to
.
.
.
:
0
I OF |
ORNLP
2108
:
.
.
.
:
V
.
:
.
_
.
9.1.25 1.1.4 1.6
MICROCOPY RESOLUTION TEST CHART
NATIONAL BURE AU OF STANDARDS 1963
ORNL-p-2/08
6648
CONF-666ci708
MAY 10 1966
NOTE: This is a draft of a paper being submitted for publication. Con-
tents of this paper should not be quoted nor referred to without
permission of the authors.
CFSTI PRICES
H.C. $2.ro ; MN 50
LEGAL NOTICE
This report no prepared u an account of Goverament sponsored work. Noither the United
Statu, Bor the Commission, nor way person acting on behalf of the Commission:
A. Makes any warranty or representation, expronsed or implied, with respoct to the accu-
racy, coinpletandis, or unofulness of the information contained in this roport, or that the we
of any information, apparatus, method, or procesi diaload to this roport may not Intringo
printly owned ricot; or
B. Asnimes say Ilabilities with respect to the une of, or for damages resulting from the
im of any inforauation, appunto, method, or process declosed in this report.
Ao wand lo the above, "persoa acting on behalf of the Communelon" includes any on.
ployw or contructor of the Commission, or employee of such contra ne, to the oxtent that
such umployu or coatractor of the Commission, or omployee of such contractor propards,
dienominator, or provides accoso to, way information purnuat to his employment or contract
with the Commission, or his employment with such contractor,
ELECTRON BEAM FLOATING ZONE REFINING OF NIOBIUM
R. E. Reed

RELEASED FOR ALKOHERCEDUNT
IN HKCIGAR SCIENCE ABSTRACTS
SOLID STATE DIVISION
OAK RIDGE NATIONAL LABORATORY
Operated by
UNION CARBIDE CORPORATION
for the
U. 8. Atomic Energy Commission
Oak Ridge, Tennessee
April 1966
ELECTRON BEAM FLOATING ZONE REFINING OF NIOBIUM*
R. E. Reed
Solid State Division, Oak Ridge National Laboratory
Oak Ridge, Tennessee
INTRODUCTION
A great deal of interest has been shown in the properties of niobium.
Such properties as high melding point, low vapor pressure, moderate density,
good fabricability, low ductile-to-brittle transition temperature, and low
thermal-neutron absorption cross section, make this metal attractive for high
temperature and nuclear reactor applications. Some of these properties, par-
ticularly mechanical properties, depend markedly upon the purity of the metal.
(continued on next page)
*Research sponsored by the Research Materials Program for the
· U. S. Atomic Energy Commission under contract with the Union Carbide
Corporation.
This investigation was fnitiated to evaluate the electron beam floating
zone refining technique as a purification method for niobium.
Several investigators have utilized electron beam floating zone refin-
ing as a method for preparing pure niobiui single crystals for mechanical
testing.
Tlie purity of these single crystals was evaluated using resist-
ance ratios, hardness measurements, some chemical analyses, and the resolved
shear stress for yielding. However, most investigators have generally used
the electron beam floating zone technique as a convenient method to obtain
pure single crystals of niobium and have not tried to identify the purifi-
cation mechanisms that may be operating. This paper describes a study uti-
lizing resistance ratios, extensive chemical analyses, and mechanical prop-
erties to assess the importance of various purification mechanisms in the
electron beam floating zone refining of niobium.
EXPERIMENTAL PROCEDURE
A. Material Sources
The niobium used in this study came from two sources. One, a 1 1/8 in.
diameter rod, 3 feet long was purchased from Wah Chang Corporation. This
rod had been formed from an electron beam melted 3 inch diameter ingot (Heat
No. 31831). The as-received rod was cold swaged down to 0.193 dia. rod for
zone refining. Second, fifty poundɛ of niobium D3 granules from Lot No. 663
vas purchased from DuPont Corporation. Arc melting with a tungsten tipped
non-consumable electrode was used to drop cast the niobium granules into a
5/8 in. dia. water cooled copper mold about 6 in. long. Unfortunately, a
tantalum starting peg was used to start the arc. Both tantalum and tungsten
impurities were increased during this process. These small ingots were then
cold swaged down to 0.193 inch diameter rod for zone refining. Table 1 gives
the vendor's analysis and the ORNL analysis for each starting material.
-4-
Table 1
Chemical Analysis of Niobium Starting Material in
Weight ppm
Element
Wah Cheng Ht 31831
Vendor
ORNL
Analysis Analysis
DuPont Lot 663
Vendor ORNL
Analysi.
s Analysis
C20
01
S10
.
.
.
<20
<40
<50
pe
.
.
Nou
Mn
.
Мо
."
Ni
.
Pb
1.
204
... more
c
si
20
<20
<20
204
Sn
1.5
o
Ta
5500
225 (296)1
n
-
T-
.
-.. .-
54(73)1
<100
27
30
102
110
1102
.
35
.
602
.
4.4
.. .....
.
Analysis taken on material after arc melting to consolidate
· gramiles.
Analysis taken from the unmelted handle of one pass rods.
Average of Neutron Activation Data for one pass rods.
*The Si analysis on the spark source mass spectrometer was.
not taken as a true value sincə there were several glass
components in the spark chamber which could act as sources
of si. It is probably much less than this mumber.
B. Electron Beum Floating Zone Refining Equipment
Two different zone refining systems were used. The first which will
be designated system LB was based on an electron beam floating zone module
(Model No. 4-EBZ-6000) purchased from Material Research Corporation. The
scanner consisted of a stationary specimen holder and an electron beam gun.
mounted on a traveling stage that moved the gun parallel to the vertical
specimen rod axis. The position of the bottom specimen holder was adjusta-
ble from outside the vacuum chamber. The specimen chamber consisted of a
glass bell jar and aluminum receiver section resting on a stainless steel
base plate. Viton O-rings and gaskats were used in all the racuum se:ls.
' A 4-inch oil diffusion pump with a water baffle and liquid nitrogen cold
trap gave a pumping speed of about 150 liters/sec.at the entrance to the
specimen chamber. The base pressure obtained in the bell jar was usually
2 x 10-7 torr. The pressure during zoning ranged from 1-2 x 10-6 torr for
the first zoning pass to about 3-5 x 10^' torr after about the fourth pass.
This was attributed to decreasing amount of outgassing from the specimen,
scanner, and specimen chamber as the zone refining progressed.
The electron beam gun was a work accelerated type with ar anmular wire
· filament as the electron source. The gun design consisted of grounded tan-
talum focusing plates 8 mm apart, a plate aperture diameter of 14.5 mm, and
a 23 mm diameter anmular filament of 0.51 mm diameter tungsten wire. The
high voltage supply had a voltage output variable to 5 KV and current to 400
ma. The emission current in the electron gun was maintained at a pre-set
constant value by controlling the filament power.
The second system which will be designated system BB was a completely
bakeable, high vacuum electron beam floating zone refiner based upon a scanner
Or

