Investigation of the Unique Cryogenic Pumping System of
the CHAFF-IV Spacecraft-Thruster Interaction Facility
-i
f\
/•}
__
r\
Andrew D. Ketsdever , Marcus P. Young , Andrew Jamison , Brian Eccles
and E.P. Muntz2
1
Air Force Research Laboratory
Propulsion Directorate (PRSA)
Edwards AFB, CA 93524 USA
2
University of Southern California
Department of Aerospace and Mechanical Engineering
Los Angeles, CA 90089-1191 USA
Abstract. Chamber -IV of the Collaborative High Altitude Flow Facility was designed to obtain high fidelity spacecraftthruster interaction data. CHAFF-IV uses a total chamber pumping concept by lining the entire chamber with an array of
cryogenically cooled, radial fins. Details of Monte Carlo numerical simulation and experimental investigation of the
radial fin target array pumping efficiency are presented.
INTRODUCTION
Interactions between propulsion system effluents and sensitive surfaces have received considerable interest
within the spacecraft community in recent years. The impact of potential interactions is becoming more critical as
mission life and payload sensitivity requirements increase. The adsorption of propellant gases on spacecraft surfaces
can change the spectral absorptivity of thermal control surfaces, alter reflectivity of optical surfaces, alter
transmission through solar cell coverglass, and induce environments that can alter scientific results. Ion electric
thrusters add further complications due to material sputtering from high energy propellant impact and the possible
alteration of spacecraft potentials.
Because of the cost and time requirements associated with space-based experiments, ground-based examination
of spacecraft-thruster interactions is necessary to compliment the limited data returned from space. In order for
ground-based experiments to be effective, facilities must faithfully and consistently reproduce the space
environment. The major limitation of ground facilities in accurately predicting the effects of thruster operations on
spacecraft systems has been driven by the facility background pressure. Although some fraction of the background
gas is composed of the residual laboratory atmosphere, the largest complication arises from the fact that the
overwhelming majority of the background pressure is thruster derived. These thruster-borne components are largely
responsible for experimental measurement errors. Minimizing the effects of these effluents on interaction results
requires extremely high pumping rates; therefore, improvements in the accuracy of ground-based interaction data
come at the expense of more efficient chamber pumping. A general review of interaction facilities and advanced
pumping concepts has been compiled by Ketsdever. [1]
Chamber IV of the David P. Weaver Collaborative High Altitude Flow Facility (CHAFF) was designed in an
effort to obtain meaningful spacecraft-thruster interaction data by maximizing the facility's high vacuum pumping.
[2] As shown in Figures 1 and 2, CHAFF-IV incorporates a total chamber pumping (TCP) concept by lining the
entire facility with an array of cryogenically cooled, radial fins. The aluminum finned arrays are contained inside a
3m diameter by 6m long stainless-steel vacuum chamber. Details of the initial investigation of CHAFF-IV s
cryogenic pumping system are presented along with results from a simple Monte Carlo numerical model. For the
experiment, the target section of the radial fin array and a simple flat panel have been independently investigated.
CP585, Rarefied Gas Dynamics: 22nd International Symposium, edited by T. J. Bartel and M. A. Gallis
© 2001 American Institute of Physics 0-7354-0025-3/01/$18.00
916
Both
Bothpumping
pumping configurations
configurations were
were cooled
cooled toto liquid
liquid nitrogen
nitrogen (LN2)
(LN2) temperatures
temperatures (77-80K)
(77-80K) with
with aa sonic
sonic orifice
orifice
introducing
a
carbon
dioxide
gas
flow
on
the
axis
of
the
chamber.
introducing a carbon dioxide gas flow on the axis of the chamber.
1.0 m Diffusion
Pumps
Radial
RadialFin
Fin
Arrays
Arrays
FIGURE
FIGURE1.1. CHAFF-IV
CHAFF-IVCryogenically
CryogenicallyCooled,
Cooled,Radial
RadialFin
FinTotal
TotalChamber
ChamberPumping
Pumping Arrangement.
Arrangement.
FIGURE
FIGURE2.2. Fabricated
FabricatedAluminum
AluminumFinned
FinnedArrays
ArraysInstalled
Installedin
inCHAFF-IV.
CHAFF-IV.
