MANJUL MATHUR

Dy.CE/C/Stores, S.C Railway

S.VENKATA KUMAR
Dy.CE/TS, S.C Railway

KAILASH SINGH
Dy.CE/West/CN/BNC, S.W Railway
 !" #$ % &

The authors acknowledge their gratitude for the valuable

guidance provided by Shri A.K. Yadav, Sr.
Professor/Bridges/IRICEN/Pune, for without his valuable
suggestions and inspiration this paper could not have
been possible.

1
 CONTENTS

1. Introduction
2. What is assessment
3. Why assess
4. Types of analysis available
4.1 Empirical models, the prime one being MEXE
4.2 Two Dimensional Equilibrium based (Mechanism)
4.3 Elastic Methods
4.4 3D models using FE
5 Method presently used on Indian railways
6 Recent development in assessment
7 Limitations of Theoretical Analysis
8 Possible damages in masonry arch
8.1 Foundation damages
8.1.1 Damages due to the element degradation
8.1.2 Corrosion of steel elements used in foundation
8.1.3 Damages due to soil- foundation degradation
8.2 Superstructure damages
8.2.1 Damages resulting from a bad resistance performance
8.2.1.1 Gravitational actions
8.2.1.2 Imposed movements
8.2.1.3 Problems caused by abutment overturn by excessive earth pressure
8.2.1.4 Bulging of spandrels
8.2.1.5 Damages in wing walls
8.2.1.6 Stepped cracking
8.2.1.7 Transversal cracking resulting – arch mechanism failure
8.2.1.8 Loss or dislocation of pieces
8.2.2 Damages caused by deficient durability
9 Strengthening & retrofitting of arch bridges
9.1 Saddling
9.2 Sprayed Concrete
9.3 Jacketing below intrados
9.4 Surface Reinforcement
9.5 Insertion of Box inside the arch
9.6 Retrofitting
10 Conclusion
11 References

2
1.0 INTRODUCTION

In general bridges have been constructed from times immemorial even before
their theoretical behaviour was known. Arch bridges were constructed in a
large number for roadways and subsequently for the railroads. Bridge owners
were empowered to restrict the maximum weight/axle load of the vehicles
crossing any individual bridge on the basis of strength assessment available
at that time. However the validity of these methods soon began to be
questioned, as vehicles far in excess of the permitted weights/axle loads
seem to be crossing some of the bridges, without any apparent ill effect on the
structure. Further, with the passage of time, the need to permit heavier
weight/axle loads also arose. This necessitated the need to assess the
inherent load carrying capacity of these bridges, especially their residual
strength and if need be their retrofitting/strengthening.

This paper is an attempt in the strength assessment of arch bridges and their
strengthening if warranted based on the current load carrying capacity. This is
of particular relevance in the present Indian Railway context wherein it is
being attempted to increase the speed and load carrying capacity of the trains
over bridges, which were constructed long ago to the then prevailing
standards.

2.0 WHAT IS ASSSESSMENT

Assessment is the quantitative determination of the capacity of a bridge to

carry static and dynamic loads. It is the combined effect of geometric form of
structure, the materials used, the structural interaction of the parts and the
condition of the structure. The objective of load assessment is to determine
the load that the bridge can carry with a reasonable probability that it will not
suffer serious damage incapacitating it to serve the intended use.

The term assessment is carefully chosen. The capacity of an arch bridge

cannot be determined, only assessed. The assessor will find it necessary to
exercise judgment at many points in the process to arrive at a sensible result.
The process of assessment may be regarded as one of developing
confidence in the structure. Particular issues over which judgment must be
exercised are the hidden details and the effects of existing damage. In many
cases, important hidden details can reasonably be assessed from an external
inspection of the bridge.

3.0 WHY ASSESS

Assessment is necessary for a number of reasons. Perhaps most important is

that bridges deteriorate with time so their capacity to withstand load also
declines. It is therefore necessary, even in a stable environment, to check that
bridges remain serviceable.

The loading to which bridges are exposed also changes with time. The axle
loads, numbers of axles and vehicle speeds increases with time, and in the
process they might infringe upon the known or unknown safety margins.

3
Masonry arch bridges form a vital part of the transport infrastructure, many of
which are historic structures constructed over 100 years ago. Since
construction, and particularly in recent years, traffic loads and speeds have
increased drastically. The structural assessment of these bridges is therefore
key to their continued service. A reliable assessment method is needed to
ensure that strengthening is used only where necessary, and is as
economical and efficient as possible.

4.0 TYPES OF ANALYSIS AVAILABLE

First coordinated research in the modern sense into the behaviour of masonry
arch bridges commenced in 1936, under Professor A.J.S. Pippard, at the
request of the Building Research Station, on behalf of the Ministry of
Transport, UK. Research. This involved development of theoretical and
empirical methods accompanied by laboratory and field testing. Over the
years, four distinct methodologies for load assessment have been developed:
1. Empirical models, the prime one being MEXE
2. Two Dimensional Equilibrium based (Mechanism)
3. Elastic Methods
4. 3D models using FE/ DE

4.1. EMPIRICAL METHODS MEXE

Professor Pippard and his colleagues developed an elastic analysis method

based on minimum strain energy principles, used earlier by Castigliano. The
Pippard elastic method involved the calculation of stress in the cross section
of the arch ring under working load conditions for different end conditions. The
dead load of the fill and the self weight of the arch ring were represented by
vertical applied loads. Analysis was carried out for a unit width and no action
was taken for any contribution to strength from the spandrel walls or the
surrounding fill.

Professor Pippard later used his method during the Second World War, to
develop tables of allowable weights for wheeled and tracked vehicles for
military use. In the years following, the war, the MEXE method (Military
Engineering Experimental Establishment) was developed from the basic
Pippard tables in the form of a readily usable nomogram. This method
enabled military engineers to quickly assess the capacity of masonry arches.

The MEXE method is empirical, based on the work of Professor Pippard in

1930s and 40s and restructured by full-scale destructive tests carried out by
Davey. The arch is first assumed to be parabolic in shape with span/ rise ratio
of 4, soundly built in good quality brickwork/ stonework, with well pinned
joints, to be free from cracks and to have abutments with adequate strength.
For such an idealized arch, a Provisional assessment is obtained from a
nomogram or the formula given below. This Provisional assessment is then
modified by factors, which allow for the way in which the actual arch differs
from the ideal.

