Causes of the Collapse of the Polcevera Viaduct in Genoa, Italy

applied sciences

Article Causes of the Collapse of the Polcevera Viaduct in Genoa, Italy

Janusz Rymsza

Road and Bridge Research Institute, 03-302 Warsaw, Poland; [email protected]; Tel.: +48-22-390-01-07

Abstract: The article investigates the causes of the Morandi viaduct collapse in Genoa. This three- span viaduct was a part of the A10 motorway leading to Savona. The structure design of the viaduct supports and diagonal stay cables was unique. The viaduct included three cable-stayed system, each supported a three-span continuous beam by two sets of prestressed concrete cables on the two sides of the pylon across the width. The pair of V shapes piers supported upside-down V pylons that supported the top ends of two pairs of diagonal stay cables. The stays were constructed from steel cables with prestressed concrete shells poured on. This article provides information on the technical condition of the viaduct and the way of strengthening the cables in the early 1990s. At that time, the author of the article visited the structure. In August 2018, the viaduct collapse occurred, and one of the structure supports collapsed. In June 2019, during the demolition process, the other two supports were destroyed. Since in Venezuela and Libya there are still two more bridges with a structure similar to that in Italy, the concept of strengthening the structure, as proposed by the author of the article 25 years ago, may still be useful.

Keywords: Morandi bridge; Polcevera Viaduct; corrosion; collapse; bridge; truck loads; collapse of Polcevera Viaduct; large pre-stressed bridge structures; Prof. Morandi

Citation: Rymsza, J. Causes of the Collapse of the Polcevera Viaduct in 1. Introduction Genoa, Italy. Appl. Sci. 2021, 11, 8098. 1.1. General Information https://doi.org/10.3390/app The Polcevera Viaduct in Genoa was designed by Riccardo Morandi (1902–1989), a 11178098 professor of civil engineering at the universities in Rome and Florence. Prof. Morandi had designed many civil buildings and he was an avid supporter of the use of pre-stressed Academic Editors: Daniel Dias and concrete in engineering structures, including bridges. In the 1960s and 1970s, he designed Giuseppe Lacidogna three similar pre-stressed bridges of unique structure. These were: bridge over Lake

Received: 23 July 2021 Maracaibo in Venezuela (235 m long; 1962), Viaduct in Genoa (~208 m long; opened on the Accepted: 24 August 2021 4 September in 1967), and the Wadi Kuf Bridge in Libya (282 m long; opened in 1971). Published: 31 August 2021 Extensive information about the viaduct in Genoa can be found in the study by Morandi in 1968 [1], while the design parameters of the three bridge structures designed

Publisher’s Note: MDPI stays neutral by Prof. Morandi were described by Troitsky in 1977 [2]. Prof. Morandi received many with regard to jurisdictional claims in international awards for his design of buildings and bridges, including the honorary published maps and institutional afﬁl- doctorate from the Technical University of Munich in 1979. iations. The bridge in Genoa consisted of a three-span viaduct and a multi-span overpass, which connected to the viaduct from the west. The viaduct was constructed over ~80 m wide Polcevera river, from which it derived its ofﬁcial name. The bridge structure was built over the railway line linking Genoa with Milan, the industrial plants, and residential

Copyright: © 2021 by the author. buildings. The viaduct carried the A10 motorway from Genoa to Savona, transferring Licensee MDPI, Basel, Switzerland. trafﬁc of a large number of heavy trucks. The carriageway had four lanes, two in both This article is an open access article directions. The view of the Polcevera Viaduct from the north-east is shown in Figure1. distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Appl. Sci. 2021, 11, 8098. https://doi.org/10.3390/app11178098 https://www.mdpi.com/journal/applsci Appl. Sci. 2021, 11, 8098 2 of 2021

Appl. Sci. 2021, 11, 8098 2 of 21

Figure 1. View of Polcevera Viaduct from the north-east (Morandi, 1968). FigureFigure 1. 1.View View of of Polcevera Polcevera Viaduct Viaduct from from the the north-east north-east (Morandi, (Morandi, 1968). 1968). 1.2. Static Scheme of the Bridge Structure 1.2. Static Scheme of the Bridge Structure 1.2. TheStatic bridge Scheme consisted of the Bridge of 11 Structure spans with the following length: 43.00 + 5 × 73.20 + 75.313 + 142.655TheThe bridge+ bridge 207.884 consisted consisted + 202.50 of +of 1165.10 11 spans spans = 1,102.452 with with the the m. following following The structural length: length: scheme 43.00 43.00 + of 5+ × 5the ×73.20 73.20 viaduct + + 75.313 75.313 and +the+ 142.655 142.655 multi-span + + 207.884 207.884 overpass + + 202.50 202.50 are+ shown+ 65.1065.10 ==in 1102.4521,102.452 Figure 2. m. m. The The structural structural scheme scheme of of the the viaduct viaduct and and thethe multi-span multi-span overpass overpass are are shown shown in in Figure Figure2. 2.

Figure 2. Structural scheme of the viaduct and the multi-span overpass (Morandi, 1968; dimensions in m). Figure 2. Structural scheme of the viaduct and the multi-span overpass (Morandi, 1968; dimensions in m). Figure 2. Structural scheme of the viaduct and the multi-span overpass (Morandi, 1968; dimensions in m). The static scheme of structures of the viaduct and the multi-span overpass consisted of The static scheme of structures of the viaduct and the multi-span overpass consisted a multi-span frame with V-type supports (Figure3a) and spandrel beams with cantilevers of a multi-spanThe static framescheme with of structures V-type supports of the vi (Figureaduct and3a) andthe multi-spanspandrel beams overpass with consisted cantile- on which the suspended spans were supported (the piers of the previously designed bridge versof a on multi-span which the frame suspended with V-type spans weresupports support (Figureed (the 3a) piersand spandrel of the previously beams with designed cantile- in Venezuela are of the X-type [3]). Overpass spans were supported at the cantilevers of vers on which the suspended spans were supported (the piers of the previously designed bridgethe V-type in Venezuela piers Nos. are 3 toof 8.the Whereas X-type [3]). the structureOverpass of spans the three were other supported piers ofat thethe viaductcantile- bridge in Venezuela are of the X-type [3]). Overpass spans were supported at the cantilevers(Nos. of 9, the 10 V-type and 11) piers was anotherNos. 3 to type. 8. Wherea A pylons the supporting structure twoof the pairs three of cablesother piers was addedof the vers of the V-type piers Nos. 3 to 8. Whereas the structure of the three other piers of the viaductto each (Nos. of the 9, framed 10 and support.11) was another The pylons type. of A thepylon viaduct supporting are shown two pairs in Figure of cables3b. Thewas viaduct (Nos. 9, 10 and 11) was another type. A pylon supporting two pairs of cables was addedthree reinforced to each of concrete the framed pylons support. of the The portal pylons shape of were the viaduct 90 m high are above shown the in terrain. Figure The3b. added to each of the framed support. The pylons of the viaduct are shown in Figure 3b. Thecross-section three reinforced of the bridgeconcrete span pylons was of a threethe portal cell pre-stressed shape were 90 concrete m high box. above The the spandrel terrain. The three reinforced concrete pylons of the portal shape were 90 m high above the terrain. Thebeam cross-section in the longest of span,the bridge between span the was forked a three columns cell pre-stressed of support concrete No. 10, hadbox. aThe length span- of drelapproximatelyThe beam cross-section in the 45 longest m,of the and span, bridge the between length span ofwas the the aforked cantileverthree columnscell pre-stressed was of ~65 support m. concrete The No. 18 10, m box. widehad The a bridgelength span- ofdeckdrel approximately wasbeam positioned in the longest45 m, at anand span, elevation the between length of of ~45 the the mforked cantilever above columns the was terrain. of~65 support m. The No. 18 m10, wide had abridge length deckof approximately was positioned 45 at m, an and elevation the length of ~45 of them above cantilever the terrain. was ~65 m. The 18 m wide bridge deck was positioned at an elevation of ~45 m above the terrain. Appl. Sci. 2021, 11, 8098 3 of 21 Appl. Sci. 2021, 11, 8098 3 of 20

(a) (b)

FigureFigure 3. 3.(a()a Overpass) Overpass support support (longitudinal (longitudinal an andd cross-section) cross-section) (Morandi, (Morandi, 1968). 1968). (b ()b Viaduct) Viaduct support support (longitudinal (longitudinal an andd cross cross section) section) (Morandi, (Morandi, 1968). 1968). Appl. Sci. 2021, 11, 8098 4 of 21 Appl. Sci. 2021, 11, 8098 4 of 20

1.3.1.3. Diagonal Diagonal Stay Stay Cable Cable Structure Structure According According to to OnOn both both sides sides of of each each pylon, pylon, two two cantilevers cantilevers of of prestressed prestressed concrete concrete and and steel steel box box werewere ﬁxed fixed to to the the pylons. pylons. These These cantilevers cantilevers were were supporting supporting the the suspended suspended span. span. The The steel steel cablescables with with prestressed prestressed concrete concrete shells shells poured poured on on had had a rectangulara rectangular cross-section cross-section and and a a formform of of a compositea composite steel steel and and concrete concrete element, element, asingle a single at at the the top top of of the the pylon pylon and and divided divided intointo two two equal equal parts parts at at the the deck deck level. level. Cross-sections Cross-sections of of the the diagonal diagonal stay stay cable cable are are shown shown inin Figure Figure4a 4a (on (on the the left-at left-at the the deck deck level; level; on on the the right-at right-at the the top top of theof the pylon) pylon) [1]. [1].