FACILITY
FACILITYBACKGROUND
BACKGROUNDPRESSURE
PRESSURE REQUIREMENTS
REQUIREMENTS
InInaatypical
typicalground-based
ground-basedfacility
facilitywith
withaasmall
smallfraction
fractionffppofofits
itsinner
innersurface
surfaceoccupied
occupiedby
bypump
pumpinlets,
inlets,the
the thruster
thruster
effluents
effluents are
aretypically
typically stopped
stopped and
and randomized
randomized by
by the
the facility's
facility’s surfaces.
surfaces. The
The random
random motion
motion of
of the
the scattered
scattered
propellantmolecules
moleculesinefficiently
inefficientlybrings
bringsthem
themtotoaapump
pumpinlet.
inlet. The
Thebackground
background pressure
pressure of
of the
the propellant
propellant gas
gas can
can
propellant
approximatedby
by
bebeapproximated
__
˙k T T
˙ kT
4M
b p
M
MkT
hb
(1)
p
=
=
Pb b = mvjA
mv ′f p As
mV˙
mV
˙ isisthe
were MM
thepropellant
propellantmass
massflow,
flow,kkisisBoltzmann's
Boltzmann’s constant,
constant, TTbb and
and TTpp are
are the
the background
background and
and propellant
propellant gas
gas
were
temperaturesrespectively,
respectively,mmisisthe
the molecular
molecular mass
mass of
of the
the propellant
propellant gas,
gas, v'v ′ isis the
the average
average thermal
thermal speed
speed of
of the
the
temperatures
backgroundgas,
gas,f pfpisisthe
thefraction
fractionofofthe
thefacility
facilityinner
innersurface
surfaceoccupied
occupiedby
bypump
pumpinlets
inlets or
or pumping
pumping surfaces,
surfaces, A
Ass is
is the
the
background
innersurface
surfacearea
areaofofthe
thefacility,
facility,and
and VV˙isisthe
the facility's
facility’spumping
pumping speed.
speed. In
In order
order to
to minimize
minimize the
the effects
effects of chamber
inner
inducedcharge
chargeexchange
exchangecollisions
collisionsininxenon
xenonion
ionthrusters,
thrusters, background
background pressures
pressures on
on the
the order
order of
of 33 xx 10~
10-66 Torr
Torr are
are
induced
917
required. [1] For a typical Hall thruster propellant mass flow rate of 5 mg/s, Eq. (1) indicates that pumping rates in
required. [1] For a typical Hall thruster propellant mass flow rate of 5 mg/s, Eq. (1) indicates that pumping rates in
excess of 2.5 x 105 5L/s are required. For a cold gas system with a nitrogen propellant flow rate of 1 g/s, pumping
excess of 2.5 x 10 L/s are required. For a cold gas system with a nitrogen propellant flow rate of 1 g/s, pumping
rates on the order of 107 7L/s are required to maintain free molecular flow in the plume backflow region.
rates on the order of 10 L/s are required to maintain free molecular flow in the plume backflow region.
Clearly a critical background number density for thruster plume interaction studies is reached when the
Clearly a critical background number density for thruster plume interaction studies is reached when the
background mean free path becomes less than or equal to the largest internal dimension of the facility Lc. Therefore,
background mean free path becomes less than or equal to the largest internal dimension of the facility Lc. Therefore,
the background number density should be
the background number density should be
1
(2)
nb ≤
2σ b Lc
where σb is the background gas collision cross section and λc is the critical mean free path.
where Gb is the background gas collision cross section and ?lc is the critical mean free path.
As Eq. (1) indicates, background gas pressure can be minimized by having large available pumping areas (fpAs).
As Eq. (1) indicates, background gas pressure can be minimized by having large available pumping areas (f As).
For a given chamber geometry, the pumping rate is maximized by increasing the fraction of the inner surface parea
For a given chamber geometry, the pumping rate is maximized by increasing the fraction of the inner surface area
which acts as a pump. This suggests that high pumping rates can be achieved when the entire inner surface of the
which acts as a pump. This suggests that high pumping rates can be achieved when the entire inner surface of the
facility is a pumping surface. The TCP concept has driven the design of several interaction facilities. [2-4]
facility is a pumping surface. The TCP concept has driven the design of several interaction facilities. [2-4]
For
the array
array that
that is
is able
able to
to return
return to
to
Forthe
theradial
radialfin
finTCP
TCP array,
array, the
the fraction
fraction of
of efflux
efflux from
from aa thruster
thruster impinging
impinging on
on the
the
thruster
vicinity
is
given
by
the thruster vicinity is given by
2
 w2
t  Do 


(3)
Fr = (1 −η) 
+

 2h(w + t) (w +t)  X 
where
the axial
direction, tt
whereηr| isisthe
thesticking
stickingcoefficient,
coefficient, wwisisthe
the radial
radial fin-to-fin
fin-to-fin spacing,
spacing, hh is
is the
the length
length of
of the
the fin
fin in
in the
axial direction,
isisthe
fin
thickness,
D
is
the
characteristic
thruster
diameter,
and
X
is
the
distance
from
the
thruster
exit
plane
to the
the
the fin thickness, Do0 is the characteristic thruster diameter, and X is the distance from the thruster exit plane to
front
edge
of
the
radial
fin
array.
