Bridge Engineering Proceedings of the Institution of Civil Engineers

Bridge Engineering 164 March 2011 Issue BE1

Volume 164 Issue BE1 Pages 314 doi: 10.1680/bren.900030
Recommendations for assessment Paper 900030
Received 03/09/2009 Accepted 26/04/2010
Eurocodes for bridges Published online 11/02/2011
Hendy, Sandberg and Shetty Keywords: bridges/codes of practice & standards/
structural frameworks

ice | proceedings ICE Publishing: All rights reserved

Recommendations for assessment

Eurocodes for bridges
&
1 Chris R. Hendy MA (Cantab), CEng, FICE &
3 Navil K. Shetty MTech, PhD, DIC
Head of Bridge Design and Technology/Director, Atkins Highways and Director, Atkins Rail, London, UK
Transportation, Epsom, Surrey, UK
&
2 Jessica Sandberg BE, CEng, MICE
Senior Engineer, Atkins Highways and Transportation, Epsom, Surrey, UK

1 2 3

The new structural Eurocodes offer increased economy in design over most existing codes of practice. However, the
assessment of existing structures is specifically outside the scope of the Eurocodes, so many of the benefits these new
codes bring cannot be directly applied when reviewing old structures. This paper investigates the areas where the
steel and concrete Eurocodes give increased resistances compared to existing codes and makes recommendations for
sections that could be directly applied to the assessment of existing structures. Areas where the Eurocodes cannot be
applied directly to existing structures are also identified, together with the reasons why, such as reliance on modern
material and execution specifications. Recommendations are then made for how the Eurocode design rules could be
modified for assessment situations, including the use of measured strengths and imperfections in calculation. Actions
and the format for combining actions are also investigated and recommendations are made for modifications to these
aspects for use in assessing bridges. Finally, an overall assessment of the scope of work required to produce an
assessment suite of Eurocodes is made.

1. Introduction gives greater scope for determining the true ultimate strength of
a structure than do more prescriptive or empirical rules.
The new structural Eurocodes offer increased economy in design
over most existing codes of practice. This is achieved through
codification of more advanced calculation methods, such as Traffic data (collected using weigh-in-motion systems for long-
non-linear analysis, and also through codified rules, which give span bridges or from known traffic counts for short-span
increased economy in many situations compared to the old bridges), measured dimensions and material properties derived
British Standard rules because they are based on much more from in situ tests on a structure can be rigorously incorporated
recent testing and non-linear finite-element parametric studies. into the limit state verification method given in the Eurocodes.
This presents great opportunities for improving sustainability in In addition, assessment criteria can be differentiated by taking
design. However, the assessment of existing structures is account of the differences in consequences of failure of bridges.
specifically outside the scope of the Eurocodes so many of the The approaches to be used for developing the relevant rules for
benefits these new codes bring cannot be directly applied when assessment are discussed in this paper.
reviewing old structures. Increased assessments of strength
would be highly beneficial in justifying that old structures with The main area of difficulty in applying the Eurocodes to
increased loading or small amounts of deterioration are still checking existing structures is that the assumptions of particular
adequate to continue in operation without costly refurbishment, material properties and workmanship inherent in the codes may
modification or reconstruction. The Eurocodes lend themselves not be borne out by the actual construction. Many of the
to adaptation for assessment of structures because the rules are enhanced resistances in Eurocode 3 for example stem from the
generally based around the application of first principles; this ductility and strain hardening characteristics of modern steels

Delivered by ICEVirtualLibrary.com to: 3

IP: 202.20.0.166
On: Mon, 24 Oct 2011 21:46:50
Bridge Engineering Recommendations for
Volume 164 Issue BE1 assessment Eurocodes for
bridges
Hendy, Sandberg and Shetty

