Phase equilibria in zirconium-rich zirconium-chromium-oxygen alloys

Jouvnd

of the Less-Common

rlfetals

Elsevier Sequoia S.A., Lausannc

PHASE OXYGEN

EQUILIBRIA ALLOYS

f~uels and Materials

IN

345

in The Netherlands

ZIRCONIUM-,RICH

ZIRCONIUM-CHROMIUM-

F. G. ELDEIZ

AND

W. RI. KLWB_XLL

-- Printed

Chulk River Nuclear Laboratories, .Stomic Energy cf Canada Llmiied,

llivision,

Chalk River, Ontario (Carzada)

(I-teceived August 5th‘

1969)

SUMMARY Phase equilibria X-ray

and electron

in the system Zr-Cr-0

diffraction,

analysis in the composition oxygen. A binary {a-zirconium

have been studied by metallography,

electron-probe

microanalysis,

hardness

and thermal

ranges o-2.25 wt.% chromium and 0-2400 p.p.m. +ZrCrz) eutectoid forms over a temperature range that

expands with oxygen concentration up to about 135°C for a zqoo-p.p.m.-oxygen alloy. Oxygen also increases the chromium concentration of the eutectoid from between 0.7 and 0.95 wt. 9/, for a o-p.p.m-oxyg en alloy and 1.25 wt.96 for a zdoo-p.p.m.oxygen alloy. The oxygen concentration

of thex solid solution has a derived maximum

of about 9000 p.p.m.

900°C.

The crystal

at approximately

structure

earlier work. The impurity

determined

for ZrCrz (M&LIZ type, a =7.2x

iron dissolves completely

has no observable effect on the lattice parameter. Outstanding microstructural features are the formation sitic product gonal

omega

in quenched phase

0.2 to 1.5 wt.:/, chromium

(a=z.oz

.A, c =3.12

in a quenched

cooling rates,

iron,

of a twinned marten-

alloys, the existence

A, cla=0.622)

chromium alloy and the appearance, at certain regions of a grain-boundary “alpha-like” phase.

.&) confirms

in ZrCra but, up to 3 wt.:;

of a hexa2.25-wt.“/,-

from the p and x +/3

I. INTRODI:CTION

The binary equilibrium although

the latter

diagram

diagrams Zr-01

and Zr-Cr”,z are fairly well established

is not too precise in the zir(~onium-rich region. Oxygen

strongly stabilizes the n-phase, while chromium mildly stabilizes the p-phase forming an eutectoid at about I wt.:/, chromium. The eutectoid composition was obtained by both HAYES et ak2 and DOMAGALA et at.3 by intersection of the (~+~)~~ and @/(@+ ZrCrz) boundary lines and reported, respectively, as 1.8 wt.9: chromium (for an alloy of about IOOO p.p.m. oxygen) and 1.0 wt.% chromium (for about IOO p.p.m. oxygen). An outline of the ternary diagram has been established by RHEE AND HOCK* for most of the system between 1200’ and 1700°C, but the diagram was only approximate J, Less-Common

Metals, 19 (1969) 345-358

W. M. RUMBALL,

346

F. G. ELDER

in the zirconium-rich corner. The present work was undertaken in the temperature range of 830°C to 1130°C, spanning the (01+p)//? b oundary in alloys of about I wt.% chromium. 2. MATERIALS

Consumable-arc-melted alloys were obtained from two sources. Alloys nominally containing 0.2, 0.5, 1.0, 1.5, 2.5 and 25 wt.% Cr and an 0 concentration of 1050+50 p.p.m. were prepared from sponge zirconium as 200-g button melts, homogenized by hot rolling and annealing, and finally rolled into strip approximately 0.32 cm thick from which specimens were cut. A I wt.% Cr/4r-p.p.m. 0 alloy was similarly prepared from iodide zirconium. Three alloys were based on a 1.25-wt.O/~ Cr alloy prepared from a 1700 lb ingot reduced to r.57-cm diameter bar stock from which specimens were taken. This alloy contained the major impurity of 0.085 wt.% Fe, in addition to 1400p.p.m. 0 and was used to produce alloys of 2400 p.p.m. and 2000 p.p.m. 0 by an oxidation technique described by KENCH et a1.5. The data of KENCH et al. were used as a guide in the selection of homogenization treatments up to 80 h in the temperature range IO~~~-II~~~C. Oxygen concentrations were determined by vacuum fusion analysis. Analysis details are given in Table I. TABLE

I

ANALYSIS

OF

ALLOYS

Zirconium grade

Nom&al wt.% Cr

wt. y0 Cr

Sponge Sponge Iodide Sponge Sponge Sponge Sponge Sponge Sponge Sponge

0.2

0.5 1.0 I.0 I.25 I.25

1.25 I.5 2.5 25.0

n.a.

