This article was downloaded by: [Politecnico di Milano Bibl]On: 19 February 2014, At: 08:25Publisher: 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

Inspecting and repairing old bridges: experiences andlessonsPier Giorgio Malerbaa

a Department of Structural Engineering, Politecnico di Milano, Piazza Leonardo da Vinci 32,Milan20133, ItalyPublished online: 12 Mar 2013.

To cite this article: Pier Giorgio Malerba (2014) Inspecting and repairing old bridges: experiences and lessons, Structureand Infrastructure Engineering: Maintenance, Management, Life-Cycle Design and Performance, 10:4, 443-470, DOI:10.1080/15732479.2013.769010

To link to this article: http://dx.doi.org/10.1080/15732479.2013.769010

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

Inspecting and repairing old bridges: experiences and lessons

Pier Giorgio Malerba*

Department of Structural Engineering, Politecnico di Milano, Piazza Leonardo da Vinci 32, Milan 20133, Italy

(Received 9 August 2011; final version received 18 April 2012; accepted 18 January 2013; published online 12 March 2013)

The concept of life cycle applied to building structures is quite recent. In the past, constructions were built to be everlasting,and the number of massive buildings and bridges still in service apparently confirms the soundness of this approach. Whenreinforced concrete was introduced, new possibilities were open to engineers, in a time when the end of a structure’s life wasstill considered as an extreme and very remote hypothesis. However, this belief was soon to be faced with the evidence of thedeterioration of the concrete surfaces and with the corrosion of steel bars and strands, so that the concept of durability had tobe introduced. It therefore became clear that the structures’ constructions have to be monitored and periodically maintained,so as to extend their service life as much as possible. This paper gives an account of studies and rehabilitation works carriedout on bridges located in the north of Italy, providing some remarks regarding the maintenance of structures of the past andthe design of structure for the future.

Keywords: bridges; maintenance; piles and piling; precast systems; rehabilitation, reclamation and renovation;strengthening

Introduction

Many bridges of Italian road and railway infrastructural

networks are now between 50 and 150 years old. The

structural typologies, the technologies and the materials

used for these bridges were very different from those used

today, and so were the traffic loads considered for the

structural checks. Although such bridges still provide

adequate functional and structural performances, never-

theless they require specific inspections and maintenance

operations, together with static and seismic adjustment

interventions.

In design practice, we used to refer to recurrence

intervals of centuries. For instance, in the European

countries, 200 years is the return period usually assumed

for floods. But, if we consider the real life of a structure,

how long do 200 years last? In fact, 200 years ago, the

material characteristics, the building technologies and the

theoretical know-how were very different from the

present ones. With regard to the materials, for instance,

the industrial process to produce cast iron was invented at

the end of eighteenth century (Cort and Wilkinson), while

a steel with modern characteristics was patented in the

middle of 1800 (Bessemer & Siemens and Martins). John

Roebling began producing wire ropes in Saxonburg in

1841.

The second half of the nineteenth century saw the birth

of new technologies, when the first experiments on

reinforced concrete were carried out (Lambot, Hyatt,

Coignet, Monier, Koenen). The formulation of the basic

criteria for reinforced concrete design (the steel works in

adherence with the concrete; concrete and steel form

together a system homogeneous on average) dates back to

1887, when the experimental studies and the remarks by

Wayss and Bauschinger were published. In 1902, Morsch

(1902) published his famous book Der Eisenbetonbau. The

first experiments and applications of pre-stressed concrete

structures were proposed by Freyssinet (1950), who

patented his ‘Procede de fabrication des pieces en beton

arme’ in 1928. In 1933–1936, Abeles (1937) carried out

the first experiments on partially pre-stressed beams and

decks.

At the same time, the nineteenth century also gave us

the basis of the strength of materials and of the theory of

structures (Timoshenko, 1953). In particular, with the

growing use of steel in structures, more complete

investigations of various types of truss structure became

necessary, and the first methods to deal with statically

indeterminate structures were proposed. The great arch

and truss steel bridges such as those of Gustave Eiffel

(Garabit Bridge, 1880–1884; Maria Pia Bridge in Porto,

1877) and of Anghel Saligny (Danube crossing in

Cernavoda, 1885–1890) date back to those times. The

first long-span reinforced concrete bridges appeared in the

first decades of the twentieth century, when Hennebique

q 2013 Taylor & Francis

*Email: malerba@stru.polimi.it

Structure and Infrastructure Engineering, 2014

Vol. 10, No. 4, 443–470, http://dx.doi.org/10.1080/15732479.2013.769010

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

built the Ponte Risorgimento (Rome, 1911) and Freyssinet

the Saint Pierre du Vauvrai Bridge (1919). A comprehen-

sive history of that pioneer age can be found, for example,

in the books of Trojanovic (1960) and Leonhardt (1982).

An epic history of the greatest cable-suspended bridges

and of their designers was written by Petroski (1996).

Dealing with old structures, we cannot avoid framing

them into the context of their time, before assuming any

sort of decision and/or choosing any sort of intervention.

The major aspects that must be taken into consideration

usually regard:

. Possible changes in the geomorphology of the

territory where the bridge is located. Topographic

surveys can detect if any settlements have occurred

both at a local and at a global scale due, for instance,

to subsidence. Another issue is represented by the

riverbed geometry, both in plan and in the shape of

bathymetric conformation and its variations.. Possible changes in the attitude of the bridge, made

evident by vertical irregularities of the road level, by

rotations of the piers, by excessive vertical

displacements of the deck or by abnormal horizontal

displacements of the bearing supports.. Conservation or damaging states of the structural

elements (foundations, piers, deck) and of the

joints, the bearing supports and the devices

intended to drain the water from the deck.

All these issues concern environmental influences on

the bridge service life, the loads consisting of self-weight,

stream flow, debris loading and wind. The distortions may

be caused by settlements of the foundations of the piers

and of the abutments. But a bridge is primarily a

transportation link and hence many bridges modifications

are caused by new traffic needs. Moreover, one remains

struck by the strong differences among the traffic loads

given by present codes and those assumed at the time of

construction: for instance, 100 years ago, the load train for

road bridges was made of a row of 16metric ton carts, 6m

long, pulled by a team of four couples of horses, 10m long

and (4 £ 1.4) ton heavy (13.5 kN/m). The maximum rail

bridge loads were 30–60 kN/m. It must be remembered

that, in those times, the speed was also lower and that

usually only two lanes were sufficient. Hence, when

possible, an actual rehabilitation of an old bridge involves

a widening of the road platform and an intervention on

those structural elements that need to be strengthened in

order to carry the new loads.

With all these considerations in mind, this paper gives

an account of studies and rehabilitation works carried out

on a group of bridges located in the north of Italy and

belonging to the main typologies used in the years between

1850 and 1970. Parts of these bridges lie on the reach of

the Po River, which delimits the southern border of

Lombardia, while other bridges cross minor rivers. Recalls

will be made on the surveys and monitoring activities and

on the problems addressed during the rehabilitation works,

while synthetic descriptions of the main interventions on

the foundations, the body of the main structure and the

special devices will be given.

Bridges and environment: survey of a group of bridges

along the Po River

Despite the modern materials employed and the refined

techniques used in the design process, when compared to

the massive structures of the past, recent bridges appear to

be more sensitive to the injuries of time. This emerged

clearly from the surveys carried out on some bridges

crossing the river Po, which will be presented in the

following sections.

The Po basin

The Po is the main Italian river. It rises in theWestern Alps

and flows to the Adriatic sea, 652 km away. Its

hydrographic basin is 74,970 km2 wide and receives 43

tributary rivers (of which, 22 are from the right riverside

and 21 from the left one), for a total length of 4500 km.

The total length of the embankments is 3564 km. It flows

across 7 regions (Valle d’Aosta, Piemonte, Liguria,

Lombardia, Emilia Romagna, Veneto and Trentino Alto

Adige), 24 provinces and over 3000 municipalities.

