University of Southampton SESA3024 and SESA6055 - Mission to the Moon LEWIS - Lunar Educational Wide Imaging Satellite Alexander Godfrey LukasG¨ossnitzer James MacCalman Ahmed El Maghraby Alessandro Melis Nicole Melzack Reetam Singh Nikolay Tenev Aur´ elien Toussaint Maria Zaretskaya April 24, 2013
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
University of Southampton
SESA3024 and SESA6055 - Mission to the Moon
LEWIS - Lunar Educational WideImaging Satellite
Alexander Godfrey
Lukas Gossnitzer
James MacCalman
Ahmed El Maghraby
Alessandro Melis
Nicole Melzack
Reetam Singh
Nikolay Tenev
Aurelien Toussaint
Maria Zaretskaya
April 24, 2013
2
Abstract
The Lunar Educational Wide Imaging Satellite (LEWIS) is designed with a
primary mission objective to image the lunar surface for education and outreach
purposes. Secondary objectives are the detection of atmospheric dust, the
analysis of the lunar radiation environment and the characterisation of the
structure of the regolith. The mission lifetime in the lunar orbit is required to
be at least six months and the launch mass must not exceed 300 kg. After being
placed in a geostationary transfer orbit (GTO) by the Ariane 5 launcher, three
different lunar transfer methods are contrasted: a direct chemical transfer, a
chemical transfer via a weak stability boundary point and an electric propulsion
low thrust transfer. A preliminary spacecraft design study is conducted for each
transfer option using a concurrent design environment. The level to which the
designs fulfill the mission objectives, or exceeds them, is evaluated. The final
recommendation for the orbiter is to use the weak stability boundary method.
The main factors leading to this decision include the relatively short transfer
time (80 days) which will minimise any degradation and debris impacts to the
spacecraft, the extended mission lifetime of 1350 days (compared with the 298
days offered by the direct chemical method) and the final orbit plane which is
Table 2.2.3: Specifications for the Piezo Dust Detector (PDD)
3 Concurrent Design Approach
Every effort was made by the team to work concurrently throughout the design
process. All decisions were thoroughly discussed and made as a team. The team
was faced with a decision on whether to use the ESA SCDE concurrent design system
or GoogleDocs. It was decided that GoogleDocs is more adequate for the reasons
explained below.
The design team is 10 people. This is a large and international group and in order
to work concurrently, the team has to be flexible to work anywhere and anytime.
13
Using the GoogleDocs allows remote access to the spreadsheets so even members who
are working from remote locations can edit the spreadsheets and contribute to the
design process.
Using GoogleDocs makes the design process more dynamic as the changes to the
configuration of subsystems are immediately visible to the each subsystem engineer.
This increases the number of iterations and optimizes the design faster. This also
makes it easier to experiment with subsystems as the system level implications are
immediately visible.
Errors can be resolved a lot faster by using GoogleDocs rather than the ESA
SCDE system. Everyone can see the spreadsheets in GoogleDocs in real time and all
subsystems are always connected. This allows mistakes to be spotted immediately in
contrast to the SCDE system when the subsystems are merged every few days.
Figure 3.0.3: Budgets tab in the concurrent design spreadsheet
One of the problems with GoogleDocs was that is was easy to fall into circular
references. To resolve this problem, one segment of the circular chain was cut and
manually equated while manually keeping track of the error in the approximation.
Another way to deal with this was by automating the process by writing scripts
which size the system iteratively. This was done for the propulsion subsystem sizing
14 3 CONCURRENT DESIGN APPROACH
for the direct chemical and weak stability boundary methods.
In Figure 3.0.3 is an example of a spreadsheet. The user can see when the last edit
was made and by whom. The current users are displayed in the top right hand corner.
The tabs at the bottom show other spreadsheets. Spreadsheets are interconnected by
linking values.
Figure 3.0.4: Attitude control subsystem spreadsheet construction
Figure 3.0.5: The implemented designprocess loop.
Figure 3.0.4 shows how each sub
system spreadsheet was structured. The
Inputs section is linked to other engineers’
Outputs section. Below the Inputs
section is a Calculation section where
all the necessary computations are done.
On the right of the spreadsheet is a
Requests section where requests can be
made by other engineers. The Outputs
section displays all the values needed by
other engineers.
To make continuous optimisation iterations as a group we used the design process
shown in Figure 3.0.5. All decisions were discussed as a group. After the discussion
changes were implemented. Then a system analysis of the design was done. Cor-
rections were made as problems arose. Then the group discussed the new iteration.
15
4 Direct Chemical Transfer Method
4.1 Mission Analysis
4.1.1 Transfer
To determine the transfer ∆V requirement for LEWIS, patched conics method was
used as a first estimate in Microsoft Excel, results shown in Table 4.1.1. It was shown
that transfer ∆V is minimized when trans-lunar ejection is performed at GTO Perigee
and lunar encounter occurs when the moon is at its Apogee. Another independent
patched conics approximation was performed in MATLAB, producing transfer ∆V of
1155 m/s. To obtain a firm result, STK analysis was performed. It was found that in
December 2013, the intersection between lunar orbit and the Earth’s equatorial plane,
which is close to LEWIS’ initial orbital plane, occurred at its highest position in the
year 2013. The transfer consists of three burns: TLE, a mid-course correction burn
is then performed to achieve the desired inclination and periselene altitude (PA) of
90◦and 100 km, respectively. LEWIS is then allowed to coast to the periselene, where
TLI is performed to achieve the desired aposelene altitude (AA). The resultant ∆V
requirement has been found to be 1194 m/s, with 5 day transfer time.
ConfigurationTLE ∆V TLI ∆V Total Transfer ∆V Transfer Time
[m/s] [m/s] [m/s] [days]Perigee to Perigee 693.7 498.9 1192.6 4.58Perigee to Apogee 704.4 466.5 1170.9 5.40Apogee to Perigee 2457.9 412.7 2870.6 5.24Apogee to Apogee 2480.7 396.3 2877.0 6.10
Table 4.1.1: Transfer∆V Results using Patched Conics Approximation
4.1.2 Final Orbit
1
2
3
4
30
210
60
240
90
270
120
300
150
330
180 0
Orientation of Final Orbit for Direct Chemical Option
Orbital Radius [Moon Radius]
South - North
LEWIS Orbit
Figure 4.1.1: Final Orbit Achieved bydirect chemical Option. AoP = 320.6◦
This choice of transfer results in a
final orbit with the highest Argument
of periselene (AoP) amongst the three
options described in this report, as shown
in Figure 4.1.1. Although it is seen from
the spacecraft system requirements that
AoP of 0◦is desirable, it was decided that
no further maneuvres to obtain AoP of
zero will be performed, since similar pho-
tographs can be taken from this orbit,
and further maneuvres would increase
the ∆V requirement by upto a factor of
1.5.
16 4 DIRECT CHEMICAL TRANSFER METHOD
The near-45◦AoP results in strong gravitational orbital perturbation, resulting in
the highest stationkeeping requirement amongst the three options, of 340m/s/year.
Stationkeeping To achieve minimum mission duration of 6 months, the PA must
be maintained well above surface. The station-keeping strategy employed for this
option is to boost the PA and AA when either PA or AA is altered by tolerance of
10 km. The inclination, Ascending Node and AoP of the orbit are not controlled, as
controlling these would require large ∆V , and maintaining these offers no benefit to
the Science Objectives, as long as orbiter stays above the surface with high variation
in orbital altitude.
Simulation in STK shows that this strategy requires 340m/s of ∆V over 1 year
since Translunar Injection (TLI); the resultant Altitude over 1 year since TLI is shown
in Figure 4.1.2. Using all available propellant on board, the designed orbit can be
maintained for 9 months since TLI. Using measured data of the lunar radiation en-
vironment, the solar array was sized to provide minimum power at this time, as
discussed in Section 4.3.
Figure 4.1.2: STK simulation on PA and AoP of LEWIS when the orbit is activelycontrolled. Keeping PA at 100 km ± 10 km for 1year requires 340m/s. By linearinterpolation given the fuel carried, LEWIS will keep the station for 9 months inlunar orbit.
4.2 Chemical Propulsion Subsystem 17
4.2 Chemical Propulsion Subsystem
The chemical propulsion subsystem generates the thrust for orbit insertion, station-
keeping and includes Reaction Control System (RCS) thrusters as external Attitude
Control System (ACS) actuators. The top level requirements for the satellite wet mass
(≤ 300 kg) and the mission lifetime (≥ 6 months) drive the sizing of the propulsion
subsystem. It will be sized such that the wet mass attains 300 kg, utilising any
unclaimed wet mass for additional expellant to extend the mission lifetime. The
propulsion subsystem restricts the mission lifetime via the expellant lifetime, i.e. the
duration for which expellants for station-keeping and RCS functions are available.
The design objective is to maximise the expellant lifetime by configuring a propulsion
subsystem that stores the maximum possible expellant mass. The expellant lifetime
is used by the systems engineer to determine the mission lifetime.
