You are on page 1of 11

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/237841725

Topographic amplification factors associated to cliff morphology: numerical


results from two case studies in Southern Italy and comparison with EC8
recommendations

Conference Paper · September 2007

CITATIONS READS

3 728

5 authors, including:

Alessandro Pagliaroli Giuseppe Lanzo


Università degli Studi G. d'Annunzio Chieti e Pescara Sapienza University of Rome
169 PUBLICATIONS   1,082 CITATIONS    134 PUBLICATIONS   1,382 CITATIONS   

SEE PROFILE SEE PROFILE

Antonio Costanzo Francesco Silvestri


National Institute of Geophysics and Volcanology University of Naples Federico II
32 PUBLICATIONS   230 CITATIONS    151 PUBLICATIONS   1,370 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

RELUIS 2019-2021 WP16.3 SOIL-STRUCTURE INTERACTION View project

Site effects View project

All content following this page was uploaded by Alessandro Pagliaroli on 04 June 2014.

The user has requested enhancement of the downloaded file.


Topographic amplification factors associated to cliff morphology:
numerical results from two case studies in Southern Italy and comparison
with EC8 recommendations

Alessandro Pagliaroli, Giuseppe Lanzo, Beniamino D’Elia


Dipartimento di Ingegneria Strutturale e Geotecnica, Sapienza Università di Roma, Rome, Italy
Antonio Costanzo, Francesco Silvestri
Dipartimento di Difesa del Suolo, Università della Calabria, Arcavacata di Rende, Cosenza, Italy

ABSTRACT
Seismic amplification of ground motion due to topographic effects at the Nicastro and Gerace cliffs (Calabria,
Southern Italy) is investigated. Linear and non linear 2D numerical analyses are carried out on representative
cross-sections using FLAC finite difference computer code. Additionally 1D analyses accounting only for
stratigraphic effect are also carried out. A topographic amplification factor at the crest is then obtained taking
the ratio of 2D over 1D horizontal peak acceleration and Fourier spectrum. The calculated amplification
factors are compared with those recommended by EC8.

Keywords: topographic amplification, cliff morphology, historical centres, EC8, numerical analyses.