Analysis

=

0.20 0.50 I.00 I.00

1.25 I.25 1.25 1.46 2.25 25.00

Not analyzed

IMPURITY

ANALYSIS

p.p.m. 0

p.p.m. Fe

1050 1050 4’ 1050 1400

zt50 *5r50

n.a. n.a.

zt50 rk5’J

n.a.

2000

&zoo

80

850 850 850 n.a. n.a. n.a.

2400 fzoo 1050 +5o 1050 f5o 1050 $50

for. OF

I.25-Wt.“/o

CHROMIUM

ALLOYS

Element:

Al

B

C

Cd

Co

Cu

H

Hf

Mg

Mn

MO

(P.P.m.)

<25


II0


<5

(25

6

96


< 10

<10

Element:

N

Nb

Ni

Pb

Si

Sn

Ta

Ti

Na

V

Zn

20

<

<20


<5

<50

(l+m.)

3.

25

EXPERIMENTAL



TECHNIQUES

<5

50

AND

200

PROCEDURES

The phase fields were established mainly by optical metallography of specimens water-quenched from the temperature of interest. Of the high-temperature phases, only p is not retained in its original state by quenching. The p-phase transforms, J. Less-Common

Metals, 19 (1969) 345-358

ZIRCONIUM-RICH ZIRCONIUM~HROMIUM-OXYGEN

347

ALLOYS

structure typical of usually depending on the rate of cooling, to a “basketweave” air-cooled @-Zr (Fig. I), or, if quenching is rapid enough, to a twinned martensitic product, K’ (Fig. 2). Roth transformation products are readily distinguishable from all other phases. Some electron diffraction examination was performed while differential thermal analysis, hardness measurements, electron-probe analysis and X-ray diffraction were used to confirm and supplement the metallographic data. In some specimens, the extent of p-transformation was determined by point-counting to establish tie-lines and phase boundaries. Specimens were held at various temperatures following stepwise cooling from temperatures in the P-region, i.e., from 1050°C for the 0.2, 0.5, 1.0, 1.25 and 1.5 wt.% chromium specimens and from r16o’C for the 2.5 wt.% chromium specimens (see Table II). The times of the isothermal anneals were based on estimates of time to equilibrium at various temperatures reported by DOMACAL.-\ et al.3 for the Zr-Cr T.‘.Bi,E

1Ia

HEAT-TREATMENTDETAILS Eqf~izibvium

tempevatuvc (“Cl ---.___-

_~__

FOR

0.2

Wt.::

(h*)

(h) _I._______-0.5 24 24 24 24 24 so 2-i 48 28 48 01

(6) _. ~___ 0.5 2s ‘25 25 25 49

0.5

0.5 24 1og “4 68 32 54 24

880 870 860

_

.-Ippvoximate total tin??

0.5 0.5

917 910 9oo 885

ALLOYS

Tiw at equilihriwn**

i)60

9.50 940 0.30 920

CHROMIUM ___..._.

Cooling time from 1050°C -

1050 970

TO I.j-“t.“& -~~

IO0

-19 I’7 0I

IO.3 I16 7.3 1’ 3

48 40

24 72

107 76 $1 48 1’3 (‘i ._ ._______ * Solution treatment was, in all cases, 0.5.-1.5 h at ro5o”C, and cooling was in approximately ~5 deg C steps down to the appropriate equilibrium temperature. ** :\fter the equilibrium treatment, all specimens were quenched in water. go 48 48

850 840 8.30

TABLE

I Ib

HEAT-TREATMENT _. ._~

Equsl~bviwP te~n~~~~t~ye

PC) -.-____ rryl 1130

III0 1090 1050 950

-.-__

DETAILS

FOR

Sol%lttnr? tempevntuue

Y°C) _._~__..