With the exception of the upper reach, until the end of

the nineteenth century the crossing of the Po was carried

out through river ferries and floating bridges. Due to the

wide span of its main branch, traditional masonry and

stone arch bridges involved limited spans resting on a high

number of piers, having basements in an insidious and

wandering riverbed. The first long-span bridges appeared

with the diffusion of the steel-truss girders having an

isostatic Gerber scheme. Many of these road and rail

bridges, although after some reconstruction work, are still

operating.

After the Second World War, thanks to new materials

and new technologies, regarding both foundation works

and piers and decks, many new bridges were built, with a

wide use of pre-stressed reinforced concrete. The main

challenge during the design and erection phases was not

represented by the spans (usually 50–70m on average),

but by the interaction with severe fluvial hydraulics,

characterised by cyclic floods, which often overtopped the

embankments, invaded the floodplains and sometimes

upset countries and villages at the two sides of the river.

One has a fairly complete knowledge of the historical

floods of the Po River since a remote age. In particular, the

main floods of the last 100 years occurred in 1926, 1951,

1994 and 2000.

P.G. Malerba444

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

A wide surveying campaign

After the 2000 flood, the Compartment for the Lombardia

region of the ANAS (Italian Agency for Roads) promoted

a campaign aimed to survey the state of the piers,

basements and foundations of the main bridges crossing

the Po and serving roads of national interest in the

Lombardia region. Such a campaign was meant to provide

a first evaluation of the bridges state and to detect possible

critical conditions of the piers and foundations, paying

particular attention to hydraulics causes of instability, such

as erosion and scour. The bridge main characteristics are

listed in Table 1 and their locations are shown in Figure 1.

Surveying and monitoring

Surveying and monitoring consisted of the following

activities:

. Preliminary examination of each bridge on the basis

of the documents (original drawings and reports)

available in the archives, examination of the results

of previous surveys and/or interventions carried out

on the bridge (widening of the main body and of the

approaching viaducts, strengthening of piers and

abutments, strengthening of the foundations, repla-

cement of joints and bearing supports, maintenance

of the drainage systems); gathering of information

from the surveillance staffs.. General visual inspection, after which a short report

and a booklet of pictures of the most significant

view and/or details, mainly regarding durability

aspects (surface state, clefts and cracks on

reinforced concrete and masonry elements; state of

the cover layer and of the reinforcement; state of the

painting on steel structures; joints and bearing

supports; parapets and accessories devices), were

prepared for each bridge.. Geometrical survey, with dimensional cross-checks

with the original drawings: when, in some cases,

these needed to be updated, the typical main

elements of the bridge were re-drawn (plan,

elevation and main sections). Detailed description

of local damage states.. Survey of the overall attitude of the bridge, with

respect to horizontal and vertical references; verti-

cality checks of the piers; levelling of the roadway;

installation of topographic datum points; detection of

slope discontinuities, reducing the travel comfort.. Bathymetric survey, in order to draw the geometry of

the riverbed, to detect possible scouring signals near

the piers and to achieve updated transversal and

longitudinal riverbed sections for the subsequent

hydraulic computations. These surveys were carried

out using a small boat. The position was determined Table

1.

Characteristicsofsomeofthebridges

exam

ined

inthepaper

(see

Figure

1).

River

Place

Yearof

construction

Total

length

(m)

Length

over

theriver

(m)

Piers

infloodplains

Piers

inriverbed

Pierandfoundation

types

1Po

Casalmaggiore

1958

1205.38

580.00

3þ

25

6Pile/piertype

2Po

Viadana

1967

1670.00

734.00

36þ

65

Hexagonal

columnsonpiles

3Po

Borgoforte

1961

1137.21

471.83

9þ

3þ

04

Pile/piertype

4Po

San

Benedetto

Po

1964–1966

613.00

613.00

54

Double-bladepiers

5Po

Ostiglia

1929/1947

511.00

511.00

15

Masonry

piers

6Po

Piacenza

1908/1947

1096.00

607.00

12þ

5þ

32

Masonry

piers

7Serio

Montodine

,1970

64.00

64.00

01

Masonry

piers

8Oglio

Pontevico

,1970

90.00

90.00

02

Pile/piertype

9Oglio

Sarnico

,1970

87.00

87.00

05

Pile/piertype

10

Oglio

Montecchio

1,1970

270.00

270.00

92

Circularpiers

onpiles

11

Oglio

Montecchio

2,1970

90.00

90.00

12

Circularpiers

onpiles

12

Oglio

Breno

,1970

403.00

403.00

15

2Circularpiers

onpiles

13

Po

PievePortoMorone

1961

1250.00

1250.00

10

5RectangularR.C.columnson

piles

14

Po

Becca

1912

1040.00

1040.00

39

Masonry

piers

Structure and Infrastructure Engineering 445

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

through a GPS system and the depth was measured

through echo soundings.On the basis of these data, the

riverbed geometry was modelled by means of a

software suitable to show plants with isohypses and

the main transversal and longitudinal sections needed

for the hydraulics computations.. Underwater surveys, aimed to detect cracks or clefts

in the submerged parts of the piers, as well as traces of

erosions or scour holes at the piers basis. In some

cases, these surveys were quite difficult, due to the

water turbidity, the flow velocity and the presence of

debris on the riverbed. The information collected

through underwater surveys was summarised in short

reports, with a description of local damage states, in

pictures and videos for each examined pier.. Vertical boreholes in the piers aimed to measure their

buried depth. Echo soundings are used to check the

actual depth of the piers above the piles.. Geognostic boreholes, with the performance of

standard penetration tests (SPT) and cone penetration

tests (CPT) and collection of disturbed and undis-

turbed soil samples, later submitted to laboratory tests.

For those cases in which previous geognostic surveys

had been documented, the new ones were distributed

so as to achieve new or complementary data.

. Partial reports explaining and summarising the results

of the surveys.

The relative cost of each survey activity is shown in

Figure 2, while the mean cost of the surveys versus the

overall length of the bridges is shown in Figure 3.

Laboratory and office activities

The results of the laboratory tests carried out on the soil

samples were used to define the load-carrying capacity of

the original foundation systems, as well as to design the

strengthening works on the insufficient ones. The office

activities concerned hydraulic engineering and structural

assessments. A first team of engineers studied the general

situation of the riverbed and its past and recent evolution,

and checked the hydraulic compatibility according to the

recent code, delivered by the Basin Authority, the former

so-called PoMagistrate, originally established in Parma by

Eugenio Bonaparte in 1808. Such a compatibility requires

the compliance with the following conditions:

. The design flood discharge must be no less than that

assumed to define the limits of the expansion

riverbed.

Figure 1. Position of the bridges listed in Table 1.

P.G. Malerba446

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

. The bridge must have a minimum hydrometric

clearance (free distance between the level of the

design flood and level of the intrados of the bridge).. The relative position of the bridge submerged

structures (piers and abutments) and the effects

provoked by the bridge interference must be taken

into account. Attention must be paid to the overflow

level and to the effects of flood debris.. The bridge must be safe under a structural point of

view.

A second team of engineers checked the structures.

Starting from the results of the geometric survey, for each

bridge the self-weight and the weight of the permanent

loads have been computed. Such loads were combined

with the traffic and hydraulic loads in order to verify the

piers and the relative foundations. The hydraulic loads

taken into consideration were the ones corresponding to

the maximum flood discharge combined with the worst

scour conditions.

Finally, for each bridge a comprehensive report was

compiled. These reports summarised all the previous

items, gave a final assessment of the actual state of the

bridge and highlighted its possible critical faults. When the

safety of the structure was at stake, suitable suggestions for

urgent interventions were provided. Each bridge was

classified and given a priority level which implied

recommendation for ordinary or extraordinary mainten-

ance activities or, in the worst cases, for radical

strengthening works.

Main results drawn from the bridge inspection

Many useful suggestions can be derived from these

inspections and assessments.

