-
I OFI
ORNL P
2203

EEEFEEEE
-
-
*
EFE
-
-
-
.
.
..
MICROCOPY RESOLUTION TEST CHART
NATIONAL BUREAU OF STAMOAMOS - 1963

ORNVP

Cowa PRICES
::
Y
2003
H.C. $2.00; MN, Evi
III.1
ORNI - AEC - OFFICIAL
Tungsten and Tungsten Alloy Tubing Development Meeting
:
JUN 27 885
May 26, 1966
Oak Ridge National Laboratory
A Summary of Progress
by
Metals and Ceramics Division,
Oak Ridge National Laboratory
Oak Ridge, Tennessee

RELEASED FOR ANNOUNCEMENT
LEGAL NOTICE
IN NUCLEAR SCIENCE ABSTRACTS
The report was prepared M M account of Govern ponsored work. Mether the United
hatos, for the Comminatam, nor w para netting a ball of the Canudatua:
A. Met my warranty or representation, car d or implied, with roapect to the nece-
ray, couple, whom of the fatormation a nd in the report, or that the wo
of my formation, part, med, at procon declaudia do raport may not being
petrataty owned rule or
. A mig Habilities with respect to the meal, or for meget rood thing from the
wne of my heartom, ruratus, method, o proces declared in the reporte
A u to whore, porno actes a hall of the Comed" medus
pogut coutrusturo Comune, or plagued we contract, tot dat
mele player or auntrietor of the Counterton, ar aplogue of well contractor propers,
damnate, or from now to, ang dormado por to Mo waplog or contract
wta Curteetan, of Memployee will mal contractor.
Work Performed Under
High-Temperature Materials Development Program
for the
United States Atomic Energy Commission
under
Contruct No. W-7405-eng-26
1..jiinOvi- l...',
1:13:35 - 53 - 1.0.
Compiled by
W. 0. Harms
.
.
-,--
'
7
'
,
quod!:
2
.
III.3
) -
itibicowa wisico:
v.
INTRODUCTION
1.10
Fabrication development and evaluation of tungsten and tungsten-base
alloys is carried out as an integral part of the High-Temperature Materials
Program at the Oak Ridge National Laboratory. Two basic fabrication pro-
cesses are under investigation. One involves large-scale extrusion and
drawing operations and the other involves direct production of the compo-
ents by thermochemical deposition of the metal or alloy. The products
are evaluated at various stages of development by metallography, x-ray
diffraction and chemical analyses, mechanical property determinations,
joining studies, and a variety of nondestructive testing techniques. This
repori describes development work in these areas for the period December,
1965, through May, 1966, with emphasis on techniques and results that are
applicable to the production of tubing.
+
EXTRUSION OF TUNGSTEN AND TUNGSTEN ALLOY TUBING
R. E. McDonald, G. A. Reimann, ana C. F. Leitten, Jr.
ORNL Extrusion Facility
TEE
1
The redesigned 3-in, tooling for the ORNL extrusion press proved to
be suitable for extrusion of tungsten and tungsten alloys, and press instru- .
mentation performed as intended. Problems arose, however, with regard to
press response ană ram speed. Changes in the hydraulic system corrected
the problem of press response, but the difficulty with slow ram speed has
not been overcome.
.
V
. Ir
WIN
.
A maximum ram speed of 2 in./sec is available at high pressure and
is suiiicient for extruding aluminum, brass, or steel; ram speeds of at
ieast 5 anå preferably 8 in./ sec are required, however, to extrude tunga a
sien ana, particularly, the tungsten-base alloys. At a ram speed of .
less than 2 in./sec, heat loss to the container and mandrel increases
the yielä strength of the billet beyond press capacity during the extru-
sion. Also, prolonged exposure to the hot billets reduces liner life and
the mandrels fail due to overheating.
.
.
.
:
-
;
-
';
.
:
*
.:
I
.
.
Billet Heater
2
The billet heater has been tested to 2300°C, and billet temperatures
to 2200°C have been employed in actual practice. The original billet
teater enclosure was not sufficiently airtight and the consequent forma-
vion of volatile billet oxide made optical temperature readings difficult.
An improved enclosure was constructed and this problem no longer exists.
&
.
....
17
ST
A substitute for the zirconia pedestal block, upon which the billet
rests in the center of the induction coil during heating, is being sought,
At temperatures greater than 1800°C the billet tends to adhere to the
0!![-BC-01:01:1
0.11 - AOC - OFFICIAL
US
-
11
Follo
E

12
U14
12
III.4
05:11.- ASC-05:101:
P.
27
"
!
pedestal block and on occasion a portion of the block has been carried
along with the billet into the press. Block facings or nafnia or yttria
have been considered, but these materials are difficult to obtain ana
quite expensive. Boron nitride has been effective in eliminating the
sticking problem but it tends to fume on heating and clouds the sight
glasses,
Tooling
2
A device was designed and constructed to hold the mandrel tooling
and graphite follower block to the stem so as to eliminate one transfer
operation and reduce transfer time from 15 to less than 10 sec. This
device is essentially an angle-iron trough fitted with air cylinders; the
backup tooling is used to load the billet into the container when the ram
is advanced and the holder is ejected from the stem when loading is
complete.
Satisfactory service has been obtained from zirconia-coated dies,
and no adverse effects on dies were noted as the result of exposures of
greater than I sec to refractory billets. Some dies machined and coated
by outside vendors have been suitable for more than two extrusions before
recoating was necessary. Others, however, have failed in the relief angle
near the die exit after a single extrusion.
There have been three instances of mandrel failure due to insułfi-
cient ram speed. Efforts are under way to improve mandrel performance
in general by grinding a taper of at least 5 mils in the 8 1/2-in. zir-
conia-coated length. The purpose of the taper is to provide relief
from the extruded material and reduce the tensile stresses generated in
the mandrel as the result of frictional drag between the mandrel and the
emerging extrusion. Experiments with the tapered mandrel will be conducted
with a view toward eliminating the molybdenum follower.
Tube-Shell Extrusions
re
lur
IWO 5 7/8-in.-diam by 12-in.-long billets of unalloyed tungsten,
weighing about 200 lb each, were extruded to 3 1/8-in.-diam round bars
at the E. I. duPont Metals Center. The extrusion temperature was 1550°C
and the base-metal oxide technique was used. Ten 3-in, -diam by 5-in.-long
tube-shell billets are being machined from these bars for subsequent
floating-mandrel extrusion experiments.
1.
2
E
.
Five extrusions of tungsten and tungsten alloys were attempted during
the present reporting period. Three of these were tungsten tube shells,
one was a 1-30 at. % Re-30 at. % Mo tube shell, and one was á molybdenum-
clad tungsten duplex tube shell. The conditions and results of these
extrusions are presented in Table III.1 (extrusions 0003, 0004, 0031, 0034,
and 0037) along with similar information on previous extrusions that were
evaluated during this period.
0011~NEC -- OFFICIEL
0?!!(-:20-
"!!1
.
.
.
.
.
:
*
*.:.'i
TE
t1
IVIDI:10 -
IV - 11880
?
"
,
1
9
.
;
!
--
-
o
*
-
-
TVIJIJIO- JY-INYO
Table III.1. Conditions and Results of Tungsten and Tungsten Alloy Tube-Shell Extrustions
Extrusion Conditions
Billet Dimensions
(in.)
Outside Inside
Diameter Diameter
Extrusion Mumber
Billet
Weight
(lb)
Temperature
(°C)
Reduction
Ratioc
Force (tons)
aximum . Minimum
Speed
Extrusl
Index
(psi)
x 103
0003
0004
0031
2.990
2.992
2.985
0.875
0.875
0.875
700
20.3
20.2
14.9
.
665
465
465
0.7
1.0
100.0
90.0
1560
15931
2.939.
2.8313
17.8
17.4
1.000
Unalloyed Tungsten Billets
1750
6.4:1.
1750
6.4:1
690 ·
1770
4.0:1 550
W-25 at. % Re Billets
2200
5.3:1 . 550
2250
4.8:1 530
1.--30 at. % Re-30 at. % MO Billets
2200
4.0:1
695
Molybdenum-Clad Tungsten Duplex Billets
1650
8.8:1
435
1900
8.8:1 505
475 •
495
10.5
10.3
89.0
97.3
0034
2.9903
1.063
17.6
17.6
0.6
d, e
1637
00370
2.987
2.985
0.8.15
0.747
14.8
14.2
425
395
5.5
1.1
57.6
63.0
Floating-mandrel technique and base-metal oxide lubricant used for all extrusions except, as noted, for extrusion 1560.
"Billets prepared by powder-metallurgy technique except, as noted, for extrusion 1593.
Based on ratio of villet and cont iner cross-sectional areas.
'Incomplete extrusion.
Mandrel failed.
Sheet bar extrusion; glass lubricant.
Swidth of sheet bar.
Thickness of sheet bar.
Arc-cast billet.
"Includes 0.010-in.-thick TCD coating.
es
.
--
HT
*
.
W
1101111--23-1;!: 0
ZŁ
e
27
bsite is ..?..
1:1511:0-3
? - ?;:10
Ənditions and Results of Tungsten and Tungsten Alloy Tube-Shell Extrustionsº
Ex'sruss.on Conditions
-
-
Billet
Weight
(lb)
Temperature
(°c)
Reduction
Ratioc
Force (tons)
Maximum Mini.muin
Speed
Extrusion
Index
(::51)
Tube-Shell Dimensions
(in.)
Outside Inside
Diameter Diameter
x 103
20.3
20.2
14.9
665
465
0.7
1.0
100.0
90.0
1.410
1.707
e
0.735
17.8
17.4
475
Uralloyed Tungsten Billets
1750
6.4:1
700
1750
6.4:1
690
1770
4.0:1
550
W-25 at. % Re Billets
2200
5.3:1
550
2250
4.8:1
530
W-30 at. % Re--30 at. % MO Billets
2200
4.0:1 695.
Molybdenum-Clad Tungsten Duplex Billets.
1650
8.8:1
435
1900
8.8:1
505
10.5
10.3
89.0
97.3
1.9158
495
0.930h
0.981
1.653
17.6
0.6
d, e
1.693
d, e
14.8
14.2
425
395
5.5
1.1
57.4
63.0
1.252
1.257
0.774
0.740
etal oxide lubricant used for all extrusions except, as noted, for extrusion 1560.
techni que except, as noted, for extrusion 1593.
: cross-sectional areas.
........!
1 17
.
W1-CSC-.IT!"!!1

