Fidelity of reduced-order models for large-scale nonlinear orifice viscous dampers

Large viscous dampers are increasingly incorporated into the designs, whether new or retrofit, of bridge structures and buildings throughout the world. These devices substantially reduce wind and seismic displacements and forces through heat dissipation. The magnitude of these civil structures and the displacement reductions required, as well as forces dissipated, require large damping devices to effectively reduce the demands imparted to the protected structure. Test specifications presently rely on sinusoidal inputs for validation purposes. This paper investigates the appropriateness of varying test stimuli and modeling complexity on parameter identification of large‐scale damping devices. The focus herein is on the specific class of devices identified as orifice fluid viscous dampers. Synthetic data sets are considered first in the analysis, with measured data from full‐scale tests following. It is shown that currently employed damper quantification tests relying on harmonic excitation, and a grossly simplified system representation, are inappropriate from the system‐identification point of view, and that experimental tests utilizing broadband random excitations allow the development of robust mathematical models for design purposes as well as computational efficiency and structural health monitoring applications. Copyright © 2008 John Wiley & Sons, Ltd.