-6-
and power supply (Model EBZ-94) also purchased from Materials Research
Corporation. The general features of the scanner were the same as that
of system LB except that it was somewhat larger and was designed for low outgas-
sing.
The specimen chamber consisted of a 18 inch diameter water cooled
stainless steel bell jar 20 inches higu mated to a receiver section with
.
A
-
i
T
a Wheeler flange using a copper wire crush gasket. All feed-through ac-
cessories were bakeable and flanged to the receiver using copper gaskets.
A liquid nitrogen trapped, 10 inch stainless steel 011 diffusion pump
hacked with a liquid nitrogen trapped 4 inch oil diffusion pump and a 425
1:lters/min. mechanical pump, delivered a pumping speed of 1750 liters/sec.
at the entrance to the specimen chamber.
The specimen chamber was generally baked at 400°C for 16 hours prior
5 x 100% torr. The pressure during zoning ranged from 2 x 10° to 5 x 10
torr during the first pass depending upon the starting material. At about
the fourth pass the preseure was 6-8 x 10° torr. It was noticed that
after cooling down the specimen chamber after zoning the bese pressure
often got down to 1-2 x 10°9 torr.
The electron beam gun design used in this unit was identical to that
used in system LB.
c. Zoning Procedures
The swaged (4.8 mm diameter) niobium rod from both sources was elec-
tron beam floating zone refined at a zoning speed of 8-10 cm/br. With the
electron gun configuration described previous ly, the molten zone was about
3.5 mm long at a power level of 280 watts. The filament was kept at as
low a temperature as possible to minimize W contamination. At a beam voltage
-7-
of 4000V and a beam current of 70 ma the filament appeared much cooler than
the molten zomne. The Wan Chang material was zoned 1, 2, 4, 6, and 12 passes
in system L. This resulted in zoned single crystal lengths approximately
19 cm. long. In addition, 1 and 4 pass rods of the Wah Chang material was
-
-
zoned in system BB. This resulted in zoned lengths of about 26 cm. Rods
--
-
-
-
of the DuPontm niobium were zoned 1 and 4 passes in both systems.
In addition, zone refined single crystal nioblum rods were seeded 80
as to produce rod axis orientations at a single orientation near the center
of the stereographic triangle where the Schmid factor on the (101) [111]
slip system has its meximum value of 0.5. These were zoned in system LB
to produce 1 3 4 6, and 12 pads single crystals from both material sources.
All rods were zoned from bottom to top and the rod was melted through
-
-
-
-
at the end of the last pass.
D. Resistance Ratios
The as-Zoned rods were first prepared for resistance ratio measure-
ments along the zoned length. Lengths of 0.51 mm diameter platimum wire
about 3 cm lang were beaded on one end by melting it in a small gas torch
flame. The breaded end of these wires were then spot welded to the zoned
rod using a wery light contact pressure. Copper wire current and voltage
leads were then soft soldered to the platimum taps. The position of these
taps are shown in Fig. I which 18 a schematic drawing of a typical as-zoned
rod from system LB. Since niobium has & superconducting transition tem-
perature of 9.2°C, the low temperature resistance measurements were made in
a liquid hydrogen bath (20.2°K). The high temperature resistance measure-
ments were made in a mineral oil bath at room temperature (296°K). The re-
istance ratios were calculated as the resistance at room temperature divided
by the resistence at liquid hydrogen temperature
.8.
-.--.-.--
E. Chemical Analysis
.
After the resistance ratios were determined, the rod was next cut into
.
sections as shown in Fig. 1. The cuts were made using a high speed abrasive
......
cut-off wheel with water coolant. Small discs 3 mm long, 4.8 mm in diameter
-
_
-
were cut from the center of each section which had a resistance ratio meas-
.--
.-
-
.
.
--
urement. Each remaining section (mumbered in sequence in Fig. 1) was bev-
--.
eled on the end toward the top of the roá where it had been melted off at
the end of the last pass. All of the sections were then given a deep chemi-
cal polish in a solution of nitric acid (70%) and hydrofluoric acid (48%)
mixed in the proportions of 3:2.
The discs were used for neutron activation analyses for Ta and W 1m-
purity. The other sections were first used to get spectrographic analyses
for all metallic impurities using a spark source mass spectrometer (Asso-
ciated Electrical Industries Model MS 7). These analyses were taken by
· striking an arc between adjoining sections taken from the rod. Thus mass
spectrographic analyses were taken at essentially the same position as the
neutron activation analyses. In fact, the W analysis as determined by neu-
tron activation was used as an internal standard for the determination of
the metallic impurities at each spark position along the zoned length.
Next, each section was again chemically polished and approximately two-
thirds of each section used for a carbon analysis and the other third for
oxygen, nitrogen, and hydrogen analyses. Carbon content was measured by
-cir..
the Leco conductometric method. Oxygen, nitrogen, and hydrogen contents
were determined using the vacuum fusion techique with a platinum bath at
2000°C.
.
..
2
... .. siecio
.
.
wwwpo
.compon
g
onom. .----
F. Tensile specimens
The seeded zone refined niobium single crystal rods that were used for
tensile specimens were not sampled for chemical analysis. It was assumed
that these rods would be similar in purity to those prepared in a duplicate
procedure and analyzed. These rods were cut into 3.8 cm long sections
using the abrasive wheel. The εections were then centerless ground to pro-
duce a gage section 1.9 cm long and 2.160 mm in diameter with a taper of
less than 0.005 mm. After grinding, about 500 microns were removed from
the diemeter by chemical polishing in the HF-HNO, solution described pre-
viously. This was sufficient to eliminate all signs of cold work in the
Laie back-refection patterns. Furthermore, the uniformity of orientation
was checked by traversing the sample axially in the X-ray beam and no as-
terism or splitting of the LÀue spots could be detected.
The perfection of some seeded one pass rods of Weh Chang niobium
grown in system LB was studied using etch pitting, Berg-Barrett X-ray to-
pographs, Borrmann anomalous X-ray topographs, and double crystal X-ray
spectrometry. The dislocation density of the as-zoned rods ranged from
10" - 10° cm
. The majority of the dislocations were in sub-boundaries.
The misorientation or these sub-boundary walls were generally less than
180 seconds. The largest sub-boundary misorientation found was 8-10 min-
utes. This work will be reported in more detail in another paper. 8
The tensile specimens were pulled in tension at room temperature at
a strain rate of 1.8 x 10-4 sec-2. A table model Instron tensile testing
machine was used. The orientation of the tensile axis as a function of
strain was determined for one pass Wah Chang niobium' by taking back re-
flection Laue X-ray patterns every 5 or 10% tensile strain increment. slip
line observations were also made at each tensile strain increment.
-10.
RESULTS
A. Resistance Ratios
The resistance ratios of the DuPont and Wah Chang starting materials
were 40 and 28 respectively. Table 1 shows that the Wah Chang material had
190 ppm more metallic impurities and 50 ppni more interstitial impurities
than the DuPont inaterial. This apparently was sufficient to cause the dif-
ference,
Resistance ratios plotted as a function of zoned length for 1, 2, 4,
6 and 12 pass Wah Chang material zone refined in system LB, are shown in
