Investigating On-Orbit Attitude Determination Anomalies for the Solar Dynamics Observatory Mission

The Solar Dynamics Observatory (SDO) was launched on February 11, 2010 from Cape Canaveral on an Atlas V launch vehicle into a geosynchronous transfer orbit. SDO carries a suite of three scientific instruments, whose observations are intended to promote a more complete understanding of the Sun and its effects on the Earth’s environment. After a successful launch, separation, and initial Sun acquisition, the launch and flight operations teams began a commissioning campaign that included, among other things, checkout and calibration of the fine attitude sensors and checkout of the Kalman filter (KF) and the spacecraft’s inertial pointing and science control modes. During KF and controller checkout, several interesting observations were noticed and investigated. Initial checkout of the fine attitude sensors indicated that all sensors appeared to be functioning properly. Initial calibration maneuvers were planned and executed to update scale factors, drift rate biases, and alignments of the inertial reference units (IRUs). After updating the IRU parameters, an oscillation was noticed in the sensor residuals and the KF estimation of the IRU bias. A concentrated investigation ensued to determine the cause of the oscillations, their effect on mission requirements, and how to mitigate them. Results of that investigation are presented in the paper. While the gyro oscillation investigation was ongoing, another unexpected behavior was noticed in one of the star trackers. The ST2 quality index displayed an unusual shape. Instead of remaining fairly constant around the expected value of 0.95 (in a range of zero to one), there was a period of approximately six hours during which the quality index showed an inverted “U” shape, dropping sharply to as low as 0.6 at the beginning of the period, climbing back towards 0.95 during the middle of the period, and then dropping again, similar to the start of the phenomenon. The poor quality solutions did not adversely affect the on-board spacecraft attitude estimate because the flight software already included logic that rejects any ST solutions whose quality index is below a database threshold. Still, the very distinctive shape of the quality index was curious enough to warrant further investigation, carried out jointly between the Attitude Control Systems team and the Attitude Ground Systems team. The methods of collecting the necessary data for investigation and preliminary results of the investigation are presented in the paper.