ALAA 93-4764

SEDS Deployer Design

and Flight Performance
Joseph A. Carroll
Tether Applications
1813 Gotham Street
Chula Vista, CA 91913

AlAA Space Programs

and Technologies Conference
and Exhibit

For permlssbn to copy or republlsh,contact the Amerlcan lnstnute of Aeronautics and Astronautics
370 L'Enfant Promenade, S.W., Washlngton, D.C. 20024
 AlAA paper 93-4764:

SEDS Deployer Design and Flight Performance

Joseph A. Carroll'
Tether Applications
Chula Vista, CA 91913
(619) 421-2100, -2111 fax
Abstract
The Small Expendable Deployment System (SEDS) was w. I SEDS-IPayload
conceived as a complement to the Tethered Satellite
System deployer for use when tether retrieval is not
required. This paper reviews the history, design, and
capabilities of SEDS, and discusses its flight performance,
based on data collected during the successful SEDS-1
flight experiment.

Introduction: SEDS Project History

In 1983, the author developed the concept of a small
expendable deployment system (SEDS) and wrote an
SBIR proposal to develop it at Energy Science
Laboratories (ESL). This led to SBIR Phase I and I1
contracts with NASA Marshall Space Flight Center
(MSFC)'. That work further stimulated NASA interest
4
in SEDS and led to two follow-on contracts: to define the
SEDS-1 flight experiment' and to design and fabricate
flight hardware for SEDS-1. In parallel with this work,
MSFC personnel developed a flight computer to control
SEDS and collect flight data. In 1989, the author split
off from ESL to form Tether Applications, which has
assisted MSFC in integrating and flying SEDS-1.

NASA was initially interested in shuttle-based SEDS

experiments, but the focus shifted to expendable vehicles
after the Challenger accident. Delta/GPS launches
seemed suitable because they left the Delta second stage Figure 1.SEDS Deployer and Payload Mounted on Delta
in suitable orbits and had significant payload margin.
The Air Force developed a policy encouraging use of this Evolution of SEDS Design
payload margin for government-sponsored payloads on a
non-interference, marginal-cost-reimbursement basis'. Figure 1shows the SEDS flight deployer mounted on the
This led to a NASA commitment to fly SEDS-1, SEDS-2, Delta. Figure 2 shows the brake in cross-section, and
and the PMG (Plasma Motor Generator) on Delta/GPS Figure 4 the deployer. The best way to explain many key
missions. AU three mount in the same place on the features is to describe the evolution that led to them.
Delta, with the same interface. SEDS-1 flew on March
29, 1993, and the PMG on June 24, 1993. Both were SEDS was conceived as a complement to the NASA TSS
successful. SEDS-2 is scheduled to fly in March 1994. (Tethered Satellite System), for cases in which tether
Other missions await determination of Delta payload retrieval is not needed. Simulations showed that
margins, launch schedules, and integration costs. deployment at low tension had two significant benefits in

- * Proprietor, Tether Applications; AIAA member.

Copyright01993 by Joseph A. Carroll. Published by the
payload boosting or deboosting: it allowed use of a
simpler brake, and it induced a wide libration that nearly
doubled the deltaV provided by a given length of tether.
AIAA with permission.

