You are on page 1of 16

52nd International Conference on Environmental Systems ICES-2023-448

16-20 July 2023, Calgary, Canada

Design of an Actively Shuttered Dust-Resilient Radiator for


Lunar Applications

Andrew S. Gibson1 and Dominic Bailes-Brown2


ESR Technology Limited, Warrington, UK, WA3 6WU

Angel Iglesias3
Almatech, CH-1015 Lausanne, Switzerland

Martin Humphries4
Spacemech Limited, Bristol, UK, BS36 1RH

Simeon Barber5
Space Science Solutions Limited, Milton Keynes, UK, MK5 7AB

and

Philipp B. Hager6
European Space Agency (ESA/ESTEC), Noordwijk, Netherlands

The Moon presents an extremely harsh environment for robotic and human exploration,
with diurnal temperature cycles spanning nearly 300C. Active thermal control is widely
regarded as a key technology to enable surface assets to survive lunar night. We report on
progress in the development of a compact and scalable, actively shuttered radiator, designed
to be resilient to the dusty environment that will be encountered by many upcoming lunar
missions, such as Argonaut/EL3. Radiator function is influenced by extreme temperature
variations, where thermal cases must consider solar input and IR heating from the surface
during the lunar day as well as heat losses during the lunar night. An actively shuttered
approach enables closure of the radiator to minimize heat losses at night and/or to protect the
radiator from contamination during events with high expected dust deposition, such as
landing, astronaut, rover or robotic activities, or the passing of the day/night terminator. The
overall approach has been to maximize functionality across a wide range of lunar cases, with
an emphasis on polar scenarios including lunar landers, payloads and rover applications. The
paper describes progress following one year of development, to adapt a heritage shutter
concept flown on Rosetta and Giotto missions, with validation of new dust-resilient elements
via component-level breadboard activities. An overview of the Engineering Model design is
provided, highlighting thermal design choices as well as the thermal verification approach.

Nomenclature
λ = Latitude
𝛩𝐿𝑅 = Radiator orientation with respect to the lander
𝛩𝐻 = Radiator orientation (azimuth angle)
𝛾𝐿 = Lander orientation with the surface

1
Head of Mechanisms, ESR Space, andrew.gibson@esrtechnology.com.
2
Mechanism Design Engineer, ESR Space, dominic.bailes-brown@esrtechnology.com.
3
Thermo-Mechanical Analysis Engineer, Almatech SA, angel.iglesias@almatech.ch.
4
Director, Spacemech Limited, martyehum52@gmail.com.
5
Director, Space Science Solutions Limited, simeon.barber@spacesciencesolutions.co.uk.
6
Thermal Engineer, Thermal Control Section, Keplerlaan 1, 2200 AG Noordwijk.
β = Louvers opening angle
𝑄̇ = Heat dissipation
α = Absorptivity
ϵ = Emissivity
i = Angle of solar incidence
𝐻𝐿 = Lander height
φ = longitude (φss = sub-solar longitude)
BOL/EOL = Beginning of life/End of life
ECSS = European Cooperation for Space Standardisation
EL3 = European Large Logistics Lander (also known as Argonaut)
ESA = European Space Agency
H = Horizontal configuration with respect to the ground
HP/LHP = Heat pipe/Loop heat pipe
IR = Infrared
LDE = Landing Descent Element (of the EL3)
LDRLR = Lunar Dust Resilient Louvred Radiator
OSR = Optical Solar Reflector
PAC = Percentage annual coverage (obscuration from regolith)
PASS = “Passenger” module that sits on top of the LDE module
ROV = Rover
Tbol = Temperature (measured bolometrically)
UKSA = United Kingdom Space Agency
UV = Ultraviolet
V = Vertical configuration with respect to the ground
VDA = Vapor deposited aluminum
VIS = Visible

I. Introduction

T hermal shutter and louver devices have been enabling technologies for many early space exploration missions,
such as the ESA-led missions, Giotto and Rosetta, as well as for American led missions including Viking, Mariner,
Cassini and Magellan (Gilmore, 2002) and others. These devices have proven to be reliable way to obtain high turn-
down ratios for temperature regulation of critical radiator elements of spacecraft, where cold survival, mass and
reliability are critical to mission feasibility and performance. The properties that were desirable for those missions are
also key for successful thermal control on the lunar surface, noting the main requirement difference, being able to
work in a dusty environment, which can affect heat rejection, as well as mechanism reliability.
With many projects targeting a soft landing on the lunar surface and extended operations there, the thermal control
challenges associated with the lunar surface environment for longer durations have become apparent. With needs

Figure 1. Giotto thermal shutter Engineering Figure 2. Thermal shutter (solid model) for use with a 360
Model. Heritage thermal shutter mechanism, mm x 600 mm radiator. Shown in a partially open state with
photo provided courtesy of Spacemech Limited VDA film depicted as transparent to aid visualization.
2
International Conference on Environmental Systems
projected related to the European Argonaut/EL3 lander as a multi-mission platform (to deliver a high mass to the lunar
surface by 2030), ESA selected a technology development proposal led by ESR Technology to develop a thermal
control mechanism which would be based on heritage thermal shutters from Giotto and Rosetta missions, tailored for
the lunar environment and scalable across a range of sizes to suit many mission profiles. This development program
has been performed under a European Space Agency contract, having commenced in September 2021 and is expected
to complete by the end of 2023. Almatech play a key subcontractor role, leading thermal and structural analysis tasks
as well as having responsibility for motor drive and film tensioning elements.
Two mock-up assemblies were built during the first phase of work to demonstrate competing concepts from the
original proposal, as described below. The build demonstrated the feasibility of the mechanisms and helped identify
challenges in the build sequences, highlighting the importance of particular flex-pivot parameters as well as some
sensitivity in alignment of the film and aspects of build needed to achieve tensioning of belts and film.