Sizing the propulsion subsystem is a recursive root-finding process, with the satellite
wet mass as the objective function. The independent variable is the stored expellant
mass at arrival in the mission orbit. With these two values, and a description of the
propulsion subsystem configuration, the propulsion subsystem is sized and its mass is
determined. Consequently the satellite wet mass is computed. This process is repeated
until the wet mass converges to the band [300−∆m, 300] kg, where ∆m = 1/10kg is
a mass increment specifying the accuracy of the calculation and also acts as the step
size during iterations.
4.2.1 Propulsion Subsystem Sizing
The required thrust levels and total impulse of the Main Engine (ME) and RCS
thrusters are small, thus pressure-feeding is acceptable to provide the necessary inlet
pressures. A trade-off study contrasting pump-fed and pressure-fed systems has not
been conducted. The mass of the propellant and oxidiser tanks are computed via
fitting a first degree polynomial to points for the tank mass and volume obtained
from readily available surface tension tanks manufactured by Astrium [14], because
sizing the fluid capture mechanisms is outside the scope of this preliminary study.
The functions obtained for Titanium MMH and MON tanks are mt = 34.35Vt +
8.387 (R2 = 0.898) and mt = 34.70Vt + 10.51 (R2 = 0.995) for spherical and
cylindrical (with domes) surface tension tanks respectively. Similarly for Titanium
surface tension Hydrazine tanks the mass is found to be mt = 79.25Vt + 1.52 (R2 =
0.780). Residual oxidisers and propellants are estimated to be five percent of the
stored mass.
For the regulated pressure-fed systems the required mass of pressurant is computed
18 4 DIRECT CHEMICAL TRANSFER METHOD
according to [17] by:
m0 =plVl
RT0
(k
1− pp/p0
). (4.2.1)
Where m0 is the initial pressurant mass, pl and Vl are the gas pressure and volume
in the liquid tanks at their depletion respectively, R and k are the gas constant and
specific heat ratio of the pressurant, pp and p0 are the pressures in the pressurant tank
at depletion of the liquid and initially respectively and T0 is the initial temperature
in the tank.
This amount of pressurant provides the required inlet pressures for thrusters and
the ME at the end of life of the propulsion subsystem (i.e. at the depletion of the
tanks). A five percent margin is included in the pressurant mass for ullage and
residuals in the lines. The tanks for the gaseous pressurants (Helium or Nitrogen) are
sized for a given initial pressure as spherical shells with a constant thickness that is
determined via the ultimate tensile strength of Titanium, including a safety factor of
1.25. The initial pressure is 200 bar [15].
The propulsion subsystem plumbing (i.e. feed lines, valves, pressure regulators)
is accounted for by mplumbing = c mPS,dry with c ∈ [7, 17.5]%. The constant c is
modified to reflect the complexity of the propulsion subsystem, i.e. higher values
for bi-propellant systems than for dual-mode systems [18]. Propulsion subsystem
instrumentation and electronics is not accounted for explicitly, it is included in the
20% subsystem margin.
4.2.2 Propulsion Subsystem Configuration
The five different propulsion subsystem configurations under consideration are a mono-
propellant system (four MEs: Table 4.2.2 (f), RCS: Table 4.2.2 (a)), four mono-
propellant MEs (Table 4.2.2 (f)) and CG RCS (Table 4.2.2 (b)), a bi-propellant
system (ME: Table 4.2.2 (i), RCS: Table 4.2.2 (e)), a bi-propellant ME (Table 4.2.2
(i)) with CG RCS (Table 4.2.2 (b)) and a dual-mode system (bi-propellant ME:
Table 4.2.2 (l), mono-propellant RCS: Table 4.2.2 (a)). These configurations are
the conclusions of numerous trials on the basis of components the majority of which
is listed in Table 4.2.2. During development, procedures size each configuration si-
multaneously (using the same propulsion subsystem requirements as inputs) and the
resulting expellant lifetime is computed.
For the sizing of the propulsion subsystem, the most important system charac-
teristic is the dry mass of the satellite’s subsystems, as it determines (in combination
with the propulsion subsystem and structure mass) the mass allocation for expellants.
4.2 Chemical Propulsion Subsystem 19
40 45 50 55 60 65 70 75 80 85 90−200
0
200
400
600
800
1000
1200
1400
1600
1800
Satellite dry mass excl. PS and structure [kg]
Expella
nt lif
etim
e [days]
Bi−propellant PS
Dual−mode PS
Mono−propellant PS
Bi−propellant ME, CG RCS thrusters
72.86 kg
272 days
Figure 4.2.1: Effect of the choice of propulsion system configuration on the missionlifetime via the subsystem dry mass (excluding the propulsion subsystem andstructure). The selected dual-mode configuration is annotated.
Figure 4.2.1 illustrates this behaviour for four different propulsion subsystem con-
figurations (the monopropellant system with cold gas RCS is omitted).It is found that
for the subsystem mass of 72.86 kg a dual-mode propulsion subsystem delivers the
highest expellant lifetime (272 days). Additionally, the use of mono-propellant RCS
thrusters in the dual-mode configuration reduces the subsystem complexity compared
to a bi-propellant system. The configuration comprising cold gas RCS thrusters,
although delivering a similar expellant lifetime, is unattractive as it does away with
having a shared propellant tank for the thrusters and the ME. A shared tank allows
the thrusters to use a propellant amount different from the design value, e.g. in the
case of having to re-purpose them for station-keeping in the event of an ME failure or
if the RCS propellant consumption is higher than expected. Therefore the dual-mode
propulsion subsystem con figuration is selected for the direct chemical transfer option.
The dual-mode propulsion subsystem is found to exhibit the highest expellant lifetime
and is thus the selected propulsion subsystem configuration. It is comprised of the
Northrop Grumman bi-propellant engine Table 4.2.2 (l) running on Hydrazine and
Nitrogen Tetroxide and 12 EADS Astrium mono-propellant Hydrazine thrusters Table
4.2.2 (a). Both share a common propellant tank. The propellant and oxidiser are
20 4 DIRECT CHEMICAL TRANSFER METHOD
pressurised by a common Helium tank via separate lines and pressure regulators.
The impulse delivered by the ME during transfer follows from the expellant mass,
engine thrust and specific impulse as 3.24 × 105 Ns, i.e. only 19 % of the MEs total
impulse capability of 1.67×106 Ns. The remainder is available for station-keeping and
RCS functions. The RCS impulse delivered during transfer is negligible in contrast to
the 1.80× 105 Ns capability of a single RCS thruster. With the annual RCS impulse
requirement of 920 Ns it is clear that neither the lifetime of the ME nor the RCS
thrusters are constraining the mission lifetime.
A breakdown of the dry propulsion subsystem in terms of mass is given in Table
4.2.1.
Component Mass Allocation[kg] (%)
Oxidiser tank 9.74 27.6Propellant tank 10.05 28.5Pressurant tank 2.38 6.7RCS thrusters (12) 3.48 9.9ME 6.03 17.1Plumbing 3.17 9.0Pressurant 0.43 1.2Propulsion system dry 35.29 100.0(incl. pressurant)
Table 4.2.1: Mass breakdown of the dry dual-mode propulsion subsystem for thedirect chemical transfer option.
4.2 Chemical Propulsion Subsystem 21
Manufacturer
Ref.
Designation
Thru
stI s
pPro
p.
Ox.
Mix.
Inlet
Mass
Acc
um.
ratio
pres.
burn
life
[N]
[s]
[bar]
[kg]
[hrs]
(a)
EADSAstrium
[13]
Mon
o-PropellantThruster
1220
N2H
4n/a
n/a
0.29
50(D
Cdual
seat
dual
solenoidvalve)
(b)
Moog
[16]
SolenoidActuated
Thruster
3.5
71.5
GN2
n/a
n/a
14.8
0.022
16666
58-118
(c)
EADSAstrium
[12]
Bi-PropellantThruster
10291
MMH
N2O
4,0.35
70(single-seat
valve)
MON
(d)
EADSAstrium
[13]
HydrazineThruster
20224
H2H
4n/a
n/a
0.395
10.5
(e)
EADSAstrium
[12]
Bi-PropellantThruster
22290
MMH
MON
0.65
(f)
NorthropGrumman
[11]
Mon
opropellantThruster
MRE-15
66228
N2H
4n/a
n/a
18.96
1.10
9.12
(g)
Ampac
ISP
[16]
MONARC-90
90235
N2H
4n/a
n/a
1.00
(h)
EADSAstrium
[10]
Bi-PropellantEngineS400-12
420
318
MMH
N2O
4,1.65
103.60
8.3
MON
(i)
EADSAstrium
[10]
Bi-PropellantEngineS400-15
425
321
MMH
N2O
4,1.65
104.30
12.8
MON,
(j)
Ampac
ISP
[16]
MONARC-445
445
235
N2H
4n/a
n/a
1.60
(k)
Northrop
[11]
Dual
ModeLiquid
Apogee
Engine
471.5
322
N2H
4N2O
41
14.13
4.763
6.71
Grumman
TR-308
(l)
Northrop
[11]
HighPerform
ance
Dual
ModeLiquid
556.0
330
N2H
4N2O
41.06
15.85
6.033
0.83
Grumman
Apogee
EngineTR-312-100YN
Tab
le4.2.2:
Selection
ofengines
andRCSthrustersusedfortheconfigu
ration
ofpropulsionsubsystem
s.
