Simplified seismic scenario analysis of existing masonry buildings accounting for local site effects

Abstract

This study discusses the effects of local sites and hazard amplification on the seismic vulnerability assessment of existing masonry buildings. In this context, a rapid seismic evaluation procedure was implemented on an old masonry building stock in the historical center Galata, located in İstanbul, to determine the seismic risk priority of the built heritage. Damage scenarios were generated for all soil classes, different moment magnitudes, and source-to-site distances to obtain more accurate results for the seismic vulnerability assessment of the studied building stock. Consequently, damage distributions estimated under nine different scenarios with/without site effects were compared and illustrated in maps to discuss changes in vulnerability owing to amplification effects. In this study, by re-examining the rapid seismic evaluation procedure by including geo-hazard-based assessment, the importance of site effects on the vulnerability and risk assessment of built heritage was underlined. The proposed framework integrating field data and local site effects is believed to advance the current applications for vulnerability assessment of masonry buildings and provide an improvement in the application of rapid seismic assessment procedures with more reliable results.

Graphical abstract

Keywords

masonry buildings / seismic vulnerability assessment / cultural heritage / local site effect

1 Introduction

The seismic vulnerability assessment of built heritages in disaster areas is crucial because they are irreplaceable assets with cultural importance. Therefore, determining the ranking of the vulnerability of these buildings with a general masonry typology has become a major concern for authorities to prioritize remedial measures. In this context, many attempts have been made to obtain potential damage distributions using seismic vulnerability assessment methods for large building stocks.

Based on the approach adopted for the seismic risk analysis of the built environment, the existing methods in the literature are mainly classified into three categories: analytical, empirical, and hybrid. Analytical methods provide a comprehensive analysis of the ground motion for specific earthquake scenarios by examining the response of the structures [1,2]. Empirical methods mainly rely on the vulnerability functions derived from the statistical analysis of post-earthquake damage data [3,4], whereas hybrid methods involve both the statistical evaluation of post-earthquake building inventory and analytical estimation of seismic damage through refined structural models using both analytical and empirical approaches [5,6].

The procedures based on the analytical approach are mainly concerned with the direct physical evaluation of structures utilizing the structural response and performance of the buildings subjected to specified seismic actions. The existing analytical procedures typically require a dynamic analysis of the structures using nonlinear methods to evaluate the seismic behavior of the concerned building typology [1,7,8]. Among these procedures, the capacity spectrum-based methodologies such as HAZUS [9], ELER [10], and SELENA [11] have been introduced for regional seismic risk analysis of built environments consisting of different building typologies. Seismic vulnerability is evaluated using analytical fragility curves derived from the analysis of refined structural models developed for generic building typologies. By implementing these standardized tools for large-scale seismic risk evaluation, the potential damage of each building is estimated for a given intensity of seismic action; therefore, the seismic vulnerability distribution of the assessed buildings is obtained for the considered region.

Assessment procedures based on the empirical approach in the literature are categorized into two groups: damage probability matrices (DPMs) [12] and the vulnerability index method [3,4]. DPMs represent the conditional probability of a building to attain a damage level under the effect of a given intensity level of ground motion. The European Macroseismic Scale (EMS-98) [13], a widely used framework in European countries, provides a basis for the calculation of DPMs using vulnerability classes and damage levels defined for seismic risk assessment of buildings. The vulnerability index method was developed by incorporating the vulnerability indices formerly introduced in Ref. [4], also referred to as the Italian method, and the vulnerability classes defined in EMS-98. This method provides a damage assessment procedure utilizing a series of intrinsic characteristics of the building comprising the typology, construction details, and organization of the structural system that affects the seismic performance of the assessed building.

Additionally, a reliable damage database generated through well-organized post-earthquake field investigations has enabled refinements in vulnerability empirical models by implementing statistical approaches. With this motivation, in recent years, studies have focused on using empirical models based on scores derived from statistical analysis for both damage and usability assessment [14–18]. In one among these previous studies focusing on the estimation of the usability of buildings, by applying statistical techniques, a usability index was computed based on the structural parameters and macroseismic intensity using a newly proposed empirical seismic usability assessment model specifically for unreinforced masonry buildings [14]. The statistical model used in that study was calibrated based on the 2009 L’Aquila Earthquake database. In other studies, the model was validated using the 2002 Molise Earthquake database, and a calibrated new model was presented, including the usability probability matrices computed in terms of peak ground velocity [15,16]. A simplified survey form was also introduced in another investigation with fewer usability parameters and categories than the survey form prepared after the 2009 L’Aquila Earthquake [17]. In that study, the correlation between structural parameters and usability performance was determined. In addition, in a recent study, census data were compared with 2009 L’Aquila Earthquake data, and census-based fragility curves were developed to forecast the usability of unreinforced masonry buildings [18].

