10th World Conference on Seismic Isolation, Energy Dissipation and Active Vibrations

Control of Structures, Istanbul, Turkey, May 27-30, 2007

APPLICATION OF BASE ISOLATION TO A LARGE HOSPITAL IN

NAPLES, ITALY

Di Sarno, L.1, Cosenza, E.2 and Pecce, M.R.1

1
Department of Engineering, University of Sannio, Benevento, Italy
2
Department of Structural Engineering, University of Naples, Federico II, Italy

ABSTRACT

The present paper focuses on the application of base isolation (BI) to a complex multi-storey
irregular structure for hospital building that is under construction in the North-East of Naples,
in South of Italy. The plan layout of the building is about 144x144m and the total height is
about 29m. The assessed structure is thus the largest application of BI in Europe; the
building will be tested through in-situ full scale uplift after the construction of the structure
is completed. The structural system utilized for the superstructure is a reinforced concrete
multi-storey framed system. The building exhibits a large mass eccentricity because of the
different height (3 and 8 storeys, respectively) of the two L-shape blocks of the
superstructure. The fundamental period of vibration of the fixed base frame is relatively high,
i.e. 1.22 sec; the horizontal flexibility of the superstructure apparently reduces the beneficial
effects of BI. Nevertheless, the case study demonstrates that BI is an effective strategy to
improve the seismic performance also for relatively flexible framed structures both at
serviceability and ultimate limit states. The isolation system employed comprises 327
circular-shaped high damping rubber bearings with different diameters (600mm, 650mm and
800mm). The design of the sample BI structure was based on the application of the rules
implemented in the Eurocode 8, which are similar to those included in the recent Italian
seismic code. Architectural and engineering aspects, including both civil and mechanical
works, are also discussed.

1. INTRODUCTION

Base isolation (BI) is an effective strategy to protect the structures of hospital buildings and
more importantly their non-structural components or contents subjected to seismic loading.
The use of BI devices as a means to guarantee the passive protection of critical facilities
under moderate-to-major earthquake ground motions relies upon the following mechanical
properties:
• Horizontal flexibility to increase structural period and reduce the transfer of seismic
energy to the superstructure (except for very soft sites);
 • Energy dissipation (due to relatively high viscous damping) to reduce lateral
displacements;
• Sufficient stiffness at small displacements to provide adequate rigidity for service
level environmental loads, e.g. wind and traffic-induced tremors.
The largest base isolated hospital buildings in the world is the San Bernardino Hospital in
California, which was built in 1993 and employs 400 high damping rubber bearings
(HDRBs) and 186 nonlinear viscous damping devices; the HDRBs have 600mm
displacement capacity. One of the early cost-effective applications of BI to medical centres
date back to the 90’s, when a total of 149 isolators were applied to the base of the lateral
resisting system of the USC University Hospital, in California (Figure 1). This building
experienced significant ground motion with a peak of acceleration of 0.49g during the
January 17, 1994 Northridge earthquake (EERI, 1994). The maximum recorded accelerations
up through the sixth floor ranged from 0.11g to 0.13g and were virtually constant over the
height of the structure. Forces were reduced by 65% across the plane of isolation and there
was absolutely no damage in the hospital. Conversely, the Los Angeles County Hospital
complex located 5/8 miles from the USC Hospital suffered about $400 million in damage.

Figure 1 – University hospital building at USC, California: perspective (left) and horizontal
acceleration profile measured during the 1994 Northridge earthquake (right).

Applications of BI systems for hospital buildings are becoming very common in

earthquake-prone regions in Europe, especially in Italy (Castellano et al., 2007). Cost-benefit
analyses based primarily on construction costs, earthquake insurance premiums, physical
damage that must be repaired and disruption cost, loss of market and potential liability have
demonstrated several advantages in using BI for many new projects, especially in zones with
moderate-to-high seismic risk.
The present paper illustrates the application of BI to a reinforced concrete (RC) multi-storey
irregular frame for hospital building which is under construction in the North-East of Naples,
in South of Italy. The layout of the structure gives rise to a relatively deformable system,
which in turn may erode the effectiveness of the isolation system. The present study shows,
however, that the use of HDRBs led to several benefits both in terms of maximum
deformations experienced by the multi-storey frame and savings for the amount of steel
reinforcement in the beams and columns of the superstructure.
2. BASE ISOLATED HOSPITAL BUILIDNG: CASE STUDY