1
The initial SEDS design had a 6x4 "ais-cross' winding The barberpole brake allowed us to eliminate squeegee
inside a shuttle GAS canister. The tether paid off one flaps and bags, but required another way to hold the
end of the Winding, with spring-loaded flaps pressing winding in place during handling and launch vibrations.
lightly against the winding to guarantee orderly payout. We used a thread tie-down at each end of the winding,
During deployment this caused large tension variations, plus a third tie-down to adjust the length outside the 4
but the absolute tension was low enough that this was not deployer. The tie-downs reliably hold the package in
considered a problem. Simulations showed that ramping place during launch vibrations, but after payload ejection
tension up near the end would save several times the they break at tether tensions of 15-30 newtons.
deployer's weight in RCS propellant, so we substituted
gas-bags and a gas supply for the spring-loaded flaps. Deployment tests done without active braking showed
that the tension term increased as the tether
Deployment tests in air and in vacuum showed that the unwound. Thus a small core maximizes passive braking
tension had a fixed component and a velocity-squared at the end of deployment. Removing the squeegee flaps
component, with the V2 term dominant at deployment and bags reduced but did not eliminate high-frequency
velocities above roughly 5 m/s. This sensitivity seemed tension variations. We found a way to radically reduce
useful because of its potential for damping undesired them late in deployment, when tension is highest. We
deployment dynamics without any need for measurement put small flanges on the core and wound the innermost
or active control. Simulations showed that this passive part of the winding parallel fashion in slightly conical
damping would be far more effective if it could be layers. The parallel pattern reduced the axial cycle
amplified near the end of deployment &e., if the brake frequency by a factor of 60,and the proper cone angle
were multiplicative rather than additive). This led us to made diameter effects cancel out axial effects. Residual
consider "tortuous-path" brakes, which impose a friction tension variations with axial position are small, as will be
that scales with the upstream tension. We selected the seen from the flight data. Figure 3 shows the effect on
"barberpole" design shown in Figure 2 because it is easy high-frequency variations. It shows 8 seconds worth of
to heat sink and has a nearly uniform exponential 62 Hz tension data from a July 1992 deployment test, at
sensitivity over a very wide range. Braking is applied by the time the winding changed from criss-cross to conical.
a small stepper motor which drives a worm gear at the
base of the brake post. A tether guide in the worm gear
wraps tether spiral-fashion around the post. Each wrap
around the post roughly triples the outboard tension.

I Tension: 0-1.5 newtons (62 Hz data) I

Figure 3. Tension due to Criss-cross and Cone Windings

Other SEDS Design Features

Deployed length and deployment rate are determined by
noting the time at which each turn deploys. The top end
of the core holds infrared emitting diodes that point at
phototransistors mounted on the top of the cansister. As
each turn unwinds, it shadows one phototransistor, then
the other. Set-reset logic distinguishes turn counts from
transverse tether oscillations, which typically add about
1% spurious counts on each channel. Other logic detects
a channel failure and switches to counting turns on the
other channel. The turn-counter was error-free on
SEDS-1: all 46,216 turns wound onto the package were
deployed and counted. Inspection of the time intervals
indicates that all counts were physically meaningful.

The SEDS-1 brake law was based on turns unwound; the

SEDS-2 brake law includes an approximate computation
of length. Post-flight data analysis uses a computer
record of the wound length at each turn, and includes
From Dcploycr
compensation for tension and temperature differences
-
Figure 2. Cross-section of SEDS "Barberpole" Brake between winding and deployment conditions.
 The sensor holder in the tap of the core also contains a Instead of modifying the brake for very heavy payloads,
reflective photo-pair that detects when the last 5% of the an alternative is to deploy at low tension, closer to the
tether begins to deploy. This is used as the brake enable horizontal. This greatly reduces brake loads and provides
sensor if the turn-counter fails. The sensor holder also slightly larger deltaVs, at the cost of modest increases in
v
contains a themister for core temperature measurement. deployment time and micrometeoroid risk.

A side view of the deployer is shown below in Figure 4. In addition to the brake proper, the brake assembly also
The deployer bolts to a shelf mounted on the side of the includes a running-line tensiometer, which uses an LVDT
Delta. It is held in place by four 1/4-28 bolts in a 2.25" load cell that measures small displacements of a sprung
bolt circle, plus (on SEDS-1) four 10-32 bolts near the tether guide. Interchangeable load cells with different
rim. The baseplate can be modified for other patterns. spring rates allow a selection of full-scale tensions from
0.15 to 150 newtons, all with 1 mm full-scale travel.
SEDS-1 and SEDSD both use a 7.5 newton range.

m The sprung tensiometer guide vibrates significantlyduring

deployment, so readings are taken at 500 Hz and digitally
averaged. A code is stored with the average to indicate
whether the digitized data includes out-of-range readings
(0 or US). Twice during deployment, 16 seconds worth
of raw data are also recorded.

Above the tensiometer is a redundant pyrotechnic bolt

cutter that has been qualified for TSS use on the shuttle.
Mounted on the cutter is a tether exit guide. Friction
1 I around that guide increases outboard tension by 1720%
10.0" per radian of tether bend around the guide, compared to
the tension measured by the LVDT.

