Jonathan Sauder - Miniaturizing Mechanical Systems for CubeSats: Design Principles and Lessons Learned
- 1. MINIATURIZING MECHANICAL SYSTEMS FOR CUBESATS: DESIGN PRINCIPLES AND LESSONS LEARNED JONATHAN SAUDER, JET PROPULSION LABORATORY, CALIFORNIA INSTITUTE OF TECHNOLOGY SPECIAL THANKS TO: NACER CHAHAT, MANAN ARYA, SERGIO PELLIGRINO,ALEX WEN
- 2. BACKGROUND Developing a number of deployment mechanisms for a number of SmallSats/CubeSats including antennas for RainCube and MarCO Review of the first decade of CubeSats revealed that 90% had deployment mechanisms. Participating in the team which updated NASA Standard 5017 for mechanisms. 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 2
- 3. MECHANICAL LESSONS LEARNED FROM SEVERAL PROJECTS/TASKS Parabolic Antennas 4/18/2023 4 Reflectarray Antennas CalTech DOLCE Solar Array © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED.
- 4. SIMPLE ARCHITECTURES + LARGE MARGINS Design with significant margins. It is not unreasonable to expect up to 200 G’s (especially for low mass parts) For actuation force 3X minimum in early stages of design 2X in later design stages. Not reliant on momentum, adequate margin to maintain motion. What is the simplest approach that can do the job? Many Smallsats use a bent tape measure and burn wire. Be wary of complex failure avoidance approaches. 9/22/2022 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. PRE-DECISIONAL INFORMATION -- FOR PLANNING AND DISCUSSION PURPOSES ONLY 5 M-cubed/COVE
- 5. APPROACHESTO INCREASE MARGINS Find “Free” Margin 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 6 Self Help Kickoff Springs Motor, Lead Screw and Leverage RainCube What costs very little to increase performance for your spacecraft? Using geometry to create or increase forces. Extra force to get motion started, but don’t rely on momentum they create. Overcenter Mechanism and Kickoff Springs DOLCE Thermal Design Margin Europa Magnetometer
- 6. ROTARY MOTION, MOMENT ARM AND STRUCTURAL DEPTH Use rotary motion over sliding Rotary motion significantly reduces the effect of friction, which a vary widely at temperatures and in vacuum. Fully use depth and moment arm Depth has a linear effect on accuracy and a squared impact on structural stress/stiffness. 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 7
- 7. IFYOU MUST SLIDE LINEARLY, ENSURE L/D>2 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 8 Inadequate L/D jammed antenna on deployment.
- 8. PROTOTYPE EARLY AND OFTEN Empathize Define Ideate Prototype Test 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 10 “They that build the most prototypes win” -Justin Koch Design Thinking Method
- 9. SOME FURTHER NOTES ON DESIGN THINKING Empathize Define Ideate Prototype Test 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 11 Don’t jump to answers, ensure you are asking the right question first.You need to spend half the time understanding the problem, to ensure you are solving the right problem. Build experience quickly by trying and failing early and often. Don’t prototype everything all at once. Know why the prototypes fail. Make the cost of failure low. High failure cost will result in sunk cost fallacy. Solve functions dependently, don’t have one design aspect performing too many jobs.
- 10. PROTOTYPE EARLY AND OFTEN 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 12 Fail Early, Fail Cheaply, Fail Often = Rapid Progress Don’t prototype everything at once and understand why the system fails.
- 11. SOME FURTHER NOTES ON DESIGN THINKING Empathize Define Ideate Prototype Test 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 13 Don’t jump to answers, ensure you are asking the right question first.You need to spend half the time understanding the problem, to ensure you are solving the right problem. Build experience quickly by trying and failing early and often. Don’t prototype everything all at once. Know why the prototypes fail. Make the cost of failure low. High failure cost will result in sunk cost fallacy. Solve functions dependently, don’t have one design aspect performing too many jobs.
