oxidizing atmosphere occurs by the dif- 
fusion of the more mobile, excess oxy- 
gen ions, whereas in a reducing atmos- 
phere a plastic or viscous flow model is 
probable. The effect of excess oxygen 
on the sinterability of UO2 was well 
demonstrated in Scott and Williams’s 
work on warm-pressing of UO: (49), 
where it was found that fine, nonstoi- 
chiometric powders could be compacted 
in metal dies to densities above 10 gm/ 
em* in 10 min at 800° C and 20,000 psi. 
When the excess oxygen was removed 
from the UO: lattice by the addition of 
powders such as iron or uranium, the 
densification was inhibited. 
The steam-sintering of wet-ball- 
milled UO: has been studied by Aren- 
berg and Jahn (50). To achieve high 
densities, the compacts were heated in 
H, to 1,400° C before introducing a 
steam atmosphere. Chalder (40), on 
the other hand, has reported that, with 
ADU-type oxide, the complete cycle of 
heating, sintering at 1,300°-1,400° C 
and cooling can be carried out in steam. 
The resulting pellets of 10.5-10.6 gm/ 
cm density usually are near UOo.15 in 
composition. 
The major difficulty with low-tem- 
perature sintering is that the product 
is nonstoichiometric, unless a hydrogen- 
cooling step is introduced. Stoichio- 
metric oxide is preferred as a reactor 
fuel for reasons that will be discussed 
in the later section on irradiation be- 
havior. If hydrogen-cooling is a proc- 
ess requirement, it is not clear that 
there is any economic advantage over 
sintering in a continuously stoked fur- 
nace in a hydrogen atmosphere at 
1,600°-1,700° C. 
Although it may be fortuitous, most 
manufacturers prepare UO: pellets 
from a ‘‘ceramic-grade” powder of 
high surface area in continuous rather 
than batch-type furnaces and in an 
atmosphere of dry hydrogen or cracked 
ammonia. Runfors et al. (1, 52) have 
reported, however, that the sintering 
temperature can be lowered about 
100° C if the hydrogen is first saturated 
with moisture at room temperature, 
although the effect of the moist atmos- 
phere on furnace life has not been 
established. 
Sheathing 
The important requirements for a 
fuel-element sheath are low neutron- 
capture cross section, good mechanical 
properties at working temperatures, 
high corrosion resistance in the hot cool- 
FIG. 5. Zircaloy-2 clad UO rod (0.56- 
in. diameter) from CR-V-e test after ir- 
radiation of 7,000 Mwd/tonne U with 
surface temperature of 400° C 
ant and compatibility with the core. 
For economic water-cooled reactors 
fueled with UOs, a zirconium alloy 
would appear to be the most suitable 
material. Stainless steel would be ac- 
ceptable but for its high neutron-cap- 
ture cross section. Aluminum-nickel 
alloys are not likely to compete with 
zirconium because of lower strength 
and lower neutron economy (53). For 
sodium-cooled systems, also, zirconium 
alloys may prove to be the best choice. 
In organic liquids, however, zirconium 
alloys are attacked by the hydrogen 
that is liberated when the coolant is ir- 
radiated. The most suitable material 
immediately available for organics at 
temperatures near 400° C is stainless 
steel. Stainless cladding is also likely 
to be used in several CO:-cooled reac- 
tors now being designed, at least until 
beryllium technology progresses. 
~ The most common design of UO: fuel 
element is typified by the first loading 
for the Shippingport reactor (41). The 
fuel assemblies consist of a bundle of 
small round rods, each rod enclosing a 
stack of right-cylindrical sintered pel- 
lets. The clearance between the pel- 
lets and the sheath is so specified that 
the pellets can be loaded easily into the » 
tubes but will expand into near contact 
with the sheath when heated in the re- 
actor. Other geometries being devel- 
oped include a nested tubular type for 
the Plutonium Recycle Test Reactor 
(37) and a plate type for the second 
loading of the PWR (41). A rod-type 
element was chosen for the first loading 
of NPD-2 because of the satisfactory 
results that had been obtained from ir- 
radiation tests on such a geometry and 
because of the inherent stability of a 
cylindrical sheath when subjected to an 
internal pressure. 
The effect of external pressure on 
sheath stability is of major concern, 
particularly in water-cooled reactors 
operating at coolant pressures of 1,000— 
2,000 psi where the fuel has a diameter/ 
sheath-thickness ratio greater than 
30/1. In NPD-2, for example, where 
the pressure is to be 1,100 psi, the pro- 
posed 1-in.-diameter, 0.025-in.-thick 
Zircaloy-2 sheath will deform plasti- 
cally at the maximum operating surface 
temperature of 300° C, forming ridges 
and dimples, if the fuel-sheath clear- 
ance is greater than 0.006 in. in diam- 
eter or 0.15 in. axially. To ensure 
accurate control of the diametral clear- 
ance, the UOz pellets will be centerless 
ground to +0.0005 in. before loading 
into carefully dimensioned sheaths. 
In addition, it is probable that an axial 
clearance of 0.2 in. will be distributed 
along the 18-in. fuel length by fabri- 
cating pellets with concave ends. 
End caps are usually attached to 
UO: fuel elements by fusion-are weld- 
ing in an inert atmosphere such as 
helium or argon. Resistance seam- 
welding combined with a Heliarc fusion 
weld is being developed to provide a 
double closure for large-diameter, thin- 
walled tubes (37). Electron-beam weld- 
ing has also been.studied as an end- 
closure method (64, 66). Since the 
latter process is carried out in vacuum, 
however, the resulting fuel element 
could have poorer heat-transfer proper- 
ties if the oxide does not expand to 
contact the sheath when irradiated. 
If the fuel elements are short, such 
as those used in PWR (41), only end 
spacers are required in the assembled 
bundle. For longer assemblies, inter- 
mediate spacers are normally incorpo- 
rated. These may be wires wound in 
a spiral around the rod, ribs integral 
with the sheath, short lugs welded on, 
or metal grids similar to the type de- 
scribed by Ambartsumyan et al. (56). 
' Irradiation Behavior 
Irradiation tests on prototype PWR 
elements clad in Zircaloy-2 have been 
reported at length (57). More re- 
cently, many UO: fuel tests in the NRX 
reactor have been reported. Thus, 
within the past two years, a much 
clearer understanding has been ob- 
tained of the irradiation behavior of 
UO,. <A brief summary of irradiation 
experience, including results from tests 
just completed at Chalk River, follows. 
Effect of heat rating. Lewis (59) 
has pointed out the economic advan- 
63 