Design
of
the
radial
fin
arrays
can
be
optimized
through
the
minimization
of
the
front edge of the radial fin array. Design of the radial fin arrays can be optimized through the minimization of the
geometric
For the
the CHAFF-IV
CHAFF-IV radial
radial
geometricterm
termininthe
the brackets
brackets of
of Eq.
Eq. (3)
(3) by
by an
an appropriate
appropriate selection
selection of
of the
the fin
fin geometry.
geometry. For
fin
target
array,
h
=
25.4
cm,
t
=
0.32
cm,
and
w
varies
from
1
to
6
cm
due
to
the
radial
nature
of
the
array.
fin target array, h = 25.4 cm, t = 0.32 cm, and w varies from 1 to 6 cm due to the radial nature of the array.
EXPERIMENT
EXPERIMENT
In
target section
array shown
shown in
in Fig.
Fig.
Inorder
ordertotodetermine
determinethe
the pumping
pumping capabilities
capabilities of
of the
the 112
112 fins
fins that
that make
make up
up the
the target
section array
3,3, LN
centerline. Carbon
Carbon dioxide
dioxide
2 was
2 gas
LN2
was used
used toto cool
cool the
the panels
panels while
while CO
CO2
gas was
was introduced
introduced on
on the
the chamber
chamber centerline.
pumping
xenon) pumping
pumping on
on 152 surface
pumpingatatLN
LN2
surface temperatures
temperatures of
of ~80
~80 K
K was
was used
used to
to simulate
simulate other
other gases
gases (including
(including xenon)
1520
K
surfaces
since
the
sticking
coefficients
are
similar
as
shown
in
Table
1.
[5]
20 K surfaces since the sticking coefficients are similar as shown in Table 1. [5]
Radial
RadialFin
Fin
Array
Array
Patterson Probe
Probe
Patterson
Sonic
SonicOrifice
Orifice
FIGURE
FIGURE3.
3. Experimental
Experimental Configuration
Configuration of
of Radial
Radial Fin
Fin Array.
Array.
For
orifice (diameter
(diameter ==
For the
the radial
radial fin
fin configuration,
configuration, the
the CO
CO2
was introduced
introduced into the chamber through a sonic orifice
2 was
0.178cm)
0.178cm)on
onthe
thechamber
chambercenterline
centerline located
located 79.1cm
79.1cm from
from the
the front
front of
of the
the array
array as shown
shown in Fig. 4(a). The density of
the
the free
free jet
jet flow
flow field
field decreases
decreases as
as the
the inverse
inverse square
square of
of the distance
distance x from the source as measured along a
streamline.
function of
streamline. The
The number
number density
density also
also varies
varies from
from streamline
streamline to
to streamline.
streamline. The number density as a function
position
positionfrom
from the
thesonic
sonicorifice
orifice isis given
given by
by [6]
[6]
918
(4)
where x is the axial distance downstream of the orifice, 0 is the angle from the orifice centerline, D0 is the orifice
diameter, and § is a constant based on the gas specific heat ratio and is 1.7 for CO2. The CO2 flow rates introduced
into the facility ranged from 10 to 120 seem although pumping rate data was obtained for mass flows up to 24,750
seem.
TABLE 1. Sticking coefficients of some common gases as a function of gas and surface temperature. [5]
The radial fin system was instrumented with five temperature sensors to ensure that the panel system was at
appropriate pumping temperatures. To determine the pressure at various locations inside the chamber, five BayardAlpert type ion gauges were used. The data presented is typically in terms of pressure differences as measured from
the facility pressure without gas flow. In this study, CHAFF-IV ultimate pressures were typically below 1.5 x 10"6
Torr. One of the ion gauges was attached to a Patterson Probe allowing 360 degree rotation along the chamber
length and a tunable distance from the side wall to the chamber centerline, r, as shown in Fig. 5. The angle a is
measured from the probe centerline, and a = 0" indicates that the probe is pointing towards the cryogenic arrays. In
order to accurately assess the data obtained by the Patterson probe, the flow near the probe orifice must be free
molecular. Based on Eq. (4), the flow from the sonic orifice can be considered free molecular near the Patterson
probe at flow rates below 100 seem. Much of the discussion for this research will be confined to flow rates in this
range.
A cryogenic flat panel pumping surface was also tested in the facility in an attempt to assess any pumping
improvement afforded by the finned geometry. Figure 4(b) shows the experimental set up for the flat panel array.