that are produced to the specification EN 10025 (BSI, 2004a). objective of assessment is to evaluate the safety of an existing
Older steels, or indeed wrought and cast irons, may not comply bridge quickly and with a minimum of effort. However, the
with the requirements of this specification. Similarly, all assessment rules and criteria need to be established rigorously
steelwork designed to Eurocode 3 should be fabricated and and judiciously. If assessments are unduly conservative,
erected in accordance with EN 1090-2 (BSI, 2008a). This structures will be unnecessarily strengthened, or needless load
document sets out, among many other things, the tolerances for restrictions will be imposed. Conversely, if the rules are too lax,
steelwork such as bow imperfections in columns. The strut the safety of bridges could be compromised.
curves in Eurocode 3 assume that these bow imperfections will
not be exceeded but this may not be the case for old bridges, Eurocodes employ characteristic actions and material properties
which were not constructed to the same tolerances; it is combined with load and material partial factors to ensure an
necessary to use the actual tolerances in calculation. The appropriate level of safety and reliability. These factors guard
Eurocodes may not even cover design with certain materials. against extreme variations in design parameters (e.g. material
Plain round bars, for example, are not covered by Eurocode 2 properties and applied loads) which could occur during service.
because of their poor and unpredictable bond characteristics. In order to ensure that the design rules are simple for routine
An additional factor is that the Eurocodes do not directly deal use, the values of the partial factors have been chosen such that
with reduction in strength due to deterioration arising from they cater for a wide range of structure/component types and
unexpected behaviour or lack of maintenance. failure modes. It follows, therefore, that the rules tend to be
conservative for the majority of bridges and the level of
In the absence of assessment versions of the Eurocodes, a conservatism varies considerably from structure to structure.
number of problems and inconveniences will arise for engineers
in the UK. First, there is the obvious problem that engineers The objective of the assessment code would be to produce a
will need to remain conversant with two design standards, more realistic assessment of a structure than is possible at the
namely the Eurocodes and the old Highways Agency (HA) design stage. This in part is achieved through using current
assessment standards, which were heavily based on BS 5400. state-of-the-art analysis techniques and design rules based on
While this is not unmanageable, it is unfortunate because a physical testing where appropriate, and in part by taking
prime driver for the Eurocodes was to avoid the need to use advantage of the information available to an assessing engineer
more than one code of practice. Second, there is the more in respect of the loading, material strength, geometrical
significant problem that structures designed to the Eurocodes properties and imperfections, which necessarily have to be
will mostly fail their assessment when the old HA standards are conservatively estimated at the design stage. The assessment
applied to them because resistances to Eurocodes tend to be code must therefore enable account to be taken of:
higher on average. The avoidance of this problem leads to a
potential third problem. Inevitably, engineers will seek to use (a) traffic loading appropriate to vehicle flows over the
large parts of the Eurocodes in their assessment as a departure bridge
from assessment standards, thus creating a pseudo assessment (b) bridge specific partial factors accounting for conse-
version of the Eurocodes, but without the control, expert quences of failure
input, consistency and consensus that would be present if an (c) measured imperfections and sizes different from those
appropriate committee had assembled the assessment code. assumed in the code
Technical approval authorities will have an important but (d) measured material strengths
difficult role in checking that such departures are appropriate. (e) construction not fabricated or erected in accordance with
the relevant current execution specifications
The remainder of this paper investigates areas where the steel (f) material not complying with the relevant current product
and concrete Eurocodes give increased resistances compared to standards
existing codes and makes recommendations for sections that (g) general configurations and shape limitations not com-
could be directly applied to the assessment of existing plying with the limits in the design code
structures and the modifications required to the sections that (h) minimum stiffnesses and strengths not complying with the
could not be directly applied. Recommendations are also made design code
for the modifications needed to action calculation and the (i) outmoded forms of construction not covered by the
combinations of actions when assessing bridges. design code
(j) deterioration of structural members through condition
2. Key features of an assessment version of factors or modification to material factors.
a Eurocode
The process of assessment is of crucial importance for A feature of previous UK assessment codes has been the
maintaining bridges in a safe and serviceable condition. The introduction of modified, and usually significantly more

4 Delivered by ICEVirtualLibrary.com to:

IP: 202.20.0.166
On: Mon, 24 Oct 2011 21:46:50
Bridge Engineering Recommendations for
Volume 164 Issue BE1 assessment Eurocodes for
bridges
Hendy, Sandberg and Shetty

complicated, calculation rules whereby some of the simplifying load partial factors given in the UK national annex to Eurocode
and conservative assumptions in the equivalent design for- BS EN 1990: Annex A2 (BSI, 2002) can be used unchanged with
mulae are removed. The need to do this in a Eurocode the BSALL.
assessment standard would often be negated because the rules
are usually already more general and computer modelling can For bridges of span less than 50 m, live load reduction factors to
often be used directly to obtain the required results. the load models LM1 and LM2 given in the UK national annex
to BS EN 1991-2 (BSI, 2003) can be derived for assessment
3. Overview of modifications to account for purposes in a similar manner to those given in BD 21 (HA, 2001)
bridge specific loading and resistance depending on traffic density and road surface roughness for
information different levels of assessment loading. The surface roughness
In assessing an existing bridge, account may be taken of a factors in BD 21 may be readily applicable but the live load
number of factors that are specific to a bridge and the reduction factors will require some calibration.
assessment criteria modified to reflect these, for example:
The load models LM3 (abnormal loads) and LM4 (crowd
(a) bridge specific loading loading) given in the UK National Annex to BS EN 1991-2 can
(b) bridge specific resistance properties (material and be used for assessment of existing bridges without further
geometrical) modification, although the derivation of bridge specific crowd
(c) consequences of failure. loading models is also possible.

The work needed to develop assessment provisions for the Where the structural dimensions and the thickness of road
above are discussed in this section. The remainder of the paper surfacing are measured for an existing bridge, the measured
discusses the evaluation of the ultimate strength of bridges dimensions, together with measured values of material density,
based on modifications to Eurocodes 2, 3 and 4. can be used to calculate more realistic values of self-weight and
surfacing load for the bridge. For surfacing, this will allow the
3.1 Bridge specific assessment loading tolerances on range of surfacing depth given in the Eurocodes
The procedures for the derivation of bridge specific assessment to be reduced, provided controls are in place to limit changes
live load (BSALL) for long-span bridges with span greater than to surfacing depths in the future. Depending on the level of
50 m are already well established in BD 50 (HA, 1992), and can variability in measured values and the controls applied, the
be readily extended to Eurocode traffic loading. BSALLs are partial factors for self-weight, superimposed dead load and
often vital in justifying continued use of bridges without the surfacing load given in the UK national annex to BS EN 1990
need for strengthening to meet full design standard loading (see annex A2 could be reduced further.
Figure 1). The derivation of BSALL involves collecting traffic
data at the bridge site through weigh-in-motion systems and 3.2 Bridge specific resistance parameters
using these data to simulate extreme load distributions. The live Where material properties, such as yield strength of steel or
cylinder strength of concrete, have been obtained through
intrusive in situ testing on an existing bridge, the in situ
characteristic strength can be calculated using the Bayesian
updating methodology given in section D7.2 of BS EN 19906.
Further details of this methodology are given by Shetty and
Chubb (2001). The calculation of characteristic strength values
takes into account the number of test results, variability in test
results and prior knowledge about the statistical distribution of
material properties. The in-situ characteristic strength values
can then be used together with the material partial factors
given in the UK national annexes of material Eurocodes EN
1992, EN 1993 and EN 1994 for bridges. Where the ultimate
resistance of a component or product is established through
testing, the design resistance can be calculated using the
procedures given in annex D of BS EN 1990 (BSI, 2002).

