Вы находитесь на странице: 1из 86

AUSTROADS INTERNAL REPORT

Guidelines for Road Network Condition


Monitoring: Part 3 Pavement Strength

IR-88/05

GUIDELINES FOR ROAD NETWORK CONDITION


MONITORING: PART 3 PAVEMENT STRENGTH
(SEALED GRANULAR PAVEMENTS)

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength


(Sealed Granular Pavements)

First Published 2005

Austroads Inc. 2005


This work is copyright. Apart from any use as permitted under the Copyright Act 1968,
no part may be reproduced by any process without the prior written permission of Austroads.

Austroads Internal Report

Austroads Project No. AS1122


Austroads Publication No. IR88/05

Project Manager
Ron Ferguson, RTA NSW

Prepared by
Tim Martin, ARRB Group Ltd
L.B. Dowling & Associates

Published by Austroads Incorporated


Level 9, Robell House
287 Elizabeth Street
Sydney NSW 2000 Australia
Phone: +61 2 9264 7088
Fax: +61 2 9264 1657
Email: austroads@austroads.com.au
www.austroads.com.au

Austroads believes this publication to be correct at the time of printing and does not accept
responsibility for any consequences arising from the use of information herein. Readers should
rely on their own skill and judgement to apply information to particular issues.
This report has been produced as an Austroads Internal Report and whilst not confidential or
specifically restricted, it is not intended for public release or general circulation.

GUIDELINES FOR ROAD NETWORK CONDITION


MONITORING: PART 3 PAVEMENT STRENGTH
(SEALED GRANULAR PAVEMENTS)

Sydney 2005

Austroads profile
Austroads is the association of Australian and New Zealand road transport and traffic authorities
whose purpose is to contribute to the achievement of improved Australian and New Zealand road
transport outcomes by:

undertaking nationally strategic research on behalf of Australasian road agencies and


communicating outcomes

promoting improved practice by Australasian road agencies

facilitating collaboration between road agencies to avoid duplication

promoting harmonisation, consistency and uniformity in road and related operations

providing expert advice to the Australian Transport Council (ATC) and the Standing
Committee on Transport (SCOT).

Austroads membership
Austroads membership comprises the six state and two territory road transport and traffic
authorities and the Commonwealth Department of Transport and Regional Services in Australia,
the Australian Local Government Association and Transit New Zealand. It is governed by a
council consisting of the chief executive officer (or an alternative senior executive officer) of each
of its eleven member organisations:

Roads and Traffic Authority New South Wales


Roads Corporation Victoria
Department of Main Roads Queensland
Main Roads Western Australia
Department of Transport and Urban Planning South Australia
Department of Infrastructure, Energy and Resources Tasmania
Department of Infrastructure, Planning and Environment Northern Territory
Department of Urban Services Australian Capital Territory
Commonwealth Department of Transport and Regional Services
Australian Local Government Association
Transit New Zealand

The success of Austroads is derived from the collaboration of member organisations and others in
the road industry. It aims to be the Australasian leader in providing high quality information, advice
and fostering research in the road sector.

Acknowledgement
Austroads wishes to acknowledge that this document is based on ARRB TR Contract Report No
RC2410/1 dated March 2004 prepared by Tim Martin, and on work by Paul Robinson from 1999 to
2001 for Austroads under Project BS.AC.007 and by Tim Martin in 2002 for Austroads under
Project BS.AC.025, as summarised in the Austroads Technical Report Pavement Strength in
Network Analysis of Sealed Granular Roads: Basis for Austroads Guidelines (Austroads 2003b).

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

EXECUTIVE SUMMARY

Austroads recognises network level data on road pavement strength as one of a number of
important inputs to a range of asset management decision tools, and in a number of
corporate performance indicators used by road agencies.

The purpose of these guidelines is to promote consistency and improved quality in


estimating, reporting and using pavement strength in network level asset management
throughout Australia and New Zealand.

In these guidelines, the term pavement strength refers to the ability of a pavement structure
to resist wheel loads that are applied to it, and is generally synonymous with structural
capacity.

The guidelines outline a 7-step process for estimating pavement strength parameters,
starting from a decision that network level strength information is needed.

Network level pavement strength parameters are estimated primarily from measurements of
surface deflection using standard loading and other standard test procedure details. The
guidelines describe the estimation from surface deflection data of Modified Structural
Number (SNC) and Adjusted Structural Number (SNP) as the most commonly used network
level pavement strength parameters.

The frequency of network level deflection surveys is covered in the guidelines in Section C.

The guidelines recognise the importance of the longitudinal sample spacing between
deflection tests in a network survey. The proportion of a network to survey is also covered in
Section C.

While these guidelines are intended to be as independent as possible of the technology used
for measuring surface deflection, they have been prepared in the context that three
methodologies are used for measuring surface deflection in road network surveys in
Australia and New Zealand, viz Benkelman Beam, Deflectograph, and Falling Weight
Deflectometer.

The guidelines discuss the relative merits of these devices, and contain suggested default
relationships between deflection data collected by the three different devices, with a warning
that the relationships should used with caution and ideally verified by experimental
observations.

The guidelines include preferred procedures for verification of deflection data, to ensure
quality data from each network survey.

Details in these guidelines for distance verification are the same as in other guidelines in this
Austroads series on road condition monitoring at a network level.

Recognising that current and remaining structural capacity can best be assessed in relation
to a defined terminal structural condition, the guidelines suggest indicative investigation
levels for deflection, rutting, roughness and cracking as an interim surrogate for terminal
structural condition, as these are the common distresses most likely to be associated with
structural deterioration.

For information rather than for practical use, Appendix 3 in these guidelines describes an
Interim Model for predicting structural deterioration of sealed granular road pavements. The
model has been postulated using preliminary information. More comprehensive and longterm pavement performance monitoring and data collection is necessary to enable the
Interim Model to be tested, confirmed or modified, and calibrated for practical use.

Austroads 2005
-- i --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

A glossary of terms used with information on network level assessment of road pavement
strength is in Section B.

The technology of road condition monitoring worldwide is continuing to develop. Austroads


encourages innovation, and promotes the coordinated introduction of improved practices.
These guidelines are therefore expected to be subject to ongoing review.

Austroads 2005
-- ii --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

TABLE OF CONTENTS
Page

SECTION A

ROAD PAVEMENT STRENGTH................................................................................1

A.1 Introduction

.............................................................................................................1

A2. The Need and Application of Network Pavement Strength ...............................................3


A3 Objective

.............................................................................................................4

SECTION B

GLOSSARY OF TERMS USED IN THE NETWORK-LEVEL ASSESSMENT


OF SEALED GRANULAR ROAD PAVEMENT STRENGTH ................................5

SECTION C

GUIDELINES FOR NETWORK ASSESSMENT OF ROAD PAVEMENT


STRENGTH ...........................................................................................................8

C.1 What is Pavement Strength? .............................................................................................8


C1.1
Guidelines .....................................................................................................8
C1.2
Background Notes ........................................................................................9
C2 Equipment for Measuring Pavement Deflection ..............................................................12
C2.1
Guidelines ...................................................................................................12
C2.2
Deflection measuring equipment general ................................................14
C2.3
Benkelman Beam (BB) ...............................................................................14
C2.4
Deflectograph (DEF) ...................................................................................16
C2.5
Falling Weight Deflectometer (FWD and HWD)..........................................18
C2.6
Relative merits of Benkelman Beam, Deflectograph and FWD ..................21
C3 Frequency of Pavement Deflection Surveys....................................................................23
C3.1
Guidelines ...................................................................................................23
C3.2
Background notes .......................................................................................23
C4 Scope of Pavement Deflection Surveys ..........................................................................25
C4.1
Guidelines ...................................................................................................25
C4.2
Background notes .......................................................................................25
C5 Relationships between Measures of Pavement Deflection..............................................32
C5.1
Guidelines ...................................................................................................32
C5.2
Background notes .......................................................................................33
C6 Verification of Distance Measurement .............................................................................36
C6.1
Guidelines ...................................................................................................36
C6.2
Background notes .......................................................................................36
C6.3
Recording deflection test locations .............................................................37
C7 Verification Testing for Deflection ....................................................................................39
C7.1
Guidelines ...................................................................................................39
C7.2
Background notes .......................................................................................40
C8 Repeatability and Bias .....................................................................................................42
C8.1
Guidelines ...................................................................................................42
C8.2. Background notes ................................................................................................42

Austroads 2005
-- iii --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C9 Data Reporting
...........................................................................................................43
C9.1
Guidelines ...................................................................................................43
C9.2
Background notes .......................................................................................44
SECTION D

SUMMARY

...........................................................................................................46

D1 Sampling and Measurement of Pavement Deflections ....................................................46


D2 Estimation of Network Level Pavement Strength Parameters .........................................48
D3 Interim Structural Deterioration Model for Sealed Granular Pavements..........................48
REFERENCES AND BIBLIOGRAPHY ...........................................................................................49
APPENDIX 1:
1.1
1.2
1.3
1.4

ESTIMATING PAVEMENT STRENGTH PARAMETERS


FROM DEFLECTION DATA................................................................................54
Modified Structural Number, SNC ...................................................................................54
Adjusted Structural Number, SNP ...................................................................................55
Structural Adequacy Indicator, SAI..................................................................................57
Relative Pavement Strength (RPS) Indicator ..................................................................58

APPENDIX 2: COST 336 PROCEDURES FOR REPEATABILITY TESTING WITH FWDs ......59
2.1 COST 336 Protocol U2-1999: FWD Short Term Repeatability Verification ....................59
2.2 COST 336 Protocol U3-1999: FWD Long Term Repeatability Verification.....................62
APPENDIX 3:

INTERIM STRUCTURAL DETERIORATION MODEL


FOR SEALED GRANULAR PAVEMENTS .........................................................66
3.1 Background to Interim Model Development ...................................................................66
3.2 Basis of Interim Structural Deterioration Model ...............................................................67
3.3 Model Calibration ...........................................................................................................71

Austroads 2005
-- iv --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

TABLES
Page

Table 1

Process for estimating pavement strength parameters


for network level asset management planning ...................................................... 2

Table 2

Indicative investigation condition


(interim surrogate for terminal structural condition) ............................................. 11

Table 3

Summary of the main features of Benkelman Beams,


Deflectographs, and Falling Weight Deflectometers............................................ 13

Table 4

Target FWD test loads and corresponding surface stresses............................... 20

Table 5

Suggestions for performance indicators


to undertake discrete network level sampling...................................................... 30

Table 6

Summary of deflection relationships.................................................................... 35

Table 7

COST 336 procedures for FWD calibration and verification................................ 41

Table 8

Extract from report on network level FWD survey ............................................... 44

Table 1.1

Sample values of mean characteristic maximum deflection (D0) and


corresponding SNC for unbound sealed granular pavements ............................. 54

Table 3.1

Impact of granular resheeting on pavements ...................................................... 67

FIGURES
Figure 1

Pavement deflection bowl (not to scale) .................................................................. 9

Figure 2

General view of Benkelman Beam (BB) with load truck and trolley,
and sketch of BB arrangement .............................................................................. 15

Figure 3

Benkelman Beams (BB) with automated and manual deflection recording ........... 15

Figure 4

Host truck with loaded rear axle and Deflectograph sled in front of rear axle ....... 17

Figure 5

DEF with RWP beam shortly after starting a deflection bowl measurement.......... 17

Figure 6

Deflectograph with LWP beam positioned


about the middle of a deflection bowl measurement ............................................. 18

Figure 7

Deflectograph with RWP beam positioned


near the finish of a deflection bowl measurement ................................................. 18

Figure 8

Schematic diagram of a FWD................................................................................ 19

Figure 9

FWD loading plate ................................................................................................. 19

Figure 10

General view of a FWD.......................................................................................... 19

Figure 11

Rear view of a FWD showing loading plate and geophones ................................. 19

Figure 12

FWD deflection sensors (geophones) ................................................................... 20

Figure 13

FWD Sampling locations on a single carriageway two lane road .......................... 27

Figure 14

FWD Sampling locations on a dual carriageway road ........................................... 28

Figure 15

Extract from a graphical report from a Deflectograph survey ................................ 45

Figure 1.1

SNC (Paterson) vs SNP (others) ........................................................................... 56

Figure 3.1

% SNC Deterioration vs Pavement Age / Design Life


(varying Pavement Design Life and constant Deterioration Factor ....................... 70

Figure 3.2

% SNC Deterioration vs Pavement Age / Design Life


(fixed Pavement Design Life and varying Deterioration Factor) ............................ 71

Austroads 2005
-- v --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

ABBREVIATIONS AND ACRONYMS


AADT

Annual average daily traffic (volume), measured in vehicles per day (vpd)

AASHO

American Association of State Highway Officials (forerunner of AASHTO)

AASHTO

American Association of State Highway and Transportation Officials

ACT

Australian Capital Territory

ADT

Average Daily Traffic

ALF

Accelerated Loading Facility

ARRB

ARRB Group Ltd, a research organisation based in Melbourne, Australia.

BB

Benkelman Beam

CAP

Traffic load capacity

CBR

California Bearing Ratio

COST

Cooperation in Scientific and Technical Research (Europe)

COV

Coefficient of Variation, the standard deviation of a population dived by the mean


(expressed as a percentage).

D0, , D200, etc

Deflection measurements forming a deflection bowl. D0 is at the load point (the maximum
deflection measurement), D200 is 200mm away in the direction of travel, etc.

DEF

Deflectograph

DGPS

Differentially Corrected Global Positioning System

DIER Tas

Department of Infrastructure, Energy and Resources, Tasmania (a member of Austroads)

DIPE NT

Department of Infrastructure, Planning and Environment Northern Territory


(a member of Austroads)

DL

Design life

DUS ACT

Department of Urban Services, Australian Capital Territory (a member of Austroads)

Eqn

Equation

ESAs

Equivalent Standard Axles a measure of traffic loading (mass)

FHWA

Federal Highway Administration (part of the USA Federal Department of Transportation)

FWD

Falling Weight Deflectometer

GPS

Global Positioning System

HDM

Highway Development and Management (formerly Highway Design and Maintenance


Standards) models, software and documentation initially developed by the World Bank
and released in 1979, based on the Highway Cost Model produced by the Massachusetts
Institute of Technology in 1971/72. Managed by PIARC from the late 1980s.

HDM III

A version of HDM models, software and documentation introduced in 1987. Only HDM-4
is supported since 2000.

Austroads 2005
-- vi --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

HDM-4

A new version of HDM models, software and documentation developed in the


International Study of Highway Development and Management (ISOHDM), managed by
the World Road Association (PIARC). PIARC released HDM-4 (Version 1) in 2000. As
at early 2005, Version 2 is undergoing development and testing.

HDM
Technology

A generic term referring to the collection of published products released by the PIARC
ISOHDM project, comprising HDM-4 knowledge, model algorithms, and the HDM-4
software.

HSD

High Speed Deflectograph

HWD

Heavy Weight Deflectometer

IRI

International Roughness Index, a measure of roughness developed in the 1980s by the


World Bank and adopted by the World Road Association (PIARC) and Austroads.

ISOHDM

International Study of Highway Development and Management

kN

Kilonewton

kPa

Kilopascal

LTPP

Long Term Pavement Performance (program)

LWP

Left wheel path the wheel path nearer to the verge (because traffic in Australia and
New Zealand drives on the left side of the road), sometimes referred to in the literature as
the outer wheel path.

MA

(Austroads) Member Authority

MR

Resilient modulus

MRWA

Main Roads Western Australia (a member of Austroads)

NAASRA

National Association of Australian State Road Authorities (forerunner of Austroads)

NRM

NAASRA Roughness Meter, or NAASRA Roughness Measure (NRM, counts per


kilometre, an alternative to IRI as a measure for roughness).

OH&S

Occupational Health & Safety

PaSE

Pavement Strength Evaluator (a Deflectograph owned and operated by VicRoads)

PMS

Pavement Management System

QDMR

Queensland Department of Main Roads (a member of Austroads)

RTA NSW

Roads and Traffic Authority, New South Wales (a member of Austroads)

RWP

Right wheel path the wheel path nearer to the middle of the road (because traffic in
Australia and New Zealand drives on the left side of the road), sometimes referred to in
the literature as the inner wheel path.

SAI

Structural Adequacy Index

SHRP

Strategic Highway Research Program (USA)

SN

Structural Number

SNC

Modified Structural Number

SNP

Adjusted Structural Number

TNZ

Transit New Zealand (a member of Austroads)

Austroads 2005
-- vii --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

TRL

TRL Limited, Berkshire, UK is part of the Transport Research Foundation group of


companies (formerly Transport Research Laboratory (UK), and Transport and Road
Research Laboratory (UK))

TSA

Transport South Australia, part of the Department of Transport and Urban Planning South
Australia (a member of Austroads)

UK

United Kingdom

UNDP

United Nations Development Programme

USA

United States of America

VicRoads

Road Corporation of Victoria (a member of Austroads)

vpd

Vehicles per day

WDM

WDM Ltd, Bristol, UK, a commercial provider of services in pavement and road asset
management.

Austroads 2005
-- viii --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

SECTION A ROAD PAVEMENT STRENGTH


A.1

INTRODUCTION

Pavement strength is considered to be one of the most important characteristics defining the
general condition of a road.
These guidelines provide the necessary pavement strength data and information, at a road
network planning level for sealed granular pavements, for road owners and practitioners.
The guidelines cover the following aspects of pavement strength:
Section B

A glossary of terms used in network-level assessment of strength of sealed


granular road pavements.

Section C

Sampling, measurement and analysis of network pavement deflection


data in Australia and New Zealand with the aim of providing consistency and
acceptable quality to the reported deflection and strength parameters.

Descriptions of the Benkelman Beam, Deflectograph, and Falling Weight


Deflectometer, being the commonly used devices for measuring pavement
deflection and a discussion of their relative merits.

Relationships between deflection data collected by different devices, to


enable data from all three devices to be used together.

Appendix 1

The estimation of pavement strength parameters for use in asset


management.

Appendix 3

As information only rather than for practical use, an interim structural


deterioration model is documented, using the pavement strength parameter
SNC to predict the deterioration in network pavement strength of sealed
granular pavements. The need for testing, confirmation or amendment, and
calibration of this Interim Model is acknowledged, and methods are outlined
for future improvement of the Interim Model, based on long term pavement
performance monitoring.

These guidelines provide a consistent approach to pavement strength assessment and analysis.
With a standard basis for recording and reporting pavement strength at a network level, road
agencies, road maintenance contractors and road pavement condition monitoring service providers
will have a consistent basis to collect and analyse data and specify measures for improved road
asset management. This will lead to better identification of road deterioration characteristics and
technical measures required to ensure stronger and longer lasting pavements are built and
maintained.
These guidelines distinguish deflections (that can be measured and reported using one of three
main measurement devices) from pavement strength and estimates of values of pavement
strength parameters. Table 1 contains a 7-step outline of the process of estimating pavement
strength parameters for network level asset management planning purposes, based on data
collected in pavement deflection surveys.

Austroads 2005
-- 1 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

Table 1: Process for estimating pavement strength parameters for network level asset management planning

Steps in the process of estimating pavement strength parameters

Reference in these guidelines

Decision to assess network level strength parameters, based on a deflection survey.

A strategic asset management decision,


not covered by these guidelines.

Design sampling details by selecting a deflection measuring device, considering timing


(season) of survey, selecting the lanes and wheel paths to survey, conducting trial deflection
surveys, and establishing the optimal longitudinal sample spacing, or if necessary an optimal
sample proportion for a Deflectograph survey.

Sections C2 and C4.

Conduct network survey and obtain reports of deflection data with supporting details
(measuring device, operator, date, time, location referencing, weather, temperatures, etc).

Sections C2 to C9.

Identify and separate lengths with bound base - where pavement configurations are not known,
subject to local experience and confirmation, deflection relationships can be used as filters,
such as limits on maximum deflection (D0) and deflection ratios such as (D250/D0).

Section C1.2.1.1.

For the remaining lengths, analyse the deflection data to identify approximately homogeneous
sections (short lengths, depending on the variability of the deflection data), and calculate and
report the characteristic maximum deflection for each homogeneous section.

Section C4.2.4.

Calculate mean of the characteristic maximum deflection values for each sub-network (longer
lengths, eg, management segments (PMS segments), road links, or road types).

Section C4.2.4.

Compute strength parameter (eg, normally SNC, and could be SNP or SAI) for each subnetwork or segment, as required for analysis purposes.

Section C1.2.1.2 and Appendix 1.

Austroads 2005
-- 2 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

A2.

THE NEED AND APPLICATION OF NETWORK PAVEMENT


STRENGTH

Many pavement performance characteristics (roughness, cracking, rutting, etc) do not give an
accurate assessment of the structural condition of a pavement because they mainly assess
surface condition and not structural condition (Eijbersen and Van Zwieten 1998). These surface
condition parameters are relatively inexpensive to collect and have traditionally been used to
broadly identify suspect areas of the network for detailed structural testing and assessment at a
project level.
The need for an improved understanding of the structural condition of the whole network for
strategic planning is driven by three main trends:

increasing axle mass limits for heavy vehicles (NRTC 1996);

relatively high rates of growth of heavy vehicle traffic on strategic freight routes (Gargett and
Perry 1998); and

some road agencies deciding to include the structural condition of pavements as a network
performance indicator (Sapkota et al 2001).
The first two of these trends potentially reduce the remaining life of pavements in the network,
causing the need for earlier than expected rehabilitation treatments. These treatments are
relatively costly and have a major impact on annual road agency budgets, so it is necessary to
determine at a network level the likelihood and extent of any major rehabilitation, well in advance
of the need.
The third point above is sometimes chosen as a contractual requirement for the contractor
managing a road network on behalf of the road agency. However, this does not necessarily imply
that the network level structural condition of pavements is suitable as the sole input for managing
the network on a structural basis. For example, complementary surface distress information is
also useful for the assessment of structural condition and to represent other performance criteria
(eg, ride comfort).

Austroads 2005
-- 3 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

A3.

OBJECTIVE

The overall objective of these Austroads Guidelines for Road Condition Monitoring is to promote
and ensure a standard set of procedures are followed, so that:

Only useful road condition data is collected for analysis;

Collection and processing of road condition data is cost efficient;

Quality of road condition data is improved; and

The road condition data collected is of increased value to road owners.