In this case, the sonic orifice was adjusted to a distance of 110.2 cm from the panel to maintain constant coverage
(-90%) from the sonic orifice between the two tests.
In both experiments, one of the two available 1.0 meter diameter diffusion pumps was used to pump
incondensible gases in the facility. The effective pumping speed of the diffusion pump on CO2 gas was measured to
be between 10,600 and 11,800 L/s with a theoretical maximum of about 20,000 L/s without conduction losses.
MONTE CARLO NUMERCIAL MODELING
A computational investigation was used to further understand the pumping characteristics of the radial fin array.
The CHAFF-IV array has 112 similar wedge shaped sections. The computational domain involved only one wedge
section for simplicity. Free molecule flow was assumed in the volume between the panels which allowed individual
molecules trajectories to be followed. Pumping statistics were built from individual particle dynamics to represent
the physical problem. Molecules were emitted from a point source on the centerline of the facility a distance of 79.6
cm from the front edge of the fin array. The particles were given randomly selected velocity components in the
horizontal and vertical directions based on the distribution in Eq. (4). All surface interactions are treated as fully
diffuse implying that the molecules accommodate to the surface temperature. The sticking coefficient can be
updated based on the gas temperature and the presumed temperature of the wall from a database based on Table 1
for CO 2. Initially, the incident molecules can either strike the front thickness of the radial fin or enter the volume
between two fins. The molecules are followed until they effectively stick to the panel or cross the front plane of the
radial array in which case they are considered backscattered molecules. The sticking probability model is based on a
Monte Carlo acceptance-rejection scheme. The initial sticking coefficient of 300 K carbon dioxide molecules on a
liquid nitrogen cooled cryogenic surface is assumed to be 0.63. [5]
919
Thesurface
surfacebehind
behindthe
theradial
radialfin
finarray
arrayisismodeled
modeledasasa aflat
flatplate.
plate.This
Thissurface
surfacecan
caneither
either
cryogenically
cooled
The
bebecryogenically
cooled
to
represent
a
pumping
surface
or
at
room
temperature
to
represent
a
chamber
wall.
Results
were
obtained
using
to represent a pumping surface or at room temperature to represent a chamber wall. Results were obtained using
both
configurations
of
the
back
wall.
For
future
CHAFF-IV
investigations,
the
radial
fin
temperature
will
be
10-15
both configurations of the back wall. For future CHAFF-IV investigations, the radial fin temperature will be 10-15
suppliedby
byaagaseous
gaseoushelium
heliumcryostat,
cryostat,and
andthe
theback
backwall
wallwill
willbebeliquid
liquidnitrogen
nitrogencooled.
cooled.Therefore,
Therefore,
this
model
KK supplied
this
model
cancan
be
used
to
address
future
CHAFF-IV
configurations
without
modification.
be used to address future CHAFF-IV configurations without modification.
a)
a)
b)b)
r
r
x
x
215.9
Sonic Orifice
299,7
Sonic Orifice
Patterson
PattersonProbe
Probe
Patterson
PattersonProbe
Probe
FIGURE
Radial
FinFin
Array,
(b)(b)
Flat
Panel
Array.
FIGURE4.4. Experimental
ExperimentalPosition
PositionofofSonic
SonicOrifice
Orificeand
andPatterson
Pattersonprobe,
probe.(a)(a)
Radial
Array.
Flat
Panel
Array.
a α==90°
90°
Ion
IonGauge
Gauge
r
α = 0°
x
Cryopumping
Cryopumping
Arrays
Arrays
Probe
ProbeOrifice
Orifice
-90°
−90°
FIGURE
5. Patterson Probe Geometry and Experimental Configuration.
FIGURE 5. Patterson Probe Geometry and Experimental Configuration.
920
RESULTS
RESULTS
The
results of the Monte Carlo simulations for the fraction of backscattered molecules from the radial fin array
The results of the Monte Carlo simulations for the fraction of backscattered molecules from the radial fin array
and
the
flat panel are shown in Fig. 6 as a function of sticking coefficient. The solid radial fin data line is for
and the flat panel are shown in Fig. 6 as a function of sticking coefficient. The solid radial fin data line is for
simulations
with aa surface.
surface. If
If the
the molecule
molecule strikes
strikes aa
simulations which
which update
update the
the sticking
sticking coefficient
coefficient for
for each
each interaction
interaction with
cryogenic
surface
and
is
not
pumped,
the
sticking
coefficient
becomes
unity
if
the
molecule
should
hit
another
cryogenic surface and is not pumped, the sticking coefficient becomes unity if the molecule should hit another
cooled
surface.