Along with the structural characteristics considered in vulnerability index-based methodologies, the soil-foundation relation is also considered in terms of the location of the building, type of foundation, field topography, and foundation base soil stiffness property. Implementation of the method for building stocks shows that the vulnerability class and the calculated vulnerability index of an individual building remarkably change depending on whether the structure rests on top of a rigid rock/stiff soil or soft soil site, thus defining a base boundary condition. As soft soils can significantly amplify ground motions corresponding to certain frequencies, the local site effects, in turn, can cause significant unforeseen damage to the structures on top, if ignored. The concept of seismic waves traveling upward through the surficial soil layers and being modified, which is referred to as the site response, provides the ability to predict the manner how particular local site conditions will impact ground motions coming from the rock below. Taken together, because the local site conditions considerably affect the dynamic response of the structures, recent research has focused on improving the existing vulnerability index-based methodologies accounting for ground motion characterization.

In a pioneering study, a simple and reliable period-dependent site amplification methodology, specific to regional applications, including the contribution of the site effect as an increment to the conventional macroseismic intensity factor, was proposed [19]. Following this reference study, emphasis was placed on the evaluation of site effects in seismic risk analysis based on empirical methodologies. Recent case studies have enabled to define local amplification values for different soil categories and topographic classes in the time and frequency domains to modify the mean damage degree and macroseismic intensity, respectively, by comparing the corresponding local response of soil with respect to engineering rock [20–22]. Further diversity in the results is also provided considering different scenarios for earthquake magnitude, source-to-site distance, and attenuation approaches, thus providing a comprehensive seismic-induced damage assessment for old masonry building stocks. The site-specific response approach is beneficial for quantifying ground motion intensity to provide a more realistic geo-hazard-based vulnerability and risk assessment, beyond a rough seismic zonation approach.

The aim of this study is to determine the seismic vulnerability of old masonry buildings with the integration of local site effects and to discuss the contribution of geo-hazard-based assessment to the seismic vulnerability assessment of buildings. With this motivation, preliminary work was conducted for the historical study region in Galata, İstanbul, to evaluate the seismic vulnerability of masonry buildings using the vulnerability index method considering the effect of local sites. The vulnerability indices were calculated for 40 historical masonry buildings and changes in the vulnerability indices induced by local site effects were compared. Damage distributions under different scenarios were acquired using the mean damage grade obtained for each building depending on the vulnerability indices, and mapping was performed in terms of damage grades to provide the seismic priority distribution of the buildings in the studied district.

2 Methods

The methodology used in this study is divided into two sections. The first section provides a brief overview of the rapid vulnerability assessment method for masonry buildings, and the second section considers the site effects and describes the methodology for geotechnical hazard amplification to be considered in the seismic evaluation procedure.

2.1 Rapid seismic vulnerability assessment

Many methodologies have been proposed for large-scale seismic vulnerability assessments to determine the priority risk level of buildings in a particular region under a specific intensity. In this study, the vulnerability index method developed by the Italian National Group for Defense from Earthquakes (GNDT) was used, and a rapid seismic assessment procedure was implemented for masonry building stocks [23]. The vulnerability index method uses parameters affecting the vulnerability of buildings, and the seismic capacity is expressed in terms of a vulnerability index. Using the proposed seismic evaluation form presented in Tab.1, consisting of 15 vulnerability parameters, information about the structural system is gathered from each masonry building in the stock. Each vulnerability parameter corresponds to a vulnerability score (s) and weight (w), as listed in Tab.1. Vulnerability scores range from A to D denoting the lowest to the highest vulnerability class of each parameter, respectively, and the weight ranges from 0.25 to 1.5 based on the importance of each parameter on the vulnerability. Subsequently, as given in Eq. (1), a vulnerability index (I_v) is calculated by obtaining the sum of the multiplication of the vulnerability scores (s_i) with the corresponding weights (w_i).

