IN THE FIELD OF TECHNOLOGY DEGREE PROJECT
VEHICLE ENGINEERING
AND THE MAIN FIELD OF STUDY MECHANICAL ENGINEERING, SECOND CYCLE, 30 CREDITS STOCKHOLM SWEDEN 2019,
Preparation for a Thermal Balance Test of the MIST student satellite
FREDRIK UNELL
KTH ROYAL INSTITUTE OF TECHNOLOGY
The MIST project was founded in 2014 at KTH and has been worked on by roughly 100 students since.
Thermal engineering has been an important part of the project from start and a thermal model has been developed to predict flight conditions. The thermal model created by previous students has to be corre- lated by performing a thermal balance test. This test is going to take place in a thermal vacuum chamber at KTH Albanova during the period November 2019 to February 2020. This thesis describes the process of preparation for a Thermal Balance Test for MIST. Included in the preparation of the test are; developing the test specifications and definition of ground support equipment needed for the Thermal Balance Test in order to successfully correlate the thermal model.
To perform the test some of the equipment has to be replaced with replacement components. The battery and the solar panels are two expensive and crucial systems for the mission and has been chosen to be replaced with mass dummies. These mass dummies are used in the thesis both to keep the correct thermal capacity and approximate the conductive properties. The solar panel dummy plates can be drilled into, to make sure the cables for thermocouples and heaters can get inside the satellite. Heaters will be mounted on the satellite, both to create gradients across the satellite to be able to measure them, and to use as safety heaters to prevent the temperature to get to low and damage components. The heaters will be mounted on the battery replacement board, the solar panel replacement plates and the bottom cover plate.
The thermocouples to be used in the test are already installed in the vacuum chamber, which are of Type T class 1 with a temperature accuracy of ±0.5 ◦C. The chamber has 50 thermocouples installed where 24 are planned to be used to measure gradients across interfaces to be able to correlate them with the thermal model.
In this thesis the thermal model is simulated in the new vacuum chamber environment to investigate what temperatures can be expected in the test and find components and interfaces of interest to place thermocouples on. The test will start from ambient temperature of 20◦C, then change the temperature of the chamber to 0◦C, with a temperature ramp of 1.5◦C/min. The chamber will stay at 0◦C until the temperature of the satellite has stabilized within the chosen stability criteria of 0.5◦C over two hours. This set point is chosen to take measurements for the hot case. When measurements are made, the chamber will go to -50◦C and stabilize at that temperature to take temperature measurements for the cold case.
The temperature gradients measured in both the hot and cold case will be used to correlate the thermal model with the physical model.
Abstract i
Sammanfattning
MIST projektet grundades 2014 p˚a KTH och har arbetats p˚a av drygt 100 studenter sedan projektet startade. Termisk kontroll har sedan projektets start varit av h¨og vikt och en termisk modell har tagits fram f¨or att f¨orutse temperaturer i omloppsbana. Den termiska modellen skapad av tidigare studenter m˚aste korreleras genom att utf¨ora ett termiskt vakuumkammartest. Det testet kommer att genomf¨oras i en vakuumkammare p˚a KTH Albanova under perioden November 2019 till Februari 2020. Den h¨ar avhan- dlingen beskriver processen att ta fram ett termisk vakuumkammartest f¨or att korrelera modellen.
F¨or att genomf¨ora testet m˚aste vissa komponenter bytas ut till ers¨attningskomponenter. Batteriet och solpanelerna ¨ar dyra och kritiska system f¨or uppdraget och har valts att ers¨attas med ers¨attningskom- ponenter. Dessa ers¨attningskomponenter anv¨ands i den h¨ar rapporten f¨or att bebh˚alla korrekt termisk massa och approximera de konduktiva egenskaperna. Solpanelsers¨attningspl˚atarna kan borras i f¨or att l¨attare f˚a in sladdar f¨or termiska sensorer och v¨armare i satelliten. V¨armare kommer att monteras p˚a satelliten, b˚ade f¨or att skapa en temperaturgradient ¨over satelliten f¨or att kunna m¨ata dessa, samt som s¨akerhetsv¨armare f¨or att undvika att temperaturen blir f¨or l˚ag och skada komponenter. V¨armarna kommer att vara monterade p˚a batteriers¨attnings komponenten, solpanelspl˚atarna och p˚a den undre t¨ackplattan.
De termiska sensorerna valda f¨or detta test ¨ar av Typ T klass 1 med en noggrannhet p˚a ±0.5◦C. Vaku- umkammaren har 50 installerade termiska sensorer och 24 av dem ¨ar planerade att anv¨andas i testet f¨or att m¨ata temperaturskillnader mellan komponenter f¨or att kunna korrelera kopplingen mellan dem i den termiska modellen.
I den h¨ar avhandlingen kommer den termiska modellen simuleras i den nya vakuumkammarmilj¨on f¨or att unders¨oka vilka temperaturer som kan f¨orv¨antas och hitta komponenter av intresse f¨or att korrelera och placera termiska sensorer p˚a. Testet kommer att starta p˚a en rumstemperatur p˚a 20 ◦C, sedan ¨andra temperaturen p˚a kammaren till 0◦C, med en hastighet p˚a 1.5◦C/min. Kammaren kommer att stanna p˚a 0◦C tills att satelliten har stabiliserats med det valda stabilitets kriteriet att temperaturen inte f˚ar ¨andras med mer ¨an 0.5◦C ¨over tv˚a timmar. Denna temperatur ¨ar vald f¨or att ta m¨atningar f¨or det varma fallet.
