Lim Tze Yee
Year 4 Mechanical Engineering
A0264670B
3. Thermal Analyses and Certification
3.1 Scope of Project
3.1.1 Problem
CubeSats are exposed to a large range of temperatures during orbit. While deep space has an approximate temperature of 4K, heat fluxes incident on the CubeSat due to the Sun and Earth can drastically heat a CubeSat up, creating massive temperature differentials. However, most of the CubeSat’s internal components have a limited operating temperature range, and must remain within these limits throughout the CubeSat’s orbital life cycle to avoid component failure that may compromise the mission.
3.1.2 Objective & Design Statement
This portion of the project seeks to obtain critical thermal telemetry to inform a thermal management system for the CubeSat that ensures all internal components remain within their allowable operating temperatures throughout the orbital life cycle. It shall provide information about potential thermal pitfalls that must be accounted for in the CubeSat’s thermal design to minimise mission risk.
This project shall firstly characterise the thermal environment, the CubeSat’s mechanical configuration and the operating temperature limits of its components. It will then evaluate the mission’s parameters and its influence on the CubeSat’s thermal behaviour. Subsequently, thermal simulations will be performed to assess the thermal feasibility of the CubeSat and inform appropriate thermal management strategies.
3.2 Context of the Problem
This section lists the values used to quantify the boundary conditions a CubeSat is exposed to during orbit and the temperature range within which each satellite component / subsystem has to stay within to keep operational.
3.2.1 The orbital environment
Galassia - 5’s orbital parameters:
- Altitude: 550km (on average)
- Orbit: Sun - Synchronous Orbit (SSO)
- In SSO orbit, the satellite passes over the same part of Earth at roughly the same local solar time each day (24 hour period).
- Period: 96 min
- Sunlit phase: 64 min (direct exposure to the Sun)
- Eclipse phase: 32 min (completely hidden from the Sun)
For the purpose of thermal simulations, the following values, attained from Mr Chua Tai Wei from ST Satellite Systems, will be used to define the orbital environment:
| Environmental Parameter: | Physical Meaning: | Value: |
|---|---|---|
| Solar Constant | The amount of solar energy incident on a surface with a direct line of sight to the Sun. This parameter is only applied during the sunlit phase of the orbit. | 1361 W/m2 |
| Earth Albedo | The portion of incoming radiation from the Sun that is reflected by Earth’s surface back into space. This parameter is only applied during the sunlit phase of the orbit. | 408 W/m2 |
| Earth IR (infrared) | The infrared thermal radiation that Earth naturally emits due to its temperature. Unlike Solar Constant and Earth Albedo, Earth IR is present throughout all phases of orbit. | 220 W/m2 |
| Deep Space | The cold background environment surrounding a CubeSat, which is approximated as a radiative heat sink of constant temperature. It acts as the primary source to which the CubeSat loses heat due to thermal radiation, and is applied on all exposed surfaces of a CubeSat. | 4K |
Table 3.2.1. Low Earth Orbit Environmental Constants
3.2.2 System requirements
Below is a diagram of the mechanical architecture containing key components of the Galassia-5 CubeSat, followed by a list of components with their respective functions and operating temperatures.