These guidelines form part of the series of Pavement Condition Monitoring Guidelines and provide
the basis of specifications and recommendations for collecting, analysing and reporting information
on pavement strength.
The document is structured around the following key sections:

Section B -

Section C
Guidelines for Network Assessment of Road Pavement Strength
(Sealed Granular Pavements)

Glossary of Terms used in Network-Level Assessment of Road Pavement


Strength (Sealed Granular Pavements)

Section C1 What is Pavement Strength

Section C2 Equipment for Measuring Pavement Deflection

Section C3 Frequency of Pavement Deflection Surveys

Section C4 Scope of Pavement Deflection Surveys

Section C5 Relationships between Measures of Pavement Deflection

Section C6 Verification of Distance Measurement

Section C7 Verification Testing for Deflection

Section C8 Repeatability and Bias

Section C9 Data Reporting

Section D Summary

Appendix 1 Estimating Pavement Strength Parameters from Deflection Data

Appendix 2 COST 336 Procedures for Repeatability Testing with FWDs

Appendix 3 Interim Structural Deterioration Model for Sealed Granular Pavements.

Each of the topics in Section C is structured to show separately:

The specific Austroads guidelines; and

Background notes on the basis for the guidelines.

Austroads 2005
-- 4 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

SECTION B GLOSSARY OF TERMS


USED IN THE NETWORK-LEVEL ASSESSMENT OF
SEALED GRANULAR ROAD PAVEMENT STRENGTH
Term

Interpretation

Adjusted Structural
Number (SNP)

A pavement strength parameter, being an enhancement of the Modified Structural Number (SNC), developed by
Parkman and Rolt (1997) to address difficulties experienced with the use of SNC for the description of pavements
which incorporate lower layers of selected subgrade, or had very thick sub-base or lower sub-base layers. The
SNP applies a weighting factor, which reduces with increasing depth, to the subbase and subgrade contributions so
that the pavement strength for deep pavements is not over predicted. For pavements less than 700 mm thick the
Modified Structural Number (SNC) and the Adjusted Structural Number (SNP) are virtually the same.
(Also see Modified Structural Number (SNC) and Structural Number (SN).)

Benkelman Beam (BB)

An instrument for measuring the deflection of the surface of a pavement caused by the passage of a dual-tyred
single-axle carrying a standard axle load (AS 1348:2002).

Bias

A statistical term to indicate whether a device is systematically measuring high or low when compared to a
reference set of measures.

Condition monitoring

Continuous or periodic inspection, assessment, measurement, reporting and interpretation of resulting data to
indicate the condition of a specific asset in order to determine the need for and nature and timing of maintenance.
(Also see condition survey.)

Condition parameter

A quantifiable expression of a specific parameter of an asset. For example, roughness, rutting, surface texture,
cracking, deflection, etc, are pavement condition parameters.

Condition survey

The process of collecting data on the condition of an asset, eg the structural or functional condition of a pavement.
(Also see condition monitoring.)

Curvature

The difference between the maximum deflection (D0) at a test site and the deflection (D200) at a point 200 mm along
the road from the point at which the maximum deflection was produced. Curvature gives an indication of the
pavement stiffness and therefore the fatigue performance of the pavement.

Deflection

See Pavement deflection.

Deflection bowl

The depressed shape produced at the surface of a pavement when a load is applied (AS 1348:2002).

Falling Weight
Deflectometer (FWD)

A device to measure the surface deflection of a pavement under a dynamic load in order to evaluate its structural
adequacy (AS 1348:2002). FWDs are generally capable of imparting a load up to 150 kN.
(Also see Heavy Weight Deflectometer HWD).)

Granular pavement

A pavement which obtains its load spreading properties mainly by intergranular pressure, mechanical interlock and
cohesion between the particles of the pavement material, which is gravel or crushed rock graded so as to be
mechanically stable, workable and able to be compacted, and generally with a particle size no smaller than sand
(adapted from AS 1348:2002).

HDM

Highway Development and Management (formerly Highway Design and Maintenance Standards) models, software
and documentation initially developed by the World Bank and released in 1979, based on the Highway Cost Model
produced by the Massachusetts Institute of Technology in 1971/72. Managed by PIARC from the late 1980s.

HDM-4

A new version of HDM models, software and documentation developed in the International Study of Highway
Development and Management (ISOHDM), managed by the World Road Association (PIARC). PIARC released
HDM-4 in 2000.

Heavy Weight
Deflectometer (HWD)

A device similar to the Falling Weight Deflectometer (FWD), but capable of imparting a greater load, up to 250 kN.
(Also see Falling Weight Deflectometer (FWD).)

Austroads 2005
-- 5 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

Term

Interpretation

ISOHDM

The International Study of Highway Development and Management, an international project for the development of
HDM-4, based at the University of Birmingham, UK, administered by PIARC in Paris, and funded by the World
Bank, Asian Development Bank, British Department for International Development, Swedish Government and
others.

Indicative investigation
condition levels

A suggested interim set of condition levels (Table 2) for use as an interim surrogate to define terminal structural
condition or the onset of pavement failure, for the purpose of determining the remaining structural capacity or
structural life of a pavement. Planning for maintenance intervention at these condition levels is intended as a
means of managing the risk of accelerating deterioration.

Lane

That portion of a carriageway occupied by a single file of traffic travelling in one direction, hence containing two
wheel paths. A lane is generally between 3.0 and 3.5 m wide. A single carriageway road normally has at least one
lane in each direction.

Link (or road link)

A length of road defined for strategic and reporting purposes, generally of the order of 100 km to 300 km, but can be
longer in remote areas (eg, Katherine to Alice Springs (1,100km) and Port Headland to Broome (600 km)).

Management segment

A length of road pavement that is relatively uniform in treatment history, current condition, terrain, and traffic usage,
with length generally between 0.5 km and 1.75 km (or up to 5 km in remote areas).
(Also see segment.)

Modified Structural
Number (SNC)

A pavement strength parameter, being a refinement of the AASHO Road Test estimation of pavement strength
(Structural Number), which directly takes into account the subgrade contribution to pavement strength (Hodges et al
1975). The Modified Structural Number (SNC) is equal to the Structural Number (SN) that would be required if the
pavement were to be designed to carry the same traffic on a subgrade with a CBR value of 3%.
(Also see Structural Number (SN) and Adjusted Structural Number (SNP).)

Network level

A type of road condition survey or data analysis where the main purpose is to monitor network performance or
assist with network asset management decisions, as distinct from project decisions.

Pavement

The portion of the road placed above the subgrade for the support of and to form a running surface for vehicular
traffic. A pavement usually comprises subbase, base and wearing surface layers.

Pavement deflection

The vertical elastic (recoverable) deformation of a pavement surface between the tyres of a standard axle.
(This definition is used in pavement design, and relates to Benkelman Beam and Deflectograph.)
The elastic (recoverable) vertical movement at the surface of a pavement due to the application of a load
(AS 1348:2002).

Pavement stiffness

The resistance to deflection of the pavement structure.

Pavement strength

The ability of a pavement structure to resist the traffic vehicle wheel loads that are applied to it. Pavement strength
is often seen as synonymous with structural capacity.

Project level

A type of road condition survey or data analysis where the main purpose is to assist with decisions about proposals
for a specific project on a short length of road, as distinct from network decisions.

Repeatability

A statistical term to indicate the extent of variation in outputs about the mean for a single operator using the same
method. Repeatability is the standard deviation of measures obtained in repeat tests using the same measuring
device and operator on a single, randomly selected road.

Reproducibility

A statistical term to indicate the extent of variation in outputs about the mean for multiple operators or measuring
devices using the same method.

Road type

Road types are approximately homogenous sections of road with similar condition, carrying a similar traffic load
under similar climatic and subsoil conditions. Consequently, a road network can be made up of a number of road
types, the number being dependent on the accuracy required of the analysis and the available computing power to
undertake the analysis.

Austroads 2005
-- 6 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

Term

Interpretation

Segment

The length of pavement considered as a separate entity in a pavement management analysis process.
(Also see Management Segment.)

Structural capacity

A descriptive term indicating the capacity of a pavement to carry traffic before the onset of structural failure or
before the pavement deteriorates to a defined terminal condition.
(Also see Pavement Strength.)

Structural Adequacy
Indicator

A pavement strength parameter developed by Eijberson and Van Zwieten (1998), and described in Appendix 1.3.

Structural Number (SN)

A pavement strength parameter, developed during the AASHO Road Test (Highway Research Board 1962). SN
simply describes the structural capacity of a pavement in a single number, regardless of the details of the materials
in the pavement. SN is related to the change in cumulative traffic loading and functional condition of the pavement
(AASHTO 1993). AASHTO (1993) estimates of SN for a given traffic load and functional condition account for the
contributing support of the subgrade through the use of the resilient modulus, MR, for soil support.
(Also see Modified Structural Number (SNC) and Adjusted Structural Number (SNP).)

Surfacing

The uppermost part of the pavement or bridge deck specifically designed to resist abrasion from traffic and to
minimise the entry of water. Sometimes referred to as the wearing surface.

Verification test.

A standardised procedure to test the validity of test results from a measuring device.

Wearing surface

Same as Surfacing.

Wheel path

That portion of the pavement that is subject to passage of and loading from vehicle wheels during trafficking. There
are two wheel paths per trafficked lane referred to in this Guidelines document as the left wheel path (LWP),
nearer to the verge, and the right wheel path (RWP), nearer to the middle of the road (because traffic in Australia
and New Zealand travels on the left side of the road).

Austroads 2005
-- 7 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

SECTION C GUIDELINES FOR NETWORK ASSESSMENT


OF ROAD PAVEMENT STRENGTH
C.1

WHAT IS PAVEMENT STRENGTH?

C1.1 Guidelines
Pavement strength is a measure of the ability of a pavement structure to resist the wheel
loads that are applied to it. Pavement strength can be estimated from surface deflection
data, though the deflection induced in a pavement by a wheel load is really a measure of the
stiffness of a pavement or the ability of the pavement structure to resist that deflection
rather than its strength.
This difference between strength and stiffness is particularly important for flexible pavements
because their mechanism of failure varies and the magnitude of the measured deflection (and
hence stiffness) of bound (asphalt and cemented) pavements is usually significantly lower
than unbound granular pavements. As a result, the relationship between the stiffness and
structural performance (and hence strength) is very different.
To estimate the strength of unbound granular pavements using deflection data, pavements
with cemented bases should be excluded from the survey or the analysis. Knowing which
segments should be excluded from a survey can be difficult, however, because the pavement
structure of many road segments is unknown. In such cases, using maximum deflection (D0)
and deflection ratios such as (D250/D0) (see Figure 1), specific deflection relationships for
identifying cemented bases need to be developed and confirmed from deflection
measurements at locations where pavement structures with cemented bases are known.
At the network level, the strength of assumed homogeneous sections of pavement can be
estimated from surface deflection data using a number of indices based on Structural
Number or Structural Adequacy (see Appendix 1).
The structural capacity of a pavement, and its remaining capacity, should be assessed in
terms of a defined terminal structural condition. While terminal structural condition would be
ideally defined in terms of limits on distress levels (deflection, roughness and rutting), for the
interim pending a better understanding of terminal condition, these guidelines refer to the
indicative investigation levels in Table 2 as the terminal structural condition. The levels in
Table 2 are related to the intended level of service or functionality of the pavement.
The remaining structural life (years) can then be estimated based on the difference between
the existing level of distress and the terminal structural condition if reliable predictions are
available for the rates of deterioration of the terminal structural condition parameters.
Relationships between remaining structural capacity and traffic levels for Australasian
conditions have been developed (Martin 1998, Loizos et al 2002). They can be used to
predict the structural condition of pavements. These predictions are the basis for reaching
informed decisions regarding the remaining life of the pavement and the necessity and timing
of structural intervention (eg, rehabilitation).

Austroads 2005
-- 8 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C1.2 BACKGROUND NOTES


C1.2.1 Derivation of pavement strength parameters from deflection data
C1.2.1.1

The effect of bound pavements

Pavement strength can be estimated from surface deflection data, although Paterson (1987) notes
that deflection measures stiffness rather than strength. Pavement strength is defined as the ability
of a pavement structure to resist the traffic vehicle wheel loads that are applied to it, while
pavement stiffness is defined as the resistance to deflection of the pavement structure
(Koniditsiotis and Kosky 1996). Pavement strength is often seen as synonymous with structural
capacity.
The difference between strength and stiffness is particularly important when assessing flexible
pavements that vary in their mechanism of failure. The magnitude of the measured deflections
(and hence stiffness) of pavements with cemented and unbound bases would usually be
significantly different. Pavements with cemented and unbound bases also have different
relationships between stiffness and the structural performance that relates to pavement strength.
To ensure the validity of assessed values of granular pavement strength parameters, which are
based on deflection testing data, sections of pavement with cemented bases should be excluded
from the survey or the analysis. In practice, selective network testing can be difficult when the
pavement configuration details of many road segments are unknown.
However, if a deflection data set is likely to include some tests conducted on pavements with
cemented bases, it may be possible to identify and remove these by considering both the
magnitude of the maximum deflection, D0, and the ratio of the D250 deflection to the maximum
deflection, D250/D0 (see Figure 1). Relatively low maximum deflections are associated with
cemented pavements and it would be unusual for these deflections to exceed 0.35 mm (using a
Benkelman Beam with a 40 kN test load at a nominal surface stress of 550 kPa) even when the
strengths of these configurations are rated as poor. The D250/D0 ratio for Benkelman Beam
deflections is > 0.8 for cemented base or asphaltic pavements (Scala 1979). These trends should
also apply to Falling Weight Deflectometer and Deflectograph deflections wherever a proportionate
relationship with Benkelman Beam deflections is adopted (refer Section C5).
1500mm

1200mm
900mm
600mm
250mm

D0

D300

D200

D250

D600

Figure 1: Pavement deflection bowl (not to scale)

Austroads 2005
-- 9 --

D900

D1200

D1500

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

In summary, the strength determination and performance characteristics of bound and unbound
pavements differ significantly. For those road networks that include bound and unbound
pavements, the deflection data relating to sealed unbound granular pavements should exclude the
lengths of pavement where it has been confirmed by sample testing of cemented bases that the
measured maximum deflection, D0, and the D250/D0 ratio are less than or greater than specified
values.
The network strength parameter Modified Structural Number (SNC) of pavements with and without
cemented bases can be estimated from deflection data using the specific relationship in
Appendix 1.1, which uses different coefficient values for cemented and uncemented (unbound)
bases.
C1.2.1.2

Network level pavement strength parameters

At the network level, the strength of assumed homogeneous sections of pavement can be
estimated from surface deflection data using a number of indices based on Structural Number or
Structural Adequacy (see Appendix 1).
Initially the Structural Number, SN, was developed during the AASHO Road Test (Highway
Research Board 1962) to define the structural capacity of the Road Test pavements. Pavements
with different materials and layer thicknesses and built on the same subgrade and with the same
remaining traffic capacity (ESAs) would have the same SN. SN has the advantage that it is related
to the change in cumulative traffic loading and functional condition of the pavement (AASHTO
1993). AASHTO (1993) estimates of the SN for a given traffic load and functional condition
account for the contributing support of the subgrade through the use of the resilient modulus, MR,
for soil support.
The AASHO Road Test estimation of pavement strength was further refined by the introduction of
the Modified Structural Number, SNC, which directly took into account the subgrade contribution to
pavement strength (Hodges et al 1975). The estimation of SNC was enhanced by the
development of the Adjusted Structural Number, SNP (Parkman and Rolt 1997), although for
pavements less than 700 mm thick the Modified Structural Number, SNC is virtually the same as
the Adjusted Structural Number, SNP (Roberts 2000b).
Appendix 1 outlines several approaches that can be used to estimate SNC and SNP from bowl
deflection data only. Comparisons of the various means of estimating SNC and SNP using either
the maximum bowl deflection, D0, or a range of bowl deflections (D0, D900 and D1500) suggest that
network level assessment of SNP or SNC could be based on the D0 deflection without any
significant loss in accuracy. This outcome also agrees with the findings of Martin and Crank
(2001) that the bowl deflections other than D0 do not improve strength parameter estimation with
the current strength and deflection relationships.
Other parameters, such as the Structural Adequacy Indicator, SAI, also provide a numerical value
for comparing pavements mainly based on their deflection data regardless of their initial structure
or degree of deterioration (Eijberson and Van Zwieten 1998). Simple relationships, such as the
Relative Pavement Strength indicator, RPS, are useful guides to preliminary intervention and
testing (Roberts 2000a).
Relationships between remaining structural capacity and traffic levels, such as the SNP with traffic
load capacity, CAP, have been developed to predict the structural condition of pavements (Martin
1998, Loizos et al 2002). These predictions are the basis of the analysis for informed decisions
regarding the remaining life of the pavement and the necessity and timing of structural intervention
through rehabilitation.

Austroads 2005
-- 10 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C1.2.2 Assessment of structural capacity from the pavement strength parameter


Current structural capacity, and therefore the remaining structural capacity, should be assessed in
relation to a definition of when terminal structural condition is reached. Terminal structural
condition is defined by its associated limiting distresses (deflection, roughness and rutting) that
depend on the levels of service, or functionality, required of the pavement. These levels of service
and their limiting distress values are based on avoiding rapid or catastrophic failure and its
consequences. This means that lower distress limits are maintained for heavily trafficked
pavements relative to lightly trafficked pavements. However, the limiting distresses associated
with the terminal structural condition are not well defined so it is recommended, in the interim, that
these distresses be defined as the indicative investigation condition levels.
Table 2 outlines some possible distress limits for defining the indicative investigation condition of
the pavement in a management segment.
Table 2: Indicative investigation condition (interim surrogate for terminal structural condition)
Typical Operating Conditions
Typical Road Function

Nominal traffic ranges


ESAs

Surface
Deflection
(D0 (mm))1

Roughness
Limit
(IRI (m/km))2

% Road Length with


Rut Depth > 20 mm3
(1.2 m straight edge)

> 30,000

> 2 x 107

0.8

4.2

10%

100

5,000 30,000

3 x 106 2 x 107

0.9

4.2

10%

80

1,000 5,000

4 x 105 3 x 106

1.1

5.4

20%

1.6

See Note 4

30%

Speed
(km/h)

AADT (vpd)

100

Highways and main roads


Highways and main roads

Freeways, etc

Other sealed local roads


Notes: 1
2.
3.
4.

Indicative Investigation Condition

Various

< 1,000

<4x

105

Based on Figure 6.5, Austroads (2004b).


Based on Table 3.1, Austroads (2004b).
Based on Table 3.2, Austroads (2004b).
In accordance with the relevant local asset management strategy.

From the defined terminal structural condition, with the distress limits as shown in Table 2, the
remaining structural life can be estimated based on the difference between the existing distresses
and the limiting distresses (or the indicative investigation condition levels). The deterioration rate,
the time rate of distress from the current distress level to the limiting distresses, is needed to
estimate the remaining life (years). The deterioration estimation needs to account for the factors
influencing deterioration such as traffic loading, climate, construction quality, maintenance, etc.

Austroads 2005
-- 11 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C2.

EQUIPMENT FOR MEASURING PAVEMENT DEFLECTION

C2.1 Guidelines
There are three main types of testing devices for measuring surface deflection at a network level:

Benkelman Beam (BB);

Deflectograph (DEF); and

Falling Weight Deflectometer (FWD) or Heavy Weight Deflectometer (HWD).

These devices all produce half deflection bowls.


The devices and the associated test
methodologies are described in Sections C2.2 to C2.6 and summarised in Table 3.
Benkelman Beam (BB)
Benkelman Beams measure the rebound deflection of a pivot beam relative to a base beam at a
point on the pavement while the test wheel load (from a loaded truck, with standard wheel
configuration, axle load and tyre pressure) moves away at creep speed along the pavement
surface. A Benkelman Beam is a lightweight, low cost tool that is widely used for project level
deflection testing, and is suitable for network level deflection surveys of small networks. Deflection
recording is usually manual, but automated recording is available.
Deflectograph (DEF)
A Deflectograph consists of two beams mounted on a sled under the chassis of a host truck. The
DEF beams measure the downward deflection in each wheel path while the test wheel load
approaches. A DEF measures and records deflections in 50 mm increments up to 900 mm from the
centre of the test load (referred to as D900), in a series of tests at longitudinal spacings of 3 to 7 m,
at a constant speed of 3 to 4 km/h.
Two DEF versions are commonly used in Australia. Both are based on the La Croix Deflectograph
that was developed and proven in Europe. One version built by RTA NSW operates in New South
Wales, Queensland and Tasmania. The other version, built and operated by VicRoads, is called
the Pavement Strength Evaluator (PaSE). TSA expects to take delivery in mid-2005 of a
Deflectograph constructed by WDM Ltd (UK).
DEFs initially applied a nominal surface stress of 550 kPa (tyre pressure). They now operate
mainly at a nominal surface stress of 750 kPa. The PaSE applies a 10 tonne (98 kN) axle load,
whereas the other Australian DEFs apply the standard axle load of 8.2 tonne (80 kN). The beam
lengths and wheelbase of the PaSE are longer than the other Australian DEFs.
Falling Weight Deflectometer and Heavy Weight Deflectometer (FWD and HWD)
A Falling Weight Deflectometer (FWD) measures surface deflection at offsets ranging from 0 mm to
a user-defined maximum offset (normally 1,500 mm, but up to 2,400 mm) from the centre of an
impulse test load. A series of geophones is used to measure the deflection. The number of
geophones varies, but generally seven are used in normal applications. The magnitude of the load,
which is applied through a circular loading plate 300 mm is diameter, is varied by selecting from a
range of drop heights. The FWD uses a load cell to measure the actual applied load.
A Heavy Weight Deflectometer (HWD) can also be used in applications such as pavements on
heavy duty roads, at airports, container terminals, and other industrial areas. HWDs are available
with capacities up to 250 kN, whereas the capacity of a FWD is typically up to 150 kN.

Austroads 2005
-- 12 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)
Table 3: Summary of the main features of Benkelman Beams, Deflectographs, and Falling Weight Deflectometers
Features

Benkelman Beam (BB)

Deflectograph (DEF)

Falling Weight Deflectometer (FWD)

Applied test load

A moving wheel load, applied by the dual wheels of a slow moving truck, with
standard wheel configuration, tyre pressure and axle load (variable and usually
40kN).

A moving wheel load, applied by the dual wheels of a slow moving


An impact load (up to 150kN for a FWD and up to 250kN for a HWD) in
truck, with standard wheel configuration, tyre pressure and axle load the form of a falling weight with a variable drop height, applied through a
(variable, usually 40kN, but 49kN for the VicRoads PaSE).
standard circular loading plate normally 300mm in diameter.