4
To summarize the method involves calculation of a provisional axle load
based on the geometry of the arch. The geometric data required are:

(i) span L in metres

(ii) the rise ‘rc’ of the arch barrel at the crown (in metres).
(iii) the rise ‘rq’ of the arch barrel at the quarter points (in metres).
(iv) the thickness ‘d’ of the arch barrel adjacent to the key stone
(in meters)
(v) The depth ‘h’ of fill at the arch crown (in meters).

These geometrical characteristics have been shown in Fig. (1).

Fig.1
A provisional axle load (PAL) is calculated either from a nomogram given in
BA 16/97 or from the following expression.
740 (d + h)
2

PAL = 1. 3
tonnes.
L
The following factors are then applied to the calculated value of PAL to take
into account variations from the idealized arch.

(1) Span rise factor Fst: It is assumed that flat arches are not as
strong as deep arches. A span rise ratio of 4 or less is given a factor
of 1.0 and this is reduced for span rise ratios greater than 4.

(2) Profile factor Fp: The ideal arch profile is assumed to be

parabolic and for this shape the rise at the quarter points is given by
rq = 0.75 rc.

5
 rq
The profile factor for ≥ 0.75 is taken to be unity and is less than
rc
rq
unity for ≤ 0.75.
rc

(3) Material factor Fm: The material factor is a combined factor which
takes into account the estimated strength of the arch material and
strength of the fill over the barrel.

(4) Joint factor Fj: The strength and stability of the arch barrel depends
to a large extent on the size and condition of the joints. The joint
factor is determined from the width and depth of the joints and the
quality of the mortar.

(5) Condition factor Fc: The estimation of the preceding factors is based
upon quantitative information available from a close inspection of the
structure. The factor however depends more on an assessment of the
importance of the various cracks and deformations that may be
present. The value of Fc varies between zero and unity.

The five modifying factors are then applied to the value of PAL to give a
modified axle load as follows:

Modified axle load = Fst * Fp * Fm * Fj * Fc * P.A.L.

Pippard made the following assumptions in his elastic analysis:

(1) The arch was parabolic

(2) The arch was two pinned
(3) The span: rise ratio was 4.0
(4) The arch section increased from the crown to the abutments
(5) The load was applied as a point load at midspan
(6) Fill density was same as the masonry density.
(7) Fill had no structural strength
(8) A limiting compressive stress of 1.39 N/mm2 was assumed.
(9) The maximum tensile strain permitted was 0.69 N/mm2.

An obvious curiosity of the method is that the method evaluates the effect of
total thickness of ring and fill (d +h) and this can lead to excessive
compressive stress. However, this method was easy to use but now it is
considered to be conservative particularly for long spans. It also has an
additional shortcoming in that the spans are limited to 18m and distorted
arches cannot be assessed.

A detailed review of the MEXE method is in hand. Until it becomes available,

it should be known that recent experience at Network Rail suggests that the
method may not be conservative, when applied to bridges to:

6
 (i) Small span
(ii) Modest cover, and

Suffice it is to say that the majority of old railway arch bridges, however, fall in
the above categories.

4.2. TWO DIMENSIONAL EQUILIBRIUM BASED MECHANISM

METHODS

The oldest approach to arch behaviour is also the simplest one -

demonstrating that a line of thrust exists within the arch, which can sustain the
applied loads. In this approach the load, which just transforms the arch into a
hinged mechanism, is evaluated. The following assumptions are made in this
method:

a). no tension
b). Infinite compressive strength
c). Infinite elastic modulus and
d). No sliding between voussoir.

The mechanism approach to arch collapse looks for the minimum load,
needed to introduce the number of hinges at the arch intrados and extrados
large enough to transform the arch into a mechanism. The collapse load can
be found by statics and it depends on the geometry of arch, the position of the
hinges and the weight of the blocks into which the hinges have divided the
arch. For a two-hinged arch, two more hinges will be required to be formed to
convert the arch into a mechanism.

The mechanism method involves lengthy iterative calculations and is gaining

in popularity as more and more computer programs are developed. The
method provides no information on stresses and deformations and as such
the load test is of no value in this procedure.

However the important conclusion of above theory is that any isolated cracks
appearing in the arch barrel need not be the cause for panic. While

7
ascertaining their cause and finding appropriate remedial measure are
necessary, nevertheless the arch will not fail until four hinges are formed. This
is also confirmed from the results of experiment carried on actual arches,
which have shown that the loads causing failure of arch are nearly 2 to 4
times the loads at which, first crack appeared.

S.No. Bridge Condition Load for Load for Ratio

First Collapse
crack (Tonnes)
(Tonnes)
1 Arch 76 near Bare arch with load at 40.6 132 3.25
Sonepur, 1.68 m spacing over a
NER BG sleeper
2. Arch 73 near Arch intact with 0.91 m 71 >233 3.28
Sonepur, cushion
NER
3. Bridge 41C Bare arch of 0.61 n 71 >330 4.65
Kota- Bina stone masonry and 0.25
Section, WR n gravel and 0.18 m
water bound macadam
pavement
4. Bridge Arch 7, bare (that is, 24.4 61 2.57
across without cushion
Uppodai parapets and spandrel
(Madras) &X walls)
9.75 m span
5 Bridge Arch 4, bare 30.5 91.4 2.50
across
Uppodai
(Madras) &X
9.75 m span

4.3 ELASTIC METHODS

Elastic models have been used for masonry arch analysis, since Castigliano
solved the problem of analysing indeterminate structures. Castigliano
developed an elastic analysis by assuming that the arch material could not
take tension. This was a reasonable assumption particularly arches built in
lime mortar which may have been in service for more than hundred years. He
calculated that line of thrust by the method of minimum strain energy and if it
fell outside the middle third of arch ring, the material under tension was
removed and calculation repeated until no tension was found at any point. At
this point the line of thrust is accepted and the stresses are calculated with
normal methods like MEXE. The ultimate load carrying capacity is defined as
the load, which produces the combined dead and live load axial and bending
stresses at the critical section, equivalent to characteristic compressive
strength of the masonry. This method was later developed into a
computerized assessment.

