Professional Documents
Culture Documents
Advances in Cryogenic Engineering
Advances in Cryogenic Engineering
Advances in
Cryogenic Engineering
VOLUME 3
K. D. TIMMERHAUS, Editor
Chemical Engineering Department
University of Colorado
Boulder, Colorado
Distributed by
PLENUM PRESS, Inc., NEW YORK
1960
ISBN-13: 978-1-4684-3107-0 e-ISBN-13: 978-1-4684-3105-6
DOl: 10.1007/978-1-4684-3105-6
Softcover reprint of the hardcover I st edition 1960
FOREWORD
i
ACKNOWLEDGMENTS
ii
CONTENTS
PAGE
ACKNOWLEDGMENT . . . .. . . . . . . . . . . . . . . . . ii
PAPERS PRESENTED
Cryogenic Proces se s
V. J. Johnson
CEL National Bureau of Standards
iii
PAGE
iv
PAGE
Cryogenic Applicactions
Cryogenic Equipment
vi
PAGE
Cryogenic Equipment (continued)
F. K. Dearborn
Air Fa rce Cambridge Research Center
vii
PAGE
Cryogenic Propertie s
viii
PAGE
Cryogenic Propertie s (continued)
R. J. Richards, R. F. Robbins,
R. B. Jacobs, and D. C. Holten
CEL National Bureau of Standards
J. G. Althouse
Lukens Steel Company
Introduction
-1-
Parahydrogen
-2-
He (col / gm) • f (P, T, C) • Hn- 1.68 (C-25)
.,-" .,-"
PRECOOL.ANT PRECOOLANT
A B
PRECOOLING TEMPERATURf; I_ se K
EXCEPT AS NOTED
SIMPLE l1NOE CYCLE
60 00
YIELD' 19~ YIELD· 16~ INLET PRESSURE - ATM
Figure 1. Figure 2.
, 15ATIoI H, ,1
"'IroAl.RA Hz
,
"'.9''Iro~VERS'i
0' I'-.. AT 20K
.. He PRESSURES
20~
"AT
10AT
.0
~
•
.0
•
. . .
HELIUM· HYDROGEN
CONDENSING CYCLE CLAUDE CYCLE
0 2
Ht PRESSURE· ATM
00
" 10101040110 00
INLET PRESSURE· ATM
Figure 3. Figure 4.
-3-
inlet pressure, plotted along the abscissa. The work of
liquefaction in kw. hr. / gal. of liquid is plotted as a function of
the inlet pressure for several conditions of para conversion
and heat exchanger .6. TIs.
-4-
almost superimposed on each other indicating the insensitivity
of the cycle to fluctuations in helium pres sure in this range.
Also of interest is the flatness of the curve between 3 and 10
atmospheres of hydrogen pressure indicating extremely stable
performance in this region.
Claude Cycle
-5-
below 65 0 K it is necessary to use a heat exchanger between the
precoolant bath and the engine inlet.
-6-
3.0
95't,PARA HZ
58 -95'1.CXJNVERSION I50 ATM '60 ATM\
·t
40ATM ,
AT 20'K~ ..
) l ......" ..
7~
5
-
......[7'" 80 ATM
:;,:::;;
, 100ATM
0
HEAT EXCHANGER AT. I' K
PRECOOLING TEMPERATURE· 60' K
PRODUCING NORMAL H2
0
DUAL PRESSURE CYCLE
o 10 20 30 40 50 60
INTERMEDIATE PRESSURE -ATM
Figure 5,
2.0
90'l.PARA H2 ~:~.) ......
- ~ ~:O'K
ca-9S'l.CXJNVERSION
AT 20·K-~"'II'" ... ot."" .. ""1
",. 30-K
6
~ '7'
2
8
INLET PRESSURE' 100 ATM
PRECOOLING TEMPERATURE '65'K
PRODUCING NORMAL H2
.4
Figure 6,
7.5
I
~
...J
<3..... VSIMPLE LINDE
I
----P~ESSURE
a: 6.0
l:
~, ILHe-~ CON::ING
/OUAL
t5 4.0
~
~
~
o
3.0
/ r---. ................
[CLAUDE - 6S'K ENGI NE
\
DUAL PRESSURE
:J WITH EXPANDER
u.
o
:!It: I.S
I
a: PRECOOLING TEMPERATURE· 6S'K
~ HEAT EXCHANGER AT, I'K
WORK INCLUDES PRECOQLANT
o 20 40 60 80 100 120
INLET PRESSURE -ATM
Figure 7.
-7-
3) For the production of parahydrogen the work of
liquefaction increases about 25 per cent while
retaining the same general characteristics.
Conclusions
-8-
When these other factors are considered,the dual pressure
cycle becomes more interesting. It has the advantage of
simplicity combined with a moderate efficiency. There are no
moving parts at low temperatures, and the system is relatively
insensitive to minor fluctuations in pressure, temperature,
and flow.
Discussion
-9-
Question by M. P. Sweeney: Did the Claude cycle shown on
the last graph expand into the two phase region?
-10-
A-2
V. J. Johnson
CEL National Bureau of Standards
Boulder, Colorado
~t:~I-~I------j l
I
l
VACUUM
REFRIG.
DRIER
ADSORPTION
PDR u:
LIQUEFIER I
~S
H I
r ----J
8'
OIL a
WATER I
SEPARATOR
I~:
-~I~II
__ ~_ ...JI
I-COMPRESSOR NZ VACUUM PUMP
I
• r....,
A~4~fr 4IE~~i,
FRO TO
COM. LIQUEFI ER
REFRIGERATION ADSORPTION
GAS SUPER- HYDROGEN SUPPLY DRIER PURIFIER
HOLDER CHARGER
1000
"ii
0'
0
(.)
......-....
I/)
I-- ~
100
Go
.... ....
I/)
....
......-
<II
Z
10
EXPERIMENTAL DATA
--EXTRAPOLATION BASED ON DATA
> OF DREVING ET AL.
I 11111111 I I 1111111 I I II f
1.0
RELATIVE PRESSURE. P/po
-12-
nitrogen impurity. The pure gas then returns through the
heat exchanger and goes to the liquefier. It will be noted
in Figure 2 that the total flow in this system thus passes
through the purification train. In some systems the purifica-
tion is done on just the supply gas. The recompression and
removal of oil from the recirculating gas is handled
separately.
1) Impurity level,
3) Pressure.
• 13-
Figure 4. Enhancement factors for nitrogen vapor in
hydrogen as computed by W. T. Ziegler.
-14-
.
/
-'~r---r---.----r---.---.----r---.---_r---r---'
t,
1
~-----j'-'--'---'-
•
II I---I--l---+--+---t---I-----l-".
'0
" .1.0.-::'
.
1:1
..
~ 1~1----+_--_+--~~.---,-~.-~~~~--~----+---~.-_-.--.-4---1
I
,0
..... - .....
'/ "~ 125r_--+-~·~"t----r~_,~-+.-~·~--'r_--+_--_+----I_--+_--_J
.0
/ V ~
"C--..-,. .-+-_.Jr-- r--....!2:.!.
".'>., ,/
" - I11
100
/ ,/ S ~ r___
i,/;""- 't!:!! I'---
~,:::::--
~k~ kr-::::::
. --
:
10
~
1 l."."V
V ~ 15111f'~,'L-+----+--~1---+---~--~----+---~---4----4
0 ~O~--~---L--~~--~.--~,OL---~IZ----1---~..L---l---~.O
• 10 12 II zo
PRESSURE. psi )( 162 " PRESSURE. pil X 1d'1
Figure 4. Figure 5.
__r_--_.
, .........
ZOOr---r----.----.----~--_,---_r--
............. ,.
". f-----I------Ih-·.:"c.'--·+----h ",.",--'I------l------t------l
"
II
1
"" ">/
~
I'-...
~ ............
- --
\0--.. V .-/'
• ../' V
"'-- < 11."---
o
o •
PRESSURE • psi X ICS1
10 12 ." I. I. zo " S
CONCENTRATION OF NITROGEN. ppm X 10. 1
• 7
Figure 6. Figure 7.
-15-
It has been found that a condensed phase (liquid or
solid) exhibits a higher vapor pressure when it is exposed to
a high pressure gas. This results in a higher concentration
of the vapor in the surrounding high pressure gas. Dokoupil 2
has made theoretical studies and measurements of this effect
for nitrogen and carbon monoxIDe in hydrogen and has
expressed his results in terms of lIenhancement factor,lI the
ratio of the observed concentration of the nitrogen in the vapor
phase to that computed from the ordinary vafor pressure-
temperature relations for nitrogen. Ziegler has extended
Dokoupil ' s theory to consider the enhancement of the equilibrium
pressure of adsorbed nitrogen in the presence of high pressur e
hydrogen.
-16-
Figure 7 plots the adsorptive capacity of silica gel for
nitrogen in hydrogen in concentrations up to 8000 ppm for four
specific conditions of temperature and pressure. The solid
lines are the capacities computed using the enhancement
theory. Note the order of capacities, i. e., A> B> C>D. The
dashed line s are the capacities computed from the adsorption
isotherm, Figure 3, for a value of p equal to the vapor
o
pressure of nitrogen at 77.4 o K. Note here D>B, i.e.,
greater capacity for greater total pressure. The circled points
are measured capacities of the CEL-NBS purifier resulting
from some twenty tests at the conditions indicated by the
identifying letter. The dotted lines show the trend of the tests.
In all cases the relative positions of the points agree with the
enhancement theory and not with the partial pres sure considera-
tion. The fact that the total amount adsorbed did not come up
to the predicted amount for high capacity silica gel could
possibly be from aging of the gel (used since 1952), not as
much silica gel in the coil as calculated, or possibly full
adsorption equilibrium was not reached.
Discussion
-17-
Question by L. Garwin, Kerr-McGee Oil Ind.: Is the
enhancement factor calculable from the effect of increased
pressure per se on the fugacity of the nitrogen impurity,
since the nitrogen is regarded as adsorbed as a condensed
phase?
-18-
A-3
D. Aronson
Worthington Corporation
Harrison, New Jersey
Introduction
-19-
Cycle Analysis
-20-
1000
900
v
800
q,¥
700 I MULTI-STAGE
COMPRESSORS
II 1t
600 -
0
1/ ,:;
Jf
500
TEMPER ~TURE
il,nKin =oF.4160
tl
II
tIl
~l
300
fI
~ A~ 1/
V rfj' !I
tJ W
Tc
V
;r--r-- EXPANDER
!\ / J
200
Te
/
V ) ~--- 7 8
Figure 2. Low temperature
100 35
/ 4.0 4.5
/~
5.0 5.5 6.0 6.5 7.0
gas cooling flow diagram
ENTROPV-S-Bt.u/(lbOR)
600 60
Torr ~ (:7 Tertii ~. ..~-
/ r_Th 55 0
500 IJ.V Jil-r..
500
rJ.y /1
,
/ ~/-
~ 45 0 c'
I 40
7 ~7
400
T E~PER! TURE / / o rt~~PE ,,,=OF+
35 o nKi
ATUR
~ 7 \ /
anKine =OF... 4 0
/ V / 300
/ \V/
0
I
/ / J
If
--;f&
/ ii/I /
250
/ V/
-
200
-/ i
/ L
20 0 If
I(
T· I TI. J I"-1f
II
150
II 8l-S""
~'
V
~ li'l.uid
\/
liquid 8 a S.
100 / 100
/ V
90 I Ie 90
80
II II 80 f
it
b
703.5 4.0 4.5 5.0 5.5 6.0 6.5
703.5 4.0 4,5 5.0 5.5 6,0 6.5 7.0 7.0
E NT ROPY-S- Btu/(ibOR) ENTROPY-S-Btu/(ib OR)
-21-
lower temperature limit. However, the extra power associated
with the use of the Brayton cycle for the higher temperature
part of the cycle is small compared to the overall plant power
requirements. In such a case the designer may choose to
omit the vapor cycle refrigeration for the sake of equipment
simplicity.
C. O. P. • (la)
-22-
compressors and the heat abstracted from the process,
assuming an ideal refrigeration cycle having no irreversibilities,
equation (1) reduces to
Heat Abstracted from Process
(Ideal) Coefficient of Performance = Power Supplied to Compressors
( 1b)
C.O.P.= Tp (Ic)
To
(2)
T
P
= (3 )
(4)
-23-
Efficiency Terms
(6b)
-24-
Ehe = (8a)
(9a)
(lOa)
-25-
The significance of some of these efficiency terms is
almost self-evident. Compressor efficiency is usually dictated
by choice of type of compressor, the volume of gas flow,
specific speed, and similar considerations which permit only
limited choice. Obviously, the highest efficiency consist ent
with capital costs is sought. By limiting the tempe rature
rise per stage, one can hold Toto the minimum practical
temperature level and so keep down compressor power, as
indicated in equation (1).
-26-
Process Efficiency
154-92 1 120.2
(liquefaction)E pr = (In (154/92») ( 150.0) = 150.0 = 0.802
(12)
"
( coo I lng d own)E =(222 - 135) (In(500/150) ) = 174.6 =0.600
pr In (222/135) 500 - 150 291
( 13)
-27-
ENTROPY,S ,Btu 1(lbOR) for PROCESS GAS
1000 -100 102.030 405060 7.08090100 ILO 12013014015016017.0180 190
800
600 ...'0 -~.
'<11
IiV"'
J'
17
I~
A
200
'1
/
J V"8
Up
b L L
I
I-t-
~
J
B
TEM ~ERAT ~I :E, -<- iTw
°Ran '\ine" ~ 6 p.~ V;
100
V~ ,KV- ~. I
I
I ~(
~RO :ES S c: AS
80 IT ID
I PJ l'l ....
1\ V
60
IV
"
'Ii ...t I I /
.,~ VJ ",-0
40
F06'R~ ~~ I I IF E
~.
I I
20
2.0
II r.
3.0
I I
4.0 5.0
I6.0I 7.0 8.0
ENTROPY, S,Btu/(ibOR) foy REFRIGERANT GAS
b
BATTERY OF COMPRESSORS, Psu.
INTERCOOLERS.
AND AFTERCOOLERS
~---~~~~~~~~~---~ Pdi
To
u. h c a
TWO ONE
J d b
(f)
0::: W
""00:
z::» THREE
<0
Q.,,-
)(
e
"" w v
F COOLER
A
-28-
As shown here, the gas has a constant specific heat over the
range of temperature ~rom 500 0 to 150 0 R, and liquefaction
occur s at constant temperature of 1500R.
= O. 722 \16)
-29-
Multiple Expanders
-30-
heat exchanger s with almost equal mas s flow of the two
streams,
6
D. Aronson, Trans. A.S.M.E., Vol. 72, 675 (1952).
-31-
A-4
Introduction
WR = AH - ToASR (1)
W = W R + :I;ToASIRR (2)
where
reversible work of separation
and/ or liquefaction
W = actual work
-33-
60H = difference in enthalpy between
final and initial states
reversible work
Process efficiency (TJp) =
work for "ideal"
cycle
work for "ideal"
cycle
Practical efficiency (TJA) =--
TJT
= actual work
TJP
-34-
A number of air separation cycles have been analysed
in the above manner and it is proposed in the next section, as
an example, to consider the Heylandt cycle for liquid oxygen
production.
TABLE I
Ideal Actual
Cycle Cycle
-35-
1------
AIR
~lillWLH~+L. o.
---- ......
-36-
In the following table, the results of thermodynamic
analyses of the ideal and actual cycles are tabulated.
TABLE II
Ideal Actual
Cycle Cycle (A - I )
(I ) (A)
Reversible work of
separation and lique-
faction 8.00 8.09 + o. 09
Rectification column
system (including
condenser evaporator
and liquid nitrogen sub-
cooler) 2.65 4.21 +1. 56
-37-
From a study of Table II, it is apparent that,
Ideal Actual A - I
Cyc1e Cycle
(I) (A)
Reversible work of
separation and lique-
faction 15.60 15.77 +0.17
TABLE IV
-40-
It will be observed that the figure for total irreversi-
bility suffered in the ideal case differs slightly from that
given in Tables II and III. This is because in Table IV
separation into the actual products has been as sumed for the
ideal unit, in order to facilitate direct comparison with the
actual rectification system, whereas previously, in comparing
the complete cycle, an ideal separation was assumed, giving
rise to a higher liquid oxygen yield.
