2017 Ischia Earthquake: Macroscale Typological and Damage Assessment of Masonry Churches

Available online at www.eccomasproceedia.org Eccomas Proceedia COMPDYN (2019) 1482-1500

COMPDYN 2019 7th ECCOMAS Thematic Conference on Computational Methods in Structural Dynamics and Earthquake Engineering M. Papadrakakis, M. Fragiadakis (eds.) Crete, Greece, 24–26 June 2019

2017 ISCHIA EARTHQUAKE: MACROSCALE TYPOLOGICAL AND DAMAGE ASSESSMENT OF MASONRY CHURCHES

Claudia Casapulla1, Francesca Ceroni2, Antonio Formisano3, Piera Salzano2*, Andrea Prota4 1 Department of Structures for Engineering and Architecture, University of Napoli Federico II Via Forno vecchio, 36, 80134 - Napoli, Italy [email protected]

2 Engineering Department, University of Napoli ‘Parthenope’ Centro Direzionale is. C4, 80143, Napoli, Italy {francesca.ceroni, piera.salzano}@uniparthenope.it

3Department of Structures for Engineering and Architecture, University of Napoli Federico II Piazzale Tecchio, 80, 80125, Napoli, Italy [email protected]

4Department of Structures for Engineering and Architecture, University of Napoli Federico II Via Claudio, 21, 80125, Napoli, Italy [email protected]

Abstract

st On August 21 2017 an earthquake of magnitude Mw = 3.9 (Md = 4) occurred in Ischia Is- land (Napoli, Italy), causing numerous damages to ordinary and monumental buildings. In September 2017 teams from the University of Napoli Federico II carried out usability checks on churches. The damage evaluation of churches was performed by filling the II level survey form (A-DC). The surveys allowed constructing a significant database of 27 churches. A detailed analysis of the inspected churches was performed in order to identify: i) the percentage of usability; ii) the most recurrent structural typologies; iii) the most recurrent damages and failure mechanisms. Based on the collected information, statistical inquires were carried out for the inspected churches as a prodromal step for subsequent vulnerability analysis.

Keywords: Churches, Seismic Vulnerability, Risk Mitigation, Cultural Heritage, Structural Damage.

1482 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

1 INTRODUCTION st On August 21 2017, an earthquake of magnitude Md = 4.0 struck the Ischia Island with epicenter in Casamicciola Terme. This event caused two fatalities and many injured people. Despite the low magnitude, the earthquake produced significant damages to masonry and re- inforced concrete (RC) buildings. The event in fact showed many deficiencies of the building stock in Casamicciola Terme and Lacco Ameno municipalities[1][2]. Above all, the churches of the island have shown numerous damages. Past studies [3]-[12], indeed, highlighted that monumental buildings, in particular churches, are considerably vulnerable to dynamic actions due to their intrinsic vulnerabilities (open plan, greater height-to-width ratio, projecting parts, large openings, slender bell towers, absence of proper transversal connections). Ischia Island is located in the Gulf of Napoli, about 30 kilometres far from the city of Na- poli (Italy). It is a volcanic island with a surface area of 46.3 square kilometres. The island is very densely populated, with around 60,000 residents. There are six municipalities on the island: Ischia, Barano d’Ischia, Casamicciola Terme, Forio, Lacco Ameno, and Serrara Fontana. The island has always been historically characterized by a significant seismicity, especially in Casamicciola Terme area. The intrinsic seismicity is linked to the nature of the island. Is- chia is, indeed, a volcanic island that has been frequently characterized by low-magnitude earthquakes (located in the northern part of the island, mostly beneath the town of Casamicci- ola Terme at very shallow depths). In 1881 and 1883, Casamicciola Terme was hit by two of the most devastating earthquakes of the seismic history of the Ischia Island. Table 1 summarizes the main events that shocked the island since the XIII century, with the corresponding moment magnitude and maximum macro-seismic intensity recorded according to MCS scale [15]. Figure 1 shows a map of the epicentres of the historical seismic events that hit Ischia Island since the XIII century.