Outline of test
method

A BB comprises a pivot beam and a fixed beam, enabling rebound deflection to be


measured at each test point (which can be at any desired spacing) at a series of
user-defined offsets (or continuously with an automated recording system) along
a wheel path behind the applied standard test load. Measures a half deflection
bowl while the applied load moves away at creep speed. A BB can be used to
measure full bowl deflections if required, however this is not relevant to network
surveys.

A DEF is similar to a short BB mounted on a sled under the chassis


of a host truck. As the truck moves forward, the DEF is placed for
testing, raised and re-placed in a continuous cycle. DEF records
downwards deflections in a half deflection bowl in both wheel paths
at intervals of between 3m and 7m and at a series of offsets 50mm
apart up to at least 900mm in front of and a short distance beyond
the applied standard test load.

A FWD measures deflections (in a half deflection bowl at a series of userdefined radial offsets normally up to 1,500 mm from the applied load)
and measures the actual applied load at each test point, which can be at
any desired spacing, and at any point on a pavement, whether in a wheel
path or not. Deflections are measured by deflection sensors (usually
geophones) located at the user-defined radial offset positions. Seven
geophones are usual, and more are possible.

Transporting the
testing device

A BB is a lightweight instrument that can be carried manually, or on a small trolley. The DEF is permanently mounted under the chassis of a dedicated
host truck, in front of the rear wheels.

FWDs are usually mounted on a small trailer towed behind a light vehicle
eg, large car or commercial van. FWDs are relatively small and more
readily transportable than DEFs.

Progression along
the road during
testing

Tests at discrete test points. A brief stop is necessary to put the BB in position
before each test. Production depends on traffic conditions, the complexity of
traffic control, longitudinal spacing between test points, method of measuring and
recording deflections (manual or automated), number of wheel paths being tested
(1 or 2), and the mode of transport for operators (walk or ride). Typical rates of
progression are of the order of 0.6km/h (test points at 10m spacing) to 1.3km/h
(100m spacing). Typical production for a shift of approximately 7 hours ranges
from 4 to 7 lane-km.

Continuous movement at 3 to 4 km/h. Typical production for a shift


of approximately 7 hours is of the order of 20 to 30 lane-km (viz up
to about 15,000 deflection bowls), depending on traffic conditions
and the complexity of traffic control.

Tests at discrete test points. The device is stationary during a test.


Typical network survey production for a shift of approximately 7 hours is
150 to 200 test points, depending on traffic conditions, the complexity of
traffic control, and the longitudinal spacing between test points.
Deflections and actual applied load are usually recorded only for the third
in a series of three drops at each test point.

Availability of
deflection testing
equipment in
Australia and NZ

A BB is a simple, low-cost and lightweight instrument. Many road agency depots,


local government bodies and testing organisations own at least one BB. Suitable
trucks are readily available in Australia and NZ.

As at December 2004, there are five DEFs based in Australia (two


owned by RTA NSW, and one each by VicRoads (known as PaSE),
QDMR and DIER Tas), and none in NZ. TSA expects to
commission a new DEF during 2005.

As at December 2004, there are approximately 14 FWDs based in


Australia and NZ (3 owned by ARRB TR, 2 by QDMR, 1 by MRWA, 1 by
TSA, 3 by PMS Pty Ltd, and a small number owned by various LGAs and
private Consulting firms). FWDs based in Australia and UK have been
used in NZ.

Suitability for
network level
deflection surveys

Because of the low number of tests per shift, a BB is suitable only for small
network surveys. Particularly suitable for project level deflection measurement.

Suitable for network level surveys, though the rate of coverage is


limited. Also suitable for project level deflection measurement.

Suitable for network level surveys, particularly for large networks where
relatively homogeneous pavements enable long spacings between test
points. Also suitable for project level deflection measurement.

Personnel
required

A truck driver, a BB operator and a data recorder (plus personnel specifically


allocated to traffic control duties).

A truck driver and a DEF operator (plus personnel specifically


allocated to traffic control duties).

A driver / FWD operator, or a driver and a FWD operator (plus personnel


specifically allocated to traffic control).

Advantages

Local availability, low establishment costs, and simple technology. No limit on


maximum offset for deflection readings, enabling absolute deflection values to be
reported.

Close spacing of test points and deflection measurements at each


test point, large number of test points per day.

Quickest device for covering a large network, provided sample spacings


are long. Applied load can be readily varied. Accurate measurements.

Practical
limitations

Personnel are on the road pavement. Relatively slow overall rate of progress with
a network deflection survey, and is therefore suitable only for deflection surveys of
small networks. Loading from the other rear wheels may affect deflection
measurements.

Relatively slow overall rate of progress (distance) with a network


deflection survey. Loading from other wheels (front and rear) and
vibrations associated with the moving truck may affect deflection
measurements. Maximum practical offset is 900 mm.

Can be difficult to test the left wheel path where seal or shoulder is narrow
and shoulder or edge is not in sound condition. Can be difficult to test the
right wheel path, because FWD trailer may encroach on the adjacent
traffic lane.

Austroads 2005
-- 13 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C2.2 Deflection measuring equipment general


To reduce costs and speed up the analysis process, surface deflection data from test loads is
currently the most commonly used method for assessing pavement strength. Estimation of
network strength parameters using deflections, as discussed in Section C1.2.1.2 and Appendix 1,
does not require any prior knowledge of the pavement layer structure. However, knowledge of the
pavement structure would permit back-analysis of deflection data to estimate moduli, etc for use at
a project level.
Three main device types are used for collecting deflection data at network level:

Benkelman Beam (BB) tests at discrete points, suitable only for small networks due to its
relatively slow rate of testing;

Deflectograph (DEF) conducts almost continuous testing at a constant speed of 3 to


4 km/h; and

Falling Weight Deflectometer (FWD) or Heavy Weight Deflectometer (HWD) discrete test
points.

Table 3 contains a summary of the main features of Benkelman Beams, Deflectographs, and
Falling Weight Deflectometers.
All of these deflection measuring devices can be fitted with a Differentially Corrected Global
Positioning System (DGPS or GPS) receiver that locates the survey test point to within specified
horizontal and vertical tolerances (eg, 2.5 m horizontal and 5 m vertical (TNZ 2002)) of its actual
location.
In addition to these three deflection measuring device types, there are other options, such as
geophysical techniques combined with cored samples for later laboratory testing, but these
methods are slower and often have sampling and interpretation problems greater than those of
deflection testing.
In many instances the availability of deflection measuring devices may govern the choice of the
device used. The frequency of network sampling (years) may also be influenced by the choice of
deflection measurement device - this is discussed in Section C3.

C2.3 Benkelman Beam (BB)


A Benkelman Beam (BB) measures relative deflection of a pivot beam to a base beam at a point
on the pavement while the test wheel load (normally the rear axle of a rigid truck) rolls slowly
(nominally at 4 to 5 km/h) along the pavement surface. The testing process is relatively slow
compared with a Deflectograph. During BB testing, the operator walks behind the truck, which
stops at each test point while the BB is set in position for the test.

Austroads 2005
-- 14 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

Figure 2: General view of Benkelman Beam (BB) with load truck and trolley, and sketch of BB arrangement

A Benkelman Beam can be used to measure full bowl deflections if required. This ability is unique
to Benkelman Beams, but collecting full bowl deflection data is not necessary in network surveys.
For a BB, the standard test load consists of an 80 kN dual wheel single axle applying a surface
stress of 550 kPa (the tyre pressure) under the test loading. Some road agencies, such as
DIPE NT, apply a 750 kPa surface stress with the Benkelman Beam to approximate current tyre
pressures used in heavy vehicles. Because of practical difficulties in maintaining tyre pressures on
an operational vehicle, variations of up to 5% during a shift are likely. The Benkelman Beam
measures the relative deflection up to any desired distance from the moving test wheel load,
although this distance is often limited, for example to 2,700 mm (TNZ 1977, RTA NSW 1982) or
1,200 mm (VicRoads 1986), depending on the time available for recording.
Improvements such as electronic measurement and recording of deflections improve accuracy of
the data and the efficiency of the survey operation. Two Benkelman Beams can be used
simultaneously, one in each wheel path. The Benkelman Beam provides relatively reliable
deflection data at a point.

Benkelman Beam
with automated
deflection recording

Benkelman Beam
with manual
deflection recording

Figure 3: Benkelman Beams (BB) with automated and manual deflection recording

Benkelman Beams are used widely for project level deflection testing on arterial roads.
Benkelman Beams may be useful for small scale network level assessments of pavement strength,
but due to the slow rate of testing, a Benkelman Beam network survey is unlikely to be
economically viable.

Austroads 2005
-- 15 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

Sampling
Deflection testing should be initially conducted every 50 to 100 m along the heaviest loaded
pavement lane in the left wheel path as a minimum requirement (Roberts and Martin 1996). The
longitudinal sample spacing can be increased to an optimal value once the investigation
processes outlined in Sections C4.2.1 to C4.2.4 are carried out. As noted in Sections C4.2.3 and
C4.2.4, this process includes the issue of how many trial sampling lengths are needed along each
road link and what is an acceptable level of variation from the representative structural condition in
order to establish the optimal sample spacing.
Measurement
For sealed granular pavements relative deflections should be measured and reported, as a
minimum, at distances of 0, 200, 300, 600 and 900 mm from the centre of the moving test load.
These deflections are referred to as D0, D200, D300, D600, and D900, respectively (see Figure 1). As
noted above, deflections can be measured more intensively and also at much greater distances
from the test load. Deflections are desirable as far as possible from the centre of the applied load
and preferably beyond the 900 mm offset, because deflections at large offsets increase the
likelihood of recording the full extent of the bowl. For stiffer pavements, such as asphalt and
stabilised materials, a measurement at 1500 mm (D1500) is usually required.
The D0 deflection represents the point of maximum deflection directly under the centre-line of the
applied load (normally the rear wheel of a rigid truck). The D200 deflection is used to indicate bowl
shape in calculating curvature. The D250 deflection, if not measured, is estimated as the mean of
D200 and D300 (see Figure 1), and can be used in filtering out of the sampled deflections any
sections of pavements with cemented bases (see Section C1.2.1.1).

C2.4 Deflectograph (DEF)


A Deflectogragh (DEF) is similar to a pair of short Benkelman Beams mounted on a sled under the
chassis of a host truck. As the truck moves forward at a fairly constant speed of about 3 to 4 km/h,
the DEF is placed for testing, raised and re-placed in a continuous cycle. The test load is applied
through the moving dual wheels on the rear axle of the host truck. DEF records downwards
deflections in half deflection bowls in both wheel paths with almost continuous sampling at
longitudinal spacings of between 3 m and 7 m and at a series of offsets 50 mm apart up to at least
900 mm in front of the applied standard test load.
For network level deflection surveys, a DEF achieves a greater rate of progress along the road and
records a larger number of half deflection bowls (ie, shorter longitudinal sample spacings) than a
Benkelman Beam.

Austroads 2005
-- 16 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

Figure 4: Host truck with loaded rear axle and


Deflectograph sled in front of rear axle

Figure 5:
DEF with RWP beam shortly after
starting a deflection bowl measurement

Two versions of this truck mounted device are in common use in Australia. Both are based on the
Lacroix Deflectograph normally referred to as the Deflectograph. RTA NSW designed and built
one version, and three State road agencies own and operate four of these Deflectographs, viz
RTA NSW (2), QDMR (1) and DIER Tas (1)). QDMR refers to its Deflectograph as PAVDEF.
VicRoads designed, built, owns and operates the other Australian Deflectograph, called the
Pavement Strength Evaluator (PaSE). TSA is expected to take delivery of a Deflectograph
constructed by WDM Ltd (UK) in June 2005.
The main differences in the two Australian versions are that the PaSE has a longer wheel base
and beam lengths (PaSE beams are 2.4 m, cf others 1.2 m). Also, the PaSE typically applies a
10 tonne (98 kN) axle load whereas the other Australian Deflectographs typically apply the
standard axle load of 8.2 tonne (80 kN). However, axle loadings can be varied on each device.
When they were first introduced, the Deflectographs applied the axle load to the moving wheels in
each wheel path at a nominal surface stress of 550 kPa (tyre pressure), as for the Benkelman
Beam. Deflectographs are now mainly operated at a nominal surface stress of 760 kPa, which is
the usual tyre pressure of most heavy vehicles (Austroads 2004c). However, these guidelines use
750 kPa as the preferred Deflectograph tyre pressure, for consistency with the Austroads
Pavement Design Guide (Austroads 2004a, Section 7.2), to eliminate any inference of a need to
maintain tyre pressures to such a high degree of accuracy, and in view of the very small difference
in practical terms between 760 kPa and 750 kPa in the context of a Deflectograph survey.

Austroads 2005
-- 17 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

Figure 6: Deflectograph with LWP beam positioned about the


middle of a deflection bowl measurement

Figure 7: Deflectograph with RWP beam positioned near the


finish of a deflection bowl measurement

Sampling
A Deflectograph measures and records half deflection bowls at longitudinal spacings of between
3 m and 7 m. For example, RTA NSW and QDMR use 4.0 m and 4.5 m respectively as standard
longitudinal spacings in Deflectograph surveys.
The characteristic maximum deflection measurements (85th percentile) in both wheel paths are
reported every 100 m (Ferne 1997) because of the nature of the sampling from this device.
With a Deflectograph, a network survey can also be conducted using sampling based on surveying
a portion (eg, 10% by length) of a defined road link. A useful guide to optimal longitudinal
sampling (described in Section C4.2.2) under these conditions could be 100 m per 1 km, or
500 m per 5 km. However, such a sampling approach would need to be validated before the
survey by more intensive testing over long lengths of the road link. In addition, the time lost in
stopping, securing the DEF equipment for travel, travelling to the start of the next sample length,
and re-establishing the DEF equipment may reduce if not eliminate any benefit, depending on
circumstances such as traffic conditions.
Deflectographs were routinely used in England for comprehensive assessment of network level
pavement strength by measuring and recording half deflection bowls at 3.5 m spacings over 100%
of the network length, at intervals of 3 to 5 years. However, in recent years there has been a move
in England away from comprehensive network deflection testing towards targeted monitoring at
selected locations (see Section C4.2.5.2).

C2.5 Falling Weight Deflectometer (FWD and HWD)


Falling Weight Deflectometers (FWD) and Heavy Weight Deflectometers (HWD) are trailer
mounted devices (Figure 7) that record half deflection bowls at discrete test points on the
pavement surface by measuring surface deflection at distances ranging from 0 mm to a userdefined maximum (normally 1,500 mm, but up to 2,400 mm) from the centre of an impulse test
load, which is applied to the pavement surface through a standard loading plate normally 300 mm
in diameter (Figure 6) by a falling weight with a variable drop height while the FWD or HWD device
is at rest. The FWD produces an essentially half-sine single impact load 25-30 ms in duration,
which corresponds to a moving wheel load.

Austroads 2005
-- 18 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

FWDs and HWDs are essentially the same and are distinguished mainly by their load capacity. A
typical FWD has a load capacity of the order of 120 or 150 kN, and a HWD has a load capacity up
to 250 kN.

Falling Mass
Rubber Buffer
Load Cell

Deflection Sensor

Deflection Bowl
(not to scale)

Figure 8: Schematic diagram of a FWD

Figure 9: FWD loading plate

Source: COST 336

Source: COST 336

Figure 10: General view of a FWD

Figure 11: Rear view of a FWD showing


loading plate and geophones

With FWD and HWD testing, various test loads can be applied to the pavement, the most common
target loads for granular pavements being 40 kN and 50 kN. However, TNZ uses 35 kN for flexible
pavements with chip seal surfacings. As a result of local conditions (eg, longitudinal grade and
crossfall at the test site), the actual load may not be exactly the target load. The FWD uses a load
cell to measure the actual load. Deflections from FWD testing are normalised to the relevant
target load by multiplying the measured deflections by the ratio of the target load to the actual load.
With a standard 300 mm diameter loading plate, each target load corresponds to a specific surface
stress, as shown in Table 4. For practical purposes, these guidelines adopt rounded values for
surface stresses, as also shown in Table 4. The reasons for rounding surface stresses include
removal of any implication of practical benefits from more precise figures, and the very small effect
of the differences between the precise and rounded values on the resulting reported deflections.

Austroads 2005
-- 19 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)
Table 4: Target FWD test loads and corresponding surface stresses
Target Test Load

Corresponding surface stress

Rounded surface stress

35 kN

495 kPa

500 kPa

40 kN

566 kPa

550 kPa

50 kN

707 kPa

700 kPa

It is also necessary to ensure that target test loads are selected appropriately to assess the
strength and stiffness of the pavement. Stresses of 550 kPa and 700 kPa are typical FWD targets,
and actual test load stresses need to be within 15% of the relevant target to achieve a reasonable
accuracy within the deflection analysis. It should be noted that typical FWD test load stresses are
different to the test load stress of 750 kPa in DEF surveys, and in some BB surveys.

Figure 12: FWD deflection sensors (geophones)

Sampling
The sampling approach outlined above for the Benkelman Beam is also a useful guide for FWD
sampling.
Measurement
For sealed granular pavements, bowl deflections are measured and typically reported at distances
of 0, 200, 300, 450, 600, 900 and 1500 mm from the centre of the impulse test load. These
deflections are referred to as D0, D200, D300, D450, D600, D900 and D1500, respectively. The deflection
D0 represents the point of maximum deflection under the impulse load, and is used in estimating
network pavement strength (see Appendix 1.1). The D200 and D300 values are used to estimate
D250 which is used in the ratio D250/D0 to eliminate cemented base pavements from the estimate of
network pavement strength because of the different performance characteristics of these
pavements (see Section C1.2.1.1). The D600, D900 and D1500 values are used in estimating the
Adjusted Structural Number (SNP) as described in Appendix 1. The additional cost of recording
and storing D450 can mostly be warranted on the basis of the potential to use this data in more
detailed analysis.
Deflections are desirable as far as possible from the centre of the applied load and preferably
beyond the 900 mm offset mentioned above, because deflections at large offsets increase the
likelihood of recording the full extent of the bowl. FWDs are capable of measuring deflections at
offsets up to 2,400 mm.

Austroads 2005
-- 20 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

Three repeat drops are usually conducted at each test point, with the data from the third drop
normally used for reporting and analysis purposes. The measured deflections are then
normalised to the appropriate surface stress, to correspond with operating tyre pressures.

C2.6 Relative merits of Benkelman Beam, Deflectograph and Falling Weight


Deflectometer
All forms of deflection testing require strict compliance with Occupational, Health and Safety
(OH&S) requirements, particularly with respect to traffic control. On a road that is open to traffic,
there is more risk associated with the use of a Benkelman Beam (BB) and a Falling Weight
Deflectometer (FWD) than a Deflectograph (DEF) because of the need for personnel to be on the
pavement in close proximity to road traffic. A Deflectograph is also inherently safer than a BB and
a FWD because of its uniform rate of testing and relative size and visibility.
All forms of testing require lane closures around the test area and immediately beyond it.
Adequate traffic controls are also needed to protect the operators and equipment when they are
stationary. Because deflection surveys are either of a stop-start nature (BB and FWD), or they
proceed slowly (DEF), it is often only possible to conduct testing on some sections, particularly
those in urban areas, at non-peak traffic periods incorporating both lane closures and traffic
control. These practical limitations can influence sampling decisions.
The following discussion about the relative merits and accuracy of strength testing devices
excludes comments about OH&S requirements because they are common to all devices.
Benkelman Beam (BB)
A BB network testing process is relatively much slower than with a DEF or a FWD. Production
depends on traffic conditions and the complexity of traffic control, whether deflection measurement
and recording is manual or automated, whether one or both wheel paths are being tested, and the
mode of transport for operators (walk or ride). Typical rates of progression are of the order of
1.3 km/h for test points at 100 m spacings, 1.1 km/h (50 m spacings), 0.9 km/h (20 m spacings),
and 0.6 km/h (10 m spacings).
The Benkelman Beam provides relatively reliable and accurate deflection data at a point. Because
of its slow testing speed, a Benkelman Beam is suitable only for small scale road networks for
network level strength assessment, and it is particularly suitable for project level assessment.
In addition to traffic control personnel, testing requires the use of a driver, a beam operator and a
data recorder, although data recording can be undertaken and stored by electronic means
reducing the testing team to two personnel.
Deflectograph (DEF)
With a testing speed of 3 to 4 km/h and sample spacing along the pavement from 3 to 7 m, this
device is superior to the BB and FWD in terms of intensity of sampling.
In addition to traffic control personnel, testing requires a driver and a Deflectograph operator who
can record significant features (seal change, change in seal width, drainage, etc) that could
influence pavement strength. Because of its relatively fast testing speed, a DEF is more suitable
than a BB for large scale road networks for network level strength assessment.

Austroads 2005
-- 21 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

Falling Weight Deflectometer (FWD)


Testing at each test point takes 1 to 2 minutes and usually, on average, 150 to 200 test points can
be tested each working day in a network survey. This rate of testing is lower than the
Deflectograph but higher than the Benkelman Beam.
The FWD impact test load, surface stress and deflections are recorded. The deflections are
measured using geophones (accelerometers) that measure the surface acceleration so that the
deflection can be estimated by double integration of the acceleration. This appears to be a more
precise measurement process than with either the Benkelman Beam or Deflectograph. However,
the FWD test method is inherently different from the BB and DEF methods, and is unlikely to give a
more accurate indication of network level strength. Anecdotal sources suggest that the impact test
load of the FWD can cause very low deflections to be recorded when saturated moisture
conditions occur in the pavement.
It should be noted that the FWD test load is an impact load compared with the slowly moving wheel
load of the DEF and BB. Also, deflections measured by BBs and DEFs are relative deflections and
are not absolute values for example, with a BB and a DEF the recorded maximum deflection D0
is the difference between the measured deflection at the applied load and the most remote
measured deflection. Further the BB measures rebound deflections while the applied load slowly
moves away, whereas the DEF measures downwards deflections while the applied load
approaches. The influence of these differences between the survey devices is presumably
accounted for by relationships between deflections measured by the BB, DEF and FWD, such as
those in Section C5.
Largely because of the relative ease of increasing the applied load, a FWD (or HWD) may be more
suitable than a BB or DEF for heavy duty pavements, especially rigid and heavily stabilised
pavements.
A FWD is normally mounted centrally on a towed trailer, and so the trailer must straddle the test
point during a test. Difficulties can arise for left wheel path (LWP) tests in keeping the trailer
reasonably level where the pavement is narrow, or the edge or shoulder are not in sound
condition. For right wheel path (RWP) tests, the potential for safety risks associated with the FWD
trailer or towing vehicle encroaching on the adjacent lane must be carefully assessed. In some
circumstances, it may be necessary for test points to be located between wheel paths.
In addition to traffic control personnel, FWD testing requires a driver for the vehicle towing the
FWD. The driver usually also operates the FWD, monitors the automated recording of test data,
and notes significant features (eg, seal change, change in seal width, drainage, etc) that could
influence pavement strength. Under some conditions, a separate person is used to operate the
FWD and record the data and other features.