THE

2.25.Wt.Oj,

Sol2‘&m time

(h)

CHROMIL’M

ALLOY _____

~

Gooli~zg time

Total

temprvatuve (h)

(h)

fPOrn SolzrfioRtirrze

1160

I

I

II50 IIyl

I 2.5

2

0.5 I

I.5 3.5 18 20. j 21

1150

16

1150

22.5

I IjO

.______

18.5

2

4 2.5

* All specimens were quenched into water after equilibrium treatment. ,I. Less-Commcn

Metals, 19 (1969) 345-358

W.M.RUMBALL,F.G.ELDER

348

system. Following the anneal, specimens were water-quenched to room temperature. All heat treatments were carried out in a dynamic vacuum of less than x0-6 torr. Specimens were water-quenched within two or three seconds of breaking this vacuum. Metallographic specimens were prepared by grinding on Sic papers, polishing on 6-,um-diamond pads, chemical polishing with a mixture of 45 parts lactic acid, 45 parts nitric acid and 8 parts hydrofluoric acid, and finally anodizing in a ro%oxalic-acid solution. The specimens were examined under normal illumination and polarized light. There is insufficient intermetallic compound present in alloys with < IO wt.% Cr to permit identification by X-ray diffraction. Thus, electron diffraction examinanation was performed on precipitates extracted from the specimens after etching with the chemical polish mentioned above but containing water instead of lactic acid to facilitate deeper attack. A conventional two-stage replication was employed with direct evaporation of carbon on to the plastic, followed by dissolution of the plastic. These carbon-extraction replicas were examined in a Hitachi HU-II electron microscope at an operating potential of IOO kV. 4. RESULTS

4.1. Metallography The locations of the phase boundaries were established by metallography: Table III summarizes the observations. The (LX-t-p +ZrCra)/(ix +ZrCrz) and /?/(a +,8 + ZrCrz) boundary temperatures, as determined by metallography, hardness testing TABI,E SUMMARY

III OFMETALLOGRAPHIC

Temp.*

wt.:4

Cr:

("ci

p.p.wt.0:

OBSERVATIONS _____..-________....

.

0.20

0.50

I.00

1050

1050

‘+I.00

B

B

1.00 _.

1050

II50

1130 x110 1090

1050 1000 970 960

950 940 930 920 9=7 9x0 900 890 885

f a+B a--tB =+B a+@

s a+B a+B a+B

880

a+B

a+P

5

m-tB 870 860 *+B 855 n+p 850 a+@ 840 a+j3 830 ~.~~~_ -. * Temperatures estimated at J. Less-Common Met&,

19

&t-B &x+/3 + Zrcxz + ZrCr2 + ZrCr2 & 5°C.

(1969)

345-358

a+p -1 ZrCrz a+/3 -+ ZrCr2 n+B -+ ZrCr2

B a i f!3+ ZrCr2

a+/l

+ ZrCrl

ZIRCONIUM-RICH

ZIRCONIUM-CHROMIUM-OXYGEN

ALLOYS

349

and thermal analysis, are compared in Table IV. Quenching of the /3 solid solution produces a distinctive fine “basketweave” or martensitic structure (Figs. I and z), (depending on quenching rate6*7), together with a small amount of a phase (having the appearance of a) forming at the j3 grain boundaries (Fig. 3). or as a halo around (x grains already present at the solution temperature (Fig. 4). Proeutectoidn precipitates as thick lathes in one or more specific orientations within a single B grain (Fig. 5), while hypereutectoid alloys form large crystalline ZrCrz particles, initially at ,!l grain boundaries (Fig. 6). Quenching the z.5-wt.%Cr alloy produced a structure taken, at first sight, to be retained /I (Fig. 7), but which, by X-ray diffraction, later appeared to contain the hexagonal w-phase, reported as a metastable decomposition product of /3 in other zirconium alloyss. It was difficult to eliminate entirely the possibility of retained @ being present as many ,!I and OJ Hragg reflections almost coincide. However, several reflections were unambiguously due to IO and all lines were closer to Bragg angles expected for 01 than for /3. Fully-transformed structures of the alloys showed the (LY+ZrCrs) to be lamellar or rod-like in appearance (Fig. 8). Figures 9-12 show the structures of fully-transformed specimens quenched from 830°C just below the (OL+ ZrCrb) /( x +fi + ZrCrz) temperature. The quantity of the intermetallic compound increases with increasing chromium concentration. 4.2. Quantitative metallogra$hy The amount of /I transformed

as a function

of annealing

temperature

I.$)