N¨ar m¨atningarna ¨ar gjorde kommer kammaren att g˚a ned till -50◦C och stabiliseras p˚a den temperaturen f¨or att ta m¨atningar f¨or det kalla fallet. Temperaturgradienterna m¨atta i b˚ada fallen kommer att anv¨andas f¨or att korrelera den termiska modellen med den fysiska modellen.
Acknowledgements
I would like to take this opportunity to thank my supervisor Mr. Andreas Berggren for his continuous support and feedback throughout this thesis. His guidance has been essential for helping to steer the thesis in the right direction and providing his technical expertise regarding thermal control.
I would also like to thank MIST Project leader Dr. Sven Grahn for the opportunity to do my master thesis for MIST. It has been exciting and an invaluable experience working on a real satellite in its developing phase.
Furthermore I would like to extend my gratitude to Ms. Linda Eliasson for all the help regarding the vacuum chamber. She has helped me gather all the technical specifications for the vacuum chamber and thermocouples and helped me understand how the vacuum chamber works.
I would like to thank Mr. David Mainwaring och Ms. Borbala Bernus for their support regarding me- chanical solutions. I would also like to thank the entire MIST team for an exciting year where I have been proud to be a cog in the machine where great strides have been taken for getting the MIST satellite ready for launch.
Finally i would like to thank my friends and family for their emotional support and encouragement through- out my entire time at KTH. I would probably not be where I am today without you.
Contents
Abstract . . . . i
Acknowledgements . . . . ii
List of figures . . . . vi
List of tables . . . . vii
Abbreviations . . . . viii
Physical Constants . . . . ix
Symbols . . . . x
1 Introduction 1 2 Theory 2 2.1 Radiation . . . . 2
2.2 Conduction . . . . 3
3 MIST Experiments 4 3.1 NanoProp . . . . 6
3.2 SEUD . . . . 7
3.3 CUBES . . . . 7
3.4 MOREBAC . . . . 7
3.5 Pietzo LEGS . . . . 7
3.6 RATEX-J . . . . 7
3.7 SiC. . . . 8
3.8 Camera . . . . 8
4 Thermal Model Pre-Study 9 4.1 Nanoprop . . . . 10
4.2 Solar panels . . . . 14
5 Thermal Balance Test 17 5.1 Vacuum chamber . . . . 18
5.2 Thermal model . . . . 19
5.2.1 Battery replacement . . . . 20
5.2.2 Solarpanel replacement . . . . 21
5.3 Thermal Hardware . . . . 22
5.3.1 Thermocouple . . . . 22
5.3.2 Heaters . . . . 24
5.3.3 Power supply . . . . 29
5.3.4 Electrical feedthrough . . . . 30
5.3.5 Coaxial feedthrough . . . . 30
5.4 Stability criteria . . . . 30
5.5 Test case . . . . 31
5.5.1 Vacuum chamber test case. . . . 31
5.5.2 TBT simulations . . . . 32
6 Conclusions 41 6.1 Conclusions . . . . 41
6.2 Future work . . . . 41
6.3 Self reflection . . . . 42
Bibliography 44
Appendix A: Test Specification and Procedure - Thermal Balance Test 45
List of Figures
3.1 MIST configuration during the fall of 2018. . . . 4
4.1 Temperature distribution for BOL conditions from the start of the fall 2018 semester . . . . 9
4.2 FSM tape [21]. . . . 10
4.3 Node definition for NanoProp . . . . 10
4.4 Temperature distribution of Nanoprop tank over one orbit without any treatment. . . . 11
4.5 Temperature distribution of Nanoprop tank covered in FSM over one orbit . . . . 11
4.6 Temperature distribution of Nanoprop tank covered with MLI over one orbit . . . . 12
4.7 Temperature distribution of Nanoprop tank, with the tank covered with Cho-Foil and the protroding part covered with MLI over one orbit . . . . 13
4.8 Temperature distribution of Nanoprop tank, with a thin washer added between the struc- tural rib and NanoProp mounting support over one orbit . . . . 14
4.9 Power generated from power simulation [9]. . . . 15
4.10 temperature of solarpanels BOL. . . . 16
4.11 Power generated from Thermal model BOL . . . . 16
5.1 Vacuum Chamber at KTH Albanova . . . . 17
5.2 Vacuum chamber front view . . . . 18
5.3 Vacuum chamber top view. . . . 18
5.4 Rail in the Vacuum Chamber . . . . 19
5.5 Vacuum chamber model in Systema . . . . 20
5.6 Battery replacement Dummyboard . . . . 21
5.7 Dummyplate -X. . . . 22
5.8 Dummyplate -Y. . . . 22
5.9 Dummyplate +Y . . . . 22
5.10 Dummyplate +X . . . . 22
5.11 Thermocouple location in the top stack . . . . 23
5.12 Thermocouple locations in the bottom stack. . . . 23
5.13 Heater location on the battery replacement board. . . . 25
5.14 Circuit diagram for the battery replacement plate heater . . . . 25
5.15 Heater location on the side cover plates . . . . 26
5.16 Circuit diagram for the side cover plate heaters . . . . 26
5.17 Heater location on the side cover plates . . . . 27
5.18 Circuit diagram for the side cover plate heaters . . . . 27
5.19 Variable Power supply . . . . 29
5.20 Electrical feedthrough in the vacuum chamber. . . . 30
5.21 Excel graph representation of the TBT test . . . . 31
5.22 Temperature profile for the top stack during TBT . . . . 33
5.23 Temperature difference over 2 hours for top stack . . . . 34
5.24 Expected temperature gradients across interfaces in top stack at T=7 Hr . . . . 35
5.25 Expected temperature gradients across interfaces in top stack at T=21 Hr . . . . 35
5.26 Temperature profile for the middle stack during TBT . . . . 36
5.27 Systema Thermal model of the antenna board . . . . 37
5.28 CAD model of Antenna board and HDRM. . . . 37