8
The following observations of RDSO during the Static and Dynamic tests
conducted on arch bridges, published under various reports (C72- 76, C 79-
80, C83-84) corroborates the elastic behaviour of masonry arches:

(i). The graph showing deflection versus load is almost straight up

to 325 tonnes.
(ii). The deflection values were practically same under repetitive
loading.
(iii). There were no residual deflections for applied loads not
exceeding 200 MT.
(iv). For loads up to 244 tonnes, the residual spread was
insignificant.
(v). Tensile strain occurs at crown level along the Arch Vault, where
as at all locations compressive strains are recorded in
accordance with the elastic theory.

The elastic approach with a cracking model offers the advantage that the
effect of deflection can be studied by analysing the deflected shape at each
iteration. The method however, demands a high level of confidence in the
material properties used, a fact, which is likely to be overlooked by many
users.

On Indian Railways as per the arch bridge code, for the analysis of new
masonry or concrete arches, the elastic method of analysis should preferably
be adopted. Either purely analytical methods or a combination of analytical
and graphical methods may also be used.

4. 4 3D MODELS USING FE/ DE

Conventional assessment of strength is either based on semi-empirical

methods such as the MEXE method, elastic method or mechanism method
While these methods are useful for initial assessments and quick to use, they
can only be applied to certain arch configurations, for example single spans
and give no information on displacements. Numerical simulation by the FE/
DE method permits a more comprehensive analysis for predicting the
behaviour of masonry.

The accurate representation of masonry for structural analysis is a complex

problem. In addition to the materially and geometrically non-linear behaviour
of the masonry blocks themselves, contact-gap-friction effects at the joints
and the evolution of further joints due to fracture are important. The ability to
model post failure behaviour in order to verify analysis against physical tests
is also useful. Many of these features can be modelled to a degree using
continuum finite element methods with sophisticated material models and
arrays of gap elements. Such methods require pre-defined contact definition
and post failure behaviour cannot be observed. A more natural approach to
the problem is to use DE.

The Distinct method was first developed in the late 1960’s to address
geotechnical and granular flow applications. The basic concept is that separate

9
parts are modelled, possibly connected at boundaries, with friction, cohesive and
adhesive contact. Initially the interaction of rigid parts was considered but later
developments led to what is now termed the DE method where deformation and
fracture can be included. The computer program, have been developed to model
the arches with the above concept.

For application to masonry arch bridges, typically discrete parts using non-
linear material properties where appropriate and allowing non-linear geometric
deformations, are used to model the abutments, piers, fill, any backing, and
representative vehicle axles. The masonry of the arch barrel is modelled by a
number of individual blocks, each normally representing a number of real
masonry parts. Multiple rings can be modelled where appropriate, allowing the
effect of features such as existing ring separation to be taken into account.
Multi-span structures can also be modelled and existing defects such as
mortar loss and weathering can be included. The forces between all
component parts are automatically calculated, both under initial dead loads
and then required traversing live load conditions. In this manner the complex
non-linear behaviour of the masonry is accurately represented at a
fundamental level. Fig. Below shows DE and FE meshes used for the analysis
of a typical two span, multiple ring arch bridge. Both the analysis technique
and the efficacy of the strengthening system have been verified by full-scale
tests.

Fig. : Discrete and finite element meshes for a two span, multiple ring bridge

5.0 METHOD PRESENTLY USED ON INDIAN RAILWAYS:

On Indian Railways, the Load Carrying capacity of arch bridge is assessed by

carrying out a Load test on the representative sound arch, where theoretical
overstress exceeds 100% and on a distressed bridge, after pressure grouting
of cracked masonry. The parameters observed during the load test are the
deflection and spread of the arch. The Safe Load is stipulated to be one,
which does not produce a deflection of more than 1.25 mm at the crown and
spread of more than 0.45 mm for arches with span between 4.5 m and 15 m.

The method derives its roots from the “Assessment by Load Testing”
proposed by Building Research station, UK.. Building Research Station
carried out load tests on typical actual arch and other older bridges duly
applying the vehicle loads of different configurations for assessment
purposes. Based on tests, the Building Research Station recommended the
following criterion for the assessment of arch bridge:

10
If, under a 20-ton single axle, the spread and deflection of the arch do not
exceed 0.015 inch (0.38 mm) and 0.05 inch (1.27 mm) respectively for
bridges upto 45 feet (13.71 m), it can be assumed that the bridge can safely
carry a 40 ton two- axle bogie.

The above criterion was derived from the load deflection characteristics
observed in a number of collapse tests on the basis that pins (hinges) did not
form until such deflections were reached. Thus, the BRS criterion
recommended by Load testing was primarily based upon the consideration of
crack development, and hence, was a form of serviceability criterion. It did not
provide any indication of the ultimate capacity of a bridge, which can be much
more than the assessed load. The load tests have shown that significant
tensile cracks appear long before the ultimate capacity of arch is reached.
The Tests have also shown that deflections increase rapidly after the level of
approximately half the ultimate load. Since repeated opening and closing of
joints under moving loads may seriously affect the integrity of the bridge, a
serviceability limits have been laid down for the deflections.

To determine the safe load, by the method used on Indian Railway, Influence
Line Diagram of the bending moment at crown of arch is first drawn. The
proposed loading is taken as EUDL. The arch is divided into convenient
number of segments. The value of ordinates of influence line (Mc) is given by:

L/2
Σ MR x(S/I)- Hc x 2Σ (yxS/I)
Mc= 0_____________________
L/2
2 Σ S/I
0
The position of proposed loading to give the maximum bending moment at
crown is determined ( point A and E in the Fig. below). Maximum Bending
moment caused by the proposed loading will be product of EUDL and net
area of ILD corresponding to the load position mentioned above.

The equivalent Single load or the Test load is then determined by equating
the bending moment at crown, produced by this single load (assuming 45%
dispersion through arch fill) to that produced by the proposed load.

11
 TEST LOAD

CUSHION
45°

DISPERSION
LENGTH

EFFECTIVE SPAN

EUDL FROM BRIDGE RULES

CORRESPONDING TO SPAN

POSITION OF LOAD
D
FOR MAX. BM

POINTS OF
CONTRAFLEXURE

A C E G

B F

INFLUENCE LINE DIAGRAM

FOR BM AT CROWN

12
6.0 RECENT DEVELOPMENT IN ASSESSMENT:

Recently UIC has recommended the use of an software application page

called RING for analysis and assessment of load carrying capacity of masonry
arch bridges. RING idealises a masonry arch structure as an assemblage of
rigid blocks and uses computational limit analysis methods to analyse the
collapse state only. Although limit analysis, or ‘plastic’/ ‘mechanism’ analysis
techniques were originally developed for steel components and structures, it
has since been shown that these can be applied to masonry gravity
structures, such as piers and arches.