22 4 DIRECT CHEMICAL TRANSFER METHOD
4.3 Power Subsystem
4.3.1 Requirements
To provide continuous power to the subsystems and payloads, their demands must
be accurately estimated. Since not all instruments are constantly operating at their
peak conditions, a power plan of the spacecraft subsystems must be compiled over one
average orbit. To accomplish this, a table is constructed containing all instruments
on-board in a row, and a time axis in a column in 5-minute interval over the course
of one orbit. Each cell in this table contains the power level from 0% to 100% of
the peak power, indicating the operational state of each instrument at a particular
time. This allows the calculation of the total instantaneous power requirement of the
spacecraft over the course of one orbit, plotted in Figure 4.3.1.
The average figure of the instantaneous power requirement was then found to be
115.2 W. To account for information lost by using discreet time, a contingency of 10%
is applied to this value. This is also to slightly oversize the power subsystem to ensure
the battery being fully charged after each sunlit phase. This, Therefore, gives the
subsystem power requirement (PSR) of 126.7 W. To carry on with sizing the batteries
and solar arrays, it is assumed that the spacecraft steadily consumes this amount of
power over the course of one orbit.
Figure 4.3.1: Power profile outlining the power demand of an average orbit and thewosrt-case sunlight condition.
4.3.2 PCDU
To provide the payload with stable electric power, the voltage and current fed into the
payload and subsystem must be regulated. To do this job, a Power Conditioning and
Distributing Unit (PCDU) is used. The schematics of power subsystem is described
in Figure 4.3.2.
4.3 Power Subsystem 23
Power produced by the array is regulated by a PCDU. Since the subsystem power
requirement is 126.7 W, and the solar array, as described below, only produces 180
W, Small Satellite Power System produced by SSTL can be used, which is scalable
upto 1.6 kW. [19]
This process occurs with some finite DC-DC efficiency, (µPCDU). Since the product
description for Small Satellite Power System did not include this efficiency, a value is
taken from a similar PCDU by Thales Alenia, which is 94%. [21]
25% of subsystem mass has been allocated for cables.
Figure 4.3.2: A schematic on the power subsystem. During daylight the solar arraypowers the payload and subsystems and charge the batteries, via a PCDU. Duringnight time, the batteries power the load via the PCDU.
4.3.3 Battery Sizing
It has been found that the power subsystems require (PSR) is 126.7 W. The battery
must be sized to enable satellite operation during all encountered eclipses. To do this,
the longest possible eclipse time is used. Given AoP of 320.6◦, the maximum eclipse
time (Teclipse) was found to be 3023 seconds. The minimum energy capacity of the
on-board battery Emin is therefore given by:
Emin =PSR × Teclipse
DoD × µPCDU
(4.3.1)
It was decided that VES 180 Li-ion batteries produced by SaftBatteries would be
used for their high specific energy of 175 Wh/kg. This battery is capable of achieving
60,000 cycles at 20% DoD. By linearly scaling this figure, it is found that even if the
battery were to be 100% discharged, it would be capable of achieving 12,000 cycles.
Since in the 10-month life the battery is going to discharge about 1350 times, the
battery should not wear by the cycles at 100% DoD. To ensure that the battery does
not completely discharge in an eclipse, however, DoD of 80% has been chosen. [22]
24 4 DIRECT CHEMICAL TRANSFER METHOD
Manufacturer Saft Batteries
Name VES 180
Dimensions
[mm]
53 ϕ x 250
Discharge
Voltage
3.6 V
Capacity 50 Ah
Mass 1.11 kg
Max Discharge
Current
100 A
Max cycles at
20% DoD
60,000
Figure 4.3.3: Specifications of VES180 Li-ion battery produced by SaftBatteries
Using these figures in Equation (4.3.1),
it was deduced that the battery must
at least contain 154 Wh. Since one
VES 180 battery features 180 Wh, one
battery is enough to power the mission.
The spacecraft peak power requirement
is also within the maximum discharge
power of 360 W, and since charging
time is longer than discharge time, the
battery charging current will not exceed
the discharge current.
Hence, one of this battery is chosen for
the spacecraft.
4.3.4 Solar Array Sizing
During daylight, the solar array must
provide for both PSR and charging the
battery. This is the minimum power that would allow the spacecraft to nominally
operate, and hence this power is used to define the End-of-Life power (PEOL). PEOL
is then given by:
PEOL =PSR
µPCDU
+DoD × Emin
µBattery × Tdaylight
(4.3.2)
where µBattery is an efficiency figure for charging and discharging the battery. For this
study, this efficiency is set to 1.
It is shown that minimum power required by the solar arrays is 162.2 W.
Over the course of the mission, the solar arrays will be subject to damaging charged
particle fluences. Data has been collected on such fluences in the trans-lunar space,
and a study has shown that most of such particles are trapped in the Van Allen
radiation belts. [25]
Using particle fluences measured in the radiation belts, and the product speci-
fication by emcore, damage done during the 21-day commissioning sequence in GTO
has been found. Using coverglass of 152 µm thickness, it was shown that during com-
missioning, SA degradation P/PBOL is 90%. [24, 20] Degradation at lunar radiation
environment was estimated to be 0.1% per year.
Using these figures, Beginning-of-Life power requirement is calculated:
PBOL =PEOL
Dtransfer × (1−Dlunar)Tmission
1year
(4.3.3)
4.3 Power Subsystem 25
where Dtransfer is a one-off degradation during transfer, and Dlunar is the degradation
per unit-time spent in lunar orbit.
Using Tmission = 10 months, this has produced BOL power requirement of 181.8 W.
As shown below, the solar arrays will be body-mounted on five faces. This means
that at any given moment, more than one face will be illuminated. To ensure that
PBOL is at all times produced when the satellite is illuminated, it has been decided
each face must be capable of producing PBOL when illuminated perpendicular to the
sunlight, whih is when the total projection-area of all solar arrays combined would be
at its minimum.
ZTJ PV Cell produced by emcore has an efficiency of 29.5%. [20] Using these cells,
the Sun-projection area required to produce PBOL is:
ASA =PBOL
µcell × Φsun
(4.3.4)
Here, packing efficiency is not considered, as the number of cells required are to be
found. Using Φsun = 1,350 W/m2, this has produced sun-projection area of 0.46 m2.
4.3.5 Solar Array Configuration
It had been recognized that the deployed arrays would introduce a single point of
failure, and would strain the AOCS, as it would increase the moment of inertia of
the spacecraft by a factor of as high as 2. As it had been found when sizing the
power subsystem for the low thrust option, the increased moment of inertia meant
that a heavier set of reaction wheels (by a factor of 3) had to be used, as explained
in Section 6.6.3. The mass-increase in choosing the heavier wheels would not allow
the spacecraft to carry enough fuel, consequently it would force the elimination of
secondary payloads. The low thrust option was able to accommodate these wheels
due to the smaller fuel mass required, however, this would not be the case for chemical
options. It has therefore been decided that this configuration will only be employed
when the sun-projection area of the array exceeds the area of one side of the spacecraft.
Since the satellite is cubic with 1 m side-length, as shown in Figures 4.8.1 and 4.8.2,
this limit would be 1 m2. This situation has been avoided, hence body-mounted
solution is employed.
The solar arrays must be placed so that, when in daylight, are producing enough
power to power the subsystem and charge the batteries. Placing the arrays on all six
sides of the spacecraft would produce a fully power-safe spacecraft, however, it would
strain the thermal control subsystem, as GaAs arrays have high emissivity, rendering
the spacecraft very cold during eclipse, as shown in Section 4.7. It was found that
placing five arrays would also provide power-safety, as well as reduce the strain on the
thermal control subsystem. Hence solar arrays are placed on five faces of the satellite.
26 4 DIRECT CHEMICAL TRANSFER METHOD
4.3.6 Power Profile
Having selected the subsystem components, the power profile over one orbit is simulated.
Power produced by the arrays and power demanded is compared, such that the battery
is charged when the power is in surplus, and discharged when the power is in deficit.
The result is plotted in Figure 4.3.1 The dip in the SA power production is due to
an eclipse; the longest eclipse at aposelene is used to produce this graph. After half
a sideral lunar rotational period, the shortest eclipse is observed at the opposite side
of the moon.
4.4 On-Board Data Handling Subsystem (OBDH)
The On Board Data Handling (OBDH) subsystem takes care of spacecraft command
and control by means of a micro-controller. The actions to be performed are defined
both by a software stored in the system and by command up-linked from ground
stations.