A. Thermal Shutter Concept


The thermal shutter mock-up was developed by
Spacemech, having produced a working model, driven by
a brushed motor/gearbox combination via stainless steel
belts. Wheels, motor mounts and radiator supports were
3D-printed, and a black polycarbonate plate installed
internally to represent the radiator. With the radiator
integral to the assembly, this configuration may be
considered for thermal control of smaller payloads,
whereas the baseline configuration involves independent Spreader bar
mounting of radiator and shutter onto the spacecraft, rover
or instrument.
The cover of the spreader bar (refer to Figure 3) and
roller tube were made from carbon fiber, with belt tension Figure 3. Functional Shutter Mock-up (½ scale
provided by a 3D-printed flexure within the spreader bar shutter, with internal radiator plate 180 x 300 mm).
(under cover), with the film tensioning provided by rotary Reflective black surface represents mock radiator,
springs in the spool sub-assembly. integral to the unit.

B. Thermal Louver Concept


The louver mock-up developed by Almatech drives
via a 4-bar linkage which pushes/pulls a tie rod, on which
is wound a metal band. This band surrounds, links and
activates the flexible pivots to rotate in unison. This
system allows synchronous movement of the louvers
using a single motor without elaborate electronics.
The main advantage of the proposed flex-based
louver system is that it can be configured to provide a
robust and reliable active thermal control system for
extreme environments. It is well-suited for lunar
applications where dust contamination presents a
considerable risk to reliable operation.
The use of flexures to support the moving louver
blades has the advantages of not requiring any lubrication
and being insensitive to thermal extremes including
Figure 4. Functional Louver Mock-up (Reduced version
cryogenic temperatures, with the potential for high
with 3 vanes out of the nominal 11, blade length=150mm,
performance and a long lifetime.
width=44mm). Photo courtesy of Almatech.
C. Summary of Requirements
In general, the radiator design is driven strongly by environmental requirements, to be able to provide heat rejection
and operate across a wide range of temperatures while protecting the radiator from solar (UV/VIS) and lunar infrared
heat flux, as well as dust, with the ability to open partially. The radiator design/analysis must consider pointing

3
International Conference on Environmental Systems
variations due to variability of terrain of the order of ±15.
The capability to minimize heat losses is fundamental during
cold phases, primarily for lunar night survival, but also during
transit to the Moon.
Parameter Requirement Details/Values

Lifetime Related 1 Earth year [goal of 2 years with an


estimated 10000 cycles (ECSS factored)],
30 krad radiation

Thermo-Optical ϵ = IR emissivity < 0.2 (BOL)


Properties (closed)
α = UV absorptivity < 0.2 (BOL)
< 0.25 (EOL)

Effective Radiator ϵ =IR emissivity > 0.80


Properties (open) α =UV absorptivity < 0.20 (BOL)
< 0.25 (EOL)
Figure 5. Typical Configuration (shutter shown in
Mass Specific mass of < 9 kg/m2
gold) Configurations may be vertical or horizontal. + 0.5 kg/m2 for harness
The requirement to provide active control of the Scalability Adaptable for radiator areas in the range
louver/shutter was mainly driven by the presence of of 0.06 to 2 m2
dust on the lunar surface, being able to regulate
temperature of the payload with varying dust coverage. Agreed Baseline Width x Length x Depth
It was considered relevant to be able to actively open dimensions 420 x 700 x 45 mm (shutter)
(scalable–up/down) 360 x 600 mm (radiator)
or close the louver/shutter whenever necessary, i.e.,
potentially during the last part of the landing, the Temperature Range -150°C to +70C (shutter operating)
crossing of the terminator, local activities of humans, -150°C to +120C (shutter survival)
the movement of rovers or robotic arms. -223 to 120C (surrounding surface)
The shutter must be scalable within a wide range of
sizes, where larger surface areas may be obtained with Power < 5W/ m2 (no allowance for heating)
multiple radiators, as these would logically be
distributed around a lander or larger installation to cope Table 1. Summary of Key Requirements for the Dust-
with variable illumination, due to rotation or changing Resilient ‘Louvered’ Radiator.
sun vector.
Specific requirements assert that only one actuator sub-system may be employed in the system, which limits
redundancy options to the benefit of mass and complexity. Scaled requirements for resources translate to a maximum
power of 1.08 W and maximum mass of 1.945 kg for the baseline design (0.216 m 2). In terms of specific verification
requirements, closed-loop control of the radiator must be demonstrated to provide controllability to ±1C with
adjustment of the protective film.
Interfaces are to be tailorable for mounting to aid heat transfer from heat pipe or loop heat pipe transport systems,
directly to payloads or electronic boxes, etc. The louver/shutter was designed to be mostly agnostic to the radiator it
is protecting, providing flexibility to adapt to different client-specific hardware. The radiator itself could then be
further equipped with a thermal switch or loop heat pipe to reduce the heat losses. This is also the rationale for the
potential low non-operational temperature range of the louver/shutter.

II. Lunar Environment Scenarios and Initial Study Results


The shuttered radiator is being developed primarily for lunar exploration missions, specifically targeted for robotic
or crewed modules and rovers. These three application types present multiple-use configurations related to mission
parameters such as landing site or exploration latitude and specific location of the radiator on the module with respect
rejection performance of the latter while limiting as much as possible direct or indirect exposure to the Sun due to
reflections and solar trapping.to the Sun. The specific topology of the terrain around the landing site presents a second
order of magnitude effect. The precise location of the radiators must be selected according to the specific application.

4
International Conference on Environmental Systems
The thermal surface
conditions shown in Figure 6 were
considered, which have been
produced from the average of 24
bolometric thermal mappings

T bol [K]
generated at 15° increments of

Latitude
subsolar longitude normalized to
the subsolar point (0°, 0°).
For the thermal cases studies in
progress, we have chosen to assess
performance at three latitudes
(Polar λ ≅90°, λ =45°, equatorial λ
≅0°) scenarios to provide a general
overview of the radiator ϕ-ϕss

performance boundaries for all Figure 6. Moon surface thermal map versus latitude, corrected longitude with
expected use conditions. Because respect to sun longitude illumination: ϕ-ϕss (Jean-Pierre Williams, 2016) .
the study cases are transient over
the lunar day, the longitude is unimportant, and we utilise 0° as the default.

