Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

A single copy of this document is licensed to

[email protected]

On

13/12/2017

This is an uncontrolled copy. Ensure use of the

most current version of the document by searching
the Construction Information Service.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

. ..
I.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Acknowledgements
All members of theTask Group contributed significantlyto the publication but special mention must be
made of the joint editors, Ceorge SomeM'lle OBE, a past Chairman of CBDC, and Bob Lark at Cardiff
University.

Both expended much time and energy in transformingthe information and findings into the finished article
and CBDC would like to express its gratitude to these two and all the other membersof the Group.

Publishedfor and on behalf of the Concrete Bridge Development Group by

The Concrete Society
Riverside House, 4 Meadows BusinessPark,Station Approach, Blackwater,Camberley,Surrey CUI7 9AB
Tel: +44 (0)1276 607140 Fax: +44 (0)1276 607141
www.concrete.org.uk

CCIP-024
PublishedJune 2007
ISBN 1-904482-35-X
0 Concrete Bridge Development Group
Order reference: CBDG/TG9

CClP publications are produced by The Concrete Society on behalf of the Cement and Concrete Industry
Publications Forum -an industry initiativeto publish technicalguidance in support of concrete design and
construction.

CClP publications are available from the Concrete Bookshop at www.concretebookshop.com

Tel: +44 (0)7004 607777

All rights reserved. Except as permitted under current legislation no part of this work may be photocopied, stored in a retrieval
system, published, performed in public, adapted, broadcast, transmitted, recorded or reproduced in any form or by any means,
without the prior permission of the copyright owner. Enquiries should be addressed to the Concrete Bridge Development Group.

Although the Concrete Bridge DevelopmentGroup (limited by guarantee)does its best to ensure that any advice, recommendations
or information it may give either in this publication or elsewhere is accurate, no liability or responsibility of any kind (including
liability for negligence) howsoever and from whatsoever cause arising, is accepted in this respect by the Group, its servants or
agents.

Printed by Cromwell Press, Trowbridge, UK

I
~ Guidance on t h e Assessment
of Concrete Bridges
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Contents
Foreword X

Abbreviations and acronyms xii

Notation xiv
1. The assessment process 1
1.1 The need for assessment 1
1.1.1 Asset management 1
1.1.2 Effectiveness of current programme 1
1.1.3 Network management 2
1.1.4 Sub-standard bridges 2
1.1.5 Assessment documentation 2
1.2 Management of the assessment process 2
1.2.1 Experience 2
1.2.2 The brief 3
1.2.3 Existing information
~
3
1.2.4 Access 4
1.2.5 Site work 4
1.2.6 Level of assessment and analysis 5
1.2.7 Management of resuks 5
1.3 Staged assessments 5
1.3.1 BA79: The management of sub-standard highway structures 6
1.3.2 BA79: Advice on assessment
~
6
1.3.3 Levels of assessment 7
1.3.4 Conclusions 8
1.4 Assessment standards 8
1.4.1 Highways Agency documentation 8
1.4.2 DeDartures from standards
~~
10
1.4.3 Other bridge owners 10
 1.5.1 Whole life assessment 11
1.5.2 Asset management 11
1.5.3 Consequences for assessment
~
13
1.5.4 A continuous assessment Droeramme 13
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

1.6 Assessment of the serviceability limit state 13

1.6.1 Serviceability assessment 14
1.7 Historical aspects of assessment 15
1.8 European developments 17
1.8.1 The BRlME proiect 17
1.8.2 Structural assessment 17
1.8.3 Deterioration 18
1.8.4 Chloride ingress 18
1.9 References 18

2. Technical approval and documentation 19

2.1 General Drocedures 19
2.2 General principles 19
2.3 Obiectives 20
2.4 Key stages 21
2.5 ‘Departures’ and ‘Aspects not covered by Standards’ 25
2.6 Reference 26

3. ~
Inspection for assessment
~
27
3.1 Process and eeneral Drocedures 27
3.1.1 Background and introduction 27
3.1.2 Reauirements for insDection 27
3.1.3 Future requirements for inspection 28
3.1.4 The inspection process 29
3.1.5 Risk assessment and safety issues 30
3.2 Condition assessment 30
~~
3.2.1 BD2l approach 30
3.2.2 CSS bridge condition indicator 30
3.3 Network Rail’s Structures Condition Markine Index (SCMI) 31
3.4 Other work 32
3.5 Preferred approach 33
3.6 References 33

ii
 4. In-situ and laboratory testing 34
4.1 Background and Introduction 34
4.2 Current testing practice 34
4.3 In-situ sampling and testing 35
4.4 Laboratory testing 35
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

4.5 Concrete parameters for assessment 35

4.6 Reinforcement parameters for assessment 37
4.7 In-situ stress measurement - concrete 37
4.8 In-situ stress measurement - steel reinforcement 38
4.8.1 Blind hole drilling 38
4.8.2 Cut bar method 38
4.9 Geophysical Techniques 39
4.10 References 39

5. Loading 40
5.1 General principles 40
5.1.1 Scope of live loading 40
5.1.2 Background t o assessment live loading 41
5.2 Bridge specific Loading 42
5.2.1 Bridge specific probabilistic loading model 42
5.3 Highway surfacing effects 43
5.4 Road traffic and abnormal loads 44
5.4.1 Purposes of highway structure assessment 46
5.4.2 Design and assessment 46
5.4.3 BD86 - Assessment of highway structures for the effects of 46
Special Types General Order (STGO) and Special Order (SO)
Vehicles
5.4.4 BD86 Load Models - SV Vehicles 47
5.4.5 BD86 - HB Conversion Charts 47
5.4.6 Assessment of Abnormal Load Movements using BD86 47
5.5 References 48

6. Analysis for assessment 49

6.1 Assessment principles 49
6.1.1 General principles 49
6.1.2 Analytical procedure 49
6.1.3 Lower bound and upper bound methods 50
6.1.4 Ductility 50
6.1 .S Differences from design 50
 6.1.6 Structural condition 51
6.1.7 Global and local analvsis 52
6.2 Simple methods 52
6.2.1 Strip method 52
6.2.2 BA16 method 52
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

6.2.3 Upper bound limit and check 52

6.3 Elastic methods 53
6.3.1 Elastic grillage 53
6.3.2 Elastic finite elements 54
~~

6.3.3 Westergaard and Charts 56

6.4 Plastic equilibrium methods 57
6.4.1 Plastic redistribution 58
6.4.2 Hillerborg strip method 58
6.5 Yield-line analysis 59
6.5.1 Principles 59
6.5.2 Upper bound solutions 59
6.5.3 Slab decks 60
6.5.4 Beam and slab decks 62
6.5.5 Box culverts and retaining walls 62
6.5.6 Limitations of yield-line methods 63
6.5.7 Computer analysis 64

~ ~~~
6.5.8 Shear 65
6.6 Strut and tie action 66
6.7 Serviceability limit state 66
6.7.1 Inspections 66
6.7.2 Plastic methods of analysis 67
6.7.3 Special inspections 67
6.8 Soil structure interaction 67
6.9 Conclusions 68
6.10 References 68

7. Hidden strengths 69
7.1 Reinforcement anchorage and bond 69
7.1.1 First principles 69
7.1.2 Reinforced concrete beams 69
7.1.3 Sub-standard cover and deteriorated concrete 71
7.1.4 Bond a t laps 72
 7.1.5 Haunched sections 72
7.1.6 Bent-up reinforcement 72
7.1.7 Inclined links 74
7.1.8 Bearing clamping of reinforcement 75
7.2 Shear 76
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

7.2.1 Shear in reinforced concrete slabs 76

7.2.2 Effect of varying section size 77
7.2.3 Shear enhancement a t supports 77
7.2.4 Shear in mestressed flawed beams 79
7.2.5 Shear in webs of post-tensioned prestressed beams 79
7.3 Deck surfacing 79
7.3.1 Properties of surfacing materials
~ ~
79
7.3.2 Discussion 80
7.4 Moment field analysis of reinforced concrete slabs 80
7.4.1 Moment fields 80
7.4.2 Wood-Armer equations 81
7.4.3 Moment field approach 82
7.4.4 Torsionless grillages 82
7.5 ComDressive membrane action
~ ~~ ~
82
7.5.1 Local strength of deck slabs and restraint 83
7.5.2 Global behaviour of beam and slab decks 84
7.5.3 External restraint 84
7.5.4 Non- Iinear nu mer ical a nalvs is 85
7.6 Piers 85
7.6.1 Assessment of piers
~~
85
7.6.2 Slender piers 86
7.6.3 Global restraints 87
7.6.4 Buckling of multiple piers 88
7.7 Redundancy of elements 89
7.8 Parapet edge stiffening 89
7.9 Width of sumorts 90
7.10 Foundations 90
7.1 1 References 91

8. Specific structural forms 92

8.1 Reinforced concrete slabs 92
8.2 Reinforced concrete beams 93

V
 8.3 Half joints 94
8.4 Hinge joints 94
8.5 Box culverts 95
8.6 Precast pre-tensioned beams 96
8.7 Reinforced concrete arches 97
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

8.7.1 Elastic analysis 97

8.7.2 Mechanism analvsis 98
8.8 Post-tensioned structures 99
8.9 References 99

9. Soecific material and assessment factors 100

9.1 Introduction and general principles 100
9.2 Alkali-Silica Reaction (ASR) 101
9.2.1 Background 101
9.2.2 Structural assessment: current practice and reference 102
documents
9.2.3 Kev Doints in structural assessment 103
9.2.4 Structural sensitivity: elements most a t risk 103
9.2.5 Indicative modifications to design models for strength 104
assessment
9.2.6 Assessment of Highway Structures - BD44/95 and BA44/96; 105
BA52/94; BD21/01
9.3 Aggressive chemical attack 105
9.3.1 Background 105
9.3.2 ‘Conventional’ form of sulfate attack 107
~~

9 . 3 . F Thaumasite form of sulfate attack (TSA) 108

9.4 Freeze-thaw action 109
9.5 High Alumina Cement (HAC) 110
9.6 Supersulfated Cement (SSC) 111
9.7 Chlorides 112
9.8 Carbonation 113
9.9 Steel corrosion 115
~

9.9.1 Introduction and background 115

9.9.2 Effects of corrosion 115
9.9.3 Corrosion rate 116
9.9.4 Prestressed concrete 116
9.10 Fatigue 117
9.1 1 Sub-standard reinforcement detailing 118

vi
 9.12 Deteriorated reinforced concrete structures 120
9.12.1 Background and introduction 120

~
9.12.2 Effects of deterioration, and principles for assessment
~
121
9.12.3 Sources of detailed information 122
9.12.4 Procedures in accordance with Highways Agency 122
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

documentation
9.13 References 123
10. Load testing 126
10.1 General principles 126
10.1.1 Background 126
10.1.2 Reasons for presence of reserves of strength 126
10.1.3 Types of load test 127
10.1.4 The aims of load testing 127
10.1.5 Documents in the Design Manual for Roads and Bridges 128

~~
10.1.6 Safety ~
128
10.2 Supplementary load testing 129
10.2.1 Status 129
10.2.2 Methodology 129
10.2.3 Further guidance 129
10.3 Proof load testing 129
10.3.1 Background 129
10.3.2 Methodology 130
10.3.3 Application 130
10.3.4 Current status 131
10.3.5 Safety 131
10.4 Full-scale or model-scale testing to failure 131
10.4.1 Background 131
10.4.2 Broad principles
~
132
10.4.3 Extent of testing 133
10.4.4 Application 134
10.5 Hybrid and other forms of load testing 134
10.5.1 Examples 134
10.5.2 Load testing for monitoring 134
10.5.3 Reliability and updating 135
10.6 Dvnamic load testing 135
10.6.1 Structures where dynamic response is critical 135
10.6.2 Dynamic tests for health monitoring 136

vii
 10.6.3 Applications for assessment 136
~~

10.6.4 Case studies and research in the literature 137

10.7 References 138

11. Reliability and risk-based techniques 139

1 1 .I Background 139
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

I 1.2 The appropriate application of reliability and risk-based assessments 140

1 1.3 Overview of reliability-based assessments 141
11.3.1 General 141
11.3.2 Uncertainties in bridge assessment 141
11.3.3 Reliability Index 142
1 1.3.4 Interpretation of Reliability Analysis Results 143
11.4 AcceDtance criteria
~~
144
11.5 Bridge specific risk-based assessments 146
11.6 References 147

12. Bridge management and assessment 148

12.1 General principles 148
12.1.I Managing sub-standard bridges 148
12.1.2 The BA79 Advice Note 148
12.1.3 Managing the results of assessments 149
12.1.4 Immediate risk structures 149
12.1.5 Interim measures during assessment 150
12.1.6 Interim measures on completion of assessment 150
12.1.7 Assessment and maintenance 151
12.2 Monitoring 151
12.2.1 Appropriate monitoring 152
12.2.2 Classes and frequency of monitoring 153
~~
12.2.3 Application of monitoring 153
12.2.4 Permanent monitoring 154
12.2.5 Monitoring specifications 154
12.2.6 Methods and took for monitoring 155
12.3 Cracking and crack widths 156
12.3.1 Crack widths 156
12.3.2 Types of crack and diagnosis 157
12.4 Deterioration rates 158
12.4.1 The causes of deterioration 158
12.4.2 Timescales and rates of deterioration 158

viii
 12.4.3 Structural assessment 161
~

12.4.4 Structural amraisal 162

12.5 Legal matters 163
12.5.1 Historical background 163
12.5.2 Current situation 164
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

12.6 The assessment of bridges for abnormal loads 167

12.6.1 Abnormal load categories 167
12.6.2 Management of abnormal indivisible load movements 167
12.7 References 170

Appendix Relevant historical references to design and materials 171

specifications and standards used in concrete bridge
construction

ix
 Foreword
This publication contains guidance on a variety of topics related t o the assessment of
concrete bridges.

It was prepared by the Assessment Task Group (ATG) of the Concrete Bridge Development
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Group and is a development of the report of the ATG published in 1997. Each Chapter is
based on an early draft prepared by a member of ATG, which was subsequently reviewed
and extended by other members. The whole publication was then edited by Professor
Ceorge Somerville and the Convenor of the group, Dr R. J. Lark (Cardiff University).

At the time of publication, the Assessment Task Group consisted of the following
members:

Bob Lark (Convenor, 2001 t o date) Cardiff University

Graham Cole (Convenor, 1997 t o 2001) Surrey County Council
David Bone Royal Haskoning
Mike Chubb Atkins Highways & Transportation
Robin Church Essex County Council
Steve Denton Parsons Brinckerhoff
lan Frostick Network Rail
Paul Jackson Gifford and Partners
Martin Lynch Highways Agency
George Somerville Consultant

The input of Brian Barton of the Highways Agency and David Cullington, formerly of the
Transport Research Laboratory, is also acknowledged.

The current assessment programme effectively started in 1984 with the publication of
Departmental Standard BD2l. Substantial progress has been made since then and it
might appear that there was little need for guidance on assessment. However,
substantial numbers of sub-standard structures still exist, many of which may benefit
from further assessment. There is a growing realisation that a continuous assessment
process forms an essential part of the sustainable maintenance of the national bridge
stock. The purpose of this Guide is t o both record the experiences of past years and t o
give examples of best practice for the future. It should be used with due care and
attention, as they may not be applicable t o a specific project. The intention is that they
should be used as supporting documentation in conjunction with the appropriate
standards and, therefore, have been written based on the assumption that the reader
will have some experience and a basic knowledge of bridge assessment.

The Concrete Bridge Development Group (CBDG) and the members of the ATG assume
no responsibility for the adequacy of the advice given, nor for the legal, contractual or
financial consequences of its use.

The intention of this Guide is t o address the assessment of all types of concrete bridge,
irrespective of their use or ownership. It is recognised nevertheless that a significant

X
 proportion of such structures are reinforced concrete highway structures, and that many
of the standard procedures that are currently used are founded in the requirements of
the Highways Agency and/or, when road over rail, Railway Group Standards. As such,
much of the guidance that is given here is also based on these requirements, but
wherever possible reference is also made t o new research and alternative points of view,
which it is believed will both aid and advance the assessment process.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

References and discussion of the advice and requirements contained in other

publications and Standards, particularly the Highways Agency’s Design Manual for Roads
andBridges, will inevitably become out of date, as these documents are themselves
updated from time to time In drafting this guidance, the task group has endeavoured t o
focus on the principles that underpin best practices in the assessment of concrete
bridges, and therefore t o provide guidance which, whilst set in the context of current
requirements, should hopefully remain useful even i f these requirements change

Comments and feedback on this Guide would be welcomed. They should be sent t o the
Secretary, Concrete Bridge Development Group, at the address given inside the front
cover.

R. J . Lark
July 2006
 Abbreviations and acronyms
AIL Abnormal Indivisible Load
AIP Approval in Principle
ALL Assessment Live Load
AS R Alkali-Silica Reaction
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

AW Authorised Weight
AWR Authorised Weight Regulations
BCA British Cement Association
BCI Bridge Condition Indicator
BRIME Bridge Management in Europe research project
BSALL Bridge Specific Assessment Live Loading
BSLL The Basic Static Live Load effect
C&U Construction and Use
CBDG Concrete Bridge Development Group
CEB Comite Euro-International du Beton (Euro-International Concrete
Committee)
CDM Construction, Design and Management
COBRAS Concrete Bridge Assessment Package
CONTECVET EU project on the Service Life of Concrete Structures (BE 4062)
css CSS (Formerly the County Surveyors Society)
DAF A Dynamic Amplification Factor
DfT The Department for Transport
DMRB Design Manual for Roads and Bridges
fib Federation internationale du beton (The International Federation for
Structural Concrete) - Formed by the merger of CEB & FIP
FIP Federation lnternationale de la Precontrainte (International
Federation for Prestressing)
FTA Freeze-Thaw action
HA The Highways Agency
HAC High Alumina Cement
HACC High Alumina Cement Concrete
HCV Heavy Goods Vehicle
HRM High Speed Road Monitor
HSE The Health and Safety Executive
IAN Interim Advice Note
N DT Non-destructive testing
NLFE Non-linear finite element analysis
NSCLTB National Steering Committee for the Load Testing of Bridges
NVQ National Vocational Qualification
Pfa Pulverised fuel ash
REHABCON EU project on Strategies for the Maintenance and Rehabilitation of
Concrete Structures
RH Relative Humidity
SCC Stress corrosion cracking
SCMI Structure Condition Marking Index

xii
 sls Sewiceability limit state.
so Special Order
SRPC Sulfate-Resisting Portland Cement
ssc Supersulfated cement
STG 0 Special Types General Order
TA Technical Approval
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

TAA Technical Approval Authority

TA F Technical Appraisal Form
TAS Technical Approval Schedule
TRL Transport Research Laboratory
TSA Thaumasite form of sulfate attack
UDL Uniformly distributed load
U Is Ultimate limit state
VSE Vehicle Standards and Engineering
wcs Worst Credible Strength

xiii
 Notation
Ash Area o f bent-up bars
ASV Cross-sectional area of vertical stirrup legs at each stirrup location
B Force in bent-up reinforcement
C Live load capacity factor/Compression force
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

C Diameter of loaded area

D Width of support
6 Deflection
E Elastic modulus
F", The total ultimate anchorage force in the longitudinal reinforcement
fCl
Concrete strength at transfer
fC" Concrete cube strength
fsh Stress in bent-up bars
ft Concrete tensile strength
fY
Yield stress
fsv Yield strength of reinforcement
H Depth of section
h Slab thickness
I0 The performance level of a structure at the time of construction
I,,,,. The critical performance level of a structure
1, The performance level of a structure at the time of an assessment
K Load reduction factor
I,1, 4 Linear measures/span
M, m Values of moment
M* Resistance moment
P Value o f point load
Q Applied force
R Reaction force
Rtt A random multiplier t o model the uncertainty in a load effect due t o the
static weight of vehicles
Radius
Stirrup spacing along a beam
Slab span
Live load effect (e.g. bending moment or shear force)
Tensile force
Wall thickness
Shear force
The fully anchored shear capacity
Shear capacity
Concentrated point load
Uniformly distributed load
Linear measure
Effective depth
Angular measures
Factor t o allow for enhancement of bond due t o transverse pressure

xiv
 P Factor used in the calculation of bond stress influenced by bearing clamping
action
Yi Loading partial safety factor
Ym Material partial safety factor
A Joint displacement
P Coefficient of friction
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

P Transverse bearing clamping pressure

r Reduction factor
 1. The assessment process

1.1 The need for The highway bridge assessment programme in the United Kingdom effectively began
assessment with the publication of Departmental Standard BD21 in 1984. (Frequent reference will
be made throughout this Guidance document to relevant Standards (BDs) and Advice
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Notes (BAs) produced by the Highways Agency. A full list of these is given in Section
1.4,) This Standard replaced the previous document, BE1/73, and introduced the limit
state concept based on the Bridge Code, BS 5400.

The requirement to permit the passage of vehicles of up to 40 tonnes and with 11.5
tonne axles was introduced by directive 85/3/EEC. The United Kingdom and Ireland
obtained derogation from the directive until 31 December 1998.

Despite having 14 years to complete the assessment programme, work was not
completed by the due date and risk management measures had to be introduced.
Current estimates are that the initial programme will take a further 1 to 10 years to
complete, depending on the bridge owner. Nevertheless,the main area for future work,
under current vehicle regulations, is with re-assessment.

1.1.1 Asset tTlanagement The emphasis today is very much on asset management. This means moving away from
reactive maintenance (simply spending roll forward budgets), and towards targeted
schemes which increase the overall value of the bridge stock (the asset), but at lowest
whole life cost.

Key activities in asset management are the inspection and assessment process. It is
necessary to determine the general condition and structural safety of each item in
the overall asset, i.e. each bridge. Having established the current state, it is also
necessary to estimate deterioration rates, so that some idea of the residual life of the
structure can be determined. This work should be done to a consistent standard, so
that investment decisions can be made with some degree of confidence across the
network.

Use of this Guide will assist in the effective management of an important national
asset.

1 . I .2Effectiveness O f Current The assessment programme has been carried out by a very large number of
programme organisations to varying standards. The actual capacity of structures which were given a
40 tonne rating is not known. This approach was satisfactory a t the time but does not
lend itself to determining the residual life of a structure and, ultimately, the value of the
asset. Many bridges that just passed assessment in 1984 may now have become
sub-standard.

Previous page
is blank 1
 A t the beginning of the assessment programme it was anticipated that the majority of
the problems would occur with shear capacity in bridges built before 1973 when design
rules changed. Although this did occur there were a significant number of theoretical
failures in flexure. The assessed levels of loading were also found, in many cases, to be
well below 32 tonnes, which was the previous loading standard.

A significant amount of experience has been acquired during the assessment

Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

programme, much of which has been highlighted in a review of the programme

undertaken by Parsons Brinckerhoff on behalf of the Highways Agency’ ’. This Guide
draws on this experience and summarises good practice.

1. I .3 Network management The Highways Agency, in particular, is looking a t route management strategies and the
management of abnormal load vehicles. For this, it will be necessary to know the
assessed capacity and condition of a bridge to a consistent standard.

1. I .4 Sub-standard bridges Bridges that fail assessment are described as sub-standard. In a simplistic approach,
these ‘failures’ then become the strengthening programme for future years. There are
thousands of these bridges across the country. It may be that a careful application of
further assessment techniques could remove these structures from the programme
saving millions of pounds in both direct and indirect costs. Use of this Guide will assist in
the maintenance of a safe highway network.

1.1.5 Assessment There is information on assessments contained in Highways Agency publications, text
documentation books, conference proceedings, learned journals, research papers, etc. This Guide
provides easy reference to much of this source material.

1.2 Management of the This Section sets out some basic concepts that need to be taken into account in the
assessment process management of the assessment process and serves as an introduction to later, more
detailed, guidance. It is based on the conclusions of the work of the first Concrete Bridge
Development Group (CBDG) Assessment Task Group, which was completed in 199712.

I .2.1Experience Bridge assessment requires the use of experienced engineers and inspectors who can
interpret defects present in a bridge and who understand the background to Codes of
Practice. The Highways Agency and the CSS (formerly the County Surveyors Society) are
considering the implementation of an accreditation scheme for bridge inspectors. Other
bridge owners already run NVQ schemes. The client should ensure that the staff
proposed for the project have sufficient experience.

2
 1.2.2The brief The initial stages of the assessment programme were primarily used t o identify those
bridges which, when using conservative assumptions, could be shown t o be satisfactory.
More sophisticated, and expensive, techniques could then be concentrated on the 30% or
so of the bridge stock that failed this initial assessment. This concept of staged assessment
has now been incorporated in HA Advice Note BA79. More recently, the need t o carry out
assessments t o a consistent minimum standard, say Level 3 of BA79, has been proposed
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

t o assist with asset management. Therefore, the requirements for a brief have changed
since the assessment programme began with the first publication of BD2l in 1984.

In inviting tenders for assessment, the bridge owner is required t o produce contract
documents that include the scope of the work and available existing information. A key
objective should be t o allow the assessing engineer t o concentrate on the investigation
and analysis of the bridge concerned by removing, as far as possible, the uncertainties of
the process. Typical examples of uncertainties are:

0 Existing information
0 Statutory undertakers
0 Access
0 Possessions
0 Testing
0 Analysis
0 HB, C&U and abnormal load assessments
0 Health and Safety.

For a lump sum bid, the assessment engineer requires a clear brief, free of requirements
t o make guesses at the tender stage on the extent of work required. The high level of
risk, which exists during the desk study and site stage, tends t o reduce money available
for assessment calculations with the inevitable consequences.

The approach which should give best value is t o use a mixture of lump sum and time
charge work. The lump sum should be restricted t o work that can be defined clearly.
Time charge work could be set against a pre-agreed budget that should not be exceeded
without the client’s permission. However, this requires a professionally qualified client
who can enter into technical discussion with the assessment engineer. It does allow
more of a team approach and the clients are much more able t o ensure that the end
product meets their requirements.

1.2.3 Existing information The existence of good records of design, alterations and previous assessments, as well as
drawings, considerably assists the assessment engineer. The procurement of drawings is
sometimes left t o the assessment engineer. This may be an easy task or may involve
considerable time and effort. This is a risk item in a lump sum bid. The client is generally
in the best position t o obtain this information.

The assessment engineer is often requested t o contact the statutory authorities t o

determine the approximate location of existing services. The client should obtain this

I
3
 information. If it is necessary to determine precisely the position of services by
constructing trial pits, then this should be incorporated in a comprehensive site testing
and investigation programme.

1.2.4 Access Frequently, assessment engineers are requested to make their own arrangements for
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

access across private land. This is a risk item since the difficulty of making such
arrangements cannot generally be determined at tender stage. Local Authorities have
powers under the Highways Act 1980 to obtain access for the purpose of inspecting,
surveying and maintaining highway structures. They are also responsible for making
compensatory payment if damage is caused to private property. Details of adjacent land
owners and access arrangements should be provided by the client leaving the
assessment engineer only to obtain access.

The need to arrange for road closures can be assessed and undertaken by the
assessment engineer, but adequate time should be allowed for the process when setting
a contract programme.

A further complication is introduced when dealing with railway tracks. Obtaining track
possession can be a protracted and time-consuming business, which is a risk item for the
assessment engineer. Track possessions should be booked and paid for by the client,
with the assessment engineer being responsible for undertaking the work within the
possession and for the cost of any overrun. This cost must be stated clearly in the tender
documents. The possessions must also be a reasonable period within which to undertake
the work required. The costs associated with cancelled possessions should be
recoverable. Similar considerations can apply to lane closures on motorways.

Health and Safety is becoming ever more important and the Construction, Design and
Management (CDM) regulations impose certain obligations on the client. Bridge
inspection work often falls outside the regulations. However, the client should ensure
that adequate Health and Safety measures are proposed by the assessment engineer.
Providing adequate Health and Safety is expensive and in the past, attempting to
minimise the tender price, assessment engineers have sometimes taken unacceptable
risks. The client should carry out risk analysis of each bridge in a commission so that
access arrangements and methods of working are common to all tenderers.

1.2.5 Site work Having carried out a desk study, collected existing information, obtained the necessary
approvals and established a safe method of working, the inspection and possible testing
of the structure can be carried out.

It is important to realise that an inspection for assessment differs from other types of
inspection.Given limited possession times, it is essential that effort is concentrated on areas
that will be critical to the assessment calculations.It may be useful to carry out a
preliminary analysis of the structure prior to site works, so that these critical areas can be
identified.

4
 The extent of testing required is often left t o the discretion of the assessment engineer.
This is an unacceptable practice in a lump sum bid as it leaves the assessment engineer
with the dilemma of how much t o include and the client with lump sum bids which may
vary widely depending upon the assumptions made by the assessment engineer. The
requirements for testing can generally be best determined after or during the inspection
and should be paid for based upon a schedule of rates t o be expanded after discussion
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

between the client and assessment engineer on the testing required t o achieve the
clients objectives. Alternatively, the client must define precisely what testing is t o be
undertaken in the contract documents.

1.2.6 Level O f aSSeSSment The expectations of the client on the degree of sophistication of analysis must be
and analysis defined clearly. The client should not, as was frequently the practice, leave the
assessment engineer with the requirement t o include for more sophisticated analysis if
the bridge fails under simple methods. This approach requires the assessment engineer
t o determine at tender stage whether a particular bridge is going t o fail a simple
assessment and, therefore, t o include for further work. In view of the introduction of
BA79, and the growing interest in asset management, the client should state the
minimum level of assessment. Therefore a clearly defined approach t o analysis is
required with payments for the work undertaken. Non linear computer work, special
modelling and reliability analysis in accordance with a Level 4 or 5 assessment in BA79
should be paid for o n a time basis, or by an agreed lump sum when the extent of the
work can be defined clearly.

Generally, an analysis of the assessment of a bridge t o carry abnormal load vehicles

should be carried out if a 40 tonne Assessment Live Loading is obtained. There may be
bridges on certain minor roads where this process cannot be justified. The assessment
live loading model has changed from HB loading t o Special Types General Order (STGO)
loading (see Section 5.4).

1.2.7 Management Of reSU[tS Once an assessment has been completed t o an appropriate level of sophistication,
which has been agreed with the client and Technical Approval Authority (TAA), see
Chapter 2, then the results need to be managed. If the result is less than 40 tonnes then
the bridge is defined as sub-standard and should be managed in accordance with BA79.
If the bridge has achieved 40 tonne Assessment Live Load (ALL) t o an agreed minimum
level of analysis then the result can be used t o assist in determining a long-term
maintenance regime for the structure. Finally, it is necessary t o determine a date by
when the bridge is re-assessed.

1.3 Staged aSSeSSmentS The traditional approach t o assessments has been t o use the simplest method of
analysis in order t o achieve a 40 tonne Assessment Live Loading. Indeed, this technique
is still referred t o in paragraph 6.2 of the 2001 version of BD2l as follows:

5
 0 me assessment process

‘I. the choice of the appropriate method (of analysis) will depend upon the structural
..
form of the bridge and the required degree of accuracy. The simple methods, although
conservative, are quick to use and should be tried initially where appropriate, before
progressing to more accurate but more complex methods. ’I

A staged approach t o assessment has been defined in BA79

Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

1.3.1 BA79: The This Highways Agency (HA) Advice Note was published t o provide guidance on the
management of sub- implementation of interim measures for structures that were assessed to be sub-
standard, or provisionally sub-standard, in order t o maintain the safety of the highway
standard highway structures network.

It provided further advice on assessment methods and, for the first time, set out formal
definitions of levels of assessment in its Appendix B.

It was also intended that the Advice Note would provide guidance applicable t o “future
cyclical assessments within maintenance management programmes”.

The Advice Note was drafted by a working group consisting of representatives from
a large number of bridge owners, consulting engineers, safety experts and the
Transport Research Laboratory. I t s use is applicable t o virtually all highway bridges
and, for the first time, gives the bridge owner a sound method for dealing with the
problem of managing bridge and other highway structures that have failed
assessments.

1.3.2 BA79: Advice O n The details given in paragraph 2.1 of the BA79 are important and are repeated here in
assessment full.

“The process of assessment and subsequent action is of crucial importance for ensuring
that all highway structures remain in a safe and serviceable state. The assessment rules
and criteria need to be applied rigorously and in a consistent manner. Ifassessments are
unduly conservative, structures will be unnecessarilystrengthened, using up scarce
resources and causing traffic disruptions. A t the same time, if the rules are lax, or applied
unevenly, some structures where the margins of safety are unacceptable may be left
without appropriate measuresbeing implemented. I’

The Advice Note provides a flow chart t o illustrate the assessment process and related
measures. It also recognises the importance of recording the various deliberations
associated with the management of sub-standard structures and provides a typical
proforma. This Advice Note has more references t o the role and requirements of the
Technical Appraisal Authority than probably any other HA Standard or Advice Note.
These topics are discussed further in Chapter 12.

6
 Paragraph B1.3 of Appendix B of the Advice Note states:

“Structural failure is not acceptable t o the public hence the order of probability of failure
inherent in the assessment criteria is very small. When a structure is assessed to be sub-
standard, it does n o t mean therefore that it will necessarilyfail or collapse. However, if
such structures were left in large numbers without remedial action, there may be an
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

unacceptable risk that a collapse in service would occur. The assessments are based on
probabilities and therefore it is impossible to know beforehand which bridges would
actually fail in practice. ”

1.3.3Levels O f aSSeSSment It is possible t o carry o u t assessments at five distinct levels, as defined by BA79. Existing
assessment Standards and Advice Notes are only applicable t o Levels 1, 2 and 3. Any
assessment carried out t o Levels 4 and 5 will require the detailed approval of the TAA
concerned. Each level of assessment has been summarised below. Reference should be
made t o Appendix B of BA79 for full details.

Level 1 - Simple
The simplest level of assessment, using simple analytical methods, which will give a
conservative estimate of load capacity.

Level 2 - Refined
This level of assessment involves the use of more refined analysis and better structural
idealisation. More refined analysis may include grillage analyses and finite element
analyses. Non-linear and plastic methods of analysis may also be used. This level also
includes the determination of characteristic strength for materials based on existing
available data, but not specific site testing.

Level 3 - Bridge Specific

A bridge specific assessment live loading may be derived rather than the use of values
from BD21 or BD50 (see Section 1.4). This approach is unlikely t o be justified for short
span bridges. This level allows the use of specific material testing t o determine Worst
Credible Strengths.

Level 4 - Modified Criteria

This level involves the use of partial safety factors that would be specific t o a particular
bridge. It should be noted that the derivation of structure specific partial safety factors
involves applying the fundamental principles of reliability theory. The Advice Note offers
the following warning:

“The background knowledge and engineeringjudgement requiredfor this level of

assessment is of a high order and hence, the TAA needs to be closely involved.”

Draft rules have been prepared but not issued by the HA although advice can be sought
from the HA if required.

7
 Level 5 - Reliability
This level involves reliability analyses that require probability data for all the variables
defined in the loading and resistance equations. The Advice Note states:

“This type of assessment requires specialist knowledge and expertise and is only likely to
be worthwhile and possible in exceptional cases.”
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

1.3.4COnChSlOnS For bridge owners with a small number of structures on mainly local roads it would be
sensible to continue to use the simplest level of assessment that is appropriate.
However, for larger owners with a more complex stock of bridges on primary routes or
motorways, there are good economic reasons for developing whole life costing
maintenance strategies. In these circumstances it is necessary to carry out assessments
to a consistent standard, say Level 3, rather than just determine ‘pass’ or ‘fail’.

1.4 Assessment standards The Highways Agency provides the principal source of Standards applicable to the
design and assessment of highway bridges in the United Kingdom. These documents are
contained within the Design Manual for Roads andBridges. For completeness, the
documents current at the time of writing are listed here; the Reader should check that
he/she is using the current versions and any standards issued subsequently. The
requirements of other bridge owners are also considered.

1.4.1 Highways Agency The documents listed below form part of the Design ManualforRoadsandBridges
documentation (DMRB) in Volumes 1 and 3. They are logged on website:-
http://www.standardsforhighways.co.uk/dmrb/index.htm, which should be consulted for
revisions and updates. Throughout they are referenced as Highways Agency publications
but in this context this normally includes the other overseeing departments - The
Scottish Executive, The Welsh Assembly and The Department of the Environment for
Northern Ireland. For details refer to the publications themselves.

Volume 1
Section 1 Approval Procedures
BD2/05 Technical Approval of Highway Structures on Motorways and Other Trunk
Roads Part 1 : General Procedures

Volume 3
Section 1 Inspection
Part 2 BD54/93 Post-tensioned Concrete Bridges. Prioritisation of Special
Inspections
Part 3 BA50/93 Post-tensioned Concrete Bridges. Planning, Organisation and
Methods for Carrying Out Special Inspections
Part 4 BD63/07 Inspection of Highway Structures

a
 Section 2 Maintenance
Part 1 BD62/07 As Built, Operational and Maintenance Records for Highway
Structures

Section 3 Repair
Part 1 BA30/94 Strengthening of Concrete Highway Structures Using Externally
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Bonded Plates
Part 2 BA43/94 Strengthening, Repair and Monitoring of Post-tensioned Concrete
Bridge Decks
BD27/86 Materials for the Repair of Concrete Highway Structures
BA35/90 The Inspection and Repair of Concrete Highway Structures
Part 3 BA83/02 Cathodic Protection for use in Reinforced Concrete in Highway
Structures

Section 4 Assessment
Part 1 BD46/92 Technical Requirements for the Assessment and Strengthening
Programme for Highway Structures. Stage 2 - Modern Short Span
Bridges
Part 2 BD50/92 Technical Requirements for the Assessment and Strengthening
Programme for Highway Structures. Stage 3 - Long Span Bridges
Part 3 BD21/01 The Assessment of Highway Bridges and Structures
Part 4 BA16/97 The Assessment of Highway Bridges and Structures (Incorporating
Amendment No. 1 dated November 1997 and Amendment No. 2
dated November 2001)
Part 5 BA38/93 Assessment of the Fatigue Life of Corroded or Damaged
Reinforcing Bars
Part 6 BA39/93 Assessment of Reinforced Concrete Half-joints
Part 7 BD48/93 The Assessment and Strengthening of Highway Bridge Supports
Part 8 BA54/94 Load Testing for Bridge Assessment
Part 9 BA55/06 The Assessment of Bridge Substructures and Foundations,
Retaining Walls and Buried Structures
Part 10 BA52/94 The Assessment of Concrete Structures Affected by Alkali Silica
Reaction
Part 1 BD56/96 The Assessment of Steel Highway Bridges and Structures
Part 2 BA56/96 The Assessment of Steel Highway Bridges and Structures
Part 3 BA51/95 The Assessment of Concrete Structures Affected by Steel Corrosion
Part 4 BD44/95 The Assessment of Concrete Highway Bridges and Structures
Part 5 BA44/96 The Assessment of Concrete Highway Bridges and Structures
Part 16 BD61/96 The Assessment of Composite Concrete Highway Bridges and
Structures
Part 17 BA61/96 The Assessment of Composite Concrete Highway Bridges and
Structures
BD34/90 Technical Requirements for the Assessment and Strengthening
Programme for Highway Structures. Stage 1 - Older Short Span
Bridges and Retaining Structures
BA34/90 Technical Requirements for the Assessment and Strengthening
Programme for Highway Structures. Stage 1 - Older Short Span
Bridges and Retaining Structures

I 9
 1 Uhe assessment process

Part 18 BA79/06 The Management of Sub-Standard Highway Structures

(Incorporating Amendment No. 1 dated August 2001)
Part 19 BD86/04 The Assessment of Highway Bridges & Structures For the Effects of
Special Types General Orders (STGO) and Special Order (SO)
Ve hicles
Part 20 BD81/02 Use of Compressive Membrane Action in Bridge Decks
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

1.4.2 Departures from The relevant Highway Authority’s procedure for agreeing structural ‘Departures from
Standards Standards’ and ‘Aspects not covered by Standards’ is an important part of the Technical
Approval procedure but will vary between different Authorities. It is usual for the
procedure to require the assessor/designer to recommend a Departure and, with the
agreement of the checker, submit this with a justification to the Technical Approval
Authority (TAA). The TAA is required to consider the proposal and should seek advice
from an appropriate specialist, who has a particular expertise in the subject of the
Departure. Usually, final approval and endorsement requires a positive recommendation
from both the TAA and the relevant specialist. See Chapter 2 for further information on
the Technical Approval process.

1.4.3Other bridge Owner8 This Guide is written primarily for highway bridges. They are equally applicable to
motorway bridges and bridges carrying local roads. Therefore, there will be a large
number of bridge owners involved besides the Highways Agency.
~

Another major national bridge owner, Network Rail, is responsible for a significant
number of road over rail bridges although these are mostly of metal or masonry
construction. Network Rail issues Railway Group Standards that are similar to, and make
extensive reference to, Highways Agency Standards. Network Rail also issues a number
of Current Information sheets, which effectively provide information on seeking
Departures from Standards.

There are a large number of autonomous local authorities, some of which have specialist
bridge teams. The Bridges Group of the Engineering Committee of the CSS (formerly the
County Surveyors Society) provides a network support group for local authority bridge
owners. CSS Bridges Group also co-ordinates local authority input into Highways Agency
Technical Project Boards. The Department for Transport (DfT) is embarking currently on
a programme of providing national advice to local authority engineers - including those
responsible for the bridge stock. This is particularly important as the HA Standards have
generally been focussed towards the high-speed motorway and trunk road network.
Normally, local authorities would follow Highways Agency Standards and Advice Notes,
particularly with regard to assessment and design Standards. However, few Standards
are written specifically with local roads in mind and fully justified Departures from
Standards may be necessary, particularly for geometric problems.

10
 1.5 Assessment and asset It has been acknowledged by the Highways Agency (HA) that the assessment of the
management bridge stock is likely t o become a ‘continuous’ process. This is so that bridge
maintenance strategies can become safety related and based on whole life performance
techniques rather than solely based on condition. It is interesting t o note that Network
Rail has a domestic assessment programme based o n an 1 8 year cycle. This Section is
based substantially on a paper entitled Whole life Performance BasedAssessment Rules -
’.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

BackgroundandPrinciples’ It provides useful background although some of the details

have changed since publication of the paper.

1.5.1 Whole life assessment The procedures proposed by the Highways Agency are intended t o ensure that present
levels of reliability o f existing bridge stock should be maintained at levels considered t o
be socially and economically acceptable, without the economic consequences of undue
conservatism. To provide marginally excessive reserves of safety in a design adds little t o
first cost, but the provision of unnecessary margins in assessing existing bridges for
increased loading or deterioration can be disproportionately expensive.

For ease of use in conjunction with current practice it was proposed t o relate reliability
t o the load capacity factor, K,on HA loading as defined in BD21, which a structure in
service may be assessed t o achieve. That factor takes into account any deterioration in
structural resistance. ‘Profiles’ of diminution of ‘available’ K factors with time may be
derived from surveys of representative bridges and empirical predictions of deterioration.

The procedure proposed by Flint and Das’ was developed into a draft Highways Agency
Advice Note BA81 entitled Whole life assessment of highway bridges andstructures. This
was trialled by a number of HA Maintaining Agents during 2000.

The requirement was t o determine the performance level of the structure at the time of
assessment, I,, and compare this t o the Critical Performance level, I,, ,. The present
assessment process (BD21) is such that I , has t o be greater than I,, , t o achieve an
assessment ‘pass’. If I , is less than I,, , then the structure has t o be managed in
accordance with the advice given in BA79 (see Figure 1.I).

The next stage in the development of any asset management plan is t o attempt t o
determine when structures, where I , is greater than I,,,,, will reach a Critical Performance
level. In order t o achieve this goal i t is necessary t o have an understanding of the
deterioration rate. Unfortunately, published deterioration models need t o be treated
with care and considerable judgement is required in their use.

1.5.2 Asset management If I , and I,,,, can be reliably determined, together with a reasonable approximation of
future deterioration, then it is possible t o evaluate various maintenance options using
whole life costing techniques (see Figure 1.2). If the maintenance options for individual
structures are combined in a comprehensive bridge management system then it is
possible t o determine a strategy t o maximise the overall value of the asset.

11
 t No. of years from construction
I, Performance level a t time of construction
I, Performance level of the structure being assessed a t time t
lCri
Critical performance level
Whole life

- -
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

- ... performance profile

shown dashed

\
\
- - - - - - - - - - > - - -\ -
\

0 t
No. of years from construction

Structure (a)
Option Maintenancestrategy
1 strengthen the structure now
2 weight restrict structure/implement interim measures and monitor

Structure (b)
Option Maintenancestrategy
1 strengthen the structure now
2 undertake minor repairs to abate deteriorationhonitor the structure
3 do minimum work now and replace at reduced life t,

0 t, t2
No. of years from construction

The simplistic annual review of maintenance budgets is now outdated. Bridge owners
need to be able to demonstrate the effects of various levels of funding on the value of
the asset. The CSS produced a useful paper’ on recommended levels of maintenance
spending and this approach is probably satisfactory for owners of small numbers of
bridges. However, a substantial amount of work is now under way to link the production

12
 of performance indicators, asset management plans and asset valuation models, and
although the discussion of these topics is beyond the scope of this Guide, i t is clear that
there will be a substantial role for effective bridge assessment techniques t o play in the
overall process.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

1.5.3COnSeqUenCeS for If comprehensive asset management plans are t o be developed based on whole life
assessment costing techniques and asset management plans, then it will be necessary t o
accurately determine I,, I,,,, and deterioration rates. Use of this Guide should assist in
this process.

1.5.4A COntinUOUS Currently, the Highways Agency is considering proposals t o re-assess a proportion of its
a s s e s s m e n t programme bridge stock every year t o ascertain their adequacy t o support imposed loads. Such
reviews would be undertaken when significant events occur that could increase the
imposed loads above those previously assessed and/or reduce the load-bearing capacity
of structures. Typically, a re-assessment every 12 years might be anticipated. It will be
for other users t o determine whether t o follow a similar approach.

1.6 Assessment of the An alternative approach t o the notion of a continuous assessment programme is the
serviceability limit state development of procedures for assessment at the serviceability limit state (sls).
Although currently some checks are required at the sls, particularly when plastic
methods of analysis are used (see Section 6.7),the UK approach t o assessment has
safety and the ultimate limit state (uls) capacity of a structure as its focus. Proposed
by Lark and Mawson’ 5! the aim of assessment at the sls would be t o create a more
positive link between assessment procedures and the sls criteria that drive
maintenance decisions, and t o provide a framework for recording and incorporating
the information that has allowed engineers in the past t o assess a structure on the
basis of engineering judgement. Although this concept is closer t o what is currently
continental practice, there is evidence that the latter is moving more towards the
contemporary UK approach of ultimate checks and so the demand for a serviceability
approach has yet t o be proven. Nevertheless, it is also recognised in the UK that many
of the structures that have been assessed as sub-standard have been so classified,
n o t because of deterioration but because of the enhanced requirements of the
technical standards. The response of Collins16 of the National Assembly of Wales has
been therefore t o call for the development of asset management policies based on
the whole life cost of maintaining the function of highway structures rather than just
their strength, an approach that is more compatible with the concept of assessment
at the sls. However, if such an approach is t o be used in the UK, further work is required
t o support its general implementation and therefore this Section is provided for
information rather than direct guidance, as is the case in Chapters 2 t o 12.

13
 1.6.1 Serviceability The justification for proposing an alternative approach t o assessment based on
assessment serviceability limit state criteria is that certification of a structure today is no guarantee
of its performance tomorrow. To address this, what has been suggested is a procedure in
which the everyday response of a structure is monitored and then projected forward,
and a ‘change management’ procedure is adopted whereby it is the variation of the
reliability of the in-service structural and environmental response of the structure with
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

time that is of interest.

It has been shown that the reliability of a structure with respect t o an environmental
serviceability limit state, such as the onset of corrosion, can be expressed as a reliability
index, the time-dependent variation of which can then be derived by assuming a
deterioration model. This, typically, is a function of the chloride levels in the concrete’ ’.
The latter is a measurable parameter. Therefore by monitoring chloride levels, the
likelihood of corrosion being a problem can be anticipated and, by comparing the
development of this likelihood with time, with that predicted on the basis of the design
parameters, a rational decision on the need for preventative action can be made.

Likewise, but not yet explored t o the same extent, it is also possible to calculate the
reliability of a bridge with respect t o structural serviceability limit states. The variation of
this reliability with either the applied loads or the material resistance can be determined
therefore, and if these in turn are related t o time, (for example due t o changes in
loading regimes and the perceived likelihood of extreme loading events, or t o changes in
material resistance due to the effects of creep, shrinkage and/or deterioration, etc.), then
the variation of this reliability with either these changes themselves or with time can
also be examined. As with the environmental example given above, the significant
benefit of adopting a sls approach is that the analytical procedure and limit state models
adopted in this procedure can be verified by monitoring the response of the structure.
Then, if found t o be inappropriate, they can be modified t o reflect the actual behaviour
of the structure and, when satisfactory, can be updated in response t o the past
performance of the structure. In other words, the actual service of the structure can be
used as a ‘proof’ load and by monitoring the structure’s behaviour the uncertainties
associated with its assessment can be continually refined and reduced.

The challenge of this approach will be t o identify appropriate criteria that can express
capacity a t the sls. This is not something that will be easy to do with confidence, because
existing definitions of sls behaviour are generally stiffness related whereas uls behaviour is
strength related and currently there are no uls criteria for effects such as those due t o
shear. Nevertheless, what is being proposed is a procedure which seeks to make use of an
understanding of the reliability of bridge structures applied in such a way that it can
support bridge managers’ engineering judgement of the response of their stock, and can be
enhanced by both traditional and future bridge monitoring techniques. For example, strain,
deflection and/or the nature of cracking patterns may be used currently to characterise the
response of a structure to both loads and movements from which its structural reliability
can be identified. Likewise chloride levels, moisture content, porosity and degree of
corrosion, etc. are measures of environmental reliability. In the future therefore, it can
easily be envisaged that defect, vibrational and other NDT monitoring techniques may also
provide suitable methods for gauging the ‘in-service’ reliability of a structure.

