Dissipative unimodal structural damping identification

Usually, in order to model the secondary equipment of French nuclear power plants, the linear behaviour hypothesis is used. This method entails two ma jor processes. The first process is to estimate the conservative part of th e model by an updating step and the second one is to use a 5% uniform dampi ng ratio to account for the dissipative part of the model. This paper prese nts a more accurate procedure for identifying a model of an industrial stru cture subjected to synthesised input acceleration time histories. Firstly, a frequency-domain curve fitting together with autoregressive identificatio n methods, are used to indicate the dependency of the response to the magni tude of the excitation. Afterwards, a hysteretic restoring component of Bou c-Wen type, is guessed to match for the non-linearity detected with the afo rementioned techniques. Finally, using statistical linearisation methods, a n equivalent linear system together with the estimated damping and ductilit y factor is calculated. (C) 2000 Elsevier Science Ltd. All rights reserved.