Change in the trends of the flood regime

Floods are a common experience for the people who live at

the Po riversides. Now, comparing the recent surveys with

the historical data, it is possible to observe a rising trend of

the maximum flood levels and, in recent years, a higher

frequency of the flood events. According to the experts of

environmental hydraulics, such a phenomenon is mainly

due to anthropic factors, such as a progressive water-

proofing of the basins, due to urban and infrastructural

growing, the removal of expansion zones and the increase

of river reaches confined by embankments. Climate

changes may also have contributed to these effects.

Figure 4 shows the maximum hydrometric levels recorded

by the Piacenza measuring station (around location 6 in

Figure 1) during the last two centuries and confirm these

remarks. As a consequence, the Po Basin Authority (1999)

recently prescribed to widen the floodplains and to adopt

more restrictive rules for the use of soil near the rivers.

Another element which was confirmed is the depth of

the scour in the rapid transient phase as determined

according to the recent Po Basin Authority specifications,

which agree with the most widely recognised formulations

Figure 2. Cost distribution for different types of tests.

Figure 3. Mean cost per meter of the surveys versus overalllength of the bridges.

Figure 4. Maximum hydrometric levels registered at thePiacenza gauging station in the last two centuries.

Structure and Infrastructure Engineering 447

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

(Hamill, 1999). These values of scour depth appeared

quite higher than those assumed in the past (Figure 5a,b)

and strongly condition the load carrying capacity

computation of the foundations and of the piles. As

regards the body of the piers, it was not found in a bad

situation. Some masonry piers presented losses of mortar

among the masonry courses. Both masonry and reinforced

piles presented traces of collision with the small boats

and ships that sail on the middle and final reaches of the

river.

For the main bridges, the position and the orientation

of the piers with respect to the main flow were judged

correct. In some minor bridges, spanning over tributary

rivers, some cases of wrong foundation basement were

found (Figure 5c,d). An insufficient deepening of the

basement and an orientation causing the maximum

interference obviously increase the erosion and the local

scour effects. Another general consideration regards the

soundness of our probabilistic design procedures, tuned to

values of return periods (100–200 years). The surveys of

the most recent bridges, for which it was possible to

compare the present riverbed profile with that of 30–40

years ago, showed cases of strong riverbed changes, with

the movement of the main current from one alignment to

another and also with the growth of temporary islands

downward the old main current. The underwater surveys

reported that, even after a long time from the end of the

flood event, a systematic encumbrance of debris remained

at the basis and along the body of the piers. This is a

problem of ordinary maintenance. But, who is in charge of

the debris removal? Is it the river authority or the bridge

authority? More simply: who has to pay?

Figure 5. Scour effects and debris action on the piers in the riverbed: (a), (b) collapses caused by the scouring action in bridges crossingriver Po’s tributaries; (c), (d) effects of position, orientation with respect to the water flow and depth of the foundations in a small river;(e), (f) multi-columns piers acting as grids in the formation of debris rafts.

P.G. Malerba448

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

Foundation stability

The most critical checks regarded the carrying capacity of

the foundations. In many cases, these checks, in the

presence of both the heavier hydraulic loading conditions

and the scour hole, revealed the need of strengthening

interventions. This latter phenomenon is particularly

severe because the removal of soil at the basis of the

pier reduces the carrying capacity of the piers itself and/or

of the piles underneath them, and contemporarily increases

the surfaces exposed to flow forces.

Deck structures

Although the main aim of the campaign was surveying the

state of piers, basements and foundations, quick checks

were carried out also on the upper structures. Some

masonry arches of the approaching bridges in the

floodplains showed cracks and clefts, mainly due to

foundation settlements. In general, such damages do not

have structural relevance. Reinforced and pre-stressed

concrete decks and structures, built in the years 1960–

1970, appeared in worse conditions, showing the usual

damage pattern: carbonated surfaces, spalling at the

corners of beams and columns and at the bottom of the

slabs, and bars at different level of corrosion. Local

damages due to an inefficient drainage system were very

frequent. Bearing supports and joints usually needed to be

removed and repaired or replaced by new ones.

Lessons and suggestions from the survey campaign

According to what was observed along the Po River and its

affluents, environmental changes, provoked by many

anthropic and climatic factors, are leading to a progressive

growth of the maximum flood levels and to a higher

frequency of the flood events. In this context, a first notice

regards the riverbed shapes (position and orientation of the

main stream and overall discharge), which may change

faster than the return periods of one or two centuries

usually assumed in the design practice.

A second remark concerns the serviceability of old

bridges, having masonry piers and caisson foundations,

and that of more recent ones, made of reinforces concrete

and resting on deep piles. It was observed that, with

respect to more recent piers, due to their shape and

proportions, old piers allow a better flow of solid debris

along the pier height, showing accumulation only at the

basis. The shapes of many modern piers tend to rake

branches, trees and any sort of solid debris, thus reducing

the flow section and increasing the drag force and the

vortex regime around the pier (Figure 5e,f). Moreover, the

old massive piers are less sensitive to surface erosion

actions and better fitted for strengthening interventions: in

fact, their wide body can be drilled and host integrative

piers helping the original foundations. In contrast, modern

piers present all the vulnerabilities of reinforced concrete

members and their repairing usually compels to work

around the original body, with all the problems posed by

the need to effectively link the new parts to the original

ones.

The needed repair intervention

Once the data acquired during the inspections and surveys

had been analysed, different repair interventions were set

up. These interventions were carried out on:

. the foundation systems (piers, basements and piles),

. the main structures of the bridges and

. their complementary or special devices.

Interventions on foundations

Nowadays, the need for an intervention on the foundations

arises mainly after their stability has been checked

according to new codes and regulations, which lead to

higher depths of the estimated scour holes and higher

values of the acting forces. This brings to refurbishing and

strengthening the original foundation systems in order to

obtain a suitable safety level. The strengthening interven-

tions are usually based on two contributions. The first one

consists of protecting the area surrounding the piers by

means of big bags containing massive stones. Such a work

stabilises the riverbed and leads to less severe expected

scour depths. The second intervention consists of

strengthening the pier basement. Usually, the strengthen-

ing is carried out in one of the following two ways:

. when the body of the pier is sufficiently compact

and massive, like in the case of masonry piers, new

piles are driven across the body itself;. when this is not possible or when previous repairing

interventions occurred in the volume of the pier,

new piles are driven around the perimeter.

A common problem of both these type of interventions

is the accomplishment of a robust connection between new

parts and old structures. Examples of solutions for

different foundations problems are presented in the

following.

Strengthening of the foundations of a bridge of thenineteenth century

The first stable crossing of the Adda River in Lodi dates

back to 1158, when Frederick I Redbeard authorised the

building of a bridge north-east of the city. The builder was

Muzio della Gatta and, thanks to this new link, Lodi

became one of the most prosperous cities of the hinterland

Structure and Infrastructure Engineering 449

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

of Milan. In the following centuries, more bridges were

erected and later destroyed. We have to notice that on 10

May 1796 Napoleon Bonaparte, fighting against the

Austrian army, crossed in Lodi with a wooden bridge

made of 57 spans and 200m long. In 1859, during the

Second Italian Independence War, this bridge was burnt.

In the years 1863–1864, the Milanese architect Gualini

built the present masonry bridge, made of nine shallow

arches spanning 18.90m each, for a total length of 175m

(Figure 6).

In 1970–1971, the deck of the Lodi Bridge was

refurbished in order to comply with the new traffic loads,

and a weir was built 200m downstream in order to protect

the basement of the bridge. In the following years, the weir

caused some floods involving the town riverside. This

brought, after new hydraulic assessments carried out in

2006, the decision of lowering the weir. This lowering,

which is scheduled for the summer of 2010, will increase

the flow velocity and therefore expose the foundations to

more severe service conditions. Thus, a geotechnical

assessment, based on laboratory tests on soil samples, and

a structural assessment were carried out, in order to

ascertain whether the bridge would sustain such worsen-

ing. These assessments found that the bridge was not safe

even before the change in the hydraulic regime, and

needed strenghtening works in any case.

