Extensive flight testing was performed at the university’s flight test facility. These tests showed the difficulty of testing such a small and light airframe due to wind and airframe noise, as well as the limitations caused by lack of feedback received by the remote pilot. The pilot was aided by the flight control system to achieve good trim conditions, and pre-recorded input sequences, similar to the dynamic wind tunnel tests, were used to excite the longitudinal and lateral dynamics of the aircraft.
One particular finding during the test campaign was that there are never totally calm conditions for this scale of airframe. Therefore, the wind speed and direction was estimated with the EKF and the data-sets with the most steady conditions were selected for the system identification. Two significant findings in the longitudinal axis are the high correlation between the pitch damping term and the pitching moment due to elevator, making it im-possible to determine both at the same time, and that in flight the inertial properties of the test aircraft include the added mass compo-nents, while during the dynamic wind tunnel tests the `vacuum’ inertias have to be used.
If both findings are included in the data processing, close agreement between flight and ground test data has been achieved. Laterally, there is still more work required to determine the optimal input sequences and therefore only some preliminary data can be reported. These preliminary results, however, already confirm that the added mass contributions need to be included for the lateral motion as well to obtain a good correlation with the wind tunnel data.