A. Thermal Control Considerations for


Potential Lunar Missions
The international space community has a
clear interest in extended human and robotic
exploration activities on the Moon. Several
landers are currently proposed by commercial,
national and international organizations, varying
in shape and capabilities, all of which seek to
transport hardware or humans to the lunar.
Landers, instruments, rovers, and infrastructure
alike will need to cope with a harsh lunar surface
environment, most notably a wide temperature
range and the presence of dust. Thermal control
solutions might vary from radiator/heater-based Figure 7. Argonaut lander delivers a combination of cargo items,
passive designs for short-term missions to more scientific payloads and small robotic assets (artist impression -
optimised active control systems in line with ESA Terrae Novae 2030+ Strategy Roadmap, June 2022.)
sustaining a longer-term presence.
Where the mission profile requires operation in a wide variety of environmental conditions, a range of operational
scenarios and is exposed to lunar dust, a shutter or louver system is an attractive option for thermal control, enabling
variable emissivity and the means to protect the radiator. Potential applications are anticipated, arising from missions
studied at ESA, such as Argonaut (formally known
Azimuth angle H
as the European Large Logistics Lander EL3)
landing descent element and the payload hosted on
Radiator orientation top of Argonaut. Potential payloads range from a
wrt the lander LR Radiator
suite of scientific instruments, technology
demonstrators to unpressurized rovers, or cargo for
Lander Height HL Artemis, as outlined in the ESA Terrae Novae
2030+ roadmap. Science payloads may remain
Lander orientation attached to the lander or be off-loaded to be stand-
wrt lunar surface L alone on the surface of the Moon. All such payloads
Lunar surface
require thermal control and might individually
benefit from a shuttered radiator, either alone or in
wrt = with respect to Landing Site at Moon Latitude  combination with other thermal control hardware
such as thermal switches, loop heat pipes, or
Figure 8. Design parameters for concept comparison pumped fluid loops.

5
International Conference on Environmental Systems
B. Initial comparison of shutter and Day Cases
louver performance Case Latitude Panel Shutter: Potential Louver: Potential heat
To evaluate as many usage scenarios as λ orientation heat dissipation – dissipation – maximum
possible for both designs (shutter and maximum [W/m2] [W/m2]
louvred), a simplified 2D model has been 1 0° Vertical 0 232.1 (90° opening)
(no heat rejection) 166.9 (67.5° opening)
developed in Python code including the 76.4 (45° opening)
parameters of orientation and possible 2 0° 45° 167.7 108.6 (112.5° opening)
position of the radiator on the lander and 44.6 (90° opening)
latitude on the lunar surface (refer to 0 (67.5° opening)
Figure 8). The code also allows to 3 0° Horizontal 248.0 0 (any opening)
evaluate the performance of the radiators 4 45° Vertical 0 70.8 (112.5° opening)
with respect to the degradation of optical (no heat rejection) 0 (90° opening)
performance mainly due to the deposition 5 45° 45° 146.8 0 (any opening)
of regolith dust, by assuming 6 45° Horizontal 322.4 0 (112.5° opening)
proportionate emissivity changes with 88.3 (90° opening)
obscuration. For the purpose of the study, 138.2 (67.5° opening)
to enable comparison of options, the lunar 7 45° -45° 400.4 341.3 (112.5° opening)
351.3 (90° opening)
surface absorptivity and emissivity 322.4 (67.5° opening)
applied were 0.93 and 0.97 respectively. 8 90° Vertical 212.5 0 (any opening)
Radiator surfaces were assumed to have 9 90° 45° 311.9 0 (112.5° opening)
emissivity of 0.90 and absorptivity of 75.1 (90° opening)
0.20. The emissivity of retractable parts 129.6 (67.5° opening)
(VDA and louvers surfaces) was assumed 10 90° Horizontal 501.6 360.4 (112.5° opening)
to be 0.10, while absorptivity of 380.3 (90° opening)
retractable parts was assumed to be 0.20. 360.4(67.5° opening)
The main conclusions of this Night Case
preliminary study are that the radiator 11 - - 8.3 8.3
design with the shutter has the best heat Table 2. Shutter versus louver heat rejection performances
rejection performance without
considering external heat sources due to larger effective emissivity. For this reason, this design was chosen for the
development of the Engineering Model (EM).
Table 2 shows that heat rejection capability facing the Sun is possible for the shutter design in horizontal
configuration but not for the louver type as it acts as a Sun trap. In the vertical configuration (w.r.t. the ground) heat
rejection is not possible for the shutter for most of the latitude cases. With low emissivity surfaces, the cold case
(Night) can be handled readily for both designs with shutter and louvers assumed to be completely closed. For polar
locations, the shutter does not require a baffle.
To be able to use the same radiator design
for all of the desired use cases, it was
necessary to consider a hybrid design
including a shutter with fixed baffles. In
addition to this measure, the radiators in
vertical configurations would need to be
considered for use in pairs, placed on the
opposite sides of the LDE in such a way that
when one is facing the Sun in a closed
configuration, the other one is able to reject
heat by remaining completely open.

C. Lander/Rover Scenario Assumptions


For the detailed ESATAN thermal
analyses, three geometrical application Figure 9. Use case configurations a) V-LDE = vertical, side-
scenarios have been considered (see Figure 9). mounted location on lander/descent module,
In the first scenario the radiator module center b) H-PASS = horizonal, top-mounted to passenger payload,
is placed vertically with respect to the ground c) H-ROV = horizontal rover mounted location.

