High resolution modal analysis using poles obtained at a single location

Abstract

A method, termed the fixed-poles method, is proposed to increase the spatial resolution of experimental modal analysis results so that the resolution of the results is typical of that found in finite element modelling. The fixed-poles procedure estimates the structure’s poles from a high frequency density, driving point frequency response function (FRF) measured with an accelerometer. Mode vectors are estimated from low frequency density, tightly spaced FRFs measured with a scanning laser velocimeter (SLV) using the poles estimated from the accelerometer FRF.

In experiments with a beam, the SLV was used to measure the frequency response of the beam at 31 frequencies and 472 locations. A single, 800 frequency line FRF was measured with an accelerometer, and the locations of the system’s poles were estimated from the accelerometer FRF using the Rational Polynomial algorithm. Mode vectors were then estimated from the 472 SLV FRFs for the first 6 modes (1 rigid body + 5 flexible). A comparison of the results was given between the proposed method and a standard global parameter estimation technique. An investigation of the effect of the locations of the 31 measurement frequencies relative to resonances was also given.

Conclusions were that the proposed method gives comparable results to a global approach in which the system’s poles were calculated only from the SLV measurements. The computational time required for the fixed-poles analysis proved to be significantly less than the global analysis, and the best results were obtained using near resonance SLV measurement frequencies.