U
III.6
-277-lin
23
an
Lubrication Development
me te ocide ayer formea on w.alloycä uungsten billets auring
Par
i s Oetween cre villet Leater and cheess container has .
OVC Co bé un eclective lubit, the cuerity od Oria Iormed on
billeis om ke 1-25 at. Se sa 1-30 ar. Re-30 át. Mo alloys is not
Suricio or accoüüte lubrication. To overcome this lubrication
Ceri.
com.cy ude ümov billets were coated with approximately 10-mil layers
c. Cuirtea 37 The Chermochemical deposition (TCD) process. Tois layer
Proviâus ä copious supply ož tungsten oxide during billet transier.
ou deposition system shown schematically in Fig. III.1 consists of
Pour Cuisine secüicmai
. ovision for regulation or the reactani gases,
ü cü heater,
3. & vacuum system,
à system for processing the gaseous by-prcüucts.
i
.
021LWS 63-42794
H2O3
G-S PEIXER.
C.
NATURAL GAS --
6:3 PSIHIATOS
ini
H2O
Local/
Curso
W.CVUT: CHA:283
104
Valive
SCRUISER
o
-3!LLES
5017
m.
I
*10'OTANCE CO!
2013:
GALS: -
..
.
BILLET
POSITIONER
-
SCAVENGER
21
in 56 0230
.
30
.
.
.
E VF
..
.
-
.
Ya
,
-
.
.
„
..
0!
-
MECHANICA: VACUUM PUD
7:121:
S
.O
.O
...
::...
:
C
a
.
..otor
T
;-:
Cuisine.X
.
S
.
-1:170
istih

III.7.
OHL-::-05:!clill
2
3
In this system, WF6 is premixed with hydrogen and is passed over the sur-
face of the billet to form an adherent metal coating with the subsequent
liberation of the ga seous by-product.
Thxee alloy tube-shell billets have been coated under the conditions
shown in Table III.2. Coating thickness uniformity was obtained by rais-
ing the billet relative to the induction coil in l-in. increments
during the deposition process.
-....
.
Table III.2. Conditions for Coating Tungsten-Alloy Extrusion
Billets by the TCD Process
Billet
Numbe
Composition
(at. %)
Flow Rate or
Reactant Gases
(cm3/min)
WF .. Ha
Billet
Temperature
(°C)
System
Pressure
(torr)
1592
1593
0034
80
W-25 Re
W—25 Re
W-30 Re-30 MO
115
1.15
115
115
3000
3000
3000
550
580
3000
550
170
-
-
-
-
Unalloyed Tungsten
Ivo billets of unalloyed tungsten were available for tube-shell extru-
sions. Because of slow billet transfer and slow press response, more than
60 sec elapsed before high pressure was brought to bear on the first bil-
let (extrusion 0003) and the press stalled. This billet was recovered
and remachined for subsequent use in extrusion 0031. Although transfer !
time and press response were improved somewhat during extrusion of the "
second billet (extrusion 0004), the mandrel. failed in four places. This ,
tube shell is shown in the bottom of Fig. III.2. The billet salvaged .
from the first attempt was transferred in 15 sec and produced a satis- !
factory tube shell (extrusion 0031), shown at the top of Fig. III.2.
- - - - - - - - - - - - - - ----- - - -
W-25 at. % Re Tube Shells
I.
AI
Five W-25 at. % Re billets were received from the Albany Research
Center, U.S. Bureau of Mines. Three or these billets were fabricated
by powder-metallurgy (PM) techniques and two were arc cast. A repre-
sentative microstructure of the PM billets is shown in Fig. III.3a.
in Fig. III.3b for comparison and is discussed later. The numerous
voids present in the microstructure are common to PM products and are
:
OPNL-i! - OrrICIAL
T

UR
-.
-
E
+
-
.
.
-
-
-
-
-
-
. **-***--...
.
-
M
-
-
r
-
!
.-
..
-.
-'
„
.
.
.
'
-
=
CMS
1
2
OAK RIDGE NATIONAL LABORATORY
3
5 6 7 8 9 10 11 12
..!!!! ! !luluillud
*
CV
Fig. III. a
A
Fig. III.2. Tube-Shell Extrusions of Unalloyed Tungsten. Tube shell at bottom (extrusion 0004)
contains the failed mandrel. Tube shell at top (extrusion 0031) is satisfactory. (Section at right
of extrusion 0031 is the extruded molybdenum follower block.) Original photo reduced 9%.
OPNL-HCC - OFICIAL
"
Orrib- AEC - OFFICIAL
S
i
Teri
EZ
ii
-".
. 1. ir


1
.
+-;..
..'-
reduce
3:1213.3-2!--11
-
r
XOOX
---
.
-

pad
TN 100x
;
.
.
.
-
-
Fig. III.3. Microstructures of W-25 at. Re and 130 at. % MO
Powder-Metallurgy Billets. Murakami's Etchant. 100x. (a) W-25 at. % Re
alloy. (b) W-30 at. % Re-30 at. Mo alloy.
OKUL-MC-Oi!CIAL
St
14.0.
****
P
III.10
*
3V - 1:!: 2
normally removed during extrusion. Nose cones of TZM were electron-beom
weided to the PM billets, which are now ready for TCD tungsten coating.
The arc-cast billets were submitted for electrodischarge machining of the
center portion for accommodation oñ' the mandrel. These billets will also
be provided with nose cones and coated with TCD tungsten.
.
..*.
**
27
..--..
-
..
rei
--
:
-.
.
.
-'--
Previous extrusions of PM and arc-cast billets were compared metal-
Lographically in the as-extruded conèition and arter various annealing
heat treatments to 2000°c. Representative microstructures are shown in
Fig. III.4. Tube shell extrusion 1593 was produced from an arc-cast
biliet eated to 2250°C and extruded at a reduction ratio 0I 4.8:1, while
sheet- extrusion 1560 was produceà Irom a FM billet at 2200°C at a
reuuction ratio of 5.3:1. In the as-extruded condition, most of the PM
products was recrystallized while very little recrystallization occured in
the arc-cast material. A l-hr vacuum anneal at 1700°C was sufficient to
compiete the recrystallization in the FM Extrusion, but I hr at 1800°C
was recuired to remove the last traces of wrought structure in the extru-
sion iron. the arc-cast billei. Ariztealing at higher temperatures produced
markea zein growth in the arc-cast material but no noticeable growth in
the PM sample. Impurities may have promoted nucleation and inhibited
grain growtà in the PM material.
...-=-'
..
.
"
-
-
.
-
-
.
W-30 at.
Re-30 at. % Mo Tube Shells
. .
. . .
. . . .
. . ."
in singie billet ol W-30 at. Re-30 at. % MO was available for extru-
sion. A TZ nose cone was electron-beam welded to one and on this PM
billet, a TCD coating or tungsten was applied, and the billet was iinish-
moura co remove the wela bead and to ensure that the TCD coating was
üniform. This billet (extrusion 0034) was heated to 2200°C and during
üzensier to the press a section of the zirconia pedestal about l-in thick
aurerea to the base of the billet but went unnoticed. The normal trans-
iex time of 15 sec was extended to 38 sec due to difficulties in penetra-
ting the zirconia with the mandrel. Even with the breakthrough load of a
near maximum 695 tons and a ram speed of less than I in./ sec, however,
sufficieri material was extrued to yield a section of usable tube shell,
ază material for metallographic examination. The partial extrusion is
som in Fig. III.5. A considerable portion of the high extrusion loed
is attributed to the presence of the zirconia block which generated high
Iricional forces when particles of zirconia became lodged between the
foilover material and the container liner. The obvious scoring of the
molybdenum follower and the dummy block (hardness, Rockwell C 52) is cited
as evidence of this effect. The unextruded portion of this billet wes
salvaged by remachining ana will be electron beam welded to one of the
relatively short arc-cast W-25 at. % Re (described in ühe previcus section)
in place of a TZM nose cone for use in a subsequent tube sheli extrusion.
Attri
*** FE
L IETOA
. .
II
*..
a
Three additional billets of the terrary alloy were received irom the
Albany Research Center. Nose cones of TZ have been welded to these
billets and they are awaiting application of TCD tungsten coatings. As
shown in Fig. III.6, the microstructure of this billet is nearly identical
to that of the W-25 at. % Re PM billet.
1:5211:0-DIV-INO

•
Powder Metallurgy Billet
Arc-Cast Billet
119:110- 237 - INFO
chritte EC-OFICI!!.
-
--
-
a
:)W
ios
.
v
om
ne
il
C
.::
coba..
mmmm
Que
-0.035 INCHES
xooi ny
DO
.
.
.
.
.
A
O!:HL-REE - OFICiiil
L
Fig. III.4. Comparison of Microstructures 01 As-Extruded and Annealed
Tungs, from Powder-Metallurgy Billet (Extrusion 1560) anà Arc-Cast Billet
(Extrusion 1593). Murakani's Etchant. (a) As extruded (2200°C),
(o) annealed 1 hr at 1700°C, (c) annealed 1 hr at 2000°C, (a) as extruded
(2250°C), (e) annealed 1 hr at 1800°C, (f) annealed i min of pnnnon
w211.

