Leaning Tower of Pisa: Recent Studies on Dynamic Response and Soil-Structure Interaction

Abstract

The Leaning Bell Tower of Pisa has been included in the list of the World Heritage Sites by UNESCO since 1987. Over the last 20 years, the Tower has successfully undergone a number of interventions to reduce its inclination. The Tower has also been equipped with a sensor network for seismic monitoring. In this study, preliminary results on the dynamic behavior of the monument are presented, including a review of historical seismicity in the region, identification of vibrational modes, definition of seismic input, site response analysis, and seismic response accounting for soil-structure interaction. This includes calibration of the dynamic impedances of the foundation to match the measured natural frequencies. The study highlights the importance of soil-structure interaction in the survival of the Tower during a number of strong seismic events since the middle ages.

Full Journal Paper (Published on Earthquake Spectra): https://bit.ly/2XlWe91

Full text

LEANING TOWER OF PISA: RECENT STUDIES ON DYNAMIC
RESPONSE AND SOIL-STRUCTURE INTERACTION
Gabriele FIORENTINO
1
, Davide LAVORATO
2
, Giuseppe QUARANTA
3
, Alessandro
PAGLIAROLI
4
, Giorgia CARLUCCI
5
, George MYLONAKIS
6
, Nunziante SQUEGLIA
7
, Bruno
BRISEGHELLA
8
, Giorgio MONTI
9
, Camillo NUTI
10
ABSTRACT
The Leaning Bell Tower of Pisa has been included in the list of the World Heritage Sites by UNESCO since 1987.
Over the last 20 years, the Tower has successfully undergone a number of interventions to reduce its inclination.
The Tower has also been equipped with a sensor network for seismic monitoring. In this study, preliminary results
on the dynamic behavior of the monument are presented, including a review of historical seismicity in the region,
identification of vibrational modes, definition of seismic input, site response analysis, and seismic response
accounting for soil-structure interaction. This includes calibration of the dynamic impedances of the foundation to
match the measured natural frequencies. The study highlights the importance of soil-structure interaction in the
survival of the Tower during a number of strong seismic events since the middle ages.
Keywords: Dynamic response, Leaning Tower, Soil-structure interaction, Seismic hazard assessment
1. INTRODUCTION
The Bell Tower of Piazza del Duomo in Pisa (Italy) was built during a period of two centuries. Its
construction began in 1173 and was completed in 1370 with the erection of the belfry. The periods of
construction are summarized in Figure 1 (a). At the beginning of construction the Tower started leaning
to the north, reaching a maximum tilt of about 0.5°. The leaning gradually switched to the South reaching
a maximum tilt of about 5.5° in the early 1990s with brought the structure to the verge of collapse..
The total height of the Tower is equal to 58.4 m measured from the foundation. Its cross-section is ring-
shaped, with an external diameter of 19.6 m at the base. The current tilt of the structure is about 5°
towards the South, leading to an offset at the top of about 5 m (Squeglia and Bentivoglio, 2015). The
estimated weight of the Tower is 14500 tons and the elevation of the center of mass is about 23 m from
the base, as indicated in Figure 1 (b).
1
Research Associate, Department of Architecture, Roma Tre University, Rome, Italy,
gabriele.fiorentino@uniroma3.it
2
Assistant Professor, Department of Architecture, Roma Tre University Rome, Italy, davide.lavorato@uniroma3.it
3
Assistant Professor, Department of Structural and Geotechnical Engineering, Sapienza University of Rome, Italy,
giuseppe.quaranta@uniroma1.it
4
Associate Professor, Department of Engineering and Geology, university of Chieti-Pescara, Italy,
alessandro.pagliaroli@unich.it
5
PhD, Department of Science, Roma Tre University, Rome, Italy, giorgia.carlucci@uniroma3.it
6
Professor and Head of Earthquake & Geotechnics Research Group, University of Bristol, U.K.; Professor,
University of Patras, Greece; Adjunct Professor, University of California at Los Angeles (UCLA),
g.mylonakis@bristol.ac.uk
7
Assistant Professor, Department of Engineering, University of Pisa, Italy, squeglia@ing.unipi.it
8
Full Professor and Dean, College of Civil Engineering, Fuzhou University, Fuzhou, China, bruno@fzu.edu.cn
9
Full Professor, Department of Structural and Geotechnical Engineering, Sapienza University of Rome, Italy,
giorgio.monti@uniroma1.it
10
Full Professor, Department of Architecture, Roma Tre University Rome, Italy, camillo.nuti@uniroma3.it

