JOURNAL
430
THE EFFECT
OF THE
OF ANNEALING
LESS-COMMON
METALS
TEMPERATURE
AND FRACTURE D. M. R. TAPLIN
ON
THE
DUCTILITY
OF URANIUM
AND
J. 1%‘. MARTIN
Department of Metallurgy, 0,zford University (Great Britain) (Received August 27th, 1963)
SUMMARY Commercial purity uranium (total impurities< 500 p.p.m.) was subjected to a rolling and annealing schedule to produce a fine uniform grain structure. Samples were rolled 28% at 300°C and annealed for 24 h at a range of temperatures between 240’ and 75o’C and tensile-tested. The tensile properties, were correlated with the metaltogether with hardness measurements, lography of fracture and a hypothesis for the two ductility peaks is presented in terms of the change in fracture characteristics.
INTRODUCTION
Previous work has indicated that the temperature of annealing of deformed uranium can have a profound effect on ductility. DE LIBANATI AND DE TANI+ found that partially recrystallized alpha uranium, obtained on annealing at 350°C was much more ductile than material fully recrystallized at 630°C. The results described below investigate this effect over a wide temperature range and attempt to correlate it with metallographic observations of fracture. MATERIAL
AND
EXPERIMENTAL
The uranium used was magnesium-reduced virgin billet of the following approximate composition: C 300; Al 50; Fe 50; Si 60 p.p.m. The cast billet, which was g in. in diameter, was cut into six equal longitudinal segments and a rolling rod 1.25 in. diameter was machined from each segment. This procedure eliminated the axial pipe from the rods and considerably reduced any segregation effects. One rod was then subjected to a mechanical and thermal treatment to break up the cast structure, weld up any porosity, and produce a fine uniform grain size. After annealing in vacuum at 600°C the rod was subjected to a final rolling reduction of 28% at 300°C to 0.423 in. diameter. Tensile specimens with a cross-sectional area of 0.025 sq. in. and a gauge length of 0.632 in. were machined from the rod and annealed for 24 h at the following temperatures; 24z”, 348”, 420°, 4g6”, 545”, 613O, 641OC (alpha treatments), 680°C and 752°C (beta treatments), in a dynamic vacuum of 10-6 torr and furnace-cooled. Tensile specimens were tested at room temperature on a standard Hounsfield Tensometer at a strain-rate of 1.25% per minute and the stress-strain curves plotted. After fracture the specimens were sectioned longitudinally and examined by optical J. Less-Common Metals, 6 (1964) 430-433
EFFECT
OF ANNEALING
TEMPERATURE
microscopy. Grain size determinations and Vickers carried out on the undeformed shoulders.
ON u
hardness
431 measurements
were
RESULTS
Figure I shows the effect of annealing temperature on per cent elongation and per cent reduction in area at fracture, the experimental points between 300’ and 600°C being the mean of two results. Good reproducibility was obtained. The per cent
/
I
250
100
200 Annealing
Fig. I. Annealing
300
400
tcmperoture
A-J
I
500
Km
80%5
700
*C(24hours)
temperature V~YSUSpercentage elongation and percentage with predominant fracture modes indicated.
reduction
in area
elongation curve shows two peaks, one at N 350°C and one at N 5oo”C, but in the per cent reduction in area curve there is a sipgle broad maximum. This is attributable to the specimens tested after annealing at 348°C and 420°C exhibiting local necking; above 450°C the specimens fractured without necking. The hardness and U.T.S. results are given in Fig. 2 and indicate a progressive decrease in hardness and strength with increasing annealing temperature. Figure 3 demonstrates the effect of annealing temperature on the true fracture stress. The grain sizes are given in Table I. The banded structure observed was similar to those previously reported in uranium (e.g. ref. 2.). After annealing at 348°C the microstructure was composed of approxi-
150,
100
200 Anneo,mg
Fig. 2. Annealing temperature
300
I 400
temperoturc
II
500
I 600
A, 700
00040
*C (24 hours1
WYSUSU.T.S. and Vickers hardness modes indicated.
with predominant
J. Lrss-Common
Metals, 6
fracture
(lg64) 430-433
432
D. M. R. TAPLIN, J. W. MARTIN
mately 25% recrystallized grains and 75O/”deformed, elongated grains. After annealing at 42o’C 50% of the microstructure was composed of recrystallized bands.