1 INTRODUCTION Conventionally, in the numerical studies, the


effects of the topography can be estimated through
The “stratigraphic” amplification of earthquake the ratio of 2D (accounting for both stratigraphic and
ground motion, related to the variation with depth of topographic amplification) over 1D (depending only
seismic impedance due to subsoil heterogeneity, is on stratigraphic amplification) seismic motion
nowadays routinely modelled through numerical parameters. Usually this ratio is calculated in terms
one-dimensional (1D) analyses. However, the of peak acceleration (Ashford and Sitar, 1997),
seismic amplification of ground motion can be also Fourier (Kallou et al., 2001; Assimaki and Gazetas,
affected by the surface topography, inducing two- 2004) or acceleration response spectra (Bouckovalas
dimensional phenomena such as the focusing of and Kouretzis, 2001). In frequency domain this ratio
seismic waves, the diffraction of body and surface has been defined as the “topographic aggravation
waves and the resonance of the topographic feature. factor” (TAF) (Kallou et al., 2001).
The overall effects related to these phenomena are With few exceptions (e.g. PS92, 1999), the
often summarised as “topographic” amplification. topographic effects are usually neglected in seismic
The occurrence of such effects is mainly reported as codes, because of the inherent difficulties in
macroseismic observations (e.g. concentration of quantifying the amplitude of topographic
structural damage at hilltops), while only few amplification. In particular, Eurocode 8 (2003)
instrumental data are available (Pagliaroli et al., incorporates these effects specifying a frequency-
2007). independent topographic amplification parameter ST
Quantification of topographic amplification is a (Part 5, Annex A), which multiplies as a constant
very difficult issue, as it depends from many factors, scaling factor all the ordinates of the elastic design
including the frequent coupling between response spectrum. The suggested values for ST are
stratigraphic and topographic effects. Also, reported in Fig. 1 for some typical topographic
especially for the experimental studies, the difficulty features (slopes, isolated cliffs and ridges) having an
of finding a reference station free of site effects (i.e. average slope angle β ≥ 15°. Typically, for β < 15°
located on a stiff, unweathered and horizontal the topography effects can be neglected. Regarding
outcropping bedrock) can further complicate the the spatial variation of ST, the highest values apply
estimation of topographic amplification. These near the top of the slopes, while the amplification
above mentioned difficulties probably explain the factor can be assumed to linearly decrease towards
quantitative differences between theoretical the base, where it becomes unity. The suggested
predictions and observed topographic amplifications. amplification factors are increased by at least 20% in
case of an existing superficial zone more than 5 m
thick.
acceleration up to 35% higher than those suggested
Isolated cliffs and slopes by EC8. Further, the Authors remarked that ST
should be considered as frequency-dependent. The
same point of view was shared by Pitilakis (2006)
ST ≥ 1.2 who suggested ST values increasing with frequency.
The estimation of topographic amplification
factors of tall and steep cliffs is especially relevant
β > 15° for the Italian territory, where such morphological
configurations are particularly widespread. The
engineering relevance of this problem is even
greater, considering that several historical towns
Ridges with crest width significantly with precious heritage are often located at the cliff
less than the base width crest (Costanzo et al., 2007a). The aim of this study
is to quantitatively investigate the topographic
amplification of two of such configurations, namely
ST ≥ 1.2
Nicastro and Gerace cliffs located in Southern Italy.
The topographic amplification factors, computed by
15°<β< 30° the numerical analyses, are compared to those
suggested by EC8 and by other numerical and
analytical literature studies.
ST ≥ 1.4
2 THE CASE HISTORY OF NICASTRO
β > 30°
Figure 1. Suggested values for the topographic amplification 2.1 Site description and subsoil model
factor ST in Eurocode 8 (Part 5 Annex A).
Nicastro cliff is located in Calabria (Southern Italy),
In literature, comparisons between topographic one of the most active seismic regions of the
amplification factors by EC8 and those calculated Mediterranean area. The cliff is elongated in the NE-
from numerical and analytical studies can be found, SW direction, about 60 m high; the width at the
among others, in Paolucci (2002) for ridges and crest varies between 30 and 60 m moving from NE
cliffs, Bouckovalas and Papadimitriou (2005) and to SW, while it is about 200 m at the base (Fig. 2).
Klimis and Anastasiadis (2002) for slope- and The ridge flanks are quite irregular, with slope
canyon-like topography, respectively. angles varying generally between 20° and 50°, but
More recently, in the Athens 2006 ETC12 almost vertical in the northern side. Precious
Workshop, six contributions dealing with the issue remains of an ancient Norman-Swabian castle lie on
of topographic amplification were presented (see the the crest.
summary report by Pitilakis and Papadimitriou, The cliff is constituted by metabasitic rocks and
2006). In particular, two of these papers regarded schists characterised by extremely variable degree of
cliff morphology (Pagliaroli et al., 2006, Paolucci weathering and jointing, which strongly affect the
2006). Paolucci (2006), based on the results of 3D mechanical behaviour of the cliff. From 1998 to
numerical analyses carried out on four selected 2003, several geotechnical investigations were
italian sites characterized by different morphological carried out (Fig. 2), including 31 boreholes drilled to
and geologic features, proposed some tentative a maximum depth of about 40 m, 5 seismic
changes for the Annex A of EC8 part 5. In detail, a refraction surveys, and 2 down-hole tests (Pagliaroli,
topographic amplification factor at the crest ST ≥ 1.7 2006). From the sections 1-1 and 2-2 reported in Fig.
was suggested for 3D configurations having height > 2, it can been seen that the cliff is locally covered by
60 m, similar dimension of the width W and length fill or debris, mainly gravel in a silty matrix, with a
L of the crest (1/3< L/W< 3), average and maximum maximum thickness of 5 m. Despite the high degree
slope angle higher than 30° and 60° respectively. of heterogeneity, it is possible to recognise some
It should be pointed out that Paolucci (2006) “macro-zones” from bottom to surface: namely,
adopted a homogeneous model for the numerical moderately weathered and jointed rock, moderately
analyses and defined a frequency-independent weathered and highly jointed rock, highly weathered
topographic amplification factor as the average and jointed rock. The physical and mechanical
value in the period range of interest (1-3 s). material properties of the above materials are
Conversely, Pagliaroli et al. (2006) modelled a real reported in Table 1. The values of shear wave
steep cliff as a 2D heterogeneous medium, finding velocity VS were taken from down-hole tests results
amplification factors in terms of peak ground (Pagliaroli et al., 2006).
2.2 Reference input motion
The historical seismicity of the area was
characterised by setting up a site-specific seismic
catalogue, based on the intensities either felt at
Nicastro, or estimated through attenuation
relationships (Pagliaroli, 2006). The most
destructive event was the earthquake of March 27,
1638 (M=7.1), producing a site intensity IS = XI
MCS. Two reference seismogenic zones were
identified: the Catanzaro strait (near-field condition)
characterised by a minimum source-to-site distance
d = 10 km and maximum historically observed
magnitude MS = 7.1; the Reggio Calabria fault
system (far-field condition), characterised by d =
100 km and MS = 7.3. The reference peak ground
acceleration on rock outcrop, aout, resulted 0.35g and
0.04g, using the attenuation relationships by
Ambraseys and Douglas (2003) and Ambraseys et
al. (1996), for the near-field and far-field conditions
respectively. The corresponding estimated mean
response spectra and their standard deviations are
shown in Fig. 3, together with those relevant to ten
accelerograms (five for near- and five for far-field
conditions), selected from European and world-wide
records on rock outcrops, matching the required MS-
d values and scaled to the expected aout.

