A08322

The Organizing Committee was composed by:Francesco Benedettini, Dante Galeota, Vincenzo Gattulli, Antonio Rovelli,Antonello Salvatori, Marco Tallini, Gianfranco Totani, Gianfranco Valente.

The papers presented in this Proceedings were selected by the Scientific Committee:Dario Albarello, Arrigo Caserta, Alberto Delladio, Gino D’Ovidio, Aladino Govoni, GiulianoMilana, Fabrizio Marra, Marco Mucciarelli.

The Workshop received supports by:University of l’Aquila and its Dipartimento di Ingegneria delle Strutture Acque e Terreno(DISAT) , Società Roma Metropolitane s.r.l., Istituto Nazionale di Geofisica (INGV), Agenzianazionale per le nuove tecnologie, l’energia e lo sviluppo economico sostenibile (ENEA).

Assistance for international relationships:Aki Takizawa

The picture in the frontal covering represents: the Church of “Santa Maria” in Paganica(L’Aquila) damaged by the carthquake on 6 April 2009. It was extracted from the enclo-sed paper by:Prof. Mauro Dolce.

PROCEEDINGS OF THE WORKSHOP DISS_10,L’AQUILA, 19 MARCH 2010

THE DYNAMICINTERACTION OF SOIL

AND STRUCTURE

Edited byGino D’Ovidio

Yutaka NakamuraAntonio Rovelli

Gianfranco Valente

Copyright © MMXARACNE editrice S.r.l.

www.aracneeditrice.itinfo@aracneeditrice.it

via Raffaele Garofalo, 133 A/B00173 Roma(06) 93781065

ISBN 978–88–548–3693–8

No part of this book may be reproduced in any form,by print, photoprint, microfilm, microfiche, or any other means,

without written permission from the publisher.

1st edition: February 2011

5

Index

7 Introduction

9 The H/V technique and example of its application for l’Aquila and

Rome areas Yutaka Nakamura

23 Modelling of ambient vibration wavefield for passive seismic

prospectingDario Albarello, Enrico Lunedei

37 Site characterization by seismic dilatometer (SDMT) in the area of L’AquilaSara Amoroso, Diego Marchetti, Silvano Marchetti, Paola Monaco,Ferdinando Totani, Gianfranco Totani

63 Base isolated buildings: economic evaluatations Paolo Clemente, Giacomo Buffarini

75 Seismic response of two couples of similar buildings in Pettino-L’Aquila during the 6 April 2009 earthquakeFrancesco Benedettini, Adriano De Sortis, Giuliano Milana

89 Dynamic interaction of soil and structure by rail traffic Arrigo Caserta, Maurizio Cerone, Francesco Crisi, Alberto Delladio, Gino D’Ovidio, Aladino Govoni, Fabrizio Marra, Yutaka Nakamura, Antonio Rovelli, Gianfranco Valente

109 The 6 April 2009 Abruzzi EarthquakeMauro Dolce

123 Seismic site response in L’Aquila city center and in suburban areas Giuliano Milana, Riccardo Mario Azzara, Francesco Bergamaschi, Etienne Bertrand, Paola Bordoni, Fabrizio Cara, Rocco Cogliano, Giovanna Cultrera, Giuseppe Di Giulio, Anne-Marie Duval, Antonio Fodarella, Sandro Marcucci, Stefania Pucillo, Julie Régnier, Gaetano Riccio

137 Extensive seismic survey with passive prospecting techniques: the L'Aquila earthquake example Dario Albarello, Maria Rosaria Gallipoli, Marcello Bianca

153 Earthquake Early Warning and Realtime Earthquake Disaster PreventionYutaka Nakamura

6

Vincenzo Gattulli, Marco Lepidi, Francesco Potenza, Alfredo M. Ceci

181 Damage level vs fundamental frequencies of damaged RC buildings Rocco Ditommaso, Maria Rosaria Gallipoli, Marco Mucciarelli, Marco Vona

189 Analisi meccanica del danno di elementi non strutturali Antonello Salvatori

195 In-depth understanding of 2009 L'Aquila earthquake: the contribution of Gran Sasso physico-chemical groundwater data Raffaele Adinolfi Falcone, Antonella Amoruso, Valentina Carucci, Luca Crescentini, Antonella Falgiani, Mauro Manetta, Andrea Marchetti, Marco Paolessi, Barbara Parisse, Marco Petitta, Sergio Rusi, Donato Sciannamblo, Michele Spizzico, Marco Tallini

219 Authors index

167 Optimization of viscous coupling between adjacent structures excited by ground motion

7

Introduction

It is a great pleasure for me to open this workshop on the dynamic interaction of soil-structure.

