This article was downloaded by: [Universita di Padova]On: 29 August 2014, At: 10:37Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Structure and Infrastructure Engineering:Maintenance, Management, Life-Cycle Design andPerformancePublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/nsie20

Contribution of in situ and laboratory investigations forassessing seismic vulnerability of existing bridgesCarlo Pellegrinoa, Mariano A. Zaninia, Paolo Zampieria & Claudio Modenaa

a Department of Civil, Environmental and Architectural Engineering, University of Padova,Via Marzolo, 9, 35131 Padova, ItalyPublished online: 24 Jul 2014.

To cite this article: Carlo Pellegrino, Mariano A. Zanini, Paolo Zampieri & Claudio Modena (2014): Contribution of in situ andlaboratory investigations for assessing seismic vulnerability of existing bridges, Structure and Infrastructure Engineering:Maintenance, Management, Life-Cycle Design and Performance, DOI: 10.1080/15732479.2014.938661

To link to this article: http://dx.doi.org/10.1080/15732479.2014.938661

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Contribution of in situ and laboratory investigations for assessing seismic vulnerabilityof existing bridges

Carlo Pellegrino*, Mariano A. Zanini1, Paolo Zampieri2 and Claudio Modena3

Department of Civil, Environmental and Architectural Engineering, University of Padova, Via Marzolo, 9, 35131 Padova, Italy

(Received 13 November 2013; final version received 20 January 2014; accepted 16 April 2014)

This work describes the main results of an in situ experimental campaign carried out in the context of the seismic assessmentof 71 existing road bridges as representative of the main structural typologies located in the Veneto region and pertaining tothe regional roadway managing authority Veneto Strade SpA. An initial in situ and laboratory experimental campaign onbasic materials and an overall geometrical survey were carried out for each bridge, using different tests according to theconstruction typologies, with the aim of characterising the main mechanical parameters of the significant structuralelements. Structural assessment has been subsequently carried out according to the parameters derived from the abovecampaign and the design ground motion action. The results have been compared with those deriving from a preliminaryanalysis conducted on the basis of some assumptions on materials’ mechanical characteristics related to the original designdocuments and the construction practice at the time of the structures’ edification. This comparative analysis could give someinsights on appropriate and rational planning of inspections on existing bridges and reliability of basic assumptions for theirseismic assessment.

Keywords: bridges; seismic analysis; maintenance and inspections; management; assessment

1. Introduction

In transportation networks, bridges arewillingly consideredamong the most vulnerable elements, also for having astrategic role in terms of civil protection aims and first aid inemergency states and immediate post-earthquake rescueoperations. These infrastructures, mostly built or recon-structed after Second World War (Pellegrino, Pipinato, &Modena, 2011), have a significant age and have beendesigned without considering seismic criteria. The devel-opment of new or improved seismic regulations in the pastyears has led to pay greater attention to seismic designissues, foreshadowing urgent need of seismic assessment ofsuch existing structures for road and railway managingauthorities. In light of these reasons, managing authoritiesare nowadays interested in increasing the knowledge levelof their structures’ conditionwith the aim of using such datafor a proper seismic assessment of the bridges (Biondini,Camnasio, & Palermo, 2014; Zanini, Pellegrino,Morbin, &Modena, 2013) and the network in which they are included(Carturan, Pellegrino, Rossi, Gastaldi, & Modena, 2013;Carturan, Zanini, Pellegrino, & Modena, 2014).

In most of the bridge management systems (BMSs)currently in use worldwide, in situ and laboratoryinvestigations are not typically foreseen and rationallyplanned, because of the huge amount of structures tomanage. Referring toEuropean experiences, in situmaterial

characterisation aimed to a proper seismic assessment oflarge stocks of bridges is often not planned in countries withnegligible seismic hazard (Hofmann, Kuhn, Friebem, &Winkler, 2012; Kamya, 2012; Soderqvist & Veijola, 2012;Torkkeli & Lamsa, 2012). Increasing attention is insteadfocused in areas with more significant seismic activitywhere, however, the experience in BMSs related to largestocks of bridges (Yue, Pozzi, Zonta, & Zandonini, 2012;Zonta, Zandonini, & Bortot, 2007) typically provide theexecution of limited visual in situ inspections finalised toupdate the condition state of the bridges withoutquantitative assessment of the mechanical characteristicsof the basic materials.

In this respect, this work shows the main results relatedto an agreement between the Department of Civil,Environmental and Architecture Engineering of theUniversity of Padova and the Veneto’s (North EasternItaly) regional roadway managing authority, VenetoStrade SpA aimed at the seismic assessment of arepresentative stock of existing bridges located in themain roadway lines in the Veneto region. First, apreliminary seismic analysis of 335 existing bridges wascarried out. In most cases, the main mechanicalcharacteristics of bridge elements were unknown andhence some assumptions about significant properties of thebasic materials were done on the basis of the original

q 2014 Taylor & Francis

*Corresponding author. Email: carlo.pellegrino@unipd.it

Structure and Infrastructure Engineering, 2014http://dx.doi.org/10.1080/15732479.2014.938661

Dow

nloa

ded

by [U

nive

rsita

di P

adov

a] a

t 10:

37 2

9 A

ugus

t 201

4

design documents, guidelines on assessing existingstructures and construction practice at the time of thestructures’ edification.

