“ SITE EFFECTS EVALUATION IN CATANIA ( ITALY ) BY MEANS OF 1-DNUMERICAL ANALYSIS „

A probabilistic seismic hazard analysis (PSHA) determines the probability rate of exceeding of various levels of ground motion in a specified period of time, in a given area. On the other hand, the neo-deterministic seismic hazard assessment (NDSHA) is based on modeling techniques, developed from physical knowledge of the seismic source process and of the propagation of seismic waves, which can realistically simulate the ground motion due to an earthquake by means of synthetic seismograms. The NDSHA confirms that peak ground acceleration values are larger than those given by the PSHA in areas where large earthquakes are observed and in areas identified as prone to large earthquakes, such as in the case of the city of Catania (Italy). The city of Catania, located on the eastern part of Sicily, is one of the most seismically active areas of Italy. The Val di Noto earthquake of January 11, 1693 is considered one of the biggest earthquakes which occurred in Italy. Site response analyses have been developed for a group of 3 test sites in the city of Catania (Italy). On the sites are localised some strategic buildings to upgrade against seismic risk. The analysis of seismic ground response at the sites was conducted using a numerical method, which is developed in three main phases: the definition of the geometric, geological and geotechnical model of the subsurface, the definition of the seismic input (synthetic or recorded), the choice of one or more computer codes to use and process the results. The reconstruction of the geological and geotechnical model of the subsurface has highlighted a morphology quite irregular especially with regard to the covers. One-dimensional local site response analyses have been performed assuming that all geologic boundaries are horizontal and the response of soil deposits is predominantly caused by waves propagating vertically from the underlying bedrock. Equivalent linear analysis using EERA code has been performed in this study. One of the targets of the paper is the development of measures for the seismic retrofitting of buildings. Seismic retrofitting and/or improving have to be definitely considered a multidisciplinary subject, which depends in fact on many factors, such as: local site effects and the dynamic interaction between the foundation soil and the structure. The accurate investigation on the structure and the surrounding soil is the first fundamental step for a realistic evaluation of the structure seismic performance. of the shear modulus G and damping ratio D with depth. Seismic Dilatometer Marchetti Tests (SDMT) have been also carried out in the area of the “National Institute of Geophysics and Volcanology” building (INGV) and in the “Madre Teresa di Calcutta” building school (CD MTC), with the aim of an accurate geotechnical characterisation, including the evaluation of the shear wave velocity Vs profile, as well as the profile of the horizontal stress index Kd. Shear wave velocity has been measured by different tests. The soil profiles in terms of the shear modulus G0 and in terms of the shear wave velocity Vs have been evaluated and compared by different in situ tests. The redundancy of measurements is very useful, for instance, for site response and liquefaction analyses, which can be based either on Vs values or Kd values. Boreholes were driven and undisturbed samples were retrieved for laboratory tests. Because of their relevance on the estimation of local ground shaking and site effects, data from in-hole geophysical surveys (Down-Hole and Seismic Dilatometer Marchetti Test) have been examined with special attention, particularly for S wave velocity measurements. It must be noted that the shear wave velocity Vs was evaluated on the basis of both empirical correlations with in situ or laboratory tests and a few direct measurements. One-dimensional local site response analyses have been performed assuming that all geologic boundaries are horizontal and the response of soil deposits is predominantly caused by waves propagating vertically from the underlying bedrock. Among the programs that adopt the equivalent linear analysis EERA code [Bardet et al., 2000] has been used in this study. The three test sites in Catania are two school buildings in the province of Catania, the Catania division of the National Institute of Geophysics and Volcanology (indicated by the INGV acronym) and finally the fourth test site is the Messina’s Regional Department of Civil Defence (indicated by the DRPC acronym). In this paper, the results of local seismic response analysis at the INGV site in Catania and at the DRPC site in Messina are presented.