(a) (b)

FigureFigure 4. (4.a )(a Cross-sections) Cross-sections of of the the concreted concreted cable cable (Morandi, (Morandi, 1968). 1968). (b )(b Schematic) Schematic anchoring anchoring of of the the cable cable in in the the deck deck and and at at thethe top top of of the the pylon pylon (Morandi, (Morandi, 1968). 1968).

ThereThere were were 24 24 steel steel tendons tendons in the in centralthe central part part of the of cross-section the cross-section of the of diagonal the diagonal stay cable,stay eachcable, consisting each consisting of 12 strands of 12 strands with a diameterwith a diameter of 1/2-inch, of 1/2-inch, i.e., 12.7 i.e., mm 12.7 (12T13) mm (12T13) with a tensilewith a strengthtensile strength of R = 170 of kg/mmR = 170 2kg/mm(ca. 17002 (ca. MPa). 1700 Those MPa). tendons Those tendons were strained were beforestrained encasementbefore encasement in concrete in concrete and they and transferred they transferred the weight the weight of the cantilever.of the cantilever. The tendons The ten- weredons concreting were concreting in segments. in segments. Forces in Forces the tendons in the were tendons transferred were tran to concretesferred to through concrete frictionthrough and friction adhesion and forces. adhesion Additionally, forces. Additi thereonally, were 28 there steel were tendons 28 steel located tendons close tolocated the edgeclose of to the the concrete edge of cross-section, the concrete eachcross-section, consisting each of four consisting strands of with four a diameterstrands with of 1/2 a di- inchameter (4T13). of 1/2 Those inch tendons (4T13). were Those located tendons in ducts,were located strained in after ducts, hardening strained of after the concrete,hardening andof injectedthe concrete, afterward. and injected Consequently, afterward. the Cons diagonalequently, stay the cable diagonal is the composite stay cable steelis the and com- concreteposite steel element and transferringconcrete element permanent transferring loads permanent of the suspended loads of span, the suspended as well as span, loads as fromwell the as roadloads trafﬁc. from the road traffic. Thus,Thus, one one diagonal diagonal stay stay cable cable consisted consisted of of 52 52 steel steel tendons, tendons, hereinafter hereinafter referred referred to to as as thethe “internal “internal tendons”, tendons”, located located inside inside the diagonalthe diagonal stay-cable stay-cable made made of 400 of(24 400× (2412 +× 2812 ×+ 284) × 1/2-inch4) 1/2-inch tendons. tendons. The The cross-sectional cross-sectional areas areas of the of steel the internalsteel internal tendons tendons “S”, concrete “S”, concrete “C” and“C” the and stay-cable the stay-cable “B” are “B” listed are listed in Table in 1Table. There 1. There were steelwere stirrupssteel stirrups with awith diameter a diameter of 6 mm,of 6 mm, spaced spaced every every 15 cm 15 in thecm cable.in the Schematiccable. Schematic anchoring anchoring of the cable of the in thecable deck in the and deck at theand top at of the the top pylon of the is shownpylon is in shown Figure in4b. Figure 4b. This type of structural detail of the steel cables with prestressed concrete shells poured onTable had to1. Characteristics ensure the bonding of the concreted of the steel cabl internale before tendonsand after withstrengthening. concrete to the transfer of loads, and also the concrete was to form the corrosion protection of the tendons. Therefore, the loadData due to road trafﬁcUnit was to be transferred S by the C entire steel and B concrete composite T section. A (cm2) 506.7 11,449.3 11,956.0 466.5 EA (MN) 9120.6 36,637.8 45,758.4 8397.0 Appl. Sci. 2021, 11, 8098 5 of 21

Appl. Sci. 2021, 11, 8098 5 of 20

Designations: A—cross-sectional area; EA—tensile stiffness where E stands for longitudinal modulus of elasticity; S—steel internal tendons of the cable; C—concrete in the cable; B—composite steel Table 1. Characteristics of the concreted cable before and after strengthening. and concrete cable; T—long external tendons of the cable. Data Unit S C B T This type of structural detail of the steel cables with prestressed concrete shells 2 pouredA on had to ensure(cm ) the bonding506.7 of the steel 11,449.3 internal tendons 11,956.0 with concrete 466.5 to the transferEA of loads, and(MN) also the concrete 9120.6 was to form 36,637.8the corrosion protection 45,758.4 of the tendons. 8397.0 Therefore,Designations: theA—cross-sectional load due to road area; trafficEA—tensile was stiffnessto be transferred where E stands by forthe longitudinal entire steel modulus and concrete of elasticity; compositeS—steel internal section. tendons of the cable; C—concrete in the cable; B—composite steel and concrete cable; T—long external tendons of the cable.

1.4.1.4. Construction Construction of of the the Main Main Structural Structural Elem Elementsents of ofthe the Bridge Bridge in inGenoa Genoa According According to to OverpassOverpass spans spans of of about about 32 32 m m were were supporte supportedd at atthe the short short cantilevers cantilevers (about (about 8 m 8 in m in length),length), of of the the V-type piers Nos. 3 to 8. FigureFigure5 5aa showsshows thethe momentmoment ofof placing placing the the span span on onthe the cantilevers cantilevers of of piers piers Nos. Nos. 6 and6 and 7 of7 of the the overpass overpass [ 1[1].]. StructurallyStructurally independent independent pylons pylons were were constr constructeducted at atthe the V-type V-type piers piers Nos. Nos. 9, 10 9, 10and and 11. Pylon Pylon No. No. 9 9 was was executed executed first. ﬁrst. Figure Figure 5b5b shows shows the the construction construction of of the the pylon pylon and and the the two-sidedtwo-sided cantilevers. cantilevers. These These cantilevers cantilevers were were constructed constructed using using the the cantilever cantilever concreting concreting method,method, with with the the use use of of additional additional compressi compressionon with with external external pre-stressing pre-stressing temporary temporary steel tendons laid on the upper surface of the cantilevers (Figure 5c). After pylon No. 9 steel tendons laid on the upper surface of the cantilevers (Figure5c). After pylon No. was completed, two other pylons were built. Each pylon with cantilevers was built inde- 9 was completed, two other pylons were built. Each pylon with cantilevers was built pendently. Stay cables were made after the pylon and cantilevers were completed. Figure independently. Stay cables were made after the pylon and cantilevers were completed. 5d shows completed pylon No. 11—with the cantilevers and prestressed concrete element, Figure5d shows completed pylon No. 11—with the cantilevers and prestressed concrete support No. 10—with the pylon made and cantilevers but still without cables and support element, support No. 10—with the pylon made and cantilevers but still without cables and No. 9—with the pylon made but still without cantilevers. support No. 9—with the pylon made but still without cantilevers. After the completion of the pylon and the cantilevers, the cantilevers were suspended After the completion of the pylon and the cantilevers, the cantilevers were suspended from the pylon using steel tendons (those which would be located in the central part of from the pylon using steel tendons (those which would be located in the central part of the the concreted cable). After these steel tendons were tensioned, the external temporary ten- concreted cable). After these steel tendons were tensioned, the external temporary tendons dons used for the construction of the cantilevers were dismantled. Figure 5e shows sup- used for the construction of the cantilevers were dismantled. Figure5e shows support No. port No. 10 with concreted cables and support No. 11 with steel tendons before concreting 10 with concreted cables and support No. 11 with steel tendons before concreting (in this (in this support the external temporary tendons of the cantilevers were still used). The V- support the external temporary tendons of the cantilevers were still used). The V-type pier type pier with two-sided cantilevers suspended to the pylon and concreted cables formed with two-sided cantilevers suspended to the pylon and concreted cables formed integrated integrated support for the viaduct. The spans of about 36 m were supported on the canti- support for the viaduct. The spans of about 36 m were supported on the cantilevers. levers. Figure 5f shows the moment of placing the span on the cantilevers of piers Nos. 10 Figure5f shows the moment of placing the span on the cantilevers of piers Nos. 10 and 11. and 11.

