EERI SPECIAL EARTHQUAKE REPORT -DECEMBER 1997

2 EERI SPECIAL EARTHQUAKE REPORT -DECEMBER' 997

The epicenter of the first tions at Colfiorito for the first earthquake was reportedly near shock were 0.38g and O 44g the village of Colfiorito. The horizontal and 0.38g venical; and earthquake sequence started on at 0.56g clnd 0.33g September 26, 1997, with a shock horizontal, 0.20g vertical. The of magnitude of Ml = 5.5/Mw = 5.7 peak ground accelerations for the at 2:33 AM local time (epicenter second shock at Colfiorit J were 43°01.75'N 12°51.08'EI. A larger 0.28g and 0.17g horizontal, and earthquake, ML = 5.8/Mw = 6.0, oc- 0.31 g vertical; and at Nccera curred about 10:40 AM local time, Umbra 0.56g and 0.50g hori- again with the epicenter at zontal, and 0.42g vertical.2 Colfiorito. Geotechnical Framework The epi- These earthquakes were part of an central area is in the Ape nnines, ongoing sequence of shocks. which run generally nortll-south in Colfiorito was the epicenter for a central . The general area ML=5.1/Mw=5.3 event on Octo- consists of folded and thrusted ber 3, 1997, and again on October limestones, cherty limestones, 7 (ML=5.31. On October 12 the marly limestones, marly I:lays, and epicenter shifted to where marls which are part of t1e a Mo =4.5 event occurred, fol- - unit. The forma- lowed by a ML = 5.4/Mw= 5.7 tions, which outcrop in elongated event on October 14. The most occurrences, date from t 1e Upper Figure 3 -This buflding in Nocera recently reported earthquake was a Triassic-lower Miocene clnd consti- Umbra illustrates typical damage to Mo=4.5 event on November 9, tute the bedrock. The m()re recent stone buildings in the region. 1997, with the epicenter in the formations are Upper Pleistocene- (photo: Tobriner) Sellano- area. Holocene materials. The" are discontinuous to the pre11iously Ground Motions: The as yet noted material and occur typically the mountain range. In the broad unofficial peak ground accelera- in the topographic low troughs of epicentral area the geologic forma- tions are cut by a number of faults of general NNW-SSE direction. These are parallel to the main direction of the formations. There is a second system of faults with a transverse strike and important horizontal component of movement.

Several geotechnical phenomena occurred. There were several surface ruptures in the epicentral area and in villages near the epicenter. The ruptures have a general orientation NNW-SSE and display a vertical displacement of from 2 to 15 cm.

A number of rock falls occurred as a result of the earthquake. Many of these caused minor temporary disruption of traffic on roads, such Figure 4 -In , significant cracks had opened in the Piazza Inferiore as the closure of highway 209 on di S. Francesco, indicating earthquake-caused subsidence. The cracks the slope of Mt. Galloro. Near had been covered with plastic (foreground}. In this photo, soldiers are , authorities were afraid an beginning to string cables to stabilize the arcade to the right in the photo. entire hillside, the site of a (photo: Comerio) hermitage and a beautiful forest,

