Keywords: structural system identification, model calibration, modal analysis, non-structural appendages, tuned vibration absorbers, modelling errors.

Accurate finite element (FE) models are required in many applications of civil engineering, ranging from health monitoring to structural control, from damage detection to structural evaluation and assessment [1]. Non-structural elements (NSEs) often interfere with the main structure, altering its stiffness and consequently its modal signature. Neglecting such interaction at the modelling stage, although a common design practice, may lead to unreliable predictions of future events as well as to biased interpretations of in-field dynamic tests.

Simplifying, this interaction may alternatively obey two main paradigms, according whether the NSEs work either in parallel or in series with the main structure. In-parallel NSEs (e.g. masonry infills in building structures, pavements or railings in bridges and footbridges) enter the stiffness matrix of the FE model as a mere additive term, modifying frequencies and modeshapes without affecting the model order. In-series NSEs (e.g. non-structural appendages like chimneys, parapets, tanks, ornamentations, but also partitions and facades in their out-of-plane modes) introduce further degrees-of-freedom into the FE representation, augmenting the modal model, and possibly inducing local or global modal-coupling effects if the appendages are in resonance with some structural mode.

Looking at the current literature, the role of NSEs in vibration-based structural system identification (SSI) are well documented for in-parallel NSEs [2,3] but totally unexplored for in-series NSEs.

In this paper, to partially fill in the gap, numerical and experimental case studies are analysed to demonstrate the influence of "resonant" appendages on the dynamics of the main structure and the need to properly account for them when performing the two steps of modal identification and model calibration. At first, the basic principles of local or global modal-coupling between the main structure and resonant elements are illustrated for in-series appendages whose frequency is close to the frequency of some structural mode, and the implications of neglecting such modal-coupling during modelling are numerically exemplified. Then, numerical simulations are produced to show that: (i) if modal-coupling is not correctly identified, the successive steps of the model-updating may be severely invalidated; (ii) if modal-coupling is properly recognized, two successful approaches can be followed, respectively based on omitting or including the appendages into the finite element model of the structure. These findings are finally confirmed by a real case study, namely the experimental SSI of a large-scale laboratory building structure containing in-series NSEs by chance in resonance with a structural mode. Resonant appendages are eventually proven to significantly influence system dynamics, and useful guidelines are provided to successfully pursue SSI in their presence.

1

M.I. Friswell, J.E. Mottershead, "Finite Element Model Updating in Structural Dynamics", Kluwer Academic Press, Dordrecht, the Netherlands, 1995.

2

J.M.W. Browjohn, A.A. Dumanoglu, C.A. Taylor, "Dyanamic investigation of a suspension footbridge", Engineering Structures, 16, 395-406, 1994.

3

J.Y. Kim, E. Yu, D.Y. Kim, S.-D. Kim, "Calibration of analytical models to assess wind-induced acceleration responses of tall buildings in serviceability level", Engineering Structures, 31, 2086-2096, 2009. doi:10.1016/j.engstruct.2009.03.010

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £65 +P&P)