2.25

1050

IOJO

(Fig. 13)

tl

B

B

/!I +

B x+fi a+/? n+/l a+fl %+/I

zrcrr

/I f ZrCr2 /I + ZrCr2

,Y+B + ZrCr2 ,cx+B + ZrCrz

J /? B p fl J p @

x4-p x+b ,a+/3 x+fl t+b x+fi x+-/3 %+/II

+ + + + + + + +

ZrCr2 ZrCr2 ZrCra ZrCr2 ZrCri ZrCr2 ZrCr2 ZrCrB

+ + + + + + + +

ZrCr2 ZrCr2 ZrCr2 ZrCr2 ZrCr2 ZrCrz ZrCr2 ZrCr2

,x+/3 n+b x+/j s+fi ,%+p a-t/3 ,z+fi r+fl+ %+/II s+j? LX+/?

s

B + -I+ + + $+ + + + + +

ZrCr2 ZrCr2 ZrCr2 ZrCrz ZrCra

ZrCr2 ZrCr2 ZrCr2 ZrCri ZrCr2 ZrCr2 ZrCr2 ZrCr2 + ZrCr2 + ZrCr2 f ZrCr2

s fl f /3 + /sl 4 /J + x+/j n+/3 ,x+/3

fi + ZrCr2

ZrCr2 ZrCr2 ZrCra ZrCr2 + ZrCr2 + ZrCr2 + ZrCr2

/I + ZrCr2

s-t/3 + ZrCrz

J. 1~~Conwno1~ Metals, 19 (1969) 345-358

Fig. I. “Basketweave” structure ( x 500, polarized light)* Fig. 2. Martensitic structure (X 500, polarized light)*

Fig. 3. Grain boundary

produced

by air cooling

produced by water quenching

“a” formed on quenching

Fig. 4. Halo of “01” around proeutectoid polarized light)*

* All photographs

in z.z5-wt.%

(>zooo deg C/set) from the p-region.

from the @-region. ( x 500, polarized light) *

LXfollowing quenching

Fig. 5. Morphology of proeutectoid il: in 1.25.wt.% quenched. (x IOO, normal jllumination)* Fig. 6. ZrCro precipitates normal illumination)*

(NZO deg C/set) from the p-region.

Cr ahoy

from the (LX+ fl)-region.

(x 500,

Cr alloy slowly cooled to 900°C and water quenched

have been reduced in reproduction

x 0.6.

from

(B -I- ZrCrz)-region.

(X 1000,

ZIRCONIUM-RICH

Fig. 7. Structure

ZIRCONIUM-CHROMIUM-OXYGEN

containing

Fig. 8. E:quilibrium

E‘ig. g. Structure

structure

theto-phase

ALLOYS

351

in a Zr/z.z5 wt.“< Cr alloy.

( X 500, normal illumination)*

p.p. m. 0 alloy.

(X 100, normal illumination)*

of 1.25-wt.S;Cr/2qoo

of o.z-wt.qd Cr/rogo-p.p.m.Oalloy

quenched from83o”C.

(X 5oo,polarized

light)*

Fig. IO. Structure light) *

of 0.5-wt.:/,

Cr/Io50-p.p.m.

0 alloy quenched

from 830°C.

(x 300, polarized

Fig. il. light) *

of 1.0~wt.06

Crj1050-p.p.m.

0 alloy quenched

from 830°C.

(x 500, polarized

Fig. 12. Structure light) *

of I.5-wt.D/, Cr/Io50-p.p.m.

0 alloy quenched

from 830°C.

(x ym, polarized

* All photographs

hax

Structure

been reduced in reproduction

X 0.6.

W. M. RUMBALL, F. G. ELDER

352 TABLE

1V

TRAN~FOXMATION TEMPERATURES

OF

ZljI.2~-Wt.“/o

CHROMIUM/2400-P.P.lIl.

_-._

.. ..~_ gho + 10°C 965 fIO”C 965 &5”C

Hardness Thermal analysis Metallography

OXYGEN

ALLOY

_~~

830 f 10°C 831 *z’C 835 It5OC

860 -

640.