To help understand why limit analysis theory is applicable, compare and

contrast the response of a steel column with uniform plastic cross-section and
a weakly mortared masonry pier, both subject to a horizontal load F, as shown
below

Laterally loaded (a) steel column, (b) masonry pier, and (c) idealised response
curves

It can be deduced that:

• while the tensile and compressive strength of the steel column endow it
with a finite plastic moment of resistance, Mp, the absence of tensile
strength means that the masonry pier does not possess a comparable
(i.e. strength derived) moment capacity;
• however, the thickness and self weight of the pier mean that there is
some resistance against overturning and the masonry pier could
conceptually be considered as possessing a moment capacity, albeit
one that varies with height (equal in magnitude to the normal force at a
given cross-section multiplied by half the pier thickness);
• furthermore, provided pier displacements do not become large, the
resistance of the masonry pier against overturning at a given cross-
section will remain broadly constant;
• hence the response of the pier can be considered ‘ductile’, an
important requirement for a limit analysis theory to be applicable.

13
The behaviour of arch bridges is more complex than the one shown in the
above example. RING uses rigorous mathematical programming techniques
to identify the most critical of numerous possible failure modes.

6.1 Validation

In Bolton, UK, in the early 1990's a number of 3m and 5m span bridges were
tested in the laboratory. A key advantage of these tests over field tests was
that the internal constructional details and material properties were known.
RING was originally developed to assist with the interpretation of the results
from these laboratory tests.

In the Table shown below, sample RING 1.5 analysis results are presented
along with the experimental test results (only bridges with detached spandrel
walls are included since these behave in a two dimensional manner).

Bridge Description Expt. Limiting Effective Theoretical Theoretical/

Collapse load classical collapse Experimental
Load dispersion passive load Collapse
(Kn) angle earth Load
(degrees) pressure (kN)
coefficient
3-1 3m single 540 45 4.5* 550 102%
span
3-2 3m single 360 45 4.5* 245 68%
span;
debonded
arch rings
5-1 5m single 1720+ 45 4.5* 2238 130%+
span
5-2 5m single 500 45 4.5* 463 93%
span;
debonded
arch rings
Multi- 3m triple 320 45 4.5* 358 112%
2 span

*Approx. of the full classical passive pressure coefficient indicated by

measured angle of friction of fill (60°)
+
The experimental collapse load of this bridge was reduced by the sudden
onset of partial ring separation

In 2001 TRL were commissioned to independently validate RING 1.1 and

other available masonry arch bridge analysis software. As part of the
validation process it was decided that the programs would be used to predict
the carrying capacities of 5 of the field bridges load tested more than a
decade previously. Details taken from the TRL report relating to RING for 4 of

14
the bridges are provided in Table 2 below (Strathmashie bridge was also
modelled but was in poor condition and, because ‘none of [the] defects were
modelled during the analysis, all the programs returned non-conservative
results’).

Bridge Theoretical /experimental collapse load

Torksey 81%
Bridgemill 100%
Barlae 92%
Preston 90%

Correlation between TRL field bridge test and RING collapse loads
(independently produced by TRL)

It is evident that agreement between the RING predictions and the full-scale
test results was found to be reasonably good. Thus the TRL report concluded
that RING ‘gives good results’ and that ‘RING, with some investment in an
improved solver, would be a very effective tool for most assessment
engineers…………”

Based on this evidence Network Rail have confirmed that RING is a suitable
program for use to assess masonry arch bridges on the UK rail network.

An advanced version of this software, RING 2.0 is under development.

7.0 LIMITATIONS OF THEORETICAL ANALYSIS

7.1 Boundary conditions

All structural analysis is critically dependent on boundary conditions. Arch

bridges have more boundaries than most structures and the resulting complex
relationships present a major challenge to the analyst.

(i). If an arch alone is to be analysed, the behaviour of the springings must

be represented in a reasonable way. With the simplest models, rigid
abutments can be assumed. For an elastic model of any sort, very small
movements of the springings will have a dramatic effect on the stresses
in the arch.

(ii). Representing the real behaviour of the abutments is, however, extremely
difficult, even in a situation where the regime is effectively two
dimensional (eg near the center of a long tunnel). It is well known that
loads approaching a bridge cause stresses in the fill, which push
abutments together and compress the arch. When the load is on the
bridge, this compression is reversed and the abutments spread slightly.
The soil pressure may be similar in these two cases, but is likely to be
differently distributed.

15
(iii). The edges of arches are usually bounded by spandrel walls. It is normal
to assume that these make no contribution to structural behaviour. That
assumption is likely to be conservative, but may play a major part in the
gross underestimation of capacity, which is evident in some bridges.

7.2 COMPLEX NATURE OF TRACK/ VEHICLE LOADS:

The following forces contributed by track and live load, which are difficult to
quantify, adds to complexity of assessment:

(i) Track maintenance, involving adding, removing, or just moving ballast, is

a routine operation on modern railways. The sleepers cannot be
regarded as fixed points of application.

(ii) The force exerted on the rail from a stationary vehicle is simply derived.
In most assessments, all moving loads must be treated as pseudo static.
That is their effect is viewed as static though the load may be moved to
find its worst effect

(iii). At the rail level, there are Longitudinal forces generated as a result of
acceleration or braking. The latter is usually the greater force. The
longitudinal stiffness of the track is so much greater than the stiffness of
its connection to the ballast that the force will distribute over a
considerable length and most will be carried off the bridge. There may be
some need to consider lateral effects from these forces on tall piers of
curved viaducts.

(iv). The impact, or shock loading effect of loads on an arch in good condition
is likely to be small, because the response of the foundation to track
loads will not differ greatly from that of the general surroundings.

(v). Centrifugal forces create two effects. Transfer of load from one wheel to
another on an axle and a direct horizontal component. In most
circumstances, track will be canted sufficiently to minimise the former
effect. Nosing and sway produce similar effects.