2
The last abrupt increase in inclination of the Tower was due to the digging of the Catino (Italian word
for basin) in 1838 by Alessandro della Gherardesca.
In the last thirty years the Tower became the subject of a series of successful interventions to reverse its
tilting. The inclination was reduced by approximately 0.5° following stabilization works which began
with the installation of 800 tons of lead ingots on the uplifted North side of the monument in 1993.
Figure 1. (a) Side view of the Tower of Pisa with indication of the years of construction and the definition of the
levels (Ordine); (b) cross section of the Tower with indication of the location of the center of mass.
The reduction in inclination was achieved following works of under-excavation and extraction of
approximately 40m2 of soil from the uplifted North side of the Tower between 1999 and 2001. Figure 2
shows the inclined boreholes used to realize these works.
Figure 2 Boreholes for final works of under-excavation
The seismic response of the Tower under earthquake action was first studied by Grandori and Faccioli
(1993), who presented results from dynamic analyses performed on a simplified finite-element (FE)

3
model of the structure in which the seismic input was defined in terms of a response spectrum.
Experimental assessments of the monument began in 1994, with the identification of modal parameters
by ISMES by means of forced vibrations using a vibrodyne.
In light of the severe inclination and the lack of ductility of the Tower, there is a need to study its
dynamic behavior under seismic action. The paper at hand presents an update on recent work carried
out by the co-authors (Fiorentino et al., 2017, 2018).
2. EXPERIMENTAL SEISMIC RESPONSE
The Leaning Tower of Pisa is equipped with a network of accelerometers for seismic monitoring. The
location of sensors is depicted in Figure 3.
Figure 3 Location of accelerometers on the Tower (S1, S2, S3, S4) and the Free Field.
The recorded strong ground motions have been analyzed using different methodologies, including Fast
Fourier Transforms (FFT), Continuous Wavelet Transforms (CWS) and Wavelet Cross Spectra (WCS).
These analyses (see Figure 4) allowed identification of the frequencies of four natural modes, as reported
in Table 1. The first two modes are associated with bending in N-S and E-W direction, respectively, and
have a natural frequency of about 1 Hz. The third is a vertical mode with a frequency of about 3 Hz. It
is worth noting that the only information available in the literature about the vertical mode comes from
Nakamura (1999). Moreover, a frequency of 6.3 Hz was also identified, possibly relating to a torsional
mode.
Table 1 Identified vibration modes and Experimental frequencies
Vibration mode
Experimental frequency [Hz]
1° bending mode (NS)
0.95
2° bending mode (EW)
1
3° vertical mode
3
4° torsional (?) mode
6.3