Austroads 2005
-- 22 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C3.

FREQUENCY OF PAVEMENT DEFLECTION SURVEYS

C3.1 Guidelines
The frequency of network level pavement deflection surveys can be influenced by:

the strategic importance of the road;

observed levels and progression rates of surface distresses such as roughness, rutting
and cracking as shown in Table 5;

the age of the pavement for example, it may be possible to test less frequently during
the early life of a pavement. However, other factors, such as adverse weather and
traffic conditions, local soil and drainage features, maintenance strategies, and
maintenance delivery practices may offset this and necessitate more frequent testing;

(if known) the rate of deterioration in strength after allowing for seasonal influences;

the volume of heavy traffic;

the deflection measurement device used (eg, reduced frequency with DEF); and

the longitudinal spacing between test points (especially with BB or FWD) for
example, with closer spacing, complete surveys may be necessary less frequently;

A suggested starting point for determining frequency of deflection testing is:


Freeways and heavily-trafficked arterial roads
Roads with high rates of deterioration
Arterial roads with average deterioration
Lightly-trafficked roads with low rates of deterioration

1-2 years
2 years
2-3 years
5 years

The actual frequency for network level deflection surveys should be determined after
considering the above influences and the notes in Section C3.2.

C3.2 Background notes


In determining the frequency of deflection testing of particular road types, the range of factors
affecting the potential structural deterioration of the pavements should be considered. These
factors may include testing budget, historical performance of structural deterioration, local climatic
features, local soil and drainage features, rate of heavy vehicle loading, maintenance strategies
and maintenance delivery.
The frequency of sampling (years) depends on the expected change of strength with time and the
functional importance of the pavement. For highly trafficked roads where a loss of structural
strength has a significant impact, sampling on an annual basis is recommended, although this is
not usually practical. This frequency of sampling could be reduced to a 3 year interval if the
historical annual network monitoring shows relatively minor changes (< 5%) in annual strength,
after excluding seasonal factors. However, for medium to lightly trafficked roads, where a loss of
structural strength has a relatively low consequence, it is recommended that sampling be
conducted around a 3 year interval.

Austroads 2005
-- 23 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

For large networks with limited funds it may not be possible, or necessary, to test as frequently as
suggested in these guidelines. A relationship between pavement age since construction and
deflection testing frequency would be expected to predict less frequent testing during the early life
of a pavement and more frequent testing as pavements age beyond say 10 years. These
relationships could be developed by individual road agencies to suit their particular conditions.
However, other factors, such as, unexpected weather and traffic conditions and better or worse
deterioration performance may necessitate more or less frequent testing.
Particularly for large road networks, there may be advantages in arranging a deflection survey
each year for part of the network rather than surveying the whole network at longer intervals. The
portion of the network surveyed annually can be determined on the basis of the desired overall
frequency of deflection surveys. Examples of potential advantages include flexibility to conduct
targeted testing on sections of particular interest while survey equipment is in the area, more
frequent (perhaps annual) surveys of selected short sections to compile a time series of deflection
data for use in supporting network-level monitoring of the rate of change of strength, more efficient
deployment of equipment and specialist operators, and regular rather than spasmodic budgetary
provision for deflection testing.
On multi lane roads where deflection surveys are usually limited to the lane with the heaviest traffic
loading (by mass not necessarily by volume) in each carriageway, it may be desirable occasionally
to assess the structural condition of other lanes in the same carriageway. Deflection surveys of
other lanes would normally be conducted less frequently than surveys in the lane with the heaviest
loading.
Because of the more intense sampling coverage in a deflection survey using a Deflectograph, it is
generally considered that less frequent surveys are necessary with a Deflectograph than with a
FWD or a BB. For example a Deflectograph deflection survey every 3 to 5 years is likely to be
satisfactory for most freeways and arterial roads.

Austroads 2005
-- 24 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C4.

SCOPE OF PAVEMENT DEFLECTION SURVEYS

C4.1 Guidelines
The aim of a network deflection survey is to collect sufficient data so that road segments that
are homogeneous with respect to pavement strength can be identified, and their
representative strength parameters assessed on the basis of acceptable limits on the
variation of pavement strength within each segment.
If it is known that the traffic is significantly heavier (by mass not necessarily by volume) in one
lane compared to an adjacent lane, then it may only be necessary to collect data in the more
heavily-trafficked lane.
Collection of deflection data in both the left and right wheel paths of each surveyed lane can
assist in identifying seasonal effects. If sampling in only one wheel path is preferred, then it
should be conducted in the left wheel path, where both loading and environmental effects
have the greatest effect on pavement deterioration.
Discrete network sampling of specific sections of pavement may be initiated on the basis of
observed surface distress levels for given road types. Suggested indicators are in Table 5.
Ideally all definable road links should be subject to sufficient testing to confidently assess the
variation in strength along a road link, and to identify the optimal longitudinal sample
spacing.
The optimal sample spacing is the longest distance between each test that would keep the
variations from the representative structural condition within acceptable limits. It is necessary
to define acceptable limits on variation of the strength parameter within each approximately
homogenous segment, in order to establish the optimal longitudinal sample spacing. A
parameter such as the coefficient of variation (COV) the standard deviation divided by the
mean is useful in this context because it provides an understanding of the reliability of the
mean. A COV of 0.25 is considered appropriate for this purpose.
Especially for long networks, the budget for a network deflection survey should allow for initial
testing along each road link to establish the optimal longitudinal sample spacing.
With a BB or a FWD, an initial uniform sample spacing of 50 to 100 m can be undertaken
within the network during the first year of network strength assessment. Where continuous
sampling devices are used, such as the Deflectograph, the optimal sample spacing may be
based on the sampled portion (% road length) of a defined road link (road length).

C4.2 Background notes


C4.2.1

Principles of sampling

The aim of a network deflection survey is to collect sufficient data on a sampling basis to enable
road segments that are homogeneous with respect to pavement strength to be identified and their
representative strength parameters estimated, on the basis of acceptable limits on the variation of
pavement strength within each segment. Such strength parameters include characteristic
maximum deflection and modified structural number, and their values must account for the
variability of pavement strength along the road and within each segment.

Austroads 2005
-- 25 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

Although continuous sampling is obviously the ideal, currently it is usually not financially feasible to
continuously measure deflections along a pavement for a whole road network. Consequently
strength can be assessed by sample deflection measurements at discrete points separated
longitudinally by a test spacing length. Sampling ensures that the minimum amount of data is
collected to assess representative values of pavement strength parameters, taking account of the
variability of pavement strength.
The approach to assessing the variability of strength depends on whether network pavement
strength is being assessed or detailed project level assessments of strength are being made.
Whilst there may be marked variations in pavement strength over short distances along a
pavement, these are not usually relevant for a network level assessment of strength unless these
variations are over significant lengths of pavement in the context of network lengths of pavement.
The longitudinal sampling interval, or test spacing length, for network pavement strength
assessment is therefore much longer than for a project level strength investigation. Project level
investigations seek to detect local weaknesses in pavement strength and therefore the test
spacing length for deflection measurement needs to be set to detect variations in strength over
relatively short lengths of pavement, say less than 20 m.
The aims of sampling the strength variability for longer links in network level assessments of
pavement strength parameters are:

to assess the representative structural condition of the pavement for strategic analysis (eg,
budgets, strategic planning for maintenance and rehabilitation intervention) and contractual
performance purposes; and/or

to detect significant lengths of pavement (eg, sections > 100 m long) where the
representative strength as assessed using sampling is approaching a value where further
detailed structural investigation and structural intervention may be warranted.

Strength variations occurring within a spacing of 50 to 100 m along the pavement lane are not
expected to be identified by a network level deflection survey with long spacings between test
points. Establishing the appropriate sample spacing along the pavement lane should be based on
achieving a reasonable value of the representative structural condition parameter of the pavement
along defined links of the network.
C4.2.2

Longitudinal sample spacing

A uniform longitudinal sample spacing of 50 to 100 m for deflection measurement over several
kilometres of a defined road link can be adopted within the network as a trial in the first year of
network strength assessment. The influence of the longitudinal sample spacing on the assessed
representative structural condition of the pavement can be determined by eliminating the strength
data obtained at various sample spacings. The optimal sample spacing would be the longest
interval showing an acceptable variation from the representative structural condition based on a 50
to 100 m measurement interval. Piyatrapoomi et al (2003, 2004) have described a case study in
Queensland demonstrating this concept.
Where continuous deflection measurement devices are used such as a Deflectograph or a
Benkelman Beam with automated recording, optimal longitudinal sampling may be based on the
sampled portion (% road length) of a defined road link (road length). A useful guide to optimal
longitudinal sampling under these conditions could be, for example, 100 m per 1 km or 500 m per
5 km. However, this would need to be validated by extensive testing over long lengths of the road
link, and as discussed in Section C2.4, the cost-effectiveness of this type of sampling should be
carefully assessed.

Austroads 2005
-- 26 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

Regardless of the measurement device used, the chosen optimal longitudinal sample spacing
would be expected to vary in accordance with the variability of the deflections measured along the
defined road link. The defined road link has to be sectioned into discrete lengths (segments) so
that the variation in structural condition within each segment is within acceptable limits and a
characteristic value can be assigned to each segment. This segmenting process aims to capture
the major longitudinal variations in strength.
The above process carried out in the first year of network strength assessment should be repeated
on all road links of the network to establish an appropriate longitudinal sample spacing for each
link. Several sampling trials within each road link may be needed if the strength variation
characteristics along the link vary due to significant changes in traffic load and local climatic and
geological differences.
Inherent in the sampling process is an acceptance that transverse strength variations (across the
pavement lane) are not sampled. Transverse variations in strength are ignored when sampling
along the left wheel path in the heaviest trafficked lane where both load and environmental effects
have the greatest influence on pavement deterioration. The left wheel path is considered to be
weakest transverse location in the pavement and therefore is an appropriate sampling location
where sampling in one wheel path only is preferred.
For two lane roads, strength sampling (regardless of the testing device used) in the left and right
wheel paths of both lanes or a single lane is a potential means of assessing seasonal effects if
they are considered relevant. However, except with a DEF, measuring deflections in both wheel
paths is not routine for a network survey, partly because of the cost.
Figure 7 shows an evenly spaced sampling arrangement that is commonly used where the
deflection measurement device is a FWD. It may be appropriate to sample in the wheel paths of
only one lane where its traffic is significantly heavier (by mass not necessarily by volume) than in
the adjacent lane.
A Deflectograph or a pair of Benkelman Beams tests both wheel paths in one lane. A FWD or a
single Benkelman Beam can measure deflections in the left wheel paths at the same spacing
between consecutive deflection measurements along each of the two lanes when testing on twolane roads.

Equally spaced staggered left wheel path sampling locations (not to scale)
Figure 13: FWD Sampling locations on a single carriageway two lane road

Austroads 2005
-- 27 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

Strength sampling along multi-lane roads may be limited to the wheel paths of the known heaviest
loaded lane along each carriageway (where Deflectograph testing is used). For FWD testing,
strength sampling normally occurs only in the left wheel path as shown in Figure 8. However,
strength sampling of other lanes in the same carriageway may be needed occasionally, to
ascertain their structural condition relative to their lighter traffic loading.

Median

Equally spaced staggered left wheel path sampling locations (not to scale)
Figure 14: FWD Sampling locations on a dual carriageway road

It should be noted that the above sampling process is an example of how a sampling regime can
be developed. Any sampling processes must be able to identify significant differences in
pavement/subgrade strength along defined road links. This will allow these links to be sectioned,
or segmented, into relatively homogenous sections for traffic loading, climate and condition
(accounting for both the surface and structural state of the pavement). This requirement for
homogenous sections should be the basis for developing an appropriate sampling process.
In setting a practical longitudinal sample spacing the trade-offs between sample spacing (length),
the accuracy of the deflection measurement at each sample point, and the frequency of the
sampling (years) need to be considered to ensure that the assessment of long term strength
change at the network level is based on a consistent data series.

Austroads 2005
-- 28 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C4.2.3

Budgeting to establish optimal longitudinal sample spacing

The budget for a network deflection survey should allow for initial trial testing along each road link
to establish the optimal longitudinal sample spacing. The number and extent (road length) of
sampling trials along each road link for establishing the appropriate longitudinal sample spacing for
each particular section within the road link may depend on the available budget for the sampling
process. However, ideally all definable road links should be subject to sampling trials that cover
sufficient road length to confidently assess the variation in the strength parameter along each link
in establishing the optimal sample spacing. Particularly for large networks, therefore any budget
constraints on the sampling process should be reviewed in the light of the estimated total cost of
the survey, the marginal cost of collecting additional data, and the expected benefits of the survey.
The budget available for the optimal sampling along each road link should also allow for different
levels of sampling (test spacing length) for each road type. Higher levels of sampling (shorter test
spacing length) are recommended for the higher functional road classes and the more heavily
trafficked roads.
C4.2.4

Assessing strength parameters

The strength parameter that represents the structural condition of the pavement along
approximately homogenous segments that comprise each road type that forms the network for a
strategic network analysis (strategic planning for annual maintenance and rehabilitation budgets
and interventions) is usually represented by the mean strength value. This mean strength value for
each road type is arrived at by taking the mean of the characteristic strength values (85th
percentile) found for the defined individual segments within each road type. The characteristic
strength is used to best represent the strength sampled in each defined individual segment.
It is necessary to define the acceptable limits on variation of the strength parameter within each
approximately homogenous segment, in order to establish the optimal longitudinal sample
spacing. The parameter for strength variation could be either the standard deviation (s) from the
mean strength value or the coefficient of variation (COV, the standard deviation divided by the
mean strength) as these parameters provide an understanding of the reliability of the mean. A
COV of 0.25 would generally be suitable (Sapkota et al 2001).
C4.2.5

Practices in deflection measurement surveys

C4.2.5.1

Current practices

Transit New Zealand and some Australian State and Territory road agencies have used some form
of network strength assessment using BBs, DEFs or FWDs, which are described in Section C2.
For example, Transit New Zealand (TNZ) has conducted an annual network level deflection
surveys of part of the NZ State Highway network since 1999 (and plans to continue this practice)
using a FWD with test points at staggered spacings of 100 m on two-lane roads (ie, 200 m spacing
in each lane) where AADT exceeds 2,000 vpd, covering the network each 3 years. Similarly,
MRWA has conducted network level deflection surveys every 1 or 2 years since 1998 of its road
network, using a FWD with test points at staggered spacings of 400 m on both lanes of two-lane
roads (ie, 800 m spacing in each lane). For the States and Territories with smaller networks (eg,
Tasmania and the ACT), pavement strength assessments were made from deflection surveys
using relatively close sample spacings at moderate testing cost. However for data collection, cost
effectiveness is usually more important than total cost, and so project level collection methods are
usually not warranted at a network level particularly for large networks, as the surveys would be
both costly and time consuming. A project level approach to sampling network strength means
intensive data collection at relatively low speed which can reduce the availability of strength data
for a given program and budget.
Austroads 2005
-- 29 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C4.2.5.2

Targeting a prioritised subset of a network

Network strength surveys aim to achieve an acceptable representation of pavement strength over
each road segment for a whole road network comprising many segments. Where there are
limitations on testing budgets or the need for network strength assessment is not critical, network
strength surveys may be based on testing a prioritised subset of segments of the road network
rather than sampling all segments comprising the whole network. With the current limitations of
strength assessment devices, strength surveys often only proceed where other performance
indicators, such as roughness, cracking and rutting, suggest potential or actual loss of structural
strength.
Table 5 suggests values and progression rates for roughness, rutting and cracking, based on
monitoring long term pavement performance (Tepper, Fossey and Koh 2002), as a practical guide
for initiation of discrete network level deflection testing of pavement lengths for given road types or
for typical functional road classifications.
Table 5: Suggestions for performance indicators to undertake discrete network level sampling

Typical Road Function

Suggested Performance Indicators and Values

Typical Operating
Conditions

Rutting
(1.2m straight edge)

Roughness

Cracking

Speed
(km/h)

AADT
(v/day)

Limit
(IRI (mm/km))

Rate
(IRI/yr)

Limit
(mm)

Rate
(mm/yr)

Limit
(% area)

Rate
(% area/yr)

100

> 30,000

3.5

0.05

10

0.3

0.1

100

> 10,000

4.2

0.08

10

0.5

0.1

Medium trafficked arterial

80 - 100

2,000 10,000

4.2

0.2

15

0.6

0.5

Low trafficked arterial or main road

Various

< 2,000

5.4

0.3

20

0.8

10

Freeway
Highly trafficked arterial road

Note:

1.

Deflection survey to occur when either or of any of the above performance indicator limits or rates is exceeded.

A further example of network sampling is the use of frequent strength monitoring of defined road
lengths, or sites, where it may not be practical to sample at a full network level. When selecting
these defined road lengths it is critical to ensure that they are representative of the road links they
are located on. Historical pavement performance indicators (roughness, rutting, strength and
cracking) as well as construction and maintenance records are useful in selecting such sites.
C4.2.5.3

Possible future developments

As at December 2004, there is no device for continuous and robust direct monitoring of pavement
strength at a network level. Such a device would be expected to reduce the current cost of
network level surveys for pavement strength, thereby enabling increased survey coverage (scope).
However, a High Speed Deflectograph (HSD) is under development by Greenwood Engineering to
measure pavement deflection at highway speed (Rasmussen et al 2002). The HSD measures the
velocity of the pavement deflection, rather than the displacement, using laser Doppler sensors
mounted on a heavy vehicle. The displacement, or deflection, is estimated from the velocity which
is the first derivative of the displacement. The results from early field trials with the HSD compare
favourably with FWD test results.

Austroads 2005
-- 30 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

A representative measure of the moisture content of the pavement/subgrade system under


different seasonal conditions and its relationship with the measured deflection would be useful
information to pursue. However, as moisture content variation within the pavement/subgrade
system is a highly variable and complex phenomenon in relation to seasonal conditions, it is
unlikely that a meaningful relationship between measured deflection and moisture content is
achievable.
C4.2.6 Measurement details
C4.2.6.1

Assessing seasonal variations in pavement strength

Ideally any potential variations in the assessed values of pavement strength parameters due to
seasonal variations should be minimised or quantified. Network assessment of pavement strength
should be conducted when the strength is judged to be at its expected lowest annual value.
Alternatively, pavement strength could be assessed from two surveys, one when the strength of
the network is judged to be at its expected highest and the other at its lowest annual values, to
gain the overall variation from a mean network strength estimate. This may not be practical for
reasons of cost and logistics, although these limitations may not apply to small networks.
Measurement of surface deflections in both the left and right wheel paths along lanes is another
approach used to gain an assessment of the seasonal influence on network strength. This
approach assumes that the right wheel path strength estimate is the upper bound value that is less
influenced by seasonal variation and is therefore relatively higher. On the other hand, the
estimated strength along the left wheel path may not necessarily represent the lower bound value
unless the survey was conducted when the expected lowest annual strength value occurred. It
may also be possible to estimate the factors that give rise to potential seasonal variations in
strength and account for these by a prediction model (Loizos et al 2002).
C4.2.6.2

Need for standard measurement procedures

A consistent test method should be used with each device so that the resulting estimates of
pavement strength are comparable between different devices and on different sections of the
network. A standardised testing procedure is necessary with documentation of the deflection
relationships between different devices to aid the conversion of the deflection data into a standard
strength parameter (see Section C5).

Austroads 2005
-- 31 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C5. RELATIONSHIPS BETWEEN MEASURES OF PAVEMENT


DEFLECTION

C5.1 Guidelines
Deflection data from the Benkelman Beam (BB), Deflectograph (DEF), and Falling Weight
Deflectometer (FWD) deflection measuring devices can be related, and can be used to
estimate pavement strength parameters.
Relationships between deflections from the three different measuring devices appear to vary
with pavement type and structure. The following relationships are based mainly on studies
on sealed granular pavements in Australia, and are suggested as default relationships where
local comparative data is not available.
For FWD data, the following relationships are based on using a 300 mm diameter loading
plate in the FWD tests (see Table 4).

These relationships must be used with caution, and


should ideally be verified by local experimental
observations.

FWD and BB:


Deflection FWD50 kN

= 1.25 Deflection BB40 kN/550 kPa

Deflection FWD40 kN

= 0.91 Deflection BB40 kN/550 kPa (when D0 < 1 mm)

FWD and DEF:


Deflection FWD50 kN

= 1.5 Deflection DEF40 kN/550 kPa

Deflection FWD50 kN

= 1.5 Deflection DEF40 kN/750 kPa

BB and DEF:
Deflection BB40 kN/550 kPa

= 1.2 Deflection DEF40 kN/550 kPa

Deflection BB40 kN/550 kPa

= 1.2 Deflection DEF40 kN/750 kPa

where:
Deflection FWD50 kN

= Deflection (maximum value, Do)


measured using a FWD with a 50 kN impact load

Deflection FWD40 kN

= Deflection (maximum value, Do)


measured using a FWD with a 40 kN impact load

Deflection BB40 kN/550 kPa

= Re-bound deflection measured using a BB


with a 40 kN wheel load and 550 kPa tyre pressure.

Deflection DEF40 kN/550 kPa

= Deflection measured using a DEF


with a 40 kN wheel load and 550 kPa tyre pressure

Deflection DEF40 kN/750 kPa

= Deflection measured using a DEF


with a 40 kN wheel load and 750 kPa tyre pressure.

Austroads 2005
-- 32 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C5.2 Background notes


C5.2.1

Deflection relationships between Benkelman Beam, Deflectograph and


Falling Weight Deflectometer

It is useful to be able to relate deflection data from different deflection devices so that the
pavement strength parameters can be estimated using data from various forms of deflection
collection.
Relationships between deflection data from FWD, Benkelman Beam and Deflectograph devices
were investigated for VicRoads (Clayton and Jameson 2001) and MRWA (Goh and Begg 2000).
In summary these relationships appear to vary with pavement type and structure and other factors,
such as differing ranges of accuracy for different devices. This makes the conversion of deflection
from one device to another problematic (Lang 2002). These deflection relationships may be
limited to specific pavement types and structures and ranges of deflection.
Documented relationships are set out below. However, they must be used with caution and ideally
should be verified by experimental observations.
C5.2.2

Existing relationships between Falling Weight Deflectometer and Benkelman


Beam

Based on the relationships found by Goh and Begg (2000) in Western Australia, DeBeer (1992) in
South Africa and the interim recommendations made by Jameson (2000), the following relationship
between FWD deflections and Benkelman Beam deflections is proposed:
Deflection FWD50 kN = 1.25 Deflection BB40 kN/550 kPa

(Eqn C1.1)

where:
Deflection FWD50 kN

Deflection (maximum, D0) determined from a Falling Weight


Deflectometer with a 50 kN impact load (refer to Table 4); and

Deflection BB40 kN/550 kPa

Rebound deflection determined from a Benkelman Beam with a 40 kN


wheel load and 550 kPa tyre pressure.