2.1. Building description

The layout of structure of the “Ospedale del Mare” has a squared plan configuration with
smoothed corners as shown in the aerial photo of the site of construction in Figure 2. The
building is part of a major complex including several other buildings: emergency unit block,
hotel, mall, management building and warehouse. The lateral resisting systems of the latter
buildings are designed utilizing conventional capacity design rules; the hospital is protected
seismically through base isolation.

Figure 2 - Aerial view of the site of construction.

The superstructure of the hospital building includes two blocks with different heights,
namely high- and low-rise blocks as shown pictorially in the longitudinal cross-section in
Figure 3. The former block has eight storeys; the total height is about 29m. The latter is 13m
high and has 3 storeys. The square layout is about 144x144m in plan.

Figure 3 - Longitudinal cross-section of the sample building.

The interstorey for the low-rise block is 3.90m for all but the basement; for the latter the
interstorey height is 3.50m. The second-to-fifth floors in the high rise block have a 3.90m
interstorey. The upper floors are 3.60m high. The basement of the high-rise block is 3.50m.
The hospital is located in a seismic region characterized by a ground acceleration equal to
0.25g as further discussed in the next sections. The lateral resistant structural system utilized
for the building is a reinforced concrete multi-storey frame. The latter exhibits a large mass
eccentricity because of the different height (3 and 8 storeys, respectively) of the two L-shape
blocks of the superstructure. The slabs for all storeys are 40cm thick; they employ special
foam blocks and are partially cast in situ. The grade of concrete is C25/30. The steel for
reinforcement bars is FeB44k, with yield stress equal to 430MPa. The concrete grade used
for the deep foundations is C20/25. A solid RC concrete slab was utilized for the foundation
of the superstructure; the thickness of such slab is 50cm. The foundation of the substructure
consists of a mat with a solid slab with uniform thickness of 120cm; this mat rests on piles.
The piles are drilled; they are 18m long and have a diameter of 0.80m. The isolation system
employed comprises circular-shaped HDRBs with different diameters as further discussed
later.

2.2. Design actions

The structural design of the hospital building was carried out in compliance with the partial
safety factor method. For gravity load, the (force) partial safety factors for the ultimate limit
state γg and γq are assumed equal to 1.4 and 1.5, respectively. For the serviceability limit
state, the above factors are equal to unity. The combination factor ψ0i varies between 0.3 and
0.6 for hospital buildings. Three combinations were utilized to perform the checks for the
serviceability limit state, i.e. rare, frequent and quasi-permanent combinations. The
coefficients ψ1i and ψ2i are equal to 0.6 and 0.3. The above assumptions are compliant with
the provisions included in the current codes of practice (OPCM 3431, 2005; EC8, 2006).
The structural analysis of the multi-storey RC frame used for the hospital building was
carried out by means of modal analysis with response spectrum. The national seismic hazard
map provides the acceleration at the bedrock to perform the dynamic analysis of the RC
multi-storey frame; the peak ground acceleration (at the bedrock) for the site of construction
is equal to 0.25g. Static and dynamic in situ investigations were performed to define
accurately the physical and mechanical properties of the soil of the construction site, as
further discussed in Di Sarno et al. (2006). The soil profile is uniform and sub-horizontal
throughout the building site. It is assumed that the soil type is B, i.e. it consists of deposits of
very dense sand, gravel, or very stiff clay, at least several tens of metres in thickness,
characterized by a gradual increase of mechanical properties with depth. The average shear
wave velocity, denoted as vs,30, ranges between 360 and 800 m/sec and the NSPT is greater
than 50. It may be assumed that the construction site and the nature of the supporting ground
is free from risks of ground rupture, slope instability and permanent settlements caused by
liquefaction or densification in the event of an earthquake. The soil factor S, which accounts
for site amplification effects, is set equal to 1.25. The value of the corner period TD, that
defines the beginning of the constant displacement response range of the spectrum, is equal
to 2.50, because the structure to be assessed is base-isolated (OPCM 3431, 2005). The
spectral ordinates for T>4.0 sec are assumed constant and equal to the value of the ordinate
at T=4.0sec. The topography of the site does not require further amplification factors, i.e.
ST=1.0. The value of the behaviour factor q is assumed equal to 1.5, because the
multiplication factor αu/α1 is equal to 1.3 and the lateral resisting system is multi-storey,
multi-bay RC frames. The importance factor γI is equal to 1.4 because the hospital is a
building whose integrity during earthquakes is of vital importance for civil protection. The
value of the damping correction factor η is equal to 0.71, in fact the rubber devices utilized
for the base isolation systems may provide values of viscous damping ξ equal to 15%.
The damage limitation state (DLS) assumes a seismic action having a larger probability of
occurrence than the design seismic action corresponding to the ultimate limit state (ULS). In
this study, the reduction factor which takes into account the lower return period of the
seismic action associated with the damage limitation requirement is assumed equal to 2.5 in
compliance with the national seismic code (OPCM 3431, 2005). The latter figure
corresponds to the value of the factor ν recommended by the European standards (EC8,
2006) for buildings with importance factor IV, i.e. ν=0.4.
The code displacement spectra derived are displayed in Figure 4 for different values of
viscous damping ξ. It is shown that for periods greater than 2.5 seconds the spectra are
characterized by constant spectral displacements. The target period used for the design of the
base isolated hospital is equal to 2.5 seconds. The maximum lateral displacements are thus
about 20cm. The target values utilized for the period of the base isolated structure is similar
to those suggested in the literature (e.g. Naeim and Kelly, 1999).
50