If a bridge is carrying loads without apparent distress and the analysis says it
shouldn’t, it is incumbent on the engineer to express a rational view as to why
the analysis falls short. On the other hand, if loads substantially below the
assessed capacity distress a bridge, the whole analysis becomes
questionable. Therefore an analytical approach needs to be validated with
physical symptoms.

From the above theories of analysis, it is evident that the load assessment of
existing masonry arch bridges is dependent on their physical state. The
Physical state has to take in account the possible damages in the masonry
arch. The possible damages in the arch bridges are discussed as under.

16
8.0 POSSIBLE DAMAGES IN MASONRY ARCH:

The appearance of damages in structures, is as inevitable as inexorable is the

passage of time, partly due to action of nature and aging of the masonry
materials and partly due in increased traffic loads, with the passage of time. In
masonry arches, it is not strange that some cracks appearing due to normal
structural behaviour may be classified as serious defects whereas the defects,
which can trigger the collapse, are overlooked or not given the due
consideration. Further the defect appearing in a particular component of arch
may not necessarily have its root in the component itself. It may be the
manifestation of defect in the other part. For example a crack in arch barrel
may have its origin in deficient foundation behaviour. It is therefore essential
to understand the different degradation mechanisms that take part on the
bridges to establish an accurate diagnosis and prescribe an efficient therapy.

The possible defects in the Arch bridges can be broadly classified into two
categories:

1. Foundation Damages
2. Structural Damages

8.1 FOUNDATION DAMAGES:

The main problem of ascertaining the damages in foundations is the difficulty

to inspect them. If this inspection is made in summer, it would be easier to
detect degradation problems of the structural element, foundations and even
problems related with soil foundation degradation provided the riverbed is dry
during the summers. For the bridge where standing water exists throughout
the year it is difficult to inspect, the as sub-aquatic inspection even using
underwater equipment, is a difficult task to perform. Therefore, in practice, the
first useful stage to detect problems of bad behaviour of foundations is to
observe and analyze symptoms that eventually are shown on the super-
structure as a consequence of rotation, or differential movement on
foundations. The following are the damages in foundations, which are usually
observed on Arch Bridges:

8.1.1 DAMAGES DUE TO THE ELEMENT DEGRADATION

The above category of damages is not only because of the degradation of the
foundation due to weathering but also due to damages in the elements
supposed to protect them. Most of the old masonry arches were constructed
with lime mortar. Lime dissolution leads to mortar and old lime concrete
elements disintegration, very usual on pile caps and masonry bridges
foundation piles. River water tends to dissolve the lime, free the mortar,
bringing them to loose sand condition. The resultant effect is the formation of
cavities or even the complete disintegration of the foundation section.

17
Mortar is a material perfectly plastic when it is placed. It hardens few hours
after being placed and imparts rigidity with time. Mortar has two main
functions inside the masonry:
- Regularize the setting among the blocks and distribute the loads
uniformly.
- Transfer the horizontal trusts from individual blocks to abutments.

It is therefore imperative that the repairs in case of loss of mortar should

preferably be carried out with originally used mortar or with mortar having
composition nearest to above. In case of arches constructed with lime mortar,
the repairs should be carried out using lime mortar and in case of non-
availability cement mortar can be used. Use of epoxy as is usually being
done, will alter the load transfer mechanism and hence should be avoided.

Fig. Masonry arch bridge footing degradation example.

Dragging of gravel and sandstone in aggrading rivers results in damages in

foundations, footings and plinths elements, together with erosive action and
degradation of mortar. These processes may also be due to due to the
increase of speed caused by the narrowing of the riverbed or by the
longitudinal profile modification. The final result is loss of material, affecting its
strength. Figure below shows the case where the foundation truss is exposed
as a consequence of the above-mentioned effect.

18
 Fig. Damages caused by lime dissolution on footing and socket.

8.1.2 Corrosion of steel elements used in foundation

During the latest third half of the 19 th century and first half of the 20 th
century metallic caissons have been profusely used to execute pile foundation
on river beds, especially if the river bed was deep. On other occasions, sheet
piling has been executed with metallic elements that were suitable as
protection against the stream. On these structures with the passage of time
and change in humidity, damages due to corrosion, with important material
loss, appeared. The importance of above factor lies in the fact that the
material, which was protected earlier, is now exposed to erosive action of
water and wind. Corrosion has resulted in serious damages to stability of
structure and therefore needs to be given a serious consideration.

8.1.3 Damages due to soil- foundation degradation

Riverbeds excavations were notably accelerated in recent times as a

consequence of the increase of the sand extraction on the river shores.
Extraction of sand on the upstream of the river is dangerous. River tries to fill
up these man made depressions, by dropping the sediments it carries. When
the solid contents of the river depletes, the flowing current digs the riverbed to
make up the loss of sediments, resulting in scouring around the piers and
abutments.

Fig. General and Local River bed undermining.

19
Local undermining next to the supports – piers and abutments – is basically
an erosion of the bottom of the river bed as a consequence of the formation of
horizontal axis whirlwinds that are developed with ringlets forms around these
elements. Riverbed materials get scoured from the bank river by the vertical
flow component, lifted and propelled by the current water. Thus, it is forms a
conic shaped hole in case of soils without consistency where the deepest
point is upstream of the pier.

Fig. Local action of the current water on foundation

If any of the above mentioned damages are observed following action should
be taken for:

- Visual inspection of the foundation

- Estimation of the dimensions and typology of the foundations
- Inspection of the soil around the foundation
- Estimation of the longitudinal profile of the river
- Estimation of the transversal section of riverbed
- Determination of the cleaning and general state of the river

8.2 SUPERSTRUCTURE DAMAGES

Structural damages can be broadly classified in two categories those resulting

from bad resistance performance and those resulting due to durability
problems which have potential of reducing the resistant capacity of the main
structural elements.

8.2.1 Damages resulting from a bad resistance performance

The structural damages can be better understood by knowing the processes

of deterioration and its associate damages in the bridge in general rather than
knowing the variety of damages that an element of arch, can possible suffer
during its service life.

8.2.1.1 Gravitational actions

These actions are due to the weight of the structure (piles, abutments, vaults,
spandrel, backfill) and due to the correspondent dead loads (ballast, rail,
parapets, etc).

