ML19344F209
| ML19344F209 | |
| Person / Time | |
|---|---|
| Site: | 07000258 |
| Issue date: | 12/31/1978 |
| From: | Boody F ILLINOIS, UNIV. OF, URBANA, IL |
| To: | |
| Shared Package | |
| ML19344F208 | List: |
| References | |
| 03387, NUDOCS 8009120614 | |
| Download: ML19344F209 (27) | |
Text
.
hyp A -
A UF HANDL1M SYSIW 6
FOR l
140 CLEAR-PUNPLO LASER EXPERIMENTS F*
4 Frederick P. Boody j
6
)
Fusion Studies Laboratory University of Illinois December 1978 4,
- 4 M
4
~
seesn*65
,l.
IntrodJction A.
Background:
Considerahli' interest exists in pumpimi nui. lear-pumped losers (NI'Ls)
U distributed throughout with the energetic f ragments produced by fission of i
on gas lasers is vapor [1].
The quenching ef fect of UF6 the later gas as Uf6 unknown but is expected to be considerable. L'e are preparing to perform the
.~ first controlled nuclear pumping experiments with UF6 vapor contained in the lasercell[2].
On a hardware level, the gas fill system currently used for NPL experiments is dt the Nuclear Peactor Lab must be evacuated af ter each pulse because it Filling the long, relatively karge diameter (for also the vacuum line.
reasonable conductance) vacuum line each pulse is very wasteful of the expensive, ultra-pure rare gases (such as Xenon) that are used for these 2
experiments.
S.
Objectives:
The raain objective was to design a UF5 han.11ina syster.:
i that could be used in conjunction with the existing NPL Group vacuum system i d at the Nuclear Reactor Laboratory to perform the experiments descr be A secondary objective was to separate the gas fill system from the.
above.
vacuum system and thus greatly reduce its volume.
C.
Design Considerations:
As can be seen, UF sa I
Figure 1 is a phase diagraci for UF.
6 6
solid at room temperature. Additionally UF contains copious amounts of F 6
Finally, U (esp. 235 ) possession and usage is 0
which is very corrosive.
i regulated by the NRC.
For these reasons UF6 presents special problems
(
l for the gas laser experimentalist.
1G 2.D r
o 4
\\
Centrol No.O 3 >
a Our usual gas handling method involves connecting high pressure bottles of the gases of interest to a manifold, through copper lines, and then through another copper, brau, or 55 line to the laser, which is also connected to a l
vacuum system. The laser, manifold, and copper lines are pumped down to the residual gas pressure desired, then the desired amounts of laser and buffer gases are added. This method is not satisfactory for UF, obviously, since 6
it is a solid at pressures above a couple hundred torr (at room temperature) and very corrosive.
Mercury and other metal vapor lasers face a somewhat similar problem.
However, there are major differences. fluclear-pumped' lasers at the TRIGA reactor are >.4 m from any possible location for an oven vapor source and the vapor der.sities required are much greater,10 - 100 torr vs. m torr.
For Hg lasers, the oven problem was solved by putting small amounts.of Hg in reservoirs within the laser envelope and the laser, which must be heated anyway, became the uven. This is unsatisfactory for UF because the excess 6
Uf w uld still react with neutrons if enriched. Also, the UF c uld be 6
6 cracked, producing UF,(x<6) solids, or react with the fluorine bearing laser in the laser. Only the precise amount gases, contaminating all of the UF6 uf UF required as vapor may be present in the la er at any given time.
6 Therefore, in addition to heating and temperature contial of the laser cell, vapor over a distance of several a method is required to transport the UF6 vap r source to the laser cell.
reters from the Uf6 transport relies on A method that has been used successfully fur UF6 the vapor pressure-temperature relationship shown in Figure 1.
By imposing a temperature differential one also imposes a pressure different141 causing vapor to be transported from the region of higher temperature and the UF6 i
thus pressure to the region of lower temperature.[3] Such a transport system is practical only for a material which'goes from solid,tg vapor (and back) s
,-.-n.-.,.,,,
r
10 0 s
I
~
gm O.2 9 j
LIQUIDt__-.m0.t l0 SOLID l
- 0.05 l
lE TRIPLE POINT j
- 0.02
- ~ ~ ~ ~
H I5 ATM
[ -
- 0.01 4
337.4 'K go
- -... -./
na tr
- 0.005 3
am y
-\\,
l
- 0.002 m
n.