Year Municipality Lat Lon Mw IMCS,max 1275 Ischia 40.743 13.942 4.01 VIII-IX Ischia Not 1557 40.721 13.953 3.5 Available Casamicciola 1762 40.746 13.909 3.5 VI-VII Terme Barano d’Ischia Not 1767 40.735 13.919 3.5 Available Casamicciola 1796 40.746 13.909 3.88 VIII Terme Casamicciola 1828 40.745 13.899 4.01 IX Terme Casamicciola 1841 40.749 13.899 3.25 VI Terme Casamicciola 1863 40.746 13.909 2.87 V Terme Casamicciola 1867 40.746 13.909 2.99 V-VI Terme Casamicciola 1881 40.747 13.895 4.14 IX Terme Casamicciola 1883 40.744 13.885 4.26 X Terme 1980 Serrara Fontana 40.718 13.89 4.37 V Table 1: Historical seismic events in Ischia (http://comunicazione.ingv.it/).

1483 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

1841 1881 1863-1867-1762 1275 1883 1828

2017 1767

1557 1980

Figure 1: Historical seismic events in Ischia Island.

2 SEISMIC EVENT On August 21st, 2017 at 18:57:51 UTC (20:57:51 local), an earthquake of a duration magnitude Md = 4 struck the Ischia Island with the epicentre in Casamicciola Terme (40.74° lati- tude and 13.90° longitude). The earthquake produced significant damages, especially to churches and masonry buildings [1].

2.1 Macro-seismic intensity surveys Immediately after the earthquake, macro-seismic surveys [16], [17] were performed by the emergency group QUEST (QUICK Earthquake Survey Team) of INGV, in collaboration with ENEA, aimed at assessing damages to buildings and estimating the macro-seismic intensities according to the European EMS scale [18]. According to the surveys, a “red zone” was outlined in the district of Casamicciola Terme, close to the epicentre. In this area, most of the buildings were made of masonry, with absence of reinforcing elements (i.e. tie rods). The most commonly used material was tuff, a soft volcanic stone typical of Ischia subsoil and, more in general, of the whole Campania region. Moreover, an overall high vulnerability was found for old masonry buildings. This was at- tributable not only to the poor quality of masonry used, but also to significant structural changes performed in the buildings over the time. Figure 2 reports some examples of the damage observed on ordinary masonry buildings in the red zone of Casamicciola Terme [19]. Table 2 shows the list of investigated locations by the QUEST group and the assigned macro-seismic intensities according to the European EMS scale [18]. The macro-seismic characteristics of the 21st August 2017 earthquake show a concentra- tion of damages in the epicentre area with a very strong intensity attenuation, even at very small distances. This is a typical result of shallow earthquakes in volcanic areas (generally associated to low magnitude earthquakes). Furthermore, the localized distribution of damage within the red zone highlights a possible

1484 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota site effect in the hilly area of Casamicciola Terme, in agreement with what was observed after the 1883 destructive earthquake [16].

a) b)

c) d) Figure 2: Damages recorded in the red zone of Casamicciola Terme (Ischia, Napoli) immediately after the Au- gust 21st 2017 event [19]: a) façade overturning in Via Ottringolo, Piazza Maio; b) partial collapse of a masonry building in Via Montecito, Piazza Maio; c) corner overturning in Via Montecito, Piazza Maio; d) shear cracks on masonry external walls in Via Montecito, Piazza Maio.

Location IEMS Location IEMS Casamicciola Terme (red zone) VIII Forio IV-V Bagni VII Perrone IV-V Fango VII Piedimonte IV-V Marina di Casamicciola VI Buonopane IV Fontana V-VI Ischia IV Cretaio V Panza IV Lacco Ameno V Sant'Antuono IV Ciglio V Campagnano III-IV Fiaiano V Cuotto III-IV Monterone V San Domenico III Serrara V Sant’Angelo III Barano IV-V Succhivo III Table 2: EMS intensity for different areas of Ischia Island [16], [17].