This work mainly focuses on the process of updatingpreliminary seismic analyses for a cluster of 71 bridgestaken as representative group of the initial 335 bridges.A series of in situ and laboratory investigations andgeometrical surveys have been developed on these 71bridges with the aim of defining main physical andmechanical characteristics of the materials constitutingbridge structural elements. Figure 1 shows the geographi-cal distribution of the above 71 bridges representing themost common bridge typologies in the Veneto’s regionalroadway network. The considered bridges belong to themost common structural typologies as shown in Tables 1–3for reinforced concrete structures and in Table 4 formasonry/stone bridges.

Comparing the results of the seismic assessmentderived from the preliminary and the updated analysiscould give a unique opportunity for obtaining someinsights on appropriate and rational planning of inspections

on existing bridges and reliability of basic assumptions fortheir seismic assessment in actual situations. The main aimof this work is to demonstrate that the use of properlyplanned in situ and laboratory investigations on represen-tative bridge structural elements may represent a usefulinstrument for the execution of careful seismic vulner-ability assessment of large stock of bridges and allowavoiding too conservative or unconservative assumptions.

2. In situ and laboratory tests

The investigation campaign had as main objective theknowledge of the principal geometrical, physical andmechanical properties of the basic materials constitutingthe 71 examined bridges. According to the data of the abovecampaign, it will be possible to build a structural finiteelement (FE) model as close as possible to the realcharacteristics of the considered structures. A preliminaryinvestigation plan was carried out, subdividing the 71bridges in two main categories depending on theconstruction material: masonry/stone bridges andreinforced concrete bridges (including also pre-stressedconcrete bridges). The subdivision by constructionmaterialhas eased to organise the sequence of in situ tests to becarried out for each category for properly planning thelogistic and operative aspects for the execution of suchsurveys. The investigation campaign has been madeperforming destructive methods taking into account thatany restriction, e.g. due to specific architectural value

Figure 1. Localisation of the 71 bridges considered for the execution of the investigation campaign in the Veneto roadway network.

Table 1. Structural typologies for the reinforced concretebridges.

Reinforced concrete bridges – structural typologies

Simply supported decks Continuous decks Arched decks

38 6 3

C. Pellegrino et al.2

Dow

nloa

ded

by [U

nive

rsita

di P

adov

a] a

t 10:

37 2

9 A

ugus

t 201

4

emerged and according to the better accuracy of destructivemethods for the characterisation of the main mechanicalparameters.

With regard to masonry/stone bridges, the investi-gation consisted in the extraction of masonry/stonesamples (Figure 2) at the abutments and arches’ positionto be subsequently tested in laboratory, measurement ofthe stress state in situ through the use of flat-jacks(Figure 3) and in situ evaluation of the capacity throughthe use of a system of double jacks. Table 5 summarises allinvestigations conducted on 24 masonry/stone bridges: foreach structure ‘X’ means that the test has been effectivelycarried out, whereas ‘O’ means that the test has not been

carried out due to execution difficulties or inaccessibilityof the site.

Regarding reinforced concrete bridges, the investi-gation campaign has been characterised by the executionof pachometer tests in each structure for the geometricalidentification of the reinforcing bars location in piers andabutments (Figure 4). Once defined the bars’ location, theextraction of rebar samples has been performed to testthem in laboratory and evaluate the main steel mechanicalcharacteristics (Figure 5). Finally, a series of concretesamples were taken from abutments and piers (Figure 6)with the subsequent estimation of the reinforced concretemechanical characteristics with laboratory tests (Figure 7).Table 6 summarises the tests carried out on reinforcedconcrete bridges.

Referring to the surrounding soils characterisation andtheir influence on the seismic behaviour of the bridges,specific geological analyses were not developed but soilclassification (based on shear wave velocity parameter)provided by the Italian code (Italian Ministry ofInfrastructures, 2008) is considered. The characteristicsof the soil are related to the amplification factor taken intoaccount in the definition of the seismic demand.

3. Procedure for the seismic assessment ofmasonry/stone bridges

The high stiffness of the resisting structural system ofmasonry/stone arch structures ensures the substantial

Table 2. Number of spans for the reinforced concrete bridges.

Reinforced concrete bridges – number of spans

Single span Two spans Three spans Four spans Five spans Greater than five spans

19 6 10 2 2 8Average span length ¼ 16.65m

Table 3. Pier typologies for the reinforced concrete bridges.

Reinforced concrete bridges – piers typologies

Column piers Framed piers Wall piers

5 8 15

Table 4. Number of spans and deck material for the masonry/stone bridges.

Masonry/stone bridges – number of spans Material

Single span Two spans Three spans Masonry Stone

17 4 3 6 18Average span length value ¼ 9.95m

Figure 2. Masonry sample taken out from bridge no. 4.Figure 3. Application of flat jacks for the in situ stress stateestimation in the abutment of bridge no. 39.