(1)I_{\mathrm{v}}= {\sum }_{i=1}^{15}{s}_i{w}_i.

Accordingly, the normalized vulnerability index (V_I) is defined as follows.

(2)V_{\mathrm{I}}= \left[\frac{I_{\mathrm{v}}-\left({\displaystyle {\sum }_{i=1}^{15}{s}_{\mathrm{m}\mathrm{i}\mathrm{n}}.w_i}\right)}{{\displaystyle {\sum }_{i=1}^{15}\left[\left(s_{\mathrm{m}\mathrm{a}\mathrm{x}}.w_i\right)-\left(s_{\mathrm{m}\mathrm{i}\mathrm{n}}.w_i\right)\right]}}\right].

In this study, a macroseismic approach in which the vulnerability of the buildings is presented by a model that provides the probable damage level as a function of intensity was used [24,25]. Macroseismic intensity is a widely used parameter that describes the seismic input related to vulnerability methods using the observed damage. Given a macroseismic scenario, the vulnerability curves link the intensity to the damage grades varying between D0–D5 which are explained by the mean damage grade ({\mu }_{\mathrm{D}}). {\mu }_{\mathrm{D}} is calculated using the analytical expression given by Eq. (3), in terms of the macroseismic intensity (I) detailed in EMS-98, normalized vulnerability index (V_I), and ductility index (Q) proposed as 2.3 [26].

(3){\mu }_{\mathrm{D}}= 2.5\left[1+\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{h}\left(\frac{I+6.25V_{\mathrm{I}}-13.1}{Q}\right)\right].

Subsequently, {\mu }_{\mathrm{D}} values calculated using Eq. (3) for each building are rounded to the next integer to identify five discrete damage thresholds (D0–D5). Herein, D0, D1, D2, D3, D4, and D5 denote no damage, slight damage, moderate damage, significant damage, severe damage, and collapse, corresponding to the mean damage grade intervals of 0 < {\mu }_{\mathrm{D}} < 0.5, 0.5 ≤ {\mu }_{\mathrm{D}} < 1.5, 1.5 ≤ {\mu }_{\mathrm{D}} < 2.5, 2.5 ≤ {\mu }_{\mathrm{D}} < 3.5, 3.5 ≤ {\mu }_{\mathrm{D}} < 4.5, and 4.5 ≤ {\mu }_{\mathrm{D}} < 5.0, respectively.

To obtain I in terms of EMS-98, the attenuation relationship given in Eq. (4) based on the moment magnitude (M_w) and source-to-site distance (D) is used [20].

(4)I=1.45M_{\mathrm{w}}- 2.46\mathrm{l}\mathrm{n}D+8.166.

2.2 Geo-hazard-based assessment

The local hazard conditions should be determined based on the seismicity of the region for more accurate damage distribution predictions. Seismicity parameters that directly affect structural performance under earthquake loads must be evaluated according to the geotechnical site characterization of the related field. The Turkish Seismic Code, TBSC-2018 [27], proposes assessing seismic characterization with geographical coordinate-based site-specific parameters through a seismic hazard map. This map considers the fault groups, earthquake characteristics of the region and source to fault distance for a specific location.

In the seismic code, the design spectral response accelerations (S_DS and S_D1) are calculated for a 5% damping ratio as follows.

(5)S_{\mathrm{D}\mathrm{S}}=S_{\mathrm{S}}\times F_{\mathrm{S}},

(6)S_{\mathrm{D}1} =S_1\times F_1,

where S_s and S₁ denote the spectral acceleration factors at t = 0.2 s (short period) and t = 1.0 s periods, respectively, and F_s and F₁ denote the local soil effect factors. The spectral accelerations, S_s and S₁, are originally defined with respect to a specific reference soil condition with an average shear velocity (V_S)₃₀ = 760 m/s for the upper 30 m of the profile. Therefore, these spectral accelerations for soil conditions different from the reference must be modified by multiplying them with the corresponding local soil impact factors F_s and F₁ according to the soil classes specified in TBSC-2018, as for ZA (V_S)₃₀ > 1500 m/s, ZB (V_S)₃₀ = 760−1500 m/s, ZC (V_S)₃₀ = 360−760 m/s, ZD (V_S)₃₀ = 180−360 m/s, and ZE (V_S)₃₀ < 180 m/s. Therefore, the local soil conditions and the ground motion level of the geographic location directly affect the seismic behavior of the buildings.