(a)

Figure 5. Cont. Appl. Sci. 2021, 11, 8098 6 of 20 Appl. Sci. 2021, 11, 8098 6 of 21

(b)

(c)

(d)

(e)

Figure 5. Cont. Appl. Sci. 2021, 11, 8098 7 of 20 Appl. Sci. 2021, 11, 8098 7 of 21

(f)

FigureFigure 5. 5. ((aa)) Placing Placing the the span span on on the the cantilevers cantilevers of of su supportspports Nos. Nos. 6 and 6 and 7 of 7 the of the overpass overpass (Morandi, (Morandi, 1968). (b (b) )Construction Construction of of pylon pylon and and cantilevers cantilevers of support of support No. No.9 of the 9 of viaduct the viaduct (Morandi, (Morandi, 1968). 1968).(c) Pre-stressing temporary tendons laid on top of the cantilevers during the construction of the canti- (c) Pre-stressing temporary tendons laid on top of the cantilevers during the construction of the can- levers (Morandi, 1968). (d) Viaduct supports at different stages of construction: No. 9—completely executed,tilevers (Morandi, No. 10—with 1968). the (d two-sided) Viaduct cantilevers supports at made different and No. stages 11—with of construction: the pylon made No. 9—completely (Morandi, 1968).executed, (e) Viaduct No. 10—with supports the at two-sided different cantilevers stages of co madenstruction: and No. No. 11—with 10—with the concreted pylon made cables (Morandi, and No.1968). 11—with (e) Viaduct steel tendons supports before at different concreting stages (Morandi, of construction: 1968). (f) No.Placing 10—with the span concreted on cantilevers cables of and supportsNo. 11—with Nos. 10 steel and tendons 11 of the before viaduct concreting (Morandi, (Morandi, 1968). 1968). (f) Placing the span on cantilevers of supports Nos. 10 and 11 of the viaduct (Morandi, 1968). 2. Methods and Materials of Strengthening of the Polcevera Viaduct 2. Methods and Materials of Strengthening of the Polcevera Viaduct 2.1. Technical Condition of the Polcevera Viaduct after 25 Years of Service 2.1. Technical Condition of the Polcevera Viaduct after 25 Years of Service Several years after the bridge opening to the regular service (in 1967), damage to concrete Severalin the structure years after was thediscovered. bridge opening In the 1980s, to the extensive regular renovation service (in works 1967), were damage car- to riedconcrete out on in the the structure. structure The was study discovered. of Italstra Inde the in 1980s, 1988 [4] extensive presents renovation the scope of works the ren- were ovationcarried work out on carried the structure. out until TheFebruary study 1988. of Italstrade The work in concerned 1988 [4] presentsprimarily the the scope repair of of the concreterenovation in pylon work of carried the support out until No. February 11 (located 1988. in the The lower-left work concerned corner in primarily Figure 1). the It was repair theofconcrete first of the in pylonpylons of built, the support and it was No. expe 11 (locatedcted to insee the more lower-left damage corner than in Figurethe other1). It pylonswas the that first were of the constructed pylons built, later. and it was expected to see more damage than in the other pylonsThe that renovation were constructed works included later. removal of loose parts of concrete, cleaning of the surfaceThe using renovation water lance, works anchoring included the removal steel mesh of loose (with parts a thickness of concrete, of 4 mm cleaning and mesh of the sizesurface of 10 using × 10 watercm), on lance, selected anchoring concrete the surfaces, steel mesh application (with a thickness of cement of mortar 4 mm andwith mesh a thicknesssize of 10 of× ~210 cm cm), and on protection selected of concrete the surface surfaces, with two application anti-corrosion of cement coatings: mortar an epoxy with a coatingthickness and of a ~2 polyurethane cm and protection coating. of Scaffolding the surface withwas constructed two anti-corrosion at the support coatings: to an carry epoxy outcoating the repair and a works. polyurethane Figure 6a coating. presents Scaffolding the scaffolding was of constructed support No. at 11 the for support repair works to carry plannedout the repair only at works. its lower Figure part6 a(state presents as of thethe scaffoldingmid-1980s [4]. of support No. 11 for repair works plannedDue onlyto the at observed its lower poor part technical (state as ofcondition the mid-1980s of the concrete [4]). in the lower part of the pier, Dueit was to decided the observed to build poor scaffolding technical at conditionthe entire height of the concreteof the pylon in the(Figure lower 6b) part al- of lowingthe pier, also it was the decidedassessment to buildof the scaffoldingtechnical condition at the entire of the height upper of part the pylonof the support, (Figure6 b) whichallowing was also difficult the assessment to access until of then. the technical Apart from condition damage of to the the upperconcrete, part steel of thecorrosion support, waswhich found was in difficult structural to access elements until of then. the viaduct, Apart from in particular damage toin thethe concrete,upper part steel of corrosionthe di- agonalwas found stay cable. in structural elements of the viaduct, in particular in the upper part of the diagonalIn January stay cable. and March of 1992, two reports [5,6] were drafted by Prof. Manfred Wicke from Inthe January Technical and University March of of 1992, Innsbruck, two reports Austria. [5,6 ]The were first drafted report byassessed Prof. Manfred whether Wickethe repairfrom the works Technical carried University out on the of structure Innsbruck, were Austria. sufficient The and first discussed report assessed corrosion whether of the the strandsrepair works in the diagonal carried out stay on cable, the structurewhile the other were report sufficient evaluated and discussed whether corrosionthe design offor the astrands strengthening in the diagonal of the diagon stay cable,al stay while cable thewas other correct. report evaluated whether the design for a strengthening of the diagonal stay cable was correct. Appl. Sci. 2021, 11, 8098 8 of 21

The following is the information contained in the first report by Prof. Wicke of Janu- ary 1992. According to this report, some of the strands in the diagonal stay cable were broken, and some others were heavily corroded. “The worst damage was to be seen at the upper end of the north-western stay-cable from support 11 and was examined by the expert at the time of the visit on 14 January 1992. He was informed then that about 30 to 40 strands were broken or had become slack and that a similar number were badly corroded. This damage is regarded as being of a serious nature”. The strands were visible in a completely deteriorated connection between the cable and the pylon. At the top of the pylon, water was penetrating through the cracked and missing concrete to tendons causing corrosion. “The above assessment applies primarily to the northern stay-cables of support nos. 9 and 11. In the opinion of the expert, the replacement of the remaining stay-cables will prove to be (…) necessary”. The poor technical condition of the cables resulted in a decrease in the yield strength of the steel in the internal tendons, and further penetration of the water into the cables would result in the breaking of successive strands. Prof. Wicke assessed that the damaged internal tendons should be replaced by new ones. “This rigorous requirement is called for mainly because the stay-cables have to bear Appl. Sci. 2021, 11, 8098 the load of a whole section of the bridge and the consequences of failure would be cata-8 of 20 strophic (…). The condition of the steel reinforcement will not permit any postponement as a speedy implementation is required”.

(a) (b)

FigureFigure 6. (a 6.) Support (a) Support No. No. 11 11 with with scaffolding scaffolding in in the the lowerlower part. ( (bb) )Support Support No. No. 11 11with with scaffolding scaffolding at full at height full height (Italstrade, (Italstrade, 1988). 1988).