5.29 Temperature difference over 2 hours for the middle stack . . . . 37
5.30 Temperature profile for the bottom stack during TBT . . . . 38
5.31 Temperature difference over 2 hours for the bottom stack . . . . 39
5.32 Expected temperature gradients across interfaces in bottom stack at T=7 Hr . . . . 40
5.33 Expected temperature gradients across interfaces in bottom stack at T=21 Hr. . . . 40
List of Tables
3.1 Power schedule [19] during orbit . . . . 5
3.2 Experminet temperature limit [8] . . . . 6
4.1 Assumptions for solar panels . . . . 15
5.1 Thermocouples for the top stack . . . . 23
5.2 Thermocouples for the bottom stack . . . . 23
5.3 Heater data [1] . . . . 24
5.4 Battery replacement heater . . . . 25
5.5 Side Cover plate heaters . . . . 26
5.6 Bottom Cover plate heaters . . . . 27
5.7 Temperature limits of experiments and sub systems. . . . 28
5.8 Heater power depending on applied voltage . . . . 29
5.9 Power supply data [6] . . . . 29
5.10 TBT thermal simulation settings . . . . 32
Abbreviations
AC Alternating Current
AWG American Wire Gauge
BOL Beginning Of Life
CUBES CUbesat x-ray Background Explorer using Scintillators
DC Direct Current
EOL End Of Life
ESA European Space Agency
FSM First Surface Mirror
HDR High Fynamic Range
HDRM Hold Down and Release Mechanism JUICE JUpiter ICy moon Explorer
IGIS ISIS Generic Interface System ISIS Innovative Solutions In Space KTH Kungliga Tekniska H¨ogskolan MIST MIniature Student saTellite MLI Multi Layer Insulation
MOREBAC Microfluidic Orbital REsuscitation of BACteria
OBC On Board Computer
PCB Printed Circuit Board
RATEX-J RAdiation Test EXperiment for JUICE SEUD Single Event Upset Detector
SiC Silicon Carbide
TBT Thermal Balance Test
TC ThermoCouple
Physical Constants ix
Physical Constants
Stefan-Boltzmann constant σ = 5.67037321 · 10−8W m−2K−4
Symbols
A Area m2
Emissivity
Q Heat W
T Temperature ◦C
α Absorbtivity
L Length m
k Thermal conductivity W/m
G Thermal Conductance W/K
hc Thermal Contact conductivity W/m2K
C Thermal capacity J/K
P Power W
U Voltage V
I Current A
R Resistance Ω
Chapter 1
Introduction
The thermal model for MIST has been developed by several previous students and is a theoretical com- parison of the physical model. All material data and conductive couplings across interfaces are based on best estimates and are prone to be estimated as the worst case scenario. Multiple small errors in the estimations can add up to big temperature differences across the spacecraft.
To make sure the model is as close to the physical model as possible a thermal balance test has to be performed and compare the thermal model with the results of the test. When the thermal model has been updated with the data from the test, the model should represent a better estimation and give more accurate prediction of the temperature variations during flight.
The objective of this thesis is to prepare the existing thermal model of MIST for upcoming thermal vacuum chamber tests. The current thermal model of MIST represents the flight model and is applied in orbit around earth to predict the flight conditions and simulate temperature variations for MIST. When MIST is to be thermally tested, the current model has to be modified both in regards to the environment and the configuration of the spacecraft. As some components of the spacecraft is especially sensitive and is critical for the success of the mission, some components will have to be replaced with relevant ground support equipment to be able to perform the test. During the test the temperatures has to be recorded in strategically selected locations to be able to correlate the thermal model. All gathered information is to be presented in a test specification and test procedure for model correlation in accordance with ESA standards.
Theory
Thermal transfer is made up by three modes, radiation, conduction and convection [7]. Convection is heat transferred from a solid to a fluid, and as the only fluid present in orbit and in the test is the propulsion fuel, convection is not going to be considered in the thermal analysis.
2.1 Radiation
Thermal transfer through radiation is governed by energy from electromagnetic waves transferred from an object to its surroundings. The thermal energy emitted from an object is described in equation2.1.
Qenb= ArσT4 (2.1)
Where σ is the Stefan-Boltzmann constant, T is the temperature in Kelvin, A is the radiating area and
is emissivity. The emissivity is the portion of thermal energy transferred compared to if the object would have been considered as a blackbody which is by definition the most efficient emitter. For an object to be in thermal balance it has to absorb as much energy as it emits. The amount of absorbed energy is dependent on the incoming energy over a surface and the absorbtivity α, where α is the portion of absorbed energy.
qa= αQi
Ai (2.2)
During the design phase of a spacecraft the absorbtivity and emissivity can be chosen to alter the thermal behavior. For a component that should act as a radiator one has to chose a material or coating that has a higher emissivity than absorbtivity, while a component that has a higher temperature limit may need to have higher absorbtivity than emissivity to be thermally stable at a higher temperature.