6
International Conference on Environmental Systems
on the octagonal surface of the LDE module at approximatively 2.25 m from the lunar surface. In the second
application, the radiator is placed horizontally with respect to the ground on top of the PASS at approximately 5 m
from the lunar surface. Finally, the radiator module is placed horizontally with respect to the ground at 1m from the
lunar surface on a simple box shape representing a rover application.

III. Thermal Performance Prediction

A. Analysis approach and Analysis modelling parameters


The analysis approach involves generating a model of a representative lunar environment using the ESATAN
software to evaluate potential heat rejection and radiative interactions for key use cases. For this, a planetary surface
radiative case was considered with the orbit of the earth. This modelling is approximate because:
• the effective epicycloid trajectory around the sun with an oscillation of the +/- 400,000 km with respect to the
Earth reference orbit is not considered directly.
• to cover cold and hot cases, the Sun flux is adjusted between the minimal value of 1315 W/m2 and the
maximum value of 1421 W/m2.
• the slight inclination of the lunar orbit with respect to the plane of the ecliptic is only considered as a
modification of solar declination parameter.
• for initial studies the topology of lunar surface around the LDE and the ROV is assumed to be flat.
A finer mesh is considered for the modules and the lunar surface around them, in order to capture as much as
possible the interaction effects of the surroundings elements with the Radiator, referring to Figure 10.

B. Use-Case Studies
At radiator level, for each lunar latitude there are hot and
cold cases which depend on the location of the radiators on the
module or the Rover, as indicated in Figure 10.
The hot worst-case occurs when the Sun is facing
perpendicular to the radiator and the surrounding environment
around the radiator is also exposed to orthogonal solar flux.
This unfavourable scenario occurs for the H-PASS or H-ROV
layout during an equatorial mission scenario. This scenario is
unlikely because, for such missions, the preferred location of
the radiators would be vertical to the LDE to maximize heat
rejection performance. More realistic operational cases are a)
defined from Polar to 45° of latitude location.

b)

Figure 10. a) ESATAN model of lunar surface


Figure 11. Thermal shutter as modelled in H-PASS showing finer localized mesh and coarser mesh
configuration (top of lander) toward the horizon, b) Temperature map [oC]
approaching sunset at a latitude of λ=45°

7
International Conference on Environmental Systems
For the cold case, the most critical configurations are at the Poles, where the radiator at the anti-Sun side of the
spacecraft will be in darkness, shielded from the light by the spacecraft body.
For the calculations, it is assumed that there is no thermal conductive coupling between the radiator and
surroundings (i.e., LDE, PASS or ROV) depending on the application considered. Only a single node dissipating 20W
is connected to the radiator surface. To be conservative, in terms of radiative coupling, the thermo-optical properties
of the surrounding elements are set to values typical for a solar array (ϵ=0.84, α=0.91) for the LDE, PASS and ROV
cases. Due to poor thermal conductivity of the Regolith, only radiative coupling is considered (with the following
thermo-optical properties for the lunar surface: ϵ=0.97, α=0.82).
The transient cases to be studied are focussed on the mechanism itself. The main objective is to assess effects on
radiator radiative performance when the radiator is affected by short transient phenomena such as actuation of the
shutter, sunrise or sunset or when the radiator is shadowed by an external element such as a solar array or an astronaut.
These transient cases will not be presented in the frame of the current paper.

C. Preliminary Analysis Results


The highest temperatures are reached for the
equatorial case when the Sun is at its zenith, as indicated
in Figure 12. This hot case is considered extreme and
results obtained considering the dissipating node show
that it is generally not possible to have effective heat
rejection if the radiator is not always in the shadow.
For the cold case, we consider the case where the
radiator is placed horizontally on the top of the module
(PASS) at a latitude close to the lunar pole. The chosen
evaluation period is just at the end of night. In this case,
we want to assess whether or not the shutter will limit the
emissivity of the radiator so that heat loss is not
excessive. It has been noted that considering each of the
three use cases (defined in Figure 9), there is no Figure 12. Equatorial thermal map (Figures in C)

significant difference in the temperature range despite


changes in attitude and sunlight incidence. Only the parts
of the module which are more exposed to radiation show
significant variations between cases.
During lunar night, the goal is to limit as much as
possible the overall heat rejection in order to minimise
survival heater power required by the landing module or
the Rover. Indeed, during night all external surfaces act as
a radiator. Table 3. Key thermo-optical properties used in
the thermal model of the lunar dust resilient radiator,
showing the maximal overall power heat rejection of the
radiator in Watts.
The shutter can also be used during the day to perform
thermal control via stepper motor position selection, if the
Figure 13. Cold case during the night close to Polar supported unit needs to be regulated to remain within a set
latitude (Figures in C) range of temperatures.
EOL values of α and ϵ are computed using EOL
primary values of the coating and considering a degradation of properties due mainly to Regolith percentage of
coverage (PAC) (see Table 3). Final ϵ is then calculated as:
𝜖𝐸𝑂𝐿−𝑅𝐸𝐺 = 𝜖𝐸𝑂𝐿 ⋅ (1 − 𝑃𝐴𝐶) + 𝜖𝑟𝑒𝑔𝑜𝑙𝑖𝑡ℎ ⋅ 𝑃𝐴𝐶 considering a PAC equal to 10% for a one-year mission.

Table 3The maximal heat rejection capabilities occur during day with BOL thermo-optical properties and during
night for EOL properties.

8
International Conference on Environmental Systems
Results presented in Thermo-optical
α BOL ϵ BOL α EOL ϵ EOL α EOL ϵ EOL
Table 4 Table 4. Effective Properties
corrected corrected
radiative heat rejection for
the Polar caseclearly OSR 0.200 0.90 0.20 0.90 0.262 0.906
demonstrate the advantage of Beta Cloth 0.248 0.86 0.267 0.86 0.322 0.870
a shutter during the night to White paint 0.15 0.24 0.20 0.24 0.262 0.312
cover the radiator, reducing VDA film 0.07 0.04 0.07 0.04 0.145 0.132
heat rejection. Indeed, the Regolith 0.82 0.96 - - - -
radiated power can be Table 3. Key thermo-optical properties used in the thermal model of the lunar
reduced by more than a dust resilient radiator
factor of 15, which may be
enabling for the survival of equipment not rated to survive cryogenic conditions. Comparing results, we see that the
performance in the H-ROV cases differ slightly to those of the H-PASS, while both are horizontally oriented. This is
mainly due to alignment of the radiator (long side aligned to the sun path) and adjacent surfaces which are less
impactful compared with larger analogous surfaces for the H-PASS case.