-41-
attributable to design factors, e.g., heat leak, pressure
drop, temperature difference across the condenser-evaporator.
The largest single penalty is, however, again due to the
adiabatic nature of the rectification process, the magnitude
of which should be virtually identical with that suffered in
the ideal case, as is the penalty due to greater than
minimum reflux.
-45-
Discussion
-46-
A-5
H. M. Scofield
Linde Company
Division of Union Carbide Corporation
Tonawanda, New York
Introduction
Defining Equations
-48-
(V -L)
Material Balance:
T
Yn VYn = LX n + (V -L) Y~
(n =1,2,3 ... , q)
Equilibrium Condition:
Yn =K
xn n
L,xn V'Yn (n • I, 2, 3 ••• , q)
TABLE I
Liquid Liquid
Condensed Evaporated
-49-
/'~<!O COH.-oSI710N
SHTU",,,ZtSO l/""PM
MD4 .1Ib
£rHY,IfAlt! rltl'fc.
cqUIt.JSI(luM ErN/9AJ'! )(e .SO
/{ct."ZI0IV5HIP5 ?t'fDPyt.EAlc.
.~
1(3 :: 1/ 1$ "'V 0.6814
..90
I'I~: 5,., XI O·7~81 A£Vt!"'fSI8L£ CASE
.ID /TNO SeTs R;J
/03)C.· ~.S694-~
r ./~
-110,---------------------,
-100
",,,,-
.'I -..,~--------=-=-~--=.-:::...-~-
...
:It I
K""" \
------------- -, ~ \
\
/
I
~ -70 \
\
/ ~ ~-------------------
,,
~--------------~'
---.:1lfru-"lfr4D v",,",,'" h4D -.sa
----MIXED ,F4EO
~--------------------~
-~~~~--~--~--~--~-~.S~--.~.~--.~7--~.8 -~0'L--~./L--~.e~--.~~~--.4~--.S~--.~6~--J.7
LlqUII:J ~QW t.,QUID how
(a..) (b)
-50-
complete reversible separation. The ternary column
separates the feed into a binary mixture of ethylene-ethane
at the top and a binary mixture of ethane-propylene at the
bottom. The top and bottom binary columns then complete
the reversible separation of the respective binaries.
-51-
The liquid composition changes resulting from these
sets of calculations are shown as dashed lines on Figure 4
where the individual lines are labeled with the set number.
With constant liquid flow throughout the column, the course
of the composition changes lies well above the reversible
case at every column level. When the liquid is adjusted to
10 per cent greater than the reversible value (Step 2), the
composition changes in the top section are already equal to
those for the reversible case within the accuracy of this
plot. The composition changes in the bottom section for the
three sets show a steady approach to the reversible values.
It may be concluded that it is indeed possible to approach as
closely as desired to reversible conditions by increasing the
number of plates and the number of stepwise adjustments of
heat flow.
Conclusions
-53-
Application of the Defining Equations of
Reversible Multicomponent Separation
F
n = I, 2, 3, ..... , q
x
F
n -t--,-
L,xn V'Y n
I P
= Kn x n (2)
-54-
Realizing that W = V -L and defining L/V =R equations
(1) may be solved for R giving
W
Yn - Yn
R =
W
n = 1, 2, 3, ... q (1 a)
xn - Yn
W W
Yn - Yn Yq - Yq
= n = 1, 2, 3, • . . (q-l) (lb)
W
Xq - Yq
_ yW
= n = 1, 2, 3, . . . (q-l)
n
(3)
q
where L: x = 1
n=l n
KF F _ yW KF x F _ yW
x
n n n q q q
=
_ yW
xF
n n
xF
q - yW
q
n = 1, 2, 3, • • . (q-l) (3a)
yW
H
= 0
q
where 1:
n= 1
Xn =1
Nomenclature
-57-
A-6
1
T. M. Flynn, D. H. Weitzel, K. D. Timmerhaus,
P.C. Vander Arend, and J. W. Draper, Proceedings of the
1956 Cryogenic Engineering Conference, p. 39, Boulder,
Colorado, September, 1956.
2E. S. Sellers and D. R. Augood, Trans. A. I. Ch. E.34,
(London), No.1 (1956).
-58-
of our earlier work. Two parameters especially important
in the design of packed towers are Height Equivalent to a
Theoretical Plate, or HETP, and holdup.
1- l"
PACKED PACKED
43" 38ee
A.B.C.D.E.
SAMPLE TAPS
30 THEORETICAL PLATES
3/4' DIAMETER
100ee LIQUID FEED
THERMO' ~
COUPLE~ -- --REBOILER
~:"'::::=::i.::::t
HEATER
1.2% HZ
PACKED COLUMN 2.7 "I. HT
-
25
1 1 _I T I.
J
--
AVG. COMP. LESS THAN 0.1% atomic T
24 I - - -
V
2
'~
HOLDUP
AVG. COMP.
50% atomic T
22
2 II---
2 0'---
'"
!;( 1.0 H
oJ _
HD ..,..............11....... ....... DT"/~
I " Tz!.!
-J
oJ , Z
!;i .........
"" •••••
............ Dz --+--
is
...o
\
0.8Il---+~\.--Ih/...=-+--+--t--i---t...:.,.........
\ 1/
ZZ
00
Et
0.6~-I--\lri·\--I--+--+--I---+-\T\I--t--H
~: )
~ ; O.4Il----l--...-j+----t--+---If---+--t--j+t.\--+--
(112 i / \.02. •
~ ~~~~-+--+--+--+-_+r_~~-'
~; 0.2Ji~~z~~··~~····~·E"-..~tHT~~~~~:j;:D:!.J-'lv~~~\~;.
~" .~ !tiD \.~~~j
..... !.g.••••
~
(II
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Z
FRACTION OF CHARGE DISTILLED
Figure 3.
-61-
distillation, and correspond to intervals of slowly changing
distillate composition. These plateaus may be considered
as individual products. Thus a temperature sensing element
may be used to indicate when a new product is coming off.
Discussion
-62-
Comment by author: This effect, increase in HETP with
increase in diameter, is fairly well known. HETP is, after
all, an empirical value and should not be expected to hold
for systems radically different from those for which it was
measured.
-63-
B-1
Introduction
TABLE I
-65-
HYDROGEN LlOUEFACTION CYCLES WITH THREE EXPANDERS
PART A
EXPANDER
I
«
EXPANDER
BASIC
Figure 1.
EXPANDER
EXPANDER
EXPANDER
EXPANDER
EXPANOER
EXPANDER
BASIC WITH
CONTI~ERSION
Figure 2.
"f--+--+--+--+---1
Figure 3.
-66-
1) Basic Cycle: The basic three expander cycle is a
moderately efficient cycle which offers the advantage
of low gross work at low pressure (calculated
results shown in Figure 3 are for 30 atmospheres
working pressure) and no external refrigerant
requirement. Possible disadvantages of this cycle
are the complexity involved in the use of three
expansion engines, and its relatively low yield
which results in large flows and equipment. The
functioning of the basic cycle is conventional, with
each expander supplying refrigeration required to
balance the heat exchanger which it brackets.
Final cooling is effected in a Joule-Thomson heat
exchanger, and liquefaction is by Joule-Thomson
expansion. A dual Joule-Thomson valve arrange-
ment is employed to prevent recirculation of
partially converted gas.
EXPANDER
Figure 4.
EXPANDER
BASIC WITH
CONT~ERSION
~
CONTINUOUS CONVERSION
Figure 5.
~ OF LlaUID NITROGEN
I
DER CYCLES
..
YIELD VS !u WORK VS EXPANOER
'4- I ~~
/IilOOlFIEO-
CONTINUOUS
" / II
.~
~ 21
/
--
/MODIFIED
S ~slc
.
.0
c~~
-- ~ 4.5
/ /e&~~~~U s ~
"
'0
--
/ ~
" / / /
• 4~
--:
'0
".
//USIC
~
...
~ CONTINUOUS
MOOIFI
f
/'
"
" . .. '.0 I
EXPANOE ..
70
" 00
EffiCIENCY
os 00 O•
tKPANDER
10
"
EFFICIENCY
.0 O.
Figure 6.
-69-
2) Modified Single Expansion Engine Cycle: This
cycle 1S obtained by the addition of a heat exchanger
between the low temperature nitrogen precooler and
the expander heat exchanger. The modification
allows the expander heat exchanger to "float",
permitting application of auxiliary refrigeration from
the expander at the most efficient point. Functioning
of the modified cycle is the same as that of the basic
cycle.
-70-
Conclusions
Discussion
-72-
B-2
-73-
100
"-.
~
- ..........
.......
~I :
" '"
90
%PARA
80
25'1'. PARA FEED/
~ )-
5L5'Y,PARA FEED/
70
60
o 1000 2000 3000 4000 5000
SPACE VELOCITIES, MIN-I
(STP GAS VOLUMES PER MIN PER UNIT VOL OF CATALYST)
r-- -- .....
-
60
j"-Oo..
% PARA ~<"1r
......
-.
-
50
~/(
r-- r- r- r--.
""- 8S~6"1r
-....
- !--
300
400 800 1200 1600 2000 2400
SPACE VELOCITIES, MIN-I
(STP GAS VOLUMES PER MIN PER UNIT VOL OF CATALYST)
CATALYST
THERMO-
COUPLE LIQUID
N2
-74-
throughout the catalyst bed. This problem is not so difficult
in the case of liquid phase conversion because heat of
conversion merely vaporizes some of the liquid without any
change in temperature. The equilibrium para concentration,
and therefore the driving force for conversion, remain
essentially constant. Heat of conversion can then be
relTIoved downstream of the converter to recondense the vapor
fraction.
i'- '"
.0
r- F::
3/.~ L-
......::.. "'6a.:!,/( 3ot.8
50
-.: ....... .........., r-=: - ~.
-... r- .........
40
30
- r-
-~
r-
85.6'/(
?S.81oo.
r---o
60
~ r---- .........
.....
........ ....... ......... i"o..
..........
r-
-
.....•. ....... ....... r-:.. 39.'~ ........
50 ...... 6a.:!,/(
....... .......
-.:
- --r-r-
"I. PARA ~ """-0 ~ ........ .±
r--- -... r- •••.
7~-:~ t--
-I(
40 ~
!-85,6'/(
30
o 400 eoo 1200 1600 2000 2400
SPACE VELDCITIES, MIN-I
=-
100
--
~~ -'r-::: F-
I--- ~ ~ ~
I~ f-.., ~
""- ......... r-.. '/(
-
:!3.S~
90
..... r-; :::::
-
r-::- ~ <6.3
-
~
loA ~ ~ "-
r-. ........ r-.... r--- -3<.,/ ~ =-~
;!Q:!
---.....
--
BO ~
7.0
"-. ~ "---~
,"PARA " - ~ r---t.
r- -....:;
f-e... r- 7.-
a.,
I"- '--I-
-- ---
~
70
-;- r-.. r- Sto -I-
...... r--. .S
60
-.. 89.S
'I(
-78-
FLOW METER
ANALYSIS IN
He
LIQUID
N2
90
\
SO \
,\
~
-- 26.7°K INLET
\ ....... ,
70
% PARA
60 \. "-
.. -. • -
"~ jl.o"~ IN~ET i
40 .... - -'-1'-'~'_'
T5,S"I K I~LET
I',
~XH~USTI -
60 r--
MIDDtE I
40
.I. J
IjLETj- r--
-80-
Discussion
-81-
Question by R. Katzen, Air Products, Inc.: Has any study
been made of the effect of the configuration of the catalyst
chamber on ortho-para conversion. Would a high L/ D
vessel show any difference from a low L/D vessel?
-82-
49
48
~
z
w
u PARA ~ PRESSURE
a:
w
..
Hydrous Iron Oxide,
0..
7cm. 3 , 20-30 Mesh,
25"0 Pora Feed, 76°K,
a: Constont Flow 7.4 Liters/min (STP)
~
47
46
0 50 100 150 200 250 300 350 400
PRESSURE, PSIG
Figure A.
52
46
~
z
w
u
a:
w 44
0..
..
.;
a:
0..
42
40
38
36
0 5 10 15 20 25
FLOW, LITERS IMIN. (STP)
Figure B.
-83-
Question by L. T. Hendrix, Richfield Oil Corporation: Have
uses for parahydrogen, other than the advantage in storage,
been found?
-84-
B-3
-85-
together with new make-up gas, passes through the gas phase
converter again. The net result of this process is to raise
the para level of the entire system, including the feed gas to
the vapor phase converter. This effect complicated our
estimation of the quantity of catalyst required because we do
not as yet have space velocities for a wide variety of feed
concentrations. We wanted to avoid over design since this
was to be an experimental test as well as a practical
improvement in our liquefier capacity.
531-L LVoAf
AP =
D 2
P
where
-86-
L = depth of packing, ft.
SAMPLE
THERMOCOUPLE WELLS
-88-
Since installing this vapor phase, ortho-para converter,
a number of liquefaction runs have been made with quite
satisfactory results. Analysis of the hydrogen gas at
various stages of the process revealed that the following
average para levels were obtained:
-89-
of analysis were, as mentioned before, 0.25 psi, 0.85 psi,
and 1.89 psi. Since the normal total pressure drop across
the purifier and drier is in the order of 150 psi, this small
additional drop acros s the catalyst chambers is negligible.
Discussion
-90-
Apparently the catalyst has not broken down or dusted during
its use. This conclusion is made from the fact that the
degree of conversion and liquefying rate has not changed
during its use indicating that the quantity of catalyst is the
same as it was initially.
-91-
B-4
T. W. Schaffers
Philips Electronics, Inc.
Mount Vernon, New York
Introduction
-92-
1 PHASE 1. The medium is in the
compression chamber where it
is compressed by the piston.
-93-
Running speed, 1725 rpm
Cooling water, 220 gal/hr
Cooling temperature, 50 0 F
Dew point, 50 0 F
Average helium pressure in working
area, 310 psi
Mechanical Design
Installation
1) No cooling water
-95-
Figure 2. Discription and function of the various components
of the gas liquefier
-96-
Figure 2.
o ,.
Figure 3.
-97-
Applications
Discussion
-98-
B-5
F. G. Brickwedde
North American Philips Company
Mount Vernon, New York
-99-
Air Feed
"-(20·C)
(rio)
Liq. N2
Product·
Rote, ~
Purity, ~
Liq. Woste
02+N2 tA
x.
0.9 0,9
~
'0 0,8 Liquid Air_
0,8
liquid
~ 0,7 Concentration
Ranoe of
0,7
i -----.
Equilibrium
0,5 with Air 0.5
0,4 293·
250·
200·
150·
100·
0,9 0,9
~ LIquid
Concentration 0.8
'0 08
Rano! of
~ 07 Rectifier 0,7
! 0.6 06
~
'" 05 0,5
04 293·
Proportional
250"
to Insulation
loss
200-
I!!IO·
100-
300 396 t
T·K
Role of Air Fud to Generator (Qrom moles/hr.)
-101-
rate of air feed, i. e., through the heat exchanger-purifier
losses. For generators with ideal heat exchangers, and
generators with exchangers having constant refrigeration
losses (independent of rate of air feed) the liquid product
delivery rate of the generator is independent of the rate of
air feed. As the rate of air feed to the generator is varied,
it is the amount of waste gases that varies--not the liquid
product capacity of the generator which is determined by the
refrigerating capacity of the Norelco liquefier.
-102-
The 6. T los s at the warm end of the exchanger (6. T ew) is
proportional to the rate of discharge of waste gases. The
maximum driving force for transfer of heat in the exchange
(6. T at the cold end) decreases as the rate of air feed is
increased. Table I shows this dependence of the losses on
rate of air feed, and also the relative importance of the
different contributing losses. The calculated data in Table I
are for an air feed at 20 0 C with a 100 per cent relative
humidity, for a 2 inch thickness of hairfelt insulation around
the heat exchanger which is built around the rectifier column,
and for a 6. T of 100C at the warm end of the heat exchanger.
The losses are compared for flow rates of 400 and 300 mols
of air per hour. The higher rate corresponds to the
maximum that the rectifier column can accommodate, using a
Norelco refrigerator that can liquify 5.7 liters of liquid air
per hour under favorable laboratory conditions. The lower
rate of air feed (300 mols/hr) is still in the range in which
nearly pure liquid nitrogen is obtainable from the generator.