1485 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

3 INSPECTIONS Immediately after the earthquake, several teams from University of Napoli Federico II carried out, under the coordination of the Ministry of Cultural Heritage (MiBACT), the usability checks on churches and monumental buildings, whose aim was to define provisional safety measures for the safeguard of buildings and their artistic content. A total of 27 surveys were performed on churches in the most affected areas of the island. The most damaged municipalities were: Casamicciola (6 inspected churches), Forio (17 inspected churches) and Lacco Ameno (4 inspected churches). Figure 3 shows the map of the island with the epicentre of the earthquake and all the inspected churches. In order to perform the inspections, the A-DC survey form [20] was adopted. The main aim of this form is to directly provide during the emergency phase usability checks for heritage buildings hit by an earthquake. The application of this form is described in Section 4.

001 SMBC Santa Maria del Buon Consiglio 015 SFA San Francesco d'Assisi 002 SMP Santa Maria della Pietà 016 SMS Santa Maria del Soccorso 003 MSI Santuario Maria SS Immacolata 017 SMAP San Michele Arcangelo (del Purgatorio) 004 SAP Sant'Antonio di Padova 018 SSFP Santuario San Francesco di Paola 005 SPB San Pasquale Baylon 019 SV San Vito 006 SMMP Santa Maria Maddalena Penitente 020 CSSA Congrega SS Annunziata 007 SG San Gennaro 021 SRPM San Rocco Pio Monte S. Anna 008 SCB San Carlo Borromeo 022 SSA Santissima Annunziata 009 SLE San Leonardo 023 CDA Congrega dell'Assunta 010 SFS San Francesco Saverio 024 BSR Basilica Santa Restituta 011 SML Santa Maria di Loreto 025 SMA San Michele Arcangelo 012 SMLO Santa Maria di Loreto - Oratorio 026 SLU Santa Lucia 013 SS San Sebastiano 027 SD San Domenico 014 SGT San Gaetano

Figure 3: Inspections performed in churches of Ischia Island after the August 21st 2017 event.

1486 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

3.1 Typological characterization of inspected churches The inspected churches were classified according to a set of significant typological information, in order to properly analyze the typological features and the related sources of vulnerability. A total of 27 surveys were performed on churches in three Ischia municipalities: Casamic- ciola (22%), Forio (63%) and Lacco Ameno (15%). Figure 4a reports the distribution of inspected churches subdivided by municipalities, showing that most of the churches (63%) were located in Forio municipality, quite far from the epicentre. Figure 4b, on the other hand, reports the distribution of churches according to the EMS macro-seismic intensity assigned [16], [17]. It is worth noting that a consistent part (78%) of the inspected churches fall in the class IEMS ≤ V, meaning that most of the churches were affected by slight damages. This is consistent with the damage recorded overall on the island, since it was noticed that a very strong attenuation effect was found even at small distances from the epicentre. The remaining churches were affected by higher values of macro- seismic intensities, even if only 2 churches were located close to the red zone and have IEMS = VIII. In order to define homogeneous structural classes related to the seismic vulnerability, the inspected churches were examined according to plan shape, type of façade, type of bell tower, volume of the construction and type of masonry. Six types of plan shape were considered as shown in Table 3. A first subdivision is performed according to the number of naves, then a further classifica- tion is carried out considering the presence of apse and/or transept. A circular (central) plan church class is also considered. The distribution reported in Figure 4 is summarized in Table 3. According to this subdivision, the most common typology within the inspected churches is a one-nave church with apse (16 out of 27, i.e. 59% of the total). A typical example is the ‘Santa Maria della Pietà’ church, located in Casamicciola Terme (Figure 5).

a) b) Figure 4: Distribution of inspected churches for a) belonging municipalities and b) EMS macro-seismic intensity recorded [16], [17].

1487 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

1 2 3 4 5 6

Plan type Church type No of churches % 1 1 3.7% One-Nave churches 2 16 59.3% 3* 3 11.1% 4** 2 7.4% Three-Nave churches 6 3 11.1% Circular plan 5 1 3.7% Not available 1 3.7%

Total 27 100% * 1 church with timber-framed structure ** 1 church with iron-framed masonry structure Table 3: Distribution of plan shapes.

a) b) c) Figure 5: Santa Maria della Pietà: a) façade; b) nave; c) bell gable.