If
the
molecule
strikes
a
chamber
wall,
the
sticking
coefficient
is
reset
to
the
original
orifice
cooled surface. If the molecule strikes a chamber wall, the sticking coefficient is reset to the original orifice
expansion
value.
The
dashed
data
line
is
for
a
constant
sticking
coefficient
throughout
the
simulation.
The
radial
fin
expansion value. The dashed data line is for a constant sticking coefficient throughout the simulation. The radial fin
arrays
outperform
the
simple
flat
panel
for
free
molecule
flow
for
sticking
coefficients
less
than
approximately
0.55.
arrays outperform the simple flat panel for free molecule flow for sticking coefficients less than approximately 0.55.
As
Figure 66 shows
shows the
the utility
utility of
of the
the
As expected,
expected, the
the two
two radial
radial fin
fin results
results converge
converge for
for large
large sticking
sticking probabilities.
probabilities. Figure
radial
fin
arrays
for
the
pumping
of
high
energy
propulsion
system
flows
where
the
initial
sticking
coefficients
are
radial fin arrays for the pumping of high energy propulsion system flows where the initial sticking coefficients are
expected
to
be
extremely
low.
expected to be extremely low.
i.o
1.0
Flat
Pane!
Flat Panel
- Radial
Radial Fin
Fin -- Constant
Constant Sticking
Sticking
Radial Fin
Fin -- Updated
Updated Sticking
Sticking
Radial
0.80
Pumping Efficiency
0.80
0.60
0.60
0.40
0.40
0.20
0.20
0.0
0
0.2
0.4
0.6
0.8
1
Sticking
Coefficient
Sticking Coefficient
FIGURE
as aa Function
Function of
of Sticking
Sticking Coefficient.
Coefficient.
FIGURE 6.
6. Monte
Monte Carlo
Carlo Results
Results for
for Pumping
Pumping Efficiency
Efficiency as
The
opposite end
end of
of the
the pumping
pumping
The measured
measured pressure
pressure in
in the
the sonic
sonic orifice
orifice backflow
backflow region (i.e. measured at the opposite
arrays)
was
used
to
determine
the
pumping
rates
for
the
radial
fin
array.
As
a
function
of
mass
flow,
the
calculated
arrays) was used to determine the pumping
flow, the calculated
˙ = 10 sccm)
˙ ==24,750
facility
seem) and 2.6 x 10
1066 L/s
L/s ((M
24,750 sccm).
seem).
facilitypumping
pumping speed
speed was
was between
between 0.65 (M
(M
M
Figure
of mass
mass flow
flow rate
rateatatthe
thecenter
centerofofboth
both
Figure 77 shows
shows the
the pressure
pressure measured
measured at the Patterson probe as a function of
pumping
arrays
(r
=
0)
and
at
the
edge
of
the
arrays
(r
=
98
cm).
The
general
trend
indicates
that
the
radial
fin
array
pumping arrays (r = 0) and at the edge the
indicates that the radial fin array
pumps
seem on the chamber centerline.
centerline. For
For higher
higher mass
mass
pumps more
more efficiently
efficiently for
for mass
mass flows
flows up
up to
to approximately 60 sccm
flow
radial fins
fins are
are outperforming
outperforming the
the flat
flat
flow rates,
rates, the
the flat
flat panel
panel array
array outperforms
outperforms the radial fins. The fact that the radial
panel
below the expected
expected value
value of
of
panelindicates
indicates that
that the
the effective
effective sticking coefficient
coefficient for the experimental results may be below
0.63
since the
the array
array temperature
temperature was
was measured
measuredinin
0.63 (Table
(Table 1)
1) based
based on
on the
the Monte
Monte Carlo simulations. This is possible since
the
theexperiment
experimentto
tobe
bebetween
between 81
81 and
and 88 K.
Above 60
60 seem,
sccm, collisions
collisions between
between plume molecules and backscattered
Above
backscattered molecules
molecules from
from the
the pumping
pumping arrays
arrays
beginto
tobecome
become important.
important. Molecular
Molecular collisions act to return molecules to the arrayss
begin
arrayss that
that are
are not
not initially
initially pumped.
pumped.
For the
the flat
flat panel
panel array,
array, the
the return
return flux
flux to
to the
the array is at a somewhat lower temperature
For
temperature having
having interacted
interacted with
with the
the
pumping surface
surface once
once before
before making
making the
the effective
effective sticking coefficient higher as
pumping
as indicated
indicated in
in Table
Table 1.1. For
Forthe
theradial
radial
finarray,
array,the
thereturn
return flux
flux once
once again
again passes
passes through the fin geometry and interacts with the chamber
fin
chamber back
back wall
wall atat 300
300
K. The
The molecules
molecules partially
partially accommodate
accommodate to the chamber wall temperature
K.
temperature (depending
(depending on
on the
the temperature
temperature
dependent accommodation
accommodation coefficient);
coefficient); thus, there is not a significant increase
dependent
increase in
in the
the effective
effective sticking
sticking coefficient
coefficient
for the
the radial
radial fin
fin array.
array. Since
Since the
the Monte
Monte Carlo
Carlo results assume a free molecule condition,
for
condition, this
this effect
effect would
would not
not be
be
reproduced
in
the
model.
