Session
Technical Session VIII: Frank J. Redd Student Competition
Abstract
Characterization testing of reaction wheels is necessary for requirement verification and to verify manufacturer specifications. Torque accuracy verification techniques include wheel speed based methods that assume perfectly made wheels or the use of torque transducers, which are expensive and difficult to set up. A low-cost optical torque characterization method is being developed to solve these issues. In the setup the reaction wheels are placed on a frictionless spin table, commanded an output torque, and then a Pixy-Cam optically tracks the angular position of the table. The data is curve-fitted to obtain angular acceleration and, in turn, the torque outputted by the wheels. In all complete trials the acceleration curves has R2 values of >.97 indicating accurate characterization of the torques. This setup benefits from the Pixy-Cam’s built in GUI and ability to interface with Arduino microcontrollers. While these results are promising, further development is required. Improving the nature of the test setup so that the center of mass of the reaction wheels can be easily located, and characterizing the error in the Pixy-Cam, are areas for future improvement. Despite these issues, this method of torque characterization still presents a promising, low-cost method for use in small satellite programs.
A Low-Cost Method for Reaction Wheel Torque Characterization in Small Satellites
Characterization testing of reaction wheels is necessary for requirement verification and to verify manufacturer specifications. Torque accuracy verification techniques include wheel speed based methods that assume perfectly made wheels or the use of torque transducers, which are expensive and difficult to set up. A low-cost optical torque characterization method is being developed to solve these issues. In the setup the reaction wheels are placed on a frictionless spin table, commanded an output torque, and then a Pixy-Cam optically tracks the angular position of the table. The data is curve-fitted to obtain angular acceleration and, in turn, the torque outputted by the wheels. In all complete trials the acceleration curves has R2 values of >.97 indicating accurate characterization of the torques. This setup benefits from the Pixy-Cam’s built in GUI and ability to interface with Arduino microcontrollers. While these results are promising, further development is required. Improving the nature of the test setup so that the center of mass of the reaction wheels can be easily located, and characterizing the error in the Pixy-Cam, are areas for future improvement. Despite these issues, this method of torque characterization still presents a promising, low-cost method for use in small satellite programs.