Fig. 2. These data indicated that the finish end of the zoned length was
slightly purer than the start end. The first pass resulted in the greatest
resistance ratio change per pass. Subsequent passes had a contimual de-
creasing effect upon the amount of increase in the resistance ratio.
The DuPont niobium was also zone refined in system LB for 1 and 4
passes. The resistance ratios measured along the zoned length for this
material, compared to the 1 and 4 pass Wah Chang niobium are shown in Fig.
3. The DuPont niobium had much higher overall resistance ratios and also
exhibited a greater variation in resistance ratio from the start to the
...--
finish of the zoned length. The finish end again had a higher ratio.
..
...
The vacuum existing in the specimen chamber during zone refining also
-
.
had an effect upon the final resistance ratio. Fig. 4 is a plot of resist-
ance ratio versus zone length for Wah Chang niobium zone refined 1 and 4
passes in systems LB and BB. The high vacuum system obtained a larger in-
crease in the resistance ratio per pass for the same material.
.ll.
B. Chemical Analyses
1. Metallic Impurities
The major metallic impurities in both starting materials were Ta and
W. Neutron activation analyses for Ta and W in the Wah Chang niobium Zone
refined in system LB gave the results tabulated in Table 2. For all rods
tested, there was no significant variation of the tantalum or tungsten con-
tent along the zone length. In addition, there was no difference in tan-
talum content as a function of number of passes. The arithmetic mean (X)
of the tantalum contents reported for all discs from the Wah Chang niobium
was 367, 369, and 367 wt. ppm for 1, 6, and 12 pass rods respectively. How-
ever, there is some indication of a slight tungsten impurity increase as a
function of member of passes possibly due to contamination from the elec-
tron gun tungsten filament. For this case, the arithmetic mean (X) for
tungsten content was 227, 229, and 242 wt. ppm for 1, 6, and 12 pass mate-
rial respectively. Thus there was about a 6% increase in the amount of
tungsten in the zoned refined rod after 12 passes. The same result was
found for the Wah Chang niobium zoned in system BB. However, no signifi-
cant W increase could be found between 1 and 4 pass material.
The neutron activation analyses for Ta and w in the DuPont niobium
zone refined 1 and 4 passes in both systems also showed no significant
variation along the zone length. There was no difference in Ta and W con-
tent as a function of member of passes (up to 4 passes). The results showed
the DuPont niobium to have about 300 ppm Ta and 70 ppm W.
The only other metallic impurities above 2 ppm in the Wah Chang start-
ing material were Cr, Cu, Fe, Mg, and Zr. The first pass reduced all of
these metallic impurities except Zr to below 1 ppm. There was no significant
-
-
-
-
•12-
-
.
.
.
Table 2
Neutron Activation Analyses for Ta and w in Wah Chang Niobium
Zone Refined in System LB
1 Pass
6 Pass
12 Pass
Distance from
start of zone
(cm)
364_216358
Handle
1.5
3.4
363224
231
221
243
377
232
260
244
369
53
369
365
372
366
216
225
359
238.
365
220
242
240
374
224
1230
227
7.2
9.1
11.0
12.9
14.8
16.7
267
| 369
362
363
| 367
367
239
227
317
8
235
368
357
221
362
382
| 366
| 232
365
238
372
230
223378
250
242
Mean (X)
Std. Dev.(a)
367227
4.
73 .
7
7
369
.7
229367
15.1 6.
5
7
.7
一一一一一一一​.……
..
-13-
trend along the zoned length to indicate a zone refining action. However,
the data for the Zr impurity did indicate a movement of Zr with the zone
direction. In addition, each zoning pass lowered the overall Zr content
of the niobium. Fig. 5 18 a plot of the Zr content along the zone length
for 1, 6, and 12 pass Wah Chang niobium in system LB. This figure shows
that after 12 passes the Zr content was reduced from 27 ppm to about 2 ppm
with a slightly higher level of Zr at the finish end of the rod.
Other than Ta and W, the other metallic impurities above 2 ppm in the
DuPont aiobium were ca, Cr, Fe, Mg, and Ni. After one pass of this mate-
rial in system BB, all of these impurities were below 1 ppm. There was no
significant trend along the zoned length for any of the metallic impurity
analyses.
2. Interstitial Impurities
The vacuum fusion analyses for hydrogen in the Wah Chang niobium zoned
in systera LB showed no significant variation of hydrogen content either
along the zoned length of individual rods or among zoned rods which were
zone refined 1, 6, or 12 passes. The arithmetic mean (X) for all analyses
was 4.0 wt. ppm with a standard deviation of 1.1 wt. ppm.
The vacuum fusion results for oxygen and nitrogen contents in the Wah
Chang niobium zone refined in system LB are best summarized in Figs. 6 and
7. Figure 6 is a plot of the oxygen content in wt. ppm as a function of
distance along the rod for 1,6 and 12 pass material. There was a marked
drop of oxygen content from X = 116 wt. ppm in the handle to a level of
X = 30 wt. ppm for the one pass rod. The arithmetic mean (X) of the oxygen
contents in the zoned region of the 6 and 12 pass rods were 25 and 24 wt.
ppm respectively. There was such scatter in the results that no variation
-14-
in oxygen cortent along the zone length could be found. In fact, some of
the oxygen analyses were not used in calculating the arithmetic means since
they were so far from the general trend of the curves. Howeve:', there was
the indication that the first pes3 6!eat reduced the oxygen contert and
subsequent zoning passes reduced this oxygen level only slightly.
The nitrogen content analyses are shown in Fig. 7. In this figure,
the nitrogen content in wt. ppm 18 plotted versus distance along the zoned
..
rod for 1, 6, and 12 pass material. This impurity behaved in a manner simi-
lar to that of oxygen. The first zoning pass decreased the nitrogen content
from X = 61 wt. ppm in the handle to X = 24 vt. ppm. Increasing the number
of passes to 6 and 12 decreased the nitrogen content to X = 9 and 5 wt. ppm
respectively. In this case, increasing the number of passes decreased the
nitrogen content quite clearly although the scatter still prevented assign-
ing a significant variation of nitrogen along the zone length.
The carbon analyses made by the Leco conductometric method showed that
for the Wah Chang niobium zone refined in system LB there was no significant
variation of carbon content along the zone length. Table 3 lists these re-
sults. However, as the number of zoning passes increased the carbon content
-
.-
-.-
..
__
.
increased. The arithmetic mean of the carbon content in the zoned section
.
of the l, 6, and 12 pass rod was 11, 15, and 45 wt. ppm respectively.
...
*
There was a major difference in the [c] / [0) ratios between the Wah
.
-
Chang and DuPont starting material. These ratios were 0.09 and 0.78 re-
-
-
spectively. This caused a difference in the purification behavior of these
.::
.
two niobium starting materials. Table 4 lists the C, Og, Ng, and H, analysis
for 1 and 4 pass DuPont niobium zone refined in system LB. In this case, the
o, content after four passes dropped to less than 5 ppm with a decrease in
the carbon content from 55 ppm to about 23 ppm. The niobium content also
........
.
- -
rom
-
. .
.
.
.
-15-
Table 3
Leco Conductometric Analyses for Carbon in Wah Chang Niobium Zone
Refined in System LB (Weight ppm)
1 pass
6 pass
12 pass
Distance from
start of zone
(cm)
Handle
Handle
0.7
2.5
4.3
6.2
8.1
10.0
11.9
Il Eū ooo ooo E
13.8
15.7
17.6
12
Mean (X)
Std. Dev. (a)
3.2
HV
wwwwwwwwwwwwwwwww
-16-
Table 4
Interstitial Impurity Analyses in DuPont Niobium Zone
Refined in Sys tem LB (Weight ppm)
ST