Losses are expressed as fractions of the refrigerating capacity
capacity Q c of the Norelco liquefier used as the condenser.
TABLE I
Rate of Refrigeration Loss
-103-
Question by G. C. Szego, General Electric Company:
Would it not prove advantageous for a purer bottoms product
to insert the feed above the bottom and provide the necessary
boil up by an auxiliary heat exchange loop with the incoming
air feed?
-104-
approximately 192/(rate of air feed, mols/hr) for Figure 2
and l15/(rate of air feed, mols/hr) for Figure 3. Figures 2
and 3 may be consulted for the further evaluation of these
fractions.
-105-
C-l
S. C. Collins
Massachusetts Institute of Technology
Cambridge, Mas sachusetts
-106-
......
.-.-~~
Figur e 1.
. /"
~\-~/ / D~
/ \ /.-'----
~>--.
'- ..
11M ..
v..1II~"Id.~. ,... ~t_~
*"M'VN ,..,. ,.-J
r•. .....,. ..... I ·~ ...,.
""",...,..,CIAII,. ........,'/ .., .... ~T <-*,'... ~I 4""ltl-
1\"'-'" ...... M - ,.0.
.. .
Figure 2.
-107-
half is returned counter cur rent to the feed in the column,
and in the exchanger. Since the waste stream ente rs the
heat exchanger in the liquid state, the refrigeration available
from it will produce almost saturated feed air at the inlet
to the distillation column. Such a design insures adequate
purification of the inlet air during normal operation.
Heat Exchange
Column
-108-
discussed by A. W. Fisher and R. J. Bowen, 1 was
developed by H. O. McMahon for air borne oxygen plant?
and other applications where little space, light weight and
high efficiency are considerations. HETP's obtainable with
McMahon packing are on the order of 2 to 3 inches, as shown
on Figure 4.
Start-up
-109-
-_ ..
.-.:I~.'"
,_,,...,oH.(",.,.
...-, 4"
, .....
.l-
- .................
(N , .
".y
_ t ... ~ . _ _ 1
~ . ~ , ---..-
.J~""""",,"...-Lo,t
'-1 '" - I I _1-
l.'-_ _ _ _ _ __ _ __ _
• ~ • j • , I ~ • • ., ...
1,1,\\ . '~.'-
-\\\,~ ,\, ~\\, ~ .\,'
Figure 3.
1-
!
.,.
...
.. l
- f ..-
... -..!.~
-
__ ______
- -
o~---~---~- ~
o
MoP"l6ItTU ~
... ""'"
~f!t,c"....,.
Figure 4 .
.... "'.....
ItwR\ ...... f.J, ......
~-~- ~ ~
Control
-111-
Further Development
Commercial Unit
-112-
Acknowledgement
Discussion
-113-
C-2
R. E. Brown*
Bureau of Aeronautics
Department of the Navy
Washington, D. C.
Introduction
-llS-
General Economic Principles
-116-
previous illustration apply. By the same processes, in the
case of compressed cylinder helium and its liquid counterpart,
with liquid helium the cost of transportation plus that of
liquefaction would be twice the liquefaction cost, in compari-
son with five times the liquefaction cost if shipped com-
pressed. Three-fifths of the former shipping cost is saved.
On the other hand, in the case of compressed tank car
helium and its liquid counterpart, with liquid helium the cost
of transportation plus that of liquefaction would be 1. 2 times
the cost of liquefaction, in comparison with equal to the cost
of liquefaction if shipped in compressed form. In this case
liquefaction might be worth while for cylinder helium, but
would not be as economical as compressed helium for tank
car helium.
-117-
Strategic Advantages
-118-
During a war emergency coastal stations are called
upon for large helium shipments which might justify lique-
faction for economy in transportation if rapid shipment by
air is wanted. In times of peace, however, there is little
such bulk shipment and most shipments are in small lots
to many completely separate and uncoordinated end use
destinations. Liquefiers at these stations might be moderately
successful in attaining desired objectives during a war
emergency. Under normal conditions, though, it would be
virtually impossible to regiment enough of the many end use
activities to want liquid helium to obtain the large liquefaction
rate necessary to make liquefaction economically worth while.
For these reasons liquefaction at coastal Naval Air Stations
is not being currently recommended.
-119-
Tank Car Transportation
-120-
a unit, with an appropriate number of tank cars, is
recommended as the first venture. Additional units can be
added as found desirable.
Semi-Trailer Transportation
-121-
cars, or eighty per cent of the planned liquid helium tank
car, in transportation capacity. It is believed that the
charges for hauling a semi-trailer will be about equal to
those for hauling a tank car. In that case the unit cost for
hauling liquid helium by semi-trailer would be one-half that
for compressed helium tank car transportation and two and a
half times that for liquid helium tank car transportation.
Adding the liquefaction cost in each case, including amortiza-
tion and maintenance charges of the liquefaction plant,
results, for liquid helium semi-trailers, in a total liquefaction
and transportation cost which seems to be somewhat higher
than that for liquid helium in tank cars and somewhat lower
than the transportation cost for compressed helium in tank
car s for the long trip, Exell to Lakehur st.
-122-
Conclusion
Discussion
Answer by author: The first 1,200, 000 cu. ft. liquid helium
tank car will probably cost $180, 000. Later cars in lots of
three or more each will probably cost $165, 000 each. Presert
compressed helium cars, holding only one-fifth as much
(240, 000 cu. ft.) cost over $100, 000 each.
A general design for a 1,200, 000 cu. ft. liquid helium tank
car has been developed to the stage necessary for bidding
purposes and for calculating performance.
-123-
is 10 atmospheres, and this is sufficient for no-loss operation
within the United States.
-124-
C-3
-125-
Cycles with expansion engines in which the gas
performs work are in principle more efficient than cycles
in which the gas does not perform external work. A
cascade cycle with liquid nitrogen (or liquid air) and liquid
hydrogen also is somewhat less efficient, because of the
los s e s in handling of both liquid pr ecoolants.
-126-
COMPRESSOR SYSTEM
, . . - - - - VENT
EXPANDER NO. I
'2 = 15 otm
I+d
O.03L--~'----7---6O------;:-----:';:IO'--""'1!;:2-
X
DESIGN CRITERIA
I) THERMODYNAMIC EFFICIENCY
2) COMPONENT AVAILABILITY
3) COMPONENT RELIABILITY
4) SIMPLICITY
5) DOWN -TIME
-127-
Liquid nitrogen or liquid air precooling of the high
pressure helium gas provides refrigeration at a temperature
of the liquid ' s boiling point at one atmosphere and makes it
possible to reduce the quantity of helium circulated per unit
volume of liquid produced. The use of this precoolant liquid
is somewhat in conflict with the previous statement that
elements of a cascade cycle reduce the efficiency of the
process. However. liquid nitrogen or liquid air can be
handled with relatively small losses and can also be pro-
duced fairly efficiently. Rather large quantities are required
to insulate the liquid helium during shipment in the transport
container; its use in the helium liquefier. therefore. does
not mean that the liquid precoolant is to be produced for the
helium liquefaction process only.
-128-
contemplated in this study. The cycle is extremely simple
and should require little maintenance and downtime. It
should be pointed out here that the cycle can be made more
efficient by the addition of expanders but it was felt that
the added complexity is not warranted to achieve a decrease
in the power cost since this cost was only a small percent-
age of the total. Direct labor costs and overhead costs are a
large item in the overall cost picture and were estimated for
the area in which the plant was to be erected. Amortization
of the capital investment, which includes depreciation and
interest, was also charged to the process and included in the
cost of liquefaction. Initial maintenance costs were estimated
to be somewhat higher than in a well established production
process. The cycle described is one that, when evaluated
using these design criteria, gave a low production cost and
still required a relatively low initial capital investment. The
total estimated cost to condense helium showed that the
project is economically feasible for the application West
Texas to Lakehurst.
SHIELD
10
'"'"
II:
'"
X .
...
:l;
o
'"
I-
'!;
'"
II:
HEAT INPUT:: 10 WATTS
~ 4
'"g:
5 ~ 15 W ~
-130-
The actual large scale transport of liquid helium
will be accomplished in tank cars. The thermodynamic
properties of helium near the normal boiling point make it
possible to seal off the transport container during the entire
time of shipment. The pressure will rise in the container
because of the heat transfer from the atmosphere, but the
final pressure can be kept within reasonable limits using
existing insulation techniques. Figure 5 shows a plot of
pressure rise versus transport time for a 16,320 liter
container filled to various levels. It is estimated that a
one way trip from West Texas to Lakehurst, New Jersey,
will require a minimum of 12 days. At the end of this
period the pressure in a 90 per cent full conto.iner will be
about 6 atmospheres. If the container is designed for a
pressure of 10 atmospheres, there would be a factor of
safety of almost two to take into account delays en route.
-131-
Discussion
-132-
Answer by author: Several materials are suitable.
Examples are stainless steel types 304L, 316, 321 and 347.
Aluminum or copper are also pos sibilities.
-133-
Question by M. A. Dubs, Linde Company: What is the
relative volume of liquid nitrogen carried on the 12 day trip
and 90 per cent full container?
-134-
Question by L. Lazare, AInerican Cyanamid Company:
Liquid helium transported will probably arrive at its restina-
tion in a supercritical dense phase. Are pumps contem-
plated, able to handle this dense phase material as well as
liquid helium?
S. Greenfield
Rocketdyne
A Division of North American Aviation
Canoga Park, California
Introduction
-136-
the evaluation of rocket engine components. In Figure I,
a photograph of a typical small thrust rocket research test
stand is shown during a firing of a thrust chamber. This
test stand uses pressurized feed systems shown schematically
in Figure 2. In contrast to this research test stand, which
is designed to accommodate thrusts of less than 10,000
pounds, Figure 3 shows a test firing on a typical large
rocket engine component test stand. The thrust chamber
shown here develops well over 75,000 pounds of thrust.
Again, this system for large engine component testing
ut ilizes pressurized feeding of both the oxidizer and the
fuel. To a lesser degree liquid nitrogen dilution can occur
during the tests of complete rocket engine systems. Here
the propellants are fed to the thrust chamber by the use of
a set of turbine driven pumps, one for each of the oxidizer
and fuel. A slight pressurization of the tanked propellants
in the missile airframe is required in order to provide
sufficient suction head to the pumps. But in this case, the
pressures are low enough so that a problem with nitrogen
dilution does not usually exist.
-137-
--~I-r(~---.L...---{><h-!>:r-- ..Iff
UOUID fut:L
OXYG(N
"'Hie
'AN.
IL". --!>:r--I
COOI..oVIT OUT
-r--------,---------.------~
-138-
translated into the effect on payload and range of a missile,
we find that such small losses in specific impulse are most
important. As an example, it can be calculated that a one
per cent decrease in specific impulse of a typical ballistic
missile can result in a nine per cent decrease in payload,
or a six per cent decrease in range.
-139-
Another scheme, which has been suggested and used,
involves the use of helium for the initial pressurization of
the liquid oxygen to the desired tank pressure level and then
switching to gaseous nitrogen for pressure feeding during
the test run. This procedure reduces the extent of the
condensation and dilution. However, it involves a complica-
tion of the pressure feed system to a point where the gains
are not economical.
-140-
by the heat capacity and conductivity of the metal walls.
On emptying the tank, such effects as sloshing or vortexing
of the fluids will naturally enhance the condensation and
dilution.
-141-
PARAMETERS AFFECTING NITROGEN DILUTION OF OXYGEN
RATE OF PRESSURIZATION
ULLAGE VOLUME
VIBRATION
SLOSHING
VORTEXING
. .
I.ClIiI " .... --O'<r---...L..--l!~i>'<J-..,.--:....:;=----------rrn--
~ ~-""""'"
10
\ 1IIIOCllffl.e{.1
I
-142-
The gas chromatography analysis involves passing the gas
through a column filled with an adsorbing material consisting
of small, closely packed solid particles. As the sample
passes through the solid packing, each component is adsorbed
to the packing preferentially. The column is then desorbed
of the sample, and, since the oxygen and nitrogen were
adsorbed at different rates, they will be desorbed in the
reverse order. As the gases leave the column, the thermal
conductivity is measured and recorded. In this manner, the
amount of each component present in the sample can be
determined to an accuracy and precision of ±. 1 per cent.
-143-
Figure 9. Photo of Poroloy diffuser
element
OJ .. os
~{. .i / / 1/ /
.~~.;> I V V
I---- f- ,. "~4. V
I, /
V
7
/ I)
r-~< I..'~ v e' I 1/ v . J
V
..
1
~~ ~/;' ~"
t="~
' #.
.
I
~
- ~~ ~ I. V
/
..oj " V II
.!'s:~. I. / /
~~
V
/
/
7 l
I
,{!<' '/1 V V V1I •
~ // )
VI
· ~. l lillL
L*rr-
/ 1 1
V L
/ V L L Ii
I, / J
V
I V /1
AIR FLOW (Gl- 'M,I .. Mot.
-144-
photograph of a typical diffuser manufactured from the
Poroloy element is shown. Figure 10 gives the relation
between strength, flow, relative density, and pressure for
Poroloy with 35 0 crossover angle. This unit is cylindrical
in shape and is about eight inches long and one and one-
fourth inches in diameter. The mesh of the metallic
portion provides for 50 per cent open area with an interstitial
opening of 10- 4 to 10- 5 square inches. At about the time
that the Poroloy diffuser elements were obtained, the nitrogen
dilution program was interrupted by more pressing problems
and, unfortunately, the results of dilution using such diffusers
cannot be reported at this time. However, since the installa-
tion of these elements have been made in the Rocketdyne
Research Test Area, we have had no incidents of nitrogen
dilution of liquid oxygen under the conditions of test with
which we are now involved. Pressurization times much less
than those reported in Figure 8 for the standard diffuser can
be obtained with the Poroloy diffuser, with the end tank pres-
sure level reaching as high as 1500 psi without qualitative
evidence of nitrogen dilution.
Concluding Remarks
-145-
surface area of the liquid oxygen is large compared to the
volume of the tank. On the large rocket component test
stands at Rocketdyne, where liquid oxygen capacities will
run to several thousand gallons, we find that the problem of
dilution with nitrogen is not quite as prevalent. Nevertheless,
incidents have been reported which bear all of the earmarks
of nitrogen dilution characterized by an unusual increase in
liquid oxygen temperature and a corresponding decrease in
the combustion device performance. Measurements have been
made of nitrogen dilution on these large test stands and
values of less than five per cent have been recorded. The
results of our research program on mechanical diffusers
have been applied to both the large test stands and our
smaller research test stands, and the incident of nitrogen
dilution has been measurably reduced.
Discussion
-147-
Question by J. B. Gardner, British Oxygen Research and
Development, Ltd.: Did you study the effect of the tempera-
ture of the pressurizing nitrogen?
W. J. Scharle
Air Products, Inc.
Allentown, Pennsylvania
-149-
on-site liquid oxygen generators. The U. S. Air Force,
therefore, decided to install its own liquid oxygen producing
facilities at points of contemplated large consumption.
2) Minimum of utilities.
4) Transportability.
-150-
I
I
I
I
I
,I ,
~~
CA"ION DIOXIDE
SCRUIBER
IAI~
1\ 1\
3000
V
~ 1\ / ~ ~v 2000
\ 1\ 16 \ 1500
~ 1'- il 1\ V IPsE~ V
1000
900
~ X 800
t-
~~g: t-
TOO
II &00
.--<>---- ~o
'00
.,/ '-II II 1I\k' ~+'vI
CASE 1- NO REHEAT ,,,,v IX rY ~ ~...'"t'./ ~
400
300
..
~ ~~ ~~
RECIPROCATING EXPANSION
ENGINE 6H' 1480 ']1 200
iii
Q.
V\ II IYV '(/11 W
«
::>
..
150
ct?,,~ 'J /
CI)
'"«
RECIPROCATING EXPANSION
ENGINE .. AH' 1260
..II!
"
~
100
90
80
Q.
TO
CENTRIFUGAL ~ 10
--
I-
EXPANDER . . AH· 420 II '0
1680
..
~ r-t 40
/~
~ ~
~
b
~I
~
0
! ~ co ..
I t- I ~ r-" 30
,
INCREASED REFRIGERATION I
I I
-151-
compression, the air is passed through a caustic scrubbing
system to remove carbon dioxide from the air stream.