The spacecraft has to work in three different phases: Geostationary Transfer Orbit
(GTO) parking, transfer, and lunar orbit. Among these, the third is the more com-
putational demanding since the science data processing will be added to all the usual
tasks (e.g., data housekeeping, solar arrays and antennas pointing). Therefore, the
On-Board Computer (OBC) is sized to meet lunar orbit computation demands.
4.4.1 Overview
The most important system duties are:
• Science data collection and processing.
• Monitoring of the spacecraft health via sensors data.
• Act as communication hub between subsystems, and between the ground station
and the spacecraft.
A system capable of handling these data can be designed as depicted in Figure 4.4.1.
The OBC receives and decodes commands which are then directed toward the payloads
and other subsystems by operating a digital to analog (D/A) conversion where needed.
The subsystems operate the required operations and return science and telemetry
data. These are marked with a unique time-stamp using the internal clock and then
processed by the on board software. Accordingly to ground control directives, storage
and down-link capabilities, the collected data are selected, usually giving priority to
science ones. In order to optimise the down-link process, different types of data are
merged and possibly compressed through the multiplexing step. Eventually all the
compressed packages are stored until the next available communication window when
the information are encoded and sent to the ground station.
4.4 On-Board Data Handling Subsystem (OBDH) 27
..
...Earth’sGroundStations
. . . ...Moon
.
. ...Receiver ...Decode
. . . . ...Payload
. . . ...Analog/DigitalConverter
. . . . ...Subsystems
. . . ...Clock
. ...DownlinkCapabilities
...DataSelection
. . ...Multiplex ...Storage
. ...Transmitter ...Encode
.
commands
.science
.
telemetry
.
Communications
.
On-Board Data Handling
.
Lunar Orbiter
Figure 4.4.1: OBDH system scheme showing data flows between subsystems as wellas between the spacecraft and both ground stations and the Moon’s environment.
4.4.2 Requirements
The OBC should be able to handle and store all the information gathered in between
at least two down-link sessions i.e., one lunar orbit. In Table 4.4.1 the types of data
and their size are reported.
For each orbit at least 10 pictures per camera are taken.
4.4.3 Components
This section focuses on OBDH components and their selection. In Figure 4.4.2 the
ensemble of system parts is organised to show their relationships.
The OBC has been assembled using only off-the-shelf components. Therefore,
each part described in the following sections comes from the catalogues of actual
manufacturers. Although several catalogues have been searched for each component,
only the chosen component is reported, along with a brief description.
28 4 DIRECT CHEMICAL TRANSFER METHOD
Source SizeHigh Res. Camera 100MB/pictureLow Res. Camera 44.352MB/pictureDust Detector 150KB/dayRadiation Detector 400KB/daySpectrometer 36Mb/orbitTelemetry 300 b/s
Table 4.4.1: Data sources and sizes. The raw file size is taken into consideration.
..
. ..COP . . . . . . ..DigitalI/O 1
..Real-TimeClock
. . . . . ..Driver
..PROM . . . ..CPU
..SRAM . .. . . . ..Driver
. ..EDAC . . . . . . ..DigitalI/O n
..FLASH . ..
..AnalogInput 1
.. . . . . .. .. ..AnalogOutput 1
. . ..Muxer ..A/D . ..D/A ..Demuxer
..AnalogInput n
.. . . . . . .. ..AnalogOutput n
Figure 4.4.2: OBDH scheme in details. The most important links between componentsare shown.
The processor has been selected paying attention to its performance, available
cache, power consumption, radiation resistance, and heritage. In addition, two fun-
damental requirements were the space operation qualification and a 32-bit architecture.
The picked micro-processor is the RAD750 produced by BAE Systems [35, 37].
The decision has been driven primarily by its computational power (400MIPS) and
in second instance by its heritage. Indeed, it has been employed in missions like the
Mars Science Laboratory (MSL) one, where it proved high reliability and the capacity
of meeting high computational demands thanks to its high clock frequency (200MHz),
which is supported by 1MB cache.
It is not unusual that for a specific micro-processor architecture a particular
operating system is written. Indeed, the RAD750 is often used in combination with
4.4 On-Board Data Handling Subsystem (OBDH) 29
the Wind River VxWorks Real Time Operating System (RTOS) [34]. This Operative
System (OS) has been used to coordinate the landing operations for the MSL, an
operation which required high accuracy and reliability. VxWorks system requirements
By considering the total number of subsystems and science instruments, it is clear
that there is the need for more than four ports. This is why a SpaceWire router has
been looked for. The router itself consists in only a small chip, indeed the assembly of
the network facility is left to the spacecraft engineers. Being this not a feasible option
an off-the-shelf router has been found. The SPA SpaceWire Network Router [33],
manufactured by Design Net Engineering LLC, is configurable with up to 16 ports,
despite no more than 8 ports would be needed (Figure 4.4.3). The router size is given
as 127mm×178mm×50mm [33]. By using this value a rough mass approximation can
be done. Since it takes up space of about 1.5 OBCs, then its weight is approximated
to 1.5 kg in order to take into account its shielding.
Cabling mass has been considered as the 20% of the total OBDH mass.
To complete the connection between the different devices, only an analog interface
is needed. In particular, a signal converter analog/digital/analog is manufactured by
Honeywell in form of on-board chip [31], and can be added to the SBC in the same
way the FLASH unit has been integrated.
The hardware has now been selected completely, the final configuration and re-
dundancy computations are reported in Table 4.4.3. In particular, to take into account
the redundancy, the entire system has been duplicated.
4.4.4 Software
The development of the spacecraft software has not been considered. Indeed this
process may take several years of code optimisation. Also, one of the main OBC
task, the attitude regulation, is performed by a separate computational unit which
is integrated into the Attitude and Orbital Control System (AOCS). Hence, the only
assumptions done are about the compression of science and telemetry data for com-
munication purposes.
32 4 DIRECT CHEMICAL TRANSFER METHOD
Images are compressed initially by using the JPEG format. Besides the JPEG
compression is not loss-less, a small compression ratio, say 1 : 10, introduces a small
amount of noise in the picture, nearly invisible to the human eye. Therefore, all the
images taken are reduce to one tenth of the raw size. Subsequently, a loss-less com-
pression algorithm is used to apply a further compression ratio of 1 : 1.76 to all the
collected data [29]. Eventually the total data size is reduced to 110.27MB/orbit. This
value represents the output given to the Communications subsystem to estimate the
down-link requirements.
4.4.5 Considerations for Direct Chemical
The direct chemical transfer option does not require any additional component aside
from those reported in the general OBDH configuration.
In order to assure a proper radiation shielding, the results obtained by the spacecraft
Chandrayaan-1 during its journey to the Moon have been used [27]. Initially a time
period of two weeks in GTO has been assumed. The continuous passage through
the Radiation belts will provide a total amount of 1.0848 krad. The short five days
transfer will add 0.144 rad. Eventually, during the six month of orbital mission the
amount of radiation is estimated to be of 669.071 rad. The total amount of radiation
for the entire lifetime is 1.754 krad.
Single Unit RedundancyPerformance 400MIPS @ 200MHzMemory 1MB cache
128MB SDRAM256KB SUROM4MB NVRAM8GB FLASH
Mass 2.6 kg 5.2 kgDimensions 127×178×110mm 127×178×220mmPeak Power 22.8WRadiation 100 kradOperating Tem-perature Range
−55◦C +70◦C
Cabling 20% subsystem massTotal mass 2.76 kg 5.52 kg
Table 4.4.3: OBDH configuration. Redundancy is taken into account by duplicatingthe whole subsystem.
4.5 Communication Subsystem 33
4.5 Communication Subsystem
The communications subsystem will carry out a number of key functions.
• Transmit housekeeping data to flight controllers.
• Receive instructions from flight controllers.
• Transmit photographs.
• Transmit scientific data.
4.5.1 Link Budget
Evaluation of the link budget is performed by selecting components and planningoperations such that required data rates can be met while minimising mass, andpower demands to an acceptable level.The link budget is calculated using Equation 4.5.1.
10 log10
(C
N0
)= 10 log10(PTGT ) + 10 log10
(GR
TR
)− 20 log10
(4πρ
λ
)− 10 log10 (LA) − 10 log10 (k) (4.5.1)
Where
CN0
carrier power to noise densityratio
PT transmitter RF power
GT transmitter gain TR receiver equivalent noise tem-perature
ρ slant range from transmitter toreceiver
λ wavelength
LA atmospheric loss k Boltzmann’s constant
The following passages will progressively define each of the terms in this equation.
The CDE spreadsheets are used to calculate a solution for each mission phase, where
the final output in each case was the RF power required for the transmitted signal.
By considering communication windows imposed by the mission orbit and attitude
selection, the quantity of scientific data gathered was optimised to satisfy mission
requirements while keeping this power within the capabilities of selected components.
4.5.2 Link Quality
To define the CN0
, a typical bit error rate of 10−5 was chosen. This gives Eb
N0of 10dB
[39]. Carrier power to noise density ratio is obtained by multiplying this density by
bit rate (Rb) according to Equation 4.5.2.