. 11101970-
+
III. 12
081l - AEC - OFFICIAL
:30
fiz. Ils
وہ کہہ کی -
F:a. III.5. Partially Extruded W-30 at. % Re-30 at. % Mo Tube Shell
(Extrusion 0034) with Zirconia Block and Failed Mandrel.
N.
Mandrel Type Duplex Tube Shell Extrusions
!
The design and extrusion of duplex billets have been reported previ-
Ousiya. The first duplex billet extruded (extrusion 1637) contained an
inner molybäeriun liner thich insulated the tungsten tube shell from the
mandrei. During the present report period, a tungsten tube shell was
estruccà by the duplex technique but no inner molybdenum liner was used
(extrusion 0037). Although slow cam speed caused mandrel failure a
12-in. jengin of usable tungsten tubing was obtained ahead of the first
randrel separation. This extrusion is shown in Fig. III.6 alongside a
positive radiograph that was made so that usabie tubing could be located.
Mazarel failure is not evident from the external appearance of the extru-
sion but four separations may be easily observed in the radiograph.
The mandrel section was removed from the forward portion of the
extrusion (right side oi Fig. III.6) and the outer molybdeaum covering
was leached away. A crack was initiated where the tubing shrank around
tre mandrel section during cooling and propagated axially during dissolu-
tion oż the molybdenum.
•
F.
The surfaces oz this material after dissolution of the molybdenum
are compared wiön those oi the previous auplex product in Fig. III.7.
.
ORK-AEC - OFFICIAL
UCL
WW
w
SP
-2. E. McDonald and G. A. Reimann, High-Temperature Materials Program
Quart. Progr. Rept. Oct. 31, 1965, ORNL-TM-1350, pp 5–8. (CLASSIFIED)
0
- TZM NOCE
STUDE SHELL EXTRUSIOI!
.
F-FAILED ZrO, COATED ANDREI
SORTION O. ¿O, FURNACE DETAL
S MOLTODENUM FOLLOW
r-GRAPHITE FOLLOI?
pDUI. DIOCK
EMES
HECHOS
Gum
:1:
OAK RIDGE NATIONAL LABORATORY
3 4 5 6 7 8 9 10
is
12
01.11-;{C-(SICIAL
Ada
.
.
.
1. -
3
E'
.
0 :::!1 -.[C-OPTICI:!
.
.
.
-
-
2
:



home.
com
... -
-
-----
0 :::!- filc - OFFICI:!.
111.13
III. 13
ORAL ALC - C:FICIAY.
.

AZ
.
Scozure TIL. 6
Photo 82663
Fig. III.6. Duplex Tungsten Extrusion 0037 with Positive Radio-
graph Showing Section or Usable Tubing and Location of Mandrel Failures..
Both the internal and external suriaces of the tube extruded without
the molybaer.i linel' are clearly superior, indicating that extrusion of
tungsten äirectly against an unyielding mandrel surface may be preferred.
Microstructures of these extrusions are also compared in Fig. III.7.
It is agerent that direct contact with the mandrel had a chilling effect
in extrusion C037 and an essentially wrought structure was obtained; the: ......
other hari, allowed complete recrystallization of tungsten.
do third duplex billet containing a W-25 at. % Re tube shell section
arom entrusion 1593 has been prepared. Extrusion of this billet has been
deier: cá until the ability to obtain the necessary ram speed has been
demonstrated.
WOCHEMICALLY DEPOSITED TUNGSTEN AND TUNGSTEN ALLOYS
Farication or tungsten and tungsten alloy components by the TCD pro-
cess oiiers several potential advantages over more conventional fabrication
techniques. These include lover processing temperatures, relatively simple
and inexpensive equipment, high purity, and minimal finishing operations
Tor both regular ana irregular shapes. The TCD techniques can also be used
for joining refractory metals and alloys. Current studies on thermochemical
deposition and evaluation oi tungsten and tungsten-rhenium alloys are
described in this section.
Deposition Studies
urgsten
(F. 7. Patterson, W. C. Robinson, and C. F. Leitten, Jr.)
Deporting early imgsten deposition studies at ORNL involving hydrogen
câücuiba 0:2 Wa', small changes in ieposition conditions often caused
wüily observable changes in chickness distribution and grain morphology
Oz the deposits. Similar effects occurred during scale-up of the
ORHI-KEC - OFFICIAL
.

نما
ارد
. .
درندے -
من
.
ه
*
:
۴
:
نی
رزبن
:.
:
.:
.
3
:
44
م
. . . . . . .
=
ة
. .
"
.
.
...
.
.
.
.
.
. مرد :
:
: د . م
*د
-
غ
ه مه
داد
..
...
. . . .
.
مے
.
.
.
.
.
037
]
..
... .......
...
ر
-
-
.
....
ا
-
جد
ا
ا -
.
-
=
=
.
.
-
-
-
ر
..
ا ا
(
ه
"
رد
ب
با
یے
ہے۔
م / ز ب
- در .67
نار
- liisa

-
S
.
limis,ni);
.
III.15
115101319-- 18:0.
it
deposition process for production of sheet material for mechanical prop-
crties evaluation. These results indicated that a basic understanding
ci facüors affecting deposition rate, purity, and grain morphology was
required for successful application of the process to various geanetrical
Sapes and sizes. In recent months, therefore, tungsten deposition
studies have been directed for the most part towards an understanding of
these factors.
The first part of this study involves a correlatica of deposition
rate with temperature, pressure, and composition of the reacting gases. ".
Dhe experimental program has been planned statistically with the aid of
a computer and requires 81 experiments divided into three phases of 27
units each. Phase I or the study is a general survey covering the range
of experimental parameters listed below in large increments.
Temperature, °C
Pressure, torr
Tota) flow rate, cm3/min
WT. i'low rate, cm3/min
· 450 to 820
5 to 20
750 to 6010
50 to 380
Scenas indicated by results of experiments in Phase I will be reevaluated ::
in Phase II which covers the middle portion of the ranges cited above in
Greater detail. Phase III covers remaining intermediate points and the
extremes on these ranges.
In the experimental portion or the çrogram, tungsten is deposited
inside 0.81-in.-diam copper tubes heated uniformly over a length of 15 in.
Laickness variations are measured by an x-ray attentation method in nine
increments along the length of each tube. Developer work in connection
with täis nondestructive evaluation is presented later in this section.
The experimental portions of Phases I and II have been completed and
Phase III is in progress.
In futuro stucies, an attempt will be nađe to correlate gas stream
characteristics and deposition rate. A second computer program has been
writüen for calculating composition, I low rate, and Reynolds number or
the gas stream as a function of distance along the deposit. Information
derived from this study should be helpful in elucidating the nucleation
and growth mechanisms of TCD tungsten.
Cungsten-Rhenium Alloys (J. I. Federer and C. F. Leitten, Jr.)
Tungsten-rhenium alloys are deposited by hydrogen reduction of
W76-Re7's mixtures. In continuation or previous studies, tungsten-rhenium
alloys were deposited on the inner surface of type 304 stainless steel
tubing in a moving induction-heated hot cone. Alloys deposited on the
stainless steel, however, invariably contained circumferential ridges
with a nodular grain structure. These riâges have been attributed to a
0 ::!|-;:( - OFFICIAL
**
4
III.16
--
-
-
-
reaction between the stairless steel and cheriwm fourides becauce waer.
plair-carbon steel mandrels were used in subsequent tests the ridges no
longer occurred. The earlier experiments were useful, however, in that
they demonstrated the potential oi the moving hot zone technique 1oz
depositindo slloys oi uniform. Composition.
The interpretation of results or moving-2.0-zone tests was aideå Oy
prior experience with stationary hot 2020. Alloys à cosited in a sta-
tionary hot zone at about 600°C exhibit an axial variation in reniwers
cortent because Re# is more readily reduced by városun ürün is WE's and
the rheniw content is higher near the inlet of tze reacuica. Pore than
Iartner downstrean. Il the mixture passes through a temperature gradient
ir raciring the reaction zone, this eflect is magnified. Movima ühe hot
zone, vaerefore, is equivalent to superimposing numerous composicior.
profiles. As a result, the deposit is axially uninomin in composicion,
but radially nonuniform.
Thermodynamic calculations were made with the aid oz a computer to
indicate parametric regions wherein composition proricö would be mini-
nized. These calculations examine factors afecting equilibrium in the
hydrogen reduction of WFG and Refs to metal via intermediate lluorides.
The parameters included in the calculations are listed below.
Lemperature, °K (°C)
Pressure, torr
400-1500 (27--227) in 100° increments
1, 5, 10, 20, 30, 50, 100, 200, 500,
ana 2000
9:1, 15:1, 39:1
1:1, 3:1, 19:1
m2-10-(HT: * RET) ratio
W76-TO-Rero ratio
The results oi these calculations indicate that (1) the tendency of
176 to be reduceá by nyárogen approaches that of ReF6 with increasing
icoperature, decreasing pressure, and increasing values oi nyárogen-
to-(WF. + ReFo) ratio, (2) the WF6-to-ReFg ratio appears to have little
or no exiect, (3) temperature has a far greater effect than pressure or
cas composition, and (4) under equilibrium conditions the composition of
the deposit corresponds to the metal content oł the WF-Rere mixture when
the reaction occurs at about 1000°C.
In order to test the validity of these calculated results an injector
is being used to admit the WF-ReF. mixture directly into the reaction zone,
The rizst injector was water coole, and strongly influenced the temperature
profile in the vicinity of the tip. Fo: exertoie, at a nominal Jurnace cem-
perature of 800°C mandrel surrace temperatures were 475°C adjaceni to the
tip and 800°C at a point 3 ir dovastrean. I though this temperature range
is well below the thermodynamically calculated temperature on optimam
deposition, alloys containing ominally 256 Re were deposited ac io torr
and with hydrogen, WFc; and ReF. flow rates of 1000, 30, word 20 cms/min,
respectively. Figure ide.. 8 shows the microstructures or sapies prcparea
with and without an injector but under otserwise similar conditions. The
sarpie shown in Fig. III. Sa was prepared without an injector and has
definite layers whica etched at difierent rates, whereas the same sampie
-
-
-
.
.
:11-
.
....
Loison des roissiü
16-OriICIAL
..
..
..
........
7.
ti
*
................
ii.
Ti,
6
2
u
..:
:
:)!::::"
.....
.
..
6:!![-!:-OFFICI..
Fig. III.8. Comparison of Microstructures of W-25% Re Alloy
Deposited by the TCD Process in a Moving Hot Zone. (a) Without an injector
for the WF6-ReF: mixture; (b) with an injector. Etchant: 50 parts N7:07

