Skip to main content

Full text of "NASA Technical Reports Server (NTRS) 19920013118: NASCAP/LEO simulations of Shuttle Orbiter charging during the SAMPIE experiment"

See other formats


N92-22 36 


NASCAP/LEO SIMULATIONS OF SHUTTLE ORBITER CHARGING 
DURING THE SAMPIE EXPERIMENT 


Ricaurte Chock 
NASA Lewis Research Center 
MS 301-3 

21000 Brookpark Rd., Cleveland, OH 44135 


ABSTRACT 


The electrostatic charging of the Shuttle Orbiter 
during the operation of the Solar Array Module 
Plasma Interaction Experiment (SAMPIE) has been 
modeled using the NASCAP/LEO computer code. 
The SAMPIE experiment, scheduled to be flown in 
the shuttle payload bay in 1993, consists of an array 
of various solar cells representing the present 
technologies. The objectives of the experiment are 
to investigate the arcing and current collection 
characteristics of these cells when biased to high 
potentials in a low Earth orbit (LEO) plasma. 
NASCAP/LEO ( NAS A Charging Analyzer 
Program/Low Earth Qrbit) is a 3-D code designed 
to simulate the electrostatic charging of a spacecraft 
exposed to a plasma at low earth orbit or ground 
test conditions. At its most extreme configuration, 
with the largest array segment of the SAMPIE 
experiment biased +600 V with respect to the 
Orbiter and facing the ram direction, the computer 
simulations predict that the Orbiter’s potential will 
be approximately -20 V with respect to the plasma. 


I. INTRODUCTION 


NASCAP/LEO simulations comparing ground test 
results with low earth orbit conditions have 
highlighted the difficulties encountered when trying 
to extrapolate solar array behavior under LEO 
conditions from vacuum chamber experiments. 
NASCAP/LEO ( NAS A Charging Analyzer 
Program/Low Earth Qrbit) is a 3-D code designed 
to simulate the electrostatic charging of a spacecraft 
exposed to a plasma at low earth orbit or ground 
test conditions (ref. 1). Using this code it has been 
found that small changes in cell geometry, such as 
allowing a cell cover glass overhang of 6 mils, will 
greatly impact the cell’s current collection behavior 
(ref. 2). In order to better understand such 


behavior, actual flight experiments are needed. One 
of these is the Solar Array Module Elasma 
Interaction Experiment (SAMPIE), scheduled to be 
flown in the Orbiter bay in 1993 (ref. 3). 

The SAMPIE experiment consists of an array of 
various solar cells, representing the present 
technologies. The objectives of the experiment are 
to investigate the arcing and current collection 
characteristics of these cells when biased to high 
potentials in a LEO plasma. These collection and 
arcing measurements will be made with the cells 
biased up to ± 600 V facing the ram and wake 
directions. 

In LEO the Orbiter’s potential will change so that 
the net current to the Orbiter from the plasma is 
zero. The potential at which this occurs is defined 
as the Orbiter’s floating potential. With the 
SAMPIE cells biased to +600 V and facing the ram 
direction there is a possibility that the Orbiter’s 
floating potential will be driven highly negative to 
balance and cancel out the incoming electron 
current. 

In order to better design the experiment so as to 
avoid possible arcing damage to the Orbiter, 
NASCAP/LEO was used to model the Orbiter’s 
electrostatic charging and obtain possible Orbiter 
floating potentials under different experimental 
configurations. From the available data (ref. 4), we 
can infer a floating potential of about -70 V for 
Skylab. This floating potential did not cause any 
problems during Skylab operations so -70 V was 
used as an acceptable floating potential. 


II. NASCAP/LEO SIMULATION 


First a finite element model of the Orbiter was 
created (see Fig. 1) using PATRAN® (a registered 
trademark of PDA Engineering), a commercially 


655 


available 3-D mechanical computer-aided 
engineering software system. The Orbiter is 
modeled as being a dielectric object whose only 
conductors are its main engine and thruster nozzles. 
The SAMPIE experiment is placed in the 
approximate center of the bay. 

The SAMPIE experiment itself is modeled as a box 
with the dimensions (.45x.45x.25 m) of the actual 
experiment and all of the top plate defined as a 
conductor. The top plate has no individual features 
such as solar cell assemblies or other experiments. 
This is because NASCAP/LEO’s resolution can’t 
distinguish individual features on the plate and still 
include the Orbiter in its computational grid. 

