ding=0 cellspacing=0 width=100%>
Yahoo! is not affiliated with the authors of this page or responsible for its content.
CONTRIBUTION OF THE JRC ISPRA TO THE INTERCOMPARISON OF ANALYSIS METHODS FOR SEISMICALLY ISOLATED NUCLEAR STRUCTURES G. MAGONETTE, V. RENDA European Commission Joint Research Centre Elsa Laboratory, Ispra, Italy
353
CONTRIBUTION OF THE JRC ISPRA TO THE
INTERCOMPARISON OF ANALYSIS METHODS FOR
SEISMICALLY ISOLATED NUCLEAR STRUCTURES

G. MAGONETTE, V. RENDA
European Commission Joint Research Centre Elsa Laboratory,
Ispra, Italy

Abstract
Aim of the work done at JRC has been essentially to investigate the potentiality of the Pseudo-
Dynamic (PsD) method to test structures incorporating anti-seismic protection devices based on
materials with a strain-rate dependent behaviour. This is of relevant importance due to the interest to
perform tests on large-scale mock-ups to assess the behaviour of realistic structure of civil
engineering interest. Two specific typologies of protection have been analysed and tested at the
European Laboratory for Structural Assessment (ELSA) of JRC Ispra. The first dealing with base
isolation and the second with energy dissipation devices. In both cases the protection devices were
based on high damping rubber material which is characterised by a moderate dependence from the
strain rate of the application of the displacements. To validate a standard procedure to test base
isolated structures by the PsD method, a collaboration was set up with the Italian Working Group on
Seismic Isolation which includes the national research centre ENEA, the national electricity board
ENEL, the industrial research centre ISMES and a manufacturer of isolators ALGA. In the framework
of this collaboration it was decided to test at the ELSA laboratory a scaled 5-storey frame structure
(provided by ENEL), isolated by means of high damping rubber bearings (HDRBs), which had been
tested on the shaking table of ISMES. This experimental activity aimed to compare the results which
can be obtained by means of the PsD testing technique with those which can be obtained by means of
a truly-dynamic test on a shaking table. To validate a standard procedure to test structures
incorporating energy dissipation devices, an international collaboration has been set up with
Industries, Research Centres and Universities in the framework of a project partially funded by the
European Commission through the General Directorate for Science and Technology. The obtained
results show once more that the PsD method, when properly applied, may reliably be used to test
structures protected by devices based on high damping rubber. This has been shown effective both in
the case of base isolation and energy dissipation devices by using a specific procedure for the
improvement of the PsD method.

1 INTRODUCTION
The European Laboratory for Structural Assessment (ELSA) of the Joint Research
Centre (JRC) is specially fitted out with up-to-date means for carrying out Pseudo-dynamic
(PsD) tests to reproduce the behaviour of large scale structures subjected to earthquake
loading.
The ELSA laboratory is at present engaged in international consortia to develop,
optimise and test innovative anti-seismic devices based on passive vibration control. To this
end, a collaboration was set up with the Italian Working Group on Seismic Isolation (Gruppo
di Lavoro sullIsolamento Sismico GLIS) which contributed to the present work [1, 2]. In the
framework of this collaboration it was decided to test at the ELSA laboratory a scaled 5-storey
frame structure (provided by ENEL), isolated by means of high damping rubber bearings
(HDRBs), which had been tested on the shaking table of ISMES.
PsD testing is, by virtue of the expanded time scale of the tests with respect to real
seismic events, normally restricted to materials assumed to behave in a rate-independent 354
manner. As regards to seismic isolation by rubber bearings, although the strain rate effect
cannot be taken into account at the experimental stage, it can be taken into account in the
numerical part of the method. A standard procedure for the PsD testing of large-scale models
of base-isolated structures has been developed and validated at the ELSA laboratory.
The experimental activity described in this paper aims to compare the results which can
be obtained by means of the PsD testing technique with those which can be obtained by means
of a truly-dynamic test on a shaking table.
The PsD test procedure includes the following steps:
Characterisation of the isolators for different frequencies to evaluate the stiffening effect
due to the strain rate and the corresponding correction that must be applied to the shear
force.
Comparison between a dynamic snap-back and a PsD snap-back.
PsD tests for seismic inputs and comparison with shaking table results.