C
N0
=Eb
N0
Rb. (4.5.2)
34 4 DIRECT CHEMICAL TRANSFER METHOD
Since the concurrent design approach meant that bit rate would vary during design,
this equation was used in the concurrent design environment spreadsheets to calculateCN0
, while bit error rate was assumed constant.
4.5.3 Ground Stations
ESA Tracking Stations (ESTRACK) will provide sufficient coverage and data rates
at typical frequency bands and bandwidths. Using STK, ESTRACK coverage was
investigated and it was found that by using tracking stations in Kiruna, Kourou, and
Perth, the spacecraft would be in almost continuous visibility (greater than 99%)
provided the Moon is not blocking the line of sight.
All subsequent investigation into the communications coverage assumes these tracking
stations are to be used, their basic information is tabulated in Table 4.5.1
Kiruna-1 Kourou-1 Perth-1
Dish diameter 15 m 15 m 15 mS-band RX band 2200 - 2300 MHz 2200 - 2300 MHz 2200 - 2300 MHz
Table 4.5.1: Performance characteristics for the Kiruna-1, Kourou-1 and Perth-1terminals [40]
4.5.4 Carrier Frequencies
High gain transmissions will use X Band. This band was selected for high gain trans-
missions because it has heritage with high data rate transmission, and is compatible
with the ESTRACK network. S-Band was chosen for low gain communications for
the same reasons.
It is clear from the data on the selected tracking stations and the data rates calculated
for the OBDH subsystem that the bandwidth required by the spacecraft for high and
low gain transmissions can be accommodated by the selected ground stations at X
and S bands.
To carry out link budget calculations in the CDE, arbitrary carrier frequencies within
the relevant bands given above have been assumed. These are 8450MHz for high gain
communications and 2250MHz for low gain communications.
4.5 Communication Subsystem 35
4.5.5 Losses
Both free space (LFS) and atmospheric losses (LA) in the beam must be considered.
Free space loss is given by Equation 4.5.3.
LFS =
(4πρ
λ
)2
, (4.5.3)
Assuming a maximum distance to be the apogee radius of the Moonfs orbit, 405400km,
we find that (LFS) is 1.43×10−11 for high gain transmissions and 3.82×10−10 for low
gain transmissions.
Using frequency bands below 10GHz, spacecraft communications will not suffer sig-
nificant clear air attenuation [41]. Assuming an atmosphere of dry air and water
vapour, zenith attenuation is approximately -0.04dB [42] for spacecraft transmission.
As well as attenuation due to dry air, some loss would be caused by precipitation.
However in this case it is not a significant factor. If it were to be considered, the
level of attenuation due to precipitation may be estimated using rainfall statistics and
the method described in various documents by the International Telecommunications
Union [39]. This method consists of sizing the communications system to handle at-
tenuation levels that will be not be exceed for some acceptable percentage of the time,
eg. 99.9%.
Over the distances concerned, free space losses occur at a much greater order of
magnitude than atmospheric losses. Atmospheric losses are around 2dB at 11GHz for
a link reliability of 99.9% [39], while free space losses are 223dB. For this reason it is
deemed an acceptable simplification not to calculate a value for attenuation due to
precipitation, and instead use only dry air attenuation in calculating the link budget.
4.5.6 Component Selection
Given the high free space loss compared with missions to LEO, the transmitter will
need to have sufficiently high power and gain in order to reach the required data rate.
After reviewing available hardware options, the following communications architecture
was selected.
• 1x X-band steerable high gain antenna.
• 1x X-band transmitter.
• 6x S-band low gain antennas.
• 2x S-band transmitters.
• 2x S-band receivers.
36 4 DIRECT CHEMICAL TRANSFER METHOD
Figure 4.5.1: Zenith attenuation of signal vs frequency [42]
4.5 Communication Subsystem 37
All of these components are available from SSTL, where both the X-band components
and S-band components have flight heritage together. The specific hardware and their
key data are given in 4.5.2, 4.5.3, 4.5.4, 4.5.5, 4.5.6.
Mass 3 kg for 15dBiC, 3.3 kg for 18dBiC
Power1.3 W static
3.9 W dynamicFrequency range 8000 - 8500 MHz
Gain 15 dB or 18 dBSlew rate < 20◦/s
Azimuth Range +/- 270◦
Elevation range +/- 110◦for 15dBiC, +/-80◦for 18dBiC
Table 4.5.2: Key data for the SSTL X-band high gain antenna pointing mechanism[43].
Mass 4 kg
Power demand65 W at 5 W RF
120 W at 12 W RFFrequency range 8025-8400 MHz
Data rate 10.0 - 500.0 Mbps
Table 4.5.3: Key data for the XTx400 X-band transmitter [44].
Mass 3 kg for 15dBiC, 3.3 kg for 18dBiCFrequency range 2000 - 2500 MHz
Gain -5 dB at 90◦off boresight
Table 4.5.4: Key data for the SSTL S-band patch antenna [45].
Mass 1.8 kgPower demand at 4 W RF < 38 W
Frequency range 2200 - 2250 MHz
Table 4.5.5: Key data for the S-band downlink transmitter [46].
Mass 1.3 kgPower demand 1.5 W
Frequency range 2025 - 2100 MHzData rate 9.6 kbps or 19.2 kbps
Table 4.5.6: Key data for the S-band uplink receiver [46].
These pieces of hardware were selected as they would roughly satisfy RF powers
calculated using the link budget spreadsheets in the CDE derived from data rates
38 4 DIRECT CHEMICAL TRANSFER METHOD
demanded by the OBDH subsystem. Their data was input and OBDH figures were
adjusted to keep the required RF power within the limits given in these tables.
The system will employ redundancy in S-band communications, but not in X-
band. This a mass saving measure; the need for redundancy was traded off against
mission life since some of the mass saved can be used for station keeping fuel, allowing
longer orbit maintenance and therefore prolonging the mission. It is possible that the
lack of redundancy for high gain communications could result in a severely reduced
transmission capability in the event of failure of the X-band antenna or transmitter.
In the case of the Galileo probe, when the high gain antenna failed to deploy the
mission controllers were forced to use low gain antennas to receive all scientific data.
A similar contingency is examined later in this chapter.
4.5.7 Investigating Communications Windows
Using Kiruna, Kourou and Perth as ground stations, STK provides information on
typical transmission windows while in the final mission orbit around the Moon.
Most access windows last more than 5000 seconds. However there were also many
windows closer to 3000 seconds. Transmission will thus occur once per orbit, and a
typical worst case transmission window is assumed to be 2700 seconds (45 minutes).
Due to the elevation constraint for the X-band antenna at maximum gain, see Table
4, and the pointing requirements of the instruments, the communications subsystem
places a demand on spacecraft attitude control. Twice each month, when the plane
of the mission orbit is almost normal to the vector to the Earth, the limit in antenna
elevation requires the spacecraft to slew for each transmission. This repointing is
necessary as long as the orientation of Earth relative to the orbit places the ground
stations outside the field of view of the antenna, using STK it was found this slew
is required for 8 orbits during these periods. Communications slewing will therefore
occur for a total of 16 orbits per lunar month, once during each of these orbits. The
degree of slewing will be 15◦to ensure that the antenna can point directly at the
ground station being used.
Due to the almost constant ground station visibility during transfer, transmission
windows were not investigated for this phase of the mission. It was also assumed that
since no scientific data is gathered during transfer, only low gain communication would
be used. By placing an S-band patch antenna on each spacecraft face the need for
communications slewing is eliminated during transfer. The small mass contribution
of multiple S-band patch antennas (80g each) was accepted to reduce the fuel mass
that would be required if the spacecraft were to be slewed during this mission phase.
4.5 Communication Subsystem 39
Figure 4.5.2: STK model of satellite communications. Cone shows field of view
4.5.8 Contingencies
Since the high gain capability of the communications subsystem will have no re-
dundancy, its possible failure, and the feasibility of using the low gain antennas to
transmit scientific data to earth, has been examined.
An alternative link budget was produced. The budget used for normal operation
takes data rates as an input and given information on the transmitter, receiver, and
some losses it will calculate the required RF power. The contingency link budget
reverses this process. It assumes that the low gain antenna will be running at full RF
power (4W) and calculates a maximum bit rate. This was found to be 36.45 kbps. This
rate was multiplied by the transmission window length to give a maximum quantity
of down linked data. How this quantity is divided between housekeeping and the
different types of scientific data is discussed further in section 4.4.
40 4 DIRECT CHEMICAL TRANSFER METHOD
4.5.9 Final Results
The following table shows the final set of data rates to be used in the case of the direct
chemical transfer method. These values were calculated using the CDE spreadsheets
and are considered to be the final output for the communications segment.
Mission phase Data rate
Transfer orbit 300 bpsMission orbit (low gain) 1.37 kbpsMission orbit (high gain) 256.02 kbpsContingency data rate incase of high gain antennafailure
36.45 kbps
Table 4.5.7: Final key values for the direct chemical option.