This choice was confirmed also by the presence of some

foundation settlements already observed in the past. In fact,

in 1947, each foundation was strengthened by driving

(9 þ 9) piles along its sides, at a depth of 12m from the

bottom of the basement, plus 2m long (16 þ 16) auxiliary

piles. At the front and at the rear of the basement, (11 þ 11)

short piles, 5m long, completed a continuous curtain,

protecting the basement against the erosion and scour. The

heads of the piles were connected to the existing structure

by means of a hooping crown made of concrete, cast around

the pile perimeter, stitched along the sides through ties of

ordinary steel. The soil, confined by the continuous curtain,

was then injected by cement mortar.

The new intervention substantially recalls the old one.

Twenty new micropiles, having a diameter of 0.20m and

24.70m long, were drilled alternatively at a ^ 58 angle

along the sides of the basement (Figures 7 and 8). The

micropiles were reinforced by valved steel tubes having a

diameter of 127mm and a thickness of 10mm (Figure 8).

Their position was defined so as not to interfere with the

old piles. The old concrete crown was demolished step by

step, caring not to cut the reinforcing bars. The demolition

was accompanied by the contemporary reconstruction of

reaches of the new, wider crown. Twenty horizontal

threaded bars, having a diameter of 32mm, passing across

the pile body and placed at two different levels (Figures 7

and 9), strongly connect the two opposite sides of the

crown. Figure 10 shows an intermediate phase of the

works. During the strenghtening works, the bridge was

periodically checked by means of topographical surveys,

to detect possible displacements.

Strengthening of foundations of piers eroded at theirbasis: the Piacenza Bridge

The second intervention concerns the masonry piers of the

Piacenza Bridge, across the Po River, shown in Figure 11.

The bridge was opened in 1908 and identically rebuilt after

a bombardment during the Second World War. The bridge

is 1097m long. The central reach, 607m long, is made of

eight steel trusses spanning 75.25m and supported by

masonry piers. Figure 12 shows piers Nos. 1 and 2. Such

piers are placed in correspondence of the lowwater channel,

which is about 160m wide and is concentrated in only one

branch. The bathymetry in that zone is quite irregular, and

shows marked variations in the riverbed depth. In fact,

about 20m upstream of the bridge, a groin, protruded

towards the river, conveys the flow against pier No. 1,

which was found surrounded by the big hole shown in

Figure 12. The force intensities and the scour depth, as well

as the estimated load-bearing capacity of the foundations,

suggested strong interventions on piers Nos. 1 and 2, and

less extensive, though still heavy, interventions on the

remaining five piers and on the abutments.

The intervention on pier No. 1 first consisted of the

river bed stabilisation: the hole around the pier was filled

by means of fibre bags filled with sand, having a volume of

2m3 each, over which a 2-m thick layer of stone riprap,

having characteristic diameter of 0.70m, was deployed. In

the oval area immediately around the pier, the riprap

diameter was increased to 1.05m. The total volume

deployed around pier No. 1 was 14,125m3.

The original foundations had already been strength-

ened after the 1994 flood, when the soil characteristics

were improved by means of jet grout columns, reinforcedFigure 6. The Lodi Bridge (1863–1864).

P.G. Malerba450

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

by steel valved tubes, reaching a depth of 12m under the

foundation basement. The intervention designed after the

surveying campaign and shown in Figures 13 and 14

consists of creating a crown made of 33 micropiles placed

around the pier body. The volume inside the body was

already occupied by the previous jet grouting works. The

micropiles have a diameter of 0.40m and are reinforced by

valved steel tubes of diameter 127mm and thickness

14.2mm. The valves have a pitch of 0.50m in the first

20m and of 0.30m for the remaining length of 1.50m. The

piles were driven to a depth of 22.5m under the riverbed

level; this means a distance of 15m from the bottom of the

original basement. The total length of the new piles is

33.85m. The set of micropiles was protected by a sheet

pile driven to a depth of 12m under the estimated scour

depth (Figure 14). The volume between the sheet piles and

the pier was then filled by concrete. Special attention was

paid to linking the added foundation elements to the main

body of the piers and making the two parts cooperating in

carrying any added load, as shown in Figures 13 and 14.

Strengthening of foundations subjected to settlementmovements: the San Benedetto Po Bridge

The San Benedetto Po Bridge underwent strenghtening

interventions both on the foundations (Figures 15, 16 and

17) and on the deck (Figures 18). The foundations suffered

from settlements, which caused relative displacements and

Figure 7. The Lodi Bridge: transversal section showing the new deck, made of precast pre-stressed beams, the position of the new pilesand that of the threaded connecting bars.

Structure and Infrastructure Engineering 451

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

rotations between adjacent piers. The intervention

regarded four double-blade piers and consisted of the

following phases:

. driving of four new piles inside the perimeter

defined by the eight piles already existing

(Figure 17);

. casting of new transverses between the couples of

blades of the original piers;. pre-stressing of the transverses, obtained by means

of short cables (Figure 17) passing across the blades

so as to create a connection frame;. placement of four 2500 kN hydraulic jacks in the

niches left between the transverses and the piles

(Figure 17), in order to force the two parts against

each other;. locking of the heads of the piles against the

connection frame by using a steel spacer;. removal of the jacks and casting of the niches.

In this way, the new piles were forced to cooperate

with the pre-existing structure for new added loads.

Strengthening of excessively slender piers: theBorgoforte Bridge

In the Borgoforte Bridge, it was necessary to intervene on

three piers which showed excessive slenderness, vibration

Figure 8. The Lodi Bridge: arrangement of the micropiles at thetwo sides of a pier.

Figure 9. The Lodi Bridge: details of the R.C. hooping crown.

Figure 10. The Lodi Bridge: photograph of an intermediatephase of the works in 2010.

P.G. Malerba452

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

sensitivity and unsafe foundation conditions (Figures 19

and 20). The intervention consisted of adding two new

pairs of piles upstream and downstream (Figure 20) and

connecting them, by means of an arch-shaped frame, to the

original pier assembly (Figure 21).

Main structures of the bridges

The interventions needed on the main structures of a

bridge may be summarised as follows:

. strengthening of the carrying structure due to some

lack in their load-bearing capacity, caused, for

instance, by settlement effects, or by corrosion of

the reinforcing bars or the pre-stressing steel in

critical sections;

. refurbishment and adjustment of the structure to

new codes prescriptions, involving, for instance,

heavier load conditions;. changes in geometry, due to the need of widening

the road platform.

In the following subsections, the cases of refurbish-

ment regarding three different types of arch bridges and of

a pre-stressed cantilever bridge are examined.

Strengthening of a deck by means of a new hiddenstructure: the Lodi Bridge

In 1970–1971, the Lodi Bridge, described in a previous

paragraph with reference to the strengthening of the

foundations, underwent a refurbishment aimed to adapt the

Figure 11. The Piacenza Bridge (1908/1947): detail of the longitudinal view of the bridge, showing piers 1 and 2.

Figure 12. The Piacenza Bridge: the bathymetry, showing the holes in correspondence of piers 1 and 2.

Structure and Infrastructure Engineering 453

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

deck to the new traffic loads. In this case, the adaptation of

the original structure, made of nine shallow masonry

arches, to the new static requirements appeared a hard

task. The solution adopted by Professor Martinez y

Cabrera was a compromise between the necessity to obtain

a crossing suitable to deal with the intensity of new traffic

loads and to respect the original masonry structure, dating

back to the nineteenth century and subjected to the

supervision of the Local Department for the Preservation

of Historical Buildings.

The body of the nine arches was emptied of the

infilling material and the masonry structure was cleaned

and restored. In correspondence with the masonry piers

axes, new R.C. walls were cast. These walls, parallel to the

long sides of the piers and connected to their original

structures, were built in order to elevate the base for new

bearing support. Then, inside the free volume of the empty

masonry body, a new carrying structure, made up of eight

prefabricated pre-stressed concrete beams, was inserted

(Figure 7) and then made continuous by the casting of the

upper slab. It must be outlined that the new deck is slightly

higher with respect to the old structure. With the exception

of zones around the supports, the space existing between

the original structure and the new carrying structure makes

them totally independent. This also allowed to widen the

lateral sidewalks.

