec la phase ,5 (b.c.c.) sont
(0001)~ 11 (lll)p
(ario), 11 (iOl),.
Linterpretation dessais aux rayons X a partir dalliages riches en elements en solution trempes a partir de la phase B, indique que la phase w form&e au tours de la trempe contient de nombreuses fautes dempilement. La structure o dans cet &tat est indiquee comme phase w diffuse. Un modele pour la transformation /? + o qui satisfait lensemble des effets de diffraction observes, a 6ti: developpe. 11 est base SUP le glissement des plans {112}p dans des directions (111)~ suivant une sequence definie. La grandeur de la composante du glissement peut &tre alteree, ce qui explique les differences dintensite 101-s de la diffraction de la phase diffuse w ou de la phase w massive. DIE o-PHASE IN LEGIERUNGEN rlUF ZIRKON-BASIS
Die metastabile o-Phase van Legierungen auf Zirkon-Basis erwies sich als echt hexagonal mit c/n = 0,622 or 0 002. Die Orientierungsbeziehungen in Bezug auf die P-Phase (kubisch-raumzentriert; aus ihr entsteht die w-Phase) sind (OOOl), i: (111)~ (2iTO), /i (iOl)fi.
Die Deutung der Rontgenmessungen an Legierungen mit hohem Zusatzgehalt, die van der P-Phase abgeschreckt wurden, zeigt, da6 die beim Abschrecken gebildete o-Phase sehr fehlerhaft ist. Die wStruktur wird in diesem Zustand als diffus bezeichnet. Es wird ein Model1 fur die Umwandlung /l - o Es basiert darauf, da13 {112},entwickelt, das die beobachteten Beugungseffekte zur Geniige erklart. Ebenen in bestimmter Reihenfolge in (111)~.Richtungen gleiten. Die GrGSe der Gleitkomponente wird der Wirklichkeit so angepa& da13 die Intensitatsdifferenzen der Interferenzen van diffuser und massiver o-Phase erkliirt werden.
INTRODUCTION
stages
of
ageing.
This interpretation
/?-phase of certain
advancement
1. THE
of a possible mechanism
CRYSTAL STRUCTURE METASTABLE o-PHASE
titanium and zirconium alloys transforms via an intermedia,te phase, designated CO,on ageing to the equilibrium phase may equilibrium. state. Under certain conditions the oa be obtained in a state of metastable
Previous
workers
have
made
crystallographic
detailed crystallographic study of the metastable Wphase in a Zr-7 at.% vanadium alloy. The second part of the paper is devoted to the interpretation of the diffuse co diffractions observed in high solute content alloys in the as-quenched condition and in the early
studies of the m-phase as it occurs in titanium base alloys. Austin and Doig(l) suggest that the cu_phaseis b.c.c. with side three times that of the basic /I b.c.c. cell. Silcock et CL@) and Bagaryatskii et ~2.~ independently proposed a hexagonal structure based on the hexagonal (rhombohedral) unit cell, which is an alterThe crystal native unit cell for a b.c.c. lattice.
structure
8, AUGUST
576
ACTA
METALLURGICA,
VOL.
8,
1960
presented here is based on that of Silcock and Bagaryat&ii and their interpretat.ion for a b.c.c. lattice are related. bohedral hexagonal is briefly described. Figure l(a) shows how the three possible unit cells The b.e.c. lattice is outlined by the tine lines, the heavy lines show the rhomcell and the hatched cell. lines the base of the of the atoms cell which Fig. l(b) is a projection
differ in that the atoms near the plane z = 4 lie in the plane for Silcocks structure but alternately 0.02 above and below in Bagar~atskiis the four hexagonal pretations hexagonal ments. occur structure. c/a = 0.613 and that have developed to the interIt can be shown that provided orientations
same extent, then both the cubic and hexagonal of w will show cubic Laue symmetry. structures provide by detailed intensity method
on to a flllla
Thus it is only possible to distinguish between the cubic and measurewill of disby in However which if c/a # 0.613 extra diffractions an alternative
(2TTo)hex / 1(ioi)b.c.c.
of the hexagonal
cell,
criminating
each with a (0001) hex parallel to a (lll),,,~,, 8. The cell parameters are related as follows: a hex Chex = & id3 ab.c.c. ab.c.c.
of the
Preliminary D-base
the present authors showed that the o ~~ra~tions cell with c/a > 0.613 and this interpretation verified in detail by the single crystal described.