Structure and Infrastructure Engineering 3

Dow

nloa

ded

by [U

nive

rsita

di P

adov

a] a

t 10:

37 2

9 A

ugus

t 201

4

absence of relative displacements until the seismicacceleration reaches values able to transform the structureinto a mechanism, through non-dissipative plastic hingesleading to the structure collapse. The key parameter for thedetermination of the structural response is the peak groundacceleration for the site in which the bridge is placed.Several studies have underlined how the assessment ofmasonry/stone arches can be properly conveyed by limitanalysis (Clemente, 1998; Clemente, Occhiuzzi, & Raithel,

1995; Heyman, 1966 1982) also in compliance with currentItalian regulations (Italian Ministry of Infrastructures,2008). A collapse mechanism must be defined for theimplementation of limit analysis of masonry/stone archbridges; the type of mechanism mainly depends on thegeometrical characteristics of the bridge itself. Variouscollapse mechanisms can be considered according to thenumber of spans of the arch bridge and in relation of theslenderness of piers and abutments.

Single-span bridges generally have massive abutmentsthat can be assumed as infinitely rigid constraints. Underthese assumptions, the most vulnerable element to seismicaction in longitudinal direction appears to be the arch,which can be subjected to antimetric collapse mechanismsthrough the creation of four plastic hinges. In this case, theseismic response of the bridge is mostly influenced by thegeometric characteristics of the arch presenting a localfailure mechanism, in particular by f/L and s/L ratios,where f is the arch rise, L the span length and s the archthickness. For single-span bridge abutments, when theratio between the abutments height Ha and the abutmentsbase width Ba in longitudinal direction is sufficiently large,the most vulnerable mechanism in longitudinal direction isa global failure mechanism that involving both thespringers and the arch. In general, the structure is lesssusceptible to seismic actions in transversal direction, butthe most vulnerable elements are those referred to thespandrel walls, which can undergo to an out-of-planecollapse. Typically, this kind of mechanism does not affectthe global structural safety but can cause instability of theroad surface.

In multi-span bridges, the seismic vulnerability inlongitudinal direction is mainly influenced by pierslenderness Hp/Bp (the ratio between pier height Hp andlongitudinal pier base width Bp): for growing values of thisratio, a collapse mechanism involving also the piers due tothe creation of plastic hinges at their bases can arise(Figure 8(a)). For multi-span bridges with thick piers, the

Table 5. In situ investigations carried out on masonry/stonebridges stock.

Masonry/stone bridges’ in situ tests

Bridgeno.

Abutment pull-outsamples

Arch pull-outsamples

Singleflat jack

Doubleflat jacks

2 X X X X4 X X X X6 X O X X9 X X O O14 X X X X15 X X X X21 X O X X22 O X X X25 O O X X26 X X X X27 O O X X30 X X X X31 X X X X32 X O X X33 O O X X39 X X X X40 X X X X52 O X O O54 X X O O55 X X X X56 X X X X61 X O X X64 O X X X66 X O X X

Notes: X ¼ the test was carried out; O ¼ the test was not carried out.

Figure 4. Pachometer test execution on the pier of bridge no. 1.

C. Pellegrino et al.4

Dow

nloa

ded

by [U

nive

rsita

di P

adov

a] a

t 10:

37 2

9 A

ugus

t 201

4

most fragile element is represented by the arch: a potentialantimetric collapse mechanism can occur in each span,similarly to what happens for single-span bridges. Theseismic behaviour in transversal direction is thereforeinfluenced by piers slenderness and bridge (and piers)width. The local collapse mechanism due to spandrel walloverturning, previously described for single-span bridges,remains to be cited for multi-span bridges (Figure 8(b)).Seismic vulnerability can be more significant in transversaldirection than in longitudinal one for multi-span bridgeswith slender piers: in these cases the assessment could beconducted by means of nonlinear static analyses throughthe implementation of a specific FE model.

In the specific case of the masonry/stone bridges’ stockanalysed in this work, 92% of them belong to the single-span typology with stiff abutments and multi-span bridgeswith thick piers, whereas slender multi-span bridges arethe 8% of the cases. For most of the considered structures(92%), the seismic analysis consisted in the evaluation ofthe seismic capacity of the arch through the study of themain kinematic mechanisms, whereas for the remaining

cases, it was necessary to develop a nonlinear staticanalysis.

A mechanism was assumed a priori, and the horizontalforce enabling that mechanism was calculated determiningthe related coefficient a0 through the application of theprinciple of virtual works as defined below:

Pv þ a0Ph ¼ 0; ð1Þ

where Pv represents the vertical load and Ph are thehorizontal forces acting on the structure. The plastichinges are placed where the pressure line becomes tangentto the arch intrados or extrados. The pressure line can beiteratively calculated assuming a four hinges initialconfiguration: if this line passes through the hinges andis completely contained in the arch thickness, the assumedhinges’ disposition is the correct one; if otherwise the lineremains outside the arch thickness the procedure has to beiterated.

This approach has allowed evaluating the horizontalaction that the structures are able to carry during themechanism evolution (Figure 9(a)). Applying the principleof virtual works to a deformed configuration of the samestructure, it can be observed a gradual reduction of thecoefficient a until reaching the limit case beyond whichthe structure collapses due to equilibrium loss. Accordingto the above concepts, the capacity curve has beenconstructed for each analysed bridge, with the followingexpression:

a ¼ a0 12dkdk0

! "; ð2Þ

where dk0 is the maximum displacement of the controlpoint, dk is the displacement of the control point, functionof the a coefficient (Figure 9(b)).

In the overall seismic vulnerability assessment of thebridges, the specific vulnerabilities of the structuralelements constituting masonry/stone arch bridges (arch,spandrel wall, piers, abutments, foundations) have been

Figure 5. Extraction of steel bars’ samples and followinglaboratory tensile test execution for the pier of the bridge no. 1.