-152-
pressure to high pressure column pressure. As seen from
Figure 2, an increase in total expander refrigeration of about
14 per cent can be realized by this scheme. Also, very
significant, is the fact that the reciprocating expansion
engine is considerably smaller in physical size because of
the higher density of the exhaust air from the engine.
Minimum of Utilities
Transportability
Equipment
-154-
Figure 3. Compressor alr to air Figure 4. 75 T /D Lox generator,
inter- and aftercooler multiple unit facility
-155-
expansion engine is directly coupled to the compressor motor
so that power is returned directly to the system.
-156-
It is estimated, for the subject facilities, that
storage tank evaporation losses are about 1 per cent per
day. Although it may be possible to reduce the storage
tank evaporation losses to about 1/2 per cent per day by the
use of vacuum type storage containers, the use of on-site
liquid oxygen generators and the low production cost of
liquid oxygen make the payout period for the increased
storage equipment cost too long. It is also of importance
that, with the use of on-site generators, the evaporation
losses normally associated with liquid oxygen semi- trailer
travel time are eliminated.
Conclusion
Discussion
-157-
Answer by author: Activated alumina was used.
-158-
Answer by author: Yes, this figure includes all auxiliary
power requireInents.
-159-
C-6
W. W. Aton
Linde Company
Division of Union Carbide Corporation
Tonawanda, New York
Introduction
-160-
To
D , llr"'- ;6U/";DW
OT"ut U..4 P""AJTS
/
,/ '"
/
/
'"
,PD~/T;¥t:
O'.:I~4"'~4M~T
8~t"AJt:; /1QU(,
-161-
absolute continuity of supply is not critical, no liquid oxygen
unit is provided. Oxygen is billed to the customer by a
positive displacement meter having ~ 2 per cent accuracy
over a 50 to I flow range.
Safety
-163-
Figure 3. Simple process diagram, air
separation unit, 80 million cu. ft. /month 02
-164-
condenser- reboiler. Operating experience has shown that
this system, properly designed, provides 100 per cent safe
conditions.
Instrumentation
-165-
control room. Succes sful unattended operation requires
fir st rate instrumentation engineering and a generous develop-
ment program preferably on a pilot plant.
-167-
•
.
't.Ol L CiIol' _ _ _
• •
1 - - - - - - - - - 1 1 0 V . I I . C . - - - - - - - - -- -t
/lUXI(. 11'/1'. Y
COAJT"I(:T 0.<.1 7iHi!E COAJT"'CTo,i(
OeLl''''' )feL"'Y (/.5eo )'O~
WHIC# Ol"'e.<.l$ 3 ShuTT/Alt; Dow"';
S~CCNOS /9'TI!H CG/UIPM,E,..,,:
S~uToOWAl
SIt/AlAioion.)<i
MCAII TC/'t/A./t:;
54Avlce, eTC.
/NOICnT,Alq~~--~---~----------i
C/~C:LI/T
B,I(el9xtV?,
-168-
but power is disrupted after a short time delay to prevent
other breakers from actuating due to secondary factors.
When the safety control relay drops out it first opens the
compres sor motor circuit to effect the shutdown and cau&es
all electrical and mechanical controls to be deener gized and
subsequently adjusted automatically to the safe position. It
then opens the circuit to the comrnercial monitoring service
causing a signal to be transmitted, lights a pilot light, and
finally opens a contact in its own coil circuit to prevent
possible reenergization due to the monitoring device closing
again.
Field Operation
-169-
oxygen supply is always maintained. The plants were
designed with an expected outage of less than 4 per cent
of which one-half of this is scheduled. The average outage
time for all plants is currently below 4 per cent and it is
expected that with further operation the outage will be
further reduced. Complete safety has been demonstrated and
the plants have been accepted with great enthusiasm by the
operating personnel and the customer.
Discussion
-171-
D-I
-172-
D-2
LIQUID HELIUM TEMPERATURES IN
AN ATOMIC REACTOR
-173-
Cooling for these experiments was supplied by a
refrigerator using helium gas as the refrigerant and employ-
ing a regenerative heat exchanger and two expansion engines
to remove energy from the gas and convert it to work in a
hydraulic mechanism. The refrigerator was designed for
another purpose by Arthur D. Little, Inc. and was modified
for our use by the designers. A detailed description of the
operating rrinciples of the refrigerator was given by A.
Pastuhov. The refrigerator has operated over 8000 hours
with very little maintenance.
3
A. Pastuhov. Proceedings of the 1954 Cryogenic
Engineering Conference. p. 21. Boulder. Colorado. September
1954.
4 R • R. Coltman and T. H. Blewitt. Proceedings of
the 1956 Cryogenic Engineering Conference. p. 110. Boulder.
Colorado. September 1956.
-174-
Figure 1. Figure 2.
-·-·-or.c.o "'~oo(L
-175-
In the liquefier section the high pressure gas enters
a regenerative exchanger at room temperature where it
gives up its heat to the gas returning to the compressors;
this exchanger is a main heat exchanger used in the Collins
helium cryostat except that the standard tap-off points were
omitted. The gas then passes through a short length of
exchanger in which the temperature is reduced to between
11 and 15 0 K by gas from the low pressure side of the
main refrigerator.
From the liquefier cold case the gas pas ses through
another exchanger where it will be cooled further by the gas
returning from the expansion valve. This exchanger will be
made up of small, thin stainless steel tubes, one within the
other. Because of the small flow of gas it is necessary
to provide a low longitudinal conductivity path for the
exchanger to operate satisfactorily. This stainless steel
exchanger is located within the cryostat in the shield section
of the reactor. This shield is of concrete and is seven feet
thick. The exchanger assembly will consist of four parallel
exchangers about 5 feet long. The gas will then enter a
J -T expansion valve for further cooling after which it will be
routed to a small gas chamber surrounding the sample
chamber. The sample chamber will be about 1/2 inch in
diameter and the cooled portion will be about 6 inches long.
The gas chamber will be about 12 feet below the J -T valve
and will be in the center of the reactor lattice.
-176-
Since the gamma heat input to a radiation shield will
exceed the heat input from thermal radiation, no thermal
radiation shields are incorporated in the reactor portion of
the circuit.
-177-
2 -1/4 inches thus reducing the effective LID ratio and at
the same time keeping longitudinal conduction to a minimum.
-178-
D-3
H. J. Smith
The Aro Equipment Corporation
Bryan, Ohio
Introduction
-179-
The complete gaging system consists of the indicating
device, intermediate device, an integral capacitance probe,
and interconnecting cables. The usual configuration has
the indicating and intermediate devices combined in one
package.
Integral Probe
-IS0-
r------l r--- -------- - - - - - - - - - - - - - - ,
I CONTAINER, I I ..'~OICATOR. LlQUEFIEO GAS QUANTITY I
I LIOUEFIEO GAS I I I
I I I I
I I I I
I I I I
I I I
I
I : I
I I I
I _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .JI
L
LEAOY,IR[ CONNECTION
~OVAIILE
/
/
-181-
and the mounting brackets. Figures 2 and 3 show side and
top views respectively of the integral probe.
(1)
or (2)
-182-
C = C
e
+ 5.796 V0 ( 3)
m
TABLE I
5 63.5
10 123.5
25 303.5
TABLE II
Volume (Liters) Measured Capacity Liquid Height
mmf (inches)
0 63.5 0
1 69.30 2.35
2 75.09 3.55
3 80.89 4.65
4 86.68 5.75
5 92.48 7.30
-183-
The capacitance of the probe is dependent upon the
dielectric constant and height of the liquid present between
its electrodes. Therefore, to give a linear change of
capacitance with respect to liquid volume the capacitance
of the probe must be characterized to the configuration of
the container. Figure 4 shows the military requirements of
measured probe capacitance with respect to liquid volume.
Figure 5 shows the relationship of height and liquid volume
for a spherical 5 liter container.
-184-
90.0
80.0
u
z
V
~ 70.0
LIQUIO -VOLUME-LITERS
HEIGHT {INCHES)
.. .8 1.0 1_2
BAFFLE DISTANCE - INCHes
I.' 1.5 1.8
-185-
The need for close tolerance of the probe placement
is further accented when we consider the reduction of probe
capacitance due to the electric field interference that occurs
whenever a ground potential surface approaches the extreme
ends of the probe. The capacitance reduction for a probe
used in a spherical 5 liter container is shown as a function
of the proximity of the fill or vent fittings in Figure 6. The
same relationship is shown for a "flat" 5 liter container in
Figure 7. A comparison of the two reduction curves show
that the capacitance reduction is a function of fitting or shell
wall proximity and will vary with the container configuration.
This requires close coordination of the probe manufacturer
and the container designer with regards to the expected
end clearance of the probe. This then may be interpreted
in terms of capacitance reduction and proper compensation
made in the manufacture of the probe.
-186-
$
~.5"
~
<,~
~A
~
~
~'
~
~Z
~
".~
~ J
.2 .if B 1.0
BAFFLE DlSTANC.E -INCHES
I.il I.' 1$ I.e
~
"-
0 .•
~
~
! 0.'
u
z
~
Of
~
.
!
0.'
~
0.1 U
DISSIPATION FACTOR
INTEGRAL PROBE
TO CAPACITANC£ GAGING
REMOVABLE PROBE
-188-
probe should have the same service life as the container.
Discussion
-190-
D-4
r--- ---l
B+ I
I
CELL
I B-
PUSH TO l---
CAL. I
A B
TO CELL
TO VENT
PURGE
Figure 1. Conventional
thermal conductivity
gas analyzer
REFERENCE A
, '
, \ .. ~J)
REFERENCE - - - ~ , NORMALIZER
SOURCE l<~.!.:
NORMALIZER
50.
HEll POT
3.3k Ik
L..JW'-~MJ""""--+
TO RECORDER
L-_ _ _..;.IO;..;.O II.V.
Figure 3. Phase-sensitive
rectifier
-192-
If the cell is excited with alternating current
several advantages become apparent for very low level use.
Small a-c signals can be measured in the presence of la-rger
parasitic d-c potentials. The amplifier is very simple with
an a-c system since a chopper is not required and standard
components can be used as in any conventional, low level,
a-c amplifier. Peaking the amplifier at the bridge excitation
frequency improves the signal to noise ratio, improving the
low level characteristics even further. The output of the
amplifier passes through a phase-sensitive rectifier to
operate a d-c recorder or output meter. The large magni-
tude of the d-c output allows the use of an inexpensive
recorder. A wiring diagram showing cell connections is
shown in Figure 2 and the phase-sensitive rectifier is shown
in Figure 3.
-193-
The d-c output is obtained from the phase -sensitive
rectifier, which gives polarity reversal when the bridge
passes through balance. The resulting signal thus obtained
is the same as though the cell were excited with d-c.
Without this phase-sensitive rectifier it is impossible to tell
whether it is the sample or the reference which has the
higher thermal conductivity as all deflections will be in the
same direction. A conventional phase-sensitive rectifier
using triodes will show some drift. To minimize this a
rectifier circuit is used which requires only diodes for
operation. This type of phase-sensitive rectifier has very
small zero drift and is linear.
Discussion
-194-
Question by G. H. Caine, Pesco Products: Would there be
any advantages in using a frequency other than 60 cycles per
second, say 1,000 or 2,000 cycles per second?
-195-
D-5
R. H. Ball
Sundstrand-Denver
Denve r, Colorado
-197-
Figure 1. Assembly drawing of Sundstrand 3 cubic inch
pump
-198-
Figure 2.
Figure 1.
--.... ,....-----
Figure 4 .
Figure 3.
- ~ ...L.
· ~. ~ .
·
· ~ ..
:s ~. .
~
~
· ....
I
· ....... ,..,
· 1"-
· .
· ~ un
· .
Figure 6.
Figure 5.
-199-
I
,",
,,' I
I
Figure 7. Figure 8.
",
-· -
,I··
)I....
.;: tt--..
· ~") ~ ...
~
.
~.
\
II ··
~.
.]>•."
.. \
· ......
.........."!;. ~
· .r - ....
· ......
· ,
Figure 11.
Figure 12.
-200-
Figure 14.
Figure 13.
JS .
-
· """ .. ~
· -ro- r-'
/ ~~
·
.
0
.-~
1--. -- :..- -..,
IP:~ e'_
°F_
-~ ;;= 1--
I-
0
·
·
Figure 17. Figure 18.
-201-
and valve surfaces were all in good condition. The pump
shaft roller bearing nearest the mounting pad showed signs
of warpage and was lightly discolored. This was caused by
its location in a dead space with no circulation of cold gas
or liquid.
-202-
bearings, hard steel pistons, iron and aluminum castings,
and shafts appeared operable. The laminated plastic was
found to be a satisfactory rubbing surface material for the
valve faces, and leather shaft seals were satisfactory.
Some corrosion was encountered due to intermittent pump
exposure to air and moisture between test runs. It should
be stated that a series of occasional difficulties prevented
any of these pumps from pumping liquid nitrogen with
reliability.
-203-
continuous chart. The recorder requires a digital to voltage
conve rter for rpm, strain gage type pressure pick-ups, an
electrical recording head on the flow meter, and copper-
constantan thermocouples. Two types of pump drives are
used, one a hydraulic motor regulated by a flow control
valve, and the other a variable speed belt drive unit.
-204-
width. Considerable trouble was encountered in getting a
good adhesive bond between the .032 inch plastic and the
metal cylinder block. Several thermo setting and epoxy
adhesives were tried until one was found that was adequate.
In conjunction with this problem, it has been necessary to
initiate a basic study of possible valve face materials and
bonding methods. A study contract covering this phase has
been assigned to the Denver Research Institute. This
program is still in progress and involves literature search,
investigation of materials and methods, and strength testing
of samples.
-205-
failed after only 4 hours of operation. Our overall exper-
ience using several types, sizes, and loading conditions on
anti-friction bearings would indicate this to be one of the
lesser probleIn areas except for corrosion resistance.
-206-
Cryogenic Laboratories and the NACA has been of great
value. In the design of mechanical devices involving stress,
friction and wear, heat transfer, corrosion, instrumentation,
etc., many unanswered questions have been raised. In
several cases finding the answer by test has been difficult.
We would like to suggest and encourage an even more
extensive pursuit of basic cryogenic engineering data and the
further dissemination of information such as this conference
provides.
Discussion
-207-
Question by J. E. Boretz, Stratos -Division Fairchild Engine
and Airplane Corporation: What was the suction specific
speed for this pump and what was the minimum net positive
suction head required?
The seals we used were the spring loaded lip type, genuine
leather sealing element type. They are identical to the more
commonly used composition sealing element units generally
used in automobile front wheel bearing seals. These are
available at hardware or automotive supply houses and are
standardized sizes. At least three companies manufacture these.
-208-
D-6
-209-
Bearing Data
-210-
DRIVE
PULLEY
SHAFT -;---t-I-...-
PUMP
DISCHARGE
VALVE
ORIFICE
HEAT
EXCHANGER
BYPASS
VALVE
CATALYST
CHAMBER
TEST B~~RIIIIGS·1'-..
-211-
retainer. In fact, a set of these bearings is still being
used in the pumping apparatus. This set of bearings was
installed February 18, 1957, and the lower bearing was
changed on March 12, 1957. The upper bearing has operated
about 60 hours while the lower bearing has operated a little
over 52 hours. The bearings still operate very smoothly.
-212-
in liquid nitrogen, hydrogen, and oxygen could be investi-
gated. The type 27 instruments have been very satisfactory
and have the same full scale accuracy as the type 9 instru-
ment.
-213-
In the preceding discussion it was assumed that the
liquid being pumped and the liquid in the vapor bulb have
the same composition. This is not necessarily true; diffi-
culty was experienced in this respect with both liquid
hydrogen and liquid nitrogen. With liquid hydrogen a number
of runs indicated a negative NPSH. Therefore, the vapor
pressure in the vapor bulb was greater than the pressure of
the liquid head above its pump suction. The liquid level
tube and the static NPSH tube were interchanged to check
proper functioning of the instrumentation lines; the same
negative NPSH reading was observed. This negative reading
was believed due to a difference in ortho-para composition
of the liquid hydrogen in the vapor bulb and pump dewar.
Catalyst chambers were therefore installed in the vapor bulb
and the pump dewar to assure that liquid of the same compo-
sition was in both places. A small negative NPSH value was
still observed indicating that complete conversion did not
occur.