4.6 Attitude Determination and Control Subsystem (ADCS)
The ADCS system will satisfy orbiter pointing requirements and steady the payload
so that science goals can be accomplished.
4.6.1 Control Modes
Firstly the control modes of the mission must be defined. Identifying these modes
will provide the requirements of the ADCS.
1. Initial Parking Orbit - The period between the orbiter being released from Ariane
5 into it’s initial Earth orbit and the beginning of the lunar transfer phase.
This will require ADCS control for testing equipment. These requirements are
negligible in comparison to the rest of the mission and so shall not be considered.
2. Transfer Slew - During this mode the Thruster will perform three burns; an
initial burn, a mid-way burn and a retro-burn. To ensure accuracy the orbiter
may need to be re-orientated.
3. Lunar Orbit Insertion and Acquisition - Initial determination of attitude and
stabilisation of vehicle upon arrival to the Moon. This mode may also be used
to recover from potential power upsets or emergencies. In order to account for
this, a safety factor will be included.
4. Nadir Data Collection - This mode occurs when the primary camera is nadir
pointing. During this mode the ADCS must provide a stable platform from
which pictures can be taken.
4.6 Attitude Determination and Control Subsystem (ADCS) 41
5. Off-Nadir Data Collection Slew - To achieve the orbiter’s science goals, it will
need to perform small slewing movements in order to image areas of interest
that are not covered by it’s ground track.
6. Communications Slew - The orbiter will have to slew back and forth from it’s
regular data collection mode in order to facilitate communications with the
Earth.
7. Eclipse Observation Slew - The orbiter will have to perform several large slews
to observe both the Moon and the Earth during the eclipse phase outlined in
the science goals.
8. Contingency Mode - A safety setting that can be implemented in an emergency.
Requirements for the four slewing modes are outlined in Table 4.6.1. For suitable
payload performance, the ADCS must fulfil the requirements given in Table 4.6.2 for
attitude determination and control whilst imaging. These requirements are applicable
to the data collection modes, and are driven by the primary payload.
Control Mode Slew Angle/◦ Slew Time/s Frequency Rate/◦s−1
Communication 15 60 0.54 per day 0.250Transfer 180 90 3 total 1.500Eclipse Observation 180 60 5 per year 2.000Off-Nadir Collection 5 60 10 per day 0.083
Table 4.6.1: Requirements for the four slewing modes - Mode 2, 5, 6 and 7.
Parameter RequirementAccuracy of Determination and Control 20 arcsecsRange Over Which Accuracy is Met ±180◦
Maximum Allowable Jitter 0.1◦/minDrift Allowance 1◦/hourSettling Time 1 min
Table 4.6.2: Camera Determination and Control Requirements
4.6.2 Selection of Attitude Control Method
In order to provide three-axis accuracy of determination and control to an accuracy of
20 arc seconds (0.00556◦), either a zero momentum Reaction Wheel Assembly (RWA)
consisting of three wheels or Control Moment Gyroscopes (CMGs) could be used.[48]
Other types of torquer simply wouldn’t be able to provide the same level of precision.
CMGs have historically been used for satellites greater than several tonnes such as the
42 4 DIRECT CHEMICAL TRANSFER METHOD
ISS.[52] There has been a recent push to miniaturise CMG technology because they
offer a high mass and power-to-torque efficiency. However, currently the smallest,
reliable, off-the-shelf products such as Astrium’s CMG15-45S and Honeywell’s M50
CMG are designed to provide pointing requirements for satellites of roughly 1000 kg
and so would still be oversized for LEWIS.[49][50] Some CMG prototypes exist for
satellites nearer 300 kg, such as the University of Surrey’s SGCMG, but they are still
in development and so can’t be considered for this mission.[51] Therefore only a zero
momentum three wheel reaction wheel assembly can be considered for attitude control.
4.6.3 Quantifying the Disturbance Environment
This transfer method only requires us to quantify the disturbance environment around
the Moon. The greatest disturbances will be due to solar radiation pressure and
gravity gradient effects caused by the Moon. The orbiter will also be affected by the
gravity gradient effects caused by the Earth and the Earth’s magnetic field, however
those effects are negligible once in lunar orbit.
The worst case disturbance torques during lunar orbit orbit must be estimated. The
torque disturbance due to the gravity gradient effect on the orbiter, Tg is calculated
using,
Tg =3µ
2R3|K| sin(2θ) (4.6.1)
where R is the orbit radius (m), θ is the maximum deviation of the z-axis from the
local vertical, (to simulate the worst case scenario this will be fixed as 45◦) and K is
the greatest difference between the moments of inertia about z, y and x axes in kgm2
after transfer. The orbit radius will change over time, but the rest of the elements of
the equation are known, giving,
Tg =3× 4.90× 1012
2R3|3.31| sin(90◦). (4.6.2)
This equation can now be used to find the average disturbance torque due to the
gravity gradient effect of the moon yielding,
Tg = 8.40× 10−7Nm. (4.6.3)
The disturbance torque due to solar radiation pressure, Ts can be quantified by using,
Ts =Fs
cAs(1 + q) cos(I)(cps − cg) (4.6.4)
where Fs is the solar constant, 1367W/m2, c is the speed of light (3× 108m/s), As is
the largest surface area plane (0.5m2 for this transfer method), cps is the location of
solar pressure, cg is the centre of gravity, q is a reflectance factor (this is assumed to
4.6 Attitude Determination and Control Subsystem (ADCS) 43
be 0.5) and I is the angle of incidence of the sun (assumed to be 0◦ for the worst case
value). A value of 0.1 m is used for the difference between cg and cps for preliminary
design. This should account for changes in the geometry or for the selection of ad-
Table 4.10.1: Final mass budget for the direct chemical transfer option.
58 5 WEAK STABILITY BOUNDARY TRANSFER METHOD
5 Weak Stability Boundary Transfer Method
5.1 Mission Analysis
5.1.1 Transfer
An STK scenario was constructed to study the ∆V requirement to perform a WSB
transfer to the moon. Two different methods have been investigated for this type of
transfer: a conventional WSB transfer via Sun-Earth L1 and another that utilizes a
lunar flyby to reach to the Sun-Earth L1 point, both illustrated in Figure 5.1.3
The two methods required very similar ∆V requirement and transfer time. They
have, however, produced a considerable difference in the final orbit: the conventional
method produced AoP of 320.0◦, compared to 4.0◦for the lunar flyby. This would
mean that the traditional WSB transfer would result in a final orbit with station-
keeping ∆V requirement similar to that of the direct chemical option, whereas for the
lunar flyby, it would require about fourth of that figure, meaning that the mission
lifetime could be extended by a factor of four. Therefore for the WSB option, the
lunar flyby method was chosen.
Approximately the same launch date was chosen as the direct chemical option. Upon
arrival, however, LEWIS is allowed to swing by. The TLE epoch is controlled to allow
control of the highest apogee altitude near L1. This allows control of the magnitude
of sun’s gravitational assist. At apogee, one targeting burn is fired to achieve the
desired PA and inclination.
5.1.2 Final Orbit
1
2
3
4
30
210
60
240
90
270
120
300
150
330
180 0
Orientation of Final Orbit for WSB Option
Orbital Radius [Moon Radius]
South - North
LEWIS Orbit
Figure 5.1.1: Final Orbit Achieved byWSB Option. AoP = 4.0◦
The final orbit achieved by WSB transfer
is near flat with AoP of 4◦. This
results in a much lower ∆V requirement
compared to the direct chemical option,
of 78m/s per year, as simulated by STK,
results shown in Figure 5.1.2.
staionkeeping The same stationkeeping
strategy as the direct chemical option is
employed: PA and AA are controlled as
soon as they drift by 10km. Since less
fuel is expended in transfer compared to
the direct chemical option, and less fuel
is required for stationkeeping per year,
LEWIS would stay on the designed orbit
for much longer, 42 months by linear extrapolation.
5.2 Chemical Propulsion Subsystem 59
Similarly the power subsystem is sized to provide minimum power at this time.
Figure 5.1.2: Keeping PA at 100km ± 10km for 1year requires 78m/s. By linearextrapolation given the fuel carried, LEWIS will keep the station for 42 months inlunar orbit.
5.2 Chemical Propulsion Subsystem
The chemical propulsion subsystem for the WSB option covers the same functions as
for the direct chemical option. The two propulsion subsystems are identical in terms
of their configuration apart from the the tank dimensions. The methodology and
sizing process employed for the direct chemical option (see Section 4.2) are directly
applicable to the propulsion subsystem for WSB.
The expellant lifetime for WSB is 1252 days in contrast to 272 days for the direct
chemical option. This can be attributed to the lower expellant mass required for the
transfer and lower station-keeping delta-v in the mission orbit.
The impulse delivered by the ME during transfer is 1.13 × 105 Ns, i.e. 18.7 % of
the MEs total impulse capability of 1.67 × 106 Ns. The remainder is available for
station-keeping and RCS functions. As for direct chemical the impulse delivered by
the RCS during transfer is negligible in contrast to the 1.80 × 105 Ns capability of
a single RCS thruster. Furthermore, with the annual RCS impulse requirement of
920 Ns, neither the lifetime of the ME nor the RCS thrusters impose constraints on
the mission lifetime.