Relative
Position on
Zoned Length
Il pass
4 pass
CO2 Ng
Na Ha
55 70 71 51 88 8 3.
| 25 14 <5 1 19 <5 «5 <1
Handle
m
Near Start
End
Near Finish
nd
1127
<5
<5
1
27
35
<5' 2
"
:
..
..
.
.
.
q
ve
.
.
.
.
mo
.
..
. ...
m imo
.
.
-
----
.-...
..
ooo....so..
-.
-..
EU
.. ... ----
TA
-17-
decreased to less than 5 ppm. Table 5 lists the C, Og, Ng, H, analyses for
the same material zoned one pass in system BB. The higher vacuum in this
system apparently caused the same interstitial impurity decrease in one zone
refining pa88.
C. Tensile Testing
One pass Wah Chang niobium single crystals with a starting tensile axis
orientation shown in Fig. 8 (position at of tensile strain) deformed on the
(101) (111) primary slip system up to 35-40% tensile strain. At point A on
Fig. 8, the tensile axis rotation deviated from the predicted path after
"overshooting" by some 8º the symmetry boundary where the conjugate slip
system was expected to start operating. Also, at 35-40% tensile strain,
slip lines due to a second slip system began to appear on the specimen sur-
face. Figure 9 18 a photomicrograph of such a one pass Wah Chang niobium
single crystal pulled to 30% tensile strain. There was no indication of
slip on a second system at this tensile strain.
The tensile curves agreed in their general features with those publi-
shed by Mitchell, Foxall, and Hirsch. There was a yield point followed
by a region with an initially high hardening rate (stage 0) decreasing to
the low rate characteristic of stage I. At about 30 to 40% tensile strain
(approximately 0.6 to 0.8 shear strain) a transition region began which
preceded a stage II hardening rate which was considerably higher than stage
I. Stage III was characterized by a decreasing work hardening rate leading
to fracture. All fractures were "chisel point" type with almost 100% reduc-
tion in area.
Figures 10 and 11 are the tensile curves up to 1.0 shear strain for 1,
4, and 12 pass material zone refined in system LB using Wah Chang and DuPont
ICEL
CD-
.
-18-
Table 5
Interstitial Impurity Analyses in DuPont Niobium Zone
Refined One Pass in System BB (Weight ppm)
Relative Position