..:
*
III. 18
;.
-
-
sitten
*
.
I.
UN
prepared with an injector (Fig. III.80) słows less evidence or a nenium
gradient through the thickness. The axial variacion in rhenium conterii in
these experiments was 26.1 * o: be over a 6-1a. lengüio.
:.'..
P
-
-
.
E
Terimants with roncooleá injector are in progress to aceis the
erfects circuccion terperatures in the calculaica Domnul rango oro 800
to 1100°C. At these ter:peratures nicken and piain-cazijo:. stiel injectors
Ere corroded by the fluorides. If this corrosion resuiüs from a coob.r.com
tior. oi che metais with Iluorine themally decomposed Iron üne iLuorides,
then mixing hydrosen with the fluorides in the injector may minimize
corrosion by forming stable hyārogen-iluoride with available Iluorine.
.
S
I
Ile deposition of W-Re alioys in the exuernal coating apparatus have
continuca. This apparatus, shown schematically in 7:6. III.G consists
oão mimi chamber and an internal.ly 'heated tubular manazel. Mixini ve
the reacties zases in the volume around the mandrel minimizes tze depica
tion éirecü tzat occurs during deposition inside relatively 2.ong, carro:
Tübes. loys of W-25 et. % Re wica soca Exiano unitormity were separed
in his aggaratus; however, the deposits usually veze nocular near whe
met ana cad aiainetrically opposite trick and which rezionis. The apparu-
tus was moairiea to charge the gas-220w pautes 12 I:OZüco improve
tickness uniformity. The gases were causea co chave de posicion
c oer iaxou a narrow arnulüs around the sciana To exit
è similar arnulus. Deposits coritaining about 55 Re were maen neparea
üü 600°C, 10-torr pressure, and gas I'low rates 01 100), 55, are 5 cms/min
Ožia: Fó, ara Reis, respectávely. Under these couition crie tinickness
uniority was not substantially az'iected but the degree or cowocess 12.5
L ovea; composition unii'ormity, on the oüzer: märä, was aaverseny arzectea.
perenüly the apparatus modification caused the gases to stream along the
manarel, resuiting in a higher rhenium content near the gas inlet than
Tartner downstream. These results are similar to those for internal
āeposits in long, narrow tubes.

9 .
Tej III
روزی/ حنا را یکی دوسری
.
Fig. III.9. Schematic Representation or soparatus for External
Deposition of Tungsten-Răenium Alloys by üze ICD Process.
I
.
.
A
GLE
.
.-
.
06::'.-.0:
-::272
MIXING CHAMBER
HEATER
-
-
-
-
.
.-
-
.
GAS OUTLET
DEPOSITION MANDREL
WWWMWWWW
-
-
-
-
-
.--
-
-
-
-
GAS INLET
20910
-
-
-
- Ur
-
-
**
MA
?
-:
'
--


III.19
Rhenium-rich alloys with rhenium contents as high as 72 Eave been
depositeå in. The watexnal coating apparatus at 600°C. These deposits
contain regions or nocular grorchs and dark, porcus coatings. The poor
(
pality of these deposits suggesis that tae apparatus used is useful
cray for deposition of turgsten-sich deposits.
,
20. descuctive Evaluation
(R. W. YcCluns)
Wenty-seve.. ICD tungsten deposits or CODICI tubes were submitted
20% noosvructive evaluation oIckness variations in colection with
9.2 stavistica de posicion. süucies described earlier. The cooper tubes
were 15 in. 10w-, 3/4, 12. -O. D., with a nomira.. wall thickness oi 0.031 in.
za turisten büă been deposited in varying sicknesses along this length
Orche wübes with the maximur. thickaess eins about 3/32 ir.. Since
There was no "on-the-she 12" technicue available 1o this evaluacion,
Caree gizierent metacas vere evaluated or iedsäciliny: ühece echocs
weze bazea on eady-currents, ultrasonics, anü ä сays.
bit
1
mae eday-current approach was selected as being the most crickly
available. In this technique the inner and outer diameters om tube
e ces red, and the dizierence vas attributed to the aupiex couper-
Ceresien wand. In the ultrasonic agroacia, resozance pricine was
hocü coeasure the entire tube wali with apropiatie correction i'o 22
ve-city comercice between coope: änä tunnen oor airlicuities were
encountered in each oi che ci coac.es, but it appeareä taat these coula
to overcome. It became evicviat, .07ever, wiat ühe 1721. Taickness Com the
c er wbing was not conscant as originally assunca. It was recognizeú
what both the eccy-current aid ümbrasociacic mechis would have beer. subject
to error: as great as the variation in copper Tickness. Developerc work
ana probe iebrication would have allowed a dirierer's eduy-current tech-
ricue to soave the problem, buü ünis mouth was not pursued in view cz
coca results from the x-ray method described below.
A through-transmission x-ray aiteration technique was used in vacich
the radiation detector was placea in the bore or the tube. The racially
directed at rays were attenuateä as they passeä сhrough the cupiex well,
but the tungsten had the greatest erieco per unit of thickness, üzus mini-
mizing the effect of the variable cooper. Tam nickness. The raciation
source was á Torelco -150 x-ray mit ocean ved at 150 KVC? ano 3 A wita
0.018 ini. Oi copper libration. the rilication was required to reduce
the overa.. beam intensity anë absorb the low-energy x rays. The cray
beam WeƏ collinated to 0.078 in. in diameter. De radiation atjector was
a via I (1) crystal, 1/4 in. in diameter and 1/2 in. 102:8, opticaly coupleü
co a c ercial ziber optic ligat pipe. The light pipe was there coupież
co an RCA 6342A protomatisier tube. he output of the photo plie:
Tube was monitored by means or a digital voltmeter.
4
24
The calibration curve was obtaineü ay piotting the voltmeter reeding
(representing radiation attentiation) vs the thickness of rolleå türssien
sheet specimens. The detector was de la stationary in the beam aná
i
-
-
!.
:-05:1";
III.20
C. ::-.
positioned or maximum signci. Mae tube to be measured vas me.slated
with the detecca zerraining on t.e axial center. Thus, only one wall
Ackness as O ored at any giver time. The measurements were obtained
au four circwania, points at 90° rotation and nine .0:
stueiral
osicions for a total oi 36 measuremerts per tube. On se on the cubes
deposiüed at high temperatures, tie cooper Thickness Vestation was approx-
steny 0.005 in. A CO"per thickness charge on this magnitucie would be
cruivüleri ::. 47 0.0005-in.. tungsten thickness and wound produce tra'
tolerance or. ie irai trick.css determinations. It was assumed 3.20
ac reäiation anderuation for TOD tungsten. Was he same as for rolled
Curiosica. I, även metallographic thickness measurements are obtained
this eSSwior is shown to be noi valia, the candoration curve can be
revised vea aurropriate corrections for differences in radiation
attenus. tion.
Deformation Straies
J. E. Spruieli2
B. F. Shulers
Süüdes 1:27€ vegur. to determine the effects of warme zoming ara
heart water on the texture, microstrun, Gil ducting C o
Cum secü. As-aepositéc... otec. örny consists
o nly
orienea,
c a r mains david3 (100 ..me to the deposicior. sub-
Surace. Mividual frams, owever, are o catec izouga larga ES
about iconi. Ten working by "olling and come types oz heat treatment
may cance microstructures anci textures mai are more cesirabie for
come c acations than those oz. The Es-dipositeå räieriai.
-
sien sheet was depositců usias reviously establis:ea ceposicion
conüitions. 5 Four sheets were obtainta iro... a single de zosition run,
cuch weet measuring 1.5 in. ice X 12 . 100.8 wica a thickness
0.092 : 0.003 in. These settó ere izan electrcaiscoarge machines (DVD)
to a ciasconess or 0.070 0.001 in. Toj sheets were ground to remove
une pizsü-cepositeü (rine-grain.) suriace to avoid any süoscoe wisturb-
bez ezi'ecüa that might be caused by manis cir. üyer. The sizeets Mere
üzen ciü into 3 X 1.25-in. coupons 1o rohing. An equal muide0.:
Couporow as cut i'rom 0.065-in Thick wrought powaermetalürey tungsten
Sheet. These coupons were recrystalized at 1500°C Ior Iar and were
subsequency rolled and heat treated in the same manner as ise TCD materiai
to serve as a basis for comparison of results.
2Consultant from the University v mennessee.
S er emioyee zona üne University o: mennessee.
heestara, J. i. y'acerer, C. . Teriter, Jr., Pres :0:2 ari
Tel:
por-32208.1ve... CRNE-3662 (cüsusi 190.77.
7. T. Patterson, W. C. Robinson, Jr., and C. . Iditten, uin,
-evene Materials P:00 am Güstü. 302. Rept. April 30, 1956.
Ülow- 40, PP. 59-99 (Classified).
NI
III.21
OINI - AC - OPEKCIAL
ORNL - ABC - OFFICIAL
..
.::
Rolling Schedule
All specimens were rolled and heat treated according to the schedule,
shown in Table III.3. The specimens were rolled at 1200°C using the frame-
and-cladding technique with Bastelloy B frames apil Hastelloy w cladding.
An aluminum oxide powder coating was placed on the tungsten samples before ,
cladding to facilitate specimen removal after rolling. Stress relieving
was carried out at 1200ºC for 1 hr.
Table III.3.
Rolling and Heat-Treating Schedule
for TCD Irmgsten Sheet
Heat-Treatment Group
-
-
· Specimen
Desig-
nation
-
IS
III
.
Total
Reduc.
tiona, b
. (*)
20 Stress'
relieve
40 Stress
relieve
GI
Anneale
MUM
Anneal
Stress
relieve
Anneal Anneal after
50% reduction;
Lufinal anneal
.
•
4A -
!.
Stress relieve.
after 50%
reduction;
Pinal stress
relieve . .
Stress relieve
after 50%
reduction;
final stress
relieve :
Anneal
Stress
relieve
Anneal after
50% reduction;
final anneal
Rolling was done by the conventional frame-and-cladding technique,
I using Hastelloy B frames and Hastelloy w cladding, at 1200°C.
Specimens were rolled in one direction only, with ends reversed -
between passes, except for specimen IV-4, which was inadvertently rolled
in the wrong direction after the intermediate anneal.