3 EERI SPECIAL EARTHQUAKE REPORT- DECEMBER 1997

4 EERI SPECIAL EARTHQUAKE REPORT -DECEMBER 1997

Similarly, hospitals continued to and properties. They WEre con- For the approximately 900 private treat newly injured people over a cerned about vandals arid deteri- buildings damaged in Assisi, the long period. This was particularly orating conditions, and 1:hey were value of the damage was esti- difficult in Foligno because the afraid that once they m()ved from mated to be about $1 billion. main hospital was damaged and the tents, the more perr1anent Extrapolating this to the roughly was operating from a temporary prefabricated housing would not be 9,000 private buildings damaged in Red Cross facility. At the same given to them. the region, the value of damage is time, buildings had to be re- approximately $10 billion, not inspected and re-evaluated for General Damage counting the cost to repair public safety multiple times, and relief buildings and churches. workers from the Protezione Civile Inspection statistics as ()f October could not ask people to go back 14, 1997, were made a vailable to into their homes until they felt the reconnaissance tean1. In 65 Past Mitigation Efforts secure that the aftershocks were towns (communes) in U rnbria, diminishing. This was not the case 30,985 private building~i had been The patina of age and our delight until the end of October, more inspected. Of thsse, 8,:: 96 were with the picturesque medieval than six weeks after the first structurally damaged and unin- nature of Umbria can be deceptive. earthquake. habitable. This is roughl'1 the Even the most famous monuments equivalent of the Ameri(:an "red should not be regarded as having Temporary Housing: About 13,000 tag" buildings. The Italic ns use a been built in a single campaign and people were housed in tents and rating system with 5 ca':egories: embalmed as perfect examples of small camper trailers. That number A = inhabitable their time. They have been altered is about 1 2 percent of the B = inhabitable wit 1 some and adapted, repaired and re- 110,000 inhabitants in 15 provisions shaped. When the church of S. communes (city/county areas). The C = partially usable Chiara threatened to collapse in government is building 28 sites for D = needs re-revie~v 1351 , flying buttresses were prefabricated housing, as people E = unusable. added to save it. They might be cannot stay in the tents and missed if they were now removed, trailers in winter. At the end of Public buildings, school~, and but the buttresses were added as October, the services for these churches were inspectelj and an ad hoc expediency. Similarly in sites were not complete and the tallied separately. Data 'IVere the old town of Nocera Umbra, central government would not available for 18-22 tOWIIS in the brick buttresses and abutments, allow the prefab homes to be region, and are described in the now beautifully weathered and inhabited without services. They table below. It is import3nt to note textured, were added to brace attempted to move the tent that about 24 percent o F the public buildings which were at risk. dwellers to hotels temporarily, as buildings and 1 7 percen t of the the weather turned cold and it schools were seriously

Damage Observed: The area of major shaking was a rural region in the mountains. There were some newer one- and two-story concrete frame buildings, but no damage to them was observed.

The buildings were typically resi- dential structures, some with commercial uses on the first floor . The observed damage was limited to cracking between different materials, i.e. the concrete and the infill hollow clay tile (HCT). The damage was minor and in no case was out-of-plane movement of the HCT infill noted.

In in the Marche Region, several four- to six-story structures Figure 5- Exterior anchors and tie bars on masonry buildings c~rea in public housing projects (repre- distinctive feature of Assisi vernacular architecture. These eally senting 300-500 units) had this mitigation efforts demonstrated their effectiveness in this quaJ~e. In this type of damage and the buildings photo, the wall on the right, with anchors and tie bars, is intat:t and had been vacated (Figure 6) .One plumb, while the un-anchored wall on the left is cracked and l'owing of the vacated buildings in Fabri- outward. (photo:Tobriner) ano had a vertical crack in a column. This might have been a pending spall. At a beam-column More modern mitigation is often Typical modern concret9 construc- joint, the reinforcing steel was hard to observe or evaluate. Many tion in the area consist~; of a exposed and spalling occurred. of the roofs of churches in Umbria concrete frame with a rnasonry have been strengthened by the use infill of hollow clay tile :HCT), In Foligno, we observed an eight- of reinforced concrete, a practice called terra cotta. The ~ICT was story concrete hotel building. We which is presently being debated in visible on the exterior Vlralls. In noted one column that had a verti- Italy. In the Basilica of S. Fran- many cases the bare H(:T was the cal crack at the second floor line. cesco, a damped steel beam was exterior finish. There WoiS no This crack may have been a pend- added to the nave side walls to seismic separation or joint between ing spall. It did not appear to be prevent cracking in earthquakes. the HCT and the concrEIte frame. due to deformation of the beam- Some masonry structures were We were unable to obsl~rve any column joint. In addition, minor retrofitted with reinforced concrete interior wall construction. spalls due to pounding were noted. foundations, ring beams of rein- Minor pounding was noted be- forced brick or concrete, and steel There were a number of industrial tween other concrete buildings in reinforced apertures. type buildings with con [;rete the built-up urban areas. frames and walls. Thes,~ buildings were constructed of prl!-cast Most of the damaged concrete Engineered Structures girders, roof and wall p 3nels. Other buildings appeared to have been industrial buildings werl! constructed in the 19605 or early In the affected zone there were no constructed of concretet frames 19705. No information was high-rise buildings. Most tall with walls of hollow clciy tile infifl . obtained on the type of joint and structures were mid-rise, in the These buildings were ol1e and two column detailing used in that era. four- to nine-story range. All large, stories in height. In the general No shear wall buildings were multi-story buildings were con- observation of these buildings no noted. structed of reinforced concrete. No damage was observed ;Ind none large, multi-story steel frame was reported to us. Thl!re is one Most of the concrete buildings buildings or wood buildings were steel frame building in ~;poleto, but were a significant distance from located in the area principally no reports of damage Vllere the epicenter. The magnitude and affected by the earthquake. received. frequency of the earthquake