The SEDS flight computer is an 8-bit computer based on

the NSC800 microprocessor. It was designed by Chris
Rupp of MSFC for use with the SEDS deployer, and was

. 0 0
built at MSFC. The same type of computer was used to
collect data on the SEDS-1 deployed payload. The
computer has a multiplexed E-bit A/D, 12EK or 512K of
mass RAM,and interface circuitry for the turn-counter,
tensiometer, tether cutter, and telemetry output to the
Figure 4. Side View of SEDS Deployer with Cabling Delta. The computer collects the following data:

On SEDS-1, the computer started the brake motor after 1. Time at which each turn of tether deploys (+ 1 ms)
41,026 turns had deployed (95% of the tether length). 2. Time brake-enabling sensor triggers (5% left)
The brake rate was fixed at 1 wrap/minute, and did not 3. Tension (500 Hz, summed down to 1 Hz)
use feedback to compensate for the faster-than-nominal 4. Temperatures (0.1 Hz): core, can, brake, computer
deployment. This plus a friction coefficient slightly lower
than expected decreased braking effectiveness. On The SEDS Tether
SEDS-2, braking will begin much earlier, and feedback
will be used to ensure the desired deployment rates. The SEDS deployer holds up to 7.5 kg of Spectra tether
or the same volume of other materials. The tether is
The brake can apply loads up to the full breaking pre-twisted during winding so it will pay off the spool
strength of the baseline 0.75 mm tether (750+ newtons). without residual twist. In cases where controlled payload
Much of the frictional heating goes into the tether itself spin is desired, this can be changed. The criss-cross part
rather than the brake. This gives the SEDS brake a total of the winding includes four different patterns (4, 5, 6,
heat-dissipation capacity > 100 W,14 times that needed and 7 turns per axial cycle) for package stability.
for SEDS-1 and 5 times that needed for SEDS-2, without
reaching a temperature that can damage the tether. The SEDS-1 tether was braided from Spectra 1000, a
Much higher capacities are possible by modifying a few highly oriented polyethylene fiber made by Allied Fibers.
v
brake components, or by wrapping the tether once or It has the highest strength-to-weight ratio and best
twice around a futed heat sink outboard of the brake. impact resistance of any commercially available fiber.
Spectra also has low friction, low abrasion sensitivity, low A dipole array was also embedded at the midpoint of the
dust generation, and better UV resistance than Kevlar. tether length. It used 10 m of nylon monofilament with
Spectra has limited heat resistance. It will fail if exposed 100 25-mm long aluminized sections at 100 mm intervals.
to steady-state aeroheating at altitudes near 130 km. (On Unfortunately, C-band radar tracking opportunities on
SEDS-1 ended before the array deployed, so we obtained d
SEDS-1, the tether probably remained attached to the
payload down to about 110 km due to a rapid descent.) no fight data on the array. But such an array should
A Teflon overbraid is needed if atomic oxygen exposure have a C-band radar cross section > 10 m2 at 7 discrete
exceeds roughly an hour at 150 km or a week at XW km. angles. Glint timing should allow tether cone angle
estimation to better than 1". Polarization of the return
We have recently learned that small amount: of coconut should provide complementary clock angle data. On
oil are used as a "spinning a i d in producing and braiding future experiments we may put 30-cm lengths of 42
Spectra. Deployment tests show that increased amounts AWG wire in the braid to ensure visibility to UHF radar.
of oil correlate with unacceptably high deployment
tensions and with an increased tension variation with Integration and Launch Operations
temperature. We are testing supercritical CO, extraction
to remove this oil. This should reduce outgassing and SEDS-1 experiment costs were dominated by integration-
may make SEDS useful in a wider range of applications. related activities, so integration and launch operations
will be discussed briefly as an guide to understanding
The SEDS-1 tether was 20 !an long and 0.75 mm in possible constraints on future SEDS experiments.
diameter. (There was room for another 3 km of tether
in the deployer, but 20 km was sufficient to de-orbit During Delta P stage fabrication over a year before
payloads from Delta orbits having perigees up to 280 flight, structural and electrical modifications were made
km.) The 0.75 mm diameter was large enough to limit to the 2"6stage electronics bay for SEDS. Holes were
micrometeoroid r i s k to about 0.1% during a one-orbit drilled and special cable harnesses were built up and
experiment. This diameter is also strong enough to allow installed. SEDS hardware was sent to Pueblo, CO for a
use with full Delta primary payloads, since the 900 kg fit-check when the 2" stage was nearly complete. Three