Fig 3.2.2. Mechanical Architecture of Galassia-5’s internal components
| Subsystem Component | Function |
|---|---|
| ADCS (Attitude determination and control system) | Determines and controls the Satellite’s orientation in Orbit. Uses sensors and actuators to maintain correct pointing for communications and operations |
| Battery + heater pack | The battery stores electrical energy for use when the CubeSat is not receiving solar power, while the heater prevents the battery (and other components) from falling below its lower operational temperature limit by emitting radiative heat. |
| EPS (Electrical Power System) | Receive electrical energy from the solar panels, charges the battery and delivers the required power to each subsystem within the CubeSat. |
| GPS (Global Positioning System) | Determine the satellite’s position, velocity and time throughout the orbit. |
| HPA (High Power Amplifier) | Amplify the radio frequency signal before sending it from the satellite to the ground station. |
| OBC (Onboard computer) | CubeSat CPU (Central Processing Unit), executes mission software, controls all other subsystems and stores payload data |
| Skunkworks Imager + Payload camera | Payload camera with its own OBC for control of the payload. |
| Solar Panels | Convert sunlight into electrical power that powers the CubeSat |
| TT\&C (Telemetry, Tracking and Command) | Enable two-way communication between the satellite and the ground station on Earth. |
| Upconverter | Electronic component that shifts a signal from a lower frequency to a higher frequency for transmission. |
Table 3.2.2.1. Subsystems and their functions
| Subsystem Component | Operating Temperature Range (°C) |
|---|---|
| ADCS (Attitude determination and control system) | -20 to 80 |
| Battery + heater pack | 0 to 45 (charging) -20 to 60 (discharging) |
| EPS (Electrical Power System) | -35 to 85 |
| GPS (Global Positioning System) | -30 to 85 |
| HPA (High Power Amplifier) | -40 to 55 |
| OBC (Onboard computer) | -30 to 85 |
| Skunkworks Imager + Payload camera | -20 to 55 |
| Solar Panels | -100 to 100 |
| TT\&C (Telemetry, Tracking and Command) | -40 to 75 |
| Upconverter | -40 to 85 |
Table 3.2.2.2. Subsystems and their operating temperature ranges
3.3 Methodology
###
3.3.1 Simulation Methodology
In the space industry, thermal studies are carried out through simulations using engineering software (this project uses SolidWorks) to attain a range of expected temperature values during different phases of orbit. While simulations are inherently known not to be a complete reflection of the actual environment, it is still an important step in the CubeSat development process. When done properly, thermal simulations provide the necessary data for a well-constructed thermal management plan.
The gist is simple – create a simplified CAD model of the CubeSat, input space’s boundary conditions and internal heat loads for different orbital and mission phases. However, the challenge lies in balancing model fidelity and efficiency (having a higher fidelity model means more accurate results but also much longer simulation processes, and at times the software itself is unable to run with too many input parameters) and ensuring that boundary conditions and thermal loads are correctly applied.
3.3.2 Verification of Simulation Methodology
###
3.3.2.1 Consultation with Dr Menendez from Taiwan
Building on from my work done up till the Interim Presentation, I consulted with an industry expert, CubeSat Project Manager & Systems Engineer Dr Angel Menendez from the National Cheng Kung University in Taiwan, to obtain a CubeSat’s on orbit thermal telemetry. His CubeSat, Lilium-1, which was of size 3U, was similar in size to Galassia-5’s. More importantly, it had the same type of orbit (SSO), which meant that it was exposed to similar orbital parameters. The Lilium-1 cubesat also utilised similar commercial off the shelf components and had a similar thermal output to Galassia-5’s.
Although the information about Lilium-1 is not fully comprehensive (due to sensitivity issues, a breakdown of the mission modes, power budget and payloads could not be shared), data regarding the range of temperatures on the external surfaces of Lilium-1 in SSO was provided. To provide confidence in my simulation methodology, it would be applied to the Lilium-1 CubeSat to obtain predicted external temperatures of the CubeSat, which will be compared against on orbit telemetry. The objective would not be to reproduce the exact on orbit data, as it would require precise knowledge of all environmental conditions in orbit and implementation into engineering software, many of which are either unknown or cannot be modeled in full fidelity. Instead, the comparison was used to assess whether the methodology could produce results of a comparable temperature range.
Below are the information provided by Dr Menendez regarding Lilium-1:
Fig 3.3.2.1.1. Dimensions of Lilium-1
Fig 3.3.2.1.2. Orbital orientation of Lilium-1
Fig 3.3.2.1.3. Arrangement of components in Lilium 1 and list of components
3.3.2.2 Lilium-1 CAD model
In building the Lilium-1 CAD model, upon consultation with Dr Menendez and understanding of industry practices, the main objective was to attain a simplified but representative CAD model that could be reliably used for thermal analysis. The focus was on preserving the external dimensions of the CubeSat, physical layout and major interfaces on which heat loads (especially boundary conditions) would be applied.
Fig 3.3.2.2.1. Isometric View of Lilium-1 CAD Model
Fig 3.3.2.2.2. Side view of Lilium-1 CAD model showing its internal components
3.3.2.3 Setup of Boundary Conditions for SSO
While the heat sources and the heat sink are the same for all CubeSats, different values of the boundary conditions must be applied in accordance with the CubeSet’s orientation during orbit, and rotation of each face on the CubeSat with respect to either the Sun or Earth (or both) must be accounted for.
