A comparison of frequency-domain transfer function model estimator formulations for structural dynamics modelling

The purpose of this paper is to present the results of a comparative study of various formulations of a frequency-domain least-squares (LS) estimator. This study was the basis for the optimization of the LS algorithm for modal parameter estimation. Nowadays, modal analysis studies the dynamical behaviour of very complex structures, which requires both robustness for high model orders and efficient numerical properties to analyse extensive data sets that are obtained during high channel-count modal testing. This paper focuses on the optimal choice between various transfer function models and a weighted formulation of the LS problem by taking also a measure for the noise on the data into account. The evaluation is based on a practical case study, discussing the dynamical modelling of slat tracks of an Airbus A320 commercial aircraft. This safety critical component is characterized by a complex modal behaviour requiring both high model orders and a high spatial resolution in order to achieve an accurate model.