CHAPTER 2. THEORY 3
The thermo-optical properties of materials will vary during the life time of a satellite due to degradation of the materials. The properties of the materials before degradation is often called Beginning Of Life (BOL) condition, where no major changes of the material has occurred. During the life cycle of the satellite, it will be exposed to environment that will degrade the materials. Quoting A. de Rooij [3];”The degrading en- vironment for spacecraft materials includes atomic oxygen, Ultraviolet (UV) radiation, ionizing radiation, ultrahigh vacuum (UHV), charged particles, thermal cycles, electromagnetic radiation, micrometeoroids, and man-made debris, micrometeoroids and orbital debris.” At End Of Life (EOL) of a satellite, the thermo-optical properties of the materials will be significantly altered which need to be taken into account when designing the thermal control system. The EOL conditions are usually seen as an increase of the absorbtivity, causing the temperature of the satellite to be higher than during BOL.
2.2 Conduction
Conduction is transfer of thermal energy through materials. The amount of thermal energy transferred through conduction is determined by the difference in temperature across the material and the thermal conductivity k of the material.
Qk = kAk
L (Th− Tl) (2.3)
Where Ak is the area of the material conducted through and L is the distance between the two points Th and Tl. While the conductivity k is a material property, kAk/L is the conductive capability of an object and is called the conductance G of an object.
G = kAk
L (2.4)
Thermal energy is not only conducted through a solid material but is also conducted between two solid objects if in contact. Instead of the material property conductivity k, the parameter contact conductance hc is used.
Gc= hcA (2.5)
Where A is the area of the interfacing surface. The property hc is dependent on the contact pressure between the two surfaces, surface roughness and deformations of the interfacing areas. The combined total conductance between two objects and its interface is calculated as below.
Gtot= 1
1 G1+G1
c+G1
2
(2.6) Where G1and G2are the conductance of different components, and Gcis the contact conductance between the two components.
MIST Experiments
MIST was originally designed to carry seven experiments. During the fall of 2018 MIST was redesigned, reducing the number of experiments to five. The two experiments removed was MOREBAC and RATEX- J. It was at the same time decided to reconfigure MIST, moving experiments between the top and bottom stack and duplicate the CUBES experiment to better make use of the free space. For this thesis the original configuration will be considered as the new configuration was still not defined at the time of this thesis. When the new model is completed, the simulations in this thesis have to be repeated and updated to work with the new model.
Figure 3.1: MIST configuration during the fall of 2018
CHAPTER 3. MIST EXPERIMENTS 5
MIST has a limited amount of power available per orbit and data storage capability. For this reason the experiments will only be run a limited amount of time and divided to run separately. The different groups of experiments to be run are divided into 6 different cases where each case is studied individually to see how the thermal profile of the satellite is during each different case.
Experiment Case 1 Case 2 Case 3 Case 4 Case 5 Case 6
Nanoprop X
MOREBAC X
RATEX-J X
CUBES X
Camera X
Pietzo LEGS X
Sic X
SEUD X X X X X X
Sub systems X X X X X X
Table 3.1: Power schedule [19] during orbit
Table 3.1 shows the different thermal cases designated, where X marks the defined case for each exper- iment. The experiments and sub systems are run with their defined internal thermal dissipation. The power schedule also defines during what periods of the orbit the experiment is supposed to be run.
The different experiments and sub systems have their own temperature limitations that have to be taken into account for the design of the thermal control system. This can put a limit on the operation time and run period of the experiments during an orbit to avoid violating the thermal operational and non operation temperature limits.
Experiment Operational (◦C) Non-operational (◦C)
CUBES -20 +30 -20 +80
Pietzo LEGS +10 +40 -30 +70
MOREBAC +20 +30 +4 +30
Nanoprop 0 +40 -10 +50
RATEX-J -30 +60 -40 +70
SEUD 0 +85 -65 +150
Sic -40 +105 - -
Camera -30 70 - -
Sub systems Operational (◦C) Non-operational (◦C)
ISIS IGIS -30 70 - -
ISIS TRXVU -40 60 - -
OBC -20 60 - -
NanoPower -40 85 - -
Battery -5 45 - -
Torquer rod -40 70 - -
Antenna Board -30 70 - -
Table 3.2: Experminet temperature limit [8]
3.1 NanoProp
NanoProp is a propulsion system for cubesats designed by NanoSpace [13]. The purpose of this experiment is to get heritage of the flight module and demonstrate that precision maneuvers can be performed by cubesats with the system. NanoProp is positioned at the top of MIST, +Z direction, and has its tank protruding the cover plate. The experiment includes a heater on the tank and the four thrusters, and the requirement to run the experiment is that the temperature is between 0 and +40◦C. As can be seen in Table 3.2,The operational temperature limit is more strict than the non operational temperature limit.
During eclipse the top part of the satellite gets cold as it is exposed to deep space, this has led to a planned operation schedule where the experiment is run during sunlight only to limit the heating power needed to reach required temperature. The heater on the tank activate 30 minutes before the experiment is planned to operate and the heaters on the thrusters activates two minutes before operation. As the heaters and thrusters require a lot of power, NanoProp will be planned for a two day period where only SEUD is run at the same time.
CHAPTER 3. MIST EXPERIMENTS 7
3.2 SEUD
SEUD is designed by KTH department of Electronics and Electronic Systems based in Kista and de- signed to measure Single Event Upsets caused by radiation and its self healing capabilities in the space environment[15]. SEUD is planned to operate at all time.
3.3 CUBES
CUBES is an experiment created by the Particle and Astroparticle department at KTH[10]. It is designed to study radiation in low earth orbit using three detectors made up of photomultiplyers and scintillators made out of three different materials. The scintillators will be made out of GAGG (Gd3Al2Ga3O12), BGO (Bi4Ge3O14) and plastic. During the fall of 2018, it was decided to add another copy of the CUBES experiment to gather twice the amount of data. This change is made for the new model and will not be considered in this thesis. The experiment requires to be exposed to deep space at all time and not be illuminated by the sun. This requirement means that it has to be placed facing the side that is always going to be pointing away from the sun, -X direction.