The geotechnical site conditions are implemented using the vulnerability index method using a local site amplification factor (f_ag) determined for each soil class. The local site amplification factor is defined as the ratio of the elastic design spectrum value obtained for a soil class (S_ae(soil)) to the generated elastic design spectrum value for the reference bedrock ZA (S_ae(rock)) as in Eq. (7).

(7)f_{\mathrm{a}\mathrm{g}}=\frac{S_{\mathrm{a}\mathrm{e}\left(\mathrm{s}\mathrm{o}\mathrm{i}\mathrm{l}\right)}}{S_{\mathrm{a}\mathrm{e}\left(\mathrm{r}\mathrm{o}\mathrm{c}\mathrm{k}\right)}}.

Once f_ag is determined, to define the macroseismic intensities including the site effects corresponding to different magnitude level earthquakes at different source-to-site distances, the seismic intensity increase parameter, ΔI, is calculated using Eq. (8), where coefficient C₂ is estimated to be 1.82 [20].

(8)\mathrm{\Delta }I=\frac{\mathrm{l}\mathrm{n}{f}_{\mathrm{a}\mathrm{g}}}{\mathrm{l}\mathrm{n}{C}_2}.

As the intensity difference is calculated, the increase in the seismic vulnerability index, ΔV, is obtained using Eq. (9).

(9)\mathrm{\Delta }V=\frac{\mathrm{\Delta }I}{6.25}.

2.3 Seismicity of the region

The studied region, Galata, a historical district of İstanbul hosting mid-rise brick masonry structures, is located close to the North Anatolian Fault Zone (NAFZ) that produces large earthquakes. The general seismotectonic characteristics of the site are outlined by defining the segmented fault system assigned to the NAFZ. Historical earthquakes of different magnitudes that occurred around the Marmara Sea and Istanbul dating back to the 1500 s are examined with source-to-site distances less than approximately 70 km, also including the intensity levels [28–32].

The west end of the NAFZ ruptured as a strike-slip fault during the Kocaeli and Düzce earthquakes in 1999, forming a fault segment known as the İzmit segment. The northern strand of the NAFZ then continues beneath the Marmara Sea, forming relatively short, discontinuous, offshore segments, such as Prince’s Islands and Çınarcık, located in the Central and Çınarcık submarine basins known as the North Marmara Fault Systems (NMFS). Finally, the NMFS connects to the strike-slip Ganos Fault at further west, completing the segmented faulting system that dominates the seismicity of central Istanbul. Because Galata is located a few tens of kilometers north of the main NMFS system, the historical earthquakes were chosen among the seismic events having both great (M_w≥ 7) and small magnitudes (M_w < 7) that occurred on the defined NAFZ segments and NMFS zone.

Based on these evaluations, for the damage scenario analysis, the following historical earthquakes were selected for the case study: the February 12, 1991, earthquake (M_w = 4.8, Prince’s Islands segment); September 26, 2019, earthquake (M_w = 5.7, Central Basin); and August 17, 1999, earthquake (M_w = 7.4, İzmit segment). These scenarios are also associated with macroseismic intensity data obtained from USGS ShakeMaps [33]. The selected seismic sources enable to define the seismic behavior of the studied region of the NAFZ, comprehensively including different magnitudes, source-to-site distances, intensities, and mechanisms to be used for a quantitative approach during damage assessment.

3 Results and discussion

In this study, 40 historical brick masonry buildings located in Galata were examined using the vulnerability index method for seismic vulnerability assessments. The studied buildings, with a maximum of six stories, were designed and constructed adjacently, representing a common structural typology of the region. Based on the seismicity of the study region, the seismic risk analysis of the inspected building aggregate was performed under nine different damage scenarios considering three different moment magnitudes (M_w = 5, 6, and 7) and source-to-site distances (D = 10, 20, and 30 km). The effect of local site conditions was incorporated in the seismic vulnerability analysis by evaluating the local site amplification factors for different types of soil classes defined in TBSC-2018 [27]. Estimating the increase in macroseismic intensities and vulnerability indices induced by local site amplification, the seismic vulnerability distribution, and expected damage grades of the inspected buildings were determined considering the damage scenarios generated for the study region.

