From 1988 to 1993 I was involved in the "Attitude and Orbit Control System" ( AOCS ) of the ISO satellite. Fokker ( FSS ) did the attitude control system as a subcontractor to AéroSpatiale. My job was to run the "control simulations". I implemented the spec for the onboard ADA software in FORTRAN, and ran it with dynamical models of the satellite's orbit, sensors, and actuators.
This was interesting enough in its own right, but at the end of the project I also contributed a technical improvement in the attitude ( pointing ) control law, which came to be known as the "dual control law".
ISO was an ESA scientific ( astronomical ) satellite. The acronym stands for "Infrared Space Observatory". The satellite was basically an infrared telescope mounted in a toroidal, two story high helium tank.
The helium was allowed to boil off through a tiny calibrated orifice, cooling the telescope to near zero Kelvin. There was enough helium on board for a two year mission, after which the satellite would have to be discarded.
The satellite was brought into a highly elliptical, 24 hour orbit. The apogee was at about six Earth radii away from Earth. The perigee was only a few hundred miles high. This orbit was cheaper than a full circular orbit far away from Earth, but it had the disadvantage that Earth filled almost half of ISO's field of view once per day.
The satellite whipped through perigee every 24 hours, taking about half an hour to make a low pass around Earth and be on its way back towards apogee again. The majority of each orbit was spent far away from Earth, allowing the telecope to point in most directions without looking towards Earth, or the sun or moon.
Getting even the rim of the Earth ( "the Earth limb" ) over the edge of the sun shade would boil off the helium very quickly, and getting the sun in view would be catastrophic. One of the sides of the satellite, let's call it its "back", carried the solar panels. This side should always be oriented towards the sun. The edge of the large conical sun shade slanted away from this direction so the satellite could be tilted to avoid the other constraints, without the sun shining inside the cone.
Even so, there were times of the year when the satellite was caught in perigee between the sun standing directly over the Earth, leaving a pointing corridor of one or two degrees between the constraints of the Sun and the Earth, occasionally even involving the Moon.
- sun shade.
- one or two degree corridor between Sun and Earth albedo, and occasionally the Moon.
- 180° slew in half an hour every day during the low perigee.
- arcminute precision pointing.
- poor reaction wheel sensor design.
- no post-operation tests.
- poor requirement setting ( settling time, pointing precision vs. acuity ).