noise. To reduce this noise, new FRF were synthesized from modal parameters obtained using modal parameter estimation in LMS Test.Lab. While modal parameter estimation and FRF synthesis removes all of the noise, very slight shifts in antiresonances are produced in the synthesized FRF. These shifts are detrimental to the resulting FRF produce by the decoupling techniques, as shown in Figure 5. The Inverse FBS approach is most sensitive to this perturbation, whereas only minimal error is produced in the Impedance and Mobility results. -30 -20 -10 0 10 20 30 40 50 60 70 dB (m/s2)/N HBcc - Reference HBcc - Mobility/Impedance HBcc - Inverse FBS 0 128 256 384 512 640 768 896 1024 -180 -90 0 90 180 Phase o Hz Figure 5. System AB-TR Estimation of Beam B Translation FRF at Node 6 using Synthesized Perturbed FRFs. Decoupling was performed on model AB-TT in a similar manner. As with AB-TR, all techniques produced accurate resulting FRF when true system and component FRF are used. However, using the perturbed synthesized FRF for model ABTT produces greater error than with model AB-TR, as shown in Figure 6. In this case, Mobility and Impedance contain spurious peaks which occur near antiresonances in the component and system FRF. The Inverse FBS approach produces overall poor results for this case, with minimal correlation. 178
RkJQdWJsaXNoZXIy MTMzNzEzMQ==