A
direct
simulation
Monte
Carlo
technique
is
required
at
higher
flow
rates.
reproduced in the model. A direct
required at higher flow rates.
The fraction
fraction of
of impinging
impinging flow
flow on the pumping arrays which is backscattered
The
backscattered to
to the
the Patterson
Patterson probe
probeisisshown
showninin
Fig. 8.
8. The
The ratio
ratio is
is obtained
obtained by
by dividing the measured Patterson probe pressure
Fig.
pressure at
at αa == 0°
0° by
by that
that atat αa ==180°
180°(i.e.
(i.e.
pointed in
in the
the direction
direction of
of the
the orifice
orifice expansion).
expansion). At 15 sccm,
pointed
seem, the backscattered fraction
fraction from
from the
the radial
radial fins
fins isis
approximately 0.052.
0.052. This
This agrees
agrees well
well with a free
free molecule backscattered fraction
approximately
fraction of
of 0.055
0.055 derived
derivedfrom
from Eq.
Eq.(3).
(3).
Figure 99 shows
shows aa radial
radial profile
profile for
for the
the fin
fin and
and the
the flat
flat panel
panel pumping
pumping configurations
Figure
configurations for
for aa mass
mass flow
flow of
of 20
20 sccm.
seem.
This data
data indicates
indicates that
that the
the finned
finned array
array is
is more
more effective
effective than
This
than aa flat
flat panel,
panel, and
and in
in some
some cases,
cases, the
the pumping
pumping ratio
ratio
reaches aa factor
factor of
of more
more than
than 4.
4. Similar
Similar results
results were
were obtained
reaches
obtained for
for flow
flow rates
rates up
up to
to approximately
approximately 65
65sccm.
seem. The
Thedata
data
indicates relatively
relatively high
high backscattered
backscattered flux
flux from
from the
the center
center of
indicates
of the
the radial
radial fins.
fins. This
This is
is due
due to
to aa maximum
maximum in
in incident
incident
921
flux on
on the
the centerline
centerline from
from the
the sonic
sonic orifice
orifice expansion
expansion and
and aa lack
lack of
of pumping
pumping surface
surface in
flux
in the
the center
center of
of the
the finned
finned
array due
due to
to construction
construction tolerances.
tolerances. The
The increase
increase in
in pressure
pressure as
as the
the probe
array
probe radial
radial position
position increases
increases (i.e.
(i.e. tends
tends
towards the
the edge
edge of
of the
the fins)
fins) can
can be
be caused
caused by
by several
several factors.
towards
factors. First,
First, the
the width
width of
of the
the radial
radial fins
fins increases
increases as
as aa
function of
of distance
distance from
from the
the chamber
chamber centerline.
centerline. Second,
Second, the
the temperature
temperature at
at the
the far
far edge
somewhat
function
edge of
of the
the fins
fins is
is somewhat
warmer (~
(~ 6-10
6-10 K)
K) than
than that
that near
near the
the center
center of
of the
warmer
the array
array due
due to
to heat
heat transfer
transfer issues
issues making
making the
the pumping
pumping less
less
efficient.
Finally,
there
is
some
fraction
of
the
orifice
mass
flow
that
does
not
impinge
directly
on
efficient. Finally, there is some fraction of the orifice mass flow that does not impinge directly on the
the pumping
pumping
arrays due
due to
to the
the high
high angle
angle gas
gas expansion
expansion that
that can
can act
act to
to increase
increase the
arrays
the background
background pressure
pressure at
at the
the array
array edge.
edge.
a)
b)
2.0
a) 1.5
b) 2.0
∆P (Torr x 106)
o
X
0.5
1.0
1.0
0.50
Radial Fins
Flat Panel
0
0
20
20
40
40
60
60
80
80
100
100
Radial Fins
Flat panel
120
120
0.0
0.0
40
60
40
60
Mass Flow (seem)
Mass Flow (sccm)
20
0
20
Mass Flow (seem)
Mass Flow (sccm)
80
80
FIGURE 7.
7. Patterson
Patterson Probe
Probe Change
Change in
in Pressure
Pressure as
as aa Function
Function of
of Mass
Flow, (a)
FIGURE
Mass Flow.