In some cases, construction tolerances (imperfections) or

Figure 1. BSALL is used for the operation of the Forth Road Bridge impact distortions may also affect the structures strength
and will therefore need to be considered explicitly if they are

Delivered by ICEVirtualLibrary.com to: 5

IP: 202.20.0.166
On: Mon, 24 Oct 2011 21:46:50
Bridge Engineering Recommendations for
Volume 164 Issue BE1 assessment Eurocodes for
bridges
Hendy, Sandberg and Shetty

greater than those assumed in the Eurocode design rules; special order loads in excess of 180 t gross weight), warning of
Figure 2 shows an extreme case of post-impact deformation. failure related to structural form and material and quality and
Greater discussion on this is included later in Sections 4 and 5. frequency of inspections.

3.3 Consequences of failure It is, however, important that the reliability based assessment
Consequence of failure reflects both the economic impact of procedures are fully standardised, including the probability
the structural failure (in terms of rebuild costs and disruption distributions for the load and resistance random variables and
costs) and also the potential for loss of life. The UK national the target values of reliability index. The necessary background
annex to BS EN 1990 annex A2 recommends that all bridges research in this regard has already been undertaken during the
up to 200 m span should normally be regarded as medium calibration of load partial factors and load combination factors
consequence structures (consequence class 2). Designing for a for use in the UK national annex to BS EN 1990, annex A2.
higher consequence class may be considered for bridges with
span greater than 200 m. However, small culverts or bridges 4. Overview of the economic benefits of
with span less than 10 m or medium-span bridges on a minor Eurocodes 3 and 4 and the modifications
road with light traffic may be regarded as low consequence required to produce a steel assessment
structures (consequence class 1). The annual target reliability code
(as defined in BS EN 1990 annex B) for a low consequence The following discussions are based around Eurocodes EN
structure can be taken as 4?2 as opposed to 4?7 used for a 1993-2 (BSI, 2006a) and EN 1994-2 (BSI, 2005a) and the other
medium consequence structure. As a result the partial factors Eurocodes they cross-reference and identify; in outline, the
on live load can be reduced by a factor of 0?9 for consequence modifications required to produce an assessment standard. The
class 1 as given in table B3 of BS EN 1990, annex B. In various sections of the codes are reviewed one by one.
assessment, such a reassessment of the consequences of failure
may be appropriate, but very careful consideration would be 4.1 General
needed. The scope of EN 1993-2 in its section 1 would need fairly
extensive modification because of the number of cross-
3.4 Reliability based assessment (Level 5 methods) references to modern product and execution standards, but
The Eurocodes allow the direct use of probability based this would mostly involve deletions. The terms, definitions and
structural reliability analysis methods for the design of symbols defined in the section, however, would mostly be
structures. It is important that their use is permitted for the retained. Guidance would have to be provided also on
assessment of existing bridges as they provide significant inspection for assessment and the information on material
benefits for assessment. These methods are referred to as Level properties, imperfections and condition that should be
5 methods in the UK as discussed by Shetty et al. (1998). In obtained and recorded. This could largely be taken from
addition to accounting for bridge specific loading, resistance existing documentation such as BD 56 (HA, 1996a).
parameters and consequences of failure, the reliability based
assessment methods enable account to be taken of a bridges 4.2 Basis of design
previous exposure to known abnormally heavy loads (e.g. The material on the principles of limit state design and design
assisted by testing would require very little amendment for use
in assessment; indeed design assisted by testing would be a
useful assessment tool.

4.3 Materials
The section on materials would require significant modifica-
tion. This section of the Eurocode sets requirements for
minimum ductility and strength. The relevant parameters
would need to come from as-built records and specifications or
from intrusive investigation and testing of the structure as
discussed in Section 3.2. Non-compliance with the prescribed
limits on these parameters, however, would mean that some of
the remaining Eurocode resistance rules could not be used, so
the implications of non-compliance on the remaining design
Figure 2. Imperfections in longitudinal stiffeners following an rules would need to be addressed. This will require either
impact modifications to the strength equations and procedures or the
use of non-linear analysis using known material properties for

6 Delivered by ICEVirtualLibrary.com to:

IP: 202.20.0.166
On: Mon, 24 Oct 2011 21:46:50
Bridge Engineering Recommendations for
Volume 164 Issue BE1 assessment Eurocodes for
bridges
Hendy, Sandberg and Shetty