6

EERI SPECIAL EARTHQUAKE REPORT -DECEMBER 1997

8

~ EERI SPECIAL EARTHQUAKE REPORT -DECEMBER 1997 hall church upon which a second more lofty hall church was built.

This Italian Gothic building is puzzling in many respects, perhaps the most curious being its buttressing system. The nave is buttressed by cylindrical towers and these are buttressed in turn by @- flying buttresses. The combination of the stubby cylinders and the elegant flyers represents two different buttressing strategies and aesthetics.

The roof of the church is sup- ported by purlins which rest on stone arches. Below the roof is a series of quadripartite rib vaults ~ which cover the nave. On the J~~ ,/ walls of the nave are a cycle of " " ./'. " " paintings by Giotto which are ---"--"-- " '.,. considered a turning point in ".. western art from Byzantine iconic ~ " imagery to more naturalistic "" ",;' " " ---" ,-i representation. The main (eastern) --7",I, " 1.-'; facade and transept facades are "" ",~ planar, with few projections, and :::::::1 ", I are capped by simple triangular .\",; . ,"" ; " II " '/' " II II *- ,..II II " II pediments. .~- ",~ ..~c ~ ' " I I m.~ '", " " , " The church had been seismically " "- ~ I& retrofitted. After an earthquake in -~--») ":'-~,~-p:1fJ 1984, prominent cracks appeared O 60 feet IIi r I I I I in the walls of the upper church O 20 meterB threatening the Giotto frescos. The LowerChu~h seismic retrofit included a damped Figure 9 -The upper and lower churches of S. Francesco in Assisi steel beam inserted along the Diagram key: (1) Collapsed vault adjacent to entrance portal (where four length of the nave below the people were killed). (2) Collapsed vault at crossing. (3) Major cracks in clerestory windows and above the vault. (4) Cracked pediment, now braced by steel cage. frescos. The beam was designed only to restrain out-of-plane bending in earthquakes and otherwise to "give" with the building. The roof had also -been quadrant of the vault immediately s. Chiara: (St. Clare) The design retrofitted with concrete purlins. in front of the crossing, the sup- for this Italian Gothic church in porting cross vault, and the web- Assisi (built between 1255 and The worst damage occurred at the bing of the east quadrant of the 1265) was based on the Basilica ceiling and roof levels. The crossing vault, all fell (Figure 11) . of S. Francesco. St. Clare was a webbing or infill of the east follower of St. Francis and estab- quadrant or cell of the quadri- Large cracks appeared in the lished the women's Franciscan partite vault adjoining facade broke centers of the nave vaults which Order, the Poor Clares. The in the second earthquake (Figure still survive. The south transept aisleless hall church exhibits the 10) .On the vault was a priceless pediment cracked. The question is same damage pattern as S. Fran- Cimabue fresco which crashed to whether the roof and side wall cesco, only less severe. There is a the floor killing four people. Down retrofits helped the structure or crack between facade and nave the nave, the webbing of the contributed to the local failures. arch. They have separated from