First of all I would like to welcome and thank Professor Yutaka Nakamura.

Professor Nakamura is one of the world’s leading scientists in engineering seismology.

He is President of the System and Data Research Company and a visiting professor at the Tokyo Institute of Technology.

Our special thanks go to the Organizing Committee and in particular, Professor Gianfranco Valente, who made this important event possible.

One of the most important causes of damage to structures is inaccuracy in estimating earthquake excitation characteristics during the design phase.

It is well known that many of the buildings in L'Aquila were built well over a hundred years ago. Hence the earthquake acted upon a variety of structures, some only designed for structural loads whilst others with no structural design at all. It is therefore reasonable to expect that the majority of structures experienced some damage.

The soil conditions of the area and the geological formation had a significant effect on the distribution of the structural damage.

The information that can be gathered from studies, symposia and workshops on microzonation and dynamic soil-structure interaction will contribute to more accurate estimations of expected ground motion caused by earthquakes and clarify governing principles for the structural rehabilitation of damaged buildings.

This is the main focus of this important workshop. In concluding my speech, I would like to wish you all a fruitful and

pleasant stay in our University.

The Director of Dpt DISAT at l’Aquila University: Dante Galeota

8

This book is a collection of papers from the workshop “The Dynamic Interaction of Soil and Structure” (DISS_10) held in the city of L’Aquila approximately a year after the destructive earthquake of 6 April 2009. The workshop was organized by the University of L’Aquila, and supported by the Società Metropolitane di Roma, INGV and ENEA, with two important aims: the first to promote dialogue between scientists from differing disciplines (geologists, seismologists, engineers) to establish a preliminary picture of the earthquake and its impact on the territory; and the second to tackle more general problems concerning the safety of historical cities. In the debates that followed, unresolved problems were defined and innovative directions for new research and new perspectives proposed, with particular reference to monumental heritage. One of the greatest challenges undertaken by seismologists and earthquake engineers in the second half of the last century was the study of ground motion radiated by large earthquakes and its interaction with building structure. The slow but significant advance of research into experimental measurements, theoretical explanations and numerical modeling has provided the scientific community with extremely useful tools for the third millennium. Following the earthquake in April 2009 in L’Aquila, it has been shown that a significant amount of important information can be gathered after an earthquake. This information was presented during the workshop and formed the basis of a number of constructive discussions. Amongst the innovative tools discussed was the use of ambient vibrations which are particularly important in seismology and earthquake engineering because of their low cost and availability. However, the definition of a largely accepted theoretical basis is still controversial. This issue monopolized a significant percentage of presentations but further innovative ideas and models are also proposed in this book. The broad spectrum of contributions from across the scientific community presented in this book are a reflection of the specialized approach that is being adopted in this field.

The Papers from the Proceedings of the Workshop are published in the same order as they were presented during the Workshop. The Editors: Gino D’Ovidio, Yutaka Nakamura, Antonio Rovelli and Gianfranco Valente L’Aquila, 30 November 2010.

_______________________________

1 System and Data Research, Tokyo, Japan

2 Tokyo Institute of Technology, Yokohama, Japan

9

The H/V Technique and Example of its Application for

L’Aquila and Rome Areas Yutaka Nakamura1, 2

Abstract The H/V technique is an estimation tool for the local site effect as not only the predominant frequency but also the amplification factor. Both microtremor and strong motion consist of various kinds of waves as P-wave, S-wave, Rayleigh wave, Love wave and so on, so the spec-trum derived from microtremor or strong motion shows many peaks corresponding to these wave components. From the viewpoint of the seismic disaster mitigation, it is important to pick out S-wave from microtremor and investigate its characteristics because the severe damage is mainly caused by S-wave. Although it is fine way to esti-mate the S-wave structure of the surface ground layers from the char-acteristics of Rayleigh wave, this technique requires not only a precise setup of the measurement points but also complex analysis. On the other hand, the H/V technique is established as a direct approach for the site amplification estimation, so it is easy to apply for the various ground condition. This paper shows the scheme of the H/V technique and its application for L’Aquila and Rome areas.