-
mass of the empty Delta 2"' stage limits the loads that months before launch, the SEDS computer was hooked
can be imposed on the tether. Hence SEDS-1 was a test up to the Delta during checkout in Florida. About 10
of an operationally useful system, not a scale model. days before launch, SEDS hardware was installed on the
Delta, checked out, and powered down. It remained in
SEDS-1 tether fabrication involved braiding 8 strands of that state until the primary Delta mission was complete.
fibers into a bollow braid. After inspection, several such
segments were spliced together end-to-end to form a 20 In parallel with this, tests, simulations, and analyses were
km flight tether. Segments need not be identical. For done on the SEDS deployer and payload, and frequent
example, the first meter of the SEDS-1 tether was a thick meetings and telecons reviewed the status of interface
Kevlar braid. The part of the Kevlar exposed before and integration issues. To reduce these costs in the
deployment was covered by heat shrink. This provided future, it may make sense to make SEDS experiments
better heat resistance during baast and 3" stage firing. electrically autonomous, even though this requires
This section of tether was given a sinusoidal set when the rcliable protection against premature payload ejection or
heat s h r i i was cooled, to provide a weak spring to take radio transmission.
up slack before deployment. The heat shrink may also
increase passive damping of payload attitude oscillations. Launch occurred in a window set by the GPS satellite.
Three launch scrubs delayed launch a total of 11 days: 2
Objects can be embedded in the tether during braiding. for weather and 1 for Delta ground support hardware.
For example, the SEDS-1 tether had a 100 m length of Aftcr launch, the solid motors fell away in two groups, 1
solder embedded in the last part of the tether to deploy, and 2 minutes after liftoff. The 1" stage fell away a few
as a test of passive braking. In addition, 3-meter lengths minutes later, and then the fairing. Ten minutes after
of solder were embedded 1.3, 1.7, and 2.1 km from the liftoff, the 2"' stage reached a 185x185 km parking orbit.
end. Each section caused a 1/4 second doubling of the
tension while it deployed. This injected a tensile wave Near the first equator crossing, the 2"' stage raised its
into the 18-19 km of deployed tether. As indicated in apogee to 740 km provide a small assist to the 3d stage,
ref. 4, each wave was detected at the payload end about which is undersized for GPS Block 2 satellites. It then
2.6 seconds later. When the waves returned to the spun up and released the 3d stage, which boosted the
deployer, they appear to have been entirely absorbed by GPS satellite into a transfer orbit. The GPS satellite
its low dynamic impedance. (This can be seen in the circularized its orbit at apogee. A plume shield protected
flight data shown in Figure 7.) the short exposed length of tether and most of the other -
SEDS hardware from the 3" stage exhaust.
 After 3"' stage ignition, activity on the 2"dstage could no F i e 5, below, shows second-by-second length, rate, and
longer affecl primary mission success. However one item tension data, plus the times of various events. Tension is

- remained a propellant depletion burn to safe the stage.

On SEDS-1, this was done after the SEDS experiment,
but on SEDS-2, the depletion burn will be done fust, to
allow the experiment to be open-ended. Cold gas will be
shown on two scales with full-scale values of 0 5 and 8
newtons, to make the data readable over a wide range.
Various features shown on Figure 5 are discussed below.
The period after 4300 seconds is shown on an expanded
horizontal scale in Figure 6, and is discussed later.
retained for attitude control during the experiment.

SEDS-1Mission Scenario and Performance The tension was non-zero before deployment. This was
due to the weak heat-shrii "spring". Tie-down breakage
Shortly before the first apogee of the 185 x 740 km orbit, induced 3 large but brief tension-spikes which are
the Delta Z* stage rolled until the SEDS payload faced estimated to have had peak tensions of 15-40newtons
earth. It then provided power to the SEDS computer, at based on ground tests. After payload ejection, the Delta
3720 seconds after launch. The SEDS computer took 0.7 rolled until the SEDS deployer was on the bottom. The
second to initialize. (Experiment times below are based average tether tension was only 0.03-0.04newtons, but
on this SEDS computer time.) A minute later, the Delta this was enough to reduce the deployment rate from 1.6
fued pyrotechnic cutters that severed the bolts holding to 0.75 m/sec within 20 minutes. The slight reduction
the Marman clamp together. Four springs ejected the shown in measured tension during the fust lo00 seconds
SEDS payload at 1.6 m/sec. Payload ejection broke the may be due to driit in the tensiometer electronics or it
three tie-downs holding the tether inside the deployer. may be real, due to reduced effects and/or tension
The resulting jerks induced a payload attitude rate of averaging through increasing periods of slackness. (The
several deg/sec. This induced payload oscillations about turn-count data indicate intervals between turns of up to
the tether attachment point, which caused transverse 3 seconds.) After 1500 seconds, orbital dynamics effects
tether oscillations and significant variations in deployment became dominant and deployment sped up. This made
rate. These lasted for about 3ooo seconds. The rate slip-stick behavior less common and increased VZ
variations were not predicted before flight, but have since damping. The result is a rapid reduction in deployment
been reproduced in detailed simulations by John Glaese.5 rate scatter, starting near 2000 seconds.