- 12. AVOID GOING STRAIGHT TO CAD… SKETCH FIRST Hand sketches Higher Fidelity Options Hybrid Hand Sketches (w/ CAD constraints ) PowerPoint CAD 11/28/2023 14 © 2022, Jonathan Sauder Image: NASA/JPL/Caltech
- 13. Hot impacts electronics, cold impacts mechanism grease TEST ASYOU FLY, AS EARLY AS PRACTICAL Thermal vacuum has broad impacts Testing with S/C like interface early is critical 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 15 Current draw was recorded in only 2 out of 7 occurrences due to non-flight software. ✓ X ✓
- 14. TEST ASYOU FLY EXAMPLE 9/22/2022 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 16 DynamicsTest Electrical EMI/EMCTests ThermalVacuum Test FunctionalTest
- 15. CONCLUSIONS When you are building miniaturized mechanisms on a tight budget, following good design practices is imperative. Several key “rules of thumb” are: Design for simple architectures with large margins To increase margin, use kickoff springs, self help, and find inexpensive margin in the design. Use rotary motion and moment arms to your advantage, with structural depth. If you must slide linearly, ensure an L/D > 2 Prototype early and often, to fail quickly cheaply, to rapidly learn and develop. Test as you fly as early as possible. Involve external reviewers early, as they can see things which may not be obvious. 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 17
- 16. Questions? The research was carried out at the Jet Propulsion Laboratory, California Institute ofTechnology, under a contract with the National Aeronautics and Space Administration (80NM0018D0004).The cost information contained in this document is of a budgetary and planning nature and is intended for informational purposes only. It does not constitute a commitment on the part of JPL and/or Caltech. Special thanks to: RainCube/KaPDA:Brian Hirsch, Brian Merrill (SMMS), Jeff Harrell, Pedro Moreira,Taryn Bailey, Hugo Rodriguez, Leah Ginsberg (intern), David Hunter (intern),Ted Steiner (intern) OMERA/MarCO: Nacer Chahat, Manan Arya,Vinh Bach, Phil Walkeymeyer, EllenThiel, Sean Dunphy, Mengjan Shi, Greg Agnes,Tom Cwik, Brian Merrill (Spectrum Marine and Model Services), MichelWilliam, andTyler Luchik who assisted in building the antenna, and Dr. Jefferson Harrell who performed the RF tests. DOLCE:AlexanderWen,Alan Truong, Charles Sommer,Terry Gdoutos, Sergio Pellegrino 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 18
- 17. BONUS CONTENT
- 18. SOME FURTHER NOTES ON DESIGN THINKING Empathize Define Ideate Prototype Test 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 20 Don’t jump to answers, ensure you are asking the right question first.You need to spend half the time understanding the problem, to ensure you are solving the right problem. Build experience quickly by trying and failing early and often. Don’t prototype everything all at once. Know why the prototypes fail. Make the cost of failure low. High failure cost will result in sunk cost fallacy. Solve functions dependently, don’t have one design aspect performing too many jobs.
- 19. CONCEPT FOR AVENUS ROVER 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 21
- 20. DON’T PROTOTYPE EVERYTHING ALL AT ONCE 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 22 Generator SiC or GaN Transceiver SiC and GaN Instruments Wind Turbine Distributed Mechanical Sensors Array NaCl Batteries Reversing Gearbox Front Wheels Rear Wheels Steering Generator Cond. Battery Cond. M U X Gear Ratio (down) Differential Gear Ratio (up) Clutch Power Controls Clock Voltage Regulator Mechanical Electrical Short-Term Course Correction Long-Term Guidance Motor Wind Vane Stored Spring Energy
- 21. PROTOTYPED ROVER 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 23
- 22. WHY START ROUGH? Lets say I want to measure the diameter of a quarter, I could use: A yardstick: Cost: $1s Time: 5s A calipers: Cost: $10s Time: 15s A micrometer: Cost: $100s Time: 30s A laser microscope:Cost: $10,000s Time: Min. Which do you choose? Depends where you are 11/28/2023 24 © 2022, Jonathan Sauder
- 23. SOME FURTHER NOTES ON DESIGN THINKING Empathize Define Ideate Prototype Test 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 25 Don’t jump to answers, ensure you are asking the right question first.You need to spend half the time understanding the problem, to ensure you are solving the right problem. Build experience quickly by trying and failing early and often. Don’t prototype everything all at once. Know why the prototypes fail. Make the cost of failure low. High failure cost will result in sunk cost fallacy. Solve functions dependently, don’t have one design aspect performing too many jobs.