ξ=5%
Spetral Displacement (cm)

40 ξ=10%

30 ξ=15% ξ=20%

20

10 Constant Displacement

0
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00

Period (sec)

Figure 4 – Code displacement spectra corresponding to different values of equivalent viscous damping.

The combination of the horizontal components (Ex and Ey) of the seismic action were
computed with respect to the rules in the seismic codes (OPCM 3431, 2005; EC8, 2006).
The vertical component of the ground motions was initially not accounted for in the seismic
design of the base isolated structure as it is considered not compulsory by the Italian seismic
standards (OPCM 3431, 2005) for such structures.

2.3. Structural modelling

The sample structure was modelled through a spatial multi-storey framed system employing
elastic frame and shell elements. The finite element (FE) software SAP2000 (CSI, 2005) was
employed to perform the dynamic analyses with response spectrum. A commercial FE
program was utilized to perform the structural checks and plot detailing drawings with
reinforcement bars, stirrups and other details for RC members and joints.
The slabs of the multi-storey frame are modelled as flexible diaphragms because of the large
openings as pictorially shown in Figure 5. In-plane stiff behaviour was assumed for each
sub-assemblage of the slabs. As a result, in the FE model utilized to perform the dynamic
analyses, the seismic masses are lumped at each structural node. Accidental torsional effects
are not accounted for because the centre of mass cannot be defined at each storey level.

Figure 5 – Typical plan layout of the hospital building.

The isolation system was simulated through equivalent elastic springs at the foundation level.
Such springs are characterized by horizontal stiffness khx=khy; the vertical stiffness kv is not
accounted for in the FE model because kv>800kh. Moreover, dashpots are not included in the
model because the effects of the damping is already included in the seismic action through
the use of the correction factor η. Additional checks with nonlinear springs for the HDRBs
were also carried out.
The foundation slab was modelled through shell elements with high in-plan stiffness (rigid
foundation slab). An additional model was utilized for the structural assessment of the
foundation system: the foundation mat is discretized through shell elements, while piles are
modelled with beam elements. The beams rest on elastic horizontal springs simulating the
stiffness of the soil.