. :"Heat treatments listed followed rolling operation unless otherwise
indicated.
"stress relief: 1200°C for 1 hr. i
Annealing was done at 2000ºC for 1/2 br in vacuum.
DINI - AIC - OFFICIAL
Fin...
ORNL - AEC - OFFICIAL
:: AUS!TICATION
11. 9:!!! A llti

7
**********
**

:
'
TTON
III.22.
WIJ1130 - DJV- INIO.
• AIC-OSICIAL

In lower reductions, the Hastelloy B frames contracted more than
did the tungsten specimens on cooling, thus causing dishing of the
tungsten and some cracking. (Linear coefficients of expansion are
4.6 min. in.: •cº? for tungsten and about 27 uin. in.? •cºl for the ':.
Hastelloys.) High reductions, on the other hand, caused the frames to,
move away from the specimens enough to prevent this situation. The :-
specimens that received a total reduction of 80%, for example, remained
flat and sound.
erin...,i..,on
Microstructural Hrects
$
The microstructure of the 23-deposited material is shown in Mg. III.1
from 0.010 to 0.015 in. of the first deposited surface (top of photo-
micrograph) was removed prior to rolling. Warm rolling to reductions
less than 60% bad little effect on the as-deposited microstructure.
Fig. III. 11 shows the microstructure of a specimen that was reduced 60%
at 1200°C and subsequently stress relieved for 1 hr at the same tempera.
ture. . Considerable breakdown of the original columnar structure is
:
.
RIGHIN
.
..EO.
.
- DESIRED
MAXIMUM

1
25
,
CENTER LINE
;********-*
110
tig. IZZro
7:65614
.... .. .------
......
----
-••-YPICIAL.
mig. III. 20. Microstructure of TCD Tungsten Sheet Uued in Deformatiod
Studies. The initial deposition surface 18 at the top of the paotanico-
graph. Etchant: 50 parts NHL OH and 50 parts H202.
END TYPING
AIC - OFFICIAL
ok
ASSISICATIC IV
Vi APL CABLE)
.
-
Lau

CLASICATION
lor Bisnis ir MALE)
III.23
ORNL - AEC - OFFICIAY
"OTTOM OF
I LINE OF TOXT
CHAPTER TITLE
1
5.
...
.
tis. zz
,li.
.
:.:.7-66706
- MARGIN ---
RIGHT. ARG
MARGIN
- DESIRED
MAXIMUM
Fig. III.ll. Microstructure of TCD Tungsten Specimen after Total
Reduction of 60% by Rolling at 1200°C and Stre88-Relief Anneal at 1200°C
for 1 hr. (Plane of section 18 parallel to the rolling direction.)
Etchant: 50 parts NHAOH and 50 parts H2O2.
.
.
...
.
apparent and there is some evidence of substructure formation. Fig. III. 12 "
shows the microstructure of a specimen reduced 80% and stress relieved. -- -DES
The columnar structure was completely broken down in this case and some +MAX
evidence of substructure formation and fibering is apparent. The grain ;
growth evident at the surfaces may have resulted from nucleation at .
craters formed during the original EDM machining of the specimen.
Fig. III. 13 is a photomicrograph of a sample reduced 80% and subsequently
annealed at 2000°C for 1/2 hr. Ertensive grain growth is evident.
.
.
V
+
.
.
.
.
................
Texture Determinations
Rolling texture determinations are being made by an x-ray diffraction
technique that requires spherical samples. Thin sheets of the rolled
material vere cemented together using an electrically conducting epoxy .
resin, keeping the rolling direction of laminae alifand. Sphero, mea nuring
1.0.250 10. in diameter were then prepared by electrodischarge machining, ... i
-
.:.
-
ORNI - ACC - OFFICIAL
e
i
E".6Li K. Jetter and B. 8. Borie, Jr., "A Method for the Quantitative -
Deterrination of Preferred Orientations," J. Appl. Phys. 24., 532–35 (1953).
VARGINAL
..
.
. MASI
,-
-
-
01.67 ," TION
11 i 11. LEI
III.24.
TORNI - AIC - OFFICIAL

Et
ORNI - Attirnich
hit. If to!
PTER 1! Ti"
ose minimalne
Pogi Itt, 12
_
ARGIN
4-61.7546
RIGHT
MARGIN
---DESIRED
MAXIMUM
-
--
.
Fig. III.12. Microstructure of TCD Tungsten Specimen after Total :
Reduction of 80% by Rolling at 1200°C and Stre88-Relief Anneal at
1200°C for i hr. (Plane of section 18 parallel to the rolling direction.):
Etchant: 50 parts NH,OH and 50 parts H2O2.
.:.
:
.::.
...
...
The texture produced after a 20% reduction by rolling of as-deposited
material is shown in Fig. III. 14; the positions of (110) poles for
"ideal" orientation after deformation are indicated by black triangles.
Several body-centered cubic materials, including PM tungsten, develop
textures close to the (100) (011) "ideal" texture after rolling.' The
relative intensities of (110) reflections, indicated by mmbers on the
contour lines in Fig. III.14, are higher in the vicinity of the black ,
triangles representing the "ideal" deformation texture than elsewhere...
Thus, the "ideal" texture is beginning to develop in TCD tungsten arteri
only 20% reduction. The (110) pole figure in Fig. III.15 shows that the
rolling texture 18 rapidly approaching the (100) [ou] "ideal" texture
as the reduction increases to 40%. Determinations of (110) and: (200) pole.
figures for 60 and 80% reduction are in progress.
.
:
.
-
-,-,..Lola.
7J. W. Pugh, the Temperature Dependence of Preferred Orientation
in Rolled Tungsten," Trans. ATME, 212, 637–42 (1958). ::
:
ORNI - AEC - OFFICIAL
· END TYPING
CLASSIFICATION :
(IF APPLICADLE)

C
.
C
.;:;.TION
III. 25.
· ORNL - AIC - OFFICIAU
.
BOTTOM OF
CiT l.INC. OF TEXT
CHAPTCH TITIS
Fig. III. 13.
4-66966
RGIN
TMARGIN...
RIGHT
MARGIN
- DESIRED
MAXIMUIM!
..
............
Fig. III. 13. Microgtructure of TCD Tungsten Specimen After Reduction
of 80% by Rolling at 1200°C and Anneal at 2000°C for 1/2 br. (Plane of
. section is parallel to the rolling direction. ) Etchant: 50 parts
NH, OH and 50 parts H202.
-
-

Physical Metallurgy of TCD Tungsten
A. C. Schaffhauser K. Ferrell
Low-Temperature Ductility
Investigations of the ductility of TCD tungsten sheet were continued.
A complete evaluation of the effect of heat treatment and specimen condi-
. tion on the ductile-to-brittle transition temperature. (DBTT) of deposit "
.:: PWL has been reported previously. 8,9 Evaluation of two other deposits
! (PW18 and PW44) produced under slightly different conditions is under way.
fi The deposition conditions for these deposits are presented in Table III.4
" and the chemical analyses are listed in Tables III.5 and III.6.
•
No 11:55
TYPE -.
i
ri roc
8A. C. Schaffhauser, High-Temperature Materials Program Quart. Prok.
Rept. Oct. 31, 1965, CRNI-TM-1350, pp 88-89 (Classified).
A c. Schaffhauser, High Temperature Materials Program Quart. Progr.
i Rent, Jan. 31m1966, ORAL-TM-1455, pp. 73-83 (Classified),
-ORNI - AEC - OFFICIAL
e
CLASSIFICATION
i TIF ANDA

Pin:
"ION
i dit !!16 ILE)
III:26
10
it! TEXT
i MA
GIN
:RIGHT
: MARGIN
wir II14
-- DESIRED
MAXIMUM!
Orny aug 66
66-4291
.
.. CENTC
...
...
..
Fig. III.14 The (110) Pole Figure for TCD Tungsten Sheet Reduced
20% in Thickness by Rolling at 1200°C. Triangles indicate the (110) (0)
Y?
"1deal" texture.
.
.
END TYPING
ORNI - AEC - OFFICIAL
·
111..10TION . .
li ::..
L ES) .
*
*
.....
..
....

siinilom
.
.
.III27
.
OI DE
NE UR 1
TERTIT
.
- OFFICIAL
RNL - AICC
OTTONI or
-LINE OF TEXT
HAPTER TITLE
ARGIN-
MARGIN-.
RIGHT
HARGIN
- DESIRED
MAXIMUM ;
:
: Open
Ourg 66-4792
wor*
.
..; LINES
YPE ...
Pig. III.15 The (110) Pole Figure for TCD Tungsten Sheet Reduced
; 40% in Thickness by Rolling at 1200°C. Trianges indicate the (100) (012)
"ideal" texture.
ORNI - AEC - OFFICIAL
CASSIFICATION
lif:
A
n ille!