mean from attenuation relationship


mean ± standard dev. from attenuation relationship
real accelograms
real accelograms average
1.6
near-field a)
1.4
Spectral acceleration, Sa (g)

1.2
1
0.8
0.6
0.4
0.2
Figure 2. Plan view and sections of the Nicastro cliff.
0
0.25
Table 1: Geotechnical properties of Nicastro subsoil. far-field b)
Spectral acceleration, Sa (g)

0.2
Material γ VS D0
ν
(kN/m3) (m/s) (%) 0.15
cover 21 0.35 350 1
highly weathered 23 0.25 700 1 0.1
and jointed rock
moderately 24 0.25 900 -1000 0.4
weathered, highly 0.05
jointed rock
moderately 26 0.25 1200 0.4 0
weathered and 0.01 0.1 1
jointed rock Period, T (s)
bedrock 27 0.25 1500 0.1 Figure 3. Real accelerograms selected for (a) near- and (b) far-
field conditions.
2.3 Seismic response analyses section 2-2. Overall, such factors are 15% higher
than those calculated for the homogeneous model.
The seismic response of the cliff to both reference
earthquakes was simulated by the finite difference 3.5 1.4

normalized topographic profile


code FLAC (Itasca, 2005). The input motions were 3
homogeneous model
1.3
heterogeneous model
applied to the base of the model as vertically 2.5 1.2

amax/aout
incident SV waves. Due to the high stiffness of the 2
1.1
materials, only linear analyses were carried out. The 1.5
1 1
dynamic response of the cliff was investigated for 0.5 0.9
both sections in Fig. 2, characterised by different 0
topographic profile
0.8
shape ratios H/L (where H is the average cliff height 0 40 80 120 160 200 240
and L is the half-width at the base), equal to 0.55 distance (m)
1A
and 0.8 for sections 1-1 and 2-2 respectively. The 1-1 cross-section

steeper section 2-2 is also characterised by stiffer a)


material with respect to the 1-1 section. A simple
homogeneous model was examined first; thereafter,
a more realistic heterogeneous model was analysed.
In the homogeneous model, the values of γ=26
kN/m3, VS=1200 m/s, ν=0.25 and D0=0.4% were
assumed as representative of the main properties of
the subsoil and the half-space beneath the cliff. The 3.5 1.4

normalized topographic profile


homogeneous model
3
finite difference meshes adopted for the heterogeneous model 1.3
2.5
heterogeneous model are shown in Fig. 4.The related 1.2
amax/aout
2
geotechnical properties are reported in Table 1. 1.5
1.1
The horizontal profiles of the peak surface 1 1
acceleration, amax, normalized to the corresponding 0.5 0.9
topographic profile
value at the outcropping bedrock, aout, are reported 0 0.8
in Fig. 4 a,b for both sections and stratigraphic 0 40 80 120 160 200
models. The profile for the heterogeneous model 2A distance (m)
was computed averaging all the analyses outputs, 2-2 cross-section

without distinguishing between near- and far-field b)