14
 i

Whichever ‘measure’ is adopted, it is the variation of the resulting reliability from that
anticipated a t the time of design which is significant; with improvements representing
an enhanced understanding of the structure’s behaviour and reductions typically being
due to changes in loading and deterioration, although they can also be due to changes
in understanding of the performance of the structure. What matters is that these
variations in reliability are managed. Thus, if the in-service reliability of a structure is
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

reduced because of deterioration or an increase in loading, the question that must be

answered is can it be restored through a knowledge of the actual behaviour of the
bridge or by tighter control of the traffic using it, or are remedial works required.
Unlike assessment a t the uls, the adequacy of the response to this question can be
monitored and the bridge manager can be assured that a margin of safety against total
collapse still exists. The suggestion is that this provides an opportunity to manage the
risk posed by the structure and, in each period, to address the changes which have
occurred in both the response and loading of the structure, whether due to traffic,
environmental or deterioration effects; to assess the consequences of these changes
and, on the basis of this assessment, to implement the necessary monitoring, controls
or remedial action.

Further research is required to identify exactly how the above will be achieved and the
data that is needed to implement it.

1.7 Historical aspects O f This Section is a summary of a paper entitled A review ofbridge assessments through
assessment time, presented by Barry Mawson, Gwent Consultancy, to the South Wales branch of the
Institution of Highways and Transportation on 10 February 2000, and has been
reproduced by kind permission of the author.

To review developments in bridge assessment it is necessary to consider the history o

vehicular loading and the link with design standards. The French appear to have been the
first to produce a design loading code for highway bridges in about 1850. In 1894 the
Dutch produced a similar loading standard consisting of two alternative carts, a
lightweight and a heavyweight cart, both drawn by a single heavy horse. Nineteenth
century bridge designers in the UK were still allowed some freedom. A uniformly
distributed load of Icwtlft’ together with the local effect of a 10 tonne wheel load was
commonly applied. A typical British bridge design towards the end of the 19th century
would accommodate a single 30 tonne vehicle on four wheels.

In 1922 the Ministry of Transport produced its first standard load train for highway
bridges. It consisted of an engine pulling three heavy trailers. In 1931 the Ministry of
Transport devised a standard loading for highway bridges, referred to as the equivalent
loading curve. For any particular loaded length, this curve represented an equivalent
uniformly distributed load, which, when combined with a specific point load located
within the loaded length, would produce the same bending moment and shear force as
the original loading train. This simplified quick analysis of smaller and standard bridges in
the design process.

15
 The assessment process

Axle weights and gross vehicle weights continued to increase and by 1973 a 10 tonne
axle and 32 tonne maximum gross vehicle weight was permitted. The Ministry of
Transport published a loading standard BE1/77, which was in parallel with the revised
loading standard B S I 53. The loading consisted of two types, HA loading which was
similar to the earlier equivalent loading curve, and an additional HB loading model to
represent abnormal vehicles.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Overloading of vehicles tended to increase, traffic densities increased, and in May 1983
gross vehicle weights were permitted under ‘Construction and Use Regulations’ to
increase from 32.5 tonnes to 38 tonnes provided they were on 5 or 6 axles. Then EC
Directive 85/3 required a minimum standard to be allowed in all EC countries, i.e. 40
tonnes on 5 axles, and that axle weights should be permitted to a level of 11.5 tonnes.
A derogation allowed the UK to delay implementation until 1 January 1999.

In 1984 the Department of Transport embarked on a 15 year programme for

rehabilitation of highway structures. The assessment programme was organised in three
stages. Stage 1 covered older short span bridges which were not designed for, or known
to be designed for, present day loading requirements, i.e. pre-1922. Stage 2 of the
programme covered modern short span bridges designed prior to changes in design rules
introduced in 1973, notably for shear in concrete. Stage 3 included long span bridges
where increase in long length loading was important, due to the effects of a larger HCV
proportion in traffic queues.

The history of bridge assessment goes back well into the 19th century for railway
bridges. The London North Western Railway commenced a programme of assessment in
the 1880s. This was followed by a more general assessment of railway bridges in the
1920s and led to the Bridge Stress Committee during the period 1923 to 1928, which
specifically covered the dynamic effects on railway bridges.

The need to assess road bridges was first properly considered in the 1960s with the
move towards the 32 tonne vehicle. This was implemented with the Assessment
Standard BE4 and was aimed a t the pre-1922 bridge stock. The further increase in lorry
weight to 38t was again covered in 1984 by the Department of Transport Standard
BD21/84. This was amended with the introduction of the potential for the 40t lorry in
1987 and consolidated into the standard in 1993 (BD21/93 - now BD21/01).

A considerable number of highway bridges were in the ownership of the railways and
their duty was limited to a design capacity of ‘the traffic of the day’. This was defined
clearly in the Statutory Instrument 1705 in 1972 which set the railway’s loading
obligation as that defined by assessment standard BE4. No allowance was made for the
reference in BE4 to the fact that it covered bridges:

0 Not designed to HA loading and so considered as sub-standard and having a limited

life.
0 That their replacement should not be unduly delayed.
0 That structures should be critically examined a t such intervals as their conditions
require but a t least once every three years.

16
 The standard loading curves used in assessment had been developed from the design
standard and allowed for the likelihood of impact, bunching and potential overload of
vehicles. Research showed that this was unduly conservative for assessment and the
introduction of assessment standard BD21/97 gave reductions of up to 35% of load effect.

Following the new vehicular standards of January 1999, modifications required to the
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

assessment standard were introduced to cover the weight limits of 18 tonnes to 26

tonnes, which allow a 11.5 tonne axle. Further changes may be necessary if the general
use of 44 tonne, or heavier, vehicles is permitted.

1.8 European It is beneficial to study the process of bridge assessments overseas, particularly in Europe.
develop ment s A European research project, known as Bridge Management in Europe (BRIME), has been
undertaken to develop an outline framework for a bridge management system for the
European road network. This Section summarises five papers, which were presented at
the Fourth International Conference on Bridge Management a t the University of Surrey in
April 2000. The original papers are detailed in the References below.

1.8.1 The BRIME project The project was co-ordinated by the Transport Research Laboratory in the UK in
collaboration with five European partners. The project was divided into eight work
packages as follows:

1. Classifying the condition of a structure.

2. Assessing the load carrying capacity of existing bridges, including the use of risk-
based methods.
3. Modelling of deteriorated structures and affect of deterioration on load carrying
capacity.
4. Modelling of deterioration rates.
5. Deciding whether a sub-standard or deteriorated structure should be repaired,
strengthened or replaced.
6. Prioritising bridges in terms of their need or repair, rehabilitation or improvement.
7. Reviewing systems for bridge management and development of a framework for a
bridge management system.
8. Project co-ordination.

The deliverables from the BRIME project may be viewed on the TRL website, a t
http://www.trl.co.uk/brime.This Section summarises the contents of work packages 1
to 5 1 8 ' 0 ' 1 2

1.8.2 Structural assessment In most of the partner countries, assessment is only carried out on specific structures
when there is a need to carry an exceptional load or where the bridge has been
subjected to deterioration, mechanical damage, repair or change of use. The UK is

17
 unusual in that it has carried out a comprehensive programme of assessment where
bridges have been built t o outmoded design standards and have not been assessed t o
current standards.

The rules used in bridge assessment are provided mainly by design standards with
additional standards relating t o testing methods including load testing. The UK is
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

unusual in that it has a comprehensive set of assessment standards. Generally, limit

state methods are used, although Germany still uses allowable stress techniques.

1.8.3 Deterioration The current practice is for all of the partner countries t o take account of deterioration in
some way. Except for the UK, taking account of deterioration is, in general, carried out
on an adhoc basis and depends on the knowledge and experience of the assessment
engineer. The UK documentation tends t o be general in nature and contains little
quantitative guidance. At present there are few procedures available for taking
deterioration into account in a structural assessment.

1.8.4 Chloride ingress In order t o estimate the remaining service life of a bridge, it is necessary t o predict
deterioration rates once the existing condition of the bridge has been assessed. The
objective of this work package was t o consolidate and improve existing knowledge of
chloride penetration in concrete, which is the principle mechanism of deterioration.

The work package was not complete a t the time the paper was written. However the
initial conclusion was that the technique was not suitable for general use and that only
experienced engineers or corrosion experts should be allowed t o act upon the results
generated by a chloride ingress model. The authors stated that although chloride ingress
models could be used for determining future maintenance, they should not be used for
assessing structural capacity or deterioration.

I .9References 1.1 PARSONSBRINCKERHOFF, TechnicaiAuditoftbeappiication ofBA79 and A rewewofbridgeassessmentfailureson the

Motorway and TrunkRoadnetwork, Final Reports HBR80616 - Highways Agency Contract 2/419, Parsons Brinckerhoff,
December 2003
1.2 CONCRETE BRIDGE DEVELOPMENT GROUP, ibeassessment ofconcrete bridges, 1, CBDG, Camberley, 1997
1.3 FLINT, AR and DAS, PC. Whole life performance based assessment rules - Background and principles, SafetyofBndges
Conference, Thomas Telford, London, 1996.
1.4 CSS, FundingforBridge Maintenance, CSS Bridges Croup, February 2000.
1.5 LARK, RJ and MAWSON, BR, Assessment at the serviceability limit state. BridgeManagement 4, (Ryall, MJ, Parke, CAR
and Harding, JE, eds), Thomas Telford, 2000, pp 426-433.
1.6 COLLINS, TJA, "Highway Function" approach t o bridge maintenance and a more flexible approach t o design loading and
dimensional requirements, Surveyors'Annuai Bridge Conference, March 2002
1.7 CEHLEN, C and SCHIESSL, P, Probability-based durability design for the Western Scheldt Tunnel, Structura,'Concrete,
Vol P1, No 2, 2000.
References 1 8 to 1 12 below were all in BndgeManagementl (Ryall, MJ, Parke, CAR and Harding, JE, eds), published by
Thomas Telford Ltd in 2000.
1.8 WOODWARD, RJ, VASSIE, PR and CODART, MB, Bridge Management in Europe (BRIME) overwew of project and review
of bridge management systems.
1.9 BLANKROLL, A eta/, BRIME -Chloride ingress and bridge management
1.10 KASCHNER, R, HAARDT, P, CREMONA, C, CULLINCTON, D and DALY, AF, Bridge Management in Europe (BRIME)
structural assessment
1.11 DALY, AF, Bridge Management in Europe (BRIME). modelling of deteriorated structures
1.12 BERC, L eta/, Bridge Management in Europe (BRIME) condition assessment of bridge structures

18
 2. Technical approval and documentation

2.1 h l e r a l procedures This Chapter is intended t o introduce the Technical Approval procedure for the
assessment, design of strengthening and refurbishment of highway structures t o
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

engineers with little previous experience of the subject. It includes:

0 General principles of the procedure

0 An explanation of its objectives
0 A description of the key stages of the procedure
0 Flow charts and figures.

The Chapter is not intended t o give the full details of every aspect of the procedure. Nor
can i t cover in detail how i t is implemented by different Highway Authorities, and the
many different types of contract and procurement methods currently in use. However, i t
should give assessing and checking engineers basic information about the procedure
and, most importantly, explain the need t o involve the Technical Approval Authority
(TAA) right from the start of the assessment process. The details included refer
specifically t o the Highways Agency procedures. Local Authority and other bridge
owners (e.g. British Waterways and Network Rail) have similar Technical Approval
procedures using the principles given in this Section. One key difference is that normally
the TAA function is not separated from the in-house engineers.

2.2 General principles The Technical Approval procedure is a form of Quality Assurance for Highway Authorities
and the owners of highway structures used by the general public. The procedure was first
introduced into the UK following a number of bridge failures in the early 1970s. These
failures were blamed in part on the use of inappropriate design Standards and also because
designers of the time were adapting the Standards for use on structures that were outside
the limits for which they could be used safely. The procedure was introduced following an
official inquiry into these failures and has been modified and adapted over time.

The procedure is a mandatory requirement for all Department for Transport (DfT) funded
highway structural design work of any significance* '. This includes designs for new bridges
and other highway structures, bridge assessments and structural modifications,
refurbishments and repairs; the standard procedure for the assessment and strengthening
of highway structures and its relationship with the design approval procedure is illustrated
in Figure 2.1. The procedure is also applied t o any temporary structural works on the
highway that could affect the safety of the public. Some form of Technical Approval or
Appraisal Procedure is a requirement of all other Highway Authorities and rail structure
owners in the United Kingdom. The detailed requirements of the procedure in use by the
Highways Agency are set out in the standard BD2/02 (see Section 1.4).

Technical Approval of highway structures is an ongoing process, which starts at the

concept or feasibility stage. It continues through design, construction and on through the

19
 Design standards Assessment standards

Departures from standard to be used

in assessment agreed with TAA
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

E Assessment report
i
Non-compliance with standard
revealed in assessment

Departures to be retained in
the structure agreed with
Highway Authority

1 -I
I Strengthening/ refurbishment AIP a I
Departuresfrom Standard
(ULS & SLS) proposed for
I

Implement agreed action refer to New works to current design strengthening / refurbishment works
BA 79/98 Management of standards plus ULS departures in agreed with Highways Authority
Substandard Structures existing elements agreed with the
Highway Authority

I
design

1
1
Strengthening / refurbishment
-..
.. +-.

nt

Figure 2.1
Standardsfor assessment and strengthening of
highway structures.
- life of the structure, including steady-state assessment and any strengthening, repairs or
modifications. The process is only concluded finally with the demolition of the structure.
The various documents, which are a product of the process throughout the structure’s
life, form a very important part of the structure records and should be carefully retained.

2.3 Objectives The original, and still the primary, objective of the procedure is to ensure the safety of
highway structures by ensuring adequacy of structural design, assessment, construction
and of steady state maintenance. Recent CDM legislation and court findings have made
it even more important for Highway Authorities and other structure owners to maintain
a robust procedure for ensuring structural safety.

20
 The procedure for assessment has had to be developed and extended to incorporate
resulting strengthening, modification and refurbishment of the structures. Highway
Authorities and bridge owners have introduced additional objectives with the aim of
ensuring good value for money, minimising environmental impact, improving
sustainability and aesthetics. The recent introduction of new forms of contract for the
maintenance and procurement of highway works has led to the need to further modify
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

how the procedures are applied.

2.4 Key Stages The detailed application of the Technical Approval (TA) procedures for the assessment of
highway structures depends on the particular requirements of the Highway Authority or
structure owner. However, this does not alter the general principles of the procedure,
which can be adapted to apply, in most instances, to all forms of structure.

Technical Approval or TA should start a t the concept planning or feasibility stage of any
assessment, strengthening, modification or refurbishment work involving highway
structures. The earlier the assessment engineer involves the Technical Approval
Authority (TAA), usually the better.

The procedure requires that the TAA have a responsibility to ensure that only competent
and experienced engineers carry out the assessment of highway structures and the check
of the structural work.

The precise nature of the check and the relationship between assessor and checker is
dependent on the category of the structure. As soon as an assessor is designated or
appointed, it is usual for discussions to take place between them and the TAA on the
various assumptions to be made and options available for inspection and testing for the
assessment. The assessor will also agree the category of the structure with the TAA,
based on the complexity, size, value or risk of the project.

BD2 defines four categories of structure: 0, 1, 2 or 3

0 Category 0 structures are typically buried structures of less than 3 m clear span and
having more than 1 m cover, multi-cell buried structures where the cumulative span
is less than 5 m and having more than 1 m cover or single span simply supported
structures with a span less than 5 m.
0 Category 1 structures are those with a single simply supported span of less than 20 m
and having less than 25" skew or buried concrete box structures with a clear span of
less than 8 m.
0 Category 2 structures are those that are not otherwise defined as Category 0, 1
or 3.
0 Category 3 structures are typically complex structures, which require sophisticated
analysis or with features such as high structural redundancy, unconventional, novel or
esoteric design aspects, a span exceeding 50 m, skew exceeding 45", moveable
bridges, bridges with suspension systems, post-tensioned concrete structures with
internal grouted tendons.

21
 2

However, BD2 does note that the category boundaries are not rigid. If in doubt, the
assessor is required t o make the decision in consultation with the TAA having regard t o
the wider issues of the value of the structure and the potential consequences of failure.

In the case of Category 3 structures, an independent consultant is required t o carry out

a check of the assessment. For Category 2 structures, this check may be carried out by a
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

checking team, which may be from the same organisation but must be independent of
the assessment team, while for Category 0 and 1 structures, although an independent
check is still required, it may simply be carried out by another engineer from the original
assessment team.

The procedure for obtaining Technical Approval for an assessment is given in figure 2.2
and can be described as follows. The assessor completes an Approval in Principle
document, which is generally called an AIP, or Technical Appraisal Form (TAF) depending
on the procurement route being adopted, the latter typically being more used on design
and build or related projects. This sets out the technical proposals for the assessment or
design of strengthening and includes all the structure geometry, ground conditions and
nature of the existing foundations. It also includes assumptions on material properties or
proposed testing, methods of analysis and assessment loading criteria, etc.

Any non-compliances or ‘Departures from Standard’ or ‘Aspects not covered by

Standards’, which are being proposed at this stage should be identified and recorded on
the AIP. Advice should be sought from the Highways Agency when considering
Departures from their Standards or Codes of Practice. Usually, it will be necessary for the
assessor t o discuss the proposed ‘Departures’ with the Highway Authority designated
representative and/or TAA, who normally are required t o accept any such proposals. A
description of the usual procedure for agreeing ‘Departures’ and ‘Aspects not covered by
Standards’ is given below. A list of relevant approved Standards that are t o be used in
the assessment is attached in the form of the Technical Approval Schedule (TAS). It is
important to note that the AIP or TAF is a ‘living’ document subject t o revision and
amendment during the assessment or design process. These are important record
documents, which need t o be retained for future reference.

The signed AIP is submitted t o the TAA for acceptance and to the checker for comment
and agreement. The significance of a personal signature by the individual design or
assessment team leader is to ensure ‘ownership’ and responsibility for the accuracy of
the information in the AIP. Acceptance by the TAA will be dependent on their
confidence in the assumptions, the method of analysis and material properties t o be
adopted in the assessment or design. Usually, the TAA will be required t o seek approval
for any ‘Departures’ from the Highway Authority’s designated representative if distinct
from the TAA a t this stage.

The detailed calculations required for the assessment or the design of the structure can
then start, followed shortly by the check. At any stage the assessor/designer, with the
agreement of the checker, may consider that further ‘Departures from Standards’ could
safely be applied t o the particular structure t o improve the assessment. The introduction
of any material changes or the introduction of further ‘Departures’ would require an

22
 -----

Figure 2.2
Technical approval procedure for assessment.
4- Assessment AIP submitted toTAA

I0

I Proposed departures from standards* 1

agreed by Highway Authority’s
designated representative? Initial or
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

higher level of assessment agreed?

I I

L TAA sign AIP/Addendum Addendum t o AIP

identify departures
and/or higher
level of assessment

L Carry out assessment and check

Amendment t o AIP identified?

tyes

Assessment required at next level?

no

Acceptable non-compliances with

standards revealed by assessment?

Assessment Assessment and check certificates with reference

and check t o assessment report (where non-compliances with
certificates standards revealed by the assessment are recorded)

+ +
Accepted criteria for departures from standard* entered in the AIP with an addendum
endorsed by the TAA and the Highway Authority’s designated representative (departures

+
and endorsement could be recorded on certificates i f required by the Highway Authority)

I
Assessment and check certificates accepted by TAA

TAA return a facsimile of the assessment and check certificates and retain the originals

and aspects
not covered
by standards

addendum to the AIP, recording all the changes to the original. This applies equally to
any significant changes that might occur during the construction of strengthening works,
etc. The approval procedure for strengthening and improvement works is similar to that
above and is detailed in Figure 2.3. It is quite common for it to become necessary for a
strengthening design to be modified during construction. This can result from a number
of causes; the most common being unforeseen conditions necessitating a change to the

23
 Non-compliance with standards* revealed
by assessment of the existing structure?
1

.
-
Engineer considers each non-compliance
with standards* and proposes whether
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

it can be retained in the structure,

or upgrading (to current design
standards) t o be carried out

no I
I I Non-compliance with standards* revealed
by the assessment t o be retained in the
I
structure agreed by the Highway Authority’s
1 designated representative?
I
4- Strengthening / improvement AIP

I I
Proposed departures from standards*
t o be used in the design agreed by
the Highway Authority

TAA sign AIP

Carry out the design and check

Design and check certificates

Accepted criteria for departures from standards* entered on the assessment and check
certificates and endorsed by the Highway Authority’s designated representative

I
Assessment and check certificates accepted by the TAA

I
TAA return a facsimile of the assessment and check
certificates and retain the originals for their records

24
 proposed strengthening details. It is most important that any changes affecting the
structure are subjected t o the same rigorous examination of the TA procedure as the
original proposals.

The final part of this stage of the TA procedure is the submission by the assessor/
designer and checker of the design and check certificates and their acceptance by the TAA.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

In the case of an assessment the assessor would normally submit a detailed Assessment
Report, which will include a comprehensive extract of the Inspection Report and any test
results. The Report will contain full details of the assessment findings and the assessed
capacity of the structure. The assessment certificate and assessment check certificate
should state the assessed capacity of the structure. As with the AIP and reports, these are
important documents and need t o be carefully retained with the structure records.

2.5 'Departures' and It has been suggested that the application of the TA procedure will inevitably stifle
'Aspects not covered by innovation. This can happen with the rigid application or misinterpretation of
assessment Standards. In contrast, careful application of the procedures by experienced
Standards' bridge engineers has been demonstrated t o encourage innovation by the use of
'Departures' or 'Aspects n o t covered by Standards'. The acceptance of non-compliances
t o Standards, which can be retained safely, has been shown t o avoid unnecessary
restrictions and expensive strengthening. It should be noted that most Highway
Authority bridge engineers, who are designated as or by the TAA, are involved with the
approval of the assessment of many structures. They are very well placed therefore t o
promulgate best practice and safely consider 'Departures from Standards'. The full effect
of any 'Departure from Standard' or 'Aspect not covered by Standards' is considered and
formally notified to, and has t o be accepted by, the TAA. This includes any 'Departures',
which can be justified by the use of higher levels of assessment. The procedure for the
management of standards is detailed in Figure 2.1.

The full effect of any 'Departure from Standard' or 'Aspect not covered by Standards'
has t o be considered by the TAA including any 'Departures' arising from the adoption of
different levels of assessment. The Highways Agency's Advice Note BA79/98 sets out
the philosophical approach t o assessment and the principle of different levels of
assessment. It identifies where 'Departures' are required for the higher, more complex
levels of assessment. The assessment process can be carried out in 5 distinct levels with
Level 1 being the simplest t o Level 5, the most sophisticated:

0 Level 1 assessments are carried out using simple analysis methods, in accordance with
the assessment Standard BD2l and the accompanying Advice Note BA1 6.
0 Level 2 assessments involve more refined analysis and the use of characteristic
strengths of materials.
0 Level 3 introduces the concept of Bridge Specific Assessment Live Loading and makes
use of material test results t o determine worst credible strength values.
0 Level 4 assessments include a reduction in partial safety factors if these can be
justified particularly where dead load or superimposed dead loads are high.
0 Level 5 assessments involve complex reliability analysis for particular structures.

25
 The relevant Highway Authority’s procedure for agreeing structural ‘Departures from
Standard’ and ‘Aspects not covered by Standards’ is an important part o f the TA
procedure but will vary between different authorities. It is usual for the procedure t o
require the assessor/designer t o recommend a departure and, with the agreement of the
checker, submit this with a justification t o the TAA. The TAA is required t o consider the
proposal and should seek advice from an appropriate specialist, who has a particular
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

expertise in the subject of the departure. Final approval and endorsement usually
requires a positive recommendation from both the TAA and the relevant specialist. In
the case of most Highway Authorities, the designated representative usually only
approves ‘Departures’ and ‘Aspects not covered by Standards’.

It is the responsibility of the TAA t o resolve any differences of opinion between the
assessor and checker and this includes the use o f specific ‘Departures’ or ‘Aspects not
covered by Standards’. This is particularly important in relation t o any ‘Departures from
Standard’, which may be accepted by the Highway Authority, as all parties are required
t o agree.

2.6 Reference 2.1 HIGHWAYS AGENCY, Highways structures Technical Approval, Easy Guide, Advice and Technical Appraisals Group, May
2002

26
 3. Inspection for assessment

3.1 Process and general

procedures
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

3.1 . I Background and Inspection is the first step generally in the assessment process, with the double objectives
introduction of establishing the condition and dimensions of the structural components and of verifying
the form of construction. The extent of the work will depend on what is already known
from archive and maintenance records, and therefore physical inspection is integrated with
desktop studies. It may also be necessary to carry out in-situ and laboratory testing to
establishkonfirm assessment parameters, and this is covered in Chapter 4.

Even a t this early stage of the process, it must be remembered that the purpose of
assessment is t o evaluate current and future structural capacity. Generally, this is done
via analytical methods (Chapter 6), while taking advantage of any hidden strengths
(Chapter 7). A major objective of inspection and testing is therefore t o determine input
parameters for this analysis, particularly with regard t o geometrical and mechanical
properties; where these properties are affected by different forms of deterioration, then
the guidance given in Chapters 3 and 4 is supplemented by that in Chapter 9.

For concrete bridges, the process and general procedures involved are those given in
relevant Standards and Advice Notes (see Section 1.4), with BD21, BA50, BD and BA44,
BD and BA63 being particularly relevant. Chapter 3 follows these procedures, although it
should also be recognised that the situation is undergoing continuous change, as
represented by the references at the end of this Chapter.

This Chapter therefore deals with inspection for assessment, the management of the
inspection process and the application of inspection results t o both one-off and
continuous assessments. Section 3.2 covers condition assessment and specific test
methods and techniques are discussed in more detail in Chapter 4.

The concept of inspection, or in an ongoing context, condition monitoring, is to make best

use of past, present and future performance-based data and observations to verify the
current integrity, and more ambitiously t o predict the future performance, of a structure.

3.1.2 Requirements for BD21 states that inspection for assessment:

inspection
. . is necessary to verfy the form of construction, the dimensions of the structure and the
‘I.

nature and condition of the structural components. Inspection should cover not only the
condition of individual components but also the condition of the structure as an entity and
especially noting any signs of distress andits cause.”

The standard also gives specific requirements for Inspection for Loading and for
Inspection for Resistance.

27
 BD2l refers t o Volume 3, Section 1 of the Design Manua[forRoadsandBridges, which
essentially means BD63 and BA63 for concrete bridges. BD63 defines the following four
main categories of inspection:

1. Superficial involving a cursory check for obvious deficiencies.

2 . General involving a visual inspection of representative parts of the structure t o
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

ascertain its condition and note any items requiring special attention.
3. Principal involving a close examination of all parts of the bridge at six-yearly, or
exceptionally up to ten-yearly, intervals. Originally, Principal Inspections solely comprised
visual examinations, but more recently limited testing has also been included.
4. Special Inspections involving a close inspection or testing of a particular area or
defect. They are carried out for specific reasons such as: following up on a defect
identified during an earlier inspection; checking for any effects caused by particular
events (e.g. flooding, vehicle impact, etc.); and before and after the passage of
abnormal loads.

BD2l notes that General Inspections are unlikely t o be adequate for assessment
purposes and also states that “all constituent parts ofthe superstructure shall be inspected
to determine their respective strengths”. As a consequence, inspections for assessment are
normally carried out in conjunction with Principal Inspections, but may also involve
Special Inspections.

With regard t o substructures, foundations, retaining walls and wing walls, BD2l requires
that all accessible parts be examined and any defects noted. This should also take into
account any evidence of ground movements around foundations and behind abutments
and wing walls. Underwater inspection of any submerged parts of structures should also
be carried out.

Special Inspections are also required for post-tensioned concrete bridges. Unusually
these involve more invasive inspection techniques, the main aim of which is t o identify
corrosion or the propensity for corrosion of the prestressing tendons. Details for the
planning, organisation and methods for carrying out such inspections are given in
BASOl93.

3.1.3 Future requirements for The Highways Agency proposed the introduction of new procedures for bridge
inspection inspections, aimed at those aspects of a structure that are most relevant t o its load
carrying capacity and durability. The objective of these proposals was to make more
effective use of testing, monitoring and the growing range of improved non-destructive
testing techniques, and t o improve the quality and consistency of both the inspections
themselves and the way in which they were reported. This was t o be done by combining
regular visual inspections of the whole structure with a programme of more detailed
investigations concentrating on known or suspected areas of deterioration. For each
structure, a unique schedule was t o be developed that was specifically designed t o
provide the information needed t o assess the structure adequately. However, these
proposals have not received universal support and discussions are currently ongoing as
t o how t o ensure that this information is acquired.

28
 To provide a starting point for assessment, all structures should be subject t o a
detailed inspection, the purpose of which should be t o establish precisely the condition
of each structure a t the time of its assessment and t o provide a reference against
which any future deterioration of the structure can be monitored. Such inspections
should record the location of all imperfections and defects on scale drawings
complemented with photographs and text. It is anticipated that regular inspections
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

should then be carried out a t programmed intervals, concentrating on the condition of

particular parts of the structure where deterioration has been found before or where it is
suspected.

3.1.4 The inSpeCtiOtl process The inspection process starts with the Client’s Brief and should include some or all of the
following activities:

0 Review archive and maintenance records for pertinent information (e.g.as-built

drawings, soils data, previous inspection reports, Health and Safety File, etc.).
0 A preliminary inspection t o determine access requirements and strategy, particularly
in respect of critical elements.
0 Organise traffic management, access equipment, inspection, testing and diving teams,
as appropriate.
U Carry out element and segmental inspections, recording main dimensions,
carriageway and soffit levels (if possible), confirm structural section sizes and
condition.
0 Record width, length, location, orientation and pattern of all cracking; size, depth and
location of spalled areas; size and location of areas of suspected delamination; size
and location of areas of scaling; size, location and description of deposits and staining,
especially rust staining; loss of section of any exposed reinforcement; and all other
defects and damage.
0 Excavate trial holes t o expose buried members where doubt exists with regard t o
detail, dimensions, condition, etc.
0 Carry out in-situ testing and sampling as required (see Section 4.3), implement
condition monitoring if appropriate.
0 Record all information relating t o the inspection on sketches, drawings and in
photographs, identify areas of specific concern and implement Special and Particular
Inspections.
0 Receive, assess and report laboratory and ongoing monitoring results.
0 Establish schedules for implementing and reporting condition monitoring, Particular
Inspections and ongoing assessments. Summarise all of the findings from the
inspection for assessment in a report which will form the basis for the assessment
itself.
U Consider a post-assessment inspection, concentrating on those areas where
cracking/defects would occur under the serviceability limit state (see Section 6.7).