The railway bridge across the Gaggione River

The structure

The railway bridge across the Gaggione River was built at

the end of the nineteenth century and connects the city of

Milan to the city of Varese, 60 km away. The bridge,

shown in Figure 22(a), has a total length of 130m and is

made of a sequence of seven stone barrel arches, which

rest on six stone piers. The barrel arches have internal

radius of 5.82m and their centres are 14.00m apart. The

piers are slightly tapered: the longest one, 32.60m high,

has a section that varies from 10.00m (width) by 5.50m

Figure 13. The Piacenza Bridge: details of the connection between the pier and the new micropiles.

P.G. Malerba454

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

(front) at the bottom to 4.50m by 2.50m at the top. The

free height of the piers is interrupted by an intermediate

service deck made of five shallow arches.

In 1985–1986, the bridge was strengthened in order to

carry new, heavier train loads. The main interventions

regarded the basements of the central piers, the node at the

intersection between the piers and the intermediate deck,

and the strengthening of the intrados and spandrels of the

arches. The intrados and spandrels were enveloped in a

layer of shot concrete, 180mm thick, and reinforced with

14mm bars arranged so as to form a mesh with a pitch of

200mm both ways. After a strong hydro-sand blasting in

order to improve the chemical–mechanical adherence, the

added layers were linked to the stone surfaces through a

Figure 14. The Piacenza Bridge: details of the anchorage; horizontal section of a pier, showing the concrete infilling between the sheetpiles and the old pier; photograph showing the completed intervention.

Figure 15. The San Benedetto Bridge (1964–1966). The completed intervention on the piles can be seen.

Structure and Infrastructure Engineering 455

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

uniform curtain of pins and confined against the original

walls through 22mm threaded bars, piercing the body of

the arches. Other works regarded the river bed

stabilisation, obtained with the introduction of embank-

ments and weirs.

The problem

It is quite obvious that, considering this type of massive

bridges, any intervention which confines tightly the

masonry and, at the same time, does not add excessive

loads to the original structure leads to an increase in the

safety level. It is less immediate to define the actual static

scheme, which leads the forces at the basements, and how

old and new structures contribute to the overall bridge

robustness. Probably, the bridge was built following

empirical rules used in the past centuries to define the

thickness of the arch rib as a function of its span and of its

rise. It is possible that some more refined study was carried

out through the Mery method, widely used in the past to

assess the arch stability.

Nowadays the arch is analysed according to elastic and

limit theories (Heyman, 1982; Strassner, 1927). However,

the arch is still considered the bearing structure, while the

superstructures (the walls and the filling up, made usually of

light non-cohesive material) are considered as dead loads.

But, why is it so common to see clefts around the arches

keystones and even, in some cases, the loss of central bricks?

The bridge behaviour

In December 1986, at the end of the strengthening works,

severe load tests were carried out on this bridge. It was

loaded by the locomotive shown in Figure 22(c) and by a

Figure 16. The San Benedetto Bridge: longitudinal view of the bridge. The area where the intervention on the foundations was carriedout is highlighted.

Figure 17. The San Benedetto Bridge: longitudinal and transversal sections showing the positions of the new transverses, the new pilesand the niche where the 2500 kN hydraulic jack were housed.

P.G. Malerba456

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

convoy. Several load positions were examined. The

vertical displacements were measured through a set of

mechanical displacement transducers (6 upstream and 10

downstream), having a sensibility of 1/100mm. These

gauges measured the relative displacements between the

bridge extrados and a rigid, simply supported, reference

beam. The reference beam supports were placed in

correspondence of the vertical axis of the piers (local

reference system). An electro-mechanical displacement

transducer measured the vertical displacements at the

crown with respect to the ground (absolute reference

system). At the intrados and at the lateral walls of the

central section, eight electromechanical strain gauges

(relative displacement transducers) were placed. Their

sensitivity was 1mm over a basis of 300mm ( 3.3m1.These gauges were placed as shown in Figure 22(b).

Figure 22(c) shows the displacements obtained with the

train in the central position.

The maximum vertical displacement given by the

transducer with respect to the absolute reference system

was vabs ¼ 0.560mm. The corresponding displacement

with respect to the local reference system was vloc ¼0.405mm. The difference Dv ¼ 0.155mm is mainly due

to the piers deformation. A finite element analysis (FEA)

with plane stress elements gave at the top of the piers vFEA¼ 0.220mm. The results shown in Figure 22(c) are

rectified with reference to the straight line that connects

the top of the piers, as computed from FEA analysis.

Figure 22(b) shows the vertical distribution of the

horizontal strains compared to the stratigraphy of the

section: the added reinforcement layer and the arch rib are

in tension; the spandrels above the arch top and the infill

material are compressed. A fairly good matching between

numerical and experimental data was reached. A

consequent comparison in term of stresses gave Ds ¼(20.03) 4 (20.04) N/mm2 (FEA Ds ¼ 20.09N/mm2)

Figure 18. The San Benedetto Bridge: vertical and horizontal layout of the new external pre-stressing cables used in the strengthening ofthe cantilevers in the floodplains.

Figure 19. The Borgoforte Bridge (1961): longitudinal view of the bridge.

Structure and Infrastructure Engineering 457

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

at the top (transducer No. 5) and Ds ¼ 7.1 4 13.0N/

mm2 (FEA Ds ¼ 5.46N/mm2) at the bottom (transducers

No. 1, 6 and 7). Hence, during the service life and for the

applied loads, the crown of the bearing arch works in

tension and the superstructure in compression.

Analogous behaviour was found through an FEA for

the effects of self-weight. The theoretical final results of

the analysis (self-weight plus the weight of the E626

locomotive) gave, for the most compressed fibre, s ¼20.28N/mm2, a relatively small value which can be

sustained also by a moderately compacted soil. These

assessments, which trust the cooperation with materials

that cannot be defined as structural in the strict sense of the

word (i.e. the filling of the spandrels), cannot be used for

safety evaluations. Safety derives from the certainty that

the evolution from the service to the ultimate state would

involve the crushing of the filling material, while the line

of the thrust lowers until it reaches the extrados of the

bearing arch, which finally works according to the usual

interpretation of its behaviour. For the sake of history, this

latter situation was checked also through the Mery

method. However, this experience provided many sugges-

tions for similar interventions.

The strengthening of the San Benedetto Po Bridge deck

In a previous section the San Benedetto Po has been

discussed with reference to the reinforcement of the

foundations. Now, for the same bridge, the strengthening

of the deck will be considered. The bridge characteristics

and the location are presented in Table 1 and Figure 1. The

bridge was built in the years 1964–1966 by means of a

technology widely employed at that time, which used

cantilevers 28.5m long, supporting 10.0-m long closing

suspended girders (Figure 16). The length between two

Figure 20. The Borgoforte Bridge: (a) transversal section showing the original pier/pile assembly; (b) transversal section showing thetwo piles added at the sides of the existing ones and the arch-shaped frame; (c) longitudinal section after the completion of theintervention.

Figure 21. The Borgoforte Bridge: photograph taken after thecompletion of the works.

P.G. Malerba458

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

Figure 22. The Gaggione Bridge (1885): (a) longitudinal view of the bridge; (b) transversal section after the strengthening, layout of therelative displacement transducers and distribution of the strains in the depth of the section during the loading tests; (c) comparisonbetween theoretical and experimental vertical displacements.

Structure and Infrastructure Engineering 459

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

piers of a typical reach is 67.00m. The road platform is

11.00m wide, and it is subdivided into two 3.25m wide

traffic lanes and in two 1.75m wide sidewalks (Figure 17).

The piers are made of a couple of parallel reinforced

concrete blades, having adequate bearing capacity and a

flexural deformability suitable to allow the shortenings

and the elongations of the deck.