input motions, because minor differences were
observed (Pagliaroli, 2006). The shaded area
represent the average ± one standard deviation
values. For the homogeneous model, the normalised
surface profile was obtained using only one
representative input motion. The topographic
profiles of the sections are also plotted in Fig. 4, by
cover (VS= 350 m/s)
normalising each elevation with respect to that at the
left boundary of the mesh. highly weathered and jointed rock (VS= 700 m/s)
For both sections, the amplification increases moderately weathered, highly jointed rock (VS= 900-1000 m/s)
from the base to the crest of the cliff, with maximum moderately weathered and jointed rock (VS= 1200 m/s)
values generally varying between 1.3-1.6 for the
bedrock (VS= 1500 m/s)
homogeneous model, and increasing to 2-2.5 for the
heterogeneous subsoil. Note the similarity of the Figure 4. Normalised horizontal peak accelerations computed
resulting profiles to the surface topography, with the at the surface of 1-1 (a) and 2-2 (b) cross-sections for the
homogeneous and heterogeneous models.
normalised elevations slightly lower than the
homogeneous amplification factors.
In the frequency domain, the topographic
In order to scale out the magnitude of the
amplification for the heterogeneous case was
stratigraphic amplification from the heterogeneous
estimated by the so-called TAF function, expressed
model results, 1D analyses were carried out for the
in terms of the ratio between 2D over 1D smoothed
soil columns corresponding to two representative
Fourier spectra at the crest nodes 1A and 2A (Fig.
nodes at the crest of the cliff (1A and 2A in Fig. 4).
5). The curves were computed as average values for
In the time domain, the topographic amplification
all the selected input motions, as they did not
factor, ST, was then evaluated as the 2D over 1D
significantly affect the variation of TAF with
ratio of amax. The resulting average values for all the
frequency (Pagliaroli, 2006). For comparison, the
selected near-field and far-field accelerograms were
transfer functions evaluated in the same nodes for
about 1.5 for section 1-1, and 1.9 for the steeper
the homogeneous model are reported. The maximum
peak spectral amplification occurs between 3-5 Hz
for both sections. The amplitude of spectral marls, which in turn rests on a relatively thinner
amplification at crest is about 1.8 and 2.5 for layer of interbedded sands and conglomerates. The
sections 1-1 and 2-2 respectively, thus increasing slab floats on a deep layer varicoloured clay shales,
with H/L ratio, analogously to amplification factor with an estimated thickness of about 200m. The
ST. These amplification coefficients are up to 30% current geomorphological setting is the result of the
higher than those computed on the homogeneous intense erosion processes, producing the progressive
model. In Fig. 5 the fundamental frequency of removal of the soils overlying the clay formation
vibration (f2D) of the cliff, estimated by the Rayleigh (Monteleone, 1993).
method applied to homogeneous triangle-shaped The longitudinal and transversal subsoil models
asymmetric ridges (Paolucci, 2002) and the for the seismic response analyses (see Figs. 7a,b)
fundamental frequency of potential triangular- were characterised combining data collected from
shaped sliding mass (fS) proposed by Bray (2007) previous investigations and experimental results of
are also reported. It can be noted that peak spectral on-purpose dynamic in situ (Down-Hole, MASW
amplification satisfactorily matches the fundamental and SASW) and laboratory tests (RC and TS)
frequency of vibration of the cliff (f2D). Therefore, (Costanzo, 2007). A synthetic picture of physical
topographic amplification can be interpreted as a and mechanical material properties, used in the
two-dimensional resonance phenomenon. analyses, is reported in Table 2.
4
3.5 H f2D = (0.6-1.0) VS/(2L) 3.2 Reference input motion
(Paolucci, 2002)
3 L
VS = cliff average shear wave velocity
fS = 0.38 VS/H In less than sixty days during 1783, since February,
2.5 (Bray, 2007)
1A - homogeneous 5 to March, 28, Southern Calabria was struck by five
TAF

2 1A - heterogeneous strong earthquakes, with magnitudes between 5.9


1.5 and 6.9.
1
0.5 2A - homogeneous
2A - heterogeneous
0
0.1 1 10
Frequency, f (Hz)
Figure 5. Average TAF (heterogeneous model) and transfer
functions (homogeneous model) computed at the crest of
sections 1-1 (node 1A) and 2-2 (node 2A).