It is important to note that among classes 3 and 4, two particular structural typologies have been found: a timber-framed masonry church (Santuario Maria SS Immacolata in Casamiccio- la Terme) and mixed timber-framed and iron-framed masonry churches (Santa Maria Madda- lena Penitente in Casamicciola Terme). Figure 6b reports the timber frame of Santuario Maria SS Immacolata church. This structural typology was significantly widespread in the past, especially in seismic regions like Calabria and Campania that suffered some catastrophic events, given its adequacy to resist earthquake. The combination of two different materials (wood and stones), indeed, ensured a good performance under seismic actions. Figure 7b,c report both the timber and iron frames found in Santa Maria Maddalena church, which is a unique example of mixed iron-masonry and timber-masonry structures in the national construction heritage. The church was totally rebuilt after its collapse under the strong 1883 earthquake. The presbytery and apse areas contain the timber-framed structure, while the nave and the transept are characterized by an iron-framed structure, both embracing the masonry structure made of local green and yellow tuff stones.

1488 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

a) b) Figure 6: Santuario Maria SS Immacolata: a) façade, b) timber frame.

c) d) e)

Figure 7: Santa Maria Maddalena Penitente: a) façade, b) iron frame, c) timber frame.

The second subdivision of the collected database considers the façade type (Figure 8). In particular, three types of façade were identified: salient (Figure 8b), quadrangular/rectangular (Figure 8c), and gabled shaped (Figure 8d). The most common façade has a quadrangular/rectangular shape (56%), followed by a typical gabled shape (37%). As for the bell tower, bell gable (Figure 9a) is found in most of the churches (44%), as reported in Figure 9. The information about the bell tower is not available in only 1 case (4%) while in the remaining cases (52%) a bell tower is present. Typically, the bell tower is partially or completely integrated (Figure 9c, d) within different portions of the church (about 52%, 37% integrated bell tower and 15% partially integrated bell tower). About the masonry types encountered during the inspections, all churches were made of stone masonry, mostly of great dimensions and good quality, such as tuff (Figure 10a). More- over, poor anti-seismic protection devices, such as tie rods (Figure 10b), were found. The size and the geometrical characteristics of the inspected churches were also taken into account. In Figure 11, the mean plan surface encountered is reported for each church plan shape. Nevertheless, it is important to note that, according to Table 3, only class 2 is significantly populated of a sufficient amount of churches. Table 4 summarizes the mean dimensions found for each plan shape. It can be concluded that the typical church in Ischia Island is a one-nave church with apse, a quadrangular or polygonal façade with a bell gable. The common height is 8 m with a surface of 120 m2. The identification of a typical church could be a useful tool in the evaluation of detailed vulnerability analysis.

1489 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

b) c) d) Figure 8: a) Statistical distribution of façade typology. Examples of: b) salient façade in San Leonardo church (Forio); c) quadrangular or polygonal façade in ‘San Francesco d’Assisi’ church (Forio); d) gabled shape façade in ‘Santa Maria della Pietà’ church (Casamicciola Terme).

b) c) d) Figure 9: a) Statistical distribution of bell tower typology. Examples of: b) bell gable in ‘San Carlo Borromeo’ church (Forio); c) partially integrated bell tower in ‘San Francesco di Paola’ sanctuary (Forio); d) integrated bell tower in ‘Congrega SS. Annunziata’ church (Forio).

1490 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

a) b) Figure 10: a) tuff masonry in Santa Maria Maddalena Penitente church in Casamicciola Terme; b) tie rods in the façade of Santa Lucia church in Forio.

400

350 Plan shape Average dimensions ]

2 2 300 type [m ]

250 1 50 2-3-5-6 100 200 4 350 150

100 Table 4: Mean dimension for each plan shape. Mean plan [m surfaceplan Mean 50

0 01234567 Plan shape Figure 11: Mean plan surface for each church plan shape.