2.4. Performance criteria

The structural performance of the sample RC base isolated multi-storey frame was assessed
with respect to local and global strength and deformation response parameters. Such
parameters were used to check the no-collapse and damage-limitation requirements. In so
doing, base and storey shears ad bending moments were computed for the different load
combinations considered. The levels of axial loads in the columns were also utilized to
perform the safety checks of the column cross-sections and design the HDRBs. The
deformation response parameters were monitored in terms of inter-storey drifts (d/h). The
threshold values utilized to check the fulfillment of the damage-limitation states are
d/h=0.005, which is given in (OPCM 3431, 2005) and EC8 (2006) for buildings having
non-structural elements of brittle materials attached to the structure.
2.5. Base isolation system

The base isolation system employed for the sample frame comprises 327 circular-shaped
HDRBs with three different diameters, namely 600mm, 650mm and 800mm. The design of
the BI structure was based on the application of the rules implemented in EC8 (2006) and in
the recent seismic national code (OPCM 3431, 2005). It is assumed to allocate a single
HDRB under each column of the superstructure. The design of the BI system can be found in
Di Sarno et al. (2006). The mechanical properties of the rubber devices used for the sample
structure are summarized in Table 1.

Table 1 - Mechanical properties of the seismic isolation devices used for the sample frame.
Diameter of Number of Shear Viscous Horizontal Vertical
isolators isolators Modulus Damping Stiffness Stiffness
(mm) (MPa) (%) (kN/mm) (kN/mm)
600 115 0.80±0.12 15 1.51 1802
650 124 1.40±0.21 15 2.98 2472
800 88 1.40±0.21 15 4.89 3949

Figure 6 shows the hysteretic behaviour of the three types of HDRBs used for the sample frame; the
response curves plotted in the figure correspond to the third cycle of loading, which is utilized to
compute the mechanical properties of the devices.

2.00

HDRB φ650
1.60
Shear Stress (MPa)

1.20
HDRB φ800

0.80

0.40
HDRB φ600
0.00
0.00 0.20 0.40 0.60 0.80 1.00 1.20

Shear deformation (γ)

Figure 6 – Hysteretic behaviour of high damping rubber bearings (HDRBs) used for the sample frame.

Qualifying cyclic tests were carried out on a sample of HDRBs as prescribed by the national
seismic code (OPCM 3431, 2005). Figure 7 shows the SI-H650/156 during the qualifying
test carried out in the laboratory of the manufactures of the devices. It is noted that the
devices exhibit stable energy dissipation response at large shear deformations under cyclic
loads.
 Figure 7 - Tests on the seismic isolation device with diameter equal to 650mm.

2.6. Seismic performance assessment

The estimated natural period for the fixed base building is equal to 1.22 seconds, while the
base isolated structure has a period of 2.37 seconds; the period lengthening is thus about 2.0.
The modes of vibrations for the system with base isolation are predominantly translational
along the principal directions of the plan layout; the modal masses are about 80% and 95%
for the first and second (translational) modes.
Simplified analyses were initially carried out at both ultimate and damageability limit states
by using the response spectrum in terms of horizontal accelerations to assess whether or not
the BI is beneficial for the sample frame. The spectral acceleration of the BI structure,
computed for T=2.37seconds, was compared to that of the fixed base structure (for T=1.22
seconds) at ultimate limit state. For the former limit state, the behaviour factor q was
assumed equal to 1.5 and the equivalent viscous damping ξ=15% (the correction factor
η=0.71). As a result, the estimated spectral acceleration is 0.078g. On the other hand, for the
fixed base building, it may be assumed that q=4.09, provided that the multi-storey frame is
capacity-designed; the spectral acceleration is 0.078g. It follows that there are apparently no
differences in the dynamic response of the superstructure. Nevertheless, there is a significant
difference in terms of damage. Indeed, the capacity-designed (fixed base structure) possesses
ductile behaviour under moderate-to-intense earthquakes; in turn, structural damage at the
ultimate limit state occurs in the dissipative zones, e.g. beam ends and base columns. For the
damageability limit state, the spectral acceleration computed at the fundamental period of
vibration is equal to 0.094g for the base isolated structure. By considering the fixed base
layout, the spectral acceleration is 0.256g. Therefore, a reduction of about 170% was
estimated for the deformation parameters, namely interstorey drifts of the multi-storey frame.
The maximum interstorey drifts of the base isolated structure on average do not exceed the
50% of the threshold value of d/h=0.5%, as shown in Figure 8.
 0.010 0.010
II III IV V VI VII VIII II III IV V VI VII VIII

0.008 γI = 1.4 Seismic load 1.0 X + 0.3 Y DLS 0.008 γI = 1.4 Seismic load 0.3 X + 1.0 Y DLS
Interstorey (d/h)

Interstorey (d/h)
0.005 0.005
Threshold value
Threshold value

0.003 0.003

0.000 0.000
0 100 200 300 400 0 100 200 300 400
Structural nodes (number) Structural nodes (number)

Figure 8 - Interstorey drifts under seismic loading conditions.