3 THE CASE HISTORY OF GERACE

3.1 Site description and subsoil model


Gerace is located at about 480 m a.s.l. on a cliff
oriented in direction NW-SE (Fig. 6), between the
basins of two rivers, characterised by continuous
gully erosion affecting the slope instability of the
Figure 6. Geological map of Gerace (modified after
valley borders. The upper part of the cliff is formed Monteleone, 1993).
by a soft rock slab, constituted by calcarenites and
sandstones, overlying a deep formation of clayey
Table 2. Geotechnical properties of Gerace subsoil.
Material γ VS ν G D0 φ' c' γr γv
(kN/m3) (m/s) (MPa) (%) (°) (kPa) (%) (%)
Fractured calcarenites 16.69 400 0.250 267 0.5 35 7
Linear
Calcarenites 16.69 750 0.250 939 0.5 40 700
Sandstones 16.94 400 0.322 271 0.5 35 54 0.100 0.048
Clayey marls 18.05 800 0.332 1155 1.5 32 34 0.140 0.060
Conglomerates 19.51 750 0.285 1097 0.5 30 1000 0.300 0.160
658 0.400 909
Varicolour clay (cliff) 21.00 1.5 25 0 0.164 0.068
725 1104
454 0.361 430
Varicolour clay (valley) 20.87 3.0 25 0 0.103 0.040
643 863
Bedrock 22.00 1500 0.300 4950 Linear
The sequence caused the destruction of many towns orthogonal sections of the cliff, and EERA (1D), at
(Carbone-Grio, 1887), as well as widespread their crossing point.
environmental damage (Cotecchia et al., 1986), Both heterogeneous and homogeneous modelling
including significant ground movements and maybe was considered, as for Nicastro case study. Linear
slope failure in the historical town of Gerace analyses were carried out on the homogeneous
(Romeo and Delfino, 1997). subsoil model, characterised by a VS of 675 m/s,
From the analysis of the site seismic history computed as the mean value with reference to the
(Costanzo, 2007), the first event of the sequence
thickness of the different formations along the cliff
could be assumed as the maximum historical
earthquake. The reference seismic motion was axis. Conversely, non-linear analyses were carried
simulated through two methods: one based on out on the heterogeneous model.
macroseismic data (estimated epicentre and Fig. 8a,b report the horizontal profiles of surface
magnitude) and the other on seismogenic parameters amplification, for both stratigraphic models, along
(position and geometry of the possible sources). The with the normalized topographic profiles. For the
acceleration and frequency content of the reference heterogeneous models, the coloured dots indicate the
motion were evaluated by means of different stratigraphic contacts between different soils.
attenuation relationships, yielding very close On the average, the amplification factor profiles
evaluations of the motion amplitude. The estimate of seem to reproduce the trends of the corresponding
synthetic ground motion parameters allowed to sort topographic profiles, with oscillations and
out a set of compatible acceleration time histories discontinuities more pronounced for the non linear
(Costanzo, 2007). heterogeneous model.
On the basis of the reference parameters, fixing The surface motion amplitudes along the NW-SE
suitably restricted ranges of magnitude and distance, section (Fig. 8a), characterised by a smoother
compatible seismic records were selected as possible topographic profile, show that the amplification is
input motions from on-line databases. To select the visibly affected by the impedance contrasts, and
most reliable accelerogram, the approach suggested enhances across the more deformable sandstone
by Bommer and Acevedo (2004) was followed, layer. Also, the interaction between incident and
computing the amplitude scale factor, Fsc, and the diffracted waves induces oscillating amplification
root mean square parameter, Drms, that supplies a factors and out-of-phase motion. In the narrower
quantitative evaluation of the similarity between the SW-NE section (Fig. 8b), wave focusing at the
frequency contents. The most suitable seismic hilltop induces a pronounced topographic
motion was the record of the Landers earthquake of amplification; the ST factor, again evaluated as the
1992 (M=7.5, de=21km) at the Morongo Valley - 2D over 1D ratio of horizontal peak acceleration,
Fire Stn 461 (Fig. 7), that showed the better spectral results of the order of 1.6.
compatibility to the reference response spectrum, As for the case of Nicastro cliff, the net
estimated through the relationship by Pugliese and topographic effect at the hilltop was evaluated by the
Sabetta (1989). TAF function, again defined as 2D over 1D Fourier
spectral ratio (Fig. 9); the arrows indicate the mean
1.0
2D fundamental frequency for the homogeneous
spectral acceleration, Sa (g)

0.8
P&S(89) mean models evaluated by means of both Paolucci (2002)
P&S(89) mean ± σ and Bray (2007) relationships. Whatever the section
0.6
LANDERS 1992
considered, in the homogeneous models the TAF
shows a fundamental peak amplification around
0.4
0.8Hz, i.e. falling between the simplified estimated
0.2 values. For the heterogenous models, the
amplification results more pronounced at higher
0.0 frequency values, i.e. in the range between 2 and 5
0 1 2 3 4 Hz, and an apparent tendency to increase beyond 6
period, T (s) Hz is shown. Likely, these effects are due to the
excitation of higher vibration modes, associated to
Figure 7. Landers earthquake of 1992: spectral compatibility seismic impedance contrasts related to the presence
with the reference motion.
of the uppermost layer of fractured calcarenites and
3.3 Seismic response analyses of the intermediate soft sandstones. At the crest of
the longitudinal cliff section (NW-SE), the TAF is
The numerical simulations of the seismic response 2.8 for a frequency of about 5 Hz; at the apex of the
of the cliff to the first event of the sequence, have narrow section (SW-NE), the TAF shows a peak of
been carried out by FLAC (2D), along the two 3.4 for a frequency around 3 Hz.
(a)
3.0 1.8
homogeneous