. **
i(%f&. Both
e/a = &(3/2)
= 0,613. cell are (O,O,O); chose the above new Single graphs crystals with were oriented coincident coincident from with Laue the photo-
(OOl),
oscillation
hexagonal
X-ray
Since (u is oriented with respect to @, a beam. specimen which is fully transformed to u) still shows cubic Laue s~rmrnetr~ and can thus be oriented in terms of the original /3. Oscillation photographs chromatic MoKrx radiation with and Weissenberg the (loo), beam. if m were oscil-
(O,O,O): Ct($>t,Bf (O,O,O); &(-&$,0.52) respectively for the w-phase. These two structures
lating from 0 to 40 to the direct X-ray The diffractions cubic or hexagonal line material straight layer lines as would Since Debye-Scherrer be expected
recorded in this way did not lie on with the ideal value of e/a = 0.613. photographs from polycrystalcell with state, with a hexagonal
~/a > 0.613 and equal to 0.622 in the metastable the single crystal of a hexagonal three indices times were compared those which would be obtained
were first indexed on a b.c.c. cell with lattice parameter were then transformed to t,he hexagonal to hexagonal
corresponding
orient~atiolls with
(OOOl), ;; (ill),
FIG. 1. (a) Three possible unit cells for a body-caked cubic structure. ---b.c.c. lattice Rhombohedral lattice --i-j-i-/ Base of hexagonal lattice (b) A projection of the atoms in a b.c.c. structure onto
(ziio~, ij (ioi ja
c, = @,J%! ap by the matrix
x Atom at 34% %f& + Atom at 384 ff& The atoms numbered are common to both (a) and (b).
HATT
AND
ROBERTS:
w-PHASE
IN
Zr ALLOYS
577
with respect to
/ 1 (TOl),.
P2 values for the atomic
gives
the
(O,O,O); i (g&B). Apart from angular factors measurements indicates observed movements these values are proporNo detailed calculated agreeidea of the by Bagaryfor the
have been made but visual the intensities Some give satisfactory
that
positions
intensities.
in the positions
&(I,+,*)
atskii can be seen from Table 2, where the intensities of the zero layer diffractions atomic to following positions (O,O,O); original (12,0,0), have been calculated *($,Q,0.52). Hence
(cubic indices) which consists of the when c/a = 0.622 (11.92, (11.92, (11.92, (11.92, 0.08, 0.08) 0.08, 0.08) 0.08, 0.08) 0.08, 0.08) positions (O,O,O); to This of the
corresponding
to the
e, = 0.622 u~ # +l/3 where aIs is the lattice parameter and the transformation ,-----+ matrix is Cubic 1 1.97 1.97 1.97
as
2(b),
Table 2,
will have zero intensity &($,$,+) intensity adjacent fraction although split. 0.44 for atomic positions
for atomic
proportional
(4223) diffraction.
i
2 The difference between
2
1
It is thought to indicate
sponds to a very small amount of p. Thus there is no evidence of the atoms near the plane z = 4. The X-ray examination of aged Zr-Nb alloys
indicates the change in the positions of the diffractions resulting from c/a changing from 0.613 (ideal) to 0.622. On oscillation photographs about
of single crystal specimens in the range 8-20 at.% Nb cc)-phase has an identical above for the Zr-7 on
ratio change shows as small deviations of the diffractions from the straight layer line positions, and the separation of superimposed diffractions. Fig. 2 is a film taken with CuKa of a Zr-7.0 at.94 V specimen aged 8 hr at 400C and consisting of -98 per cent co. Table 1 gives the analysis of this 6lm. Here the (hkl) cubic indices for c/a = 0.613 are compared with those for c/a = 0.622 together with the observed lindices. The agreement between the observed values and those corresponding to c/a = 0.622 is considered adequate
6-W
The axial ratio was found to be 0.622 & and ageing temperature.
0.002 but the lattice spacing values are dependent alloy composition
Additional evidence supporting the hexagonal version of the crystal structure has been obtained from a Zr-8 at.% Nb single crystal specimen which had been strained at 400C while in the metastable cu condition. On examination at room temperature it was found that the cu developed on only two of the four possible orientations.
to four hexagonal
PP.)
