Robust energy-based model updating framework for random processes in dynamics: Application to shaking-table experiments

Abstract

This paper presents a robust model updating strategy for correcting finite element models from datasets acquired in low-frequency dynamics. The proposed methodology is based on the minimization of a modified Constitutive Relation Error (mCRE) made of two terms: (i) a Hermitian data-to-model distance written in the frequency domain enriched with (ii) a CRE residual accounting for model bias with strong mechanical content. An automated L-curve based methodology is derived for tuning the relative weight of the two terms and improving the algorithm robustness to noise level. An extended formulation of the mCRE in terms of Power Spectral Density is also proposed: a data windowing preprocessing step ensures statistical consistency of the updated parameters when dealing with noisy random processes. The methodology is applied to two earthquake engineering examples. The performances of the methodology are assessed using synthetic measurements from a plane frame subjected to random ground acceleration. Actual measurements from the SMART2013 database are next processed to observe the eigenfrequency drop of a reinforced-concrete structure submitted to a sequence of gradually damaging shaking-table tests. In this last application, the corrected model predictions are in good correlation with former data-driven subspace-based identification results.