4
D. H. Weitzel and L. E. White, Rev. Sci. Instr.
26, 290 (1955).
-214-
Operation of the Test Facility
Discussion
-215-
Answer by author: The loading on these bearings is not
exactly known but about 88 per cent of the total running
time on the bearings with non-metallic retainer was done
in liquid nitrogen where the motor was pulling between 500
and 600 watts. The severe conditions described, however,
are those described in the conference paper where moisture
is condensing on the bearings causing some corrosion. In
addition, any particles of moisture which are not evaporated
freeze to the bearing surfaces and cause the bearing to be
very rough.
-216.
D-7
D. W. Seavey
Cambridge Corporation
Lowell, Massachusetts
-217-
D-8
C. D. Arrick
Linde Company
Division of Union Carbide Corporation
Tonawanda, New York
Introduction
-218-
classification of them has been developed with an eye toward
the potential hazard from the various categories. They have
been divided into three classes called the "volatiles," the
"non-volatiles, \1 and the "acetylenic" hydrocarbons.
TABLE I
Classification of Hydrocarbons
Potential Hazard
Class Materials Included in Typical
Liquid Oxygen
-219-
Table 1 shows the simple classification of the hydro-
carbons with respect to their hazard. These categories were
established by test. The analysis for volatiles consists of
the determination of all hydrocarbons having appreciab Ie vapor
pressures (above a few microns) at about _180 o C. The non-
volatiles are those hydrocarbons with low vapor pressures at
-180 o C. The acetylenic hydrocarbons are specifically deter-
mined separately.
TABLE II
Hydrocarbons in Atmospheric Air at Plant Sites
-220-
is made to keep extraneous combustible contaminants out of
the process air by strategic location of the plant air intake
and by regulations preventing contamination of the process
air with combustibles such as paint or solvent fumes, etc.
TABLE III
Typical Solubility
Class Concentration Limits
(At about _l80 o C
Production Equipment
-221-
high and low pressure columns. It operates with a small
temperature difference between the condensing nitrogen and
boiling oxygen. The liquid oxygen product, constituting
about 20 per cent of the process air, is withdrawn from the
body of boiling oxygen in the main condenser. This limits
the average concentration of hydrocarbons in this liquid to a
maximum of about 5 times that in the air processed,
assuming no reduction prior to the main condenser. This
is basically a safer system than in a gaseous oxygen plant
where the product is withdrawn as vapor and additional
hydrocarbon removal provisions, such as silica gel traps,
are required to limit the combustibles in the main condenser
liquid. Concentration by a factor approaching five normally
is quite safe and results in the typically clean liquid shown
in Table III. There is, however, the pos sibility of unlimited
local concentrations of combustibles in this heat exchanger.
The liquid must not be evaporated to dryness or unduly con-
centrated at any point since solubilities at that point will be
exceeded and there is the danger of local accumulations of
solid hydrocarbons. It should be stressed that even in a
liquid oxygen plant, where a substantial quantity of liquid is
continuously drained from the main condenser, considerable
experience is required to establish safe design and operating
procedures. At no point should a condition be created which
would permit the concentrating of combustibles to levels
appreciably above the average concentration in the body of
liquid oxygen.
-222-
02 Vflf"'O~
h~ ,.d~
-223-
foreign material such as solid carbon dioxide, water, etc.
(which can plug a condenser tube) the type of equipment
failure would be relatively mild and would rupture only a
few tubes, with no danger to personnel.
Distribution Equipment
-224-
evaporation is used. Large temperature differences in
stearn, hot water, or atmospheric vaporizers flash the
liquid and the contaminants so that the latter are not concen-
trated and solid combustible s cannot be deposited on the heat
transfer surfaces.
Discussion
-225-
E-l
J. E. Tracey
Arthur D. Little, Inc.
Cambridge, Massachusetts
F. K. Dearborn
Air Force Cambridge Research Center
Bedford, Massachusetts
-226-
the nose of an "Aerobee" rocket. A weak light signal is
passed through the atmosphere and received by the photo
tube while the rocket is in flight, and from the signal
emissions of the phototube a spectrum can be constructed.
1
R. W. Engstrom, J. Opt. Soc. Am., 36, 6 June
1947.
-227-
0
VOlTSPCASTAG[
TU8E T'fPC IP21
~IOO r _l I .! Iotre
j
10
8" OW1 0l7 or r:'tAl.'fSEJtf!j= I, ~ Us.
IIlIb R['-C~[NCt Odb =1 IIIICR0\0I'0\.1 ACROSS I
~ La...
L.04t)
.....
....
"!t1GN.Al.. OUTPVT • I ,
or- .cATlooc ,LU-"..1,lA1J
IV :I
0.
LUM[N,!.". 10
C.OLOR TE t.'P[fU,TURC :.a81cf'1'C
CHopprO . 90CJ'A f ~ ,
.....
....
f
O-
----
I/~ ,
0 ~
V'
/
'0
, 3 · "1"
"IfC
",Noo.
tlASltc T
Figure 3.
' 4 0/1 IIUT
(T.O LlNESI
Figure 4. Photo-
multiplier cryostat
-228-
significantly by refrigeration of the tube. The curve of
Figure 2 is a result of that work. It shows that this form
of dark current may be virtually eliminated, and except
for random emis sions, there will be little or no nois e to
interfere with accurate measurement of very low light fluxes.
-229-
In removing the refrigerated tube apparatus from the
laboratory to the nose of an llAerobee ll rocket, the construc-
tion requirements such as those for vibration, shock,
acceleration, weight, and particularly space were strict.
This entailed some changes in the laboratory model, the
r~sults of which are shown in Figure 4.
Discussion
-232-
POWDER PIPING ACCESS L UID LINE
"[SERVOtR PluG
Figure 1.
Figure 2.
....i
r HEAT lEA~ TRENC
6000 LITER DEWAR
"'lilt \" .... ,,,",,au," ftol,
tlllr.lL,&fIO.ll ;IIIr.T
OPERATING PEfitFORIt4-A CE
BEECH AIRCftAf" T CORPOIltA TlQN
6000 LITER DEWAR
O.'::o-----,~:-;.-- .O'--"':-;---'~O;--.~
[Ptt ER vESSEL r)IAMETE:R
•.--------;,.O
rUt ·}·--~"-T-'~-~~·'-.-~-'''~'--~·O'--~~'--~'O'--"~
Figure 3. Figure 4.
-233-
performance without the added complexity of a radiation
shield.
The vent line, liquid line, and all gage lines enter
the inner tank through a 10 inch 1. D. access manhole. All
lines are thus readily acces sible in case repair or alteration
is necessary. This feature is especially desirable for an
experimental vessel as changes are often required.
-234-
The large access manhole is a potential source of
considerable heat leak. The heat leak is minimized by
channeling boil-off gas over a series of baffle plates below
the manhole plug. Heat entering the vessel is absorbed by
the boil-off gas as it is discharged from the vessel. This
system has proved very effective in reducing loss through
the manhole.
Discussion
-236-
Question by V. J. Johnson, CEL National Bureau of
Standards: What was the maximum pressure build-up in the
dewar? Was the first unit tested with liquid hydrogen?
Answer by author: No. The dewar has only one vent system,
making this procedure impossible.
-237-
E-3
H. A. Eichstaedt
Ronan and Kunzl, Inc.
Marshall, Michigan
-238-
but years, without re-evacuation. Repairing had to be a
simple and easy operation. Trouble shooting had to be kept
to a minimum.
-239-
The use of aluminmn made possible a substantial
reduction in the weight of the container. The low tempera-
ture properties of aluminum and the low working pressure
requirement (50 psig) of the Air Force admirably suited
aluminum for this weight requirement.
-240-
Figure 1. Valve and gage unit Figure 2. End view of
viewed from the container unfinished container
-241-
Figure 3 shows how the valves, gages and piping
can be removed from the container without disturbing the
vacuum and guaranteeing interchangeability from vacuum
jacketed container to vacuum jacketed container.
-242-
Manual, to pull the handle to make sure the pressure relief
valve is free prior to pressurizing the unit.
-243-
Figure 5. Completed 50 gallon
USAF type liquid oxygen
container, trailer mounted
-244-
Discussion
-245-
Comment by H. L. Johnston, H. L. Johnston, Inc.: Our
observations at the Ohio State University some years ago
established the superiority of Santocel over other insulations
then tested. However, since that time we have found that
some perlites may be obtained that considerably exceed the
performance while other samples of perlite are inferior.
This is understandable since both perlite and Santocel are
chemically and physically almost alike. Both are exploded
silica gels. Santocel is produced by release of pressure from
silica gel forms with high boiling alcohols in an autoclave.
Perlite is formed in nature by volcanic action. As may be
expected the natural product may vary widely in its
properties.
-246-
Answer by author: The 48 hour vacuum pres sure test is
conducted conforming to the requirements of Air Force
Specification MIL-T-4829, which states that each tank will
have a partial filling of liquid oxygen with an absolute pres-
sure in the inner tank of 65 psi, and the annular space shall
show a vacuum of 100 microns or less. The container under
pressure must then stand for 48 hours without showing an
increase of pressure in the vacuum space to exceed 1
micron an hour. As an aside here, may I state that our
Air Force Quality Control Inspector has suggested to the Air
Force that this be changed to a micron increase per hour
during the vacuum test, as he bases his observation on more
than 100 containe rs which showed no increase of pres sure in
the vacuum space while the containers were undergoing the
48 hour vacuum and pressure test.
-247-
E-4
E. G. Eeles*
National Re search Council of Canada
Ottawa, Canada
Introduction
X-Ray Camera
-249-
HEATER +THERMOCOUPLE WIRES
~~ AND GAS THERMOMETER CAPILLIARY
SEALED AT EXIT.
~ LIQUID N2 DEWAR
LIQUID H. DEWAR
RADIATION
SHIELDS
HEATER
-250-
The specimen holder is connected to the base of the
refrigerant container through the thermally adjustable coupling
shown. This coupling is adjustable through a wide range of
conductivities by means of the brass and copper sleeves and
by the pressure of helium gas in the annular volume. By
control of the voltage applied to a heater which is wound on
the specimen holder any specimen temperature above that of
the primary refrigerant can be maintained with the minimum
consumption of the refrigerant.
-251-
measurements show that, at room temperature, this thick-
ness of cellophane transmits about 10 per cent of the thermal
radiation falling on it. Experiment has shown that in service
the cellophane attains a temperature not more than 100C
above that of the shield. The use of cellophane as the window
material would appear to lower the heat leak to satisfactory
limits, the as sumption being that the absorption coefficient
of the cellophane has a value at lower temperatures similar
to that at room temperature. Experiment suggests that this
is so since removal of the cellophane from the outer windows
results in a measurable increase in the temperature of the
center radiation shield.
Conclusions
-252-
The work described here was performed during the
tenure of a postdoctorate fellowship of the National Research
Council of Canada.
Discussion
-253-
E-5
R. D. Goodwin
CEL National Bureau of Standards
Boulder, Colorado
Introduction
-254-
figure. 1 The two bridges, of course, become identical when
all arms are equal. In the following discussions, the ratio
arms always will be taken of equal resistance. This condi-
tion simplifies the problem of compensation of resistances in
the connecting wires between bridge and thermometer.
Qualitative Properties
5
L. Page and N. 1. Adams, Jr., Electricity, D.
Van Nostrand, New York, New York (1949).
-255-
f }® I·
A IJ
/ ~ ~
~
~ .e!
t ~®r.
~
~~.
~
~ !! m
G:f
tr
4
10
I02r----.-----,----~
I~'YII
lo'l-----t--- +d
~ A a
rEl
a
g
a> R
?
IWR(2. (I +.8. +~) 8E I3T
.!! " B- o a g 81n R / aT
a b
-256-
The line, W, repre sents the powe r dis s ipa ting
capacity of any thermometer, which results from the thermal
conduction behavior of insulating materials. For very wide
ranges of temperature it would be both convenient and prac-
tical to employ a bridge connection which would produce this
qualitative power behavior automatically with temperature.
-257-
select bridge circuit A, with ratio arm resistance greater
than the maximum thermometer resistance. The properties
of this circuit are as follows:
-258-
100,----.----,----,----~--~
10~---+----4----+----4----J
THERMOMETER ./
POWER, m w B b
rE1
9
1.0 t-++---?L-f-----,~~---t--~ b < R b RsEMI
Figure 6. B~idge
50 100 200 300
for thermistors
TEMPERATURE. oK
Figure 5. Metallic
thermometer power
a:
~102~______' -______- r______- '
R=104n BRIDGE B
b=loon
THERMOMETER
POWER, m W
I L -__~~~~__~~~__~
BE/BT =0.001 V;oK d ~4
I DETECTOR I NPUT IMPEDANCE, n
4 10 20 40
TEMPERATURE, OK
Figure 8. Relative
Figure 7. Carbon thermistor power
thermometer power
-259-
Bridge for Thermistors
-261-
Summary of Re sults
Mathematical Appendix
rR + cC + bB = rR + aA + gG + bB =
rR + aA + xX = v, (ld, Ie, 1£)
-262-
Figure 9. General bridge circuit
G-
-263-
B + G + X = const.
r a (a+ x+ r) v (2a)
(b+ c+ r) -c r v (2b)
(b+ r) (a+ g) (a+ r) v. ( 2c)
D = M+ Ng =P + Qx ( 3)
where the coefficients are defined as
GD/v = cx - ab ( 5)
-ml g = I + d In gl d In G ( 6)
m = MIN ( 7)
-264-
Bridge Sensitivity. A specific bridge sensitivity may
be defined as
S = (l/v)dE/d In x
where E = gG ( 8)
S = gcx/D (9a)
o
U = (b/c)/(I+b/c)2 (lOa)
(lOb)
L =I + m o /g (IOc)
where
i = (1 + n)/2n; and j =(1 + n)/2 (lId)
-265-
Bridge B is defined by fixed arms band c of Figure
9. With c= nb, arm a is adjusted to a nx. Hence, =
Ub = n/(l + n)2 = 1/4ij (1Za)
Zb =1 + (r /2 b)( 1 + b / x) /j ( 12b)
(13)
becoming in particular,
W
o
= H(dE/dT)2/ (d In x/dT)2 (14)
-266-
It follows that
1 when (x / a) ~ 1. (ISb)
>
1/2 1/2 -1
dEI/dT = J o W 0 K[ g (1 + l/q)] d In x/dT (IS)
1/2
K
a
=x /(1 + x/a) Fa = 2x/(1 + x/a) (1Sa)
1/2
Kb =x /2 Fb = (b + x)/2 (lSb)
q = 1. (19)
-267-
when the relative power requirement therefore becomes
(21 )
Discussion
-268-
E-6
CALIBRATION OF THERMOCOUPLES
AT LOW TEMPERATURES
Introduction
Apparatus
-269-
20000
EXCHANGE
GAS,-+---f.
~rt-+-- UPPER
TEMPERING
10000 I
BLOCK 6000
4000 i7
UPPER VENT
TUBE
2000
i
:/'
, i
GAS
1000
iii
THERMO-
METER +--*-4--1+ 600
400
1-/~I,-H-'lI--- MEASURING
BLOCK l 200
iI
)
~ 100
"- 0
7
~
FILL UJ 40
TU BE -+-----jP"l ~-+-- VACUUM
I)
1
0
)
0
VACUUM THERMOELECTRIC
CONTAINER
6 I, FORCE OF
4 u-2.1 AT". Co VS Cu
2 V 4 6 10 20
T TT
40 60 100 200 300
~-+-+-- LOWER TEMPERATURE. oK
TEMPERING
BLOCK
Figure 2.
Figure 1. Thermometer
calibration apparatus
6= /
o
I 10-
ELECTRICAL l
RESISTIVITY
E
:.: 50 .r: - COPPER WIRE
o
o
.... /
,:
~_ 40 -
a:
UJ
-
-
i--
I--
I--
THERMOELECTRIC
POWER OF
Au - 2.1 AT%.
Co VS Cu
-- - t::
>
;::
'!? 10- 7
il
3: (J)
o0.. - - UJ
30 a:
u
I
a:
~ /
~
.J
UJ
o
20
/ 8 /
:;; 1/
~ 10
./
r
~
_I""
o 9
I 34 6810 203040 6080100 200 300
10 50 100 300
TEMPERATURE. °K TEMPERATURE. OK
Figure 3.
Figure 4.