the structure. Other requirements for minimum toughness to Non-linear analysis, particularly of slender structures, will often
guard against brittle fracture would need less modification, but give the most accurate and realistic picture of true structural
the Charpy energies for use in calculation would again need to strength and is thus a very effective assessment tool. The model
be determined via the means above. The sections on brittle in Figure 3, for example, predicted the physical test failure load
fracture referenced in EN 1993-1-10 (BSI, 2005b) permit the to within 4%. Non-linear analysis methods are, however,
use of fracture mechanics methods to check adequacy, so influenced by imperfections (residual stresses and tolerances
actually lend themselves to assessment. on straightness and flatness) and those given in Eurocode 3 may
not be appropriate for an existing structure. Requirements for
The corresponding section in an assessment code could usefully non-linear analysis would therefore require either magnitudes of
give key material characteristics such as yield strength, ultimate equivalent imperfections to be provided for old structures or for
tensile strength, ultimate strain, toughness and modulus of imperfections to be measured in the existing structure and
elasticity for historic steel grades. This could be readily residual stresses determined by calculation based on, for
provided by extracting the required properties from old example, weld details. Often doing the latter will in any case
superseded steel specifications. An assessment national annex be beneficial compared with the alternative of using the design
would be a good place for this information to be recorded. code values. The rules would need to provide guidance on the
application of imperfection patterns to the structure. Guidance
4.4 Durability would also be needed on the material properties to use in
analysis where the material behaviour does not conform to the
Most of the design requirements involving durability would simply
assumptions and limitations in the design code, but non-linear
be deleted in an assessment version of the standard, as inspections
analysis is, in principle, the best form of analysis provided that
would determine to what extent the design had succeeded in that
the material characteristics are known and can be modelled.
regard. Deterioration would need to be accounted for via a
condition factor in the resistance equations in the same way as is
currently required in UK assessment standard BD 21. Plastic analysis would also be more readily justifiable for use in
assessment since in-service performance can be visually
assessed. The Eurocodes set out the requirements for plastic
4.5 Structural analysis
analysis so only a relaxation in the conditions of use is likely to
Much of what is written on structural analysis would require be needed in the assessment version, provided that the
very little amendment as the principles given are to model the requirements for ductility are met.
structure in a way that reflects its behaviour; this applies
equally to design and assessment. Elastic analysis itself, with
4.6 Ultimate limit states
subsequent member cross-section and codified buckling
checks, needs little adaptation except that facility would be 4.6.1 Material factors and cross-section resistances
needed to enable the buckling curves to be modified to allow An important feature of the design code is the application of the
use of measured imperfections. partial safety factor for material strength, cM. Eurocode partial

Y
Load case =
X 24
Increment 24 Le factor = 1.00318
Results file = 1
Displacement
Z
Contours of dx

-15.4089
-14.0081
-12.6073
-11.2065
-9.80567
-8.40486
-7.00405
-5.60324
-4.20243
-2.80162
-1.40081
0
1.40081
2.80162
TGV 8 4.20243
5.60324
Max 6.367 at Node 7938
Min -16.05 at Node 10935

Figure 3. Physical test and non-linear model of a plate girder

Delivered by ICEVirtualLibrary.com to: 7

IP: 202.20.0.166
On: Mon, 24 Oct 2011 21:46:50
Bridge Engineering Recommendations for
Volume 164 Issue BE1 assessment Eurocodes for
bridges
Hendy, Sandberg and Shetty

factors could generally be used with the characteristic material

properties determined from as-built records or from tests.
Additionally, EN 1990 sets out means of determining material
resistances and factors in its section on design assisted by testing.
This aims to fix material factors such that a base level of (i) Accumulated stresses determined on (ii) Accumulated stresses
gross cross-section used to determine an
reliability is obtained when using the assessment rules, based on effective steel cross-
the true resistances obtained in tests. The rules in EN 1993 utilise section for recalculation
of accumulated stresses
relatively few values of cM for simplicity. Background papers
(e.g. Johansson et al., 2007) are, however, available in a number
of cases to support the use of reduced values in several instances Figure 4. Illustrative procedure for determining effective cross-
where adequate reliability would still be obtained and advantage section in Class 4 composite beams
of this could be taken in assessment.

Class 4 beams with longitudinal stiffeners are treated in the

The Eurocode cross-section resistances make some allowance
same way as beams without longitudinal stiffeners in EN 1993,
for strain hardening in a number of areas (for example in
unlike in BD 56 where a completely different approach to
determining the reduction of strength due to holes and in the
calculation was employed. In BD 56, individual panels and
shearmoment interaction curves) so it would be necessary to
stiffeners are checked for buckling once stresses were
set out what to do if the ductility limits were not met. The
determined in them, generally using gross cross-sections, other
simplest situation in this case would be to revert to the more
than for flange plates where some allowance is made for loss of
conservative provisions of BD 56, adapted to make them fit the
strength due to local plate buckling. There is therefore limited
general Eurocode terminology in terms of notation. A similar
load shedding between components and a single overstressed
exercise has already been done in the drafting of PD 6695-2
component can govern the design of the whole cross-section. In
(BSI, 2008b) for salvaging non-contradictory material from BS
EN 1993-1-5, effective widths are again used to allow for
5400 part 3 (BSI, 2000), to be used in design with Eurocode 3.
buckling of web and flange elements, as for unstiffened class 4
cross-sections, but the same approach is also used for stiffeners
4.6.2 Bending cross-section resistance
Figure 5. This effectively allows load shedding between all the
For class 1 to 3 cross-sections, the calculation procedure and
various elements such that their combined strengths are
results for bending resistance are very similar to those for
optimally used. This represents a significant change from
previous practice. One small area of economy for Class 3 cross-
previous UK practice and can give rise to a significantly
section design is that where the bending resistance is based on
increased prediction of load-carrying resistance. These rules
first yield at an extreme fibre, EN 1993 defines an extreme fibre
could be used with very little modification other than to permit
as the mid-thickness of the flange, rather than the outer surface
adjustment to the reduction factors for plate buckling to allow
as was previous UK practice. This gives a small increase in
for measured imperfections.
economy over BD 56, particularly for shallow beams with
thick flanges. For steelconcrete composite design, EN 1994
4.6.3 Shear buckling resistance and web transverse
employs a similar rectangular stress block for concrete for
stiffeners
plastic design as is used in BD 61 (HA, 1996b), although the
The rules for shear buckling in EN 1993-1-5 (BSI, 2006b) and
resisting compressive stress is slightly higher. This leads to
BD 56 are based on quite different theories but produce similar
slightly increased resistances.