In the case of Lilium-1, due to its mission parameters, its payload camera is meant to always point towards Earth. Hence, these surfaces (+Z) will receive constant Earth Albedo (during sunlit phase) and IR (throughout) heat fluxes, but other surfaces (+X, -X and -Z) will have varying incident Solar Constants due to the angle at which the face faces the sun at different phases of the orbit. In SolidWorks transient thermal simulations, the applied heat fluxes from Table 3.2.1 are varied across a period of 5760 seconds (96 mins), the duration of one full orbit. The Sunlit Phase is set from 0s to 3360s and 5281s to 5760s, while the Eclipse phase is from 3361s to 5280s. Details of the adding of the heat loads with respect to time during each orbit can be found in Appendix C.
Fig 3.3.2.3. Schematic of the Lilium-1 CubeSat in orbit, moving in the direction of the +X axis, with the +Y axis coming out of the plane and the +Z axis pointing towards the Earth at all phases of orbit
3.3.2.4 Simulation results
With the above boundary conditions used, the simulations are repeated until two consecutive simulations show the same temperature range (to a precision of 4 significant figures), which is taken to be the steady state orbital thermal performance of the CubeSat.
Below are the results:
Fig 3.3.2.4.1. Lilium-1 Sim Hottest Results (31°C maximum)
Fig 3.3.2.4.2. Lilium-1 Sim Coldest results (-4°C minimum)
These results show that temperatures on the CubeSat’s exterior ranges between -4 and 31°C.
3.3.2.5 Verification of Simulation results
Dr Menendez’s thermal telemetry shows that external temperatures recorded on the 6 surfaces of the primary 3U structure ( +/- X/Y/Z) on 9 separate orbits ranged between -22°C and 45°C, with a temperature difference of up to 62°C within an orbit.
| Orbit Number | Minimum Temperature (°C) | Maximum Temperature (°C) | Temperature Difference (°C) |
|---|---|---|---|
| 1 | -21.9286 | 19.0186 | 40.9475 |
| 2 | -17.9099 | 11.0620 | 28.9719 |
| 3 | -19.9665 | 26.0153 | 45.9818 |
| 4 | -8.9380 | 26.0153 | 34.9533 |
| 5 | -16.9754 | 43.0246 | 60.0000 |
| 6 | -16.9754 | 43.0246 | 60.0000 |
| 7 | -16.9754 | 44.0527 | 61.0281 |
| 8 | -1.9386 | 42.0901 | 44.0287 |
| 9 | -21.9286 | 19.0060 | 40.9346 |
Table 3.3.2.5. Temperature ranges on the Lilium-1’s external surface across different orbits
My simulated results were found to have a narrower range than the recorded telemetry, which I (and G5’s industry partners) believe are reasonable, as the simulation was conducted under idealised assumptions, such as perfect orientation of the CubeSat throughout orbit. It does not account for other variables such as imperfect orientations or thermal loads from payloads that could affect the actual temperature range. The comparison therefore supports the methodology as a useful predictive tool to guide G5’s thermal management, although a safety factor should be included in the thermal design to account for variances in actual orbit.
Further simulations in Appendix D showcase that other orbital orientations could explain the temperature extremities, as actual orbital conditions are unlikely to be as ideal as simulated.
3.4 Fabrication of Prototype
3.4.1 Development of Galassia-5 CAD Model
Using the verified methodology from testing with the Lilium-1 CubeSat, a simplified CAD model for G5 was developed.
Detailed CAD of G5:
Fig 3.4.1.1. Isometric view of G5
Fig 3.4.1.2. Side view showing G5’s internal components
Simplified CAD model to be used for Simulations:
Fig 3.4.1.3. Isometric view of the simplified G5 CAD
Fig 3.4.1.4. Side view showing G5’s internal components
3.4.2 Application of Boundary Conditions
The same values of the thermal boundary conditions were applied to G5’s respective faces. However in the case of G5, unlike that for Lilium-1, G5’s solar panels would be facing the Sun throughout orbit. Hence, Solar Constant on the solar panels is kept constant at 1361 W/m2 while the incident Earth IR and albedo values change according to orbital position. Similarly to that of Lilium-1, emissivity values of external-facing surfaces were set at 0.88, while that of inwards-facing surfaces were set at 0.03.