29

~~~
 3 Onopeaion for assessment

3.1.5 Risk assessment and The main Health and Safety Issues to be considered in risk assessments for bridge
safety issues inspections are:

U Working a t height
0 Working in confined spaces
0 Working adjacent to traffic
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

0 Working with and adjacent to 'live' services

0 Working from a boat
0 Working on a railway
U Working a t night
0 Working near water (leptospirosis)
0 Use of mechanical access equipment
0 Use of small tools
0 Fire
0 Presence of asbestos3' - (movement joints, waterproofing, pipework, permanent
formwork, etc.)
U Lead paint.

This list is not exhaustive and risk assessments should be carried out for each and every
bridge site. For additional advice see HSE best practice In particular where
destructive testing is being used the specific and special risks associated with such
procedures must be considered.

3.2 Condition assessment The actual condition of a concrete structure a t the time of the inspection for assessment
could have a major influence on the assessed capacity of the structure and therefore
must be taken into account in that assessment.

3.2.1 BD21 approach BD2l states that the assessment resistance shall be determined from the calculated
resistance multiplied by an overall condition factor. It further states that the condition
factor, determined on the basis of engineering judgement, shall represent an estimate of
any deficiency in the integrity of the structure and may relate to a member, a part of a
structure or the structure as a whole. This approach is considered inappropriate for
concrete bridges.

3.2.2 CSS bridge condition A review carried out by the CSS Bridges Group concluded that to assist the maintenance
indicator management of a stock of bridges it is essential to have a 'Condition Indicator' which
can be used to determine whether the overall condition is deteriorating over time or
not. Their view is that such trends will indicate whether adequate funding is being
provided for structures' maintenance work. It was also recommended that all authorities
should use a single system in order to ensure consistency and credibility. Therefore, CSS
commissioned the development of two guidance documents:

30
 1. A guidance note on the inspection of highway bridges33.
2. A guidance note on the evaluation of a Bridge Condition Indicator based on element
condition data collected during these inspection^^^.

The CSS inspection system is based on a fixed list of 38 elements that occur on highway
bridges and a pro forma containing these elements and other data collection
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

requirements has been produced. If an element on the list is present on the bridge then
the inspector must record its condition. The condition is recorded as a combination of
the Severity and Extent of the defects on the element. The Severity scale ranges from 1
(no defect) t o 5 (failed) and the Extent scale from A (no extent) t o E (extensive). The
inspection document provides sound guidance on the classification of defects, with
tables and photographs of defects t o assist the inspector.

The severity/extent ratings provided by the inspections are used t o evaluate the Bridge
Condition Indicator (BCI) score. The BCI score is built up as follows:

1. The element severity/extent alphanumeric ratings are translated to an Element

Condition Score which has a scale of 1 (best condition) to 5 (worst condition).
2. The Element Condition Score is translated to an Element Condition Index, still on the
1 t o 5 scale, which takes into account the importance of the element. (N.B.All 38
elements on the CSS pro forma are classified as being of very high, high, medium or
low importance.)
3. The weighted average of all the Element Condition Indices on a bridge gives the
Bridge Condition Score, where 5 is the worst possible score and 1 is the best possible
score for a bridge.
4. The Bridge Condition Score is translated t o the BCI, the BCI is on a 0 (worst
condition) t o 100 (best condition) scale for ease of external reporting.
5. The weighted average of the BCls for a stock gives the Bridge Stock Condition Index,
retaining the 0 t o 100 scale. (N.B. The individual BCls are weighted by deck area.)
6. Guidance is provided on how t o interpret and present the BCI data.

The BCI does not represent the true functionality/safety requirements of a bridge, i.e. t o
safely carry the required service loads, and as such should not be used in isolation. The
BCI is just one component of a full set of Asset Management tools currently being
developed by the CSS and Highways Agency in partnership for highway bridges, which
includes Reliability and Availability Indicators and Asset Valuation procedures. The BCI
will, in the long term, be used in the Asset Management of highway structures for setting
targets and the prioritisation of maintenance work. However, in the interim it is being
used as a high level management tool t o monitor the trend of bridge condition over time.

3.3 Network Rail's Network Rail's Structures Condition Marking Index (SCM1)34was developed by external
Structures Condition consultants for Network Rail as a means of providing a more objective measure of
recording the condition of its structures. The scoring takes place during each detailed
Marking Index (SCMI) examination, which occurs every six years. The scoring system and associated algorithm
are currently in use for bridges but have also been developed for other structural assets,
such as tunnels and retaining walls.

31
 The system requires on the first visit the key components of the structure to be
identified and the defects associated with that component to be scored based upon
their extent and severity. At the following detailed examinations only the scoring of
defects is required. The extent and severity scores are material specific.

Each structure (or each span in multi-span structures) is scored out of 100, with 100
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

being perfect condition.

The average score for the whole population scored is published by the company in its
annual Business Plan and is a requirement of the Office of the Rail Regulator.

Extensive trials were undertaken with bridge engineers and examiners to validate the
output of the system to ensure that the score accurately reflects the condition of the
structure. It does not give a direct indication of serviceability as no account is taken of
the assessed loading capacity.

The system was not developed to manage directly either safety or work prioritisation,
although clearly condition influences decisions in both these areas. The scores are to
provide Network Rail with information on the condition of its bridge stock or other asset
type primarily nationally, regionally, by route and material type. This information can
then used to assist in determining national policies on the management of structures
and provide one of a number of measures input into tools used to manage work
prioritisation, future funding and route upgrade policies.

3.4 Other work The BRIME project35identified that in Europe generally current methods of condition
assessment use two different approaches for both the assessment of individual elements
and of the structure as a whole. The first uses a cumulative condition rating which is
derived from the condition of individual elements, while the second uses the condition
rating of the bridge element in the worst condition as the condition rating of the
structure itself. The condition assessment is based on bridge inspection and similar
procedures, although with different intervals between inspections, being used in
different countries.

An alternative procedure is to use damage classification systems36.More recently, this

approach has been developed as part of the CONTECVET project37and the findings
have been implemented by the British Cement Association as part of a structures
management package for the Hong Kong Buildings Department. The BRIME project also
investigated the use of damage categorisation models as a basis for an expert system to
assess damage and deterioration, although they concluded that:

‘I. . . further
research is needed to improve the methods developedfor categorising
damagedareas, , , this requires the development of a larger database and the addition of
new parameters. ’I

32
 3.5 Preferred approach Where an inspection for assessment of a concrete bridge highlights areas of
deterioration such as spalling, loss of reinforcing bar cross section, etc. then that
deterioration must be measured and recorded to enable it to be taken into account in
the assessment of individual elements. Procedures for doing this have been investigated
by WS at kin^^^ in work for the Highways Agency and by Webster3’ as part of the
CONTECVET project3’. Recommendations as to the most appropriate methods for
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

bridge assessment are given in Chapter 7.

Similar advice to that given above appears in a technical audit of BA79 undertaken by
Parsons Brinckerhoff310for the Highways Agency, and the latter are currently considering
these and other recommendations as draft amendments to BD44. It is intended that
these amendments will permit various forms of deterioration to be taken into account in
the condition assessment of individual elements.

Chapter 9 gives information on specific material factors that may need to be taken into
account.

3.6 References 3.1 HIGHWAYS AGENCY, Draft for Comment, Interim Advice Note on Asbestos Management Application to the Strategic
Road Network, (2004)
3.2 HEALTH & SAFETY EXECUTIVE, Bestpracticeguides
3.3 CSS, Bridge Condition Indicators Volume 1 CommissionReport, Volume 2 Guidance Note on Bridge Inspection Reporting
(Addendum No 1, May 2004), Volume 3 Guidance Note on EvaluationofBridge ConditionIndicators (Addendum No 1,
May 2004)
3.4 NETWORK RAIL Structures Condition Marking Index, Company code of Practice RTICEICI041
3.5 WOODWARD, RJ eta/, Bridge Management in Europe (BRIME), Deliverable D14 Final Report, Contract N o ’ RO-97-
SC.2220, BRIME. March 2001.
3.6 SODERQUVIST,The Finnish practice and experience regarding bridge inspection and management, The managementof
Highwaystructures,Highways Agency, 22-23 June 1998.
3.7 BRITISH CEMENT ASSOCIATION et a / , A vabdatedusers manualforassessing therexidualservicelife ofconcretestructures,
EC Project Ref. IN 309012, CONTECVET, BCA, Camberley, 2001
3.8 WS ATKINS, Deterioratedstructures- The assessmentofdeteriorated reinforced concrete bridge structures, Highways
Agency Final Report, November 2000.
3.9 WEBSTER. M. The assessment of corrosion-damagedconcretestructures, PhD Thesis. Birmingham University.July 2000.
3.10 PARSONS BRINCKERHOFF, TechnicalAudit ofthe applicationofBA79 and A review ofbridge assessmentfai/ures on the
Motorway and Trunk Roadnetwork,Final Reports HBR80616 - Highways Agency Contract 2/41 9, Parsons Brinckerhoff,
December 2003

33
 a

4. In-situ and laboratory testing

4.1 Background and Testing is closely associated with the Inspection process, in terms of
introduction establishing/confirming assessment parameters (Section 3,1), and in particular for
condition assessment, where some form of deterioration is involved (Section 3.2 and
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Chapter 9).

It is important t o have clear objectives in selecting test methods and in developing a

test programme t o ensure that these are met, with an acceptable level of confidence.
Precision of tests and sampling regimes are crucial in this regard, particularly for the key
stage of data interpretation.

1 Testing may be required for any of the following reasons:

1. To establish/confirm representative mechanical properties for the concrete,

reinforcement or prestressing tendons, for use in subsequent structural analysis.
2. To locate the position of reinforcement or prestressing tendons, and t o establish the
extent of any section loss or concrete spalling.
3. To identify possible deterioration mechanisms, and t o establish which one is
dominant.
4.To quantify relevant material properties and characteristics, relevant to current and
possible future deterioration, under the measured/assessed local micro-climate and
general environmental conditions.

In all of this, the purpose is t o augment and enhance the overall Inspection for
Assessment process, and t o produce data on the general condition of the structure
and/or t o serve as input for subsequent structural analysis.

The above objectives require the availability of a wide range of test methods, which
may either be non-destructive or intrusive, carried out on site or in the laboratory.
This is an evolving field, but details of current relevant test methods are given at the
end of this Chapter, together with some guidance on choice, and application in
practice.

This Chapter can only give a brief review, and this is done in the context of practice and
recommendations in relevant Highways Agency documents, most notably in BD21, BD
and BA44, BD and BA63, and the Bridge Inspection Guide.

4.2 Current testing practice The emphasis here is on practice in determining assessment parameters for subsequent
structural analysis in accordance with Highways Agency guidance documents, i.e. on
objectives 1 and 2 in Section 4.1

34
 L

4.3 In-SitU sampling and The extent of in-situ sampling and testing on any particular structure varies from
testing structure to structure, depending on the existence or otherwise of as-built and
maintenance records, the form of the structure and its general condition. Sampling and
intrusive investigations should be carried out in areas which will not impair the capacity
or the durability of the structure, and such sites should be made good using appropriate
materials.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

As a minimum, in-situ sampling and testing should include a cover survey, verified by
local breakout if possible. This is often carried out in conjunction with half cell potential
measurement, sampling for chloride ingress profiles, and depth of carbonation
measurement, all of which help to give an indication of the condition of the structure
and the potential for future problems. Occasionally, resistivity testing is also undertaken
to provide information regarding the likelihood of ongoing corrosion.

Further in-situ operations can include: exposure of an area of reinforcement to

determine size, type, spacing and condition; the recovery of cores; and the recovery of
steel reinforcement samples. In areas of concrete where delamination is suspected, e.g.
deck or beam soffits, a tapping or sounding survey with a light hammer is useful in
detecting potential spalling and loss of cover.

Information regarding in-situ stress measurement of concrete and steel is given later in
this Chapter.

’
Durability of concrete structures - Investigation, repair, protection4 contains information
on planning, test and inspection techniques and on the interpretation of condition
surveys. Further information is contained in CBDG Technical Guide 2 - Guide to testing
and monitoring the durability of concrete

4.4 Laboratory testing Laboratory tests may include the determination of chloride and cement content plus the
determination of estimated in-situ cube strength. Steel tensile testing may also be
carried out. Testing to identify possible deterioration mechanisms and related material
properties (objectives 1 and 4 in Section 4.1) involves significant laboratory testing.

4.5 ConCrete parameters The worst credible strength (WCS) for concrete in an element can be determined from a
for assessment minimum of three estimated in-situ cube strength results using a formula given in BA44,
where estimated in-situ cube strengths are determined from concrete core tests.

It is important to recognise that the BA44 formula was derived on the basis of the
variation in the estimate of concrete strengths from core test results at a particular
‘location’, within which the variation of concrete strength is small, and not across an
entire structure. The use of this formula to estimate concrete strengths for an entire
structure or even element must be approached therefore with caution since test results at
one location may well be unrepresentative of the material properties a t another.

35
 4 In-situ and laboratory testing

Non-destructivetechniques such as Schmidt or Rebound hammer may be used to

demonstrate that core locations are representative of the properties of wider regions of
a structure or to identify the need to take additional cores. Although Schmidt or
Rebound hammer tests can be calibrated against a core test they are no real substitute
for core tests in their own right when feasible.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

It is advisable that the value obtained from the BA44 formula is compared with values
obtained from other methods and an appropriate value selected by engineering
judgement. Table 4.1 illustrates four different approaches that might be considered,
with example values based upon an actual set of 5 core test results with estimated
in-situ cube strengths of 84.5, 86.0,97.0, 83.5 and 83.0N/mm2.The mean value of
this set of test results is 86.8N/mm2and the standard deviation 5.82N/mm2.This is
greater than the standard deviation assumed in deriving the BA44 formula and
illustrates that in this case it may be prudent to use a worst credible strength somewhat
lower than the value determined using the BA44 formula. In fact, it actually shows that
the cores do not all come from a ‘location’ within which the variation of concrete
strength is small. When larger numbers of cores are taken, it is common for some to be
below the WCS. Strictly, the BA44 formula gives an indication of the worst likely mean
strength, and is only valid if either the variation in concrete strength is small or the
behaviour is sensitive to mean concrete strength, rather than the strength in critical
local areas.

It is emphasised that the second approach below, based on using the BA44 formula in
conjunction with the lowest recorded core strength, is included only as an aid to
establishing a sensible concrete strength for elements when the number of cores taken is
limited. It might be appropriate when it was considered that only a particular core was
likely to be representative of the concrete in a critical area.

In practice, it is often necessary to use cores such as these, which do not come from one
‘location’, to obtain a single WCS for the whole structure or element. Strictly, the most
correct statistical approach is to use the T distribution which allows for uncertainty in
determining the standard deviation as well as the mean. However, when the number of
cores is small, it becomes necessary to use more judgemental approaches, considering
such aspects as the lowest core result or, when results are similar, the possibility that
this may be coincidence.

Table 4.1
Examplesof four different approaches.
IMethod
I Description
I Estimated
WCS (N/mm*
 4.6 Reinforcement The reinforcement parameters required for assessment are type, size, condition, spacing,
parameters for assessment location and yield strength. AS previously noted, type, size and condition are often best
determined by breaking open a small area of concrete for examination, whilst spacing
can be determined using a cover meter.

When it comes to yield strength, BA44 states that a minimum of three samples should
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

be taken for tensile testing. These samples should be of similar type and diameter and
should be taken from low stress areas. Steel samples taken from concrete cores are not
recommended for tensile testing.

Determining the worst credible strength from a set of three samples requires engineering
judgement. BA44 states that the formula for concrete strengths may also be used for
reinforcing steel. However, the characteristic strength of the same three samples can be
lower than the value calculated by the BA44 formula, depending on the spread of results,
and the formula therefore does not provide the worst credible strength. Statistical
analysis of a large number of samples tested by the TRL indicates that the worst credible
strength tends towards the mean strength with a sufficiently large population.

Indeed, testing reinforcing bars can show a sizeable variation and may also show
significantly greater yield strengths than the characteristic values specified during design.
It is noteworthy, however, that variations in reinforcement yield strengths within a
structure will essentially be averaged because the number of bars that collectively
contribute to the flexural strength of a section or yield-line. In such cases therefore, the
use of the WCS, which is actually an indication of the mean strength, is more correct
than in cases where behaviour is sensitive to the local strength of individual bars. In
dealing with variations in reinforcement strength, some comfort can also be drawn from
the fact that, typically, assessments are based on reinforcement yield strengths, whereas
before flexural col[apse occurs some strain hardening of the reinforcement will generally
occur, particularly for significantly under reinforced sections and mild steel bars.

4.7 In-SitU Stress There are two basic techniques that have been widely used: slot-stress and hole drilling.
~

measurement - concrete Both rely on measuring the strain relief around a hole which is cut into the concrete. The
semi-circular slot, typically about 150 to 300 mm diameter, is cut with a diamond saw
and the hole, generally about 75 to 100 mm diameter, with a diamond tipped corer.
Both techniques are semi-destructive and care has to be taken to ensure that the

In the case of slot-stress, the strain can be restored by jacking the slot apart. This jack
load can be related to the initial stress state. It is also possible to determine the initial
stress state by using results of finite element analysis backed up by laboratory testing
from the strain releases across and on either side of the

For hole drilling, then the relaxation in strain as the hole is progressively drilled can be
related to the initial strain and, hence, stress state by semi-empirical methods. These are
established and correlated by both finite element analysis and laboratory testing. As part

37
 4

of this work, a circular jack was also developed with can be used to determine the in-
plane elastic modulus of c o n ~ r e t e ~ ~ , ~ ~ .

Care should be taken to allow for the variation in concrete physical properties as well as
the internal stress state. The strains and stresses in a cross-section can be affected by
locked in stresses due to differential shrinkage and creep resulting from construction and
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

subsequent load history as well as transient temperature fluctuations. The technique has
been successfully applied to a number of prestressed concrete bridges. The field accuracy
of the technique is plus or minus 2N/mm2.

4.8 In-situ stress

measurement - steel
reinforcement

4.8.1 Blind hole drilling A similar principle 3 the slc stress technique can be used for measuring steel resses
and is sometimes referred to as blind hole drilling. This entails drilling a small hole in a
reinforcing bar or prestressingtendon and measuring the strain relief around the hole.
This can be calibrated to give the steel stress and hence force.

One of the problems with this technique is that reinforcing bars have built in residual
stress profiles which are caused by the rolling process. Prestressing wires also have
drawing residual stresses. This stress profile has to be known, hente the need to calibrate
against an identical or a t least similar bar -which would need to be taken from the
structure. Additional complications can arise from the stress concentrations caused by
the deformations on deformed bars although these can be ground off before strain
gauging and drilling. The accuracy is unlikely to be less than plus or minus 20N/mmZ
which is often of little use for reinforced structures. Tests carried out in the mid-1980s
by WS Atkins to explore this technique were not pursued for this reason.

4.8.2 Cut bar method The application is to expose a bar in the structure, strain gauge it,&$ then to cut it
(with minimum heat). The strain relaxation then gives the in-situ stress in the bar. The
bar is then reinstated with a full penetration butt weld which would re-tension the bar
due to weld shrinkage.

This technique was referred to in a paper presented in 199446in respect of a box culvert.
The significance for culverts is that the actual soil pressures are somewhat indeterminate
and the readings allow the capacity available for carrying live load to be estimated.

The drawbacks are that dead load stress levels can be low; the stress measured after
exposing the bar over sufficient length is the average between points where it is properly
anchored; actual stresses are dependent on the amount of tension contained by the
concrete. However, if the reinforcement stresses are predicted to be very high but there
are no signs of cracking to confirm this then using the technique could be justified.

38
 I

4.9 Geophysical techniques The most commonly used geophysical technique in the investigation of concrete bridges
is surface impulse radar. This technique can be used t o determine hidden detail in
abutments or other elements and can also be used t o detect the presence of
reinforcement adjacent to remote faces. However, specialist interpretation is required,
quite often coupled with trial holes or cores for calibration.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

4.10 References 4.1 MAYS,cc (ed.),Durabi/ityofConcreteStructures-Investigation, Repair, Protection. E&FN Spon, London, 1992
4.2 CONCRETE BRIDGE DEVELOPMENT CROUP, Guide to testing andmonitoring the durabiOtyofconcretestructures.
Technical Guide 2, The Concrete Society, Camberley, 2002
4.3 FORDER, S. Ca/ibrationofsaw cutting techniquefor insit ustress determination MEng Project Dissertation, University of
Surrey, 1992
4.4 MEHRKAR-ASL, 5, Direct measurement ofstresses in concretestructures,PhD Thesis, University of Surrey, 1988.
4.5 MEHRKAR-ASL, S. Concrete stress-relief coring theory and application, ProceedingsofFP Symposiumon Post-tensioned
Concrete Structures, London, 25-27 September 1996, pp 569-576.
4.6 CHALKLEY, C, Proof load testing - a recent proof load test, and its results, Proceedingsofa Surveyorconferenceon
Hidden Strength-Load Testingfor Bridge Assessment, London, February, 1994.

39
 D Loading

5. Loading

5.1 General PrinCipkS Loading rules are given in BD2l for short and medium span bridges and in BD50 for long
span bridges. These documents are fundamental to the assessment loads for all types of
bridge and reference should be made to them for the rules and explanatory notes. The
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

main purpose of this Chapter is to explain the principles and point to references that
provide further background details.

There have been several changes to loading rules in BD21 since it was first introduced,
one being to account for the increase in gross vehicle and axle loads. The most recent
changes allow the use of bridge specific live loading for short and medium span bridges,
where good surface condition and low traffic flow lead to less onerous requirements
(See Section 5.2).

Assessment loading is limited generally to the application of dead, superimposed dead

and Type HA live loads. (Dead loads are the weight of the structural (i.e. load bearing)
elements whereas superimposed dead loads are the weight of the non-structural
elements, such as surfacing, parapets, services and miscellaneous street furniture.) Type
HA live loading consists of a uniformly distributed load (UDL) together with a knife-edge
load (KEL), which can be determined as described in BD21. For assessment purposes, the
values specified by this Standard have to be factored to give the Assessment Live
Loading (ALL). Although BD2l only covers permanent and Type HA live loads, most
bridges that can carry the 40t ALL have also been checked for Type HB loading and,
therefore, should now also be checked for STGO loading (See Section 5.4).

5.1.1 Scope O f h e loading The Type HA (design) loading given in BD2l allows for the effects of 40 tonne vehicles
and includes a contingency margin for unforeseeable changes in traffic patterns. For
assessment, reduction factors are applied to this Type HA loading to give the various
ALL levels with no contingency provision. The 40 tonne ALL covers the effects of normal
vehicles (Authorised Weight Regulation) of up to 40 tonne gross vehicle weight
[including 41 tonne 6 axle lorries, 44 tonne 6 axle bimodal articulated lorries and draw
bar trailer combinations and 44 tonne 6 axle general haulage lorries. (Although the gross
weights of these vehicles exceed 40 tonnes, the load is distributed over six axles, rather
than the five axles of the 40 tonne vehicle, as well as a longer wheelbase. Thus, the load
effects generated by these vehicles are lower than those caused by the 40 tonne
vehicle.)] and 11.5 tonne axle weight.

For cases where structures are found to be incapable of carrying the full 40 tonne ALL,
loading criteria are given which correspond to specified limits on gross vehicle weights
Special loading criteria are also given for fire engines.

The Type HA UDL and KEL given in BD2l are generally only suitable for modelling
longitudinal load effects. Alternative loads are given for the effects of vehicles on trough

40
,

decks, short span masonry arches, decks with main members that span transversely
including skew slabs with significant transverse action, and buried concrete box
structures with cover greater than 0.6 m. A recent change requires longitudinal elements
to be checked using specific vehicles where there is a low capacity for transverse
distribution.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

5.1.2 Background to The Type HA ALL that is given in B D 2 l for short spans (2-50 m length) is derived from
assessment live loading an ultimate or extreme loading as opposed to a working load. This ultimate load was
derived from first principles and is based on the assumption that the worst credible load
that can reasonably be expected to occur in the lifetime of the bridge will be equivalent
to some multiple of Type HA loading. It has been shown that this extreme load has a
return period of 200,000 years or a 0.06% chance of occurring in 120 years. The values
of nominal HA loading that are given in BD2l were determined therefore by dividing the
extreme loading by 1.5.

I Four elements have been used to generate the extreme loads. They are:

1. Loading from Authorised Weight (AW) vehicles

2. Impact
3. Overloading
4. Lateral bunching.

Details of these are given in Annex C of BD21

The loading has been derived for a single lane only. It has been assumed that if two
adjacent lanes are loaded there is a reasonable chance that they will be equally loaded

The various factors that have been used in determining the loading are span dependent.
(The exception is impact for the single vehicle case.) The derived loading has taken
account of the possibility of convoys of eight or more HCVs, but, on more lightly
trafficked routes, the probability of having a bridge completely filled with heavy vehicles
is small. The loading is conservative therefore for lightly trafficked medium span
structures.

The impact factor that has been used was derived from measurements taken on
motorway overbridges, which were of modern construction and where the road surface
and bridge joints should have been in good condition. The overload factors were derived
from a sample survey of approximately 3500 vehicles and thus can be assumed to be
typical of what may occur at any time, or in any place in the country.

The loading that is derived using these procedures is considered therefore to be fairly
universal in its application and to reflect situations that can occur a t any bridge site.

41
 D Loading

5.2 Bridge Specific loading For short and medium spans, bridge specific live loading is derived by the application of
a load reduction factor (K) t o account for different traffic flows and road surfaces.
Bridge Specific Assessment Live Loading (BSALL) for long span bridges is based on
traffic surveys and is described in BD50. Specific guidelines for Trunk Roads are given
in BD2l and appropriate K factors can be determined from graphs [the K diagrams) for
loaded lengths between 2 and 50 m and six combinations of traffic flow and surface
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

condition.

If a bridge is found t o be inadequate at the 40 tonne load level, the permissible weight
restriction level is the highest capacity for which the K value in the appropriate K diagram
is less than the live load capacity factor C, derived using the available live load capacity.

For bridge specific reliability-based assessments a different loading model is required. For
the Highways Agency this has been developed in a project described by Coopers’, the
result of which is a probabilistic loading model which accounts for traffic flow, road
surface characteristics and the dynamic response of the bridge under consideration.

5.2.1 Bridge Specific For a site-specific reliability-based assessment, the load effect [bending moment or
probabilistic loading mode[ shear force) 5,, a t any section can be taken t o be a random variable expressed as:

S,, = BSLL x R,, x DAF

Where BSLL = the Basic Static Live Load effect [defined below)
R,, = a random multiplier t o model the uncertainty in the load effect due t o
the static weight of vehicles
DAF = a Dynamic Amplification Factor t o account for the dynamic amplification
of the static load effect due t o the impact of axles and the dynamic
response of the bridge.

The Basic Static Live Load effect at the critical section is determined by loading the
relevant influence surface with the following:

0 A uniformly distributed load (UDL) of 27 kN/m run across a width equal t o the
smaller of the notional lane width or 3 m,placed centrally within the lane.
0 Two axle loads, each 300 kN, with a 1.2 m spacing between axles, placed anywhere
within the lane t o maximise the load effect at the section considered.

This follows the geometrical configuration of the load model in Eurocode 1, Part 3
because it was felt that the tandem axle configuration provides geometrically realistic
force distributions on the structure.

Each axle consists of two wheels a t a 2 m track width placed centrally within the lane.
The same loading is used for both Lane 1 and Lane 2 but a reduced UDL of 7 kN/m run
and two axle loads of 100 kN each are used for the remaining lanes. The carriageway is
divided into notional lanes according t o BD21.

42
 The random multiplier R,, is assumed to have a Cumbel distribution, the parameters of
which are given separately for bending and shear in Cooper” and they depend on the
span and traffic flow rate. The maximum value of this extreme type distribution also
depends on the number of repetitions of a load event within the given reference time.

Bridge natural frequencies are considered to be closely correlated to the span length.
The Dynamic Amplification Factor (DAF) is assumed therefore to have a Normal
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

distribution, the parameters for which are based on span, the number of loaded lanes
and the pavement condition.

5.3 Highway Surfacing Significant research has taken place to enable Bridge Specific Loadings to be determined
effects (see Section 5.2). Application of this work first appeared in the 1997 version of BD21.
Not only were the notional lane widths modified from previous calculation methods but
the number of HCV movements and the road surface condition were also taken into
account. Three traffic flow categories (High, Medium and Low) and two road surface
categories (Good and Poor) were defined. This gave six possible combinations of effects,
e.g. Hp (High-Poor) and Mg (Medium-Good), and six separate graphs of Reduction
Factor, K, against Loaded Length, I ,were produced. The category Hp gave the worst
loading condition whereas the category Lg gave the least onerous loading condition. The
assessment engineer was left with the task of determining whether the road surfacing
was ‘Good’ or ‘Poor’ quality.

The three traffic flow categories are defined as follows in BD21:

1. High (H) Flow > 70

2. Medium (M) 70 > Flow > 7
3. Low (L) 7 > Flow

Flow is equal to the total annual two-way HCV traffic over the bridge divided by 8760
where HCV is defined as goods vehicles that are over 3.5 tonnes maximum permissible
gross vehicle weight.

The 1997 version of BD2l required road surface categories to be determined using
equipment described in TRL Report L R I 125”such as the High Speed Road Monitor
(HRM). The acceptance criteria quoted approximated to the mid threshold value for
roads with 50 mph speed limit or greater as given in Highway Standard HD2gS3.This
technique was reasonable for application to high speed roads which were regularly
checked using equipment such as HRM, but proved to be problematical for local roads
where the use of such specialised vehicles was rare and the criteria quoted were
inappropriate for roads with lower speeds and greater rates of change in gradient.

The 2001 version of BD2l simplified the process by allowing the subjective determination
of road surface quality. For instance, it is stated that “Motoways and trunk roads may
generally be consideredas ‘good’surface category if they are maintainedand repairedbefore
they deteriorate to ‘poor’surface”. The ‘poor’ classification is applied if, when driving a
vehicle over the bridge in free flow traffic condition, any of the following applies:

43
 5 Loading

0 Subsidence - vehicle bounces

0 Sub-base deterioration - vehicle pitches locally
U Surface deterioration -visual deterioration of surfacing or steps in expansion joints

Additionally, where practicable, the assessment should be confirmed by observation of

HGVs crossing the structure (a full description of the criteria is contained in BD21).
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

If a quantitative assessment of the road surface condition is carried out then it is now
permissible t o use the criteria in HD29 appropriate t o the speed of the road carried by
the bridge. ‘Good’ roads are those in categories 0 and 1 as given in HD29 whereas ‘Poor’
roads are those in categories 2 and 3. Measurements are required t o be taken between
points 20 m beyond the bridge. This increase over the previous figure of 5 m is because
road profile unevenness at bridge abutments is a major cause of significant dynamic
loading on bridge structuress4.

The results of these changes mean that for minor roads the maintenance of a ‘Good’
category may require maintenance at more frequent intervals than would be demanded
by consideration of the pavement alone. It may also be necessary t o enter into
agreements between the bridge owner and the highway maintenance authority t o
ensure that adequate standards of maintenance are achieved. Specialised inspection
requirements should be included on bridge files.

5.4 Road traffic and There are two main classifications of road traffic in the UK. These are:
abnormal loads
1. Normal vehicles: The large majority of vehicles using the highway network are
regarded as ‘normal’ traffic, which covers cars, light goods vehicles, rigid and
articulated vehicles and heavy goods vehicles up to a gross weight of 44 tonnes.
These vehicles comply with the Road Vehicles Construction and Use (C&U)
Regulations 1998 and Authorised Weight (AW) Regulations 1998.
2. Abnormal vehicles: These are vehicles, either empty or laden, which do not comply
with C&U Regulations. This non compliance with the Regulations could be because:

0 The load carried exceeded dimension limits

0 The vehicle and load exceeded weight limits
0 The vehicle had been designed for carrying outsized loads and/or for particular purposes.

These vehicles include mobile cranes, construction plant and low loaders carrying
exceptional industrial loads (e.g. electrical transformers, machine presses, etc.). These
vehicles and payloads are commonly referred t o as an Abnormal Indivisible Load (AIL)

There are two types of abnormal vehicles:

1 , Special Types General Order (STGO) vehicles: This group includes vehicles that do not
comply with the AW Regulations but comply with The Motor Vehicles (Authorisation
of Special Types) General Order (STGO Regulations). Under Article 18 of the STGO
Regulations, the maximum gross vehicle weight and maximum axle weight are 150
tonnes and 16.5 tonnes respectively.

44
 Loading 5

2. Special Order (SO) vehicles: This group includes vehicles that do not comply with the
AW Regulations or STCO Regulations and is covered by Section 44 of the 1988 Road
Traffic Act. SO vehicles have maximum axle weights of greater than 16.5 tonnes or
gross vehicle weights in excess of 150 tonnes.

Both types of Order work in conjunction with the Road Vehicles (Construction and Use)
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Regulations. The movement of an AIL is covered by the provisions of Section 44 of the

Road Traffic Act 1988 and is required to be notified to the appropriate authorities
(police, Highway Authorities and bridge owners).

Indivisible load movements that take place under the STCO do so under a number of
conditions. Some of the C&U Regulations apply and, depending on the size and weight
of the load, some special conditions. The STCO divides vehicles into three categories
depending on their weight. The chief criteria are summarised in Table 5.1.

If the Category 1 or 2 axle limit is exceeded, then the load must be moved under the
higher category 2 or 3. All vehicles must carry a 'Category Sign'.

In addition to the above, the STCO also makes special provision for 'Engineering Plant',
which may be self-propelled or towed. For example: a mobile crane or bulldozer,
which cannot meet some of the C&U Regulations because of the task it is designed
to do.

Any vehicle moving in excess of 150 tonne gross vehicle weight, 16.5 tonne axle weight,
5.0 m wide, or 27.4 m long (excluding the length of the tractor unit) requires the
permission of the Highways Agency. Two types of AIL require signed authorisation, as
shown in Table 5.2.

Table 5.1 Category Maximum Maximum Minimum Speed Other

'

I
Categoriesof Special Types General Order vehicles.
'9
gross weight axle weight number limits conditions
of combined of axles (mph)
vehicle

45

- -
 Dimensions and weights 5 m < width < 6.1 m > 150,000 kg
and or
< 150,000 kg >6.1 m wide
and or
<27.4 m long >27.4 m long

Required authorisation STGO Special order under Section 44 of the

Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

VSE - Form VR1 RTA 1988 from the AIL Section of the
VSE

5.4.1 Purposes O f highway The prime purpose of highway structure assessment is to verify that the structure is
structure assessment capable of carrying the intended road traffic safely. The secondary purpose is to
determine the carrying capacity or the critical pinch point of the route from assessment
of the structures on it, if all structures on that route were assessed.

5.4.2 Design and assessment Since 1973 (BE5) it has been a requirement to design ‘other public roads’, ‘Principal
roads’ and ‘Trunk Roads and motorways’ for 30 (120 tonne), 37.5 (150 tonne) and 45
(180 tonne) units of HB loading respectively.

These design load requirements are now embodied in the current UK highway loading
Standard BD37.This Standard defines Type HA and Type HB load models, which are
intended to allow for the loading effects of ‘normal’ and ‘abnormal’ vehicles
respectively. UK highway bridge assessment loadings for normal traffic are given in
BD2l. Until 2001 there was no assessment load model for abnormal vehicles, so the
Type HB load model in BD37 was also used for assessment.

The HB model was derived in the late 1940s. An assessment load model for AILS was -’

developed because the HB model was found to be unduly conservative when used in
assessment of short span structures. Furthermore, the HB model could not be readily
correlated to the current fleet of abnormal vehicles. Consequently, assessments in HB
units do not fully meet the prime purpose of assessment.

5.4.3 BD86 - Assessment of For short span structures with loaded lengths of less than 50 m, load assessment
highway structures for the models for abnormal vehicles have been developed based on the load effects from
actual STCO vehicle weights and configurations and traffic data. These models can be
effects Of Types used to assess the load effects from STCO vehicles more accurately than the HB load
General Order (STCO) and model in BD37.
Special Order (SO) Vehicles
In addition to meeting the purposes of AIL assessment, the BD8655load models enable
the following augmentations:

Attainment of higher load capacity ratings, particularly for structures with loaded
lengths of less than 10 m.
I I
0 Flexibility t o modify various factors such as dynamic amplification factor, overload
factor, etc. t o suit a specific structure.
Consistent levels of safety for structures of different spans and for different STCO and
SO vehicle movements.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

5.4.4 BD86 Load Models - Five load models (SV vehicles) were derived t o represent the load effects from the axle
sv Vehicles arrangements permitted under the STCO Regulations, which produce the most severe
load effects. The five SV vehicles are:

1. SV80 covering Category 2 STCO vehicles 3.0 m wide up t o 80 tonnes.

2. SVIOO covering Category 2 STCO vehicles 3.0 m wide up t o 100 tonnes.
3. SV150 covering Category 3 STCO vehicles 3.0 m wide up t o 150 tonnes.
4. SV-Train consisting of the SV150 STCO vehicle drawn by a 3.0 m wide tractor unit
(locomotive) .
5. SV TT covering military Tank Transporters 3.7 m wide up t o 105 tonnes.

The SV80, SVIOO and SV150 vehicles have axle spacings of 1.2 m, with a variable length
at the centre between two bogies of 1.2 m, 5.0 m or 9.0 m. The 1.2 m spacing reflects
the findings of the database studies and the STCO Regulations. The 5.0 m and 9.0 m
spacings are only used above specified loaded lengths for load effects with two or more
peaks in the influence line (e.g. for continuous span structures). Inevitably this may
require a large number of load cases t o be considered and therefore general guidance
has been given in BD86 t o assist in reducing the number.

The most significant difference between the SV load models and HB model is that the
latter tends t o be over conservative for spans less than about 10 m where the two
adjacent heavy axles tend t o dominate. This is because STCO vehicles have a greater
number of lighter axles, which tend t o spread the load more than HB loading.

5.4.5 BD86 - HB Conversion A number of simple conversion charts based on influence lines have been derived t o
Charts facilitate the conversion of SV rating t o HB units. These enable the methods contained
in BD86 for the assessment and management of STCO movements t o be used for most
bridges that have previously been assessed for Type HB loading. Examples of the use of
the conversion charts are given in the Standard, together with some limitations on
applicability.

5.4.6 Assessment O f Guidance is given in BD86 to allow abnormal load movements t o be managed where a
Abnormal Load Movements structure has already been assessed for SV vehicles, or where HB conversion charts may
be used. The first stage of the assessment for the particular notified STCO vehicle is a
using BD86 simple screening assessment. The screening compares the vehicle type, gross weight,
axle weight and axle spacing of the notified STCO vehicle with limits appropriate t o the

. I
 SV vehicles. The STCO vehicle may pass if its gross weight is less than the gross weight
of the appropriate SV vehicle multiplied by the corresponding Reserve Factors for the
load effects being considered. The original SV assessment would have given the Vehicle
Ratings and Reserve Factors for the structure. Prior t o a rigorous SV assessment, the HB
t o SV conversion charts may be used t o obtain Vehicle Ratings and Reserve Factors.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Note that the Vehicle Rating for a structure is defined as the most onerous SV vehicle
that can safely pass over the structure (i.e. the vehicle with the smallest Reserve Factor
greater than 1.O). Reserve Factor is defined as the factor on the assessment SV load
required t o reach the first failure.

When a STCO vehicle does not satisfy the axle weight and spacing characteristics of the
SV vehicles, a detailed assessment is carried out. The load effects caused by the STCO
vehicle are compared with those of the SV vehicle, multiplied by the Reserve Factors,
using one or more influence lines most appropriate for the bridge.

If load effects due t o the STCO vehicle still exceed those of the factored SV vehicle,
further refinements such as reductions in dynamic amplification factor and/or overload
factor and associated loading, can be made where justifiable. These reductions depend
on the extent to which the movement can be regulated with regard t o keeping the
structure clear of associated normal traffic, restricting the speed of the STCO vehicle
over structures and certification by the haulier of the load carried.

5.5 References 5.1 COOPER,DI, Development of short span bridge-specific assessment live loadlng In /nternationa/symposiumonsafetyof
bridges (Das, PC, ed.) Institution of Civil Engineers and the Highways Agency, London, 4-5 July 1996, Thomas Telford,
1997.
5.2 TRANSPORT RESEARCH LABORATORY, Measurementandassessment ofunevenness on malorroads. Report LR1125, TRL
Ltd, Crowthorne, 1984.
5.3 HIGHWAYS AGENCY, HD29 Structural Assessment Methods. Design ManuallorRoadsandBridges, Vo1.7, Section 3, The
Highways Agency, London. 2001
5.4 JORDAN, P, Unpublished TRL Report PR/CE/18/98, Comment on Departmental Standard BD21/97. RoadSurface
Categories, 1998
5.5 HIGHWAYS AGENCY, BD86. The Assessment of Highway Bridges for the Effects of Special Types General Order (STGO)
and Special Order (SO) Vehicles, Design ManualforRoads andBridges, The Highways Agency, London, 2003.

48
 6. Analysis for assessment

6.1 Assessment principles

6.1.I General PrinCipkS Reserves of strength can exist but remain unrecognised if full advantage is not taken of
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

the range of analytical methods available for the assessment of concrete bridges. As a
result, in some instances, assessments can be excessively conservative and adequate
structures may be condemned as unsafe.

Analytical methods appropriate for bridge assessment vary greatly in their complexity
and in their cost of application. Therefore, it is appropriate generally to begin an
assessment with a simple method and then to extend the analysis if a shortfall in
capacity is identified.

Details of specific analytical approaches are set out in this Chapter. Usually, it is the strength
and safety of a structure that is the principal concern when an assessment is undertaken.
The guidance provided is focussed therefore on assessment a t the ultimate limit state.

6.1.2 Analytical procedure Typically, the assessment of concrete bridges is undertaken using the following simple
procedure:

U Establish load effects, using an analytical method chosen by the assessing engineer.
0 Compare load effects with capacities, determined in accordance with BD44 and
’,
BA446 ‘.

Improvements in assessment ratings may be achieved therefore through judicious

selection of analytical methods or by recognising unnecessary conservatism in the
capacities determined. However, this simple approach to assessment does not recognise
one of the principal reasons why concrete bridges are able to carry far greater loads than
might be expected; their capacity for redistribution. For a concrete bridge to collapse it
is necessary for a failure mechanism to form. Therefore, the maximum load a structure
can carry safely is not necessarily governed by the load that causes the first ‘region’
within a structure to reach its capacity (i.e. to yield), as often assumed in design. If a
structure is sufficiently ductile and provided alternative load paths exist, additional
loading can be carried safely by the structure through redistribution until sufficient
‘regions’ are yielding for collapse to occur. Generally, concrete bridges have both
adequate ductility and redundancy to enable significant redistribution so that there is a
sizeable difference between the load a t ‘first yield’ and the ultimate collapse load.

If the benefits of redistribution are to be realised in assessment, the simple approach of

separating load effect calculations and capacity calculations cannot be retained generally.
Instead, a form of analysis that accounts for both load effects and capacities in
combination must be adopted, such as plastic methods or non-linear numerical methods,
usually non-linear finite element analysis (NLFE). Methods that can account for

49
 6 Analysis for assessment

redistribution require careful application, in particular, in understandingthe implications of

the ductility of concrete structures. However, if distress occurs at the serviceability limit
state, a structure may be considered to be ‘unsafe’ and usage may be restricted as a result.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

6.1.3 Lower bound and Upper The methods used for the analysis of concrete structures a t the ultimate limit state are
bound methods underpinned generally by Plastic Theory (see, e.g., Clark6.3).When an elastic analysis or a
strip method is used, the resulting distribution of stresses is in equilibrium with the
applied loads and therefore, provided the stress nowhere exceeds the capacity of the
structure, a lower bound assessment of the strength of the structure is obtained.
Numerical methods, such as grillage and finite element analysis, can be used to find
distributions of stresses that satisfy equilibrium approximately. The results of applying
such methods can be considered generally as lower bound assessments, although strictly
this is not the case since equilibrium is not satisfied exactly.

Methods, such as yield-line analysis, identify a failure mechanism but do not check
stresses everywhere. They are upper bound methods therefore. They can be particularly
useful in bridge assessment since they account directly for a structure’s capacity for
redistribution.

The degree of complexity of different methods of analysis and uses are shown in
Table 6.1.

6.1.4 Ductility Plastic methods implicitly assume that structures are ductile. Generally, concrete
structures are sufficiently ductile for this assumption to be valid and some guidance on
limits is given in standards. However, establishing precise bounds on the-ductility
requirements for plastic analysis to be valid is not ~traightforward~.~,~.~.
Non-linear
numerical methods, such as non-linear finite element analysis, can be used to account
‘2
for the effects of limited ductility. However, such methods can be highly complex (and
therefore costly) to apply, and particular care and expertise is required to ensure that the
results are reliable.

.
6.1.5 Differences from design Superficially, assessment is similar to undertaking analysis and checking sections in
design. There are, however, important differences which should alter the approach
fundamentally. If this is not appreciated, it can result in a significant waste of resources

Table6.1 - 1 Increasing complexity I Analysis approach I Example

 on unnecessarily sophisticated analysis. More seriously, it can result in structures being
strengthened or even condemned when, in fact, they are satisfactory.

Even quite minor repairs, strengthening or modifications to existing bridges are very
expensive. This contrasts sharply with the situation in design. A structure that does not
yet exist, is very easy to change. Therefore, the designer chooses a convenient analytical
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

approach and then adjusts the structure to comply with the results of the analysis.
In assessment, in contrast, the approach should be to alter the analysis to suit the
structure. There are a wide variety of analyses normally which will give different but safe
assessments. The further the structure is from being statically determinate, the greater
the range of possible solutions.

The analytical approaches used in design are often conservative. Alternative approaches
are available which give more realistic results. These approaches are often, although not
always, more expensive. The high cost of strengthening existing structures makes it
more likely to be worth using them in assessment. However, whilst the designer aims for
a consistent margin of safety, and hence uses consistent analytical approaches, the
margin of safety in existing bridges may vary wildly. Because of this, there are also cases
where a very simple conservative analytical approach is quite adequate to show that a
structure is satisfactory.

Another difference between design and assessment relates to the relative importance of
the different limit states. In design, serviceability is often critical. In assessment, the first
and most important responsibility is always to ensure safety, to ensure that the ultimate
strength is adequate. BD2l says that pre-1965 bridges will not normally be checked for
serviceability.The serviceability checks in BS 5400 Part 4, particularly the stress check,
impose a severe limitation on the benefit that can be gained from using non-elastic
analyses. Without this limitation, the scope for using plastic and other inelastic methods
is much greater.

The effect of these differences is that a wider range of analytical approaches are used in
assessment than in design. Faced with a beam and slab bridge to analyse for design,
most engineers would use a linear elastic analysis, probably using a grillage model and
section properties based on gross-concrete sections. They would tend to use the same
form of analysis for all such bridges. In assessment, it may be appropriate to use a
simple static analysis, a grillage based on a variety of section properties, a three-
dimensional elastic finite element model, a plastic analysis, a non-linear finite element
analysis, a load test or a combination of these approaches.

6.1.6 Structural condition One of the most important differences between design and assessment lies in the fact
that assessments are undertaken on actual structures that typically have been in service
for some years. As a result their condition may not be as it was intended when designed
either as a result of construction errors or subsequent deterioration. Assessments should
be undertaken on the basis of the actual properties of the structure, as built and as
matured. Further guidance on such issues is included in Chapter 9.

51
 6.1.7 Global and local In assessment, as in design, it is usual (although not universal) to consider global and local
analysis analysis separately. In a beam and slab type deck, the global analysis gives the moments
and forces in the beams and diaphragms. If there are no intermediate diaphragms, it will
also give significant transverse moments in the slab. These moments, which are known as
global transverse moments, have to be considered in the slab design.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

The distinction between global and local analysis does not arise in slab type bridges. The
overall or global analysis of these can be treated in a similar fashion to global analysis in
beam and slab bridges. However, particularly for smaller bridges, it will often be easier to
use analytical approaches more often used for deck slab analysis.

6.2 Simple methods

6.2.1 Strip method The simplest way to determine forces and moments in the beams of a beam and slab
structure is to use a static load distribution ignoring the capacity of the deck to
distribute loads. The approach can also be used to determine the moments in a section
of a slab deck, although when axle loads (rather than uniform lane loads) are considered,
it will be necessary to assume some degree of spread of the individual wheel loads.
Because it is so simple, this method is useful as an independent check when more
sophisticated methods are used.

A static load distribution provides a set of forces which are in equilibrium. It is therefore
a lower bound (safe) method according to plastic theory. Hence, if a structure has
adequate strength according to such an analysis, it is safe. If only a strength assessment
is required, there is no reason to go to a more sophisticated analysis. The analysis also
gives conservative figures for serviceability assessment of longitudinal members.
However, it cannot provide any information about transverse moments since it ignores
them. Hence, if serviceability assessment of transverse members is required, a more
sophisticated analysis is needed.

6.2.2 BA16 method BA16 provides a simple assessment method using empirical charts. This approach is
more restricted in application than a static distribution and cannot be used for HB
loading, but it can be less conservative. There is also a similar approach for transverse
members but this is still more restricted in application.

6.2.3 Upper bound limit and The opposite extreme from a static load distribution is to assume the distribution is
check perfect and compare the total longitudinal moment and vertical shear across the width of
a bridge with the total available capacity. This is an unsafe method (it is related to yield-
line analysis) but it does give useful information. If the bridge fails this test, no amount of
improved load distribution will make it work. It provides an upper bound to strength and
also another useful check on other analyses. The distribution analysis should always give
results which fall between those obtained using a static and perfect load distribution.

52
 If both the section and the loading are fairly uniform across the width of the bridge,
there may be little difference between a static and ideal load distribution, in which case
there is little advantage in doing more advanced distribution analysis.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

6.3Elastic methods
6.3.1 Elastic grillage If the simple methods do not show the bridge to be adequate, the usual next step is to
use the type of elastic methods used in design. Before embarking on such an analysis, it
is worth reviewing its limitations. The increase in strength comes only from improving
the distribution of load between the beams. This distribution comes as a result of the
transverse moments. The distribution analysis cannot increase the total moment
capacity of the bridge. This means that if the bridge has no transverse moment capacity,
or if all the beams are fully loaded in the simple analysis, a distribution analysis cannot
increase the capacity.

Section properties
When elastic analysis is applied to reinforced concrete structures, BD44 (like 855400:
Part 4) allows considerable freedom in the choice of section properties. It allows them to
be calculated from either cracked or uncracked sections. Either approach leads to
satisfactory, but sometimes significantly different, results. In design, uncracked
properties are used almost invariably. The main reason for this is that cracked section
properties can only be calculated after the reinforcement has been designed, whilst the
primary object of the analysis is to determine the required amount of reinforcement.
Reinforcement can be designed directly using uncracked concrete properties whilst, if
cracked properties are used, the process becomes iterative. In assessment, this
disadvantage of using cracked properties does not arise.

Cracked v uncracked properties

In a slab with isotropic reinforcement, it makes no difference to the predicted moments
whether cracked or uncracked properties are used as the longitudinal and transverse
stiffnesses are equally affected. However, most bridges are far from being isotropic. In
the longitudinal direction they may be heavily reinforced or they may have prestressed
concrete or steel beams. In any of these cases, the choice of approach will have little or
no effect on stiffness. Transversely, however, they may be very lightly reinforced and the
uncracked stiffness can be as much as ten times greater than the cracked stiffness. This
is often the case in older structures which were designed using a static load distribution
and so have very little transverse steel. When such structures are assessed using a
conventional grillage based on gross-concrete properties, the transverse steel often