*zoo I * 10

20

30 %

Fig. 13. Transformation

40

50

60

Transformation

70

60

90

100

by volume

vs. temperature curves for alloys with o.z-1.25 wt.% Cr

was established by a point-counting technique, having an estimated accuracy of & 5:/o. The important feature of these curves is the marked acceleration of the rate of transformation with temperature at around 890°C. This change in transformation rate must be related to the shape of the binary eutectoid valley and the geometry of the &Zr solid-solution phase field. 4.3. X-ray and electron diffraction ZrCrz may exist in either Laves-phase crystal structures of C14 (MgZnz, hexagonal) or CI~ (M&US, cubic)s. SHEN AND PAASCHE 10 have recently confirmed the findings of ALISOVA et al.11 that the cubic form is stable up to 1550°C. The cubic modification has been found by X-ray diffraction in the present investigation on the 25wt. %-Cr alloy (powder annealed in the (/3+ ZrCrz)-region) and by electron diffraction on the low-chromium alloys, annealed in the range 7oo”-92o’C. The lattice parameter is in good agreement with that previously reportedg, viz., a=7.21 A. Iron present as an impurity in the I.z5-wt.% Cr alloys segregates to the intermetallic compound (see below) but does not appear to affect the lattice parameter. The w-phase observed in the z.25-wt.% Cr alloy has lattice parameters very close to those of the w-phase observed in Zr7 wt.% Nb alloys (see Table V). J. Less-Cornvmm

Metals, 19

(1969)

345-358

ZIRCONIUM-RICH

ZIRCONIUM-CHROMIUM-OXYGEN

353

ALLOYS

4.4. ELec~~o~-~~cy~~yobeanalysis Several specimens were analyzed in order to locate phase boundaries and the results are summarized in Table VI. The intermetallic composition corresponds closely to the 53.1 wt.% Cr of stoichiometric ZrCrz. The segregation of the iron into the intermetallic compound has been referred to above. The analyses were made using the alloys quenched from the p-region as internal standards but, even so, the instrument was probably working at the lower limit of its accuracy in the ,x solid solution.

INTERPLANAR IN Zr/e.ag-wt.76

SPACINGS

FOR Co PHASE

CHROMIUM

ALLOY

-.____ PEnw

d-Sp~c~~gs

jAkE) -~

Calc.*

II"

2.j13

*.jIL

201

1.785 1.455 I.327

1.781

121

112 220 ‘31 302 222 312 * Calculated

TABLE

(if)

--__ 06s.

1.456 1.326

I.256

I.*jy

1.125 1.063

1.127 1.063

0.979 0.954

0.978 0.953

d-spacings

based on the hexagonal

cell a= 5.022 .k, c=3.124

_A, clazo.622.

VL

ELECTRON

MICROPROBE

0.2

1050

ANALYSIS

OF QUENCHED

zr-cr-0

ALLOYS

~- . 950 917 900 880

0.30

0.25

-_

--

0.23

-

0.40 0.40

870

840

0.5

10.50

950 917 900 880 870 840

1.0

1.25

1.25

1050

2400

rq.00

0

0.55

-

0.50 0.50

-

0.02 -

0.47 0.65

0.02

0.65

49.8

49.1 I.00

9=o 870 840

0.95 0.70

52.9

920

0.17/0.10

0.08

0.88

0.13

580

0.10

0

0.85

O.fO

860

0.10

0

0.80

0.20

840

0.10

0

0.70

0.20

870

-49 -49 “49

0.80

J. Less-Common

X.9 1.9 2.3

51.1

Metals,

19 (1969)

345-358

W. M. RUMBALL,

354

I;. G. ELDER

Both chromium and iron concentrations are considered to be accurate to only fo.05 by wt.%. The alloy containing 1.25 wt. “/ Cr and 2400 p.p.m. 0 was determined metahography to be close to the binary eutectoid composition and the composition of /3 remaining during the binary eutectoid separation will follow the binary valley. Table VI and Fig. 14 show that the chromium concentration of the eutectoid falls with increasing temperature below that of the I-wt .“i, Gr binary-eutectoid composition quoted by DOMAGALA et al.3 for a roe-p.p.m. 0 alloy. However, the 4I-p.p.m. O/ r.o-wt.“/o Cr alloy contained no proeutectoid ZrCra but a very small amount of a, suggesting a eutectoid composition for this oxygen concentration of about 1.0 wt.% rather than about 0.75 wt.94 from Fig. 14. The formation of (x alone at low undertoolings, rather than (a+ZrCra), as the eutectoid, may be attributed to the difficulty of ZrCrz-precipitation in these alloys. Certainly, there is a decrease in chromium of the eutectoid with decreased oxygen concentration in the alloy and it may be confidently concluded that the binary Zr-Cr eutectoid composition lies in the range 0.7-0.95 wt.o/Cr. 1000

111

o’s

yrr, 7.0 wt. %

Fig. 14. Projection

( 1.5

2,o

1 2.5

chromium

of (a + ZrCm) binary eutectoid valley.