Figure 6. Extraction of concrete samples from the abutment of bridge no. 3.

Structure and Infrastructure Engineering 5

Dow

nloa

ded

by [U

nive

rsita

di P

adov

a] a

t 10:

37 2

9 A

ugus

t 201

4

investigated. The spectral acceleration a*0 has beencompared with the elastic spectral acceleration ag(PVR)corresponding to T ¼ 0, amplified by the soil factor S. Theelastic spectral acceleration is scaled by a structurereduction factor q at the ultimate limit state (ULS):

a*0 $ agðPVRÞ%S; ð3Þ

a*0 $agðPVRÞ%S

q: ð4Þ

Furthermore, a comparison between the ultimatedisplacement d

*

u and the displacement demand obtainedfrom the displacement spectrum SDe corresponding to thesecant period Ts (Equation (6)) was made:

d*

u $ SDeðTsÞ; ð5Þ

T s ¼ 2p

ffiffiffiffiffid

*

s

a*

s

s

; ð6Þ

where d*

s is the secant displacement (equal to 0.4d*

u) and a*0

is the seismic acceleration derived from the capacity curvecorresponding to a displacement value equal to d

*

s . Theresults obtained above have been expressed by means ofsafety factors FCi for each one of them (Equations (7)–(9)), calculated as the ratio between a resistance parameterR and a demand parameter E:

FC;1 ¼R

E¼ a*0

agðPVRÞ%S; ð7Þ

FC;2 ¼R

E¼ a*0

ðagðPVRÞ%SÞ=q; ð8Þ

FC;3 ¼R

E¼ d

*

u

SDeðT sÞ: ð9Þ

Once verified the arches, the procedure provided, assecond step, to the evaluation of the structural capacity ofthe spandrel walls, i.e. the most vulnerable elements whenhorizontal forces are acting in transversal direction. Theapplication of the principle of virtual works has led toevaluate the horizontal load multiplier coefficient aturning the structure into a mechanism. In this case, themechanism consists in a rotation of the spandrel wallaround a cylindrical base hinge. Seismic assessment hasbeen performed considering a rectangular wall withconstant thickness t, height Z and unitary length accordingto Rota, Pecker, Bolognini, and Pinho (2005). The acoefficient for seismic loads has been calculated by meansof the following rotation balance equation:

aPh

2þ aPSOILZSOIL þ aNhþ Sh

Z

3

2 Svt2 ðN þ PÞ t2¼ 0;

ð10Þ

where P represents the spandrel wall weight, S representsthe lateral ground horizontal action, N represents thevertical force applied at the top of the wall and aPSOIL

takes into account the inertia of the filling materialadjacent to the spandrel wall. Safety factors FCi have beencalculated also for the spandrel wall.

Finally, assessment of the abutments has been done. Ithas been evaluated if the arch collapse mechanism involveor not the abutments. If a mechanism is localised only inthe arch (as in almost all of the bridges analysed), a safetyfactor FCi as the ratio between the resisting bendingmoment MRd of the abutments and the relative demandbending moment MSd was evaluated. If abutments areinvolved in the collapse mechanism a limit analysis forevaluating the global collapse mechanism and the relativecoefficient multiplier a0 has been performed. The safetyfactor FCi against sliding of the foundation has beencalculated as the ratio between the frictional resistanceFRd and the base shear force FSd. The safety factorsobtained for the abutments and foundations (Equations(11) and (12)) have then been used for estimating theresisting peak ground acceleration PGARES,i (Equation(13)) as a function of the demand peak ground accelerationPGAacti:

FC;ABUTMENTS ¼ R

E¼ MRd

MSd; ð11Þ

FC;FOUNDATION ¼ R

E¼ FRd

FSd; ð12Þ

PGARES ¼ FC;i%PGAact: ð13Þ

Finally, the remaining 8% of slender multi-spanmasonry/stone bridges have been studied with specific

Figure 7. Laboratory compressive strength tests on the concretesamples taken from the abutment of bridge no. 3.

C. Pellegrino et al.6

Dow

nloa

ded

by [U

nive

rsita

di P

adov

a] a

t 10:

37 2

9 A

ugus

t 201

4

3D models and a macro-modelling approach. MidasFEAsoftware (2009) was used andmasonry/stone was modelledas homogeneous continuum, with eight- and six-node

elements. The numerical models have been characterisedby material and geometrical nonlinearity. The total straincrack model implemented in MidasFEA was used as

Figure 8. Masonry arch multi-span collapse mechanisms for bridges with slender piers in longitudinal (a) and transversal (b) directions.

Table 6. In situ investigations carried out on reinforced concrete bridges stock.

Reinforced concrete bridges’ in situ tests

Bridge no.Pachometerabutment Pachometer pier

Abutment pull-outconcrete samples

Pier pull-outconcrete samples

Abutment extractionsteel samples

Pier extractionsteel samples

1 O X O X O X3 X X X X X X8 X X X X O X10 X X X X X X11 X X X X X X12 X X X X O X13 X O X O X O16 X O X O X O17 X X X X X X18 X O X O X O19 X X X X X X20 X O X O X O23 X X X X X X24 X X X X O X28 X X X X X X29 X O X O X O34 X X X X X X35 X O X O X O36 X O X O X O37 X O X O X O42 X X X X X X43 X O X O X O44 X X X X X X45 X X X X X X46 X O X O X O48 X O X O X O49 X O X O X O50 X O X O X O51 X X X X X X53 X O X O X O57 X O X O X O58 X X X X X X62 X X X X O O63 X X X X O O67 O X X X O X68 X X X X X X69 X X X X X X70 O X O X O X71 X X X X X X

Notes: X ¼ the test was carried out; O ¼ the test was not carried out.