\\ CORRIDOR OF-OPERATING tr O.1 O
GAS CONDITIONS Q-Mc-3 5 10 14 KG OF U-235 0.01 0.00l 200 300 400 500 600 TEMPERATURE, *K Figtre 1 State Diagram for UF6 O.
4 b
..,4
4 with the attendent 1.trge changes in specific voimiei,
lhus the laser gases must be transported by the usual methods:
fillint; the laser cell from a high pressure reservoir and evacuating the laser cell with a vacuum pump to the 6 ylinder, rather
.ttmosphere. Since the UF is to be returned to a spent UF c
6 than vented to the atipsphere, and because it would most likely badly mess up must be separated from the laser gases.
the vacuum system, the UF6 Table I con.7 ares the boilina or st.blir.iination teciperatures for the and CU.
It can be seen that a dry laser gases of interest with that of UF6 2
rom XeF* laser gases.
(But not from ice trap can be used to separate UF6 any byproduct HF that is formed. This can be renoved in a separate cperation at room temperature.)
Table i Boiling or Sublimation Temperatures of Interest a t 1 Atm.
i Subs tance Temp (K)
N 77 g
Ar 84 F
85 2
NF I44 3
166 Xe 195 CO2
!! I ~
793
-~
329 UF6 l
(
l
,e 4
g.
5
?
Fi;ure 2 is a block diapron of the.., er:d UF t
6 handlJng systeus. Tlio arrows indicate flow direction.
A typical experiment would be run as follows (a detailet operating procedure is given Ln section three):
(1) the laser cell is evacuated to the desired residual gas pressure 1
,j (determined bf, allowed impurity concentration, (2) the laser is filled with the desired mixture of laser gases and UF '
6 (3) the reactor is pulsed and data to en,
'j (4) the "used" UF is returned to the Uf 6
6 s'spply system and the contaminatt (with chemical reaction products) laser gas exhausted.
ts (5)
If it is desired to change the laser gas mixtu'rb for the next the gas fill line and manifold are evacuated.
g t
s a
l l
I 1
l l
1C S.'.'.D 4
Centml tic.0. 2 o.
6 FSL-78-285 l(MODIFICATION) lI r
(NEW~; DESIGNED l GAS FILL FOR COMPATI SYSTEM 4" DIFFUSION)
LASER l
l VACUUM CELL l
l SYSTEM BILITY WITH l
UFsSUPPLY l
(EXISTING)
UFa SUPPLY l
SYSTEM l
SYSTEM)
(NEW) l Figure 2 Illock Diagram of Apparatus for UF -NPL Experiments 6
e.
e.
4
. fl.
System Design
The mudified The re-design cl the gas fill system w.v. h.c.ically trivial.
4 system design is shown in Figure 3.
The nodification required is the addition of the components to the lef t of the dotted line. A new valve will be connected s
to the existing gas fill manifold.
From this va1ve the laser gasses will pass 1
SS tube to through a dry ice trap and a long, small diameter (1/8" - 1/4" 00) g b
the laser cell.
The UF Supply System design problem was somewhat more demanding.
It was 6
The first part was the design of the vacuum envelope, divided into two parts.
which assumes the presence of a temperature control system; the second part was the design of the temperature control system.
II.A.
Vacuum Envelope supply system.
Figure 4 is a schematic of the vacuum envelope of the UF6 In this section, the details of the vacuum en.'elupe will be ' considered.
In the next scction we will consider temperature control, shown cross-hatched in i
Figure 4.
taser Cell The most critical part of the vacuum envelope design is the laser cell where the actual experiment takes place and which is exposed to nuclear radiation.
