Analysis of Sources and Quantification of Uncertainty in Experimental Modal Data

Abstract

The use of uncertainty propagation analysis in conjunction with the finite element method has become more and more popular in recent years due to increasing computer power. Among other things, uncertainty propagation methods are utilized to ensure the robustness of the computational model, or respectively, of the structural design. The determination of fuzzy response data relies heavily on the assumptions made for the variability of the uncertain parameters, mainly because it is impossible to express the scatter of uncertain parameters, like for example joint stiffness parameters, in a mathematical way. Consequently, an inverse approach is required to quantify parameter uncertainties in a finite element model. Computational Model Updating (CMU) techniques offer the possibility to identify model parameters from measured response data. However, statistical test data is required for uncertainty identification by using finite element models and computational model updating techniques. Thus it is necessary to analyze the impact of different possible sources of uncertainty on the test data variability prior to CMU. The paper comprises the results gained from multiple vibration tests performed on a replica of the well known GARTEUR SM-AG19 benchmark structure at the German Aerospace Center (DLR) and subsequent extractions of modal parameters. In particular, the different sources and the corresponding magnitudes of uncertainty are illustrated. This means that special care is taken to uniquely quantify the variability introduced on modal data with respect to different operators, different modal analysis methods, structure dependant scatter, and differences in phase separation testing and phase resonance testing. As a result, a comparison is made to express the variances of the different sources in the experimental modal analysis process which allows for a classification of the error sources (ranking with respect to the relative impact on the variability of the test data). The aim of the identification of the different variances of the miscellaneous sources is to provide a realistic database for subsequent stochastic finite element model updating and to support the selection of meaningful stochastic model updating parameters.