9 Static Calibration of Microelectromechanical Systems (MEMS). . . 93 Fig. 9.2 A picture of the 4.5 m long wind turbine blade showing the positions of the accelerometers, the specially designed rig and anti-vibration pads used to support the test structure ADXL335 MEMS accelerometers were glued to the blade as shown in Fig. 9.2, in order to measure movement within the three defined degrees of freedom (DOFs) of a wind turbine blade (flapwise, edgewise and torsion). This method was used to discretise the motion of a wind turbine blade as it would be impractical to deploy sensors throughout all material points of the blade to measure its motion. Flapwise DOF is the up and down deflection of the blade when in the horizontal position as shown in Fig. 9.2. Edgewise DOF describes the side deflections perpendicular to flapwise DOF. Torsion is assumed to be described by rotation of the chord about the pitch axis [34, 35]. Accelerometers were wired to a 16-channel NI USB-6,251 data acquisition system (DAQ) [36] set to read 5 s of data at a rate of 20 k samples/s using LabVIEW SignalExpress software [37]. MATLAB [38] was used to analyse measured data. 9.3 Static Calibration In static calibration, the output voltage of a uniaxial accelerometer for instance, is a measure of the angle™ [rad] between the sensitive axis of the device and the direction of gravity. The parameters, sensitivity and offset of a uniaxial accelerometer can be obtained by applying two different angles to the device: thus, two equations with two unknowns are obtained. Therefore, a triaxial accelerometer could be calibrated by keeping each axis under two different known angles ™ with respect to gravity, as shown in Fig. 9.3 and summarised in Table 9.1, thus obtaining six equations with six unknowns. This is the minimum required set of equations to determine the three different sensitivities Sx, Sy and Sz [V/g] and offsets Ox, Oy, Oz [V] of the device [39]. For wind turbine blades, it is time-consuming and inefficient to take measurements at all six different positions for each of the accelerometers positioned along the continuously varying blade surface profile. Maximum and minimum static acceleration will occur at different angles of orientation relative to gravity for each of the axes of the five accelerometers on the blade. Therefore, a global coordinate system was generated using the orientation of the wind turbine blade via the specially built and rotatable test rig. The blade was assumed to be fixed along the x-axis of the accelerometer and have no effect on the measured resultant acceleration as the blade is rotated. For this reason, positions ™5 and™6 were ignored. ™1 was set to0ı (when the blade lays flat as in Fig. 9.2), ™2 was set to 90ı (perpendicular blade position to ™1), ™3 to 180ı (opposite blade position to™1) and™4 to270ı(opposite blade position to ™3). The relationship between the normalised accelerometer measurements (Ax, Ay and Az) and the accelerometer raw measurements (Vx, Vy andVz) [40, 41] can be expressed as;
RkJQdWJsaXNoZXIy MTMzNzEzMQ==