4
Figure 4 Location of sensors on the Tower (left) and CWT of the response recorded at S3 (right). S3-E = East
West component; S3-N = North-South component;
3. SEISMIC INPUT AND SOIL CHARACTERIZATION
3.1 Historical seismicity
The area around Pisa is characterized by moderate seismicity. The main seismic sources in the region
are located in the area of Pisa hills which are responsible of the 1846 M 6 Orciano Pisano earthquake)
and the Garfagnana area, which released the 1920 M 6.5 Garfagnana earthquake. According to the Italian
Database of historic earthquakes (Rovida et al., 2016), from year 1117 to 2018 eight earthquakes with
intensity levels (IMCS) equal or greater to 6 struck Pisa. This is considered a lower bound for inflicting
structural damage according to the EMS98 damage scale (Grunthal, 1998).
Seven of these earthquakes took place after the completion of the Tower in 1370, therefore the Tower
has withstood a considerable number of earthquakes with IMCS ranging from 6 to 7. Figure 5 and Table
2 summarize estimated IMCS values (≥ 5) from historical earthquakes in the area.
Regarding the damage to the local building stock, an extensive study was carried out by Moroni and
Stucchi (1993) and published as an appendix in the work by Grandori and Faccioli (1993). The historical
sources cited in that study reported moderate to heavy damage to buildings.
The earliest information dates back to 1322 (IMCS = 5-6), when one source reported the fall of a marble
plate depicting a Madonna from the Duomo façade.
A greater amount of information is available for more recent events. With reference to the 1767
earthquake of Versilia, the historical documents mention heavy damage including damage to chimneys,
balcony collapses, many cracks in buildings and some collapses in perimeter walls.
Cracks in the majority of masonry residential buildings and cracks on the church of S. Giovanni in Pisa
were reported during the 1814 Livorno earthquake (IMCS = 6).
The most damaging earthquake which took place during the life of the Tower was the August 14 1846
M 6 earthquake, known as Colline Pisane or Orciano Pisano earthquake. The maximum intensity was
assigned to the village of Orciano Pisano, where the earthquake resulted in the partial or total collapse
of the majority of the buildings (IMCS = 9). An Intensity IMCS = 7 was assigned to Pisa, where extensive
damage was observed on masonry buildings and on some public buildings and churches. No damage
was reported on the Tower of Pisa.

5
Figure 5 Macroseismic intensities in Pisa from year 1100 A.D. to year 2018 as reported in the CPTI15 catalogue
of Italian Earthquakes (adopted from Locati et al. 2016, Rovida et al. 2016).
Damage observed in Pisa during the 1846 earthquake is reported in a historical document put together
by Professor Leopoldo Pilla (1846), a famous geologist of the University of Pisa (and a martyr in the
first war of Italian Independence in 1848). The vault of the church of S. Michele collapsed, there was
damage to a vault in the church of S. Francesco. Cracks were reported on the Clock Tower of Palazzo
Pretorio, in the columns of the peristyle.
In the Piazza del Duomo, one cross of the roof and a marble square stone of the outer wall of the Duomo
fell down. Some light cracks were observed in the Camposanto (Cemetery) and the Battistero
(Baptistry). No damage was observed on the Tower of Pisa (Campanile).
Table 2 Historical earthquakes around Pisa with macroseismic Intensity IMCS ≥ 5.
Meizoseismal area
Date (DD/MM/YYYY)
Intensity
in Pisa
Maximum
intensity
Monti Pisani
03/01/1117
7-8
-
Pisa
10/01/1168
5-6
5-6
Pisa
? / ?/1322
5-6
5-6
Colline Metallifere
07/08/1414
6
7-8
Monti Pisani
06/02/1481
5-6
5-6
Appennino settentrionale
17/08/1536
6-7
6-7
Garfagnana
06/03/1740
5
8
Livorno
27/01/1742
5
6
Lunigiana
21/01/1767
6-7
7
Costa pisano-livornese
03/04/1814
6-7
6-7
Colline Pisane
14/08/1846
7
9
Lucca
27/10/1914
6
7
Mugello
29/06/1919
5
10
Garfagnana
07/09/1920
6-7
10
Appennino settentrionale
25/10/1972
5
5
The 1920 earthquake of Garfagnana was very damaging in the meizoseismal region, but only slight
damage was observed in Pisa.