4 DAMAGES RECORDED

4.1 A-DC form During the emergency phases, the A-DC survey form [20] is aimed at providing usability checks for heritage buildings affected by an earthquake [21]. Immediately after the 2017 Is- chia earthquake, several inspections were carried out on churches. A total of 27 surveys were performed on churches in the most damaged municipalities (Casamicciola Terme, Lacco Ameno and Forio). Figure 12a shows the usability outcomes distribution for the inspected churches of the database. Within such inspections, 56% of the structures resulted safe, 33% safe with precautions and the remaining 11% resulted unsafe. Figure 12b shows the distribution of usability outcomes as a function of the macro-seismic intensity observed. It is clear that for increasing values of macro-seismic intensity there is an increase of unsafe conditions. Moreover, the number of inspected churches, N, for each value of macro-seismic intensity, IEMS, is reported in the graphs. It is worth noting that most of the inspected churches were characterized by IEMS = V, about 50% of them resulted safe and about 40% were safe with precautions. Only 2 churched were unsafe.

1491 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

N=5 N=16 N=3 N=1 N=2

11%

Safe Safe with precautions 33% 56% Unsafe

b) Figure 12: a) Usability outcomes for all inspected churches; b) usability outcomes as a function of macro-seismic intensity.

4.2 Damage index The A-DC survey form [20] analyses 28 damage mechanisms (i.e. the façade, the colon- nade, the vaults, the chapels, the apse, the transept, the dome and the bell tower) in order to verify their activation and the level of damage reached (D1-D5), defined according to the Eu- ropean macro-seismic scale [18]. The survey form also provides a formulation for evaluating a global damage index according to the following equation:

(1)

where n is the number of activated mechanisms and dk is the level of damage recorded for each mechanism that varies from 0 to 5. The obtained values of damage index id for the whole population of analyzed churches are shown in Figure 13. For all 27 analyzed churches of Ischia Island, the mean value of damage index is 0.12, with maximum values of 0.30 in only three cases. Most of the churches are, indeed, characterized by a low damage index, lower than 0.25. Figure 14 shows the correlation between the damage index and the macro-seismic intensity according to ESM scale [18]. About 80% of churches (i.e. 21 out of 27) are characterized by IEMS ≤ V. For higher macro-seismic intensity the database is not sufficiently populated for making reliable statistical consideration. Nevertheless, it is also clear from Figure 14 that two churches characterized by the highest macro-seismic intensity (i.e. VIII) present the highest damage index.

1492 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

Figure 13: Damage indexes for the 27 inspected churches in Ischia.

Figure 14: Correlation between the damage index and the macro-seismic intensity.

4.3 Activated mechanisms The A-DC survey form [20] analyses the possibility of activating 28 mechanisms. The listed mechanisms are reported in Table 55. M1-Overturning of the façade is one of the most studied out-of-plane mechanisms [22], [23] (see Figure 15).

1493 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

M1 Overturning of the façade Façade M16 Overturning of the apses Apses M2 Overturning of the gable M17 Shear failure in the apses M3 Shear mechanism in the fa- and presbytery walls çade M4 Porch and narthex. M18 Vaults of the apses and of M5 Transversal vibration of the Naves the presbytery nave M6 Shear mechanism in the M19 Hammering and damage in Roof nave lateral walls the nave roof M7 Longitudinal vibration of the M20 Hammering and damage in central nave the transept roof M8 Vaults of the central nave M21 Hammering and damage in the apses roof M9 Vaults of the lateral naves M22 Overturning of the chapels Chapels walls M10 Overturning of the transept Transept M23 Shear failure in the chapel façade walls M11 Shear failure in the transept M24 Collapse mechanism in the walls chapel vaults M12 Vaults of the transept M25 Interaction between elements of different behaviour M13 Kinematic chain in the tri- Triumphal M26 Overturning of the standing Bell tower umphal arches arches out elements M14 Collapse of the dome and Dome and M27 Global collapse of the bell the Tiburio Tiburio tower M15 Collapse mechanism of the M28 Mechanism in the bell cell lantern Table 5: Damage mechanisms and macro-elements defined in A-DC survey form [20].

b) Figure 15: Mechanism M1, overturning of the façade: a) scheme, and b) real case in San Sebastiano church (SS).