578
ACTA
METALLURGICA,
VOL.
8,
1960
TABLE 1
--:;:~qzY&:&,
_
Zero layer, left-hand side of direct beam
3 4 1, 2, 3, and 4 1 0
B
25
0 4 0 4 9
930 930 930 930 370 __12,0,0 12,0,0 12,0,0 12,0,0 400 12,60 12,60 12,so 12,60 420
, 1
0*
G
Weak diffractions on some films
: : 9 1 4 1 :
--
--.
~___
~ 1
0.10 0.100* 0
side 741 lO,i,l 10,&l 10,5,1 11,2,1 ll,~,l 11,4,1 13,2,1 13,2,1 13,4,1 11,8,1 6.94 9.94 9.94 9.96 10.94 10.94 10.96 12.94 12.94 12.91 10.91
I
2 2172
3.s
0.92
0.94
0.94 5.04 1194 1.94 4.04 2.06 2.06 3.91 7.91
0.96
1.06 1.06 1.04 0.94 0.94 0.91 1.09
0.92 0.93
1.05 0.93 0.87 1.09
7a b
4313 4133
1
i
side 4 4 1 3 4 side 903 8.94 0.06 0.06 0.06 0.06 4 1 x
3 4
1 i
903
3 4 1, 2, 3 and 4 3b I 3033 3522 2352 5501 3252 3303 0557 5322 x 3 4 1 : 4 4 4 9 : 0 4 0 4 1 0 903 903 301 12,33 12,33 12,33 12,33 12,33 12,33
;
I
8.94
8.96 8.96
3.00
a b
;4
* Mean of two overlapping 3a. diffractions with I indices f 0.04 for diffractions with Identity
HATT
AND
ROBERTS:
w-PHASE
TABLE 2
IN Zr ALLOYS
~~~_. ___~
579
Ident. No.
c/a = 0.613
hkl cubic
hkl
F= z = 0.52
__~.
1 2 3 4
~
0.22) 0.22 1 044 .
b a c
990 990
990
990
3 4
i.12
0.04 0112 0.04
1 9 1 4 1 4 1
2a b a c
I
3 4
~ i
0.12
Rrom the above results it may be concluded (1) The metastable has a hexagonal lattice constants c/a = 0.622 & 0.002 G, = 3.00 A, varying crystal for structure all alloys with axial
in the diffuse o which precedes the development metastable cc) structure. 2. THE DIFFUSE o-STRUCTURE Section 1 of this paper has been concerned with the a+phase in its metastable when alloys of increasing state. solute content
of the
are approximately
a, = 5.02 and
entirely However,
and ageing temperature. (2) The w-phase is oriented phase according to the relationships
quenched from the p-field a characteristic structures is observed.(5) complete transformat,ion to the h.c.p.
cannot be suppressed. On increasing the solute concentration the w-phase together with the u-phase is observed while at slightly higher concentrations it is possible to obtain almost 100 per cent Further increase of the solute concentration (r) phase. results in
a certain amount of Zr-,3 phase being retained together with cc) which does not produce sharp X-ray diffractions. The a-phase in this state is designated diffuse W. The diffuseness of the diffractions increases with increasing solute content although the amount of diffuse cc)present may still be about 70 per cent of the total volume. In certain alloy systems where the solid solubility of
580
ACTA
METALLURGICA,
VOL.