1. Introduction The H/V technique is an estimation tool for the local site effect as not only the predominant frequency but also the amplification factor. Both microtremor and strong motion consist of various kinds of waves as P-wave, S-wave, Rayleigh wave, Love wave and so on, so the spec-trum derived from microtremor or strong motion shows many peaks corresponding to these wave components. From the viewpoint of the

The dynamic interaction of soil and structureISBN 978–88–548–3693–8DOI 10.4399/97888548369381pag. 9–22 (febbraio 2011)

Yutaka Nakamura

10

seismic disaster mitigation, it is important to pick out S-wave from microtremor and investigate its characteristics because the severe damage is mainly caused by S-wave. Although it is fine way to esti-mate the S-wave structure of the surface ground layers from the char-acteristics of Rayleigh wave, this technique requires not only a precise setup of the measurement points but also complex analysis. On the other hand, the H/V technique is established as a direct approach for the site amplification, so it is easy to apply for the various ground condition. 2. Brief story of the H/V technique

Originally, the H/V spectral ratio was found from strong motion re-cords at various sites in Japan. Although the horizontal motion shows larger value than the vertical motion on the soft ground, the horizontal and the vertical motion are similar in both the maximum value and the shape of the waveform on the hard ground. At first, the horizontal to vertical ratio for the maximum value was derived and compared with the softness and the amplification factor of ground. As a result, the horizontal to vertical ratio of the maximum value highly corresponds to the ground characteristic on the same site [1]. Then the horizontal to vertical spectral ratio (H/V spectral ratio) of microtremor was estab-lished to estimate the predominant frequency and the amplification factor and confirmed that the estimation result is stable not depended on the measured time and season [2]. 3. Outline of the H/V technique There is no reason to be amplified the amplitude of a particular fre-quency range or direction at the hard and uniform ground. It is natural that the vibration in the hard ground is uniform for wide frequency range or each direction. The measured results of both microtremor and earthquake suggest it. On the other hand, it is important for a soft ground deposition over the hard basement to grasp the amplification characteristics of the horizontal motion. At the frequency range the horizontal motion is amplified, it seems that the vertical motion is not amplified. And at a ground with that kind of the amplification charac-

The H/V Technique and Example of its Application for L’Aquila and Rome Areas 11

teristics, it is important to consider the Rayleigh wave propagation on the ground surface. Earthquake disaster is caused mainly by the body wave and if the Rayleigh wave causes any damage, the damage must be a minor damage. That is to say, it is important to understand the phenomenon of the trapped energy of the body wave inside the soft surface layer. The phenomenon shows the amplification characteris-tics of earthquake motion by so called the multiple reflections. From this viewpoint, the Rayleigh wave exists as a noise and it is necessary to reject the effect of the Rayleigh wave. The effect of the Rayleigh wave for the amplification characteristics R (= Ahs/Ahb, the surface (Ahs) to the basement (Ahb) of the horizontal motion) of the surface ground is estimated by the surface (Avs) to the basement (Avb) ratio of the vertical motion E (= Avs/Avb). Then the corrected amplifica-tion characteristic Am is shown as follows;

Am = R/E = (Ahs/Ahb)/(Avs/Avb) = (Ahs/Avs)/(Ahb/Avb) Ahs/Avs This is the H/V spectral ratio [2]. Figure 1 shows the concept of the H/V technique. The H/V spectral ratio shows the amplification charac-teristics by the multiple reflections of the SH wave at least around the predominant frequency F0, and also shows the characteristics con-taminated by the Rayleigh wave around 2F0. In case of smaller effect of the Rayleigh wave, it is possi-ble to estimate not only the preliminary peak but also the secondary peak of the amplification characteristics caused by the multiple re-flections from the H/V spec-tral ratio. Please see [3] or [4] in detail. 4. Average velocity, thickness and vulnerability index of surface layer

F0 2F0

For the H/V technique,

: Measuring Point

Figure 1. Concept of the H/V technique

Yutaka Nakamura

12

The H/V spectral ratio of both microtremor and strong motion can provide at least fundamental frequency and its amplification factor. Based on the predominant frequency Fg in Hz and the amplification factor Ag derived from the H/V spectral ratio, the average velocity AVS in m/s, the thickness h in m for the surface ground are defined as fol-lows,