Fig. 3. Annealing temperature V~YSUS true fracture stress with type of grain structure and predominant mode of fracture indicated.
TABLE Annealing temperature (“C)
I
Grain size (average diameter in mm)
Comments
242
-
I
348 420
-
banded structure in direction of rolling with recrystallized (grain size 0.01 mm) I and deformed regions
545 613 496
0.015 0.04 0.01
64’
0.06
some secondary recrystallization
680 752
0.75 0.9
beta-transformed grains
as rolled
deformed elongated grains
uniform smooth grain structure 1
structure; some large
Examination of all the fractured specimens revealed the expected mixed ductileintercrystalline-cleavage fractures. The coarse-grained beta-treated structures exhibited a considerable amount of cleavage, whereas the alpha-recrystallized structures exhibited a largely intercrystalline fracture. The unrecrystallized and banded structures exhibited a ductile fracture with much less evidence of intercrystalline or cleavage failure. DISCUSSION
The ductility values obtained in the present work are considerably higher than those generally quoted (e.g. ref. 4) ; this is considered to be due to the particular precautions taken in the preparation of the material. J. Less-Common
Metals,
6 (1964)
430-433
EFFECT OF ANNEALING
TEMPERATURE
OS Lv
433
There are two changes in fracture characteristics with increasing annealing temperature in the range investigated; (a) At annealing temperatures below 450°C a predominantly ductile fracture occurs, whilst after annealing above 450°C intercrystalline fracture predominates. Since no intergranular precipitate has been observed this suggests that above 450°C diffusion and segregation of some as yet unidentified element. takes place which lowers the grain boundary fracture stress. There is, however, evidence that the ductile-brittle transition temperature of some metals can be lowered by small amounts of deformations and the above fracture transition might also be accounted for in terms of the progressive removal of the deformed structure on annealing. (b) The second change in fracture mode occurs at the alpha/beta transformation temperature; above the transus fracture is predominantl~~ transcrystalline cleavage. Beta-treatment, with furnace cooling, results in a very large grain size and it is considered that this gives rise to an increase in the ductilcGrittle transition temperature in uranium, as is the case with other metals5. Figure z indicates that significant softening begins after annealing between zoo’ and 300°C’ and this is associated with increased necking which causes the rise in the reduction in area curve, and a small increase in the total elongation (Fig. I). After annealing at 42oC necking starts earlier so that it actually reduces the total elongation. After annealing at 500°C recrystallization is coml>lete and this fine-grained homogeneous structure results in a high ductility. However, this heat treatment also lowers the intercrystalline fracture stress and thus, although a high uniform elongation is obtained, fracture occurs without necking and the reduction in area curve starts to fall. \{‘ith grain growth ductility is reduced and both the elongation and reduction in area curves fall. Beta-treatment results in a considerable softening and a hcterogeneous coarse-grain structure but cleavage fracture now predominates and there is a consequent abrupt fall in ductility. The abrupt fall in fracture stress at the betatransus (Fig. 3) is explained by the large increase in grain size giving an effect similar to that observed in iron6 and in zinc 7. This grain size effect also accounts for the progressive fall in fracture stress between 450c-65o’C. The observed peak in the fracture stress curve around 350°C may be related to the banded nature of the microstructure of the specimens. XCKNOWLEDGEMESTS
The authors arc indebted to the U.K.A.E.A. for financial support and to Professor W’. HUME-ROTHERY, Isaac Wolfson Professor of Metallurgy in the University of Oxford, for the laboratory facilities made available. Thanks are also expressed to Mr. B. R. BUTCHER of A.E.R.E. for helpful discussions.
1 X.N. A.UE LIRANATI AND S.V.DETANIS,CO?@~. Kend.,z_s_j (19) (1962) 2435-2437 2 1;.G. FOOTE, Progr. h’ucl. Energy, Ser. \‘, I (Igj6) 99. 3 1). M. R. TAPLIN AND J. W. MARTIN, I. h’ucl. k’ater., IO(Z) (19631 zrg-225. 4 I,. GRAINGER, Uiranium and Thorium, George Kiewnes, London, 1957, p. IOI. 5 I).MCLEAN, Grain Boundaries in Metals, Oxford University Press, 1957, p. 307. 6 N. J, I'ETCH,J. Ivan and Steel Inst. (London), 174 (1953) 25. 7 G. \2:. GREENWOOD AND A. G. QUARRELL,~. Itzst. JL’etals,82 (1954)551.