PSI - Issue 44

Laura Giovanna Guidi et al. / Procedia Structural Integrity 44 (2023) 1284–1291 Laura Giovanna Guidi et al. / Structural Integrity Procedia 00 (2022) 000–000

1285

2

1. Introduction and motivation of the study

Seismic isolation is a design strategy to mitigate forces transferred to the structure during earthquake, decoupling it from the ground motion by using flexible elements at the base (or at the middle story of the buildings). Looking at the evolution of base isolation system (BIS), described by De Luca and Guidi (2019) (2020) and discussed in Design Recommendations for Seismically Isolated Buildings by AIJ (2016), rubber devices are commonly used for this aim since the 1980's (Kelly, 1993) (Kelly and Konstantinidis, 2011), being widely tested also during destructive events. Thanks to their configuration, that combines steel plates and rubber layers mutually vulcanized, they are the most suitable solution to bear gravity loads and contemporary to cover large horizontal deformations. Nowadays, considering the unpredicted spectral values of accelerations and displacements recorded during the latest destructive events worldwide, the IS-strategies are looking towards a Next Generation of Design ( Miyazaki, 2008), to cover even wider horizontal displacements at long vibration periods. At the expected large deformations, stability issue becomes a primary issue to consider for bearing design, as argued by Warn et al. (2007), Warm and Whittaker (2008), Weisman and Warn (2012). Undergoing large horizontal displacements, a significant reduction of bearings’ critical axial-load capacity occurs. In accordance to Sanchez et al. (2013), this theoretical critical load corresponds the value of axial force for which the horizontal stiffness is reduced to zero. In view of the analytical formulation, later argued, this limit value tends to zero when the horizontal displacement is equal to device diameter, i.e. when the overlapping area between the upper and lower base is null. However, some experimental results from literature (Buckle and Liu, 1993) have showed a certain reserve of axial load carrying capacity also for horizontal deformation equal to bearing width: in this case bearing experience also large vertical deformation due to the interaction between vertical load and horizontal offset. Underlining the limits of theoretical formulations, Buckle et al. (2002) justify this capacity considering bearing nonlinearities that occur at high shear strain. Looking at stability issue for rubber devices, in this paper data from an experimental campaign on full-scale HDRBs are discussed. To analyse the interaction between vertical pressure and large shear deformations and to examine possible instability mode for bearings, static shear tests have been considered. As partially described by the authors (De Luca et al., 2022), the test program assumes an increasing γ from 175% to 250%, while a combined vertical stress act (passing from 6MPa to 14MPa). In the following dissertation, experimental results are matched with theoretical formulations and provision from building codes. The elaborations provide also for the use of dimensionless parameters, as in particular the ratio γ /S 2 (= d H / ɸ ), to easily compare data from different bearings. Such type of representation by using dimensionless factors is the prelude to the definition of stability domain. 2. Method Stability is one of the most relevant aspects governing the seismic response of such kind of devices, as studied by authors (Lucibello et al., 2011) (Montuori et al., 2016). This topic concerns both the bearing capacity under long term gravity loads in unreformed configuration, and the stability under short-term vertical stress (due to gravity + seismic loads) at large deformations. In particular, when vertical stress and high shear deformation act, the bearing capacity of elastomeric isolators progressively degrades while horizontal displacement grows. In this paper stability issues for elastomeric bearing is argued through the elaboration of empirical data from an experimental campaign on full-scale high damping rubber bearing, tested for a maximum horizontal displacement of 88% of diameter. Before checking device stability by experimental data, a brief overview form technical literature is proposed, with the aim of highlighting the main factors that can influence the evaluation of the critical load. 2.1. Main theoretical aspects on critical load In seismic condition, when vertical stress, due to gravity plus seismic loads, acts at large lateral deformations, the behaviour of elastomeric devices becomes highly nonlinear (Vemurum et al., 2014): in these conditions, device instability occurs when horizontal shear stiffness decreases and tends to zero. This aspect has been widely discussed in literature, from the earliest studies by Gent (1964) and Derham (1981) that referred to the original Harings’ theory (1949) and predicted the decrease in horizontal stiffness with the increase of axial load. Koh and Kelly (1987) proposed the first mechanical model, analysing a buckling type of instability similarly to that of an ordinary column, dominated by the low-shear stiffness of a bearing. As argued by Naeim and Kelly (1999) and later by Kelly and