8,
1960
the P-phase is limited, it is not possible to observe the complete sequence of as-quenched structures described in the and for this reason diffuse w is not observed Zr-Nb and Zr-U systems. suggests that the a)-phase The term the regular relative to here to describe lattice planes
Zr-Cr and Zr-V systems but is readily detected in the The mode of formation martensitic movement lographic of is taken particular
adjacent planes of the parent matrix in definite crystaldirections, the movement resulting in a new crystallographic structure. The interpretation of the diffuse w study presented below leads to the formulation of a mechanism, accounts not involving diffraction of bulk cc). or streaks coalesce and pattern becomes by solute of the diffusion, which for the observed effects and will
FIG. 4. Weissenberg photograph of the zero layer perpendicular to (110)~ from the same crystal as used for Fig. 3. also present are some weak diffractions from another phase designated ,!?.* Fig. 4 is the Weissenberg photo-
On ageing alloys which form diffuse o during the quench, the diffuse diffractions the characteristic w apparent. enrichment This coalescence diffraction
is accompanied
diffuse w-pha,se. Experimental The diffuse single crystal results w-phase technique has been examined by the using monochromatic radia-
graph of t,he zero layer perpendicular Fig. 5 is the reciprocal one quadrant symmetrical common to
to [liO],,
and
being shown since the other three are The diffuse u) diffractions this. both in the (liO), In Fig. 5 these and hence
tion from a curved quartz monochromator. Figure 3 shows a single crystal oscillation photograph from a wai,er quenched sharp p diffractions Zr-20 at. y0 Nb alloy with the The photograph with shows areas of diffuse crystal oscillating about [ liO],. together intensity,
can be divided into two groups, viz. to [ 11 lla in [112], and hence in group (A) to to [ 11 lip. as disks of intensity perpendicular perpendicular make angles of 120 in group (B)
(A) Streaks which are elongated in [n2], perpendicular perpendicular can be interpreted to [ill], [ill], (B) Streaks which are elongated
Thus in reciprocal space the diffractions or three rods of intensity which due to symmetry Similarly
diffractions
cular to [ 11 lla or three rods of intensity perpendicular to [ill],. Table 3 gives the indices and F2 values for those bulk o diffractions which can occur in the reciprocal lattice section shown in Fig. 5. The table shows that apart from those diffractions which overlap p diffractions section; only w orientations 2 and 4 diffract in this to the reciprocal lattice points corresponding
these orientations are shown by crosses and dots, respectively. Thus each of the streaks belonging to
FIG. 3. Oscillation photograph of a single crystal of Zr-20 at,.% Nb with monochromatic MoKa radiation. The crystal oscillated about (110)~ with the (001)~ oscillated through 0 + 45 to the direct X-ray beam. The crystal was water quenched after 2 hr at 920C and the photograph shows diffractions from B, diffuse o and ,!I. * The @-phase has a b.c. tetragonal structure with a = 3.52 A and c/a = 1.13 and is orientated to the B-phase having (OO1)b.c.t. I/ (001)~ and (lOO)b.,.t. 11 (100)~. The phase exists in small amounts (less than 5 per cent of the total volume) in high solute content niobium alloys.
HATT
AND
ROBERTS:
o-PHASE
IN
Zr ALLOYS
5x1
4. However the intensities of the streaks do not agree with the corresponding atomic movements The diffuse diffuseness bulk CO intensities, but as will if small be shown later better agreement give a structure intermediate of the COas compared is obtained
are given to the atoms in bulk w to between diffraction j3 and bulk CO. streaks from
Laue conditions
has not been possible to resolve whether the intensity in reciprocal space is a disk of intensity perpendicular to (111 )s or three rods perpendicular to (111 )B both possible interpretations will be developed. If the intensity distribution orientation 2 correspond corresponds to disks in reciprocal lattice space those disks associated with cc)in either to rods of CO in orientation 2 in the crystal where the axes of the rods are group (A) above can be associated with a diffraction to parallel to [ill],, orientation direction. as is shown in Fig. 6(c) or to COin only in the [ill], for the disks applies from bulk o in orientation group (B) with a diffraction 2 and those belonging 2 with exact periodicity A similar argument
_~
__ ___ _.~__
I 1 hkl
cubic
TABLE 3
Ident.
No.
hkil
hex.
I
Orient.
_
observed
Intensity
_I__112 222 442 552 772 882 10,10,2 11,11,2 13.13,2 114 224 444 554 774 884 10,10,4 11,11,4 13,13,4 118 228 448 558 778 888 lO,lO,S l,l,lO 2,2,10 4,4,10 5,5,10 7,7,10 8,8,10 10,10,10 lOi 0001 202T 1012 3032 2023 4043 3034 50% 1011 2010 0002 3037 1013 4042 2024 5053 3035 3031 2022 4040 1073 5051 0004 6062 3032 4041 2023 5050 _ 1014 ~ 6061
I P2,.,, I IP,.,,
0 1 4 : 4 4 9 0 4 0 9 1 : 0 4 4 1 : 4 : 9 9 1 4 0 0 0.6 1.0 4.3 1.0 0.4 7.8 CO.1 ~ 9.0 1.0 I 0 4.3 0.6 ; ;:r: 9.0
hlcl cubic
I!
hlcil
Intensity
Orient.