9

EERI SPECIAL EARTHQUAKE REPORT -DECEMBER 1997 cracked. Vaults in interior rooms on tourist dollars. "strengthening" attempts in their were damaged, frescos on the ceil- BOO-plus year lifetimes. rt is ing of the refectory fell, and The picturesque quality of the hill important to ask whether, over diagonal cracks appeared in cells towns is one of the key reasons such a long period, earthquakes on the southeastern side of the tourists come to the region, yet have contributed to a degradation complex. S. Pietro's Benedictine the decisions about how and when of materials and the structure's abbey had severe diagonal cracks to repair properties will tie either ability to resist earthquakes. and was perhaps subsiding down ( 1) in the hands of outsiders (in the Questions about the maintenance the hillside. case of the hill towns), (2) in the of the buildings also playa role. hands of the Catholic Church (in deciding a priority for restoration Extensive intrusions by engineers Housing, Social, and Cultural of monuments), or (3) in the hands may not be warranted in monu- Impacts of government officials attempting mental buildings. Although current to provide housing solutions for a thinking suggests that historic In categorizing the series of poor and rural population (in the buildings be strengthened to a moderate earthquakes which case of the rural villages). In fact, achieve a greater level of struck Umbria and the Marche the short-term decisions to protection than regu'ar buildings, in terms of lessons relocate roads around damaged the damage to churches in Umbria learned, the series might well be villages and to build prefabricated suggests intrusions should be kept considered one of" architectural housing in the flat agricultural to a minimum, and buildings impacts" earthquakes rather than areas will hurt the limited tourist should be allowed to crack and be one which had significant engi- economies in the small villages, as repaired. Although there is much neering impacts and lessons. What these settlements will lose both research to be done on analyzing is unusual and special about this access to local business and the the causes of the vault failure in S. disaster is the substantial social, charm of village life. Thus, the Francesco, or the separation of cultural, and economic impacts of long-term secondary economic roofs and facades in other what would normally be impacts of the earthquakes could churches, it seems clear that considered a very moderate series be much greater than the initial heavy engineered interventions will of earthquakes. Tourism and costs of building repair and not be compatible with historic religious pilgrimage are the primary restoration of cultural monuments. materials and building forms. While economy, not only in Assisi, but in it is clear that efforts to preserve the entire region. Locals estimate the priceless art work on walls and Important Issues and that 5 million people per year visit ceilings must be undertaken, the Assisi. With virtually all the major Recommendations work must be done with concern churches closed to the public for for whether the retrofit and some time into the foreseeable In Italy. modern design standards construction will harm the future, there will clearly be a are clearly comparable to those in paintings in the long run. slowdown in visitors even if the United States. and so it is not hotels, restaurants. and other surprising that the team found very Ultimately, the Umbria-Marche private facilities are repaired and limited evidence of significant earthquakes are significant in that re-opened . damage in engineered structures. they represent the considerable Similarly. stone vernacular build- impact that moderate earthquakes Hill towns such as Assisi and ings behaved much the same as do can have on important cultural Nocera Umbra are reported to vernacular structures around the properties and on the long term house only 500-600 local inhabi- world. failing due to inadequate economic well-being of the region. tants. Property values are high anchorage of walls to roofs. with Umbria without its churches and {with real estate estimated at heavy roofs and chimneys pushing its hill towns has very little $400 per sQuare foot), and houses outward. In these cases. simple economic raison d'etre. Solutions in town are owned as second and traditional retrofit techniques to housing a poor and rural homes by foreigners. It is not clear do work for life safety. Most population and to restoring the how Quickly these owners will act damaged buildings can be repaired. monuments need to be seen as to repair their properties, but it is but not upgraded significantly. parts of a whole. The churches clear that a long-term slowdown in need the villages and the villages tourism will hurt the locals who By contrast, monumental struc- need the monuments to restore the live in more modern structures in tures, particularly churches, have economic health of the region. new flatland developments, but been through a number of whose jobs and businesses depend earthquakes and a number of

