Rideshare-Initiated Constellations: Future CubeSat Architectures with the Current Launch Manifest

Transcription

1 Rideshare-Initiated Constellations: Future CubeSat Architectures with the Current Launch Manifest Joseph Gangestad, James Wilson, Kristin Gates, and John Langer The Aerospace Corporation National Space Symposium Colorado Springs, CO April 2015 The Aerospace Corporation 2015

2 CubeSats: Commoditizing Access to Space Launch costs driven by supply and demand The P-POD frees CubeSats from a specific launch vehicle. 2

3 Building Constellations via Rideshare Walker Constellation Rideshare-Initiated Constellation Symmetry provides regular coverage Built up one rideshare at a time How does a rideshare-initiated constellation compete on traditional coverage metrics: revisit time, daily visibility? 3

4 Coverage of the Aerospace Constellation: Max Revisit The Aerospace Corporation has 6 CubeSats on orbit, launched between 2012 and 2014, and 6 more will launch in Addition of 2 polar satellites (AeroCube-6) dramatically reduces maximum revisit time. 4

5 Rideshare Manifest: # Date Vehicle Apogee [km] Perigee [km] Inclination [deg] LTAN 1 Apr 2015 Falcon N/A 2 May 2015 Soyuz :00? 3 May 2015 Atlas V N/A 24? 2018 Atlas V N/A Many rideshare opportunities acknowledged in 2015 and Few confirmed for 2017 or Manifest is in constant flux due to slips, addition or removal of rideshares, and refinement of orbits. 5

6 Building a Rideshare-Initiated Constellation Maximum Revisit Time First approach: launch 1 or 2 CubeSats on every launch in the manifest. Maximum revisit time is highly variable for rideshare-initiated constellations: The randomness of the orbits ensures an occasional bad revisit. 6

7 Building a Rideshare-Initiated Constellation 95 th Percentile Revisit Time First approach: launch 1 or 2 CubeSats on every launch in the manifest. 95 th percentile revisit time is more stable. Lesson: randomness introduces handful of bad apples with high maximum revisit, but attractive performance 95% of the time. 7

8 Comparison against a Reference Constellation Walker Constellations The comparison constellation is a pair of 6-satellite sub-constellations: Walker 6/6/4 (520 km, 24 deg) and Walker 6/6/2 (720 km, 72 deg) The symmetry of Walker constellations provides consistent coverage over time. The 95 th percentile revisit time is little different from maximum revisit time: hence it is not used as a metric with traditional, symmetric constellation designs. 8

9 Walker Compared to a Rideshare Constellation Maximum Revisit Hard to beat the symmetry of a Walker constellation on maximum revisit 9

10 Walker Compared to a Rideshare Constellation 95 th Percentile Revisit 28 CubeSats 28 CubeSats on the 2015 manifest match 12-satellite Walker revisit time at 95 percentile. 10

11 Resilience of Rideshare-Initiated Constellations Impact on Coverage from Losing 1 or 2 Satellites After losing one plane (2 satellites), the rideshare constellation is largely unaffected at 95 th percentile revisit. The symmetry of Walker constellations makes them vulnerable to satellite attrition. The hole in coverage must be mitigated via time-consuming re-phasing of the constellation. 11

12 Judicious Selection of Rideshares Finding Equally Optimal Rideshare Scenarios Flying 1 or 2 satellites on every rideshare is a simple strategy but may involve redundant launches that do not benefit coverage. Rideshare manifest Launch Date Altitude Inc. A DMSP-22 3 May x 830 km 98 deg B DSCS-7 9 Jun x 300 km 35 deg Z TESS 1 Feb x 650 km 80 deg Use an evolutionary algorithm ( GRIPS ) to explore tradespace of rideshare selections. Metrics GRIPS Optimization Example Mission GRIPS finds rideshare selections along Pareto front of: 1) Maximum revisit time, 2) 95 th percentile revisit time, 3) Average revisit time, and 4) Number of CubeSats. # of Sats per Launch Revisit Time Launches: {A,B,E,H,K,Z} 12

13 Rideshare Selection Strategies from GRIPS 95 th Percentile vs. Average Revisit Each point on the plot represents a sequence of rideshares. Beyond ~7 satellites, little gain in average revisit time. Solution #91 Solution #6 2 solutions (#6 & #91) have been singled out for further illustration. 13

14 Solutions from GRIPS: #6 and #91 GRIPS finds combinations of rideshares that provide largely uniform coverage across latitudes. Revisit time can be halved at the cost of going from 5 to 12 CubeSats. 14

15 Influence of Different Rideshares This plot shows the rideshare selections for all 96 of GRIPS optimal solutions. Six rideshares are favored: 4 Sun-synchronous and 2 low-inclination. 15

16 Conclusions Dedicated orbits aren t always necessary. Many coverage metrics can be achieved with the rideshare opportunities already available. Global 95 th percentile revisit times of ~90 min are possible by flying 1 CubeSat on each upcoming rideshare opportunity. Global average revisit times ~10 min flying 2 per rideshare opportunity. Challenging to achieve attractive maximum revisit time with rideshares only: too much randomness. However, 95 th percentile revisit time performs well. 95 th percentile may be adequate for a CubeSat paradigm that is not intended to be 99% reliable or available. Randomness of rideshares introduces some resilience against attrition. Diversity of rideshare opportunities is important: Sun-synchronous orbits are popular, but Other inclinations are needed to keep rideshare a viable option for achieving global coverage. 16

17 Backup 17

18 Coverage of the Aerospace Constellation: Avg Revisit The Aerospace Corporation has 6 CubeSats on orbit, launched between 2012 and 2014, and 6 more will launch in How do rideshare-initiated constellations perform if we take advantage of the entire rideshare manifest? 18

19 Using a Subset of the Manifest US-only launches, ISS Deployments Decay after 1 Year Second approach: launch 1 or 2 CubeSats on a subset of the manifest: US launch vehicles only, and assume ISS deployments decay after 1 year. ISS deployments begin to leave the constellation Note: Many of the Sun-synchronous rideshares are on Russian launch vehicles, making it hard to cover high latitudes with US-only launches. 19

20 Walker Compared to a Rideshare Constellation Maximum Revisit Hard to Beat the Symmetry of a Walker Constellation on Maximum Revisit 20

21 Walker Compared to a Rideshare Constellation 95 th Percentile Revisit 28 CubeSats 20 CubeSats 20 CubeSats 20 CubeSats begin to match 12-satellite Walker revisit time at 95 th percentile. 21

22 Resilience of Rideshare-Initiated Constellations Impact on Coverage from Losing 1 or 2 Satellites Both Walker and the rideshare constellation suffer in maximum revisit when losing 1 or 2 satellites. After losing one plane (2 satellites), the rideshare constellation is largely unaffected at 95 th percentile revisit. The symmetry of Walker constellations makes them vulnerable to satellite attrition. The hole in coverage must be mitigated via time-consuming re-phasing of the constellation. 22