Rearranging Equation C1.1 for the Benkelman Beam deflections, assuming linear elastic deflection for a
reduced FWD test load of 40 kN:
Deflection BB40 kN/550 kPa = 1 Deflection FWD40 kN

(Eqn C1.2)

where:
Deflection FWD40 kN

Deflection (maximum, D0) determined from a Falling Weight


Deflectometer with a 40 kN impact load (refer to Table 4).

Tonkin and Taylor (1998) recommend the following approach be used as an interim relationship
between Benkelman Beam deflections and 40 kN FWD load deflections based on New Zealand
flexible pavements.
Where maximum deflections are less than 1 mm, adopt the following:
Deflection BB40 kN/550 kPa

1.1 Deflection FWD40 kN

(Eqn C2.1)

0.91 Deflection BB40 kN/550 kPa

(Eqn C2.2)

Alternatively
Deflection FWD40 kN

Austroads 2005
-- 33 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

where:
Deflection FWD40 kN

Deflection (maximum, D0) determined from a Falling Weight


Deflectometer with a 40 kN impact load (refer to Table 4).

The other terms are as defined above. Equation C2.1 is the same as that adopted for the
Austroads Design Procedures for Flexible Overlays on Flexible Pavements (Austroads 2004c).
Where maximum deflections exceed 1 mm, the following relationship applies:
Deflection BB40 kN/550 kPa

1.1 ( Deflection FWD40 kN )0.4

(Eqn C3.1)

( 0.91 Deflection BB40 kN/550 kPa )0.4

(Eqn C3.2)

Alternatively
Deflection FWD40 kN

where the FWD and BB deflections are in mm.


If a FWD load of 50 kN and plate pressure of 700 kPa is used, the relationship shown in
Equation C2.2 becomes the following, assuming the deflection is linear elastic:
Deflection FWD50 kN

1.13 Deflection BB40 kN/550 kPa

(Eqn C4)

Equation C4 is approximately the same as Equation C1.1. Equation C2.2 and Equation C3.2 imply
that there may be a tendency for non-linear deflection behaviour for New Zealand pavements when
deflections exceed 1 mm.
Having regard to all the possible variations and influences,
Equation C1.1 appears to be the more robust for a wider range of application, and so
Equation C3.1 and Equation C3.2 are not recommended.
Queensland Department of Main Roads (QDMR) found the following relationship between the
QDMR Benkelman Beam and FWD deflections (Baran 1994) when using a 40 kN test load for both
the Benkelman Beam and FWD:
Deflection BB40 kN/550 kPa

A Deflection FWD40 kN

(Eqn C5)

where:
A = coefficient depending on level of deflection
= 0.9 (deflections > 1 mm)
= 0.95 (deflections 1 mm)
The other terms are as defined above.
C5.2.3

Existing relationships between Falling Weight Deflectometer and Deflectograph

Research in the Netherlands confirmed the feasibility of establishing relationships between


deflection bowl data from a FWD and a Deflectograph (Hoyinck et al 1992). This research was
mainly conducted on asphalt pavements rather than unbound pavements. In addition, while FWD
deflections were normalised to a 50 kN applied load, the axle load for the Deflectograph was
100 kN, not 80 kN as used with most Deflectographs in Australia. Consequently, more extensive
fieldwork is required to establish a robust relationship for Australasian conditions.

Austroads 2005
-- 34 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

However, both QDMR and the RTA NSW have found the following relationship for the deflections
between their similar Deflectographs and the Benkelman Beam (Baran 1994):
0.85 Deflection BB40 kN/550 kPa

Deflection DEF40 kN/550 kPa =

(Eqn C6)

Similarly, the Austroads Design Procedures for Flexible Overlays (Austroads 2004c) used the
following relationship between the Deflectograph and Benkelman Beam:
Deflection BB40 kN/550 kPa = 1.2 Deflection DEF40 kN/550 kPa

(Eqn C7)

If Equation C7 is substituted into Equation C1.1 then the following relationship between FWD and
DEF deflections results:
Deflection FWD50 kN = 1.5 Deflection DEF40 kN/550 kPa

(Eqn C8)

If a higher surface stress (tyre pressure) of 760 kPa is used for the Deflectograph, Equation C8 is
nominally the same for the relationship between the FWD and DEF deflections because deflection
is directly proportional to load rather than surface stress, and may be expressed as:
Deflection FWD50 kN = 1.5 Deflection DEF40 kN/760 kPa

(Eqn C9)

In reviewing Equations C9, C8 and C1.1, the following relationship is proposed between the BB
and DEF deflection results that is the same as Equation C7, except for the Deflectograph tyre
pressure:
Deflection BB40 kN/550 kPa = 1.2 Deflection DEF40 kN/760kPa
C5.2.4

(Eqn C10)

Summary of deflection relationships

On the basis of the most robust relationships, the following relationships in Table 6 are
recommended to relate deflection data from the FWD, Benkelman Beam (BB) and Deflectograph
(DEF).
Table 6: Summary of deflection relationships
Convert Deflections From Devices

Equation

Relationship

BB deflections to FWD

C1.1

FWD50 kN = 1.25 BB40 kN/550 kPa

BB deflections to FWD (when D0 < 1 mm)

C2.2

FWD40 kN = 0.91 BB40 kN/550 kPa

DEF deflections to FWD

C8

FWD50 kN = 1.5 DEF40 kN/550 kPa

DEF deflections to FWD

C9

FWD40 kN = 1.5 DEF40 kN/760 kPa

DEF deflections to BB

C10

BB40 kN/550 kPa = 1.2 DEF40 kN/760 kPa

DEF deflections to BB

C7

BB40 kN/550 kPa = 1.2 DEF40 kN/550 kPa

Note:

All FWD deflection parameters in this document are based on use of a 300 mm diameter loading plate,
and so a 50 kN applied load corresponds to a road surface stress of 700 kPa, etc, as set out in Table 4.

Austroads 2005
-- 35 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C6.

VERIFICATION OF DISTANCE MEASUREMENT

C6.1 Guidelines
The accuracy of distance outputs from each deflection measuring device should be verified:

before each network deflection survey commences;

at intervals not exceeding one month during each survey;

after the completion of each survey; and

immediately after any change is made to the distance measuring features of the
deflection measuring device or its host vehicle during a survey.

The preferred means of checking distance outputs is by the deflection measuring device
making 5 runs on a test lane between clearly marked stations where the distance between
them is at least 1.0 km, and has been measured to an accuracy of within 0.005%
( 50 mm/km) using precise ground survey techniques.
The distance recorded on all runs should be within 0.1% ( 1 m/km) of the distance obtained
from precise ground survey techniques.

C6.2 Background notes


The above guidelines are similar to corresponding advice in other Austroads guidelines in this
series. However, higher accuracy is achievable than the 0.1% ( 1 m/km) in these guidelines. For
example, TNZ requires distances recorded by deflection measuring devices to be within 0.05%
( 0.5 m/km) of the distance obtained from precise ground survey techniques (TNZ 2002).
Accurate correlation of road condition data with physical location on the road network is pivotal to
the success of almost all uses of road condition data. Australian experience indicates that almost
all errors in reports from road condition surveys relate to distance measurement, particularly when
comparing time-series data from a number of surveys. Reliable distance measurement is therefore
an essential aspect of all road condition surveys.
Road agency databases should generally be supported with robust and unambiguous physical
reference points and permanent features such as roadside markers, bridge abutments, side road
intersections, etc. Transit New Zealands State Highway Location Referencing Management
System Manual (TNZ 2004) is an example of a detailed set of procedures for establishing and
maintaining markers to support a road reference system. Road agency databases should also
contain the distances between the various reference points and permanent features. A limited
number of these distances should be measured to a high degree of accuracy (0.005% or better)
using precise ground survey techniques, and clearly marked. The precisely measured sections
should be sealed, relatively straight, without sags or crests, and between 1 km and 15 km long.
There are a number of ongoing uses for measured sections, including calibration and checks of a
range of road condition survey host vehicles and equipment.
Preparation for a road condition survey should include inspection and any necessary maintenance
of roadside markers at reference points.

Austroads 2005
-- 36 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

In general, a distance verification check should be undertaken before and after each survey, and at
intervals during lengthy surveys, either monthly or after completion of specific sections within the
survey. For a relatively small road condition survey, because of the cost of establishing a distance
verification measured length, a road condition measuring device may be accepted for use without
further verification provided:

a thorough and successful distance verification has been undertaken within the past
3 months; and

there have been no changes to hardware or software since the last verification.

Condition and pressure of tyres on the host vehicle can affect distance measurement outputs.
Experience has shown that distance outputs are very sensitive to the temperature of tyres, and that
variations in temperature are typically the most common cause of inaccurate distance
measurement in road condition surveys. It is therefore desirable to operate the host vehicle at the
normal operating speed for up to 15 minutes before undertaking a distance check or commencing
survey operations.
Currently available deflection measurement devices (BB, DEF and FWD) do not operate at high
speed, as described in Section C2. Distance checking of deflection measurement devices should
therefore be done from a standing start, and by coming to rest at the end. The risk of errors in
distance measurement increases with high acceleration and braking.
Traffic control may be necessary during distance checking to ensure safety.

C6.3 Recording deflection test locations


For monitoring the change in network strength with time a practical means is necessary to identify
test points (BB and FWD) and testing locations (DEF) so that subsequent testing can be
conducted as near as possible to the previous testing points and locations. Especially in network
surveys by a FWD and occasionally with BB testing, it is most desirable that test locations be
recorded accurately, and a means should be used to accurately re-locate each actual test point on
the pavement some time (years) after the test has been performed, using road referencing
systems in combination with GPS coordinates. This is particularly important for FWD testing where
the sample spacing can be long. In this way, each new deflection test report can more accurately
indicate the change in network strength with time.
In some circumstances, paint marks on the road surface may be appropriate to locate deflection
tests undertaken using a FWD or a BB. However, paint marks will be lost when the next
resurfacing occurs.
As noted in Section C2.2, all deflection measuring devices can incorporate a Differentially
Corrected Global Positioning System (DGPS), often referred to as a GPS receiver, that locates
each survey test point to within a specified distance (eg, 5.0 m) of its actual location. TNZ
specifies an accuracy of 2.5 m in this context (TNZ 2002).

Austroads 2005
-- 37 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

In using location referencing devices such as GPS, it is necessary to consider:


1.

the accuracy of the location fixing of each test location has to be consistent with accuracy
required at a network level which could range from 1 m to 5 m, depending on the
variability and structural condition of the network; and

2.

the GPS location, or reference location, must be capable of being readily and accurately
converted to a linear link-node referencing system, as commonly used by the larger road
agencies.

Austroads 2005
-- 38 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C7

VERIFICATION TESTING FOR DEFLECTION

C7.1 Guidelines
These guidelines cover verification of deflection surveys on road networks with sealed granular
pavements carried out by Benkelman Beams (BBs), Deflectographs (DEFs) and Falling Weight
Deflectometers (FWDs).
Detailed verification testing is necessary to assess a new system before it is commissioned for use, after
changes to hardware and software, immediately after damage may have occurred, and as described
below in association with network surveys.
For all deflection measurement devices, the manufacturers instructions should be followed for periodic
(daily, weekly, monthly, etc) checks and all necessary adjustments made to keep the device correctly
calibrated.
Periodic harmonisation tests for similar devices are also advisable, similar to the DEF harmonisation trials
conducted in the early 1990s by RTA NSW with the four DEFs that RTA built.
Verification of deflection measurements using a Benkelman Beam or a Deflectograph:
Verification of deflection measurements using a BB or DEF should be undertaken monthly, and involves:

checking and adjustment of the components usually in accordance with manuals provided by the
supplier (eg, BB dial gauge or the equivalent automated deflection measuring equipment, DEF
signal conditioner, DEF test frame, etc, and for both BB and DEF the test load, tyre pressures,
temperature measuring equipment, etc); and

fresh surveys of a section of road with known deflection characteristics.

Verification sections are required within each network being surveyed, for use before, during, and after
each deflection survey. Verification sections for deflection surveys on sealed granular pavements:

should be generally representative of the nature and condition of the pavements being surveyed;

should exhibit the complete expected range of maximum pavement deflections, especially between
0.5 mm and 1.5 mm;

for a DEF deflection survey, should be at least 500 m long; and

for a BB survey, should be long enough for at least 10 test sites (eg 100 m with 20 m spacings, or
500 m with 50 m spacings).

Acceptance is based on judgement of the reasonableness of any differences between deflections


measured either in repeat runs on the same day, or between tests at the same test locations in successive
surveys, after allowing for strength changes due to seasonal and other effects.
Verification of deflection measurements using a FWD (or HWD):
Because of the relative complexity of a FWD (when compared to a BB or DEF), a complete calibration and
verification process for a FWD involves separate calibration and verification of a number of components
as well as the complete FWD unit. For FWD users, the main issues include the accuracy of the offset of
deflection sensors from the load point, the repeatability of load and deflection measurements both shortterm and over time, and the relative calibration of the different sensors.
These guidelines favour the calibration and verification testing protocols published by the European
COST 336 documentation for FWD verification (Chapter 6 in COST 2003), as listed in Table 7.

Austroads 2005
-- 39 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C7.2 Background notes


For any asset management planning process, it is most desirable that the input data is accurate
and unbiased. Inaccurate and biased pavement deflection data can lead to incorrect conclusions
about the condition of a road pavement, and poor estimates of the remaining service life. For
example, a systematic error of 5% in deflection may result in an error of 25% in estimated
remaining pavement life.
Common sources of error in deflection surveys include errors in location referencing (see
Section C6), incorrectly loaded trucks (BB and DEF), improper calibration of measuring equipment
(all devices), incorrect placement of deflection sensors (FWD), inappropriate adjustments
(normalisation, FWD), and inappropriate adjustments for the duration of loads (FWD).
For FWD, COST 3361 prepared a set of procedures for calibration and verification of individual
components as well as for the FWD as a complete unit (COST 1999). COST identified three levels
of calibration and verification, viz:

Manufacturer level, for testing before delivery to a purchaser and after servicing, to ensure
high quality standards are met (eg, M1-1999 in Table 7);

User or FWD operator level, for periodic checks during testing operations (eg, U1-1999 to
U6-1999 in Table 7); and

Calibration Station, which is an intermediate level, involving a degree of impartiality and


testing of a complexity that would usually be beyond the capability of an operator or user (eg,
C1-1999 to C5-1999 in Table 7).

Table 7 lists these COST procedures, the suggested frequency of their application, the aims and
brief descriptions. Appendix 2 contains the COST procedures for short term repeatability and for
long term repeatability, which can be conducted together. The complete COST 336 documentation
is at http://62.242.229.98/fog/fwd/cost336.htm.
Corresponding
FWD
information
from
the
USA2
is
at
http://www.tfhrc.gov/pavement/ltpp/resource.htm (Law PCS 2000).
This was developed for
research purposes as part of the US Long Term Pavement Performance (LTPP) program.

The European Commission's COST Transport Program undertook a project on the use of FWDs in pavement
evaluation (COST 336), involving 20 European nations from June 1996 to December 1999. COST 336 involved three
main tasks, including the applicability of FWDs at road network level, and FWD calibration. The report describes the
agreed common European code of good practice for the use of FWDs in pavement evaluation.
2

The USA Long Term Pavement Performance (LTPP) program was established in 1987 as part of the Strategic Highway
Research Program (SHRP). As at 2005, the the Federal Highway Administration (FHWA) manages the USA LTPP
program.

Austroads 2005
-- 40 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)
Table 7: COST 336 procedures for FWD calibration and verification
COST 336 designation and name

Suggested frequency

Aim and brief description

C1-1999

Dynamic reference
calibration of FWD load cell

At least once every two years, or more often


as considered necessary by the owner.

Aim is to ensure that the load cell measures the peak value of the load pulse accurately. With the load cell mounted to the
loading plate, various drop heights are applied in a series of drops on the FWD plate and a reference load cell platform.

C2-1999

Laboratory reference
calibration of dismounted
FWD deflection sensors

At least once every two years.

Aim is to verify and calibrate the signal processing electronics of the FWD that produce deflection values. Various
amplitudes and durations of deflection pulses are applied to the FWD deflection sensor and a reference instrument.

C3-1999

In-situ reference calibration of


dismounted FWD deflection
sensors

At least once every two years.

Aim is to verify that the FWD defection sensors produce correct peak values of deflection with the FWD as the load
generator. The test uses reference instrumentation, and checks that vibrations caused by release of the falling mass do not
affect the recorded deflections, by mounting the sensor on a test rig some distance form the FWD.

C4-1999

In-situ reference verification


of mounted FWD deflection
sensors

Optional, pending more experience.

Aim is to verify that the FWD deflection sensors at various distances from the loading plate produce correct peak values of
deflection. Deflections are recorded for a range of drop heights by the FWD sensor and a reference instrument. The test is
repeated for each of the FWD deflection sensors.

C5-1999

In-situ FWD harmonisation


procedure

Optional, suggested at least once every two


years.

Each FWD is compared with others. Aims are to verify repeatability, to derive harmonisation factors, and to improve
reproducibility of deflection recording.

M1-1999

Static reference calibration of


FWD load cell

At least once every two years, and if


COST 336 C1 cannot be performed, and
when a new load cell is mounted on the FWD.

Aim is to ensure that the load cell measures the peak value of the load pulse accurately. Can be used as a substitute for
COST 336 C1. Reference load cell and the FWD load cell under test are subjected to a range of loads.

U1-1999

Verification of FWD deflection


sensor position

At least once every month, and each time the


deflection sensors are moved.

Sensors may be moved between FWD jobs, due to differing client needs. Aim is to ensure that sensors are positioned to a
tolerance of 4 mm + 0.5% of the radial distance (eg, 10 mm for the 1200 mm sensor).

U2-1999

FWD short-term repeatability


verification

At least once every month, or more often as


considered necessary by the operator.

Aim is to verify that the FWD produces consistent results at a specific test point. Deflections and loads are recorded from
all but the first two of 12 drops. Deflections are normalised and checked for consistency (SD of loads < 2% of mean load,
and SD of normalised deflections < 0.2 mm).
(Note: This procedure is in Appendix 2.1 herein.)

U3-1999

FWD long-term repeatability


verification

At least once every month, or more often as


considered necessary by the operator.

Aim is to indicate possible problems with FWD deflection results with a standard target load. Deflections and loads are
recorded from all but the first two of 12 drops at an established marked test site. Deflections are normalised and checked
for consistency with previous results at the same test site. Consistency is judged after taking account of seasonal and
temperature conditions.
(Note: This procedure is in Appendix 2.2 herein.)

U4-1999

Relative calibration of FWD


deflection sensors

At least once every six months, or more often


as considered necessary by the operator.

Aim is to ensure that all sensors on the FWD are in calibration with each other. Known as stacking test, all deflection
sensors are stacked coaxially so that they are exposed to the same deflection in each drop in a series of 10 drops.
Recorded deflections are checked for consistency (range < 0.4mm, and ratio of mean deflection for each sensor to the
mean deflection for all sensors between 0.995 and 1.005 inclusive).

U5-1999

Reference calibration of FWD


temperature probe

At least once every year, or more often as


considered necessary by the operator.

Aim is to ensure that the FWD temperature probe measures the air and pavement temperature accurately. A reference
thermometer is used. Applies to thermo-couples and other contact type temperature recording devices, but not to infrared
temperature sensors.

U6-1999

Reference calibration of FWD


odometer

At least once every six months, or more often


as considered necessary by the operator.

Aim is to determine calibration factors for the FWD odometer or distance measuring instrument. The FWD distance reading
is recorded in 2 or 3 runs over a precisely measured 500 m length travelled at low speed.

Austroads 2005
-- 41 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C8.

REPEATABILITY AND BIAS

C8.1 Guidelines
Repeatability is a statistical term to indicate the extent of variation in outputs about the mean for a single
operator using the same method. Repeatability is the standard deviation of measures obtained in repeat
tests using the same measuring device and operator on a single, randomly selected road.
Bias is a statistical term to indicate whether a device is systematically measuring high or low when
compared to a reference set of measures. Bias error is the average of the errors from many
measurements throughout the survey. Bias should be no greater than 1%.
Repeatability runs and bias checks for surveys using a FWD (or HWD):
Repeatability and bias for deflection surveys using Falling Weight Deflectometers (FWDs) are covered in
Section C7 and Appendix 2.
Repeatability runs and bias checks for surveys using a BB or a DEF:
Repeatability runs and bias checks of BB and DEF deflection survey systems should be conducted over a
clearly defined section of a lane of at least 10 km total length with a range of characteristic maximum
surface deflections at the 100 m reporting interval level of between 0.5 mm and 3.0 mm.
Five (5) repeatability runs should be conducted for each survey team undertaking the network survey at
the start of the survey. Subsequent bias checks and further repeatability runs are then conducted on the
test section throughout the survey at intervals of no more than 3,000 lane-km.
The standard deviation of characteristic maximum surface deflections for individual 100 m reporting
intervals should be less than 3%. In addition, the r2 correlation should be at least 0.95 when the individual
maximum deflection values for each reporting interval are regressed against the mean maximum
deflection values.
Non-compliance with the above recommendations is a matter for each road agency to decide
independently, as these recommendations may not always be achievable in practice.

C8.2. Background notes


Repeatability and bias are important factors in ensuring that changes in strength, as derived from
surface deflections, can be identified over relatively short time spans. Repeatability indicates
variation in measures about the mean, whereas bias error is potentially a more serious problem
that can distort perceptions of the condition of the road network and indicate that the pavement
strength is better or worse than it really is. Bias error can be caused by errors in equipment
calibration or physical damage to the measuring device such as may be sustained survey
operations. The bias error indicates whether the measurement is systematically high or low when
compared to a reference set of measurements.
For a FWD, the COST U2-1999 and U3-1999 procedures cover repeatability and bias testing.
Appendix 2 contains the COST procedures for short term repeatability (COST 336 U2-1999) and
for long term repeatability (COST 336 U3-1999), which can be conducted together.

Austroads 2005
-- 42 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C9.