appears overstressed. However, a more realistic assessment using uncracked transverse
stiffnesses will show the transverse steel to be less highly stressed a t the expense of
giving a less favourable distribution between beams.

53
 Varying section properties
Reducing the transverse stiffness can result in an analysis that suggests that the
longitudinal members are inadequate.Sometimes analysis using cracked properties gives
overstressed longitudinal members, whilst analysis using uncracked properties suggests
that the transverse steel is inadequate. In such cases, an analysis with intermediate
properties could show that both are satisfactory. Using intermediate properties is also a
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

useful technique when the top and bottom steel is different, and it is not obvious whether
the section will be in sagging or hogging. It will be safe to use section properties calculated
from the greater steel area. The strength is then checked using the appropriate area.

Torsionless gri llages

Another property that frequently has to be varied in analysis for assessment is the
torsional stiffness. The beams in many older structures were not designed for torsion
and will appear to have inadequate torsional strength. However, a grillage analysis using
reduced or zero torsional stiffness will still produce a safe assessment.

This is a useful technique for assessing skew slabs, particularly where the transverse
reinforcement is either skewed or very light. A torsionless skew grillage can be used with
the elements running in the reinforcement directions. The moments from this can be
used directly to check the reinforcement without the need for post-processing, The
approach is safe but it is not valid for serviceability assessment. In particular, it will not
predict the top cracking which can arise in the obtuse corner of simply supported slabs.

Slab reinforcement
Whilst beams can be checked directly from grillage results, more interpretation is required
in slabs, particularly skew slabs. The reinforcement in these is normally designed, and often
assessed, using Wood-Armer equations, usually via post processors on the grillage or finite
element program. The solution is derived by determining the minimum total steel to resist
the moments a t the point. This approach has limitations in assessment but is almost
always conservative as is explained in more detail in Section 7.5.

6.3.2 Elastic finite e[efTlentS Elastic finite element analysis is used in various ways in assessment and the main ones
will now be considered.

Plate models of slabs

Plate finite element analysis is sometimes used for geometrically complicated slab decks
because, with modern programs, it is easier to assemble the models. Finite elements give
a better representation of the behaviour of elastic plates than grillage models. However,
as a reinforced concrete slab cracks, the moments tend to approach the distribution given
by a grillage using cracked section properties. If isotropic plate elements from simpler
programs are used this loses the ability to take advantage of cracked properties or of
reduced stiffnesses to allow redistribution. In extreme cases of highly skewed slabs with
light transverse reinforcement placed parallel with the abutments, rather than
perpendicular to the main steel, the cracked elastic stiffness in the stiffest and least stiff
direction can differ by a factor of 100. It is common for a conventional assessment based

54
 on isotropic elastic finite element analysis to give such a structure a significantly lower
assessed capacity than a static load distribution. This problem can be avoided by using
more sophisticated programs which allow orthotropic properties.

When plate analysis is used with either point loads or point supports, some care is needed
in interpreting the results. The peak moment intensity under a point force in an elastic thin
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

plate finite element analysis is very ‘mesh sensitive’; it tends towards infinity as the element
mesh is made finer. Because the real slab has a finite thickness, and the load has a finite size,
this peak does not arise in practice. The problem can be relieved by modellingthe size of
the load or support. However, for the ultimate limit state it is considered appropriate to
spread the peaks over a reasonable width; perhaps three times the slab depth.

Another consequence of mesh sensitivity is that if the elements are too big in relation to
slab thickness, it is possible to get unsafe results with the analysis failing to pick up
moment peaks which can be important.

Beam and slab decks

In a finite element model of a beam and slab deck, the beams and the parapet upstands
can be modelled using offset beam elements with the slab modelled with plates. This is
more realistic than a grillage analysis in the sense that it gives a more realistic
representation of the interaction between string course, slab and beams. It also enables
the real behaviour of intermediate diaphragms (or, more commonly, steel cross girders)
which are separate from the slab to be modelled. However, a finer than normal element
mesh is required to achieve this.

If this type of analysis is done with a fine enough element mesh, the results can be used
directly to check the slab reinforcement, avoiding the requirement for a separate local
analysis. This approach used to be very expensive but, as computer power becomes
cheaper, it becomes more economic to use it in a wider range of circumstances.

It is preferable to use a program which enables the plate elements to be made

orthotropic, even if the initial analysis is undertaken using isotropic properties. If, as
often happens, initial analysis gives excessive stresses in the transverse steel, the analysis
can then be adjusted by changing the element properties. This is much easier than
transferring to a completely new analysis.

Three-dimensional models
Full three-dimensional finite element analysis of box girder and cellular structures is a very
powerful and, with modern hardware and software, reasonably economic way of obtaining
realistic serviceability assessments. However, its use in strength assessment is more
restricted. It does have application on major structures where, for example, it can be useful
in determining the implications and causes of observed stresses and behaviour.The
difficulty in strength assessment is that code approaches for determining bending and
shear strength work on whole sections. Thus, having undertaken a sophisticated analysis
which gives elastic stresses throughout the section, moments and shears for comparison
with code predicted capacities can only be obtained by integrating the stresses over the
section. The results are very similar usually to those obtained from a simple beam analysis.

55
 8 Analysis for assessment

6.3.3 Westergaard and Global analysis

Charts As noted above, it is possible and now sometimes economic to analyse a whole bridge
with a finite element model with a fine enough mesh to use the results directly to check
the slab reinforcement. However, it is more usual to use less detailed analyses and these
call for a separate analysis to determine the local moments in the deck slab due to
wheel loads. These local moments have to be combined with the global transverse
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

moments from the global computer model: the moments induced in the deck slab by its
action in distributing load between beams. Only coexistent values (values occurring in
the same piece of slab under the same load case) need to be combined.

Nodal loading
In order to avoid the combination of global and local moments being over-conservative,
it is necessary to ensure that the results in the global model do not include any
component due to local moments. This is ensured if there are no nodes between the
beams and the loads are applied to the nodes rather than using the option available in
many programs to apply them to the members.

Slabs
The methods used to determine local moments can also be used to analyse slab structures
and are usually quicker to use than a grillage for short span structures where only a few
wheels have to be considered. Conversely, on major bridges with wide spans between
beams or webs, it may be easier to analyse the deck slab using finite elements or a grillage.

Westergaard
The simplest approach is that due to Westergaard66.This considers a one-way spanning
simply supported slab allowing for its finite thickness. It gives the maximum moment
per unit width under a single wheel as:

M, = 0.2107 [0.4825 + Io~(s/c,)]

My= M, - 0.0676P

Where P = value of point load

s =slabspan
C, = 2[,/(0.4c2+ h2)]

c = diameter of loaded area

h =slab thickness

Strictly, 5 should include the spread through the surfacing allowed by 8037 but not the
spread through the slab as this is allowed for in the analysis.

In reality, most slabs are built into beams and are continuous. An approximate
correction for this is to reduce M,and Mygiven by the above equations by 0.07P and
0.1065P respectively.

When the slab span is longer than a critical value, normally 1.7 times load spacing, the
effect of more than one wheel has to be considered. Westergaard derived solutions for
this and they are also given by R o ~ ande ~elsewhere.
~

56
 Pigeaud Charts
Older bridges often had intermediate diaphragms so that the slab was two-way
spanning. A simple elastic way of analysing these for concentrated loads using charts has
been derived by Pigeaud and the necessary charts are given by R o ~ and e ~by ~Reynolds
and Steedmad8.They can be used for one-way spanning slabs as well but other
methods are usually easier.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Influence surfaces
The other common way to undertake elastic analyses of slabs is using published
influence surfaces. These are the two-dimensional equivalent of influence lines. They
plot the bending moment intensity (for example) at a particular point in a slab due to
point loads applied a t all positions in the slab. To use them the wheel load area,
including the spread allowed by the code, is plotted on a copy of the chart and a
numerical integration of the influence factors over this area undertaken. For simply
supported, one-way spanning slabs, the approach gives very similar answers to
Westergaard. However, for continuous slabs it has the advantage of giving a more
realistic account of the effect of continuity, giving reduced bottom steel requirements,
and of giving a value to check the top steel against.

The best known influence surfaces are due to Pucher6’. There are others, however,
including some which enable slabs with haunches to be considered, as well as skew slabs.

6.4 P h S t i C equilibrium The lower bound theorem of plastic limit analysis states that if a distribution of stresses
methods can be found in equilibrium with an applied loading, and which nowhere exceeds the
capacity of the structure, then that load can be safely carried by the structure. If a
structure is statically indeterminate (i.e. it has some redundancy), then different
equilibrium distributions of stresses can be found and any one of these would be
acceptable as the basis of a plastic lower-bound analysis. This theorem underpins
many of the approaches used in the design of structures, particularly elastic methods,
see, e.g., Clark63.It is important to recognise, however, that this theorem is strictly
only applicable to ductile structures and cases where displacements are small.
Furthermore, it is essential that equilibrium is satisfied everywhere throughout the
structure.

The ductility of reinforced concrete sections in flexure can be assessed from their
rotation capacity. Considerable research has been focused on establishing the rotation
capacity of reinforced concrete sections, a good review of which is given in CfB/NP
Bulletin d’lnformation No. 2436’0.The rotation capacity of concrete sections is
governed either by concrete crushing or by reinforcement fracture. Criteria are included
in BD44 for both of these cases, although the expressions given do not appear to be
consistent with the latest research in this field, as described in CfB/NP Bulletin
d’lnformation No. 243. Particular care should be taken when considering structures
which are either heavily reinforced or which are lightly reinforced with reinforcement
that itself has limited ductility. Structures that are moderately reinforced typically have a
high degree of ductility, making them suitable for plastic analysis.

57
 If lower-bound plastic methods of analysis are to be used, it is also necessary to consider
carefully the shear capacity requirement for the structure. Typically, this may be done by
ensuring that the shear capacity of the structure is sufficient for both an elastic analysis
and for the lower-bound equilibrium stress distribution.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

6.4.1 Plastic redistribution If an elastic analysis is undertaken for a structure with a particular applied loading, the
resulting distribution of stresses will be in equilibrium with the applied loading, and
therefore provide a plastic lower-bound stress distribution. further lower-bound stress
distributions may then be found be adding any self-equilibrating stress distribution to
the elastic result. (A self-equilibrating stress distribution is a non-zero distribution of
stresses throughout a structure, in equilibrium with itself and with zero external applied
loads, except for support reactions. Parasitic moments that develop in continuous
prestressed structures are an example of a self-equilibrating stress distribution.)

This approach is commonly used in the design of continuous beams in buildings, where
moments are redistributed from mid-span to supports, or visa-versa, by adding a
moment distribution consisting of straight lines between supports (i.e. a self-equilibrating
moment distribution). Such moment distributions can be derived from considering the
effect of a small rotation occurring through the formation of a plastic hinge.

A similar approach can be used in the analysis of reinforced concrete bridges. This is
most simply undertaken by considering the effect of a small rotation in a yield line
across the supports if there is a hogging deficiency a t this location or a t mid-span if
there is a shortfall in sagging capacity. The degree of redistribution permitted by BD44
can be checked by comparing the rotation required in the yield line with the specified
limits. More sophisticated self-equilibrating stress distributions can be developed.

6.4.2 Hillerborg Strip method The Hillerborg strip method6”,612is an alternative lower-bound plastic method in which
a distribution of stresses in a structure, in equilibrium with an applied loading, is found
by dividing the structure into longitudinal and transverse strips, and assigning a
proportion of the loading applied to the region where a longitudinal and transverse strip
intersect, to each of the two strips. The sum of the loads applied to the longitudinal and
transverse strips in the intersecting region must be equal to the applied loading.

The method is very convenient for designing two-way spanning slabs in buildings. The
approach is less well suited however to the assessment of one-way spanning structures
(such as most bridge decks) with complex loading patterns. Yield lines are usually a
more appropriate approach as is outlined in the following section

58
 6.5 Yield-line analysis

6.5.1 P r i n C i p k S Yield-line analysis is a long established method of using the plasticity of reinforced
concrete slabs in order to obtain greater capacity than that obtained by elastic analyses.
It was pioneered by K W Johansenin his doctorate thesis in 19436’3.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Geometrically compatible plates of a bridge deck are deflected under load to simulate a
failure mechanism. Each plate is bounded by straight lines and the boundaries form
plastic hinges with the reinforcement yielding such that a mechanism forms. The work
done in deflecting the load is equated to the work done in yielding the reinforcement
along the plate boundaries. See Figure 6.1.

6.5.2 Upper bound SO[UtiOnS This method provides an upper bound solution so it is necessary to examine all conceivable
combinations and permutations of plate configurations to determine the failure mode

Figure 6.1
General principles of yield-line analysis for a simply

7
supported slab considering one simple mechanism.

I (I Plate A

Plate B
I
L Sagging yield line

Plan
Slab simply supported on t w o sides
subject t o uniformly distributed load

Unit displacement

Section A-A
External work done = Load on Plate X displacement of centroid

Section A-A
Internal work done =Yield moment of reinforcement crossing the yield line X (OA + OB)

59
 Awalysis for assessment

which gives the lowest ratio of applied work to internal work. Therefore, higher partial safety
factors are used usually for yield-line analysis in order to allow for this uncertainty.

An upper bound to the load capacity is derived from equating external and internal work
done.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

6.5.3 Slab decks Yield-line analysis is a simple and effective analytical technique for slabs. Care is needed
in the analysis as any given mechanism gives an upper bound for strength. Published
and affinity theorems can be helpful in finding the worst mechanism and
the corresponding strength, particularly for short spans where fan mechanisms around
individual wheels have to be considered in order to identify the worst case. In the simple
example shown in Figure 6.2, the circular fan is a significantly worse mechanism than
the easier to analyse square.

Work equations for deflection of 6 (uniform slab sagging capacity M/unit width, hogging
capacity m/unit width):

Circle
P ~ = I -d4+m(?)
~Mr6
r
:. P = 2~ (M + m)
Square

:. P=8 ( M + m )
+
~ T ( M m)<8(M + m )

i.e. the circular yield-line pattern gives a lower capacity than the square pattern.

The affinity theorem? l 4 allow standard solutions for right isotropic slabs to be applied
to skewed and orthotropic slabs. This is particularly helpful as the worst case solutions
for orthotropic slabs include elliptical fans.

For some slab bridges, it is relatively easy, if laborious, to write the work equation which
gives upper bounds to the failure load (see Clark63)for yield-line mechanisms such as

Figure 6.2
Yield-line mechanisms for a point load on a slab.

60
 the ones shown in Figure 6.3 and for a skew bridge in Figure 6.4. However, care must
always be taken to ensure that the most critical mechanism has been found, and
thorough account must be taken of the influence of reinforcement curtailment,
variations in section properties and support conditions. Computer methods can be
extremely helpful for rapidly analysing more complex structures to identify critical yield-
line mechanisms (See Section 6.5.6).
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Figure 6.3
Typical yield-line mechanisms for slabs.
(uniform steel giving moment capacity M
per unit width)

/ / / / / / / / / / / I

-Sagging yield line

._____ Hogging yield line

Figure 6.4 Simple support

Typical yield-line pattern for a slab bridge
(from Clark63).

Footway
-
Footway
Hogging yield line
Sagging yield line

61
 8 Analysis for assessment

Yield lines v. Elastic analyses

Yield-line analysis is generally only used when elastic analysis has failed to show adequate
strength. However, for simple slabs subjected to wheel loads, yield-line analysis is often
quicker than elastic analysis so it may be appropriate to start with yield-line analysis. The
method can also be applied to continuous and trapezoidally-shapedslabs.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

6.5.4 Beam and slab decks Principles

Yield-line analysis can also be used for beam and slab decks. However, this requires
more judgement as methods of analysis are less established. Recent work on developing
a computer program to analyse beam and slab decks has been carried out by
Middleton6”. The work equation is written and solved in the same way as for slabs. As
with all yield-line analysis, it is necessary to check shear strength separately. Ductility
problems are also much more likely as beams can often be over-reinforced.

Torsional hinges and transverse hinges in slabs

Some engineers have included work done by torsional hinges in the beams and also
moment hinges in slabs acting in conjunction with beam hinges. However, when
Jackson616investigated the approach both theoretically and by comparison with tests,
he concluded that it was better not to do this. The deformations required to develop full
torsional resistance are unrealistically large whilst the transverse section of the slab acts
as the top flange of the beam. It cannot simultaneously be fully stressed in compression
under global moments and acting as a yield line. Ignoring work done against these
hinges makes the analysis only marginally more conservative but considerably easier.

6.5.5 BOX Culverts and Yield lines can be particularly helpful in the assessment of box culverts and portal
retaining walls structures in order to take full account of their transverse and longitudinal capacity

Transversely, the full support and mid-span section capacity can be used although
redistribution methods described in Section 6.4.2 can also be used. Care must be taken
to consider plastic hinges that can occur where corner bars are curtailed either in the top
slab or part way down the walls, see Figure 6.5
Figure 6.5
Yield lines for the transverse analysis
of box structures.

Uniform reinforcement Effect of curtailment in top slab Effect of curtailment in Effect of curtailment in
wall reinforcement (1) wall reinforcement (2)

62
 Figure 6.6
Yield lines for the longitudinal analysis of box

1
structures.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Capacity of this yield line may be

affected by transverse hinge location
as shown in Figure 6.5

Longitudinally, more conventional slab yield lines can be used to increase HB or BD21
vehicle assessment capacities, assuming the walls provide moment restraint. The slab
capacity a t the wall joint can also be limited by curtailment of the corner bars, see
Figure 6.6.

Yield lines can also be used to analyse retaining walls and wing walls as is described by
Clark63.

6.5.6 Limitations O f yield-line There are a number of important points to be considered when undertaking yield-line
methods analyses.

Ductility
It is essential that reinforced concrete slab sections have sufficient ductility to justify
using the method. However, as discussed by Denton 6 5 , it is difficult to quantitative
precisely such ductility requirements since they can be dependent upon the structure
and the loading arrangement. Certainly, the sections should be under-reinforced such
that reinforcement yields substantially before concrete crushing occurs.

Codes generally allow the application of yield-line analysis when the neutral axis depth
is less than of the section depth. Such a limit will be safe, and may be quite
conservative in some cases. Jackson616has proposed significantly less onerous rules such
that if the strain in the outermost layer of reinforcement under ultimate capacity, as
defined in BD44/90, exceeds 0.0024 + 1.2fy/€,, the ductility is adequate. Similar
requirements are given in the draft BD on the use of yield-line analysis. This will not be a
problem unless the slab is heavily reinforced or concrete is of very low strength. It is
often worth taking cores for structures which may have low strength concrete in order
to justify higher strengths and ductility. Problems with ductility may also occur in very
lightly reinforced slabs with brittle reinforcement.

Edge beams
It may be necessary to be conservative when estimating any benefit from edge upstands
as the composite section may be non-ductile. Torsional resistance of edge or parapet
beams is also best neglected as is discussed in Section 6.5.4.

63
 6 Anallyis for assessment

Anchorage and curtailment

Bars must be fully anchored on each side of the yield line as the reinforcement must be
able to reach its full yield strength. Therefore, curtailment of the reinforcement or
inadequate laps can influence significantly the failure mechanism.

Concrete deterioration and corrosion

Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Delamination of the cover concrete adversely affects the bond strength of reinforcement
and extends the effective curtailment lengths. It can also reduce the effectiveness of laps
such that full yield cannot occur. Moreover, pitting corrosion of reinforcement can cause
premature yielding on a reduced section, which greatly reduces the overall ductility. In
general, plastic and yield-line analyses are not recommended for deteriorated concrete
structures.

Lift off at supports

When analysing heavily skewed slabs, it may be necessary to consider mechanisms in
which the slab lifts off the supports. Mechanisms that rely on corners being held down
are potentially unsafe.

Slab voids
For voided slabs, it is necessary to check that excessive local bending moments are not
developed in the top and bottom flanges through transverse shear.

Prestressed slabs
The ductility of prestressed slabs can be different to reinforced concrete and
serviceability is more important. Therefore, the method is not recommended generally
for these types of structures.

Local and global effects

Further to the points discussed in Section 6.5.4, when yield lines are used for beam and
slab bridges, due to redistribution, global flange forces or global transverse moments in
the deck slab can be ignored when assessing local strength except in slabs which are
highly stressed under both global and local effects. In such cases, the brittleness of local
failure modes can prevent redistribution. When the slab is used to enhance the plastic
moment capacity of the beam, the longitudinal moment capacity in the slab should be
ignored in the work done against yield lines in the slab.

6.5.7 Computer analysis Yield-line analysis can require a large number of calculations to check all the possible
failure configurations. Computer programs and spreadsheets have been written to speed
up the analysis process, for example, as described by Denton617.

Work by Middleton a t Cambridge University in conjunction with the Highways

Agency6’’ has led to the development of the computer program COBRAS6’*for bridge
deck yield-line analysis, which is now commercially available. The program is easy to use
and undertakes a large number of yield-line analysis patterns with the minimum of data.

64
 COBRAS is understood to be the most sophisticated commercially-availableyield-line
analysis program and enables a wide range of yield-line configurations to be selected
and analysed. The program varies each yield-line configuration in turn on the bridge
deck in small increments, so finding the critical failure geometry for each configuration.
The optimum solutions of each of the various configurations are then compared with
each other and the overall optimum solution obtained for the particular loading case. In
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

this way, the assessed capacity of the bridge deck is determined.

6.5.8 Shear Conventional yield-line analysis only assesses the ultimate bending capacity. It does not
deal with shear capacity. Normally, shear capacity is checked both by punching shear
and by the shears output from a standard elastic analysis, for example grillage or finite
element analyses. Research has concentrated on plastic methods for the assessment of
shear, but the work is quite

6.6 Strut and tie action Strut and tie action has been used for the analysis of pile caps and anchorages. The
principles are set out in Reference 6.20 and an example of use is given in Reference 6.21.
It can be used also for the analysis of diaphragms and deep beams. The principle is
based on establishing compressive struts of concrete and tensile ties of reinforcement.
The size of the compressive struts is dependent on the allowable compressive stress
for the grade of concrete and the force carried. The method usually requires iteration
to establish the appropriate geometry of the strut and reinforcement in the ties
(see Figure 6.7).

Plane frame or space frame analyses can be carried out to solve two and three-
dimensional strut and tie problems. These are linear-elastic analyses of a non-linear
problem and serviceability needs to be considered as discussed in Section 6.7.

Example - Column head

A reinforced concrete column of diameter D supported a steel box crossbeam carrying a
motorway deck through a steel rocker bearing. Below the top solid section of depth H of
the column, the column was hollow with reinforced concrete walls with a thickness of t.
Hoop reinforcement was provided in the solid section as anti-bursting reinforcement.
From strut and tie action, a calculation of triangle of forces provided the transverse
tension Tand, hence, the load in the hoop reinforcement, TI2 as shown on the plan. It
can be shown that the force in the hoop reinforcement is given by:

T=P x (D- t ) / H

In this case the reinforcement was partly corroded and the upper part of the column was
strapped with steel rings to compensate for the loss of capacity.

65
 Analysis for assessment

Figure 6.7
Example of strut and tie action. TI2 Hoop force

Hoop
reinforcement
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Radial tie and 1

strut forces

PLan

J ,
Bearing load P

Compression struts

1" O
Tie provided by hoop steel
0
0

+ --t
D
4

-
Strut and tie action on the top of a hollow section

6.7 Serviceability limit

state

6.7.1 Inspections BD21 assumes that for older structures an inspection gives the best indication of
serviceability and therefore quantitative analysis is n o t required. For newer structures,
serviceability assessments are recommended because there is insufficient service
experience to ensure that problems will be apparent from inspection. The same arises
for structures where the loading is being increased significantly, such as due t o road
widening or increased railway loads, although this is n o t mentioned in the BD.

BD21 and BD44 do not give serviceability criteria. The usual approach is t o report
against the criteria used in design. However, remedial action is not always required for
'failures'. Experience shows that structures that are predicted by elastic analysis t o suffer
serious yielding and cracking often show no signs of distress. Where it is unclear if
remedial action is needed, a quantitative elastic assessment can be used t o identify
areas for closer inspection and future monitoring.

66
 6.7.2 Plastic methods of When elastic methods of analysis with representative section-stiffnesses are used t o
a na ly s is assess the ultimate limit state loading effects, this generally does not result in excessive
stresses and deformations at the serviceability limit state. However, the same is not
necessarily the case for yield-line, strut and tie action and other plastic methods of
analysis which assume that yielding of reinforcement can occur as the ultimate loading
levels are approached. Consequently, if plastic methods of analysis are used it is prudent
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

t o consider separately the serviceability performance of the structure, either using elastic
methods or special inspection.

6.7.3 Special inspections If structures fail the serviceability limit state assessment, it is recommended that a
special inspection of the parts of the structure that are expected t o show signs of
distress is carried out. It is best that such inspections are carried out after assessment as
the pre-assessment inspections often do not identify the problems or do not record the
information in sufficient detail for its significance t o be appreciated. In practice,
inspections have shown that slabs failing their ultimate capacity by elastic analysis but
passing by yield line, may not be severely cracked, and their serviceability condition may
be deemed t o be acceptable based on inspection. In such cases, continued monitoring t o
ensure durability is advised.

6.8 Soil structure Soil structure interaction can be modelled by finite element analysis and finite difference
interaction programs such as FLAC to realistically assess the soil forces which act on the structure.
Typical examples are reinforced concrete arches and buried box and portal structures.
Particular care is required, however, with such an approach due t o the variable
properties of the soil, and the risk that the relieving effect of the soil on the structure
may not be as great as predicted. Advice from geotechnical engineers is recommended
and a range of Sod properties used t o take account of the uncertainties.

The distribution of bearing pressures under spread foundations depends on the type of
soil and the duration of loading. Cohesive soils which are 'soft' because of long-term
settlement give uniformly varying bearing pressures under bases. Cohesionless soils are
'hard' and slight deformation of the structure, such as the hog of the base slab of boxes,
can lead t o parabolic bearing pressure distributions with peak values at the hard spots,
which in the case of the box would be under the walls. This effect can be analysed with
simple plane frame analysis of the box with springs representing the soil. Again, a range
of soil stiffnesses should be used.

Embedded retaining walls such as contiguous and secant piles walls or diaphragm walls
should be analysed with suitable soil structure interaction programs such as FREW and
WALLA P.

67
 8 Analysis for assessment

6.9COndUSiOnS Many detailed points have been made in the previous sections. However the important
principles to note are:

1. In contrast to design, the details of a structure being assessed are fixed. It is necessary
therefore to choose an analysis which is appropriate to the structure. The structure
cannot be altered to suit the analytical results; the analysis has to be adjusted to suit
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

the structure.
2. When an analysis suggests that something is inadequate, its significance should
always be questioned. It may be a real problem but, frequently, there are other
analytical methods available that can prove that it isn’t. Even if this analysis is very
expensive, it is still likely to be much cheaper than resorting to strengthening the
structure.
3. There is no reason to use an analysis that is any more expensive than the minimum
required to prove the structure adequate. If a more sophisticated and expensive
analysis is used (compared with a static load distribution) then there should be logical
reasons for the choice.
4. A simple, non-linear or plastic analysis may not fully satisfy serviceability
requirements. Serviceability checks and inspections can be used to assess the
potential risks of durability problems arising during the remaining life of the structure.

6.10References 6.1 HIGHWAYS AGENCY, BD 44 The Assessment Concreteof Highway Bridges andStructures, The Highways Agency, London.
1995
6.2 HIGHWAYS AGENCY, BA 44. The use ofBD44 forthe Assessment ofconcrete Highway BridgesandStructures. The
Highways Agency, London. 1996.
6.3 CLARK, LA, Concrete bridge design to 55 5400, Construction Press, London, 1983.
6.4 MORLEY, C and DENTON. SR, Modified plasticity theory for reinforced concrete slab structure of limited ductility,
StructuralEngineering, Mechanics and Computation,Vol. 2, 2001
6.5 DENTON. SR, Theanalysis ofreinforcedconcreteslabs, and the implicationsoflimiredductility, PhD thesis, Cambridge
University, 2001
6.6 WESTERGAARD. HM, Computation of stresses in bridge slabs due to wheel loads, PublicRoads. Vol. 2. No 1. March
1930, pp 1-23.
6.7 ROWE, RE, Concrete Bridge Design. Applied Science Publishers ttd, London, 1962.
6.8 REYNOLDS, CE and STEEDMAN, JC, Reinforcedconcrete designers handbook, Viewpoint Publications. (Various editions)
6.9 PUCHER, A, Influencesurfacesforelasticplates, Springer Verlag, Vienna and New York, 1977.
6.10 CE8 Bulletin 243, Strategiesfor testing andassessment ofconcretestructures, CEB/FIP. Lausanne. Switzerland, May, 1998.
6.11 HILLERBORG, A, The Advanced Strip Method - A simple design tool, MagazineofConcreteResearch, Vol. 34, No. 121.
December 1982
6.12 HILLERBORG, A, Strip MethodDesignHandbook,E and F Spon, 1996
6.13 JOHANSEN,KW. YieldLine Formulaefor Slabs, Cement and Concrete Association, Camberley, 1972.
6.14 JONES, LL, WOOD, RH, Y,eld-LineAnalysisofSlabs, Thames and Hudson, Chatto and Windus, 1967.
6.15 MIDDLETON, CR, Case Studies from the UK using Yield-Line Analysis for Concrete Bridge Assessment, Austroads Bridge
Conference 1997, Sydney, Australia Vol 1, 1997.
6.16 JACKSON, PA, The global and local behaviour of bridge deck slabs, TheStructuralEngineer, Vol 68. No. 6. March 1990,
pp 112-116
6.17 DENTON, SR. Hidden Strengths, Highways, November 1995.
6.18 COBRAS Concrete Bridge Assessment Package User Manual, University of Cambridge Department of Engineering
6.19 NEILSEN, MP, LimitAnalysisandConcretePlasticity, Prentice Hall, USA, 1984
6.20 LEONHARDT, F, Reducing the shear reinforcement in reinforced beams and slabs, Magazine ofConcrete Research, Vol
17, No. 53, 1965.
6.21 At-SHAWAF, S and JACOBS, P, The Basingstoke Canal Aqueduct, Proceedingsof Institution of Civil Engineers. Vol. 126,
No. 1, February 1998, pp 19-30,

68
 7. Hidden strengths

The large number of structures which have failed assessment to current standards yet do
not display any signs of distress, suggests that there are reserves of strength which are
not taken into account in the normal assessment process. This Chapter explores some
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

factors that affect the strength of structural elements and may be of use in justifying
increased capacities.

7.1 Reinforcement
anchorage and bond

7.1. I First PrinCipkS Structures can often fail assessment because they do not comply with detailing rules
for curtailment and laps. In reinforced concrete beams, the apparent shear strength can be
very low when the ‘additional longitudinal reinforcement’ is small, i.e. when the main
bending reinforcement is highly stressed. In these circumstances, it is worth considering
the anchorage and bond of reinforcement from first principles. The following method is
based on satisfying bending and shear equilibrium in reinforced concrete elements.

7.1.2 Reinforced Concrete Sometimes called Regan’s method because of his publication7’, it generally applies to
beams reinforced concrete beams where links are provided to carry shear. It is also called
‘variable angle truss analogy’ and is included in EN 199272.It can be illustrated simply in
a cantilever situation as shown in Figure 7.1 (a) but it can also be applied to simply
supported and continuous beams.

A shear failure crack is assumed to form at an angle 8 to the neutral axis of the beam such
that the shear force Vis just carried by the capacity of the links crossing the shear plane.

For vertical equilibrium:

V = A,,f,,(Zcot 8 /5)

Where A,,= the cross-sectional area of all the (vertical) stirrup legs a t each stirrup
location
5 = the stirrup spacing along the beam
fsv = the yield strength of the stirrup reinforcement divided by ’ym
Z =the effective depth.

Therefore: cot 8 = (V/ASvfSv)(S/Z) (Equation 7.1)

Cot 8 is generally taken as not less than 1, or greater than 3 (i.e. 45 degrees > 8 >
17.5 degrees). However, EN 1992 restricts cot 8 to the range 1 to 2.5.

69
~
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

4
4 Example of Regan's method

/-rpi
W
Force
1 Force in bar
Applied force due t o shear

Force in bar due E

2 cot H

-
t o moment Z

A
Distance along cantilever

b
J Force to be resisted by main reinforcement
to maintain equilibrium (applied force)

Reinforcement Reinforcement
W
capacity capacity

\ I
1
Aggregate of bond
strength along bar

A Bond from bend

Distance along cantilever

4 Force which can be developed in reinforcement

from anchorage and bond (capacity)

- 7.1
Figure The load on the cantilever tiD W is iust SuDDorted bv a force in the main reinforcement f a t
I ,

Example of Regan's method (variable angle truss

analogy) for reinforced concrete beams
the crack. By considering the equilibrium of the cracked cantilever ABDE by taking moments
appliedtoa cantilever. about the centre of concrete compression C above E, and here neglecting dead loads, then:

vz
wy + Zcot 8) = FZ + -cot 8
2

The bending moment M a t the point in the reinforcement where F is measured is equal
to Wl. Also W = V.

70
 Therefore: M + VZ cot 6'= FZ + (VZ/2) cot 0

F = (M/Z) + Vcot 6'- (V/2) Cot 6'

And this reduces to:

Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

F = (M/Z) + (V/2) Cot 0 (Equation 7.2)

The force in the reinforcement is a function of both the moment and the shear force,
which is the basis of the additional longitudinal reinforcement requirements and rules
for curtailment of reinforcement.

The maximum force in the reinforcement at supports however cannot exceed MJ,Z as
the angle of the shear failure plane is limited by the physical presence of the support, i.e.
cot 0 must be zero. The maximum bending moment in the span is also never exceeded
either as a t that point the shear force Vis zero. See Figure 7.l(b).

Having calculated the force in the reinforcement, the adequacy of the anchorage t o the
curtailment point can be determined by summing the bond strength along the bar from
the point at which the force is measured t o the end of the bar. See Figure 7.1(c). The
assessment Standard gives values for the bond strength and further guidance for
deteriorated structures is given in Section 7.1.3. Transverse pressures applied t o the
reinforcement (e.g. from bearings) can also increase anchorage strength. This is discussed
in Section 7.1.8. Hooks and bends can be taken into account in providing extra bond.

Providing the maximum shear stresses are checked in accordance with BD44, there
should be no problems with compressive shear failure of shallow shear planes with low
angles of 0.

The benefit of this method is that it calculates the force in the reinforcement at any
point by a more exact method than is given in BD44. The method has been used
previously and 'Departures from Standard' have been accepted, drawing upon Clause
5.8.7 of BD44, which allows the use of rigorous analysis for calculating curtailment
lengths and anchorages of bars.

7.1.3 Sub-standard Cover and Fundamental t o Regan's method is the development of the force in the reinforcement,
deteriorated concrete which is dependent on its bond and anchorage. For undamaged structures the bond
strength given in the assessment Standard BD44 can be used to calculate the developed
force. Should the cover be less than the normal cover required for fully developing bond
(i.e. less than the bar diameter) then reduced values should be used. In the case of
damaged or deteriorated structures, the cover can be delaminated and this reduces the
bond substantially. Typically, the bond in deformed bars with delamination t o half-barrel
depth can be 50% of code values provided they are well constrained by links. However,
the bond of plain round bars can reduce t o 10% of the code value. Figure 7.2 illustrates
the effect of reduced bond. Guidance on this is given in a revised draft of BD44, which
takes account of deteriorated structures. See also Chapter 9.

71
 Deficient capacity
Delaminated length

1
.. < Reinforcement capacity
for full bond
Force capacity of bar limited - .,--
Reduced bond
by 0.87 X yield strength
7Full bond
.. -
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Applied force -J

__t
reinforcement
Distance along cantilever

Applied force in main reinforcement

compared with capacity

Figure 7.2
Delaminated bond of reinforcement.

7.1.4 Bond a t laps If the bond strength is reduced significantly there may be insufficient anchorage for
reinforcement and laps may be deficient. Regan’s method will identify lap deficiencies as
shown in Figure 7.3.

If anchorage or lap lengths are less than those required for the bars to yield then
ductility will be compromised and the distribution of load in indeterminate structures
needs to be carefully considered. See also Chapter 6 on general analysis and yield-line
analysis for comments on ductility.

7.1.5 Haunched Sections Regan’s method of calculating the force in the longitudinal reinforcement resulting from
the combination of bending and shear can be used for varying depth sections such as
haunched beams by calculating the appropriate value of cot 8 for the section
considered, as shown in Figure 7.4.

7.1.6 Bent-up reinforcement Where there are bent-up bars towards the beam end, these can be incorporated in the
Regan analysis to carry additional loading as shown in Figure 7.5. The angle 8 of the
shear failure plane is derived such as to provide sufficient reinforcement crossing the
shear failure plane to carry the applied shear. This reinforcement consists of the shear
links crossing the failure plane together with the vertical component of the capacity of
the bent-up bars.

Using the same principles and terms as Section 7.1.2, Cot @canbe found as follows:

(Equation 7.3)

72
.. .. .. ... .. . - . ... . - .. .- ... ' . . .. 1 . _.:~~

Figure 7.3
Deficient lap lengths identified by Regan's method. A
A
,
-- 1L
1,

- L

//
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

1,

LH bars '
Plan of bar laps

Combined capacity

Resistance Capacity of LH bars Capacity of RH bars

force

d
Distance

Capacity of bars with full bond

Reduction
Combined capacity
f in capacity

Resistance
force t of i H bars of RH bari
d
Distance

Reduced capacity due to reduced

bond caused by delamination

Figure 7.4
Shear of haunched beams.
V
t

V =Q-Csin@
0 = Found by trial and error based on number of links t o carryV

Where B = force in the bent-up reinforcement

a = angle of the inclined bars t o the vertical
A,, = area of bent-up bars
fsb = stress in bent-up bars =yield stress divided by ym
(VIZ) = reduction factor t o account for the inclined bars not developing their
yield strength where they cross the shear plane

73
 Figure 7.5
Regan’s method applied to bent-up reinforcement.

2 cot 0

Vertical
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

links
Bent up
reinforcement

R U” 1 2 cot 0

Reduction from
inclined bars

Force

A d
Distance

Force to be resisted by
main reinforcement

The anchorage of the bent-up bars must be such as t o enablef,, t o be developed.

Several different locations of the shear failure plane must be analysed t o provide the
critical capacity of the continuing main bars. Where the bottom of the crack reaches the
face of the support, further locations of the crack should be considered in the form of a
fan with increasing angles of 8.

Taking moments about the intersection of the crack and the compression force C,f is
calculated from:

The force F must not exceed the capacity of the continuing main reinforcement acting
at 0.87 of its yield stress. The anchorage bond of the main reinforcement continuing to
the end of the beam must be sufficient t o develop the force f in the bars at their
intersection with the crack.

7.1.7Inclined links The above examples only consider vertical links but the principles equally apply t o beams
with inclined links. Both the horizontal and vertical components of the link forces are used
in checking equilibrium in much the same way as bent-up reinforcement described above.

74
 7.1.8 Bearing Clamping O f Transverse clamping pressure on reinforcing bars can increase the bond strength and can
reinforcement compensate for inadequate anchorage, lap or curtailment lengths which might arise as a
result of under-design, increased loading, or delamination or spalling of cover. Transverse
pressure may be provided on the soffit of beams or slabs by supporting bearings, piers or
abutments. On the top surface of beams, transverse pressure may be provided by the
loading through deck beam bearings, see Figure 7.6.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

The code rules for the anchorage of bars at simply supported ends can be significantly
less severe than simple first principles calculations would suggest, because they allow
empirically for bearing clamping effects.

The increase in bond strength due to transverse clamping pressure depends on a number
of different factors including:

0 Depth of cover (the effect is greater as the cover reduces)

0 Whether the concrete is delaminated under the bearing
0 Concrete strength
0 Type of reinforcement - plain round or deformed
0 Bar diameter and the amount of transverse pressure
0 Containment of the bars by links and side cover
0 Type of bearing - rigid or flexible.

References 7.3 to 7.6 give results of work carried out on this subject and which were
used to derive rules for the draft revised BD44 dealing with deteriorated structures.

The effect on bond of transverse pressure where the concrete under a bearing is
delaminated is more significant than for non-delaminated cases but the unclamped
delaminated bond value is very low. For clamped delaminated cover over Type 2
deformed bars in tension, the bond strength derived from clamped pull-out tests a t the
r ~ ~be found as follows, provided the side cover is not
University of W e ~ t m i n s t e can
Figure 7.6
Bearing clamping forces and application susceptible to failure:
to Regan's method.
(Example based on Figure 7 1)

Bearing
clamping
force
.. 7Reinforcement capacity

Force capacity of bar limited

by 0.87 X yield strength
without Applied
. . force 1 - --..- A
clamping
1 Bond from bend
A
Distance along cantilever

Force which can be developed in reinforcement

from anchorage and bond (capacity)

75
 Bond stress over the clamped length of bar is:

Where p = 0.25@ /,/feu ) but not less than 0.7

p =transverse pressure in N/mm2
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

f,, = cube strength in N/mm2

and p /,/feu > 0.05

In a delaminated anchorage length where some bars are not in stirrup corners, either the
delaminated bond stress for bars not in stirrup corners should be used for all bars
including those clamped, or the bond stress should be taken as zero for all bars or part
length of bars not in stirrup corners and not clamped.

Care must be taken, when allowing for clamping increasing bond strength in beams, that
failure of side cover does not weaken the assumed bond strength. Vertical pressure from
the clamping of the bearing can promote failure of the side cover and then reduce rather
than increase the bond strength of the sidebars.

7.2 Shear

7.2.1 Shear in reinforced The rules for the assessment of shear cover two extreme cases:
concrete slabs
0 Punching shear a t pier supports or a t points of application of concentrated loads
U Average shear across sections

They do not give specific guidance for the following problem areas:

U Concentrated loads close to free edges

0 Concentrated loads close to supported edges
0 Concentrated loads on wide one-way spanning slabs (such as slab bridge decks)
0 Shear a t obtuse corners of slab bridges
0 Shear on sections inclined to the directions of the reinforcement.

The following, however, does aid tackling these problems:

0 Unlike BS5400 Part 4, BD44 allows enhancement of shear stresses on short shear spans.
0 It is recommended generally that the distribution of shear is determined from elastic
grillage or finite element analyses. However, analyses tend to give very high local
values and the amount of averaging is not defined.
0 In most cases, it is accepted that it is reasonable to distribute the shears over a width
of up to 2 to 3 times the section depth. However, all forms of distribution or
spreading of the load must satisfy equilibrium.
0 Shear capacities on planes that are not normal to the directions of reinforcement should
be calculated using the effective reinforcement crossing the plane using cos4 0 terms.
This is based on stiffness requirements rather than strength as proposed by Clark

76
 Regan has carried out testing of reinforced concrete slabs covering concentrated loads
on free edges and half-width loading and has demonstrated reasonable correlation with
the test results using lower bound plastic analysis with checks on the corresponding
shears. Hillerborg’s method 78t0710demonstrates this approach. This gives confidence
that simple elastic analyses will give safe results.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

A reasonable approach t o the problem areas identified above, therefore, is t o carry out
grillage analyses specifically orientated t o suit potential shear failure planes with
member widths of 2 t o 3 times the section depth. This will give the shears appropriate
t o average over the section and ensure that equilibrium is being satisfied.

7.2.2 Effect O f Varying If the longitudinal section of an element varies then the relative angle between the tension
section Size and compression forces resisting bending will decrease or increase the shear carried by a
section. The effect is most pronounced on haunched members as is shown in Figure 7.7,
where the shear carried by the section is reduced by the compression force in the bottom
of the section. Similarly, simply supported fish-bellied sections reduce the shear carried
because of the tensile force in the bottom flange. However, haunches at simple supports
and reverse haunches at continuous supports can increase the shear carried, and this needs
t o be taken into account.

7.2.3 Shear enhancement a t Enhancement of shear at supports requires anchorage development lengths of 20
supports diameters according to BD44. However, tests have been carried out at Queens
University Belfast, TRL, Birmingham University and Bath University, which indicate that
intermediate values can be u ~ e d ~ ~ - ~Further
” , ~ ’ *details
. of this work are included in
Section 9.1 1. Regan’s method in combination with bearing clamping as described in
Section 7.1.8, can also be used t o justify increased shear capacities a t supports providing
links are present.

Figure 7.7
Shear carried by variable section deck. Continuous bridge deck
(also applies t o Tension force
simply supported) Inclined neutral axis in top flange

+-\- t
4
.--.--/-
I--.--- /--

I /
,;*?
- , straight haunch

Compression force
in bottom flange

77
 7 Hidden strengths

As an alternative to the method in 7.1.2, a method has been developed by Shave eta1713
that enables the shear capacity t o be calculated in cases where the longitudinal
reinforcement has a short anchorage length. The method is based on a lower bound
approach, and has been simplified in a form that would be compatible with the
general approach of 8044. However, for bridges with short anchorage lengths,
the method is much more accurate and less conservative than the existing rules in
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

BD44.

For members with shear reinforcement, the theory is based on a variable angle truss
approach. The angle of the truss 0 may be chosen such that the following inequalities
are satisfied:

V < AJYY (z/s) cot0 (ensures that the stirrup stresses do not exceed yield)

V < (bzvf,,) / (cote+ tan0) (ensures that concrete compression does not
exceed crushing limit)

V < 2[Fub- (M/z)]/cot0 (ensures that anchorage failure does not occur)

These inequalities are consistent with the approach in EN 1992, and may be optimised
t o determine the best lower bound that does not exceed the relevant limits.

The second inequality requires the concrete effectiveness factor v t o be calculated.

Various proposals for effectiveness factors have been made72, 14, ”, The approach in
Shave et alis based on the formula for effectiveness factor in Reference 7.14.

So long as the beam is not over-reinforced in shear, the solution for shear capacity
varies with the square root of the anchorage force f u b .

For members loaded a t short shear spans and members without shear reinforcement, a
strut and tie approach is used. The variation of shear capacity with anchorage force is
more complicated, but may still be approximated by a square root function up t o the
fully anchored capacity.

These models give results that are consistent with upper bound analyses, and have been
verified by a comprehensive programme of experimental testing.

BD44 suggests that a t short anchorage lengths, the shear reinforcement is deemed t o be
ineffective, and that the enhancement of shear capacity at short shear spans is not
allowed. However, the research by Shave et al indicates that this is not the case. Even a t
short anchorage lengths, the stirrups contribute t o the shear capacity and shear
enhancement occurs a t short shear spans.

The shear capacity is reduced by a reduction factor that accounts for the anchorage
length. This reduction factor is written as:

78
 - -

where a is factor to allow for enhancement of bond due to lateral pressure, and fubmax is
the required anchorage force to give the fully anchored shear capacity This is defined as

Fu, max = 65svcbd

using the notation of BD44/95 The shear capacity is then found by multiplying the
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

standard fully anchored shear capacity (including shear reinforcement and shear
enhancement at short shear spans) by the reduction factor Finally, there are some
maximum and minimum limits placed on the shear capacity, which are the same as the
existing limits in BD44/95 This method gives predicted shear capacities that have been
verified by test data, and is much more accurate than BD44 at short anchorage lengths

7.2.4 Shear in prestressed The standard equations for shear strength in BS 5400 Part 4 and BD44 are based on
flanged beams rectangular sections where the shear stress is 50% higher than the average stress Generally,
this is conservative for I section beams which are uncracked in flexure and rigorous checks
can be carried out a t the top and bottom web to flange connections as well as intermediate
points including the neutral axis, using the appropriate section properties to calculate the
principal tensile stress The code rules only check a t the neutral axis Due to the effect of
flexural tensile stress, the rigorous check can give a lower strength when the bending
moment is significant In such cases, the method does not give any advantage but the code
rules can be used since they have been justified by comparison with test results

When using this rigorous method on post-tensioned beams, then the reduction of web
width for the presence of ducts can take account of their actual position rather than a
general reduction See also below for grouted ducts

7.2.5 Shear in webs O f The reduction in effective width of webs due to the presence of ducts is specified in both
post-tensioned prestressed BS5400 Part 4 and BD44 as 0 67 times the diameter of the duct Tests have been
carried out to suggest that this is conservative providing that the ducts are adequately
beams grouted and reductions to say 50% of the diameter can provide significant benefit

7.3 Deck Surfacing The use of asphalt surfacing to provide additional stiffness and strength to bridge decks
has so far only been applied to extend the fatigue life of orthotropic steel decks A paper
on testing mastic asphalt to find a suitable mix for that purpose cites the use of an
indirect tensile strength test, an indirect tensile creep test and an indirect tensile fatigue
test7 Whether surfacing can be justified in increasing strength is considered below

7.3.1 Properties O f Surfacing Asphalt surfacing IS known widely as flexible surfacing Its success IS based on its ability
materials to flow and to seal up cracks, which form due to temperature and other weather effects
and due to the action of wheel loads Clearly, as such, it is not a rigid material - it

79
 Hidden strengths

moves under load. It does not appear feasible therefore to derive an E-value for asphall
material since creep under load is such a significant factor.

Temperature effects are highly significant, asphalt being much softer in high summer
temperatures and much more rigid in cold winter temperatures. Therefore, it is a very
variable material.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Asphalt materials vary widely. While the cube strength of concrete will largely determii
its contribution to the strength of a bridge deck, and its E-value will not vary
significantly for this purpose even as the strength varies, this cannot be said for asphalt
The tests for asphalt surfacing to determine its suitability are generally compaction
related rather than specifically strength related. Therefore, if asphalt were to be used t c
contribute directly to the strength of bridge decks, new tests would most likely have to
be devised to ensure that adequate strength and stiffness were being provided.

The bond between the asphalt surfacing and the bridge deck cannot be guaranteed.
There are many cases of surfacing peeling away from the deck and therefore there
would be a question of reliability if the surfacing were to contribute to the integrity of
the bridge deck. The content and laying conditions (temperature and humidity) of bott
the waterproofing membrane and the surfacing base course will affect the bond both
between the deck and the membrane and between the membrane and the base course

7.3.2 Discussion It is unlikely that normal depth surfacing on bridge decks will be beneficial to the
ultimate strength and capacity. There may be serviceability benefits as for orthotropic
decks but these can be judged by performance on most bridges.

7.4 Moment field analysis

of reinforced concrete slabs

7.4.1 Moment fields If an element of a slab is subjected to a combination of bending and twisting moments (i.c
a moment field), the bending moment applied about any axis within the plane of the slab
can be determined from equilibrium considerations. The applied moment field is given by:

M, = M,cos' a + M, sin' a
Where M ,and M, are the principal moment^^'^,^^^ and a is the direction relative to th
principal moment M,.This applied bending moment field is illustrated in Figure 7.8,
using a polar co-ordinate plot.

Similarly, the moment capacity of the slab will vary depending upon the axis about
which it is calculated. Assuming that the reinforcement orientated in different directior
acts independently, which is reasonable provided the slab is lightly reinforced and the
reinforcement is not heavily skewed, the resistance moments vary by a similar
relationship to the applied moments, as follows:

M*$ = M*,COS' $
 Figure 7.8
Applied moment field.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Where M*, is the resistance in the direction of the reinforcement and 4 is the angle t o
the direction being considered.

Again, assuming the reinforcement in different directions acts independently, the

contribution from all the reinforcement directions can be combined t o give the
Resistance Field by superposition. This approach is illustrated in Figure 7.9.

7.4.2
Wood-Armer When assessing slabs using grillage or finite element methods, i t is necessary t o
equations demonstrate that the bending moment capacity about any axis in the plane of the slab
exceeds the moment applied about that axis. In the design of reinforced concrete slabs
this requirement is satisfied generally through the use of the Wood-Armer equations
(Figure 7.10)7'9,720.The equations ensure that the capacity of a slab is not exceeded in
flexure by an imposed loading, whilst minimising the total amount of reinforcement
required. However, the use of these equations for assessment leads generally t o a

Figure 7.9
Resistance moment fields.

81
 Figure 7.10
Wood-Armer resistance field.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Wood's
resistance

conservative estimate of structural capacity in all cases where there is spare capacity in
one of the reinforcement directions at the critical point. Ignoring this can lead t o highly
conservative assessment ratings.

7.4.3 Moment field approach Alternative approaches are available, based on the same fundamental principles as the
Wood-Armer equations, but which are directly applicable t o assessment; see
Denton717, '*. These directly compare the applied moment field with the resistance field
of the reinforcement provided in the structure and avoid the conservatism of the
Wood-Armer equations, see Figure 7.1 1. The LEAP5 post processor RASP also follows
the same principles but a simple computer program or spreadsheet which checks the
applied moment field against the resistance field in all directions can also be used.

7.4.4 Torsionless grillages As a further alternative, a torsionless grillage can be run based on the actual
reinforcement directions and the stiffnesses varied iteratively if necessary, in order t o
