Experimental assessment of temperature effects on the performances of a direct-drive servovalve for primary flight actuators

The paper deals with the experimental study of the temperature sensitivity of the performances of fly-by-wire hydraulic actuators. The activity is carried out by performing an extensive test campaign on the direct-drive electrical motor of a modern fly-by-wire actuator in environmental control conditions. The direct-drive technology, essentially based on the use of rare-earth magnet electrical motors for controlling the valve spool position, demonstrated to be a strategic area for the enhancement of the performance and the reliability of modern fly-by-wire flight control systems. Despite of several advantages in terms of system architecture optimisation, fluid contamination resistance, and control design flexibility, the use of direct-drive servovalves arises design criticalities that were negligible with the traditional flapper-nozzle solution, such as the chip shear force requirement, the sensitivity to electrical failures, as well as the temperature sensitivity of electrical motor performances (caused by the variation of motor air gaps for thermal dilatation, and by the variation of permanent magnet properties). A dedicated experimental set-up is arranged, by integrating a thermal chamber with a real-time actuator control system developed in the Matlab-SimulinkxPC Target environment. Both the static and the dynamic performances of the direct-drive motor are concerned, evaluating the basic equipment characteristics such as threshold, motor gain, open-loop and closed-loop frequency responses. The tests are performed at five temperatures: extreme hot (71°C), extreme cold (-54°C), ambient, and two intermediate cold conditions (-20°C and -40°C). Experimental results are reported and discussed, providing a physical interpretation of the temperature sensitivity phenomena. The conclusions are finally substantiated by means of a detailed model of the direct-drive motor dynamics previously developed and experimentally validated by the authors with reference to the ambient temperature. The model is adapted for taking into account the temperature effects, and a satisfactory matching between simulation and experiments is obtained.