Structure and Infrastructure Engineering 7

Dow

nloa

ded

by [U

nive

rsita

di P

adov

a] a

t 10:

37 2

9 A

ugus

t 201

4

constitutive law for masonry. Specific checks/verificationson the reliability of the elements and constitutive laws usedin the models are included in the documentation availablein MidasFEA (2009) user manual.

4. Procedure for the seismic assessment of reinforcedconcrete bridges

The significant number of reinforced concrete bridgesanalysed in this work has justified the creation of a specificseismic assessment procedure. The major seismic vulner-abilities have been attributed to the vertical elements suchas piers and abutments. The structural capacity of theabutments has been assessed in terms of resistance with theaim of checking that they remain elastic up to ULS.A specific FE model has been realised for each abutment,and a static analysis has been developed. The abutmentshave been modelled by means of Strand7 v.2.4 software(2010), discretising the structureswith bi-dimensional plateelements and subsequently performing linear staticanalyses. Hence, adopted FE models are rather simple and

currently used in the common practice. Specific checks/verifications on the reliability of the elements and analysesused in the models are included in the documentationavailable in Strand7 v.2.4 (2010) user manual.

These analyses have been performed to obtain the entityof the acting forces in the abutments in terms of bending atthe wall base, bending moment at the top of the wall, shearfailure at wall bottom and shear failure on the lateral walls(Figure 10). A safety factor FCi has been calculated bydividing the bending moment/shear resisting forces by therelative bending moment/shear acting forces obtained:

FC;i ¼R

E¼ MRd

MSd; ð14Þ

FC;i ¼R

E¼ FRd

FSd; ð15Þ

PGARES ¼ FC;i%PGAact: ð16Þ

Regarding the piers, once defined the materials’ mainmechanical characteristics and the localisation of the

Figure 10. Assessment of the reinforced concrete bridge abutments.

Figure 9. Arch progressive collapse mechanism (a) and capacity curve (b) of the bridge no. 52.

C. Pellegrino et al.8

Dow

nloa

ded

by [U

nive

rsita

di P

adov

a] a

t 10:

37 2

9 A

ugus

t 201

4

reinforcing steel bars, the moment–curvature diagram ofthe element section has been found, calculating yieldingFy

and ultimateFu curvature and their corresponding bendingmoment values.

Moment–curvature diagram has been calculated inboth principal directions for each pier due to the differentlongitudinal and transversal structural response. Oncedefined the length of the plastic hinge Lpl, rotation andtheir relative displacement capacity values have beenevaluated for different limit states, considering the top ofthe pier as control point. The obtained displacementcapacity value SRd has then been compared with thedisplacement demand SSd, calculated from the displace-ment elastic spectrum for each limit state considered andcomputing the safety factor values (Equation (17)). Theobtained results have been converted in terms of resistingaccelerations, according to Equation (18):

FC;i ¼SSLi

Rd

SSLi

Sd

; ð17Þ

PGASLi

RES ¼ FC;i%PGAactSLi : ð18Þ

The acting shear FSd at the pier bottom has beencompared with the corresponding resisting shear TRd. Thefailure mechanisms described above have been consideredbecause they are representative of the main potentialseismic vulnerabilities for the structural typologiesincluded in the reinforced concrete bridges’ stock of thiswork. Other potential vulnerabilities for reinforcedconcrete bridges have been considered, but any potentialsignificant issue has not been observed. In particular, acomparison between deck displacement demand andavailable bearing length has been performed and bearingloss and potential pounding effects between adjacent

simply supported spans have been checked withoutshowing any problems.

5. Results

5.1. Mechanical characteristics of the basic materials

In situ and laboratory investigations have led to theevaluation of the main materials’ mechanical character-istics and have allowed the comparison between the valuessupposed during the first part of the study and thosedetected during the investigation campaign, obtaininginformation about the accuracy of the preliminaryassumed hypotheses.

In most of the cases, the values collected with theinvestigation campaign were greater than those prelimi-nary estimated, whereas only a small sample of bridgeswas characterised by values from the investigation lowerthan the assumed ones. Figure 11 shows in abscissa firstlyassumed values and in ordinate in situ measured values ofthe concrete compressive and steel yielding strengths,expressed in MPa.

Figure 12 presents the variation of the probabilitydensity functions (pdfs) describing assumed and measuredcompressive concrete strengths (Figure 12(a)) andassumed and measured tensile steel yielding strengthvalues (Figure 12(b)). Table 7 depicts the average valuesand the relative standard deviations of the materials’mechanical properties. Regarding concrete, the averagevalue of the measured compressive strength is higher thanthat assumed. Instead, the average value of the measuredsteel yielding strength is greater than that assumed.

Regarding masonry/stone bridges, although the pro-cedure is substantially based on limit analysis and bridges’geometrical configuration, similar comparisons have beenmade because the values of the materials’ mechanical

Figure 11. Mechanical characteristics for the considered reinforced concrete bridges: concrete compressive strength (a) and steelyielding stress (b).

