Uncertainty in Power Flow Due to Measurement Errors in Virtual Point Transformation for Frequency-Based Substructuring

Experimental substructuring by means of virtual point transformation (VPT) can be utilized to implicitly account for rotational degrees of freedom (DOFs) at the coupling boundary of a given substructure. Measured frequency response functions (FRFs) are cast onto a virtual node containing three translational and three rotational DOFs via a projection matrix, which is determined by geometric relations between impact and response locations on the structure. In this study, the uncertainty of power flow due to measurement errors in the projection matrix and FRFs is quantified by means of Monte-Carlo simulation. This is performed on a numerical model of two beam structures and is then compared to experimentally obtained data using an impact modal test. Upper and lower bounds on the broadband power flow are created using these data. It is shown that closely space modes lead to high variability in the calculation of power flow near resonance even for small measurement errors. Metrics for analyzing the quality of the virtual point transformation are discussed, as well. This work is beneficial to understanding how experimental errors manifest in the calculation of power flow between coupled structures.

Automated summary

This study investigates the impact of measurement errors in the projection matrix and FRFs on the uncertainty of power flow, using both numerical models and experimental data. The results show that closely spaced modes lead to high variability in the calculation of power flow near resonance, even for small measurement errors.