6
3.2 Seismic input
Seismic hazard assessment was performed by the authors by combining a probabilistic (PSHA) and a
deterministic approach (DSHA). To this end, SP96 (Sabetta and Pugliese, 1996) and AB10 (Akkar and
Bommer, 2010) Ground Motion Predictive Equations (GMPE) were employed. Uniform Hazard Spectra
(UHS) on rock were computed for Return Periods (RPs) of 130 and 500 years. These values are based
on the correlations between MCS intensity and RP, already used by Grandori and Faccioli (1993).
Disaggregation results were employed to identify controlling earthquakes. Based on the Italian seismic
catalogue CPTI15 (Rovida et al., 2016), it was possible to identify two key earthquake scenarios: a M
5.2 event with distance from source of about 20 km associated with a return period (RP) of 130 years
(e.g. Livorno 1742), and a M 5.7 earthquake with a distance from source of about 20 km (e.g. Orciano
Pisano 1846) associated with a return period of 500 years. These are related to MCS intensities VI and
VII, respectively. The target response spectrum for EC8 class B site was evaluated by means of the
Akkar and Bommer GMPE, including the subsoil term associated with Vs,30. Eight accelerograms were
selected for each RP from the European Strong Motion Database (Luzi et al. 2016), considering
5<M<5.5 for 130 years RP, and 5.3<M<6.2 for 500 years RP. The selected components of the horizontal
accelerograms were scaled in such a way so that the average spectrum of each set of accelerograms
approximates well the target spectrum for Soil B. This task has been accomplished using In-Spector
software (Acunzo et al., 2014). The scaling was carried out in the range of the fundamental periods 0.3-
1.1 s in order to take into account the periods of the first two bending modes (about 1 s) and that of the
third (vertical) mode (about 0.3 s) of the structure, thus obtaining the proper scaling factor SF for each
record. To obtain the vertical time histories on Soil B, each original vertical record taken from the
European Strong Motion Database was scaled with the corresponding SF (see Figure 6).
Figure 6. Spectrum-compatible horizontal acceleration time histories for EC8 Soil B with 500 years Return
Period: horizontal (left) and vertical (right) components
3.2 Geophysical tests
A 2D geophysical array (SESAME 2005) was deployed to provide a shear-wave velocity profile of the
soil underlying Piazza del Duomo. The layout of the instruments is displayed in Figure 7 (a). This kind
of test can reach depths of approximately 100 m, which is significantly larger compared to the depths
reached by available Down-Hole and Cross-Hole tests. The measurements were performed using nine
REFTEK130 stations equipped with Lennartz 3D 5s velocimeters and deployed in a triple equilateral
triangle configuration. The central station was located near the Baptistery. The set-up is depicted in
Figure 7 (b) and 7 (c).

7
Figure 7. (a) Array2D test in Piazza del Duomo in Pisa: a) layout of the instruments in the square; (b) Installation
of one the seismic stations; (c) Measurement station formed by REFTEK130 seismic stations and Lennartz 3D
5s velocimeters
The software GEOPSY was employed in the analyses. The inversions revealed the presence of a rigid
layer (VS500m/s) at a depth of about 100m (Figure 6). A single station analysis was performed within
the same test to evaluate H/V spectral ratios, from which a resonance peak at 1.3Hz, associated with an
interface at 40m depth was identified. Another peak was identified at 0.3Hz, which is possibly associated
with an interface at 500m depth or so. The H/V ratio and the associated resonance peaks are shown in
Figure 8.
Figure 8. Comparison between inversions from Array2D test and old DH/CH tests (left graph) and H/V spectral
curve for a single station (right graph)
3.3 Subsoil model and site response analysis
The soil model adopted for the site response analyses is reported in Table 2. It is based on the proposal
by Viggiani and Pepe (2005) and takes into account geotechnical investigations carried from 1970 to
1993. In this model three distinct horizons are identified: A (sandy and clayey silt), B (marine clays)
and C (dense sand), which can be further subdivided into the strata described in Table 3. Thickness and
unit weight for each stratum were estimated according to the data reported in the above study. The