normalized topographic
2.5 1.6
heterogeneous
2.0 Topographic profile 1.4
amax /aout

profile
1.5 1.2

1.0 1.0

0.5 0.8

0.0 0.6
-1800 -1600 -1400 -1200 -1000 -800 -600 -400 -200 0 200 400 600 800 1000 1200 1400 1600
distance, d (m)

0 0.5 1 km
C1
500 m NW
fractured calcarenites (435ms.l.m.)
calcarenites
sandstones
clayey marls SE
250
conglomerates & sands

varicolour clay shales


0

(b)
3.0 3.5
homogenous

normalized topographic
2.5 heterogeneous 3.0
topographic profile
2.0 2.5
amax/aout

profile
1.5 2.0

1.0 1.5

0.5 1.0

0.0 0.5
-1600 -1400 -1200 -1000 -800 -600 -400 -200 0 200 400 600 800 1000 1200 1400 1600

distance, d (m)
0 0.5 1 km
C2
500m
(435m s.l.m.)

SW NE
250

0
Figure 8. Horizontal peak acceleration profiles along NW-SE (a) and SW-NE (b) sections, for the heterogeneous
and equivalent homogeneous models.

4
SE-NW homogeneous
SE-NW heterogeneous

3
SW-NE homogeneous
SW-NE heterogeneous 4 COMPARISON WITH EC8 PROVISIONS

The topographic amplification factors derived from


Paolucci, 2002
TAF

2
numerical and analytical studies available in
1 literature were compiled and analysed by Pagliaroli
et al. (2006). The gathered data, referred to the crest
0
Bray, 2007
of the topographic features, are displayed in Fig. 10.
0.1 1 10 In particular, the variation of the topographic
frequency, f (Hz) amplification factors versus the shape ratio H/L is
Figure 9. TAF computed at the crest of sections NW-SE and SW- shown in Figg. 10a and 10b in terms of horizontal
NE, for the heterogeneous and equivalent homogeneous models. peak acceleration and peak spectral values,
respectively. It is apparent from the figures that analogously, ST is 35% higher than EC8
spectral amplification data are more limited and recommendations for the SW-NE section of Gerace
scattered (Fig. 10b). Linear regression analyses of cliff. The underestimation provided by the EC8
the data was carried out and the resulting best trends specifications is even greater, if the comparison is
are plotted in Fig. 10 as dashed lines. Fig. 10a also made in terms of spectral values (Fig. 10b).
reports with a solid line the analytical solution by
Sanchez-Sesma (1990) for a triangular shaped ridge
subjected to incident SH waves, which satisfactory 5 CONCLUSIONS
approximates the whole set of data. Note that the
topographic amplification factors ST suggested by Seismic amplification of ground motion due to
EC8, for ridges with crest width significantly less topographic effects was investigated through
than the base width, fall at the lower bound of the numerical analyses carried out on two cliffs located
literature data. in Southern Italy (Gerace and Nicastro). These
effects were estimated via the 2D over 1D ratio of
ground motion parameters, namely amax and Fourier
spectra.
The two cliffs are characterized by different
geometric configurations, with Gerace cliff having
higher dimensions than the Nicastro hill. The
mechanical properties are also quite different:
Nicastro cliff is essentially constituted by rocky
materials with an average shear wave velocity
VS=1200 m/s, while Gerace cliff is formed by a
succession of soft rocks and soils with an average
VS=675 m/s.
These geometric and mechanical differences lead
to different physical phenomena responsible for
topographic site effects. In the case of Nicastro, the
predominant incident wavelengths are comparable
with the base width of the cliff, thus resulting in a
2D resonance with an amplification increasing from
the toe towards the crest. In the case of Gerace, the
incident wavelengths are significantly smaller than
the base width; therefore focusing and interaction
between direct and diffracted waves, shown by
enhanced fluctuations of ground motion surface
profiles, are probably responsible for topographic
effects.
In both sites, the topographic amplification
factors calculated for the heterogeneous models are
higher than those obtained for the homogeneous
analyses, especially for the Nicastro cliff. This
behavior can be attributed to the more complex
wavefield which develops in an heterogeneous
