ESA GNC Conference Papers Repository

A well-known and powerful tool for developing, testing and verifying an attitude control system?(ACS) on the ground is a test bed with an air bearing. It can almost perfectly simulate the low-torque situation of a satellite in space. The use of such an air bearing test stand is related to the test philosophy of a satellite project. It is the trade-off between purely analytical and software examination of attitude determination and attitude control algorithms or the use of hardware-in-the-loop simulation (HILS) and software-in-the-loop simulation (SILS) in conjunction with an air bearing test stand. Experience has shown that the tests on the air bearing test stands tend to reveal errors in the ACS algorithms, that are not necessarily found with pure software verification and helped to avoid such surprises later in space. The physical design of modern air bearing test stands is a combination of several components. There is the air bearing itself, the simulation of the satellite's moments of inertia and the matching of the center of gravity (CoG) with the air bearing’s center. The bearing itself and the deviations of the real center of gravity from the ideal position introduce a residual disturbance torque, but at a level comparable to typical disturbance torques for small satellites in LEO orbits. The test bed can be extended to include geomagnetic field simulation for a given orbit and initial conditions, solar simulation, GNSS simulation, and even star tracker simulation. An external reference system can also be included, that allows precise tracking of the movement of the device under test (DUT). In the last decades, there has been a rapid development of such test beds for different classes of satellites from µ-satellites (moments of inertia up to 20 kg*m²) to nano-satellites and 1 U picosatellites. This paper shortly introduces the testbed and describes the challenges of including the external reference system and the star simulation into the system. Also, the development of the NanoSat testbed is described and test results concerning the disturbance torques are presented. Additionally, autocalibration of the CoG with a model-based approach is described and test results are shown.