The internal details of the laser cell are set !. 'aser physics considerations One of and will be discussed here only as they effect the vacuum envelope.
these considerations is that, in addition tn being compatible with fluorine compounds (NF. UF ) and ne6 tron radiation, to permit the inclusion of electrical 3
6 Teflon (TFE) and poly-pumping the laser cell materials must be insu?ators.
As designed vinyl chloride (PVC) were identified as suitable materials.
the cell uses PVC for tubing and TFE for parts machined from rod because "Torrseal" low vapor pressure they are most available in these forms.
Fused silica, wnich naintains epoxy is used-for attaching the PVC to the TFE.
/
4 O
n.
~.
FSL-78-290 PRESSURE b VACUUM FILL LINE SYSTENI TO ASER
} I J
DRY i 1-
ICE TRAP V
V NF Xe 3
Figure 3 Gas Fill System S
9 9
9 5
4 4
g 9
G w.
--,-,,e
.n_,..-
,----.,n--,m,,--,.ey ea--g
.e~-w,-new.,,w--.+,--v g--
l fst-78-2a-i r~ N F.lLL LINE x
,+
O I
SYSTEM AWNWNWWNN NWNN SOLENOID VALVES
/
N LASER CELL x ' ' s W.WWNKN NNWWNWW] }
/
TRANSFER LINE
)---.-
\\ b\\\\\\\\N INSUL ATION AND :
RLUS TO RESERVOIR HOT N2 HEATER TAPES PRESSURE
\\
GAUGE N\\\\\\\\NNN\\N VACUUM x {E-( ___
-*--) TR ANSFER LINE hh\\\\\\N N\\ M N hNN-
-x x SYSTEM TO LASER Q k k Q i,' h j,
xgq SUq s
NM NM
['-
i DRY NEW USE LN2
' COILS
'O g A UFs e
UFs SUPPLY
'E RETURN CYLINDER
/
CYLINDER Figure 4 UF Handling system schematic 6
i in its >991 transmisivity under irradtion (because of its high purity) is used for windows, and viton is utilized for the gaskets.
figure
- 5. shows various views ni the laser cell. A prototype cell has been constructed using acrylic in place of TFE.
~
Iransfer Line For a very long Vacuum line (such as is required at the TRIGA) the ultinate pressure obtained at the end of the vacuum line (i.e. the laser cell) is f
l i dependent on the inverse of the cube of the diameter of the vacuum line j d t',J t
dnd independent of the pump capabity (unless the pumping speed is very low).[4]
Thus it is desirable to make the effective diamet. of the vacuum line as large
~
as possible. However, limited space in the throughpurt n.kes a very large diameter line.into the laser cell impossible. The diameter of the throughport is 6".
A 1" 10 cylindrical volume,. concentric with the centerline of the throughport, is reserved for the output from the laser cell. Thus a 21/2" thick annular cross-sectional area is available for the laser cell envelope $)d the hardware handling systems.
for connecting the laser cell to the gas fill, vacuum, and UF6 As shown in Figure 6, the largest diameter circular cross-section transfer line that can be accomodated is 1"-l 1/4." Any geometry other than circular would probably be too expensive to fabricate and would not have a much more favorable surface area to volume ratio. The vacuum line currently used is 1" in dia.
and residual gas concentrations have been found to be satisfactory.
The transfer line vacuum &nvelope will be made of SS tube because of 1.ts availability (relative to Aluminun and lionel).
ee 4
r c~
1-,.
-m..
h
/
v jj m
g)
O O
03 1.
um e-1
.J r--
e o
o
. i _ _ _ _ P rm D
=
e mm
\\
O O
s
{ \\
O a
\\
.b M
4
.V z
s
- o Q-
\\
Eg w sa
\\=~
.U. W mC G
s bW
\\
- k f
j U e s
'\\
ea b
oa CA
(-
g-
.=
U L 3 0 35 %=
m G
L 3
Q b
.s l
~
e
~
w I
~
i
_.r-.
s 1
se d
J O
'h l\\.*
,/x.s
/
/x OG.)