c
on
N,
H
on Zoned Length
Handle
74
«5 <1
...
Near Start
13
<5
-
L
:
End
-
Near Middle
of Zoned Length
<5
«5
<1
Near Finish
End
22
<5
35
1
imouthoo.com
.
-..
.
...--
-
w
.
-..
.....-
ww
+
.simo
.
-19-
niobium respectively.
Table 6 lists the resolved shear stress at the lower
yield point as a function of numher of passes in system LB for both starting
materials. This shear stress decreases with increasing number of passes for
both starting materials. The arop in the shear stress per pass decreases
as the number of zoning passes increases.
The DuPont niobium had the lowest
shear stress of 1.19 Kg/m after 12 passes in system IB.
The yield points shown in Figs. 10 and 11 were extremely interesting
when compared to the interstitial impurity analyses. A wide, yield point
and large yield drop with an extensive stage O (from a stream strain of 0.05
to 0.35 in Fig. 10) was characteristic of the Wah Chang 1 pass niobium single
crystals. The 12 pass Wah Chang niobium sinc:le crystal tensile curve cxhibited
a much sharper yield point and a smaller yield drop with a shorter stage o
(Fig. 10). The chemical analyses were 11 ppm C, 30 ppm 02, 24 ppm N, for the
1 pags n.obium and 45 ppm C, 24 ppm Oz, 5 ppm N, for the 12 pass. This suggests
that a sharp yield point and short stage O may be associated with the carbon
impurity while the wide yield point and extensive stage 0 are due to oxygen
and/or nitrogen.
The results on the DuPont single crystal niobium (Fig. 11) support this
suggestion. In this case, the one pass DuPont niobium single crystal ex-
hibited a much sharper yield point, smaller yield drop and a less extensive
state O than the one pass Wah Chang niobium. The 12 pass material had es-
sentially no yield point and a very short stage 0. In this material there
was ~ 26 ppm C, w 10 ppm Oz, < 5 ppm N, after one pass and ~ 23 ppm C, < 5
ppm 0g;: < 5 ppm N, after four, passes. The twelve pass material has not been
· analyzed yet. Again, the s'jaro yield point and short stage O could be as-
sociated with the carbon interstitial impurity. Further work will be done
on this point.
-20-
Table 6
Resolved Shear Stress at the Lower Yield Point for Niobium
Zone Refined in System LB
Starting
Material
Number of Zone
Refining Passes
at 10 cm/hr
Resolved Shear Stress at
Lower Yield Point in
Kg/mm
2.48
1.96.
Wah Chang
Ht. 31831
Niobium
1.72
1.57
1.49
1.32
ñ a Friul afwn
1.79
DuPont
Lot 663
Niobium
1.57
1.40
1.28
1.19
-21-
DISCUSSION
The present investigation generally agrees with the previously published
results on the purification of niobium by electron beam floating zone refin-
ing · Votava) after electron beam floating zone refining niobium from Murex
Company at a zone velocity of 84 cm/hr in a maximum vacuum of 3 x 20°° torr,
concluded that the purification largely takes place by vaporization of impu-
rities with only a small zone refining effect. This was largely based on
resistance ratios (R 273°K/R 77°K) taken on the zoned length. There was a
slight indication that the finish end of the zóne was purer than the start.
The same trend was found in the present investigation. However, the re-
ported resistance ratios (R 273 *K/R 77°K) of Votava (3.86, 5.05, 5.15, 5.10
for 0, 8, 16, and 27 passes respectively) indicated that the 27 pass mate-
rial was more impure than the 16 pass rod. Since niobium metal was used
throughout for the electron gun construction, this contamination must have
been interstitial impurities.
A resolved lower yield shear stress of 2.6 kg/mm was reported by
Votava for 3 pass (at 84 cm/hr) Murex niobium with a tensile axis orienta-
tion and tensile strain rate similar to the present investigation. A Vickers
hardness number of 44-48 was also indicated. His gas analysis for this mate-
rial was 0.2-1.5 ppm H2, 12-17 ppm NZ, and 10-26 ppm Oz. No other analyses
were given. This material compares with one pass Wah Cheng niobium zone
refined at 10 cm/hr in system LB. Here, the Vickers hardness number was
2.5 Kg/mom, and the gas analysis was 24 ppm NZ, 30 ppm Oz, and 4 ppm Hg.
The Murex niobium generally has a rather high carbon content (~ 10 ppm)“
and high Ta content (1200-3000 ppm) 3,9 compared to the material sources used
in this study. This could explain the difference between the flow stresses
-
-
-
-
-
- 22-
reported by Mitchell, Foxall, and Hirsch and those given in Table 6. They
reported values for the resolved shear stress at the lower yield point
ranging from 3.5 to 1.8 Kg/mm for 1 to 6 pass Murex niobium. This was
obtained on material zoned in < 5 x 10° torr vacuum at 30 cm/hr for all
passes except the last one which was at 12 cm/hr. The tensile axis orien-
tation was similar to that used in this study but the strain rate was
lower
(4.5 x 10
sec). In addition, Mitchell, Foxall, and Hirsch.
reported that after 6 passes the niobium picked up extensive Ta and W
contamination from the electron gun while the On, Ng, and H, content de-
creased only slightly. No carbon analyses were reported.
It 18 unfortunate that the major metallic impurities that are gen-
erally found in niobium, namely Ta and W, are also the metallic impu-
rities that can not be removed from niobium by electron beam floating
zone refining. The present investigation has shown that these two impu-
rities did not vary along the zoned length and that Ta did not decrease
with increasing nuniber of passes. In addition, the W content increased
slightly as the number of passes increased probably due to contamination
from the tungsten filament in the electron gun. This occurred despite an
effort to keep the tungsten filament considerably cooler than the molten
zone in the niobium rod. The tantalum focusing plates evidently were not
a source of tantalum contamination. It was noticed that the edges of the
focusing plate aperture quickly became coated with niobium metal which
possibly prevented the sputtering of Ta from this area by electron bom-
bardment.
With the exception of the Zr impurity in the Wah Chang niobium and Ta
and W as discussed above, all other metallic impurities were reduced to
levels below 1 ppm in one zone refining pass at 10 cm/hr. In fact, the
-23-
levels became so low that it as r t
rrising to find that sou sorrice!.t
trend could be found in the spark source mass spectrometric data along ti:
zone length. In the case of the Zr impurity in the Wah Chang niobium, the
data (Fig. 5) suggested that the major mode of purification was still evapo-
ration accompanied with a small zone refining effect which indicated that
the impurity moved in the direction of the moving zone.
The data from the resistance ratio measurements (Figs. 2-4) indicated
that a slightly purer metal generally existed at the finish end of the zoned
length. Thus the small zone refining effect on the Zr impurity which would
move the 2r to the finish end has little to do with the overall purification.
It 18 concluded that with the exception of Ta and W, the metallic im-
purities in niobium are removed mostly by evaporation to levels less than
1 ppm by the electron beam float:rg zone technique.
•.
There was a good correlation between the resolved shear stress at the
.-
...
.
a..
..
lower yield point and the resistance ratio of the nioblum single crystals.
The shear stress for yielding in niobium is strongly influenced by the in-
terstitial C, 09, Ng, and H, impurity content. Therefo.re, it is reasonable
to assume that the resistance ratio was also strongly influenced by the in-
terstitial inpurity content. The resistance ratios were generally constant
with zoned lengte with a slightly higher ratio at the finish end (Figs. 2-4).
This suggests that there was no significant trend in interstitial impurity
content along the zoned length. The chemical analyses also indicated that
this was probably so (Figs. 6-7, Tables 3-5).
-coco
.
..cu
..-
However, the overall resistance ratio of a zoned length depended upon
the purity of the starting material, the vacuum level during zoning, and
the mumber of zone refining passes (Figs. 2-4). The overall impurity level