(a) rr =
= 00 cm,
cm, (b)
(b) rr =
= 98
98 cm.
cm.
0.08
0.08
Pressure Probe Flux Ratio
o
£
£
£g
PH
|
5a
0.07
0.07
0.06
0.06
0.05
0.05
0.04
0.04
0.03
0.03
CO
£
0.02
0.02
- Radial Fins _
Radial
Fins
- Flat
Panel
Flat Panel
0.01
0.01
0
0 0
0
40
60
40
60
Mass Flow (seem)
Mass Flow (sccm)
20
20
100
100
80
80
FIGURE 8. Fraction of Backscattered Molecules from Pumping Arrays as a Function of Mass Flow.
FIGURE 8. Fraction of Backscattered Molecules from Pumping Arrays as a Function of Mass Flow.
2.5
2.0
∆P (Torr x 106)
∆P (Torr x 106)
1.5
1.5
1
I
1.5
1.0
53
0.50 0.50
0.0
0
Flat Panel
Radial Fins
20
20
40
60
40
60
Radial Distance (cm)
Radial Distance (cm)
80
80
100
100
FIGURE 9. Patterson Probe Change in Pressure as a Function of Radial Position for a Flow Rate of 20 SCCM.
FIGURE 9. Patterson Probe Change in Pressure as a Function of Radial Position for a Flow Rate of 20 SCCM.
922
100
100
Figure
of the
the Patterson
Patterson probe
probedata
dataatatvarious
variousfinned
finnedarray
arrayradial
radialpositions
positionsforfora a
Figure 10
10 shows
shows the
the angular
angular variation
variation of
mass
the pressure
pressure isis minimized
minimized for
foran
anangle
angleofofαa ==0˚0°where
wherethe
theprobe
probeisispointed
pointed
mass flow
flow of
of 20
20 seem.
sccm. As
As expected,
expected, the
directly
There is
is very
very little
little angular
angular dependence
dependenceofofthe
thebackscattered
backscatteredmolecule
moleculepopulation
population
directly atat the
the pumping
pumping surfaces. There
for
the edge
edge of
of the
the finned
finned array.
array. At
Atthis
thislocation,
location,the
thepumping
pumpingisisknown
knowntotobebeless
less
foraaposition
position of
of 98
98 cm
cm which
which is
is near
near the
efficient
efficientas
as discussed
discussed previously.
previously.
1
0.8
∆P (Torr x 106)
€
X
0.6
V
13 - -cr"
0.4
-e—r r= =9898cmcm
- B- r- =
r =57.5
57.5cmcm
-n- r- r= =0 0cm
cm
0.2
0.2
00
-100
-100
-50
-50
00
5050
'
100
100
Angle
Angle(degrees)
(degrees)
FIGURE 10.
10. Patterson
Patterson Probe
Probe Data
Data as
2020
SCCM.
FIGURE
as aa Function
Function of
of Sampling
SamplingAngle
AngleatatSeveral
SeveralRadial
RadialPositions
PositionsforforFlow
FlowRate
Rateofof
SCCM.
CONCLUSIONS
CONCLUSIONS
Numerical and
and experimental
experimental comparison
Numerical
comparison of
of the
the pumping
pumping efficiencies
efficiencies ofofthe
theradial
radialfinfinarray
arraywith
witha asimple
simpleflatflat
panel array indicates that the radial fins are more efficient for free molecular flows (i.e. mass flows less than 60
panel array indicates that the radial fins are more efficient for free molecular flows (i.e. mass flows less than 60
sccm on the chamber centerline). The radial fins consistently outperform the flat panel at off-axis positions (i.e. r >
seem on the chamber centerline). The radial fins consistently outperform the flat panel at off-axis positions (i.e. r >
0) for the range of experimental mass flows investigated in this study. It is expected that the radial fin array will
0) for the range of experimental mass flows investigated in this study. It is expected that the radial fin array will
significantly outperform a flat panel array for thruster efflux at elevated temperatures and for energetic ion flows.
significantly
outperform a flat panel array for thruster efflux at elevated temperatures and for energetic ion flows.
For neutral gas at elevated temperature and energetic ion flows, the radial fin array backed by a LN2 shield
For
neutral
gas at elevated temperature and energetic ion flows, the radial fin array backed by a LN2 shield
should perform optimally. The sticking coefficient increases for a given surface temperature as the gas temperature
should
perform
optimally.
Thefin
sticking
for by
a given
temperature as the gas temperature
decreases. When
the radial
array coefficient
is at 15-20increases
K backed
a LNsurface
2 shroud, the energetic molecules will
decreases.