of the shear modulus G and damping ratio D with depth.Seismic Dilatometer Marchetti Tests (SDMT) have been also carried out in the area of the "National Institute of Geophysics and Volcanology" building (INGV) and in the "Madre Teresa di Calcutta" building school (CD MTC), with the aim of an accurate geotechnical characterisation, including the evaluation of the shear wave velocity V s profile, as well as the profile of the horizontal stress index K d .
Shear wave velocity has been measured by different tests.The soil profiles in terms of the shear modulus G 0 and in terms of the shear wave velocity V s have been evaluated and compared by different in situ tests.
The redundancy of measurements is very useful, for instance, for site response and liquefaction analyses, which can be based either on V s values or K d values.
Boreholes were driven and undisturbed samples were retrieved for laboratory tests.Because of their relevance on the estimation of local ground shaking and site effects, data from in-hole geophysical surveys (Down-Hole and Seismic Dilatometer Marchetti Test) have been examined with special attention, particularly for S wave velocity measurements.It must be noted that the shear wave velocity V s was evaluated on the basis of both empirical correlations with in situ or laboratory tests and a few direct measurements.
One-dimensional local site response analyses have been performed assuming that all geologic boundaries are horizontal and the response of soil deposits is predominantly caused by waves propagating vertically from the underlying bedrock.Among the programs that adopt the equivalent linear analysis EERA code [Bardet et al., 2000] has been used in this study.
The three test sites in Catania are two school buildings in the province of Catania, the Catania division of the National Institute of Geophysics and Volcanology (indicated by the INGV acronym) and finally the fourth test site is the Messina's Regional Department of Civil Defence (indicated by the DRPC acronym).In this paper, the results of local seismic response analysis at the INGV site in Catania and at the DRPC site in Messina are presented.

THE STUDY AREAS
The first test site is the building of the INGV (National Institute of Geophysics and Volcanology), that is a public scientific institution which carries out research, monitoring and surveillance in the fields of geophysics and volcanology [Abate et al., 2016a;Abate and Massimino, 2016].The building (Figure 1) is located at the historical center of the high seismic risk city of Catania [Cavallaro et al., 2008].Among the main earthquakes that struck Catania, the most important is the 1693 event, also know as the "1693 Val di Noto earthquake" that struck also part of southern Italy, Calabria and Malta on January 11 at around 21:00 local time.
This earthquake was preceded by a damaging foreshock on January 9.It had an estimated magnitude of 7.4 on the moment magnitude scale and a maximum intensity of XI on the MCS, destroying at least 70 towns and cities, seriously affecting an area of 5,600 square kilometers and causing the death of about 60,000 people [Grasso and Maugeri, 2009a, 2009b, 2012, 2014;Castelli et al., 2016a, Maugeri et al., 2012].The earthquake was followed by tsunamis that devastated the coastal villages on the Ionian Sea and in the Straits of Messina.Almost two thirds of the entire population of Catania were killed.The extent and degree of destruction caused by the earthquake resulted in extensive rebuilding of the towns and cities of southeastern Sicily, particularly the Val di Noto [Castelli et al., 2016b[Castelli et al., , 2016c[Castelli et al., , 2016d]], in a homogeneous late Baroque style, described as "the culmination and final flowering of Baroque art in Europe".
Landsliding events triggered also by rainfall during earthquakes should be studied in these cases [Castelli and Lentini, 2013;Castelli et al., 2017], including also lateral spreading and liquefaction [Castelli et al., 2016e].
The second test site is the building of the Messina division of the DRPC (Regional Department of Civil Defence), located at the city center of Messina (Figure 2).
The most important historical earthquake was the 1908 Messina earthquake (also known as the 1908 Messina and Reggio Calabria earthquake), with a moment magnitude of 7.1 and a maximum MCS of XI.The cities of Messina and Reggio Calabria were almost completely destroyed and between 75,000 and 200,000 lives were lost.
The earthquake's epicenter was in the Strait of Messina which separates the busy port city of Messina in Sicily and Reggio Calabria on the Italian mainland.Its precise epicenter has been pinpointed to the northern Ionian Sea area close to the narrowest section of the Strait, the location of Messina.It had a depth of 5-6 miles (8-10 km).
At least 91% of structures in Messina were destroyed or strongly damaged and some 75,000 people were killed in the city and suburbs.Reggio Calabria and other locations in Calabria also suffered heavy damage, with some 25.000 people killed.Reggio Calabria historic centre was almost completely erased.The number of casualties is based on the 1901 and 1911 census data.The ground shook for about 30 to 40 seconds, and the damage was widespread, with destruction felt within a 300-kilometer radius.In Calabria Region, the ground shook violently from Scilla to the south of Reggio Calabria and caused landslides in the Reggio Calabria area.