.:
Y:
<:,-.
,., .
Unwinding rate: 0-70 tumdsec ,e

Tension: 0-0.5 newtons 0-8 newtons ~

. - - w 7.':
. -.. , .
' .

]+Start deploymelit Dipole+/ Rraking A+End Cut +I

I Winding panern: cnss-cross, 4 1urnsJcycle
1
-5 - - ) 6-t7JConel
. ..

0 sec 1OOO zoo0 3000 4000 500(1

v

Figure 5. SEDS-1 Deployer Flight Data (I-sec averages)

Perigee oocurred near 2900 seconds. Air drag on the The transition from aiss-cross to conical parallel wind
tether near 185 km was enough to induce a bow in the occurred at 4380 seconds. The deployment rate was
tether. This caused the slight bulge in the deployment slowly approaching its peak of U m/s. The conical
rate w e that is centered on 2900 seconds. layers of this final part of the winding caused a large
triangular-waveform variation in turncount rate. The W

At 3450 seconds, the criss-cross Winding pattern changed amplitude was small at first, since it involved partial
from 4 turns per axial cycle to 5. This has little effect on layers. The amplitude grew and the frequency dropped
deployment tension. At 3665 seconds, the Delta ZPdstage as the conical layers reached further along the axis.
switched from pointing in the flight direction to a rate- Inspection of the data indicates the tension does not vary
damping mode, to limit cold-gas usage as tether tension much on the 10-second time scale of a typical axial cycle.
increased. Shortly thereafter, the midpoint of the tether This indicates that the cone angle is about right. (By
deployed with its embedded dipole array. Two adjacent comparison, deployment tests with cylindrical rather than
1-second tension values stand above the rest. This is conical layers show a relative variation in tension that is
compatible with lab test data which indicate a 25% comparable to that shown here in the unwinding rate.)
tension increase during dipole deployment.
By 4500 seconds after ejection, tension had increased
Shortly after dipole deployment, a very low frequency passively by a factor of 20, enough to overcome the
step-like oscillation began in the deployment tension, increasing orbital dynamics effects and prevent further
superimposed on an otherwise smooth tension ramp. acceleration of deployment. The three short lengths of
This is most clearly visible near 4ooo seconds, but can solder embedded in the tether deployed at 30-second
also be seen near 4500 seconds. The slight drops at the intervals, injecting tensile waves into the tether which
start of each new winding pattern were expected, but the were damped when they returned to the deployer.
rest of this variation needs further study.
Figure 7 shows 16 seconds of high-rate tension data,
The rest of the experiment is better studied on Figure 6, starting just before deployment of the 1" solder segment.
which shows the same data beginning at 4300 seconds, The 90 Hz data shown eliminates the free resonance of
stretched horizontally by a factor of 4. The data are still the tensiometer by averaging groups of 5.55 samples, with
one-second averages, but they are spread further apart. fractional weighting used for values at group boundaries.

-z
\
\
\ Deployment rate: 7-13 meters/sec
................................................................... . . . . . ..........................................................

...........................