-24-
of C, Og, Hg, and N, also depended upon these variables (Figs. 6-7, Tables
3-5). The effect that these experimental variables have upon the purifi-
cation mechanisms operating in the case of C, 00, Ng, and H, impurities can
be seen more clearly by relating them to more basic variables. These are
as follows:
L
A. Time at a high temperature--This is directly related to number
of passes and/or zoning velocities for a given electron gun
design and beam power.
B.
Partial pressures of C, 0,., No, and H, above the hot niobium
metal--This is directly related to the total pressure in the
specimen chamber during zoning. Also, it 18 related to the
purity of the starting material if it is realized that outé
gassing from the niobium is a major part of the vacuum pumping
load especially in the first few passes.
--
There are three probable purification mechanisms operating during the
tena baan 97405 ZIE 5 C I ET IP : ther-
stitial impurity level attained during the process. These mechanisms are:
-
'
1. The carbon monoxide and/or carbon dioxide reactions.
2. Attainment of equilibrium concentrations of 0g, Ng, and H, in
the solid niobium adjacent to the molten zone.
3. Removal of O, from the solid niobium adjacent to the molten
zone by the formation and volatilization of atomic oxygen and/
or niobium suboxices.
The role of these purification mechanisms in the electron beam floating zone
-------
technique will be discussed separately.
-
Moi
cora: ... ..
1. The Co or co, Reaction
The different [c]/[02] ratios of the starting materials resulted in a
difference in the final levels of the C and on content in the zoned niobium.
The results show that for a [c]/[02] ratio of 0.78 the oxygen content decreased
rapidly to below 5 ppm after 4 passes in system LB. For å ratio of 0.09 the
sin..me.com. con me.. S
-
--n
:
.....
donosi na sio
-25-
oxygen content only dropped to 25 ppm after 6 passes in system LB. The pos-
bible mechanism acting here 18 the formation of co or co, gas in the molten
zone which then escapes from the molten metal into the vacuum chamber. A
low initial (c]/[0,] ratio in the metal would prevent much oxygen removal
by this process. A high ratio could leave excess carbon. It appears that
the proper ratio should be about [c]/[0 ] = 0.5. In addition, there is
probably a carburization or decarburization effect occurring on the solid
niobium rod depending upon the temperature ard the partial pressures of Co
and co, existing in the vacuum system. This could account for the carbon
pick up after 12 passes of the Wah Chang niobium in system LB. This effect
would be kinetically blower than the reaction occurring in the molten zone
because the rate controlling process is probably diffusion of the inter-
stitials in the metal or surface adsorption-desorption at the surface.
There is some indication that the molten zone is protected from any
interaction with gases in the vacuum environment. Allentº using the pendent-
drop method to measure surface tensions of the refractory metals, found that
wide variations in vacuum from 2004 to 10°7 torr as well as leak rate varia-
tions had no measurable effect on the surface tension determinations. Be-
cause of the sensitivity of the surface tension to interstitial impurities,
he concluded that the vaporizing metal protected the surface from impingement
of impurity gas molecules. Therefore, any pick up of interstitial impurity
during zoning is probably due to the interaction of the solid niobium rod
with the residual gases in the vacuum chamoer. Co and co, are conmon residual
11
gases in oil diffusion pumped systems.
Outgassing from the molten zone and
heated solid niobium metal would also contribute co and co, residual gases.
-26-
2. Attainment of Equilibrium Concentrations of 00, Ng, and H,
Pasternak
has extensively studied the oxidation and nitridation of
niobium in vacua in the 10% to 10° torr range. For the case of nitrogen,
he proposed that absorption and desorption are the slowest steps in the over-
all sorption mechanism and thus deternine the observed kinetics. The pro-
· posed reaction steps were:
NA (gas) + A - NA
NA
+ 0 = NO + A
* N. (gas )+ @ = NC
The symbols A and
represent surface and interstitial sites respectively.
in the niobium. At equilibrium the following relationship holds:
Co = p7/2 exp [Coin/
where Ce = equilibrium concentration of nitrogen in niobium (atom %)
P = partial pressure of nitrogen in equilibrium with metal (torr)
AHO2n = heat of solution
Q = entropy factor (torr-1/2. atom %):
R = gas constant
T = temperature
From the experimental kinetics of nitrogen pickup and outgassing and
assuming the above reaction stops, Pasternak was able to obtain values of a
and AH that agreed well with the values obtained under equilibrium conditions ,
These values were: d = 1.2 x 10-5 torr-1/2. atong; AH = 53,500 calories.
-27-
..
.
The equation for the equilibrium concentration shows that a low partial
pressure of nitrogen and a high temperature contribute to a low concentration
of nitrogen in the metal. However, the kinetics for the attainment of this
equilibrium concentration 18 controlled by the adsorption and desorption of
nitrogen at the metal surface. This indicates that the lowest nitrogen con-
centration will be attained by heating the niobium to a temperature near the
....
.
-
melting point in a high vacuum for a time lorig enough to reach equil?brium
and then cool the sample fast enough to prevent adsorption and desorption
processes from maintaining an equilibrium concentration as the temperature
decreases.
The slight continual increase in the resistance ratio as a function of
zone length generally observed in this study (Figs. 2-4) can be explained by
progressed particularly on tbe first pass 2:1 system L3 which was an
ibakaabie
syster (Fig. 3). Thus the partial pressure on N, and o, probably decreased
as zoning progressed, resulting in lower equilibrium concentrations of oxygen
and nitrogen. The vacuum fusion analyses for oxygen and nitrogen unfortunately
exhibited too much scatter among sections along a zoned length to test this
suggestion.
3. Formation and Volatilization of Atomic Oxygen and/or Niobium Suboxides
In the oxygen-niobium system, Pasternakalso found the same general
sticking probabilities for oxygen on niobium below 1000°C as for nitrogen.
However, at higher temperatures in addition to adsorption and desorption,
two additional processes apparently occur which result in high "apparent
sticking probabilities." The first reaction is the dissociation of the oxy-
gen molecule on the hot surface followed by irreversible sorption of the
-28-
oxygen atoms on a cold surface in the system. The second reaction is the
formation of NO at the hot niobium surface followed by the evaporation of
the NbO molecule from the surface to a wall of the system.
Gebhardt, Fromm, and Jakob s also suggested that their results from
degassing experiments on niobium could be explained by assuming that oxygen
leaves the metal in the form of a volatile niobium suboxide such as Nbo or
Noo, at high temperatures.
Again it 18 apparent that a low partial pressure of oxygen in the sys-
tem 18 very important. However, the presence of the competing reactions in
addition to the sorption process for the case of oxygen suggests a method of
purification. This method would consist of heating niovium at very high tem-
.
.
-..
peratures in a good dynamic vacuum with low partial pressures of nitrogen and
oxygen. This would reduce the equilibrium concentration (C) of both impu-