TheSince following the actual is thetechnical information condition contained of the tendons in the firstwas unknown, report by Prof.at least Wicke the follow- of January ing provisional measures concerning road traffic were recommended: 1992. According to this report, some of the strands in the diagonal stay cable were broken, and• some Prohibition others wereof overtaking heavily corroded.by trucks and “The non-standard worst damage vehicles. was to be seen at the upper end• of Limitationthe north-western of the traffic stay-cable speed to from50 km/h. support 11 and was examined by the expert at the time“A of limit the on visit the on permitted 14 January weight 1992. for He individual was informed lorries thenis not that required. about Exceptional 30 to 40 strands wereloads broken should, or however, had become no longer slack be and permitte that ad. similar With respect number to werethe uncertainties badly corroded. of the This damageactual damage is regarded situation as being and ofthe a fact serious that nature”.a failure of The the strands stay-cable were might visible occur in awithout completely deteriorated connection between the cable and the pylon. At the top of the pylon, water was penetrating through the cracked and missing concrete to tendons causing corrosion. “The above assessment applies primarily to the northern stay-cables of support Nos. 9 and 11. In the opinion of the expert, the replacement of the remaining stay-cables will prove to be ( ... ) necessary”. The poor technical condition of the cables resulted in a decrease in the yield strength of the steel in the internal tendons, and further penetration of the water into the cables would result in the breaking of successive strands. Prof. Wicke assessed that the damaged internal tendons should be replaced by new ones. “This rigorous requirement is called for mainly because the stay-cables have to bear the load of a whole section of the bridge and the consequences of failure would be catas- trophic ( ... ). The condition of the steel reinforcement will not permit any postponement as a speedy implementation is required”. Since the actual technical condition of the tendons was unknown, at least the following provisional measures concerning road traffic were recommended: • Prohibition of overtaking by trucks and non-standard vehicles. • Limitation of the traffic speed to 50 km/h. “A limit on the permitted weight for individual lorries is not required. Exceptional loads should, however, no longer be permitted. With respect to the uncertainties of the actual damage situation and the fact that a failure of the stay-cable might occur without warning, at least the reduction of traffic as supposed above should be imposed. The temporary measures are put forward in the context of the speedy implementations of the repair work”. Appl. Sci. 2021, 11, 8098 9 of 21

Appl. Sci. 2021, 11, 8098 warning, at least the reduction of traffic as supposed above should be imposed. The tem-9 of 20 porary measures are put forward in the context of the speedy implementations of the repair work”. In the suspended 36 m long span, that contained 10 tendons, and each tendon consisted ofIn 18 the wires suspended with a 36diameter m long of span, 7 mm, that some contained of the 10 wires tendons, were and broken each and tendon some consisted were corroded.of 18 wires “The with damage a diameter is comparable of 7 mm, with some that of the stay-cables, wires were but broken it was and reckoned some were to becorroded. less serious “The on account damage of is comparablethe degree of with degradation that of the being stay-cables, manifestly but smaller, it was reckoned plus the to differentbe less position serious on of accountthe damaged of the area degree in relation of degradation to load flow”. being manifestly smaller, plus the different position of the damaged area in relation to load ﬂow”. 2.2. Strengthening of the Concreted Cables at Support No. 11 of the Polcevera Viaduct between 2.2. Strengthening of the Concreted Cables at Support No. 11 of the Polcevera Viaduct between 19921992 and and 1994 1994 TheThe design design of of strengthening strengthening of of the cables waswaselaborated elaborated by by Francesco Francesco Pisani, Pisani, the the engi- engineerneer from from the the Studio Studio Pisani Pisani in in Rome Rome (formerly (formerly associated associated with with Prof. Prof. Morandi, Morandi, responsible respon- siblefor for the the static static calculation calculation of theof the structure). structur Thee). The design design of strengthening of strengthening was was described described in the in workthe work by Camomilla by Camomilla [7], which[7], which is co-authored is co-authored by the by designer.the designer. TheThe assumptions assumptions that that were were taken taken in indesigning designing of ofstrengthening strengthening of ofthe the diagonal diagonal stay stay cablescables were were as asfollows: follows: • • PreservingPreserving the the architecture architecture of ofthe the structure. structure. • • HighHigh risk risk is associated is associated with with the the demolition demolition of ofthe the diagonal diagonal stay stay cables, cables, e.g., e.g., with with the the lacklack of ofpossibility possibility to tomaintain maintain the the stiffness stiffness of ofthe the concrete concrete support. support. • • NoNo possibility possibility of ofclosing closing traffic trafﬁc through through the the bridge. bridge. • • EnsuringEnsuring such such compressive compressive stress stress in in concrete concrete after strengtheningstrengthening that that the the steel-concrete steel-con- cretecross-section cross-section transferred transferred loads load ass aas composite a composite cross-section. cross-section. AccordingAccording to the to the design design by Pisani, by Pisani, the diagon the diagonalal stay staycable cable should should be strengthened be strengthened by 12 bylong 12 external long external tendons. tendons. Each tendon Each tendonwas designed was designed to consist to of consist 22 strands of 22 with strands a diam- with a eterdiameter of 15 mm of (22T15). 15 mm (22T15).The tendons The tendonswere located were along located two along vertical two surfaces vertical of surfaces the diago- of the naldiagonal stay cable, stay six cable, on each six side on each (Figure side 7). (Figure The area7). of The the area steel of cross-section the steel cross-section of long external of long tendonsexternal (“T”) tendons is given (“T”) in Table is given 1. The in Tabletendons1. Thewere tendons laid in steel were brackets laid in steelon the brackets diagonal on staythe cable diagonal in the stay spacing cable of in 4.0 the m. spacing The tend of 4.0ons m.were The anchored tendons to were the anchoredsteel elements, to the the steel so-calledelements, cap the at the so-called top of capthe atpylon, the top and of to the the pylon, transverse and to beam the transverse at the deck, beam which at theis also deck, connectedwhich is to also the connected short external to the shorttendons. external There tendons. were six There of those were short six of tendons, those short and tendons, each consistedand each of consisted31 strands of with 31 strands a diameter with of a 15 diameter mm (31T15). of 15 mmThey (31T15). were located They wereat the located bridge at deckthe level, bridge along deck the level, two along horizontal the two surfaces horizontal of the surfaces cable at of the the length cable atof theabout length 6.5 m. of The about short6.5 tendons m. The shortwere tendonsneeded due were to neededthe adopted due tomethod the adopted of strengthening method of the strengthening diagonal stay the cable.diagonal stay cable.

FigureFigure 7. Strengthening 7. Strengthening of diagonal of diagonal stay stay cable cable according according to tothe the design design by byF. Pisani. F. Pisani.

TheThe method method of ofstrengthening strengthening was was as asfollows: follows: • • TheThe concrete concrete was was removed removed in inthe the lower lower part part of ofthe the cable cable at atthe the length length of of6.5 6.5 m. m. • • ShortShort tendons tendons were were installed installed and and concreted concreted at atthe the length length of of~6.0 ~6.0 m mwith with high-strength high-strength concreteconcrete (leaving (leaving the the non-concreted non-concreted 0.5 0.5 m mgap). gap). • Short tendons were strained to the maximum value and all old, corroded strands of internal tendons (pre-tensioned and post-tensioned) in the non-concreted 0.5 m gap were cut off. Appl. Sci. 2021, 11, 8098 10 of 21

Appl. Sci. 2021, 11, 8098 10 of 20 • Short tendons were strained to the maximum value and all old, corroded strands of internal tendons (pre-tensioned and post-tensioned) in the non-concreted 0.5 m gap were cut off. •• TheThe stresses in shortshort tendonstendons werewere successively successively reduced reduced while while increasing increasing the the stress stress in inthe the long long external external tendons tendons to to the the design design value. value. TheThe strengthening strengthening of of the the diagonal stay ca cablesbles at support No. 11 was implemented betweenbetween 1992 1992 and and 1994. 1994. First, First, ca cablesbles located located at at the the northern northern side side (from (from the the mountains) mountains) werewere strengthened, strengthened, and later the southern side (from the seaside). In In addition, addition, the the method method ofof fastening fastening of of the the cables cables was was changed changed at at the the top top of of pylon pylon No. No. 10 10 by by providing providing a a steel steel cap. cap. ThisThis type type of of element element on on pylon pylon No. 11 11 is is shown shown in in Figure Figure 88d.d. PylonPylon No.No. 9 was not repaired oror rebuilt. rebuilt.