Moon Configuration Shutter Fully Shutter Fully Rejection Ratio


Latitude Open [W/m2] Closed [W/m2] Open/Closed
Day (max) 92.3 8.78 10.5
H-PASS
Night (max) 89.3 9.28 9.60
90° Day (max) 51.9 18.3 2.80
V-LDE
(Polar) Night (max) 50.5 18.8 2.70
Day (max) 88.3 5.24 16.8
H-ROV
Night (max) 85.2 5.44 15.7
Table 4. Effective radiative heat rejection for the Polar case

a) H-PASS config shutter closed b) V-LDE config. shutter closed c) H-ROV config. shutter closed

d) H-PASS config shutter open e) V-LDE config. shutter open f) H-ROV config. shutter open
Figure 14. Comparison of radiator surface temperatures as calculated for open and closed configurations in
the three use-case configurations - Polar Latitude case when the Sun is on the Zenith.
9
International Conference on Environmental Systems
Similar results have been obtained for other latitude cases, with general agreement in line with trends noted from
Table 2. Values obtained in Table 4 do vary according to the radiator orientation with respect to the Sun’s rays. It was
also noted that the surface of the radiator is evidently well linked to the dissipating node which induces almost constant
temperature over the radiator surface when not directly exposed to the sun.

IV. Thermal Shutter Mechanical Design


The design of the thermal shutter is akin to a solar blind, designed to withstand extreme temperature variations
and resilience to dust exposure. The sections below explain the heritage link to early thermal shutters flown on ESA
deep space missions and provides descriptions of the construction and logic behind the mechanical/material
construction and the approach taken to deal with related risks associated including electrical charging effects.

A. Thermal Shutter Heritage


Giotto was Europe’s first deep space mission, designed and built by British Aerospace (BAe) at Bristol, UK, to
study Halley’s Comet. In 1982, it was determined that the baseline thermal design of the Giotto spacecraft was
insufficient to ensure compliance to thermal design requirements, which led to the incorporation of a variable radiating
area. The Giotto spacecraft shown in Figure 15 below included a useful radiating area of 0.060 m2 per shutter
mechanism (out of an area of 0.086 m2), using goldized Kapton film and OSRs for radiator surfaces (F.Felici, 1983).

1. Frame Top Member 6. Pulley


(detached for clarity) 7. Kapton Film
2. Stepper Motor 8. Spreader Bar
3. Synchronisation Shaft 9. Micro Switches
4. Bevel Gears 10. Frame
5. Kevlar String
Figure 15. Giotto Heritage Shutter (indicated by red arrow) - built under ESTEC funding, noting that
M.Humphries of BAe (now with Spacemech) served as ‘Shadow Engineer’ assisting Fokker/CASA, (F.Felici G. , 1983)

These were developed and tested by Fokker, with the support of BAe.
The shutter was driven by a stepper motor, covering and uncovering radiators mounted on curved body panels and
was operated from ground control. The mechanism mass was about 1kg, and incorporated pulleys in tension that used
Kevlar cords that underwent radiation testing to prove stability. The mechanism was qualified with 150 full operational
cycles and 75 in thermal vacuum. A development model was tested from -40oC to +80oC, followed by a later solar
simulation test, measuring the heat exchange control capabilities at different solar aspect angles, checking the
functional characteristics under realistic gradients (F.Felici, 1983).

B. Material Selection (including finishes) for Thermal Performance


The majority of the shutter body is to be made from aluminum alloys, with attention to coefficients of expansion,
to mitigate against bulk dimensional changes over the 220C operational span of temperature from cold to hot plateaus.
Most of the frame is made from an aluminum alloy. Along the length, the grade of aluminum has been selected to
provide a nominal thermal expansion match to the belt/wheel drive system. This has the benefit of reducing the
predicted swing in belt tension preload over temperature. The baseline VDA selected is a double-sided aluminum
coating onto a 50 micron thickness of polyimide (BOL ϵ=0.04, α=0.07), while this has yet to be fully confirmed
10
International Conference on Environmental Systems
through thermal analysis as sensitivity studies are ongoing. Options will be studied to maintain the polyimide film
within its safe temperature range, where net heating results from the absorptance value being significantly higher than
the emissivity. Polyimide film has an excellent resistance to radiation and typically survives to temperatures over
350C, therefore it is a reasonable candidate with considerable heritage. This material will be tested for a minimum
of 200 cycles of operation at the base temperature of -150C.
It is being considered to coat internal surfaces of the shutter walls and motor with a reflective paint with a relatively
low α/ϵ ratio to avoid overheating, where this is being analysed as part of the ongoing phase of work. The effect is
hypothesized to avoid the walls being a significant heat trap and could increase peak heat rejection by increasing the
effective field of view of the radiator by means of reflection.

C. Thermal Interface Approach


The interface is assumed to mount via eight M4
fasteners with effective isolation provided with the
use of 2 mm thick insulating washers, with 3 along
each of the longer side supports and 1 at each end, to
stiffen the motor mount. It is assumed that these 8
fasteners provide the mechanical connection to the
spacecraft, rover or other payload structural mounts.
The radiator itself is independently mounted to the
spacecraft structure. Mounting bolts are accessible
from the top, while the covering MLI will need to be
pulled back to reach these holes. The outside of the Figure 16. Mechanical Features and Mounting
shutter will be covered with MLI. The current Arrangement (underside of thermal shutter showing isolating
development will utilise a cryogenic blanket layup spacers, indicated by red arrow, implemented at mounting
of polyester, with double-sided layers will be points).
installed for vacuum testing/model correlation,
which is suitable as no solar simulation testing is planned at this time. The effect of other exterior coverings is being
considered within the thermal analysis to establish a baseline recommendation for flight. The design of an optional
fixed baffle design is at the concept stage, with the volume inside of the belts reserved for this sub-assembly.