When
the
radial
fin
array
is
at
15-20
K
backed
by
a
LN
shroud,
the surface
energeticwith
molecules
will
2
accommodate (at least partially) to the 77 K surface. After scattering from
the LN
a velocity
2
accommodate
(at
least
partially)
to
the
77
K
surface.
After
scattering
from
the
LN
surface
with
a
velocity
2
distribution function characteristic of the surface temperature, the effective sticking coefficient on the radial array
distribution
function
the surface
temperature, the effective sticking coefficient on the radial array
will be lower
therebycharacteristic
increasing theof
pumping
efficiency.
will The
be lower
thereby
increasing
the
pumping
efficiency.
pumping rate for the radial fin array ranges from approximately 0.65 to 2.6 x 106 6L/sec depending on the
The
pumping
forexperimental
the radial finconfiguration
array rangesutilized
from approximately
10 L/sec
depending
source mass flow.rateThe
approximately0.65
5.6%toof2.6thex total
CHAFF-IV
radialonfinthe
source
Thethat
experimental
configuration
utilized1.0
approximately
the total
CHAFF-IV
radial
array. mass
This flow.
indicates
pumping rates
of approximately
to 4.5 x 107 75.6%
L/sec of
should
be possible
for cold
gasfin
array.
Thisthe
indicates
that pumping
of approximately
1.0 pumping
to 4.5 x rates
10 L/sec
should be
for cold
gas
flows with
entire facility
pumpingrates
configuration
active. The
for energetic
ionspossible
are expected
to be
flows
with
the
entire
facility
pumping
configuration
active.
The
pumping
rates
for
energetic
ions
are
expected
to
similar. Although the rate should increase due to the flux energy to the pumping surface, pumping efficiencybe
similar.
the due
ratetoshould
increaseofdue
to the gases
flux energy
to the
pumping
surface,
pumping
efficiency
decreasesAlthough
are expected
the sputtering
condensed
and facility
material
(graphite
protective
layers).
decreases are expected due to the sputtering of condensed gases and facility material (graphite protective layers).
REFERENCES
REFERENCES
1. Ketsdever, A., “An Overview of Ground Based Spacecraft-Thruster Interaction Studies: Facility Design Issues,” AIAA paper
th
2000-0463,A.,
38"An
Aerospace
Sciences
Meeting,
Reno,
NV, 2000.
1. Ketsdever,
Overview
of Ground
Based
Spacecraft-Thruster
Interaction Studies: Facility Design Issues," AIAA paper
th
2. 2000-0463,
Lutfy, F., Vargo,
S., Muntz,Sciences
E.P., Ketsdever,
“TheNV,
David
P. Weaver Collaborative Flow Facility’s CHAFF-IV for Studies
38 Aerospace
Meeting,A.,
Reno,
2000.
of Spacecraft
Propulsion
and Contamination,”
AIAAP.paper
98-3654,
34th JointFlow
Propulsion
Conference,
2. Lutfy,
R, Vargo,
S., Muntz,Plumes
E.P., Ketsdever,
A., "The David
Weaver
Collaborative
Facility's
CHAFF-IVCleveland,
for Studies
1998. Propulsion Plumes and Contamination," AIAA paper 98-3654, 34th Joint Propulsion Conference, Cleveland,
ofOH,
Spacecraft
3. Dettleff,
G., Plahn, K., “Initial Experimental Results form the New DLR-High Vacuum Plume Test Facility STG,” AIAA
OH, 1998.
paper 97-3297,
1997.
3. Dettleff,
G., Plahn,
K., "Initial Experimental Results form the New DLR-High Vacuum Plume Test Facility STG," AIAA
4. Stephens,
J.B., “Space
paper 97-3297,
1997. Molecular Sink Simulator Facility Design,” NASA TR-32-901, 1966.
Welch, K.,J.B.,
Capture
Pumping
Technology,
Pergamon
Press,Design,"
1991. NASA TR-32-901, 1966.
4.5.Stephens,
"Space
Molecular
Sink Simulator
Facility
Ashkenasn,
H., Sherman,
“StructurePergamon
and Utilization
of Supersonic Free Jets in Low Density Wind Tunnels,” in 4th
5.6.Welch,
K., Capture
PumpingF.S.,
Technology,
Press, 1991.
International
Symposium
on
Rarefied
Gas
Dynamics,
ed.
J.
Academic
Press,inNew
1966,Wind
pp. 84.
6. Ashkenasn, H., Sherman, F.S., "Structure and Utilization de
of Leeuw,
Supersonic
Free Jets
LowYork,
Density
Tunnels," in 4th
International Symposium on Rarefied Gas Dynamics, ed. J. de Leeuw, Academic Press, New York, 1966, pp. 84.
923