satisfy both strength and equilibrium as discussed in Section 6.4.2.

7.5 COmpreSSiVe Tests on slabs restrained around the edges have recorded strengths, which are far
membrane action greater than predicted by conventional flexural analysis. The strength is also far less
sensitive t o reinforcement area so that the ratio of actual strength t o conventional
prediction, which typically is around four with conventional amounts of reinforcement,
but can be ten or more with very light reinforcement. The reason has been identified as
the presence of inclined compressive struts in the concrete providing an arch or dome
effect. The phenomenon is called Compressive Membrane Action. Although there are a

82
 .. ... - --.. .__7r__1__-__

Figure 7.11
Resistancefield which demonstrates satisfactory
resistance. Wood’s moments moment field
(but ‘failure’ to comply with Wood-Armer
resistance requirements)
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

I moment field I Imoment field I

number of references that explain the principles, the methods are under used because of
the lack of usable methodologies and guidance. However, the Highways Agency has
produced guidance on the ~ubject’~’. Key aspects will be considered here.

7.5.1 Local strength O f deck BD81 is primarily concerned with the deck slabs of beam and slab type bridges. It gives a
slabs and restraint simple way of assessing the local strength of the slab based on work by Kirkpatrick,
Rankin and Long722,

An even simpler alternative is to use charts in the Ontario Highway Bridge design
code723.Using either method, for most practical cases, the strength will be very high so
that its precise value is not important. The major practical restriction on these
essentially empirical methods is that they depend on nominal rules to ensure that the
restraint required to develop the compressive force is adequate. BD81 imposes
restrictions on span, span to depth ratio and edge details as well as requiring support
diaphragms. These rules are necessarily conservative.J a ~ k s o n ‘ for
~ ~ ,example, has shown
that the effect can still work without support diaphragms. However, there is no simple
way of assessing the strength of deck slabs that do not comply with the rules. The only
way is to use some form of non-linear numerical analysis. Non-linear finite element
analysis has proved able to model the behaviour well in some cases. The predictive
capabilities of non-linear numerical methods is less well established than their use in
analysing experimental findings. Their use in assessment therefore requires specialist
expertise. However, at the present state of the art, it is by no means foolproof and great
care is required. Key relevant aspects are considered in 7.5.4.

83
 7.5.2 Global behaviour O f Compressive membrane action has the effect that deck slabs, even very lightly
beam and slab decks reinforced ones, behave as though they were heavily reinforced. They have both the
strength and the lack of ductility of heavily reinforced slabs. Because of this, the ‘global
transverse moments’ induced in the slab by its action in distributing load between
beams cannot necessarily redistribute away from the areas where there are high local
moments. It has been found7” that this can greatly reduce the local strength of deck
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

slabs. This will not be a problem normally in bridges that have intermediate diaphragms
or cross frames as these take the moments and minimise the effect on the slab. If the
main beams are very stiff and have large reserves of strength, the effect is not likely t o
be significant either. BD81 allows the effect t o be ignored if the main beams are
prestressed and comply with the serviceability requirements of BD24. For all other cases,
it is necessary t o consider the influence of global effects on the slab and the only way of
doing this is again with non-linear numerical analysis. In assessment, it is important t o
realise that this includes cases where the prestressed beams have adequate ultimate
strength but, for example, are found t o be Class 3 under service loads.

Although it has been suggested that compressive membrane action could also contribute
t o a slab’s ability to resist global transverse moments, this has not been proven and there
is evidence that the effect may not be developed a t all. It is recommended therefore that
the slab should be checked in the normal way t o ensure that the reinforcement is
adequate t o resist these moments. BD81 suggests that the moment could be obtained
from a conventional uncracked grillage analysis. However, it says that in the analysis t o
determine the moment in the beams, the stiffness of the transverse slab members should
be halved t o allow for cracking. This would require two separate analyses. A conservative
alternative t o avoid this is t o undertake only the analysis with the halved transverse
stiffness but t o check for double the transverse moments this gives. Although this is
conservative, it will often still show slabs t o be adequate.

7.5.3 External restraint BD81, and the majority of the extensive research on compressive membrane action in
bridges, concentrates on structures where the restraint is internal, coming from the
structure surrounding the critical areas. Moreover, it is possible t o get significant
advantages from external restraint from abutments. This can increase significantly
the strength of concrete slabs and sometimes beam and slab structures. This effect,
which is sometimes known as flat arch action, has been investigated by Das7” and by
Jackson726.

It has been found that the actual enhancement is often substantial. However, in
assessment, it is necessary t o use a conservative estimate of the stiffness of the available
restraint. This will give a far lesser enhancement but it can still be significant. Primarily, it
is likely t o be of benefit in older reinforced concrete bridges where there was often no
provision for expansion. More modern structures, at least prior t o the return t o integral
bridges, are provided normally with expansion joints and are less likely t o benefit. An
exception, however, is culverts and portal type structures. Even where the restraint
stiffness is relatively low, there may be a permanent compression in the deck due t o its
action in propping the abutments.

a4
 As well as enhancing the global longitudinal capacity, restraint can greatly improve the
distribution properties by increasing the transverse bending capacity. This has proved
particularly beneficial in filler beam727
and other bridge types where the transverse steel
can be very light or even non-existent. Again, in assessment, it is necessary to use
conservative estimates of the restraint which may only come from friction. However,
-- the best predictions of test results have been obtained with what are essentially upper
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

bounds to likely

7.5.4 Non-linear numerical For all but relatively simple cases, non-linear numerical analysis is the only way to obtain
analysis realistic analyses allowing for compressive membrane action. It is not possible to
consider all the difficulties of non-linear analysis here and only aspects related
specifically to the problem considered are discussed.

When considering the behaviour of restrained slabs, the biggest variable is nearly always
the extent of restraint. Both the strength and stiffness of the restraint are critical to the
strength of the restrained slab. Therefore, there is little point in undertaking a relatively
sophisticated analysis of a slab model using assumed restraint. Where the restraint is
internal, it is much better to model enough of the bridge to model the source of the
restraint directly. It will also be necessary to model a t least a span of the bridge when
global effects are likely to be important.

A potential problem is that the tensile strength of concrete could be a major factor in
providing the restraint. This, however, may be unreliable due to cracking caused by
anything from early thermal effects to previous applications of different load cases.
Therefore, it is safer generally to use only nominal tensile strength to ensure this is not a
significant factor.

If the slab is reasonably thin compared with its span, certainly if it is outside the limits
prescribed in BD81 for use of simple methods, the non-linear analysis should include
large displacement effects. This is because the deflection reduces the effective lever arm
of the restraint force.

7.6 Piers
7.6.1 Assessment Of piers Normally, piers can be analysed by very simple hand calculation methods. However,
when they are assessed to modern codes, they frequently appear inadequate. This is
often due to increases in the design transverse load, particularly the vehicle collision
loads for which piers adjacent to carriageways now have to be checked but also some
other loads, such as bearing friction.

For very weak piers, the assessment should take account of the loads effects due to the
instantaneous removal of one or more supports. Instantaneous load effects can be up to
twice that of a gradually applied load and these effects should be redistributed to the
remaining supports, as appropriate for the particular superstructure. Equivalent gradually

85
 applied loads should then be calculated from these redistributed and instantaneous
support reactions and the ULS load capacity of the superstructure checked under the
application of this equivalent load, with yfLtaken as 1.00.

7.6.2 Slender piers

Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Another reason for assessment failures, which applies primarily t o slender piers, is the
allowance for slenderness moments. These can also cause problems t o foundation
assessments as well as pier stems. Past practice was t o allow for buckling by using a
reduction factor on axial load, whilst modern practice is t o use added moments. The
'effective heights' used are also different, particularly for cantilever piers729.For slender
piers, conventional assessment methods are almost invariably conservative. There are
two reasons for this. The first is the way the added moment used t o allow for buckling
was derived730.This was based on a non-linear analysis, which aims t o predict the
deflection of the pier at failure. The added moment is simply this deflection multiplied
by the axial load. The analysis, which was calibrated against tests, gives reasonable
predictions of the deflection atfailure of a pier which is loaded t o failure under
displacement control. However, the peak load is reached a t a much smaller deflection.
This is illustrated in Figure 7.12, which shows a load-deflection plot for an actual pier
that was assessed using a non-linear analysis.

To determine the actual capacity of a slender pier subject t o axial load and applied
moment and/or transverse force, it is possible t o perform a non-linear analysis of the
pier. The non-linear analysis must model both the geometric (large displacement) non-
linearity and the material non-linearity of the reinforced concrete section. Commercial
software is available t o model both of these. An alternative to using a package which
includes material non-linearity is t o perform an iterative analysis in which the displaced
shape is updated at each iteration t o take account of large displacements and the
section properties are revised as elements become plastic. The iterations are continued
until the displacements converge. This is tedious however and time-consuming. The

Figure 7.12
Non-linear analysis of a slender pier.
Safe
Applied deflections

Safe _,_ Unsafe Applied loading

Transverse
force
F

max

Concrete Concrete
cracks crushes

Transverse deflection A '

86
 column should either be modelled with an initial bow or the load should be given an
initial eccentricity t o account for the actual pier construction tolerances. The
eccentricity could be obtained from the assessment standard or it could be obtained by
precise survey of the pier.

The example shown in Figure 7.1 2 illustrates the benefit of a non-linear analysis. A
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

simple cantilevered pier is subject t o both axial load, P, and a transverse load, f , at the
top. The load P is applied as a constant load and the top of the pier is deflected under
deflection control in small steps.

Initially, the behaviour is linear as the concrete is approximately linear in compression.

As the force F increases, the concrete cracks in tension and its tangent modulus in
compression progressively reduces. Initially, the moment resistance of the section
increases faster than the increase in moment resulting from the growing eccentricity of
the load. Consequently, the load f increases with deflection. However, as the pier
becomes more flexible, there comes a point at which the rate of increase of moment
resistance equals the rate of increase in the moment resulting from the growing
eccentricity. This represents the peak in load F on the above graph.

A real pier with this combination of applied axial and transverse load would fail at this
peak load point. The modelled pier loaded under displacement control, however, would
exhibit a drop in transverse load f until final failure occurred when the concrete reached its
crushing strain. This final failure point is likely t o be close t o the failure load predicted by
BD44. Depending on the pier geometry and reinforcement, the real failure load at the peak
of the graph can be considerably higher than that predicted by BD44. A partial factor of
safety appropriate t o upper bound methods should be used with this method of analysis.

An example of the use of this method gave a peak load 70% greater than that given by
a conventional assessment. Because of this, the analysis saved many thousands of
pounds in strengthening works and associated disruption cost. This provides another
illustration of how analytical techniques, which are considered too expensive t o use in
design, can be highly economic in assessment.

7.6.3 Cloba restraints When piers appear t o be 'unsafe' due t o lateral loads, a careful consideration of the real
articulation of the bridge is often revealing. For example, consider a simple two span
bridge, which is fixed a t the central pier but free a t both ends. Initially, most engineers
would assume that any longitudinal load, that is load parallel t o the span, which was
applied t o the pier or the deck would be resisted by flexure in the pier. However, bearing
friction on the end supports would be reversed and considerably reduce the force
applied t o the piers. See Figure 7.1 3.

If this force is exceeded, there has t o be a very significant displacement of the pier top
and the deck For the pier t o fail. This would almost certainly exceed the movement
available a t the expansion joint and a more realistic analysis would assume the load went
via the curtain wall into the soil. The pier could then be checked for this amount of

a7
 Figure 7.13
Pier buckling restraints. Longitudinal force - braking, traction, wind
-f
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

-F
I Conservative
approach
R R
-F

-f
I I Realistic
e- approach
ILR
(friction)
-f - 2pR
t
-f Joint closes 4bA
I Ultimate
R
I
7 case

Pier checked for A deflection Abutment force

(ignore resistance developed) oppose by soil
resistance

displacement. See Figure 7.13. If the design load was applied to the pier rather than the
deck, the connection between the pier and the deck may fail first. However, it may be
cheaper and easier t o strengthen this connection rather than t o strengthen the pier itself

7.6.4 Buckling O f multiple When a structure has more than one fixed pier then the buckling of the combined
piers system needs t o be considered. There is a simple relationship for the elastic bucking of
multi-leg systems: instead of the buckling of a single column being governed by:

P / PEulei (where PEuler

= Euler buckling load)

multiple pier buckling is dependent on:

There is a need therefore t o check the global behaviour of the combined system and the
local buckling of individual piers. The local buckling capacity of individual columns is
then governed by fixed ended buckling rather than sway buckling. However, this will
occur a t about four times the sway buckling load and is very unlikely t o be a problem.

88
 For global buckling of the pier system, the non-linear behaviour has t o be taken into
account. This approach is exactly the same as that described in Section 7.6.2 except that
the model is more complex as it should take account of all the piers and the
combination of eccentricities that may occur, see Figure 7.14.

Figure 7.14
S S F F F S S
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Stability of multi-pier supports.

I
F - Fixed
S - Sliding
U U U U U

Friction forces Buckling restricted

may reverse due by all fixed piers
to temperature and friction on
System stability sliding bearings

An added complication is that friction on the sliding bearings of multi-column bridges

can both resist and assist buckling. Thermal expansion and contraction combined with
braking and traction loads need t o be taken into account in the most adverse direction
by considering worst-case sequences of loading and effects. This can be carried out by
hand methods rather than non-linear analysis t o obtain the loads resisted by the
individual piers and then these applied using the standard (conservative) rules. If these
are not sufficient then it will be necessary t o carry out a non-linear analysis.

7.7 Redundancy Of Structures have redundancy occasionally which is not allowed for in straightforward
e [ements analysis.

For example if a bridge deck beam is damaged, the adjacent beams may be able t o carry
the load shed by the damaged beam distributed by the surrounding transverse beams. It
can be advantageous t o consider alternative load bearing mechanisms or secondary load
paths in bridge systems. See also Section 6.1. I .

Piers can be vulnerable t o severe damage if subject t o heavy vehicle impact. However,
the removal of a single column or pier may not result in collapse of the deck.
Reassurance may be obtained by modelling the deck without the support of the
vulnerable column or pier. The structure may be shown t o support its own dead weight
in these circumstances, indicating that catastrophic collapse may not be inevitable.

7.8 Parapet edge Stiffening Parapets often act t o stiffen the edges of decks and can be included in the analysis
provided there is sufficient longitudinal shear connection t o ensure that they act
compositely with the structure.

89
 7.9 Width O f SUppOf?.S The finite width of bridge piers is a useful means of limitingpeak moments in bridge beams
and slabs supported monolithically by columns and piers respectively (see Figure 7.15).

Figure 7.15
Effect of finite width of supports.
WD
M, = -
3r
For circular bearing area
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

WD
M, = -8
For rectangle bearing area

I
ri”
Dispersion through concrete
does not satisfy equilibrium dispersion through
- ignore - bearing plate
W
Reduction in peak support moment
due to finite width of support

7.10 Foundations The forms of failure that occur in these are likely to be visible. Because of this, normally,
they are not analysed in routine assessments unless either the structure is showing
visible signs of distress or there are reasons for believing that the load on it has
increased significantly since design. An example of the latter that has occurred was a
retaining wall; it appeared that a multi-storey building had been built close behind it
without anybody having investigated the consequences to the wall.

Substructure design is even less of an exact science than superstructure design. Until
quite recently, many earth retaining structures were designed only for active pressure
which, under current rules, would be designed for at-rest pressure. This has the effect
that a simple check to modern standards is likely to fail many structures. However, the
observed condition is often sufficient to prove that at-rest pressure is not experienced. It
will often be helpful to use lower pressures in assessments and, usually, geotechnical
specialists can justify this for specific structures.

When calculations suggest that abutments are inadequate, it is always worth reviewing
the way they would behave if they moved. Normally smaller bridges decks, particularly
slab bridges, will work quite satisfactorily when they are propping the abutments, even if
they were not designed to work that way.

90
 7.11 References 7.1 RECAN, PE, Shear, Current Practice Sheet No 105, Concrete, November 1985, pp 25-26.
7.2 BRITISH STANDARDS INSTITUTION, BS EN1992-1,Designofconcretestructures, Part 1 CeneraiRuiesandRulesfor
Buildings, BSI, London, 2004.
7.3 BATAYNEH. MK, The effects of lateral compressionon bondbetweendeformedreinforcing bars andconcrete, Doctoral
Thesis at Oxford Brookes University. November 1993.
7.4 UNTRAUER, RE and HENRY, RL. Influence of normal pressure on bond strength,ACi/ournal, May 1965, pp 577-585
7.5 NAVARATNARAJAH,V and SPEARE, PRS, An experimental study of the effects of lateral pressure on the transfer bond of
reinforcing bars with variable cover, Proceedings institution ofcivil Engineers, Part 2, Vol 81, December 1986, pp
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

697-175.
7.6 RECAN, PE, Jests ofdelaminatedanchoragessubiect to transverse pressure, University of Westminster, March 1999.
7.7 CLARK. LA, BALDWIN. MI and CUO, M, Assessment of concrete bridges with inadequately anchored reinforcement,
Bridge Management3,(Harding. JE. Parke. CAR and Ryall. MJ,eds), E&FN Spon, London. 1996. pp 225-232.
7.8 HILLERBORC,A, Strip MethodofDesign, Viewpoint Publication, 1975
7.9 HILLERBORC. A, The Advanced Strip Method - A simple design tool, MagazineofConcreteResearch, Vol 34, No. 121,
December 1982.
7.10 HILLERBORC, A, StripMethodDesignHandbook. E and F Spon 1996
7.11 CLELAND, D, Tests on supports ofsiab bridge decks with inadequateanchorage,Report for the Roads Service, N Ireland,
Queens University Belfast
7.12 CULLINCTON, DW, DALY, AF and HILL, ME, Assessment of reinforced concrete bridges. Collapse test on Thurloxton
Underpass, Bridge Management 3, (Harding, JE, Parke, CAR and Ryall, MJ, eds), London, E&FN Spon, 1996. pp 667--674.
7.13 SHAVE, JD, IBELL, TJ and DENTON, SR, A new assessment model for shear in reinforced concrete bridges, Bridge
Management 5 (Parke, CAR and Disney, P, eds), London, Thomas Telford, 2005
7.14 IBELL, TJ, MORLEY, CT and MIDDLETON, CR, A plasticity approach to the assessment of shear in concrete beam and slab
bridges, The StructuralEngineer, Vol 75 I No. 19, 1997, pp 331-338.
7.15 NIELSEN, MP. limitAnalysisandConcretePlasticity, 2nd ed , CRC Press, 1999.
7.16 .MOHAMMED. LN and PAUL, HR, Design of SMA Mixture for an Orthotropic Bridge Deck Resurfacing,ProceedingsRoads
96 Conference. New Zealand, 1996.
7.17 DENTON, SR, The assessment of reinforced concrete bridge decks, Bridge Modification 2. Proceedings oflnternational
Conference, (Pritchard, B, ed.), Thomas Telford Ltd, London, 1997, pp 40-54
7.18 DENTON, SR and BURCOYNE. C, The assessment of reinforced concrete slabs, TheStructuralEngineer.Vol 74, No 9.
May 1997, pp 147-152
7.19 WOOD, RH, The reinforcement of slabs in accordance with predetermined field of moments, Concrete, Vol. 2, No. 2,
February 1968, pp 69-76
7.20 ARMER, CST, Discussion of the above, Concrete, VoI 2, No. 8, August 1968. pp 319-320
7.21 HIGHWAYS AGENCY, BD 81/02, Use ofcompressivemembraneaction in bridgedecks, May 2002, (Corrected August
2002) Design Manual for Roads and Bridges. Vol. 3, Section 4, Part 20.
7.22 KIRKPATRICK, J, RANKIN, GIB and LONG, AE, Strength evaluation of M-beam bridge deck slabs, The Structural Engineer,
Vol 626, No 3, 1984, pp 60-68.
7.23 ONTARIO MINISTRY OF TRANSPORTATION AND COMMUNICATIONS, Ontario HighwayBridge Design Code,
Downsview. Ontario, Canada, 1983 (and later editions). p 175.
7.24 JACKSON,PA, The global and local behaviour of bridge deck slabs, JheStructuralEngineer, Vol 68, No. 6, March 1990,
ppll2-116
7.25 DAS, PC, Load carrying characteristics of flat arches and their implications for the design assessment and strengthening
of bridges, Proceedings offifth internationalconference on structuralfaults and repair, Engineering Technics Press,
Edinburgh, 1993, pp 315-320
7.26 JACKSON, PA. Flat arch action, Arch Bridges. Proceedingofthe first internationalconference on arch bridges, Thomas
Telford, London, September 1995, pp 407-415
7.27 LOW, AMcC and RICKETTS, NJ,Theassessmentoffillerjoist bridges without transverse reinforcement, Research Report
Special 383, Crowthorne, Transport Research Laboratory, 1993.
7.28 JACKSON. PA The analysis and assessment of bridges with minimal transverse reinforcement, Bridge Management 3
(Harding, JE, Parke, CAR and Ryall, MJ,eds), University of Surrey, April 1996, E&FN Spon, pp 779-785.
7.29 JACKSON, PA, Slender concrete bridge piers and the effective height provisions ofBS 5400: Part 4: 1984, Publication
42 561, Cement and Concrete Association (now British Cement Association), Camberley, June 1985
7.30 CRANSTON, WB, Analysis anddesign ofreinforcedconcrete columns, Publication 41.020, Cement and Concrete
Association, (now British Cement Association), Camberley, 1972.

91
 63

8. Specific structural forms

Chapter 8 considers particular difficulties in assessing the loading capacity of a number

of common structural forms. These are:
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

0 Reinforced concrete slabs, commonly used as both bridge decks in themselves and
supported by steel or reinforced, prestressed/post-tensioned beams. They are
frequently assessed as having inadequate capacity under local wheel effects.
0 Reinforced concrete beams and slab bridges decks suffer from problems of inadequate
shear and anchorage bond at the ends of the beams. Care needs t o be taken that
what appears t o be a reinforced concrete beam is not a concrete encased steel beam
such as the 'Preflex' type manufactured by Boulton & Paul in the 1960s.
0 Half Joints, used particularly in bridge construction t o provide a span greater then the
span provided by standard prestressed concrete beams. The detailing of the half joint
reinforcement is important in the assessment of these joints. Hinge joints have similar
problems t o those of half joints.
0 Box culverts are a common form of construction for culverts and subways. They are
often assessed as inadequate t o carry the imposed loads despite no visible signs of
distress. Such culverts may be in situ or more often of precast concrete construction.
They are sensitive particularly t o analysis of the application of earth pressures and the
support t o the base slab.
0 Precast pre-tensioned beams have and continue t o be used t o provide bridge decks
for spans of up t o about 30 m. They have been a very successful form of construction
but early designs suffer in particular from an absence of shear reinforcement.
0 Arches are more commonly constructed in masonry, however arches are also
constructed in both in-situ and precast reinforced concrete and in mass concrete. The
mass concrete and lightly reinforced concrete arches rely upon maintaining
compression throughout the section as with a masonry arch. More heavily reinforced
sections are designed t o carry load in both tension and compression and the method
of analysis should take notice of this. They include a wide variety of structural forms.
U Post-tensioned structures have particular features that separate them from other
structural forms.

Each of the above is considered, particular problems are described with reference t o
other sections of this Guide and additional information is given on how t o consider the
issues.

8.1 Reinforced ConCrete Reinforced concrete slabs are used frequently as bridge decks themselves for small span
slabs structures and as deck slabs typically supported on steel or reinforced, prestressed/post-
tensioned concrete beams. The large number of structures that conventional
assessments fail t o show t o have adequate load capacity but which, in practice, show no
signs of distress indicates that there are hidden reserves of strength that these methods
do not account for.

92
 Guidance on appropriate methods of analysis is given in Chapter 6 and areas o f hidden
strength are given in Chapter 7. Advice on in-situ and laboratory testing t o provide
improved understanding of the material properties is given Chapter 4 and advice o n load
testing is given in Chapter 10. Further advice on in-situ and laboratory testing is given in
the Concrete Bridge Development Group publication on testing and monitoring the
durability of concrete structures”.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

The key problems areas are typically:

1. Limited reinforcement and inadequate anchorage lengths

2. Cantilever slabs failing in flexure under accidental wheel loading
3. Inadequate transverse reinforcement
4. Inadequate anchorage for shear enhancement at the ends of decks
5. Local wheel effects on footways and verges.

Possible solutions:

1. Alternative approaches t o the problem o f inadequate anchorage length are discussed

in Section 7.1.
2. The assessed capacity o f cantilever slabs can be increased by the use of Yield- line
Analysis, see Section 6.5.
3. The apparent problem of inadequate transverse reinforcement is often as a result of
the use of uncracked linear elastic grillage and finite element analysis. Possible
solutions are discussed in Sections 6.4 and 6.5.
4. Inadequate anchorage for shear is discussed in Sections 7.1 and 7.2.
5. Analysis of local wheel effects on footways and verges are often undertaken using
Westergaard or Pucher Charts. These methods are discussed in Section 6.3.3.
Improvements can be achieved by the use of yield-line or finite element analyses.

8.2 Reinforced ConCrete Advice on the analysis for decks including reinforced concrete beams is given in Chapter
beams 6 and on the hidden strengths that may be present in Chapter 7. Key problem areas are
inadequate anchorage and bond and shear at the ends of beams, and failure of early
beams in both flexure and shear discussed in Sections 7.1 and 7.2. In some beams there
is a deficiency in the amount of longitudinal reinforcement, which results in the shear
links being ineffective.

Some early bridge beams used the Hennibique system of reinforcing82.In this system,
50% o f the bars were bent up at about the 1/3rd point of the span t o enhance the shear
capacity at the end of the beam. The design was not based upon current reinforced
concrete theory. Later designs have also used the concept of bent-up bars. Advice on the
calculation of stress in bent-up bars is given in Section 7.1.6.

93
 8.3 Half joints Half joints were introduced into bridge decks as a means of simplifying design and
construction operations. This form of joint is vulnerable to deterioration in the event of
deck expansion joint failure, where chloride-rich seepage through the joint can cause
corrosion of the reinforcement and concrete deterioration. Loss of reinforcement section
through corrosion, or associated concrete spalling can induce higher stresses and reduce
significantly the safety margins expected of serviceable structures. Half joints are a
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

particular concern because they are not easily accessible for inspection or maintenance
and they are mostly located over or under live traffic lanes. They are particularly difficult
to repair.

Bridges owned by the Highways Agency (HA) incorporating half joints are well
distributed throughout England. Many have been subject already to visual inspection,
and will have been prioritised for maintenance by the Agents working on behalf of the
HA on the basis of their external condition. Some may already have been repaired
and/or strengthened. An Interim Management Strategy for all structures of this type has
been set out by the Highways Agency in IAN53/0483.

This document sets out a risk-based strategy which is necessary to ensure that all
structures of this type, which are vulnerable particularly to deterioration and difficult to
inspect, are recorded, specially inspected, and remedial works planned, and to allow the
future maintenance funding requirements to be identified.

Assessments should be based on assumed levels of reinforcement corrosion because the

actual level of deterioration is very difficult to establish. This sensitivity analysis is then
used in the risk assessments.

8.4 Hinge joints Hinge joints have similar problems to half joints and can occur in decks and columns.

In particular, deck hinge joints are vulnerable to deterioration as they often are
restrained from moving by the adjacent supports and shrinkage causes a crack to open
up a t the joint. De-icing salts can penetrate directly to the reinforcement and cause
pitting corrosion. An Interim Management Strategy for hinge joint structures has been
issued by the Highways Agency (IAN51/0384)which has a very similar risk-based
approach to the IAN for half joints. A draft BA is in preparation to supersede the IAN but
currently is not ready for issue.

As for the half joints, assessments should consider different levels of reinforcement
corrosion to determine the sensitivity to deterioration. Fatigue of the reinforcement is a
concern especially when the hinge bars crossing the joint are corroding. It is possible to
carry out intrusive investigations by water jetting short sections oftthe joint but the
variability to deterioration is still difficult to estimate.

Hinge joints in piers are less vulnerable but are often in a severe environment in splash
zones adjacent to carriageways. The principles for assessment are similar to that given in
IAN51/03 which is to use strut and tie models.

94
 8.5 BOX CU1VertS The inspection of box culverts reveals structures that are generally in a good condition
with little signs of distress. However, the initial assessment of such structures often
suggests that they have inadequate load capacity.

Assessment failure is generally due t o bendinghhear failure of the top slab and t o
inadequate detailing. The consideration of the following may provide an increased
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

assessment capacity:

0 Shear enhancement at the corners due t o the presence of the chamfers.

0 The calculated shear strength may be enhanced using methods given in Section 7.2.
0 The lateral distribution of load through the roof slab may be more realistically
modelled using yield-line theory. This has been proven t o provide significant increases
in capacity. Details on the application of yield-line theory are given in Section 6.5.4.
0 The use of membrane action within the roof slab can often provide significant
increases in strength; guidance is given in Section 7.5.
0 For culverts with no fill above, the deck surfacing may provide additional stiffness t o
the deck slab. Whilst this may not provide an increase in capacity at the ultimate limit
state, it may benefit the serviceability limit state; although in practice it is found that
it is difficult t o develop this benefit. Further discussion of this is given in Section 7.3.
0 The fill material may arch over the culvert thereby relieving some of the live loading.
Further advice is given in Section 6.8. It should be noted that Clause 4.2 of BD31/0Ia5
includes for an additional 15% superimposed dead loading t o be included t o allow for
negative arching effects of the fill material.

BA55/00 Section 386 provides further advice on the assessment of buried box type
structures particularly in relation t o the application of earth pressure.

Further advice on the problems of inadequate detailing particularly in relation t o

anchorage and bond, which are the most common problems, is given in Section 7.1.

Advice on highway loading effects that may influence the assessed capacity is given in
Chapter 5.

The spring constants used t o model earth pressures in computer frame models can have
a significant effect on the moments and shears derived. These or the stiffness of the
bottom slab should be reviewed and the sensitivity of the model t o changes in spring
constants considered. Some assessment engineers have measured the deflection of the
culvert roof slab under known load and then back calculated the spring constants t o
produce the same deflection in the model.

Due t o the difficulties of estimating the soil pressures the measurement of the stress in
the bars using blind hole or stress relaxation technique^^','^ could provide information
on the real stresses present in the structure, although both methods should be treated
with caution.

95
 8.6 PreCaSt pre-tensioned Precast pre-tensioned beams by the nature of the design principles and the process of
beams manufacture have proved generally t o be very durable. The difficulties associated with
the assessment of such structures are often an absence of information regarding the
prestressing present and calculated shear failures.

Often prestressing details are not included in the bridge records and then it is necessary
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

t o establish the strand patterns. This can be achieved by breaking out the ballast wall of
the end spans and exposing the beam-ends. The strands can then be seen on the beam-
end face usually covered with bitumen paint, which can be removed by wire brushing or
water jetting. The diameter and pattern can then be measured and presence of
debonding identified. The extent of the latter cannot be established easily without
resorting t o radiography techniques but, generally, the beam capacity is not particularly
sensitive t o the debonded length.

The shear behaviour of precast pre-tensioned concrete beams is complex and is still not
fully understood. Shear strength in BS 5400 and BD44 is defined using two parameters,
V,, and V,,, V,, relates t o failure when the principal tensile stress in the web exceeds the
ultimate tensile strength of the concrete, whilst V,, involves combined bending and
shear. The values given in both BS 5400 and BD44 for V,, are empirically determined.
The frequently used form of construction, using closely spaced beams, means that
evidence of shear problems are not visible on site with the exception of the external
beams which carry reduced live loading.

Early precast pre-tensioned beams often contained no shear reinforcement and this was
adopted in BS 5400 Part 4 1978 which allowed the use of beams with no reinforcement
under specific circumstances, in particular that the applied shear was less than 50% of
the shear resistance. In some early pre-tensioned beams, particularly Tee beams, the
spacing of the shear links is too great for them t o be considered effective.

Tests were undertaken by the Transport and Road Research Laboratory on a number of
precast pre-tensioned beams with no link r e i n f o r ~ e m e n t ~These
~ , ~ ~indicated
. that the
beams possess reserves of strength due t o the conservative assumptions in the method
of calculation and the high strength of the concrete. The tests also indicated that
cracking due t o overload would have been visible on the bottom flanges before failure as
bending cracks appeared first. These would only be visible however upon close
inspection. As noted above, the web cracks would often not be visible due to the close
proximity of the beams or concrete infill to provide composite construction. The shear
resistance as determined by the tests was at least twice that calculated. Failure of an
individual beam is unlikely due t o the ability of the deck t o redistribute load transversely.
This may be further enhanced by the use of transverse stressing of the beams, a practice
particularly associated with railway overbridges constructed as part of the electrification
programme in the 1960s, although in many of these decks inadequate waterproofing
means that the transverse stressing has now failed.

Some precast pre-tensioned decks may also be constructed as shear key decks where
interlocking reinforcement protruding from adjacent beams provides a shear key
between the beams when the concrete is placed. Careful consideration is required in the

96
 analysis of such decks t o model the shear keys and in the determination of the adequacy
of the shear keys. There are no specific guidelines for the assessment of such decks and,
therefore, reference is required back t o the design guidance in particular BE238’0. Books
on bridge deck analysis by Hambley”’ and others give advice on the analysis of such
decks.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Further tests for shear in under-reinforced, precast pre-tensioned concrete beams were
undertaken in 2002 at the Transport Research Laboratory (TRL) for Railtrack. This
included the assessment of infill decks comprising precast pre-tensioned beams placed
side by side with infill concrete placed between them and over them t o form a solid
slab. The existing code requirements in BA44 are based upon previous tests undertaken
at TRL. These allow the assessed shear capacity of infill decks t o be based upon the
arithmetic sum o f the shear capacity of both the beam and the infill concrete as
calculated separately. Generally, the determination of shear strength is based upon
the minimum percentage steel area (0.15%) whether the section is reinforced or
not. Information Sheet No 2g812issued by Railtrack for use on Bridgeguard 3
assessments allows shear enhancement t o be considered when calculating the shear
capacity of the infill concrete even when no reinforcement is present. This document
also provides a justification for the use of short shear span enhancement in pre-
tensioned beams.

8.7 Reinforced Concrete Reinforced concrete arches include a wide variety of structural types. Principally, they
arches are three hinged, t w o hinged or fixed ended. They may have open or solid spandrels. The
arch may support a bridge deck above via columns or transverse walls. The bridge may
comprise a series o f parallel ribs supporting a bridge deck above. Some reinforced
concrete arch bridges are constructed as bow string girders with the deck resisting the
horizontal thrust from the arch. Where the deck or carriageway is supported on fill
above the arch they are termed closed or solid spandrel, other types are classified as
open spandrel.

BA55/00 advises that reinforced concrete closed spandrel arch bridges should be
assessed using the same requirements of other types of concrete bridges, but that the
restraining effects of the surrounding fill should be taken into account in the analysis as
appropriate.

The analysis of arches may be undertaken using either elastic or mechanism methods
Computer programs have been developed for both.

8.7.1 Elastic analySlS Hand analysis using the work of Castigliano can be undertaken t o calculate the areas of
tension and therefore cracking. The arch thickness and inertia is then reduced t o take
account of the area in tension, and the calculation repeated until balance is obtained
between the loading and the thickness of the arch not in tension. The above does not
allow for the effects of the surrounding fill t o be considered and ignores the presence of
any reinforcement.

97
 An alternative approach is t o use a structural frame program with the restraining
effects of the fill applied as spring constants. As with reinforced box culverts, the
derivation of the applied spring constants requires careful consideration as,
significantly, these may affect the derived moments and shears. It is also found that
unless the stiffness of the arch is reduced (e.g. by using cracked properties or including
hinges) this approach often gives a lower strength than treating the arch as
un-reinforced.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Soil interaction can be modelled using programs such as FLAC813 if the arch is thin and
flexible.

The simplest arches for analysis are three hinged arches as they are statically
determinate. Formulae are available in many structural textbooks to calculate the
moments and shears a t any section.

Formulae are available for calculation of two hinged arches which, like the three hinged
arch, only transmit thrusts t o the abutmentsBi4. Typically, many smaller arches forming
culverts are two hinged arches as the connection t o the abutments cannot be
considered t o provide full fixity. The abutments of such structures may also often be
inadequate t o resist rotation under load.

The effects of temperature, creep and shrinkage need t o be considered with statically
indeterminate arches. O f these, temperature is the most easily considered as the effects
of creep and shrinkage may be affected by the means of construction.

Reynolds and Steedman814suggest that shrinkage can be considered as equivalent t o a

fall in temperature of 8.3”C. BD91 gives some advice on when temperature and
shrinkage need t o be considered in masonry bridges. This advice should be conservative
for RC arches.

Formulae are available for the determination of stresses in fixed arches of any profile,
although these are difficult t o apply and generally use is made of frame analysis
programs. A particular type of fixed arch is the parabolic arch where the thickness of the
arch increases as a parabolic function from the crown t o the springing, when only the
bending moment and thrust a t the crown and springings need t o be considered and
formulae are available t o determine these.

8.7.2 Mechanism analysis Mechanism methods for the analysis of arches have been proposed particularly by
Heyman8” for masonry arches and can be applied t o reinforced concrete arches in
which the material behaves plastically. The mechanism method is based upon plastic
theory, which is discussed in Section 6.4, and the question of ductility of reinforced
concrete is discussed in Section 6.1. For reinforced concrete arches, failure will occur
on the formation of four hinges. For this t o occur the section must be sufficiently
ductile t o allow considerable rotation such that one section does not fail
prematurely.

98
 The analysis of lightly or non-reinforced concrete arches may also be undertaken using
computer programs designed for the assessment of masonry arches such as RINC816and
ARCHIE817. Both analyses give lower bound results.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

8.8 Post-tensioned Because, primarily, the design of post-tensioned structures is undertaken at the sls,
structures again, the assessment of such structures is associated generally with checking those
sections that behave as reinforced concrete slabs and beams, as discussed in Sections
8.1 and 8.2. However, care must be taken t o ensure that these sections have a
satisfactory ductility and are appropriately detailed. In particular, tendon anchorages
should be checked as reinforced concrete sections subject t o bursting forces, although
unless there is evidence o f distress such sections are unlikely t o be problematic.

8.9 References 8.1 CONCRETE BRIDGE DEVELOPMENT

GROUP, Guide to the testing andmonitoring ofdurabiiiryofconcrete bridges, CBDC,
Camberley, 2000.
8.2 MULLER, G and RANKIN, G, The Hennibique System. a Renaissance, ProceedinqsoftheInstitutionofCinlEngineers, Part
1.1991, pp 179-187
8.3 HIGHWAYS AGENCY, IAN 53-04, ConcreteHa/~-~ointDeckStructures,
February 2004.
8.4 HIGHWAYS AGENCY, IAN 51-03, HingeDeckStructures,July 2003
8.5 HIGHWAYS AGENCY, BD 31/01, The Design of Buried Concrete Box and Portal Frame Structures, DesignManualfor
RoadsandBndges,Vol. 2, Section 2, 1987
8.6 HIGHWAYS AGENCY, BA 55/00, The assessment of bridge substructures and foundations, retaining walls and buried
structures, Design Manua/forRoadsandBridges, Vol. 3, Section 4, 2000
8.7 ANDERSON, M, Dead load stress measurement, ConcreteEngineering Internationa/,July/August 2000.
8.8 CULLINGTON, D, Assessment of pre-stressed bridge beams, First InternationalConference on Bridge management,
inspection, maintenance, assessmentand repair, Surrey University, 1930, pp 447-457
8.9 CULLINGTON, DW and RAGGETT, SJ Shear Strength ofsorne 30year o/dprestressedbeamswithout [inks, Research Report
327, Transport Research Laboratory, Crowthorne, 1991
8.10 HIGHWAYS AGENCY, BE 23/71, Technical Memorandum (Bridges) Shear key decks, DesignManua/forRoadsandBridges,
Vol. 1, Section 3
8.1 1 HAMBLEY, EC, Bridge DeckBehaviour, E&F Spon, Various Editions.
8.12 RAILTRACK,Bridgeguard3,Information Sheet No 29, October 2000.
8.13 ITSACA CONSULTING GROUP INC. FLAC-Fast LangranianAnalysisofContinua, 1998
8.14 REYNOLDS, CE and STEEDMAN, JC, ReinforcedConcrete Designers Handbook (Various Editions), Viewpoint Publications.
8.15 HEYMAN, J, TheMasonryArch, Ellis Norwood, 1980.
8.16 UNIVERSITY OF SHEFFIELD, RING.
8.17 OBVlS LID UK, ARCHIE, 1999-2001.

99
 8 Specific material and assessment factors

9. Specific material and assessment factors

9.1 Introduction and This Chapter considers specific factors, which may influence the strength, stiffness, stability
general principles and serviceability of concrete bridges, and which may require consideration in structural
assessment.The sub-sections that follow may be categorised broadly as follows:
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

(A) Factors affecting concrete strength and stiffness

9.2 Alkali-Silica reaction
9.3 Aggressive chemical attack
9.4 Freeze-Thaw action.

(B) Special material issues

9.5 High Alumina Cement
9.6 Supersulfate Cement

(C) Corrosion issues

9.7 Chlorides
9.8 Carbonation
9.9 Steel corrosion

(D) Structural issues

9.10 Fatigue
9.1 1 Sub-standard reinforcement detailing
9.12 Deteriorated reinforced concrete structures

The basic purpose of assessment is to evaluate current and future structural capacity. This
is usually done via analytical methods (Chapter 6), while taking account of any hidden
strengths (Chapter 7). The process and procedures are as outlined in Chapters 1 and 2. The
accuracy of outputs from analysis is highly dependent on the quality of the input
assessment parameters. Generally, these are established during the Inspection and Testing
phases (respectivelyChapters 3 and 4). Geometrical and mechanical properties are
particularly important. In this Chapter 9, guidance is given, to augment that in Chapters 3
and 4, on how the factors itemised above might affect the assessment parameters.

By broad category, these are:

(A) Factors affecting concrete strength and stiffness.

These three items are deterioration mechanisms, which directly attack the
concrete. Stiffness can be reduced, which can influence the distribution of action
effects in the structure as a whole. Strength may also be reduced, either by loss of
section or by a reduction in mechanical properties; this would reduce resistance to
imposed loads a t critical sections. In determining the magnitude of these
reductions, it will be necessary to establish the un-deteriorated strength of the
concrete in the structure, which will have increased with time from that assumed
a t the design stage.

100
 (B) Special material issues

Some elements may be found, which are made with special cements, such as High
Alumina Cement or Supersulfate Cement. Sections 9.5 and 9.6 give guidance on
how the general performance of concrete may be modified when these special
concretes are present.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

(C) Corrosion issues

The most common form of deterioration in concrete bridges is corrosion of

reinforcement, usually due t o chlorides from de-icing salts, but also possibly from
carbonation of the concrete. This can lead t o loss of rebar cross-section, due either
t o general corrosion or local pitting; the resulting loss of strength of section (in
bending, shear, torsion, bond, etc.) will depend on the location of the corrosion and
on the quality of the reinforcement detailing. Corrosion may also lead t o a loss of
ductility in the reinforcement. Usually, there are visual signs of corrosion at the
surface of the concrete, in the form of staining, cracking, or, in extreme cases,
delamination; this latter factor can result in a loss of concrete section, and a
reduction in bond or anchorage.

(D) Structural issues

A treatment for fatigue is proposed in Section 9.10. Reinforcement detailing may

also be sub-standard, and this can increase the sensitivity of structural elements to
the effects of deterioration, particularly corrosion; this is dealt with in Section 9.1 1.
Finally, current approaches t o evaluating the structural effects of deterioration are
brought together in Section 9.12.

Considerable engineering judgement is required in evaluating the structural

effects of deterioration in individual cases. In general, damage classification and
condition assessment methods, while useful in giving a qualitative measure of
the overall state, are no substitute for quantitative methods, having either a
deterministic or probabilistic basis. Real strength is not proportional generally t o the
apparent level of deterioration; great care is necessary therefore in establishing
values for the assessment parameters, which reflect both the level of deterioration
and the sensitivity of the structural form. Particular care may be necessary where a
structure has been repaired in the past, since the effectiveness of the repair may be
questionable; frequently, conservative assumptions are necessary.

9.2 Alkali-Silica Reaction

(ASRI
9.2.1 Background Alkali-Silica reaction is a reaction between the hydroxyl ions in the pore solution and certain
forms of silica occasionally present in significant quantities in the aggregate, in the presence
of moisture in the concrete. The product of the reaction is a gelatinous silica hydrate,
containing silica, sodium, potassium, calcium and water, The gel can hydroscopically absorb
pore solution, and swell; this swelling can induce internal stresses within the concrete, which

101
 9 Specific material and assessment ffaaors

may be sufficient to cause internal cracking, expansion and visually severe cracking. The
reaction is finite, and its duration depends on the levels of alkalis present, on the size, nature
and reactivity of the aggregate, and on sufficient moisture.

The causes of ASR are now well recognised and understood, and recommendations have
existed since 1983 for minimising the risks due to it, in new construction; References 9.1
and 9.2 are typical. It is unlikely, therefore, that structures built in recent years will suffer
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

from ASR.

The correct diagnosis of ASR as the dominant deterioration mechanism is of great

importance, since, if not done, there is a risk of an erroneous assessment, and the
specification of inappropriate remedial measures. Gels can form from other deleterious
reactions, and alkali-silica gel itself can be formed without causing expansion. The level
of current and likely future expansion is crucial for subsequently assessing possible
structural effects, and therefore is an essential element in the diagnosis procedure.
Consensus guidance on diagnosis is availableg3,and for greater detail, see Reference 9.4.

Much work has now been done on the assessment of the possible structural effects of
ASR. While published recommendations may vary in detail, the principles are broadly the
same. The key to all of them is the estimation of expansion during the diagnosis stage.
Depending on the magnitude of that, methods are presented for predicting the effect on
geometrical and mechanical properties, prior to estimating the influence on relevant
action effects, such as bending, shear, axial compression, torsion, bond and anchorage.
Mostly, these procedures are well documented. The approach in this Guide is to give an
elaborate reference system in the sub-section which follows, while subsequently
highlighting important issues in current procedures or the assessment of concrete
bridges in the UK. Structural sensitivity, and the vulnerability of the resistances to
particular structural actions, are important issues in this regard.

9.2.2 StrUCtUra( aSSeSSment: The actual references are listed a t the end of this Chapter. They are grouped here under
current practice and a series of different headings, together with brief explanations.

reference documents a) General consensus guidance documents - structural assessment

ISE Reportg5.This was the first authoritative document in the UK, with a strong
dependence on estimating expansion, and emphasising the importance of detailing
sensitivity. It should be read in the wider context of the Institution’s more general report
on structural appraisalg6.

BA52/9497.This Advice Note refers specifically to Highway Structures, and comes from
Section 4, Volume 3 of the Design ManualforRoadsandBridges. It should be read in the
context of other relevant BDs and BAs (see Section 1.4), but particularly BD21/01, and
BD44/95 and 8A44/96.

CSA-A86A-00g8andCONTECVET Manualg9.These two documents introduce an

international slant, whiie updating UK p r a c t i ~ e ~ ’Both
, ~ ~ are
. based on modifying design
equations for assessment purposes.

102
 Clarkglo.This report is included in this category, since it identified the types of element most
at risk, and made recommendationsfor modifying design strength models in BS 5400. It
was the basis for much that appeared in later guidance documents listed above.

b) Research reports on the effect of ASR on various structural actions.

The eight references placed in this c a t e g ~ r y give
~ ” test
~ ~ data
~ ~ on
~ the structural
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

performance of concrete elements affected by alkali-silica reaction. The list is by no means

exhaustive, but they represent core data obtained in the early 199Os, used in developing
assessment methods for some of the guidance documents in category (a) above.

c) Diagnosis, test methods, expansion and restraint, etc.

The four references in this give detailed information on test methods
during the diagnosis stage. Of particular importance is the estimation of expansion, as
affected by restraints in the actual structure, and as measured on cores stored a t
different temperatures. This information has been integrated into the general
assessment procedure contained in Reference 9.9.

d) Synergetic effects
Where more than one deterioration mechanism is acting, one will usually dominate, and
the diagnosis phase should clearly identify which one. However, the effects that the
primary mechanism can produce may be exacerbated by defects due to other causes.
This is treated in Reference 9.9 for combinations of ASR, corrosion and freeze-thaw.
References 9.23 and 9.24 give further information.

9.2.3 Key points in StrUCtUra[ This Section merely highlights some of the important points from the reference
assessment documents in the previous sub-section. This Guide deals with Highway Structures, and
therefore, in terms of detail, the emphasis is on relevant BDs and BAs (BA52/94;
BD21/01; BA44/96 and BD44/95).

For structural assessment, the general approach will be analytical (Chapter 6) or will
involve the re-evaluation of critically loaded sections using modified design models. The
primary objective, based on input from the Inspection and Testing stages, is to devise
assessment parameters for analyses, or modifications for design models, which represent
the physical damage due to ASR and the effect of the estimated expansion. In the first
instance, this involves estimating any reductions in mechanical or geometrical
properties, which may be followed by introducing any further necessary modifications to
the design models themselves.

9.2.4 StrUCtUra[ SenSltlVity: Clarkgloindicated that the structural elements or forms most a t risk to the effects of
elements most a t risk ASRwere:

0 Over-reinforced beams
0 Columns and other compression members

I 103
 9 Speciffic materia! and assessment ffaaors

U Anchorage zones of both reinforced and prestressed members, where there is

insufficient restraint provided by links
c] Unreinforced or lightly reinforced members
U Decks composed of sub-elements of webs, flanges and diaphragms
0 Supporting elements with poor reinforcement detailing, which does not restrain
expansion.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

In assessing these and other types of element, it is also important to consider sensitivity
to structural form and to reinforcement detailing. Restraint to the expansive action is
significant, whether created by structural redundancy or by confinement provided by
reinforcement. Reference 9.9 provides data on how mechanical properties may be
affected by different levels for free expansion, and adds modifying factors for a range of
restraints. In addition to reductions in mechanical properties, it may also be necessary to
consider geometrical properties, where physical damage has led to severe cracking,
spalling or delamination.

9.2.5 Indicative modifications Most approaches to strength assessment are based on modified design models. The
to design models for strength approaches in References 9.7 to 9.10 are broadly the same in principle, but differ in
detail, including some assumptions made, in the interests of conservatism. Table 9.Ig9
assessment gives an indication of the type of amendment proposed. The specific case of Highway
Structures, covered by BD44/95, is dealt with in a later Section.

Possible limitations to assessed strength, unique to ASR-affected elements

In general, the approach to assessment, via analysis or modified design models, is focussed
on a re-evaluation of the primary action effects such as bending, shear, axial compression,
torsion, etc. Bond and anchorage may be more significant in assessment, due for example
to reductions caused by severe cracking or delamination. It may be necessary also to
consider alternative load bearing mechanisms, induced uniquely by the deterioration,
which may lead to reductions in strength, stiffness, stability or serviceability.

For ASR in particular, the following possibilities are identified and may require
consideration:

U Possible buckling of cover concrete if this becomes delaminated.

0 Possible buckling of longitudinal bars in compression members, if the cover concrete
becomes delaminated. This is only likely in old structures, with very few links which
may be anchored inadequately (link spacing greater than 44 diameters for mild steel
and 32 diameters for high yield steel.
U Possible instability effects in heavily over-reinforced beams.
0 Crushing of concrete, if the difference between the failure strain in the concrete and
the yield strain in the compression reinforcement, is less than the ASR expansion strain
U Some of the indicative amendments in Table 9.1 are dependent on the anchorage
bond being unaffected. A careful study of deterioration in anchorage zones of both
reinforced and prestressed elements is required.

104
 Table 9.1
Indicative amendments to be made t o structural
codes t o allow for the effects of ASR.
Flexure U Use ASR-affected concrete strengths derived from uniaxial (core or cylinder)
compressive strengths
0 Include ASR-induced expansion such that reinforced members are treated as
prestressedmembers
0 Only use 60% of the ASR-induced (restrained) expansion in determining the
additional strain in the reinforcement or tendons
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Column behaviour 0 Use ASR-affected concrete strengths derived from uniaxial (core or cylinder)
compressive strengths
Shear 0 Use ASR-affected concrete strengths derived from uniaxial (core or cylinder)
compressive strengths
Ci Include ASR-induced expansion such that reinforced members are treated as
prestressedmembers. For some codes this will mean using the prestressed
concrete shear rules
0 Only use SO% of the ASR-induced (restrained) expansion in determining the
additional strain in the reinforcement or tendons
U Ascertain the effects of any reductions in anchorage bond strength on shear strength
Torsion 0 Use ASR-affected concrete strengths derived from uniaxial (core or cylinder)
compressive strengths
0 Carry out an extra equilibrium check to ensure that the stress in the concrete
compression struts is not exceeded
Punching shear 0 Use ASR-affected concrete strengths derived from uniaxial (core or cylinder)
compressive strengths
0 Include ASR-induced expansion such that reinforced members are treated as
prestressed members. For some codes this will mean using the prestressed
concrete shear rules
U Only use SO% of the ASR-induced (restrained) expansion in determining the
additional strain in the reinforcement or tendons
0 Ascertain the effects of any reductions in anchorage bond strength on shear strength
Bearing 0 Use ASR-affected concrete strengths derived from uniaxial (core or cylinder)
compressive strengths or (preferably) direct tensile strength measurements