Od: III28: 10min
be
1:.
AEC - OFFICIA
ORNI - AEC - OFFICIAL
том :)
NE OF TEXT
TER TITLE -
':
.
:
Table III.4. Deposition Conditions of TCD Sheet
Average
.
Deposition Rate
(als/hr):
Deposit
Mumber
PWL.
PWl8
'PWhedon
Gas Flow
(cm3/min) Pressure
82 W Ar (torr)
1500 300 500 50 .
3000 200 :: 5 :
3000 200 .. . 5 .
(°C)
570
560
560
LARGIN- .
RIGHT
MARGIN
ANINIR?U:$'!
DESInline
DE
BMM:
TABLE III.5. Interstitial Impurity Analysis of TCN Tungsten Sheet
o Chemical Analysis (promos
!
Chemical Analys:
Deposit
Humber
Sample
Location
·
PW4
......5
...
-. -. Inlet
Outlet
Iniet
Outlet
Inlet
... 14
,
: 5
i
Pwi8
ooi wou on
........---.00TEB ::-.
vvvv".
no ang
10
VV
a 990nm
1
PW44
Midale
Outlet
.
.---
-
.-
ini
:
: 40
-:-
-
-
15.
.
.
..
1.3.
2
GO :
i
.
.
..ENC TYPING...
Srce
OORNL - AEC - OFFICIAL
L
CA
. ;;INN
!1 .. opel ir 1!!!!)
.
Hi
11,5
04
.

III.29
ORNL - AIC - OFFICIAL
OTTOM OF
LINE OF TEXT
APTER TITI,
!
Den
.
Table III.6. Analysis of Impurities in TCD Tungsten as
Determined by Spark-Source Mass Spectrometer
Deposit Puzlepurity Content
Impurity
Alio
0.4
.
0.02
.
.
B
Ca
ca
0.2
V.
0.8
cu
- MARGIN
...>
Impurity Content“ (ppm)
Deposit PW18 Deposit PW4A
Annealed I har
Annealed. I bor
Ab-deposited at 1800°C As-deposited As-deposited at 2500°C.
0.1
: 0.4
0.4
0.1 :
0.02
0.06
0.2 .
0.06
0.6
< 0.05 < 0.05 i 2 . < 0.05
< 0.05
0.8
0.8 < 0.05
< 0.05
< 0.05 . < 0.05 . 0.05
0.8
10
.. 10
10 . ..
33
. 3. La
200
60
0.6 0.4
0.4
60
0.1: **** o.i
< 0.1
<
9
39
39
<0.2
<0.2
. 3
3
: 10
0.03 < 0.1 <0.1' : < 0.1 0.1
10.!.
Pero
.............. 1717:30
V v
O
...
.
MO;
-
60
O
....
NA
0.05
.-
Si
^
Sn -
Ta
<0.2
< 0.2
< 0.2
TITI
30
T:V
-- Semi quantitative analysis - values reported accurate to within a
factor of 2 of actual value.
icons -
:*-::
-40.
10:..
.
-
C
NG LINO3
TYPE ---
SICIAL
.
-
-
- -
-EVO TYPING
ORNI - AC - O
..
.
CLASS ATIO
illi Mid;"! IGA ILI
.
:..
i
-
..


· III.30
IS
PRA
11:00. Microstructures of as-deposited and heat-treated specimens are shown in
Figs. III.16 and III.17.
ORNLALC - OFFICIA
ORNL • AIC -
The random grain structure of deposit PW18 (Fig. III.16) is not
typical of TCD tungsten. The cause of the more equiaxed grain structure....
in this particular deposit is not known. It has been reported that
vanadium contamination of the WF. foed gas might be responsible for the best in
more random grain structure, 10 The cadmium detected in deposit PW28 may
have produced similar results. Also, intentional additions of oxygen,
nitrogen, and methane to the hydrogen reed gas have been shown to produce !
a fine as-deposited structure. 11,12 The lack of grain stability during 1..
heat treatment of this material has been discussed previously.
-.
.
The preliminary data given in Table III.7 indicate that this randou !
as-deposited grain structure is not effective in lowering the DBIT. This ;
observation is consistent with previous observations that the DBTT is
· "not significantly affected by removal of the fine-grained, first-deposited:
surface normally produced during deposition. Also, our studies on the:
low-temperature deformation behavior of recrystallized PM tungsten show ! RIGHT.
HARGIN
little dependence of DBIT on grain size..13
----DESIRED
MAXIMUM!;
1°R. L. Heestand, J. I. Federer, and C. F. Leitten, Jr., Preparation !
and Evaluation of Vapor-Deposited Tungsten, ORNL-3662 (August 1964).
11Studies of Thermionic Materials for Space Power Applications,
NASA CR-54715 (Oct. 15, 1965). Classified.
12 Studies of Thermionic Materials for Space Power Applications,
NASA CR-54779 (Dec. 20, 1965). Classified.
13K. Farrell, A. C. Schaffhauser, and J. 0. Stiegler, High-Temperature
Materials Program Quart. Progr. Rept. April 30, 1966, ORAL-TM-1520
(to be published).
--
TITLE
......
............
Table III.7. Ductile-to-Brittle Transition Temperature of
.. TCD Tungsten Sheeta
....
Ductile-to-Brittle Transition Temperature (°c)
Annealed 1 hr
Annealed 1 hr
As-Deposited i at 1400°C . . at 2500°C
.
Deposit
Number
......
..
cons
-
.
PW4
PW28
PWL
.
. 270
250
400
185
I
225
250
!
St.
1
350
::L
Test conditions: punch radius 4 T and deflection rate of
0.1 in./min.
... "Defined as the lowest temperature at which a 90° bend could
be obtained. .
.
*
ORNI - AFC - OFFICIAL
OINI - AEC - ORFICIAL

Cli;;. i 710
.
e
-Ė - DE
:
.
,
..................
dijou ni Ö -- inü
WIJIS10 - 13-
-
O
Y-6770
Y 66079
omadu..do wes niet som er
v
---

.
O
r die mens vid bommais do what are commodo no noto.....teorier oca
Fug. m
AS-DEPOSITED
IHR AT 1800°C
TIC
.
Y67732
Y 66083
ch
.
100 HR AT 1800°C
I HR AT 2500C
.
" No Reductions i
ORNI - ACC - OFFICIAL
..ORNL - AIC - OFFICIAL
m
i
III
3


RRent - Handmannadanan


ence
:.
.--.
Y-67701 :
Y-62455
W
S
u
.
*
.............
..
.,
1
.
**..
...::
.
...
-
i prendere
Un mo..E
2
----
C
ort
....
s
AS-DEPOSI TED
1 HR AT 2500°C
%" No Reductric
Fig. III.17. Effect of Heat Treatment on the Microstructure of ICD Tungsten
Deposit PWhite. 100x. Etchant: 50% NH,OH-50% H202.



ICINO
III.33
.
ORNL - AIC-ORIKCIAL
1.1.1:
ORNI - AB
The effect of annealing temperature on the DBTT of 2.posit PWlk 16
aloo given in Table III.7. The reason for the very high DBTT of the as-
deposited specimens and specimens annealed for I ler at 2500°C 18 not
apparent. The deposition conditions, chemical analysis, and microstruc- :
ture of PWhole are very similar to those of PWh. The spectrometric analysis
for trace impurities (Table III.6) indicates a much higher potassium con-
tent in PWhite but the significance of this is not known. Purther tests are
under way to determine the cause of the poor ductility of this deposit. i.
Studies on Void Formation
-
-
-
-
ir NARGIN.
Massive void formation has been encountered during extended operation
of some TCD tungsten materials at 1800 to 2000°C. The problem of massive !.
void formation appears to be related to the high fluorige content (100 to
200 ppm) in material deposited at high rates (2 30 mils/hr) and high
pressure (0.5 to 1 atm), or to the effects of impurities in the WF6 feed
fa 8.
RIGHT
MARGIN
The TCD tungsten produced at ORNL is depositer. at low deposition
rates and pressures from purified feed gases (see Table III.4). The
- DESIRED
fluorine content of this material is normally less than 30 ppm and the MAXIMUM
concentration of other impurities is low (Tables III.5 and III.6). There !
18 relatively little void formation in this high-purity material after
exposure at temperatures as high as' 2500°C (Figs. III. 16 and III.17).
However, it has been shown that these small voids formed on annealing
grow tremendously under stress at high temperature. 14
.--.........
We are investigating void formation in high-purity TCD material to
determine under what conditions the voids would be detrimental in service.
We also hope to gain a better understanding of the source of the voids :
and to develop methods to eliminate them. The formation and growth
characteristics of voids during high-temperature annealing are considered
in this section. Void formation and growth under high-temperature creep
conditions are described in the next section.
--.
-...
----
------
-
-
-
A typical microstructure of voiás formed in high-purity TCD tungsten
18 shown in Fig. III. 17 and at high magnification in Fig. III.18.' The
diameters of voids in the columnar grain boundaries range from 1 to 3 H.
and voids within the grains are 0.5 u in diameter or less, after l-hr r
treatment at 2500°C. Fewer voids are present after 1 hr at 2200°C, and'
voids are usually not observed in material treated below 2000°C. It is :
interesting to note that in deposit PW18, where a large amount of grain .
growth occurred, the voids within the grains are larger than those in i
material showing no grain growth (Figs. III. 16 and III. 17). Also, the :)
voids are observed at lower temperatures when grain growth has occurred
even though this material is of higher purity. The voids observed after
-
.
..
..
..
-
-
... - -
-
ORNI - AIC - OFFICIAL
i
14H. . McCoy, High Temperature Materials Program Quart. Progr.
. Rept. Jan. 31, 1966, ORRI-11-1455, pp. 84-101 Classified).
ORNI - AEC - OFFICIAL