Structure and Infrastructure Engineering 9

Dow

nloa

ded

by [U

nive

rsita

di P

adov

a] a

t 10:

37 2

9 A

ugus

t 201

4

characteristics derived from in situ investigation cam-paigns could be useful for the construction of appropriateFE models of the bridges. Masonry/stone compressivestrength and elastic modulus have therefore been firstlyassumed equal for all the analysed bridges according to thesuggestions of the Italian regulations about the estimationof mechanical parameters of masonry/stone structures.After in situmeasurements of those parameters, the resultshave been compared with the assumed ones, highlightinghow the effective material resistances were often greaterthan those preliminarily assumed (Figure 13). Average andstandard deviation values of the main materials’ propertieshave been evaluated also for masonry/stone bridges’ stockand listed in Table 7 and shown in Figure 14.

The results have highlighted higher compressivestrength values in the second analysis than thosepreliminarily assumed. In particular, the average elasticmodulus is around 10,000MPa (even if the highest one,related to a stone bridge, was about 26,000MPa). It hasto be stressed that the stock is mostly composed by stonebridges (Table 4) typically having higher elastic modulithan masonry bridges. Measured quantities have beencompared with the assumed ones, evaluating their ratiosfor each structure and thus evidencing the situations forwhich in situ investigations show conservative (ratiolower than 1), correct/neutral (equal to 1) or unconser-vative (higher than 1) assumptions. Figure 15 illustratesthe fractions of conservative, correct/neutral and

unconservative ratios between measured and assumedvalues of the main mechanical materials’ parameters.It can be observed that more correct/conservativeassumptions have been made for mechanical character-istics of the reinforcing steel than for concrete due amore accurate quality control during the steel productionprocess.

5.2. Masonry/stone bridges

The results obtained in first and second analysis are nowshown for the possible collapse mechanisms. Figure 16shows the results related to the analyses of the arch(Figure 16(a)), spandrel wall (Figure 16(b)) collapsemechanisms considering safety limit state (SLS) and ULS,abutment and foundation (Figure 16(c)). The diagramallows to visualise the safety factors FCi, as defined inEquations (11)–(13), calculated in first and secondanalysis for each masonry/stone bridge, highlighting alsothe relative improvement or the worsening of each FCi

between the first and second analyses. The dots on theinclined lines are characterised by unchanged safetyfactors between first and second analysis.

The results highlight how the most vulnerable collapsemechanism is that of the spandrel wall in transversaldirection, with the majority of the analysed bridgescharacterised by safety factors FCi lower than 1. It can beobserved how safety factors for the arch and the spandrel

Table 7. Average and standard deviation values for concrete, steel and masonry/stone main materials mechanical parameters.

Average value (MPa) Standard deviation (MPa)

Materials Assumed value Measured value Assumed value Measured value

Abutments – concrete compressive strength 29.42 35.9 4.91 14.84Abutments – steel tensile yielding stress 426.13 418.97 21.2 30.31Piers – concrete compressive strength 31.3 46.35 4.83 17.05Piers – steel tensile yielding stress 421.65 405.55 27.85 44.4Masonry/stone – compressive strength 4.2 4.25 – 0.72Masonry/stone – elastic modulus 6000 9989.13 – 6529.4

Figure 12. Pdfs of compressive concrete strength (a) and tensile steel yielding stress (b) of the whole reinforced concrete bridges’ stockas firstly assumed and after in situ investigations.

C. Pellegrino et al.10

Dow

nloa

ded

by [U

nive

rsita

di P

adov

a] a

t 10:

37 2

9 A

ugus

t 201

4

wall at SLS are bigger than those obtained at ULS. Theresults have evidenced how the geometrical survey hasplayed a useful role in the whole structural assessment, inparticular referring to spandrel wall failure mechanismestimations (as shown in Figure 17). A significantpercentage of unchanged results have been obtainedsince, in these cases, assumed bridges’ geometricalcharacteristics have been confirmed during the in situgeometrical surveys. Figure 18 presents the assessmentoutcomes and relative changes from the first to the secondanalysis for the structural elements of the masonry/stonebridge stock. The diagrams show how spandrel wall seemsto be the most vulnerable element according to the relatedunsatisfied outcomes.

5.3. Reinforced concrete bridges

Figures 19 and 20 depict the results of the assessment ofthe group of reinforced concrete bridges considered inthis work for abutments (Figure 19) and piers (Figure 20),

respectively. The results show the comparisons betweensafety factor values obtained in the first study (on thebasis of the original drawings, where available, orassumed according to the past practice) and those derivedfrom the in situ measurements. Figures 19 and 20 showthe safety factors FCi calculated in first and secondanalysis for each reinforced concrete bridge, showing therelative improvement or the worsening of each FCi

between first and second analysis. The results havehighlighted a general increase of the positive outcomesmoving from the first to the second analysis, mainly dueto the conservative assumptions adopted in the firstanalysis.

Figure 21 illustrates the fractions of increased,unchanged and reduced safety factors FCi for the firstand the second analysis of reinforced concrete bridges.The diagrams highlight how safety factors for bending inabutments and piers are characterised by a substantialincrease of the FCi values. Figure 21 also shows how thein situ investigations on materials’ properties have led to a

Figure 13. Mechanical characteristics for the considered masonry bridges: masonry compressive strength (a) and elastic modulus (b).