The treatment of class 4 cross-sections and beams with

longitudinal stiffeners (which are treated as class 4 cross-sections
in EN 1993) differs significantly, however, from that in BD 56
and brings significant benefit. Class 4 beams without stiffeners
are treated by making reductions to the compression areas and
then checking stresses against yield when calculated on the
resulting reduced cross-section. The procedure for composite
beams is first to calculate accumulated stresses on the gross
cross-section, following the construction sequence, and then to
determine the effective areas of the compression elements based
on this stress distribution. Finally, the accumulated stresses are
recalculated using the reduced effective steel cross-section at all Figure 5. Typical effective cross-section for design to EN 1993-1-5
stages of construction. This is illustrated in Figure 4.

8 Delivered by ICEVirtualLibrary.com to:

IP: 202.20.0.166
On: Mon, 24 Oct 2011 21:46:50
Bridge Engineering Recommendations for
Volume 164 Issue BE1 assessment Eurocodes for
bridges
Hendy, Sandberg and Shetty

results for shear resistance. Little change would be required for permitting almost full web shear resistance with full
an assessment version of the Eurocodes when the required bending resistance, as shown in Figure 6, which reflects
material properties were met in terms of ductility; the shear the findings of recent non-linear parametric finite-element
resistance of webs is not very sensitive to imperfections so no studies.
adjustment would be necessary.
The rules do rely on ductile behaviour and an amount of strain
The implications of the different theories are, however, more hardening being available. If relevant limits for these were not
significant in the design of transverse stiffeners as Hoglunds met, the simplest resolution in this case would be to revert back
theory (used in the Eurocode) places less demand on their to the more conservative provisions of BD 56, adapted to make
strength. This is reflected in EN 1993-1-5, which allows lighter them fit the general Eurocode terminology.
transverse shear stiffeners to be designed than would be
permitted to BS 5400 part 3. The rules in EN 1993-1-5 have still Longitudinally stiffened cross-sections are treated in essen-
been shown to be conservative and could be refined further in tially the same way in EN 1993-1-5 as for unstiffened cross-
the assessment Eurocode as suggested by Presta et al. (2008). sections, so the same economic benefits can be obtained. To
BD 56, the check of the cross-section would have to be
4.6.4 Shearmoment interaction performed on a panel by panel basis in such a way that any
EN 1993 produces a more economic check of shear and shear stress at all has the effect of reducing bending strength.
moment interaction than does BD 56 and consequently has One important caveat is that longitudinal stiffeners must not
been used in assessment to justify not strengthening existing be susceptible to torsional buckling if this method is to be
bridges. It is more economic for three reasons, as follows. used, otherwise adequate ductility does not exist. If stiffeners
are susceptible to this mode of buckling, a method analogous
(a) Shear does not interact with lateral torsional buckling to that in BD 56 would again be required. EN 1993-1-5
resistance. It only interacts with cross-section resistance. section 10 provides such a method. The rules in EN 1993-1-5
(b) The interaction diagram is a continuous curve, rather for checking susceptibility to torsional buckling are them-
than a series of straight lines as was the case in BD56, as selves a little conservative as they ignore restraint from the
shown in Figure 6. parent plate to which they are attached. Rules could easily be
(c) Even if the cross-section is in class 3 or 4 (so that the provided to include this restraint, as well as measured
bending resistance is limited to first yield), the interaction imperfections, and make a less conservative estimate of
is performed using the plastic bending resistance. The torsional buckling strength. Modified versions of the rules
interaction is truncated by the requirement to limit the in BA 56 (HA, 1996c) could be used.
moment to the elastic moment. This has the effect of
4.6.5 Lateral torsional buckling and distortional
V buckling
While BD 56 gave extensive empirical guidance on lateral
torsional buckling, EN 1993 takes a more theoretical approach.
Vbw,Rd EN 1993, as a general approach, gives only an expression for
slenderness
s
Wy fy
. lLT ~
Vbw,Rd
Mcr
2
where Mcr is the elastic critical buckling moment. No guidance is
given on the calculation of Mcr, which tends to lead the designer
towards performing a computer elastic critical buckling analysis
for its determination. The advantage of this in an assessment
context is that it usually produces a less conservative prediction
Mf,Rd Mel,Rd Mpl,Rd M of slenderness and strength reduction than does a simplified
hand calculation Figure 7. The reduction factorslenderness
curves themselves would need to be adapted to allow use of
Figure 6. Typical interaction diagram for shear and moment to EN measured imperfections in the assessment code, but this would
1993-1-5 be straightforward to do and could be adapted from provisions
in BD 56. For quicker estimates of strength, EN 1994 gives

Delivered by ICEVirtualLibrary.com to: 9

IP: 202.20.0.166
On: Mon, 24 Oct 2011 21:46:50
Bridge Engineering Recommendations for
Volume 164 Issue BE1 assessment Eurocodes for
bridges
Hendy, Sandberg and Shetty

