By E. Creutz, Head, Department of Physics, Carnegie Institute of Technology
and D. Gurinsky, Head, Division of Metallurgy, Brookhaven National Laboratory”
Peres sas of carly interest to the
nuclear physicist working on the fission
chain reaction for several reasons. It has an
extremely low capture property for the ther-
mal neutrons which produce the fission of
uranium. Being of rather small atomi
weight, it is useful as a medium for slowing
down the fast neutrons from fission (neu-
trons having energies of about one to two
million electron volts) to thermal energy
(about 1/40 electron volt), at which energy
Fabrication of
Beryllium Metal
level they have a large probability of pro-
ducing fission in uranium. Elements of
greater atomic weight simply scatter off the
neutrons without taking up much of their
energy. Also beryllium is of interest because
of its rather good resistance to corrosion.
Unfortunately, very little was known
about the fabrication possibilities of beryl-
lium in 1944. Casting techniques had been
developed by the Brush Beryllium Co. and
others, but attempts at forging and extrud-
ing had not been very successful because of
its poor duetility.
Extrusion — The first successful attempt
to extrude beryllium took place at the Wol-
verine Tube Div., Calumet and Heela Copper
Co. in May 194, with the cooperation of
J.P, Schumar and J. F. Rodgers. Since it
was believed that failure on previous attempts
might have been due to insufficient pressure,
a special container and die were made to take
a Jin, diameter billet on which the full
600-ton force of the press could be applied
if necessary.
‘The beryllium billets were enclosed in
covered graphite crucibles and heated in an
air-filled furnace. After heating to 1900? F.
for as long as 10 hr., little change in the
¥*This work was carried out while the authors were
members of the Metallurgical Laboratory, University of
Chicago, operating under the Manhattan Project,
METAL PROGRESS; PAGE 82
beryllium surface took place other than some
small yellow spots indicating very sligh:
carbide formation, Immediately before es-
trusion, the container was lined with a mag-
nesium eup which was to melt and serve as
lubrication, since heryllium does not alloy
readily with magnesium. By this technique
rods were successfully extruded down to *:
in, diameter, It was found that a thin liner
of graphite was at least as satisfactory as the
magnesium, This broke up during extrusion
but served to keep the beryllium from alloy-
ing with the stecl of the die and container
Pressure required for a nominal reduc-
tion in area (about 16 times) was about
60,000 psi. for a billet of 4% in, diameter
sheathed in yj, in. of graphite. Higher pres-
sures were required with smaller containers
and smaller billets. Best results were ob-
tained with the billet temperature in excess
of 1995°F. although temperatures down tc
1650° F. were used.
‘The good lubricating properties of the
graphite liner are shown by the excellent
flow properties of the beryllium. billets.
Figure 1 shows that the flow lines start
almost perpendicularly from the container
wall (billet surface), indicating relatively
small surface frietion. ‘This condition exists
to some extent even inside the die, provided
that the included angle of the die is not tov
large. This does not happen, however, in a
flat die nor one in which the included angle
is as large as 160°.
A result of poor lubrication, and prob-
ably also of too high billet temperature, is
the rattlesnake effect on a billet that was
extruded at 2055°F,, Fig. 2. Since the metal
for the billet was not pure (it may have
contained as much as 0.1% aluminum), hot
shortness was to be expected.
‘Tubing was extruded using drilled billets.
Several I-in. id. tubes were made without
damage to the mandrel, but a %-in, mandrel
was pulled in two. However, this fact does
not necessarily te a limitation on the
size of tube which can he extruded. A lower
billet temperature could probably be used
with sufficient pressure, which would result
in less weakening of the mandrel.
A bell-mouth die appears desirable for
good extrusion of beryllium, If the ineluded
angle is too large (120°), there is a tendencyExtrusion of Beryllium
sion—that is, a 4%-in.
billet was extruded to 2
in. and then to qin.
‘Tubes extruded with small
area reduetions (2 to 8
times) tend to split lon-
gitudinally while cooling.