Figure 14. Pdfs of masonry/stone compressive strength (a) and elastic modulus (b) of the whole masonry/stone bridges’ stock as firstlyassumed and after in situ investigations.

Structure and Infrastructure Engineering 11

Dow

nloa

ded

by [U

nive

rsita

di P

adov

a] a

t 10:

37 2

9 A

ugus

t 201

4

Figure 15. Fractions of conservative, neutral and unconservative ratios between measured and assumed values of the main mechanicalmaterials’ parameters.

Figure 16. Results for the analysed masonry bridges: arch collapse mechanism (a), spandrel wall collapse mechanism (b) and masonryabutments and foundations (c).

Figure 17. Fractions of reduced, unchanged and increased safety factors FCi for the first and the second analysis of masonry/stonebridges.

C. Pellegrino et al.12

Dow

nloa

ded

by [U

nive

rsita

di P

adov

a] a

t 10:

37 2

9 A

ugus

t 201

4

Figure 18. Assessment outcomes and relative changes from the first to the second analysis for the structural elements of the masonry/stone bridge stock.

Figure 19. Results for the abutments of the analysed reinforced concrete bridges: top and base bending moments (a); shear at the baseand shear in the walls (b).

Figure 20. Results for the analysed reinforced concrete bridges: framed piers (a) and wall piers (b).

Structure and Infrastructure Engineering 13

Dow

nloa

ded

by [U

nive

rsita

di P

adov

a] a

t 10:

37 2

9 A

ugus

t 201

4

Figure 21. Fractions of reduced, unchanged and increased safety factors FCi for the first and the second analysis of reinforced concretebridges.

Figure 22. Assessment outcomes and relative changes from the first to the second analysis for the structural elements of the reinforcedconcrete bridge stock.

C. Pellegrino et al.14

Dow

nloa

ded

by [U

nive

rsita

di P

adov

a] a

t 10:

37 2

9 A

ugus

t 201

4

significant percentage of cases in which reduction of thesafety factors for shear if compared with the first analysisis observed.

Figure 22 suggests that in situ investigations hasallowed to find a reduction of the FCi values for framedand wall piers in the longitudinal direction. Figure 22shows the assessment outcomes and relative changes fromthe first to the second analysis for the structural elementsof the reinforced concrete bridge stock. The diagramconfirms previously described considerations, in particularrelated to the utility of in situ investigations for shearassessment of the piers (in more than half of the casescharacterised by unsatisfied outcomes) and, on the otherhand, shows how for bending assessment the resultssubstantially remain still satisfied. Referring to abutmentsand foundations outcomes, despite for these cases thein situ investigations have not led to relevant changes inthe assessment outcomes, the results have shown a generaldiffused vulnerability for these elements (still unsatisfiedresults: 55% for abutments combined axial force-bendingat SLV (Safe Life Limit State), 35% for abutments topbending at SLV, 39% for abutments base shear at SLV).

6. Conclusions

The paper provides insight on planning seismic vulner-ability assessment of large stocks of bridges with a realapplication to the road network of the Veneto Region,Italy. In the first part of the study, a preliminary seismicanalysis of 335 existing bridges was carried out. In thisfirst phase, the main mechanical characteristics of thestructural elements were often unknown and, as in theprofessional practice, some reasonable assumptions weredone according to the construction practice of the firstconstruction period.

The second part of the study, described in this paper,was focused on updating the preliminary seismic analyseson a cluster of 71 bridges considered as representative ofthe main typologies related to the initial 335 bridges.A series of in situ and laboratory investigations alloweddefining the main physical and mechanical characteristicsof the materials constituting the main structural elementsof this cluster of masonry/stone and reinforced concretebridges. Measurements related to in situ and laboratoryinvestigations have allowed to characterise in a morerealistic way the analysed bridges and improve seismicassessment. For each bridge, the significant safety factorsFCi have been calculated.

Measured quantities have been compared with theassumed ones, evaluating their ratios for each structureand thus evidencing the situations for which in situinvestigations show conservative (ratio lower than 1),correct/neutral (equal to 1) or unconservative (higherthan 1) assumptions. The fractions of conservative,correct/neutral and unconservative ratios between

measured and assumed values of the main mechanicalmaterials’ parameters have been calculated.

The results have provided evidence how the geometri-cal survey has played a useful role in the whole structuralassessment, in particular referring to spandrel wall failuremechanism that seems to be the most vulnerablemechanism for the masonry/stone bridges. The assessmentoutcomes and relative changes from the first to the secondanalysis for the structural elements of the bridge stockhave been shown with the aim of highlighting the caseswhen the inspections were more useful.

Regarding reinforced concrete bridges, safety factorsfor bending in abutments and piers are characterised by asubstantial increase of the FCi values after the inspections,whereas in situ investigations on materials’ properties haveled to a significant percentage of cases in which reductionof the safety factors for shear if compared with the firstanalysis is observed. Referring to abutments andfoundations outcomes, despite for these cases the in situinvestigations have not led to relevant changes in theassessment outcomes, the results have shown a generaldiffused vulnerability for these elements.