3.4 MOREBAC
MOREBAC is a biological experiment where dried microorganisms is resuscitated to measure growth in orbit[12]. The experiment was removed from MIST during the fall of 2018 and will not be included in the new model, but it will still be considered in this thesis.
3.5 Pietzo LEGS
Piezo LEGS is a motor based on piezoelectric elements designed to move a drive rod using pietzoelecric legs[11]. This principle is today widely used in optical equipment as they can move very small precise distances. The purpose of this experiment is to investigate how this kind of motor works in vacuum and high radiation environments for future space missions. Piezo LEGS is planned to run at the same time as SiC as the voltage step up module for Pietzo LEGS is located in the SiC experiment.
3.6 RATEX-J
RATEX-J is an experiment proposed by the Swedish Institute of Space Physics in Kiruna[14]. It is a radiation test experiment for the ESA JUICE mission and is composed of three different particle detectors.
A solid state detector, a multi-channel plate and a ceramic channel electron multiplier. The experiment requires to be exposed to deep space at all time and not be illuminated by the sun. This requirement means that it has to be placed facing the side that is always going to be pointing away from the sun, -X
direction. The experiment was removed from MIST during the fall of 2018 and will not be included in the new model, but it will still be considered in this thesis.
3.7 SiC
SiC is an experiment containing three transistors[16]. The transistors are made out of silicone, graphene and Silicon Carbide. The purpose of the experiment is to flight test the Silicon Carbide transistor and take long term measurments helpfull for future research. SiC can be run seperatly, but has to be run at the same time as Pietzo LEGS as the voltage step up module required to run Pietzo LEGS is located on the SiC board.
3.8 Camera
The camera is located at the bottom of the satellite, -Z direction, which is pointing in Nadir direction at all time[17]. The camera will take high quality pictures of earth to be sent down to earth. The images will be captured using HDR technology and will be compressed and reconstructed using a new learning based compression method. The camera will be connected with SEUD to utilize it to process the images.
Chapter 4
Thermal Model Pre-Study
The thermal model for MIST has been developed by previous student teams since 2015 and implemented in the software Systema Thermica. The model has to be continuously updated to reflect any changes made to the physical model and to implement updated solutions for thermal control. All experiments and subsystems have given temperature limitations that have to be taken into account when designing the thermal control system. As MIST has limited amount of power, there is low excess power available for active thermal control. Passive control has to be designed to work around the given limitations.
Figure 4.1: Temperature distribution for BOL conditions from the start of the fall 2018 semester
Figure4.1shows the temperature fluctuations for each Thermal case compared to the operational and non- operational limits from the start of the semester [23]. As can be seen in Figure 4.1, several experiments are outside temperature limits. Nanoprop is one of the experiments which is most problematic and needs further investigation.
4.1 Nanoprop
Nanoprop is located at the top of MIST, and has its tank protruding the structure. This causes big temperature fluctuations as the tank is exposed to direct sunlight during the phase where the satellite is illuminated, and exposed to deep space during eclipse. As the tank is made of Titanium, which has extremely low thermal conductivity, big temperature gradients is seen across the tank where some parts is exposed to space and some parts are inside the satellite. To modify the temperature profile of the tank, it can be coated with different materials to change the absorbtivity, emissivity and conductivity.
The different coatings that where investigated were First Surface Mirror (FSM), Multi-Layer Insulation (MLI) and CHO-Foil.
• FSM tape is a coating where the reflecting surface is placed above the substrate. As shown in Figure 4.2, the reflec- tive surface, Aluminum, is placed above the Polymide sub- strate. FSM has low emmisivity to absorbtivity ratio and was considered to increase the overall temperature of the tank.
Figure 4.2: FSM tape [21]
• MLI is built up by multiple thin layers of Kapton or Mylar, separated by a non-metallic mesh to reduce the thermal conduction through the blanket. MLI acts as an insulator to reduce the heat exchange with the environment both in sunlight and eclipse to reduce temperature fluctuations.
• Cho-Foil is an aluminum tape with high thermal conductivity [2]. It is considered to be used to decrease temperature gradients across the tank as the tank is made out of Titanium which is a low conducting material.
The different implementations were made to investigate the temperature profile of the NanoProp tank where the graphs will show one curve per node. The node definitions used in the simulation software Systema are illustrated in Figure4.3. The results are shown from simu- lations using Case 1 where Nanoprop and SEUD are used, during End Of Life conditions. The time period shown are from 30 minutes before NanoProp is supposed to be used until the satel- lite starts to enter eclipse.
Figure 4.3: Node definition for NanoProp
CHAPTER 4. THERMAL MODEL PRE-STUDY 11
Figure 4.4: Temperature distribution of Nanoprop tank over one orbit without any treatment
As can be seen in Table3.2the temperature limit for Nanoprop is −10◦C while not operational and 0 ◦C while operational. With current configuration the lower temperature is well below the temperature limit and the goal is to increase the lower temperature without increasing the top temperature.
To increase the temperature, FSM tape was investigated as coating for the part of the tank that is protruding the structure. The FSM tape changes the thermo-optical properties of the surface to increase the absorbtivity and absorb more of the incoming solar flux.