Fig. 15. Vertical pseudo-binary Zr-Cr section at 1050 p.p.m. 0. 5. DISCUSSION AND CONSTRUCTION OF PHASE DIAGRAM

The addition of oxygen to binary Zr-Cr alloys results in the separation of the eutectoid (n+ZrCrz) athermally rather than isothermally, Most of the observations are conveniently presented in Fig. x5 : a vertical section at 1050 p.p.m. oxygen, which shows the extent of the (LX+fl+ZrCrs)-region for alloys of varying chromium concentration at an oxygen level typical of commercial zirconium alloys. By the use of the metallographic data in Table ITI, electron probe analysis results (Table VI), 74 transfornlation curves (Fig. 13), and the lever rule, and by assuming ZrCrz dissolves no oxygen* (Fig. 16), fairly accurate isothermal sections may be constructed. Details of the calculations involved are given in the Appendix. In addition, the slope of the binary eutectoid valley may be derived from the metallo-

* This was assumed by RHEE AND HOCH”. Also, PEBI.ER AND C;ULBRANSEN~~found negligible hydrogen solubility under comparable conditions, and it is reasonable to assume zero solubility for the larger interstitial oxygen atom. J, Less-Cowmolz

Metals,

q

(1969) 345-358

ZIRCONIUM-RICH

0

1.0

05

ZIRCONIUM-CHROMIUM--OXYGEN

1.5

2.0

wt. % ChrOmlum

0

500

1000 Oxygen

1500

355

2.5

Fig. 16. Isothermal sections of Zr-G-0 85oYz.

6201

ALLOYS

2000

concentration

phase diagram at (a) 95o”C‘, (b) qzo°C, (c) 890°C and (d)

2500

3

10

(p.p.m.)

Fig. I 7. The effect of oxygen concentration on the (a + j3 + ZrCrz)-region in the binary eutectoid alloy.

graphic data (see Appendix) and Fig. 17 shows the excellent agreement between the derived shape and experimental observations. The relationship between the (LX+/?+ZrCr,)/,J and (LX+/3+ZrCrz)/(ti +ZrCrz) boundaries is amenable to thermodynamic treatment if it is assumed that the eutectoid (E) may be regarded as a single component in a quasi-binary systemla. The two boundaries involved then are (#I+ E)@ and (#?+ .s)/e and the Can’t Hoff equation may be applied to their initial slopes, dC, dC, ---=---dT dT

AH RT,2

(1) J. Less-ComanzonMet&,

'9 (196~) 345-355

W. M. RUMBALL, F. G. ELDER

356

where K&T and &I&IT are the slopes of the (/?+&)/B and (/?+E)/F boundaries, respectively, in atomic fraction/deg K, AH is the heat of transformation for the eutectoid reaction ,4 + F, R the gas constant (=1.g8 cal/deg K) and T, is the binary eutectoid temperature ( =IIoS’K). The AH value may be estimated additively14 from the AH value for the ol-%//I-Zr transformation and AH of fusion for ZrCra. The latter figure is relatively insignificant compared with the former, due to the low weight-fraction of the compound present in the eutectoid (about 0.02) and may be ignored. AH then is set at 1280 cal/mol and, measuring the initial slope dCpldT from Fig. 17 (0.652 x10-~) and using eqn. (I), gives a value of 5.912 x10-4 for ac,/aT which is 9470 p.p.m./“C. Thus at 2400 p.p.m. 0 (the highest-oxygen alloy studied) the eutectoid transformation should be complete at 835 (binary eutectoid temperature) +z400/g470°C =835.25%. Thus, oxygen is predicted to have an insignificant effect on this temperature, as was indeed observed, the eutectoid reaction always being completed within the experimental error of 835” 25°C. An interesting feature emerges from the derivation of the oxygen concentration in the a-phase present in the three-phase region, namely, that, with decrease in temperature, the oxygen concentration first increases and then decreases sharply below about 900°C (Fig. 18). The complex curve in Fig. 1.8in fact represents a vertical section through the ternary diagram at the assumed constant-chromium content of the a-Zr solid solution. The &/(a +/I) surface is virtually horizontal as it intersects the Zr-0 system but is more nearly vertical as it intersects the Zr-Cr system. The a/(& +/?) surface must, therefore, be buckled in the ternary system and it is this buckle which produces the kink in Fig. 18. The consequence of this “kink” is the sharp change in rates of transformation with temperature noted in Fig. 13.