D. Mechanism-Driven Idler
Design Aspects wheel
The design fundamentally
consists of a belt drive, Film
actuating a tensioned film spool
using an electric motor Stepper (outer) Film
driving through a gear box motor & tensioning
and magnetic coupling. An gearhead Drive provided
advantage of this approach is linkage by large-
(Inner) angle flex
that a degree of scalability can Magnetic
pivots
be achieved very easily either coupling
by adjusting the film width or
the actuation length, which
only requires a different Belt
length of belt. As stated Figure 17. Thermal Shutter drive and film tensioning system
above, the breadboard is
being designed for an assumed radiator size of 0.60 m by 0.36 m; making either dimension smaller would be straight-
forward to implement, as would increasing the width. Increases in actuation length would require a modification to
the film tensioning system but is otherwise simple for smaller scaling factors. Greater scalability factors are very
possible but further modifications would be required to the motor (and gear box), film tensioning system and overall
structure to maintain a reliable and well mass-adjusted system. As an example of this, a version designed to fit onto
one side of a 1U CubeSat has been developed, demonstrating the possible range of scalability.
Figure 17 is a cross section diagram of the design highlighting several key features. The belts interact with sprocket
wheels, which rotate on metallic bushings. A number of dust mitigation measures have been implemented, from simple

11
International Conference on Environmental Systems
techniques such as selecting geometry to minimize dust ingress into sensitive areas, through to more elaborate
measures such as lightly preloaded polymer seals used to keep dust out of the wheel bushings.
To address the potential dust sensitivity of the gearbox, a magnetic coupling
has been implemented, allowing the motor and gear box to be in a completely
sealed enclosure. The magnetic coupling itself is then also protected with
sliding seals but is inherently more resilient to small amounts of debris and
designed to accommodate gapping such that it would take significant dust
ingress to begin to reduce performance. The motor and gearhead are
completely sealed and will therefore be insensitive to dust exposure.
The tensioned film is supported by a spool, suspended on large angle flex-
pivots produced by Almatech and made from a high-performance engineering
plastic (refer to Figure 18); these are to be tested for low-cycle fatigue effects
at low temperature and were selected for resilience to dust. Snubbers are
implemented to ensure the motion of the tube is limited axially and radially. Figure 18. Large Angle flex-pivot
supports for tensioning of VDA
E. Management of Electrical Charging Effects film roller
Charge management has been carefully considered by design, to ensure that
all metal components that might accumulate significant charge are electrically connected to a ground strap. The
bushing system for the roller film, motor and idler wheels are intended to retain electrical conduction with wheel
rotation. The bushings are made from a bronze material running against KolsterisedTM stainless steel spigots with an
advanced sputtered lubricant coating, referred to as MoS 2-202, provided by ESR Technology. This lubricant is
expected to provide adequate electrical conduction through the bushing assembly to allow charge bleeding (while
maximizing lifetime), whilst a backup solution would be to utilise MoS2-301, which includes a metallic dopant that
could potentially increase conductivity. Wheels are coated with PVD silver for both tribological and conduction
enhancement for contact with the drive belts. The VDA film is bonded to the aluminum support roller using an
electrically conductive adhesive, such that charge accumulated in the tube can be bled through the VDA film via the
spreader bar, drive belts, wheels, bushings and main structural supports via a dedicated grounding strap. The
verification approach for bushing contact conductance is described below.

F. Component Prototyping
Component prototypes have been Electrical Spool
Flex coupling
designed and are being tested to validate new (driven via
aspects of mechanism design, with a focus on feedthrough)
Preload
dust resilience and stable friction Masses Wheel
characteristics over the lifetime of the device. Preload Spring
Bushing
Two key early validation tests have been Spigot
designed, to de-risk critical elements of the Electrical
design; these are a vacuum life test and in-air Insulation
dust resilience test. Both of these activities High Accuracy
Triaxial Torque
will be testing the bushing, wheel and belt Force/Torque Transducer
interactions and are largely tribological tests, Transducer
though they will also include dust seals. This
will allow the main tribological elements of
the design to be validated both for general
vacuum behavior and for dust resilience. Figure 19. Test Rig for Bushing Lifetime and Conductivity Study
(cross-section of driving and driven bush assemblies to mimic shutter
1. Vacuum Life-test of Conductive belt drive, with electrical resistance and torque monitoring).
Bushings and Bushing Dust Seals
As depicted in Figure 19, a vacuum life test is configured with a stainless steel belt connecting two wheels where
both will be tribologically representative. Both the bushing and dust seals will be fitted and loaded to replicate the
design, with the belt tensioned accordingly. This will allow to characterize the torque and conduction behavior of the
system, assessing changes over lifetime. Tests will be conducted at ambient temperature with representative forces
applied forces representing worst-case thermal conditions.
The device total factored lifetime is 10,000 cycles; one cycle being defined as an opening and closing of the device.
Each pulley wheel completes approximately 4 rotations for a half cycle; thus, the total lifetime of each axle stub is
12
International Conference on Environmental Systems
approximately 80,000 rotations. Four full rotations will not be possible for the component prototype due to constraints
of the available test chamber, so the approach will be to apply equivalent rotations to that of the system level design.
The planned drive speed of the pulley wheel is 2 rpm; this is based on achieving the required actuation time with
margin (< 5 minutes), involving the assumption of completing 4 rotations for a half cycle. Note that for dry lubricants,
it is valid to accelerate the rotation speed for the life-test; therefore, a nominal drive speed of 15 rpm is baselined.