AVS = Vb/Ag, h = AVS/(4Fg) = Vb/(4AgFg),

where Vb is velocity in basement ground. On the other hand, earthquake damage of members of structures occurs at the time of exceeding the destruct limit of the strain caused by deformation, and it may cause a collapse if the stability of the structure is lacked. Then the vulnerability index was de-fined focusing on the strain [5]. K value in 10-6/cm/s2 is defined to estimate the strain in the unit of 10-6 at particu-lar parts of the structures, multiplying the possible maximum acceleration at the engineering basement (see Figure 2). These K val-ues are defined individually for ground and various structures, as Kg value for ground, Kj value for em-bankment, Ks value for rigid frame structure, Kb value for building and Kd value for derailment. For example Kg value is defined as follows (see Figure 3).

Figure 2. Vulnerability indexes K values

Figure 3. Vulnerability index Kg value for surface ground

The H/V Technique and Example of its Application for L’Aquila and Rome Areas 13

Kg = Ag2/Fg

If maximum accelera-tion is 100 cm/s2 at the basement and Kg value is 20, the average strain at the surface ground layer is estimated 2000 (x10-6) and this strain may cause some damage to this ground. Figure 4 shows an exam-ple of the Kg value distri-bution at Marina district in San Francisco after the 1989 Loma Prieta earthquake. Large Kg (Kg > 20) were observed at the severe liquefied area. 5. Microtremor measurements in L’Aquila and Rome

5.1. Location and method Figure 5 shows the measurement area and locations in L’Aquila and Rome area. L’Aquila area was measured before the workshop with its par-ticipants. Rome area was measured after the workshop. Measure-ment locations around the Colosseum were set as extensions of meas-urement on 1998. Microtremor was measured at each site three times and its

Figure 4. An example of comparison between Kg value and damage

Figure 5(a). Overview of the measurement in L’Aquila

LRG

ULG

SBG

HOG

AQA

AQV32 sites in L’Aquila measured

during 17th and 18th of March

2010 before L’Aquila Workshop

held at 19th of March 2010

Yutaka Nakamura

14

length was 40.96 seconds in 100 Hz sampling using NewPIC+ of SDR (see the detail on http://www.sdr.co.jp). Fourier Spectrum and H/V spectral ratio were calculated for each 40.96 seconds data with twenty times Hanning window and then aver-aged the three data at each site. Peak frequency Fg and peak value Ag were read from the averaged spectral ratio H/V, and Kg value were calculated with these values.

Figure 5(b). Measured sites in ULG and SBG

ULG1

ULG2

ULG3

ULG4

ULG5

ULG6

ULG7

ULG8

Measurement of ULG in 2010

SBG1 SBG2

SBG4

SBG3

SBG5

Measurement of SBG in 2010

Figure 5(c). Measured sites in LRG

The H/V Technique and Example of its Application for L’Aquila and Rome Areas 15

Figure 5(d). Measured sites in HOG, AQV and AQA

CMGW

11 sites in Rome measured during 20th and 21st of March after L’Aquila Workshop held at 19th of March 2010

RHG

ARG1 ARG2

ARG3ARG4

ARG5

ARG7

ARG8

ARG6

G1 G2

G3G4 G5

G6

RH412

Measurement around Colosseum in 1998 markedMeasurement around Colosseum in 2010 marked

Figure 5(e). Measured sites in Rome

Yutaka Nakamura

16

5.2. Results Figure 6 shows the aver-aged spectral ratio H/V for each sites. Please note that the point LRG7 shows abnormal result because of a strong wind, so this data was omitted from the discussions follow-ing. Table 1 summarizes the measurement result for Fg, Ag, Kg, h and AVS. According to this table the predominant fre-quencies in L’Aquila area dis-tributes totally wide range as 1-9 Hz, and distributes in 2-3 Hz in the mountain area (LRG). And that of Rome area distributes in 1-2 Hz. AVS as-suming Vs = 600 m/s shows large value in Rome and less than 100 m/s for many sites in L’Aquila, and this suggests that there may exists a soft ground in L’Aquila. The Kg value shows almost small amount in Rome and most part of L’Aquila except the mountain area (LRG), and that in the mountain area shows obvi-ously large value. Figure 7 shows the distribution of the Kg value. In this figure, the diameter of the circle for each point corresponds to the amount of its Kg value. It is found that the Kg value differs widely from each other. The campus of L’Aquila University in the old town (ULG) was suffered se-