medium. Moreover, this comparison between
Figure 10. Topographic amplification as a function of shape
ratio H/L in terms of (a) horizontal peak acceleration and (b)
homogeneous and heterogeneous subsoils
spectral values. highlighted the capability of TAF function to capture
the main features of the topographic amplification,
For comparison, Fig. 10 also reports the topographic i.e. the frequency range in which it occurs and its
amplification factors obtained in this study at the magnitude.
crest of the cliffs, for the heterogeneous models. In From a quantitative point of view, computed
Fig. 10a these data are located in correspondence of amplification factors are generally higher than those
the average literature trends (and therefore well suggested by EC8, up to 35% in terms of amax. This
above those specified by EC8), with the exception of underestimation is even greater, if the comparison is
the amplification computed on the SW-NE section made in terms of spectral values. This happens
of Gerace cliff. The topographic amplification because the topographic amplification in frequency
factors ST obtained for the Nicastro hill are about domain is band-limited, being concentrated around
25% and 35% higher than the EC8 recommended frequency values depending on the physical
values for section 1-1 and 2-2, respectively; phenomena involved.
Therefore, a more realistic topographic caso di Gerace”, Ph.D. thesis, Università della
amplification factor should be defined as a Calabria (in Italian).
frequency-dependent parameter. However, because Costanzo, A., D’Onofrio, A., Lanzo, G., Pagliaroli,
of the inherent difficulties in generalizing the effects A., Penna, A., Puglia, R., Santucci De Magistris,
of topography due to the several and complex F., Sica, S., Silvestri, F., Tommasi, P. (2007a).
associated physical phenomena, the definition of a “Seismic response of historical centers in Italy:
related spectrum seems premature. In this respect, selected case studies” 4th International
significant improvements might be obtained with
Conference on Earthquake Engineering (ICEGE),
experimental data, that can be extremely helpful to
validate predicted topographic amplification factors. workshop #2 on “Geotechnical Earthquake
Engineering related to monuments and historical
centres”, Thessaloniki (Greece), June 25-28.
6 REFERENCES Costanzo, A., D’Onofrio, A. and Silvestri, F.
(2007b) “Numerical simulations of the ground
Ambraseys, N. N., Simpson, K. A., and Bommer, J. deformation recorded in the historical town of
J. (1996) “Prediction of horizontal response Gerace during the seismic events in Calabria
spectra in Europe” Earthquake Engineering (1783)”, IV International Conference on
Structural Dynamics, Vol. 25, pp. 371-400. Earthquake Geotechnical Engineering,
Ambraseys, N. N., and Douglas, J. (2003) “Near- Thessaloniki, Greece, paper no. 1613.
field horizontal and vertical earthquake ground Costanzo, A., D’Onofrio, A., Silvestri, F. (2007c)
motions” Soil Dyn. and Earth. Eng., Vol. 23, pp. “Analisi dei danni registrati nel borgo di Gerace
1-18. durante gli eventi sismici della Calabria del
Ashford, S. A. and Sitar, N. (1997) “Analysis of 1783”, XII National Conference “L’Ingegneria
Topographic Amplification of Inclined Shear Sismica in Italia”, Pisa, ANIDIS (in italian).
Waves in a Steep Coastal Bluff” Bull. of the Cotecchia, V., Guerricchio, A. and Melidoro, G.
Seismological Society of America, Vol. 87, pp. (1986) “The Geomorphogenetic Crisis Triggered
692-700. by the 1783 Earthquake in Calabria (Southern
Assimaki, D., and, Gazetas, G. (2004) “Soil and Italy)” Symposium: Eng. Geology Problems in
topographic amplification on canyon banks and Seismic Areas, 6, Bari.
the 1999 Athens earthquake” Journal of Earthq. Eurocode 8 (2003) “Design Provisions for arthquake
Eng., Vol. 8, pp. 1-43. Resistance of Structures — Part 5: Foundations,
Bommer J. J. and Acevedo A. B. (2004) “The use of retaining Structures and Geotechnical Aspects”
real earthquake accelerograms as input to ENV 1998-5, CEN European Committee for