rities to the lowest practical value.
This should be done in a cold wall
system such that the hot specimen can "see" the vall. Thus the competing
reactions for oxygen would result in the deposition of oxygen atoms and NbO
molecules irreversibly on the cold wall. Fast cooling from temperature would
retain the low equilibrium concentration of N, and oz.
Taylor and Christian used essentially this treatment by resistance
heating a 3 mm diameter zone refined niotium rod to ~ 2400°C in ~ 1 x 10°
torr vacuum. This treatment improved the resistance ratio R 300°K/R 20°K of
the as-zoned rod from ~ 100:1 to ~ 2000:1. Their value of resolved shear
stress for yielding in compression at a strain rate of 6 x 10-4 sec - was
0.85 Kg/com. Using the data of Cost and Wertł4 for the equilibrium solubility
of nitrogen in niobium and extrapolating to high temperatures and low pres-
sures, Taylor and Christianº explained the purification effect as due to the
attainment of a low equilibrium concentration of oxygen and nitrogen. It is
-29-
suggested here that some of the purification was due to the format ca and
deposition on a cold surface in the system.
Wronski and Fourdeux? used a "flash annealing" technique which proba-
bly accomplished the same result on polycrystalline niobium sheet. They
obtained a resistance ratio (R 295°K/R 77°K) equal to 5.68 and a resolved
shear stress for yielding of ~ 2 Kg/mm2. Their vacuum was 1 x 100 torr
at a time to temperatures near the melting point.
In the electron beam floating zone technique for the purification of
niobium, all three of these mechanisms probably affect the interstitial im-
purity content of the zoned rod. The kinetics of these mechanisms, i.e. the
CO, co, gas reaction, the attainment o
culiibrium concentrations cf impu-
rities in the metal, and the formation a:, volatilization of atomic oxygen
and/or niobium suboxides, depends upon such variables as time, temperature,
and the partial pressure of the appropriate gas above the metal surface.
Experimental factors that affect these variables are purity of starting mate-
rial, quality of the vacuum environment, number of passes, zoning speed,
electron gun design, and beam power.
SUMMARY
The results reported here on electron beam floating zone refined niobium
indicated that very little purification was due to movement of impurities by
the zone refining action. Instead, most metallic impurities were reduced by
evaporation. Tantalum and tungsten were exceptions to this in that there
was no significant reduction in the content of either impurity. In fact,
tungsten showed a 6% increase at the end of 12 passes probably due to con-
tamination from the electron gun tungsten filament. The [c]/[0,] ratio of
-30-
the starting material was important and affected the final level of these
.
two impurities. For one starting material, the carbon content increased
during zone refining possibly due to a carburization reaction occurring
among the residual gases in the vacuum environment. The initial low hydro-
gen content showed little significant change as a result of zone refining.
Oxygen and nitrogen levels in the starting materials were greatly lowered
by the first pass made during zone refining. Subsequent zoning passes had
a smaller effect but the chemical analyses, resistance ratios, and tensile
properties did indicate a decreasing oxygen and nitrogen content with in-
creasing number of passes. This was attributed to a lower partial pressure
of oxygen and nitrogen in the vacuum environment as zone refining proceeded.
Adsorption and desorption of nitrogen at a heated, solid niobium surface
was suggested to be the rate controlling mechanism for nitrogen purification
during zone refining. The same held for oxygen except that the additional
reactions of dissociation of the oxygen molecule and the formation of NbO
on the hot niobium metal surface with subsequent vaporization of oxygen
atoms and NbO molecules to deposit on cold surfaces in the system also play:
an important role in the purification process.
-31-
BIBLIOGRAPHY
1. E. Votava, Phys. Stat. Sol. 5, 421-434 (1964).
2. E. Votava, J. Less-Common Metals, 9, 409-415 (1965).
3. T. E. Mitchell, R. A. Foxall, and R. B. Hirsch, Phil. Mag. 8, 1895-1920
(1963).
4. B. Harris, J. Inst. Met. 22, 89-92 (1963-64).
5. B. Harris, J. Less-Common Metals 1, 185-196 (1964).
6. G. Taylor and J. W. Christian, Acta. Met. 13, 1216-7218 (1965).
7. J. W. Christian and B. C. Masters, Proc. Roy. Soc. (London) A, 281, 223-
239 (1964).
8. R. E. Reed, H. D. Guberman, and T. 0. Baldwin, to be published.
9. M. J. Leadbetter and B. B. Argent, J. Less-Common Metals 3, 19-28 (1961).
10. B. C. Allen, Trans . AIME 227, 1175-1183 (1963).
11. H. Inouye, ORNL Report No. 3674, September, 1964.
12. R. A. Pasternak, Stanford Research Institute Report SRIA 132, November,
1964.
13. E. Gebhardt, E. Fromm, and D. Jakob, 2. Metallk. 55, 432-444 (1964).
14. J. R. Cost and C. A. Wert, Acta Met. 11, 231-242 (1963).
15. A. Wronski and A. Pourdeux, J. Less-Common Metals 7. 205-211 (1964).
16. F. W. Young, T. O. Baldwin, A. E. Merlini, and F. A. Sherrili,
Advances in X-ray Analysis, Vol. 9, Plenum Press, New York, 1966.
ens. ATM
-32-
FIGURE CAPTIONS
Fig. 1
Schematic of a Typical Zone Refined Niobium Rod showing Sampling
Fig. 2
Procedure ior Chemical Analysis. Note: Spark source mass spec-
trometric, Leco conductometric, and vacuum fusion analyses were
made on the numbered sections as described in the text.
Resistance Ratio along a Niobium Single Crystal Rod as a function
of the Number of Floating Zone Passes. Source: Wah Chang Corp.;
zone speed: 10 cm/hr; diameter: 4.8 mm (nominal); system: LB.
Resistance Ratio along a Niobium Single Cryetal Rod as a function
of the Starting Material. Zone speed: 10 cm/hr; diameter: 4.8
mm (nominal); system: LB.
Resistance Ratio along a Niobium Single Crystal Rod as a Function
of the Vacuum Level. Zone speed: 10 cm/hr; diameter: 4.8 mm
(nominal).
Fig. 3
Fig. 4
.:
Fig. 5
The Zirconium Content along a Niobium Single Crystal Rod as a
Function of Number of Passes. Zone speed: 10 cm/hr; diameter:
4.8 mm (nominal). Note: This was data from a spark source mass
spectrometer using the W 184 mass line as an internal standard.
Fig. 6
Oxygen content along a Niobium Single Crystal Rod as a Function of
the Number of Floating Zone Passes. Source: Wah Chang Corp.;
zone speed: 10 cm/hr; diameter: 4.8 mm (nominal); system: LB.
Niobium Content along a Niobium Single Crystal Rod as a function
of the Mumber of Floating Zone Passes. Source: Wah Chang Corp.;
zone speed: 10 cm/hr; diameter: 4.8 mm (nominal); system: LB.
Fig. 7
-33-
FIGURE CAPTIONS (cont.)
Fig. 8
Rotation of the Tensile Axis Orientation as a function of Tensile
Fig. 9
Strain for a Niobium Single Crystal. Source: Wah Chang Corp.;
zone speed: 10 cm/hr; number of passes: 1; system: LB.
8lip Lines on a Niobiu Single Crystal after 30% Tensile Strain.
Source: Wah Chang Corp.; zone speed: 10 cm/hr; number of passes:
1; system: LB.
Resolved Shear Stress--Shear Strain Curves for the Tensile Deforma-
tion of Niobium Single Crystals as a function of the Number of
Floating Zone Passes. Source: Wath Chang Corp.; zone speed: 10
cm/hr; system: LB. Note: The curves are only plotted to a shear
Fig. 10
strain of 1.0.
Fig. 11
Resolved Shear Stress--Shear Strain Curves for the Tensile Deforma-
tion of Niobium Single Crystals as a function of the Number of
Floating Zone Passes. Source: DuPont Corp.; zone speed: 10 cm/hr;
system: LB. Note: The curves are only plotted to a shear strain
of 1.0.