AQV

0.1

1

10

0.1 1 10 100AQA

0.1

1

10

0.1 1 10 100Frequency in Hz

Am

plifi

catio

n Fa

ctor

Am

plifi

catio

n Fa

ctor

Figure 6(a). Spectral ra-tios H/V for AQV and AQA

Fg Ag Kg h AVSsite Hz /Gal m m/s

SBG1 4.32 3.38 2.6 10 178 SBG2 3.86 3.33 2.9 12 180 SBG3 2.49 3.03 3.7 20 198 SBG4 4.05 5.29 6.9 7 113 SBG5 2.78 3.05 3.3 18 197 ULG1 5.71 11.70 24.0 2 51 ULG2 6.35 10.80 18.4 2 56 ULG3 6.57 14.30 31.1 2 42 ULG4 8.72 14.50 24.1 1 41 ULG5 6.47 3.78 2.2 6 159 ULG6 9.20 6.37 4.4 3 94 ULG7 6.13 2.54 1.1 10 236 ULG8 5.62 3.16 1.8 8 190 LRG1 2.59 7.40 21.1 8 81 LRG2 3.30 7.70 18.0 6 78 LRG3 2.12 7.10 23.8 10 85 LRG4 3.15 6.18 12.1 8 97 LRG5 1.56 3.79 9.2 25 158 LRG6 2.15 5.49 14.0 13 109 LRG7 N/A N/ALRG8 2.61 5.93 13.5 10 101 LRG9 2.61 4.61 8.1 12 130

LRG10 2.95 4.58 7.1 11 131 LRG11 3.03 5.71 10.8 9 105 LRG12 2.93 5.16 9.1 10 116 LRG13 3.10 7.27 17.0 7 83 LRG14 2.08 3.74 6.7 19 160 HOG1 1.54 3.20 6.6 30 188 HOG2 4.17 4.25 4.3 8 141 HOG3 7.69 4.71 2.9 4 127

AQV 2.93 5.44 10.1 9 110 AQA 8.67 7.04 5.7 2 85

CMGW 1.37 2.76 5.6 40 217 RHG 2.59 4.19 6.8 14 143

ARG1 2.51 2.99 3.6 20 201 ARG2 2.15 3.13 4.6 22 192 ARG3 1.54 2.60 4.4 37 231 ARG4 1.59 1.97 2.4 48 305 ARG5 2.12 1.81 1.5 39 331 ARG6 3.69 1.82 0.9 22 330 ARG7 1.68 3.57 7.6 25 168 ARG8 1.15 2.79 6.8 47 215 98G1 1.37 2.30 3.9 48 261 98G2 1.12 2.39 5.1 56 251 98G3 1.42 2.20 3.4 48 273 98G4 1.44 3.23 7.2 32 186 98G5 2.25 2.69 3.2 25 223 98G6 0.49 2.86 16.7 107 210

N/A: affected by strong wind

Table 1. Summary of measurements

The H/V Technique and Example of its Application for L’Aquila and Rome Areas 17

vere damage. Espe-cially in the courtyard filled by the ancient debris, the embankment higher than surrounding roads liquefied severely and flew out to the sur-rounding roads. Hence Kg values are still over 20 even after recon-struction of the em-bankment, Kg values can be estimated in the same range before the reconstruction and the shear strain in the sur-face ground seems to be over 1/200 caused by the several hun-dreds Gals strong mo-tion in the basement ground and the severe liquefaction may be occurred. There were also subsided road may caused by the liquefac-tion, so the Kg value seems to be large, too. Thus, advance calcula-tion of Kg value with the microtremor meas-urement can help tak-ing the proper coun-termeasure previously with choosing the pos-sible damage area of