dynamic analysis,” Journal of Earthq. Eng., 8, Standardisation, Brussels.
special issue 1:43-91. Itasca Consulting Group (2005) ”Flac 5.0 USER’S
Bouckovalas, G. D., and Kouretzis, G. (2001) Manual,” User’s Guide, Itasca Consulting Group,
“Review of soil and topography effects in the Minneapolis, USA.
September 7, 1999 Athens (Greece) earthquake” Kallou, P.V., Gazetas, G., and Psarropoulos, P.N.
Proc. of 4th Int. Conf. on Recent Advances in (2001) “A case history on soil and topographic
Geot. Earthquake Engineering and Soil effects in the 7th September 1999 Athens
Dynamics, San Diego, California. earthquake” Proc. of 4th Int. Conf. on Recent
Bouckovalas, G. D., and Papadimitriou, A. G. Advances in Geotech. Earth. Eng. and Soil
(2005) “Numerical evaluation of slope Dynamics, San Diego, California.
topography effects on seismic ground motion”, Klimis, N. S. and Anastasiadis, A. J. (2002)
Soil Dynamics and Earthquake Engineering, Vol. “Comparative Evaluation of Topography Effects
25, pp. 547-558. via Code Recommendations and 2-D Numerical
Bray, J.D. (2007) “Simplified seismic slope Analysis” 12th European Conference on
displacement procedures” , IV International Earthquake Engineering, Barbican Centre,
Conference on Earthquake Geotechnical London, UK.
Engineering, Thessaloniki, Greece, Invited Monteleone, S. (1993) “Deformazioni gravitative
Lectures. profonde e grandi frane indotte dalle argille vari-
Carbone Grio, D. (1884) “Terremoti di Calabria e colori scagliose a Gerace e liquefazione nei terre-
Sicilia nel secolo XVIII,” Barbaro editore, Oppi- ni del territorio di Locri,” D.Sc. thesis, Università
do Mamertina (RC) (in Italian). della Calabria (in Italian).
Costanzo, A. (2007) “Analisi di fenomeni deforma- Paolucci, R., (2002) “Amplification of earthquake
tivi di pendii e rilievi in condizioni sismiche: il ground motion by steep topographic
irregularities” Earthquake Engineering and
Structural Dynamics, Vol. 31, pp. 1831-1853.
Paolucci, R. (2006). Numerical investigation of 3D
seismic amplification by real steep topographic
profiles and check of the EC8 topographic
amplification coefficients. Workshop of ETC12
Evaluation Committee for the Application of
EC8, Athens, January 20-21.
Pagliaroli, A. (2006) “Studio numerico e sperimen-
tale dei fenomeni di amplificazione sismica locale
di rilievi isolati” Ph.D. Thesis, Università di Ro-
ma “La Sapienza”, Roma, (in italian).
Pagliaroli, A., Lanzo, G., and D’Elia, B. (2006).
Numerical study of the topography effects at the
Nicastro (Southern Italy) cliff and comparison
with EC8 recommendations. Workshop of ETC12
Evaluation Committee for the Application of
EC8, Athens, January 20-21.
Pagliaroli, A., Pitilakis, K., Chávez-García, F.,
Raptakis, D., Apostolidis, P., Ktenidou, O.-J.,
Manakou, M., Lanzo, G. (2007). Experimental
study of topographic effects using explosions and
microtremors recordings. Proc. 4th International
Conference on Earthquake Engineering (ICEGE),
Thessaloniki (Greece), June 25-28, paper #1573.
Pitilakis, K. and Papadimitriou, A. (2006). Topo-
graphy Effects. Workshop of ETC12 Evaluation
Committee for the Application of EC8, Athens,
January 20-21, General Report.
Pitilakis, K. (2006). Suggestion and comments.
Workshop of ETC12 Evaluation Committee for
the Application of EC8, Athens, January 20-21.
PS-92 (1999). “Règles de construction parasismique:
Règles PS applicables aux bâtiments“, Normes
NF P 06-013, Troisième Tirage.
Pugliese, A. and Sabetta, F. (1989) “Stima di spettri
di risposta da registrazioni di forti terremoti ita-
liani”, Ingegneria Sismica, 6:2 (in Italian).
Romeo, R. and Delfino, L. (1997) “CEDIT: Catalo-
go Nazionale degli effetti deformativi del suolo
indotti da forti terremoti,” Technical report
SSN/RT/97/04, Department of Civil Protection,
Rome (in Italian).
Sanchez-Sesma, F. J. (1990) “Elementary solutions
for response of a wedge-shaped medium to inci-
dent SH and SV waves”, Bulletin of the Seismo-
logical Society of America, Vol. 80, pp. 737-742.

View publication stats

You might also like