Z
Y X

Figure 7. Elastic critical buckling analysis of historic bridge edge girder

simpler methods of calculation of slenderness without the need 4.9 Fatigue

for computer analysis. The approach to calculation of fatigue life is unlikely to be
different between design and assessment, but it may be
4.6.6 Flexural buckling of struts appropriate to use a different vehicle spectrum for assessment
Flexural buckling slenderness is similarly written in terms of a than for design based on measured traffic conditions. EN 1993-
critical buckling force Ncr so that 1-9 already allows reduced factors of safety to be used where
s damage tolerant conditions exist (i.e. the structures details are
Npl regularly inspected for fatigue cracks and there is some
. l~ redundancy available in the event of a component failure) and
Ncr
this approach is particularly suitable for assessment where
monitoring of the structure is possible. It would also be possible
to incorporate modifications to the SN curves where fracture
Once again, this allows the assessor to use a computer elastic toughness properties were measured directly in the structure.
critical buckling analysis, which usually produces a less
conservative prediction of slenderness and strength reduction
than does a simplified hand calculation of slenderness,
5. Overview of the economic benefits of
particularly in cases where otherwise simplified effective
Eurocode 2 and the modifications
lengths would need to be used. A tapered arch with hangers
required to produce a concrete
and transverse bracing (Figure 8) provides a good example of a
assessment code
case where an effective length would be difficult to determine The following discussions are based around Eurocodes EN
without a computer analysis. The reduction factorslenderness 1992-2 (BSI, 2005c) and its cross-references to EN 1992-1-1
curves themselves would again need to be adapted to allow use (BSI, 2004b) and identify, in outline, the modifications
of measured imperfections in the assessment code, but this required to produce an assessment standard. The various
would be straightforward to do and could be adapted from sections of the code are reviewed one by one.
provisions in BD 56.
5.1 General and basis of design
4.7 Serviceability The modifications required for the general and the basis of
Most of the design requirements involving serviceability would design section would be similar to those discussed for steel above.
be retained in an assessment version of the standard, but
inspections would determine to what extent the design had 5.2 Materials
succeeded in that regard. The section on materials would require modification for
concrete to remove the need for compliance with current
4.8 Fasteners and welds execution standards and reference would need to be added to
Most of the provisions for bolts and welds would be able to the determination of characteristic material properties as
remain in an assessment standard but greater coverage of the discussed in Section 3.2, where no as-built records are
resistance of other connectors, such as rivets, would need to be available. The section on creep and shrinkage would also
added from older codes such as BD 56. require some modification, as usually the detailed concrete

10 Delivered by ICEVirtualLibrary.com to:

IP: 202.20.0.166
On: Mon, 24 Oct 2011 21:46:50
Bridge Engineering Recommendations for
Volume 164 Issue BE1 assessment Eurocodes for
bridges
Hendy, Sandberg and Shetty

Figure 8. Typical tapered arch

composition information required for these calculations will remain valid for assessment. For non-linear analysis, however,
not be available for older structures. In such cases it would be measured imperfections would need to be used if the actual
possible to give recommended values based on common construction tolerances used were not known. Non-linear
concrete mixes that were used at the time of construction. analysis is a particularly appropriate assessment tool because it
produces the most realistic prediction of structural behaviour
For reinforcement, requirements for compliance with current and resistance. However, the practical application of the
execution standards will need to be removed and the properties technique is more limited for concrete structures than for steel
assumed for validity of the design rules, such as ductility, ones because of the added complexity. For members where the
clearly stated. Where reinforcing steels do not comply with EN predicted failure is a flexural one, such as for beams and
1992, alternative rules such as those found in BD 44 (HA, slender piers (Figure 9), the analysis is relatively straightfor-
1995) may need to be used, although generally it should be ward. For more complicated cases where concrete failure under
possible to modify the EN 1992 rules to suit. This might a multiaxial stress field is concerned, the analysis can be much
include removing consideration of strain hardening in the more time consuming, iterative and difficult to agree with
flexural resistance check where the existing steel is brittle. checkers because of the number of concrete models that exist.
Guidance will also need to be included for the treatment of Guidance on the latter could usefully be taken from fib bulletin
plain round bars, which are not covered by EN 1992 because 45 (fib, 2009) among other sources.
of their very poor bond characteristics. The properties of
commonly used historic reinforcing steels could usefully be Plastic analysis would also be more readily justifiable for use in
given in this section. An assessment National Annex would be assessment since in-service performance can be visually
a good place for this information to be recorded. assessed. The Eurocodes set out the requirements for plastic
analysis so only a relaxation in the conditions of use is likely to
5.3 Durability and cover be needed in the assessment version, provided that require-
As for structural steel, most of the design requirements ments for rotation capacity are met. To that extent, it would be
involving durability would simply be deleted in an assessment necessary to give guidance on rotation capacity for reinforcing
version of the standard, as inspections would determine to steels not covered by EN 1992.
what extent the design had succeeded in that regard. The
implications of low concrete cover would, however, need to be
5.5 Ultimate limit states
considered at the ultimate limit state where this adversely
affects anchorage and lap lengths; the EN 1992 rules assume a 5.5.1 Bending
minimum cover equal to the bar diameter in this regard. The rules for flexural resistance would require little modifica-
Deterioration would need to be accounted for by way of a tion but reinforcing steels would need to be classified for
condition factor in the resistance equations, in the same way as ductility; there is a small amount of benefit that could be
is currently required in UK assessment standard BD 21 or by gained from these rules because of the consideration of
other in situ measurement of the residual material available. reinforcement strength beyond yield, although this would
depend on the ductility of the reinforcement and the strain
5.4 Structural analysis hardening characteristics. Moment redistribution would also
As for structural steel, the structural analysis section would provide benefit, as allowed by EN 1992, although there was
require little modification as many of the assumptions would previously allowance in BD 44 for this as well.