(a)

(b)

Figure 8. Cont. Appl. Sci. 2021, 11, 8098 11 of 20 Appl. Sci. 2021, 11, 8098 11 of 21

(c)

(d)

(e)

Figure 8. Cont. Appl. Sci. 2021, 11, 8098 12 of 20 Appl. Sci. 2021, 11, 8098 12 of 21

(f)

FigureFigure 8. 8. (a(a) )Support Support No. No. 11 11 with with strengthened strengthened north- north-easterneastern concreted concreted cable cable (photo (photo by by Rymsza; Rymsza; 04.1993).April 1993). (b) Strengthening (b) Strengthening of the of the concreted concreted cable cable with with long long external external tendons—view tendons—view from from the the deck deck (photo(photo by by Rymsza; Rymsza; 04.1993). April 1993). (c) The (c) north-western The north-western concreted concreted cable cableafter strengthening—view after strengthening—view from thefrom top the of toppylon of pylonNo. 11 No. (photo 11 (photo by Rymsza; by Rymsza; 04.1993). April (d) 1993).Steel cap (d) on Steel pylon cap No. on pylon 11 (photo No. 11by (photoRymsza; by 04.1993). (e) The south-western concreted cable before strengthening—view from the top of pylon Rymsza; April 1993). (e) The south-western concreted cable before strengthening—view from the top No. 11 (photo by Rymsza; 04.1993). (f) View from the top of pylon No. 11 to pylon No. 10 with of pylon No. 11 (photo by Rymsza; April 1993). (f) View from the top of pylon No. 11 to pylon No. scaffolding (photo by Rymsza; 04.1993). 10 with scaffolding (photo by Rymsza; April 1993).

2.3.2.3. Characteristics Characteristics of of the the Diagonal Diagonal Stay Stay Cable Cable before before and and after after Strengthening Strengthening AlthoughAlthough the the report report by by Prof. Prof. Wicke Wicke from from Ma Marchrch 1992 1992 [6] [6] mostly mostly concerned concerned the the design design ofof strengthening strengthening ofof thethe diagonal diagonal stay stay cables cables at supportat support No. No. 11, it11, also it containedalso contained information infor- mationon the on stay the cable stay beforecable before strengthening. strengthening. According According to that to that report, report, the staythe stay cable cable atthe at thetop top of theof the pylon pylon was was tensioned tensioned with with the the force force of of 19.119.1 MNMN (26.1–7.0)(26.1–7.0) under permanent permanent load,load, whereas whereas concrete concrete was was comp compressedressed with with the the force force of of 7 7 MN MN and and the the internal internal tendons tendons tensionedtensioned with with the the force force of of 26.1 26.1 MN. MN. AsAs the the result result of of the the strengthening, strengthening, long long external external tendons tendons were were strained strained with with the the force force ofof 13.5 13.5 MN. MN. That That force force resulted resulted in inthe the increase increase of the of thecompressive compressive force force in concrete in concrete by 10.2 by MN10.2 and MN reductionand reduction of the of tensile the tensile force force in the in the corroded corroded internal internal tendons tendons by by 2.1 2.1 MN MN (the (the remainingremaining force force of of 1.2 1.2 MN MN was was transferred transferred to to the the cantilever cantilever beam, beam, causing causing torsion torsion and and bendingbending moments). moments). Thus, Thus, after after strengthening strengthening and and under under permanent permanent load, load, the the concrete concrete in thein thediagonal diagonal stay stay cable cable was was compressed compressed with with the the force force of of 17.2 17.2 MN MN (7.0 (7.0 + + 10.2) 10.2) and and the the corrodedcorroded internal internal tendons tendons were were tensioned tensioned with with the the force force of of 24.0 24.0 MN MN (26.1–2.1). (26.1–2.1). While, While, underunder live live load, load, the the diagonal diagonal stay stay cable cable stre strengthenedngthened with with external external tendons tendons was was tensioned tensioned withwith the the force force of of 6.0 6.0 MN, MN, whereas whereas the the tendon tendonss were were tensioned tensioned with with the the force force of of 5.1 5.1 MN MN andand external external tendons tendons with with the the force force of of 0.9 0.9 MN. MN.

3.3. Results Results of of Strengthening Strengthening according According to to the the Author Author 3.1. Characteristics of the Diagonal Stay Cable According to the Author 3.1. Characteristics of the Diagonal Stay Cable according to the Author In 1993, the author of the paper was a participant of the 31st IRI (Istituto per la In 1993, the author of the paper was a participant of the 31st IRI (Istituto per la Ricos- Ricostruzione Industriale) course organized in Italy for bridge experts from different truzione Industriale) course organized in Italy for bridge experts from different countries. countries. During several months of the course, he had the opportunity to familiarize During several months of the course, he had the opportunity to familiarize himself with himself with the work of design and construction companies in Italy. In April 1993, the work of design and construction companies in Italy. In April 1993, he stayed at the he stayed at the Italstrade company in Milan. Since the company was involved in the Italstrade company in Milan. Since the company was involved in the reconstruction of the reconstruction of the Polcevera Viaduct, the author was sent to Genoa as part of the visit to Appl. Sci. 2021, 11, 8098 13 of 20

the company. During this period, the diagonal stay cables of the viaduct were strengthened according to the design by Pisani. Table1 summarized data characterizing the structure of the diagonal stay cable before and after strengthening with the use of the long external tendons. These are cross-sectional area and tensile stiffness. The cables were not strengthened at two supports of the viaduct, Nos. 9 and 10. The structure of the stay cable at those supports was constructed according to the design by Prof. Morandi. Assuming that: • internal forces speciﬁed in the report of Prof. Wicke (Wicke, March 1992) are correct, • the longitudinal modulus of elasticity of the tendon E = 180,000 MPa, • the modulus of elasticity of concrete E = 32,000 MPa, • the length of the cable is 92.2 m (based on Figure3b), • the cross-sectional area and the tensile stiffness are calculated and speciﬁed in Table1, it can be calculated that in the non-strengthened cable: Under permanent load: • the tensile stress in internal tendons was 515.1 MPa (26.1:0.05067), • the compressive stress in concrete was 6.1 MPa (7:1.14493), • the elongation of the cable transferring the load as a steel-concrete section was 38.5 mm (19.1 × 92,200:45,758.4), • the elongation of the cable transferring the load only through internal tendons would be 193.1 mm (19.1 × 92,200:9120.6); Under live load equal to 6.0 MN: • the tensile stress in internal tendons was 23.6 MPa (6.0:0.05067 × (9120.6: 45,758.4)), • the compressive stress in concrete was 4.2 MPa (6.04:1.14493 × (36,637.8: 45,758)), • the elongation of the cable transferring the load as a steel-concrete section was 12.1 mm (6.0 × 92,200:45,758.4), • the elongation of the cable transferring the load only through internal tendons would be 60.7 mm (6.0 × 92,200:9120.60). In summary, in the non-strengthened cable under total load (permanent and live load): • the tensile stress in internal tendons was 538.7 MPa (515.1 + 23.6), • the compressive stress in concrete was 10.3 MPa (6.1 + 4.2), • the elongation of the cable transferring the load as a steel-concrete section, was 50.6 mm (38.5 + 12.1), • the elongation of the cable transferring the load only through internal tendons would be 253.8 mm (193.1 + 60.7), • the increase in elongation of the cable transferring loads only through internal tendons would be 203.2 mm (253.8–50.6); such an increase in elongation of the cable would cause vertical displacement equal to 108.9 mm (sin32.4o × 203.2; the value of the angle was assumed based on Figure3b). The non-strengthened diagonal stay cable can be characterized as follows. Under the total action of loads, the stresses in the internal tendons were 539 MPa (which is ca. 32% of the tensile strength of steel (539:1700 × 100)) and 10.3 MPa in the concrete (which represents ca. 30% of compressive strength of concrete). The elongation of the cable transferring the load as a composite steel-concrete section was 50.6 mm. Assuming that the concrete is not compressed, and the load is transferred only through the internal tendons, the elongation of the cable would be 253.8 mm. An increase in elongation of the cable transferring the load only through internal tendons would result in vertical displacement of the end of the cantilever equal to 108.9 mm. According to Prof. Wicke (Wicke, March 1992), “vertical displacement of more than 100 mm will seriously damage the main bridge girder at the central pier”. Appl. Sci. 2021, 11, 8098 14 of 20