O C( V
~
m s,
I i
i Y
(
l Figure 6 Diagram of Laser Carriage for TRIGA Reactor Thru-port with Maximum Diancter Transfer Lines Indicated (Full Scale) 4 g,
-,--e.,-,-
-, ~ -, - -
,,--,-av-,,-,-,
. Ui SUPPLY g
As can be seen from figure 1, any reasnn.ib' g <lmirr:d can be achieved to the laser icit.
T'.ii, is not true fnr (by heating) for transporting Uf6 contained in the the reverse process. Also only a small fraction of the UF6 supply cylinder must be transported to the laser cell. However, we wish to fr m the cell'~to the return cylinder.It return essentially all of the UF6 w
wd r what size to make the f
- It is not obvious hat cryopumping the UF 6 system.
Reference 3 used four 1-1 monel cylinders. The only "cryopumping rate" i" "'
that they report is the return of 20(+.5)o of 25(f.5)g of UF6 hour. However, their " canister" is 851 in volume vs.1/2 L for our laser anticipated to be in cell (-21/2 L for the transfer line) and the most UF6 the laser cell at one time is -l g.
Thus, our pumpdown rate should be somcwhat better.
The maximum cryopumping speed can.be estimated by assuming that all in the return cylinder are trapped.
.)
UF molecules impinging on the cold UF6 6
Theimpingementrateisgivenby[4]
22 y = - nv' = 3. 5 x 10
())
The inside diameter of the cylinder can be taken as 6 cm. Thus, the pumping
' peed of the cylinder can be cakulated from[4]
s,,,. g. > ^ u. 3.6 g A.
(2)
Substituting the M value for UF, 350 X, and the area corresponding to a 6 6
Cm diameter gives l h 4
a
- e 5,
IN t/s.
lhis is a pumping speed c.omparable to a small dif fusic: purrip.
density (or partial Estimation of the pumpdown time to a given UF6 pressure) is more difficult because it requires knowledge of the surface properties of the various materials involved for UF. However, several
~
6 simplified approxiriations can be inade.
The crudest approximation is to neglect surface effects (and the finite conductance of the transfer line) altogether. Thus we are essentially 2
considering a 3t volume with an -30 cyropumping area.
From reference 4, dil 8kT UF VF
(
~ ^cyl
- "Y^
- ~
OH V
dt UF SYS **
6 or dI30F kN A
T 6,,
T__
cyl dt = - 36.4 gg----
d t.
(4)
SYSt**
6 UF 6
6 F(r T = 350K,
l UF dN l
6 = - 36.4 dt NUF6 so that 1
f'gt% %
~ ~. s* e e u!
s
"UF s
_39 y,
(5) u 'rTOT""
u 6 far - = 10~0, t is about.5 sec.
N O
s While this result is 'sery crude, pumpdow times more than three orders of magnitude greater would be acceptable.
Another crude approximation is to treat the system as a normal vacuum system made of materials with high outgassing rates
- S I aM molecular Taking Seff::
th max.' " N # '
S
- S e
flow regime, and neglecting p lt.; the pumpdown time is given by [ 4 ]
u Q
9-p (6)
In t = 5,3 l
Q g
pg-S mar.
It has previously been determined that V is -3 t. and 5,,, is -100 t/s.
can be approximated by taking a high value of 10-5 g
O = q *A and q g
g system g
2 (torr-1 iter)/(S-cm )[4].
Then
~2 L "~ " l"#
10-5 torr-liter 4 x 10 4 x 10 cm =
g 2
sec 9
sec-cm This says that the lowest attainable pressure is torr-liter
= 4 x 10-4 torr.
P
)
liter /sec
=
=
jo,,3g An infinite amount of time would be required to achieve this pressure. Using equation (6), the time to pump down from 100 torr to 5 x 10-4 torr would be e
4
9' 3L 100 4,,,,
in
'4 t = 100 t/s
- r. x 10~4 4 x 10
~
Again it appears that puisipdown times will be very titisfactory. And, wlille a UF pressure of b x 10-4 torr is not ideal, it represents about 10 99 for our 6
system volume -- a negligible amount, even for flPLs.or the NRC.
The result of these crude calculations and the comparison to reference 3
~
J cylinder wall is that it appears that cryopumping using a single 1-1 UF6 satisfy the needs of the experiment quite well.