NASCAP/LEO is a modular code. Each module is 
a program, or collection of programs, which solve 
a particular aspect of the spacecraft charging 
problem. A call to the CURRENT module, for 
example, will calculate currents from the plasma to 
the spacecraft. Other available modules are 
RDOPT and IPS. In the RDOPT (Read Options) 
module the user can input parameters such as 
plasma temperature and density, spacecraft speed 
and conductor potentials among others. The IPS 
(Initial Potential Specification) module calculates the 
electrostatic potentials of the spacecraft’s surfaces 
and it’s surrounding space environment. To run 
any given simulation one calls each of the modules 
individually. 

The floating potential of the Orbiter with the 
SAMPIE experiment in operation was calculated 
using the RDOPT, IPS, and CURRENT modules. 
The procedure is straightforward. One performs 
several NASCAP/LEO code runs varying the 
Orbiter’s potential until the net current to the 
spacecraft is negligible. The potential at which this 
occurs is then taken as the Orbiter’s floating 
potential. 

This procedure would only take into account sheath 
generated particles but by including three QUICK, 
CHARGE, POTENT cycles in the simulation we 
can take into account ambient particles as well. 
QUICK, CHARGE, and POTENT are other 
modules available from the NASCAP/LEO code. 
All the computer runs for this paper were done on 
a Celerity 1200 mini computer running Accel 4.2 
UNIX. Further details on this procedure or about 
the NASCAP/LEO modules can be found in the 
NASCAP/LEO User’s Guide (ref. 5). A sample of 
NASCAP/LEO input is shown in Fig. 2. 

At the end of each simulation a CURRENTS utility, 
not to be confused with the CURRENT module, is 
run. The output from CURRENTS consists of the 
electric current values to the Orbiter/SAMPIE 


surfaces. This output is divided into material and 
conductor surfaces. In the present simulations one 
can read individual current values to the Orbiter, 
thruster nozzles, bay area, body (wings, 
empennage, cabin), the top of the bay doors, and 
the SAMPIE plate as well as the total current to the 
Orbiter/SAMPIE object. 

In this paper, the simulations consist of the four 
experimental configurations listed below: 

Case 1: SAMPIE in the Orbiter’s ram, biased to 
+600 V. 

Case 2: SAMPIE in the Orbiter’s ram, biased to 
-600 V. 

Case 3: SAMPIE in the Orbiter’s wake, biased to 
+600 V. 

Case 4: SAMPIE in the Orbiter’s wake, biased to 
-600 V. 

All SAMPIE biases are with respect to the Orbiter 
potential. Orbiter potentials are with respect to 
plasma ground. 


III. FLOATING POTENTIAL 
DETERMINATION AND SIMULATION 
RESULTS 


Case 1 is the most critical. With the SAMPIE 
experiment biased highly positive and facing the 
incoming ram particle flux, one can expect 
SAMPIE to draw large negative currents from the 
plasma. To cancel out this current the Orbiter will 
charge negatively in order to repel electrons and 
attract ions from the plasma. Depending on the 
magnitude of these currents the Orbiter may charge 
highly negative, thus exceeding safety limitations 
and interfering with the successful completion of the 
experiment. 

When SAMPIE is biased negative and facing the 
ram direction it will collect ions proportional to the 
ram ion flux on its surface. Ram ion flux for LEO 
is in the order of lO^-lO' 3 A/m 2 (ref. 5) so one 
would expect small currents for case 2, thus a low 
floating potential. 

One would also expect low floating potentials for 
cases 3 and 4 because of the reduced plasma density 
due to wake effects. A spacecraft flying through 
the plasma at a typical LEO velocity of 7500 m/s 
creates a region behind it in which both electron 
and ion densities are reduced in comparison with an 
undisturbed plasma. This spacecraft velocity is 
about six times larger than the ion thermal velocity 
(using a .1 eV oxygen ion) so a spacecraft would 
travel a distance equivalent to several of its own 
radii before the ions could fill in the region behind 
it. Electrons are more mobile and can fill in the 


656 


I 



Fig. 1: NASCAP/LEO model of the Shuttle orbiter with the SAMPIE experiment on the 
center of the bay. 


rdopt 5 

temperature .1 
rbo l.lell 
eirlim 0.1 

pcond 1 -200 
bias 2 600 

satvel 7500 0 0 
ionmass 16 amu 
sbeatb boundary 1 
end 
wake 

ips 5 

all matl -.3 

end 

quick 

charge 

potent 

quick 

charge 

potent 

quick 

charge 

potent 

cunent 

end 


# Read computational options 

# Plasma temperature (eV) 

# Plasma density (#/m*) 

# convergence error parameter 
for IPS module 

# Sets conductor 1 to -200 V 

# Biases conductor 2 600 V positive with respect 
to conductor 1 

# Object speed (x y z) m/s 

# Oxygen ion mass 

# Sheath defined at the 1 V contour 

# End of computational options 

# Calculates reduced ion densities due to wake 
effects 

# Initial Potential Specification 

# Assigns a surface potential of -3 V to material 
mat! as an initial guess 

# End of IPS options 


# 3 QUICK, CHARGE, POTENT cycles 


# Calculates currents to spacecraft 

# End of NASCAP/LEO run 


Fig. 2: NASCAP/LEO sample input. 