Figure 4.5: Temperature distribution of Nanoprop tank covered in FSM over one orbit
As can be seen in Figure 4.5, the temperature of the protruding part of the tank is increased, but the bottom part of the tank is not affected as much. This creates a big temperature gradient across the tank which is a behavior that is not favorable. As well as creating a temperature gradient across the tank, the overall temperature is still below temperature limits and needs further investigation.
In attempts to reduce the gradient across the tank and reduce the fluctuations, a 10 layer small blanket MLI to cover the protruding part of the tank was investigated. The use of MLI insulates the tank from the outside environment which removes most of the radiative heat loss when in eclipse and blocks solar flux when illuminated.
Figure 4.6: Temperature distribution of Nanoprop tank covered with MLI over one orbit
The use of MLI increased the lower temperature, but created even larger temperature gradients across the tank. The conductive properties of titanium is very low, which means that the heat flux travels slowly across the tank. The biggest part of the heat flux is now through the conductive coupling between the tank and the mounting board, which is applied near the bottom of the tank. To reduce the gradient, the conductive properties across the tank have to be increased. This can be achieved by adding layers of Cho-Foil[2] aluminum tape on the tank. Aluminum has a thermal conductivity of 230 which is much higher than Titanium, that has a thermal conductivity of 7. By adding two layers of Cho-Foil the heat flux across the tank is doubled.
CHAPTER 4. THERMAL MODEL PRE-STUDY 13
Figure 4.7: Temperature distribution of Nanoprop tank, with the tank covered with Cho-Foil and the protroding part covered with MLI over one orbit
As can be seen from Figure 4.7, The temperature gradient is significantly decreased by increasing the thermal conductivity. However, the temperature fluctuation is still to large. To lower the temperature fluctuation across the tank, the incoming heatflux from the structure has to be decreased. A thin 0.5mm Titanium washer was added in the four interfaces between the structural ribs and Nanoprop mounting support to isolate Nanoprop conductively from the structure. By adding a washer between the rib and nanoprop, a new smaller interface is created between the two parts which limits the flow of energy. As well as limiting the interface area, the Titanium washer has very low thermal conductivity to further lower the conductive heatflux.
Figure 4.8: Temperature distribution of Nanoprop tank, with a thin washer added between the structural rib and NanoProp mounting support over one orbit
The biggest influence on tank temperature turns out to be the conductive coupling to the structure.
Covering the protruding part of the tank with different coatings has a minimal influence in comparison.
The interface can also easily be modified to alter the thermal conductivity by changing either the thickness or changing the material of the washer. The model that is used throughout this thesis is with the updated interface and the tank covered in CHO-Foil and MLI.
4.2 Solar panels
Power generation of MIST is a subject that has been investigated by the power team, where simulations has been made to get an estimation of the total power generation over one orbit. The power team need to get an estimation of how much power is available, but it is also important for the thermal model to get an accurate estimation. The amount of energy generated by the solar panels is transformed to electrical energy. In order to reflect this in the thermal model the amount of generated electrical energy need to be subtracted from solar panels as this absorbed heat flux will not contribute as a heat source on the solar panels.
CHAPTER 4. THERMAL MODEL PRE-STUDY 15
Figure 4.9: Power generated from power simulation [9]
Figure 4.9 shows the power teams power simulation for BOL conditions, where the The MPPT 1 curve is the amount generated from the solarpanel in +X direction, MPPT 2 in -Y/+Y direction and MPPT 3 from the deployable solar panels. The ideal curve represents the power generated if each cell is individually controlled at its maximum power condition, and no losses are assumed. The total power generated is 12.6 [W·hr] for the ideal case. The power generated in the thermal model is assumed ideal and with an efficiency that is temperature dependent.
BOL EOL
Solarflux [W/m2] 1322 1414
Efficiency at 28◦C 0.3048 0.2561 Efficiency change per degree -0.00070 -0.00064
Table 4.1: Assumptions for solar panels
Figure 4.10: temperature of solarpanels BOL
The temperature of the all the solar panels except the deployable are always below 28◦C, resulting in a slightly higher efficiency than assumed in the powersimulation.
Figure 4.11: Power generated from Thermal model BOL
The total power generated from the thermal model is 13.9 [W·hr], which is slightly higher than the calculated power generation from the power simulation. The power simulation is a more detailed estimation of the solar cells than the thermal model and a small discrepancy is expected.
Chapter 5
Thermal Balance Test
To make sure the thermal model is as close to the physical model as possible a thermal balance test has to be performed and compare the thermal model with the results of the test. When the thermal model has been updated with the data from the test, it should represent a better estimation and give more a accurate prediction of the temperature variations during flight.
The test will be performed in the vacuum chamber that was installed at KTH Albanova in the spring of 2019. The test is planned to take place during the period November 2019 and February 2020 to be able to meet the MIST projects launch date in December 2020.
Figure 5.1: Vacuum Chamber at KTH Albanova
5.1 Vacuum chamber
The vacuum chamber is 2.3m long on the inside with a diameter of 1.4m and has a thermal plate that can be controlled independently from the shroud. The inner lining of the shroud is made of a high emissivity material to increase the radiative transfer of energy with the satellite. The vacuum chamber is controlled by a software called Winkratos and can be set within the temperature range -150 ◦C to +150 ◦C and can reach a pressure of 10−7 mbar. The maximum temperature ramp possible in the vacuum chamber is 3◦C per minute. By using the maximum ramp speed the chamber is prone to overshoot the assigned temperature, and it is therefore recommended to use 1.5◦C per minute as maximum ramp speed to avoid large overshoot.