Fatigue 0 No guidance can be provided yet
Bond U Use ASR-affected concrete strengths derived from uniaxial (core or cylinder)
compressive strengths or direct tensile strength measurements
il Use the Tepfers bond expression, but see Section 9.12

9.2.6 Assessment Of Highway The principles and general approach in the above sub-sections apply, but detailed
Structures - BD44/95 and requirements are contained in the above Highways Agency documents. The indicative
amendments in Table 9.1 then become particular. Table 9.2 lists relevant equations for
BA44/96; BA52/94; BD21/01 assessment in BD44/95, with suggestions being made for reductions in assumed
material strengths, as affected by ASR

9.3 Aggressive chemical

attack

9.3.1 Background In the UK, sulfates in soil and groundwater are the chemical agents most likely to
attack concrete. Another possible cause of concrete deterioration is groundwater acidity,
and this is sometimes linked with the presence of sulfates. The extent to which sulfates and
acidity affect concrete is linked to their concentrations,the type of ground, the presence of
groundwater (and its mobility), the type of concrete, and the form of construction.

105
 Table 9.2
Suggested material strengths for use in equations given in BD44/95. for ASR.

Uniaxial f,, replaced by fJO.8 4.3.2 Material Figure 1 Cross references

Compressive Properties Subclauses 5.3.2.l(b), 5.5.3.2(b)
Strength (fc) and 6.3.3.1(b)
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

5.3.3.1 Shear stress 0.92J V c J Y m J

5.3.3.2 Shear capacity Equation for V,
5.3.2 Resistance moment of beams Equations 2, 3, 4, 5 and 7
5.5.3 Short columns subject to Equations 14 and 15
axial load and bending
about the minor axis
5.5.4 Short columns subject to Equation 17
axial load and either bending
about the major axis or biaxial
bending
5.6.1.1 Definition (Reinforced 0.1f'"AC
concrete wall)
6.3.3 Ultimate limit state: flexure Equation 27
6.3.4.3 Sections cracked in flexure Equation 29A First component of the equation
only

indirect or Splitting If,,replaced by fJO.3 5.3.4.3 Stresses and reinforcement Equations for V,,
Tensile Strength V,) and Vt,
5.8.6.3 Anchorage P,ifcJYm, If reinforcement is present that
Bond crosses the potential splitting
cracks then the fcU value taken as
the value for the ASR concrete
should be used
6.3.4.2 Sections uncracked in flexure Equations for M,,and
maximum principal tensile
stress
6.4 Slabs Equation for maximum
principal tensile stress
6.7.4 Transmission length in kt@/,ifO f,, is the concrete strength at
pre-tensioned members transfer
f,, replaced byf,/O.ll 5.8.6.9 Bearing stress inside bends Equation for bearing stress
inside bends
7.2.3.3 Bearing stresses Equations for bearing stresses

Conventional sulfate attack has been recognised for many years, and dealt with by
defining classes of aggressivity and recommending concrete specifications for each, in
terms of ingredients and mix proportions. As more has become known, these have been
refined in successive British Standards; BS85OOgz5,is the most recent example.

In recent years, the situation has become more complicated by several developments:

0 The 'discovery' of new forms of sulfate attack, particularly thaumasitegz6.

0 The greater use of brownfield sites, possibly increasing the range of aggressive chemicals.
U Recognition that some sites may contain sulfides, such as pyrite, which may increase
sulfate content from oxidation following ground disturbance.

106
 -?.

Specific materia! and a

This has led to a re-classification of aggressive chemical with

corresponding material specifications.

All of this means that, while provisions to deter sulfate attack have existed in
specifications for many years, they have not always worked or have been misapplied.
Therefore, assessment engineers may be faced with deterioration due to some form of
sulfate attack, and the need to take this into account in assessment. The purpose of this
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Guide is to outline the key features involved.

9.3.2 ‘COnVentiOnal’ form O f This term applies to sulfate attack, which leads to the formation of ettringite and
su[fate attack gypsum in susceptible concretes. For this to occur, the following must be present:

0 A source of sulfates, generally from sulfates or sulfides in the ground

0 The presence of mobile groundwater
0 Calcium hydroxide and calcium aluminate hydrate in the cement matrix

In the alkaline pore solution provided by the sodium, potassium and calcium hydroxides,
liberated during the cement hydration reactions, the sulfate ions from the ground react
with calcium hydroxide to form calcium sulfate (gypsum). This, in turn, reacts with
calcium aluminate hydrate to form calcium sulfo-aluminate hydrate (ettringite). In
Sulfate-Resisting Portland Cement (SRPC), the tricalcium aluminate (C,A) level is kept to
a minimum, so reducing the extent of this reaction.

As for all deterioration mechanisms, careful diagnosis is important; if the dominant

mechanism is not identified, the assessment of the likely consequences may be wrong,
leading to inappropriate remedial Sulfates in the ground are usually of
natural origins and are those of sodium, calcium and magnesium. Usually they occur in
significant quantities, in weathered or disturbed zones (possibly also in backfill). This
means that the most likely locations for sulfate attack are in concretes from
approximate ground level to a depth of about 10 m. The extent of the damage, and the
rate of the reaction, is dependent on the porosity and permeability of the concrete, and
on the mobility of the groundwater. The first effect of the reactions is to increase the
strength and density of the concrete, as the reaction products fill the pore space. When
it is filled, further ettringite formation induces internal stresses in the concrete, which
may disrupt, causing expansion of the affected region. Together with white crystalline
I accumulations, this is characteristic of the ettringite/gypsum form of sulfate attack.

1 As a result, affected concrete elements may suffer from:

0 Disintegration of the concrete due to expansion, followed by erosion

0 Softening of the concrete
0 Increase in ingress of moisture and aggressive salts, possibly leading to corrosion of
any reinforcement present.

I For assessment purposes, these effects may lead to:

0 Loss of effective concrete cross-sectional area

0 Loss of concrete strength

107
 8 Specific materid and assessment ffaaors

U Loss of cover t o reinforcement

0 Loss of bond between concrete and reinforcement
0 Possible loss of reinforcement cross-sectional area, due t o subsequent corrosion

The significance of this will depend on the type of construction, and on the nature, size
and shape of the structural elements. This can only be fully assessed by a risk analysis
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

typified by that in Reference 9.26.

9.3.3 Thaumasite f o r m O f This form of attack is dealt with in detail in References 9.26 and 9.27.
sulfate attack (TSA)
Generally, several factors must be coincident for TSA t o occur in susceptible concrete:

U A source of sulfates, generally from sulfates or sulfides in the ground.

0 The presence of mobile groundwater.
0 A source of calcium silicate hydrate, mostly derived from cementitious calcium
silicate phases present in Portland cements.
0 The presence of carbonate, generally in coarse and/or fine concrete aggregates.
U Low temperatures, since thaumasite formation is most active below 15°C.

The significant difference from 'conventional' sulfate attack is the presence of carbonate
and the involvement of calcium silicate hydrates with the sulfate ions, to produce
thaumasite. The calcium silicate hydrates provide the main binding agent in Portland
cement, so that this form of attack weakens the concrete, and, in advanced cases, the
cement paste matrix is reduced eventually t o a mushy incohesive white mass. SRPC
concretes can be vulnerable, but concretes containing ground granulated blastfurnace
s(ag have better resistance.

Other factors known t o influence the severity of TSAgZ8include the concrete mix design,
workmanship, changes t o ground chemistry and water regime, and the type, depth and
geometry of the buried concrete.

The key effects of thaumasite on structural strength are:

0 Loss of concrete cross-section and strength. This is the most critical feature. Concrete
affected by TSA should be disregarded in strength calculations. This may lead t o
overloading in the remaining section, or t o instability and buckling.
0 Loss of cover t o reinforcement. Most substructures are in compression. If the depth of
the affected concrete is less than the cover, generally compression bars will reach
their ultimate design strength provided that the detailing of the links is sufficient t o
provide restraint.
0 Loss of bond. In compression members, provided that buckling of the main steel is
avoided, this may be less important. For other elements, including piles and pile caps,
both flexure and shear will require consideration
0 Synergetic effects, including a possible increase in the risk of corrosion, may also
require consideration.

108
 9.4 Freeze-thaw action Concrete exposed t o cycles of alternative freezing and thawing can undergo progressive
deterioration, which starts a t the surface and moves deeper into the concrete with
increasing number of cycles. Its primary effects are surface scaling and internal
mechanical damage due t o the expansive action of ice formation.

Water expands on freezing and in concrete this creates pressure that causes internal
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

damage when greater than the tensile strength of the concrete. Nevilleg2’ identifies two
mechanisms by which this occurs; the freezing of water in voids and of that diffused into
small pores by osmotic pressure generated by the process of freezing. The latter is
thought to be particularly important in causing damage t o concrete because of the
greater pressure that is generated in the small pores. Osmotic pressure can also arise in
concrete that has been exposed t o de-icing salts.

Repeated cycles of freezing and thawing cause cumulative damage. The destructive
expansion of water turning into ice in each cycle allows water in the following cycle t o
migrate t o new locations and t o cause further damage t o the concrete on freezing.

The factors affecting the severity of damage can thus be summarised as follows:

0 The degree of saturation; fully saturated concrete is most vulnerable

0 The level of exposure t o the atmosphere
0 The frequency and number of freeze-thaw cycles
0 The level of exposure t o de-icing salts
0 The quality of concrete.

A bridge is more susceptible t o damage by freeze-thaw action the further north it is

located in the UK because of the increasing severity of the environmental factors.

In structural assessment terms, the key factors to be considered are as follows:

0 Reductions in concrete cover due t o surface scaling. In its simplest form, this may
increase the risk of subsequent corrosion. In more severe cases, as scaling and
crumbling of cover extends laterally and into the concrete, bond and anchorage may
be affected, particularly in negative movement regions or in reinforcement designed
for punching shear.
0 Internal damage will affect (reduce) the mechanical properties of the concrete.
Guidance on this is given in Reference 9.30.

If there are other destructive mechanisms acting simultaneously, such as the penetration
of chlorides, then the different types of interaction between them may need t o be
consideredg30.

It is suggested that the assessment of frost-damaged structures should be divided into

two. Firstly, a qualitative one based on visual inspection t o estimate the aggressiveness
of the environment and consequences of structural failureg3’, and secondly, a
quantitative one t o analyse structural capacities of the critical elements using lower
bound value of concrete properties. In the latter, it is important t o assess the present
and future structural capacities and safety levels t o determine a strategy covering
inspection, further investigation and future repair and strengthening workg32.

109
 9.5 High Alumina Cement High Alumina Cement (HAC) is manufactured from limestone and bauxite and contains
principally alumina with smaller amounts of ferrous and ferric oxides, silica and traces of
(HAC)
titanium oxide, magnesia and alkalis.

HAC concrete (HACC) has a very high rate of strength development, achieving 80% of
its ultimate strength within 24 hours of placement, and, because of this, was widely
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

used in the 1950s and 1960s to maximise the output of precast concrete building
components.

Because production runs were small, HAC was less widely used in the manufacture of
precast bridge beams a t this time.

HAC had been used as early as the 1920s for its sulfate resisting properties in the
construction of bridges and was known then as ‘fondu’ or ‘lightning cement’.

In the early 1970s a number of roofs made with HACC collapsed. Following this, the
Building Research Establishment carried out a major investigation into HACC and
established that under normal service conditions the compressive strength of this
concrete could be halved over a long period of time.

The investigation confirmed previous findings that the hydrate produced by the
hydration of HAC is chemically unstable a t normal and high temperatures and converts
to a more stable form. This conversion results in a greater porous structure of the
concrete and leads to a loss of strength.

Other facts about HACC that have been noted include:

0 The loss of strength is greatest for mixes made with high water/cement ratios and
exposed to high temperatures.
0 The general pattern of loss of strength versus time is similar for any mix proportion of
concrete.
0 The minimum strength reached is dependent on the water/cement ratio used and can
be as low as 0.4 times the strength a t day one for a free water/cement ratio of 0.6.

Because the problem of HACC is one of greatest concern to the building industry, little
work has been done on its effect on bridges. However, because bridges in general are
more conservative in their design than that of buildings, the loss of strength of HACC
has less impact.

The process of conversion can take many years to complete, but can be assumed to
have been fully completed in bridges in the UK where the structural use of HAC was
withdrawn from British Codes in the early 1970s.

It is recommended, because of the large variability in the strength of HAC concrete, that
cores are taken for the measurement of strength. The estimate of cube strength from
core tests is considered to be conservative but offers the best method for HACC because
of the background experience in its interpretation.

110
 The risk and the rate of corrosion of reinforcement are determined as in Portland
cement concrete mainly by a combination of the depth of carbonation, chloride and
sulfate levels and environment. However, because of its greater porous structure, these
could be higher in converted HAC concrete. Some guidance is given in References 9.33
to 9.38.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Salts
Contact with salt solutions is more likely to cause deterioration through corrosion of
reinforcement rather than attack of the concrete itself.

Carbonation
HACC are subject to carbonation in the same way as concrete made with Portland
cement. However, because the chemistry and mechanism of carbonation is different
from that in Portland cement concrete, the method of testing for carbonation in HACC
is different. Petrography is considered to be the only cost effective way to establish the
depth of carbonation, as the phenolpthalein test for assessing the depth of carbonation
in Portland cement concretes is unreliable for HACC

Sulfates
HACC is more vulnerable to sulfate attack if its conversion has been rapid, thus
substantially increasing its porosity and permeability, and if its water/cement ratio is high.

9.6 Supersulfated Cement Supersulfated cement (SSC) is manufactured from granulated blastfurnace slag (80% to
85%), calcium sulfate (10% to 15%) and Portland cement clinker (up to 5%).
(ssc)
It is resistant to attack in aggressive conditions and, consequently, has been used in the
manufacture of concrete pipes for placement in contaminated, acidic or sulfate-rich
ground, and in the construction of bridges over railway lines during the steam train era.

Because of its low heat of hydration, it has been used in large concrete pours to
minimise thermal cracking.

Because of its high slag content, concrete made with SSC requires to be wet cured for a
significant time to prevent exposed surfaces from becoming friable or powdery.

Carbonation causes a greater loss of strength in concrete made with SSC than Portland
cement and can also lead to friable or powdery surfaces.

SSC has a low alkali content and therefore any reinforcement, embedded in concrete
made with it, will be more susceptible to corrosion if chlorides are present or if
carbonation has taken place.

Lime released during hydration of SSC made with higher contents of Portland cement
will interfere with the reaction between the calcium sulfate and the slag.

111
 Lime-base sealers, known to have been applied as standard practice in the past, and lime
from adjacent concrete made with Portland cement have been observed to damage
concrete made with SSC.

SSC is no longer made in the UK but is quite commonly used in Belgium and to a lesser
extent in France.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Carbonation of Portland cement concrete results in an increase in strength, whilst that

made with supersulfated cement on carbonation suffers a loss. However, since this
applies only to the outer skin of the concrete the effect is not normally structurally
significant. The use of supersulfated cement does not affect the assessed capacity of
reinforced concrete.

9.7Chlorides Sections 9.7, 9.8 and 9.9 are related, in covering different aspects of the major cause
of the deterioration of concrete bridges - reinforcement corrosion. This Section
deals with possibly the prime cause of corrosion - chlorides from various sources;
similarly, Section 9.8 covers the other significant cause - carbonation of the
concrete. Section 9.9 then relates to the corrosion process itself. Finally, the possible
structural consequences of corrosion are covered in Section 9.12, in terms of bridge
assessment.

The mechanism of corrosion, in aqueous media, is electrochemical in nature; zones of

different electrochemical potential develop (anodes and cathodes). The corrosion
process mainly proceeds by the formation of numerous microcells. The nature of these
cells can vary, and, under certain circumstances, macrocell effects can develop; this
depends on the resistivity of the concrete, and, in the case of chlorides, on the
uniformity of the penetration of the chlorides through the cover concrete. Any
investigation of corrosion needs to look at the nature of the cells involved, since the
galvanic current in macrocell action can lead to local pitting, rather than general
corrosion, which is potentially more serious in structural terms. Local pitting is more
likely where corrosion is initiated via chlorides, but it is by no means universal.

The essential action of chlorides, when they have penetrated the concrete and reached
the reinforcement, is to disrupt the protective passivity layer of oxide on the surface of
the reinforcing bars, which was formed by the hydrating cement paste, and, in the
presence of moisture and oxygen, to create corrosion cells.

For concrete bridges, the sources of chlorides can be either internal or externai.
Internally, they can emanate from contaminated aggregates or water. Until the
mid-I 970s, they could also come from admixtures such as calcium chloride, which
was used as an accelerator, until banned in Codes and Standards. Internal sources of
chlorides are much less likely in modern structures due to improved standards and
controls for the production of concrete. External sources come from the use of de-icing
salts during the winter months in the UK, and from salts in marine and maritime
environments.

112
 Before the action of chlorides can come into play in creating corrosion cells, they must
first penetrate the concrete cover and reach the reinforcement. There are three features
which are important in this process:

1, The build up of chloride concentration in the surface of the concrete. This depends on
the concentration and continuity of the chloride sources. It also depends significantly
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

on the local environment and on the microclimate that is created due t o the location
and orientation of the structural element, e.g. some surfaces are directly washed by
driving rain and run-off, others are not. Cycles of alternative wetting and drying are
also important.
2. The penetration of the chloride ions through the cover concrete. In submerged zones
or in fully saturated concrete, chlorides penetrate by diffusion. In other cases,
particularly where wetting and drying cycles are involved, capillary absorption may be
involved, and this is a faster mechanism. A great deal of research effort has gone into
the prediction of this process, usually involving some form of diffusion model; the
essentials of these may be found in References 9.39, 9.40 and 9.41.
3. When the chlorides reach the reinforcement, a critical amount has t o accumulate, t o
induce the onset of corrosion; this is known as the threshold level. This is not a fixed
single value, but depends on:

0 Type of cement: fineness, amount of C,A, amount of gypsum, presence of pfa or

slag, etc.
Watedcement ratio (porosity)
0 Curing and compaction (porosity)
0 Moisture content and variation
0 Type of steel, surface roughness and condition
0 Oxygen availability.

The same concrete may have very different threshold levels therefore in different
situations, and this has been found in practice. One approach has been t o assume a
critical chloride threshold level, conservatively at that given in codes for mixing water
(0.4% of cement weight), but values in excess of 1.O% of cement weight have been
measured in

The above brief review of the action of chlorides has been presented t o make the point
that, in an assessment situation, where the presence of chlorides is suspected but
corrosion has not started, in-depth investigation may be worthwhile (via measurement
and diagnosis) t o predict how long it might be before corrosion might start.

9.8Carbonation The alkalinity of hydrated cement paste is important in providing protection t o

reinforcement against corrosion. This is reduced by carbonation of the concrete.
Carbonation of concrete occurs when carbon dioxide from the atmosphere reacts with
the calcium and alkaline hydroxides and cement phases, leading t o a reduction of the
pH value in the pore solution t o values near neutrality. As the carbonated front reaches
the reinforcement, the protective passive film on the steel surface may break down, and
the corrosion process can start, in the presence of water and oxygen.

113
 8 Specific material and assessment factors

Carbonation is a diffusion process, and therefore similar models can be used t o predict
its advance as those used for chloride penetration 9 3 9 , 9 4 0 , 9 4 1 . A general review of
carbonation in concrete is provided in Reference 9.43, and Reference 9.44 provides data
on carbonation depths measured on site. The rate of carbonation depends on the
characteristics of a particular concrete. It does not occur in fully saturated or very dry
concrete, and is probably most intense at a RH value of about 50%. Alternative wetting
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

and drying is another critical environmental condition, in alternatively providing

conditions for high rates of carbonation and those for higher corrosion rates.

The soundness and depth of cover influences the rate of carbonation. Older structures
with low covers to reinforcement and low-strength concretes inadequately cured are
more susceptible t o the effects of carbonation.

Concrete made with blended cements if not properly compacted and cured can sustain
rapid rates of carbonation. Carbonation is more rapid in the presence of fly ash but is
counteracted by the dense structure of the hardened cement paste produced by the
reaction between the pozzolanic silica and calcium hydrate. Concrete made with sulfate-
resisting cement sustains a higher rate of carbonation than one using Portland cement.
Carbonation takes place also in high alumina cement paste with carbon dioxide reacting
with the calcium aluminate hydrates. Conversion of HAC increases the rate of
carbonation.

The rate of carbonation is not uniform and is slowed by the reduced porosity of
carbonated concrete; carbonation produces products of increased volume, and water
then enables the hydration of hitherto unhydrated cement to take place. Shrinkage due
t o carbonation may contribute t o crazing (shallow cracking) that allows in moisture,
contaminates and accelerates carbonation ingress. Loss of strength does occur with
carbonation but, because this applies only t o the surface zone of structural members, is
not significant.

Corrosion caused by carbonation is general with expansive rust products causing

delamination and spalling. Carbonation accelerates chloride-induced corrosion of
reinforcement. It has been shown that sequential carbonation and chloride attack can be
a very much worst degradation problem than either process alone.

Corrosion caused by carbonation not only cause reinforcing bars to lose their cross-sectional
area but their cover and bond through the spalling of concrete. Further discussion on the
effects of corrosion on reinforcement is given in Section 9.9.

The key effects on structural strength are:

U Loss of concrete cross-section area

0 Loss of reinforcement cross-sectional area
0 Loss of cover t o reinforcing bars
U Loss of bond between concrete and reinforcement in affected area

These are discussed in Section 9.1 2

114
 9.9Steel corrosion
9.9.1 Introduction and The mechanism of corrosion is an electrolytic process, which, in concrete structures, is
background initiated by carbonation of the concrete or by the penetration of chlorides. A brief
description is given in Section 9.7, which stresses that different types of corrosion cell
can occur, leading t o either general corrosion around the perimeter of the bars or
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

t o local pitting which penetrates more deeply into the bar cross-section. In assessment,
therefore, the nature of the corrosion cells has t o be looked at closely. General
background t o all of this is provided in References 9.29, 9.39, 9.40 and 9.41.

The significance of corrosion in engineering terms, was first given shape by Tuuttig4’,
who proposed a two-phase model:

1. An initiation period, which consists of the time from the erection of the structure until
the aggressive agent (chlorides or the carbonation front) has reached the reinforcement
and the conditions are right for corrosion t o start (see Sections 9.7 and 9.8).
2. A propagation period from the depassivation of the reinforcement until a certain
unacceptable level of deterioration has been reached. For this period, corrosion rate is
of great importance in assessment, in estimating the nature and magnitude of rebar
section loss - and, hence, its effect on residual structural capacity.

9.9.2 Effects O f COtTOSiOn The effects of corrosion in structural assessment terms are as follows:

1. Reduction ofrebar cross-section: This may be either general, around the bar perimeter,
or local, penetrating into the cross-section. The structural effect is t o reduce capacity
(bending, shear, etc.), although this is not always proportional t o the loss of rebar
section, depending on detailing and on where the corrosion is occurring.
2. Ductility ofreinforcement: There is evidence that corrosion can reduce rebar ductility,
which may be important if plastic or non-linear analytical methods are used for
assessment.
3. Cracking, spalling, delamination: The formation of expansive corrosion products can
lead t o cracking along the line of the reinforcement. At corner sections, this can lead
t o local spalling. In extreme cases, delamination of concrete cover can occur over
significant areas, either for tension or compression steel. In compression members,
the concrete cover can be rendered ineffective in load carrying terms; in extreme
cases, short elements may be rendered slender.
4. Loss ofbondandloranchorage: There are two aspects here:

0 Formation of expansive corrosion products may alter the basic bond characteristics.
0 Spalling and delamination reduces the effective bond perimeter, and residual bond
will depend on reinforcement detailing (especially of links).

Loss of anchorage is more critical than loss of bond.

The development of corrosion, and the magnitude of its effects, may also depend on the
general quality of the design and construction, and, in particular, on the presence or
otherwise of defects due t o other causes. Cracking, in general, is particularly significant
in this regardg46,9 47.

115
 Assessment models have been developed t o predict all of the effects itemised above. These
are a t different stages of development, and have varying amounts of calibration against test
data. Most are based on the principle of modifying the corresponding design models.

9.9.3 COrrOSiOn rate

Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

This is a highly variable phenomenon, much influenced by the moisture state and the
availability of oxygen, in the inner concrete environment where the reinforcement is
located. This, in turn, is affected by the local microclimate at the perimeter of the
concrete element, which is subject t o daily and seasonal variations in moisture levels
and temperature. How this outer environment affects the inner environment depends
on the quality and thickness of the concrete cover.

Methods of varying effectiveness do exist for estimating corrosion rates in the field,
either directly or indirectly. However, unless they are used t o take measurements under
the different local internal and external environments that might exist, they may not
give values for loss of rebar section, which are representative over significant periods of
the remaining life of the structure.

This is a major dilemma, since an underestimate of the rate could lead t o excessive
damage prematurely, and (more likely in practice) an overestimate can lead t o
premature remedial measures. Until knowledge and technology improves, the solution
has t o lie in a mix of good site measurements and risk analysis, involving probability
distributions for sensitive locations. A method of obtaining representative values for
corrosion rate is given in Reference 9.40.

9.9.4 Prestressed Concrete Pre-tensioned concrete elements, typified by standard bridge beams, have a good track
record over periods up t o 40 years, in terms of steel corrosion. Possibly, this is due t o
high cover values and t o concrete quality dictated more by strength at transfer rather
than strength in service.

The situation is different for post-tensioned concrete bridges, where corrosion has
occurred a t poorly formed joints and a t poorly protected anchorages. Inadequate
grouting of tendon ducts has been another major factor.

Local corrosion leads t o the breaking of individual wire and eventually failure of tendons.
The effects of tendon failure vary from one structure t o another and are influenced by:

0 Bridge deck geometry

0 Prestressing system
0 Integrity of grout within the duct
0 Extent of tendon corrosion.

Because it is difficult with present inspection methods t o discover tendon failure and t o
determine adequacy of grouting, and then t o model these defects analytically, the
assessment of post-tensioned bridges is problematical.

116
 Recent work by Cavell and Waldron a t Sheffield U n i v e r ~ i t yhas,
~ . ~however,
~ provided a
better understanding of the behaviour of corrosion-damaged post-tensioned concrete
bridges. This found that:

0 Significant levels of deterioration do not always compromise flexural strength to the

same degree.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

0 Presence of grout voids alone may not significantly affect flexural strength.
0 Presence of voids a t or near to a tendon failure will affect the ability of a failed
tendon to re-anchor fully and thus profoundly affect residual strength, particularly
when a duct contains long or continuous voids.
0 Length and distribution of grout voids along the beam are important factors when
considering the loss of strength.

Stress corrosion cracking (SCC) is another form of corrosion of tendons that has been
reported. It is a brittle failure a t relatively low constant tensile stress of an alloy exposed
to a corrosive environment.

In a survey of French bridgesg4’, SCC was found to only affect quenched hot-rolled wires
containing more than 0.1% of copper that were used from 1950 to 1965. SCC has not
been reported as a problem to bridges in the UK.

Fortunately examples of post-tensioned bridges collapsing in the UK are rare, because

most have been designed as ‘fully’ prestressed and would have to lose a substantial
percentage of their tendons before they became unsafe.

The assessment of concrete structures affected by steel corrosion is discussed in

Section 9.12.

9.10 Fatigue Fatigue performance depends upon the strength of the member and the load spectra.
The key elements of the load spectra are the amplitude and frequency of load
applicationgs0.Investigation of actual fatigue parameters can be undertaken using
weigh-in-motion measurements to provide information on the loading and frequency of
heavy loading a t specific sites.

Fatigue damage is particularly a problem with short span bridges, where a high
proportion of the induced stresses in the reinforcing bars will be due to live loading. The
location of the reinforcing bars is also important with those located beneath a slow lane
being subjected to a greater amount of heavy vehicle loading.

The prediction of fatigue life can be undertaken using the rigorous method given in
BS 5400 Part 10. However, the Advice Note BA38/93”’ provides guidance on a
simplified method, which should cover most cases. The latter document notes that
corrosion or damage results in shorter fatigue life of reinforcing bars. The Appendix in
this document is useful in determining the loss of section of corroded bars based upon
the observable surface.

117
 8 Specific materia[ and assessment factors

BD44 does not include a requirement to undertake a fatigue check, as this is considered
to be a serviceability check, and concentrates on analysing load carrying capacity a t
ultimate limit state. BA38/93 states that fatigue assessments are only appropriate for
bridges that are less than 25 years old. This is because for structures older than this, the
loading and fatigue history is uncertain and the future life may be limited.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Where bars are mechanically damaged, then the assessment of the fatigue life may be
based upon the same principles as for corroded bars.

The strength and modulus of elasticity of concrete can reduce significantly under cyclic
loading. In reinforced members, concrete fails in the compression zone through micro
cracks. Whilst in steel bridges fatigue changes can be monitored by observation of visible
cracking, those in concrete bridges cannot, as cracking is internal.

9.1 1 Sub-standard Sub-standard reinforcement detailing is a particularly difficult subject to consider in the
reinforcement deta i1ing assessment of bridges.

One of the most common reasons for a concrete bridge to fail its assessment is the
inadequate detailing of the reinforcement. Usually, this is associated with an apparent
bond or shear failure. Fortunately, the incidences of bridges failing in service as a
consequence of inadequate detailing are rare, and this suggests that the current
assessment code is conservative in dealing with this problem.

The current assessment code (BD44) is based almost in its entirety on the design
code with some technical changes to the clauses associated with shear, bond and
detailing to make it less conservative where possible. The design code covers complex
structural behaviour with empirical formulae derived from strictly controlled test
situations, for example in the anchorage of reinforcement. These formulae do not
allow for interaction between concrete, longitudinal and shear reinforcement. Whilst
in the design of a new structure the conditions of these tests can be replicated, they
are unlikely to be so in an older structure being assessed. Consequently, the application
of these formulae in their assessment format is restricted to a limited number of
situations.

Bond failure is an extremely complex phenomenon being a function of many variables

including, in particular, cover. The assessment code does acknowledge this, but gives no
guidance, except to give direction to a number of referenced papers.

It has been suggested as a consequence of testing undertaken by Clarkgs2a t Birmingham

University, that the effective area of an inadequately anchored bar is a function of the
ratio of the actual lengths, and not of their squares, as implied in the bridge assessment
code. This conclusion is also reached by Cleland It is assumed in both papers
that this effective steel area can be used in the standard assessment formula for shear in
BD44/95. However, this formula is taken from the design Standard BS 5400, and applies
to the effect of fully anchored reinforcement, which is largely governed by the stiffness

118
 of the steel rather than by the ultimate bar force. The approach of reducing the effective
area in the formula to allow for reduced anchorage is inconsistent therefore with the
rationale behind the formula.

For loads that are applied close to a support, there is usually an enhancement in shear
capacity to model the effect of the steeper angle of compression. Clark et a/954
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

recommend that the ‘shear enhancement’ factor in BD44/95 for short shear spans be
allowed, even for inadequately anchored reinforcement. This contrasts with the existing
rule in BD44/95 that an anchorage length of 20 bar diameters must be present for shear
enhancement to be used in assessment.

The decrease in anchorage length was also shown by Clark to be accompanied by a

reduction in reliability of the anchorage. Clark suggests that allowance for this could be
made by deducting an unreliable length of bar, this being a function of the number of
bars a t the section under consideration.

A new approach that deals with the mechanics of the problem in a more rational way
has been developed by Shave e t a / 9 5 5based on test results and theoretical analyses. The
effect of the anchorage length on the shear capacity is modelled by considering the
anchorage force that may be developed in the steel. The shear capacity is then reduced
from its standard fully anchored value by a factor that is a function of the anchorage
force.

The existing approach in BD44/95 assumes that shear reinforcement is totally

ineffective a t low anchorage lengths, and does not allow enhancement of shear a t short
shear spans unless an anchorage length of at least 20 bar diameters has been provided.
Shave’s method is based on the fully anchored capacity, including the effects of shear
reinforcement and enhancement of shear a t short shear spans, and so these effects are
considered even a t low anchorage lengths. The proposed method is based on the
following expression:

where Vu =shear capacity

V,,, = fully anchored shear capacity including shear reinforcement and shear
enhancement as calculated using BD44/95
a, = a factor to allow for enhancement of bond due to transverse pressure
F,, =total ultimate anchorage force in the longitudinal reinforcement and the
other parameters are as defined in BD44/95.

Shave’s method agrees well with test data and is considerably less conservative than
BD44/95, especially when the effects of shear reinforcement and shear enhancement
are considered.

A large number of reinforced concrete bridges in the ownership of local authorities were
built in the earlier part of the twentieth century when there was an incomplete and,

119
 in some situations, incorrect understanding of the factors that affected strength. The
detailing of reinforcement in these structures also differed significantly from current
practice. Bussellg 56 describes the various propriety reinforcement systems in use a t the
beginning of the twentieth century.

Interestingly, the introduction of the first British codified design and construction
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

guidance for reinforced concrete a t the end of the Great War led to a period of poor
quality construction. The contractors were said to be less aware of good design and
working practices than the specialist concrete firms who built many of the pre-war
buildings and bridges.

This highlights the need to identify as far as possible all details of a reinforced concrete
bridge before commencing its assessment. The empirical code rules are severely limited
in their applicability and a more realistic assessment approach is required that looks a t
the detailing and design of bridge individually.

The use of plastic theory as an alternative to the empirical rules, which, more
adequately, would consider the influence of detailing on the load capacity of bridges,
has been proposed. This requires a detailed knowledge and understandingof plastic
theory that many practising engineers may not possess or may be reluctant to use
because it is not covered by the current code.

Consequently, there is a need to improve the current bridge assessment code to ensure
that its conservatism is reduced significantly while remaining safe, particularly in the
consideration of sub-standard reinforcement detailing.

9.12 Deteriorated
reinforced concrete
structures

9.12.1 Background and Much work has been done on the mechanisms of the different deterioration processes,
introduction and most are now well understood. Less work has been done on the effects that
deterioration can have on structural capacity; the result of this is that assessment
methods in the literature are highly variable, ranging from theoretical reliability methods
to assuming that structural capacity is somehow proportional to the general condition
of the structure. The first extreme has its difficulties, primarily due to the confidence
levels in the essential input and the problems of taking account of structural sensitivity
and reinforced detailing. The second extreme can be seriously misleading, depending as
it does on qualitative judgement, with reductions in structural capacity often not
proportional to loss of section.

The general approach to assessing the effects of ASR is now fairly well established
(Section 9.2),in terms of using modified design equations, with inputs derived from test
data. A similar approach seems valid for the effects of freeze-thaw action (Section 9.4).
The major difficulty is with corrosion. While the same approach is again valid in

120
 principle, the effects of corrosion are more wide-ranging, and there is a dearth of valid
test data, to permit proper calibration of the modified design models. There is the
further complication that the deterioration itself can, in some circumstances, create
failure mechanisms where no corresponding design models exist.

In this Section, only a brief issue of the key issues can be given. This is focused on:
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

0 Effects of deterioration, and principles for assessment

0 Sources of detailed information
0 Procedures in accordance with Highways Agency documentation

9.12.2
Effects O f For corrosion, these are itemised in Section 9.9, and listed here for convenience:
deterioration, and Drincides
I I
0 Reduction of rebar cross-section
for assessment 0 Ductility of reinforcement
0 Cracking, spalling, delamination
0 Loss of bond and/or anchorage.

The real effect of corrosion on all of these will depend on the quality of the design and
construction, on structural sensitivity and reinforcement detailing, on where the
corrosion has occurred, and on the type of structural element involved. It will also
depend on the particular action effect under consideration.

Structural assessment will be initiated generally by some form of structural analysis a t a

level described in Section 1.3, with reference to BA79; available analytical methods are
covered in Chapter 6. The first task in this procedure is to derive assessment parameters
as inputs, which take account of the corrosion effects listed above. In general, this will
involve reduced values for geometrical and mechanical properties, ideally derived from
the inspection and testing phases on site - but making use of general guidance in
available authoritative documents as back up. It will then be necessary to look a t
individual action effects at critical sections, such as shear, bond, anchorage, axial
compression - inevitably using modified design models. Strength will be the dominant
concern, but some issues such as spalling may require consideration where safety is
involved.

There are techniques and models available for doing this (see the References in the
following Sections), which are at different stages of development, with varying amounts
of calibration against test data. In applying these, it is important to develop an
understanding of the structural behaviour, and of how this might be modified by the
effects of deterioration - and of the sensitivity of the structures to these effects.

1 121
 9 Specific material and assessment ffaaors

9.12.3 Sources O f detailed In offering these sources, the emphasis has been put on consensus documents, where
information the guidance and the proposed models have been calibrated against test data. There
four primary sources:

1. HighwaysAgencydocumentation: This is dealt with in more detail in the following

sub-section, since it is the source that will generally be used in assessing concrete
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

bridges and highway structures in the UK.

2. fibsources: The starting point here is Reference 9.57 This updates, and summarise:
the work of the former CEB and FIP over a period of 20 years, with greater detail
being contained in the references to the report. This is strong on technical issues a
modelling.
3. BRlME sources: Reference to this is made in Section 1.8. There is a strong emphasis
European bridge management systems.
4. CONTECVET: Following an earlier European-funded project (BRITE 4062), on the
science of deterioration, CONTECVET focused on the development of models for
structural assessment. For corrosion, the details are in Reference 9.40, with
companion manuals on ASRg9 and on freeze-thawg30.Currently, the work is
continuing, in linking assessment to optimising repair and strengthening solutions
(acronym - REHABCON). CONTECVET technology is more focussed on bridges in
Reference 9.58.

9.1 2.4 Procedures in As may be seen from Section 1.4, Volume 3 of the DesignManualforRoadsandBridgc
accordance with Highways contains a number of Standards and Advice Notes that relate to the assessment and
repair of bridges. On assessment, these are either:
Agency docu mentat io n
1. General. BD21; BA16; BD and BA 44; BA79, or
2. Related to a particular form of deterioration. BA52; BA51; BA38, or
3. Related to a particular type of element. BA55; BA39.

All of this represents the current methodology for the assessment of concrete bridge:
As with other approaches, it is constantly being updated, as more test data appears, i
as experience is gained. From 1997 to 1999, a comprehensive test series was conduc
a t the University of Westminster in association with W S Atkins. The results are recor
in References 9.59 to 9.73. As a result, amendments have been proposed for BD44, b
details of which are given below. Until BD44 is formally updated, the use of any of th
would only be by agreement between the approval authority and the assessing engini
Key proposed amendments are:

0 Making an allowance for the future trends of deterioration in assessments (this is

achieved by the ongoing monitoring and modelling of deterioration).
0 Reducing the contribution of concrete to shear strength by 25% in the presence of cri
up to 2 to 3 mm in width (for wider cracks it is suggested that shear strength be base(
on either the tensile strength of adjacent stirrups or dowel action of the main bars).
0 Limiting the scope for redistribution of moments because of the reduced ductility (
deteriorated members.

122
 0 Taking account of deterioration in determining effective sections.
0 Taking account of corrosion in calculating the cross-sectional area of reinforcement.
0 Determining the effectiveness of compression reinforcement (it is suggested that the
surface layer be ignored and other layers be considered only 80% effective if the
section is delaminated in the compression zone and corroded cross-sectional area of
shear reinforcement does not meet minimum r e q u i r e m e n t s ) .
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

0 Ignoring shear and torsion links that are lapped in locations that are d e l a m i n a t e d .
0 Taking the effective loss of the area of a shear link to be half of the actual area, providing
that the beam depth is adequate to develop bond along half the leg of the link.
0 Describing a more rigorous method of calculation to the shear strength of defective
shear l i n k s .
0 Doubling the limiting bond stress for the leg of l i n k s .
0 Modifying the method of calculating enhanced shear strength of sections close to
supports consideringthe applied force in the main reinforcement and the type of l o a d i n g .
0 Using the varying angle truss approach to assess shear strength.
0 Checking torsion in combination with s h e a r .
0 Discounting longitudinal reinforcement in columns where links fail to meet minimum
design r e q u i r e m e n t s .
0 Using Regan’s method to calculate forces in bars due to the combined effects of
bending and s h e a r .
0 Modifying the calculation of bond stress to account for the presence of links, position
of bars in relation t o links, amount of delamination (flush with the top or mid barrel
of the reinforcement), transverse pressure and combinations of bars.

Regan and others have carried out flexural testing of reinforced concrete beams with
unbonded main reinforcement and, from this, have suggested formulae to calculate
ultimate moments similar in form to that in B D 4 4 . These formulae, which take account
of arching action, are applicable to simply supported beams that are adequately
reinforced against shear and anchorage f a i l u r e s .

9.13 References 9.1 BuitDiNG RESEARCH ESTABLiSHMENT (ERE), Alkali-silica reactioninconcrete, Digest 330 (4 parts), ENE, Garston, UK,
1997.
9.2 THE CONCRETE SOCIETY, Alkali-silica reaction.Mmimising the risk ofdamage to concrete, Technical Report 30 (3rd Ed ),
The Concrete Society, Camberley, 1999
9.3 BRITISH CEMENT ASSOCIATION, The diagnosis ofalkali-silica reaction: Report ofa Working Party, Report 45.02 (2nd Ed ),
BCA, Camberley, 1992
9.4 HOBBS, DW, Alkali-silica reactionin concrete, Thomas Telford Ltd. London, 1988
9.5 INSTITUTION OF STRUCTURAL ENGINEERS, Structural effects ofalkali-silica reaction: Technics/ guidance on the appraisal
ofexistingstructures. (2nd Ed.), IStructE. London, 1992.
9.6 INSTITUTION OF STRUCTURAL ENGINEERS, Appraisalofexistingstructures (2nd Ed.), IStructE, London, 1994.
9.7 HIGHWAYS AGENCY, BA52/94, The assessment of concretestructures affected by alkali-silica reaction, 1994.
9.8 CANADIAN STANDARDS ASSOCIATION INTERNATIONAL, Guide to the evaluation andmanagementofconcrefe
structures affectedby alkali-aggregate reaction, Document No CSA-A86A-00. CSA, Ottawa, Canada, February 2000.
9.9 BRITISH CEMENT ASSOCIATION et a/ A validatedusers manualforassessing theresidualsemcelife ofconcretestructures
affectedby ASR, EC Project Ref. IN 309012, CONTECVET, BCA, Camberley, 2001
9.10 CLARK, LA, Criticalreview of the structuralimplications of the alkali-silica reaction in concrete, Transport Research
Laboratory (TRL), Contractor Report 169,TRL. Crowthorne. 1989
9.11 CHANA, PS. Bondstrength ofreinforcement in concrete affected by alkali-silica reaction, Transport Research Laboratory
(TRL), Contractor Report 141, TRL. Crowthorne, 1989.
9.12 CHANA, PS and KOROBOKIS, g, The structuralperformance ofreinforced concreteaffected by alkali-silica reaction, Phase
I., Transport Research Laboratory (TRL), Contractor Report 267, TRL, Crowthorne, 1991.
9.13 CHANA, PS and KOROBOKIS, g, The structuralperformance ofreinforced concreteaffectedby alkali-silica reaction. Phase
11, Transport Research laboratory (TRL), Contractor Report 233, TRL, Crowthorne, 1991

123
 9 Specific material an8 assessment ffaaors

9.14 CHANA, PS. and KOROBOKIS.g. The structuralperformance ofreinforcedconcrete affectedby alkali-silica reaction, Phase
Ill, Transport Research Laboratory (TRL). Contractor Report 311, TRL, Crowthorne, 1992
9.15 COPE, RJ and SLADE L. The shear capacity of reinforced concrete members subjected to alkali-silica reaction, Structural
EngineeringReview. No 2, 1990, pp 105-112.
9.16 NG, KE and CLARK, LA. Punching tests on slabs with alkali-silica reaction, TheStructuralEngineer, Vol. 70, No. 14, July
1992, pp 247-252.
9.17 COPE, RJ, MAY, IM and WEN, H, Prediction ofstress distributions in reinforcedconcrete members affectedbyalkali-
aggregate reaction, Transport Research Laboratory (TRL). Contractor Report, TRL, Crowthorne, 1993
9.18 CAIRNS, J and ZHAO, 2. Behaviour of reinforced concrete beams with exposed reinforcement,Proceedingsofthe
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

InstitutionofCivilEngineers. StructuresandBuildings, Vol. 99, May 1993, pp 141-154.

9.19 JONES,AEK and CLARK, LA, The practicalities and theory of using crack width summation to estimate ASR expansion,
Proceedingsof the Institution of Civil Engineers; Structures andBuildings, Vol. 104, 1992, pp 183-1 92.
9.20 JONES.AEK and CLARK. LA, The effects of restraint on ASR expansion of reinforced concrete, MagazineofConcrete
Research. December 1995.
9.21 CHANA, PS and HOBBS, DW, The suitability ofcoresfor determining expansion rate andfuture expansion in concrete
adverselyaffectedby ASR. BRITE-EURAM Project 4062, Contract BREU-CT92-0591,January 1995.
9.22 FRENCH, WJ, Concrete petrography a review, Quarterlylournal ofEngineering Geology. Vol 27, 1991, pp 17-48.
9.23 KAWAMURA, M, TAKEMOTO, K and ICHISE. M. Influence of the alkali-silica reaction on the corrosion of steel
reinforcement in concrete, Proceedingsof the 8th InternationalConference on Alka/i-Aggregate Reaction, Kyoto,Japan,
Elsevier Applied Science, London, 1989, pp 115-1 20.
9.24 BOLTON, RF and WANE, J, Secondary effect of ASR on the durability of concrete Freeze-thaw, Proceedingsofthe9th
InternationalConference on Alkali-Aggregate Reaction, London, 1992. Vol. 1, The Concrete Society. Carnberley, pp
117-1 26.
9.25 BRITISH STANDARDS INSTITUTION, 858500-1:2006, Concrete - comp/ementaryBritish Standardto BS EN206- 1 . Part 1,
Methodofspecify,ng, andguidancefor the specifier, BSI, London, 2006.
9.26 DEPARTMENT OF ENVIRONMENT, TRANSPORT AND THE REGIONS, The thaumasiteform ofsulfateattack. risks,
diagnosis, remedialworks andguidance on new construction, Report of the Thaumasite Expert Group, DETR, London.
January 1999.
9.27 BUILDING RESEARCH ESTABLISHMENT (BRE), BRE Special Digest 1, Concretein aggressive ground (in 4 parts), BRE,
Carston, 2001
9.28 HIGHWAYS AGENCY, IAN48/03, Measures to minimise the risk ofsurfate attack (including thaumasite) - New structures
andstructuresunder construction. 2003
9.29 NEVILLE, A, PropertiesofConcrete, Longman, 1999
9.30 BRITISH CEMENT ASSOCIATION et a/ A validatedusers manualforassessing the residualservice life ofconcretestructures
affected byfrost action, EC Project Ref. IN 30901 2, CONTECVET, BCA, Camberley, 2001
9.31 CONCRETE BRIDGE DEVELOPMENT GROUP, Guide to testing andmonitoring the durabilityofconcretestrucfures,
Technical Guide No. 2, CBDC, Camberley, 2002.
9.32 PEARSON, S and PATEL, R, Repair ofconcretein highway bridges -apracticalguide. Application Guide AC43, Transport
Research Laboratory, Crowthorne, 2002.
9.33 DEPARTMENT OF THE ENVIRONMENT (DOE), Highalumina cement concrete, Technical Memorandum BE5/74. DOE,
London, 1974
9.34 INSTITUTIONOF STRUCTURAL ENGINEERS, Reportofa Working Pariyonhigha/umina cementconcrete, IStructE,
London, 1976.
9.35 CURRIE, Rand CRAMMOND, N, Assessment of existing high alumina cement construction in the UK, Proceedingsofthe
Institution ofCivil Engineers -StructuresandBuildings, ICE, London, February 1994
9.36 CURRIE, Rand CRAMMOND. N, Assessment of existing high alurnina cement construction in the UK. Discussion
on Reference 9 35. Proceedingsof the Institution of Civil Engineers - StructuresandBuildings, ICE, London, February
1995.
9.37 BUILDING RESEARCH ESTABLISHMENT, Assessment ofexisting high alumina cement constructionin the UK. BRE Digest
392, BRE, Garston, March 1994.
9.38 THE CONCRETE SOCIETY, Diagnosis of deterioration in concrete structures, Technical Report 54, The Concrete Society,
Camberley, 2000
9.39 BUILDING RESEARCH ESTABLISHMENT (BRE), Corrosion ofstee/inconcrete, Digest 444 (4 parts), BRE, Garston
9.40 CEMENT ASSOCIATION et al. A validatedusersmanualforassessing the residualservice life ofconcretestructures affected
by corrosion. EC Project Rei IN 309012, CONTECVET. BCA, Carnberley, 2001.
9.41 BAMFORTH, P, Guidance on theselection ofmeasuresforenhancingreinforcedconcrete durability. Report for DETR PI1
Project C139/3/376, Department of Environment,Transport and the Regions, London, 1999.
9.42 PETTERSON, K, Chloride threshold value of the corrosion rate in reinforced concrete, Proceedingsoffhe International
Conference on corrosion and corrosionprotection ofsteel in concrete, University of Sheffield, 1994. (Also available from
the CBI, Stockholm, Sweden)
9.43 PARROTT, LJ,A review of carbonation in reinforcedconcrete, British Cement Association, Camberley, 1987.
9.44 BUILDI NG RE S EA RC H ESTA B LI S H ME NT, Carbonation depths in s tructural quality concrete: an assessment of evidence from
investigation ofstructures and othersources, Paper BR75, ERE, Garston.
9.45 TUUTTI. K. Corrosion ofsteel in concrete, Swedish Cement and Concrete Research Institute (CBI), Stockholm, Sweden,
1980.
9.46 THE CONCRETE SOCIETY, The relevanceofcracking in concrete to the corrosion of reinforcement, Technical Report 44,
The Concrete Society, Camberley, 1995
9.47 THE CONCRETE SOCIETY, Non-structuralcracksin concrete, Technical Report 22 (3rd Ed.), The Concrete Society,
Carnberley. 1992

124
 9.48 CAVELL, D and WALDRON. P, Parametric study of the residual strength of deteriorating simply supported post-
tensioned concrete bridges, Proceedings ofthe Institutionof CivilEngineers- Structures andBuildings, ICE, London,
November 2001
9.49 HIGHWAYS AGENCY et a/, Post-tensionedconcretebridges. Anglo-French liaison report, Thomas Telford Ltd. London,
1999.
9.50 LAMAN, JA and NOWAK, AS, Fatigue Load Models for Girder Bridges,ASCEJournal ofStructuralEngineering, Vol 122,
NO.7, 1996.
9.51 HIGHWAYS AGENCY, Advice Note BA38/93, Assessment ofthefatigue life ofcorrodedordamagedreinforcing bars, 1993.
9.52 CLARK, LA, Assessment of Concrete Bridges, National workshop on new conceptsfor the management ofhighways
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

structures. Institution of Highways and Transportation. Leamington Spa, 1990.

9.53 CLELAND, DJ,CUMMINCS, SJ, RANKIN, GIB, TAYLOR, Sand SCOTT, RH, Influence of reinforcement anchorage on the
bending and shear capacity of bridge decks, JheStructuralEngineer,Vol 79, No 16, 21 August 2001, pp. 24-31
9.54 CLARK, LA. BALDWIN, MI and CUO. M, Assessment of concrete bridges with inadequately anchored reinforcement.
Bridge Management 3, University of Surrey, April 1996.
9.55 SHAVE, JD, IBELL, TJ and DENTON, SR, Shear Assessment of Concrete Bridges with Poorly Anchored Reinforcement,
Proceedings Structural Faults andRepair2003, Engineering Technics Press, Edinburgh, 2003.
9.56 BUSSELL, MN, The era of the proprietary reinforcingsystems, Proceedings ofthe InstitutionofCivi/ Engineers - Structures
dndBuildings. ICE, London. August/November 1996.
9.57 fib, Management, maintenance andstrengthening ofconcrete structures. fib bulletin 17, fib, Lausanne, Switzerland, 2002.
9.58 WEBSTER. MP, The assessment of corrosion-damaged concretestructures, PhD thesis, University of Birmingham, July
2000.
9.59 WS ATKINS. DeterioratedStructures Phase 2A Report, Document Reference AR40980/1/4/1, March 1997.
9.60 REGAN, PE, Tests ofslabs withfullyorpartlyexposedmain reinforcement. University of Westminster, July 1997.
9.61 REGAN, PE, Further tests ofslabs with exposedmain steel, University of Westminster, November 1997
9.62 RECAN, PE, DeterioratedStructures Jests ofRectangular Beams with Partially ExposedMain Steel, University of
Westminster. November 1997.
9.63 RECAN, PE, Effects ofDelamination ofcover on the Shear Strengths ofReinforced Concrete Beams Jest Series S G T,
University of Westminster, January 1998.
9.64 RECAN, PE, Flexure of UnbondedReinforcedConcrete Beams, University of Westminster, January 1998
9.65 RECAN, PE, Jests ofReinforcedConcrete Beams with Partially ExposedCornerBars, University of Westminster, July 1998
9.66 RECAN, PE, Jests ofReinforcedConcrete Columns with loss ofcover. University of Westminster, July 1998.
9.67 RECAN. PE, ExploratoryJests ofthe Punching ResistanceOfSbbS with Partially ExposedReinforcement, University of
Westminster, July 1998.
9.68 RECAN, PE, Effects of Prior Cracking on the Shear Resistanceof Reinforced Concrete Beams. University of Westminster,
December 1998
9.69 REGAN, PE, Jests of Beams with Partially ExposedMainSteel Series F Cantilevers under Distributedloading. University of
Westminster, November 1998.
9.70 RECAN, PE, Shear Strengths ofRCBeams with Defective Stirrups, University of Westminster,January 1999.
9.71 RECAN, PE, Jests ofBeams with Partly ExposedMainSteel Series C, RestrainedBeams,University of Westminster, January
1999.
9.72 RECAN. PE, Jests ofDelaminatedAnchoragessubject to Transverse Pressure. University of Westminster, March 1999.
9.73 RECAN, PE, Tests ofBeams with Partly ExposedMainSteel Double Cantilever Beams, University of Westminster. May 1999

Further reading
D DALY, A, Bridge Management in Europe Modelling of Deteriorated Structures, Bridge Management 4 (Ryall, MJ, Parke,
CAR and Harding, JE, eds), Thomas Telford, 2000
0 KENNEDY-REID, I, The Strength of Deteriorated Reinforced Concrete Bridges, Bridge Management 4 (Ryall, MJ,Parke,
CAR and Harding, ]E. eds), Thomas Telford, 2000

125
 - ~~

10. Load testing

10.1 General principles

10.1. I Background
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

The load testing of bridges is carried out in many countries and for a variety of purposes
This document is concerned only with its use in assessment in the UK Although concrete
bridges are the subject of this document, the principles apply to other types except where
there are specific references to concrete details and behaviour such as cracking

The urge to carry out load tests for assessment is based on a common view that
analytical methods often underestimate bridge strength. Load tests on redundant
bridges to collapse show that this impression is not without foundation. However, it
probably applies less to modern bridges than to older short span bridges, which are less
likely to have proper bearings and well-defined load paths. Older bridges are often much
stiffer in response to live loading than indicated by calculation. Although they can also
be stronger, the ratio of stiffnesses is not a reliable guide to the ratio of strengths. These
points are discussed by Cullington and Beales”.’.

Load testing can be regarded as supplementary to the normal process of assessment, used
when calculation alone is not enough to demonstrate an adequate load-carrying capacity.
Although this intention is clear enough, opinions are divided on how it should be applied.
The assessing engineer should be aware of that fact when considering load testing

Opinions on load testing are divided because assessment is carried out a t the ultimate
limit state. (An exception to these principles is the load testing of bridges in which cast
iron members are the critical elements, which are assessed for resistance to fatigue