-270-
adjusted by a heater and to which are attached the measuring
junctions of the thermocouples. The measuring block contains
a gas thermometer bulb and a re-entrant well for a capsule
type platinum resistance thermometer. In addition, certain
temperatures can be established by measuring the vapor
pressure of pure liquid that can be introduced into the upper
chamber. The wall temperature of the upper chamber is
controlled to about the same value as the measuring block
by a differential thermocouple and heaters. This container
is neces sarily filled with exchange gas from the refrigerant
liquid. With exception of the lower tempering block, which is
of brass, the other major components are of copper. However,
the tube connecting the upper and lower containers is of
stainless steel to minimize heat flow to the refrigerant. A
copper heat short is connected from the refrigerant container
to the upper vent tube in order to bypass to the refrigerant
bath any heat leak from the surroundings to the upper
container.
-271-
copper. It should be noted that no measurements were made
below 4 o K. Figure 3 showing the thermoelectric power versuS
temperature is more informative. This shows the sensitivity
of the couple at various temperatures. The sensitivity of the
gold-2.l atomic per cent cobalt versus copper couple is 15
microvolts/oK at 20 0 K as compared to 6 microvolts/oK for
constantan versus copper couple at the same temperature.
Table I gives the thermoelectric force and the thermoelectric
power for the gold-2.l atomic per cent cobalt versus copper
couple for the temperature range from 0 to 300 o K.
Conclusion
-272-
TABLE I
Temp. emf dE/dT Temp. emf dE/dT
( OK) (mv) (mv/oK) ( -K) (mv) (mv/oK)
Discussion
-273-
Question by E. T. Benedikt, Douglas Aircraft Company: How
long doe s the helium procedure of calibration take?
-274-
F-1
Introduction
-276-
ExperiTI1ental Methods
-277-
The other type of "flow" system is called the vapor-
recycle method. In this approach a two-phase mixture is
placed in a vessel at controlled temperature. Vapor is then
withdrawn from the cell and recirculated through the liquid.
This contact between vapor and liquid provides a fairly rapid
approach to equilibrium. After equilibrium is reached,
samples of the liquid and vapor are withdrawn and analyzed.
In this method also, some provision must be made to keep
the pres sure constant during sampling. The vapor -recycle
method was used in this investigation.
Experimental Apparatus
-278-
caUl TlIAP
SCHEMATIC DIAGRAM OF EQUIPMENT TEMPERATURE CONT1IOL SYSTEM
Figure 1. Figure 2.
PRESSURJ
DIA~~
NITROG
.':'''! ~
SYSTE
2 I
/ ''\
f: ~ \
REGlLAT~
' ,
IDOD
V '\ "\
l~ )
vr 1--- LJ
SCHEMATIC DIAGRAM aF
EQUIPMENT
CHRCIIIOTOGAAfIHIC 0
MOL FRACTION HVDROGEN
--- 08 1.0
Figure 3. Figure 4.
-279-
It was necessary to dampen the movement of the
differential gage to prevent rapid movement of gases in and
out of the differential gage. This was done by inserting about
two feet of 17 gauge wire into the tubing leading to the differ-
ential gage. This made the gage completely insensitive to the
pressure surges caused by the recirculating pump but still
permitted it to indicate the liquid level.
-2BO-
FrolTI the pUlTIP the vapor continues to a stainless
steel cell, terlTIed herein the "variable volulTIe cell," designed
for 10,000 psi. The vapor passes into the cell through a
perforated tube and rises through a series of baffles to exit
at the top. These baffles provide turbulence necessary
for periodic flushing of the vapor in the cell. The volulTIe
of this cell is approxilTIately 350 cc and provisions were lTIade
for adding lTIercury to the cell through the bottolTI, providing
a lTIeans of pressure control.
-281-
By controlling the pressure difference between the
storage flask and the surge drum in the evaporated nitrogen
line, the rate of refrigeration could be controlled quite
accurately.
-282-
calibrated up to 2000 psia by comparison with the lower
range gage, and also showed no error as large as one scale
division.
-283-
Operation of Equipment. While still at room temper-
ature, the system .;as evacuated for 1/2 to I hour and then
charged with either nitrogen or carbon monoxide to five or
six hundred pounds. The equilibrium cell was then cooled to
the temperature desired and more nitrogen or carbon monoxide
was added until the differential gage indicated that the equilib-
rium cell was 1/2 to 3/4 full. Hydrogen was then charged
until the desired pressure was approached. All the gases
were charged at a slow rate to insure as complete a removal
of contaminants as possible.
The liquid sample line was then purged and the liquid
sample was collected. The liquid sampling time was usually
about five minutes. The valve in the recycle outlet line was
-284-
then closed and the sample system from the containers to
the sample valves was re - evacuated. The vapor sample
valve was then opened and the pressure on the external
portion of the system was allowed to drop about ten psi
before closing the valve again. The sample system and
vapor sample container were re-evacuated and then the
sample was collected.
-285-
The equilibrium cell and cooling system described
above were used only to obtain the data at -310 0 and -280 0 F.
For the data at -240 0 F, the equilibrium cell and cooling
system described by Price lO were used.
Analytical Methods
-286-
sample of gas would reproduce the "per cent of full scale
deflection" to within one part per thousand even though the
zero point of the cell and/or the full scale deflection varied
slightly. The full scale deflection and zero point would be
determined each day that samples were to be analyzed.
-287-
simplify the separation of the components as well as the
determination of composition. The nature of the two compo-
nents indicated that adsorption chromatography offered the
best answer to the separation. Activated charcoal (60-80
mesh) was chosen as the separating medium after several
adsorbents were investigated. The gas chromatography
apparatus is diagrammed in Figure 3.
-288-
Experimental Results
-289-
4000
'"
i'
PRESSUR - COM OSITION
DIAGRAM QROGEN-
CARBON M~~~IO~ SYSTEM
50
f - - - --
/ \ ~~t. ~M P:OE~S~~"GEN
NITRpGEN BINARY I
!
3000
1\ ~
ill 200
0.. f
,
iO
- \ I \~
, ...
0.
~
\ \
0 \
/~
\
I
:11'
",
t ..
I.
'<!
"~...
- OOF
/
1000
f '7
" '.to
~
",
" r---
V
-V
0.1 f--
V LJ
o0 0.2 0.' 0.6 0 1.0 o.0 20 50 100
MOL FRACTION HYDROGEN 1000 5000
PRESSURE - PSIA
Figure 5. Figure 6.
I--
HYDROGEN Ie - \ALUE VS MOL
FRACTION NITROGEN IN LIQUID AT -~
-
"'""-.).. ~
0
30
~~
...
0
~ --. r----c o I
Sf A
i
'"3
'\
\.
'\, '\,
\ "\ 10 '-l!. --. I
~
"
.0
0 --, --
0'--
-~ 800 SIA
0
'.
. "- /"
I,
"" I .
I I
--,-I--
--
~ ~
o 1100 SfA. I I
"-" / i
f -~
0
0.1
t- PSIA
2D
i
I"" '0' .20 0.40 0.00 0.80 1.00
PRESSURE PSI A MOL FRACTION NITROGEN IN UQUID
Figure 7. Figure 8.
-290-
(KP) FOR NITROGEN VS MOL FRACTION
NITROGEN IN LIQUID AT -280· F
1000
90 0 I
X-KP AT INFINITE DILUTION EQ(l5
I
-
800
1400 ~SIA
700
SO 0
j I
I I I
500 '"
I IIbo
I
z SIA
III 40 I
o i I
I
0
a: I
~300 I
~~ISIA i •
a: ! .,
...
o
I !
a:: 20O r - r--.:
!
~ r-- ' 5)0 P IA
"ID'
",.
Figure 9.
F
o ,....---:-
:::0 40
z ~
o
m \\~ ~
~300.....-
---
o
a: -'""
~ 800 PSIA
~
+
_20
-
Q.
'" D"
50~ 1----"7" 0
I 315 PSIA ~
0
0
100
90 X-KP AT INFINITE
80 DILUTION EQ. ( J
70
60
o 0.20 0.40 0.60 0.80 1.00
MOL FRACTION NITROGEN IN LIQUID
Figure 10.
-291-
TABLE I Smoothed Data
Code No. Temp. Press. Liquid Compos1t1on Vapor Composition K-Value
op psla. m.f.H2 m.f.N 2 m.f.CO m.f.~ m.f.N 2 m.f.CO H2 N2 CO
-292-
It is apparent from the K-x plots that the K-value of
nitrogen at constant temperature and pressure is fairly
independent of composition except in the low pressure-low
nitrogen composition region, while the carbon monoxide
K-value becomes more concentration dependent as the pressure
increases, and it always increases with decreasing carbon
monoxide concentration. The hydrogen K-value seems to be
fairly uniformly independent of composition in the region where
the carbon monoxide composition is low •
. Discussion
-293-
F-2
D. I. J. Wang
Linde Company
Division of Union Carbide Corporation
Tonawanda, New York
-294-
samples of vapor and liquid. In studying the system argon-
oxygen which boils at approximately 90 0 K at one atmosphere,
several considerations lead one to conclude that a still of
somewhat different design might be desirable:
7
J. C. Chu, et a1., Preprint No. 22, Annual Meeting
A. I. Ch. E., St. Louis, Missouri, December 1953.
-295-
In the final design of a still possessing the above
characteristics, the colwnn consists of six trays, one inch
in diameter, with vapor sample taps located beneath each
tray. The trays, furthermore, have a high liquid depth and
low vapor velocity to insure good vapor-liquid contact, a
relatively small nwnber of perforations on the tray to prevent
the coalescence of bubbles, and a very short horizontal path
of the liquid flow to eliminate the existence of a concentration
gradient in the liquid across the tray. Heat for vaporizing the
column liquid is supplied at the reboiler through an electric
heater; liquid oxygen is used for condensing the colwnn vapor.
Proper insulation is provided to minimize heat leak.
8
B. F. Dodge and A. K. Dunbar, J. Am. Ch. Soc.,
49, 591 (1927).
-296-
concerned was then calculated. Reproducibility of the data
was within ± 2 per cent of In a over the entire composition
ranfe. The accuracy of the experimental method is estimated
at - 3 per cent of In a. A small correction factor to the
relative volatility a, in the order of 0.01, is applied to each
observed point in order to bring them all to the arbitrarily
chosen temperature of either 90.7 0 or 95 0 K.
-297-
0·7
d.'"
0 1~·7~.
0.>
x "P"X.
1 d·f ~
~
1"'-........
N ~--~ 1/1°·7 0"".
~l
~ O.~
~ f:::::-<,..
~ 0.2
15 ';1( v -- ~
0./
o
~ t:=::..
--
o /0 20 ,30 1'0 sO GO 70 80 f" )00
o
o 10 ,?O.$o 'f0 SO GO 7" 80 7'0 /0"
-298-
TABLE I
In (-'!....!..) dx =- 0 (2)
Y2 T
s: In (2 ) dx
Y2
, where lIn ....!...L I is the
1 '(2
'I 1
.2L I dx absolute value of In(--)
So lIn
'{2 Y2
TABLE II
Thermodynamic Consistency of Vapor-Liquid Equilibrium
Relations for the System Ar gon-Oxygen at 90. 7 0 K
-300-
Redlich-Kister criterion, since this criterion had to be
satisfied in the formulation of that correlation. It is also
noted that the 4 per cent "unbalance" in the author's experi-
mental work at 95 0 K appears essentially to be all in one
direction and is, of course, roughly equal to the amount of the
deviation of the present work from the Linde Company
correlation at the temperature in question.
(a-I) 0 = ( .!..2'..) ]
x- a x p x-O ( 3)
-301-
TABLE III
Calculated and Observed Relative Volatilities of the
System Argon-Oxygen at Infinite Dilution
Conclusion
-302-
2) For syste111s c0111paratively easy to separate,
equilibriu111 data over the entire range of
c0111positions 111ay be deter111ined si111ultaneously
by 111easuring the c0111position on successive
trays.
-303-
Sufficient data have not been gathered in this work
for correlation over the entire range of pressures and
temperatures of interest. This paper has instead been pre-
sented chiefly to suggest the possibilities of the novel
equilibrium still employed in these studies and to bring to the
attention of investigators in the field this rather abnormal
behavior in the argon-oxygen system with the hope that more
extensive and elegant endeavors may thus be stimulated.
Discussion
-304-
F-3
Introduction
-305-
These adhe sives were two component systems
consisting of a liquid amine type curing agent and an epoxy
resin paste containing an inorganic filler material. The filler
was predominantly alumina and asbestos totalling approximately
55 per cent of the resin-filler mixture. The curing agent
used was diethylaminopropylamine, a relatively slow reacting
type recommended fOr elevated temperature curing.
Test Procedures
-306-
t. 0
,
V
/
~
-.,...."..\
, /'
V
,........
.mc..<I) -.cJutl.n "-
0
""'
0000
'[
--- ~IOU
"l
1\ 0000
. ,
'"
e-
~; I
/ """""""
"""" ¥
'1t
o
- ....~ f..--
.......---
V
Figure 3. Strength of a filled epoxy o
resin adhesive bond vs. exposure
temp. prior to testing at room temp.
-307-
Specimens subjected to elevated temperatures prior
to testing were maintained at the temperatures noted in the
results for eight hours in a helium atmosphere. Those
subjected to low temperatures prior to testing were alter-
nately cooled and warmed twice between 76 0 and 300 0 K by
immersion in liquid nitrogen and warming in ambient air to
room temperature. Cooling and heating rates were conse-
quently quite high.
Results
-308-
a 90 per cent cohesive failure with the remainder of the
failure occurring adhesively. The specimens which had been
o
subjected to 43S K (165 C) failed onl} about 50 per cent
0
cohesively.
-309-
sharply diminished strengths caused by elevated temperatures
are clearly shown.
Thermal Stress
2
N. A. de Bruyne, J. Appl. Chern., 6, 303 (1956).
-310-
that this shrinkage is not negligible. Within the family of
epoxies, shrinkage can be minimized by controlling the degree
of cross-linking during cure and by adding fillers to the
resin. But minimizing cure shrinkage would be desirable, of
course, only if it did not involve a more serious compromise
with other properties of the cured adhesive.
-311-
UU.-JLL[.n
r.00)()I Qt."IIt-J
(~!> ,. AL.U M '1J"1-7-r~"T/""7-r-r?,""""/"'7"""71
r'L L.ED
A"" A"'B~">TO'5)
~t>OXY Rt';'''1
hLUUIUUU
copp~o
'\04. ...T4IULr....,,>
">Tr.r.L
8000
I
T~tJC;IOt.J
coooo
rr
Cl
-
~
1- 4000
I.!J
::2
LJ
CJ
tr ., 1.1I~.h.12
<!OOO
-312-
The above argument based on thermal stress is
probably insufficient to account for the entire amount by
which the observed shear strength is reduced at low temper-
atures. Neither does this argument account for the lower
room temperature values observed after a thermal cycle to
76°K. It is not expected that any chemical or other irrevers-
ible changes are produced in the metal or adhesive by the
temperature cycling, so the most logical explanation of these
effects seems to lie in the consideration of the transient
stresses produced by the cooling rate. Even if adhesive and
adherend possess similar expansions between the temperature
extremes, the adhesive will lag the metal in most changes
of temperature because of its lower thermal conductivity and
the contact resistance presented at the interface. The stress
produced in cooling and warming may thus be sufficient to
cause local failures of the bond. This would explain the
observed residual effect on shear strength. (See again Figure 3.)
-313-
Epstein 3 mentions that the strength of adhesives which
do not liberate gase s during cure show little dependence upon
glue-line thickness. He states, however, that thickness
should not exceed ten mils for optimum strength. Figure 6
shows strength versus thickness at 300 0 K of bonds which
received no treatment other than the usual curing cycle prior
to test. The slight curvatures are probably better attributed
to experimental error than to any real effect.
Conclusions
-314-
5) Given a strong adhesive bond at room tempera-
ture, the match between thermal expansion of
adhesive and adherend is a good criterion of its
strength at low temperatures.
Discussion
-315-
F-4
Introduction
Specimens
1,2,3 A • S • T • M ., D eSlgna
' t'Ion E 23 - 47T •
-316-
With the exception of the OFHC copper, the notch of
each of the copper alloys was milled with a one tooth cu~ter.