Fig 3.4.2. Schematic of G5 in orbit, showing its constant orientation with respect to the Sun while different surfaces face Earth at different phases of the orbit
3.5 Simulation testing and results
Thermal simulations were conducted for G5 to determine the maximum temperature range it may experience over the course of its mission throughout different mission and orbital phases. To ensure a conservative assessment, the situation parameters were selected to represent the most extreme hot and cold cases that G5 may encounter in orbit. Although actual on-orbit temperature range is expected to be narrower, using these results provide a prudent basis for informing the CubeSat’s thermal management plan.
For all G5 simulations, the hottest and coldest temperatures have repeatedly shown to be at the end of the Sunlit and Eclipse phases of the orbit. Hence, the temperatures at those points will be used to define the orbital temperature range of each subsystem.
3.5.1 Autonomous Simulations
In G5’s autonomous mode, its Solar panels are always facing the Sun for maximum charging. Heat fluxes from the Sun and Earth are focused on the CubeSat’s larger surface areas (Solar panels and 6U surfaces respectively), causing the CubeSat to receive a large amount of heat from its environment.
The heat loads applied are shown in Appendix E.
Fig 3.5.1.1. Simulation results showing the hottest temperatures
Fig 3.5.1.2. Simulation results showing the coldest temperatures
Fig 3.5.1.3. Simulation results showing the temperature range of Solar Panels during orbit (from -5 to 70°C)
3.5.2 Simulations for hottest scenario (Mission Mode)
For these simulations, it is assumed that all the components are turned on at the same time, and operate together during the hottest phase of the CubeSat’s autonomous orbit (end of the sunlit phase) to simulate the hottest possible scenario for G5.
Fig 3.5.2. Maximum expected temperature (77.3°C)
3.5.3 Simulations for coldest scenario (Orbit without any thermal management)
In these simulations, they account for conditions in which the CubeSat fails to maintain its orientation towards the Sun. Without the CubeSat’s larger surfaces facing the heat sources directly, the CubeSat is expected to gain less radiative heat flux, and have lower temperatures.
Without thermal management (optimal orientation of the CubeSat), these simulations assume perfect rotation such that each face gets equal amount of exposure to each heat source throughout orbit.
Fig 3.5.3.1. Simulation results showing the hottest temperatures
Fig 3.5.3.2. Simulation results showing the coldest temperatures
Fig 3.5.3.3. Simulation results showing a much lower temperature range of Solar Panels during orbit (from -46 to 7°C)
3.6 Analysis
These results show that under the current orbital environment, background heat sources are sufficient to keep many of G5’s subsystems above their respective minimum operating temperatures, and that overheating is not an issue for most of the architecture.
3.6.1 Analysis of Cold temperature limits
On the lower end however, the battery pack and heater as well as the primary payload (Skunkworks imager) fall below their operating temperatures, while the High Power Amplifier comes close to its thermal limit (simulated -33°C, with its lower limit being -40°C). Under autonomous orbits without an internal thermal management plan, the battery pack may drop below 0 degrees during autonomous orbit, preventing it from charging effectively, potentially affecting the mission should it be unable to store energy from the solar panels for future redistribution. While the imager heats up sufficiently during the Sunlit phase, as it is not within operating temperatures throughout the orbit, it may not be able to reliably perform its function.
In consideration of this, further simulations were carried out under a simple thermal management plan, with the heater being turned on during the Eclipse phase using stored power to keep the battery pack above 0°C.
Fig 3.6.1. Simulation results showing temperatures of internal components with the heater turned on
These simulations show that the heater does provide sufficient power to keep the battery pack within its charging temperature range throughout the orbit. For the payload imager, the heater is unfortunately too far to provide power to heat it up sufficiently. For context, the current mechanical architecture is decided in lieu of mechanical constraints due to existing subsystem dimensions. With regards to the thermal conditions of the payload imager, the payload team has in fact recognised that the current setup may render the imager too cold for operation, and have begun work on a separate heater within the payload.