ORAL-OWG 66-932
ZONING DIRECTION
START OF
ZONE,
HANOLE
MELTED OFF AT
END OF LAST PASS,
9
8
7
6
5
4
3
2
1
0
1
2
3
4 5 6
centimeters
7
8
9
10
11
12
13
14
15
16
17
18
19
1 * CURRENT TAP
V = VOLTAGE TAP
----- = POSITION OF CUT WITH HIGH SPEED ABRASIVE WHEEL
un = NEUTRON ACTIVATION ANALYSIS SPECIMEN

..
-
.
-.
ORNL-DWG 65-8167R
12 PASSES
6 PASSES
4 PASSES
RESISTANCE RATIO (RROOM TEMP/RLIQUID HYDROGEN)

2 PASSES
14 PASS
PASS DIRECTION-
2
0
16
18
2 4 6 8 10 12 14
DISTANCE FROM START OF ZONE (cm)

: ORNL-DWG 66-2696
- 4 PASS, DU PONT LOT 663 +
4 PASS, WAH CHANG HT. 31831
-4 PASS, DU PONT. LOT 663 —
RESISTANCE RATIO (RROOM TEMP/RLIQ HYDROGEN)
1 PASS, WAH CHANG HT. 31831
-
-
--
-
--
-
-
-
-
-
PASS DIRECTION
-
.
-
VACUUM RANGE: 3.4 x 10-6 TO 5.3 x 10-7 torr |
4
0
4 8 12 16 20 24
DISTANCE FROM START OF ZONE (cm)
28
32

ORNL-DWG 66-2695

4 PASS, 1.7 x 10-7 TO 6.3 * 10-9 torr
-
-
4 PASS, 2.5 x 10-6 TO 6.0 x 10^2 torr
RESISTANCE RATIO (PROOM TEMP/RLIQ HYDROGEN
1 PASS, 1.8 x 10-7 TO 1.1 x 10-7 torr
PASS, 2.8 x 10-6 TO 1.0x10-6 torr
PASS DIRECTION -
MATERIAL SOURCE: WAH CHANG HT. 31831
4 8 12 16 20 24
DISTANCE FROM START OF ZONE (cm)
28
32

ORNL-DWG 66-3042

STARTING MATERIAL ANALYSIS WAH CHANG HT. 31834
-
-
ZIRCONIUM CONTENT (wt ppm)
14 PASS
6 PASS
O
:
20
.
12
16
DISTANCE FROM START OF ZONE (cm)
•


ORAL - DWG 66-934
OXYGEN
X = 116; 0 =4.9 (HANDLE)
ao PASS
---+ 6 PASS
- 12 PASS
LINES CONNECT POINTS ALONG ZONED LENGTH THAT WERE USED
FOR MEAN VALUES (M) AND STANDARD DEVIATIONS (O)
ZONE DIRECTION
OXYGEN CONTENT Iwt ppm)
= 30; o = 6.3 (1 PASS)
START OF ZONE-
= 25; 0 = 5.6 (6 PASS)
X= 24; o = 2.9 (12 PASS)
10
8
6
4
2
2
4
6
8
0
12
14
16
18
20

:
.
-
ORNL-DWG 64 933
TROGEN
I
NITROGEN
1
o I PASS
+ 6 PASS
-A 12 PASS
4 REPORTED AS < -
to
AV
X = 61; 0 = 2.6
(HANDLE)
ZONE DIRECTION
LINES CONNECT POINTS ALONG ZONED LENGTH THAT WERE USED FOR
MEAN VALUES () AND STANDARD DEVIATIONS (O):
NITROGEN CONTENT ( wt ppm)
START OF ZONE-
X=24; 0 =4.0 (1 PASS)
X=9;0= 2.1(6 PASS)
X=5; 6 =0.8 (12 PASS)
10
8
6
4
2
0
2 4 6 8
DISTANCE ALONG ROD (cm)
10
12
14
16
18
20
..
.
.
.
...
.
-
...
---
$

ORNL-DWG 65-7642R
CHANGE IN ANGLE
BETWEEN TENSILE
AXIS AND SLIP DIRECTION
ad

TENSILE.
TENSILE STRAIN
004
014
Inizoane colk.000
O MEASURED DIRECTION
OF TENSILE AXIS
• CALCULATED DIRECTION OF
TENSILE AXIS FOR SINGLE
SLIP ON (101) [119]
111
SLIP DIRECTION

ORNL-DWG 66-2690
TENSILE AXIS
TENSILE AXIS

INITIAL CROSS SECTION
DEFORMED CROSS SECTION
K


.
.ro
meer...rec.
tinoms
Lambo
SLIP PLANE,
(104)-
:
:
:
:
[127), À
winter mo
n.
-ir:
...
Bangkin
..:me
;"
SLIP
DIRECTION
[111]
?
|
"*
VIEW IN À DIRECTION, 500X
.
Slip Lines in Single Crystal Niobium After 30 % Extension

ORAL-DWG 66-3043
MATERIAL: WAH CHANG HT. 34834 NIOBIUM
A PASS
RESOLVED SHEAR STRESS (kg/mm2)
4 PASS
· 12 PASS
co
0.2
0.4 0.6
SHEAR STRAIN
0.8
1.0

.. ..
.
---..-.
..
..
.. .
.. .--..

ORNL-DWG 66-3044
...
..
MATERIAL: DU PONT LOT 663 NIOBIUM
.....
........
RESOLVED SHEAR STRESS (kg/mm2)
4 PASS
-4 PASS
12 PASS
0.2
-
0.8
.
0.4
0.6
· SHEAR STRAIN
4.0
.
.

.
L
END
DATE FILMED
6 / 16 /66
-
-
-
.
.
**
.
*
*
-
i
-
-
*
1