The local site amplification factors were determined based on the horizontal design elastic spectra generated for each generic soil class and for the reference bedrock (ZA–ZE) shown in Fig.1, considering the design ground motion level with a return period of 475 years (probability of exceedance 10% in 50 years). The observed altered behavior of the spectrum enables the evaluation of the response of individual soil classes with reference to the bedrock in terms of the design spectral accelerations. Clearly, the maximum design spectral acceleration (S_DS), and therefore the highest local site amplification factor (f_ag), was obtained for soil class ZC. Because the natural vibration period of the masonry buildings belonging to the structural aggregate, as obtained by structural analysis, was in the range of 0.13–0.30 s, soil class ZC was determined as that involving the most intense spectral accelerations.

Tab.2 also presents the increase in the macroseismic intensities determined using Eqs. (8) and (9) using the amplification factors for different soil classes, where the expected damage scenarios and probable damage distributions obtained for soil class ZA refer to the reference seismic vulnerability analysis without local site effects in this study.

Tab.3 summarizes the macroseismic intensities estimated for the set of scenarios generated using the seismic attenuation relationship, as shown in Eq. (4), and the increased intensities induced by the local site amplification factors. Because of the discrete nature of macroseismic intensities in the EMS-98 scale, calculated increment ΔI = 0.189 for soil class ZB in Tab.2 resulted in the same intensity values estimated for soil class ZA corresponding to the reference rock-site condition. For soil classes ZC, ZD, and ZE, the increments owing to local site amplifications caused alterations in the EMS-98 macroseismic intensities.

The vulnerability index of each building was calculated based on an evaluation of 15 parameters representing the structural vulnerability of the assessed building. The parameters related to the structural characteristics of the buildings were compiled from plan views of the buildings [34]. Subsequently, the distribution of the estimated vulnerability index values for the entire building aggregate was obtained, as shown in Fig.2. Furthermore, because the vulnerability index was also affected by the change in the macroseismic intensities owing to local site amplification, the vulnerability index distribution incorporating the effect of local site conditions was determined for the buildings. As shown in Fig.2, the results of the analysis, including the site effects, were derived for soil class ZC as the most critical soil class, with the highest difference in intensity and vulnerability index among the soil classes considered in this study.

Based on the results of the analysis performed without local site effects, calculated vulnerability index values were distributed within a range of 0.16–0.42, and the mean index was estimated as 0.25. However, considering the local site effects in the analysis, vulnerability index varied in the range of 0.27–0.53, and the mean value of distribution was calculated as 0.35. Additionally, the increase in the vulnerability index obtained for each building by considering site amplification effects was also observed from the spatial vulnerability index distribution of the buildings determined with and without site effects, as shown in Fig.3.

The expected damage distributions obtained for the seismic scenarios generated for each soil class with variable magnitudes and source-to-site distances are shown in Fig.4. To derive the damage distributions, as shown in Fig.4, the mean damage grades calculated for all the buildings were evaluated according to the {\mu }_{\mathrm{D}} intervals specified in theory, and then, the buildings were sorted according to the discrete damage grade categories (D0–D5) corresponding to the {\mu }_{\mathrm{D}} interval. According to the methodology implemented in this study, the extent of the damage that the building might suffer under a potential seismic action was mainly associated with the building vulnerability index and macroseismic intensity expressed in terms of the moment magnitude and source-to-site distance. Because the macroseismic intensity was affected by the site amplification factor determined for each soil class, the impact of the site conditions on the expected damage grades of the buildings was clearly observed for each scenario.

Among the damage scenarios estimated for different levels of seismic hazard, the changes in the results concerning a source-to-site distance of 20 km with and without site effects were compared using the spatial distributions of the buildings according to the damage grades estimated for each moment magnitude, as shown in Fig.5.

To sum up the results represented in Fig.5, in the analysis without the local site effect for M_w = 5, 97.5% of the entire building stock was expected to suffer no damage (D0), whereas 2.5% attained a slight damage grade (D1). However, considering the local site effects in the analysis for the same magnitude, the damage grades of the entire building stock were shifted to D1 and D2 with the same distribution ratios. Analysis without site effects for M_w = 6 showed that 65% of the buildings had no damage (D0), whereas the rest of the buildings were slightly damaged (D1). Considering the local site effects in the same scenario, 52.5%, 45%, and 2.5% of the buildings attained slight damage (D1), moderate damage (D2), and significant damage (D3) grades, respectively. Additionally, in the case of the highest moment magnitude M_w = 7, without the local site effect, the majority of the buildings (80%) were moderately damaged (D2), whereas 7.5% and 12.5% of the buildings were slightly (D1) and significantly (D3) damaged, respectively. However, when the site effects were considered for M_w = 7, 45% and 55% of the buildings reached significant damage (D3) and severe damage (D4) grades, respectively.