The non-structural components, including building contents and architectural finishing, are
safely protected from the vibrations induced by frequent earthquakes. The maximum
horizontal displacements of the isolators vary between 15.9 cm and 16.7cm. The latter values
were multiplied by the magnification factor γx=1.2 as suggested for building structures by
(OPCM 3431, 2005) and (EC8, 2006) to ensure the increased reliability required for the
isolation devices. The lateral drifts under seismic loads are plotted in Figure 9. It is shown
that the lateral displacements are concentrated chiefly at the base of the structure, where the
HDRBs are located.

15.00 15.00

10.00 10.00
Displacements (cm)

Displacements (cm)

5.00 5.00
Isolators I II III IV V VI VII VIII Isolators I II III IV V VI VII VIII

0.00 0.00
0 100 200 300 400 0 100 200 300 400
-5.00 -5.00

γI = 1.4 Seismic load X - Displacement Ux DLS γI = 1.4 Seismic load Y - Displacement Uy DLS
-10.00 -10.00

-15.00 -15.00

Structural nodes (number) Structural Nodes (number)

Figure 9 - Lateral drifts under seismic loading conditions.

The displacements of the RC frame are caused mainly by lateral translations as shown by the
deformed shape of the foundation block plotted in Figure 10 for seismic loads acting in X-
and Y-direction, respectively. It also noted that the slab of the superstructure is sufficiently
rigid and the profile of the displacements remains plane in the deformed shape.
In terms of resistance, the bending moments in columns in the base isolated system is about
40% lower than for the fixed base frame; similarly for the beams at lower storeys. The
percentage of longitudinal reinforcement bars for the former varies between 1.05% and
1.94% of the gross cross-sections. The reinforcement of the deep beams is about 0.6% of the
cross-sections.

2.7. Construction issues

Some details of the piers which act as supports for the HDRBs at the ground floor are shown
in Figures 11 and 12. In the former figure, the layout of the steel reinforcement (longitudinal
and transverse) of the aforementioned piers is also provided along with the close-up view of
the anchorage bolts of the rubber devices.
1.00 10.00
γI = 1.4 Isolators - DLS Y γI = 1.4 Isolators - SLD Y
X X
9.00
0.50

Displacements (cm)
Displacements (cm)

L=14703 cm L=14703 cm
8.00
7.21 7.27 7.33 7.35 7.34 7.32 7.31 7.29
0.00
7.00 7.30 7.34 7.35 7.33 7.31 7.30 7.23
88

90

92

94

96

98

100

102

104
7.17 7.24
-0.32 -0.30 -0.30 -0.30 -0.29
-0.32 -0.32 -0.31

-0.32 -0.31 -0.30 -0.30 -0.29

-0.33 -0.33 -0.30 -0.31
-0.50 6.00 Seismic load X - Displacement Ux
Seismic load X - Displacement Uy

5.00

88

90

92

94

96

98

100

102

104
-1.00
Structural Nodes (number)
Structural Nodes (number)

10.00 2.00 1.69

1.69
1.69 1.69
1.68

1.68 1.69 1.69 1.69 1.69

1.69 1.65

8.00 L=11180cm
7.11 7.14 7.17 7.21
1.69 1.69

1.00
Displacements (cm)

Displacements (cm)
6.99 7.23 Y
6.88 6.92 6.92 6.92 6.94 6.96 7.04
7.13 7.16 7.18 X
6.00
Y Isolators - DLS
X 0.00
4.00 γI = 1.4

3
6
0
4
8
2
6
0
4
8
2
6
0
6
24
22
21
20
19
19
18
18
17
16
16
15
15
13
2.00 Seismic load Y - Displacement Uy -1.00 L=13093cm
Seismic load Y - Displacement Ux Isolators - DLS
0.00 γI = 1.4
-2.00
1 5 7 11 15 19 23 27 35 39 43 47 51 55 59 63
Structural Nodes (number) Structural Nodes (number)

Figure 10 - Lateral displacement of the foundation block: seismic loads along X-direction (top) and
Y-direction (bottom).