DATA REPORTING

C9.1 Guidelines
Reporting deflection data:
Deflection data can be reported in tabular or graphical form or both. Deflections are reported
in mm (to 0.01 mm) from BB and DEF surveys, and in microns from FWD data. (Note: 1,000
microns = 1 mm.)
For deflection surveys using a Deflectograph (DEF), summary surface deflection bowl data,
as listed below, should be reported at 100 m intervals in both wheel paths, using for network
purposes the deflection bowl with D0 nearest the characteristic value (85th percentile) for all
bowls (with 4m spacing, there are 25) in that 100m length. In addition, charts of individual
half bowls (showing deflections at 50 mm spacings) can confirm the location of the maximum
deflection and the standard offsets for reporting other deflection values.
For deflection surveys using a FWD or HWD or a Benkelman Beam (BB), all recorded
surface deflection bowl data should be reported for each test point.
As a minimum, the following surface deflection bowl data should be reported:

for a BB or DEF, measured deflections at least at distances of 0, 200, 300, 600 and
900 mm from the centre of the moving test load (ie, D0, D200, D300, D600, and D900);

for a FWD (or HWD), normalised deflections measured at distances of 0, 200, 300,
450, 600, 900 and 1,500 mm from the centre of the impulse test load (ie, D0, D200, D300,
D450, D600, D900 and D1500);

for all devices, curvature (ie, D0 - D200); and

for a DEF, the characteristic maximum deflection, D0.

Deflection measurements are desirable as far as possible from the centre of the applied load
and preferably beyond the 900 mm (BB and DEF) and 1,200 mm (FWD) offsets mentioned
above, as deflections at large offsets increase the likelihood of recording the full extent of the
bowl.
Reporting other data:
For deflection surveys using a Falling Weight Deflectometer (FWD or HWD), the target
load and applied load should be reported for each test location.
For deflection surveys using a Deflectograph (DEF) or a Benkelman Beam (BB), the load
and the tyre pressures should be reported.
Irrespective of the deflection measurement method, that is, for surveys using BB, DEF or
FWD, the location (including road, lane and wheel path or offset distance from kerb or road
edge), date, time, prevailing weather, air temperature, and pavement temperature (only
where the asphalt thickness is known to be > 50 mm) should be reported.

Austroads 2005
-- 43 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

C9.2 Background notes


Table 8 is an example of a tabular report from a FWD survey with a target surface stress of
700 mm (50 kN with a 300 mm diameter loading plate, see Table 4). While Table 8 shows
deflections at nine offsets, as outlined in Section C2.5, seven are common, and six are adequate
for network purposes.
Table 8: Extract from report on network level FWD survey

FWD Network Level Survey Report


Road name: .. SH 10 - Pacific Highway

Test device identification: FWD No xyz (Owner Pty Ltd) ..

Transverse location of test sites: Left Wheel Path


Lane
No
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1

Surface FWD
Chaina Time of
Temp Stress
ge (km) Test
(kPa)
(oC)
0.025
0.125
0.226
0.325
0.425
0.526
0.625
0.725
0.050
0.150
0.250
0.350
0.450
0.550
0.650
0.750

21:32
21:33
21:34
21:35
21:36
21:38
21:39
21:40
23:32
23:33
23:34
23:35
23:36
23:38
23:39
23:40

29
29
28
28
28
28
28
28
15
15
15
15
15
15
15
15

696
706
702
700
701
713
701
709
699
703
699
702
689
699
703
700

Target surface stress: . 700kPa

Date: 16 September 2004


Longitudinal spacing: 100 m

FWD Deflection (micron)


Deflection Offset (mm)
0

200

300

450

600

750

900

1200

1500

169
367
271
325
686
388
418
475
244
464
312
190
212
208
352
277

83
269
191
242
513
288
347
362
186
354
225
158
180
173
298
214

69
214
156
201
405
231
293
290
159
306
191
145
166
160
259
182

63
180
120
162
284
178
238
218
133
242
152
130
145
142
211
143

58
144
89
128
183
134
184
160
112
185
119
116
125
124
168
110

52
117
72
105
119
105
146
118
91
141
97
102
108
106
136
83

54
99
57
89
79
95
122
87
79
106
76
89
90
91
108
63

41
67
43
59
51
55
75
47
51
56
46
65
63
62
64
31

33
41
29
36
31
40
38
24
45
41
39
54
46
46
44
23

With deflection data from a Deflectograph survey, characteristic deflection is reported because the
longitudinal sample spacing of between 3 m and 7 m enables a Deflectograph to be used as a
project level strength assessment tool, as well as for network level purposes. In these guidelines,
the 85th percentile is regarded as the characteristic value for network purposes. This contrasts
with Austroads guidelines for project level deflection analysis, where the characteristic value
depends on the road function (Austroads 2004b, Section 6.2.2.6).
With deflection data from a Deflectograph survey, the characteristic maximum deflection and the
characteristic curvature are reported at 100 m reporting intervals, as for other pavement condition
parameters such as roughness, rutting and texture. With deflection data from a BB or FWD
network survey, the longitudinal sample spacing is likely to be too long for meaningful reporting at
100 m intervals, and so deflection bowl data including curvature is reported specifically for each
test point.

Austroads 2005
-- 44 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

A Deflectograph measures relative deflections at close spacings. Reporting options include:

tabular form at offset distances of 0, 200, 300, 600 and 900 mm from the centre of the
moving test load and referred to as D0, D200, D300, D600, and D900, respectively, so that they
correspond with deflections recorded in BB and FWD testing; and

graphical form at offset intervals of 50 mm between the centre of the moving test load and
the end of the beam (1.2 m for Deflectographs in use in Australia as at December 2004).

Figure 15 is an example of a graphical report from a Deflectograph survey. Of the four charts in
Figure 15, the top two show the LWP and RWP maximum deflections (D0) at 4.0 m longitudinal
spacings (ie 250 values) over a length of 1.0 km from chainage 0.500 km to 1.500 km. The bottom
two charts show the shapes of the LWP and RWP deflection bowls at chainage 1.462 km. Each
bowl was created from 22 deflection readings at 50 mm intervals. Note that the Deflectograph
recorded a small number of deflections beyond the maximum deflection in each wheel path,
thereby ensuring that the actual maximum deflection can be identified. As with the BB, the
reported deflections are relative values, each being the difference between the recorded deflection
and the deflection reading furthest from D0, because the furthest reading is reported as zero.

Figure 15: Extract from a graphical report from a Deflectograph survey

Austroads 2005
-- 45 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

SECTION D SUMMARY
D1.

Sampling and measurement of pavement deflections

The principles of sampling and deflection measurement in a network level context for sealed
granular pavements have been outlined with a view to providing consistency and quality of
outcomes, including a discussion of the relative merits of different devices. A number of new
procedures need to be established, reviewed and clearly defined.
These guidelines distinguish deflections (that can be measured and reported using one of three
main measurement devices) from pavement strength and estimates of values of pavement
strength parameters. Table 1 contains a 7-step overall outline of the process of estimating
pavement strength parameters for network level asset management planning purposes, based on
data collected in pavement deflection surveys.
The following procedures are proposed:
Sampling

The strength parameter that represents the structural condition of the pavement along
approximately homogenous segments suitable for a strategic network analysis, for purposes
such as strategic budgetary planning for maintenance intervention, is usually the mean of the
characteristic strength values (85th percentile) found for all individual segments within each
road type.

It is necessary to define acceptable limits on variation of the strength parameter within each
assumed homogenous segment in establishing the optimal longitudinal sample spacing as
part of the sampling process. The statistical parameter to represent strength variation could
be either the standard deviation, s, or the coefficient of variation, COV, as they provide an
understanding of the reliability of the mean. A COV of 0.25 is a suitable limit for the variation
in the strength parameter for use in establishing the optimal longitudinal sample spacing
(Sapkota et al 2001).

The available budget for the sampling process will influence the number and extent (road
length) of sampling trials along each road link for establishing the appropriate longitudinal
sample spacing for each particular section within the road link. However, ideally all definable
road links should be subject to sampling trials that cover sufficient road length to confidently
assess the variation in the strength parameter along each link in establishing the optimal
sample spacing. Particularly for large networks, therefore any budget constraints on the
sampling process should be reviewed in the light of the estimated total cost of the survey, the
marginal cost of collecting additional data, and the expected benefits of the survey.

Ideally any potential variations in the assessed values of pavement strength parameters due
to seasonal variations should be minimised or quantified. Network assessment of pavement
strength should be conducted when the strength is judged to be at its expected lowest
annual value.

Alternatively, pavement strength could be assessed from two surveys, one when the strength
of the network is judged to be at its expected highest and the other at its lowest annual
values, to gain the overall variation from a mean network strength estimate. This may not be
practical for reasons of cost and logistics, although these limitations may not apply to small
networks.

Austroads 2005
-- 46 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

Comparison of surface deflections in the left and right wheel paths along lanes is potentially
another approach used to gain an assessment of the seasonal influence on network
strength. This method is not routine, partly because of the cost of a network survey of
deflections in both wheel paths. This approach may assume that the right wheel path
strength estimate is the upper bound value that is less influenced by seasonal variation.
However, the estimated strength along the left wheel path may not necessarily represent the
lower bound value unless it is judged that the survey was conducted when the expected
lowest strength value occurs. It may also be possible to estimate the factors that give rise to
potential seasonal variations in strength and account for these by a simple prediction model
(Loizos et al 2002).

Table 5 shows suggestions for the initiation of network sampling. This is based on current
long term pavement performance monitoring, and is a practical guide to discrete network
level sampling of pavement lengths for given road types or typical functional road
classifications.

Test procedures

A consistent deflection testing approach should be used for each deflection testing device so
that the resulting estimates of pavement strength are comparable between different devices
and on different sections of the network. Where a standardised testing procedure is not
possible, documentation of the testing procedure needs to be undertaken to aid the
conversion of the deflection data into a standard strength parameter. An approved Test
Method is the preferred approach. Deflection testing should be performed using standard
applied loads and pressures to simulate the field conditions of the actual axle loads and tyre
pressures experienced.

In network level surveys, at each test point, at least the D0, D200, D300, D600 and D900 deflection
values should be recorded. Where possible, the D450 and D1500 deflections should also be
recorded. The D0 value is used in estimating network pavement strength. The D200 and D300
values are used to estimate D250 which is used in the ratio D250/D0 to filter out cemented base
and other bound pavements from the estimate of network pavement strength. The D600, D900
and D1500 values can be used in estimating the Adjusted Structural Number (SNP) as
described in Appendix 1. Deflections are desirable as far as possible from the centre of the
applied load and preferably beyond the 900 mm offset mentioned above (eg, D1500), because
deflections at large offsets increase the likelihood of recording the full extent of the bowl.
The additional cost of recording and storing D450 can mostly be warranted on the basis of the
potential to use this data in more detailed analysis.

In setting a practical longitudinal sample spacing the trade-offs between sample spacing
(length), the accuracy of the deflection measurement at each sample point and the frequency
of the sampling (years) need to be considered to ensure that a consistent long term
assessment of strength change at the network level occurs.

Location referencing

For monitoring the change in network strength with time a practical means is necessary to
identify test points (BB and FWD) and testing locations (DEF) so that subsequent testing be
conducted as near as possible to the previous testing locations. In network surveys by a BB
or FWD, a means should be used to accurately re-locate each actual test point on the
pavement some time (years) after the test has been performed at each specified location
using road referencing systems in combination with GPS coordinates. This is particularly
important for FWD testing where the sample spacing can be long.

Austroads 2005
-- 47 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

For location referencing devices such as GPS, the accuracy of the location fixing of each test
location has to be consistent with accuracy required at a network level which could range
from 1 m to 5 m, depending on the variability of the structural condition of the network.

The GPS location, or reference location, must be capable of being readily and accurately
converted to a linear link-node location referencing system that is commonly used by MAs.

D2.

Estimation of network level pavement strength parameters

Various means of estimating network level pavement strength parameters have been outlined. As
shown by comparative testing (see Appendix 1.2.4), bowl deflections other than the maximum
deflection, D0, do not significantly improve estimates of the Modified Structural Number, SNC, with
currently used relationships.
Quantification of network strength potentially allows assessment and prediction of structural
deterioration and estimation of the remaining traffic load capacity of the pavement. Network
strength parameters can be effectively used in conjunction with the following considerations:

Current structural capacity, and therefore the remaining structural capacity, should be
assessed in relation to a definition of when terminal structural condition is reached. The
actual terminal structural condition and its associated distresses will depend on the levels of
service, or functionality, required of the pavement. As the defined terminal structural
condition is difficult to quantify, indicative investigation condition levels are suggested for use
in the interim (see Table 2). These levels of service and their limiting distress values are
based on avoiding rapid or catastrophic failure and its consequences.

From a defined terminal structural condition, with the distress limits as shown in Table 2, the
remaining structural life can be estimated based on the difference between the existing
distresses and the limiting distresses. The deterioration rate, the time rate of distress from
the current distress level to the limiting distresses, is needed to estimate the remaining life
(years). The deterioration estimation needs to account for the factors influencing it (traffic
loading, climate, construction quality, maintenance, etc).

Deterioration prediction needs to be made with reference to the limiting distresses and the
existing distresses. Table 2 shows that the limiting distresses for the suggested indicative
investigation levels depend on the level of service required from the road. The variation of
the limiting distresses with level of service also provides some reserve structural capacity
needed to avoid catastrophic failure.

D3.

Interim structural deterioration model for sealed granular pavements

An interim structural deterioration model, using a pavement strength parameter, SNC, as the
dependent variable has been postulated (see Appendix 3). The Interim Model is included for
information only. This postulation has been based on the following assumptions that need
substantiation, or modification, based on further information:

The terminal structural condition of the pavement has been inferred from Austroads (2004b)
predicted deflection reductions due to the impact of granular resheeting.

The above assumption is also based on the view that the deflection reductions due to
strength recovery are equal to the deflection increases due to the loss of strength.

The network strength parameter, SNC, is based only on the measurement of maximum
deflection (D0). However, previous studies have shown that deflection (bowl shape) is a
more appropriate estimate of pavement stiffness rather than strength.

Austroads 2005
-- 48 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

REFERENCES AND BIBLIOGRAPHY


AASHTO (1993). AASHTO Guide for Design of Pavement Structures, (AASHTO: Washington
D.C., USA).
AUSTROADS (1992). Pavement Design A Guide to the Structural Design of Road Pavements.
AP-17/92. (Austroads: Sydney, NSW, Australia).
AUSTROADS (2003a). Comparison of Project Level and Network Level Pavement Strength
Assessment. AP-T21/03. (Austroads: Sydney, NSW, Australia).
(The output of Austroads Project T&E.P.N.528)
AUSTROADS (2003b). Pavement Strength in Network Analysis of Sealed Granular Roads: Basis
for Austroads Guidelines. AP-R233/03. (Austroads: Sydney, NSW, Australia).
(An output of Austroads Project BS.A.C.025)
AUSTROADS (2004a). Pavement Design A Guide to the Structural Design of Road Pavements.
AP-G17/04. (Austroads: Sydney, NSW, Australia).
AUSTROADS (2004b). Pavement Rehabilitation - A Guide to the Design of Rehabilitation
Treatments for Road Pavements. AP-G78/04. (Austroads: Sydney, NSW, Australia).
(The output of Austroads Project T&E.P.N.502)
AUSTROADS (2004c). Technical Basis of the 2004 Austroads Design Procedures for Flexible
Overlays on Flexible Pavements. AP-T T34/04. pp 149 (Austroads: Sydney, NSW, Australia).
(An output of Austroads Projects T&E.P.C.029 and PUB.PT.C.007)
BARAN, E. (1994). Developments in pavement testing and interpretation, Queensland Technology
Workshop on Low Volume Roads, March 1994.
BLAKE, P., MILAZZO, A. and ROSE, S. (1996). Rural Road Replacement Projections A
Probabilistic Method using Roughness, Proc. Roads 96 Conference, Combined 18th ARRB
Transport Research Conference, New Zealand, Part 4, pp 215-228, (ARRB TR: Vermont South,
Victoria, Australia).
CLAYTON, B. and JAMESON, G. (2001). Correlation between Benkelman Beam, PaSE
Deflectograph and FWD, ARRB Transport Research Contract Report RC2007, (ARRB TR,
Vermont South, Victoria Australia).
CLAYTON, B. and STYLES, E. (2001). Long Term Pavement Performance Study Summary of
Data Collection Activities: 1994 2001, ARRB Transport Research Contract Report RC1525-2,
June 2001, pp32, (ARRB TR, Vermont South, Victoria Australia).
COST 336 (2003). Use of Falling Weight Deflectometers in Pavement Evaluation, viewed 13
January 2005, http://62.242.229.98/fog/fwd/cost336.htm.
DE BEER, M. (1992). Developments in the Failure Criteria of the South African Mechanistic
Design Procedure for Asphalt Pavements. Proceedings, 7th International Conference on Asphalt
Pavements, Vol 3 (ISAP: Austin, Texas, USA).
DEPARTMENT OF MAIN ROADS NSW (1982). Benkelman Beam Deflection Test. Test Method
T160. (Roads and Traffic Authority NSW, Sydney, Australia).
EIJBERSEN, M.J. and VAN ZWIETEN, J. (1998). Application of FWD-measurements at the
network level, 4th International Conference on Managing Pavements, Vol 1, pp 438-450 (Dept. Civil
Engineering, University of Pretoria, South Africa).

Austroads 2005
-- 49 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

FERNE, B.W. (1997). Use of deflections at network level in England for programming and other
purposes, Cost 336 Workshop, 4-5 June at LNEC, Lisbon, Portugal.
GARGETT, D. and PERRY, R. (1998). Interstate non-bulk freight, Proc. 22nd Australian Transport
Research Forum, Vol 22 pp 19-28 (Transport Data Centre of New South Wales Department of
Transport: Sydney, Australia).
GOH, A.L. and BEGG, S. (2000). Development of a relationship between Falling Weight
Deflectometer and Benkelman Beam Deflection, Goh and Begg Consultants, MWRA Panel
Contract 1035C98.
GREENWOOD ENGINEERING (2002). High Speed Deflectograph, Denmark
(www.greenwood.dk).
HIGHWAY RESEARCH BOARD, (1962). The AASHO Road Test, Report 5, Pavement Research,
Highway Research Board Special Report 61E, pp 56 (National Academy of Sciences: Washington,
D.C., USA).
HODGES, J.W., ROLT, J. and JONES, T.E. (1975). The Kenya Road Transport Cost Study:
Research on Road Deterioration. Transport and Road Research Laboratory Report 673, pp 56
TRRL: Crowthorne, Berkshire, United Kingdom).
HOYINCK, W.T., VAN DER LOO, J.M.M., MULDERIJ, J. and WEGMEETDIENST, R.K. (1992).
Comparative Tests of FWD and Lacroix Deflectograph. Proc. 7th International Conference on
Asphalt Pavements. Vol 3 pp179 193. ISAP.
ISOHDM (International Study of Highway Development Management Tools) (1998). HDM-4
Technical User Guide Part D: Models, D-1: Road Deterioration (Flexible Pavements) (HDM-4\Vol2\D-1a) (ISOHDM: University of Birmingham, United Kingdom).
JAMESON, G.W. (1993). Development of procedures to predict structural number and subgrade
strength from falling weight deflectometer deflections, (ARRB TR, Vermont South, Victoria,
Australia, unpublished).
JAMESON, G.W. (2000). Towards improved Austroads guidelines for the design of flexible
overlays on flexible pavements, Contract Report RC91039D-2 (ARRB TR: Vermont South, Victoria,
Australia).
KONIDITSIOTIS. C. and KOSKY, C. (1996). National Uniformity in Pavement Condition Data
Definition, Proceedings Roads 96 Conference, Combined 18th ARRB Transport Research and
Transit New Zealand Land Transport Symposium, Part 4, pp 425-46 (ARRB TR: Vermont South,
Victoria, Australia).
LANG, J. (2001). Personal Communication, October 2001, Department of Infrastructure, Energy
and Resources, Tasmania, Australia.
LANG, J. (2002). Personal Communication, July 2002, Department of Infrastructure, Energy and
Resources, Tasmania, Australia.
LAW PCS. (2000). LTPP Manual for Falling Weight Deflectometer Measurements - Operational
Field Guidelines, Version 3.1. Prepared for the FHWA LTPP. (LAW PCS, Beltsville, Maryland,
USA).
LOIZOS, A., ROBERTS, J. and CRANK, S. (2002). Asphalt Pavement Deterioration Models for
Mild Climatic Conditions, Proceedings 9th International Conference on Asphalt Pavements,
Copenhagen, Denmark (not yet published).

Austroads 2005
-- 50 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

MCLEAN, J.R. and MARTIN, T.C. (1999). In search of the missing model, Road and Transport
Research, 8(3) (ARRB TR: Vermont South, Victoria, Australia).
MARTIN, T. (1998). State-of-the-art pavement performance modelling at a network and project
level, ARRB Transport Research Special Report SR56, pp 39 (ARRB TR: Vermont South, Victoria,
Australia).
MARTIN, T. (2002). Interim Network Level Structural Deterioration Model for Sealed Granular
Pavements, ARRB Transport Research Contract Report RC1702/3, pp 14 (ARRB TR: Vermont
South, Victoria).
MARTIN, T.C. and RAMSAY, E. (1996). Rural pavement improvement prediction due to
rehabilitation. ARRB TR, Research Report ARR 283, pp 23 (ARRB TR: Vermont South, Victoria,
Australia).
MARTIN, T. and GLEESON, B. (1999). The Effect of Maintenance on Pavement Performance:
The Accelerated Load Pilot Test, ARRB Transport Research Contract Report RC7094B, pp 42
(ARRB TR: Vermont South, Victoria, Australia).
MARTIN, T., GLEESON, B., JOHNSON-CLARKE, J., TREDREA, P. and FOSSEY, D. (2000). The
Effect of Maintenance on Pavement Performance: Accelerated Load Testing in 1999/2000, ARRB
Transport Research Contract Report RC90264, pp 44 (ARRB TR: Vermont South, Victoria,
Australia).
MARTIN, T., GLEESON, B., JOHNSON-CLARKE, J., TREDREA, P., LUKE, R. and FOSSEY, D.
(2001). The Effect of Maintenance on Pavement Performance: Accelerated Load Testing in
2000/2001, ARRB Transport Research Contract Report RC1739, pp 62 (ARRB TR: Vermont
South, Victoria, Australia).
MARTIN, T. and CRANK, S. (2001). Use of Pavement Strength Information in Network Asset
Management, ARRB Transport Research Contract Report RC1702, Austroads Project No
BS.A.C.025, October 2001, pp 32 (ARRB TR: Vermont South, Victoria, Australia).
MOROSUK, G., RILEY, M.J. and ODOKI, J.B. (2001). HDM-4 Highway Development and
Management, Vol 6, Modelling Road Deterioration and Works Effects, Review Draft Version 1.1,
pp 306 (World Road Association (PIARC): Paris, France).
NAASRA (National Association of Australian State Road Authorities) (1979). Interim Guide to
Pavement Thickness Design, pp 169 (NAASRA: Sydney, Australia).
NRTC (National Road Transport Commission) (1996). Mass Limits Review: Report and
Recommendations, July 1996, pp 64 (NRTC: Melbourne, Australia).
PAINE, D. (1998). The incorporation of structural data in a pavement management system, Proc.
4th International Conference on Managing Pavements. Department of Civil Engineering, University
of Pretoria, South Africa.
PARKMAN, C.C. and ROLT, J. (1997). Characterisation of pavement strength in HDM-III and
possible changes for HDM-4, Transport Research Laboratory Unpublished Report,
PR/ORC/587/97 (TRRL: Crowthorne, Berkshire, United Kingdom).
PATERSON, W.D.O. (1987). Road Deterioration and Maintenance Effects: models for planning
and management, The Highway Design and Maintenance Standards Series (John Hopkins
University Press: Baltimore, USA).