8
assumed shear wave velocity profile is based on the aforementioned seismic array inversions. However,
it should be pointed out that this profile is in good agreement with the VS values measured by other
geophysical tests in the upper strata (e.g. SDMT tests carried out in 2015 up to 40 m and a 65 m deep
cross-hole test in 1999). It should be noticed that a nominal seismic bedrock (VS > 800 m/s) is not
encountered over the explored upper 95 m.
Table 3. Subsoil model adopted for site response analyses over the upper 95 m. LT = Lithotype, H = layer
thickness, γ = unit weight, VS = shear wave velocity, NL = Nonlinear Characterization ([R] = Rollins et al.,
1998; [D] = Darendeli, 2001), SB=Seismic Bedrock
LT
ΔH
(m)
γ
(kN/m3)
VS
(m/s)
NL
LT
ΔH
(m)
γ
(kN/m3)
VS
(m/s)
NL
MG
3.0
18.5
180
average [R]
B7
4.6
18.5
RC tests
A1
5.4
19
DSDSS test S4-
C2 σ'v=65 kPa
B8
1.4
18.5
230
230
RC tests
A2
2.0
18
PI=30 σ'v=55 kPa
[D]
B9
4.0
19
230
RC tests
B1
3.5
17
RC tests
B10
2.6
19.5
RC tests
B2
2.0
17.5
RC tests
C1
27.5
20.5
340
PI=0 σ'v=350 kPa
[D]
B3
4.9
16.5
RC tests
C2
11
20.5
PI=15 σ'v=500 kPa
[D]
B4
1.2
19.5
RC tests
C3
16
20.5
PI=0 σ'v=600 kPa
[D]
B5
3.0
20
230
RC tests
SB
(C3)
-
21
500
-
B6
2.4
19
PI=8 σ'v=200 kPa
[D]
For this reason, considering the uncertainties in the VS values at higher depths, the input motion for site
response analysis was defined according to EC8 class B classification (i.e. instead of rock - class A).
Regarding the nonlinear properties, most of the strata are characterized based on resonant column (RC)
tests (Impavido et al., 1993).
Stratum A1, for which no cyclic data are available, was characterized through DSDSS (Double
Specimen Direct Simple Shear) tests conducted on a soil sample extracted from a depth of 6.3 m (S4-
C2) (D’Elia et al. 2003). The cyclic tests were conducted for different vertical consolidation stresses σ'vc
(65-130-260 kPa); corresponding results are reported in Figure 9 (left graph) in terms of normalized
secant shear modulus and damping ratio as a function of shear strain amplitude.
Given the lack of experimental data, literature curves obtained for similar soils were employed for the
rest of soil strata (Darendeli, 2001; Rollins et al., 1998) (see Table 3). Site response analyses were carried
out using the 1D frequency-domain equivalent linear code STRATA (Kottke and Rathje, 2008). The
results for a return period of 500 years are reported in Figure 9 (right graph) in terms of horizontal
acceleration response spectra computed at ground surface (averaged over all input motions employed).
The average input spectrum at seismic bedrock level is also shown for comparison. Moderate
amplification phenomena take place in the medium-to-long periods with a maximum amplification ratio
slightly higher than 2 at around 1.2 s, where the corresponding average spectral accelerations are about
0.2g. Spectral accelerations as high as 0.5-0.6g (average values) develop at ground surface in the 0.2-
0.4 s period range.

9
Figure 9. G/G0 and D curves obtained through DSDSS test on S4-C2 sample for A1 stratum (left graph); input
and output average response spectra obtained from equivalent-linear site response analyses carried out for 500
years return period (right graph).
4. SOIL-STRUCTURE INTERACTION AND DYNAMIC RESPONSE
A simplified FE stick model, depicted in Figure 10 was built considering the inclination of the Tower
in the N-S direction. For each storey of the Tower, the coordinates of the centroid were defined, based
on the study by Macchi and Ghelfi (2005). For each centroid, 3 translational masses were defined.
Figure 10. Simplified model of the Leaning Tower of Pisa. mi: mass of the ith storey; zi: elevation of the ith
storey; Kx, Ky, Kz = translational base impedances; Krx, Kry, Krz = rotational base impedances (left graph);
reduction in seismic response due to soil-structure interaction for a mode-adjusted spectrum for 500 years return
period (right graph).
Three translational springs and rotational springs were assigned at the base of the model using the tables
by Mylonakis et al. (2006). Table 3 shows the comparison between the results of the modal analysis and
the frequencies obtained experimentally. For a nominal soil shear modulus of G = 77 GPa, the
frequencies obtained by considering the foundation alone and the foundation with the "Catino" are 0.87