657 






wake region more rapidly. However this is limited 
by the space charge of electrons already present in 
the wake, so for most of the wake region the 
electron and ion densities are comparable (ref. 5). 
Results of measurements done by Murphy et al. 
(ref. 6) from within the Orbiter bay indicate a 
decrease of 3 orders of magnitude in electron 
density as a conservative estimate for the near wake 
region. 


Case 1: SAMPIE in the ram, biased +600 V with 
respect to the Orbiter 


As the Orbiter potential increases from 0 V to 
-400 V we see that currents to the Orbiter body 
and the top of the bay doors are negligible 
compared to the other currents, see Fig. 3. The 



Fig. 4: Case 1, currents to spacecraft vs. Orbiter 
potential. 



Fig. 3: Case 1, currents to spacecraft vs. orbiter 
potential. NET is the net current to Orbiter. 

door tops are shielded by SAMPIE’ s sheaths most 
of the time. The Orbiter/SAMPIE object collects 
current mainly through the Orbiter nozzles, the 
SAMPIE plate, and the bay area. The majority of 
the ions are collected by the nozzles while the 
electrons are collected by the experimental plate and 
the bay. The bay area is a dielectric surface but the 
fact that it is moving into the ram and that 
SAMPIE’s sheaths focus charge into the bay will 
allow it to collect charge up to approximately 
-2.4 mA, from then on it will not collect larger 
currents. 

At an Orbiter ground potential of -350 V the 
currents collected by the nozzles cancel out the 
current collected by SAMPIE, see Fig. 4. However 
at this potential the bay is charged up to about 
-3.4 mA. There is no positive charge large enough 


to cancel the bay charge. The positive current 
collected on the body is on the K^A range and 
these currents will not flow through the dielectric 
body to the bay. From these results one might infer 
a floating potential of about -350 V for case 1 . 

The floating potential may not be as large in reality 
because only a small area of the plate (some cell 
interconnects) will be biased to +600 V relative to 
the Orbiter instead of the whole plate surface as the 
simulation assumes. These cell interconnects would 
then be the effective collecting area. Assuming that 
the sheath through which SAMPIE current is 
collected scales proportionally to the conducting 
area of the plate we can scale the currents to 
SAMPIE by reducing the plate area. One can use 
these new currents to obtain a better estimate of the 
Orbiter/SAMPIE floating potential by reducing the 
currents to SAMPIE by a factor of actual collecting 
area vs. total experimental plate area. This should 
provide an estimate of the actual current collected 
by the experiment. These currents are then plotted 
to obtain a new I/V curve from which one can 
deduce a floating potential. 

First it is important to verify if reducing the area 
of the SAMPIE plate on the simulation by a given 
factor reduces its current collection by a similar 
factor. Upon inspection of the Orbiter/SAMPIE 
object it can be seen that one can reduce the area of 
the experimental plate to half of its original value 
and still be within the margins of NASCAP/LEO’s 
grid resolution. One can use this second 
Orbiter/SAMPIE model in a simulation and 
compare its currents to the original model. If the 


658 



currents to this new SAMPIE object are 
approximately half their original value the approach 
is correct. 


The new Orbiter/SAMPIE object will be referred to 
as SAMPIE2, where SAMPIE2’s experimental plate 
is one half the area of SAMPIEl’s plate model. In 
the SAMPIE2 simulation at high voltages the ratio 
of currents SAMPIE2/SAMPIE1 is very near to .5 
as was previously assumed while the currents to the 
nozzles remains the same (see Fig. 5). At low 




Fig. 6: Case 1 floating potential determination. 


experimental array beyond which using the shuttle 
becomes impractical and possibly hazardous. The 
simulation shows that a plate .2 m 2 in area can not 
be biased to high positive voltages without driving 
the shuttle’s floating potential highly negative. 


Fig. 5: Case 1, ratio of currents to spacecraft. 
SAMPIE2 vs. SAMPIE1 simulations. 

voltages the ratio is between .4 and .5 which is not 
in bad agreement with our assumption. It can also 
be seen that for an Orbiter ground potential smaller 
than -300 V the ratio of current to the bay is . 1 . So 
if one reduces the collecting area, the bay currents 
will become negligible. One can thus be reasonably 
confident in scaling the currents to the SAMPIE 
plate by an appropriate area factor. 