11 EERI SPECIAL EARTHQUAKE REPORT -DECEMBER 1997

1 We gratefully acknowledge the Voices from the Past assistance of our Italian colleagues Prof. Antonio Giuffr~, Prof. Luigi Decanini, Ing. Alessandro Palumbo, Estimating Seismic Risk in Italian Hill- Top Arch. Caterina Carocci, Prof . Towns, Circa 1930 Corrado Latina, Prof. Luigia Binda, Ing. Antonio Pugliese, Dr. Mario Serio, and Prof. Latina's former "The [talian Earthquake of July 23, 1930, Notes of a brief inspection students Nicola Lapi, Raffaella by John R. Freeman, Civil Engineer" as published in Earthquake Michelangeli, and Emanuela Damage and Earthquake Insurance, John Ripley Freeman, McGraw- Mollica. Special thanks are also Hill, New York, 1932. due to Signora Conti Capellor , Frate Nicola, Frate George, Frate The loss of life in this earthquake as stated in the early newspaper Bob of the convent of S. reports was so appalling (upward of 1,800) that the writer, then in Francesco, Father Fabbris don Carlo, Parroco di S. P~tro, Frate Germany on a tour of hydraulic laboratories, visited the scene as soon Angelo of S. Chiara, Frate Guiseppi as practicable, in order to learn whatever lessonsthis disaster might Bertuzzi, and Prelate Vittorio Viola present, as to safeguardsfor the information of American builders. of the convent of S. Maria degli Angeli. In this quake of July, 1930, the serious wreckage was almost wholly confined to ten closely-built hill-top communities. In general, the 2 We had no geotechnical expert damage outside this narrow zone, 40 miles in length, of hill-top on our team and are grateful to communities was relatively extremely small. The shockswere strongly Professor Luigi Decanini of the felt in Naples and caused great alarm. A few weak walls and cornices Dipartimento di ingegneria strut- fell in and near Naples and killed perhaps 5 or 10 people. The press turale e geotecnica, Universit~ dispatches of the next few days expresseda state of mind rather than degli studi di Roma "La Sapienza" the real extent of damage. The writer saw no outward signs of damage for providing us with the time of while motoring over many miles of city streets in and around Naples to the seismic events and their Pozzuoli, Castellammare, Sorrento, Amalfi, and easterly Avellino. unofficial peak ground accelera- tions. The geotechnical framework Although the newspaper reports indicated that damagewas serious all section of the report is based on a through the Neapolitan district, the writer judged that the total paper submitted to EERI by Dr. E. structural damage within 25 miles of Naples was less than 1/20th of 1 Lekkas, Assistant Professor , per cent of the sound value within this region. University of Athens, and Dr. J. Fountoulis, Lecturer, Faculty of ...A second lesson from these old records is that a very rough Science, Department of Geology, estimate, made by dividing the number of recorded destructive quakes University of Athens. in about 700 years by the number of village-groups damaged by one quake, indicates that to anyone of the hundreds of hill-top communities the chance of disaster in anyone particular year is of the order of hardly more than one chance in five hundred and that to any one particular house in anyone year, the chance of destruction is of the order of perhaps not more than one in five thousand, taking all these Italian hill-cities into the average. It would be well worth while to an economist, while considering the apparent urgent need of rebuilding these houses constructed of weak rubble masonry, to confirm by careful research the above rough estimate of chance of damage. Estimates of that kind help toward obtaining earthquake insurance of a satisfactory kind at reasonable cost.

The recurrence of damage in Melfi after the lapse of 80 years, and also the recurrence at a few other localities in Italy, illustrates that the chance of damage may be much larger for certain localities, now fairly Figure 12 -Typical damage to well defined by the record of 2,000 years, than that roughly estimated vernacular architecture in Umbria- above for Italy as a whole. Marche, Italy. (photo: Tobriner}

12