under working stresses.) To be directly relevant to assessment, a load test must provide
additional information about that limit state not obtainable by other means. The applied
test load can be high enough to show directly that the uls is satisfied or the uls
behaviour must be inferred from the response at a lower load. Purely exploratory site
testing is not recommended. Calculations should be provided in advance to estimate the
range of performance expected under the test loads and ensure that brittle failure is
avoided.

Load testing has been the subject of study and debate in the National Steering
Committee for the Load Testing of Bridges (which was originally set up informally by a
group of bridge owners, consultants and specialists and later formalised with the support
of the Building Board of the Institution of Civil Engineers) and has occupied engineers in
the papers and proceedings of several conferences.This Chapter explains the current
situation as it appears to the Task Group.

10.1.2 Reasons for presence The ultimate capacity of a bridge or component under live loading may be greater than
of reserves of strength obtained by calculation for two main reasons which should be borne in mind when
considering bridge load testing:

126
 1. The critical load effects may be less than calculated.
2. The strength of a member or the bridge may be higher than calculated

Clearly, the possibility that it may sometimes be less than obtained by calculation cannot
be ignored. In some circumstances, it may be possible t o address both cases using test
loads less than uls loads provided data are available, or can be obtained, on the behaviour
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

of similar bridges or components beyond the test loads. Unless the original model gives a
greatly underestimated capacity, non-linear behaviour is likely t o occur before the uls
loads are reached. For the first of the two cases, it is necessary t o show that the structural
model adopted remains valid or overestimates load effects occurring at higher loads. For
the second case, the resistance model must be valid or underestimate the resistance. In
theory, if the load model is inadequate, the uls may occur at a lower load than calculated
particularly for brittle types of failure, and there are examples of the measured resistance
being lower than the calculated unfactored resistance.

10.1.3 Types O f load test This Section deals with the main types of load testing that the assessing engineer may
wish t o consider. These are:

0 Supplementary Load Tests

0 Proof Load Tests (Section 10.3) (Note that the terms Proof Load Test and Proving
Load Test mean the same thing in this document)
0 Tests t o ultimate
0 Hybrid and other load tests
U Dynamic tests.

The Supplementary Load Test is a type of static load test most easily identified as
conforming with the assessment documents in the Design Manualfor Roads andBridges.
The use of other types is currently at the discretion of individual Technical Approval
Authorities.

10.1.4The aims of load Possible aims for an in-situ load test for assessment are:
testing
0 To provide information at working load levels that can be used t o revise the structural
model for further assessment calculations.
U To provide a direct indication of the lower bound ultimate capacity by applying test
loads factored for the ultimate limit state (Proof Load Test).
0 To give confidence in the performance of the structure under normal working loads.
0 To provide information that can be used for a monitoring regime.

In addition:

U Load tests that directly establish the ultimate limit state performance may be carried
out on models in the laboratory or on redundant structures on site.

127
 U In-situ dynamic testing may be helpful when an aspect of the dynamic performance is
in question. It has also been studied for its potential t o give information about
assessed capacity or changes in capacity.

For a static load test, the main ways in which it is believed a load test can help directly
are:
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

U To show that the distribution of load within elements is more favourable than deduced
from the assessment model - i.e. it produces lower load effects in critical members.
0 To show that restraints are present that reduce the load effects on structural
members - e.g. dispersal through fill, end restraints.
0 To show that the capacity of members is greater than calculated.

10.1.5Documents in the The Standards BDZIlo3and BD341°4 from the DesignManualforRoadsandBridges

Design Manualfor Roads and require more refined methods t o be applied when simple methods of assessment fail t o
demonstrate an adequate capacity. BD2l is the primary reference and is not particularly
Bridges encouraging. It states (Clause 3.29) that:

“The objective of load testing shall be to check structural behaviour under load andlor
verifv the method of analysis being used, i.e. to prove the accuracy and suitability of the
assessment model of the structure. ’I

Further guidance is given in BA54. This defines Supplementary Load Tests and Proving
Load Tests. Supplementary Load Tests receive a qualified acceptance, whereas Proving
Load Tests are not recommended. Definitions of the tests are extended in Guidelines for
the Supplementary Load Testing of Bridgeslos (NSCLTB, 1998 - see Section 10.2) t o
include Proof Load Testing and Dynamic Load Testing.

It is important t o note that in this Section the terms Proof Load Test and Proving Load
Test are used interchangeably. The specialised definition of Proving Load Tests given in
the NSCLTB publication is useful in the context of that document but at odds with other
countries practising Proof Load Testing.

10.1.6Safety Safety is an important issue and it is accepted that there is an element of risk in any
load test associated with the possibility of damage t o the structure and personal injury.
The higher the load in relation t o the day-to-day service loads and a realistic assessed
capacity the greater the risk.

The NSCLTB p ~ b l i c a t i o nprovides

’~~ a framework for assessing risk that leads t o a
ranking for the level of expertise required for the test according t o scores for probability
and consequences. Some bridges, it suggests, are unsuitable for testing and others
should only be tested by specialist consultants with considerable structural knowledge
and testing experience of the particular type of structure. Proof Load Tests are expected
t o fall into that category.

128
 10.2 Supplementary load
testing

10.2.1 Status This method of testing is described and recommended in qualified terms in BA54. It is
the type of testing referred t o in BD2l. The Supplementary Load Test is an expression
used in the UK. In other countries, the term diagnostic test is sometimes used t o mean a
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

similar test.

10.2.2 Methodology The loads applied are intended t o be of a magnitude producing load effects that do not
exceed those occurring on a day-to-day basis. Loads can usually be applied using vehicles.
They should be applied in a way that produces effects in the structure that replicate the
effects found in the assessment calculations: e.g. transverse and longitudinal bending.

The structural model used in the assessment is loaded using identical load cases, and
then revised until correspondence with the test is achieved. This requires sufficient
measurements t o be made, for instance of strains and displacements, for those
comparisons t o be made effectively. The revised model is then used with the assessment
live loading t o produce a revised assessment. However, the equilibrium model that
results from the test must be checked carefully to ensure that it is valid up to ultimate
loads. For example, connections and restraints used explicitly or implicitly in the model
need t o be assessed. If their contributions cannot be sustained t o the uls, changes must
be made t o the model accordingly. This is expected t o require the results of tests t o
ultimate carried out on similar bridges or models (see BA54’06).

It is important t o note that a linear extrapolation from the test loads t o the uls loads is
not the intention. Unless the original assessment is highly conservative, non-linear
behaviour is likely t o occur before the uls loads are reached. It is necessary t o show that
the structural model or resistance model, as appropriate, remains valid or conservative
a t the higher loads

10.2.3 Further guidance Detailed guidance is given by the NSCLTB’07.Reference should also be made t o BD21’08
and BA54. The use of testing must be approved by the Technical Authority.

10.3 Proof load testing

10.3.1 Background There are no guidelines for Proof Load Testing available in the UK except for those given
briefly in BA54, which does not recommend the method. In principle, loads have t o be
applied t o the bridge t o represent directly and fully all load cases that are critical in the
assessment (and critical to the structure in practice, if these are different). Provided the
structure remains in a satisfactory condition under all combinations of assessment live
loading, it is concluded t o have sufficient load carrying capacity.

129

L
 Inclusion of this Section in this document does not constitute a recommendation for its
general application.

10.3.2 Methodology In order t o satisfy the uls, it is necessary at least t o apply loads that are factored for the
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

uls. However, it may be sufficient t o adopt values of unity for yt,for dead load and yt3
(using standard terminology, e.g. as in BD21) on the basis that the structure is fulfilling
this function directly and requires no additional allowance for variation. This assumption
relies on ensuring that the dead and superimposed loads are not allowed t o increase over
time and during the test loading, all critical combinations of load are faithfully replicated.

Loads are applied incrementally. If the load test has t o be stopped short of the required
loading, a reduced live loading capacity is obtained by applying the partial factors for
load. Then a reduced loading is selected (from BD2l for highway bridges) consistent
with the loads and load effects proved t o be acceptable in the test.

Reference is made above to the proof loads being a t least those factored for the uls. It
can be argued that applying the uls factored loads is not completely sufficient because
the theoretical probability of these loads being exceeded in service is higher than the
acceptable probability of reaching the uls in service. This is because in an assessment by
calculation, partial factors are applied t o the resistance side of the assessment equation.
However, provided the structure is showing no signs of distress a t the maximum applied
load it may be acceptable to assume intuitively that the probability of failure at a
somewhat higher load is low. The observed performance of the structure is a
substitution for the resistance/material partial factors.

10.3.3Application Attention must be given t o the criteria for stopping the test short of the factored
assessment loading. As there is no established practice for Proof Loading in the UK, there
are no criteria for stopping the test. Selection of these criteria is structure dependent
and requires knowledge of structural performance beyond the experience of most non-
specialist engineers.

Another important consideration is the form of loading t o be adopted. If this replicates

the loading configuration of the assessment loading directly, the test may be referred t o
as a Load-Type Proof Load Test. If not it is a Load-Effect Type Proof Load Test.

In the latter case, a theoretical structural model is required to convert the loads t o load
effects. Loading is applied to replicate the load effects that would arise in the structure if the
assessment-loadingconfiguration had been applied. It is considered important to derive this
model from load test results (strains, deflections, etc.) otherwise the assumed model may
be sufficiently incorrect to invalidate the test. This issue must be carefully considered, for
example, if end restraints are thought t o have a significant strength-enhancing effect in the
performance of the bridge. In that case, the uniformly distributed load (UOL) should be
obtained using the cusp method given in BD37, which compensates for the errors arising
when a UDL is substituted for the axle configuration from which it is derived.

130
 10.3.4 Current status Although tests of this type are reported t o be used in some c ~ u n t r i e s ’ ~ ~ ,‘ they
~’~’’
appear t o be the province of specialists. Proof Load Testing is not currently in use in the
UK because there are reservations about its adequacy, safety and long-term effects, and
there are no fully researched guidelines.

In 1998, the NSCLTB issued an invitation t o bridge owners (with the support of the
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Highways Agency) t o undertake Proof Load Tests on a trial basis under the supervision
of the committee. There have been no takers t o date. It was intended that guidelines
would be produced by the Highways Agency tailored t o the bridges in question. A
structure accepted for a trial Proof Load Test had t o pass onerous qualification
conditions that were supplied with the invitation. Issues t o be addressed in the
guidelines included the distinction between a Load-Type Proof Load Test and a Load-
Effect Type.

In the event that any bridge owner or assessing engineer wishes t o follow this line of
enquiry, the NSCLTB, the Highways Agency or TRL should preferably be contacted
directly. References t o load tests carried out in other countries, notably Canada, are
available.

10.3.5 Safety The safety of the structure, the test team and the public is crucial in a Proof Load Test.
All methods of load application, measurement and monitoring during the test must be
devised t o manage the risk t o an acceptable level. It is recommended that Reference
10.5 should be consulted and the most qualified level of specialists engaged t o do the
test. See also the paragraph directly above.

10.4 Full-scale or
model-scale testing t o
failure

10.4.1 Background There are numerous published examples of tests t o failure of bridge elements or models.
Generally, these are the basis of methods of design or assessment contained in
Standards. Their use directly for the assessment of a particular structure should not be
overlooked, however. Provided the structure is of sufficient importance, or present in
sufficient numbers, the use of model or component tests can be cost effective.

Published papers can be consulted t o establish what should be done and what can be
achieved. Because of the nature of this type of test, a specialist is normally required -
if the assessing engineer does not have a suitable qualified person in the team.

This type of test is primarily applicable t o cases where the load effects are known for the
structure and load case(s) in point, but the calculated resistance is insufficient and there
is reason t o believe the calculation gives an underestimate.

131
 It may also be of assistance when the behaviour of a whole structure or large section of
it is in question a t the uls. This addresses the structural model, the load effects and the
resistance in combination. A laboratory model can be constructed, a t either full scale or
reduced scale, and loaded in the service range culminating in a test to failure.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

10.4.2 Broad principles The test specimen should be designed to reproduce faithfully the structure or
component being assessed. For example, reinforcement details should be reproduced
rather than simplified unless they lie outside the failure region.

Often, for reasons of cost and practicality, it is necessary to use a scale less than unity.
As a general principle, it is recommended that the scale adopted is as large as possible,
particularly when shear is being investigated.

The availability of reinforcement of the desired size or sizes may influence the precise
scale chosen. Often it is preferable to select the scale so that a crucial reinforcing bar
can be reproduced exactly to scale, enabling the principal features of detailing to be
reproduced accurately.

It is acceptable to model a component or part of the structure provided proper attention

is given to the boundary conditions. All load effects should be reproduced in the
combination(s) present in the structure being assessed. Particular attention should be
given to free rotation and sliding a t supports if this is the configuration being evaluated.
Sufficient measurements should be taken to demonstrate that this is the case in the test.

Applied loads and reactions are the most important quantities generally to be measured
in a test to failure. Where practical, sufficient load measurement points should be used
to provide an audit and check load distribution.

Control of the test may be by load or displacement. Commonly, load control is used for
the initial stages of the test, but displacement control has advantages in the latter
stages. It is less likely to cause a sudden failure or allow creep to failure while
measurements are being made. It may give a better indication of behaviour if the
reducing capacity and ductility are of interest.

It is common practice to record the development of cracking by marking the structure at

intervals during the test. Displacement measurements are straightforward - strain
measurements on the concrete, less so in regions of tension and cracking. Reinforcing
bars can be strain gauged internally if required, but interpretation of the results can be
difficult between the loads to cause concrete cracking in the vicinity of the gauge and
yield in the bar. Specialist knowledge is required if these results are to be used in analysis.

In designing the specimen, it is appropriate usually to carry out strength calculations

using the model dimensions, as if it were the structure being assessed. Sometimes this
results in the proposed specimen having a theoretical failure in a different mode from
the structure being assessed - a fact that is important to account for in the

132
 development of the test. For instance, in a small model the presence of the depth factor
in shear may increase the resistance t o shear above the resistance t o flexure.
Consequently, the test may not achieve its objectives because the model fails in the
wrong mode.

It is sometimes acceptable t o test specimens at a relatively young age (e.g. less than 28
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

days), particularly when several tests are required and the programme has t o be
developed and specimens cast as work progresses. However, some properties of
concrete, such as bond, are known t o develop at a different rate from compressive
strength. If it is considered that a concrete property, significant t o the mode of
behaviour in question, is affected in that way, the age of the specimen should not be
significantly less than 28 days. Otherwise, ages less than 1 4 days are not advisable.

In assessing the test results, it is customary to use the mean measured strength of concrete
and steel elements from tests on specimens, cubes in the case of concrete. If Code
expressions are being used as the basis for the evaluation, all partial factors are removed to
produce the theoretical resistance. If there are implied safety factors, these must be removed
also. In BD44, the principle has been to revise the resistance equations so that partial factors
are given explicitly. This aids the process. However, it is advisable t o examine the derivation
of the formulas when this process is carried out. BA44 is an appropriate reference as this not
only explains BD44 provisions, but also provides background references.

10.4.3 Extent O f testing An important decision that may affect the practicality of testing is the number of
specimens that have t o be tested for each required loading configuration. Generally’
random factors give rise t o a spread of test results that should be taken into account,
particularly if the tensile strength of concrete contributes t o the capacity being
measured. Guidance is given on testing in BS 5400: Part 1 (and Draft Eurocode
prEN.1900) for design purposes, but for the assessment of existing structures, the
tendency is t o reduce the number of specimens t o a minimum. One approach is t o test
one specimen in the required configuration and make a decision based on the margin of
safety it demonstrates. Some repetition is always desirable.

When it is necessary t o test several load cases, for example, different shear spans in tests
for shear resistance, this can be used t o reduce the number tested for each load case, in
some cases t o one. This is because the validity of the data can be assessed by taking the
results as a set and checking that the trends are systematic and not subject t o wide
scatter from the theory proposed t o fit the data.

If the testing is intended t o apply t o a family of structures containing parameter

variations that are t o be modelled, savings in specimen numbers may arise by planning
the test programme with that in mind. However, it is generally more difficult t o
interpret test data when several parameters are varied together.

Denton eta1”12 report a short series of tests directed a t one type of reinforced concrete
slab in which a combination of proof tests and failure tests was used to advantage in
tests carried out in the laboratory.

133
 10.4.4 Application As the testing operation itself generally requires a testing laboratory to carry it out,
advice can usefully be sought from the laboratory operators on the specimen design and
testing configuration as part of the package. This assumes that the testing organisation
can provide the necessary experience and expertise, a factor that is a significant
consideration in selecting them.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

10.5 Hybrid and other This refers to testing which does not fall wholly within one of the main types -
forms of load testing Supplementary or Proof Testing. Examples arise because, for instance, the load applied
does not comply with the main types or the interpretation of the results is different. It
must be emphasised that these are theoretical possibilities that have not necessarily
been used in practice, a t least in the UK.

10.5.1 Examples One hybrid form of load test is a type of Supplementary Load Test in which the loads are
applied incrementally beyond the day-to-day loads. This enables the structural model to
be checked a t a load closer to the ultimate factored loads and, therefore, reduces the
difference between the load test and the assessment condition. If loading continues to a
sufficiently high load, the condition for a Proof Load Test may be achieved.

Another form may occur when the test is approached from the other direction, i.e. from
the Proof Load Test. In this case, the test is planned as a Proof Load Test but the
instrumentation and load application initially conforms to the Supplementary Load Test
requirements. If the load test proceeds satisfactorily to the Proof Load Test values of
applied load, the test is interpreted as a Proof Load Test. If, however, it is necessary to
stop the test before the Proof Loads are reached, it may be possible to interpret the
results as if for a Supplementary Load Test.

Bridge owners sometimes call for testing that resembles Supplementary Load Testing in
respect of the conduct of the test, instrumentation and level of loading applied.
However, a formal assessment is not carried out a t the ultimate limit state, because, for
instance, there are insufficient data from bridges tested to collapse to enable the
behaviour of the model to be established with confidence. The best that can be achieved
with this type of test is confidence in the behaviour of the structure under service loads.
The results can be used for monitoring as described in the next Section.

10.5.2 Load testing for This type of test originates from a desire to adopt a monitoring regime in accordance
monitoring with BA79. If it is considered necessary to understand how the structure is carrying the
load without distress, a form of test similar to a Supplementary Load Test may be
considered. The intention in this case is not to re-model or re-assess the structure,
although that could be done in addition.

Using an understandingof the behaviour of the bridge based on a load test enables a
more rational decision to be taken on the monitoring regime and the nature, location

134
 and alarm levels required to indicate a significant behaviour change. For instance, if
there is moment end restraint present that is contributing to a high observed stiffness of
the deck, it may be possible to provide gauges to detect any reduction, or identify the
potential location of future cracking.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

10.5.3 Reliability and Reliability considerations may be applicable to load testing and interpretation of the
updating results. For example, a paper by Fu’O13 explains this in the context of testing in the USA.
If a full reliability analysis is being carried out for an assessment (Level 5 in BA79), it is in
principle possible to introduce the results of a load test to improve the reliability
estimate obtained by calculation alone.

A paper by Das and C ~ l l i n g t o n ’ proposes

~’~ the use of Bayesian updating methods with
load testing. This form of testing has not been used in practice at the time of writing. In
principle, this entails loading being applied as if in a Proof Load Test. However, it is not
necessary to apply the full uls loads provided information is available about the
behaviour of similar bridges. It depends on a relationship being established between the
load to cause a low level of structural distress and the ultimate load - for instance
between the loads to cause non-linearity and failure. Provided the structure does not
reach the distress level under the test load, a higher test load may be inferred having
only a small probability of causing failure. If the probability of failure under the required
assessment loading is judged low enough, the structure may remain in service with an
appropriate level of monitoring if necessary. The level of monitoring is likely to be
significantly less onerous than would be the case without the testing.

10.6 Dynamic load testing

10.6.1 Structures where Dynamic testing is occasionally required when the dynamic behaviour of the structure is
dynamic response is critical a critical factor. Examples are excitation of long span bridges by wind, and the
pedestrian excitation of footbridges. The vibration of footbridges is not a strength issue.
It has been known for the vibration of concrete bridges to be reported as questionable,
but when dynamic measurements are recorded, the stresses involved are generally small.

When dynamic testing is necessary, a specialist consultant/testingorganisation should be

contacted at an early stage. Some large consultants have in-house specialists in the subject.

Measurements of the structural response can be made during service loading or under
test loading. Typical measurements include acceleration of the superstructure and
sometimes displacement at joints and bearings. Analysis of the data is likely to require
dedicated software directed, for example, to the computation of frequencies and mode
shapes.

In order to understand the behaviour of a bridge, it is helpful to carry out dynamic

calculations for comparison. Frequencies are relatively straightforward to obtain by

135
 calculation. It is more difficult t o calculate the amplitude of response. Structural
damping values can be obtained in free decay or less reliably from the analysis o f the
frequency-domain response in the mode required.

10.6.2 Dynamic tests for

Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

The basis for this form of testing is that physical changes t o a structure generally change
health monitoring the dynamic characteristics. Natural frequencies, modes shapes and damping may all be
affected. If changes occur t o the dynamic characteristics, it may be taken as a signal
therefore that structural changes have also occurred. Routine determination of the
dynamic characteristics will bring these changes t o the attention o f the maintaining
authority and may be useful for the management o f the structure.

Alternatively, changes t o the measured dynamic characteristics may indicate that

boundary conditions or environmental changes have taken place. When environmental
changes are responsible, there is a tendency for genuine structural effects t o be
obscured. In themselves, changes caused by environmental factors are seasonal and
generally of little structural consequence.

The potential for testing o f this type has been recognised for many years. Early practical
applications occurred with offshore oil installations. Compared with bridges, the
inspection of those structures is more difficult and expensive, and the consequences o f
structural failure are more critical in terms o f financial and safety respects. For bridges,
generally, the cost o f monitoring has been t o o high for the benefits that accrue.

Experience has shown that many structures are excited sufficiently by traffic or wind.
Processing of data from accelerometers enables frequency and mode shape t o be
determined. Damping can also be estimated but with less confidence.

10.6.3 Applications for It is important t o recognise that changes t o frequency and mode shape are related t o
assessment stiffness and not strength. Direct relevance t o the assessed capacity at the uls is limited
therefore. Such changes are relevant t o assessment only if it can be shown that the
source o f the changes is detrimental t o strength. This requires a secondary evaluation,
probably involving inspections and other forms o f testing. For instance, if the shear
connectors in a composite bridge have failed, it may reduce the local stiffness, which in
turn could reduce the section stiffness over a length of superstructure and affect
frequency and mode shapes. Knowing that the connectors have failed will enable a
revised assessment t o take place. Naturally, it is necessary t o show that the dynamic
changes will be detectable when the extent of connector failure becomes significant.

Another example is the development of cracking in a reinforced concrete member. It is

well known that as applied load increases, cracking may develop and spread. This
happens in laboratory tests and can happen under increasing service loads o n a bridge
particularly in flexure. There is an accompanying loss of stiffness that can be detected by
measuring changes of frequency (at least under laboratory conditions). However, the
ultimate capacity is not directly affected. At the uls, usually, the concrete will be

136
 extensively cracked in flexure even when the failure is finally in shear. This is a case
where stiffness changes are not accompanied by a reduction in strength. It is not directly
relevant therefore to assessment at the uls. Nevertheless,the bridge manager may find
the information useful if cracking is difficult to detect by visual inspection.

A case where strength changes are not accompanied by stiffness changes is local
corrosion of reinforcement. This can reduce the areas of reinforcement significantly over
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

a short length and reduce the uls capacity without significant global stiffness change.
Reliance on dynamic monitoring would not be appropriate in that case.

Calculations can be carried out to estimate the effect of structural changes on the
stiffness and mode shape. Studies have shown that, generally, the structural
deterioration has to be large and extensive to affect frequency by as much as 5%, and
environmental effects can cause more change than this. Structural changes of this
magnitude may be visible by inspection in a bridge.

Large, long span bridges with clean lines and efficient bearings may be reasonably
stable and/or behave predictably in the presence of temperature changes. Structural
changes from strengthening may increase stiffness sufficiently to be detected above
the environmental changes. Nevertheless, in practical terms, the detection of strength
changes resulting from deterioration that could not be detected by other means is
not considered likely. Smaller bridges such as motorway overbridges are more likely
to be affected by thermal changes, particularly when the surfacing is a substantial
thickness.

10.6.4 Case studies and Conference papers and journal articles can be found that describe techniques of
research in the literature measurement and analysis, and the results of testing on a large number of structures.
For health monitoring applications, what are absent are clear-cut examples where:

0 The testing was for bridge management and not a research exercise; and was not
associated with a problem of dynamic ‘behaviour.
0 Changes to dynamic characteristics were the first indications of a potential problem
0 The changes were structurally significant and not detectable by other means, for
instance visual inspection.
0 Dynamic response was able to identify and quantify structural deterioration that
involved loss of capacity - and this was subsequently corroborated, and remedial
action taken.

I A conference paper by B r o w n j ~ h n ’ provides

the topic and an extensive bibliography
~’~ a useful summary and introduction to

137

I.,
 10.7References 10.1 CULLINGTON, DW and BEALES, c, Is your strengtheningreally necessary’ Proceedings oftheConferenceonBridge
modifcation. Institution of Civil Engineers, 23-24 March 1994, Thomas Telford. 1995, pp 270-284.
10.2 HIGHWAYS AGENCY, BA 54, Load testing for bridge assessment, Design Manualfor Roads andBndges.
10.3 HIGHWAYS AGENCY, BD 21, The assessment of highway bridges and structures, DesignManualforRoadsandBridges.
10.4 HIGHWAYS AGENCY, BD 34, Technical requirements for assessment and strengtheningprogramme for highways
structures Stage 1 - Older short span bridges and retaining structures, Design ManualforRoads andBridges.
10.5 NSCBLT 1998, Supplementaryloadtesting ofbridges. The Institution of Civil Engineers National Steering Committee for
the Load Testing of Bridges, Thomas Telford. 1998, p 73
10.6 HIGHWAYS AGENCY, BA 54, Load testing for bridge assessment, DesignManualforRoadsandBridges.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

10.7 NSCBLT 1998, Supplementary load testing ofbridges, The Institution of Civil Engineers National Steering Committee for
the Load Testing of Bridges, Thomas Telford. 1998, p 73.
10.8 HIGHWAYS AGENCY, BD 21, The assessment of highway bridges and structures, DesignManualforRoadsandBridges.
10.9 MINISTRY OF TRANSPORTATION,ONTARIO, Ontario highway bridge design code, 1992
10.10 NATIONAL CO-OPERATIVE HIGHWAY RESEARCH PROGRAMME, Manualfor bridgerating through loadtesting,
Research Results Digest Number 234, , Transportation Research Board, National Research Council. Washington DC,
November 1998
10.11 NEW ZEALAND STANDARDS, ConcreteStructuresSrandardPart I - thedesign ofconcretestructures,1995.
10.12 DENTON, SR, IBELL, TJ and POSNER, CD, Full-scale model testing to investigate the shear capacity of concrete slabs
with inadequately anchored reinforcement,In Bridge Management4 (Ryall, MJ, Parke, CAR and Harding, JE. eds),
Thomas Telford, 2000, pp 55-62
10.13 FU, C, Highway bridge rating by non-destructive proof-load testingfor consistent safety, Research Report 163, Transport
Research and Development Bureau, New York State Department of Transportation, 1995.
10.14 DAS, PC and CULLINCTON, DW, Proof load testing and Bayesian updating, Bridge rehabilitationin the UK;review ofthe
currentprogramme andpreparingfor thenext, Institution of Civil Engineers, 2-3 October 2000.
10.15 BROWNJOHN,JMW, Dynamic performance and characterisation of highway bridges in Singapore, In Currentandfuture
trends in bridge design constwctionandmaintenance. Singapore, Thomas Telford, October 1999.

138
 11. Re iabil ty and risk-based techniques

This Chapter provides a guide on the use of reliability and risk-based techniques for the
assessment for bridges.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

11.I Background The process of assessment is of crucial importance for maintaining highway bridges in a
safe and serviceable condition. The objective of assessment is t o evaluate the safety of
an existing bridge quickly and with a minimum of effort. However, the assessment rules
and criteria need to be established rigorously and judiciously. If assessments are unduly
conservative, structures will be unnecessarily strengthened, or needless load restrictions
will be imposed. Conversely, if the rules are too lax some bridges could actually fail
during service” ’.
Codes and standards for design and assessment employ partial safety factors t o ensure
an appropriate level of safety for bridges. These factors guard against extreme variations
in design parameters (e.g. material properties, applied loads, etc.), which could occur
during service. In order t o ensure that the design rules are simple for routine use the
values of the partial factors are chosen such that they cater for a wide range of
structure/component types and failure modes. O f necessity, therefore, the rules tend t o
be conservative for the majority of bridges and the level of conservatism varies
considerably from structure t o structure.

In current assessment practice, the safety is checked mainly at a component level using
in most cases elastic methods of analysis and, therefore, does not take account of the
reserve strength of the whole bridge system. Only in the case of failure of reinforced
concrete slabs is the use of yield-line analysis t o qualify the ultimate strength becoming
more common. Also the assessment criteria do not take into account the consequences
of failure and so, for example, a bridge on a motorway and a small culvert on a local
road are assessed t o the same criteria” ’.

There have been no reported cases in the UK of highway bridge collapses as a result of
traffic overloading, however element distress has been observed in some instances. The
design rules are considered therefore t o provide an adequate level of safety. However,
this does not necessarily mean that the safety level is optimal; it may well be
conservative and the extent will vary for every structure. For many bridges that have
been in service for a number of years and have shown no signs of distress, later
assessments have shown low or no apparent live load capacity! In order t o ensure that
the assessment rules provide a more consistent level of safety across the bridge stock
and t o avoid unnecessary strengthening costs due t o over-conservative assessments, it is
necessary t o develop criteria which can be tailored to each bridge by taking into account
its particular safety characteristics” 3.

To address these issues, the Highways Agency’s Advice Note BA79/98 sets out a
philosophical approach t o assessment and defines the principle of different levels of
assessment” as follows:

139
 U Level 1 assessments are carried out using simple analysis methods, in accordance with
the assessment Standard BD21 and the accompanying Advice Note BA16.
0 Level 2 assessments involve more refined analysis and the use of characteristic
strengths of materials.
0 Level 3 introduces the concept of Bridge Specific Assessment Live Loading and makes
use of material test results t o determine worst credible strength values.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

0 In addition Levels 4 and 5 were included which were based on using structural
reliability methods. However, these specific methods have'been dropped in the draft
amended version of BA79 (see Section 12.1 and Reference 12.5) and reference made
t o the general use of structural reliability methods.

Whilst structural reliability has been used in other fields, notably for offshore structures,
there have been very few applications for bridges in the UK.

11.2 The appropriate In general, reliability and risk-based assessments should be considered only after generic
application of reliability assessments have been completed and the bridge fails t o meet the general assessment
requirements.
and risk-based assessments
A reliability analysis can also be performed t o provide further assurance on the results
from lower levels of assessment.

An assessment that makes use of bridge specific partial safety factors can provide
particular benefit in the following cases:

U The consequences of failure of a bridge are significantly lower compared t o a major

bridge on a busy motorway or a bridge over a busy railway line.
U The bridge is known to have carried unrestricted traffic over a reasonable period of
time (e.g. 5 years) without damage/distress but the calculated live load capacity
factor, K, suggests a significant shortfall of capacity (e.g. K < 0.7).

In addition t o the above, a detailed reliability analysis may be beneficial in the following
cases:

U Bridge specific data are available on material properties, dimensions and traffic loads.
0 Reliable information is available on abnormal heavy vehicles passing the bridge.
0 Results of supplementary or proving load tests are available.
0 The critical component is known t o have suffered deterioration but the extent of
deterioratioddamage cannot be estimated accurately.
0 When weight restrictions suggested by more general assessments need t o be relaxed
t o minimise traffic disruption.
0 Results of model tests of adequate scale with similar section details and support
conditions t o that of the bridge element under consideration are available.

140
 11.3 Overview of
reliability-based
assessments

1 1.3.1 General The risk of structural failure depends on the uncertainties in load and resistance parameters
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

as well as other factors such as gross errors. Quality control and verification are used to
control gross errors, while limit state designs, or alternatively, structural reliability index can
be used to mitigate failure due to normal variability. An overview of the application of
structural reliability methods is given in the following Sections.

11.3.2 Uncertainties In A number of uncertainties influence the assessment of safety of a bridge and its
bridge assessment components. These uncertainties arise from inherent variability of loads and material
resistance parameters, uncertainties in deterioration over time, uncertainties in the
analysis models used for determiningthe load effects and capacities, and the
uncertainties in measurement and inspection techniques. In view of these uncertainties,
it is possible that an element could fail from an adverse combination of extreme values
of the variables. This is illustrated in Figure 11. I , in which the overlap between the
loading and resistance distributions represents the failure region and the area of overlap
between these distributions is related to the probability of failure.

In a deterministic method of assessment, the safety of a bridge component is ensured

by using a number of partial factors to guard against extreme variations of the random
parameters. In a reliability analysis, which is based on a probabilistic approach, the
partial factors are not used. Instead, the uncertain parameters are modelled using
appropriate probability distribution functions, and the probability of failure of the
component is calculated.

Figure 11.1
Illustration of the concept of probability of failure.
0.0020

0.0016

.-
h
+
E
+! 0.0012
-n.-.-
Y
2,

n
n
P 0.0008

0.0004

0.0000
0
Shear force (kN), shear capacity (kN)

141
 11.3.3 Reliability Index In order t o carry out a reliability-based assessment of a bridge, all the relevant failure
modes for each critical component of a bridge are identified and a reliability analysis is
carried out separately for each failure mode.

It is common t o express the result of a reliability analysis in terms of a ‘Reliability Index’,

p, which is mathematically related t o the probability of failure, P,:
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

where #J ( . ) is the cumulative distribution function of a standard normal variable. This

relationship is shown in Figure 11.2. Note that P,decreases as P increases.

Figure 11.2
Relationship between reliability index and
probability of failure.

I I I I I I I I \

l.OE+lO 1.OE+09 1.OEi-08 1.OE+07 1.0€+06 1.OE+05 1.OE+04 1.OE+03 l.OE+OZ l.OE+Ol 1.1

Probability of failure

The reliability index, or probability of failure, should be defined with respect t o a specified
time interval. When the reliability is evaluated for a reference interval of one-year, it is
called an ‘annual reliability index’; and when it is evaluated for the lifetime of the structure,
it is called a ‘lifetime reliability index’. The annual reliability index is used in determining
the adequacy of a bridge. The corresponding annual probability of failure represents the
likelihood of the failure mode occurring in any one year. For a non-deteriorating element
where traffic live load is the only time varying load, the annual reliability index is calculated
considering the maximum live load event that could occur within any one-year interval.

In order t o determine the structural adequacy of a component the calculated reliability

index needs t o be compared against a specified minimum acceptable value called the
‘target reliability index’.

It should be noted that the results of reliability analysis are very sensitive t o the
resistance and load effect models used, the probability distributions for the basic
random variables and, t o a lesser extent, the method for calculating probability of failure
or reliability index. To ensure that assessments are carried out on a consistent basis, it is
important that a standardised procedure is followed and for the assessment of Highways
Agency structures this might be as given below.

142
 1. Identify critical component(s) in a bridge based on previous assessment results.
Identify the relevant failure modes for this component.
2. Select a failure mode and identify the relevant basic random variables.
3. Where the structural response is linear, compute the load effects a t the critical
section using a linear-elastic grillage or finite element analysis separately for
structural dead load, superimposed dead load and surfacing load components. These
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

results can be taken from previous assessments, where available.

4. Evaluate the load effects a t the critical section due t o the ‘Basic Static Live Load’.
5. Choose appropriate probability distributions for the basic random variables.
6. Select an appropriate method for reliability analysis and any one of the available
commercial software systems, Develop sub-routines for the failure mode function
and link it with the selected computer program, as required. (This needs t o be done
only once for each bridge and failure mode.)
7. Perform an initial reliability analysis for the selected failure mode using the
probability distributions from Step 5 and any other deterministic parameters as
required. Determine the annual reliability index, P, and the sensitivity factors, Lyi, for
each random variable.
8. Design a ‘calibrator’ element such that the element satisfies the assessment
requirements exactly. Perform reliability analysis of the calibrator element using the
probability distributions from Step 5 and any other deterministic parameters as
required. Determine the annual target reliability index, PO.Modify this value based
on consequences of failure t o obtain the final target reliability index, PT.
9. Collect additional datahnformation on material strengths, loads, etc. specific t o the
bridge being assessed. The results of sensitivity factors from Step 7 should be used
t o identify what data it would be most beneficial t o collect.
10. Re-calculate the reliability of the component by incorporating bridge specific
datahnformation using Bayesian updating methods.
11. The component passes the assessment for the selected failure mode if P > &. If
not, consider if additional data collection would be beneficial and refine the
reliability estimate as in Step 10.
12. Repeat Steps 2 t o 11 for each selected failure mode and component of the bridge.

11.3.4 lnterpretatiOn O f The probability of failure computed from a reliability analysis should be treated as a
Reliability Analysis Results ‘notional’ value t o be used in a relative sense t o compare reliabilities of different
structures. It should not be interpreted as a measure of the frequency of failure that
could be expected in service, i.e. in the sense that one out of so many bridges would fail.
This is because each structure is unique and, hence, a population cannot be readily
defined. Furthermore, the probability distributions used in a reliability analysis are
intended t o represent inherent variability in design parameters that could be expected
for structures that are designed, constructed, operated and maintained t o good
engineering practice. They do not account for ‘gross errors’ in design (e.g. incorrect
calculations of loads or capacities); construction (e.g. the use of wrong steel
reinforcement, missing bars, poor quality of construction); or in operation (e.g. gross
overloading of vehicles) which are seen as the cause of most structural failures observed
in practice.

143
 aa ReLlabiRity and risk-based techniques

Nevertheless, by treating the underlying uncertainties on a consistent basis the reliability

method gives a rational basis for comparing different components and different
structures. Guided by engineering judgement and experience, the method provides a
rational basis for evaluating the margin of safety for a given failure mode and t o
demonstrate structural adequacy. This provides a formal link between Codes of Practice
and reliability requirements. The reliability analysis can also be used as a decision tool t o
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

compare technical alternatives in terms of their cost and risks. Flaig and Lark” propose
a framework for the incorporation of such an approach into a decision suppor;t system
._
for bridge management.

Reliability-based techniques provide a greater flexibility for incorporating service data

(e.g. load testing, material testing, measurement of dimensions, deterioration, etc.) and
any additional safety characteristics (e.g. warning of failure, consequences of failure) for
a specific structure and thus provide a more rational assessment of its adequacy.

11.4 Acceptance Criteria In order t o judge whether the calculated value of a reliability index for a particular bridge
is adequate, it is necessary t o specify the minimum acceptable value of the reliability
index, often termed the ‘target reliability index’.

A number of Codes and Standards and other published works specify target reliability
values for bridges and other types of structures. However, it is necessary that the
specified target values are consistent with the method of reliability analysis, in particular
the probability distributions used for the various basic variables and the models for the
calculation of capacity and load effects. For this reason, published values cannot be
readily used.

In choosing a value for target reliability, one is essentially faced with the question of
How safe is safe enough? This involves complex technical, social and economical issues
and requires value judgements.

Generally, there is a consensus among researchers that the following factors should have
an influence on the choice of target reliability index:

1. Consequences of failure
2. Reserve strength and redundancy
3. Warning of failure
4. Inspection and monitoring measures
5. Marginal cost of increasing the reliability
6. Acceptability of risk by the society.

It is important t o consider the consequences of both element and overall collapse of a

bridge. The collapse of a bridge can result in fatalities and injuries, disruption t o traffic,
environmental damage in some cases, loss of confidence in the profession, adverse
publicity, enquiry and litigation, changes t o Codes and Standards, assessment and
strengthening of other similar bridges. Element failure or distress can lead t o temporary
closure of the bridge, thereby causing traffic disruption.

144
 Factors 2 and 3, in turn, influence the consequences of failure. Inspection and
monitoring procedures help t o detect signs of distresddamage a t an early stage, thereby
preventing complete collapse of a bridge. They are only effective if significant warning is
available following an element failure.

Factor 5 brings the economic reality into the picture. The marginal cost of increasing the
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

reliability is considerably higher for an existing structure compared to new build. This
supports the use of marginally lower target reliability for existing structures compared t o
the target used for the design of new structures of the same type.

The issue of social acceptability of risk is very complex; but is known t o be influenced by
factors such as: how frequently the failures occur; the number of fatalities and injuries;
the scale of resulting disruption t o public; whether the risk is voluntary; whether the
failure is due t o man-made or natural causes; publicity from the media; the resulting
costs t o the public of increasing safety; etc. It is believed that society expects a higher
level of safety for buildings and bridges compared t o risks accepted in other aspects of
daily life.

One view is that the safety level implicit in current design/assessment practice is about
right since this has been a result of years of experience, both good and bad. Therefore
when a new approach for design/assessment is introduced, it is strongly desirable that
the resulting reliability levels in the new method are on average of the same level as
those implicit in current Codes and Standards. Any changes in reliability, where
appropriate, should be introduced in a gradual manner backed-up by careful monitoring
of selected structures over a period of time.

In principle, three approaches can be recognised for establishing target reliability values,
see for example Ditlevsen and Madsen” 6 :

1. Socially acceptable risk levels from historical data

2. Calibration t o existing Codes and Standards
3. Economic optimisation.

The first approach is limited in that it cannot easily be related t o the ‘formal failure
probabilities’ computed from a reliability analysis. Furthermore, as bridge failures due t o
overloading have not occurred in UK,direct calibration against bridge failure statistics is
difficult.

In using the economic optimisation approach, the difficulty is in the accurate evaluation
of all direct and indirect consequences of failure. This approach is satisfactory where
economic losses dominate over life, limb and social consequences; otherwise the
approach becomes controversial.

The second approach, although not free from difficulties, seems t o be the only
practicable way forward at the present time. This approach was used for deriving partial
safety factors in BS 5400: Part 3 (steel bridges) and for a number of North American
Codes for bridges and buildings.

145
 It is strongly suggested that results from a reliability-based assessment should not be
used in a strict pass/fail manner and Lark and Flaig” discuss the use of such results to
aid bridge management. The results should form inputs to the decision-making process
for choosing between alternative remediaVmitigation options for the bridge. An example
of the application of such an approach is described in Reference 11.8.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

11.5 Bridge Specific In addition to utilising bridge specific material strengths and loads in an assessment of a
risk-based assessments bridge, there may be justification for further rationalising the assessment criteria by
taking into account factors such as age of the bridge, loading history, consequences of
failure, reserve strength and redundancy, warning of failure, and inspection and
monitoring regime. In the following paragraphs, modifications to the assessment criteria
are given for consequences of failure and loading history.

Modification based on consequences of failure

Bridges with low consequences, for example those carrying a minor road over a small span,
could have a lower target reliability than a bridge carrying a motorway so that the resulting
risks are the same for the two bridge types. By maintaining the same level of target risk, R,,
the target probability of failure, P,, for a bridge with consequences of failure, C,, can be
determined. By developinga relationship between reliability index, p, and Live Load
Capacity Factor, K, the corresponding modification of the K-factor can be determined.

Modification based on loading history

Similarly, if the bridge has been in service for a sufficiently long period of time, it can be
reasonably expected that the bridge would have been subjected to some extreme
loading. Using Bayesian reliability analysis procedures, the variation in annual reliability
index for bridges of different age (or service exposure) and with different calculated K
values can be developed. The typical result is shown in Figure 11.3.

Figure 11.3
Typical variation of Kfactors based on
the age of a bridge.

KMcd ; iK
t . ..... , .: , ....
, , ,, , .... , , , , ..., , , ,, ,,,, , , , ,

Live load capacity factor K

146
 The lower curve uses no information about the service history of a bridge, while the
upper curve assumes that the bridge has been carrying unrestricted traffic for over five
years at a level similar t o the present-day traffic. Although the bridge may have been in
service for longer than five years, it is likely that the traffic in the past has been lighter
and hence this is n o t taken into account.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

It can be seen from Figure 11.3 that significant improvement in K can be obtained when
the K,,,, value is low. The benefit of Bayesian Updating reduces as the calculated K,,,,,,,
value approaches 0.7. Therefore, the increase in the K factor obtained using this method
may not be adequate in most cases for a bridge t o pass the assessment. However, this
may help in justifying a higher weight restriction o n a bridge than what would otherwise
be dictated by a more general assessment.

Reliability and risk-based assessments require considerable statistical expertise and, in

general, expert help should be sought along with client approval before using either of
these techniques.

11.6References 11.1 DAS, PC, Development of bridge-specific assessment and strengthening criteria, in SafetyofBridges, (Das, PC, ed.),
Thomas Telford, London, 1997.
11.2 SHETTY, NK, CHUBB, MS and HALDEN, D, An Overall Risk Based Assessment Procedure of Substandard Bridges, Safety
ofSridges, ICE, London,July 1997
11.3 SHETrY, NK. CHUBB, MS and MANZOCCHI,CME, Advanced methods of assessment for bridges, International
Conference on Management ofSridges, ICE, London, 1998
11.4 DAS, PC, Development of a comprehensive structures management methodology for the Highways Agency,
Management o,fh;qhwaystructures, (Das, PC, ed.), Thomas Telford Publishing, London, 1999.
11.5 FLAIC, KD and LARK, RJ,The Development of a Risk-Based Decision Support System for Bridge Management,
Proceedings of the Institution of Civil Engineers- Bridge Engineering, 2005.
11.6 DITLEVSEN, 0 and MADSEN. H 0. StructuralReliabilityMethods, Chichester, John Wiley, 1996.
11.7 LARK, R.J.and FLAIC, KD, The use of reliability analysis to aid bridge management, TheStructuralEngineer,Vol. 83, No 5,
March 2005
11.8 CHUBB M and SHETTY. NK, Reliable Bridge Assessment, Concrete Engineeringinternational July/August 2000.
~
12 Bridge management and assessment

12. Bridge management and assessment

12.1 General PritlCipkS This Chapter considers how the assessment process can be integrated within bridge
management activities, by providing advice on the management of sub-standard bridges
and the influence of assessment results on maintenance activities. An introduction to the
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Assessment Concept was set out in Chapter 1 where the Need for Assessment, Staged
Assessments and Assessment and Asset Management are particularly relevant to Bridge
Management and Assessments. In this Chapter, only general principles will be established
in relating bridge management to assessment; these should be read alongside the
guidance given in Chapters 1 and 2, with relevant Standards (BDs) and Advice Notes
(BAs) being listed in Section 1.4. Reference should also be made to the Code of Practice
for the Management of Highway Structures published in September 200512’.

12.1 .I Managing The Highways Agency Advice Note BA79, The Management of Sub-standardHighway
sub-standard bridges Structures, (the Advice Note) was published to provide guidance on the implementation
of interim measures for structures which were assessed to be sub-standard, or
provisionally sub-standard, in order to maintain the safety of the highway network. Its
use is applicable to virtually all highway bridges.

The substantial completion of the assessment programme has left a large number of
bridges that have failed the assessment process, but there is little prospect of these
being strengthened in the near future. The assessment code, BD21, requires that in the
event of an assessment failure, one of a series of formal interim measures is imposed on
the bridge. Many bridges are unsuitable for propping, and weight restrictions or lane
closures would often cause significant disruption to traffic flows. The diversion of traffic
away from a restricted bridge may create additional problems for safety on diversion
routes. Consequently, many Authorities had been operating monitoring schemes with
the intention of allowing bridges to remain open to unrestricted traffic. The publication
of the Advice Note provided, for the first time, a rational method for considering the
whole concept of sub-standard bridges and provided a framework for instigating a
robust monitoring regime in approved cases’22.

12.1.2 The BA79 Advice The Advice Note is divided into five main chapters:
Note
1. Assessment (see Chapter 1)
2. ‘Immediate Risk Structures’
3. Interim Measures During Assessment
4. Interim Measures on Completion of Assessment
5. Prioritisation for Strengthening (not considered in this Guide).

This document makes more reference to the role of the Technical Approval Authority
(TAA) than any other Highways Agency Standard or Advice Note (other than BD2). It

148
 also leaves a number of areas open for interpretation. It is necessary, therefore, for
bridge owners, Technical Approval Authorities and bridge managers t o consider their
respective duties. In practice these three roles are often combined. Some of these
area for interpretation are discussed further below (refer also t o Chapter 2). It is also
important that, for Local Authorities, locally elected members are aware of the steps
that need t o be taken t o ensure the safety of the highway network. Diagrams, derived
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

from Appendix A of the Advice Note, illustrating the procedures are given in Figures 2.1
and 2.3. Experience with the use of the Advice Note is contained in a document
produced for the Highways Agency by Parsons Brinckerh~ff’‘~.This showed that there
had been an inconsistent application of the advice and that further guidance was
required.

12.1.3 Managing the results A key aspect of the Advice Note is the need t o keep adequate records throughout the
of assessments assessment process, with details of the decisions taken at each stage. This is done by the
use of a proforma (known as Form E l ) throughout all of the processes shown on the
flow chart (Figure 2.3). The use of this proforma is recommended for use by a Highway
Authority irrespective of the ownership of the bridge. The Form E l should be kept on
the bridge file at each stage or level of the assessment process. Where the structure is
considered t o be ‘monitoring-appropriate’ then a monitoring procedure document
(see below) should be appended t o the Form E l . The following procedures are
recommended:

0 The level of assessment completed should be shown on the assessment certificate

0 Amendments t o Approval in Principal (AIP) forms are required for each level of
assessment proposed by way of an addendum t o the original AIP.
0 Sub-standard bridges must be restricted t o Authorised Weight Regulations (AWR)
vehicles if no lower limit is imposed.