10
Hz and 0.88 Hz, respectively, which are about 10% lower than the measured values of 1Hz (Fiorentino
et al 2018).
More satisfactory results were obtained for G = 95 GPa, which lead to a natural frequency of 0.96 Hz in
bending and 3.1 Hz in vertical mode. A further refinement was performed by varying the values of the
foundation springs by ±20%, to obtain improved agreements between the natural frequencies estimated
experimentally and analytically (denoted as “Calibrated” in Table 3).
Table 3. Comparison between numerical and experimental frequencies before and after model updating
Ring Foundation only
Foundation with Catino
Exp. mode
Exp. freq.
(Hz)
G=77GPa
G=95GPa
G=77GPa
G=95GPa
Calibrated
Bending N-S
0.96
0.87
0.96
0.88
0.97
0.95
Bending E-W
1
0.87
0.96
0.89
0.97
1
Vertical
3
2.8
3.1
2.8
3.1
3
Torsional
6.3
4.31
4.7
5.9
6.4
6.3
A response spectrum analysis was performed using the calibrated model, which provided some
preliminary results on the force demand as reported in Table 4. The results are compared to those
obtained for gravitational loading. Evidently, the base moment produced by seismic load with a return
period of 130 years equal to 230 MNm is about 20% lower than the gravitational one. On the other hand,
the moment obtained for a return period of 500 years is considerably larger than the gravitational one.
Table 4. Force demand at the base of the Tower in terms of overturning moment.
Load
Overturning moment [MNm]
Gravitational
280
Seismic input- mean 130 years
230
Seismic input - mean 500 years
420
The importance of soil-structure interaction on the vibrational characteristics of the tower can be
explored based on the so-called wave parameter (1/σ) introduced by Veletsos and co-workers (Veletsos
& Meek 1974; NIST 2012; Maravas et al 2014),


 (1)
H* being the height of an equivalent Single-Degree-Of-Freedom (SDOF) structure (about 23 m based
on the elevation of the centre of mass of the Tower), f its fixed base natural frequency (about 3 Hz) and
Vs the soil shear wave propagation velocity (about 225 m/s). Considering these figures, the wave
parameter is estimated at around 0.3, a remarkably high value that exceeds all available data for building
structures (Stewart et al 1999). The SSI period can be estimated in an approximate manner from the
familiar expression (Veletsos & Meek 1974; NIST 2012; Maravas et al 2014)

  

  (2)
where k is the equivalent stiffness of the superstructure modelled as a SDOF oscillator (which can be
evaluated as k = 4 π2 m f2 = (4 π2 14500x 32 = 5.2 x 106 kN/m) and Kx, Kry are the foundation stiffnesses
in horizontal translation and rocking (about 5 x 106 kN/m and 5 x 108 kNm/rad), respectively.
Substituting these values into the above equation, one obtains 
  which is reasonably close to
the measured value of 1 Hz (the difference being mainly due to the SDOF idealization of the actual

11
infinite-degree-of-freedom system). The period shift due to SSI, (
 ≈ 1/0.3 ≈ 3) is the highest known
for a structure of this height (Stewart et al 1999).
An equivalent analysis can be carried out considering that the natural frequencies in rocking oscillations
of a perfectly rigid superstructure is fry = 1/2π (5 x 108 kNm / 1.1 x 107 Μg m2)0.5 ≈ 1.1Hz, the
corresponding frequency in swaying of a rigid superstructure is fx = 1/2π (5 x 106 kN/m / 14500 Μg)0.5
≈ 3Hz, and the natural frequency of the fixed-base structure is f =1/2π (5.2 x 106 kN/m / 14500 Μg)0.5 ≈
3Hz. Combining the above results using Dunkerley’s rule:

   
 (3)
yields 
 , which in meaningful agreement with the first estimate.
The role of SSI in the seismic response of the Tower can be assessed with the help of the mode-adjusted
spectrum of Figure 10 (right graph), obtained by considering a modal participation coefficient of (2/3).
Evidently, under fixed-base conditions the spectral response is on the order of 0.4g, whereas under
flexible-base conditions it drops below 0.1g – a 400% reduction. Note that this reduction is probably a
lower bound, as it does not account for period elongation due to non-linear soil response, increase in
damping etc. The beneficial effect of SSI on the seismic response of the Tower of Pisa is massive.
5. CONCLUSIONS
The study at hand reports on a numerical and experimental characterization of the seismic behavior of
the leaning Tower of Pisa, and a brief review of historical seismicity. The definition of the seismic input
at bedrock level is based on the results of a new set of geophysical and geotechnical tests performed in
Piazza del Duomo. On the basis of these measurements, a subsoil model was compiled to perform site
response analyses, which provided the free-field input ground motion. A structural Finite-Element (FE)
model was put together based on experimental measurements to explore the earthquake response of the
Tower, including the effect of soil-structure interaction. The FE model was compiled considering the
Tower inclination which can capture overturning moments due to vertical seismic motion. The
foundation impedance matrix was evaluated based on solutions from the literature and was refined to
match the experimental data. The shift in natural period due to SSI, from about 0.3s under fixed-base
conditions to over 1s considering soil compliance (
 ≈ 3) and a corresponding wave parameter (1/σ)
of about 0.3 are the largest known for a structure of this height. The reduction in spectral acceleration
demand due to SSI is on the order of 400%, from 0.4g under fixed-base conditions, to less than 0.1g
considering soil flexibility. This reduction is probably a lower bound, as it does not account for period
elongation due to non-linear soil response and associated increase in damping. Evidently, the beneficial
effect of SSI on the seismic response of the Tower is massive and may explain the lack of earthquake
damage on the structure, despite its severe inclination, low strength and limited ductility. Apart from
SSI, the modest seismicity in the area fundamentally contributed to the survival of the monument.
6. ACKNOWLEDGMENTS
This study was funded by Opera della Primaziale Pisana and coordinated by Camillo Nuti. Special
thanks are due to Professor Luca Sanpaolesi for his comments and suggestions. The authors also thank
INGV-Rome (Dr. Giuliano Milana) for providing the equipment for the 2D geophysical array.
7. REFERENCES
Acunzo, G., Pagliaroli, A., and G. Scasserra (2014) In-Spector: un software di supporto alla selezione di
accelerogrammi naturali spettro-compatibili per analisi geotecniche e strutturali. Proceedings of 33rd conference
of GNGTS, Bologna, Italy (in Italian).
Akkar, S. and J.J. Bommer (2010) Empirical equations for the prediction of PGA, PGV, and spectral accelerations
in Europe, the Mediterranean region, and the Middle East. Seismological Research Letters, 81(2), 195-206.
D’Elia, B., Lanzo, G., and A. Pagliaroli (2003) Small-strain stiffness and damping of soils in a direct simple shear
device. Proceedings of the Pacific Conference on Earthquake Engineering, Christchurch, New Zealand.

