20 D.J. Alarcón et al. 0 -100 -80 -60 -40 2 4 Frequency [kHz] Magnitude [0 dB = 1 m/s] 6 8 10 X Y Z Fig. 2.9 Averaged FRFs plot (dB scaling) for this experiment. Note the strong in-plane components (those on the XY plane) of some modes, impossible to determine with other kinds of LDV systems Bypassing these limitations meant finally not taking into account the small defects the propeller blade presents. The solution to these issues is left for discussion as further work. Two datasets have been obtained for this paper, EMA modal parameters (eigenfrequencies, damping ratios and mode shapes) and FEA modal parameters (only eigenfrequencies and mode shapes in this case). An averaged FRF plot is shown on Fig. 2.9 displaying the modes obtained experimentally on this work. This plot serves as a first “sanity check” for the simulation results, and to know the limits of the correlation. The experiment was performed up to 10 kHz, but the complexity of the shapes increases sharply after 5 kHz. More spatial resolution on the DOFs would be needed, together with a stronger energy input in that region to better display the mode shapes beyond 5 kHz. In exchange, testing time would sharply increase, and impact modal testing should be replaced by a more mass-loading excitation method. Consequently, only modal parameters up to 5 kHz are shown in Fig. 2.11. Figures 2.10 and 2.11 show a comparison between EMA and FEA eigenfrequencies and mode shapes. The obtained FEA mode shapes have however a clear correlation with their experimental counterparts. 2.5 Conclusions and Further Work The scope of this paper is proving that establishing a correlation and validation cycle is possible by using the previously described procedures. Modal testing by means of 3D SLDV or 3D geometry scanning have multiple applications separately in several industries. This work serves as a proof, although not a full study case, that these techniques can be also applied and combined on the research and the reverse-engineering of the dynamic properties of composite aerospace structures. Modal testing via 3D SLDV has been proven to be a robust and reliable technique, capable of delivering results that can be simply compared back-to-back with those derived from FEA. Nonetheless, a full validation cycle could not be described in this paper, and remains as the most immediate future work. As it has been widely described, the simulation leg of this cycle has suffered of several drawbacks that have been solved to the best of the authors’ knowledge. However, these drawbacks are of a technical nature—they can be solved with increased training and higher computing power. Further work will be devoted to solve these technical issues while keeping computing times low. Low computing times are a priority, as the described methodology is intended to serve an industry with constantly decreasing delivery times and shorter development cycles. The presented work has been carried out outside a project framework. The authors are open to establish national and/or international research cooperation projects in order to further develop the proposals presented in this paper.
RkJQdWJsaXNoZXIy MTMzNzEzMQ==