W
- e l
e f
e e
e 54 h
4 e.
e r
___..,...,.-_,,_m
,_-..--_,m_
m_..,.
,..,_..,.__.m.
II 11.6.
Temperature Control Since the UF transfer system utilizes a wpor pressure puuping scheme 6
(the return portion of the cycle being the well known cryopumping process),
s accurate control of temperature is important. Additionally, uniformity of temperature is important since having the temperature drop below the
~
subliminating temperature at any point would result in the deposition of UF6 t
at that point. This is of special concern for the sealing surface of the valves.
He have desi nce a tenparature control systen that attempts to provide
?
maximum performance at minimum cost.
Factors included in performance include accuracy of temperature control at a point, uniformity of temperatures,
~
flexibility, and minimum of manual intervention. The temperature control system can be divided into two basic subgroups:
(1) Heating and cooling (2) Temperature sensing and control electronics 1.
Heating and Cooling Four different types of hardware must be heated. These are:
(a) the laser cell, (b) tiie transfer line, (c) the Uf cylinders, and (d) valves.
6 (a) 1.aser Cell Heating and Insulation
(
The laser cell, Figure 5, it is not very e.ont-ive to tha' ur.e of heater tapes since they would have to be applied to the outside of the capacitor. The capacitor and the " air" gap between the capacitor and ti'e laser cell provide a considerable thermal resistance. As will be explained.in the tenperature sensing section, thermistors are the. desired method of temperature sensing. However, the lifetime of thermistors in the severe radiation environment of the reactor is unknown and probably verj l
snort.
Both of these problems can be solved by'the use of flowing hot nitrogen l
e' 4
l
nit.roqen can lill the 4.ip I"tumn the capacitur anti th<-
beating.(C']
The bril laser cell and by flowing it, temperature tensin...o be rpoved outside the Nitrogen is chosen over air because it is put coctive with UF 0'
6 reactor.
pressure higher than the pressure any of the laser g'ases. By maintaining an N2 int he laser.
, in the laser, any leakage will be of N2 used for tne (b) The trans.fer line could be heated by either the hot N2 laser cell or by heater tapes as with the valves. Both methods have advantages.
lleater tapes are very easy to install, one simply wraps them around the transfer lines, and cccupy a minimum volume (if no further insulation is used),
which can enable the maximization of the cross-sectional area of the tra On the other hand, with line (the importance of which was discussed earlier).
hCater tapes, one controls the heating rate rather than the temperature so However since one end of the one must accurately measure the temperature.
transfer line is in a high radiation environment thermistors cannot be used.
Thermocouples are much less accurate. Heating and temperature measurement would probably be required in a minimum of three regions, the middle and the two ends, making the contral problem slightly more complex than it could be.
Similarly, six pairs of wires would be required.
is already required for the laser cell, no additional Since hot il2 hardware is required for its use with the transfer line other than the co-axial envelope around the transfer line required to contain the tenperature can be The two maintained more accurately and uniformly than with heater tapes.
disadvantages are 1) that any co-axial flow design will be bulkier than heater tapes witnout insulation and 2) a 4-5 r: co-axial flow line will be relatively com;. lex and expensive to build.
4
in sunwary, a heater tape system is simpler tai cunstruct but inore j
complicated to operate and allows a greater condus.tonce (and thus vacuum system) perfo:T.iance) but gives poorer teuiperature control performance than a co-axial flow system.
Figure 7 depicts the hot N approach which we have chosen.
2 (c) The UF cylinders require both heating and cooling:
heating for 6
vaporizing the UF and generating the desired pressure to transport it -5 m 6
to the laser cell and cossling, with liquid nitrogen., to cryopump the UF6 back.
In fact, for the cryopumping phase, heating and cooling will be required simultaneuusly - the inlet to the cylinder will be heated to prevent plugging.
At. shown in Figures.i and 7, copper cooline coils will be sol *Jered to the bottom half of the monel UF cylinders.