12
Darendeli, M.B.,2001. Development of a new family of normalized modulus reduction and material damping
curves. Ph.D. Thesis, Department of Civil Engineering, University of Texas, Austin.
Fiorentino, G., Lavorato, D., Quaranta, G., Pagliaroli, A., Carlucci, G., Nuti, C., Sabetta, F., Della Monica, G.,
Piersanti, M., Lanzo, G., Marano, G.C., Monti, G., Squeglia, N., and R. Bartelletti (2017) Numerical and
experimental analysis of the leaning Tower of Pisa under earthquake, Procedia Engineering, 199, 3350-3355.
Fiorentino, G., Lavorato, D., Quaranta, G., Pagliaroli, A., Carlucci, G., Mylonakis, G., Sabetta, F., Della Monica,
G., Lanzo, G., Aprile, V., Marano, G.C., Briseghella, B., Monti, G., Squeglia, N., Bartelletti, R., Nuti, C. (2018)
Leaning Tower of Pisa: uncertainty reduction for seismic risk assessment through dynamic monitoring, site
response analysis and soil-structure interaction modelling. Earthquake Spectra. (under review).
Fiorentino, G., Nuti, C., Squeglia, N., Lavorato, D. and Stacul, S. (2018). 1-D Nonlinear Seismic Response
Analysis using strength-controlled constitutive models: the Case of the Leaning Tower of Pisa Subsoil,
Geosciences. (under review).
Grandori, G. and E. Faccioli, 1993. Studio per la definizione del terremoto di verifica per le analisi sulla Torre di
Pisa – Relazione finale. Comitato per gli interventi di consolidamento e restauro della Torre di Pisa. (in Italian).
Grünthal, G. (1998). European macroseismic scale 1998. European Seismological Commission (ESC).
ISMES, Modellazione numerica della struttura della Torre – calibrazione dello stick model dell'elevazione e analisi
a spettro di risposta, Comitato per gli interventi di consolidamento e restauro della Torre di Pisa, Incarico Prot. n.
JAM2660.12 tp del 11.03.1992.
Kottke, A.R., Wang, X. and E.M. Rathje (2013). Technical manual for STRATA. Geotechnical Engineering
Center, Dpt of Civil. Architectural and Environmental Engineering, University of Texas, October 2013.
Locati M., Camassi R., Rovida A., Ercolani E., Bernardini F., Castelli V., Caracciolo C.H., Tertulliani A., Rossi
A., Azzaro R., D’Amico S., Conte S., Rocchetti E. (2016). DBMI15, the 2015 version of the Italian Macroseismic
Database. Istituto Nazionale di Geofisica e Vulcanologia. doi:http://doi.org/10.6092/INGV.IT-DBMI15.
Luzi, L., Puglia, R., and E. Russo (2016). Engineering Strong Motion Database, version 1.0. Istituto Nazionale di
Geofisica e Vulcanologia, Observatories & Research Facilities for European Seismology. doi: 10.13127/ESM.
Macchi, G. and S. Ghelfi (2005) Problemi di Consolidamento Strutturale in La Torre di Pisa – Gli studi e gli
interventi che hanno consentito la stabilizzazione della Torre di Pisa. Bollettino d’arte, vol.3 (in Italian).
Maravas, A., Mylonakis, G., and Karabalis, D. (2014). Simplified discrete systems for dynamic analysis of
structures on footings and piles, Soil Dynamics & Earthquake Engineering, 61, 29-39.
Moroni A., Stucchi M., Storia sismica di Pisa, Istituto di Ricerca sul rischio sismico, Milano. Appendix B of
Grandori and Faccioli (1993). (in Italian).
Mylonakis, G., Nikolaou, S., & Gazetas, G. (2006). Footings under seismic loading: Analysis and design issues
with emphasis on bridge foundations. Soil Dynamics and Earthquake Engineering, 26(9), 824-853.
Nakamura, Y., Gurler, E.D. and J. Saita (1999) Dynamic characteristics of leaning tower of Pisa using microtremor
- Preliminary results. Proc. of the 25th JSCE Earthquake Eng. Symp., Tokyo (Japan) 2, 921-924.
NIST GCR 12-917-21, 2012. Soil Structure Interaction for Building Structures. NEHRP Consultants Joint
Venture. Applied Technology Council & the Consortium of Universities for Research in Earthquake Engineering.
Pilla L. (1846) Istoria del tremuoto che ha devastato i paesi della costa toscana il dì 14 agosto 1846, Pisa.
Rollins, K.M., Evans, M.D., Diehl, N.B. and W.D. Daily III (1998) Shear modulus and damping relationships for
gravels. Journal of Geotechnical and Geoenvironmental Engineering, 124(5), 396-405.
Rovida, A, Locati, M, Camassi, R, Lolli, B and P. Gasperini (eds.), 2016. CPTI15, the 2015 version of the
Parametric Catalogue of Italian Earthquakes. Istituto Nazionale di Geofisica e Vulcanologia.
Sabetta F, Pugliese A (1996) Estimation of response spectra and simulation of Non-stationary earthquake ground
motion, Bulletin of the Seismological Society of America, 86(2):337-352.
SESAME European research project WP06 – Deliverable D19.06, Report on FK/SPAC Capabilities and
Limitations University of Potsdam, Germany. Derivation of dispersion curves, 2005.
Squeglia, N., Bentivoglio, G. (2015) Role of monitoring in historical building restoration: The case of Leaning
Tower of Pisa. International J. of Architectural Heritage: Conservation, Analysis, and Restoration, 9(1), 38-47.

13
Stewart, J. P., Seed, R. B., and G.L. Fenves (1999). Seismic soil-structure interaction in buildings. II: Empirical
findings. J. Geotech. Geoenviron, 125(1), 38-48.
Viggiani, C., and M.C. Pepe (2005) Il sottosuolo della Torre in La Torre di Pisa – Gli studi e gli interventi che
hanno consentito la stabilizzazione della Torre di Pisa. Bollettino d’Arte, vol.2 (in Italian).