Austroads 2005
-- 51 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

PCS/LAW ENGINEERING, (1993), Manual for FWD testing in the Long-Term Pavement
Performance Program. SHRP-P-661. Strategic Highway Research Program (SHRP) pp 129
(National Academy of Sciences: Washington, D.C., USA).
PIYATRAPOOMI, N., KUMAR, A., ROBERTSON, N., WELIGAMAGE, J. (2003). A ProbabilityBased Analysis for Identifying Pavement Deflection Test Intervals for Road Data Collection.
International Conference on Highway Pavement Data, Analysis and Mechanistic Design
Application, Volume II, pp. 291-302, Columbus, Ohio, USA.
PIYATRAPOOMI, N., KUMAR, A., ROBERTSON, N., WELIGAMAGE, J. (2004). Reliability of
Optimal Intervals for Pavement Strength Data Collection at the Network Level. Proceedings of the
6th International Conference on Managing Pavements (ICMP6), Brisbane, Australia.
QDMR (Queensland Department of Main Roads) (1993). Pavement Design Manual, (QDMR:
Brisbane, Australia).
RASMUSSEN, S., KRARUP, J.A., HILDEBRAND, G. (2002). Non-contact Deflection
Measurement at High Speed, Proceedings 6th International Conference on the Bearing Capacity of
Roads, Railways and Airfields (BCRA 2002), 24-26th June, 2002, Lisbon, Portugal.
ROBERTS, J.D. (1995). Pavement Management System: Operation Guide and System
Description. 5th ADB Road Improvement Project, Philippines, Kampsax in association with SMEC
and OPCV of Australia, (SMEC International Pty Ltd: Cooma, NSW, Australia).
ROBERTS, J.D. (2000a). A Pavement Structural Deterioration Model for HDM-4, Proceedings
First European Conference on Pavement Management Systems, Budapest, Hungary.
ROBERTS, J.D. (2000b). Project level pavement performance modelling, Road and Transport
Research, 9(4), pp. 29-47 (ARRB TR Ltd: Vermont South, Victoria, Australia).
ROBERTS, J. and MARTIN, T. (1996). Recommendations for monitoring pavement performance.
ARRB TR, Research Report ARR 293, pp 69 (ARRB TR: Vermont South, Victoria, Australia).
SALT, G. and STEVENS, D. (2001). Pavement performance prediction: determination and
calibration of structural capacity (SNP), Proceedings, 20th ARRB Conference, Part 4 pp 99-116
(ARRB TR: Vermont South, Victoria, Australia).
SAPKOTA, B., BUTKUS, F., NORRIS, B. and GOH, A.L. (2001). Main Roads Western Australias
experience in the use of the falling weight deflectometer for network pavement strength
assessment, Proceedings, 20th ARRB TR Conference, (ARRB TR: Vermont South, Victoria,
Australia).
SCALA, A. J. (1979). An Analysis of Deflection Bowls in Pavements Measured by the Benkelman
Beam Test. Internal Report No AIR 032-2. (Australian Road Research Board: Vermont South,
Victoria, Australia).
SMITH, R.B. (1985). Preliminary Evaluation of the Dynatest 8000 Falling Weight Deflectometer,
Australian Road Research, 15(4), pp. 229-238 (ARRB: Vermont South, Victoria, Australia).
TEPPER, S. and MARTIN, T. (2001). Long Term Pavement Performance Maintenance (LTPPM)
Progress Report, ARRB TR Contract Report RC1561, pp 77. (ARRB TR, Vermont South, Victoria,
Australia).
TEPPER, S., FOSSEY, D. and KOH, S.L. (2002). Long-Term Pavement Performance Study Data Report: 2001/2002 Season. Contract Report RC2008-3. (ARRB TR, Vermont South,
Victoria, Australia).

Austroads 2005
-- 52 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

TONKIN and TAYLOR (1998). Pavement deflection measurement and interpretation for the
design of rehabilitation treatments, Transfund New Zealand Research Report No.117, (Transfund
New Zealand, Wellington, New Zealand).
TRANSIT NEW ZEALAND (1977). Standard Test Procedure for Benkelman Beam Deflection
Measurements, T/1 June 1977, (Transit New Zealand, Wellington, New Zealand).
TRANSIT NEW ZEALAND (2002). Data Collection Specification, Auckland North Network,
PSMC 005, September 2002, (Transit New Zealand, Wellington, New Zealand).
TRANSIT NEW ZEALAND (2004). State Highway Location Referencing Management System
(LRMS) Manual, SM051. (Transit New Zealand, Wellington, New Zealand).
TRL (Transport Research Laboratory) (1993). A Guide to the structural design of Bitumen
Surfaced Roads in Tropical and Sub-Tropical Countries. Overseas Road Note 31, 4th Ed. (TRL:
Crowthorne, Berkshire, United Kingdom).
VICROADS (1986). Pavement Strength Evaluation and Rehabilitation, Technical Bulletin No. 33
(VicRoads: Kew, Victoria, Australia).
YODER, E.J. and WITZCAK, M.W. (1975). Principles of Pavement Design, 2nd Edition, (J. Wiley
and Sons: New York, USA).

Austroads 2005
-- 53 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

APPENDIX 1:

1.1

ESTIMATING PAVEMENT STRENGTH PARAMETERS


FROM DEFLECTION DATA

Modified Structural Number, SNC

A relationship between deflection and modified structural number, SNC, developed by Paterson
(1987) from Brazilian pavement performance data, is defined as follows for sealed granular
pavements:
SNC = Ao ( D0 ) 0.63

(Eqn 1.1)

where:
D0 = Maximum deflection (mm) determined from a Benkelman Beam
Ao = Model coefficient = 3.2 (uncemented base) and 2.2 (cemented base).
As Paterson (1987) notes, Equation 1.1 is not a good fit to the observed data (r2 = 0.56) because
SNC assesses pavement/subgrade strength and deflection measures the stiffness of the
pavement. This is the fundamental limitation of using deflection data to estimate pavement
strength.
Equation 1.1 has been applied to the maximum deflections from the Deflectograph and the FWD,
often without applying corrections to account for the differing test plate pressures. Notionally the
Benkelman Beam and the Lacroix Deflectograph test loads apply a surface stress of 550 kPa while
the FWD applies a surface stress of 700 kPa. Provided the deflection testing is within the usual
range of elastic behaviour, then the relationships in Equation C1.1 and Equation C1.2 in
Section C5.2 can be used in the interim to convert FWD deflection to Benkelman Beam deflection.
Table 1.1 gives sample values of SNC for unbound sealed granular pavements, derived using
Equation 1.1.
Table 1.1: Sample values of mean characteristic maximum deflection (D0) and
corresponding Modified Structural Number (SNC) for unbound sealed granular pavements

D0 (mm)

SNC

D0 (mm)

SNC

D0 (mm)

SNC

D0 (mm)

SNC

D0 (mm)

SNC

0.20

8.82

0.60

4.41

1.00

3.20

1.40

2.59

1.80

2.21

0.25

7.66

0.65

4.20

1.05

3.10

1.45

2.53

1.85

2.17

0.30

6.83

0.70

4.01

1.10

3.01

1.50

2.48

1.90

2.14

0.35

6.20

0.75

3.84

1.15

2.93

1.55

2.43

1.95

2.10

0.40

5.70

0.80

3.68

1.20

2.85

1.60

2.38

2.00

2.07

0.45

5.29

0.85

3.54

1.25

2.78

1.65

2.33

2.05

2.04

0.50

4.95

0.90

3.42

1.30

2.71

1.70

2.29

2.10

2.01

0.55

4.66

0.95

3.31

1.35

2.65

1.75

2.25

2.15

1.98

Note:

For most sealed granular pavements in Australia and New Zealand (viz, sealed granular pavements less than 700 mm thick),
SNC and SNP have the same values (see Section C1.2.1.2 and Appendix 1.2).

Austroads 2005
-- 54 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

1.2

Adjusted Structural Number, SNP

A number of methods are available for estimating the adjusted structural number, SNP, from
deflection data (Morosiuk et al 2001). The methods outlined below are those that are considered
appropriate for conditions in Australia and New Zealand.
1.2.1

Jamesons Method

The structural number, SN, is calculated using the following relationship derived from Jameson
(1993) from a range of pavements:
SN =

1.69 +

842.8

42.94

(D0 D1500)

(Eqn 1.2)

D900

where:
D0 = deflection (mm) at the centre of FWD test plate
D900 = deflection (mm) 900 mm from centre of FWD test plate
D1500 = deflection (mm) 1500 mm from centre of FWD test plate.
The other terms in Equation 1.2 are as defined previously. The deflections are all normalised to a
stress of 700 kPa.
The California Bearing Ratio (CBR) of the subgrade is determined using the following equation
(Jameson 1993):
Log10(CBR) = 3.264 1.018 Log10(D900)

(Eqn 1.3)

The CBR is then used to calculate the structural contribution of the subgrade, SNsg, as shown
below (Hodges et al 1975):

SNsg = 3.51 Log10(CBR) 0.85 ( Log10(CBR) )2 1.43

(Eqn 1.4)

The adjusted structural number, SNP, is determined from the sum of the structural number, SN,
and the contribution of the subgrade, SNsg, as shown below (Hodges et al 1975), assuming that
the modified structural number, SNC, is approximately the same as SNP for most pavements in
Australia and New Zealand (see Section C1.2.1.2):
SNP = SN + SNsg
1.2.2

(Eqn 1.5)

Roberts method

Roberts (1995) developed the following relationship for the structural number, SN, based on FWD
data collected in Australia and the Philippines:

SN = 12.992 4.167 Log10(D0) + 0.936 Log10(D900)

(Eqn 1.6)

with the structural contribution of the subgrade estimated using Equation 1.3 and Equation 1.4.
This allows the adjusted structural number, SNC, to be estimated from Equation 1.5.
Loizos, Roberts and Crank (2002) derived the following relationship to estimate SNP for asphalt
pavements in Greece using FWD deflections:
SNP = 167 ( D0 ) 0.57

Austroads 2005
-- 55 --

(Eqn 1.7)

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

1.2.3

Salts method

Salt and Stevens (2001) developed the following relationship for the Adjusted Structural Number,
SNP, based on FWD deflections on sealed granular pavements in New Zealand:
SNP = 112 D0-0.5 + 47 (D0 D900)-0.5 + 56 (D0 D1500)-0.5 0.4

(Eqn 1.8)

where D0, D900 and D1500 are as defined above, except that the test plate pressure is normalised to
a stress of 566 kPa. This means that if the test plate pressure is 700 kPa (more precisely 700 kPa
See Table 4), then the deflections are adjusted by a ratio of 566/707 (0.80).
1.2.4

Comparison of SNP and SNC estimation for granular pavements

A comparison was undertaken between the Adjusted Structural Number, SNP, estimated by
Jameson (1993), Roberts (1995) and Salt and Stevens (2001) using deflections D0, D900 and D1500,
and the Modified Structural Number, SNC, estimated by Paterson (1987) using the deflection D0.
The comparison was based on a sealed granular pavement network with 688 FWD tests.
Figure 1.1 compares the SNC and SNP estimates and the relationships found between SNC and
SNP for the various SNP relationships.
8.5
SNP (Jameson 1993)
SNP (Roberts 1995)

7.5

SNP (Salt and Stevens 2001)


Line of Equality (SNC = SNP)
SNC (Paterson 1987 using D0)

6.5

5.5

4.5

3.5

Sealed Granular Pavements (688 Test Samples)


SNC = -2.76 + 1.61 x SNP (Jameson: r 2 = 0.96; SE = 0.21)
SNC = -0.72 + 1.05 x SNP (Roberts: r 2 = 0.94; SE = 0.24)
SNC = -0.32 + 1.08 x SNP (Salt & Stevens: r 2 = 0.999; SE = 0.05)

2.5

1.5
1.5

2.5

3.5

4.5

5.5

6.5

7.5

8.5

SNP Prediction (using D0, D900 & D1500)

Figure 1.1: SNC (Paterson) vs SNP (others)

Figure 1.1 shows that all the relationships between SNC and SNP are closely related for this
network data set (r2 = 0.94 to 0.999), particularly the Salt and Stevens (2001) estimate of SNP.
This outcome shows that the additional bowl deflection terms D900 and D1500 do not provide a
significantly different estimate of SNP from that based on only the D0 deflection. On face value the
above data set appears to be reasonably representative of a sealed granular pavement network as
it has a wide range of SNC and SNP values. Assuming this data set is representative, network
level estimates of SNP, or SNC, could therefore be based on estimates using the D0 deflection
without significant loss in accuracy.

Austroads 2005
-- 56 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

The correlation of the Salt and Stevens (2001) estimate of SNP with the Paterson (1987) estimate
of SNC is remarkable. This is particularly so because the Paterson (1987) estimate of SNC was
based on Benkelman Beam deflections on Brazilian pavements while Salt and Stevens (2001)
used FWD deflections on New Zealand pavements to estimate SNP.

1.3

Structural Adequacy Indicator, SAI

The use of the Structural Adequacy Factor, SAI, developed by Eijberson and Van Zwieten (1998)
is described below.
To apply this model in practice three levels were developed as follows:

SAI at Level 1 = SCI600 below lower level;

SAI at Level 3 = SCI600 above highest level; and

SAI at Level 2 = SCI600 between highest and lowest level.

The Surface Curvature Index, SCI600, at the lower level is defined as follows:

SCI600 =

+ a1

a2

(Eqn 1.9)

( 1 + a3 Trucks )
where:
Trucks = Number of trucks per day per lane
a1, a2 and a3 = Constants defined in tabulation below.
Constant

Full Depth Construction

Granular Base Construction

a1

53.9

60.52

a2

134.1

269.6

a3

0.002576

0.003841

The parameters, SCI600, lower level and highest level are defined as follows:

SCI600 = D0 D600;

SCI600, lower level = see Equation 1.9; and

SCI600, higher level = 1.8 lower level.

If the SCI600 is above the highest level then the pavement is regarded as being structurally
inadequate (SAI at level 3) and if SCI600 is below the lower level then the pavement is regarded as
being structurally sound (SAI at level 1). When SCI600 falls between the higher and lower levels,
the structural condition of the pavement is uncertain (SAI at level 2).
The above was further refined by developing a Structural Distress Indicator matrix which, included
the following parameters:
1.

the SAI level (1, 2 or 3);

2.

the parameter, Diff. SCI, defined below in Equation 1.10; and

3.

the amount of visible cracking (percentage of surface area cracked) in the pavement (%
cracking).

Austroads 2005
-- 57 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

Diff. SCI =

( SCI600,traffic SCI600,untraffic )

100%

(Eqn 1.10)

SCI600,untraffic
where:

SCI600,traffic = SCI measured on trafficked area of pavement


SCI600,untraffic = SCI measured on untrafficked area of pavement.
The three parameters were then combined to develop the Structural Distress Indicator matrix of
values from 1 to 10, where 1 indicates a poor structural condition and 10 is a very good structural
condition. The layout of this matrix is below.
Visible
Cracking

<1%

SAI Level 1

SAI Level 2

SAI Level 3

Diff.SCI Diff.SCI Diff.SCI Diff.SCI Diff.SCI Diff.SCI Diff.SCI Diff.SCI Diff.SCI Diff.SCI Diff.SCI Diff.SCI
<10% 10-20% 20-40% >40%
<10% 10-20% 20-40% >40%
<10% 10-20% 20-40% >40%
10

1-10%
10-30%
>30%

1.4

Relative Pavement Strength (RPS) Indicator

The Relative Pavement Strength (RPS) indicator is a useful guide for assigning the timing of
structural intervention (Roberts 2000b). The RPS is a dimensionless unit that estimates the
current structural adequacy of a pavement. The RPS is defined as the percentage of remaining
life due to traffic load relative to the user defined analysis period.
Consequently, a pavement with adequate structural strength for the analysis period (and only the
analysis period) has an RPS value of 100%. To use this indicator each pavement is assigned an
RPS value, generally based on a 20 year analysis period, however, the analysis period can be
revised as required.
Assuming the 20 year analysis period, a pavement with only two years of remaining life has an
RPS of 10%, while a pavement with 50 years remaining life has an RPS of 250%. The timings of
investigative and structural intervention works are assigned according to the outcomes of this
simple analysis. For instance an RPS of less than 10% requires immediate site investigation for
pavement strengthening, while an RPS ranging from 10% to 40% may need a future strengthening
and a more focussed monitoring of strength within one to two years, and an RPS ranging from
40% to 80% does not require any detailed attention in the short to medium term other than some
strength monitoring within the next two to three years.

Austroads 2005
-- 58 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

APPENDIX 2

2.1

COST 336 PROCEDURES FOR


REPEATABILITY TESTING WITH FWDs

Cost 336 Protocol U21999:


Fwd Short Term Repeatability Verification
NOTES:

This protocol is issued under the fixed designation 'Protocol U2'. The
number immediately following the designation indicates the year of original
adoption or, in case of revision, the year of last revision.
Testing in accordance with this protocol should be conducted at least once
every month, or more often as considered necessary by the operator.

1.

Scope

1.1

This protocol covers the determination of the repeatability of magnitude of load and
deflections generated by the Falling Weight Deflectometer (FWD).

1.2

This protocol provides the means for any FWD user for periodic verification of repeatability of
the FWD equipment.

2.

Referenced documents
Protocol U3 - FWD Long-term Repeatability Verification.

3.

Significance and use

3.1

The objective of this procedure is to verify whether the FWD under test is capable of
producing consistent results on a specific test site. In this procedure the short-term
repeatability of a FWD is verified by using a series of twelve successive drops without lifting
the loading plate. The first two drops are omitted from the analysis. The deflections are all
normalised to the mean of the load imparted. The standard deviation of the load and
normalised deflections should agree with the specified limits. When the results do not meet
the requirements, the test should be repeated. Cases of persistent non-compliance,
invalidate data collected by the instrument under test.

3.2

This protocol must be applied as often as specified in the Calibration Scheme or more
frequently as considered necessary by the FWD user. The FWD user should keep records
of the verification as conducted using this protocol.

3.3

This procedure may be combined with Protocol U3.

4.

Apparatus

Falling Weight Deflectometer including control and signal processing electronics.

Austroads 2005
-- 59 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

5.

Procedure

5.1

Enter the following data on the calibration sheet:


FWD user;
FWD manufacturer;
FWD type/serial/ID number;
FWD deflection sensor serial numbers;
FWD loading plate diameter;
FWD deflection sensor offsets;
Current calibration factors for FWD deflection sensors;
Repeatability verification operator name;
Location of repeatability verification;
Date and time of last repeatability verification; and
Date and time of repeatability verification.

5.2

Position the FWD on a smooth, level, sound asphalt pavement structure with no visible
cracks, where a peak deflection in the order of 250 to 600 m can be produced in the load
centre when using the selected target load level.

5.3

Warm up the FWD rubber buffers and condition the test point by repeating a sequence of ten
drops until the loads and deflections that are registered are nearly uniform. The deflections
in this sequence of ten drops should not be showing a steadily increasing or decreasing
trend. If liquefaction or compaction is indicated by the warm-up data, or when the required
deflection level cannot be achieved, relocate the FWD to another pavement.

5.4

Set the drop height and drop mass to generate the selected target load level. Apply two
seating drops, for which no data is recorded, followed by ten replicate drops, for which peak
values of load and deflection are recorded. Only these last ten drops will be used in the
analysis. Do not raise the FWD loading plate during the test.

6.

Analysis

6.1

Normalise all deflections with the use of linear interpolation techniques to a reference load
level that does not depart more than ten percent from the actually applied load
(Equation 8.1). Determine the mean deflection of each deflection sensor for the set of ten
drops (Equation 8.2).

6.2

Determine the standard deviation of all loads (Equation 8.3), and the standard deviation of all
normalised deflections of each deflection sensor (Equation 8.4).

6.3

The standard deviation of the load recorded in the series of ten drops shall be less than, or
equal to two percent of the mean of the recorded values. If the actual standard deviation
exceeds the requirement, then the repeatability verification should be repeated at another
pavement.

Austroads 2005
-- 60 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

6.4

The standard deviation of the normalised deflections, recorded in the series of ten drops
shall be less than, or equal to 2 m in case the mean of normalised deflections is less than,
or equal to 40 m. The standard deviation of the normalised deflections, recorded in the
series of ten drops shall be less than, or equal to the sum of 1.5 m and 1.25 percent of the
mean of the recorded normalised values, in case this mean is greater than 40 m. If the
actual standard deviation of one or more deflection sensors exceeds the specified values,
then the repeatability verification should be repeated at another pavement. Failure to satisfy
the repeatability criteria again, necessitates closer investigation of the deflection sensors and
their holders on the raise/lower bar. The non-compliance invalidates data collected by the
FWD under test.

7.

Symbols
i
j
NK
Fi
Fref
sF
uij
dij
dj

=
=
=
=
=
=
=
=
=

Drop label
Deflection sensor label
Number of drops (= 10)
Magnitude of load at Drop i (kN)
Preselected target reference load (kN)
Standard deviation of load over all drops
Unnormalised deflection measured by Deflection Sensor j at Drop i
Normalised deflection measured by Deflection Sensor j at Drop i
Mean of normalised deflections measured by Deflection Sensor j over NK

sdj

Standard deviation of normalised deflections measured by Deflection Sensor

drops
j
over NK drops
8.