For cooling, LH2 "III D' 6
run through the coppe' coils. Heating will be provided by a. combination of beater tapes and I,ot N. The cylinders and valves will be wrapped with 2
heater tapes, in two regions, and inserted in an airtight, insulated enclosure connected to the hot N flow system. Any l!F leaks will be into this 2
6 enclosure.
(d) Five va'ves requi;e heatinn.
These tre the two U!'6 cylinder val m,
the laser cell to transfer line valves, the transfer line to transfer line
~~
l trap valve, and the transferline trap to vacuum system valve. All of these will be heated usion heater tapes. The heater taacs will be wrapped around the valves and fibergl,a,ss insulation will be wrapped around the heater tapes.
I i
l 4
l 7
Il.
i 7 q}
b g
Q, s
ne
- 5
", f, I
))-
R
'}
i E
.I gg
- O e
L
'.l, j
B
,'i{
'[ f me t
[".
s y
S J
dD E
g T
q n
S R
H M
Y S
E V
H O
d T
T n
S a
E A*
w
- c t
E 6
av F.
p H
F T
U e
t L
r x
TX s
t E
o c
V f
?
t x
E
+
=
s L
m a
v S
e t
Uv t
C T
s E
c r
o e
- U L
y L
E O
S w
R F
)
e.
,L T
R l
t O
T e
C o
rt3, ^
t d
T E
r o
S s
eaM'w s
L a
t o
t E
w n
c R
o o
L E
C t
tr M_
a O
H e
R T
e T
r N
L u
i p
O A
t r
C U
a D
r gY T
p
(
w e
i" L
w N
m e
l E
e t
t t
A T
s L
t i
s E
E s
c R
t U
7
,v t.
,o l' o S
n n
p A
e T
E r
r 1
=
C o
M u
s E
c g
a L
S E i
_r '
c E
o E R F
t o
P U t
A T
T F.h/y go A
t R
s R
E s
E P.
tni T
a a A
P w
E E
t L
o H T R
c s
L r
t l '
1
\\c i;I T
t a
L T '
Hf l
I
)
s 7.
Temperature Sensing and Control Elec.tronic <.
The temperature of greatest interest is 18.it iis l'ie 1,aser cell.
This will i
be measured using ene or buth of two methods.
lhe preferable method depends on the effectiveness of a shielding material called flex boron which has recently come to our attention (the NPL Groupl but which we have yet to test 5
in the reactor.
If effective, we will be able to use it to shield a number of i
neutron radiation sensitive components (the unkncwn is the fraction of current damage caused by y's), including thermistors, one o,f which will be located in a shielded portion of the laser cell, in either case, the inlet and outlet temperatures will be measured using thermistors and the cell temperature N7 temperatures.
Similar thermistors will be used to inferred from these N2 supply encic,sure.
measure the temperature in the UF6 The electronics for the thermistor temperature measuring system are shown in block diagram f ora in Figure P..
Such systems are highly accurate even though relatively ineypensive. A similar system is used by the Co.abustion Engineering Boronor.ieterD (desir;ned by the author). !fith a 10 TJiz V/F and a 1 second counting time, resolution was.01 C and error was <.1 C.
For this system an
.1 second counting period will used producing a resolution of.1"C and an error of
.l C.
The thermistor that will be used is a unique U
dual thermistor (manufactured by Yellow Springs Inst.) that has an output that, changes linearly with temperature. A monolithic instrument amp (such as the Analog Devices 521) raise,s,the voltage level of the thermistor output so that the m4ximum design temperature equals the full scale voltage of the voltage to frequency convertor (V/F), which functions as an analog to serial-digital convert'er. The serial-digital data is transmitted to a microcomputer where it can be converted to temperature units and displayed or used for control.
,.*se) g 6
-. - ~ * ~
4 w
l'SL - / U - 283 q
SYSTEM e
NOlJr. IN!:All SL I -POINT DUAL g' lill llMIS T OR CIRCulT THEPMISTOl'
- l TEMPERATURES
=-
TEMPERATURE l SENSING L_ _ __ _ __
t_
}
o y
\\
HESISTOR IC INST.