In summary, the cracking of concrete in the diagonal stay cable confirmed the lack of composite action of the concrete in the load distribution causing an increase in elongation of the steel cable resulting in such a vertical displacement (ca. 110 mm) that the failure could be caused due to breaking of the cable. In the case of one of the pylons-No. 11-4 cables were strengthened (whereas the new tendons cannot be strained to the required degree due to a limited pressure on deteriorated concrete). Assuming that the internal tendons can further transfer the load, the strengthened cable may be characterized as follows: Under permanent load: • the tensile stress in internal tendons was 473.7 MPa (24.0:0.05067), • the compressive stress in concrete was 15.0 MPa (17.2:1.14493); Under live load equal to 6.0 MN: • the compressive stress in concrete was 4.2 MPa [6.0:(11,449.3 + 506.7 × (180,000: 32,000))], • the tensile stress in external tendons amounted to 19.3 MPa (0.9:0.04665). After strengthening of the cable with the long external tendons, only a 9% decrease in the tensile force was achieved under permanent load in the corroded tendons ((24–26.1):24 × 100) and as much as 2.5 times increase in the compressive force in the concrete (17.2:7). Under live load, the stress in external tendons was 19.3 MPa, which is ca. 1.1% (19.3:1700 × 100) of the breaking strength of steel (assuming that steel with non-inferior properties was used compared to the designed properties of the viaduct). The stresses in concrete were considerably increased as a result of strengthening of the stay cable. Moreover, a slight decrease of stresses in the corroded internal tendons was achieved with a technically questionable assumption that both the internal tendons stressed before and after concreting will still transfer loads. Thus, after strengthening the stay cable, the corroded internal tendons will be subjected to dynamic load by road traffic. The requirement stated in the first report by Prof. Wicke (Wicke, January1992), i.e., “to release the old strands, cannot be met”. The author’s observations about the strengthening of the structure of the viaduct by compressing the diagonal stay cables: • External tendons slightly relieve strains from the corroded internal tendons—as much as 85% (5.1:6.0 × 100) of the load will still be transferred by the corroded stay cable under live load and only 15% (0.9:6.0 × 100) by external tendons. Figure8a shows pylon No. 11 with strengthened north-eastern cable. Figure8b presents strengthening of the cable with long external tendons in the view from the viaduct deck and Figure8c - in the view from the top of pylon No. 11. Figure8d shows the newly installed steel cap on pylon No. 11; • The level of prestressing the long external tendons is limited by the technical condition of the concrete in the diagonal stay cable-stresses at the level of 1% of the tensile strength of steel will be generated in external tendons. Figure8e shows the cable of pylon No. 11 before strengthening; • Compression of the curved structural element, like the diagonal stay cable, is difficult in practice (Figure8c,e). Summarizing, a different method of strengthening, than in the case of pylon No. 11, is recommended for the other two pylons (Nos. 9 and 10). Strengthening of pylon No. 10 was possible because scaffolding was already built beside that pylon (Figure8f).

3.2. Author’s Concept of Strengthening of Pylons Nos. 9 and 10 of the Viaduct from 1993 Figure 9 shows three viaduct supports. Each of them was treated differently: • the cables were strengthened at support No. 11, • a steel cap was installed at support No. 10 to connect the cables at both sides of the pylon (with scaffolding), • no works were performed at support No. 9. According to the author, the compression of the cable with external tendons at support No. 11 will be ineffective. Strengthening the curvilinear, corrosion-damaged cable could not be effective-sectional stressing was difficult for technological reasons, and the Appl. Sci. 2021, 11, 8098 stress level of the external tendons was low due to the poor technical condition15 of of the 20 concrete. However, the lack of any strengthening of the cables at supports Nos. 9 and 10 was a technical and organizational mistake.

FigureFigure 9.9.View View ofof thethe 33 supportssupports ofof thethe viaduct—No. viaduct—No. 1111 withwith strengthenedstrengthened concreted concreted cables, cables, No. No. 10 10 withwith an an additional additional steel steel cap cap and and No. No. 9 without9 without any an structuraly structural changes changes (photo (photo by by Rymsza; Rymsza; April 04.1993). 1993).

AccordingAfter all, the to the technically author, the correct compression solution of to the reduce cable withthe load external on the tendons concreted at support cable No.would 11 willbe to be add ineffective. independent Strengthening steel cables the attached curvilinear, to thecorrosion-damaged pylon, e.g., located in cable parallel could to notthe beexisting effective-sectional diagonal stay stressing cable. With was such difﬁcult a solution, for technological the concreted reasons, cable andwould the be stress con- levelsiderably of the unloaded. external tendons Additional was suspension low due to thecables poor in technicalthe structure condition allow of for the better concrete. load However,distribution the and lack redundancy-if of any strengthening one element of theis damaged, cables at other supports elements Nos. will 9 and take 10 a greater was a technicalload and andthe object organizational will not suddenly mistake. crash. AfterThe method all, the of technically strengthening correct the solution structur toe of reduce the facility the load could on be the as concreted follows: cable would• removing be to add the independent suspended steelspan-unload cables attached the support, to the including pylon, e.g., brackets, located in parallel to• therepair existing and diagonal reinforce stay of concrete cable. With supports, such aincluding solution, brackets the concreted and concrete cable would tendons, be considerably• adaptation unloaded. of the pylon Additional and brackets suspension to attach cables additional in the structure steel allowtendons for suspending better load distributionthe brackets, and redundancy-if one element is damaged, other elements will take a greater load• andfixing the additional object will steel not suddenlytendons to crash. the pylon and brackets, • Thestrengthening method of strengtheningthe suspended the span structure (or replacing of the facilityit with coulda lighter be asone) follows: and placing it • removingon the reinforced the suspended joints at span-unload the ends of the supports. support, including brackets, • repairSuch a and solution reinforce would of concreteextend the supports, life of th includinge structure brackets and reduce and concretethe risk of tendons, a sudden • failure.adaptation This solution of the is pylon muchand simpler brackets than tocompression attach additional of the concreted steel tendons cable suspending and it could be usedthe brackets,on the other two supports—Nos. 9 and 10. • ﬁxing additional steel tendons to the pylon and brackets, The author submitted his concept of strengthening the viaduct to Italian engineers, • strengthening the suspended span (or replacing it with a lighter one) and placing it on who were compressing the concreted cables at support No. 11. Figure 10a shows support the reinforced joints at the ends of the supports. No. 10 before strengthening, and in Figure 10b after strengthening. The concept of struc- turalSuch strengthening a solution wouldproposed extend by the lifeauthor of the has structure not been and accepted reduce thefor riskvarious of a suddenreasons. failure.Certainly This one solution of them-perhaps is much simpler even thanthe most compression important of theone-was concreted the need cable to and change it could the be used on the other two supports—Nos. 9 and 10. The author submitted his concept of strengthening the viaduct to Italian engineers, who were compressing the concreted cables at support No. 11. Figure 10a shows support No. 10 before strengthening, and in Figure 10b after strengthening. The concept of structural strengthening proposed by the author has not been accepted for various reasons. Certainly one of them-perhaps even the most important one-was the need to change the architecture of the structure. Additional steel cables would change the appearance of the viaduct designed by Prof. Morandi, and this was incompatible with the basic assumption to the design goal of preserving the original appearance of the viaduct. Appl. Sci. 2021, 11, 8098 16 of 21

architecture of the structure. Additional steel cables would change the appearance of the viaduct designed by Prof. Morandi, and this was incompatible with the basic assumption to the design goal of preserving the original appearance of the viaduct. At the end of the IRI course, a meeting was held with a representative of the Italian government. At the meeting, it was necessary to say in Italian what technical project in Italy had made the most impression on the participant of the course. The author stated that he was most impressed by the structure of Prof. Morandi’s Viaduct in Genoa, which should be properly strengthened as soon as possible because the current way of strengthening will be ineffective. The author presented his concept of strengthening (shown in Figure 10b) as one of the possible applications for strengthening the other two viaduct Appl. Sci. 2021, 11, 8098 supports. 16 of 20 The author stated that if the viaduct was not properly strengthened, it may collapse. The conversation took place 25 years before the collapse of the viaduct.

(a) (b)

FigureFigure 10. 10. ((aa)) Support Support No. No. 10 10 of of the the viaduc viaductt before before strengthening. strengthening. ( (bb)) Strengthening Strengthening of of support support No. No. 10 10 of of the the viaduct viaduct accordingaccording to to the the concept concept of of Rymsza Rymsza of of 1993. 1993.