20
In case of brick masonry bridges, where spandrel walls are constructed
monolithically with the arch barrel, longitudinal cracks sometimes appear
under the inside edge of spandrel wall on the intrados. If such cracks are very
fine and do not widen with time then they are mostly attributable to the
difference in stiffness between the spandrel wall, which acts like a deep
beam, and the flexible arch barrel (which results in incompatibility of
deflections at their junction). Such cracks are not considered serious, but they
must be kept under observation.

Many times, track level on the arch is raised bit by bit and new masonry
courses are added on the spandrel wall without giving thought to the
adequacy of spandrel wall cross section. This is also a cause for such cracks.

8.2.1.2 Imposed movements

Imposed movements are the result of foundations movements in abutments

and pile and their consequences, for example, the undermining. This action is
the most important and is the responsible for most of the structural damages.
Unequal settlements of the soil underneath the foundations of piers and
abutments cause differential movements among different areas in the
structure resulting in development tension forces on foundations and arch
barrels which ultimately result in formation of cracks the masonry elements.
Depending on the location and magnitude of the differential movements of the
foundations and on the type of monolithic masonry, the cracking will have a
vertical or inclined pattern. The structural importance of this defect will depend
on its stage of development whether or not it is stabilized and on the structural
characteristics of the elements.

Fig. Settlement of foundation and induced damages.

Fig. below shows the example of the transversal rotation of a foundation

about its longitudinal axis, resulting in an inclined and stepped cracking, due

21
to differential settlement of the same footing. Such type of cracks are also
observed due to local undermining of foundations.

Fig. Differential settlement and transversal rotation of longitudinal axis of pier base.

The above figure also shows the bottom view (intrados) of the crack in the
arch. Together with this type of cracking it is also possible to find joints
opening on the arch, or even relative movement between arch blocks,
particularly on the ring course that is on the same side of the settled area of
the foundation. It is also possible to detect damages on the spandrel that
depends on their rigidity and on the existing bond between spandrel and arch
ring.

In extreme case of transverse bending coupled with differential settlement, a

partial collapse of arch, as shown below can result.

Fig. Relative movement among longitudinal elevation of the bridge.

22
Relative movement between edges and central part of foundations particularly
when the barrel is wide – or it has been widened- is also observed. This
differential settlement causes transverse bending of the foundation and
therefore, vertical cracking, among joints which can also extend into arch
barrel. On the contrary, if the edges of the foundations near the cutwaters
settle more than the central part, the damage is revealed in the opposite way,
the width of the crack is very small or inexistent in or near foundations and is
wider on the arch barrel.

Fig. Vertical cracking on pile because of differential settlement between the edges
and the central part of foundation.

8.2.1.3 Problems caused by abutment overturn by excessive earth

pressure

These types of problems are infrequent, as not many cases of collapse due to
this reason have been reported. Nevertheless, it is true that, over exploitation
of these structures (increase of loads and, speeds), the progressive
deterioration of the drainage systems of the embankments and backfills,
among other factors, contribute to increase the pressure that soil and water
have on the abutments and walls. Figure below shows a sketch with the
above-mentioned failure. This problem is more frequent in structures with
short deep spans and high abutments. However the cases of the increase of
the horizontal trust due to improper drainage in backfill resulting in generation
of pore water pressure and ultimately the leaning of abutment and wing walls
are many.

Fig. Damages caused due to inward rotation of the abutment.

23
8.2.1.4 Bulging of spandrels

This problem is connected with an excessive earth pressure of the fill and
water retained in it, and also to the horizontal component of live loads. This
damage is generally seen in bridges with deep and not very wide barrels and
with big depth fill over the crown.

Fig. Bulging of spandrels

Blockage of drainage and excessive surcharge may also, sometimes, lead to

sliding forward of the spandrel wall, particularly in case of bridges where
spandrel wall and the arch barrel are not monolithically connected. The
damage happens because the external forces due to reasons mentioned
above are more than the stabilizing force of the weight of the spandrel
multiplied by the friction coefficient. It is very important to know the bond
employed to connect the spandrel and the arch barrel and the actual width of
spandrel at the bottom. This damage occurs in bridges of deep and not very
wide barrels and with a high depth fill over crown.

Fig. Sliding of spandrels

In extreme cases, the whole of the spandrel wall can be overturned, if the
destabilizing forces due to choked fill, surcharge and increased live loads are
very large.

24
 Fig. Overturning of the spandrels

8.2.1.5 Damages in Wing walls

8.2.1.5.1. Overturning and bulging

The same considerations for spandrels are also for valid for wing walls and
sidewalls. As it has been already mentioned the origin of these damages can
be a deficient drainage in the backfill, caused by weep-holes obstruction
resulting in increased horizontal pressure on the walls causing a dangerous
destabilizing conditions.

Fig. Overturning and bulging of the wing walls and walls.

8.2.1.5.2 Vertical cracking in the joint between the abutment and

wing Walls:

This damage is due to the different kinematics of the elements movements.

The abutment is joined to the spandrels and the arch barrel, while the wing
walls and sidewalls have free horizontal movement, particularly on the top.

25
Almost vertical cracking, with width increasing with height, happens in the joint
between the abutment and walls.

Fig. Vertical cracking in the joint between the abutment and the wing walls and side
walls.

8.2.1.6 Stepped cracking

This failure is due to differential settlements in the plane of the wing wall.

Fig. Cracking in stair pattern on wing walls or on side walls.

In all of the above cases following action need be taken:

- Inspection of the foundation and check for possible movements
- Check the type of bond employed between the abutment and the wall
- Inspection of the backfill of the walls
- Inspection of the drainage system of the wall and improve drainage

8.2.1.7 Transversal cracking resulting – arch mechanism failure

This is the most serious type of damage shows up in form of transverse

cracks in the intrados of arch barrel. It is generally accepted that in masonry
arch bridges, the compressive strength of stone - or even of the brick is
greater than, the level of stress in the arch. So it is said that, generally, the
collapse of a single masonry vault will happen because of the formation of
mechanism. Thus, it can be said generally that the arch must have four
alternative hinges to collapse, under the application of live load. The

26
transverse cracks in intrados indicate the presence of tensile stresses, which
if neglected, will result in formation of hinges. The damage therefore implies
the imminent collapse of the structure so urgent measures must be taken.

