Inverse quantification of epistemic uncertainty under scarce data: Bayesian or Interval approach?

This paper introduces a practical comparison of a newly introduced inverse method for the quantification of epistemically uncertain model parameters with the well-established probabilistic framework of Bayesian model updating via Transitional Markov Chain Monte Carlo. The paper gives a concise overview of both techniques, and both methods are applied to the quantification of a set of parameters in the well-known DLR Airmod test structure. Specifically, the case where only a very scarce set of experimentally obtained eigenfrequencies and eigenmodes are available is considered. It is shown that for such scarce data, the interval method provides more objective and robust bounds on the uncertain parameters than the Bayesian method, since no prior definition of the uncertainty is required, albeit at the cost that less information on parameter dependency or relative plausibility of different parameter values is obtained.