tation rate or different pH yielded a 
UO: product that sintered to a lower 
density. When the precipitation was 
carried out by continuously mixing two 
liquid streams, however, the only vari- 
able that affected pellet density appre- 
ciably was the pH of the mixed solu- 
tions; the optimum pH was 7-8. For 
conversion to UOs, thé ADU was loaded 
in trays to a depth not exceeding 2 in. 
and reduced in hydrogen for 1 hr at 
900° C. The resulting pyrophoric 
powder was cooled to room tempera- 
ture in hydrogen and stored in carbon 
dioxide for 1-2 hr, after which it could 
be handled and stored in air. After 30 
days’ storage in air, the O/U ratio in- 
creased to 2.25. 
Several tons of UOz powder have 
been produced by Eldorado Mining 
and Refining by both batch and con- 
tinuous precipitation. Sintering tests 
for 1 hr at 1,625° C on many samples 
pressed at 40,000 psi have consistently 
produced pellets of 10.4-10.6 gm/cm*. 
FE RE SEN SERB 
; TABLE 2—Sinterability of UO2 (40)* 
Sintered 
Source of density 
UO, Preparation (gm/cm$) 
Mallinck- Pyrolysis of 7.8 
rodt uranyl nitrate 
hexahydrate 
(UNH) to UOs;, 
H, reduction 
Shattuck Believed similar 8.6 
to Mallinckrodt 
National Believed similar 9.1 
Lead to Mallinckrodt 
Eldorado Pyrolysis of 9.2 
Mining UNH to UO;, 
& Refin- H, reduction 
ing 
Eldorado Pyrolysis of 10.3 
Mining UNH to UOs, 
& Refin- UO; hydrated by 
ing wet-ball-milling, 
H, reduction 
U.K.AEA, Precipitation of 10.6 
Spring- ammonium di- 
fields uranate (ADU) 
from UNH, H:2 
reduction 
Mines Precipitation of | 10.1-10.7 
Branch, ADU from UNH, 
Ottawa Hz reduction 
AECL, Precipitation of 10.4-10.6 
Chalk ADU from UNH, 
River H, reduction 
* All pellets cold-pressed at 40,000 psi 
and sintered for 30 min at 1,700° C in 
hydrogen. 
62 
Although UO, pellets of the quality 
required for fuel elements can be pre- 
pared from ‘‘ceramic-grade”’ powders 
at relatively low compacting pressures 
and sintering times, it is still not certain 
whether these will be less expensive 
than pellets made from less sinterable 
powders. For example, the Eldorado 
Company is marketing ‘‘standard- 
grade”? UOsz, prepared by the hydrogen 
reduction of UO;-2H.O in a moving- 
bed furnace, for $1.15/lb less than 
ADU-type UO2 (44). A careful cost 
analysis for each specific fuel geometry 
would be required to indicate whether 
the cost advantage of such cheaper 
powder would be offset by the larger 
investment in compacting equipment 
and lower furnace throughput. 
ADU-type UO: powder, made as de- 
scribed, offers some advantage for more 
economic fuel production in that it can 
be cold-pressed without an organic 
binder. It is necessary to apply a lub- 
ricant such as stearic acid to the die 
walls during the pressing operation to 
obtain crack-free compacts, which can 
then be charged directly to the sinter- 
ing furnace without the presintering 
treatment normally used for binder 
removal. 
Method of compaction. The most 
common method used for preparing 
UO, compacts is dry-pressing in hard- 
ened-steel or tungsten-carbide dies. 
Usually, an organic binder such as par- 
affin wax, polyethylene glycol or cam- 
phor is added to increase the green 
strength of the compact during subse- 
quent handling. A small amount of 
stearic acid, 0.2-0.4 wt%, is often 
added as a die lubricant (39). 
The extrusion process (37, 45) may 
offer some advantage where a large 
length-to-diameter ratio is required or 
for fabricating large-diameter tubes of 
UOs:, but no fabricator has yet chosen 
the technique in preference to auto- 
matic dry-pressing. Other possible 
processes such as slip-casting, isostatic- 
pressing and hot-pressing have also 
been rejected up to now. 
The compaction of UO2 powders in 
metal sheaths by rotary-swaging has 
been investigated extensively at Han- 
ford (37, 46) and at Chalk River (39, 
47). In both laboratories it was found 
that, of the many powders evaluated, 
arc-fused UO: could be compacted at 
room temperature to the highest den- 
sity, about 10 gm/cm’, in either 
Zircaloy-2 or stainless-steel sheaths. 
Much lower densities were obtained 
when powders with high surface area, 
such as ADU-type UO», were used. In 
the Hanford experiments, higher densi- 
ties resulted when the oxide was swaged 
hot at 600° C (46). 
At first sight, swaging promises a 
cheaper method of fabrication than 
the conventional sintered-pellet route. 
However, the swaging of fuel elements 
with a high ratio of diameter to sheath 
thickness may not be economically 
feasible. It seems inevitable that the 
trend in fuel-element development will 
be to decrease the sheath thickness to 
the minimum permitted by fabrication 
and irradiation experience. As an ex- 
ample, the diameter/sheath-thickness 
ratio of the UO2 fuel elements in the 
Shippingport reactor is 18/1 (41), 
whereas the target in the Canadian 
NPD-2 reactor is 40/1. Recent tests 
at Chalk River, with a ratio near the 
latter value, have shown that small 
cracks had formed on the inside surface 
of the Zircaloy-2 tube during swaging 
after a reduction in the element cross- 
sectional area of only 20%. One such 
element, which had been reduced 48% 
in area at 400° C to 0.8-in. outside di- 
ameter and 0.025-in. sheath thickness, 
was irradiated in a pressurized-water 
loop in the NRX reactor and split open 
immediately upon reactor startup. 
One end of the swaged specimen, which 
had been cut off, was metallographi- 
cally examined, revealing cracks on the 
inside Zircaloy-2 surface up to 0.002 in. 
deep. As a result of such experience, 
swaged UO: has been rejected as a first 
charge for the NPD-2 reactor. Efforts 
are continuing, however, to determine 
if an economic process incorporating 
intermediate annealing steps can be 
evolved to produce crack-free sheaths. 
The loose compacting of UO2 powder 
in a fuel sheath is simple in concept, 
but, thus far, no method has been found 
to exceed a packed density of 8-9 
gm/cm’. 
Sintering method. Ceramic-grade 
UO, can be sintered to densities above 
10 gm/cm? in neutral or oxidizing at- 
mospheres at temperatures 300°—400° C 
lower than in reducing atmospheres. 
In Thackray and Murray’s early work 
(48), for example, it was found that 
with an ADU-type oxide of composi- 
tion UO2.13, after pressing at 20,000 psi, 
a density of 10 gm/cm* was reached 
after 30 min in argon at 1,400° C. 
Belle and Lustman (2, 3) have con- 
cluded from their studies on self-diffu- 
sion kinetics that densification in an 