0.1

1

10

100

0.1 1 10 100

ULG2YULG2X

0.1

1

10

100

0.1 1 10 100

ULG1YULG1X

0.1

1

10

100

0.1 1 10 100

ULG4YULG4X

0.1

1

10

0.1 1 10 100

SBG1YSBG1X

0.1

1

10

0.1 1 10 100

SBG2YSBG2X

0.1

1

10

0.1 1 10 100

SBG3YSBG3X

0.1

1

10

0.1 1 10 100

SBG4YSBG4X

0.1

1

10

0.1 1 10 100

ULG6YULG6X

0.1

1

10

0.1 1 10 100

ULG8YULG8X

0.1

1

10

0.1 1 10 100

SBG5YSBG5X

0.1

1

10

100

0.1 1 10 100

ULG3YULG3X

0.1

1

10

0.1 1 10 100

ULG5YULG5X

0.1

1

10

0.1 1 10 100

ULG7YULG7X

Ampl

ifica

tion

Fact

or

Frequency in Hz Frequency in Hz Frequency in Hz

Am

plifi

catio

n Fa

ctor

Campus in Old town of L’Aquila (ULG) Around San Bernardino (SBG)

Figure 6(b). The H/V for SBG and ULG

0.1

1

10

0.1 1 10 100

HOG3yHOG3x

0.1

1

10

0.1 1 10 100

LRG11yLRG11x

0.1

1

10

0.1 1 10 100

LRG12yLRG12x

0.1

1

10

0.1 1 10 100

LRG13yLRG13x

0.1

1

10

0.1 1 10 100

LRG14yLRG14x

0.1

1

10

0.1 1 10 100

LRG6yLRG6x

1

10

100

0.1 1 10 100

LRG7yLRG7x

0.1

1

10

0.1 1 10 100

LRG8yLRG8x

0.1

1

10

0.1 1 10 100

LRG9yLRG9x

0.1

1

10

0.1 1 10 100

LRG10yLRG10x

0.1

1

10

0.1 1 10 100

HOG2yHOG2x

0.1

1

10

0.1 1 10 100

LRG1yLRG1x

0.1

1

10

0.1 1 10 100

LRG2yLRG2x

0.1

1

10

0.1 1 10 100

LRG3yLRG3x

0.1

1

10

0.1 1 10 100

LRG4yLRG4x

0.1

1

10

0.1 1 10 100

LRG5yLRG5x

0.1

1

10

0.1 1 10 100

HOG1yHOG1x

Frequency in Hz

Am

plifi

catio

n Fa

ctor

Frequency in Hz Frequency in Hz

Frequency in Hz

Frequency in Hz

Frequency in Hz

Am

plifi

catio

n Fa

ctor

Sites in Mountain Campus of L’Aquila University (LRG)

Sites in Hospital Area (HOG)

Figure 6(c). The H/V for LRG and HOG

Yutaka Nakamura

18

the ground fluidization or subsidence. The Kg values on the mountain area (LRG) are larger than that of other area, especially in the northern side. A leaning bust near the point LRG1 might be suffered by damaged ground. To-tally in the L’Aquila area the surface ground dam-age would be caused in the area with Kg > 10. 6. Discussion Figure 8 shows a comparison between the spectral ratio H/V of the strong motion records (PGA was around 650 Gal) and that of the microtremor at the observatory AQV. Obviously, both the predominant frequency and the amplification factor of the strong motion are sifted lower and larger, respectively. It seems that this caused by non-linearity of the surface ground. According to the results of surface layer investigation shown in Figure 9 [6], the surface ground is estimated as the sediment of silty soil of around 50m depth on limestone. However the amplitude of strong motion displacement is around 5cm derived from the double in-tegration of the acceleration records, and the shear strain will be as-sessed as 1/1000 approximately, then a non-linearity shown in Figure8 may not be occurred. Assuming the shear wave velocity (Vs) of the basement ground as 600 m/s regarding to the layer investigation, AVS = 110 m/s and h = 9 m for the surface ground are estimated from the H/V spectrum of the