2. Component-Level Dust Resilience Prototype Manual crank


Testing Dust Injection nozzle
There are two critical areas where sealing is needed
to prevent the ingress of dust – these are the sliding seal
on the bushing and at the magnetic coupling. The
component breadboard tests will evaluate the
susceptibility to dust exposure for both concepts. Dust
testing, conducted with LHS1 and LHS1D will involve
exposure of the seal to a lunar dust simulant; initially
while the coupling is static, and then when the assembly
is rotating. The performance will be assessed by
measuring changes in torque during exposure.
The first release of dust will be performed on a static
setup, deemed more representative of exposure during a Torque sensing load cell
landing event. While not specifically calibrated, this trial located beneath flange
is foreseen to involve a controlled volume of regolith Figure 20. Component Dust Test Enclosure for
simulant (of order 1 mL) being released in a critical area metered deposition in localized areas (focused on sliding
of the mechanism (potentially sensitive to dust ingress). seal locations which protect radial bushings).
After the exposure, torque characterization will be
performed to establish if the dust has had a measurable effect on the system. The process will be repeated multiple
times. Any degradation or changes in performance will be noted prior to the final stage of testing, which will involve
applying dust to a rotating system. This is considered a worst-case in terms of proving dust resilience as the rotating
system will likely induce dust paths that are not present for a static system. As this test will be in air, rotations will be
slow to minimize air disturbances that could deflect the dust from the rotating components. After the rotational dust
exposure, a final torque characterization will be performed, followed by inspection.
Resilience to dust will be tested at component level to de-risk and identify susceptibilities. Attention will be
focused on axial dust seals, and specifically the seals protecting the bushings and the magnetic coupling, which

Radiator Surface

Electro-Magnetic
Sweeper Coil
Brush Block

Figure 21. Dust Sweeper Development for Dust Mitigation. UKSA grant-funded activity, Left:section view of
showing combined magnetic/brush block assembly, Right: results of sweeping for 2 different brush heights.

contains a lightly loaded polymeric sliding seal. A dedicated test rig (see Figure 20) enables measurement of the stub
axle torque reacted as lunar regolith simulant ‘dust’ is progressively applied to the target locations in controlled
quantities. Testing will cover cases where dust is applied to a stationary system and where dust is applied whilst the
mechanism is rotating. Testing will be conducted with lunar highland regolith simulant sourced from Exolith Lab at

13
International Conference on Environmental Systems
University of Central Florida. LHS-1 is baselined, and its derivative LHS-1D containing only particles finer than 30
µm, are ‘high-fidelity’ simulants blended from terrestrial minerals to simulate typical lunar highland regolith.

3. Potential enhancements for dust management


The extent to which dust could become lodged in the rolled-up blanket will be assessed in ambient dust testing
described in section VI. If found to be necessary, future enhancements to be considered would include the
incorporation of a brush element to lightly sweep the film prior to wrapping around the roller. This would take
advantage of the existing mechanism motion and could be applied for radiators or solar panel applications. In a
separate grant-funded development (UKSA grant for exploratory funding GEI-040), ESR has developed a vacuum
compatible brush based on a fiber material described by Gaier (James R. Gaier, 2011). A brush construction using a
type of fiber identified has been trialed in a testbed similar to the shutter mockup described above, sweeping at ambient
conditions with 1 and 2 rows of bristles and with 2 levels of interference (brush compression), utilizing JSC-1 simulant.
The results in Figure 21 show potential for further advancement of the idea, as optical measurements demonstrated
the ability to remove 1 mL of JSC-1 regolith simulant at ambient. The fibres did not significantly abrade Kapton film
or VDA tape after > 200 strokes, showing potential to be used to clean the film. An electro-magnetic sweeping coil
was also successfully trialed, removing virtually all magnetic iron filings from the radiator surface within 2 passes.
Further studies are proposed to test sweeping methods in vacuum with more realistic electro-static forces present.

V. System Level Thermal and Dust Resilience Verification


The development program has prioritized key thermal, mechanism and dust resilience themes for verification by
test, whereas structural specifications are verified by FEM analysis only. The heat rejection potential and
controllability will be demonstrated by analysis and test, with correlation analysis to be performed to ensure full
understanding of results and to improve the model as a building block for a future qualification program.

A. Thermal Test Verification


Chamber lid with LN2
A test setup has been designed feedthrough (Assembly
to demonstrate heat rejection via hangs from it)
GM Cryocooler
the shutter mechanism over a Assembly
wide range of realistic
temperatures spanning a range of
radiator temperatures from -30C Cold-finger

to +70C. The setup in Figure 22


will allow to control the
temperature of the thermal shutter Dual Heat Exchanger
frame and target plate on reverse of target
Multiple thermal with thermal strap
independently, enabling to strap attachment attachment points
execute the test campaign as per points
Error! Reference source not
Support frame
found. The tests will evaluate the (shutter & radiator)
Thermal Shutter Body
sensitivity of the film control by
Radiator Plate Thermal Target
measuring the temperature (High emissivity)
w/heaters
changes for a small number of
Conductive stand-offs
motor steps in Tests 7-9 as
(x 8) for shutter
identified in Table 3. The thermal
shutter mount will be controlled Figure 22. Thermal vacuum setup for verification of heat rejection,
via a large single stage Gifford- controllability and thermal performance of mechanisms.
McMahon cryocooler, with the
temperature adjusted with the use of heaters. The target plate will be cooled for the cold heat rejection test with the
use of two in-line LN2 heat exchangers, making use of spare cooling power to improve the base temperature using
thermal straps. The radiator plate for this test is classified as ground support equipment, being a 6 mm thick sheet of