- 24. AVOID COUPLING IN SOLUTIONS 11/28/2023 26 © 2022, Jonathan Sauder Functions Design Parameters R Lever L Lever Speed X X Direction X X Functions Design Parameters Pedal Wheel Speed X Direction X Images from Consumer Reports 26
- 25. EXAMPLE: COUPLED FORM IS OK 11/28/2023 27 © 2022, Jonathan Sauder Functions Design Parameters Hot Knob Cold Knob Flow X X Temperature X X Functions Design Parameters Height Angle Flow X Temperature X Images from Amazon.com 27
- 26. SEPARATE STIFFNESS AND STRENGTH (12) Stiffness in reliant on material and geometry. Strength is reliant on material and geometry. This is a coupling issue! 11/28/2023 28 © 2022, Jonathan Sauder Functions Design Parameters Steel Triangle Stiff Structure X X Strong Structure X X 28
- 27. SEPARATING STIFFNESS AND STRENGTH Know your actual requirements A stiff structure does not always need to be strong (optics, eg. MOIRE) A strong structure does not need to be stiff (Fort Moultrie, Palmetto Logs) 11/28/2023 29 © 2022, Jonathan Sauder South Carolina State Museum DARPA 29
- 28. SEPARATE STIFFNESS AND STRENGTH (12) Just because something is stiff, doesn’t mean it is strong enough, and just because it is strong, doesn’t mean it is stiff enough. In fact, stiffness can kill your design! (CTE) 11/28/2023 30 © 2022, Jonathan Sauder 30
- 29. INTERFACE THATWAS TOO STIFF Press fit with different stainless steels. 440 series (martensitic) vs. 300 series (austenitic) 11/28/2023 31 © 2022, Jonathan Sauder 303: 17.3*10-6 mm/mm°C 440C: 10.1*10-6 mm/mm°C 31
- 30. STIFF INTERFACE: JPL EXAMPLE Stainless Steel E = 210 GPA EA 9360 Epoxy E = 3.36 GPA And we athermalized the design! 11/28/2023 32 © 2022, Jonathan Sauder 32
- 31. ATHERMALZING A DESIGN Use a material with vastly different CTE to result in a near zero change between the other materails. 11/28/2023 33 © 2022, Jonathan Sauder 303: 17.3*10-6 mm/mm°C 440C: 10.1*10-6 mm/mm°C Epoxy: 60*10-6 mm/mm°C 33
- 32. EXACTLY CONSTRAIN DESIGNS Exact Constraint can avoid the coupling between stiffness and strength. Underconstraint Issues Location is not defined Has degrees of freedom which can lead to rattling. Overconstraint Issues Tolerances need to line up perfectly or they will introduce stress to the design. CTE will wreak havoc on your design 11/28/2023 34 © 2022, Jonathan Sauder 34
- 33. CONSTRAINTS AND DEGREES OF FREEDOM 4/18/2023 © 2023 CALIFORNIA INSTITUTE OF TECHNOLOGY. GOVERNMENT SPONSORSHIP ACKNOWLEDGED. 35 A plane is really 3 contact points. A line is really 2 contact points. So when thinking about constraints, think about points of contact. The definition of the constraint line created by a constraint. Each constraint creates a line perpendicular to its contact point with the object.
- 34. WHATTYPE OF CONSTRAINT DO THE KINEMATIC MOUNTS PROVIDE? 1/11/2018 36 © 2016, Jonathan Sauder 3 Kinematic Mounting Features (6 points of contact) In Hinge Joint Nesting force from torsion spring Hinge Pin X Hinge Pin
- 35. LOOKING AT THE 2D CASE 1/11/2018 37 © 2016, Jonathan Sauder
- 36. LOOKING AT THE 2D CASE 1/11/2018 38 © 2016, Jonathan Sauder 1 2 3 1-2 2-3 1-3
- 37. LOOKING AT THE 2D CASE 1/11/2018 39 © 2016, Jonathan Sauder 1 2 3 1-2 2-3 1-3 CCW CCW CW
- 38. LOOKING AT THE 2D CASE 1/11/2018 40 © 2016, Jonathan Sauder 1 2 3 1-2 2-3 1-3 CCW CCW CW
- 39. LOOKING AT THE 2D CASE 1/11/2018 41 © 2016, Jonathan Sauder 1 2 3 1-2 2-3 1-3 CCW CCW CW
- 40. LOOKING AT THE 2D CASE 1/11/2018 42 © 2016, Jonathan Sauder 1 2 3 1-2 2-3 1-3 CCW CCW CW
- 41. WHATTYPE OF CONSTRAINT DO THE KINEMATIC MOUNTS PROVIDE? 1/11/2018 43 © 2016, Jonathan Sauder X Hinge Pin 3 Kinematic Mounting Features (6 points of contact) In Hinge Joint Nesting force from torsion spring Hinge Pin ✓ ✓ X X
- 42. THE ECCENTRIC HINGE 1/11/2018 44 © 2016, Jonathan Sauder