4 Conclusions

This study aims to reveal the local site effect on seismic vulnerability evaluation by integrating the geo-hazard-based assessment in the vulnerability index-based approach for a refined procedure that provides more accurate damage distributions. With this motivation, this study examined the seismic vulnerability of masonry buildings located in the Galata region, a historical district of İstanbul, under nine scenarios, considering all soil classes, different moment magnitudes, and source-to-site distances. Comprehensive results obtained by the seismic vulnerability assessment combined with the geo-hazard assessment proved that the accuracy of the damage distribution of the studied masonry buildings was enhanced. Therefore, for a realistic seismic-induced damage assessment of old masonry building stocks, the effect of local soil properties along with the seismic source characterization should be considered in the analysis.

In conclusion, this investigation, applying different potential scenarios considering site effects, would be deemed a valuable contribution toward more realistic damage estimation with its findings. Proving the necessity of including geo-hazard-based assessment in the procedure, this study is believed to assist both researchers and local authorities in defining pre- and post-earthquake preparation strategies and policies for the mitigation of seismic risk.

References

[1] Barbat A H, Pujades L G, Lantada N. Seismic damage evaluation in urban areas using the capacity spectrum method: Application to Barcelona. Soil Dynamics and Earthquake Engineering, 2007, 28(10−11): 851–865

[2] Silva V, Crowley H, Varum H, Pinho R, Sousa R. Evaluation of analytical methodologies used to derive vulnerability functions. Earthquake Engineering & Structural Dynamics, 2014, 43(2): 181–204

[3] Lagomarsino S, Giovinazzi S. Macroseismic and mechanical models for the vulnerability and damage assessment of current buildings. Bulletin of Earthquake Engineering, 2006, 4(4): 415–443

[4] Benedetti D, Petrini V. On the seismic vulnerability of masonry buildings: An evaluation method. L’industria delle Costruzioni, 1984, 149: 66–74

[5] Kappos A, Panagopoulos G, Panagiotopoulos C, Penelis G. A hybrid method for the vulnerability assessment of R/C and URM buildings. Bulletin of Earthquake Engineering, 2006, 4(4): 391–413

[6] CavaleriLDi Trapani FFerrottoM F. A new hybrid procedure for the definition of seismic vulnerability in Mediterranean cross-border urban areas. Natural Hazards, 2017, 86(Sup 2): 517−541

[7] Crowley H, Pinho R, Bommer J J. A probabilistic displacement-based vulnerability assessment procedure for earthquake loss estimation. Bulletin of Earthquake Engineering, 2004, 2(2): 173–219

[8] D’Ayala D, Speranza E. Definition of collapse mechanisms and seismic vulnerability of historic masonry buildings. Earthquake Spectra, 2003, 19(3): 479–509

[9] FederalEmergency Management Agency (FEMA). HAZUS-MH: Multi-Hazard Loss Estimation Methodology Earthquake Model. Washington, D.C.: FEMA, 2012

[10] Erdik M, Sesetyan K, Demircioglu M, Hancilar U, Zulfikar C, Cakti E, Kamer Y, Yenidogan C, Tuzun C, Cagnan Z, Harmandar E. Rapid earthquake hazard and loss assessment for Euro-Mediterranean Region. Acta Geophysica, 2010, 58: 855–892

[11] Molina S, Lang D H, Lindholm C D. SELENA—An open-source tool for seismic risk and loss assessment using a logic tree computation procedure. Computers & Geosciences, 2010, 36(3): 257–269

[12] Dolce M, Kappos A, Masi A, Penelis G, Vona M. Vulnerability assessment and earthquake damage scenarios of the building stock of Potenza (Southern Italy) using Italian and Greek methodologies. Engineering Structures, 2006, 28(3): 357–371

[13] GrünthalG. European Macroseismic Scale EMS-98. Luxembourg: European Center for Geodynamics and Seismology, 1998

