Site Effects Assessment Using Ambient Excitations sesame european Commission – Research General Directorate Project No. Evg1-ct-2000-00026 sesame report of the wp04



Download 131.01 Kb.
Page2/4
Date conversion18.07.2018
Size131.01 Kb.
1   2   3   4

1 6 12 18 24 30 36 42




































D04.04

D16.04

D20.04



Milestones & Expected Results:

D04.04: Homogeneous Data Set of Noise and Earthquake Recordings at Many Sites

D16.04: Report Providing Comparisons of Experimentally and Theoretically Estimated Transfer Functions with the (H/V) Spectral Ratio

Ambient Vibrations Technique and Evaluation of the Applicability of the Latter in Cases of Linear or/and non-Linear Soil Behavior


D20.04: Report Including Comparisons of Damage Distribution in Modern Urban Areas with Results from (H/V) Spectral Ratio

Chapter 2 : Correlation between (H/V) ratio and seismic damage:The case of the city of Thessaloniki [by Panou A., Theodulidis N., Hatzidimitriou P, Papazachos C., Stilianidis K.]
2.1. Data Acquisition and Processing

Ambient noise measurements were performed in the downtown district of the city of Thessaloniki (Northern Greece) (upper part of Figure 1). The measurement grid (about 150mX150m) covered the historical center of the city and in total 250 measurements were carried out (lower part of Figure 1). The records were preformed from Monday to Friday either during evening period (18:00pm-22:00pm GMT with closed market) or during night period (23:00pm-05:00am GMT) as proposed by Panou et al., (2004). The equipment used was the Cityshark 24-bits recorder (Chatelain et al., 2000) coupled with a Lennartz 3D/5s velocimeter sensor. The response of the seismometer is flat to velocity between 0.2 and 50 Hz. Data analysis was focused in the frequency range between 0.2 and 20 Hz. Ohta et al., (1978) have suggested an observation record at a point at least 10-to-20 minutes long in order to get a record good enough for analysis. In this work for each site the recording system operated continuously for 20 min with a sample rate of 100 Hz.

Ambient noise data were processed in two stages. First, for each ambient noise recording, a number of windows, having a duration of 20 sec each, were selected using the ‘window selection’ module of the JSESAME software (SESAME project, 2001), in order to exclude portions with unrealistically large amplitudes or spikes, as has been also suggested by Duval et al., (2003). Using the ‘H/V processing’ module of the JSESAME software, the following steps were applied on the ambient noise data: (a) offset correction, (b) computation of Fourier spectra in all three components (E–W, N–S, UP), (c) application of a cosine taper, (d) smoothing of the Fourier amplitude spectra by a Konno-Ohmachi algorithm (Konno and Ohmachi, 1998).

For each frequency point the horizontal recording spectrum was divided by the vertical one, separately for both horizontal components, in order to detect any significant difference between the EW/V and NS/V spectral ratios. For each measurement point a spectral ratio and an estimation of the fundamental frequency for each horizontal component (foew and fons) and the corresponding H/V amplitude level (Aoew and Aons), were provided. The fundamental frequencies calculated from the EW/V ambient noise spectral ratio (foew) versus the ones obtained from NS/V ambient noise spectral ratios (fons) were found to be almost similar. Thus, hereafter in the present study the average value of the fundamental frequency, fo, is used.

2.2. Comparison with damage distribution
The city of Thessaloniki was strongly affected by the June 20, 1978 earthquake (M6.5) that occurred at an epicentral distance of about 30km. The historical center of Thessaloniki at the time of the earthquake consisted mainly of buildings of six to nine stories height and selected as a test area for comparing the results of the (H/V) ambient noise spectral ratio of 250 ‘in situ’ measurements with damage distribution (Penelis et al., 1985). Figure 2 shows the comparison of the contour map of the fundamental frequencies (upper part) and the H/V amplitude level (lower part) with the damage distribution per building square. It is qualitatively observed that there is a correlation between them.

To further investigate this correlation the building squares were split into buildings based on the material that they were build, as is showed in Figure 3. The majority of the buildings, is mainly constructed by reinforced concrete. Then only the R/C buildings were clasified by the coloured card (Figure 4). Only 22 buildings fall in red category. At Figures 5 and 6 the cost of walls repair and the structural cost of repair is given, respectively. It is clear from this figures that although some of the buildings squares were clasiied as of high or very high damage, only a few buildings belonging to them exhibited high level damage.