Since 1988, the bridge showed irregularities in the

roadway, rotations of the piers and excessive vertical

displacements. These anomalies were partly due to

settlement of the foundations and partly to a loss of pre-

stressing mainly deriving from the effects of creep and

shrinkage. In some cases, the deflections at the tip of the

pre-stressed cantilevers reached 100mm. The bridge was

kept in service but traffic limitations were imposed. As

regards the intervention on the foundations, they were

described in a previous paragraph. In this paragraph, a

brief account will be given of the works carried out on the

deck. The aims of these works were:

. to recover part of the loss of pre-stressing;

. to compensate possible reductions of the cable

sections due to corrosion;. to recover part of the permanent deflections.

Figure 18 shows the solution adopted for the reaches

on the floodplains, which, for each cantilever, involved:

(1) the casting of three transverses that house the

anchorages of a new set of pre-stressing cables;

(2) the placement of external cables anchored in the new

transverses and placed as follows:

. No. 6 17.00m long cables, placed inside the deck

boxes and astride the piers, which connect the

couple of transverses nearest to the piers;. No. 2 41.00m long cables, placed in the space

between the two deck boxes and astride the piers,

which connect the two intermediate transverses;. No. 2 cables 53.00m long placed as before, which

connect the two end transverses;. No. 3 cables 17.00m long placed in the space

between the two deck boxes and connecting the end

transverses to those nearest to the pier.

For the reaches over the river, all the cables were

placed inside the deck boxes. These works, together with

repair of the damaged R.C. parts, of the road platform and

of the water drainage systems, allowed the bridge to re-

open to the normal traffic service.

Auxiliary and special devices

The service life of a bridge is strongly influenced by the

regular functioning of all its different components, each

one intended to carry out a specific function. A special

Figure 23. (a) Effects of leakage in a lateral sidewalk; (b), (c) worn rubber covering, torn up bolts and permanent deformation in a jointafter a few years of service.

P.G. Malerba460

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

mention must be made to the water drainage system. It has

a relatively modest cost if compared to other parts of the

bridge, but it may cause severe damages. Leakages

through any points of weakness in the waterproofing

system or in the expansion joints may lead to significant

reductions of the structural safety. Figure 23(a) shows the

intrados of a sidewalk of an old arch bridge (Cremeno

Bridge) before its repairing. Other vulnerable systems are

the expansion joints. Figure 23b,c shows the state of a

typical joint after few years of service (8 years in the case

of the photos): worn rubber covering, torn up bolts and

permanent deformations are clearly visible.

A cause of these drawbacks can be found in the

dynamic effects due to road platform discontinuities in

correspondence of the joints and to poor attention paid to

the details during the construction works. For short and

medium spans, when the horizontal excursions are limited,

a solution which strongly reduces these effects is shown in

Figure 24. The basic idea is quite simple: a steel

diaphragm, welded to one of the two edges of the joint, is

free to slide over the other edge. In this way, the gap

between the parts of the joint is protected, while the road

platform shows no discontinuities. Two small grooves at

the ends of the bituminous coating astride the joint will

reduce the risk of cracks just in correspondence of the joint

itself.

A malfunctioning in these devices can be found during

a visual inspection. Unfortunately, there are other devices

that can work in a wrong manner and that cannot be

detected without specific inspection activities or whose

repairing/substitution requires special interventions.

Substitution of the bearing supports of an old railwaybridge: the Mizzoccola Bridge

The Mizzoccola Bridge belongs to a small railway line that

connects the Italian city ofDomodossola to the Swiss city of

Locarno, 52.2 km away. The line was built in the years

1920–1923. The bridge (Figure 25) was opened in

November 1923 and afterwards it was systematically

monitored and maintained. In 2005, the Societa Subalpina

Imprese Ferroviarie (SSIF) Company, which owns the

bridge, detected traces of incorrect movements in one of the

steel bearing supports (Figure 26c,d) and then decided for

their substitution. This work implied the solution of two

different problems:

. the lifting of an old, statically indeterminate three-

span continuous beam;. the choice of suitable lifting points in the zone

occupied by the original bearings.

The first problem was solved through structural

analyses, which allowed us to determine the maximum

differential displacement between the supports the bridge

would bear without excessive stresses induced in the steel

structure. The second problem required the set up of a

technique that, without causing damages to the structure,

allowed the lifting by means of hydraulic jacks of an old

truss made of riveted steel elements, composed of thin,

non-weldable L section and plates. The problem was

solved by flanking the elements converging to the support

nodes with modern steel plates and clogging the interstices

with low-strength concrete. Over the plates, a transversal

beam was welded: at its ends, the beam received the force

conveyed by the jacks. All the auxiliary supports were

Figure 24. Sketch of a hidden joint, suitable for short andmedium span bridges.

Figure 25. The Mizzoccola railway bridge: front and plan view

Structure and Infrastructure Engineering 461

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

made at the same time. The lifting was carried out

separately for each support.

Figure 26(e) shows the plates of the original supports

after milling and restoring. Figure 26(f) shows the new

cylindrical supports. A historical comparison may be of

interest: at the load tests carried out on 6 November 1923,

the maximum vertical deflection at the middle of the right

lateral span was 10.00mm. For the same loading

condition, at an intermediate test carried out on 25

October 1985, the deflection at the same point was

11.21mm. After the recent works, these tests were

repeated on 29 April 2005 and the deflection was

11.06mm. There are no records of the original deflections

at the middle of the central span: the deflection measured

in 2005 was 16.20mm.

Substitution of the end restrains and of the joints in along pre-stressed bridge: the Grosotto Viaduct

This case considers the effect of the shortening of a

cantilevering erected bridge, which belongs to the stretch

of the national road no. 38 in the Grosotto municipality.

The bridge is composed of precast segments made

continuous over a total length of 913m (Figure 27). The

time-dependent shortening due to creep and shrinkage

provoked the disarticulation of the plate across the edges

of the joints (Figure 28c), an excessive sliding of the

external bearing supports (Figure 28b) and the detaching

of the end restraints against the abutments (Figure 28a).

The refurbishment consisted of repositioning the lower

plates of the bearing supports and strengthening the

same supports against seismic actions by means of steel

plates. The end restraints were overhauled and reposi-

tioned against a previously shimmed part of the

abutments.

The strengthening of the ties of one of the first tiedbridges: the Polcevera Bridge

According to the actual concepts of maintenance, the stays

are considered as parts that can be substituted when

necessary. Nevertheless, when the number of stays is very

limited, it is difficult not to involve the entire structure. In

the following section, the refurbishment of the ties of one

of the three A-shaped frames which characterise the

Polcevera Bridge is presented.

Figure 26. The Mizzoccola Bridge: (a), (b) the auxiliary system adopted to lift the bridge; (c), (d) the original bearing supports after theirremoval; (e) the plates of the original supports after milling and restoring; (f) the new cylindrical supports.

P.G. Malerba462

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

The structure

The Polcevera Bridge was designed by Riccardo Morandi

(Rome, 1902–1989), built in the years 1960–1964 and put

in service in September 1967. It flies over a large railway

parking lot and connects the A7 Genoa–Serravalle

highway to the A10 Genoa–Ventimiglia highway, which

reaches the French border. The bridge is 1121.4m long

and 18.00m wide. Its main stretch is characterised by two

parts (Morandi, 1967). The first part is composed of six V-

shaped piers, which sustain short cantilevered decks

supporting suspended girders 36.0m long. The second part

is composed of three A-shaped frames 90.2m high,

supporting decks 171.9, 171.9 and 145.7m long.

The decks are connected by 36.0-m long suspended

girders. At a distance of 10m from their ends, the long

decks are suspended to a couple of ties, made of pre-

Figure 27. The Grosotto Viaduct on National Road no. 38: longitudinal view, typical deck section and pier section.

Figure 28. The Grosotto Viaduct: (a) detachment of an end restraint from the abutment; (b) excessive sliding of one of the externalbearing supports and (c) disarticulation of the plates across the edges of a joint.