4. DiscussionAt the end about of the the IRI Causes course, of athe meeting Viaduct was Collapse held with a representative of the Italian 4.1.government. The Consequences At the of meeting, the Viaduct it was Collapse necessary to say in Italian what technical project in Italy had made the most impression on the participant of the course. The author stated that The collapse of the Polcevera Viaduct in Genoa, Italy took place on the 14 August he was most impressed by the structure of Prof. Morandi’s Viaduct in Genoa, which should 2018 during a violent downpour, as a result of which 43 people were killed and 12 were be properly strengthened as soon as possible because the current way of strengthening will severely injured. At the time of the viaduct collapse, there were three trucks and 10 cars be ineffective. The author presented his concept of strengthening (shown in Figure 10b) as on the structure (acc. to RAI Television). The section of the viaduct with a length of about one of the possible applications for strengthening the other two viaduct supports. 250 m collapsed, i.e., the first pylon on the side of the multi-span overpass (support No. 9 The author stated that if the viaduct was not properly strengthened, it may collapse. on Figure 1-in the upper right corner and Figure 2), together with frame support, cantile- The conversation took place 25 years before the collapse of the viaduct. vers, and the suspended spans. In October 2017, the Wadi Kuf Bridge in Libya was closed. After4. Discussion the collapse about of the the viaduct Causes in of Genoa, the Viaduct the bridge Collapse in Venezuela over Lake Maracaibo was4.1. also The Consequencesclosed for the of time the Viaduct of repairs. Collapse The initial investigation of the causes of the Mo- randi Viaduct collapse in Genoa as a contribution to the design of pre-stressed structures The collapse of the Polcevera Viaduct in Genoa, Italy took place on the 14 August 2018 was presented by the Author in previous publication [8]. during a violent downpour, as a result of which 43 people were killed and 12 were severely injured. At the time of the viaduct collapse, there were three trucks and 10 cars on the structure (acc. to RAI Television). The section of the viaduct with a length of about 250 m collapsed, i.e., the ﬁrst pylon on the side of the multi-span overpass (support No. 9 on Figure1-in the upper right corner and Figure2), together with frame support, cantilevers, and the suspended spans. In October 2017, the Wadi Kuf Bridge in Libya was closed. After the collapse of the viaduct in Genoa, the bridge in Venezuela over Lake Maracaibo was also closed for the time of repairs. The initial investigation of the causes of the Morandi Viaduct collapse in Genoa as a contribution to the design of pre-stressed structures was presented by the Author in previous publication [8].

4.2. Causes of the Viaduct Collapse According to the Author 4.2.1. General Remarks The cause of the viaduct collapse was the rupture of one of the stay cables at support No. 9 (Figure2) holding the cantilever from the side of the multi-span overpass. Pre- sumably, it was the north-western cable. It is one of those northern cables (at the side of the mountains), whose poor technical condition was discussed in the ﬁrst report of Prof. Wicke [6], and which was not strengthened. According to the author, there were three main causes of the viaduct collapse: material, structural and organizational. Appl. Sci. 2021, 11, 8098 17 of 20

4.2.2. Material Cause The diagonal stay cable, as the composite steel-concrete structural component, is not a good solution for transferring the tensile force because of its insufﬁcient redundancy. The initial value of the compressive force in steel tendons decreases over time as a result of the so-called pre-stressing losses. The decrease in the value of the compressive force to the value of the initial force may be estimated at up to 20%. Pre-stressing losses are caused by: • shrinkage of concrete (decrease in the volume of concrete as a result of drying), • creep of concrete (increase in deformation without a change in load), • relaxation of steel (decrease of the pre-stressing force at constant deformation). The greater pre-stressing losses are, the greater is the length of the compressed element. In this case, the element was very long—it had a length of more than 90 m. Moreover, the shape of the diagonal stay cable changed from straight to more curved and as a result of perpendicular force reduced the value of the stress force. The more curved the shape of the tendon, the greater the value of the perpendicular force and the greater the pre-stressing loss. Hence, the additional compression of the misaligned cable as part of strengthening caused large pre-stressing losses (the more curved the tendon, the greater the losses). The reduction of the pre-stressing force overtime and the deformation of the diagonal stay cable resulted in cracking of the concrete cover. The cracked cover was not a suitable anti-corrosion protection for the tendons, which under these conditions were subject to accelerated corrosion. Intensive local corrosion occurred in places where oxygen and atmospheric moisture were penetrating through concrete defects. Similar opinion was presented in the article by Biliszczuk [9] “weak point of this structure ( ... ) was corrosion protection for the tendons, particularly in area of crown plate at the pylon top”.

4.2.3. Structural Cause First of all, the principal rule of designing the cable-stayed structures is to use several or a dozen stay cables supporting the bridge deck to achieve redundancy, and therefore, the damage of one of the cables will not lead (in any case should not lead) to the sudden collapse. The design solution adopted by Prof. Morandi that only two structural components supported the cantilever with the suspended span, excludes any cooperation of these structural components in a situation where one of them is destroyed or damaged (for example, a collapse would also take place even if a vehicle hit the cable and destroyed it). In addition, the replacement of a diagonal stay cable—an important element of the cable-stayed structure—was technically complicated and impossible to perform without closing the traffic on the structure. Secondly, the cantilever length of approximately 65 m is too large. In the case of a traffic jam on a motorway viaduct, there may be three five-axle trucks with a weight of 40 t and a length of up to 16.5 m in the spacing of about 5 m, on a single lane. This would result in overloading a single cable with vehicles of a total weight of 120 tons, standing on one lane. Thirdly, the concrete cross-section of the pylon is too small for the cross-section of the diagonal stay cables attached to the pylon. The cables transfer tensile forces and the pylon compressive forces; therefore, buckling should be considered during its design. Conse- quently, the relationship between the stiffness of the pylon and the cables is imbalance. Fourthly, the assessment of the technical condition of steel tendons inside the concrete element is very difficult in practice, especially the assessment on the peak of the pylon at the height of 45 m above the deck level.

4.2.4. Organizational Cause The report of Prof. Wicke from January 1992 [6] showed that the poor technical condition is observed in the cables on the northern side of supports Nos. 9 and 11. Therefore, it is incomprehensible why only diagonal stay cables at support No. 11 were strengthened, and no attempt was made to strengthen the diagonal stay cables at support No. 9. Leaving these Appl. Sci. 2021, 11, 8098 18 of 20

cables without any treatment was a mistake, regardless of the ineffective strengthening of cables at support No. 11.

4.3. The Destruction Stages of the Viaduct The viaduct collapse began with the rupture of the diagonal stay cable. The mechanism of changes in the mechanical properties of the stay cable was as follows: 1. The value of the pre-stressing force in the strained cables was reduced as a result of pre-stressing losses. 2. Due to the large dead weight of the cable, its shape changed from straight to curved and this led to a decrease of the compressive force, causing tension in concrete at the bottom surface of the cable. 3. The decreasing value of the compressive force in the cable as a result of pre-stressing losses and changes in the shape of the diagonal stay cable caused cracking of the concrete, starting from the bottom surface of the diagonal stay cable. 4. Cracking of concrete resulted in a decrease of the area of the compressed concrete cross-section that transferred the tension forces. 5. The cracked concrete ceased to provide adequate anti-corrosion protection of strands in the tendons and allowed for steel corrosion. 6. The strands not protected with concrete cover corroded and successively cracked, causing a decrease in the steel–concrete cross-sectional area of cable transferring the tension forces. 7. The decrease of the cross-sectional area of the concreted cable under constant forces resulted in increased stresses in the corroded tendons. 8. With such a decrease in the value of the pre-stressing force in the cables that concrete was not compressed in no place of the cable’s cross-section, the tension forces were transferred only by the corroded tendons. 9. When the load is transferred only by the corroded internal tendons, the elongation of the cable would be greater than 20 cm. 10. The successive ruptures of the corroded strands in the steel tendons under constant force would cause an increase of stresses in other unbroken strands and their excessive elongation, until the rupture of the last strand in the tendon. Figure 11 shows the various stages of the destruction of the viaduct. The mechanism of viaduct destruction was as follows: 1. The viaduct superstructure allowed for direct loading of the cantilever with three trucks-at the time of the viaduct collapse, presumably three trucks were on the cantilever. 2. The cantilever was structurally connected with the diagonal stay cables, with poor technical condition that could not transfer such a heavy live load. 3. The overloaded cantilever resulted in a rupture of one of the two cables, loaded more in the upper part due to the dead load of the cable than in the lower part (the weight of 1 m of the cable was ca. 3 t (1.1956 × 2.5). 4. The north-western cable was broken at support No. 9 (Figure 11a). 5. The broken cable with a weight of about 280 t (92.2 × 1.1956 × 2.5) fell on the cantilever, which resulted in the destruction of its structure and the rupture of the south-western cable (Figure 11b) located on the other side of the carriageway (the force should be determined taking into account the dynamic impact factor, e.g., equal to 1.5). 6. The destruction of the cantilever resulted in a collapse of the suspended span supported on the cantilever (Figure 11c). 7. The imbalance in the load of the pylon on both sides caused its destruction, and this led to the collapse of the cables, the cantilever, and the span at the eastern side of the pylon (Figure 11d–f). Appl. Sci. 2021, 11, 8098 19 of 21

4. The north-western cable was broken at support No. 9 (Figure 11a). 5. The broken cable with a weight of about 280 t (92.2 × 1.1956 × 2.5) fell on the cantilever, which resulted in the destruction of its structure and the rupture of the south- western cable (Figure 11b) located on the other side of the carriageway (the force should be determined taking into account the dynamic impact factor, e.g., equal to 1.5). 6. The destruction of the cantilever resulted in a collapse of the suspended span supported on the cantilever (Figure 11c). Appl. Sci. 2021, 11, 8098 7. The imbalance in the load of the pylon on both sides caused its destruction, and19 of this 20 led to the collapse of the cables, the cantilever, and the span at the eastern side of the pylon (Figure 11d–f).

