Design and operation of a non-stationkept LEO constellation

Abstract

There are a number of programs either currently in progress, or proposed, which are based on constellations of numerous (greater than 10) spacecraft flying in Low Earth Orbit (LEO). The majority of these programs utilize carefully controlled constellation configurations, which rely on accurate maintenance of the orbit parameters. This orbit control function imposes requirements on several aspects of the spacecraft resources such as design, and the constellation operations concept, all of which increase the cost and complexity of the system. A mission is described herein which does not necessarily require control of a precise constellation configuration, with a resulting potential decrease in spacecraft weight, complexity, and overall system cost. The operation of such a system without stationkeeping is heavily dependent on the actual constellation definition, and requires real time analysis of individual vehicle orbits to enable payload operation. Presented are a summary of the overall system trades and performance analyses of a non-stationkept spacecraft constellation configuration. Preliminary descriptions of the impact of the lack of propulsion and stationkeeping implementation are identified, along with the potential payload impacts from the more classical, deterministic coverage. An analysis is then provided demonstrating the coverage available with a non-stationkept constellation. Of particular importance are the spacecraft initial injection conditions, and the tolerances on these conditions, as the initial conditions define constellation over the life of the spacecraft. Selection of the initial conditions is discussed and analyzed to account for dispersions in injection characteristics and bounded orbital drift characteristics. For the mission as defined, the coverage requirements include average coverage over a 24 hour period; average duration of spacecraft contacts with a ground site; and the average and maximum latency, or V/aiting period' between contacts. The coverage discussion includes all of these parameters, providing a characterization of the system performance. This leads to a discussion of the operation of such a system. INTRODUCTION A majority of the currently considered LEO systems include some level of stationkeeping to maintain spacecraft-to-spacecraft geometry. This is an intuitive philosophy, providing certain obvious advantages at the cost of several spacecraft subsystem and operational impacts. The main benefit of a stationkept constellation is the repeatability of payload coverage, a clearly desirable feature. A defined and maintained spacecraft spacing assures that coverage is available on a deterministic basis. However, the establishment and maintenance of a precision constellation configuration directly impose requirements on the launch vehicle; and the spacecraft propulsion and attitude control Copyright © 1995 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved. subsystems. These have subsequent impacts on the power, thermal, and command and telemetry subsystems. The net effect is the decreased availability of limited spacecraft resources for the payload, the primary purpose of the spacecraft.. STATIONKEEPING IMPACTS Launch vehicle performance directly impacts the requirement to maintain a non-stationkept constellation, the critical issue being the accuracy of the achieved orbit. Launch vehicle injection errors determine the impact of drift on the constellation configuration. Published errors for several launch vehicles into LEO missions are presented in Table 1. Associate Fellow, A.IAA 543 American Institute of Aeronautics and Astronautics LAUNCH VEHICLE Ariane Atlas Delta Pegasus Proton ALTITUDE (a) ERROR 4.0 km (2.2 nm) 6.5 km (3.5 nm) 18.0km (9.7 nm) 8.0 km (4.3 nm) 16.0km (8.6 nm) INCLINATION ERROR 0.06 deg 0.011 deg 0.5 deg 0.01 deg 1.0 deg TABLE 1. TYPICAL LAUNCH VEHICLE ALTITUDE AND INCLINATION ERRORS For a stationkept spacecraft, the weight penalty associated with correcting these magnitudes of altitude errors using propulsion can be less than 2% of the spacecraft mass, depending on the propulsion technology implemented. The wide variation in inclination errors, however, results in a much wider range of potential weight penalties, varying from ~23% of the spacecraft weight for the least efficient correction of a 1° inclination error, to less than 0.5% for the most efficient correction of a 0.1° error. Candidate propulsion systems could vary from cold gas systems (highest fuel weight, low cost) to Electrothermal Hydrazine Thrusters (EHT's) (high fuel efficiency, with high cost). Higher performance thrusters which might be available (such as Arcjets or pulsed plasma thrusters) are not considered here. Assuming that a viable option is to select a propulsion technology commensurate with the mission, the net fuel weight penalty associated with correcting launch vehicle errors can be assumed to be less than 3 4%. While the above described fuel weight penalty is not in itself significant, other implications which must be considered are more severe. First, is the implicit requirement that there is a propulsion system. That is, while the amount of fuel used is modest (as a fraction of the spacecraft weight) the propulsion hardware to support the function can only be reduced to a fixed amount. This could be a significant fraction of a smallsat weight budget. Control of the spacecraft attitude must normally deal with the very low level disturbance torque's associated with residual magnetic dipole, solar and atmospheric pressure, and gravity gradients. This allows for a control methodology with a similar low level of torque. Magnetics are a prime candidate for such a control system. Disturbance torque's from propulsion thrusters are typically several orders of magnitude larger. These require either a significantly different control methodology to maintain pointing (such as additional thrusters to maintain control during the stationkeeping operation); or sufficient time to allow the transients to either decay naturally or be reduced by the operational attitude control system. Note that these comments are equally applicable to any stationkeeping function performed over the life of the mission. The penalty in spacecraft complexity for the stationkeeping function is significant, and could be used to greater advantage. Finally, subsequent to initial constellation deployment, the operation of a spacecraft while being maneuvered from a launch vehicle injection state to the final orbit state must be performed simultaneously with the ongoing task of the full constellation operation. Constellations of multiple spacecraft will require continuous, periodic launches to replenish such a constellation. Inorbit sparing, which alleviates somewhat the time constraint for replacing failed vehicles, may actually increase this launch rate, as the reliability and failure statistics for in-orbit sparing are not yet well established. Further, assuming stationkeeping maneuvers would only be required once per month, there would still be a requirement to perform, on the average, two maneuvers per day. This can represent a significant operational effort. A question which must be asked is does the benefit gained by such propulsion maneuvers justify the resource cost for a service which is compatible with a non-stationkept constellation. PAYLOAD COVERAGE CONSIDERATIONS OF A NON-STATIONKEPT CONSTELLATION The primary goal of any communications system is the service provided. In the LEO spacecraft case, there are two primary types of service; bent pipe, and 'store-and-forward. Real time service requires that either the entire service (transmit and receive nodes) be located within the service area of an individual spacecraft; or that links exist between spacecraft to allow communications to be routed to other satellite service areas. 'Store-andforward' service allows for the payload information to be stored on the spacecraft for some period of time, and subsequently downloaded to the receive terminal at some later time. These distinctively different types of service result in different