Figure 16 presents the correlation between the possible and activated mechanisms in the inspected churches of Ischia. As is clear from the figure, some mechanisms present a higher percentage of occurrence. In particular, M3 and M6 (shear mechanism in the façade, Figure 17a, b, and shear mechanism in the nave lateral walls, Figure 18a, b, respectively) are the most frequent in the surveyed churches, characterized by a highest mean damage. Those mechanisms were activated in over 50% of churches.

1494 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

Figure 16: Percentage of possible and activated mechanisms in the whole database.

a) b) Figure 17: M3, shear mechanism in the façade: a) scheme, and b) real case in San Pasquale Baylon church (SPB).

a) b) Figure 18: M6, Shear mechanism in the nave lateral walls: a) scheme, and b) real case in San Michele Arcangelo (del Purgatorio) church (SMAP).

1495 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

High occurrence, though lower, is observed for M1 (Figure 15a), M13 and M17 mechanisms (overturning of the façade, kinematic chain in the triumphal arches and shear failure in the apses and presbytery walls, respectively, Figure 19a, b), that were activated in over 30% of churches.

a) b)

Figure 19: a) M13, kinematic chain in the triumphal arches and b) M17, shear failure of the apses and presbytery walls.

4.4 Damage probability matrices (DPMs) Damage Probability Matrices (DPMs) based on statistical elaborations of the inspected churches damage are also herein implemented. DPMs are aimed at performing vulnerability analysis since they provide direct relationship between the observed damage (id) and the seismic action in terms of macro-seismic intensity (IEMS). Actually, this kind of relationship is quite reliable for ordinary buildings, since using the EMS scale different classes of vulnerability are inherently correlated to the damage levels. For the churches, instead, vulnerability classes are not provided by EMS scale [18] and there are some difficulties in defining them without specific inspections, meaning that the application of the EMS scale can be inaccurate. However, a possible way is to define homogeneous groups of macro-seismic intensities and to transform the damage index id into a discrete variable, associating it to the level of damage dk, as reported in Table 6 (similarly to [24]). Figure 20 reports the DPMs obtained for different intervals of macro-seismic intensities. It has to be noted that the database is not sufficiently populated to properly fill all the macro- seismic intervals. In fact, only for IEMS=V the data interval is more substantial. Nevertheless, it is clear from Figure 16 that churches characterized by a low macro-seismic intensity (i.e. IEMS ≤ VI), suffered low damage (damage level D0), while for higher intensities, higher levels of damage (damage levels D1-D2, corresponding to a damage index id = 0.2÷0.3) are reached.

Level of damage dk D0 D1 D2 D3 D4 D5 0 1 2 3 4 5 Damage index id 0÷0.1 0.1÷0.25 0.25÷0.4 0.4÷0.6 0.6÷0.8 0.8÷1

Table 6: Correlation between level of damage dk and damage index id.

1496 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

IEMS ≤ IV - N=5 120% Data

100%

80%

60% Churches

40%

20%

0% 0 1 2 3 4 5 Damage level

IEMS =V - N=16 IEMS≥VI - N=6 60% 70% Data Data 60% 50%

50% 40%

40%

30% Churches Churches 30%

20% 20%

10% 10%

0% 0% 0 1 2 3 4 5 0 1 2 3 4 5 Damage level Damage level

b) c) Figure 20: DPMs for different values of macro-seismic intensity.

5 CONCLUSIONS The earthquake that struck the Ischia Island on August 21st 2017 caused significant damages to churches and masonry building in Forio, Lacco Ameno and Casamicciola Terme municipalities. This paper presents a general description of the event and the subsequent need of inspect- ing structures during the post-earthquake emergency phase. Numerous inspections were performed on churches and monumental buildings, with the aim of defining provisional safety measures for the safeguard of their value and artistic content. The damage evaluation was performed through the II level survey form (A-DC) filled in situ. A database of 27 churches was constructed and a detailed examination of building typologies was carried out. The main outcome of the typological analysis was to identify the most recurrent characteristics of the churches in Ischia Island. It was found that most churches have a one-nave church with apse, a quadrangular or polygonal façade, and a bell gable. The average height is about 8 meters, while the surface is around 120 m2. The identification of a typical church configuration is essential for performing further and more detailed vulnerability analysis. Consequently, a detailed analysis on the usability outcomes and the damage recorded during the inspections was carried out. During the emergency phase, most of the churches resulted safe (56%). For some churches, provisional safety measures were needed (36%), especially