Fig. Configuration of collapse by enough number of hinges. (4 hinges –arch

mechanism)

8.2.1.8 Loss or dislocation of pieces

The origin of this type of damages can be either due to strength or to

durability, or due to combination of them. If the origin of the damage is due to
a bad resistance behaviour, it is usually a symptom of separation movements
among the springing lines of the arch barrel, or rarely, by loss of axial force in
vault, or because of problems of heavy local loading, affected by impact near
the crown of the arch barrel when there is little depth of fill over the crown.

Fig. Falling of pieces on masonry vaults.

Following need to be looked into for the above damage:

- Inspection of the track (looking for track devices that can generate
vertical impacts on the barrel)
- Determination of the speed and the maximum axial load that the bridge
is carrying.
- Estimation of the type of masonry employed (limestone, sandstone,
granite, etc.)
- Inspection of the drainage system of the bridge, specially that of barrel.

27
8.2.2 Damages caused by deficient durability

Besides the above damages the other damages are damages due to climatic
conditions, weathering of constituent materials with age and damages due to
neglected maintenance.

The failures caused due to climatic conditions e.g rain, ingress, disruptive
forces due to freeing and thawing of frost, solar radiation and abrasive actions
of floating particles carried by wind, are demonstrated by erosion of surface of
arch spandrel walls, abutments, joints infilling etc. The emissions have
unfavorable impact on arch bridges and both surface of elements and joints
infillings are chemically eroded. The degree and speed of erosion will depend
on the type of material used in construction, quality of infill joint and on the
emissions concentration.

As a consequence of degradation action of the weather conditions with time

(i.e ageing of structures), the load carrying capacity of the bridge deceases
due to material and structural deterioration. The failure results due to action of
traffic coupled with the material and structural deterioration.

The common neglected maintenance, which may become the contributory

reasons for the failure, are growth of vegetation and improper maintenance of
drainage. The growth of vegetation results in deterioration of joint infill
(mortar) and consequent lossening of masonry units. When drainage is
blocked either behind spandrel walls, abutments and wing walls, the fill
material at these locations becomes water bearing exerting excessive earth
pressures which may result in sliding/ bulging and cracks in spandrels, cracks
in abutments and wings and other defects.

Fig. Presence of plants on the bridges and disorders caused by them.

The knowledge above damages is necessary for their incorporation in the

models used for assessment of load carrying capacity. Before any retrofitting
is attempted the minor damages like loss in mortar, cracks originating due to
reasons other than structural reasons etc. need to be attended.

28
9.0 STRENGHTENING & RETROFITTING OF ARCH BRIDGES

Based on the assessment of the strength of arch bridges a variety of

strengthening methods are available. These vary in effectiveness and each
has advantages and disadvantages. Some of the methods are patented and
proprietary.

9.1 Saddling

One of the simplest and most popular of the traditional methods is saddling.
This involves removal of the fill to expose the extrados of the barrel. A
reinforced or mass concrete flat or curved slab is subsequently cast in place
over the original barrel.

While saddling will undoubtedly increase the capacity of the bridge, with
minimal change to the external appearance of the bridge, it is expensive and
will cause considerable disruption to traffic and buried services like
communication cables, pipe lines etc.,. The bridge is also in a temporarily
vulnerable state once the fill has been removed, unless the barrel is
supported, which can be a costly process.

As per Para 5.3.5.2 of IRS arch bridge code, in the case of strengthening
over extrados of arch, the new arch ring should be designed to take the entire
load, viz .dead and live loads.

9.2 Sprayed Concrete

Another traditional method is the use of sprayed concrete applied to the

intrados of the arch. This may be used in conjunction with a reinforcing mesh.
Whilst this negates the need for drilling, the intrados is the part of the barrel
exposed to weathering, resulting in friability. Applying sprayed concrete can
cause moisture to be locked into the barrel. Other problems include poor
composite behaviour and incompatible of materials. Sprayed concrete has
been applied in conjunction with a corrugated metal lining.

Fig. Bridge strengthened with sprayed concrete

29
9.3 Jacketing below intrados

For strengthening weak/distressed arches, method of jacketing at the intrados

is preferable, if the resultant reduction in the waterway is permissible as per
the guidelines of IRS arch bridge code. The new arch ring should be designed
to take the entire load by itself where the existing arch has transverse crack(s)
by ensuring that a proper bond is established between the existing masonry
and new material by suitable means such as dowels and post grouting
through grout holes to be left while casting the jacket.

Also it should be ensured that cracked masonry, whether in arches or in

abutments and piers, should be grouted under pressure to plug all the cracks
before the additional material is provided.

9.4 Surface Reinforcement

There are proprietary systems available using a network of steel bars located
in slots cut into the intrados and bonded using special adhesives. Such
systems have been shown to increase the strength of the bridge. However,
access to the arch intrados is not always easy or possible. More importantly,
the application of sprayed concrete and/or externally bonded or slotted
reinforcement will have a detrimental effect on the appearance of the intrados.
This is un-desirable in many situations and unacceptable for many structures
of historic importance.

9.5 Insertion of Box inside the arch

On Indian Railways we have strengthened weak/distressed arches by

constructing boxes inside the arch, either single or multiple spans. The space
above the box upto the intrados of the arch is filled with lean concrete under
pressure. Again the effectiveness of load transfer is questionable as a small
gap left between the intrados and top of box may result in arch deflection and
consequently it’s carrying the load. This is only possible where adequate
waterway is available. This method is again not acceptable in many situations
especially for many structures of historic importance. Also strictly speaking,
this method cannot be counted as one of the means of strengthening or
retrofitting of arches.

9.6 Retrofitting

The method of retrofit no-prestressed reinforcement enables strengthening

masonry arch bridges without significant intervention into the whole structure.
The principle is to stabilize newly originated tensile stresses, to restrict/reduce
further development of existing cracks and to constrain/prevent an onset of
the new ones.

From the static viewpoint, unreinforced masonry structure is unable to transfer

tensile forces that can originate on existing structure from following reasons.

• Higher imposed load against the designed one

30
 o The load-bearing ability of a structure or its parts weakens due to the
degradation process,
o The static action/loading changes either due to the appearance of
significant cracks or due to the abutments settlement.

Another consequence of the retrofit reinforcement application into masonry

structures is the rigidity improvement. The effect is evident especially on the
structures cracked due to higher live loads.