Figure 5.2: Vacuum chamber front view Figure 5.3: Vacuum chamber top view
The use of the thermal plate in the vacuum chamber induces a conductive coupling between the thermal plate and the satellite which is not desired. The test is performed to correlate the thermal model with the physical model and the conductive coupling will only induce an uncertainty that has to be taken into account. A preferred solution is to let the satellite hang from the internal rails in the chamber. The conductive coupling between the satellite and the wires is a lot smaller than if the satellite would be placed on the thermal plate which simplifies the correlation. As the thermal plate is not desired to be used, it can be removed. While hanging, the satellite can easily be placed in the middle of the chamber to get uniform radiative heat transfer.
CHAPTER 5. THERMAL BALANCE TEST 19
Figure 5.4: Rail in the Vacuum Chamber
5.2 Thermal model
When simulating the TBT the simulation environment has to be changed. During flight simulations, MIST is in orbit around earth at a certain altitude and inclination. For the test simulation, a new environment is needed according to the dimensions in Figure 5.2 and Figure 5.3. The Thermal model of the vacuum chamber has been created as a simple model with the internal shroud as three nodes;
the inner wall, the door and the inner cylinder. These nodes are controlled together to get a uniform temperature distribution. According to the specifications for the vacuum chamber [4] the shroud has a temperature uniformity throughout the shroud of ±5◦C at steady state and a uniformity of ±15◦C at maximum temperature ramp. However the uniformity is not possible to predict and assumed to have a mean value at the set temperature.
Figure 5.5: Vacuum chamber model in Systema
When performing the test, the satellite should be configured as much as possible as it would be while launched. However, some equipment are extra sensitive and are not preferred to have installed during the thermal balance test. Two of the parts that have been discussed to remove for the test are the battery and the solar panels [18].
5.2.1 Battery replacement
The battery is very sensitive to low temperatures and is an extremely crucial subsystem. If the battery would be damaged in some way the whole mission could be compromised, therefore the battery will be replaced with a pc104 board with a mounted aluminum plate to simulate the thermal mass of the battery and act as a heat sink for attached heater. On this Aluminum plate, both thermocouples and heaters can be placed to monitor temperature and induce heating energy to the middle of the satellite. The Aluminum plate added to the PCB is 6x6x1 cm which corresponds to a thermal capacity of 90 J/◦C, which is a close approximation to the battery thermal capacity.
By removing the battery, two problems arise. The satellite has to be power by an external source, and the stack connectors attached on the battery PCB has to be replaced. A new PCB has to be designed with connectors attached to connect the part of the satellite below the battery with the part above. Designing this PCB has to be done in order to be able to remove the battery, but is out of the scope of this thesis.
CHAPTER 5. THERMAL BALANCE TEST 21
Figure 5.6: Battery replacement Dummyboard
5.2.2 Solarpanel replacement
The solar panels are not as sensitive to thermal conditions as the battery, however they are sensitive to mechanical strains. The deployable solar panels is not constructed to be deployed under earth gravity, and if deployed the HDRM would have to be reassembled. Another reason to remove the solar panels is to be able to get cables for both thermocouples and heaters inside the satellite. If the solarpanels were to be removed, they have to be replaced with something else as removing the solar panels would expose all components. It is preferred to replace the solar panels with a coverplate with the same material and dimensions as the solar panel substrate. According to the solar panel drawing [22] the solarpanels are mounted on a 1.5mm Aluminum plate. The satellite needs some way of hanging in the vacuum chamber, this can be solved by adding mounting hooks on the dummyplates to be attached to cables attaching the satellite to the rail in the top of the vacuum chamber.
Using these Aluminum plates, one can modify them by drilling holes where cables for TCs and heaters are required to limit the amount of cables running inside the satellite. As can be seen in Figure 5.7and 5.8 there is one hole in the cover plate on –X side, where cables for sensor 1, 7 and 8 shall go through.
There are 3 holes in the cover plate on the –Y side. The top hole is made for sensor 2, 3, 4, 5 and 6. The middle hole is made for sensor 9, 10, 11 and 12. The bottom hole is made for the cable for the battery replacement heater. All holes are made as 10mm holes.
Figure 5.7: Dummyplate - X
Figure 5.8: Dummyplate - Y
Figure 5.9: Dummyplate +Y
Figure 5.10: Dummyplate +X
5.3 Thermal Hardware
5.3.1 Thermocouple
The thermocouples to be used in the test are of Type T class 1 according to the IEC60584[20] standard.
The tolerances for these TCs are ±0.5◦C or ±0.004 · |T |◦C depending on which is largest. The tolerance to be used is ±0.5◦C, as the temperature is never going to be lower than -50 ◦C. These TCs are already installed in the vacuum chamber and can be read from the vacuum chamber control software Winkratos.
There are 50 Thermocouples installed in the chamber where 48 are available for desired measurements.
The choice of TC placement location is baced on interfaces between components that are of interest to correlate. The thermal transfer between two components are crucial to get correct to be able to predict what temperatures the satellite actually will experience while in orbit. The TCs have to be placed as close to the interface as possible to be able to read the gradient across the interface. If the TC is attached to far away from the interface, the temperature difference between the two sensors will not only be be- cause of the interface as the thermal energy travels through the medium. This is especially important
CHAPTER 5. THERMAL BALANCE TEST 23
for components whith a high temperature gradient across it and material with low thermal conductivity.
Interfaces of interest are where the predicted gradient is large, for these interfaces even a small deviation in the assumed thermal conduction between the components could have a large influence on the overall temperature profile of the satellite. The chosen components to place Thermocouples on is shown in Table 5.1and5.2and their physical location is illustrated in Figure5.11and5.12.