Structure and Infrastructure Engineering 463

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

stressed concrete. This scheme repeats concepts already

adopted by Morandi for the Maracaibo Lagoon Bridge

(Venezuela, 1957–1962, Consorcio Puente Maracaibo)

and then for the Rio Magdalena Bridge (Barraquilla,

Columbia, 1969–1972, Lodigiani) and the Wadi Kuff

(Beida, Libia, 1965–1971, C.M.C.) (Morandi, 1969). In

the following section, some aspects regarding the

construction phases are recalled, in order to allow an

understanding of the repairing interventions carried out on

the bridge.

The concept of stay according to Morandi

The three A-shaped frames mentioned above form

independent balanced systems, each spanning on four

supports and carrying a five-cells deck which is 171.9m

long and 18m wide (Figure 29). The two internal supports,

41.64m apart, are placed at the top of two pairs of

diverging inclined legs. The other two supports of the deck

are represented by the lateral ends of the transverses

placed at a distance of 10m from the tips of the

cantilevers. The ends of the transverses are, in turn,

suspended at the two pairs of ties. These ties are probably

the most characteristic elements of the Morandi system.

The tie sections are shown in Figure 30.

The construction sequence of the A-shaped frames can

be summarised in the following steps:

. Construction of two balanced cantilevers, made of

nine successive segments, each 5.50m long. During

this stage, the deck was supported by the external

pre-stressing action exerted by temporary bundles of

wires, anchored at the ends of each couple of

segments. These bundles spanned over the extrados

of the deck, passing over a set of 2.10m high

trestles, placed in correspondence of the two internal

supports. The trestles were adopted to increase the

internal lever arm and to give some vertical

component to the forces acting in the wires. At a

10-m distance from the end of the cantilevers, a

massive transverse received the anchorages of the

actual suspension ties. At this temporary stage, the

transverses were only partially cast, in order to limit

the weight.. After the partial completion of the cantilevers, the

temporary suspension action was transferred to a set

Figure 29. The Polcevera Bridge (1960–1964): general view of the bridge, an A-shaped frame, the antenna and the typical section ofthe deck.

P.G. Malerba464

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

of bundles of wires, passing over the top of the

antenna and anchored against the six vertical walls

of the box girder. After the activation of this new

supporting system, the end transverses were

completed.. At this stage, it was possible to spin, for each tie, the

suspension cables, arranged as shown in Figure 30.

The end transverses were partially pre-stressed and

the cables were partially tensioned.. The temporary supporting cables were removed.

The final 10-m long stretches of the cantilevers were

cast by means of a movable truss formwork. The

pre-stressing of the deck was completed and the

external temporary bundles of wires spanning over

the extrados were removed.. At this point, each suspension cable was integrated

by 28 auxiliary tendons, composed of four strands

and arranged as shown in Figure 30. The whole

bundle of main and auxiliary tendons, each of them

enveloped by its own sheath, was encased in a

concrete covering, cast in segments and having a

mean section of 0.98m by 1.22 m. It must be noted

that, at the casting time, the main tendons were

anchored at the ends of the external transverses and

partially tensioned, while the auxiliary tendons had

temporary anchorages in niches over the extrados of

the deck and were still not tensioned.. After the concrete of the case enveloping the

tendons had reached a suitable resistance, the

auxiliary tendons were tensioned in order to put

the concrete case under compression. Until this

stage, the cables remained free to move in the

concrete case, the sheaths still not being injected.. By means of suitable couplers, the auxiliary cables

lengths were extended and their ends anchored at

the transverses. The lower end of the case around

the tendons was completed and then pre-stressed.

The pre-stressing was given through the auxiliary

tendons tensioned from the side of end

anchorages.. The construction sequence was completed with the

injection of the sheaths, which made the tendons

and their enveloping case collaborating in carrying

any new added load.

Such a sequence is no doubt complicated, but has a clear

aim: to create ties that behave as a homogeneous system

made of tendons working in tension and of a pre-stressed

Figure 30. The Polcevera Bridge: composition of one tie and its sections.

Structure and Infrastructure Engineering 465

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

concrete case, working in decompression, but not in

tension, under the added loads (suspended girders, finishing

works, traffic, wind and temperature loads). In this way, the

fatigue effects in the strands were limited, thanks to the

reduction of stress variations due to variable loads, and at

the same the strands were protected against corrosion.

Needed repair interventions

After about 25 years of service, many parts of the bridge

presented severe damage states. On the ties of frame No.

11, at the Genoa side, clear corrosion traces in the strands

of the tendons appeared. Minor damages were detected on

the tendons at the top of the antenna of the nearby frame

(No. 10) and in other parts of the bridge. In 1992–1994, a

recovery programme was carried out under the guidance of

Francesco Pisani, who was one of Professor Morandi’s

aides at the time of the bridge design, and planned the

repairing intervention phases (Martinez y Cabrera &

Pisani, 1994). The main intervention concerned the four

ties of frame No. 11. The basic concept of the intervention

was to flank each original tie with a set of 12 additional

modern cables, in order to transfer the suspension action

from the ties to the stays.

The repair intervention sequence

The intervention was aimed at creating a new suspension

system, adding new elements to the original structure with

the following sequence:

. Placement of (2 þ 2) new anchorage boxes, at the

top of the antenna. Each box was made of two steel

plates, each having three niches and connected by

steel stiffeners (Figure 31a).. Placement of collar guides along the catenary of

the original ties. The collars, placed at a pitch of

4m, had 12 holes where the new cables were

housed (Figure 31b).. Placement of lower anchoring devices at the

abutment at Genoa side (Figure 32). The abutment

was drilled to allow the new cables to cross the

thickness of the original concrete structure. The

abutment structure was strengthened by means of

Titan bars.. Placement of lower anchoring devices at the ends of

the transverse beam towards frame No. 10

(Figure 33). This device was made of two steel

plates which formed a cellular ribbed structure

(Figure 31c).. Placement of a set of 12 additional cables for each tie.

The cables were made of 22/0.6 inch low-relaxation

strands, having a section of 150mm2. Each hot dip

galvanised strand, as well as the group of the strands

which formed the cable, were encased in high-

density polyethylene (HDPE) sheaths. The cables

were mounted in steps.. Placement of a set of short cables, later used to tune

the tension in the new stays and in the concrete

original ties.

Transfer of the tension force from the ties to the stays

The transfer of the tension force from the ties to the stays

was obtained with the following sequence:

. Placement and partial tensioning until 250N/mm2

of four long stays (Figure 34, tensioning phase 1).. Placement of the remaining eight long stays. Tension

reduction of the first four stays and contemporary

increase of the tension in four of the new eight cables.

The tensile stress in all theseeight cableswasequalised

at 150N/mm2 (Figure 34, tensioning phase 2).. Progressive demolition of the last reaches, 7.00m

long, of the damaged parts of the concrete ties and

partial reconstruction for a length of 6.50 m

(Figure 35). The gap of 0.50m was used for the

subsequent cutting of the old cables. For the

reconstruction, a high strength fibre reinforced

concrete, having f’c ¼ 50N/mm2, was used. In the

reconstructed part, 2 £ 40 bars having a 36mm

diameter were inserted to fix secondary anchoring

devices. These new added devices anchored one of

the ends of the new short cables. The other end

was anchored at lower terminal plates (Figures 31c

and 33).. Placement of the (3 þ 3) short stays, made of

31/0.6 inch low-relaxation strands. Tensioning of the

six short cables at 815N/mm2 (Figure 34, tensioning

phase 3), for the set at the Savona side, and at 910N/

mm2, for the set at the Genoa side.. Cutting of both long and short old cables.. Progressive tension increase in the new long cables

and progressive tension decrease in the new short

cables, according to phases 4–9, shown in Figure 34.

At the end of the tensioning phases, the long cables

were set at a stress of 430N/mm2 and the short cables

at 230N/mm2 (Savona side) and 270N/mm2 (Genoa

side).. Cable injection and filling of the gap of 0.50m at the

lower end of the ties.