For reinforcement it is possible to use

• Conventional reinforcing bars with rustproof coating, plating, etc.,

• Reinforcing bars of different special shapes either with anticorrosive
treatment or manufactured from stainless steel,
• Other advanced materials (carbon, glass and aramid rods).

These can be applied for retrofit of masonry walls, arched structure, flat
lintels, arch cords and piers.

A disadvantage of conventional reinforcement application is insufficient

protection against corrosion. The reinforcement with anticorrosive treatment
(rustproof coating, plating, etc.,) needs careful handling as it is prohibited to
use any bar with corroded surface. The problem may also arise at the ends of
bars, which are normally not protected. Some applicable reinforcing products
are presented in the below table.

Product (system) Type of Description

product
Stainless reinforcement
of special shapes
Helifix HeliBar Stainless austenitic cold rolled
steel of special helical shape
and high strength in tension
Brutt saver Brutt Saver Stainless austenitic cold rolled
Profile steel of special helical shape
and high strength in tension
Cintec Cintec CHS
Thor Helical Thor Helical
Bars
Fiber Reinforced Structural reinforcing bar
Polymer (FRP) made from filaments or fibers
held in a polymeric resin
matrix binder. The FRP bar
can be made from various
types of fibers such as Glass
(GFRP), Carbon (CFRP) or
Aramid (AFRP)
Huges Brothers Asian 100 Vinyl Ester Matrix Glass
GFRP Rebar

31
 Asian 101 Polyester Matrix Glass GFRP
Rebar
Asian 200 Carbon Fibers Reinforced
Polymer (CFRP)
Fibes-Bond system FRP 101 Glass, carbon or aramid Fiber
Reinforced Polymer
Wabo Mbrace Glass, carbon or aramid Fiber
Reinforced Polymer
Preswitt Polyplast Pre-mixed, fiber-reinforced
cement based plaster with
high tensile strength fiberglass
reinforcing mesh and specially
formulated liquid resin.

As binding (transferring) medium between reinforcement and origin masonry

has to be used special mortar (grouting substance) having following
characteristics:

• Long-terms life cycle and durability,

• Zero (or almost Zero) volume and temperature changes,
• Good bond to brick/stone/combined masonry and to reinforcing
member

Embedding of the reinforcing bars in the drilled chases is mostly realized

either with cement mortar or with mortar based resins. The disadvantage of
cement mortar is relatively long-term hardening and necessity of watering
(mortar in thin joints dries quickly). The properties of resulting composites
“reinforcement-mortar-origin masonry” are highly dependent on the physical-
mechanical properties of mortar Its strength, cohesion with the base-masonry.
The properties are significantly changeable as they have dependence on the
temperature and moisture at setting and hardening of mortar.

The performance of the existing structure and the strength enhancement

provided by reinforcement is predicted by the use of numerical simulation
using the FE/ DE method. These methods calculate the strength deficiency
and target strength that the internal reinforcement was required to achieve.

Typically, retrofitting with reinforcement can be broadly used to repair the

existing arches in cases of:

1. Soffit – barrel arch beaming and pinning

2. Pier – crack pinning
3. Abutment – crack stitching and pinning of coping stones
4. Beaming and Pinning
5. Pier – crack stitching
6. Spandrel Pinning
7. Replacement and Pinning of spalled bricks
8. Ring Separation – barrel arch pinning

32
 Fig. Retro-fitting with reinforcement

Some types of retrofitting using are shown below:

Fig. Spandrel Pinning

33
A proprietary system known as the Cintec anchor is used to enable
reinforcement to be accurately retrofitted. The reinforcement anchor consists
of three main components, a stainless steel bar, a fabric sock and a
cementitiuos grout. The stainless steel bar provides increased tensile and
compressive capacity. The woven polyester sock permits sufficient leakage of
grout during inflation to develop a chemical and mechanical bond with the
surrounding masonry whilst protecting the masonry from being displaced or
otherwise damaged by the pressurised grouting and limits the volume of grout
escaping into the surrounding masonry. The grout used is similar to Portland
cement based products.

The main steel body of the anchor is completely surrounded by a fabric sock.
The anchor is then located in an oversized drill hole joining the materials to be
anchored together. Fluid grout is then injected under pressure through the
middle of the anchor, until it reaches the remote end. There, it passes through
a series of grout flood holes into the fabric sock. The entire assembly inflates
like a balloon under the pressure. The excess milk of the grout and bonding
agent passes through the fabric sock, both fixing and providing a mechanical
and chemical bond to the parent material. Variation in the size and shape of
the individual components enable the basic method to be extended to meet
the designer's requirements.

Plan

Section

Fig. Schematic arrangement of Arch strengthening

34
10.0 CONCLUSION

The main objective of this paper was to bring out the various assessment
techniques available for masonry arch bridges and the techniques for their
repair and retrofitting. This is important since arch bridges are structures,
which have enormous load carrying capacity, and thus require their proper
assessment. Assessment of the strength of these bridges on Indian Railways
has become necessary in view of large-scale gauge conversion works (meter
gauge to broad gauge) and higher axle loads on BG with the introduction of
high-powered locomotives and other higher load carrying freight stocks.
Also, the various damages that an arch bridge undergoes have been
described in some detail, as most of these bridges are more than a century
old. This is because; the load carrying capacity of an existing bridge prior to
and consequent to repair/strengthening cannot be assessed realistically
without incorporating the damages in the models for strength assessment.
The damages are manifestation of inherent problems that are prevailing,
which are first required to be addressed, before any repair or strengthening is
attempted.

Repair and retrofitting solutions for each arch bridge have to be decided for
each bridge on an individual basis taking into account the damages it has
suffered and the load carrying capacity it has to be restored to.

11.0 REFERENCES

(1) Catalogue of damages on masonry arch bridges-UIC Draft report,

October 2.005
(2) Assessment, strengthening and preservation of masonry structures
for continued use in today’s infrastructure- Lynne Mabon, Structural
Analyst, Gifford and Partners, Southampton, UK
(3) http://www.helifix.com
(4) http://www.cintec.com/
(5) IRS- Code of practice for the design and construction of masonry and
plain concrete arch bridges-1941
(6) Proposal for the methodology of retrofit reinforcement-UIC Report
(7) Arch Bridges- Edited by Professor C. Melbourne

35