For the OFHC copper, a special jig was devised to drill the
hole of the keyhole accurately. The slot of the keyhole-notch
was cut with a saw. The notch of each of the aluminum
alloys was milled with the one tooth cutter.
Experimental Procedure
-317-
cyclohexane. Rapid transfer of the specimen from a liquid
nitrogen bath makes testing possible near 76°K.
Accuracy
Results
-318-
COPPER ALLOYS
OL-~--~---L--'=OO~~--~'OO~~--~~~~--~~~--L-~XO
I
TEMPERATURE - oK
Figure 1.
0
P---....
100
0
24S-'T4
7!5S-T6
0 00 100 150 200 250 XO
TEMPERATURE - °K
Figure 2.
100
HAYNES ALLOYS
0 1,-
trANi, U·N01CI1 ::::::l'
~_,,01C> -
V ;:::.-0 t-;;:l.LO'l~' ,
-
- - " ........-: f-"" HAST. C, V-NOTER:
Figure 3.
-319-
Copper Alloys
-320-
The Berylco 10 exhibited a very granular fracture
and broke completely at all temperatures.
Aluminum Alloys
-321-
Haynes Alloys
-322-
Conclusions
Discussion
-323-
strictly a part of the impact strength. On the other hand,
one might consider that the Charpy test is a good test for
aluminum alloys in that it enhances their property of
ductility.
-324-
F-5
Introduction
-325-
,,~
J1...o,[
Figure 1.
Figure 2 .
..
r' ..
'\ \\
..
, , ~
.~.
---
Figure 3.
-326-
c = Ae / Ae ( 1)
x t
tx = tf +
- cAt = tf ± c9 (2)
t
Mathematical Relationships
-327-
from the fin base outwards to the tip, where 8 L defines the
temperature difference there, as compared with 8 at the base.
The usual differential equation may be set up to ctover this
case, following the application of the general as sumptions
specified previously, and leading to the conventional expres-
sions,
( 3)
a = ( 4)
l/l
where n = [(P/Ac)(h/ k)] (5)
1 S L
(8dx)
__ 8t (e 2nL - 1)
- ( 8)
L 0 nL (e lnL + 1)
8 1 (e lnL _ 1)
C - 2!!. :: (9)
f, m - at nL (e 2nL + 1)
-3l8-
C is often known as .. fin effectiveness." From equation
f,m
(8), the heat flow between the constant temperature fluid and
the variable temperature fin (as controlled by its properties,
etc.) may be calculated,
= (10)
9x = ( 90 I 2)( e nx + e
-nx
) ( 11)
and (12)
-329-
PIA = 2(a + b) I (ab), for a rectangle a x b (15 )
c
8x 1 [nx n(2L-x) +
( 2h e ikn)(e n (L-x) - en(L+x)) 1
Cf = = --- e +e
(l+he/kn)e
nL
+ (l-he/kn)e
-nL
8t e 2nL +l
(18)
- 330-
It is to be noted that with tip edge h = 0, equiva-
e
lent to an insulated tip, equation (18) reduces to equation (7).
-nx
+ (l-he/kn)e ] (19)
nx -nx
(1 + hel kn)e + (1 - h e /kn)e
(20)
nL -nL
(1 + he/kn)e + (1- he /kn)e
-331-
Figure 4 shows. in simplified form. the division of a
typical fin into uniform s trips A x wide. The ener gy flow
process involves both fluid surface heat transfer
[ A% = (P A x)h 0] and fin conduction heat transfer
(21)
(22)
-332-
This step wise computation is, of course, readily repeated
in the traverse from left to right towards the base. The
base can be any number of strips distant from the t~p, with
the successive strip computations via equation (23), i. e. ,
in effect, 8 = 8 2 + 1(~8)2 + 2(~8)3' etc., up to N in
num ber. It3 is to be noted that prosecution of the calculation
procedure, leading to the ultimate evaluation of C f , may
equally well be carried out graphically.
-333-
Figure 4.
Figure 5.
10
~
V
/~
/; /
6
-- - --
~
::::V
4
/0 /
~ ~
2
- - I---""
~
00 4 6 10 12 14 16 18 20
POSITION ON FIN
Figure 6.
-334-
Rh/Rc :=(Ack)/ [(PDoX)(hDoX)] := 1/[(h/k)(P/A C )(.:;\x2)] :=
-335-
values of m serve to directly modify the corresponding
values of the resistance ratio Rh/Rc'
- 336-
FIN POSITION
Figure 7.
Figure 8.
Figure 9.
-337-
illustrated in the different case treatments, is likewise shown.
Figure 9 shows the comparative results of Columns I and G,
i. e., wedge versus rectangular shape.
Conclusion
-338-
arithmetic average of the C f values at the different fin
positions, as obtained by any of the different comparative
methods cited. em may also be further determined from the
electrical model.
(27)
as (28)
-339-
TABLE I
Acknowledgment
L = Fin length, ft
t = Temperature, of
Cf = Fin temperature difference ratio, fraction of
overall value e t
e = Temperature difference, of 1/2
n = Fin characteristic = [(PI AcHh/k)]
P = Fin perimeter, ft.
Ac = Fin cross sectional area, sq ft
Discussion
-341-
F-6
TEMPERATURE-PRESSURE-TIME RELATIONSHIPS
IN A CLOSED CRYOGENIC CONTAINER
S. C. Huntley
National Advisory Committee for Aeronautics
Lewis Flight Propulsion Laboratory
Cleveland, Ohio
Introduction
Procedure
r.==~d'I=60~ATMOS VE,T
INCREASE IN AVERAGE lI0UID TEMP _
INTERVAL OF TIME WITH CONTAINER CLOSED -
(M, A'v,
80 ~,oF;HR ~ HEAT LEAK RATE
6T~ A~
60 --r;:-)X 61. - PeAL AT .61 •
• O.H" 304 5.5. INNER SHELL
VAPOR
WITH ALUM-FOIL WRAP
PRESSURE,
HAND - CRANKED PROPELLER IN. Hg GA~5
COPPER RADIATION
SHIELD WITH
ALUM-FOIL WRAP- -_--.............
'-VACUUM JACKET
20
PRESS'>
o 5 10 15 20 25
TEMP .. OF ABOVE REFERENCE
" " 40
""
HEIGHT,
""
IN.
,-
20
.,.," "'--CALCULATED (REF. LID. N2 AT
-344-
in average liquid temperature at any time after the container
was closed was used with the curve of Figure 2 to get the
calculated pressure for the same time interval.
-346-
HEAT LEAK RATE. (6T",AV 161).9.2- F/HR
lIaUID CONTENT. 30~_
20
PRESSURE, 10
IN. HQ. GAGE
30 40
~ I-STIRRING LIQUID
LIQUID CONTENT,
10 ".
,,---:::;:::::.:::--~
~=-~--~~--~'---~80~--'IOOkn---7.'20
PRESSURE.
IN. Hg GAGE
TIME. MIN
60
40
PRESSURE.
-II- UI-
IN. HQ. GAGE
20
ELECTRIC e lECTRIC
FAN ON ....., FAN ON
I MIN 9 MIN
1IME-
-347-
2) Heat conduction downward from the uninsulated
steel lid along the inner shell and also through
the quiescent vapor, which brought heat directly
to the surface.
-348-
natural convection in the vapor causes the heated, less dense
vapor to accumulate at the top of the container. The heated
vapor, being farthest from the surface, has a beneficial
effect as long as the heat remains there, because less
conduction takes place between the cooler vapor near the
surface and the liquid layer at the surface. Altering the
vapor-temperature gradient by forced convection may result
in more heat getting to the surface, with a resultant increase
in surface temperature and a subsequent increase in pressure.
The effect on pressure of altering the vapor-temperature
gradient was demonstrated by using an electric fan to mix the
vapor. The fan was installed in a fixed position, which
required that the liquid content be changed to vary the height
of the fan above the surface.
Summary
-349-
equilibrium relation based on the average liquid temperature.
Stirring the liquid to equalize liquid temperatures caused a
reduction in pressure, which would extend the time required
to reach a limiting pressure. If the design of a closed
container will not accommodate stirring the liquid, a lower
pressure can be achieved with a given heat-leak rate by
minimizing the direct heat flow to the surface. Even a
momentary increase in heat flow dir ectly to the stable liquid
layer at the surface should be avoided to achieve minimum
pressure in a closed container.
Discussion
-350-
fan effects. Doesn't this then indicate that convection is the
major contributing factor in this phenomenon?
-351-
Question by W, Blackman, United Aircraft Corporation:
Would you expect the same type of temperature distribution
within the dewar if the dewar pres sure were raised above
the critical pressure of nitrogen, i. e" would the steep
gradient exist at what was formerly the liquid surface?
-352-
G-l
R. J. Corruccini
CEL National Bureau of Standards
Boulder, Colorado
Introduction
-353-
without intervening collisions. Hence the heat flow is propor-
tional to the pressure. It is not related to the thermal
conductivity coefficient since the latter property is meaningful
only for the gas in bulk. Free-molecule conduction usually
prevails in high vacuum insulation, though occasionally the
vacuum will be sufficiently poor that an intermediate condition
prevails.
1
E. H. Kennard, Kinetic Theory of Gases, McGraw-
Hill Book Company, Inc., New York, New York (1938).
-354-
H ::: (1)
a :::
(2)
p is the pressure
M is the molecular weight of the gas
T is the absolute temperature
a is the overall accommodation coefficient
A is the area
-355-
non-equilibrium gas which has a value intermediate between
T 1 and T Z but is not the average. The meaning of p also
needs to be examined, since, under the conditions for free-
molecule heat conduction, the thermal transpiration effect
operates and the local pressure varies with local temperatYJ;e
throughout the evacuated space according to the law, p/(T) 1!:..
= constant. In the derivation of equation (1), p is as so cia ted
with T and since p/(T)l/ Z appears identically as a factor in
equation (1), it is only necessary that values of p and T
appropriate to the same point in the vacuum system are
inserted into the formula. In principle p could be measured
by a suitable gage located between the walls where the
effective gas temperature is indeed T. In practice, however,
a value of p appropriate to an external gage at ambient
temperature is used. If the condition, L» d, also applies
to the passage connecting the gage to the vacuum space
(where d is now the passage width), then the temperature at
the pres sure gage should be inserted for T making it
unnecessary to evaluate the effective temperature in the
region of heat flow. This will be true even when the gage
temperature lies outside the limits T1 and T Z '
Mixtures of Gases
k = d
] (3)
)L
-357-
this formula for a random porous structure having mean pore
size d and applied it to the flow of heat in silica aerogel.
Use of the formula for calculating conduction acros s vacuum
insulation has been justified on the basis that the formula
should still hold for a uniform structure characterized by a
boundary spacing equal to the aforementioned parameter d.
Accepting this assumption, we can test the validity of the
formula by seeing if it reduces to accepted forms at high and
low pressures.
( 4)
p = Mp/RT ( 5c)
u = (8RT/'lTM)1/2 (5d)
m
-358-
H = 1< (T 2 -T 1) /d ( 7)
Q (8b)
H = (9a)
k(T2 -T1)
H = = (9b)
Provided that g 1 < < r l' g2 < < r 2' (and the above equation
-359-
BOUNDARY BOUNDARY
#2 #1
T2 ......
............
...... ...
)'~
~A 4f»~ DISTANCE
~ ~,q,., >-
.-.- (In Direction
«
0::
u~
~ of Heat Flow)
lLJ
Q.
::E
lLJ
I-
",,,
,,
TI
I
~g2 >1< d >I<gl~
-360-
H
N k(TZ - T 1)
r (10)
rl (In -=-z... + ~ + ~
rl rl rZ
Dushman shows that if gl = gz' equation (9a) reduces
(11)
i = I, Z
Now if k 1 ,
1
and T
'I
1
*
are taken to be suitably defined
-361-
The representation of the transition region by equation
(9a) is equivalent to multiplying k by the factor d/ (d + gl + gZ)
which for diatomic gases, small temperature difference, and
unit accommodation coefficient is approximately d/ (d + 4L).
By contrast, Kistler's equation applied to the same conditions
is equivalent to multiplying a value of K that is four times too
small by a transition factor d/ (d + L).
TABLE I
40 K 20 0 K 76°K 300 0 K
T )1/2
L = 8.6 --'!L ( - ( 12)
p M
where L is in cm. T is in deg K, T] is in poise sand p is
in mm Hg.
- 362-
Experimental Values of Accommodation Coefficients
-363-
data of the second type. 9, 10, 11, 12, 13
TABLE II
Accomodation Coefficient a 1
T, oK Helium Hydrogen Air
-364-
The experimental values have nearly all been
obtained at gas pressures of 10 microns or more, since, in
working with filaments, free-molecule conduction predominates
when the mean free path exceeds the filament diameter, and
the latter is quite small. On the other hand, the pressures
required in vacuum insulation on an ordinary scale are usually
0.1 micron or less. The pressure variation of 0.1 for various
gases on platinum near ambient temperature has been
measured by.Alndur, Jones, and Pearlman 14 for pressures
in the range 10 to 500 microns. The values are comparable
to those in the above table but show small increases with
increasing pressure. .Alndur l5 has shown that this variation
can be accounted for on the assumption that the accomodation
coefficient is a linear function of the fraction of surface
covered by adsorbed gas, the latter being given by Langmuir's
adsorption isotherm. If his expression is used to extrapolate
the accommodation coefficient to pressures of 0.1 micron or
less, values of only a few per cent are obtained, approaching
the values mentioned previously for flashed surfaces. How-
ever, unpublished work at this laboratory at pressures down
to 0.1 micron does not show such a decrease but rather
indicates values conforming approximately to those in the
table.
Working Formulas
-365-
TABLE III
NZ ~400 .0159
°z ~ 300 .0149
HZ 77 and ZO .0398
He any .OZ80
Discussion
-366-
G-2
Introduction
-367-
glued satisfactorily to aluminum for use at low temperature,
that the gage output is repeatable, and that a zero shift at
low temperatures may occur. Day indicates that the gage
factor changes with temperature; Majors shows that it does
not. Majors shows that at _320 0 F the gages are extremely
temperature sensitive, indicating a false 400 psi stress for
10F temperature change in active gage when at -320 0 F
approximately, and with the compensating gage at room
temperature.
-368-
ElflNrGMi£S
'Il#NSION
" COMP~S$IOAJ
Fli.i. WlrHNITR06£N
1000
BOD
ROOM fEMRaIRy£
av/£}EPBY J./e
(12 % CNAN6E IN
I: AT -32o-F) .1
600
.~ 400
-~
"~
~
'3
~
~' 0
,2
~
t;
"~ -.
J>'/,
\;
~ ~
~
/
/ ~
-zoo
I- ~
/ / ~ -4000
/ / <!'
-~. ..,~
~
-'000
~ -IJOOO
-/0000
Figure 3, Temperature
characteristics of ABD-l gages
-369-
3) The low temperature zero shift of the A-3
gage is large and unpredictable.
-370-
gel and securely taped at the opening. In this way, diffi-
culties with moisture were eliminated.
-371-
JODO
~ 600
't
~"
~ 600
~ ~
\i 400
~
f
'l
ZOO
..,~ 0
+20 "'40
<:. ZOO TEhlP. OF t:/CTIVE 61/6,0 MINUS
~ TEMP. OF DUMMY 64GE, o f
~ 400
~
~ ~
~ 600
~
800
1000
STRI/IN GI/GE
SELECTION
SWITCH
SR-+ STRRIN
INDICI/TOR
- 372-
secured to the tank shell by a single screw in order that
thermal strains in the compensating block would be eliminated.
Grease was liberally applied under the compensating block
for better heat conduction. Copper-constantan thermocouples,
connected in differential, were located near the active gages
and on the compensating block so that a correction to the
indicated strain due to temperature difference could be made.
- 373-
TABLE I
22 0 F
Gage Co1d- Warmup- Warm- Co1d- Warmup- Warm- Co1d- Warmup-
No. Warm Warm Warmup Warm Warm Warmup Warm Warm
\
3 +70 -5 +25 +60 0 0 +60 -10
Acknowledgment
-374-
G-3
Introduction
where
PL = density of liquid
g = gravitational acceleration
Description of Apparatus
Figure 1 is a sketch of the test section used in the
heat transfer apparatus. To simulate the de sign of the magnet.
a 7 inch stainless steel plug was accurately centered in a 5
inch stainless steel tube.
