7 Dynamic Ground Testing: Ground Vibration Tests Through Control Surface Excitation 57 Fig. 7.2 mismatch between DGT-GVT on A340-600 Numerical values for frequency and damping are compared but not displayed in this paper. Only mismatch between GVTDGT results is displayed in the table hereafter. A color code is given to assess extraction quality: green is acceptable, while red is too far from requested result quality (Fig. 7.2). Some of the requested modes could not be identified. For example, the 2NZ wing bending mode could not be properly acquired, because the a/c was on its tires which resulted in a heavy coupling between this mode and the heave rigid body mode. Pitch and yaw at engine location was not captured. An inappropriate excitation is one explanation for this un-identified mode. 7.3 Instrumentation Usually, a GVT is performed by means of hundreds of accelerometers glued on the external skins of the aircraft structure. The installation process can take several days of installation. On top of that, the instrumentation requires a scaffolding to be installed around the aircraft which is really a time consuming task (Fig. 7.3). In order to avoid the sensors and scaffolding installation, the first idea is to take advantage of the FTI instrumentation normally dedicated to the flight domain opening and control law tuning which must be installed and validated for the first flight anyway. On the A350-900, the flutter instrumentation is comprised of approximately 100 single-axis accelerometers dispatched within the structure. A similar number is foreseen for the A350-1000. Nevertheless, there is a drawback. The instrumentation is imprisoned in the airframe and cannot be easily accessed in case of failure. The analog to digital conversion is done locally, making any possible repair sensitive as the associated electronics is similarly distributed throughout the aircraft. On the other hand, the instrumentation is installed several months before the tests and can be validated in advance during the aircraft’s stay in the FAL. This set of accelerometers is capacitive, which fully complies with the GVT requirements for low frequency measurement (Fig. 7.4). The target assigned by the design office is to identify all the modes up to 15 Hz. Using 120 accelerometers instead of 500–800 sensors for a classical GVT, modal validation becomes tricky due to reduced fidelity of the MAC matrix. This risk has been mitigated by means of the A350-900 GVT results. The full modal basis was recalculated while retaining only the GVT sensors which have a corresponding FTI sensor. A study has been performed on modal extraction algorithm robustness: even with very few sensors working on aircraft (studies were carried out with as little as 25 sensors), the frequency and damping of modes can still be extracted. Nevertheless, mode naming is not possible and mathematical models cannot be sorted out from physical ones. More sensors are necessary for a proper modal identification.
RkJQdWJsaXNoZXIy MTMzNzEzMQ==