14
International Conference on Environmental Systems
Cycles (2003 VDA film Test Item Temperatures (°C) Cooling Activation
Case Description
total) (% open) Shutter Radiator Target Cold Head LN2
0 Ambient Air - Run-in 200 22 22 22
1 Ambient Vacuum – Functional Test 200 22 22 22
2 Hot Non-Operational (Survival) 0 0% 120 120 floating
3 Cold Non-Operational (Survival) 0 0% -150 floating cold  
- Repeat cases 2-3 (3 full cycles) 0
4 Hot Operational - Functional Test 200 70 70 floating 
5 Cold Operational - Functional Test 200 -150 floating cold  
6 Ambient vacuum - Functional Test 400 22 22 22
a. Heat rejection test (ambient) 0 0% 22 22
7 b. Sensitivity vs film position 0.5 50±1% 22 measure < -75 
c. Heat rejection test (ambient) 0.5 100% 22 22
a. Heat rejection test (hot) 0 0% 40 70
8 b. Sensitivity vs film position 0.5 50±1% 40 measure < -50 
c. Heat rejection test (hot) 0.5 100% 40 70
a. Heat rejection test (cold) 0 0% -50 -30 
b. Sensitivity vs film position 0.5 50±1% -50 measure (Straps to
9 < -110 
radiator &
c. Heat rejection test (cold) 0.5 100% -50 -30 target)
10 Ambient Vacuum - Functional Test 600 0% 22 22 22
11 Post Test - Ambient Air Cycles 200 0% 22 22 22

Table 5. Summary of thermal test campaign for Engineering Model thermal shutter unit (green shading
represents near ambient temperatures, yellow represents above ambient and blue represents below ambient)
Al-1050A H14 material, instrumented with heaters and platinum resistance thermistors to monitor temperature
gradients.

B. System Level Dust Resilience Verification Approach


Ambient dust testing will also be performed on the breadboard system, as the final step in its verification campaign,
following thermal vacuum testing described previously. The final choice of dust test facility is still to be confirmed
but will implement witness plates that enable a quantitative estimate of the applied dust loading to be obtained.
The primary aim is to verify the general dust resilience of the mechanism designs, particularly those aspects not
already de-risked by the component breadboard test campaign. It is expected that the test will provide insight into the
susceptibility of the film roller operation to entrapment of dust and the implications for the VDA film being damaged
through cycling and could indicate the need for further work on dust mitigation measures to ensure sufficient lifetime.
Further work will be required to study the thermal-related impacts of residual dust on the radiator surface, as well as
other important thermal surfaces, such as the optional baffles which will be required for vertical configurations.

VI. Conclusions
The baseline design of a thermal shutter device has been described and compared for potential performance with
a thermal louver, where both types are capable of providing high turn-down ratios for specific scenarios. The shutter
has been chosen due to overall simplicity, given a bias for polar operation. A programme of mechanism de-risking
has been outlined with the thermal analysis approach described. The design of the frame and mechanisms utilise
materials to minimize expansion issues and retain compatibility for use over a very wide temperature range.
Advancements have been introduced versus heritage models, to enhance dust-resilience for lunar surface operations.
FEA thermal analysis has been performed with results indicating feasibility of the shutter for operation in 3 relevant
use cases where the shutter shows benefits to reduce gradients, assuming the utilization of OSR radiator surface
material. Further, the reduction of dust on the radiator by means of the added capability to protect with the VDA film
shows promise to ensure less degradation in performance expected, particularly for hot cases, assuming the expected
high reliability of the mechanism can be verified through testing planned and subsequent qualification.
The completion of prototyping and Engineering Model test campaign are expected to provide significant risk
mitigation, to ensure lifetime of the mechanism as well as to demonstrate suitability of thermal design, to confirm
suitability for a range of applications in upcoming lunar missions. Evaluation of heat rejection limits and sensitivity
studies will be performed to assess dust deposition impacts and predicted changes with thermo-optical degradation.
An optional baffle system remains to be detailed to cope with vertical cases where IR input from the lunar surface
and albedo effects can compromise performance. Efforts are planned to validate and improve the thermal model
through test correlation following thermal vacuum testing. Dust testing at system level will be key to understand
15
International Conference on Environmental Systems
weaknesses and identity dust entrapment zones, while the fidelity of this test is seen to be an incremental step toward
qualification, as it will not replicate electro-static forces. The need for design adjustments is foreseen ahead of
qualification, to optimise cleanability, management of residual dust effects and the optimisation of external insulation.

Acknowledgments
This work has been funded via an ESA Express Procurement (EXPRO+) contract, 4000135381/21/NL/KML. The
support of various experts at ESA is much appreciated including mechanism support from Elena-Cristina Paul,
materials support from Malgorzata Holynska and radiation environment support from Fabrice Cipriani. Thanks also
to Angus Bishop, Matthew Oldfield and Olly Poyntz-Wright for their support in development of the mechanism design
as wells for test setup planning and development efforts. Also acknowledged are Stefano Nebuloni for support on
thermal and mechanical analyses and Marc Lichtenberger and Fabrice Rottmeier for management support.

References
1Felici, F., Brouwer, G.F., “The Giotto Thermal Control Shutter,” 13th ICES Proceedings, San Francisco, 1983, pp. 337-344.
2Felici, F., Lindhout, P., “Thermal Design of the Giotto Spacecraft,” 13th ICES Proceedings. San Francisco, 1983, pp. 127-137.
3Gilmore, D. G., Spacecraft Thermal Control Handbook: Volume 1: Fundamental Technologies, 2nd Edition, El Segundo,

California, AIAA - Aerospace Press, 2002, Chapter 9 (Louvers).


4Gaier, J.R., “Evaluation of Brushing as a Lunar Dust Mitigation Strategy for Thermal Control Surfaces”, NASA/TM-2011-

217231. Glenn Research Center, Cleveland, Ohio: AIAA–2011–5182.


5Williams, J-P., Paige, D.A., Greenhagen, B.T., Sefton-Nash, E., “The Global Surface Temperatures of the Moon as Measured

by the Diviner Lunar Radiometer Experiment,” Icarus 283(E8), pp.307.

16
International Conference on Environmental Systems

You might also like