Figure 11 – Details of the pier with the high damping rubber bearing device and anchorage bolts.
Moreover, the RC cross-sections of the caps have sufficient strength to sustain the overstress
due to the concentrated loads that will act on them during the up-lift of the superstructure in
order to replace the rubber devices. The action effects of a 5mm imposed uplift displacement
was used to check the capacity of the RC caps.
The steel reinforcement utilized for the cap of the piers is schematically shown in Figure 13.
The shape of the cap and the steel reinforcement were designed in order to minimize the
stress concentrations.

Figure 12 – View of the ground floor and of the piers with the high damping rubber bearing devices.

3. CONCLUSIONS

The structural design of the new base isolated hospital building under construction in the
suburb of Naples, in Italy, was illustrated in this work. The case study building employs a
reinforced concrete multi-storey frame as earthquake-resistant lateral system. The sample
frame possesses a large mass eccentricity because of the two L-shape blocks of the
superstructure with different heights. The fundamental period of the structure with fixed base
is relatively high compared to that of the base isolation system (1.22 versus 2.37 seconds).
Therefore, it may be argued that the relatively high horizontal flexibility of the
superstructure apparently reduces the beneficial effects of the base isolation system.
Nevertheless, the sample building shows that base isolation is an effective strategy to
enhance the seismic structural performance also for relatively flexible framed systems at
serviceability and ultimate limit states. It is demonstrated that although the base-isolated and
the fixed base construction may undergo the same maximum accelerations the structural and
non-structural damage are prevented in the frame resting on rubber devices. Savings of about
40% were estimated for percentage of steel reinforcement to be used for the columns and
beams of the base isolated frame.
 Figure 13 - Details of the reinforcement bars placed in the column caps.

ACKNOWLEDGEMENTS

This work was financially supported by the Italian Consortium of Laboratories RELUIS,
funded by the Italian Federal Emergency Agency. Any opinions, findings and conclusions or
recommendations expressed in this paper are those of the authors and do not necessarily
reflect those of the Consortium RELUIS.

REFERENCES

Castellano, M.G., Cosenza, E., Di Sarno, L., Guerriero, L., Lupoi, G. and Infanti, S. (2007). Seismic
isolation of hospitals: The examples of Udine, Lisbon and Naples hospitals, Proceedings of
the 10th World Conference on Seismic Isolation, Energy Dissipation and Active Vibration
Control of Structures, Istanbul, Turkey, May 27-30.
Computer and Structures, Inc. (2005). Static and dynamic finite element analysis of structures.
Advanced 9.0.1, Berkeley, California.
Di Sarno, L., De Risi, B. and Mascolo, C. (2006). Application of isolation to a large hospital building,
Proceedings of the First European Conference on Earthquake Engineering and Seismology,
Geneva, Switzerland, 3-8 September 2006, Paper No. 48.
Earthquake Engineering Research Institute (1994). Northridge Earthquake January 17, 1994.
Preliminary Reconnaissance Report. Hall, J.F. Editor, EERI, California.
Eurocode 8 (2006). Design provisions for earthquake resistance of structures. Part 1.3: General rules.
Specific rules for various materials and elements. Eur. Comm. for Standardisation, Brussels,
Belgium.
Kelly, J.M. (1996). Earthquake-resistant design with rubber. 2nd Edition, Springer-Verlag, Inc., New
York.
Naeim, F. and Kelly, J.M. (1999). Design of seismic isolated structures: from theory to practice. John
Wiley & Sons Inc., New York.
Ordinanza Presidente Consiglio dei Ministri (2005). Seismic hazard zonation and seismic design rules
(in Italian).