Tubes are satisfactorily
extruded if the billet is
provided with a hole, and
the reduction is 8 or more
times. Cans may be made
Fig, 1—Unextrnded End of Beryllium by impact extruding
Billet Showing Flow Lines That Meet around a mandrel. Here
Container Wall Almost Perpendicularly also, if the reduction is
small, cracking may oceur.
for the rod to be pulled apart and split lon- Beryllium containing 1% titanium or 3%
gitudinally. The large radial compressive aluminum ean be extruded under similar con-
force supplied by an included angle of 90° ditions. The 3% aluminum alloy has a greater
or less gives better results, However, with — tendency to give poor surfaces than the pure
high temperature and hot dies, good results metal or the titanium alloy.
have been obtained, even with the 120° dies. Forging — In order to improve the strue~
‘The combination of high temperature ture of cast beryllium and thus improve
and high pressure required poses a real die physical properties such as malleability and
problem; nothing better could be found than ductility, experiments were performed on the
Fig. 2— Rattlesnake Effect in Extruded Rod
18-4-1 high speed steel treated for
maximum toughness and hardness. + Fig. 3— Ultimate Tensile Strength of Be-
With proper lubrication and by ryliium as a Function of the Logarithm
extruding the billet while in a thin ip Eee abe 1 oF he patio ot the Pile
graphite jacket, abrasion of the die 7
is Jess than with uranium and TT
other difficult materials. A problem ot +—
is the occasional cracking of the
hard die,
As the reduction in area dur-
ing extrusion is inereased, both the
yield point and tensile strength of
beryllium are increased. ‘This may
he seen in Fig. 8 where the average
ultimate tensile strength of several
samples is plotted as a function
of twice the logarithm to base 10
of the ratio of the billet diameter
to the diameter of the extruded rod.
The last point at 512 times redu
tion corresponds to a double extru- Log.P
g
s
s
Tensile Strength, Psi.
,
re
S
DECEMBER 1952; PAGE’ 83Forging of Beryllium
forging of beryllium at Westinghouse Electrie
Corp., Bloomfield, N. J. Three centrifugally
cast billets 2 in. dis and 4 in, long were
enclosed in soft steel jackets % in, thick,
Inasmuch as previous experiments showed
that above 1920’ F, a molten alloy is formed
if iron is in contact with beryllium, a layer
of caleium oxide was tamped between the
jacket and the beryllium, and a steel cover
was welded to the jacket to completely con-
confined samples (Fig. 4) were heated to
1920” F. and forged radially under a number
of strokes; reheating was necessary to main-
tain the desired temperature. ‘The partially
confined steel-jacketed samples were forged
longitudinal
Welding — Attempts to weld beryllium
by means of the atomic-hydrogen are and
flux, the argon-shielded a.c. arc, and the
helium-shielded dc. are have been unstecess-
ful. Straight polarity (tungsten cleetrode be-
ing negative) as well as reverse polarity
Fig. 4— Beryllium Billet
fine the billet. Three billets were also pre-
pared in a similar manner except that no
cover was put on the steel jacket; some un-
jacketed beryllium billets were also forged
in these tests.
Results of the experiments indicate that
with an unjacketed billet, the best product
was obtained if the billet before forging was
somewhere between 2100 and 2190° F. and the
billet was forged with one stroke. Repeated
blows resulted in a number of radial cracks
starting from the outer edge and working in-
ward. Too low a temperature resulted in
complete cracking of the billet. ‘The totally
helium and argon-are welding methods are
successful.
Conclusion — It has been fou
to fi {¢ beryllium metal by ext
forging. With the first method it is useful
to jacket and lubricate the billet with a thin
shell of graphite, and with the second, it is
d possible
sion and
important to confine the billet with a mild
steel jacket to prevent edge cracking. Me-
chanical properties —in particular ultimate
tensile strength—are appreciably improved
over those available in cast beryllium after
treatment by either of these fabrication
methods. eS
Beryllium
Beryllium and some of its compounds
are extremely toxie when taken into the
lungs or blood stream, This toxicity is
highly specific, varying greatly with indi-
viduals. Sever
I deaths traceable to the
handling of beryllium metal powder and
powdered beryllium compounds have oc-
curred during the past few years. There-
fore, extreme precautions should be taken
Poisoning
in handling any of these, particu
carrying out operations where dust or
fumes are generated. Handling of the
metal with gloves or machining only under
adequate fluid Iubrication and with the
operator wearing a respirator are highly
recommended. ‘The effects of beryllium
poisoning frequent
ent until several years after exposui
do not become appar-
‘AL PROGRESS; PAGE 8¢