All these phases were systematically monitored. The

main controls regarded the following:

. vertical displacements at the middle and at the lower

ends of the stays;. horizontal displacements at the top of the antenna;. stress variation in three sections of the deck and in

four sections along the stays.

P.G. Malerba466

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

The repair works concerned also the other two frames.

On frame No. 10 a local repair, aimed to strengthen the

upper end of the ties, was carried out. On frame No. 9,

whose cables appeared less damaged with respect to the

previous ones, no particular interventions were adopted.

Surface protection interventions were carried out on all

three frames.

Conclusions about the repair intervention of the Polcevera

Bridge

As already mentioned above, the basic concept of the main

intervention was to flank the original ties with new cables

and to transfer the suspension action from the ties to the

stays. Through specially designed devices (the collars and

the new anchorage systems) and following the recovery

sequence (the progressive tensioning phases), the new

‘composed’ stay system, resulting from the coupling of the

old ties to the new cables, maintains its original shape,

while the stiffness characteristics remain very close to the

original ones. This is important in order to maintain the

original design behaviour of the bridge and to avoid any

change in the deflection and flexural behaviour of the deck,

with consequences also on the elements of the main frame.

Another aim of the progressive tension transfer from

the old ties to the new composed stays was to reduce the

structural risk of excessive compression stresses in the

concrete ties, avoiding potential bursting effects. In fact,

Figure 31. The Polcevera Bridge: (a) anchorage plates at the top of the antenna; (b) collar fastening of the original stays and (c) lowerend ribbed plate.

Structure and Infrastructure Engineering 467

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

the operation of transferring the carrying action from the

old system to the new one involves the cutting of the old

cables. The old cables were fully collaborating with the

concrete section of the original tie. If they were not hold

back, they would have discharged the original pre-

stressing force entirely on the concrete section. To this

purpose, the short cables served to bypass the section

where the old strands were cut and, by tuning the tension in

the new long and short cables, to graduate the load

transfer. At the end of this process, the compressive stress

in the concrete ties was about 10N/mm2, as the original

design assumed. Finally, it must be pointed out that the

interventions were carried out without traffic interruptions.

Only some traffic limitations were needed during the

demolition and tensioning phases.

General considerations on the Polcevera Bridge

The Polcevera Bridge and other Morandi tied bridges

represent an exceptional reference from the conceptual,

aesthetic and technical point of view, which is even more

relevant if related to the times in which these structures

were built. Nowadays, however, similar static schemes,

though brilliant, cannot be proposed. According to the

modern criteria of durability, the ‘pre-stressed concrete

tie’ does not appear as a safe solution for elements in

Figure 32. The Polcevera Bridge: anchorage of the tie to the abutment.

Figure 33. The Polcevera Bridge: anchorage of the stays to thedeck transverse.

Figure 34. The Polcevera Bridge: diagram showing theevolution of tensile stresses in the long and short cables, duringthe tensioning phases adopted in the strengthening works.

P.G. Malerba468

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

tension. Moreover, the suspension action entrusted to a

limited number of elements makes the whole structure

little robust and the maintenance actions quite difficult.

Modern bridge configurations, characterised by a relative

great number of stays (a ‘curtain of stays’), are designed so

that, should the failure of one of the many stays occur, the

subsequent loss of suspension action would be made up for

by other suspension elements, making the cables

maintenance and/or substitution easier.

Conclusions

At the end of this paper, the following conclusions can be

drawn. As regards the relationship between the bridges and

the environment, on the basis of what was observed, the

actual trend to avoid or to limit the number of piers in

riverbed, and to prefer an increase in the span of the deck,

is confirmed and recommended. Possible piers in the

riverbed must have strong foundations, surrounded by

suitable riverbed stabilisation devices. The piers must be

correctly placed with respect to the flow.

As far as existing bridges are concerned, it seems that

the old massive and well-shaped piers behave better than

some types of piers built in the 1960s and 1970s, which

tend to rake solid debris and are more vulnerable.

Moreover, a massive pier makes it easier to carry out

integrative works aimed at helping the original foun-

dations. Speaking about the main structures, the usual

remarks on the durability of reinforced and pre-stressed

concrete structures can be recalled. A proper choice of the

type of concrete, a correct curing process, adequate cover

and detailing of the reinforcing bars may considerably

lengthen the service life and reduce maintenance costs. A

particular attention must be paid to the drainage system.

The lack of efficiency of the drainage system is one of the

main causes of damages and corrosion both in steel and in

concrete structures. Any effort to eliminate joints or, at

least, reduce their dynamic effects must be done.

Acknowledgements

I would like to acknowledge the members of my design team:Emanuele Barbera, Paolo Galli, Chiara Malerba, Marco diDomizio, Giacomo Comaita and Giancarlo Mineo. Many of theworks presented were carried out in cooperation with lateProfessor Francesco Martinez y Cabrera, who held before me thechair of Bridge Theory at the Politecnico di Milano, to whom mymemory and my gratitude go. The hydraulic studies were carriedby the ETATEC Company, under the supervision of ProfessorAlessandro Paoletti (Politecnico di Milano). The Polceveraintervention was conceived by Francesco Pisani, under thesupervision of F. Martinez y Cabrera. All Italian authorities whopromoted special investigations and demanding repairing worksare gratefully acknowledged. Among these are as follows: theANAS (Italian Agency for Roads) Compartment for theLombardia region, Milan; The Lodi Municipality, Lodi; theTechnical Office of the Lecco Province, Lecco; FNM SpA (NorthMilan Railways), Milan; SSIF Spa (Swiss-Italian RailwayCompany), Domodossola.

References

Abeles, P.W. (1937). Versuche mit Rechteckbalken, bewehrt mitbesonders hochwertigem Stahl. Beton und EisenbetonbauNo., 17, 18.

Freyssinet, E. (1950). Souvenirs, Conference prononcee par M.E.Freyssinet a Paris. Beton und Stahlbetonbau, February.

Hamill, L. (1999). Bridge hydraulics. London: E & FN Spon.Heyman, J. (1982). The masonry arch. New York: Ellis Horwood

Series in Engineering Science.

Figure 35. The Polcevera Bridge: phases of demolition and reconstruction of the end zones of the ties.

Structure and Infrastructure Engineering 469

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4

Leonhardt, F. (1982). Brucken – bridges. Stuttgart: DeutscheVerlags-Anstalt GmbH.

Martinez y Cabrera, F., & Pisani, F. (1994). Rehabilitation ofstays of the Polcevera Viaduct. Proceedings of theInternational Symposium on Cable Stayed Bridges(pp. 640–665). Shanghai.

Morandi, R. (1967). Il Viadotto Polcevera dell’AutostradaGenova-Savona. L’Industria Italiana del Cemento No., 12,849–872.

Morandi, R. (1969). Some type of tied bridges in prestressedconcrete. Concrete Bridge Design ACI, SP-23, 447–465.

Morsch, E. (1902). Der Eisenbetonbau. Stuttgart: Wittwer.Petroski, H. (1996). Engineers of dreams: Great bridge builders

and the spanning of America. New York: Vintage Books.

Po Basin Authority (1999). Criteri per la valutazione della

compatibilita idraulica delle infrastrutture pubbliche e di

interesse pubblico all’interno delle fasce A e B. Delib. N.

2/99.

Strassner, A. (1927). Neuere Methoden zur Statik der

Rahmentragwerke und der elastischen Bogentrager. Berlin:

Verlag von Wilhelm Ernst & Sohn.

Timoshenko, S.P. (1953). History of strength of materials. New

York: McGraw-Hill.

Trojanovic, M.S. (1960). Mostovi od armiranog I prednapreg-

nutog betona do 1960. Beograd: Zavod za Izdavanje

Udzbenicka Socijalisticke Republike Srbije.

P.G. Malerba470

Dow

nloa

ded

by [

Polit

ecni

co d

i Mila

no B

ibl]

at 0

8:25

19

Febr

uary

201

4