0.1

1

10

0.1 1 10 100

G6xG6y

0.1

1

10

0.1 1 10 100

G5xG5y

0.1

1

10

0.1 1 10 100

ARG5xARG5y

0.1

1

10

0.1 1 10 100

ARG6xARG6y

0.1

1

10

0.1 1 10 100

ARG7xARG7y

0.1

1

10

0.1 1 10 100

ARG8xARG8y

0.1

1

10

0.1 1 10 100

RHGxRHGy

0.1

1

10

0.1 1 10 100

ARG1xARG1y

0.1

1

10

0.1 1 10 100

ARG2xARG2y

0.1

1

10

0.1 1 10 100

ARG3xARG3y

0.1

1

10

0.1 1 10 100

ARG4xARG4y

0.1

1

10

0.1 1 10 100

G1xG1y

0.1

1

10

0.1 1 10 100

G2xG2y

0.1

1

10

0.1 1 10 100

G3xG3y

0.1

1

10

0.1 1 10 100

G4xG4y

0.1

1

10

0.1 1 10 100

CMGWxCMGWy

0.1

1

10

0.1 1 10 100

G5xG5y

Frequency in Hz Frequency in Hz

Frequency in Hz

Frequency in Hz

Am

plifi

catio

n Fa

ctor

Am

plifi

catio

n Fa

ctor

Line EW Old and New around Colosseum

Am

plifi

catio

n Fa

ctor

Line NS Old and New around Colloseum

West side the Foot of the Column of Marcs Aurelius

Figure 6(d). The H/V for sites in Rome

The H/V Technique and Example of its Application for L’Aquila and Rome Areas 19

microtremor. Al-though the layer with the compa-rable depth can be seen in Figure 9, Vs is around 400 m/s and is not similar to this. Assuming that the estimation re-sult from the mi-crotremor meas-urement is cor-rect, if the depth of the surface layer is 9 m and the displacement at the ground sur-face is around 5 cm, the shear strain is esti-mated as 6/1000 (= 5 cm / 900 cm). On the other hand, because the Kg value at this site is 10, the shear strain at the ground surface can be estimated at most 6/1000 (= 10 x 600 Gal x 10-6) with assuming that the maximum ac-celeration at the basement ground is 600 Gal. Both result of the shear strain estimation are similar to each other. According to the G- curve for the clay arranged by Ishihara [7] shown in Figure 10, the rigidity of the soil will be reduced to 1/3 with such shear strain. This corresponds to the 0.58 times of the frequency shift and it is agreeable to the observed frequency change from 2.9 Hz to 1.7 Hz. Namely, the frequency sift shown in the strong motion is attributed to the surface ground 9 m thick and 110 m/s AVS.

SBG1 SBG2

SBG4

SBG3

SBG5

15

10

15

20

ULG1

ULG2

ULG3

ULG5

ULG6

ULG7

ULG8

ULG4

15

10

15

20

LRG1

LRG2

LRG3

LRG1 LRG4

LRG5

LRG6

LRG7

LRG8

LRG9

LRG10

LRG11

LRG12LRG13

LRG14

15

10

15

20

HOG1

HOG2

HOG3

15

10

15

20

AQA

AQV

15

10

15

20

G1 G2

G3G4G5

G6

RHG ARG1

ARG2 ARG3

ARG4 ARG5

ARG7

ARG8

ARG6

15

10

15

20

Figure 7. Distributions of Kg value: outside diameter of concentric circles correspond to 20 of Kg value

Yutaka Nakamura

20

The Kg value related to the earthquake damage to the surface ground is often more than 10 and it corre-sponds to the sedimentary layers less than 200 m/s of AVS. In Japan the ground of AVS 500-600 m/s is stiff enough to the basement ground. It is important to extract accurately the sedi-mentary ground with large Kg value and then to grasp the dynamic characteristics of that site. 7. Conclusion This paper describes the outline of the H/V technique developed by focusing on the multi-reflection phenomena in surface ground and shows its application with the data of the microtremor measure-ment at the workshop. This time the measurement was done only on the ground. Next time I would like to measure microtremor not only on the ground but also on structures or on buildings. The microtremor includes various valid information. I would like to keep challeng-ing to create proper analysis methods for various purposes

0.1

1

10

0.1 1 10 100

AQVxAQVy

AQV

0.1

1

10

0.1 1 10 100

H/V of Microtremor

H/V of Strong Motion

Frequency in Hz

Am

plifi

catio

n Fa

ctor

Am

plifi

catio

n Fa

ctor

Figure 8. Comparison between spectral ratio H/V of strong motion and that of microtremor at the site AQV

?

Figure 9. Geological profile of the site AQV (modified from [6])