Tension: 0-8 newtons. '

.........
+End reached

45da sec SO00 sec 5500 sec

Figure 6. SEDS-1 Flight Data (4300 to 5700 seconds)

~~~~~~~

i ~~ '

4485 sec 4490 4495 45uu ~

Figure 7. 16-Second High-speed SEDS-1 Tension Sample, Showing Deployment of First Solder Segment

6
 At 4663 seconds, the SEDS computer began to apply Final Notes on SEDS Capabilities
adive braking, as shown by the bars at the bottom of

- Ftgures 5 and 6. The brake ramped tension up to > 3

newtons. When the last 100 meters began to deploy,
with its embedded solder, the tension doubled again. But
tensiometer vibrations caused some readings to be off-
Mission-specific SEDS capabilities and constraints are
best assessed by using Beadsii, a tether dynamics
simulation program for PCs written in Borland Pascal'.
scale even though the average may have been on-sale.
This resulted in under-estimates of the actual tension. Many SEDS applications involve low-tension deployment
Seconds that included off-sale tension readings are followed by 3-6 minutes of braking and then a swing to
marked at the very bottom of Figure 6. The faster-than- the vertical. The largest deltaV OCCUTS if the payload is
expected deployment and lack of feedback in the brake deployed at low tension and released in the middle of a
law caused the end of the tether to be reached at a wide prograde swing (a forward swing if deploying up,
deployment rate of 7 m/s. This caused a large tether and a backward one if deploying down). This provides 4
stretch followed by a rebound. The peak tension was of m/s of deltaV per km of tether length, a deltaV which is
order 20 newtons, assuming a half-sine tension profde split between payload and host vehicle inversely with
with the right total impulse. This is 4 times the peak their mass. The deltaV is in plane, and has the largest
tension expected during the SEDS-1 experiment, but still effect on altitude 1/2 orbit after release. Near-vertical
only about 2% of the breaking strength of the tether. stabilization is also possible. It will be tested on SEDS-2,
which is now scheduled for March 15'94. SEDS-2 braking
During braking the tether began to swing down toward will start early in deployment and will w e rate feedback.
the vertical. The mean tension increased during the
swing, but due to signiticant damping in the tether plus A short tether orbit life minimizes contribution to space
energy transfer to other modes, the peak tension during debris. In a boost operation, the tether should be cut
later bounces was lower than that experienced during the loose at the vehicle end about 30 minutes after it is cut
fust bounce. Data collected by the SEDS-1 payload loose from the payload. This releases the tether near
remained on-scale during these bounces. the midpoint of a retrograde libration and can reduce the
free tether's orbit life from days to a few hours.
The tether cut time was based on a fixed delay of 1040
seconds after the start of active braking. This resulted in On Delta-based experiments, the Mannan damp limits
C.,
payload release near but not at the vertical. This simple the deployed payload mass. The existing clamp can hold
logic actually gives lower reentry location dispersions 25-35 kg, depending on CG offset from the separation
than a more sophisticated logic that tries to cut the tether plane. The clamp fits between the SEDS deployer and
right at the vertical. Release took place at 5611 seconds, an adjacent electronics bay access door, so envelope
20 seconds less than one orbit after SEDS powerup. constraints may be more severe than mass constraints.
Other mounting schemes might relax these constraints.
Since release occurred near apogee and the m k u m For example, MDAC built a clamp for Losat-X which
orbit change is half an orbit later, most of the SEDS can handle 60-80 kg, but this clamp mounts centrally
orbit change was to perigee. The Delta's perigee was between two E-bay access doors. Using it with SEDS
raised several km, and the perigee of the payload and requires mounting the SEDS hardware inside the
attached tether dropped well into the earth's atmosphere. payload. (A smaller SEDS deployer designed for this
The payload and tether reentered 1/3 orbit after release. purpose is now in fabrication.)
Pre-flight predictions of reentry location were accurate
enough that observers stationed in Mexico were able to SEDS can also boost primary Delta payloads. Since a
capture TV imagery of the nightstime reentry. primary Delta payload weighs much more than the empty
2ndstage, such a boost also provides a large deboost to
The above discussion of SEDS-1 flight performance can the 2"' stage. In cases where slightly negative payload