Assuming a worst case in which the whole surface 
area of the cell will act as a conductor, i.e. the cells 
will be "snapped over", a likely possibility with this 
high positive bias. With four Space Station 
Freedom type solar cells as the base line there is a 
total surface area of about 2.48xl0" 2 m 2 . The ratio 
of this area to the original SAMPIE plate area is 
approximately .124. Reducing the SAMPIE 
currents by . 124 and graphing them (see Fig. 6) one 
finds a floating potential of about -20 V. 

The NASCAP/LEO simulations therefore predict 
that when the SAMPIE experiment is biased 
+600 V with respect to the Orbiter and facing the 
ram, the shuttle’s floating potential will be in the 
range of -20 V. One also sees that for this type of 
experiments there is a limiting size to the 


Case 2: SAMPIE in the ram, biased 
-600 V with respect to Orbiter 


One proceeds in the same manner as described 
above. Since this case is not expected to be critical 
the original Orbiter/SAMPIE model (SAMPIE 1) 
may be used. 

One expects low currents for this configuration and 
the simulations bear this out. The Orbiter ground 
voltage was changed from -50 V to +300 V. At all 
voltages SAMPIE current collection was small, 
around 6.5x10^ A to 8.2xl0 4 A. The sheath is 
localized around the SAMPIE box and does not 
affect current collection as in case 1. In this case 
the Orbiter connects to the plasma through the 
nozzles as if no experiment were present therefore 
the net current to the Orbiter consists of the 
nozzle’s current. 

The floating potential for case 2 can be calculated 
to be in the range of -10 V, see Fig. 7. 


659 



CASE 4: SAMPIE in the wake, biased 
-600 V with respect to Orbiter 



Fig. 7: Case 2 floating potential determination. 


CASE 3: SAMPIE in the wake, biased +600 V 
with respect to Orbiter 

SAMPIE collects electron current but only on the 
order of lO^A which is small when compared to the 
10 3 A to lO^A nozzle current. So once again the 
net current to the Orbiter is the current to the 
nozzles. This current is zero for an Orbiter potential 
between -5 V and -2 V (see Fig. 8). 



Fig. 8: Case 3 floating potential determination. 


As before the only connection to the plasma is 
through the Orbiter nozzles. The SAMPIE plate in 
the wake is a poor ion collector. In Fig. 9 it may 
be seen that the Orbiter will float at about 0 V. The 
CURRENTS output seems to indicate it will float 
slightly positive between 0 V and +1 V. 



Fig. 9: Case 4 floating potential determination. 


III. CONCLUSIONS 


The NASCAP/LEO simulations predict that while 
the operation of the SAMPIE experiment will have 
an impact on the Orbiter’ s floating potential, it will 
not be a serious one. A worst case of -20 V has 
been predicted which is within the -70 V mentioned 
before as an acceptable floating potential. They 
also indicate possible limitations in similar 
experiments, for example, the same experiment 
with an active array collecting area of .2 m 2 , biased 
highly positive, would drive the shuttle’s potential 
to unacceptably large negative voltages. The 
SAMPIE experiment will not have this problem 
because the biased area is small and the platform 
upon which it is mounted, i.e. the Orbiter, has good 
contact with the plasma via the thruster nozzles. 
However, it is imperative that the Orbiter nozzles 
not be in the Orbiter’ s wake during the SAMPIE 
experiment for this will decrease the nozzles 
electrical contact with the plasma. Other high 
voltage experiments mounted on platforms which do 
not have a large exposed conductive area may 
charge up to large potentials which may then 
interfere with the experiment’s operation. 


660 


I 






REFERENCES 


1. Mandell, M.J., Katz, I., "High Voltage Plasma 
Interaction Calculations Using NASCAP/LEO”, 
AIAA-90-0725 , 28th Aerospace Sciences 
Meeting, Reno, Nevada, Jan 8-11 1990. 

2. Chock, R., "NASCAP/LEO Simulations of 
Space Station Cell’s Current Collection", 
unpublished. 

3. Hillard, B.G., "The Solar Array Module Plasma 
Interaction Experiment (SAMPIE)", SOAR ‘90 
Proceedings NASA Conference Publication 3103 
Vol. II, Albuquerque, New Mexico, Jun 26-28 
1990. 

4. Woosley, A.P., Smith, O.B., Nassen, H.S., 
"Skylab Technology - Power Systems”, THE 
SKYLAB RESULTS, AAS Publication Office, 
Tarzana, California, 1975, 559. 

5. Mandell, M.J., Davis, V.A., "User’s Guide to 
NASCAP/LEO", Draft 

6. Murphy, G., Pickett, J., D’Angelo, N., Kurth, 
W.S., "Measurements of Plasma Parameters in 
the Vicinity of the Space Shuttle” , Planet. Space 
Sci., 34, 10, 993 


661