FigureFigure 11. 11.Fragments Fragments of of the the animation animation showing showing the the probable probable course course of of the the viaduct viaduct collapse collapse (author (author ofof the the animation: animation: Kai Kai Kostack, Kostack, [ 10[10]).]).

TheThe study study on on post-collapse post-collapse analysis analysis of of Morandi’s Morandi’s Polcevera Polcevera Viaduct viaduct inin Genoa Genoa Italy Italy waswas presented presented by by others others [ 11[11].]. Their Their study study employed employed the the existing existing knowledge knowledge in in estimating estimating thethe remaining remaining life life of theof the bridge, bridge, including: including: the existing the existing design design information information about the about bridge; the post-collapsebridge; post-collapse investigation investigation report of report the ministry of the ofministry infrastructure of infrastructure and transport and [transport12]; and archival[12]; and truck archival trafﬁc truck and traffic axle load and data. axle load data.

5.5. Conclusions Conclusions InIn the the 1960s 1960s and and 1970s, 1970s, Prof. Prof. RiccardoRiccardo Morandi Morandi designeddesigned threethree bridgesbridges withwith aa similar,similar, atypicalatypical structure.structure. They were were built built in in Venezuela, Venezuela, Italy, Italy, and and Libya. Libya. During During the the viaduct viaduct col- collapselapse on on 1414 August August 2018, 2018, the the support support No. No. 9 of 9 the of the viaduct viaduct in Italy in Italy collapsed. collapsed. The The other other two twosupports, supports, Nos. Nos. 10 10and and 11, 11, were were demolished demolished on on 28 28 June June 2019 2019 and thethe PolceveraPolcevera ViaductViaduct ceasedceased to to exist. exist. However, However, two two more more bridge bridge structures structures of of this this same same design design by by Morandi Morandi still still exist.exist. It It seems seems that that the the author’s author’s concept concept of of adding adding steel steel cables cables attached attached to to the the pylon pylon could could bebe carried carried out out to to strengthen strengthen these these structures structures (Figure (Figure 10b). 10b). Since Since both structuresboth structures have ahave large a pylonlarge cap,pylon a fancap, system a fan system of cable of arrangement cable arrang (notement a harp (not system),a harp system), proposed proposed by the author by the inauthor Genoa in in Genoa 1993, in could 1993, be could used. be The used. Author The Author recommends recommends removing removing the suspended the suspended span and placing a span that would not be suspended but would be pulled in (the hinges would be removed—the static scheme of the platform would not be a Gerberian beam, but a continuous beam). In such a situation, additional steel tendons could be evenly distributed over the entire length of the span. Regardless of which of the solutions is adopted (repair of a suspended span or the use of a continuous structure), there will still be difﬁculties related to the poor technical condition of all structural elements (damage to concrete and steel corrosion) and the need to perform construction on both sides simultaneously to prevent the destruction of a pylon with too little stiffness. Fortunately, failure of the bridge structures do not happen very often; however, there are some bridge failures during the construction phase, e.g., in the pre-stress process Cie´sla[13]. The historical bridges still in service are seldom subject to failure. According to Appl. Sci. 2021, 11, 8098 20 of 20

the author, this may change in the future, and failures of the bridges would happen more frequently than before. Often, unique bridge structures, with large spans do not meet the requirements of the European Standard [14] (EN 1990, 2004), Table 2.1, concerning a safe use period equal to 100 years. The Polcevera Viaduct is an example of such a structure. Many new bridge structures—non-typical with large span lengths and a large length of river crossing—were constructed primarily in the last 20 years. Architects and engineers compete with one another in the design of architecturally attractive bridge structures with extreme dimensions. The constantly increasing live loads and weather anomalies (the temper wind and the driving rain) cause accelerated degradation of each structure (the Polcevera Viaduct collapsed during a violent downpour). On the other hand, the effects of the collapse of non-typical structures with large spans and large lengths of crossings are proportional to these dimensions and their originality. The period of safe use of structures with a unique design is, in principle, much shorter than assumed for typical structures. In addition, maintaining a good technical condition of such structures is difﬁcult in practice and very costly. It is also difﬁcult and costly to strengthen or dismantle them. The viaduct collapse in Genoa in Italy should be treated as a warning to all bridge administrations to avoid the construction of unusual and excessively large bridges.

Funding: This research received no external funding. Institutional Review Board Statement: Not applicable. Institutional Consent Statement: Not applicable. Data Availability Statement: The study did not report any data outside of this publication. Conﬂicts of Interest: The author declares no conﬂict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

References 1. Morandi, R. Viaducto sobre el Polcevera, en Genowa—Italia. Inf. Construcción 1968, 21, 57–88. [CrossRef] 2. Troitsky, M.S. Cable-stayed bridges. In Theory and Design; Crosby Lockwood Staples: London, UK, 1977. 3. Podolny, W., Jr. Concrete cable-stayed bridges. Bridge Eng. 1978, 2, 121–130. 4. Italstrade. Autostrada Genova-Savona Viadotto Polcevera: Lavori di risanamento della pila. In Lavori Eseguiti da Italstrade a Tutto il 29 Febbraino 1988; 1988. Available online: http://yadda.icm.edu.pl/yadda/element/bwmeta1.element.baztech-192d98c7-1d6b- 4fb9-9a16-263b70ddb798 (accessed on 24 August 2021). 5. Wicke, M. Interim report on the corrosion damage to the prestressing steel on the “Viadotto sul Polcevera”. In Report to ISA Italstrade Appalti S.p.A.; Innsbruck: Genoa, Italy, 1992. 6. Wicke, M. 2nd Report on the Polcevera Viaduct. Report to Autostrada S.p.A.; Innsbruck; Springer: Berlin, Germany, 1992. 7. Camomilla, G.; Pisani, F.; Martinez y Cabrera, F.; Marioni, A. Repair of the Stay Cables of the Polcevera Viaduct in Genova, Italy; IABSE Reports: Zürich, Switzerland, 1995. 8. Rymsza, J. Causes of the Morandi viaduct disaster in Genoa as a contribution to the design of pre-stressed structures. Roads Bridges—Drog. i Mosty 2020, 19, 5–25. [CrossRef] 9. Biliszczuk, J.; Teichgraeber, M. O katastroﬁe wiaduktu Polcevera w Genui, we W3oszech. Inzynieria˙ i Bud 2018, 74, 578–582. 10. Kostack, K. Fragments of the Animation “Morandi Bridge Destruction Simulation Genoa 2018|Demolition in 2019”. Available online: www.youtube.com/watch?v=Y6suQ0FIoIQ (accessed on 31 August 2021). 11. Morgese, M.; Ansari, F.; Domaneschi, M.; Cimellaro, G.P. Post-collapse analysis of Morandi’s Polcevera viaduct in Genoa Italy. J. Civ. Struct. Health Monit. 2020, 10, 69–85. [CrossRef] 12. Ministero delle Infrastrutture e dei Trasporti, Commissione Ispettiva Ministeriale. Comune di Genova, Autostrada A10—Crollo del Viadotto Polcevera, Evento Accaduto il 14 Agosto 2018; Governo Italiano: Roma, Italy, 2018. 13. Cie´sla,J.; Biskup, M.; Topczewski, Ł.; Skawi´nski,M. Cases of Bridge Structural Failure during the Process of Pre-Stressing. Roads and Bridges–Drogi i Mosty 2017, 16, 15–35. [CrossRef] 14. EN 1990. Eurocode. In Basis of Structural Design, Table 2.1; 2004; p. 20. Available online: https://www.phd.eng.br/wp-content/ uploads/2015/12/en.1990.2002.pdf (accessed on 24 August 2021).