be summarized briefly as follows: margins exist, SEDS might be used to place a payload in
a higher orbit than otherwise feasible, while at the same
1. Full deployment occurred without problems time deorbiting the 2ndstage in a controlled manner.
2. The data was all received and all made sense
Economical recovery of reentry capsules requires a small
3. Unexpected transverse dynamics caused no problems
recovery zone. This can be compatible with a wide range
4. Pauses in deployment did not induce tether fouling of low-speed deployment tensions if that range brackets
5. SEDS damped axial tether dynamics very effectively a cusp value that causes a minimum reentry range. On
6. SEDS can be robust against tension estimate errors SEDS-I the cusp was just over 0.02 newton. The actual
i
7. The brake law needs feedback for controlled stops SEDS-1 tension was over 50% higher, but that increase
did not cause large reentry errors.

7
SEDS can act like a kite tail and stabilize a payload at References
the beginning of reentry, until the Spectra melts. And
even if the payload needs to cut the tether loose before 1. SEDS: The Small Expendable-tether Deployment
this, an intentional torque imbalance in the tether can System, Final Report on NASA SBIR Phase I1 Contract
provide a several-rpm payload spin about a vertical axis NAS8-35256, Energy Science Laboratories, Inc.,
e
by the time of release. Such vertical-axis spin is suitable December 1987. Available from NASA Marshall.
for payloads reentering 1/4 to 1/3 orbit later.
2. Final Report on SEDS Experiment Design Delinition
Table 1 shows SEDS capabilities with various Spectra Contract NASB-37380, Energy Science Laboratories, Inc.,
tether braid constructions that all have the same mass February 1990. Available from NASA Marshall.
and volume. The table takes credit for payload orbit
changes only. It assumes LEO orbits, a 53" libration 3. Space Systems Division CLM/CLFP/CWh' Policy for
angle, a host vehicle with 10 times the payload mass, and Secondary Payloads on Delta 11, August 1989. Available
a tether safety factor of 4. It also assumes that from CLMD, USAF HQ Space Division, Los Angeles
micrometeoroids with >1/3 the tether diameter will cut AFB,CA 90009-2960.
the tether upon impact. Man-made debris increases these
risks only slightly. Deployment tests have only been done 4. L. Me&, Jr., Engineering Performance of the SEDS-1
with 8x375 and 8x215 braids so far, but tests with 4x215 End Mass Payload, AlAA paper 93-4769, Space
and 24x650 braids are planned in 1994. Programs and Technologies Conference, Huntsville,
September 1993.
Table 1. SEDS Capabilities with Various Tethers
5. John Glaese, Comparison of SEDS-1 Pre-Flight
Braid Dia MaxL AV MaxF'ay Impact Risk Simulation and Flight Data, AIAA paper 93-4766,Space
-
mm - Km & %
! DavsLife Programs and Technologies Conference, Huntsville,
24x650 1.7 4.3 15 30,000 0.003 1,ooO September 1993.
8x650 1.0 13 46 3,300 0.032 80
8x375 0.75 23 80 1,070 0.11 23 6. Beadsim: Beadsim is an easy-to-use tether dynamics
0.41 simulation program for PCs, written in Borland Pascal
8x215 0.57 40 140 350 6
7.0. Available from NASA Marshall or Tether
4x375 0.53 46 160 270 0.56 4 Applications on a floppy disk containing source and
4x215 0.40 80 280 88 1.9 1 executable code, sample cases, and a help tile.
Besides the tether itself, the only changes needed for
SEDS to be able to provide a modest boost to even a full Bibliography
shuttle payload of 30,ooO kg are the tensiometer and the
brake control law: Simulations indicate that control gets Delta 11 Commercial Spacecraft Users Manual, MDC
easier as the payload mass increases. (The 26 kg SEDS- H3224A, McDonnell Douglas, July 1989.
1 payload was near the minimum feasible.) Standard
"two-fault tolerant" rules may require adding redundant Delta 11 Secondary User's Guide, available from NASA
sensors and cutters to SEDS, but this appears practical. Goddard, Huntington Beach (at McDonnell Douglas).

NASA Tethers in Space Handbook, Second Edition,

Acknowledgements prepared by SRS Technologies for NASA Headquarters,
Code MD, May 1989. Overview of tether applications
SEDS development was funded under SBIR contracts and tether dynamics, with extensive bibliography.
and follow-on contracts with NASA MSFC. George von
Tiesenhausen was contract monitor on the first contract,
and Jim Harrison has been contract monitor since then.
In addition to them, I would like to thank Jim Arnold,
Ivan Bekey, George Levin,Chris Rupp, Sid Saucier, and
J.R. Thompson for their support, and my colleagues
Chuck Alexander, Mike Carpenter, Jim Clinton, Kevin
Cross, Mike Fennell, George Henschke, Marianne
Knight, Matt Nilsen, John Oldson and Eric Pulliam for
all of their work on SEDS.