weiter
Cl. S'!CTION
11.:,',ICASE)
OINI - AEC - OFFICIAL *
-
III:34
.
IV.
TO!.1 OF
NE OF MYT
POTEN 7171.1
)
-
CD
.
Figure 2.18.
4-71161
in.
MARCIN
men of
.
RIGHT
ARGIN
ESIRED
AXIMUI.
Fig. III.18. Voids in TCD Tungsten After Annealing 1 hr at 2500°C.
Deposit PW4h. Unetched.' 1000X.
annealing this material at 1800°C and the growth of these voids with time
18 shown in Fig. III. 19.
These observations indicate that the source of the voids 18 uniformly
distributed within the grains. Due to enhanced diffusion in the grain
boundaries the voids there are able to grow by coalescence. During grain :
growth the moving boundaries could sweep submicroscopic voids together
and cause them to coalescence to some critical size. The boundaries
could then move around these larger voids leaving a uniform distribution
of voids within the grain as observed in deposit PW18 annealed 1 hr at
2500°C (Fig. III.16).
VI 21200 - now and
The initial stages of void formation have been investigated by trans-i.
mission electron microscopy. This technique erables us to investigate
the possibility that an excess concentration of vacancies produced during
deposition might be contributing to void formation in the high-purity
material. Samples from deposit Pwlly annealed at various temperatures :.
between 1000 and 2000°C were examined. Figure III. 20 sbows the small
grain-boundary voida present in samples annealed at 1800°C. No voids : 1
were observed in samples annealed below this temperature nor was there
any evidence of vacancy clustering or collapsed vacancy loops as would:
be? expected 12 the voids were due to an excess concentration of
vacancies.
...
OINT - AEC - OFFICIAL
Bilio -
.

v
-
-
.
-
1
-
s
· CLASSIFICATION
("'A.':r!1 iUN
sosa i .!!1ILET
III.35
OOTTOM OF
'ST LINE OF TEXT
CHAPTER TITLE - phone
· ORNL - AEC - OFFICIAL

OH
tig. III.19
.
Alo80
6092
" MARGIN ***
RIGHT
Fig. III.19. Voids in ICD Tungsten Material that Exhibited Grain . mancin
Growth. Annealed at 1800°C. Deposit PW18. 500x. Unetched. (a) Annealed
....DESIRED
i hr' and (b) annealed 100 hr.
IRGIN
IMAXIMUM

...
.........
..
........
LiliLiSS
Fig. III. 20
YE9160
.
iri 1 INC.
*YPE:
.
Fig. III.20. Electron Micrograph Showing Initiation of Voids in
.. ICD Tungsten Annealed 1 hr at 1800°C. Deposit PWlk. Original photo
reduced 20%.
END TYPING :
.
CAEC - OFFICIAL
i
...
.
.
...............
........
.......
.
Christi ICATION
TE ANTIC.
HN

III.36
ONE Mko.
MAN! - AFC - OFFICIAL
:
1
-
M OF
OF TIVO
ER TIT!.
.
-
Direct carbon replicas of fractured surfaces were examined in order. I l
to determine the true size and shape of the voids. With this tochni que
there is no possibility for enlargement or distortion of the voids because ...,
no mechanical polishing or etching is involved. Pgure III.21 shows voids.
on a graln-boundary Tracture surface of a sample annealed at 2500°C and ī
fractured at room temperature. The polyhedral shape of the grain-boundary
voids 18 due to the variation of surface energy of the voids with orienta-
tion of the material.19 A void in thermal equilibrium has flat faces of
low surface energy. In body-centered cubic polyhedral voids formed by: !
eight (110) and six (110) faces are the most common. The equillbrium
shape of a void in a grain boundary is determined by the lowest surface" ; !
energy appropriate to each of the two grain orientations and the inter-
facial energy between the two grains. The elongated voids at the inter-
section of the fractured grain-boundary surfaces shown in Fig. III.21 are ;
believed to be due to the interfacial energy between the two grains.

-
í
MX
RGIN
...ben
..
...!
15R. 8. Nelson, et al., Phil. Mag. 11(109), 91-11 (1964).
11GHT
L
. -DERGIN
7
- DESIRED
MAXIMUM

I...
2:17
Fina71, al
469194
oo
Fig. III.21 Direct Carbon Replica of Grain Boundary Fracture Surface!
of ICD Tungsten showing Voids Formed on Annealing 1 hr at 2500°C. Deposit
PWhile. Original photo reduced 19.5%.
DINI – AEC - OFFICIAL
ORNI - AEC - OFFICIAL
::
:
END TYPING
..::
. 21
x
NA
--
CLASSIFICATION

III.37
ORNL - AEC - OFFICIAL
''OOTTO!.4 OF
AST LINE OF TI:XT
'R CHAPTER TITLE
FITOR
Creep-Rupture Properties of TCD Tungsten
OFFICIAL
H. L. McCoy
. Studies have been conducted to determine the creep-rupture properties!
of several lots of TCD tungsten. Because of the small amount of test mato
erial available from each lot and the inherent scatter in the test data,
1t has been necessary to consider the properties of the TCD tungsten only...
on a comparative basis with those of the PM tungsten.
The material used in this study was produced by deposition from a
gas mixture of Wig and hydrogen. The material was deposited in a square
mandrel as an open-ended box about 2- x 2-in. cross section and 20 in.
'long. A copper mandrel was used for the "H" series and a molybdenum man-
drel was used for the "p" series. The sides of the box were separated by
grinding, and individual specimens were machined by electrodischarge mach-
ining. The special tool developed for this purpose is showa in Fig. III.22.
About 0.022 in. was removed from the machined edges by lapping with dia. .
HARO
RIGHT
EFT MARGIN -.. bond ametne me moe
..mond abrasive. The gage' section in the test, specimen was 0.25 X 1.50 in..
MARGINT
20 i . . .
: : ....... DESIRED
MAXIMUM
no

"
.
CATE? LIE
Fig. III. 22
Photo 83112
.
.
oto
-
25
AINING, INI :
OF TYPE:
. : 2
ORNL - AEC - OFFICIAL
ORNL - AEC - OFFICIAL
Foi Mg. HII.22. Mxture for Preparing Tungsten-Creep Specimens by
..o Electrodischarge Machining.
ENOTYPING
CLASIFICTION
(1) At.1. 16. AiLE;
.
III. 38
ORNE - AIC - OFICIAL
QENI - AC - OFFICIAL
Tojo? ist
Ni lif T
ETTOR TITLE
The microstructure of one beat of material 18 sbowa in Hg. III. 23.
.Figure III. 23a shows the microstructure as seen from a plane parallel to
the surface of the sheet, and Hg. III. 23b shows a cross section of the
· deposit. The surface layer of fino grains was formed adjacent to the
mandrel. This layer of fine grains soon changed to large columar grains i
which would normally extend through the remainder of the deposit. However,
the particular heat shown was laminated once due to an interruption in thol.
hydrogen supply. The test specimens were oriented so that the applied ,
stress was normal to the columnar grain boundaries. Some of specimens
were surface lapped on both sides to remove the layer of fine grains and
the irregular tops of the columnar grains.
.....
....
The results of creep-rupture tests on TCD material are given in
Table III.8. The creep-rupture properties of the TCD tungsten are cam ! ::
pared with those of the PM tungsten in Fig. III. 24. Although there is
some scatter, the data fall in a reasonable band about the line representi
ing the properties of the PM tungsten. Figure III. 25 shows a comparison
of the minimum creep rate of TCD and PM tungsten. The minimum creep rate .
AROIN - be of the TCD tungsten is less at 1650°C and greater at 2200°C. Mgure ITI: 26 RIGI
MARG
campares the rupture ductilities of the two types of material. At 1650°C;
the TCD tungsten, with only one exception, has a ductility of 5% or less
At 2200°C the TCD tungsten shows a trend of increasing fracture ductility:
with increasing rupture life. The matter of ductility is discussed in
more detail below.
.-
.
.
-.
.
Although some of the specimens were surface lapped and others were
tested in the as-deposited condition, there appeared to be no influence
of this variable on any property measured.
Metallographic studies revealed several interesting features of the
deformation processes in the TCD tungsten. Figure III.27 shows the frac-
ture of a specimen tested at 1650°C and 4000 psi. · The general character: 1 ::
istics. of this TCD specimen are about the same as those observed in the
PM specimen shown in Fig. III.28. However, the cracks in the PM materials
are elongated and wedge-shaped whereas those in the TCD material are round. :
except where they have connected. The cracks were general throughout the
PM material, but were locater only near the fracture in the TCD specimen. ::
,
Figure III. 29 shows a specimen tested at 2200°C and 2000 psi. The
tests lasted only 21.3 br, but the network of voids has become very' general.
Figure III.30 shows a specimen tested at 1250 psi and 22000°C. This test
in lasted 789.9 br. The longer test time resulted in.growth and coalescence
of the voids and grain-boundary migration such that an equiaxdal grain.
structure was formed. For comparative purposes, the microstructure of a
• specimen of PM tungsten tested at 2200°C is shown in Mg. III.31. This :
... specimen lasted 42.1 hr and the grain size is quite large. Hence thei
grain size of the TCD tungsten 18 more stable at elevated temperatures. -
It is felt that the ductility can be rationalized in terms of the
grain structure for a material to exhibit good ductility at elevated :
1 temparatures, it is necessary that shear stresses be developed across the i
END TYPING
...ORNI -AEC - OFFICIAL
ORNI - AEC - OFFICIAL
..
.nowdenen ..........
.......
.
..
%
.
ALLE!
9.35191 ITION
to Aprilia MOLE)
Il-Mini
!..dirm