This stainless steel tubing was used as the heater; its
other dimensions were 1/2 inch O. D. and 0.010 inch wall thick-
ness. The tubing was used as a resistance to current forced
through it by 12 six volt batteries connected in parallel. The
heated part of the tubing is between the bras s end pieces which
served as connections for the battery leads and also as
supports for the Styrofoam insulation on the outside of the
heater. The insulation made certain that essentially all of the
heat was transferred to the liquid in the annular gap.
-376-
.... ......
TO"
PlO P1(Cl
TO
I/OU -
WT"
~_JJ
IOn...
COfTV."..
PIN
_AT
TO "TTon'
TMItU
Figure 2. Water-
cooled rheostat
Figure 1.
-377-
A water cooled rheostat (Figure 2) was used to control
the voltage to the heater. The rheostat is made of stainless
steel tubing, one pair of 0.020 inch wall thickness (coarse
adjustment) and the other pair 0.006 inch wall (fine adjustment).
The two rectangular pieces at the bottom of the tubes contained
woven copper wire between two pieces of aluminum bolted
together. As these pieces were moved upward the copper wire
shorted across the two stainless tubes, thereby increasing the
power input to the heater.
Discussion of Experiments
-378-
Natural Convection. The nearest agreement between
previous work and the experiments here was obtained when the
test section was immersed in a pool of nitrogen or hydrogen. In
these tests, called natural convection (no forced flow through
the heater), the power input was increased until the temperature
of the heater began to rise. In most cases the AT could not be
held stable at lOoK; it would instead rise sharply until the power
was lessened. The film boiling phenomena observed by others 2 , 3
seems to explain the sudden rise in temperature. Since there
was not a primary interest in film bOiling, no data were taken in
the film boiling region.
Results
-379-
by this equation it was nece ssary to correlate f (the vapor
fraction) and Z (the hydrostatic head equivalent) to the experi-
mental data. In the derivation of the equation Sydoriak made
the as sumption that all the heat input was used to vaporize
liquid; hence, f was calculated from the equation
Q
or f = (2b)
where V is the volume flow rate, ccl sec, and Q/PL >... is the
Q ~p 1/2
observed
Qcalculated = C(------) ( 3)
V
always true.
-380-
88
80
H2 10 MIL GAP
72
64
W
I-
C2 40
FIGURE 5
3:
0
-'
u..
32
LOS ALAMOS
2' H2 22 MIL GAP\ .•...••... ·
.......................
N2 10 MIL GAP
16
~~0----~10-0----1-4-0----,-80-----22-0----2-W-----~----~3-W-----38-0-
POWER INPUT, WATTS
Figure 5.
FIGURE 6
'"
l-
I-
<t
3:
10'
,.:
::J
"-
;;:
a:
w
3:
0
"-
,~L,--------------------~-----,O--------------------------~,'O
GAP CROSS SECTIONAL AREA, CM 2
Figure 6,
-381-
There is no obvious explanation for this discrepancy since the
measurements have been repeated and the instrumentation
seems reliable.
-382-
Appendix A
Q in watts, V in cc/sec
~ p in inches of Hg
2
A (20 mil gap) = 0.1885 cm
1/2
Q calculated = 162 A ( (~p) (Qobserved)
V
2
Q in watts, A in cm , ~p In inches Hg, V in cc/sec
-383-
TABLE I
-384-
TABLE II
Q obs
Gap ~p V Qcal Qobs/Qcal
inches Hg cc/sec watts
-385-
Appendix B
1/2
V (p )
A B
(1)
TRANSFER LINE,
114'00 ""'
t FLOW
.....-VACUUM
/ JACKET
//~:~~UDRE
/'" TAP LINES
188"10-
~.~ I~I~
5rtARP-
105"
",
EDGED
ORIFICE
PLATE
,,, :,, "
,,
SCREEN~'
FILTER I I
STORAGE
DEWAR --
+8~<tSH : I
, :
l .......L..l
~
'0
,; ~
~
~
// ~
~
N, ,
~ '.06.0r--··
~ 1---- , :
g'OV I ;
1
'0 ....t::
~t::-, '
,o~ilil
01 Ql 04 Q6 011 10
AP ACROSS ORIFICE, Inc;I'I" of HZO
Figure 9.
-387-
Our orifice was first calibrated by measuring the
time required for a given weight of water to transfer through
the line with a measured pressure drop across the orifice.
Figure 9 shows the water line drawn through points ranging
from 6.pl s of 1. 5 inche s water to 80 inches water. These
points are omitted to simplify the graph.
-388-
Discussion
Answer by author: The inlets in the two tests were the same;
they were so small in comparison with the space immediately
before the plug that we were sure of no irregularities due to
entrance effects. The plug also extended one inch past the
test section to assure uniform entrance.
-389-
G-4
Introduction
Friction, wear, and surface-failure properties of
materials sliding together in cryogenic liquid are vitally
important to the development of efficient pumps. Bearings and
shaft seals in particular have critical sliding surfaces and are
used in all types of pumps. Positive displacement pumps may
include additional parts such as vanes, pistons, or gears with
surfaces in sliding contact.
-390-
metals in sliding contact. The literature contains numerous
instances where formation of oxide films during sliding pre-
vented surface welding of mutually soluble metal couples.
Many cryogenic liquids, including nitrogen and hydrogen,
prevent the reformation or repair of surface oxide s, once
these oxide films are broken or disrupted. The probability of
metallic welding could therefore be greater when the original
oxide layer is worn away during sliding. This suggests that
the selection of slider materials in cryogenic engineering can
be more critical than for applications in room atmosphere.
Nonsoluble metal couples and couples including a nonmetallic
material might be of particular interest for cryogenic sliding
surfaces.
-391-
HELIUM SUPPl Y
Figure 3.
Figure 4A.
-392-
housing. The shaft was sealed at the top and bottom of the
gimbal mounting by flexible metallic bellows and, in operation,
the enclosed volume was pressurized with helium. Friction
force was measured by a strain-gage dynamometer ring
assembly. The deflection necessary for friction measurement
was so small «0.003 inch) that the restraining spring action
of the bellows was negligible. The spring propertie s of the
bellows were important, however, where extreme wear of
rider specimens neces sitated considerable (>0.040 inch)
displacement of the arm toward the center of the rotating speci-
men. This deflection of the bellows reduced the true load on
the wearing surface. Deflection of the top of the arm in the
direction of loading was measured with a fixed micrometer.
The runs were terminated when excessive wear caused unload-
ing of the spe cimens. Previous calibration made it pos sible
to determine true load with any amount of deflection. Load
was applied by a pulley and dead-weight system.
-393-
static seals used were made of either Teflon or Kel-F. No
trouble except cold flow has been experienced with Teflon and
Kel-F static seals. Two concentric Kel-F O-rings were used
to seal between the lower and the upper housings when they
were bolted together before each experiment. These rings
were reused 45 times or more before replacement was neces-
sary. The experimental materials used in these studies were
selected for various reasons that will be mentioned as the results
are presented. Specimen finishing was varied somewhat in the
experiments because of the great difference in materials.
-394-
displacement method. The specimens were carefully studied
and photographed at low magnification after each run. Unles s
noted, all photographs were made using oblique illumination.
Results
-395-
52100 ~IDE~ S210D DIS
Figure 4B.
A9 RIDER
l04 DI~K
Figure SA.
Figure SB.
Figure 6A.
-396-
Surface welding and transfer with the metal couples
previously mentioned indicate that the low temperature and any
boundary lubricating influence the liquid nitrogen might have
are not sufficient to make the use of these combinations
practicable under the conditions reported herein. It was sug-
gested earlier that with a surface oxide film worn away, the
mutual solubility of metal couples should be of increased
importance to the occurrence of surface welding. Silver has
little or no solubility in the primary alloy constituents of type
304 stainless steel. Therefore, runs were made with cast
electrolytic silver riders sliding on type 304 stainless steel.
The results of sliding these metals together are given in
Figure SA. Severe plastic flow of the silver was indicated
by the extensive feather edge. This detritus was separated
from the rider specimen as a result of handling after the run.
The initial friction coefficient was 0.80 and the wear was
0.63 x 10- 3 cubic inch per hour.
-397-
Observation of the plastic behavior (ductility) of the
various metal combinations during sliding in liquid nitrogen
indicates that materials normally considered brittle at room
temperature might continue to be useful in cryogenic appli-
cations. In unreported NACA experiments a titanium carbide
base cermet has been found to have low wear and to resist
surface failure when run against tool steel with temperatures
from 80 0 up to 700 0 F. Experiments were performed with the
cermet (K162B) sliding against hardened 440-C stainless steel.
High hardness (54 Rockwell C) of the mating disk was utilized
in order to obtain the minimum real area of contact and
sustain high contact stresses which should provide a continuing
severe test of a brittle material. The data shown in Figure
6A indicate that the test was too severe with the 1000 gram
normal load.
-398-
Figure 6B.
~H(NOL IHPQEGNA 1 EO
CARBON qlOfR
Figure 7A,
PHE~OlIHPRtGNATED
CAQaON RIDER
Figure 7B.
Figure 8.
-399-
Runs made with a phenol impregnated carbon (Purebon P5J)
sliding on type 304 stainless steel using an open dewar gave
very low wear (0.009 x 10- 3 cu inch/hr) as indicated in Figure
7 A. There was no surface disturbance on the disk and only
slight evidence of solid contact. When the same material
combination was run with the closed system, entirely different
results were observed (Figure 7B). The runs gave extremely
high and erratic wear values in seven duplicate runs from
34 x 10- 3 to 91 x 10- 3 cubic inch per hour. Frozen adsorbed
moisture films in the case of the open dewar is the apparent
explanation for the difference in wear rate.
-400-
influence wear and friction propertie s. The lowes t initial
friction obtained with a cotton-fabric -laminated phenolic
sliding against type 304 stainless steel in liquid nitrogen was
0.49. The addition of graphite and molybdenum disulfide
neither reduced wear nor lowered friction. For example,
Formica LN gave a wear rate of 0.11 x 10- 3 cubic inch per
hour, and the addition of graphite or MoS 2 resulted in wear
rates from 0.28 x 10- 3 to 0.42 x 10- 3 cubic inch per hour.
The additives also had an adverse effect on friction, resulting
in initial friction coefficients from 0.81 to 1. 40.
-401-
1 [fLON ~IOt R lOC DIS
Figure 9A.
Figure 9B.
.... .
"e.
o
..
1.0
::! ~
u
z
.." .,..o
.8
~ 0
~
~
~ ~ .6
v
.. . •
u
ii ~
e 0
u
...
~
c .2
~
~~ ...,
z
~
"'- II c
" 0
HETAlS NON-HET ' ~ S ON 304
-402-
were always very smooth, usually highly reflective to light,
and showed no evidence of any adverse sliding conditions. The
highly reflective character of the wear surfaces made it
pos sible to detect elastic recovery that was at a maximum in
the center of the contact area. There was no opportunity for
a quantitative evaluation of this elastic behavior. The disk
specimens had very smooth highly polished surfaces in most
cases. There were, however, thin uniform transfer films of
compositions visible on the disks after operation. Etching of
the disk surfaces that had been run with unfilled Teflon showed
that the transfer film on the disk was highly protective to the
base metal.
Conclusions
-403-
in cryogenic liquids. Failure of Teflon in slider applications
has previously been associated with thermal degradation
accentuated by poor thermal conductivity. Cooling by the
liquid nitrogen probably played a vital role in the good results
obtained with Teflon compositions.
Discussion
-404-
Question by J. M. Novajosky, Allison Division, General
Motors Corporation: Why did you choose to use a round
nosed rider rather than the more easily fabricated flat rider?
-405-
Question by R. Gilman, Precision Equipment and Company,
Inc.: Do you have any comparative data for heat treated
Be-Cu on stainless?
-406-
G-5
J. G. Althouse
Lukens Steel Company
Coatesville, Pennsylvania
-407-
G-6
Introduction
-408-
SHIELD HEAT
RING SINK
PLATINUM REFERENCE
RESISTANCE JUNCTIQH
TlIERMOIIIETER POST
VACUUII
SAMPLE CONTAINER
THERIID·
COUPLE
THERIIAL HOlDERS
SHIELD
SHIELD
HEATER
GLASS
DEWAR
SAMPLE
HEATER AND
SHIELD
DEWAR BOIL·OfF
SUPPORT HEATER
Figure 2.
Figure 1.
20 2. 0
15
<,y
5 .,.. . .•.
..
.... ."
~ ",. '
0:':
E
10
- G , 0:': I 0
7" ~
, '
'
8.0 E ·8 7 "7
~ o ~
~ 6.0
:;- '- 6
#Y
-
\ !? 1.1. ~ ,/
•3
VI
:: 4 .0 "0
/ ~ -~ / 2 I' 51 40
~ 3.0
,~~ ~ ~ ~ ~~ .. .,- ~
>-- I ~, ~v
>- 2.0
-Or
~
:;: 2 II' ..
~~ V ,~~
~ 110
:;: !-' .•. - '-'fo··
0: ~ V I--' 0:
o
g 1.0
v.; .....
::>
o .1 " II'
o ~
r
.8 ~ .0 THERMAL
o~
/. / I THERMAL VII' l CONOUCTIVITY
--
.~
.6 CONOUCTIVITY 0 .0 6
#/ I ALUMINUMS
.0 4~
1/
II' I ALUMINUM ALLOYS
.3 / I I 3
I
.6810 20 4060 100 200 JOO .0 4 6 8 10 20 40 60 100 200 JOO
TEMPERATURE, 'K TEMPERATURE, 'K
Figure 3, Figure 4.
-409-
The electrical resistivity is determined from the
dimensions and resistance of the sample. To measure the
resistance, a current of 50 to 60 milliamperes is passed
through the sample and the potential drop between the bottom
and top thermocouples is measured on a microvolt potentiome-
ter.
-410-
was lower than that of the "as fabricated" specimen.
-411-
6
4 ,,-.-
. . .,,- .
.
,,- -.-
- -
, .'"
20( 4-T4
L
*:~
515 q.-O ~
E 2
50~ 2-0
. .--
0 ./"I .
I ELECTRICAL ,
/,'/
/ .
E 10- 6 t--- RESISTIVITY // .'
.s=-
O -- /. I
I .
. I
.. 6 r---
ALUMINUMS ~
>-
t-
>
4 W. I
.,- ~
~
6063 -T5 I
- V
r--
t-
en V I
~ J
en 2
w 1100 0
Q:
.-J 10-7 J
<l:
u
-
Q:
6
V
t-
U
w 4
V
.-J
W \,Ii "T~IA r- ..-'
/
,....
10 50 100 300
TEMPERATURE, oK
Figure 5.
-412-
TABLE I
*L--Longitudinal
T --Transverse
-413-
TABLE II
-414-
Composition, Grain Size, and Hardness
Summary
Discussion
-415-
G-7
-417-
Huntley, S. C. F-6 342
Isakson, V. E. E-2 232
Jacobs, R. B. D-6 209
G-3 375
Johnson, R. L. G-4 390
Johnson, V. J. A-2 11
Kayan, C. F. F-5 325
Macinko, J. A-I 1
Mann, D. B. C-3 125
Martin, K. B. D-6 209
McClintock, R. M. F-3 305
McIntosh, G. E. B-1 64
McReynolds, W. A. G-7 416
Meyers, N. D-l 172
Mikesell, R. P. F-4 316
Pastuhov, A. G-2 367
Powell, R. L. E-6 269
Purcell, J. R. D-4 191
Reed, R. P. F-4 316
Richards, R. J. G-3 375
Rind, H. D-l 172
Robbins, R. F. G-3 375
Robinson, G. Y. C-l 106
Roder, H. M. G-6 408
Schaffers, T. W. B-4 92
Scharle, W. J. C-5 149
Schrodt, J. E. G-7 416
Scofield. H. M. A-5 47
Seavey, D. W. D-7 217
Selldorf, J. T. C-l 106
Smith, H. J. D-3 179
Smith, K. C. A-4 32
Spero. R. P. B-1 64
Timmerhaus, K. D. A-6 58
Tracey, J. E. E-l 226
Vander Arend, P. C. C-3 125
Van Valin, C. C. B-2 73
Wang, D. 1. J. F-2 294
Weitzel, D. H. A-6
B-2
. 58
73
B-3 85
D-4 191
Wilson, W. A. B-3 85
Wisander, D. W. G-4 390
-418-