Delivered by ICEVirtualLibrary.com to: 11

IP: 202.20.0.166
On: Mon, 24 Oct 2011 21:46:50
Bridge Engineering Recommendations for
Volume 164 Issue BE1 assessment Eurocodes for
bridges
Hendy, Sandberg and Shetty

Y
Y Z
Z
X X

Figure 9. Non-linear analysis of slender piers (deflected shapes

shown for various load cases)

5.5.2 Shear special case of plastic analysis, once again the rules allow
The variable angle shear truss model provided in EN 1992 models to be developed that suit the actual arrangement of
(Figure 10) provides significant benefit for reinforced concrete reinforcement provided. Some additions to the Eurocode for
sections, and using this model for assessment will provide an assessment would, however, be useful. The stress limit in struts
important tool. The resistance model is essentially a plastic where there is transverse tension is currently rather conserva-
solution that allows the resistance to be tailored to the tive in cases where the tensile strains are small and the stress
reinforcement provided. Where concrete shear stresses are limit could be refined to include the effect of tension angle in
not high, greater link resistances can be developed by using a relation to the strut axis and the tensile strain. Furthermore,
flat truss angle. Greater shear stresses can also be allowed additional guidance should be included to advise by how much
compared to BD 44, which will often give improved assessed the idealised model can depart from the elastic flow of forces,
load rating for beams with slender webs. because concrete has limited ductility and excessive departures
from such conditions will reduce the allowable strut and node
5.5.3 Torsion compression limits. Guidance on quantifying this reduction
As for shear, the rules for torsion provided in Eurocode 2 would could also be usefully added.
require very little modification and could be directly adopted for
assessment. Once again, the variable angle truss model gives the 5.5.5 Membrane rules and sandwich models
designer an important tool to tailor the resistance model to the The membrane rules in EN 1992-2 clause 6.109 would be useful
reinforcement provided both longitudinally and transversely. for assessment where shell elements are required in analysis
owing, perhaps, to the complexity of structure geometry and
5.5.4 Strut and tie the lack of applicability of other codified assessment rules.
The strut and tie rules could also be directly adopted for These rules allow stresses to be used directly from a finite-
assessment, see Figure 11. Since strut and tie analysis is a element analysis model for the assessment calculations.
However, the significant limitation of the approach is that
Compression chord (concrete in design is done element by element and hence no redistribution
Stirrups
compression zone)
between elements is possible. This makes the use of the
membrane rules conservative compared to other member
design rules (such as those for shear and torsion) and these
s
member rules should therefore always be used where they are
applicable. Within an element, however, there is some scope to
tailor the analysis to suit the reinforcement directions.
Tension chord (tensile Typical web
reinforcement) compression strut 5.6 Serviceability
Most of the design requirements involving serviceability could
Figure 10. Variable angle shear truss model from Eurocode 2 be retained in an assessment version of the standard,
but inspections are generally more useful in determining

12 Delivered by ICEVirtualLibrary.com to:

IP: 202.20.0.166
On: Mon, 24 Oct 2011 21:46:50
Bridge Engineering Recommendations for
Volume 164 Issue BE1 assessment Eurocodes for
bridges
Hendy, Sandberg and Shetty

~0.125b

0.5b
b

Figure 11. Simple application of strut and tie rules for bridge pier

serviceability performance. This again would be consistent Detailing rules relating to workmanship could simply be
with previous UK practice. deleted, as poor workmanship could be accounted for by way
of a condition factor. This would include, for example,
5.7 Fatigue minimum bar spacing requirements; non-compliance will
Checks of fatigue in concrete and steel were not covered to any potentially lead to poorly compacted concrete, but this should
significant degree by previous UK assessment standards, but be possible to identify from examining the surface or taking a
fatigue can be important in some structures where the details core sample from the concrete.
are unusual. Details that may need assessment include bars
connected with threaded couplers and bent bars used at Detailing rules which affect the applicability of design rules will
locations where the bars are highly stressed. Both these need to be maintained and guidance provided on how to
situations can lead to fatigue problems (Hendy and Smith, handle non-compliant details. For example, minimum cover
2009; Sandberg and Hendy, 2010). EN 1992 provides a realistic rules are required to ensure that the rules for bond strength can
assessment of fatigue in both these situations and would be used. Where the minimum cover requirements are not met,
require minimal modification. The coupler failure in Figure 12 additional rules will need to be provided to determine the
was predicted by the rules in Eurocode 2 for example. effects on the bond strength. Rules on minimum bend radii for
bars are also given in the design code to prevent crushing of the
5.8 Detailing rules concrete inside the bends and to limit bending cracks in the
Section 8 of EN 1992, dealing with detailing, would have to be reinforcement to acceptable levels. Where these minimum radii
modified substantially to deal with non-compliant details. are not complied with, explicit checks on concrete stress can
Detailing rules tend either to be provided to ensure adequate still be made inside the bends, but additional information may
workmanship or to guarantee compliance with assumptions in need to be provided for fatigue strength calculation of such
the design rules. bars.