In 1959 the seismic code of Greece changed drastically. Thus, buildings of the center of the city separated into 2 categories, those built before and those after 1959. The vast majority the buildings in the studied area were built after 1960 (Figure 7). In Figure 8 the histogram of the year of completion of the buildings while in Figure 9 the histogram of the buildings built before and after 1959 at each damage category is, respectively, illustrated.
Finally, based on the previous information the damage distribution due to 20/6/1978 earthquake (Penelis et al., 1985) was also converted to EMS_98 (European Macroseismic Scale, 1998). Figure 10 shows the maps of the maximum amplifications of the (H/V) spectral ratio (Lower part) observed at the fundamental frequency (Upper part) of each site, on which the four grades of EMS_98 damage distribution are superposed. Despite the observed scatter, the comparison between the observed intensity of the 1978 earthquake, with the fundamental frequency (fo) and the corresponding H/V amplitude level (Ao) of ambient noise H/V spectral ratio (Figure 11) reveals a satisfactory correlation.

References

Chatelain, J.-L., Gueguen, Ph., Guillier, B., Frechet, J., Bondoux, F., Sarrault, J., Sulpice, P., and Neuville J.-M. (2000), Cityshark: A user-friendly instrument dedicated to ambient noise (microtremor) recording for site and building response studies, Seismol. Res. Lett. 71, 698-703.

Duval, A.-M., Chatelain, J.-L., Guillier. B., and the SESAME WP02 team (2003), Influence of experimental conditions on H/V determination using ambient vibrations (noise), 13th World Conference on Earthquake Engineering, 1-6 August, 2004, (Vancouver, Canada).

European Macroseismic Scale, 1998, http://www.gfz-potsdam.de/pb5/pb53/projekt/ems.

Konno, K., and Ohmachi T. (1998). Ground-motion characteristics estimated from spectral ratio between horizontal and vertical components of microtremor, Bull. Seism. Soc. Am. 88, no. 1, 228–241.

Panou A. A., Theodulidis N., Hatzidimitriou P., Savvaidis A., and Papazachos C. B. (2004). Reliability tests of horizontal-to-vertical spectral ratio based on ambient noise measurements in urban environment: The case of Thessaloniki city (Northern Greece), Pure & Applied Geophys., 2004, (in press).

Penelis G., Stylianidis K., Stavrakakis B., (1985). Statistical evaluation of the response of the buildings in the center of Thessaloniki to the earthquake of 20 June 1978, 12th regional Seminar on Earthquake Engineering.

SESAME Project 2001, Site EffectS assessment using AMbient Excitations, http://sesame-fp5.obs.ujf-grenoble.fr.

List of Figures
Figure 1: Upper part: Map of the Thessaloniki, Northern Greece. Lower part: Location of ambient noise measurements in the downtown district of the city of Thessaloniki.
Figure 2: Comparison of the fundamental frequency (Upper part: east-west) and the corresponding H/V amplitude level (Lower part: east-west) with the damage distribution during the 20/06/1978 earthquake (Penelis et al., 1985). The categories are shown with different color symbols.
Figure 3: Distribution of damage per building square against the type of structure.


Figure 4: Distribution of damage per building square against the coloured card.
Figure 5: Distribution of damage per building square against the cost of the repair in walls.
Figure 6: Distribution of damage per building square against the structural cost of the repair.
Figure 7: Distribution of damage per building square against the seismic code.
Figure 8: Histogram of the year of completion of the buildings.


Figure 9: Histogram of the buildings that corresponds at each damage category before and after 1959.
Figure 10: Comparison of the fundamental frequencies (Upper part: east-west) and the corresponding H/V amplitude level (Lower part: east-west) with recorded intensities of 1978 earthquake (EMS, 98).

Figure 11: Correlation between fundamental frequency (fo) and corresponding average amplitude (Ao), with 4 damage grades of EMS-98 scale, from the 1978 Thessaloniki earthquake.





GREECE



Figure 1



Figure 2




Figure 3


Figure 4



Figure 5




Figure 6




Figure 7



Figure 8



Figure 9





Figure 10





Figure 11



Chapter 3: Correlation between (H/V) ratio and seismic damage: The case of the city of Kalamata [by Panou A., Theodulidis, Savvaidis A.]
3.1. Data Acquisition and Processing
In August 2003 ambient noise measurements took place in the city of Kalamata (Southern Greece) (Figure 1). The grid of measurements covered almost the entire city. The recordings were preformed late evening early morning [20:00pm to 01:00am GMT] from Monday to Friday. The equipment used comprises the Cityshark recorder 24-bits recorder (Chatelain et al., 2000) coupled with a Lennartz 3D/5s velocimeter sensor. A GPS system provided the geographic position of each measurement point. In the experiment, the recording system operated continuously for 20 min. The sample rate was 100 Hz.

