An Approach to Involving Testing and Software Computing into the Predictions of Model Aircraft Engine Performance

Autor: Arno Apffelstaedt

Projekt 2


The module of aircraft design at the University of Limerick is taught by using the annually Design/Build/Fly (D/B/F) Competition of the American Institute of Aeronautics and Astronautics (AIAA). The students are assigned to design an unmanned, electrical powered and radio controlled aircraft which satisfies best the given mission objectives. In several flight tests, it was found out that the thrust generated by the propellers does not seem to match expectations. This report deals with the question how to use testing and software computing to predict the performance of model aircraft engines accurately. An existing motor test rig was redesigned to produce accurate measurements on the power system. Two motors, the Hacker A60-18M and the Plettenberg HP370/50/A3, were chosen to be tested consecutively with two different propellers, the APC 20x10 and the APC 22x12W. The motors were powered with battery packs of identical design and capacity. The force of the propeller pulling forward was measured and recorded. Further, the rotational rate of the propeller, the speed of air exiting the propeller and the motor voltage and current were measured. The time a power system could provide a constant thrust value was calculated and named the ‘usable running time’. The measured performance from static and wind tunnel tests was contrasted to the equipment’s theoretical performance and to the estimates of computer software called Propeller Selector. It was found out that the performance measured was close to propeller and battery discharge theory. The usable running time was found to be a function of the battery discharge and independent from the motor-propeller combination used. It was further found out that the estimates of the Propeller Selector were close to the results of the static tests. A mean down deviation of 5.22 per cent was found. Measurements of the wind tunnel tests were found to be doubtable. The small diameter of the wind tunnel air stream could not provide a uniform airflow over the propeller diameter. As a result, the estimates of the Propeller Selector were far-off the measured performance. The deviation averaged out at 36.24 per cent. A CD containing test data is included in the appendix.