6. Conclusion
The new structural Eurocodes lend themselves to adaptation
for assessment of structures because the rules are generally
based around the application of first principles. This not only
means that the rules can be adapted relatively easily, but they
also provide scope for determining an accurate prediction of
the true ultimate strength of a structure based on structure
specific information on loads and material properties in a
rigorous manner. This paper has identified areas where the
Figure 12. Broken reinforcement coupler due to fatigue Eurocodes could be directly applied to the assessment of
existing structures and those which would need modification. It

Delivered by ICEVirtualLibrary.com to: 13

IP: 202.20.0.166
On: Mon, 24 Oct 2011 21:46:50
Bridge Engineering Recommendations for
Volume 164 Issue BE1 assessment Eurocodes for
bridges
Hendy, Sandberg and Shetty

has also identified areas where the Eurocodes give increased HA (Highways Agency) (1992) Technical requirements for the
resistances compared to existing codes. assessment and strengthening programme for highway
structures stage 3 long span bridges. In Design Manual for
The amount of work to produce an assessment version of the Roads and Bridges. Highways Agency, London, BD 50/92.
Eurocodes would not seem to be prohibitive, even given a HA (1995) The Assessment of Concrete Highway Bridges and
modest budget, if the scope was limited to adapting the existing Structures. Highways Agency, London, BD 44/95.
rules to assessment without refining them further for economy. HA (1996a) The Assessment of Steel Highway Bridges and
This should be the priority for the work because the Eurocodes Structures. Highways Agency, London, BD 56/96.
themselves provide a better estimate of strength than many of HA (1996b) The Assessment of Composite Highway Bridges and
the UK assessment standards and the lack of a Eurocode Structures. Highways Agency, London, BD 61/96.
assessment standard is likely to lead to inappropriate applica- HA (1996c) The Assessment of Steel Highway Bridges and
tion of the design Eurocodes to assessment. The secondary task Structures. Highways Agency, London, BA 56/96.
of updating the rules and the partial factors would be HA (2001) The Assessment of Highway Bridges and Structures.
considerably more time consuming and would mostly be Highways Agency, London, 2001, BD 21/01.
equally applicable to design. It is therefore unlikely that this Hendy CR and Smith DA (2009) Design of the Dubai Metro light
refinement would be undertaken solely for an assessment code. rail viaducts substructure. Proceedings of the Institution of
Civil Engineers, Bridge Engineering 162(2): 6374.
REFERENCES Johansson B, Maquoi R, Sedlacek G, Muller C and Beg D (2007)
BSI (British Standards Institution) (2000) BS 5400: Design of steel Commentary and Worked Examples to EN 1993-1-5 Plated
bridges. Part 3. British Standards Institution, London. Structural Elements. JRC Scientific and Technical
BSI (2002) BS EN 1990: Eurocode Basis of structural design. Reports, Italy.
British Standards Institution, London. Presta F, Hendy CR and Turco E (2008) Numerical validation of
BSI (2003) BS EN 1991-2: Eurocode 1: Actions on structures. simplified theories for design rules of transversely stiffened
Part 2: Traffic loads on bridges. British Standards plate girders. The Structural Engineer 86(21): 3746.
Institution, London. Sandberg JM and Hendy CR (2010) Replacement of the stays of
BSI (2004a) BS EN 10025: Hot rolled products of structural a major cable stayed bridge. Proceedings of the Institution
steels. British Standards Institution, London. of Civil Engineers, Bridge Engineering 163(1): 3142.
BSI (2004b) BS EN 1992-1-1: Design of concrete structures. Shetty N and Chubb M (2001) Probabilistic methods
Part 1.1: General rules and rules for buildings. British for improved bridge assessment. Proceedings of the
Standards Institution, London. International Symposium on Bridge Management,
BSI (2005a) BS EN 1994-2: Design of composite steel and Singapore, pp. 58.
concrete structures. Part 2: General rules and rules for Shetty N, Chubb MS and Manzocchi GME (1998) Advanced
bridges. British Standards Institution, London. methods for the assessment of bridges. Proceedings of the
BSI (2005b) BS EN 1993-1-10: Design of steel structures. Part International Symposium on Management of Bridges,
1.10: Material toughness and through-thickness properties. Institution of Civil Engineers, London, pp. 35.
British Standards Institution, London.
BSI (2005c) BS EN 1992-2: Design of concrete structures. Part
2: Concrete bridges. British Standards Institution, London.
BSI (2006a) BS EN 1993-2: Design of steel structures. Part 2:
Steel bridges. British Standards Institution, London.
BSI (2006b) BS EN 1993-1-5: Design of steel structures. Part WHAT DO YOU THINK?

1.5: Plated structural elements. British Standards To discuss this paper, please email up to 500 words to the
Institution, London. editor at [email protected]. Your contribution will be
BSI (2008a) BS EN 1090-2: Execution of steel structures and forwarded to the author(s) for a reply and, if considered
aluminium structures. Part 2: Technical requirements for appropriate by the editorial panel, will be published as
steel structures. British Standards Institution, London. discussion in a future issue of the journal.
BSI (2008b) PD 6695-2: Recommendations for the design of Proceedings journals rely entirely on contributions sent in
structures to BS EN 1993-2:2006. British Standards by civil engineering professionals, academics and stu-
Institution, London. dents. Papers should be 20005000 words long (briefing
fib (Federation Internationale du Beton) (2009) Practitioners papers should be 10002000 words long), with adequate
Guide to Finite Element Modelling of Reinforced Concrete illustrations and references. You can submit your paper
Structures. Federation Internationale du Beton, Lausanne, online via www.icevirtuallibrary.com/content/journals,
Switzerland, fib bulletin 45. where you will also find detailed author guidelines.

14 Delivered by ICEVirtualLibrary.com to:

IP: 202.20.0.166
On: Mon, 24 Oct 2011 21:46:50