GEOTECHNICAL SITE CHARACTERISATION
An intensive geotechnical site characterization was carried out performing in situ and laboratory tests, in order to determine the most appropriate static and dy-namic soil parameters [Ferraro et al., 2015].By means of laboratory tests the values of unit weight, natural water content, void ratio, porosity, saturation ratio have been obtained.Resonant Column Tests (RCT) and Cyclic Loading Torsional Shear Tests (CLTST) have been also performed [Cavallaro et al., 2012;Caruso et al., 2016;Castelli et al., 2016f;Cavallaro et al., 2016a;2016b].
The types of tests performed are the same for both sites (INGV and DRPC), in fact, for each site, three boreholes were carried out, with undisturbed soil sampling.The three surveys were also equipped in order to perform Down Hole (D-H) and Cross Hole (C-H) tests according to this sequence: S1 and S2 boreholes for the execution of a  Cross Hole test and S3 boreholes for the execution of the Down Hole test.Seismic Dilatometer Marchetti Tests SDMTs have been also performed in each site.

INGV CATANIA TEST SITE
Figure 3 shows the location of the three boreholes in the INGV Catania test site and Figure 4 shows the soil stratigraphy of the boreholes.The S1 and S3 boreholes have a depth of 60 meters, the S2 borehole has a depth of 80 meters.Table 1 shows the list of soil samples for laboratory tests; for the geotechnical model considered in the numerical analysis, the following four soil samples were used: S1C1 (retrieved at 4,00 -4,50 m), S2C3 (retrieved at 14,00-14,40 m), S2C5 (retrieved at 72,00 -72,50 m), S3C4 (retrieved at 5,50 -5,90 m). Figure 4 -INGV -Soil stratigraphy in the order S1, S2, S3.
Figure 5     the 800 m/s value.The SDMT test shows a dual behaviour: up to the depth of 7.5 m it seems to overestimate V s values compared to D-H and C-H tests, while from 7.5 m to 45 m it shows a linear trend with values that lie in the middle between the results of the D-H test and the C-H test results.This apparatus was also used in offshore condition by Cavallaro et al. [2013aCavallaro et al. [ , 2013b]].Results of Resonant Column Tests (RCT) and Cyclic Loading Torsional Shear Tests (CLTST) are presented in Figures 6-8 to evaluate the shear modulus G and damping ratio D variation with the shear strain level γ.These tests are fundamental for the accurate evaluation of local site response analyses, as well as for the evaluation of the dynamic behavior of fullcoupled soil-structure systems [Abate et al., 2007;2015;Massimino et al., 2015].The specimen was anisotropically reconsolidated to the best estimate of the in situ effective vertical and horizontal stress.The same specimen was first subjected to RCT, then to CLTST after a rest period of 24 hrs with opened drainage.CLTSTs were performed under stress control condition by applying a torque with triangular time history at a frequency of 0.1 Hz.
Results of Resonant Column Tests (RCT) and Cyclic Loading Torsional Shear Tests (CLTST) are presented in Figures 11-15 to evaluate the shear modulus G and damping ratio D. The specimen was anisotropically reconsolidated to the best estimate of the in situ effective vertical and horizontal stress.The same specimen was first subjected to RCT, then to CLTST after a rest period of 24 hrs with opened drainage.CLTSTs were performed  under stress control condition by applying a torque with triangular time history at a frequency of 0.1 Hz.
Figure 16 shows the

