The Torques on Hipparcos As Observed Over the Mission

In the new reduction of the Hipparcos data the attitude reconstruction has been developed as a reconstruction of the underlying torques rather than a modelling of the error angles, as was done for the published data (Chapter 10). However, this process could only be started from a preliminary reconstruction of torques as based on the old, published solutions. From the old attitude solutions the underlying torques, as well as the inertia tensor characteristics, are derived from a comparison between rates and rate changes as a function of time, implementing all elements of the Euler equation for a rotating rigid body (Eq. 8.5). The torques as derived in the new solution provide corrections for the inertia tensor, and more direct input for the torque analysis.

In the following analysis, the underlying external torques are assumed to be a continuous function up to its second derivative in time. There are two exceptions to this assumption that we know off: possible abrupt changes in the satellite’s magnetic moment due to switching on or off of electrical equipment (heaters primarily), and eclipses. The generally short pen-umbra phases of eclipses have been rejected from the data for this reason. They showed more complications in modelling than could be supported by the available data.