RAX: Lessons Learned in Our Spaceflight Endeavor Matt Bennett University of Michigan CubeSat Workshop Cal Poly, San Luis Obispo April 21 st, 2010
Background Sponsored by National Science Foundation University of Michigan and SRI International Collaboration Co-investigators: 1. 2. Dr. Hasan Bahcivan, SRI International Prof. James Cutler, University of Michigan Concept for mission developed at CubeSat Workshop 2 years ago AOSS/AERO: 582 2
Objective: Study the microphysics leading to the formation of magnetic field-aligned plasma irregularities (FAI) Why Study FAI from Space? Kilometer-scale FAI form in the ionosphere and cause scintillation of radio signals Microphysics behind FAI formation and distribution not understood FAI measurements not possible using ground radars at upper latitudes because scatter perpendicularity condition not met Mission Overview Receiving backscatter normal to magnetic field is possible in a near-polar orbit because of the wealth of approach angles
Interrogate FAI with pulses from incoherent scatter radar in Poker Flats, AK (PFISR) Pulses will scatter off the anomalies into space RAX will be in position to receive scattered signals and record them to memory for processing and downlink to Earth stations Experiments performed daily for one year Concept of Operations Data will be used to generate FAI forecast models
Where it All Started Grant received September 2008 with only 11 months to deliver a flight-ready spacecraft! All we had was an empty lab, a handful of students, and a picture
Lesson #1: Building Nanosatellite Program Infrastructure Doesn t Happen Overnight Took two months to fill lab with basic tools for design, fabrication, and testing Took four months to acquire/train the core team, harness resources First Michigan-designed boards fabricated nine months* after project start * Note: Delivery date had slipped by four months at approx. six months into schedule
Lesson #2: Utilize Professional Engineering Expertise in Early Design Stage Challenge: How do we record two independent data streams at 1 MHz sample rate with 16-bit resolution? FPGA ideal solution Dedicated hardware No latency associated w/ µ-processors running an OS Triple modular redundancy No FPGA expertise on RAX Team! Solution: Sought professional engineering expertise Paid for engineering consulting time during architecture phase Student involved in architecture; later assumed design, build, test duties This subsystem had only one revision prior to final flight product
Lesson #3: Off-the-Shelf Solutions Not Always Off-the-Shelf Benefits Burdens Saves development time and cost Build-to-Order = Long lead times Sometimes ships as tested to expected environment or have flight heritage Cannot test compatibility with other systems immediately One solution cannot meet all needs
Lesson #3: Off-the-Shelf Solutions Not Always Off-the-Shelf RAX Structure Could not accommodate active GPS antenna Assembly order demanded custom top plate PEM nuts added for additional panel security
Lesson #3: Off-the-Shelf Solutions Not Always Off-the-Shelf Bottom Line: COTS work, but allocate schedule margin for testing and revisions! RAX Power System Single COTS solution did not work Timeline demanded we design our own Final system less efficient, but very reliable
Lesson #4: ASSUME NOTHING! November Qualification Vibe: NOT required internal engineering design validation Shook RAX development unit to 14.1 g rms (GEVS Standard) Acceptance Vibe: Required shake level was 0.78 g rms Test due February 1 st, delivery May 24 th Expected ability to remove RAX from POD after test and deliver in May Challenge: Correct issues and prep all hardware for flight in 13 days (WRONG!)
The Race to Finish RAX What we had to do to make deadline: Designed new solar panels and three new subsystems boards in one night Staffed clean room 24/7 to complete board fabrication Flew panels to NASA GRC for flash testing to save time Performed integration, staking, mass property check, and checkout testing in 24 hours leading up to departure for vibe RAX Excerpt Team from Daily co-manager s Caffeine email: Intake Matt's Schedule: 5am - 8am - Sleep 8am - 10am - Board Staking 10am - 2pm - Sleep 2pm - 6pm - Integrated Assembly 6pm - 7pm - Packing for Cal Poly 7pm - 12am - Full Test 12am - 2am - Panels 2am - 5am Sleep 6am Depart to Cal Poly
The Race to Finish RAX RAX Team Daily Caffeine Intake Excerpt from co-manager s email: Matt's Schedule: 5am - 8am - Sleep 8am - 10am - Board Staking 10am - 2pm - Sleep 2pm - 6pm - Integrated Assembly 6pm - 7pm - Packing for Cal Poly 7pm - 12am - Full Test 12am - 2am - Panels 2am - 5am Sleep 6am Depart to Cal Poly
Lesson #5: Transporting Spacecraft to Test/Launch Facilities Save yourself this problem; contact Dept. of Homeland Security a month in advance before transporting via airline, or ship to test/launch site
Acceptance Vibe Successful
Lesson #6: Looped Coax Makes a Great Antenna! GPS malfunction due to coax radiating received signal in coiled configuration GPS vital to science experiment Corrective action: swap position with instrument data processing unit RAX already vibed and sealed no access permitted to fix the problem What did we do??
Lesson #7: Buy Enough Parts to Make Two Flight Units! We built another flight unit! RAX-1B identical to RAX-1A save for GPS position swap Launch slip permitted us to do this at a safe, steady pace Will qualify this unit for launch in May RAX-1A will be fixed and re-qualified for other flight opportunities or lab unit to troubleshoot 1B orbit issues
Lessons Wrap-Up What to take away from this talk 1. Take your time to build development infrastructure 2. Take advantage of professional engineering early in your design 3. Consider COTS components carefully before you buy Consider delivery timeline Test thoroughly, test early 4. Assume Nothing Demand for early vibe was costly, both financially and physically Software development time lost to building second unit Problem was avoidable if clear understanding was established in advance
Launch Launch Date: September 2010 Launch Vehicle: Minotaur IV Launch Location: Kodiak Launch Complex (AK) Orbit: 650 km circular, 72º inclination
It s All About BOLD FLIGHT
NSF: Therese Jorgensen Acknowledgements Co-PI s: Dr. James Cutler and Dr. Hasan Bahcivan Scott Schaire at NASA Wallops NASA Glenn Research Center Space Test Program Cal Poly San Luis Obispo University of Michigan supporting professors and staff The CubeSat Community Follow our progress at http://rax.engin.umich.edu