NUMERICAL ANALYSES IN THE TEST SITES
One-dimensional local site response analyses assume that all geologic boundaries are horizontal and the response of soil deposits is predominantly caused by waves propagating vertically from the underlying bedrock.So, one-dimensional numerical codes are valid for modelling plane-parallel layers along a vertical column, assuming laterally-homogeneous stratigraphy.Under these assumptions, the main factors responsible for seismic motion amplifications are: 1) impedance contrasts between ground layers, particularly with bedrock; 2) resonance effects due to the closeness between the frequencies of the motion at the substrate and the natural vibration of the deposit.Calculation procedures consider, in the solution of the dynamic equilibrium of the system, the non-linear relation through two types of analyses: equivalent linear and nonlinear analyses.The equivalent linear analysis consists in the execution of a sequence of complete linear analysis with subsequent update of the parameters of stiffness and damping until the satisfaction of a predetermined convergence criterion.There parameters depend on the state of deformation of the ground.The nonlinear analysis consists in the integration step-by-step of the equations of motion, simultaneously changing the parameter values of stiffness and damping.The first analysis provides satisfactory results at not excessive deformation of the ground, less than 1%, for higher deformation is necessary to use non-linear incremental analysis.Among the programs that adopt the equivalent linear analysis the best known and most fre-quently used is the computer code SHAKE [Schnabel et al., 1972a[Schnabel et al., , 1972b;;Idriss and Sun, 1992] and later revisions like EERA [Bardet et al., 2000]   homogeneous layers with linearized visco-elastic behaviour).
It consists of flat and parallel layers of infinite horizontal extension on an outcropping half-space corre-sponding to the bedrock, on which the input motion is applied.
Each layer is considered to be homogeneous and isotropic and characterized by the thickness h, the density, the shear modulus G and the damping ratio ξ.The input motion propagates in the direction perpendicular to the free surface.Figure 17 shows the calculation scheme for N layers with these properties: shear modulus G, damping ratio ξ, thickness h.The amplitude of incident and reflected waves in each layer, and reflected waves F N at the bedrock are unknown.The incident wave E N at the bedrock is given.The calculation method considers the motion equations applied in each layer under two conditions: 1) the congruence between displacements and stresses at the interfaces; 2) equilibrium between incident and reflected waves at the surface.
Finally, in order to take into account the non linear soil behaviour, an iterative procedure is carried out by considering the behaviour of the shear modulus G and of the damping ratio D, as functions of shear strain γ.The main steps of the iterative procedure are: 1) estimation for G and ξ in each layer; 2) ground response calculation in each layer; 3) effective shear strain calculation based on the maximum shear strain; 4) new iteration based on G and ξ new values and new ground response calculation until the desired convergence value is reached.
The results are presented in terms of acceleration time history, Fourier spectra, response spectra at the surface and in terms of maximum acceleration profiles as function of the depth [Ferraro et al., 2009].
One dimensional equivalent linear analyses have been carried out at the Catania INGV test site with different bedrock depths (200 m, 300 m, 400 m, 600m) and at the Messina DRPC test site (DRPC-S1S2 and DRPC-S3).

INGV TEST SITE NUMERICAL ANALYSES
As regards the input motion at the Ultimate Limit State (ULS), the site response analyses at the INGV test site were made using as excitation at the base of the models both synthetic seismograms and recorded ac-celerograms, the first obtained by the 1693 Val di Noto and 1818 Etna earthquakes, the second by the recordings of the December 13, 1990 earthquake, at the Sortino (a max = 1.00 m/sec 2 ) recording station.The approach of using synthetic seismograms for generating    a seismic ground motion scenario at the bedrock has been so used in the present work.The reference events were the catastrophic earthquakes that struck Eastern Sicily: on January 11, 1693, assumed as a level I maximum credible earthquake scenario; on February 20, 1818, assumed as a level II base operative earthquake scenario.
The epicentre of the source in the case of 1693 scenario is about 13 km far from the coast, which is considered as the coordinate system source.A magnitude M = 7.0, corresponding to the destructive event which struck Catania in 1693, has been taken into account.The source has been simulated through the overlaying of 5 sources placed at different depths and activated at different times, simulating approximately the propagation of the rupture on the segment fault.Seismograms       Results in terms of maximum acceleration at the surface and soil amplification factor R, for the 600 m soil model, using as input ag = 0.282 g acceleration value.
TABLE 3. Results in terms of maximum acceleration at the surface and soil amplification factor R, for the 200 m soil model, using as input ag = 0.207 g acceleration value.
TABLE 4. Results in terms of maximum acceleration at the surface and soil amplification factor R, for the 600 m soil model, using as input ag = 0.207 g acceleration value.
TABLE 5. Results in terms of maximum acceleration at the surface and soil amplification factor R, for the 200 m soil model, using as input ag = 0.282 g acceleration value.
have been drawn for each site long a set of six receivers placed at different depths, starting from the surface up to almost 170 m [Grasso et al., 2005].
In the case of level II 1818 earthquake scenario seis-    Figure 18 shows some of the synthetic seismograms by 1693 and 1818 scenario earthquakes and the recorded accelerogram of December 13, 1990 earthquake used as input for seismic response analyses.
Preliminary numerical analyses have been carried   Caruso et al., 2016].Finally, the Figure 23 shows the comparison in terms of maximum acceleration profiles, for the different soil models considered in the numerical analysis.Additional seismic response analyses have been carried out [Ferraro et al., 2016] for the 200 m and 600 m seismic bedrock soil models, by using the same input seismograms of Figure 18, scaled to the different values of acceleration of 0.207 g and 0.282 g provided by the Italian seismic code (ultimate limit state with return periods of 475 and 949 years).The numerical analyses results are shown in terms of response spectra at surface (Figures 24-27), and in terms of surface maximum acceleration and amplification ratio (Tables 3-6).