Ambient noise data were processed in two stages. First, for each ambient noise recording, a number of windows, having a duration of 20 sec each, were selected using the ‘winselect’ code of the JSESAME software (SESAME project, 2001), in order to exclude portions with forbiddingly large amplitudes or spikes, as has been suggested by Duval et al., (2003). To these data, using the ‘hvproc’ code of the JSESAME software (SESAME project, 2001), the following processing was made: (a) computation of Fourier spectra in all three components (E–W, N–S, UP), (b) offset correction, (c) application of a cosine taper, (d) smoothing of the Fourier spectra by the Konno-Ohmachi routine (Konno and Ohmachi, 1998). For each point the horizontal record spectrum was divided by the vertical one and the H/V spectral ratio were obtained.

The second step consists of plotting ambient noise H/V spectral ratio versus frequency. Each measurement point provides a spectral ratio and enables an estimation of the fundamental frequency (fo) and the maximum value of the ambient noise (H/V) spectral ratio amplitude level (Ao) at the site studied. By spatial interpolation between these points, we can deduce a map of resonance frequencies (fo) and a map of the maximum H/V amplitude level (Ao) observed at these fundamental frequencies. Figure 2 presents the variation with frequency of the average H/V ambient noise spectral ratio for three representative ambient noise recordings. The averaged H/V ambient noise spectral ratios, shown in Figure 2 display a range in complexity (e.g. ranging from a unique peak ratio to a series of peaks). The character of the H/V ambient noise spectral ratios might be related to sub-soil topography or/and to underlying geological formations.

Other researchers have also observed this variation. Luzón et al., (2001), studied the seismic response of flat sedimentary basins and carried out numerical experiments to determine the applicability of the H/V ambient noise spectral ratio in two different kinds of structure. They concluded that H/V ambient noise spectral ratio could, reasonably well, predict the fundamental local frequency when there is a high-impedance contrast between the sedimentary basin and the bedrock, except in the center of the basin. On the other hand H/V ambient noise spectral ratio could not be used, at least in sedimentary basins with low-impedance contrast with respect to bedrock. Cid et al., (2001), studied the seismic response of different sites of Barchelona through numerical modeling. Their numerical results compared with those obtained from ambient noise measurements and showed that H/V ambient noise spectral ratio predicts the fundamental frequency of the site only when there is a sharp shear-wave velocity interface in the soil column. Woolery and Street (2002), observed that a relatively horizontal, sharp shear-wave velocity interface in the soil column resulted in an H/V ambient noise spectral ratio with a single well-defined peak in the New Madrid seismic zone in the central United States.

Observations at sites with more than one sharp shear-wave velocity contrast and horizontally arranged soil layers resulted in at least two well-defined H/V ambient noise spectral ratio peaks. Furthermore, at sites where there were sharp shear-wave velocity contrasts in non-horizontal, near-surface soil layers, the H/V ambient noise spectra exhibited a broad-bandwidth, relatively low amplitude signal instead of a single well-defined peak. Taking into account he aforementioned, two sets of contour maps were produced; one with fundamental frequencies, fo > 1 Hz, and another with fo <1 Hz.

3.2. Comparison with Macroseismic Data
Figure 3 shows the contour maps of the fundamental frequency, fo, and of the maximum amplifications, Ao, respectively, when fo > 1 Hz, of the H/V spectral ratio observed at the of each site. On the same map the isoseismal intensities IMM from the 13/09/1986 Earthquake (Leventakis et al., 1992), is superposed. A quantitative comparison from the previous data is presented in Figures 4. A correlation between intensity and fo-Ao is clearly observed.

Figure 5 shows the contour maps of the fundamental frequency, fo, and of the maximum amplifications, Ao, respectively, when fo < 1 Hz, of the H/V spectral ratio observed at the of each site. On the same map the isoseismal intensities IMM from the 13/09/1986 Earthquake (Leventakis et al., 1992), is superposed. A quantitative comparison from the previous data is attempted in Figure 6. In this case, there is no correlation between intensity and fo-Ao.



References

Chatelain, J.-L., Gueguen, Ph., Guillier, B., Frechet, J., Bondoux, F., Sarrault, J., Sulpice, P., and Neuville J.-M. (2000), Cityshark: A user-friendly instrument dedicated to ambient noise (microtremor) recording for site and building response studies, Seismol. Res. Lett. 71, 698-703.

Duval, A.-M., Chatelain, J.-L., Guillier. B., and the SESAME WP02 team (2003), Influence of experimental conditions on H/V determination using ambient vibrations (noise), 13th World Conference on Earthquake Engineering, 1-6 August, 2004, (Vancouver, British Columbia, Canada) (Submitted).