960

10

-

\o

943-

\

-1.

g

p 92o1 -$900x ; 880860 840.

A: _/--.

7L

0 Forone mayonly

/-

820 0

-

2000

4000 Oxygen

h&c;*for two or mar.?

6000

8000

concentration

(ppm.)

10000

Fig. 18. Derived oxygen concentrations in 0: for the three 1.25-wt.%

Cr alloys.

ACKNOWLEDGEMENTS

The authors wish to thank Dr. PAULDIXON of Canadian Westinghouse Company for preparing several of the alloys, Dr. R. ASHLEYand Mr. T. LONGHURST for the oxygen analyses, and Mrs. T. BETHUNEfor the electron-microprobe analyses. J. Less-Common

Metals,

q

(1969)

345-358

2I~CO~I~M-RICH

ZIRCONI~~~HROMIUM-OXYGEN

357

ALLOYS

REFEREXCES 1 M. HANSEN ANI) K. ANDERKO, ~~~~s~i~~~tia~t of B?mzry AZloys, McGraw-Hill, New York, 1958. 2 1:. T. HAYES,.%. H. ROBERSON AND M. H. DAVIES,TWZS. AI&E, 194 (1952) yq. 3 R.F. DOMAGALA,D.J.MCPHERSON ANDM.HANSEN, Tvans..4IME,197 (rg53)279. 4 S. K. RHBB AND M. HOCH. Tvaws. =IIME, 230 (1964) 1687. 5 J. R. ICENIX, J. H. FOLEY ANL) S. A. XLDKIDGE, CRSL Rrpi. So. .-lECL-2623, xgG0. 6 G. F. SLATTERY.~. ~ess-Co?~l~~o~~ Metals, 16 (1968)91. 7 \Z'.31. %MB.4LL, CRAiL Re$t. So. .4ECL-3050, 1968. 8 U.J. COMETTO,G.L.HOUZE, JR. AND R.F.HEHEMANN, Tvans..41ME, 233 (1965) 30. 9 W. KOSTOKER, J. Medals, 5 (1953) 304.

APPENDIX

Derivation

and use of the 2ewr-rule

relationship

Referring to the isothermal section of Fig. AI for all alloys within the threephase tie-triangle (a +p +ZrCrs), the composition of the phases is given by the corners of the triangle, i.e., A, B and C. As C, in this instance, lies at 53 wt.?; Cr and o p.p.m. 0, the lines AC and BC may be assumed parallel to the Zr-Cr axis; this will involve errors of, at most, r “/ in, for example, the oxygen concentrations in 01.

0

05

1.0

15

wt

2.0

25

% ‘hi”rniUrn

Fig. XI. Schematic isothermal section of Zr- -Cr-0 phase diagram

Prom the lever-rule, the equilibrium temperature T (“C) is:

fraction

at temperature, T(T)

of /? transformed

(r) in the alloy X at

where x0 is the oxygen concentration in the ahoy X, xi3is the oxygen concentration in untransformed /3 (at point B) and X~ is the oxygen concentration in the x (at point A). Rearranging eqn. (,%I).

In eqn. (AZ), both xdl and x0 are unknown, while t has been determined experimentally. From the experimental data for t at various temperatures (Fig. x3), x, may be calculated for a range of temperatures from eqn. (AZ). As, at most temperatures, two J. Less-Common Metals,

19 (1969) 345-358

W. M. RUMBALL,

358

or more alloys were examined in the three-phase triangle, simultaneous be set up. For example, for two alloys (superscripts I and z) Xfl =

tlz$

F. G. ELDER

equations

may

- +X01 (A3)

tr-22

and x0 is then inserted in eqn. (A2)to give two values of xX which can be averaged, Whenever experimental data for three alloys were available, this procedure gave three values of ~0 (eqn. (A3)), which compared very well for temperatures down to 8go”C (usually to within +IOO p.p.m.). Below Sgo”C, the values for -c were not sufficiently accurate for reliable estimates of ~0. From the average values of ~0, two or three values of xtz were obtained which did not vary more than +5% from the average values plotted in Fig. 18. .J. Less-Common

Metals,

19 (1969) 345-358