PSI - Issue 44

Fabio Mazza et al. / Procedia Structural Integrity 44 (2023) 147–154

148

Fabio Mazza et al. / Structural Integrity Procedia 00 (2022) 000–000

2

1. Introduction

Hospital buildings are essential to provide emergency response during seismic events, needing to remain fully operational during and after an earthquake. Recent earthquakes have demonstrated the inadequacy of existing fixed base hospitals, from a structural and especially from a nonstructural point of view (Di Sarno et al. (2013)). Damage to important medical equipment (e.g. incubators, dialysis machines, surgical beds, lamps and surgical machines), hospital services and partition walls can cause serious problems to the operability and functionality of the hospital buildings, also considering overturning of shelves and unanchored cabinets, with the consequent possible leakage of hazardous contaminants. Seismic response of nonstructural elements has been quite studied so far; nevertheless, only recently attention has been focused on hospital services, contents and equipment (Filiatrault and Sullivan (2014)), also considering shaking table tests of base-isolated medical facilities (Sato et al. (2011)). Thus, evaluation of the seismic vulnerability of nonstructural components is of great importance. Considering what presented above, the aim of this work is to reach an integrated numerical assessment in a hospital setting, with particular focus on evaluating and classifying functionality of nonstructural elements and critical medical equipment. To this end, two scaled four-storey (fixed-base) and three-storey (base-isolated) hospital steel framed buildings, subjected to shaking table tests at the University of Kyoto (Japan), are analysed using results of a two-phase experimental campaign with numerical structural and nonstructural blind prediction, considering two earthquakes scaled at different intensity levels. A self-built C++ code is developed to account for lumped plasticity modelling of steel frame members and variability of the friction coefficient of curved surface sliding bearings. In addition, three contest nonstructural components are modelled in the fixed-base structure: i.e. elastic single degree of freedom systems representing two tanks filled with sand (top floor); elastic beam elements for piping (third floor); five-element macro-model for the in-plane-out-of-plane nonlinear mutual interaction of partition walls (first floor). In the end, a self-built MATLAB code is employed with the aim of analysing sliding, rocking and jumping motion of three contest medical equipment (i.e. incubator at third floor; dialysis machine at second floor; surgical bed at first floor), on the basis of acceleration time histories of selected structural nodes of the fixed-base structure. 2. Test structures 2.1. Fixed-base test structure A scaled fixed-base four-storey steel framed hospital building (Figure 1) is built and used for experimental and numerical investigations (Blind Prediction Contest (2020)). In-plan X dimension consists of a single bay of 7 m length, while the in-plan Y dimension is equal to 10 m, divided into two bays of the same length (Figure 1). Floor height is equal to 3.6 m for level 1, and 3.4 m for the other levels, with a total height of 13.8 m. One-way 15 cm thick concrete slabs are placed at all levels, supported by the I-shaped beams depicted in Figures 1a to 1d, whose dimensions are reported in Table 1, together with information about the six steel columns with square hollow cross-sections, where B , H , t w and t f represent flange width, overall height, web and flange thicknesses, respectively. Cross-section of steel and concrete beams is also reported in Table 1.

(a) Level 1.

(b) Level 2.

(c) Level 3.

(d) Level 4.

Figure 1. Plan views of fixed-base test structure (unit in cm).