Pre-ApprovalWe are looking to provide a systems architecture for RIT SPEX to deploy solar cells to maximize the area of solar cells exposed to the sun during orbit and flight of a 1U CubeSat
StationaryWe have decided to design a solar cell array that deploys two wings on opposites sides of the vehicle. In conjunction with the three panels already in place and assuming a constant solar angle flight, which is common, it will expose three panels to the sun constantly. This design also needs to meet flight readiness, minimize mass and reduce cost for the SPEX research group.
- Two Deployable Panels On Opposite Sides
- Undeployed State- five Panels on x+-, y+- and Z-
- Deployed has three Z- panels, and x+- panels
- Hinge mechanism integrated directly into frame, reducing part count
- Metallic rail elements are kept to maintain adherence to flight standards and provide structure support
- The structure is designed to function both with P16102's vibration test and those set up by NASA
Hot KnifeTo deploy the solar cells, the elements should rotate around a hinge point integrated into the structure with springs also installed in there. To fully stow these components, a hot knife mechanism, common in space applications will be employed through the use of fishing line to hold the wings shut. On demand, a length of high resistance Nichrome wire heats up, forcing the nylon to break. The springs will then deploy the wings to be exposed to the sun.
Major Questions to AnswerBefore this presentation there were many questions:
- What frame building material would we need to choose?
- Was there acoustic coupling not yet accounted for?
- What thermal environment would the CubeSat be exposed to on launch and in space?
- Is there a connection or relationship between heat and vibration and failure modes?
After this presentation:
- Is this design ready to be fabricated and implemented?
- Are there any concerns about the choice of design or the potential failure modes under vibration?
- Are the contingencies accounted for in the design and implementation appropriate?
Material ChoiceWe have chosen Polyetherimide, better known by its commercial name: Ultem 9085, as the material out of which to print our components. This choice was made based on numerous advantageous material properties which are detailed in the next section.
For the rails, we are planning on using 6061, which is allowed by section 3.2.15 of the CubeSat design standard.
We are planning to tap and connect using M3x0.5 screws to connect most hardware and low-profile rivets if necessary.
Metal, glass, or fractured solar cells will be used as boilerplates for the actual solar cell devices mounted on five faces of the CubeSat.
A volume fitting mass analog to be made later will occupy the remainder of the 1.33 kg mass budget for the CubeSat as internal electronics.
Design Questions From Last Review
- Eigenfrequencies and Strengthening
- Machinability and Manufacturing Changes
- Thermal Analysis (with vibration)
- Acoustic Vibration
Changes to Final Design
Several 3D printable materials were compared across several properties. Upon analysis and comparison, we found Ultem to be the best choice for our design. Ultem offers the following advantages to our project:
- Superior strength/weight ratio to all but carbon fiber
- Commonly used in aerospace applications
- Naturally low outgassing
- Great thermal properties
- Clear manufacturability
- Well understood for spaceflight and recommended by authorities
Mechanical Design and Drawings
- Top and bottom extrusions dimensioned per CubeSat Launch Initiative
- Flanges for M3 tapped hole screw location
- Hollowed out for superior volume fraction
- Mass = 20.6 g/rail
- Added ribs for structural integrity
- Hole sized to chosen solar panel design
- Hinge has built in stop for 90 degree rotation
- Mass = 37.0 g
- Structural integrity enhanced through mounted solar panels
- Upper rib reduces bending
- Chamfered edges stopping friction prevented deployment
- Mass = 12.3 g/plate
Bottom plate feature dimensions
- Rib retaining the restraining wire adds strength
- Height into volume fraction = 2 mm
- Bolt holes for fastening to rail flange
- Mass = 10.7 g
- Added additional features to hot knife retention plate
- Aides in routing of wire and routing of electronics
- Reduces reliance on small bottom plate surface geometry
Hole placementA standard hole size of M3 was chosen for all screw attachments in the assembly. This allows the selected solar panels to be attached and eliminates the need for multiple varieties of screws. Screw holes were added to the top plate and the top solar panel was rotated 90 degrees from the previous design so that the screws would not interfere with the hinges.