~
[
~~
NETWORK AMP.
l d
~
a TEMPERATURE MEASUREMENT
_ \\
EXClTATION
/
VOLTAGE l
l MICROCOMPUTER
't
- _-.-_.-. _I CONTROL 6 ADDRESS OUS
-3 ADDRCSS l
DECODER l
O O
l 8 BIT 2K R AM +
14 BIT I
CPU EPROM 12716 COUNTER i8085 c,
O 8 BIT 0
o o
clock DATA BUS 30.72 kHz v
o i
l 9
f CE b
IN COUNTER DATA g
l Sgf
\\
WG M ggy
+
i J
M ONE ZERO SHOT
~
scT nest 7 :
CROSSING l
DETECTOR FLIP FLOP d
d l
HE ATER CONTROL i
7 DRIVE Pb TRANSISTOR c
TRIAC THAtlSFORMEH AC SOLENOID CONTROL cj CONTROL l
ORIVF TRANS.
q FLIP FLOP ~
l(
}
(
}-
N}/OR
=
SOLENO!D
_a L_
HEATER Tennperature Sensing and Centrol Electronics"for UF Handlint System 6
Figure 8 1
. llems that will be cnntrolled include the tawi4 ilisert and flow rate of the hot f;,, the flow of the LN, and the various systew ' nit noid valves.
2 The solenoid valves are the easiest to i.ontiel.
The n-computer address is decoded to drive a transistor which drives ti.e solenoid directly or through j
a relay.
The LN fl w is controlled in a similar_ mar.ner using air pressure 2
~
-controlled by a solenoid to propel the LN.
(This method is used for keeping 2
the diffusion pump c'old trap filled in the vacuum system currently in use at the reactor.)
A somewhat more complex system is used to control the N2 t eater and the heater tapes. According to the heating rate desired, the 9-computer outputs an eight bit digital word, corresponding to the fraction of the 60 Hz power frequency phase required to produce that heating rate, to a latch. At the start of each 60 Hz half-cycle a counter.is reset froci the latch and counted down to zero at which point it produces a pulse which turns on the triac and allows power to Elow in the heater circuit for the
)
remainder of the half cycle.
If the word is O power starts flowing immediately.
If the word is 256 the triac will never turn on.
Precision of control will be better than 1 part in 100.
I
.e e
e.
l 0422c
Ill. Operating Pe m t dures for Ul 6 ""d Ud5 I'""IIi*3 SY5t"*5 for fil'L-UI, I'mnping Lxperin:nts g
A.
System Preparat ion (1-5 performed only at the beginning of each series of experiments) 1.
Open9,ssandUf}handlingsystemsandlasercelltovacuumsystem, pump duwn and bake out (with diffusion -pump).
cyl. and transfer line trap
~
2.
Valve-off diffusion rump, chill UF6 to O C (ir,c bath, af ter initial cooling with LN ), fill with inert 2
valves. Allow accumulated HF to be pumped out gas, and open UF6 (note:
forepump is exhausted through reactor stack).
3.
Close valve tetween transfer line trap and vacuum system. Chill and then the heat trap with a heat gun, thereby UF supply cyl. with LNp 6
from trap int cylinder.
cryopumping UF6 4.
Cool transfer line trap cnd oas fill line trap with dry ice.
5.
Close-of f cylinders from transfer line and open transfer line trap -
vacuum system valve.
6.
Open dif fusion pump valves and pump down system to desired background pressure.
B.
Transfer UF fmn supply cyl. to laser cell.
6 l.
Heat supply cyl., transfer line, and laser cell to dasired temps.:
I' laser cell and transfer line to Tsubl + 25K (for desired nUF which,whenekpanded supply cyl. to temp. which will provide a P33g into laser cell, will provide the desired r.UF '
0 6
return cyl. with LN '
2 2.
When desired temps. are achieved, close laser cell to gas fill line valve and transfer line to tras.sfer line trap valve (thus valving UF system and laser cell off from vacuum system).
6 3.
(a) Set temp. control on supply cyl. to provide temp. for sat.
pressure for desired n F, then U
(b) open UF supplycyl.valhe.
[Expansiopf UF into transfer 6
line ind laser will cool UF, T=T
, q= pM b 6
g y
4.