A breakdown of the dry propulsion subsystem in terms of mass is given in Table 5.2.1.
60 5 WEAK STABILITY BOUNDARY TRANSFER METHOD
Figure 5.1.3: STK simulaitons on WSB transfers. Conventional WSB transfer viaL1(Top): Transfer ∆V = 1177m/s, Transfer Time = 101 days. And transfer via lunarflyby (Bottom): Transfer ∆V = 1048m/s, Transfer Time = 80 days. For advantagesin final orbit, the latter method is employed.
5.3 Power Subsystem 61
Note the miniscule differences compared to the similar dual-mode system employed
for the direct chemical transfer (see Table 4.2.1 in Section 4.2).
Component Mass Allocation[kg] (%)
Oxidiser tank 9.71 27.6Propellant tank 10.04 28.5Pressurant tank 2.34 6.6RCS thrusters (12) 3.48 9.9ME 6.03 17.1Plumbing 3.16 9.0Pressurant 0.43 1.2Propulsion system dry 35.18 100.0(incl. pressurant)
Table 5.2.1: Mass breakdown of the dry dual-mode propulsion subsystem for the WSBtransfer option.
5.3 Power Subsystem
5.3.1 Requirements
To estimate the power requirement for the spacecraft, the same approach used for
the direct chemical option, described in Section 4.3, has been used. The power re-
quirement of each instrument has been has been given, and a power plan over one
orbit has been established. However, since the eclipse happens at a different section
of the orbit, the power plan looks slightly different, as shown in Figure 5.3.1.
The average power 114.7 W; and with 10% contingency, subsystem power re-
quirement (PSR) is given to be 126.2 W. To size the power subsystem, it is assumed
that the spacecraft steadily consumes this amount of power.
Figure 5.3.1: Power profile for WSB transfer option.
62 5 WEAK STABILITY BOUNDARY TRANSFER METHOD
5.3.2 PCDU
The magnitudes of power required and produced by this spacecraft are very similar to
that of the direct chemical option. Hence, the Small Satellite Power System by SSTL
is also used for this satellite.
5.3.3 Battery Sizing
It has been shown that PSR = 126.2 W. The longest possible eclipse time has also
been found to be 3620 seconds. Taking depth of discharge of 80%, The minimum
energy capacity of the battery is given by Equation (4.3.1) to be 181.7 Wh. This is
larger than the energy capacity of the VES-180 cell shown in Figure 4.3.3. Hence,
two of this battery are carried in the spacecraft, giving Ebattery = 360 Wh.
5.3.4 Solar Array Sizing
The end-of-life power for the spacecraft is given by Equation (4.3.2) to be 167.8 W.
Since the same length of time is spent for the commissioning phase in GTO, the same
amount of damage is expected during this phase as with the direct chemical option. It
is also assumed that the radiation environment in the trans-lunar space and near L1
is similar to that at the lunar orbit, the main constituent of which being the protons
from the sun and galactic cosmic rays.
Using these figures, the beginning-of-life power has been found to be 193.7 W.
The same solar cell from emcore is used for its high efficiency. Using Equation
(4.3.4), the sun-projection area of the array is found to be 0.49 m2.
5.3.5 Solar Array Configuration
Since the sun-projection area is less than 1 m2, body-mounted configuration is chosen.
As with direct chemical method, the arrays are mounted on five faces.
5.3.6 Power Profile
Using the PCDU, batteries and solar arrays, a power profile is simulated over one
orbit, results shown in Figure 5.3.1.
5.4 OBDH
Given the transfer time of about 80 days and the 10 mm thick shielding, the amount
of radiation to be received by the spacecraft is estimated to be about 5.617 krad.
5.5 Communication Subsystem 63
5.5 Communication Subsystem
Table 5.5.1 shows the final set of data rates to be used in the case of the weak stability
boundary transfer method. These values were calculated using the CDE spreadsheets
and are considered to be the final output for the communications segment.
Mission phase Data rate
Transfer orbit 300 bpsMission orbit (low gain) 2.03 kbpsMission orbit (high gain) 643.68 kbpsContingency data rate incase of high gain antennafailure
36.45 kbps
Table 5.5.1: Final key values for the weak stability boundary option.
5.6 ADCS
5.6.1 Control Modes and Selection of Attitude Control Modes
The control modes for the WSB transfer method will be exactly the same as the
direct chemical transfer method. It will also have the same ADCS requirements for
its slewing modes and camera requirements as given in Tables 4.6.1 and 4.6.2. As a
result the argument for the attitude control method will also be the same as for the
direct chemical transfer for the WSB transfer. The conclusion of this was to use a
zero momentum three wheel reaction wheel assembly.
5.6.2 Quantifying the Disturbance Environment
The final lunar orbit for this transfer method will be very similar to the direct chemical
option. The largest surface has remained 0.5m2. Only the inertia matrix has changed
due to some changes in hardware and configuration.
This means that Ts is still 6.83× 10−7 Nm. The new disturbance torque to gravity
gradient effects can be calculated using Equation 4.6.1 by using the new K value
of 2.80kgm2 and the method outlined in Section 4.6.3. This yields a Tg value of
7.11× 10−7 Nm and thus TL = 1.39× 10−6.
5.6.3 Selection and Sizing of ADCS Hardware
As the ADCS requirements for the WSB method are extremely similar to the direct
chemical method, the same hardware has been selected with the properties as listed
in Section 4.6.4.
64 5 WEAK STABILITY BOUNDARY TRANSFER METHOD
5.6.4 Hardware Application and Resultant Thruster Requirements
As the requirements on the ACDS are the same apart from the changes in the inertia,
the method and equations listed in Section 4.6.5 can be used as before. The total
momentum storage per day is 1.31 Nms, compared to 1.27Nms for the direct chemical
method. In order to stay within the RWA’s operating limits, momentum dumping
will occur similarly once per orbit (4.43 times per day). This yields the impulse and
thruster requirements given in Tables 5.6.1 and 5.6.2.
Argument of Perigee [deg] 178 Argument of Perigee [deg] 90True Anomaly [deg] 0 True Anomaly [deg] 0
Depart Date 00:00:0023/01/2013
Arrival Date 15:50:2423/01/2015
Transfer Fuel [kg] 32.7 Transfer Time [days] 730.7
72 6 LOW THRUST TRANSFER METHOD
6.1.2 Final Orbit
1
2
3
4
30
210
60
240
90
270
120
300
150
330
180 0
Orientation of Final Orbit for Low Thrust Option
Orbital Radius [Moon Radius]
South - North
LEWIS Orbit
Figure 6.1.2: Final Orbit Achieved by lowthrust option. AoP = 0◦
Low thrust transfer method has allowed
free choice of the final orbit. The desired
orbit with AoP of zero is hence achieved.
Stationkeeping As AoP of zero is
achieved, PA and AA will only slowly
drift. The first stationkeeping maneuvre
is required at the 17th week since TLI,
as shown in Figure 6.1.3.
The stationkeeping strategy for the
low thrust option differs to those
employed by the chemical options, due
to the difference in the thrust levels
achievable by the engines. For this option, only PA is controlled, and AA is allowed
to drift freely, as the stationkeeping burn would take about an hour, and it is not
desirable to use this time to control AA, which could otherwise be used to perform
observations near the periselene. Burns are performed near aposelene to keep the
periselene well above surface, as there is plenty of time when the spacecraft is orbiting
slowly.
The tolerance PA is allowed to drift is calculated by the following steps:
1. On 100 km × 3600 km orbit, find time taken for True Anomaly to proceed from
170◦to 190◦. This is the time used to perform the aposelene maneuvre.
2. Use first guess for PA tolerance (10 km was used), to calculate ∆V required at
aposelene
3. Use rocket equation to calculate mass of propellant required
4. Use mass-flow rate of the engine used to calculate the time taken for the burn,
compare with value obtained in 1.
5. Repeat 2-4 to find allowable tolerance for PA.
As a result, tolerance of 7.5 km was found. STK simulation was then performed using
impulsive maneuvres at aposelene to find how many burns are required per year, the
result shown in Figure 6.1.3. Annual ∆V requirement for stationkeeping was found
to be 8 m/s; much lower than the values for the chemical options. The difference is
due to the difference in the stationkeeping strategies, and the relative stability of the
orbit obtained by the low thrust option. By linear extrapolation with the fuel carried
on board, it is estimated that the spacecraft will stay in designed orbit for about 14
6.2 Electric Propulsion Subsystem 73
years. It may well be that other subsystem will fail before this point, however, it took
only 1 kg of fuel to achieve the 14 years, and the spacecraft weighs 280 kg in total,
allowing for additional payload or contingencies.
Figure 6.1.3: Keeping PA at 100 km ± 7.5 km for 1year requires 8.1 m/s. By linearextrapolation given the fuel carried, LEWIS will keep the station for 13.9 years inlunar orbit.