DRPC TEST SITE NUMERICAL ANALYSES
Numerical analyses have been carried out considering two soil models, based on the Cross-Hole test results (Figure 28) performed in the S1 and S2  boreholes, and on the Down-Hole test results (Figure 29) performed in the S3 borehole, by using the seven input seismograms [Amoruso et al., 2002] shown in Figure 30 obtained by a source modeling of 1908 Messina and Reggio Calabria earthquake and 1693 Val di Noto earthquake, scaled to the value of 0.332 g, provided by the Italian seismic code NTC2008 (ultimate state limit with a return period of 949 years).
The numerical analyses results are shown in Figures 31-34 in terms of time history of surface maximum acceleration, amplification ratio and response spectra, and in terms of surface maximum acceleration and soil amplification factor R values (Tables 7-8).

CONCLUSIONS
In this study it was benefited of a great availability of borehole data, geophysical surveys and laboratory tests carried out from various campaigns of geological surveys that, in different phases, have interested the area of two tests sites in the cities of Catania and Messina seismic areas.In order to study the dynamic characteristics of soils in the test sites, in situ and laboratory investigations have been carried out to obtain soil profiles with special attention being paid to the variation of the shear modulus G and damping ratio D with depth.
Local site response analyses have been brought for the site area at Ultimate Limitation State ULS by 1-D linear equivalent computer codes for the evaluation of the amplification factors of the maximum acceleration.Results of the site response analysis show high values in soil amplification effects of the soil.The horizontal acceleration at the surface obtained by the site response analysis is in some cases higher than 0.500 g, with soil amplification factors up to 2.30, higher than the soil amplification factor Ss on the basis of the average shear waves velocity (Vs30) according to the current Italian seismic code.The difference between the two soil amplification factors is due to the fact that (Vs30) is not a key parameter in the evaluation of soil amplification.Through 1-D performed numerical analyses it has been possible to evaluate the influence of stratigraphic effects in seismic response of the site.
Finally elastic pseudo-acceleration response spectra at the ground surface have been compared (graphically) with the reference ones, provided by the NTC 2008 Italian seismic code at ULS. Seismic retrofitting and/or improving have to be definitely considered a multidisciplinary subject, which de-pends in fact on many factors, such as: local site effects and the dynamic interaction between the foundation soil and the structure.The accurate investigation on the structure and the surrounding soil is the first fundamental step for a realistic evaluation of the structure seismic performance.

SITE
Figure3shows the location of the three boreholes in the INGV Catania test site and Figure4shows the soil stratigraphy of the boreholes.The S1 and S3 boreholes have a depth of 60 meters, the S2 borehole has a depth of 80 meters.Table1shows the list of soil samples for laboratory tests; for the geotechnical model considered in the numerical analysis, the following four soil samples were used: S1C1 (retrieved at 4,00 -4,50 m), S2C3 (retrieved at 14,00-14,40 m), S2C5 (retrieved at 72,00 -72,50 m), S3C4 (retrieved at 5,50 -5,90 m).Figure4-INGV -Soil stratigraphy in the order S1, S2, S3.Figure5shows the V s profiles obtained by D-H, C-H and SDMT tests and their comparison.The Down Hole test shows a V s profile that increases with depth almost according to a linear trend with values variables from 190 m/s to 1000 m/s; at the depth of 35 m the V s profile reaches the value of 800 m/s.

FIGURE 5 .
FIGURE 5. INGV -V s profiles obtained by D-H, C-H and SDMT tests and their comparison.
V s profiles obtained by D-H, C-H and SDMT tests and their comparison.The Down Hole test shows a V s profile that increase with depth almost according to a linear trend with values variables from 190 m/s to 1000 m/s; at the depth of about 50 m the V s profile reaches the value of 800 m/s.The Cross Hole test is in good agreement with the Down Hole test.The C-H test shows a V s profile that linearly increase with depth with values variable between 190 m/s and 800 m/s at the depth of about 50,00 m.The SDMT test shows V s values which are in good agreement if compared to D-H and C-H tests, up to the depth of 30 m.