1497 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

in areas where V ≤ IEMS ≤ VII was registered. The remaining churches (33%) were found to be unsafe, they being mostly characterized by higher values of macro-seismic intensity. Few significant damages were detected into inspected churches. Most churches showed damage index id < 0.25 and the mean damage recorded was equal to 0.12, consistently with the low magnitude of the earthquake. Several cases of higher damage were found in Casamic- ciola Terme close to the epicentre area, since the earthquake was very superficial with a strong intensity attenuation, even at very small distances. Finally, an accurate study of the activated damage mechanisms was carried out, showing a clear predominance of in-plane shear ones.

ACKNOWLEDGMENTS The authors acknowledge the sponsorship of the Italian Civil Protection, through the RELUIS Project - WP4: MAppe di Rischio e Scenari di danno sismico (MARS) (2019).

REFERENCES [1] C. Del Gaudio, M. Di Domenico, P. Ricci, G. M. Verderame, Preliminary prediction of damage to residential buildings following the 21st August 2017 Ischia earthquake. Bul- letin of Earthquake Engineering, 16(10), 4607-4637, 2018. [2] B. Briseghella, C. Demartino, A. Fiore, C. Nuti, C. Sulpizio, I. Vanzi, D. Lavorato, G. Fiorentino, Preliminary data and field observations of the 21st August2017 Ischia earthquake. Bulletin of Earthquake Engineering, 17, 1221-1256, 2019. [3] F. Doglioni, A. Moretti, V. Petrini, P. Angeletti, Le chiese e il terremoto. Dalla vulne- rabilità constatata nel terremoto del Friuli al miglioramento antisismico nel restauro, verso la politica di prevenzione. Trieste, Italy, Lint Editoriale Associati, 1994. (in Ital- ian) [4] S. Lagomarsino, S. Podestà, Seismic Vulnerability of ancient churches: I. Damage assessment and emergency planning. Earthquake Spectra, 20(2), 377–394, 2004. [5] S. Lagomarsino, S. Podestà, Seismic Vulnerability of ancient churches: II. Statistical analysis of surveyed data and methods for risk analysis. Earthquake Spectra, 20(2), 395–412, 2004. [6] F. da Porto, F. Silva, C. Costa, C. Modena, Macro-scale analysis of damage to churches after earthquake in Abruzzo (Italy) on April 6, 2009. Journal of Earthquake Engineer- ing, 16(6), 739-758, 2012. [7] C. Casapulla, P. Salzano, A. Sandoli, L.U. Argiento, F. Ceroni, B. Calderoni, A. Prota, Statistical analysis of the structural damage to churches affected by the 2016-17 Central Italy earthquake sequence, XVII National Conference on L’Ingegneria sismica in Italia (ANIDIS 2017), Pistoia, Italy, September 17-21, 2017. (in Italian) [8] G. De Matteis, M. Zizi, V. Corlito, Analisi preliminare degli effetti del terremoto del Cen- tro Italia del 2016 sulle chiese a una navata. XVII National Conference on L’Ingegneria sismica in Italia (ANIDIS 2017), Pistoia, Italy, September 17-21, 2017. (in Italian). [9] G. De Matteis, G. Brando, V. Corlito, E. Criber, Seismic vulnerability assessment of churches at large territorial scale: calibration of a methodology on the basis of the 2009