Bill of MaterialsOur Bill of Materials has been updated to reflect the change in printing material from carbon fiber to Ultem and actual production quotes.
Design for Manufacturing
3D Printing of UltemUltem 9850 was recently made 3d printable, supplied in a filament form, the material must be heated to a high temperature in order to make it melt. Ultem selection meets a variety of selection criteria for ductility, heat transfer, and strength, and 3D printing the material allows us to make and design complex internal geometry which allows this project to capitalize on the technology and save a great deal of weight and volume.
Hyrel 3D Printing-AM InstituteOn campus, a Hyrel 3D printer was recently purchased, and it features a high temperature head capable of reaching 350 degrees C. This printer is designed for industrial applications, and features a higher resolution and printing quality than most commercial printers at 0.1 mm layer resolution (0.4 to 1 mm resolution). Hyrel will be able to provide a professional quality print in house and at the same time, its use is restricted to projects like P16102, meaning that the opportunities to print would be often.
Buying filament from a company such as Fisher Unitech would cost approximately $650 and be sufficient for 6 to 10 CubeSats in full based on testing and potential breakage.
Incept3d OutsourcingThe opportunity to outsource the printing is possible. Quotes filed through Incept3d indicates that it would cost upwards of $250 to set up for a few prints of the base plate. The turnaround time is unknown currently and the total cost of all printing would be approximately $500.
AnalysisUltimately, we are planning to purchase the Ultem filament from an industrial supply company and then use it in the Hyrel Printer here. So far with printing, out of the seven elements we've printed and the ten total components needed to print have failed, indicating a potential print failure rate of approximately 25%.
If we were to print on campus, we'd have an upfront cost of $650, and then the ability to reprint as necessary, in light of the fact that the printed mass of the CubeSat is under 100 grams and that future projects could use the material.
Outsourcing the printing, valued at potentially $500 with a 25% chance of failure and reprint not including failure through testing leads to the following potential costs, $500 + 25%($500) = $625 + shipping, handling, and leadtime.
MachiningThe rails will be machined out, and designed for it. A bar stock of 6061, which is permissible in the CubeSat design standard will be cut to net length. The pieces will be surfaced finished on two sides, then the end sections will be machined. After this, the part will be clamped from the end rails and machined out. Once the machining is complete, holes will be drilled and tapped. The process could be automated with Prototrak or done by hand. It will take 3 to 4 hours per rail to machine.
To assemble the system, the baseplate and hinge assembly will be made and then meshed to the rails along with the top plate that houses the wire path.
Ideally, we'd use screws at this level to allow us to assemble and disassemble easily, but could use rivets to finalize the assembly.
Dis-Assembly/ER 12 Time RequirementAs it stands right now, there are two methods to removing the circuitry core of the CubeSat design. The insides are designed to meet the Pumpkin CubeSat PCB specification and there is space for mounting.
Method 1- Fall Apart Method In this case, the screws connecting the side rails are removed, allowing the sides to fall away, exposing the core with the top and bottom plates still attached with the deployables still attached. This requires only 8 fasteners to be removed.
Method 2- Core Extraction A top or bottom plate could be removed, along with the supports connecting the core to the bottom or top plates, allowing the entire core to be removed independent of the frame housing.
Qualification of Design
Systems Architecture Chart
Launch Thermal Analysis
- Rail faces in contact with PPOD were set to constant temperature taken from thermal environment parameters of Launch Vehicles: Atlas II, Delta V, and Falcon 9
- Rest was set to both radiative and convection on all exposed boundaries
- The rails were cool and at around 320 K which is expected due to contact with the larger PPOD thermal mass
- All exposed areas reached a maximum of 343 K which is under minimum safe temperature of Ultem
In Orbit Thermal Analysis
- Simulation was performed with radiation from the sun incident on the main three solar panels
- The back of the satellite was subjected to radiative heat transfer with the earth
- The results show that temperatures the satellite will be subjected to will not damage components
- Our preliminary natural frequency was relatively so design changes were added to augment the natural frequency
- Stiffness's of low natural frequency components were increased to avoid resonance issues
- Lowest frequency = 322 Hz with Rib, and 180 Hz without Rib.