(--p,--observed
i P2,,,,,
2 ;
4 2 4 2 4 2 4 2
*
w
Will
wm
SIll
SIIl -
w * *
S VW
: 4 2 4 2 4 2 4 2 ; 4 2 4 2 4 2 4 2
S S
w * * * w m w
::9
0.6 1.0 0 0.4
k 1 m
0
1 1
0006
112 222 442 552 712 882 10 10 2 __ 11 112 ___- 13,13,2 114 224 444 554 774 -__884 10 10 4 __ 11,11,4 13,13,4 118 228 448 558 178 888 lO,lO,S l,l,lO :::::: 5,5,10 7.7.10 8,8,10 10,10,10
loio 0001 2021 7012 3032 2023 4033 3034 5054 loil 2010 0002 3037 7013 4042 2024 5053 3035 3081 2012 4040 1073 5057 0004 6062 3032
2 4 2 % : 4 2 4 2 4 2 4 2 4 2 4 2 4 2 4 2 4 2 ; 4 2 4 2 4
*
w
F20.58 ~0
0.6 1.0 1.0 4.3 0.4 7.8 CO.1
:
4 0
wm wm
SIII
SITI -
w * * s
VW
I 4 ~ 9 0 4
0 9 i
9.0 I
1.0 0 4.3 0.6
s
S
iv * * -
* -
w -
m w
6.9
* Oscillation range too small to include this diffraction. Not observed. There are also o diffractions from all o orientations superimposed on the b diffractions (llO)B (220)8 (33O)p (44O)p (002)@ (112)J (222)~ (332)~ (442)~ (114)~ (224)~ w Orientation (2) [OOOl], Ij [lll]fi w Orientation (4) [OOOl], [2iio], tiA-(4 PP.)
11 [iii],
ij [iollB
[2irO],
II r10118
ACTA
METALLURGICA,
VOL.
8,
1960
A SINGLE OF o
DISC
I
A
RECIPROCALLATTICE REPftESElWiTlCN
SINGLE ROD
P2>s
glidecomponent ctf
12345612 g g 0 g SO g
345 tf0 g
6 S
[II iJ Rfpmdicular
and rod.
to page
b)
THREE
RODS
C)
A ROD
510
OF o
A P!_ANE OR 0lS.C
associated with orientation 4 but here the axis of the rods of cttwould be parallel to @il],r. Alternatively if the disks of intensity in reciprocal space are in fact three rods of intensity those associated with LC) in orientation 2 will correspond either to three disks (or more probably ablate spheroids) of w in orientation 2 in the crystal, each disk being parallel to [Ill], but making 120 with each other (Fig. 6b) or three blocks of o with orientation 2 in the crystal, each block having periodicity in two orthogonal directions but not in the third which is perpendicular to [ill],. The above interpretations may not be very different since both relate to regions of w in the crystal which are structurally perfect in only one direction, for example w with orientation 2 is structurally perfect in only the [ill], direction and any direction perpendicular to this will intersect a large number of defects.