6.2 Electric Propulsion Subsystem
The choice for the LEWIS mission is the QinetiQ-manufactured Kaufman type T5
thruster shown in Figure 6.2.1. The ion engine is has grids 10 cm in diameter. Direct
current is discharged between the hollow cathode and cylindrical anode to ionise the
Xenon. [54] A simple side view of the gridded ion thruster is shown in Figure 6.2.1.
The ion engine assembly comprises four major components, configured in the ar-
chitecture shown in Figure 6.2.2, resulting in a 30.7kg dry mass.
The gridded ion thruster T5, has the key parameters shown in Table 6.2.1.
The Proportional Xenon Feed Assembly (PXFA), designed by MOOGBradford,
regulates and maintains the flow of Xenon from its tanks to the main cathode and
neutraliser to feed the T5 thruster.[56] The key specifications the PXFA are shown in
Table 6.2.2.
The Ion Propulsion Control Unit (IPCU) designed by Astrium provides the
required voltage and current to the thrusters, as well as measuring the temperature,
and controlling the propellant flow.[57] The key specifications for the IPCU are given
in Table 6.2.3.
Xenon Propellant and propellant tank for T5 thruster - xenon has been
chosen for propellant, due to its high performance, non-toxicity and good storage
74 6 LOW THRUST TRANSFER METHOD
Figure 6.2.1: QineticQ manufactured Kaufman type T5 Thruster
Table 6.6.2: Impulse and Thrust Requirements for the low thrust transfer method.
Parameter RequirementTransfer Impulse 1909.02NsYearly Impulse 2730NsLargest Thrust 0.926NMaximum Yearly Thruster Burn Time 1368sTransfer Thruster Burn Time 1096sFrequency of Transfer Momentum Dumping 0.5 per dayFrequency of Operational Momentum Dumping 1 per orbit
Table 6.6.3: Thruster Requirements for the low thrust transfer method
6.7 Thermal Control Subsystem
6.7.1 Primary Assumptions
The assumptions made in Section 4.7.1 will also apply to the low thrust case. The
transfer method will differ however. Eclipses will be experienced during the transfer
and the assumption is that the longest of these will last two hours, where the spacecraft
will fall to its lowest temperature [66].
82 6 LOW THRUST TRANSFER METHOD
6.7.2 Method of Calculations
The method given in 4.7.2 will be used to find the equilibrium and transient tem-
peratures, and to size any heaters needed.
6.7.3 Equilibrium and Eclipse Temperatures
As the solar arrays are not body mounted for this satellite, the surface properties have
changed. This satellite will radiate far more heat than the other two options due to
the high power usage of the propulsion system.
For the spacecraft to reach acceptable equilibrium temperatures for its mission,
Multi Layer Insulation (MLI) and white paint (with a high ϵ) have been chosen as
coatings. The proportion of MLI-to-paint was determined on a trial and error basis,
and evolved as the design of the satellite changed. The properties of the MLI and
white paint are compared in Table6.7.1. The calculated temperatures for each mission
Table 6.10.1: Final mass budget for the low thrust transfer option.
90 7 EVALUATION AND COMPARISON OF TRANSFER METHODS
7 Evaluation and Comparison of Transfer Methods
7.1 Transfer time
The transfer time for the direct chemical Option is 5 days, while for low thrust option
it is 730.7 days, and for WSB option, it is 80 days.
This adds radiation and debris damage risks to the low thrust and the WSB option,
as scored below.
The propellant mass requirement for the direct chemical option is 100.1 kg, while it
is 96.6 kg for WSB option, and 34.3 kg for the low thrust option.
7.2 Final orbit
The ideal orbit chosen for the mission has argument of periselene (AoP) of 0◦, produced
with EMT software. Each transfer options has a different value for the final AoP,
varying from 4◦North for WSB transfer method to 40◦South for the direct chemical
transfer. These deviations from the ideal orbit could be corrected by executing ma-
noeuvres after the Moon insertion burn; but we decided to avoid it, as such burns are
expensive in terms of propellant mass. The ideal orbits will provide ability to take
sufficient picture quantity at different spatial resolutions. Hence, as a result we have
AoP of 4◦N for WSB, 40◦S for DC, and 0◦for low thrust.
Observable lunar landscape is different for the three options; achievable ground resolution
over the six-month period is shown in Figure 7.2.1.
7.3 Extended Mission Lifetime
Due to the moon’s non-spherical gravity, the final orbit affects the orbital perturbence
forces. Without correction, this would affect the rate of decay of orbit; with correction,
this affects the frequency of stationkeeping maneuvre required, hence affecting the
mission lifetime.
Though each option fulfils the mission requirement of 6 months, total duration of
time the spacecrafts are able to maintain the required orbit differ; it is desirable to
achieve longer mission life, as this would increase the ammount of data retrievable by
the mission.
The final orbit of the direct chemical option provides the least extension of the
three options, scoring 3 extra months. The WSB gives 36 months of life extension.
The longest extension is achieved by the low thrust option, due to the stable orbit
with AoP of 0◦and the high Isp of the engine.
The details of stationkeeping are described in Figures 4.1.2, 5.1.2 and 6.1.3.
7.4 Mass 91
-80 -60 -40 -20 0 20 40 60 800
20
40
60
80
100
120
South - Lunar Latitude [deg] - North
Ground Resolution [m]
Achievable Ground Resolution by the Three Options
Direct Chemical option
WSB option
Low Thrust option
Figure 7.2.1: Achievable ground resolution of each design option over the six-monthperiod
7.4 Mass
During the design optimisation stage, it was aimed to keep the final spacecraft mass
below the given 300 kg. It was found that the major contributor to the overall mass
were the main propulsion, AOCS and Power subsystems.
During the design process, it was decided that any mass left would be used to carry
stationkeeping fuel. An algorithm to do so has been embedded into the spreadsheets,
achieving launch mass of just under 300kg for direct chemical and WSB options.
For the low thrust option, however, it has been found that using this algorithm
would add tens of years’ worth of stationkeeping fuel. It has been decided that
this would not be necessary, as some other system would most likely fail before this
happens. As a result, the low thrust option weighs 282kg in total. This provides an
opportunity to accommodate further payloads, or redundancies, increasing the value
of the mission.
As a reference, it has been found that Ariane 5G charges $10000 per kg. Hence
the lighter the spacecraft is, the cheaper the launch would be.
92 7 EVALUATION AND COMPARISON OF TRANSFER METHODS
7.5 Risk Assessment
In the risk assessment table, Table 7.5.1, the X column quantifies the hazard on the
system, scaled from one to five, with one being minimal damage and five representing a
total loss of the spacecraft. The Y column is the likelihood of the hazard to occur. The
risk column shows the weighted factor applied on the systems, affected by the occurred
hazard. After considering the three options and its risk with following factors, it was
found that the direct chemical option possesses the least risk of total loss of mission,
that the low thrust option has the highest risk of the three options.
Failure mode Type of Hazard X Y Risk Redun-dancy?
Heaters Failure to maintain temperature; 5 0.1 0.5 YesCoating Penetration of solar flux into 5 0.3 1.5 Yesablation the on-board systemBattery Inability to store electric power 5 0.1 0.5 Nofailure from solar arraysSolar arrays Failure to generate solar power 4 0.3 1.2 Yeswiring failureSolar array No generation of power leading 5 0.5 2.5 Nodeployment to mission failurePCDU Failure to distribute power 3 0.4 1.2 YesMain engine Orbital insertion failure 5 0.1 0.5 NoValves Improper flow of propellant to the 3 0.6 1.8 Yes
engine causing variable thrustthen desired
Fuel Explosion and damage to 5 0.2 1 Noleakage / subsystems while posing a threatfreezing to thermal subsystemReaction Unable to move the spacecraft 3 0.1 0.3 Yeswheels along its three axisSensors Failure to perform orientation 2 0.2 0.4 Yes
around a reference pointThrusters Unable to perform Attitude 5 0.5 2.0 Yes
Control manoeuvresPayloads Unable to meet primary 4 0.1 0.4 Yes(primary) mission requirementPayloads Circuit imbalance and inability to 2 0.1 0.2 No(secondary) achieve secondary objectivesComputers Loss of control over a subsystem 5 0.3 1.5 YesDownlink Unable to send data back to 5 0.5 2.5 Yesfailure ground station causing on board
data handling crashUplink Unable to send command to the 4 0.2 0.8 Yesfailure spacecraft causing communication
malfunction and might lead tomission failure
Table 7.5.1: Assessment of system points of failure.
93
Transfer Options Major Failure modes causing Totalmission failure Risk
Direct Chemical Main Nozzle, Valves, Fuel Leakage,Sensors, 5.6Option Payloads, Computer, Communication, HeatersWeak Stability Main Nozzle, Valves, Fuel, Leakage,Sensors, 7.1Boundary Payloads, Computer, Communication, Coating
ablation, HeatersLow Thrust Option Main Nozzle, Valves, Fuel Leakage,Sensors, 10.8