FIGURE 19 .
FIGURE 19.Soil model with bedrock at the depth of 200 m.Time history of surface acceleration (a), amplification ratio (b), spectral acceleration (c).

FIGURE 20 .FIGURE 21 .
FIGURE 20.Soil model with bedrock at the depth of 300 m.Time history of surface acceleration (a), amplification ratio (b), spectral acceleration (c).

FIGURE 22
FIGURE 22 Soil model with bedrock at the depth of 600 m.Time history of surface acceleration (a), amplification ratio (b), spectral acceleration (c).

FIGURE 23 .
FIGURE 23.Comparison in terms of maximum acceleration profiles, for the different soil models considered in the numerical analysis.

FIGURE 24 .
FIGURE 24.Comparison between spectral acceleration at the surface obtained by numerical analyses for the 200 m seismic bedrock soil model, and input acceleration of 0.207 g, with the spectral acceleration provided by the Italian seismic code NTC2008.Labels of different colours in the legend represent different synthetic seismograms at various depths of 1693 (20m, 31m, 40m) and 1818 (40m, 49m, 56m) scenario earthquakes used for in the numerical analyses.

FIGURE 25 .
FIGURE 25.Comparison between spectral acceleration at the surface obtained by numerical analyses for the 600 m seismic bedrock soil model and input acceleration of 0.207 g, with the spectral acceleration provided by the Italian seismic code NTC2008.Labels of different colours in the legend represent different synthetic seismograms at various depths of 1693 (20m, 31m, 40m) and 1818 (40m, 49m, 56m) scenario earthquakes used for in the numerical analyses.

FIGURE 27 .
FIGURE 27.Comparison between spectral acceleration at the surface obtained by numerical analyses for the 600 m seismic bedrock soil model and input acceleration of 0.282 g, with the spectral acceleration provided by the Italian seismic code NTC2008.Labels of different colours in the legend represent different synthetic seismograms at various depths of 1693 (20m, 31m, 40m) and 1818 (40m, 49m, 56m) scenario earthquakes used for in the numerical analyses.

FIGURE 26 .
FIGURE 26.Comparison between spectral acceleration at the surface obtained by numerical analyses for the 200 m seismic bedrock soil model and input acceleration of 0.282 g, with the spectral acceleration provided by the Italian seismic code NTC2008.Labels of different colours in the legend represent different synthetic seismograms at various depths of 1693 (20m, 31m, 40m) and 1818 (40m, 49m, 56m) scenario earthquakes used for in the numerical analyses.

FIGURE 28 .
FIGURE 28.DRPC soil model based on Cross-Hole test.FIGURE 29.DRPC soil model based on Down-Hole test.
FIGURE 30.DRPC-Set of input seismograms used for numerical analyses.

FIGURE 31 .
FIGURE 31.DRPC -CH-S1-S2 model.Results at the surface in terms of acceleration time history (a), amplification ratio (b) and spectral acceleration (c).Labels of different colours in the legend represent different synthetic seismograms obtained from different source models.
the 1990 recorded input motion scaled to six different values of PGA provided by the current Italian seismic code for different ULS and the different return periods of 60, 101, 475, 949, 1898 and 1950 years: 0.083 g; 0.102 g; 0.207 g; 0.282 g; 0.392 g; 0.397 g.In Figures 19-22 the results of site response analyses are presented, for different bedrock depths (200 m, 300 m, 400 m, 600m), in terms of time history of surface acceleration, amplification ratio and spectral acceleration [

FIGURE 33 .
FIGURE 33.DRPC -DH-S3 model.Results at the surface in terms of acceleration time history (a), amplification ratio (b), spectral acceleration.Labels of different colours in the legend represent different synthetic seismograms obtained from different source models.

TABLE 7 .
Results in terms of maximum acceleration at the surface and soil amplification factor R, for the Cross-Hole S1-S2 soil model.

TABLE 8 .
Results in terms of maximum acceleration at the surface and soil amplification factor R, for the Down-Hole S3 soil model.