plane by &3/2 * aB in the (1 11>* direction contained in that plane. A gliding sequence occurring on any particular set of {112)@ planes of a particular zone necessarily implies an identical gliding sequence occurring on each of the other two sets of {112)# planes of that zone, Hence bulk w in orientation 2 can be generated by extensive gliding on the (ii2),, (i2Q or (2ii)B planes in a [ill], direction, i.e. g above is in the [f 111, direction. Similarly disks of cu will be produced if glide takes place on a limited number of the above planes. The disks will be parallel to [ill], and perpendicular to [ii2],, [121], and [Zii],. This is shown in Fig. 8 which is a projection of atoms in in b.e.c. structure onto a (liO}, plane. The disks would degenerate into rods of a parallel to [ill], if the width of disks in the (llO), directions, which lie in the plane of the disks, is restricted to a few atoms while the glide along the (lll>s has occurred for many atoms. The restriction of the width of the disks in the (llO), directions can be brought about by faulting between the disks. Fig. 9 is a projection of the atoms in a b.c.c. structure onto a (ill), plane, that is a projection at right angles to that of Fig. 8. The three sets of (112j8 planes belonging to the [ill], zone are
The experimental results given above suggest that movement of {112}, planes is involved in the /? -+ LO transformation as is the case for the /? --+ u transformation mechanism proposed by Burger@). A projection of the atoms in a b.c.c. structure onto a ( liO)B plane is shown in Fig. 7. The packing sequence of the (112), planes is given by 1,,,>,,>,, 2 3 4 5 6 1 2 3 etc. The w structure can be generated by the {112}@ planes gliding as follows:
FIG. 7. Projection of the atoms in a b.c.c. structure onto a (1101 plane illustrating the (112) packing
HATT
AND
ROBERTS:
o-PHASE
IN
Zr ALLOYS
583
of the four
4 5 6
I
of
the gliding of {112}@ planes in different zones. assumed that nucleation then the chance of faulting
0 I
If it is
2
3 4 5 6 I
9 is
0 9 3 0
0:
(112}, planes in the same zone is 2 : 1. The case of interaction of {112}P planes of different zones is complex and will not be developed in detail here but concerning it does not alter the general mechanism. Table 3 shows that the intensities sponding correlation bulk u) diffractions. if in of the diffuse w of the correa better above of When the cc) the streaks do not agree with the intensities is obtained is diffuse conclusions
w:
4 0: 0: I
However,
4 5
component
.aa) instead
Fm. 8. Projection of the atoms in a b.c.c. structure onto a {llO} plane showing the glide sequence to produce o.
positions are shown in Table 3 under F2,,,, and it will be seen that better agreement with the observed intensities is obtained. (1 la}, sulting from Since diffuse o exists as disks on that coherency strains reprevent the gliding planes it is probable transformation
sequence from going to completion. It is of interest to note that the new atomic positions reported those Ti-5%
Fm. 9. Projection of the atoms in a b.c.c. structure onto a { 11 l} plane illustrating the three (112) planes of a (111) zone. l Atoms in the plane of projection x Atoms q + Atoms q above the plane above the plane
structure for
to of
proposed Cr alloy.
Bagaryatskii
in a
However,
the heat treatment of Bagaryatskiis alloys it is not possible to relate the rumpling of the mid plane of atoms in the w-phase in Ti-5 T/oCr to coherency strains. Conclusions The model proposed for the b -+ w transformation satisfactorily accounts for the diffraction effects observed from the diffuse w-phase and implies that
numbers
primed Faulting
the gliding
sequence
suggested
is operative
during
disks if simultaneous
occurs on non equivalent (112}, planes of the [ll lls zone as defined by the glide sequence given above in which planes 1, 1 and 1 have zero glide component. If, for example, simultaneous nucleation of 0 occurs where the on planes 3 and 5, where plane 3 glides 0 and plane 5 glides g, a series of faults will develop disks merge. Many permutations and combinations of the above gliding sequence of {112}s planes which could result in faulting are possible when a large number of nucleation
considered
to acknowledge and
of to
D. Stewart, experimental
B. Bradford programme
with the
Dr. E. A. Calnan, Dr. G. I. Research Institute, and Dr. B. Butcher of Atomic Energy Harwell, for the many useful
Williams of the Fulmer M. B. Waldron and Mr. Research Establishment, discussions. Energy authors nublish. I
sites are present. Interaction must be of each of the three {112}, planes in each
This work was sponsored by the Atomic Research Establishment, Harwell, and the are ~ crateful to the Director for nermission to L
584 REFERENCES
ACTA
~ETALLURGICA,
VOL.
8,
1960
1. A. E. AUSTIN and J. R. Dora, J. Met&s, N.Y. 9, 27 (1957). 2. J. M. Sr-ccoc~, M. H. DAVIES and H. K. HARDY, in Syvqx&~na on the Mechanism of Phase Transform&ions in Metals. London, Institute of Metals (1955).
3. Yn A. BAIXRYATSKII, G. I. ROSOVAand T. W. SAGANOVA, l)akE. Akad. Nat& SSSR 105, 1225 (1955). 4. B. A. HATT, S. A. ROBERTSand G. I. WILLIAMS, Nature, Land. 180, 1406 (1957). 5. B. A. HATT and 5. A. ROBERTS,unpublished work. 6. W. G. BUROERS,Ph;ysica, &rav. 1, 561 (1934).