Once pressures and temperatures stabilize:
(a) close transfer line valve to laser, (b) start cooling supply cyl. in ohder to th transfer line back into supply cyropump extra uncontaminated UF6 cyl and, (c) start cooling transfer line trap with dry ice.
While. supply cyl. is cocling and pumping go on to...
s*
,v
C.
Iill laser cell with laser and but fer gav'.
1.
(a) Clow clas fill punifold valve f.o vas:uum system.
(b) for laser systems with small atum fraction constituents (.10) also close manifold - gas fill line valve.
E' 2.
(a) Add gases to manifold or gas fill line in order of increasing d tout fraction (add to desired pressure times volume factor).
(b)
If small atom fraction constituents, add these plus portion of buffer gas first to manifold only (using volume factor of manjfold to n.anifold + fill line + laser cell) then expand into ir.anifold plus fill line and add remainder of buffer gas to desired pressure (taking into account volume factor).
3.
Open vale between gas fill Itne and laser cell and, as soon as pressure fluctuations subside, close valve again.
[SinceUF6 pressures will be (20% of total pressure in laser cell and since volume factors should be 2 er greater, pressure in the fill line should be 10-100 times greater than the UF6 pressure in the laser cell so that UF I ss fr m the laser cell into th. nas fill 6
line should be minimal - hopefully neu'.. ble.]
4.
Af ter allowing time for temperatures to 5 tabilize and for gas mixing by diffusion, the laser cell is readv tor pulsing.
U.
Pulse reactor and take data (covered by separate procedure) 1 E.
Evacuate gases and return UF to return cyl.
6 l.
Check to see that transfer line trap is cold (dry ice cooling) 2.
Shut UF supply cylinder valve and diffusion pump valve 6
3.
Open transfer line - transfer line trap valve and pump out.
gases [UF is caught in trap] to best pressure available with 6
forepump.
4.
Close transfer line trap - vacuum system valve and open UF return 6
cylinder valve.
5.
Remove cooling front transfer line trap and heat with heat gun until all Uf and other cryopumpable residuals are returned to 6
return cylinder.
6.
Close return cylinder valve. Open vacuum system to transfer l'ine trap and laser cell to gas fill valves.
System is now ready to return to A.6 and recycle.
-r e - - -.
e-,
-..,,n
,,,---,,,,--w.-,-,s-,,
-,v.
,e
-n-v e----+--~
,v--ww-s
,-,~-ve-w m-----cr-*,
-c
---,,---mm----+w~-
v,
....e lie f erent.cs F. 'P. Un idy, et. 41, "Proqess in No..le.ir Pumped I.asers" in !<adia, tion.
1.
- ~.
1 : cujy Lenve i. in's in '.pai s', K. W. I:illrian, I rl., Vnt. 61 of Progress
(,%'2'.
S.Attronautics' and Aeronautics, AI AA, th w York,1978.
Bcoqy, proposal to NASA entitled "An lxperimental Investigation
.f. g5
. -: :.- 2.
of Uf fission Nuclear-pumped Lasers," Submitted by G.11. Miley, 6
P.l., anc f unded as US PASA NSG 1547 (1978).
cc" h -3.
J'.' F: Jaminet and J. S. Kendall, " Circulation Eystec' tor Flowing Uraiiium flexh fluoride Cavity Reactor Cxperiment," Pr. ;. 3rd Conference on. Uranium Plasmas and Applications, Princeton Univ.,10-12 June 197.6's. NASA.
Gh 4.
A.
. achiels, Lecture 11aterial presented in course entitled " Vacuum D
Technoingy," presented at the University of Illinois, Urbana, IL 61801, Junta-Aug. 1978. Many of the formulas used are found in MW..
V'acubii let.luniluy, it'. foundations, f ormulae avid Tables, Leybold-
.M* '
)!cFists Voisuum Products, fnc7(T977),~'used'as"freTeFeii'ce for the G~- :
abo M e;urse.
w.-- - -
~
e e.
O e
.g..%
=
eau-g.
S 8
. ~.
e
~
x.... D
.