Equations

8.1

Normalise deflections to target reference load level:

dij =

8.2

Fref
Fi

uij

Calculate arithmetic mean of normalised deflections per deflection sensor:


NK

d
dj =

8.3

ij

i =1

NK

Calculate standard deviation of loads:

NK
NK F - Fi
i =1
i=1
NK ( NK - 1 )
NK

2
i

sF =

Austroads 2005
-- 61 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

8.4

Calculate standard deviation of normalised deflections:

NK
NK d - dij
i =1
i=1
NK ( NK - 1 )
NK

2
ij

sd j =

9.

Report
The report should contain at least:
Data of the calibration sheet;
Principal test data used in analysis;
Analysis results; and
Declaration whether FWD or FWD component under test complies with the
specifications.

2.2

Cost 336 Protocol U31999: Fwd Long Term Repeatability Verification


Note:

This protocol is issued under the fixed designation 'Protocol U3'. The number
immediately following the designation indicates the year of original adoption
or, in case of revision, the year of last revision.
Testing in accordance with this protocol should be conducted at least once
every month, or more often as considered necessary by the operator.

1.

Scope

1.1

This protocol covers the determination of the long-term repeatability of the Falling Weight
Deflectometer (FWD) on a carefully selected test site producing deflection data with limited
annual and/or seasonal variation.

1.2

This protocol provides a procedure for easy verification of accuracy of FWD deflection output
at the home base of the FWD user.

2.

Referenced documents
Protocol U1 - Verification of Deflection Sensor Position; and
Protocol U2 - FWD Short-term Repeatability Verification.

Austroads 2005
-- 62 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

3.

Significance and use

3.1

In this procedure the long-term repeatability of a FWD is verified by using a series of twelve
successive drops. The first two drops are omitted from the analysis. The deflections are all
normalised to the target load level. This target load level may be freely chosen at the first
time of performing this repeatability verification action. In all later replicates, the same target
load level must be used. The mean of the deflections is compared to the results previously
collected at the same location. This location should preferably be selected close to the FWD
home base and shielded from climatic influences as much as possible. The objective of this
test is to detect any anomalies in the deflection output. Deflection results will not be a
constant over the year due to temperature and seasonal changes. For that reason the test
provides only indicative data. It reveals whether unexpected absolute changes of deflection
have occurred. If suspicion has risen over the output, load cell and deflection sensors should
be investigated to identify the source of the problem.

3.2

This protocol must be applied as often as specified in the Calibration Scheme or more
frequently as considered necessary by the FWD user. The FWD user should keep records
of the verification as conducted using this protocol.

3.3

This procedure may be combined with Protocol U2.

4.

Apparatus

Falling Weight Deflectometer including control and signal processing electronics


Thermometer
Clock

5.

Preparation

5.1

In the first application of this protocol, the FWD under test should be equipped, and drop
height and deflection sensor offset should be set as used in normal operation situations.
These settings will be termed as default settings. In any future use of this protocol, settings
should be identical to the default settings. Keep a record of the default settings.

5.2

Select a smooth, level, and sound pavement structure with no visible cracks on which
deflections may be measured which will hardly change with time of the year. This test site
should preferably be shielded from direct solar radiation and other climatic influences.
Deflections measured in sequences of multiple drops should not be showing a steadily
increasing or decreasing trend. If so, relocate the FWD to another pavement.

5.3

Mark the position where the loading plate of the FWD rests so that it can be relocated
precisely on the same spot at another day of testing. This may be done by paint, or by
marking a small divot in the pavement with a chisel. Also mark the direction in which the
deflection sensor beam points. Include a description of the test position in the record of the
default settings.

6.

Procedure

6.1

Enter the following data on the calibration sheet:


FWD user;
FWD manufacturer;
FWD type/serial/ID number;

Austroads 2005
-- 63 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

FWD deflection sensor serial numbers;


FWD loading plate diameter;
FWD deflection sensor offsets;
Default FWD settings;
Current calibration factors for FWD deflection sensors;
Repeatability verification operator name;
Location of repeatability verification; and
Date and time of repeatability verification.

6.2

Verify if deflection sensors are mounted at the default offsets (use COST 336 Protocol U1).
In case of non-compliance, reposition deflection sensors to the correct offset. Check
whether the FWD is equipped according to the default settings.

6.3

Warm up the FWD rubber buffers and condition the test point by repeating a sequence of ten
drops until the loads and deflections that are registered are nearly uniform.

6.4

Use the selected target load level. Apply two seating drops, for which no data is recorded,
followed by ten replicate drops, for which peak values of load and deflection is recorded.
Only these last ten drops will be used in the analysis. Do not raise the FWD loading plate
during the test.

6.5

Measure the pavement temperature at mid-depth of the asphalt concrete layer or cement
concrete slab. Register the temperature in C with one digit placed beyond the decimal
point. The accuracy of the temperature-measuring device should be 0.5C. Record day of
the year and clock time in hours (24 hour system) and minutes (eg, 14:35).

7.

Analysis

7.1

Normalise all deflections with the use of linear interpolation techniques to a reference load
level that does not depart more than ten percent from the actually applied load
(Equation 9.1). Determine the mean deflection of each deflection sensor for the set of ten
drops (Equation 9.2).

7.2

Plot the deflections against the measured pavement temperature. Examination of the
variations of these parameters with time will then provide an indication of any anomalies in
the consistency of the FWD equipment, which need to be further investigated.

8.

Symbols
i
j
NK
Fi
Fref
uij
dij
dj

=
=
=
=
=
=
=
=

Drop label
Deflection sensor label
Number of drops (= 10)
Magnitude of load at Drop i (kN)
Preselected target reference load (kN)
Unnormalised deflection measured by Deflection Sensor j at Drop i
Normalised deflection measured by Deflection Sensor j at Drop i
Mean of normalised deflections measured by Deflection Sensor j over NK
drops

Austroads 2005
-- 64 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

9.

Equations

9.1

Normalise deflections to reference load level:

dij =

9.2

Fref
Fi

uij

Calculate mean deflection per deflection sensor:


NK

d
dj =

10.

ij

i =1

NK

Report
The report should contain at least:
Data of the calibration sheet;
Principal test data used in analysis;
Analysis results; and
Declaration whether FWD or FWD component under test depart from the expected
pattern.

Austroads 2005
-- 65 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

APPENDIX 3

INTERIM STRUCTURAL DETERIORATION MODEL FOR


SEALED GRANULAR PAVEMENTS

This Appendix outlines information about an interim structural deterioration model for sealed
granular pavements. The information provided is preliminary, and substantial observational data is
needed to support and confirm the adaptation of the Interim Model for practical use. The primary
aim of the interim structural deterioration model is to provide a quantitative assessment of the
remaining structural pavement life in a network context. It is not intended as a detailed pavement
structural capacity prediction tool at the sub-network level.

3.1.

Background to Interim Model development

3.1.1

Strength parameter and terminal condition

The selection and definition of a suitable pavement performance indicator(s) or strength


parameter, to assess an approaching terminal structural condition, is needed to objectively assess
the remaining structural pavement life in a network context. The only Austroads (1992) defined
terminal condition limiting pavement life is roughness, which is a surface serviceability criterion not
a structural criterion. There is no current recommended structural failure limit in terms of the
pavement performance indicators. Current practice, at a project level, tends to use indicators such
as rut depth and surface deflection in combination with roughness to assess either the end of
pavement life, with subsequent rehabilitation, or the need and extent of surface intervention.
These indicators are useful for targeting deflection surveys.
The magnitude and severity of these indicators for terminal structural condition could also vary
depending upon the consequences of structural failure for the pavement. Therefore these
indicators may need to be lower in magnitude and severity for more heavily trafficked roads
compared with those used for lightly trafficked roads. For example, this is reflected in Table 2 and
Table 5 in these guidelines.
A 1996 study of 30 arterial rural road samples in Victoria (Martin and Ramsay 1996, McLean and
Martin 1999) showed that a majority (67%) of the studys road samples were rehabilitated partly or
fully for excessive roughness reasons rather than for excessive surface deflection alone. For
these samples, excessive roughness was defined when the roughness exceeded 4.3 IRI
(112 NRM). Excessive deflection was defined when the maximum deflection, D0, from FWD
testing exceeded 1.4 mm.
3.1.2

Inferred terminal conditions from Austroads (2004b)

Section 6.2.5 of Austroads (2004b) states that for a granular resheet there is a 6% reduction in
surface deflection for every 25 mm of granular resheet thickness. These rates of deflection
reduction are assumed to apply to the usual range of granular resheet thickness from 100 mm to
250 mm. The impact of this reduction in deflection is a commensurate increase in the pavements
modified structural number, SNC, using Equation 1.1 in Appendix 1 relating SNC to deflection.
Table 3.1 summarises this estimated impact on deflection reduction and increase in SNC.

Austroads 2005
-- 66 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

If the increases in SNC shown in Table 3.1 were assumed to be those needed to restore the
pavement to its original zero pavement age structural condition, then a simple deterioration model
for pavements needing rehabilitation could be postulated. The assumption underlying this
approach is that strength recovery is equal to strength deterioration, that is, linear elastic theory is
appropriate. For most in service conditions of granular pavements, the elastic behaviour
assumption is reasonable (Yoder and Witzcak 1975), although it is not likely to be appropriate for
pavements with marginal materials subject to excessive loading. It also may not be appropriate for
lightly trafficked thin pavements where the majority of structural deterioration may be within the
supporting subgrade.
Table 3.1: Impact of granular resheeting on pavements

Note:

Resheet Thickness
(mm)

Reduction in Deflection1
(%)

Increase in SNC2
(%)

100

24

12.7

150

36

17.6

200

48

21.9

250

60

25.6

1.
2.

Austroads 2004b (Section 6.2.5 and Figures 6.5 and 6.7).


Based on Equation 1.1 in Appendix 1.

3.2.

Basis of interim structural deterioration model

3.2.1

Model postulation

The deterioration of pavement/subgrade strength can be described either by the reduction in an


appropriate strength parameter, or increases in measured pavement deflection over the
pavements life-cycle. An examination of various deflection and traffic load capacity relationships
(Martin 2002) showed that the variations between the deflection and traffic load capacity
relationships are greater than the variations between the pavement strength parameter (such as
SNC) and traffic load capacity relationships (NAASRA 1979, Paterson 1987, Martin 1998, QDMR
1993, TRL 1993). This outcome suggests that deflection deterioration may be more difficult to
predict than strength parameter deterioration.
It is proposed to describe the deterioration of network pavement strength in terms of the Modified
Structural Number, SNC, which is defined in Appendix 1.1 and is the same as the Adjusted
Structural Number, SNP, for sealed granular pavements less than 700 mm thick, as noted in
Section C1.2.1.2. SNP is used by the Highway Development Management model, HDM-4
(ISOHDM 2000), as a pavement strength deterioration indicator.
The most practical means of currently estimating SNC at a network level is to measure surface
deflection by the various means and devices discussed earlier and as shown in Appendix 1. This
means using the relationship between SNC and deflection shown in Equation 1.1 and Table 1.1 in
Appendix 1, with all of its attendant problems of matching strength (SNC) with pavement
performance in terms of deflection that measures pavement stiffness better than strength.
Equation 1.1 in Appendix 1 uses only the maximum deflection, D0, to estimate SNC. As implied by
Martin and Crank (2001) and noted in Section C1.2.1.2, the other bowl deflections (D900 and D1500)
do not significantly improve the estimation of SNC with the relationships used.

Austroads 2005
-- 67 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

Prediction of network level pavement deterioration is an analysis with a long term time frame.
Although a significant number of Australias sealed granular pavements have lives in excess of 40
years (Blake et al 1996) there are often instances of premature structural failure due to local
defects in the pavement base or subgrade that reduce pavement design life. In other cases
reduced pavement life is due to the use of marginal materials. Nevertheless, network pavement
deterioration cannot reasonably account for local defects and marginal materials if they are just a
local problem.
3.2.2

Model definition

The following Equation 3.1 is postulated on the basis of expected sealed granular pavement
performance for use as the Interim Model for the time dependent deterioration of SNC:
SNC =

Kc SNCd { 2 EXP ( km b/DL AGE) }

(Eqn 3.1)

where:
SNC = modified structural number of the pavement/subgrade at time = pavement age, AGE
SNCd = modified structural number of the pavement/subgrade as required for the design traffic
load
Kc = calibration factor for climatic and construction quality factors that influence the initial
value of SNC (default = 1; possible range 0.8 [wet climate, poor construction] to 1.2 [dry
climate, very good construction] )
km = Calibration factor for maintenance factors (drainage and surfacing) that influence the value
of SNC during pavement life (default = 1; possible range 1.2 [poor maintenance] to 0.5
[excellent maintenance])
EXP = e raised to the power
DL = design life of the pavement (currently, 20 to 25 years) which can change with altered
traffic loading conditions
b = Deterioration factor
AGE = Pavement age (years).
Equation 3.1 meets the main boundary conditions, that is, when the pavement age, AGE, is zero,
SNC is at its highest expected value, and when the pavement age is at its terminal value (AGE
DL), SNC has decreased to its lowest expected value (actual value not specified at present). The
reducing value of SNC can be used to estimate the pavements remaining traffic load capacity by
means of existing SNC relationships with traffic load capacity (NAASRA 1979, Paterson 1987,
Martin 1998, QDMR 1993, TRL 1993) at any time (AGE) during the pavements life-cycle.
However, as noted in Section 3.2.1, there is considerable variation in the estimated remaining
traffic capacity given by these equations.
Limitations of the Interim Model
The Interim Model (Equation 3.1) predicts structural deterioration of sealed granular pavements
assuming that ongoing distress only occurs in the upper portion of the pavement base with the
majority of the remaining pavement in relatively good condition. This assumes a stable subgrade
with good drainage and no deterioration in the subgrade. Where distress occurs in lower
pavement layers, such as in very thin pavements with distress in the supporting subgrade,
Equation 3.1 may not be a useful predictor of structural deterioration. However, it still may be
possible to calibrate Equation 3.1 for these conditions.

Austroads 2005
-- 68 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

The Equation 3.1 model form for deterioration does not provide for potential improvements to
pavement/subgrade strength that may occur during a pavements life-cycle as observed by the
previous data review (Martin 2002). The reasons behind these observed strength improvements
need to be established to discover whether additional independent variables need to be included in
Equation 3.1. The longer term prognosis for all pavement/subgrade strength is for its eventual
deterioration, unless maintenance intervention markedly improves structural conditions.
The impact of increased traffic loading above the original design load estimate in Equation 3.1
could be accounted for by a pro rata reduction in the design life, DL, as follows:
Reduced DL =

DL

Design Traffic Load


Increased Traffic Load

(Eqn 3.2)

Equation 3.2 is a simplified, but practical, approach that could be used in conjunction with
Equation 3.1 until improved structural deterioration models become available.
Equation 3.1 can be refined when further evidence is available, such as the typical pavement
terminal life (years) and the prevailing terminal structural conditions (roughness, rutting and
deflection). This refinement needs a reasonable assessment of what would also be an adequate
reserve of traffic load capacity for the pavement to avoid excessive surface and structural
deterioration prior to rehabilitation. This reserve could be expressed as a percentage of the initial
value of SNC (or initial traffic load capacity), which means that the magnitude of the reserve
capacity increases in line with pavements that have an increasing value of SNC. In the interim,
Equation 3.1 can serve as a potential means of predicting the loss of SNC for pavements at a
network level over their life-cycle.
Sectio 3.2.3 provides two alternative means of defining the calibration factors in Equation 3.1. The
assumption underlying both means of defining the calibration factors is that strength recovery is
equal to strength deterioration, that is, linear elastic theory is appropriate, as noted in
Section 3.1.2. The limitations of this approach are also discussed in Sectio 3.1.2.
Deterioration prediction
Structural deterioration prediction using Equation 3.1 needs to be made with reference to the
terminal structural condition and the existing structural condition. Table 2 in Section C1.2.2 shows
that the terminal condition varies depending on the level of service required from the road. The
varying terminal condition with level of service indirectly accounts for some reserve structural
capacity needed to avoid catastrophic failure.
3.2.3

Potential use of the Interim Model

The estimation of the deterioration in SNC can be demonstrated under two extreme possible
conditions as shown below.
Deterioration model with a varying pavement design life (constant deterioration factor, b)
The deterioration predicted by Equation 3.1 is shown in Figure 3.1 for a varying pavement design
life. Equation 3.1 deterioration factor, b, in this case is based on using the increases in SNC with
various levels of rehabilitation thickness in Table 3.1 as being equivalent to reductions in SNC
during the deterioration phase.

Austroads 2005
-- 69 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

Assume initially pavement design life is reached when the minimum rehabilitation thickness of
100 mm is needed to restore the pavement and the surface deflection has decreased by 24%.
Equation 3.1 estimates a potential extension to pavement life of some 91% if pavement life is
based on the longest intervention level when the rehabilitation thickness to restore the pavement is
250 mm and the surface deflection has decreased by 60%. Equation 3.1 uses a constant
deterioration factor, b, of 0.12 in Figure 3.1 to meet the above conditions.
The predicted 91% increase in pavement life means that a design life of 20 to 25 years could
become 30 to 48 years which is similar to the range observed in some parts of Australia (Blake et
al 1996).
Deterioration model with a fixed pavement design life (varying deterioration factor, b)
Alternatively, if it is assumed that the pavement design life is fixed and that rehabilitation coincides
with the end of the design life, Equation 3.1 needs different values of the deterioration factor, b, to
predict the different rates of deterioration in SNC shown in Table 3.1. Figure 3.2 shows the range
of values for b which could also represent the varying influences like maintenance which could be
included in the calibration factor for maintenance, km, if b was held to a constant value.
The value of b in Figure 3.2 varies from 0.12 to 0.23, a range of twice the lowest value of b. This
means that the calibration factor for maintenance effects, km, could range from 1.0 to 0.5.
1
100%
12.7%
17.6%

SNC (Paterson 1987)


as a percent of the as-constructed value

SNC = Kc x SNCd x [ 2 - EXP

( km x 0.12 / DL x AGE)

21.9%

25.6%

Deterioration restored by
100 mm overlay
Deterioration restored
by 150 mm overlay
Deterioration restored
by 200 mm overlay
Deterioration restored
by 250 mm overlay

AGE / Initial Design Life (DL)

1.36

1.66

1.91

0%
0

0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

(Pavement Age, AGE) / (Design Life, DL)

Figure 3.1: % SNC Deterioration vs Pavement Age / Design Life


(varying Pavement Design Life and constant Deterioration Factor (b = 0.12)

Austroads 2005
-- 70 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

1
100%
12.7

17.6
21.9

SNC (Paterson 1987)


as a percent of the as-constructed value

SNC = Kc x SNCd x [ 2 - EXP

( km x b / DL x AGE)

25.6

b = 0.12 (restoration by 100 mm overlay)


b = 0.16 (restoration by 150 mm overlay)
b = 0.2 (restoration by 200 mm overlay)
b = 0.23 (restoration by 250 mm overlay)

0%
0

0.2

0.4

0.6

0.8

(Pavement Age, AGE) / (Design Life, DL)

Figure 3.2: % SNC Deterioration vs Pavement Age / Design Life


(fixed Pavement Design Life and varying Deterioration Factor)

3.3.

Model calibration

3.3.1

Use of performance data

There is currently performance data that can aid the calibration and the possible refinement of the
Interim Model. This performance data includes the accelerated load testing data aimed at
quantifying the influence of maintenance on pavement performance which covers the full pavement
life-cycle (Martin and Gleeson 1999, Martin et al 2000, 2001).
Additional independent variables may need to be included in the Interim Model if they can
rigorously account for long term changes in structural condition, including strength improvements
and seasonal changes. These are not readily available and some will have to come from the
performance data for the long term pavement performance monitoring (LTPP) sites and the long
term pavement performance maintenance monitoring (LTPPM) sites (Clayton and Styles 2001,
Tepper and Martin 2001), and road agency historical databases.
3.3.2

Model calibration

It is possible that the model form finally used for the deterioration of pavement/subgrade strength,
SNC, will allow for variation of both the pavement life and the rate of deterioration as observed
from the deterioration of deflection measured at a network level. It is important that the model
allows for ongoing refinement and calibration from observed performance.

Austroads 2005
-- 71 --

Guidelines for Road Network Condition Monitoring: Part 3 Pavement Strength (Sealed Granular Pavements)

In the interim, if the rate of SNC deterioration with time is observed it can be expressed in terms of
Equation 3.1 as follows:
d(SNC)
d(AGE)

= km b
DL

Kc SNCd { 2 EXP ( km b/DL AGE) }

(Eqn 3.3)

Equation 3.3 can be used to re-calibrate either Kc or b in Equation 3.1 for a given value of
pavement age, AGE, where the pavement maintenance regime has not changed. This approach
assumes that the maintenance factor calibration, km, does not need adjustment. However, the
design life, DL, may also need adjustment under these conditions. The DL may also need
reduction or an increase if there is a significant change to the design traffic loading using the
approach shown in Equation 3.2.
Where the maintenance regime has changed, the maintenance factor, km, will need calibration.
Unexpected and abrupt changes in the observed rate of SNC deterioration may demand a change
in the form of the deterioration model.

Austroads 2005
-- 72 --

INFORMATION RETRIEVAL

Austroads (2005), Guidelines for Road Network Condition Monitoring: Part 3


Pavement Strength (Sealed Granular Pavements), Sydney, A4, 86pp,
IR-88/05

KEYWORDS:
Benkelman Beam, deflection, Deflectograph, Deflectometer, monitoring, network
level, pavement, pavement management, pavement performance, pavement
strength, performance prediction, road management, road network, structural
analysis

ABSTRACT:
This document contains guidelines for and background notes on network level
measurement and reporting of deflection data, and analysis of the structural capacity
of sealed granular pavements, for road network management purposes in Australia
and New Zealand. The guidelines discuss the frequency and scope of network
deflection surveys, including issues such as selection of longitudinal sampling
intervals or sampling proportions for deflection surveys, and choice of parameters
and estimation of parameter values to represent network level pavement strength in
asset management analysis.
The guidelines are intended as a basis for a consistent approach in Australia and
New Zealand. Verification procedures for deflection measurement devices are
covered, and repeatability and bias are discussed. A glossary of terms used in
network level assessment of pavement strength is also included.
The guidelines include for information rather than use, an Interim Model that has
been postulated for prediction of structural deterioration of sealed granular
pavements, depending on influences such as traffic loading, climate, construction
quality, maintenance circumstances, and age.

Вам также может понравиться