2 PRINCIPLES OF SEISMIC ISOLATION
In the last years an important effort has been done to introduce new seismic protection
techniques, some of which are now included in design standards for seismic areas.
Traditional earthquake design methodologies use high strength or high ductility
concepts to mitigate damage from seismic effects. An alternative approach consists in
isolating the structure base from the ground by means of flexible devices, called isolators,
placed between the superstructure and its foundation [3]. A base-isolated system is
characterised by a very low frequency, such that during a strong earthquake, the superstructure
moves like a rigid body over its isolation system. Deformations and energy dissipation are
mostly concentrated in the isolators. A seismic isolator must be rigid in the vertical direction
(to support the dead load of the superstructure), flexible in the horizontal plane (to allow for
large relative displacements between the superstructure and the ground) and possibly, it must
be able to dissipate a significant amount of energy.
From the various devices proposed for seismic isolators, the laminated elastomeric
bearing is emerging as the preferred device for large buildings/structures, such as nuclear
reactors plants. A great number of experimental and numerical studies have already been
performed for all kinds of rubber bearings and several applications to bridges, buildings and
industrial plants already exist in many countries. HDRBs are formed by two end plates and
several relatively thin inner steel plates embedded in a high damping rubber matrix, to which
they are connected through bonding. These isolators can sustain large vertical loads with small
deformations due to confining effect of the inner steel plate and are characterised by a low
horizontal stiffness when subjected to horizontal loads, which allows for large transverse
deformations in severe earthquakes. In these bearings, high damping is obtained by mixing the
rubber with suitable additives (carbon, oils and resins); this allows combining in a single
element both the frequency filtering and energy dissipation capacities necessary to achieve an
effective isolation action. The HDRBs behaviour is mostly characterised by the rubber
mechanical properties, which are highly non-linear, both in terms of stiffness and damping. As
a matter of fact, the width of the experimental hysteresis loop, that determines the amount of
damping, increases with the shear strain. The horizontal force-displacement envelope is
described by an initially high stiffness that decreases to a nearly constant stiffness in the range
354
354
354
354
360 355
between 50% and 150% shear strain; finally, the stiffness increases up to isolator severe
damage, which can occur even over 400% shear strain [4].

3 ISOLATION
DEVICES

3.1 Specimen Rubber Bearings
The HDRBs used in this test campaign were fabricated with a soft compound (G = 0.4
MPa), attached to the structure with bolts and a dowel system, and provided an isolation
frequency (about 1 Hz) in the range of interest for seismic isolation. These isolators have a
diameter of 125 mm and are made of 12 rubber layers with a thickness of 2.5 mm (30 mm of
total rubber height and a shape factor S equal to 12), and 11 steel layers of 1 mm alternating
between the rubber layers. A 10 mm thick steel plate is used at each end of the bearing for
mounting to flange plates of 15 mm which, in turn, are attached to the base and the
superstructure [5]. Consequently, the total height of each isolator was 91 mm. They were
designed for a working shear strain of 100% (30 mm of horizontal displacement) and a
nominal load of 50 kN. Six of these isolators were made available to ELSA for this test
campaign.
3.2 Characterisation tests and Strain Rate Effect Compensation
For elastomeric bearings, a decrease in the testing speed of two or three orders of
magnitude, as is usual for PsD tests [7], may introduce considerable changes in the stress-
strain behaviour, especially for filled rubber bearings [8]. As reported before, these changes
may be described as a proportional force reduction.
Because high damping rubber exhibits some viscous behaviour, it is to be expected that
both the stiffness and damping of the bearings evaluated here will show some dependence on
loading frequency. In general an increase in the rat