European Macroseismic Scale, 1998, http://www.gfz-potsdam.de/pb5/pb53/projekt/ems.

Konno, K., and Ohmachi T. (1998). Ground-motion characteristics estimated from spectral ratio between horizontal and vertical components of microtremor, Bull. Seism. Soc. Am. 88, no. 1, 228–241.

Leventakis G., Lekidis V., Papaioannou Ch., Zacharopoulos S., Tsokas G. and Kiratzi A., "Equal-Intensity contour map for the city of Kalamata due to September 1986 earthquake", Proc. 1st Hellenic Conf. of Earthquake Engin. and Engin. Seismology, 1992; 2:321-330 (in Greek).

SESAME Project 2001, Site EffectS assessment using AMbient Excitations, http://sesame-fp5.obs.ujf-grenoble.fr.

List of Figures
Figure 1: Location of ambient noise measurements in the city of Kalamata, Southern Greece with the Isoseismal Intensities IMM from the 13/09/1986 Earthquake (Green line: VI, Brown line: VII, Red line: VIII, Black line: IX).
Figure 2: Plot of the average H/V spectral ratio versus frequency for three measurement points. (Red line: east-west component, Black line: north-south component).
Figure 3: Upper part: Contour map of the fundamental frequencies, when fo > 1 Hz (east-west component). The categories are shown with different color symbols. Lower part: Contour map of the H/V amplification level (Ao) at fundamental frequencies (east-west component), when fo > 1 Hz. The categories are shown with different color symbols. The Isoseismal Intensities IMM (Green line: VI, Brown line: VII, Red line: VIII, Black line: IX) from the 13/09/1986 Earthquake are also shown.
Figure 4: Upper Part: Correlation between the fundamental frequency (Upper part: east-west) and the corresponding H/V amplitude level (Ao) (Lower part: east-west component) when fo > 1 Hz, with the Isoseismal Intensities IMM (Green line: VI, Brown line: VII, Red line: VIII, Black line: IX) from the 13/09/1986 Earthquake.

Figure 5: Upper part: Contour map of the fundamental frequencies, when fo < 1 Hz (east-west component). The categories are shown with different color symbols. Lower part: Contour map of the H/V amplification level (Ao) at fundamental frequencies (east-west component), when fo < 1 Hz. The categories are shown with different color symbols. The Isoseismal Intensities IMM (Green line: VI, Brown line: VII, Red line: VIII, Black line: IX) from the 13/09/1986 Earthquake are also shown.

Figure 6: Upper Part: Correlation between the fundamental frequency (Upper part: east-west) and the corresponding H/V amplitude level (Ao) (Lower part: east-west component) when fo < 1 Hz, with the Isoseismal Intensities IMM (Green line: VI, Brown line: VII, Red line: VIII, Black line: IX) from the 13/09/1986 Earthquake.

Figure 1




Figure 2



Figure 3






Figure 4



Figure 5







Figure 6


Chapter 4: Correlation between (H/V) ratio and seismic damage: The case of the city of Rome [byFabrizio Cara, Giuseppe Di Giulio, Fabrizio Marra, Giovanna Cultrera, Andrea Tertulliani, Paola Bordoni, Luca Lenti, Giuliano Milana, Antonio Rovelli.]




















Chapter 5: Correlation between (H/V) ratio and seismic damage: The case of the city of Palermo [by Fabrizio Cara, Giuseppe Di Giulio, Giovanna Cultrera, Antonio Rovelli (1,) Riccardo Mario Azzara (2,) Maria Stella Giammarinaro, Paola Vallone (3,) Roberto D’Anna, Giuseppe Passafiume (4)]


(1) Istituto Nazionale di Geofisica e Vulcanologia, Roma,(2) Istituto Nazionale di Geofisica e Vulcanologia, Osservatorio Sismologico di Arezzo,(3) Università di Palermo – Dipartimento di Geologia,(4) Istituto Nazionale di Geofisica e Vulcanologia, Osservatorio Geofisico di Gibilmanna(Palermo)





























Chapter 6: Correlation between (H/V) ratio and seismic damage: The case of the city of Angra-do-Heroismo [by Paula TEVES-COSTA 1, 2, M. Luisa SENOS3 and Carlos S. OLIVEIRA4, Proc. 13th World Conference on Earthquake Engineering


Vancouver, B.C., Canada, Paper No. 1004, August 1-6, 2004.























Chapter 7: Correlation between (H/V) ratio and seismic damage: The case of the city of Fabriano [by Marco Pagani, Alberto Marcellini, Alberto Tento, Istituto per la Dinamica dei Processi Ambientali, CNR, Milano]



1   2   3   4


The database is protected by copyright ©hestories.info 2017
send message

    Main page