20.
From
.
-
2
•AR -
E
2
OINI - AIC - OFFICIAL
1
-
-:
ledura
wa
9035 INCHES
FOOX
-
-
ET
.
-0.035 INCHES
XOOM
64 74 Run
ORNI - AEC - OFFICIAL
ORNI - AEC - OSSICIAL
cm Tigrama
a Conditions. Lot PW 3. Etchant: 50 parts NHAUH" 200- games. com
! (a) View of plane parallel to surface of sheet. (b). View of plane normal .
:: An wintnou or whasti
EcommiTac - Sporting

..
.
.
.
ORNI - AEC - OFFICIAL
-
SNN
**************
WIDITQ-93-Wuo..
NVIDISIO - DJV-INIO
A
Test
Hember
Lot Surface
hember. Treatment
Rupture
Strain
Minimum
Creeg Rate
1650
3478
328
3556
3837
3838
3569
4045
8-16
H-16
P -3
-3
P -3
Pi-3
• Table III.8. Creep Properties of TCD Tungsten
Time to Indicated Strain, br
Temperature Stress
(°C) . (psi)
> : 2%
10% - Rupture
6000
19.2
30.5
1650
4000
5.0
35.0 185.0 390.0
1650 6000
593.7
1650
9.7
6000
4.7 11.0
1650
12.8
5000
35.0 102.0
1650 4000 20.0
225.0
413.0
2200
•1346
1.0
2.7 6.5
13.0
2200
62.9
2000
0.5
1.4
4.6
10.3
2200
1500
4.7
9.3 18.0
29.3
2200
35.7
1250
4.0 48.0 160.0
306.0
2200
789.9
2000
0.7 2.0
2200
4.3
2000
0.4 0.6
1.0
1.6
2200
1.85
2000
1.0
1.5
2.4
3.7
2200
2000
2.1
3.15
4.4
2200 21100
0.5 1.4
2.3
3.1 . 4.65
4.2.
21.0
6.3
3.1
3.5
.
176.5
:
..
Leoped
Lapped
Hone
love
Hone
Lapped
Lapped
Lapped
Hone
Lapped
Hone
Lapped
Lapped
Hone"
Lapped
3577
0.063
0.020
0.083
0.12
0.016
0.0040
0.75 .
1.1
0.19
0.026
3.2
35.0
21.9
21.3
8-16
R -3
2 -3
1-3
R-19
PW-20
PU-23
PW-23
PW-24
3680
3664
407
92
X95
530
5294
0.71.
16.6
29.2
4.2
15.6
21.8
21.9
21.6
1.2
· 5.05
5.3
2.8
1.0
0.85
1.33
Annealed 90 hr at 1650°C prior to loading.
Recluce
to &l"
896 y cience.

PS
in.
En
.
1
ORNL-DWG 66 - 2493
20,000 OTTI
10,000
• H-16
o PW-3
· PW-19
FA PW-20
to PW-23 +
1
950. L
PW-24 til
-POWDER METALLURGY
tiba
00
STRESS (psi)
2000
200°C
.:

1000
11.24
*10. 100
TIME (hr)
Comparison of the Creep - Rupture Properties of TCD and.
Powder Metallurgy Tungsten.....
'.
.
ORNI - AEC - OFFICIAL
TV101330-33V - INHO
ORNL-DWG 66-2491
20,000
LOT NO.
• H-16
o PW-3
I PW-19
A PW-20
OPW-23 -
A PW-24 +
POWDER METALLURGY
10.000
STRESS (psi)
2200 °C
2000
1000
.
.
.
lo..
50.004 . 0.04.... .0.1
3 :10
MINIMUM CREEP RATE (%/hr).
Comparison of the Minimum Creep Rates of TCD and Powder-
Metallurgy Tungsten.
.
....
.........
OINI - NIC C - OFFICIAL
.,
.
. .
. .
Calendar
.-
-
A
L
L
--
-
-
-
.
--
-
-
-
-
-
-
-



MIDO -330
T664490331V-INUO
.
.
- •
ORNL-DWG. 65-12660
.. O TCD TUNGSTEN, LOT P-W-3, NO SURFACE LAPPING
. A TCD TUNGSTEN, LOT P-W-3, BOTH SURFACES LAPPED
o TCD TUNGSTEN, LOT H-16, BOTH SURFACES LAPPED
OTCD TUNGSTEN, LOT P-W-19, NO SURFACE LAPPING
- TUNGSTEN SHEET, POWDER METALLURGY
1650°C-
ELONGATION (%)
012200°C)
(2200°C)471
2200°C
(2200°C).
(1650°C)
0(2200°C)
IL
(2200°C)
.
il.
Fig II
II 0(1650°C)
1. (1650°C) || (1650°C)
(1650°C)
. 2 5 10 20 50 100 200 500 1000
RUPTURE LIFE (hr)
Comparison of the Fracture Ductilities of TCD and Powder - Metallurgy
Tungsten.
. obtii Afet Bifiche
:

Chev = Aic Coffical
III.42
· DINI Hirneuri.

Feis. III.26 .
: Olinx. Dwg 65-12660
RIGHT
MA?GI
..OLSIACO
"WXIMU
Fig. III.26. Comparison of the Fracture Ductilities of ICD and
Powder-Metallurgy Tungsten.
'..
grain boundaries so that they have greater freedom of motion. With the
boundaries normal to the applied stress no large shear stresses can ex st.
and, therefore, only limited elongation can occur. This explains the
observation that the elongations were small for specimens tested at
1650°C and for short tests at 2200°C. With longer times at 2200°C the
grain structure became more equiaxial with the result that the boundaries
were no longer normal to the applied stress. This resulted in greater
fracture elongations.
The influence of stress on the growth of voids in the TaD material
may be seen by comparing Mgs. III.32 and III.33. The specimen in
Mg. III.32 was annealed at 2200°C for 5.3 hr in the absence of stress.
The specimen in Mg. III.33 was tested at 2200°C and 2000 poi and nailed
ORNI - AEC - OFFICIAL
ve

TD. /
III.43
• OİNL - AEC - OFFICIAL
---
---
---
---
---
---
Paj. . 27
4-63592
'DE! ;
; ;
:
.
.
RIGH
MARGAR
. - DESIRE
MAXIM
!
Pg. III.27. Photomicrograph of a TCD Tungsten Specimen Tested at
1650°C and 4000 psi. Lot PW3. Failure occurred in 4.13.0 br with 3.2%
strain. As polished.
Fig. III. 28
9.60657.
"TAININ, 1.!!
Or I Y
iphone
QINI - AL-NR.
ORNI - AEC - OFFICIAL
Mg. III.28. Photomicrograph of Powder-Metallurgy Tungsten Sheet ..
Tested at 6000 poi and 1650°C. Rupture occurred at 135 bar and 31.25"
struin. As polished.
T
I!
i
snt
111
.
:ION
III.4
.
.
.
.
* FTOMS OF
NE OF IST i
PTER TIL

.
tij II : 29
4.67559
CMARGIN... Do
RE:
RIGHT
MARGIN
::-DESIRED
A. MAXIMUN
--
-
f
.
Fig. III.29. Cross Section of ICD Tungsten, Lot PW 3, Tested at
2000 psi and 2200°C. Failed at 21.3 hr and 21.3% strain. Etchant:
50 parts NHAOH and 50 parts H2O2.

-
.
.
.
1
-
Page 22.30
.
2972
.
.
..
.
..
3. Es
INI - AEC OBFICIAL
.
...
Mg. III.30. Photomicrograph of TCD Tringsten Specimen Tested at
2200°C and 1250.pei. failure occurred ia 789.9 hr mithy29.29 strain.
Lot PW 3. Btchant: 50 parts NHL OH and 50 parts H202.
--------
-
.
LASIFICATION
1111!!,AULT)
i'.
.

CINE
ATL81
III.45
ORNL - AEC - OFFICIAL
:
T!:,:
ST o
,...
.
. . Tig. 11.3/
HARU
.4-68655
FiT SARGlis
'po
Blood
TAA.30
· ESIRI
...: Xid
Fig. III.31. Photomicrograph of PM Thingsten Sheet Tested at 1500 281
and 2200°C. Rupture occurred at 42.1 hr and 16.64 strain. Etchant:
50 parts NHAOH and 50 parts H202.
VAR


Fig. III.32
4-68718
ORNI - AEC - OFFICIAL
į , Mg. III.32. Cross Section of a ICD Tungsten Specimen. Lot PW19.
..Annealed 5.3 hr at 2200°C. Btchant: 50 parts NH OH and 50 parts H2O2.
YPIN

A
.
III.46.
CINE OF SY!
TLR TOT.
ALC - OFFICIAL
ทราบทค
01N
1
Faj. III. 33
23..
4-61652
NARUT
RIGHT
MARGIN
-- DESIRED
MAXIMUM
.
Fig. III.33. Cross Section of a ICD Tungsten Specimen. Lot Pw 19.
Tested at 2200°C and 2000 psi. Failed at 4.3 hr and 4.2% strain.
Etchant: 50 parts NH,OH and 50 parts H2O2.
Gil
SIR
TARU!!,
in 4.3 hr. Hence, it appears that the voids are nucleated during anneal-
ing at 2200°C, and that they grow dramatically under stress. It is also
apparent that these voids develop quickly under stress at temperature.
This behavior probably is due to the presence of an impurity in the grain
boundaries of the deposited product, which when heated expands and farms
pero a void. One proposal is that the impurity 18 same fluorine complex.
Another proposal is that some light elements such as aluminum or magnesium;
may be present as impurities. These elements would have a very high vaper-DES
pressure at 2200°C and could nucleate voids. As discussed in the previous + MA
section, chemical analyses are presently being made in an effort to ili
identify the impurity.
MAH

CENTE? SINE

----...--- -
DC
-------
-
mr.o
ENOTYPING
-
::::
:::16
:04
- OnASCOLICIAL nr.

:
T11131
! PROORIR
CORRECTOR

END
SA
*
!
1
*
SEE
DX?
17
SEA
7
.
ia
DATE FILMED
7/28 /66




ANA
*
D
w
WW
Kiti
,
PA