1498 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

L’Aquila earthquake. 3rd International Conference on Protection of Historical Con- structions (PROHITECH’17), Lisbon, Portugal, July 12 - 15, 2017. [10] G. De Matteis, E. Criber, G. Brando, Seismic vulnerability assessment of masonry churches through the application of probabilistic methods. 9th International Conference on Structural Analysis of Historical Constructions (SAHC2014), Mexico City, Mexico, October 14-17, 2014. [11] G. De Matteis, E. Criber, G. Brando, Damage probability matrices for three-nave masonry churches in Abruzzi after the 2009 L’Aquila earthquake. International Journal of Architectural Heritage, 10(2-3), 120-145, 2016. [12] S. Taffarel, M. Giaretton, F. da Porto, C. Modena, Damage and vulnerability assessment of URM buildings after the 2012 Northern Italy earthquakes. 16th International Brick and Block Masonry Conference (IBMAC 2016), Padua, Italy, June 26-30, 2016. [13] L. Krstevska, L. Tashkov, N. Naumovski, G. Florio, A. Formisano, A. Fornaro, R. Lan- dolfo, In-situ experimental testing of four historical buildings damaged during the 2009 L'Aquila earthquake. COST ACTION C26: Urban Habitat Constructions under Cata- strophic Events, Proceedings of the Final Conference, Naples, Italy, September 16-18, pp. 427-432, 2010. [14] A. Formisano, G. Vaiano, F. Fabbrocino, G. Milani, Seismic vulnerability of Italian masonry churches: The case of the Nativity of Blessed Virgin Mary in Stellata of Bondeno. Journal of Building Engineering, 20, 179-200, 2018. [15] A. Sieberg, Scala MCS (Mercalli-Cancani-Sieberg). Geologie der Erdbeben, Handbuch der Geophysik, 2(4), 552–555, 1930. [16] Gruppo di Lavoro INGV sul terremoto dell’isola di Ischia (2017). Rapporto di sintesi preliminare sul Terremoto dell’isola d’Ischia (Casamicciola) M4.0 del 21 agosto 2017 (6 settembre 2017). https://doi.org/10.5281/zenodo.886045. (in Italian) [17] R. Azzaro, S. Del Mese, G. Martini, S. Paolini, A. Screpanti, V. Verrubbi, A. Tertulliani, L. Graziani, A. Maramai, Rilievo macrosismico per il terremoto dell’isola di Ischia del 21 agosto 2017. QUEST-QUick Earthquake Survey Team – INGV, 2017. (in Italian) [18] G. Grunthal (ed.), European Macroseismic Scale 1998. Cahiers du Centre Européen de Géodynamique et de Séismologie, Luxembourg, 15, 1998. [19] C. D’Ambra, A. Prota, M. Di Ludovico, G. Manfredi, Rapporto fotografico relativo ai danni subiti da alcuni edifici a seguito dell’evento sismico del 21 agosto 2017 presso isola d’Ischia-M = 4.0, Report ReLUIS published on http://www.relui s.it/ (in Italian) [20] MiBACT, Direttiva 23 aprile 2015, Aggiornamento della direttiva 12 dicembre 2013, relativa alle “Procedure per la gestione delle attività di messa in sicurezza e salva- guardia del patrimonio culturale in caso di emergenze derivanti da calamità naturali. GU Serie Generale n.169 del 23-7-2015, 2015. (in Italian) [21] Consiglio dei Ministri, Direttiva del Presidente del Consiglio dei Ministri, 9 febbraio 2011, Linee guida per la valutazione e riduzione del rischio sismico del patrimonio culturale con riferimento alle Norme tecniche per le costruzioni di cui al decreto del Mini- stero delle infrastrutture e dei trasporti del 14 gennaio 2008. Supplemento ordinario alla Gazzetta Ufficiale, n. 47 del 26 febbraio 2011-Serie generale. Rome, Italy: Istituto Poligrafico e Zecca dello Stato, 2011. (in Italian)

1499 Claudia Casapulla, Francesca Ceroni, Antonio Formisano, Piera Salzano and Andrea Prota

[22] C. Casapulla, L. Giresini, P.B. Lourenço, Rocking and kinematic approaches for rigid block analysis of masonry walls: state of the art and recent developments. Buildings, 7, 2017. [23] L. Giresini, C. Casapulla, R. Denysiuk, J. Matos, M. Sassu, Fragility curves for free and restrained rocking masonry façades in one-sided motion. Engineering Structures, 164, 195-213, 2018. [24] S. Lagomarsino, S. Giovinazzi, Macroseismic and mechanical models for the vulnerability and damage assessment of current buildings. Bulletin of Earthquake Engineering, 4, 415, 2006.

1500