One objective of the Advice Note is t o limit the disruptive effect of weight restrictions
by the use of monitoring on appropriate structures. The use of abnormal indivisible loads
(AIL) on Special Type General Order (STCO) vehicles in the highway network is
increasing and the selection of suitable routes is not made any easier by the large
number of sub-standard bridges that exist. In particular circumstances, this problem can
be overcome by a special assessment for a specific vehicle carried out in accordance
with BD86/01. The existence of one AWR restricted bridge on a motorway link can
create significant problems for the alternative AIL route on the parallel local road
network. Annex D t o BD86/01 provides advice on the management of abnormal loads.

12.1.4Immediate risk These structures are where an immediate and unacceptable risk t o public safety is
structures identified (BA79). The diagnosis should be confirmed by the TAA. The Advice Note
recommends that “factors such as the nature of the structural weakness, any
corresponding signs of distress, the recent load history of the structure and the level of
assessment completed should be taken into account”.

149
 Bridge management and assessment

1 2.1.5 Interim meaSUreS The assessment process can involve several stages of work and be extended over a
during assessment considerable period of time. It is recommended, therefore, that interim measures are
implemented whilst the assessment process is completed, unless the structure is
considered t o be ‘low risk’. The use of this procedure might require, for example, the
erection of weight restriction signs and their subsequent removal without any change in
the condition of the bridge. The bridge owner should be prepared t o explain these
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

essential safeguards t o elected Councillors or Members of Parliament.

The definition of ‘low risk’ requires that there is likely t o be a considerable reserve of
strength inherent in the structure. It should be remembered that at this stage of the
process it will not have been possible t o have completed the full range of options open
t o the assessment engineer. Therefore, the definition can only be subjective. The Advice
Note also suggests that structures can be considered t o be ‘low risk’, “where
consequences of failure are low or where live load capacity factor is greater than 0.7”.

12.1.6 Interim MeaSUreS On The purpose of these measures is t o reduce the risks of a failure t o acceptable levels
comp[etion of assessment until strengthening is carried out. The Advice Note introduced the definition of
‘Monitoring - appropriate Structures’. In order t o qualify as ‘Monitoring - appropriate’,
structures need t o satisfy all of the following criteria (except in the case of small spans
generally less than 5 m):

0 No significant signs of critical distress

0 Hidden problems unlikely t o be present
U Predictable and gradual failure mode
0 Monitoring should be meaningful and effective.

The use of monitoring is treated as a ‘Departure from Standard’. Further guidance is

contained in Appendices C and D of the Advice Note.

The assessment engineer may be able t o offer alternative Formal Interim Measures. For
example, a bridge may be assessed at 7.5t with three lanes of traffic but at 18t with only
two lanes of traffic. The lower Assessment Live Loading (ALL) will handicap buses and
most commercial vehicles but permit unrestricted use by private cars. Conversely, the
higher ALL is suitable for most buses and a large number of commercial vehicles but the
reduction in traffic lanes may cause queuing of private cars a t peak times. The bridge
manager will need t o discuss the options with the highway manager and record the
outcome on Form E l . The lower the level of posted weight restriction, the more likely it
is that it will be ignored.

Very low levels of weight restriction (3t) allow an additional width restriction t o be used.
Generally, this ensures enforcement of the weight limits but the physical restriction may
need t o be substantial t o deter drivers determined t o avoid using a diversion route.
Higher levels of weight restriction are more difficult t o enforce and a risk management
exercise should be carried out t o consider the effects of possible repeated abuse.
Monitoring and further assessments may also be required.

150
 Weight restrictions are imposed by the use of temporary traffic regulation orders. These
are normally only rated for 18 months unless extensions (up t o six months) are granted
by the Secretary of State. Failing this, they either have t o be removed or made
permanent. Objections may be raised t o a permanent order that could, in an extreme
case, result in a public inquiry with the obvious programming and resource implications.
It is important that legal sections preparing these orders are aware of the structural
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

problems and do not include exemption clauses that are common in environmental
weight restrictions.

12.1.7Assessment and Section 1.1 introduced the concept of Asset Management, i.e. the need t o increase
maintenance the overall value of the bridge stock, the asset, but a t lowest whole life cost. The
concept is developed further in Section 1.5, which referred t o the work carried out by
Flint and DaslZ4and the publication of a draft advice note (BA81) by the Highways
Agency .

The assessment rating of an individual bridge should be included in the calculation of a

bridge condition index, which can then be used t o monitor the effectiveness of
maintenance over a total bridge stock. Discussion of these techniques is beyond the
scope of this Guide but the reader is referred t o the output of the BRIME project (see
Section 1.8) for further details. The reader should also refer t o the Code of Practice for
the Management of Highway Structures”’ which includes many recommendations for
good highway structure management practice set out in an asset management
framework.

In response t o the findings of Parsons Brinckerhoff‘s review of the application of BA79

123, BA79 is currently being updated in consultation with structure owners and other

stakeholders. The Advice Note BA79 will be republished as a new Standard, BD79,
incorporating much of BA79. The background t o this new Standard is described by Shave
’.
et aP2 The new Standard establishes mandatory, process orientated, requirements
together with guidance on how these can be fulfilled. Guidance on record keeping is
improved, and management processes and responsibilities have been clarified. The
following guidance is based on the requirements of the current Advice Note, BA79.
Much of the guidance will remain broadly valid when the new Standard is published. In
addition, many of the issues identified with BA79 in the text will be addressed in the
new Standard.

12.2 Monitoring Monitoring is a process that is instigated following concern over the performance or
integrity of a structure following observations made from a routine or special inspection
and/or the outcome of an assessment. The transition from inspection t o monitoring is
often blurred; due t o the dilemma over what frequency of monitoring is appropriate for
a specific set of circumstances. Monitoring is distinctly different however t o routine
inspection and should comprise two principle components; it must be either more
frequent than the norm or it must involve special targeted scrutiny. Useful general
guidance on monitoring is given in References 12.3 and 12.6.

151
 a2 Bridge management and assessment

As risk is related to the probability of an event occurring and time, the frequency of
monitoring is a key parameter that needs to be determined and this is an area that
warrants careful consideration. The factors that should be taken account of include:

0 Load assessment capacity rating

U Structural and material condition
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

U Deterioration rates
0 Interim measures
0 Road or rail classification
0 Possible failure mechanisms
0 Obstacles being crossed.

Based on these factors, a monitoring interval needs to be identified, which should range
between immediate action to a maximum monitoring interval of six months. The
monitoring frequency should be subject to review by the engineer and adjusted, either
from the outset, based on engineering judgement, or as an action following from the
results of monitoring.

12.2.1 Appropriate BA79 Appendix D’27,goes into some depth into the forms of bridge or bridge elements
monitoring that are ‘monitoring appropriate’, but in summary the following structures are likely to
meet the relevant criteria:

0 Reinforced concrete slab bridges or composite steel and composite slab bridges with
theoretical longitudinal or transverse flexural inadequacy, especially where adequate
continuity exists over the supports.
0 Bridges in which the structural inadequacy is rooted in an element or connection
whose failure would not precipitate sudden collapse and whose failure can be
observed by monitoring. The crucial feature is that the bridge will retain a substantial
proportion of its load carrying capacity following element/connection failure until the
failure is detected and safeguarding measures are implemented.
0 Bridges of small span, generally less than 5 m, which are in sound condition and
where the consequences of failure in terms of death and injury or traffic delay costs,
etc., are low.

Those bridges that are not normally ‘monitoring appropriate’ include bridges “that are
sub-standard by virtue of tension, shear or anchorage inadequacies where failure in
tension, shear or anchorage would precipitate collapse of the bridge”.

However, as recognised in Appendix D in BA79, in reality the situation is more complex.

So, these basic criteria should only be used as a guide to inform the assessing engineer’s
judgement, as to whether an appropriate monitoring regime can be initiated that is both
meaningful, effective and that will give adequate forewarning of distress or failure. The
following sections outline some of the key issues relevant to determining appropriate
monitoring.

152
 12.2.2Classes and frequency Appendix C of BA79 states that the frequency of monitoring should be as follows:
of monitoring
Class 7 - “observations are normally carried out at intervals of weeks ora few months
andshould normally be more frequent than fora structure which meets the requirement
of BD2 1I’

Class 2- “depending on the bridge, from periodic visits at intervals of several months, to
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

morefrequent visits or to continuousmonitoring”

Class3- ‘yrequent or continuous”

Although there is no maximum interval specified in BA79 for any of the classes of
monitoring any interval greater than six monthly should be carefully considered; as it is
not within the spirit of the advice note and poses the question whether such inspections
are distinguishable from the normal scope of a superficial or general inspection. Specific
care should be taken to examine the structure for signs that the identified structural
inadequacy is causing distress. The frequency of monitoring inspection should be
appropriate to the mode and likelihood of distress or failure. Experience from testing and
in-service performance of similar structures can be a particularly useful guide in this
context.

As well as considering the various timings, e.g. weekly, monthly, quarterly, etc.,
monitoring inspection during or likely unusual loading or environmental events should
also be considered.

12.2.3 Application O f The application of monitoring either in isolation or in combination with other measures
monitoring is the most commonly used method of managing sub-standard structures. However, as
is recognised in BA79, ensuring the safety of a sub-standard structure through
monitoring is a complex process and requires in-depth knowledge of its structural
behaviour. The process of monitoring should aim to verify structural behaviour
characteristics by using a variety of investigative techniques to evaluate:

U Loading
0 Structural response to loading
U Ground or foundation movement
0 Structural response to ground or foundation movement
0 Environmental factors, e.g. temperature
0 Structural response to environmental factors
U Defects, deterioration and their development
0 Bearings and other restraints
0 Structural response to bearing actions and other restraints.

Using these techniques, the monitoring process must also be properly evaluated by:

0 Gathering and reviewing records, facts and data (including in-service performance)
0 Assessment of the theoretical and actual structural response

153
 Bridge management and assessment

0 Forming an hypothesis
0 Checking results and observations
0 Drawing conclusions.

What is learnt from the above evaluation will need to be compared with the known
service and theoretical performances, whether any loading restrictions or interim
supporting measures need to be implemented and the development of a monitoring
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

specification.

The form of monitoring must be appropriate however to enable a sound assimilation of

the performance and functionality of the bridge to be formed.

12.2.4 Permanent Permanent monitoring should be used for selected, severely deteriorated bridges and for
monitoring critical bridges. Generally, such bridges would fall within BA79 Class 2 and 3, where
safety of the structure is of paramount concern.

Safety monitoring is appropriate when failure may occur through the worsening of
damage. Growth of the damage should be reliably measured by an instrumentation
system, and the measurement must allow a simple interpretation of damage severity.
Overall, the monitoring system must offer an assurance of safety.

The implementation of a safety monitoring regime requires a preliminary, five-stage

analysis:

1, Analysis of the bridge condition.

2. Study of the various structural responses and possible failure paths.
3. Choice of the failure path(s) having the greatest probability of occurring.
4. Selection of the measurable quantities that reveal an adverse evolution of damage in
the bridge.
5. Selection of threshold values of the physical effects that will require intervention in
terms of emergency repairs, restriction of loads or closure of the bridge.

12.2.5 Monitoring Within Appendix C of BA791Z7,it is stated that for each sub-standard structure being
specifications managed through monitoring, a clear, unambiguous procedural plan should be prepared
detailing the monitoring specification. Unless the monitoring is merely intended to
check that other forms of interim measures are continuing to function satisfactorily, the
plan should address the following:

0 The basis of the assessment inadequacy, clearly and concisely stated.

0 An appraisal of the reasons for the observed satisfactory service performance.
0 The anticipated mode or modes of failure.
0 A description of the parameters to be monitored and their relationship to the
predicted mode of failure and progression to that state, together with a required
frequency of observation and frequency of monitoring.

154
 0 A description of the ranges of observations which are acceptable and the values, or
other features which constitute alarm or warning levels requiring action.
U A clear set of procedures to be implemented if alarm or warning levels are reached.
0 Recording and reporting requirements.
0 Provision for review of the monitoring regime or procedures following observed
behaviour of the structure.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

The absence of a clear monitoring specification will lead to the following problems,
which have implications for the safe management of sub-standard structures:

0 Inspectors are less likely to be able to target the focus of their monitoring to
particular areas of structural inadequacy.
0 Inspectors are less likely to be aware of what range of observations are acceptable.
0 Inspectors and engineers are less likely to have or be aware of what procedures should
be followed if the observations are not deemed acceptable.
0 Information is more likely to be lost on organisational changes and particularly
changes managing agents. This is especially relevant where monitoring is by means
other than visual inspection, where any development work carried out by the
managing agent could be deemed to be Intellectual Property.

12.2.6 Methods and took for Visual inspection

monitoring A substantial part of monitoring concrete bridges will comprise visual observations, the
measurement of cracks; size, geometry, direction, number and any associated leakage of
water and other substances.

This regime (Class 1) is likely to be appropriate for monitoring structures that are sound
with no signs of distress and where the likely mode or modes of distress are likely to
progress slowly and that are easily measurable. The following types of distress may be
appropriately monitored in this fashion:

0 Bulging
0 Corrosion
0 Delamination
U Distortion
0 Flexural cracking
U General cracking
0 Inadequate concrete repairs
0 Loss of section
0 Settlement
U Spalling
0 Tilting.

Other testing and forms of monitoring

When Class 2 and 3 frequency of monitoring is required then visual monitoring will
need to be supplemented by appropriate testing and permanent forms of monitoring.

155
 Bridge management and assessment

An overview of these is given in the CBOG Technical Guide 2, Guide to testing and
monitoring of concrete xtructure~’*8, including:

U Testing within the structure maintenance process

0 General aspects of testing
0 Site tests for physical structure and response
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

0 Site tests for corrosion and probability of corrosion

0 Laboratory tests for physical structure and response
0 Laboratory chemical tests
U Automated monitoring of physical structure and response
Cl Automated monitoring corrosion and probability of corrosion

Monitoring reviews
The need and frequency of monitoring is appropriate for a specific set of circumstances
and any monitoring regime should be reviewed in the light o f the observed
deterioration rates; experience and knowledge that emerges from other bridges and
loading events such as exceptional loading from abnormal loads.

12.3 Cracking and crack The appearance of cracks in hardened concrete always gives rise to comment but,
widths irrespective of this, the structural significance of cracks, and hence their relevance to
assessed capacity, needs to be considered carefully.

To enable this consideration, as much as possible of the following information should be

gathered and recorded during an inspection:

0 Location, frequency, orientation and pattern of cracking

0 Length, width and general appearance of individual cracks
0 Estimated age of concrete when cracking first appeared.

12.3.1 Crack widths Cracks rarely have a constant width throughout their length and generally only become
noticed when they exceed 0.1 m m in width. Crack widths can aid in diagnosis of their
cause and can also become of concern in respect of durability of the concrete. Many
cracks are considered ‘live’, i.e. their width varies over time, and this may be due to
structural effects, thermal effects or a combination of those effects.

Crack widths in excess of 0.5 mm are probably structural cracks, and crack widths in
excess of 1.Omm are possibly indicative of the reinforcement having yielded and should
be investigated further.

156
 12.3.2 Types O f crack and Cracks fall into t w o general categories, structural and non-structural. The significance of
diagnosis any cracks observed during a bridge inspection should become apparent by careful
diagnosis of those cracks. Guidance o n a provisional diagnosis of the cause of cracking
and other physical defects is given in CBDG Technical Guide 212*,and Appendix D of
that document contains reference photographs and diagrams.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Structural cracks
Structural cracks are those caused by load effects or other external influences such as
movement or support settlement. They may be due t o flexure, shear or punching shear and
indicate that assessment should be carried out at the serviceability limit state in accordance
with Section 6.7. They will be recognised generally by their location in the structure, e.g.
across the soffit near the mid span of a slab deck or diagonally in the web of a beam near a
bearing, and may be reasonably wide. Consideration should be given t o the repair or
remediation of these cracks t o restore as far as possible the durability of the concrete.

Non-structural cracks
Concrete Society Technical Report 22, Non-structural cracks in c ~ n c r e t e ’ ~ ’classifies
,
three main types of non-structural crack:

1. Plastic
2. Early thermal contraction
3. Long-term drying shrinkage

An additional type of crack, and one which is of particular relevance t o assessment, is

that caused by reinforcement corrosion. Such cracks are characterised by their width,
location and associated rust staining. A typical example is given in Cracking Reference 1
of Appendix D in Technical Guide 2’28.

Plastic cracks, arising from plastic shrinkage and plastic settlement, occur in the fresh
concrete before hardening, whereas cracks due t o early thermal contraction and long-
term drying shrinkage occur after the concrete has hardened.

Figure 1 of Technical Report 22 gives a ‘family tree’ of crack types and Figure 2 shows
examples of those crack types on a hypothetical concrete structure. A classification o f
these crack types is given in Table 1 of the document. The information obtained from
the inspection should be used in conjunction with Table 1 t o classify the cracks.

Cracking, especially crazing, may be indicative of poor curing, but it may also be
indicative of alkali aggregate reactions, especially alkali-silica reaction. ASR cracking is
usually parallel t o the direction of stress. Further information o n ASR is given in Section
9.2 and in Reference 12.10.

The significance of non-structural cracks in respect of assessment is their potential effect

o n the durability of the structural element, which may need t o be taken into account as
part of the assessment of that element. Information o n the assessment of deteriorated
structures is given in Section 9.12 and methods of enhancing the durability of concrete
are given in Concrete Society Technical Report 61 12.11.

157
 72 Bridge management and assessment

12.4 Deterioration rates The main objective of programmes of assessments in the past has been t o evaluate the
load capacities of structures designed t o earlier codes and rules of thumb. Deterioration,
where pronounced, was usually taken into account but its progression and future
influence on the load capacity of a structure was rarely considered.

It became apparent t o engineers that many structures had durability and service
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

problems that eventually led t o reductions in load carrying capacities, and that for
more cost effective management of structures the future performance of structures
had t o be determined by bringing together the results of condition and structural
assessment.

This requires knowledge of:

0 The causes of deterioration

0 The effects of deterioration on the load carrying capacities of structures
0 The timescales and rates of deterioration.

It must be emphasised that there can be considerable uncertainty associated with the
modelling of deterioration. The approaches described here can be helpful in informing
decisions about the management of structures, particularly where they are supported by
’’
correlations between test results and predictions. Technical Report 61 l 2 provides more
detailed guidance on deterioration models.

12.4.1 The causes of Of the causes of deterioration of reinforced concrete, the penetration of chlorides from
deterioration de-icing salt and the advancement of carbonated zone (carbonation) leading t o the
depassivation and subsequent corrosion of reinforcement are the most important
factors. Detailed descriptions of the processes involved in deterioration are given in
Chapter 9 and how these affect the load carrying capacities of structures in Chapters 6
and 7. This Section considers the penetration of chloride only as this is the most likely
form of deterioration encountered by engineers in the UK, concentrating on timescale
and rates of deterioration.

12.4.2 Timescales and rates The determination of timescales involved and deterioration rates requires the corrosion
of deterioration of reinforcement t o be considered in two stages as follows:

1. Initiation period
2. Propagation period.

Initiation period
This is the period that an aggressive front of chloride penetration takes t o advance t o
and then t o initiate corrosion of reinforcement, frequently assumed t o occur when 0.3%
chlorides by weight of cement a t the depth of the reinforcement are present. It is
calculated mostly from models based on the mathematics of diffusion.

158
 Older structures, particularly those built before the 1950s and 60s, were built with far
lower covers than considered necessary today and consequently have had far shorter
initiation periods and correspondingly far greater deterioration problems when
compared with more modern structures. Concretes in these older structures will often
also be more permeable with low cement contents.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Rate of advance calculations

A number of models have been suggested t o predict the advance of an aggressive front
but most have adopted the same general form based on Fick's 2nd Law of
l 2 13. The relevant equation, assuming one-dimensional diffusion, is:

SC S2C
-(x,t) = D-(x,t)
6t Sx2

where C(x,t) = chloride concentration at a distance x from the surface after time t.
D = diffusion coefficient

The commonly used solution assuming constant chloride concentration at the surface of
the concrete is:

C(x,t)= C,[I- erf ( x / ( ~ , / ( ~ t ) ) ]

where erf = the error function

C, = chloride surface concentration, which is assumed constant, though in
practice surface concentrations can vary and will influence the
calculation)

More involved analyses have been developed but in most situations a pragmatic and
balanced approach using the basic solution t o Fick's 2nd Law has been found sufficient
t o support the technical and financial decisions that need t o be taken.

More modern structures may have been subjected t o surface treatments for
preventative maintenance reasons and this will need t o be taken into account in the
value used for the diffusion coefficient. Buenfeld and Zhang1214provide guidance on this.

Propagation period
This is the period from the end of the initiation period t o the time of assessment of the
structure. The initiation and propagation periods are illustrated diagrammatically in
Figure 12.1.

For the propagation period t o commence, the chloride concentration at the surface of
the reinforcement must exceed a threshold value Cth;Vu and S t e ~ a r t " . 'provide
~
guidance on this. Thus, during the period from when the threshold value is exceeded t o
the time of assessment a representative corrosion rate value (an estimate of metal loss
per unit of surface and time) is calculated. It should be noted that along with the onset
of delamination the corrosion rate changes.

159
 a2 Bridge management and assessment

Figure 12.1
Initiation and propagation periods.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Initiation Propagation
period period

Lifetime or time before repair

Corrosion rate
Corrosion rate is usually expressed as an attack penetration depth, either uniform or
localised, expressed as a depth of steel lost over one year or as a measure of electrical
current over a unit area. The former is difficult to measure where, for example, pitting
corrosion caused by chlorides has occurred and the reinforcement remains embedded in
concrete.

There are several methods used to measure electrical current but the most widely used
of these is the linear polarisation resistance method”’6. In this, a small electrical current
or voltage is applied to the reinforcement and the corresponding response respectively
in voltage or current is measured. From this the polarisation resistance and, in turn,
instantaneous corrosion rate is determined.

The instantaneous corrosion rate does vary with changes in the climate and therefore
to obtain a representative value it is recommended that measurements are taken
during:

U Dry periods with low temperatures

0 Periods of low temperatures after raining continuously during a t least one or two
days
0 Dry periods and high temperatures
0 Periods a t high temperatures one week after raining continuously during two to three
days.

The results obtained are then averaged to give the representative value.

160
 Where only single measurements are possible, the manual for assessing corrosion-affected
concrete structure^'^'^ provides guidance on how t o use these with laboratory testing of
cores taken near to the points of measurement to derive the representative values.

Different locations in a structure and different deterioration processes will also require
the calculation of different representative values.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

From the representative values, penetration depths of corrosion are determined and in
turn the residual or remaining diameters of the corroded bars are assessed taking into
account whether the corrosion is uniform or pitting.

12.4.3 StrUCtUra[ assessment The consequences of corrosion of reinforcement results in:

0 Reduction of rebar section and ductility

0 Reduction of bond between reinforcement and concrete
0 Loss of concrete integrity due t o cover cracking and/or spalling

This is illustrated in Figure 12.2

Having knowledge of when corrosion commenced and also the rate at which the steel is
corroding at a particular section in a structure, enables one t o determine the loss of
steel, the loss of concrete and, in turn, the level of performance of that section. This is
illustrated in Figure 12.3.

When determining the time a structure reaches a non-acceptable level of performance,

it is necessary t o appraise the effects on the concrete of the different levels of corrosion
of the reinforcement. The assessed capacities for different penetration depths using the
appropriate geometrical and mechanical characteristics relevant t o the different levels of
corrosion are estimated and plotted as shown in Figure 12.4. From this ‘deterioration’
curve the predicted time when the structure reaches a non-acceptable level of
performance and the residual life of the structure are determined.

Figure 12.2
I I
Consequences of corrosion.
Loss of concrete Delamination Bond deterioration

J /- A I

Mechanical properties Reduction of

(ductility) rebar area

V
I Reduction in Load capacity I

161
 In Bridge management and assessment

Figure 12.3
Determination of level of performance.
Structural Condition data
assessment

Estimate Estimate
Determine corrosion start of
critical rate corrosion
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Level of Current
performance deteriorated
* of section section

Figure 12.4
Deterioration curve.

Assessed
capacities at
Present different levels
ranariiu / of corrosion

Minimum acceptable performance

, Residual period

Time t
Present time
--
Initiation
I-
Propagation
tl
period period

The above is based on a deterministic approach but because of the large amount of
variability in the factors considered a stochastic approach is t o be preferred. Although
the paper by Tantele et aP2 concerns the effectiveness of preventative maintenance, it
provides useful guidance on a stochastic approach relevant t o the assessment of
deterioration rates.

12.4.4 StrUCtUra[ apprakal Atkins propose a general corrosion model for reinforcement based on that
developed for the Midland Links. An indication of the threshold corrosion level for
delamination a t Midlands Link is given.

162
 Bridge management and assessment 12

Hughes eta112,20in their paper on the site testing and monitoring carried out on the pier
crossheads of the M4 Elevated Viaduct in West London noted that the average losses of
reinforcement for the solid, delaminated and spalled surfaces were respectively 3.1 %,
12.2% and 16.9%.

However, there appears to be little information published about the quantitative

Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

interpretation of the effects of corroded reinforcement on concrete structures. What

information exists appears to be restricted to a few organisations that have undertaken
this type of work in the past.

Duty of justices of peace to investigate

"all manner of annoyancesof bridges broken in the highe wayes to the
damage of the Kynes liege people"

Statute of Bridges 1530

The assessment programme has encompassed bridges owned by various organisations

that either come under the category of a Highway Authority or person/body that are
liable by virtue of statue or of historical liability to be responsible for maintaining bridges
on public highways.

12.5.1 Historical background The responsibility for building and maintaining bridges in medieval times was part of the
feudal obligations placed upon landowners. Many bridges were built with money raised
through indulgences by the church; the belief being that the purchase of an indulgence -=

would save one from a period of time in purgatory before entering heaven.

However by the sixteenth century, with the dissolution of the monasteries, the role of
the church in building and maintaining bridges declined and this, together with other
factors, led to many bridges being in a poor state of repair. It was recognised that
bridges were important to the economy as a whole, that the wider community should
take greater responsibility for their maintenance.

The Statute of Bridges, 1530, placed responsibility on the inhabitants of the counties to
maintain bridges of public utility and gave power to justices of peace at the Court of
Quarter Sessions to administer the extent of this liability.

However, others remained Liable because of statute or historical liability as illustrated by

the Bridge House Estates, a charity formed over 900 hundred years ago for the
12.22.
managing and maintaining of bridges across the Thames in the City of

Throughout succeeding centuries further legislation was introduced that defined in ever-
increasing detail the Liability to maintain bridges. Interestingly, the Bridges Act 1803

163
 12 Bridge management and assement

required all bridges built for public utility in a county to be so under the direction or to
the satisfaction of the County Surveyor, who prior to the local government reforms of
the 1880s was an officer of the Court of Quarter Sessions and, subsequently, an
employee of the newly created county council; this is an early form of technical
approval procedure so familiar to bridge engineers today.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

By the Highways Act 195912.23, the county council was made the Highway Authority for
all 'county bridges' and with the Local Government Act 197212.24; the distinction
between the authority liable for the maintenance of the bridge and for the highway over
it to all intents and purposes disappeared.

The Highways Act 198012.25 does, however, recognise that not all bridges in the highway
are the responsibility of the county council.

12.5.2 Current situation The determination of this liability to maintain bridges is a major concern of highway law
therefore with respect to bridges and for assessment this is particularly important in
establishing the load bearing obligations of the owners. Table 12.1 summarises the
liability to maintain bridges.

Thus, for bridges not the liability of Highway Authorities to maintain it is necessary to
consider legislation existing at the time of their construction and that brought in
subsequently to define the responsibility of their owners.

The Transport Act 196812.26, defined the responsibilities of the inland waterway and
railway companies for their highway bridges many of which were built at the time of
the construction of the canals and railways. Significantly, this Act recognised that
the highway load bearing obligations differed from those for Highway Authorities.
Section 117 of the Act states that these shall be in accordance with an order made
by the appropriate minister, this being the Statutory Instrument 1972/1705 which
refers to Technical Memorandum BE4, which limits the load bearing obligation to
24 tons.

Table 12.1
Publicly maintainable bridges
Bridge maintenance liability.
a All bridges built since 1959 and approved and adopted by the Highway Authority
b All bridges built before 1960 and 'of public utility or benefit' except those bridges which are privately
maintainable
c All bridges built at any time and expressly adopted by the Highway Authority by agreement or under
any statute or statutory process

Privately maintainable bridges

a All bridges built since 1959 and not approved or adopted by the Highway Authority
b Any bridge built by a person or organisation who interfered with the highway and caused the need
for a bridge
c Any bridge built by a person or organisation who by prescription remains liable to maintain it
d Any bridge built by a person or organisation who by reason of tenure remains liable to maintain it
e Any bridge built under a statutory power which requires the bridge to be privately maintained

164
 Section 7 of the Trunk Roads Act 1946lZz7 and later Section 5 5 of the Highways Act
1980’225 led to the adoption of all private bridges on roads that became trunk in
accordance with these acts. Subsequently, roads losing their status as trunk roads
became with the affected bridges the responsibility of local Highway Authorities in
accordance with Section 2 of the Highways Act 1980; this has led to the anomalous
situation where some bridges having been built and formerly maintained by the railway
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

companies not being subject to the requirements of the Transport Act 1968.

For highway bridges owned by other persons or organisations, it is even more important
to investigate their histories and consider both the common law liabilities and legislation
existing a t the time of their construction. It is generally recognised in law that the
owners of these bridges need only maintain them to standards adequate a t the time of
their construction and are under no obligation to bring them up to standards
commensurate with the volume and weight of modern traffic. A Highway Authority,
however, does have the power under Sections 93 to 95 of the Highways Act 1980 to
take responsibility for a bridge in agreement with the owner if it wishes to bring the
bridge up to the current load bearing standard.

For bridges in unknown ownership the situation is unclear. However, Highway

Authorities have a duty in accordance with Section 41 of the Highways Act 1980 to
maintain their highways to an adequate standard, and in the event of bridges falling into
a state of disrepair and/or are inadequate to carry the existing traffic they would not be
fulfilling this. Therefore, it would appear that these bridges would have to be considered
as bridges owned by Highway Authorities.

The Road Traffic Regulation Act 1984’228in its Section 122 states that Highway
Authorities have a duty to:

0 Secure and maintain reasonable access to premises

U Preserve or improve the amenities of areas through which roads run

This together with the duty imposed by Section 41 of the Highways Act 1980 imply
that Highway Authorities should ensure that all bridges carrying roads are capable of
carrying all vehicles designed in accordance with construction and use regulations.
Clearly, this presents problems to Highway Authorities because of:

0 The continually increasing weights of heavy goods vehicles brought about by

commercial and political factors.
U The differing load standards that other parties are legally required to provide.
U The ever increasing ages of bridges caused by the failure by central and local
government to fund an adequate bridge replacement programme.
0 The failure of central and local government to adequately fund a bridge strengthening
programme.

The Highway Authorities are left with the choice, therefore, of imposing weight
restrictions on bridges correspondingto the results of the programme of assessments or
accepting the risks of allowing unrestricted use of weak bridges with or without other

165
 12 Bridge management and assessment

measures such as monitoring or lane restrictions to continue. Bridge engineers are under
considerable pressure from politicians to do the latter, but must use their own
judgement and experience to do the former where in their judgement the public is put
a t too great a risk.

The imposition of a weight restriction can imply that the Highway Authority is not
fulfilling its duty to maintain the highway to an adequate standard, and consequently
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

can in theory be served notice under Section 56 of the Highways Act 1980 to restore
the highway to a satisfactory standard by strengthening or replacing the weak bridge.

Culverts
Although not strictly part of the assessment programme a Highway Authority should
assess, following the outcome of the case Bybrook Barn Garden Centre-v-Kent County
Council, whether its culverts are large enough to take existing flows and, if not, whether
it has a duty to enlarge them.

Accommodation and occupation bridges

A Highway Authority has no duty to maintain bridges over or under the highway unless
the bridge or underpass is itself a public highway carrying vehicular traffic, or public
rights of way on foot or horseback, or has been provided by the authority for pedestrian
safety. However, a Highway Authority must ensure that these structures have been
adequately designed for required loadings before granting licences for new bridges under
Section 176 of the Highways Act 1980 and ensure that the design loadings are adhered
to. For older structures where design loadings are unknown a Highway Authority may
have to assess these to ensure that the safety of the highway is not impaired.

The legal obligations for accommodation and occupation bridges, particularly with
regard to Network Rail, are subject to contractual agreements and reference to statute
necessitating site-specific advice.

Generally, accommodation bridges were constructed at the request of landowners

whose lands were otherwise severed by the construction of railways, canak and
highways. As such, they can only be legally used by the successors in title to the original
landowner(s),although in some cases public footpath and bridleway rights may also
have been acquired across the bridge.

Occupation bridges carry or pass over private roads which existed before the construction
of the railways and highways, and which serve adjoining lands. In principle they can only
be legally used by the subsequent owners/occupiers of the lands originally benefiting
from access, which may now be open to interpretation in the light of changing land use.

In both cases, those entitled to the use of the bridge can invite others to exercise similar
rights provided that in so doing, the burden on the owner of the bridge to repair and
maintain the structure is not increased.

As a result of recent case law, Network Rail are obliged to maintain private bridges to
the maximum load bearing capacity of the original bridge, calculated five years after the

166
 date of its construction. A bridge’s load bearing capacity is the maximum weight that it
can bear without causing it to fail or be damaged, whether immediately or through
repetitive use. Section 6 of the Locomotive Act 1861’229may only be used to warn
users that the capacity of the bridge is insufficient. The capacity may be expressed as
axle weight or gross laden weight.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Access to property and powers of entry

Almost invariably, a Highway Authority will require access through adjacent property to
carry out the required inspection for the assessment of a structure. Sections 289 and 291 of
the Highways Act 1980 provide the powers to a Highway Authority to do this, but Sections
290, 292 and 294 also place certain obligations on it when exercising these powers.

When inspection of a structure that is over or adjacent to operational railways, canals or

navigable waterways and to watercourses or flood defences is required then Network
Rail, British Waterways, Environment Agency or other relevant body must be consulted
well in advance of it being undertaken.

12.6 The assessment of

bridges for abnormal Loads

12.6.1 Abnormal load There are two main categories of Heavy Load Vehicle, both of which are allowed to
categories operate under the provisions of Section 44 of the Road Traffic Act 1988. There are those
which move under Motor Vehicles (Authorisation of Special Types) General Order 1979
(STCO), weighing up to 150 tons, and those which exceed this limit and move under an
special individual order (SO) issued by VSE Division in DETR. It is estimated that there are
some 200,000 STCO movements on trunk roads each year, and 200-300 SO movements.

Abnormal Indivisible Loads (AILS)are moved in England after prior notification under the
provisions of Section 44 of the Road Traffic Act 1988. Notifications are made to DETR,
Highway Authorities, bridge owners and police forces.

The notification requirements for heavy load vehicles are shown in Table 12.2

Under the current Regulations, Highway Authorities have little control over STCO
movements. For example, there is no control over the time of day a t which a movement
can take place, so there is considerable potential for disruption to traffic from large
slow-moving loads.

12.6.2 Management O f Currently, the Highway Authorities utilise a variety of adhoc methods to manage
abnormal indivisible load notifications and check bridges for each notified AIL movement. It is common practice
to compare the AIL with a known HB assessed capacity of the bridges on the
movements proposed route. Some Highway Authorities maintain a grid of Abnormal Load routes on
which the bridges have been assessed or they have been known to have safely carried AILS.

167
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

Department of Transport Special Order plus eight

weeks clear notice to the HA five clear days notice
to Police and five clear days notice with indemnity
t o Highway and Bridge Authorities

I Width exceeding 2 9 m up t o 8 0 m Two clear days notice to Police

Department of Transport form VRI plus ten clear

days notice to the HA and t w o clear days notice to

Police and five clear days notice with indemnity to

Highway and Bridge authorities

I
7 4 m rigid Two clear days notice t o Police

]Vehicle combination exceeding 25 9 m Two clear days notice t o Police

A recent audit of AIL procedure carried out for the Highways Agency showed that,
although a generally professional approach was being adopted by the Highway Agency’s
Maintenance Agents, the systems and procedures needed to be improved and harmonised.

Separation of responsibilities for the maintenance of motorways and Trunk Roads from
local authority roads and the introduction of DBFO roads have significantly increased
the number of notifications that have to be made.

It has been accepted generally that in managing the Trunk Road Network, the Highways
Agency needs to make effective provision for the necessary movement of Abnormal
Indivisible Loads (AILS) on the network. This has to be achieved with minimal adverse
impact on the network infrastructureand other users. Much the same applies to other
Highway Authorities and bridge owners. Increasingcongestion on the nation’s road
system will make the movement of AILS more problematic.

This will worsen Highway Authorities’ assessment and clearance difficulties associated with
these movements. These difficulties are exacerbated because AILSare usually wide, long
and slow. Highway Authorities need an efficient and effective system for administering AIL
notifications, assessments and clearances, so that these initiatives are not jeopardised.

The current adhoc systems are potentially unsafe; do not facilitate a comprehensive
database of bridge capacity Loads, which can readily be interrogated, and there are risks
that key data could be lost when the responsibilities for highway bridges are periodically
changed.

168

.. ~ ~-
 The Highways Agency has commissioned the development of a new system called
ESDAL (Electronic Service Delivery of Abnormal Loads) using a software package
designed for MS Windows NT operation that can be accessed on its website
w.esdal.com, which will provide the facilities required to receive, check and
administer AIL notifications.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

The following principal functions will be provided by the system:

W Fax and imaging software to receive and respond to AIL notifications, with facilities
for displaying, printing, logging and archiving.
W On-screen mapping, allowing the notified route to be selected, with facilities for
enlargement, on screen selection of routes and search and find tools.
W Database, maintaining records of AIL loads and movements, standard routes,
operators (includingtheir indemnity/insurance details) AIL vehicles, bridge details and
their capacity.
Mapping, gazetteer and place name data, standard routes.
W Bridge analysis software, enabling each bridge on the route selected to be checked by
a process of comparative assessment (i.e. based on imposed weight limits and pre-
assessed capacities) for the capacity for the vehicle axle configuration notified.
W Height checks for overbridges.
W Reporting facilities that enable more detailed statistical examination of the database.
Restrictions: e.g. global, local, position of AIL, restrictions applicable with other traffic,
speed restrictions.

The development of ESDAL has been divided so far into four phases as shown in
Table 12.3.

The first phase came on line in March 2006 and the final ones are due at the end of
2007/early 2008. A guide for the owners of structures.can be accessed via
http://www.esdal.com.

Table123 -p
'hase I Description -

169
 12.7 References 12.1 THE STATIONERY OFFICE, A Code of Practicefor the Management ofHighway Structures, September 2005.
12.2 COLE. C, Managing sub-standard bridges, Bridge Management 4, (Ryall, MJ,Parke, CAR and Harding, /E, eds), Thomas
Telford, 2000
12.3 PARSONS BRINCKERHOFF, HBR80616 - Highways Agency Contract 21419, TechnicalAudit oftheapplicationofBA79
and A review of bridge assessment failures on the Motorway and Trunk Road network, Final Reports, Parsons Brinckerhoff,
December 2003.
12.4 FLINT, AR and DAS, PC. Whole Life Performance Based Assessment Rules - Background and Principles, SafetyofBridges
Conference. Thomas Telford. London, July 1996.
12.5 SHAVE, J, DENTON. Sand JANDU. A, Management of Sub-standard Structures - A new Standard, Proceedingsofthe
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

First lnternationalConference on Advances in Bridge Engineering, Brunel University, 26-28 June 2006.
12.6 ManualofBridgeEngineering: Thomas Telford Books ISBN 0 7277 2774 5
12.7 HIGHWAYS AGENCY, BA 79/98 -Amendment No 1, The Management of Substandard Highway Structures, Design
Manualfor Roads and Bridges.
12.8 CONCRETE BRIDGE DEVELOPMENT CROUP, Guide to testing andmonitoring the durabilityofconcretestructures,
Technical Guide 2, CBDC, Camberley, 2002.
12.9 CONCRETE SOCIETY, Non-structuralcracks in concrete, Technical Report 22, 3rd Ed., The Concrete Society, Camberley,
1992.
12.10 INSTITUTION OF STRUCTURAL ENGINEERS, Structuraleffects ofalkali-silica reaction - Technicalguidance on the
appraisal ofexisting structures, The Institution of Structural Engineers, London. July 1992.
12.11 CONCRETE SOCIETY, Enhancingreinforcedconcrete durability, Technical Report 61, The Concrete Society. Camberley,
2004.
12.12 COLlEPARDl, M, MARCIALIS, A, and TURRIZIANI, R, Kinetics of penetration of chloride ions in concrete, IlCemento,
1970, pp. 157-164 (in Italian)
12.13 COLLEPARDI, M. Penetrationofch/oride ionsfnto cementpasteandconcrete,American Ceramic Society, 1972
12.14 BUENFELD. NR and ZHANC. JZ, Chloride diffusion through surface-treated mortar specimens, Cement andconcrete
Research, Elsevier Science Ltd, Vol. 28, Issue 5, 1998.
12.15 VU, KAT and STEWART, MG. Structural reliability of concrete bridges including improved chloride-inducedcorrosion
models, Structuralsafety, Elsevier, Vol 22, Issue 4, 2000.
12.16 CONCRETE SOCIETY, Nectrochemical tests for reinforcement corrosion, Technical Report 60, The Concrete Society,
Camberley, 2004.
12.17 BRITISH CEMENT ASSOCIATION et al. A validatedusersmanualforassessing the residualservicelife ofconcretestructures
affectedby corrosion. EC Project Ref. IN 309012, CONTECVET, BCA. Camberley, 2001.
12.18 TANTELE, E, ONOUFRIOU, T and MULHERON. M, Effectiveness of preventative maintenance for reinforced concrete
bridges - a stochastic approach, BridgeMangement5, (Parke. CAR and Disney, P. eds), Thomas Telford, 2005.
12.19 ATKINS, C, HOCC, V, MIDDLETON, C and ROBERTS, MB, A proposed empirical corrosion model for reinforced
concrete, Proceedings ofthe lnstitution of Civil EngfneersStructures andBuildings, 2000, Vol. 140, pp 1-1 1
12.20 HUCHES, C, Smith, JSC and Warren, CP. Site testing and monitoring of pier crossheads suffering from chloride induced
reinforcement corrosion, Bridge Management 4, (Ryall, MJ,Parke, CAR and Harding, JE, eds), Thomas Telford, 2000.
12.21 PIERCE, P, Oldlondon Bridge, Review, 2002.
12.22 SAUVAIN, SJ, Highwaylaw, 2nd Ed.. Sweet & Maxwell, 1997.
12.23 Highways Act 1959, HMSO, http.//www.tsoshop CO u k l bookstore
12.24 Local Government Act 1972, HMSO, http.//www.tsoshop co.uk/ bookstore
12.25 Highways Act 1980, HMSO, http://www.tsoshop.co.uk/ bookstore
12.26 Transport Act 1968, HMSO, http://www tsoshop.co.uk/ bookstore
12.27 Trunk Road Act 1946. HMSO, http.//www.tsoshop.co.uk/ bookstore
12.28 Road Traffic Regulation Act 1984, HMSO, http.//www.tsoshop.co.uk/ bookstore
12.29 Locomotive Act 1861, HMSO, http://www tsoshop.co.uk/ bookstore

170
 Appendix. Relevant historica I references t o
design and materials specifications and
standards used in concrete bridge
construction
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

The following is indicative of the information that may be useful in bridge assessment

Bridge design 0 LECAT, AW, DUNN, C and FAIRHURST, WA, DESIGNAND CONSTRUCTION OFREINFORCEDCONCRETEBRIDCES.
Concrete Publications Ltd, London, 1948 (Revised 1957).
0 MINISTRY OF TRANSPORT, Memorandum 577, Bridgedesignandconstruction,HMSO, London, 1945,1952.
0 MINISTRY OF TRANSPORT, SpecificationforRoadandBridge Works. HMSO, London, 1951,1987,1963,1976,1986
(Amendment No 1, 1988).
0 MINISTRY OF TRANSPORT, Memorandum S77/1, Rulesfor the design anduse ofFreyssinet hinges on highwaystructures,
HMSO, London. 1966.
0 CONCRETE AND CONSTRUCTION ENGINEERING, 1906-67, Various appropriate articles on
http.//w.lStructE org/Library/advanced search.asp, Search term - conc cons engg + search term.
0 BRITISH STANDARDS INSTITUTION, BE5/75, Rulesfor the design anduse OfFreyssinethinges inhighwaystructures.1975.
0 DEPARTMENT OF TRANSPORT, DesignManualfor RoadsandBridges,HMSO, 1992, 1997.
0 DEPARTMENT OF TRANSPORT, BD24/92, Design ofhighway bridges andstructures.HMSO, London, 1992.
0 DEPARTMENT OF TRANSPORT, BD15/92, Ceneralprinciplesfor the design andconstruction ofbridges. HMSO, London.
1992.
DEPARTMENT OF TRANSPORT, BDS7/94, Designfordurability, HMSO, London, 1994.
0 DEPARTMENT OF TRANSPORT, BD58/94. Designforconcrete highway bridgesandstructureswith external and
unbondedprestressjng.HMSO, London, 1994.
0 DEPARTMENT OF TRANSPORT, BD42/96, Design ofintegral bridges, HMSO, London, 1996.

ConCrete The first national design‘code for concrete structures was introduced in 1934.
Subsequently new codes have been introduced, as indicated below (with the date of first
publication). Invariably there was some overlap a t each transition between an old and a
new code. In addition, codes are subject to revision before they are eventually replaced.
0 REYNOLDS, CE and STEEDMAN, J.C,ReinforcedConcrete Designer’s Handbook, E&FN Spon, London, 1932,1939,
1946, 1948 (revised 1951, 1954), 1957, 1961 (revised 1964), 1971 (revised 1972). 1974, 1981, 1988, 436pp.
0 BRITISH STANDARDS INSTITUTION,CP114, Thestructuraluse ofnorrnal reinforcedconcretein buildings. 1948, First
revision 1957, Reset and reprinted 1968 (including Amendment No 1). Amendment No. 2 published 25 May 1967.
0 BRITISH STANDARDS INSTITUTION, CP115, Thestructuraluse ofprestressedconcreteinbuildings. 1959.
0 BRITISH STANDARDS INSTITUTION, BS1926:1962, specificationfor ready-rnixedconcrete.
0 BRITISH STANDARDS INSTITUTION,CP116, Thestructuraluse ofprecast concrete, 1965.
0 MATTHEWS, DD et al. The draft unifedcode ofpractice forstructuralconcrete (CP7 70) Concrete, Vol 4, No. 2, February
1970, pp 66-70. Three papers covering:
The new code its background and purposes, by Matthews, DD
A discussion on the draft code
Slab design and the draft code, by Taylor, R, Hayes, B and Mohamedbhal, GTG.
0 BRITISH STANDARDS INSTITUTION, CP110, Code ofpracticeforthestructural use ofconcrete, 1972.
0 DEPARTMENT OF TRANSPORT, Technical Memorandum (Bridges) BE 1/72 Reinforcedconcretehighwaystructures,BE
2/73 Prestressedconcretehighwaystructures,1973
0 BRITISH STANDARDS INSTITUTION, BS5328:1976. Methodsforspecifying ready-rnixedconcrete.
0 BRITISH STANDARDS INSTITUTION, 555328.1981, Merhodsforspecifyingconcrete, includingready-rnixedconcrete.
0 BRITISH STANDARDS INSTUTITION, BSS400. Steel, concrete andcompositebridges, Part 4. Code ofpracticefordesign
ofconcrete bridges. 1984.
0 BRITISH STANDARDS INSTITUTION, 5881 10-1.1985, Structuraluseofconcrete, Part 1: Codeofpracticefordesignand
construction.

171
 0 BRITISH STANDARDS INSTITUTION, 855328.1 1991 (revised 1997). Concrete- Part 1: Guide tospecifyingconcrete.
0 BRITISH STANDARDS INSTITUION, 855328-2 1991 (revised 1997), Concrete- Part 2,Methodsforspecifyingconcrete
mixes.
0 BRITISH STANDARDS INSTITUTION,855328-3.1990, Concrete- Part 3 Specificationfortheproceduresto be usedin
producingand transporting concrete.
0 BRITISH STANDARDS INSTITUTION, 855328-4:1990, Concrete - Part 4. Specificationfor theprocedures to be usedin
sampling, testing and assessing compliance of concrete.
0 BRITISH STANDARDS INSTITUTION, B5 EN206-1.2000 (Incorporating Corrigenda Nos 1 and 2 and Amendments 1. i
3). Concrete- Part 1. Specification, performance, production and conformity.
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

0 BRITISH STANDARDS INSTITUTION, 858500 2002 (revised 2006), Concrete- ComplementaryBritish Standardto
ES EN206- 1
Part 1 - Method of specifying andguidancefor the specifiers
Part 2 - Specificationfor constituentmaterials and concrete.
0 DEPARTMENT OF TRANSPORT, IAN 52/04. Changes to Aggregate andConcretespecification affecting MCHWSeries
500,600, 700, 1700and Notes for Guidance NC 100, NC 500, NC 600, NC700, NC 800 and NC 1700, Highways
Agency, London, January 2004.
0 DEPARTMENT OF TRANSPORT, IAN 58/04. Changes to concretespecificationaffecting MCHWlnterim Advice Note
52/04 guidanceforstructuralconcrete, Highways Agency, London, May 2004
0 DEPARTMENT OF TRANSPORT, IAN 74/06, Revisedguidanceregarding the use ofES8500for the design and
constructionofstructures using concrete, Highways Agency, London, 2006

Reinforcement 0 MINISTRY OF TRANSPORT, Memorandum 785. Permissibleworking stresses in concreteandreinforcing barsforhighwaj

bridgesandstructures,HMSO, London, 1961
0 STEEL REINFORCEMENT COMMISSION, UK reinforcement standards 1938-1990, Concrete, Vol. 24, No. 3, March
1990, pp 40-41
0 BRITISH STANDARDS INSTITUTION, 65785.1938 and BS785:Part 1 1967, Hotrolledbarsandharddrawn wireforthe
reinforcement of concrete.
0 BRITISH STANDARDS INSTITUTION, BS4449:1969, 1978, 1988, 1997, Speci~cationforcarbonsteelbarsforthe
reinforcemen t of concrete.
0 BRITISH STANDARDS INSTITUTION, 8S11441943, 1967, Coldworkedsteel barsforthereinforcement ofconcrete.
0 BRITISH STANDARDS INSTITUTION, 854461.1969. 1978, Specificationfor cold workedsteelbarsfor the reinforcemen!
of concrete

POSt-tenSiOning Systems The commercial use of prestressing in the UK began just before the Second World War.
During and just after the war it was used to overcome material shortages. The early
designs were mainly by refugee European engineers and the systems manufactured wen
not British. After about 1950 various British systems were developed.

Detailed information on systems available between 1940 and 1985 may be found in
ClRlA Report 106, Post-tensioning systems for concrete in the UK: 1040-1985,
published in 1985. A number of the systems described ceased to be used after about thc
1960s. Thus when assessing an existing structure, identification of the type of anchoragi
used may give an indication of the age of the structure
ANON, Test of prestressedconcrete railway bridge, Concrete andConstructionEngineering, Vol. 46, No. 5, June 1951,
pp 186-188.
MORICE. PB, The theoryofdesign offullyprestressedbeams. Report CACA 130, Cement & Concrete Association (now
British Cement Association),Camberley,July 1953
MORICE, PB,A test on a 55ftspanprestressedconcrete beam, Report TRA/144. Cement & Concrete Association (now
British Cement Association), Camberley. March 1954
MORICE, PB, Concentratedloadonprestressedconcrete bridge decks, Report TRA/175, Cement & Concrete Associatioi
(now British Cement Association), Camberley,January 1955.
MORICE, PB. The ultimate strengthof two-span continuousprestressedconcrete beams as affected by tendon
transformation andun-tensionedsteel,Report TRA/186, Cement & Concrete Association (now British Cement
Association), Camberley, May 1955.
ANON, A new type of prestressed bridge deck, ConcreteandConstructionEngineering, Vol. 53, No. 3, March 1958,
pp 145-146.
ANON, An investigation of prestressed bridges, Concrete andConstructionEngineering, Vol. 53, No. 9. September
1958, pp 351-354
CIRIA, Post-tensioningsystemsforconcretein the UK.1940-7985, Report 106, 1985

172
 Precast beams 0 PRECAST CONCRETE DEVELOPMENT GROUP PC1, Standardbeam sectionsforprestressedconcrete bridges ( I ) ,
InvertedTbeamsforspans25 to 55ft. Cement and Concrete Association, October 1963 Revised as Report CSC1,
Concrete Society, Camberley, October 1967
0 PRECAST CONCRETE DEVELOPMENT GROUP, PC4, Standardbeam sectionsforprestressedconcrete bridges (2). Box
section beamsforspans 40 to 85ft, Cement and Concrete Association. November 1963, Revised as Report CSC2,
Concrete Society, Camberley, October 1967
0 PRECAST CONCRETE DEVELOPMENT GROUP, PC6, Standardbeam sectionsforprestressedconcrete bridges (31 Box
section beamsforspans 85 to l20ft, Cement and Concrete Association, April 1964, Revised as Report CSC3. Concrete
Society, Camberley, October 1967
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

0 PRECAST CONCRETE DEVELOPMENT GROUP, PC8, Standardbeam sectionsforprestressedconcrete bridges (4), I-

section beamsforspansfrom 40 to S i f t , Cement and Concrete Association, July 1964, Revised as Report CSC4,
Concrete Society, Camberley, October 1967
U PRECAST CONCRETE DEVELOPMENT GROUP, PC9, Standardbeamsectionsforprestressedconcrete bridges (S),I-
section beamsforspansfrom 85 to l20ft, Cement and Concrete Association, August 1964, Revised as Report CSCS,
Concrete Society, Camberley, October 1967

Highway loading 0 BRITISH STANDARDS INSTITUTION, 85153 Part 3A 1954, Speci~cationforsteelqirderbridges-Loads, 1954
0 MINISTRY OF TRANSPORT, Memorandum 771, Standardhighwayloadings, HMSO, London, 1961
0 BRITISH STANDARDS INSTITUTION, 855400 Part 2, Specificationforloads, 1978
U DEPARTMENT OF TRANSPORT, BD14/82 (Amendment No l ) , Loadsforhighwaybridges,1982
0 DEPARTMENT OF TRANSPORT, BD37, Loadsforhighwaybndges 1988, Clause 6 8 1 replaced by BD48 93
0 DEPARTMENT OF TRANSPORT, BD60/94, Design ofhighway bridgesfor vehicle co//isionloads, 1994
U DAWE, P, Traffic loading on highway bridges,Thomas Telford, London, 2004, 168pp

173
 CONCRETE BRIDGE DEVELOPMENTGROUP
The Concrete Bridge Development Croup aims to promote e x a b c e in the design, conmuaion and

NCRETE DEVELOPM

The Concrete Bridge DevelopmentGroup

Riverside House
4 Meadows Business Park

in concrete bridgetkign,constr~ction
and mnagMent
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.

encoume devttswnent of inmative ideasand concepts T& +44 (0)127633m,F a : -144 (0)1276 38899
m uk

m
m
m
to those who have an interest in
in types of membership are available:
Licensed copy from CIS: [email protected], Systra Ltd, 13/12/2017, Uncontrolled Copy.