[14] Zucconi M, Sorrentino L, Ferlito R. Principal component analysis for a seismic usability model of unreinforced masonry buildings. Soil Dynamics and Earthquake Engineering, 2017, 96: 64–75

[15] ZucconiMFerlito RSorrentinoL. Verification of a usability model for unreinforced masonry buildings with data from the 2002 Molise, Earthquake. In: Proceedings of the International Masonry Society Conferences and the 10th International Masonry Conference. Milan: IMC, 2018

[16] Zucconi M, Ferlito R, Sorrentino L. Validation and extension of a statistical usability model for unreinforced masonry buildings with different ground motion intensity measures. Bulletin of Earthquake Engineering, 2020, 18(2): 767–795

[17] Zucconi M, Ferlito R, Sorrentino L. Simplified survey form of unreinforced masonry buildings calibrated on data from the 2009 L’Aquila earthquake. Bulletin of Earthquake Engineering, 2018, 16(7): 2877–2911

[18] Zucconi M, di Ludovico M, Sorrentino L. Census-based typological usability fragility curves for Italian unreinforced masonry buildings. Bulletin of Earthquake Engineering, 2022, 20(8): 4097–4116

[19] Giovinazzi S. Geotechnical hazard representation for seismic risk analysis. Bulletin of the New Zealand Society for Earthquake Engineering, 2009, 42(3): 221–234

[20] ChieffoNFormisano A. The influence of geo-hazard effects on the physical vulnerability assessment of the built heritage: An application in a district of Naples. Buildings, 2019, 9(1): 26

[21] ChieffoNFormisano A. Geo-hazard-based approach for the estimation of seismic vulnerability and damage scenarios of the old city of Senerchia (Avellino, Italy). Geosciences, 2019, 9(2): 59

[22] Chieffo N, Formisano A. Induced seismic-site effects on the vulnerability assessment of a historical centre in the Molise Region of Italy: Analysis method and real behaviour calibration based on 2002 Earthquake. Geosciences, 2020, 10(1): 21

[23] Formisano A, Florio G, Landolfo R, Mazzolani F M. Numerical calibration of an easy method for seismic behaviour assessment on large scale of masonry building aggregates. Advances in Engineering Software, 2015, 80: 116–138

[24] GiovinazziSLagomarsino S. A macroseismic method for the vulnerability assessment of buildings. In: Proceedings of 13th World Conference on Earthquake Engineering. Vancouver: Venue West Conference Services Ltd., 2004

[25] GiovinazziS. The vulnerability assessment and the damage scenario in seismic risk analysis. Dissertation for the Doctoral Degree. Braunschweig: Technical University of Braunschweig, 2005

[26] GiovinazziSLagomarsino SPampaninS. Vulnerability methods and damage scenario for seismic risk analysis as support to retrofit strategies: An European perspective. In: Proceedings of the NZSEE 2006 New Zealand Society for Earthquake Engineering Annual Conference. Napier: NZSEE, 2006

[27] TBSC-2018. Turkey Building Seismic Code. Ankara: AFAD, 2018 (in Turkish)

[28] Ambraseys N N, Jackson J A. Seismicity of the Sea of Marmara (Turkey) since 1500. Geophysical Journal International, 2000, 141(3): F1–F6

[29] Ambraseys N N. The earthquake of 1509 in the Sea of Marmara, Turkey, Revisited. Bulletin of the Seismological Society of America, 2001, 91(6): 1397–1416

[30] Erdik M, Demircioğlu M, Şeşetyan K, Durukal E, Siyahi B. Earthquake hazard in Marmara Region, Turkey. Soil Dynamics and Earthquake Engineering, 2004, 24(8): 605–631

[31] Armijo R, Meyer B, Navarro S, King G, Barka A. Asymmetric slip partitioning in the Sea of Marmara pull-apart: A clue to propagation processes of the North Anatolian Fault. Terra Nova, 2002, 14(2): 80–86

[32] Parsons T. Recalculated probability of M > 7 earthquakes beneath the Sea of Marmara, Turkey. Journal of Geophysical Research, 2004, 109(B5): B05304

[33] U.S. Geological Survey. Earthquake Lists, Maps, and Statistics. 2020. Available at the website of U.S. Geological Survey

[34] ÖncelA D. Apartment: A new kind of housing in Galata. İstanbul: Kitap Yayınevi, 2010 (in Turkish)

RIGHTS & PERMISSIONS

Higher Education Press