An additional point to note for these simulations is the chosen emissivity value of 0.03, which shows a strong tendency for internal subsystems to retain heat. The actual emissivity of internal subsystems may not be as low, especially as it may be increased to facilitate dissipation of heat from heating hotspots.
To guide the colder end of the thermal management plan, measures are needed to ensure sufficient heating to keep the components within their operating temperature ranges. This is especially critical for subsystems away from the solar panels and heater, as they receive much less heat during the sunlit phase, being on the opposite end of the CubeSat facing the Sun. Furthermore, during the Eclipse phase, these subsystems are pointed at deep space, enhancing their cooling effect. While the payload imager may have its own heating measures, the High Power Adapter and subsystems around it must be considered too.
3.6.2 Analysis of Hot temperature limits
Due to the cold heat sink of deep space, overheating is less of a concern compared to excessive cooling. However, there may be local overheating within these mission-critical power intensive components during imaging and transmitting phases. Furthermore, the current simulation assumes equal heat dissipation throughout the entire subsystem, whereas in reality most of the power usage will be concentrated within a few specific components in the subsystem, and hence higher than simulated temperatures are expected. Currently, these specifications are not known yet as the components are still in development.
For these power intensive components, a heat sink will likely be required to facilitate quick cooling during certain mission phases. It must also be noted that the heat sink may also drive up internal heating requirements to keep all components within operating temperatures and its implications on cooling effects must be considered.
Thankfully, not all components are expected to run concurrently as it would be extremely power intensive and may constrain the power budget. However, accounting for this scenario in the thermal management plan ensures that all possible overheating cases are well accounted for in the CubeSat’s thermal design, minimising risk.
3.7 Concluding Statements
###
3.7.1 Conclusion
In conclusion, after multiple iterations, the G5 team has a thermal management methodology that has been verified with on orbit telemetry. Using this methodology, it firstly shows that for the external structure of the CubeSat, it assures that the Solar Panels stay within its operating temperature limits, and that the intended power draw will not be affected.
Internally, components away from the heater and solar panels will be pretty cold and likely fall below operating temperatures in G5’s current configuration. At the same time, power intensive components, especially the payload computer, may experience local overheating during specific mission phases. Moving forward, these are the 2 more important considerations when finalising the thermal management plan for G5 to ensure thermal compatibility across all mission phases.
Finally, while the current methodology can be used to guide design of a thermal management plan, it would be very conservative due to the existing limitations. In fact, such SolidWorks methodology has been used for CubeSats, most notably in G5’s predecessor, Galassia-2. However, most CubeSat teams (such as Lilium-1) prefer to use industry-standard software such as ANSYS for thermal simulations, due to its capability for higher fidelity modelling of complex thermal interactions.
3.7.2 Current limitations
The main limitation would be the CAD model’s fidelity. While it rather accurately represents the external CubeSat structure, it misses out on details for subsystems. Heating power is assumed to be radiated out evenly from all surfaces for simulations, when it is actually concentrated on specific components. Upon learning from experienced industry experts, while similar methodologies are used in the early stages to speed up the process of attaining results, limitations due to simplified CAD geometries must be accounted for in the latter stages.
Additionally, until the mechanical architecture, material properties, and subsystem placement are finalised, uncertainty remains in the modelling of internal thermal conduction, which may cause variations in simulation results. Preliminary results may not be representative, should there be major changes in the CubeSat’s internal design. Finally, a further limitation of the current model arises from the use of SolidWorks instead of software such as ANSYS, which has a higher capacity for representing thermal exchanges in a CubeSat. Therefore, while the present model does surface key focal points for designing a thermal management plan, detailed results and analysis can produce a more comprehensive thermal verification.
3.7.3 Future work
Future work should aim to improve the fidelity of the thermal model through more detailed geometry (especially for subsystems). Most importantly, as some subsystems in the CubeSat are still in its iterative phase, the thermal modelling of G5 must keep up with each iteration to inform component placement and systems architecture, especially when technical specifications such as internal contact points and surface properties are finalised. As the project nears its completion, the thermal design should also look towards physical testing in a Thermal Vacuum to validate the thermal management plan. Once it passes physical testing, the CubeSat will be ready for launch.