- Increased natural frequency of the deployables by nearly a factor of 2.
- Highest stress concentration is towards the top of the assembly.
- Not account for the chamfered deployable resting on the rails in the undeployed configuration.
- Maximum frequency from P16103 is 100 Hz which well below natural resonant frequency.
- P16103 is applying load orthogonal to the direction that would cause CubeSat to resonate at that lower frequency. Ultimately reducing risk of cube excitation of during P16103’s vibration test.
- Maximum stress = 1.7 MPa
- Load applied in direction most likely to cause resonance.
- Stress concentrated at screw locations, and at the non-ribbed deployable.
- At 100 Hz it shows that the resonant values found in the SW natural frequencies is not a concern at frequencies far from lowest natural resonance.
- Yield strength is well above maximum stress experienced from loads applied in all three directions.
- All LVs acceleration effects do not cause structural failure
- This analysis changes marginally with thermal loading during launch
- Most stress is concentrated in the hinge around 3 MPa
- Maximum stress are concentrated at screw locations into the rails
- The acoustic frequencies exceed our resonant frequencies however they are negligible in comparison to the stress induced through random vibration.
- Subject Matter Expert - Dr. Venkataraman reasoned that the acoustic pressure applied to the CubeSat would be so small that the effect of resonant can be neglected.
System Operations Flowchart
Prototype DevelopmentTo this point, the specificity of our tests as well as visualization of the overall design have been somewhat restricted. We sought to free ourselves from this restriction by making a full structural prototype. As all of our structural components have CAD models, this was accomplished using RIT's "The Construct". All parts were printed out of PLA which is free to us as students. Printing the sample parts also prepared us for the type of lead times we are likely to experience. In addition, we were able to verify the feasibility of our planned assembly method by assembling the prototype.
Previously conducted creep tests have the geometry and loading to properly reflect the conditions that the actual restraining line will experience. Using a 3D model of the CubeSat bottom plate and weights based off of the tension the actual springs will provide, a more relevant creep test was performed.
Printed Bottom PlateAs part of our plan to print a full prototype of the CubeSat, a bottom plate was printed out of PLA at RIT's "The Construct". This plate gives us a better idea of how friction and bending will effect creep.
TensionEach deployable solar panel is pushed open by two 0.05 in*lb (.0056 N*m) springs. The two springs pull on the restraining line with a lever arm of approximately 0.1 m. This provides a total of 0.1 N of tension on the line. In addition, the restraining line will intentionally be tensioned after installation for added security. A total tension of 1.6 N is assumed to incorporate this added tension. 1.6 N represents the amount of tension require to pull the fishing line taut. This tension is accomplished by hanging a 160 g mass from either side of the restraining line.
ResultsA 15.5 cm section of the restraining line was marked prior to the test. After leaving the setup at room temperature for four days, the marked section remained 15.5 cm in length. This shows a complete lack of creep and suggests that the creep we see on the final product will not exceed the 6.5 mm clearance between the resting state of a deployable panel and the PPOD wall. Although more testing will be needed next semester to further verify this result, this test suggests that the current restraining line design is acceptable and we are justified to move forward with it.
Lessons Learned from Assembly
We decided to print out the entire structure and practice putting it together. Through this experience, a lot was learned about how this assembly could be made better when released to manufacturing. Changes would include:
- Thickening Flanges of the Rail
- Wide the screw plates for tapping
- Spacing the fastener locations away from the rail
- Redesigning flanges
Risk AnalysisOur risk assessment has been updated to reflect the following developments:
- Low-Outgassing Ultem has replaced a carbon fiber - nylon matrix as out material of choice.