Finally, the use of properly planned in situ andlaboratory investigations on representative bridge struc-tural elements may represent an useful instrument for theexecution of careful seismic vulnerability assessment oflarge stock of bridges and allow avoiding too conservativeor unconservative assumptions for the estimation of themain mechanical characteristics of the materials and, as aconsequence, finding inaccurate structural safety factors.

Acknowledgements

The work has been developed in the research agreement betweenthe Department of Civil, Environmental and ArchitectureEngineering of the University of Padova and the Veneto’sregional managing roadway authority Veneto Strade SpA. Theauthors would also thank Engg. Michele Fanutti, SamueleLidron, Mauro Rampazzo and Giuseppe Ricci for theircontributions in developing some of the analysis models.

Notes

1. Email: marianoangelo.zanini@dicea.unipd.it2. Email: paolo.zampieri@dicea.unipd.it3. Email: claudio.modena@dicea.unipd.it

References

Biondini, F., Camnasio, E., & Palermo, A. (2014). Lifetimeseismic performance of concrete bridges exposed tocorrosion. Structure and Infrastructure Engineering, 10,880–900.

Carturan, F., Pellegrino, C., Rossi, R., Gastaldi, M., & Modena,C. (2013). An integrated procedure for management ofbridge networks in seismic areas. Bulletin of EarthquakeEngineering, 11, 543–559.

Carturan, F., Zanini, M.A., Pellegrino, C., & Modena, C. (2014).A unified framework for earthquake risk assessment of

Structure and Infrastructure Engineering 15

Dow

nloa

ded

by [U

nive

rsita

di P

adov

a] a

t 10:

37 2

9 A

ugus

t 201

4

transportation networks and gross regional product. Bulletinof Earthquake Engineering, 12, 795–806. doi:10.1007/s10518-013-9530-8.

Clemente, P. (1998). Introduction to dynamics of stone arches.Earthquake Engineering and Structural Dynamics, 27,513–522.

Clemente, P., Occhiuzzi, A., & Raithel, A. (1995). Limitbehaviour of stone arch bridges. ASCE Journal of StructuralEngineering, 121, 1045–1050.

Heyman, J. (1966). The stone skeleton. Earthquake Engineeringand Structural Dynamics, 2, 249–279.

Heyman, J. (1982). The masonry arch. Chichester, West Sussex:John Wiley & Sons Limited.

Hofmann, M., Kuhn, B., Friebem, H., & Winkler, B. (2012, July8–12). Pre-assessment of existing road bridges – Newprocedure for a rough but quick estimation of the capacity ofexisting road bridges. In 6th international conference onbridge maintenance, safety and management, IABMAS 2012,Stresa, Lake Maggiore, Italy. London: Taylor & FrancisGroup.

Italian Ministry of Infrastructures (2008). Norme Tecniche per leCostruzioni, DM 14/01/2008. Rome: DEI, Tipografia delGenio Civile. (in Italian).

Kamya, B.M. (2012, July 8–12). Bridge management systems:Challenge of adopting a bridge management systemappropriate to the needs of a local authority: Examplesfrom the United Kingdom. In 6th international conference onbridge maintenance, safety and management, IABMAS 2012,Stresa, Lake Maggiore, Italy. London: Taylor & FrancisGroup.

MidasFEA v2.9.6 (2009). Nonlinear and detail FE analysissystem for civil structures. Seoul: Midas InformationTechnology Co. Ltd.

Pellegrino, C., Pipinato, A., & Modena, C. (2011). A simplifiedmanagement procedure for bridge network maintenance.Structure and Infrastructure Engineering, 7, 341–351.

Rota, M., Pecker, A., Bolognini, D., & Pinho, R. (2005).A methodology for seismic vulnerability of masonry archbridge walls. Journal of Earthquake Engineering, 9,331–353.

Soderqvist, M.K., & Veijola, M. (2012, July 8–12). The newmanagement system of engineering structures in Finland. In6th international conference on bridge maintenance, safetyand management, IABMAS 2012, Stresa, Lake Maggiore,Italy. London: Taylor & Francis Group.

Strand 7 v2.4 (2010). Strand7 finite element system release 2.4.Sydney, NSW: Strand 7 Pty Ltd.

Torkkeli, M., & Lamsa, J. (2012, July 8–12). Guidelines andpolicy for maintaining and managing all engineeringstructures of the Traffic Agency. In 6th internationalconference on bridge maintenance, safety and management,IABMAS 2012, Stresa, Lake Maggiore, Italy. London: Taylor& Francis Group.

Yue, Y.C., Pozzi, M., Zonta, D., & Zandonini, R. (2012, July 8–12). Bayesian networks for post-earthquake assessment ofbridges. In 6th international conference on bridge mainten-ance, safety and management, IABMAS 2012, Stresa, LakeMaggiore, Italy. London: Taylor & Francis Group.

Zanini, M.A., Pellegrino, C., Morbin, R., & Modena, C. (2013).Seismic vulnerability of bridges in transport networkssubjected to environmental deterioration. Bulletin of Earth-quake Engineering, 11, 561–579.

Zonta, D., Zandonini, R., & Bortot, F. (2007). A reliability-basedbridge management concept. Structure and InfrastructureEngineering, 3, 215–235.

C. Pellegrino et al.16

Dow

nloa

ded

by [U

nive

rsita

di P

adov

a] a

t 10:

37 2

9 A

ugus

t 201

4