Sensor # Component Interface
1 Top coverplate 1
2 Side coverplate 4
3 Structural rib 1, 2, 3, 4 4 Mounting support 2, 5, 6
5 Rail 3
6 Thruster PCB 5
7 Mounting board 6, 7, 8
8 Tank 7
9 Hexnut spacer 8, 9
10 NanoProp main PCB 9, 10
11 Spacer 10, 11
12 SEUD board 11
Table 5.1: Thermocouples for the top stack Figure 5.11: Thermocouple location in the top stack
Sensor # Component Interface
13 Structural rib 12, 13
14 RATEX-J top panel 12
15 Rail 13
16 RATEX-J bottom panel 14
17 Spacer 14, 15
18 SiC 15, 16
19 Spacer 16, 17
20 CUBES/Legs mounting board 17, 18
21 Hexnut spacer 18, 19
22 Structural rib 19, 20, 21
23 Bottom coverplate 20
24 Side coverplate 21
Table 5.2: Thermocouples for the bottom stack
Figure 5.12: Thermocouple locations in the bot- tom stack
The TCs are made out of a twisted pair of conductors. Each conductor is made out of seven strands AWG 32 with a diameter of 0.2mm and total area of 0.44mm2. A question that has been discussed is if the TCs will influence the temperature profile of the components attached to. The amount of influence made by the thermocouples is determined based on the temperature gradient of the cable according to equation 2.3. If the cable has approximately the same temperature as the component it will have minimal impact but if it would assume the same temperature as the shroud it could leak up to 50mW of thermal energy.
5.3.2 Heaters
During the test, it is desired to create a temperature gradient across the satellite to be able to meassure the temperature gradient between two components. As the temperature profile created by the shroud would be a uniform temperature across the entire satellite, it is required to create the gradients using either internal dissipation from subsystems and experiments, or by adding heaters. The subsystems will be turned on during the test to be able to relay temperature readings from internal sensors, but it is still required more power to create larger gradients. The experiments have too fluctuating dissipation profiles to be able to get a stable temperature profile and will therefore be turned off. Instead, heaters will be added across the satellite to induce the necessary temperature gradients.
Heater X dim [mm] Y dim [mm] resistance [Ω] Max power at 28 V [W]
Minco HK6907 2.54 5.08 47.78 16.37
Minco HK6908 2.54 7.62 32.09 24.35
Table 5.3: Heater data [1]
Internal heaters
The satellite does not have many internal heaters, as all power created by the solar panels is needed for experiments and subsystems there are only a few heaters in the original design.
• NanoProp: Tank heater 1.4 W max power
• NanoProp: Thruster heaters 0.25 W x 4 max power
• Battery: 3.5 W max power
The heater on the NanoProp tank will be used as the conductive coupling between the tank and the mounting board are of interest to correlate, but the heaters on the thrusters is not beneficial to use as it would diminish the gradient through the experiment. As discussed in section: 5.2.1The battery will be replaced with a PCB dummy with an Aluminum plate. By replacing the battery, it is also beneficial to add a new heater on the PCB dummy to be able to add thermal energy inside the satellite. This heater will be used to heat the center of the satellite to create a gradient through the satellite where the temperature will decrease the further away from the center a component is. The heater chosen to add to the PCB
CHAPTER 5. THERMAL BALANCE TEST 25
dummy is Minco HK6907 as its dimensions are small enough to fit on the Aluminum plate and the power available to add is enough for its purpose. At 28 V this heater has a power density of 1.53 [W/cm2] which is above the recommended power density of 0.54 [W/cm2]. The voltage is there for limited to 16 volts.
Series number Heater Power per heater at 16 V [W] Power density at 16V [W/cm2]
Series 1 Minco HK6907 5.3 0.5
Table 5.4: Battery replacement heater
Figure 5.13: Heater location on the battery replace- ment board
Figure 5.14: Circuit diagram for the battery re- placement plate heater
External heaters
As it was decided to remove the solarpanels and replace them with Aluminum plates, it is now possible to add external heaters to be able to heat up the structure at specific locations. To create the intended gradient across the satellite it is chosen to add heaters on the coverplates on the middle of the mid stack.
One heater is placed on each side of the satellite to get a uniform distribution around the satellite. The heaters are connected as two parallel connected series of 2 heaters to get enough power available. Another reason to have them connected in parallel is to have the redundancy of two separately connected series if one of them would fail.
Series number Heater Power per heater at 28 V [W] power density at 28V [W/cm2] Direction
Series 1 Minco HK6908 6.06 0.38 -X
Series 1 Minco HK6908 6.06 0.38 +X
Series 2 Minco HK6908 6.06 0.38 -Y
Series 2 Minco HK6908 6.06 0.38 +Y
Table 5.5: Side Cover plate heaters
Figure 5.15: Heater location on
the side cover plates Figure 5.16: Circuit diagram for the side cover plate heaters
Safety heaters
To prevent damaging components due to the components being too cold, safety heaters are required. For the mid stack, where there are heaters both internally and externally, there is no need for extra heaters as the heaters in place are capable of delivering alot more power than planned to use in the test. In the top stack, the heaters on NanoProp can be used as safety heaters. However for the bottom stack, there are no heaters that can be used in case the experiments gets too cold. The heaters on the cover plates on the mid stack will help to heat the components that are closest to the heaters, but it will take some time before that heating is noticeable for the components at the bottom of the satellite. It is therefore necessary to add some heaters on the bottom of the satellite as well. The heaters added will be two Minco HK6908 connected in series.