- All components will be fastened with screws and none will be glued.
- Our available budget and predicted expenses are now better known.
- We have made the choice to try to print Ultem components in-house.
Final Customer RequirementsWhen traced back to customer requirements, we show that our design meets the needs of our customer.
|ER and Description||Minimum||Ideal||Current||Status?|
|1 Min Undeployed Area Time (cm2-orbit)||100||100||100||Satisfied|
|2 Deployed Area Time (cm2-orbit)||150||300||300||Satisfied|
|3 System Mass (grams)||420||300||~360||Satisfied|
|4 CubeSat Clearance (mm)||1||1||>1||Satisfied|
|5 Time from Command (s)||2<t<8||2<t<8||~ 7||Satisfied|
|6 Operating Temperature (deg C)||(-40,40)||(0,20)||Material Selection||Satisfied|
|7 Available Interior (mm)||96 90 85||100 100 100||96 90 ~80||Satisfied|
|8 Exterior Protrusion Space (mm)||6.5/face||6.5/face||CAD||Satisfied|
|9 Change in COM (mm)||50||>10||Simulation||Satisfied|
|10 Constant Launch Acceleration (g)||6||8||Simulation||Satisfied|
|11 Cost of Prototype ($USD)||500||400||$100-$500||Satisfied|
|12 Time to Remove Blank (s)||8||6||<8||?|
|13 Launch Vibration Test (Pass/fail)||Pass||Pass||Simulation||Satisfied|
|14 Full Documentation||Pass||Pass||Pass||Satisfied|
MSD II Plans
|Supplier||Line Items||Status||Order Form|
Table Top Deployment
Facilities Required: None.
Engineering Requirements Tested: ER5-Time from command to deploy, ER11-Cost of prototype.
Timetable for Testing: 1/26/16 - 2/1/16.
Facilities Required: Oven and Freezer.
Engineering Requirements Tested: ER5-Time from command to deploy, ER6-Operating temperature.
Timetable for Testing: 2/2/16 - 2/8/16.
High Altitude Balloon
Facilities Required: High altitude balloon.
Engineering Requirements Tested: ER2-Maximum deployed area time, ER5-Time from command to deploy, ER6-Operating temperature.
Timetable for Testing: 2/9/16 - 3/7/15.
Bottom Plate Creep
Facilities Required: Oven.
Engineering Requirements Tested: ER8-Available exterior protrusion space.
Timetable for Testing: 2/9/16 - 2/15/16.
Full System Creep
Facilities Required: Oven.
Engineering Requirements Tested: ER6-Operating temperature, ER8-Available exterior protrusion space.
Timetable for Testing: 2/29/16 - 3/10/15.
Facilities Required: None.
Engineering Requirements Tested: ER4-PPOD-CubeSat clearance, ER8-Available exterior protrusion space.
Timetable for Testing: 2/16/16 - 2/22/16.
Facilities Required: P16103 test table. Shop Air.
Engineering Requirements Tested: ER13-Launch vibration test from 16103.
Timetable for Testing: 3/25/16 - 4/7/16.
Imagine RIT PlansDue to the linked nature of our projects, out group will share a booth with P16103. At least one member from each group will occupy the booth at all times throughout the day. Our part of the display will include the final CubeSat, a descriptive poster and PLA printed CubeSat parts that visitors can handle.
ApprovalAs we come to the end of MSD I, we seek approval from our guide for the following:
- Bill of Materials
- Design meets customer requirements
- Organization of documentation package
- Validation of design (physical testing and computer simulation)
- MSD II Schedule
- Action items can be found here: Action Items
- Imagine RIT Shared Vision: Imagine Vision
- MSD II Plan: MSD II
- MSD II Test Plan: Test Plan
- Phase I MSD II Shared Vision: Phase I
- Problem Tracking: P-Tracking
- Team Critique: Critique